Cxo v103 I1 Jan

Vol. 103 No.
1
JANUARY 2020
C L I N I C A L A N D E X P E R I M E N TA L
SPECIAL ISSUE VISUAL OPTICS: LOOKING BEYOND 2020
Modern spectacle lens design

Stand magnifiers for low vision
Multifocal optics
Aberration-controlling lenses for keratoconus
Optical changes with orthokeratology
Optical regulation of eye growth
Aberrations and refractive error development
Peripheral refraction and aberrations
Aberrations and accommodation
Blur adaptation
Adaptive optics vascular imaging
Orthokeratology compression factor
in partnership with
The Hong Kong Society of
Professional Optometrists
Official Journal of: EDITOR
Emeritus Professor Nathan Efron AC
DEPUTY EDITOR
Dr Maria Markoulli
EDITORIAL BOARD INTERNATIONAL ADVISORY BOARD

Associate Professor Nicola Anstice Dr Waleed AlGhamdi
The Hong Kong Society of (Associate Editor) Qassim University, Saudi Arabia
University of Canberra, Australia Professor Raymond Applegate
Dr Lauren Ayton (Associate Editor) University of Houston, USA
The University of Melbourne, Australia Associate Professor Reiko Arita
Dr Alex Black (Associate Editor) University of Tokyo, Japan
Queensland University of Technology, Professor David Atchison
Australia Queensland University of Technology,
Impact Factor: 1.559 ISI Journal Citation Reports©
Ranking: 2018: 40/59 (Ophthalmology) Dr Mei-Ying Boon (Associate Editor) Australia
University of New South Wales, Australia Professor Sharon Bentley
Aims and Scope: Clinical and Experimental Optometry
is a peer reviewed journal listed by ISI and
Ms Lyn Brodie Queensland University of Technology,
Optometry Australia, Australia Australia
abstracted by PubMed, Science Citation Index and
Current Contents. It publishes original research Associate Professor Bang Bui Dr Paul Constable
papers and reviews in clinical optometry and vision (Associate Editor) Flinders University, Australia
science. Debate and discussion of controversial The University of Melbourne, Australia Professor Stephen Dain
scientific and clinical issues is encouraged and Dr Andrew Carkeet (Associate Editor) University of New South Wales, Australia
letters to the Editor and short communications Queensland University of Technology, Professor Steven Dakin
expressing points of view on matters within the Australia The University of Auckland, New Zealand
Journal’s areas of interest are welcome.
Clinical and Experimental Optometry also welcomes Dr Nicole Carnt (Associate Editor) Professor Erica Fletcher
papers that explore the history of optometry and University of New South Wales, Australia The University of Melbourne, Australia
vision science. Dr Holly Chinnery (Associate Editor) Professor Alex Gentle
OnlineOpen: Clinical and Experimental Optometry The University of Melbourne, Australia Deakin University, Australia
accepts articles for Open Access publication. Emeritus Professor H Barry Collin AM Professor Robert F Hess
Please visit https://authorservices.wiley.com/ (Emeritus Editor) McGill University, Canada
author-resources/Journal-Authors/open-access/ University of New South Wales, Australia Professor Michael Ibbotson
onlineopen.html for further information about Dr Cirous Dehghani (Associate Editor) National Vision Research Institute, Australia
OnlineOpen.
The University of Melbourne, Australia Professor Michael Kalloniatis
Address for Editorial Correspondence: Dr Laura Downie (Associate Editor) Centre for Eye Health, Australia
Editor, Clinical and Experimental Optometry, The University of Melbourne, Australia
Suite 101, 68–72 York Street, South Melbourne, Professor Lisa Keay
Victoria 3205, Australia. E-mail: cxo.editor@ Dr Katie Edwards (Associate Editor) University of New South Wales, Australia
optometry.org.au; Tel: +61 3 9668 8500; Queensland University of Technology, Professor Allison McKendrick
Fax: +61 3 9682 0928. Australia The University of Melbourne, Australia
Disclaimer: The Publisher, Optometry Australia
Emeritus Professor Nathan Efron AC Professor Eric Papas
(OA), New Zealand Association of Optometrists (Editor and Chairman) University of New South Wales, Australia
(NZAO), Hong Kong Society of Professional Queensland University of Technology,
Australia
Dr Konrad Pesudovs
Optometrists (HKSPO), Singapore Optometric Optometry consultant, South Australia
Association (SOA) and Editor cannot be held Dr Lesley Frederikson
responsible for errors or any consequences arising New Zealand Association of Optometrists,
Professor Jacob Sivak
from the use of information contained in this New Zealand University of Waterloo, Canada
journal; the views and opinions expressed do not Associate Professor Ian Gutteridge Professor Fiona Stapleton
necessarily reflect those of the Publisher, OA, University of New South Wales, Australia
(Associate Editor)
NZAO, HKSPO, SOA and Editor, neither does the Professor Chi-ho To
The University of Melbourne, Australia
publication of advertisements constitute any Hong Kong Polytechnic University, Hong
endorsement by the Publisher, OA, NZAO, HKSPO, Associate Professor Isabelle Jalbert
Kong SAR, China
SOA and Editor of the products advertised. (Associate Editor)
University of New South Wales, Australia Professor James Wolffsohn
Copyright © 2020 Optometry Australia Aston University, United Kingdom
Associate Professor Andrew Lam
Wiley is a founding member of the UN-backed
Hong Kong Polytechnic University, Hong Publications Manager: Jessica Donald
HINARI, AGORA, and OARE initiatives. They are
Kong SAR, China
now collectively known as Research4Life, making
online scientific content available free or at Dr Angelica Ly
nominal cost to researchers in developing University of New South Wales, Australia
countries. Please visit Wiley’s Content Access – Associate Professor Michele Madigan
Corporate Citizenship site: http://www.wiley. (Associate Editor)
com/WileyCDA/Section/id-390082.html. University of New South Wales, Australia
ISSN 0816-4622 (Print) Dr Maria Markoulli (Deputy Editor)
ISSN 1444-0938 (Online) University of New South Wales, Australia
Dr Lisa Nivison-Smith (Associate Editor)
Cover image: Lisa Breayley, MedPIC, University of New South Wales, Australia
The Royal Victorian Eye and Ear Hospital Dr Nicola Pritchard (Associate Editor)
Queensland University of Technology,
Australia
Dr Kah Ooi Tan
Singapore Optometric Association, Singapore
Dr Phil Turnbull (Associate Editor)
The University of Auckland, New Zealand
Associate Professor Stephen Vincent
(Associate Editor)
Queensland University of Technology,
Australia
GUEST EDITORIAL
Looking and seeing beyond 2020
Clin Exp Optom 2020; 103: 1–2 DOI:10.1111/cxo.12993
Stephen J Vincent PhD a combination bifocal consisting of a flint glass myopia progression in children,10 the under-
Scott A Read PhD inset of higher refractive index embedded lying optical mechanism remains unclear. Nti
Contact Lens and Visual Optics Laboratory, School of within a crown glass spectacle lens.4 While our and Berntsen11 review the optics of modern
Optometry and Vision Science, Queensland University scope of practice has expanded substantially overnight reverse geometry orthokeratology
of Technology, Brisbane, Australia over the past century from sight testing opti- lens designs including their effects on accom-
E-mail: [email protected] cians to primary health-care providers, optics modation, peripheral refraction and on axis
remains at the heart of the optometric profes- higher order aberrations. Lau et al.12 also
sion. In this special issue of Clinical and Experi- report on the change in the higher order
Submitted: 7 October 2019 mental Optometry, international leaders aberration profile by modifying the Jessen
Accepted for publication: 8 October 2019 critically examine the current state of the field, factor in paediatric orthokeratology for myo-
encompassing recent advances in ophthalmic pia control.
and physiological optics, with an eye to the Given the well-documented increase in
future, beyond 2020. the prevalence of myopia over the past cen-
The inaugural issue of Clinical and Experi-
In contrast to the Kryptok of 1919, Jalie5 tury, understanding the optical effects of
mental Optometry (then The Commonwealth
(a name synonymous with ophthalmic potential interventions to slow myopia
Optometrist), ran to a grand total of 12 pages.
optics) reviews modern spectacle lens progression and axial eye growth in children
While a modest beginning, this first publica-
designs including free-form manufacturing is a current global research priority.
tion provides important insights into the
techniques to minimise the visual impact of Chakraborty et al.13 provide a detailed over-
state of our profession in the early twentieth
spectacle lens aberrations. Carkeet6 also view of the animal model literature examin-
century. For example, the first issue publi-
examines the optics of stand magnifiers, a ing how optical factors influence eye
shed in March 1919 included a brief note,1
commonly prescribed low vision aid despite growth. The contributions of this body of
directed at optometrists, explaining the cor-
significant advances in electronic devices in work to the current understanding of how
rect pronunciation of the word ‘Optometry’;
recent years.7 A novel method utilising digi- visual experience influences myopia devel-
‘…emphasis should be put on the second sylla-
tal photography to determine the equivalent opment and progression, and the transla-
ble ‘tom’… pronunciation should be universal,
viewing distance of stand magnifiers is also tion into interventions to control myopia in
so as not to confuse the public.’
described. the human eye are discussed. Hughes
In an era of restricted scope of practice,
prior to gaining access to diagnostic and thera-
In recent years, a range of new contact et al.14 examine the potential role of higher
lens designs have emerged for the correction order aberrations upon eye growth with
peutic agents or even legislation to protect the
of presbyopia, irregular corneal astigmatism, respect to myopia development and control
public and define the profession,2 it is perhaps
and for the control of childhood myopia. Kol- through both optical and pharmaceutical
not surprising that optical matters pre-
dominated this first issue of the journal. lbaum and Bradley8 tackle the often clinically interventions in humans.
challenging task of providing clear vision over In parallel with technological advances in
Included under the banner heading ‘Visual
a range of vergence demands in presbyopic manufacturing, our understanding of the natu-
Optics’ was an overview of clinical techniques
patients and examine the strengths and limi- ral optics of the eye and the visual effects of
employed during subjective refraction,3 still
tations of the different strategies used to optical corrections has continued to evolve.
utilised to this day, such as gradual fogging to
generate multifocal optics. Another techni- Romashchenko et al.15 analyse data from over
relax accommodation in the latent hyperope.
cally challenging optical correction in contact 2,000 eyes and describe in detail the change in
Here the author emphasised the importance
of practitioners having a clear understanding lens practice is minimising the visual refraction, higher order aberrations, and image
of both the qualitative and quantitative effects sequellae of elevated higher order aberra- quality across the visual field. Del Aguila-
of imposed defocus upon the visual system: tions in keratoconic eyes despite optimal Carrasco et al.16 review work examining the
‘We must know what effects should result from standard contact lens correction. Jinabhai9 changes in higher order aberrations associated
the application of certain lenses under certain reviews experimental approaches and com- with accommodation, and reciprocally, the
conditions and check up our diagnosis thereby.’ mercially available customised contact lens changes in the accommodation response
Another section titled ‘Practical Optics’ solutions to address this issue including when specific higher order aberrations are
included an homage of sorts to the ‘Kryptok’a – customised wavefront-guided soft and scleral manipulated. Cufflin and Mallen17 also investi-
contact lenses. gate the adaptive changes in the visual system
a
A derivation of the Greek “Kryptos” meaning hid- While it is now well accepted that overnight in response to imposed blur. These papers are
den or secret. orthokeratology treatment significantly slows particularly relevant to optical interventions
© 2020 Optometry Australia Clinical and Experimental Optometry 103.1 January 2020
1
Beyond 2020 Vincent and Read
designed to manipulate the visual experience correct simple ametropia, and in the case of 8. Kollbaum PS, Bradley A. Correction of presbyopia: old
problems with old (and new) solutions. Clin Exp Optom
and slow eye growth in children which alter childhood myopia, control its progression. 2020; 103: 21–30.
peripheral refraction, higher order aberrations, New contact lens technologies are also 9. Jinabhai AN. Customised aberration-controlling cor-
improving visual outcomes for patients with rections for keratoconic patients using contact lenses.
visual quality and potentially the accommoda- Clin Exp Optom 2020; 103: 31–43.
irregular corneal astigmatism, and adaptive 10. Cho P, Tan Q. Myopia and orthokeratology for myopia
tion response.
optics continues to expand our knowledge control. Clin Exp Optom 2019; 102: 364–377.
While a century ago, the examination of 11. Nti AN, Berntsen DA. Optical changes and visual per-
of the visual system, with the potential for
the posterior eye was exclusively the remit formance with orthokeratology. Clin Exp Optom 2020;
more widespread clinical impact in years to 103: 44–54.
of the ocularist (ophthalmologist), significant
come. In 2020 and beyond, visual optics will 12. Lau JK, Vincent SJ, Cheung SW et al. The influence
developments in ocular imaging now allow of orthokeratology compression factor on ocular
no doubt remain the cornerstone of the
vision scientists to examine the retina non- higher-order aberrations. Clin Exp Optom 2020; 103:
optometric profession. 123–128.
invasively with exquisite detail, to the level 13. Chakraborty R, Ostrin LA, Benavente-Perez A et al.
of individual photoreceptors. Bedggood and Optical mechanisms regulating emmetropisation and
REFERENCES
Metha18 describe the current state of adap- 1. Optometry. Commonwealth Optom 1919; 1: 6.
refractive errors: evidence from animal models. Clin
Exp Optom 2020; 103: 55–67.
tive optics imaging to visualise both the 2. Vincent SJ. Sydney barber Josiah Skertchly
14. Hughes RPJ, Vincent SJ, Read SA et al. Higher order
structure and function of the microvascula- (1850-1926): scientist, educator and advocate for
aberrations, refractive error development and myopia
Queensland optometry. Clin Exp Optom 2017; 100:
ture of the retina, and how this technology control: a review. Clin Exp Optom 2020; 103: 68–85.
402–406.
15. Romashchenko D, Rosen R, Lundstrom L. Peripheral
can influence clinical practice for a range of 3. Cumberland JK. Subjective sight-testing. Common-
refraction and higher order aberrations. Clin Exp
wealth Optom 1919; 1: 7–8.
systemic diseases such as diabetes, stroke, Optom 2020; 103: 86–94.
4. The Kryptok. Commonwealth Optom 1919; 1: 5–6.
16. Del Aguila-Carrasco AJ, Kruger PB, Lara F et al. Aberra-
and dementia. 5. Jalie M. Modern spectacle lens design. Clin Exp Optom
tions and accommodation. Clin Exp Optom 2020; 103:
Advances in our understanding of visual 2020; 103: 3–10.
95–103.
6. Carkeet A. Stand magnifiers for low vision: descrip-
optics and numerous research discoveries tion, prescription, assessment. Clin Exp Optom 2020;
17. Cufflin MP, Mallen EAH. Blur adaptation: clinical and
refractive considerations. Clin Exp Optom 2020; 103:
and developments have changed the prac- 103: 11–20.
104–111.
tice of optometry over the past century. 7. Chong MF, Jackson AJ, Wolffsohn JS et al. An update
18. Bedggood P, Metha A. Adaptive optics imaging of
on the characteristics of patients attending the
Practitioners now have a wide range of the retinal microvasculature. Clin Exp Optom 2020;
Kooyong low vision clinic. Clin Exp Optom 2016; 99:
103: 112–122.
state-of-the-art optical solutions available to 555–558.
Clinical and Experimental Optometry 103.1 January 2020 © 2020 Optometry Australia
2
INVITED REVIEW
Modern spectacle lens design
Mohammed Jalie DSc SMSA FBDO (Hons) Ophthalmic lens design concerns the control of spectacle lens aberrations which occur
SLD HonFCOptom HonFCGI MCIM when the eye rotates away from the optical centre of the lens. The most significant aberra-
Department of Biomedical Science, Ulster University, tions are oblique astigmatism and mean oblique error (power error). A brief review of these
Coleraine, Northern Ireland, UK aberrations is given, explaining how the lens designer can control them using just the bend-
E-mail: [email protected] ing of the lens, and what results can be achieved using simple spherical and toroidal sur-
faces. Before 1985, aspherical surfaces were used only for post-cataract spectacle lenses
and high-power magnifiers. Today, aspherical surfaces are used by all major lens manufac-
turers to produce thinner, lighter and more attractive best-form lenses in the normal power
range. Aspherical surfaces are employed because the surface itself is astigmatic and the
surface astigmatism is used to combat aberrational astigmatism due to oblique incidence.
The various types of aspherical surface and how the surface astigmatism arises is
described, before considering how this feature is used to produce flatter, thinner lenses. In
the case of astigmatic prescriptions, the surface requires different asphericities along its
principal meridians and the geometry of these atoroidal surfaces is also described. The
advent of free-form manufacturing techniques requires the lens designer to convert the sur-
face description to the (x,y,z) co-ordinates needed to generate the surface. Examples of how
these co-ordinates can be obtained from the equation to the surface are given for toroidal
and aspherical surfaces. In the case of free-form progressive surfaces, the pre-determined
Submitted: 28 March 2019 z-co-ordinates must be added to the z-co-ordinates of the prescription surface to obtain the
Revised: 29 April 2019 final free-form surface. In the case of optimised prescription surfaces, on-board software
Accepted for publication: 29 April 2019 will analyse the result by ray tracing to obtain the final z-co-ordinates.
Key words: aberrations, aspherical surfaces, atoroidal surfaces, oblique astigmatism, power error
When the eye is viewing along the optical spherical surface concentric with the centre Abbe number of the material denoted by
axis of a spectacle lens, the form of the lens of rotation of the eye upon which the far the V-value, defined as:
does not matter. The image formed by the point remains. This surface is referred to as
lens is not afflicted with any defects or aber- the far point sphere. The aim of spectacle V -value = ðnd – 1Þ=ðnF – nC Þ,
rations that might affect its sharpness or lens design for distance vision lenses is to
shape. However, in practice the eye turns enable the lens to form point images of where nd, nF and nC represent the refractive
behind the lens to view through off-axis distant point objects on the far point indices of the material for yellow, blue and
visual points and it is then that the lens sphere. In reality, the images are afflicted red light, respectively. In conditions of high
form assumes importance. Ideally, the off- with various aberrations and those which contrast, its effect is to cause coloured
axis performance of the lens should be the are of significance to the spectacle fringes to be seen surrounding the image of
same as its performance at the optical cen- wearer1–7 are: a high-contrast target.
tre. In general, this is not the case, with the • transverse chromatic aberration Under conditions of low contrast, colour
off-axis images being afflicted with various • oblique astigmatism fringing may not be noticed. Instead, the
aberrations which spoil the quality of the • curvature of field effect of TCA is to cause a reduction in visual
images formed by the lens. • distortion. acuity. This effect is referred to as off-axis
blur and often gives rise to the complaint by
Transverse chromatic aberration patients that ‘the lenses are fine when I look
Spectacle lens aberrations Transverse chromatic aberration (TCA) is through the centres but are blurred when I
due to the lens material. It is caused by the look through the edge!’
As the eye rotates about its centre of rota- fact that the refractive index of the lens To a good approximation, the magnitude
tion, the macula and the far point of the material varies with the wavelength of the of the TCA at any given point on a lens is
eye also rotate, the latter tracing out a incident light. This effect is expressed by the found by calculating the prismatic effect, P,
3
Modern spectacle lens design Jalie
at the point and dividing this by the Abbe forms for spectacle lenses,9 but the authors when a subject is given new lenses of a
number, V, that is, conclude that the resulting forms differ by different form.
negligible amounts from their third-order
TCA = P=V : [1] equivalents. In any case, the arrival of the
computer has rendered these approximate Field diagrams
It is generally considered that the thresh- analytical methods redundant.
old value for TCA is 0.1Δ. TCA less than 0.1Δ The effect of oblique astigmatism is to A ‘best-form’, or ‘corrected curve’, spectacle
is unlikely to give rise to complaints. The V- produce a blurring of the image as though lens is one with surface powers that have
value for normal-index materials in which an unwanted sphero-cylinder had been been specially computed to eliminate, or at
the refractive index is in the region of 1.50 interposed between the lens and the eye. least minimise, certain stated defects in its
(for example, crown glass and CR 39), is The reduction of oblique astigmatism is very image-forming properties. Of the four aberra-
about 60 and the prismatic effect at the important in the design of spectacle lenses tions described above, transverse chromatism
visual point would need to be about 6Δ and it will be seen later that this may be can only be eliminated by constructing an
before the threshold is reached. Using para- achieved by a suitable choice of lens bend- achromatic lens; that is, a pair of lenses
xial theory, this amount of prism would be ing or by the use of an aspherical surface. bonded together, in which the chromatism of
encountered at a point 15 mm from the one component neutralises the chromatism
optical centre of a +4.00 D lens. Materials in Curvature of field of the second. Such devices are too bulky to
which V-values are in the region of 40 would In the design of a camera lens, curvature of be used as spectacle lenses. Ordinarily, chro-
give rise to 0.1Δ of TCA at a point where the the image plane should be zero, allowing matism is a function of the Abbe number of
prismatic effect is 4Δ – for example, 10 mm the image plane to correspond with the flat the lens material and is minimised for a given
away from the optical centre of a +4.00 D film plane. In the case of a spectacle lens, power by selecting a material with the highest
lens. It is for this reason that it is wise to the curvature of the image is usually insuffi- available Abbe number.
select a material with the highest available cient to match the curvature of the far point For most people, the brain readily adapts
Abbe number. sphere. There is a mismatch between the to distortion and usually, this aberration is
axial power and the mean oblique power of an ongoing problem only in cases where
Oblique astigmatism the lens, known as mean oblique error or there has been a significant change in the
When a narrow pencil of rays is refracted power error. prescription, or in the lens form. It is possi-
obliquely by a spherical surface, the ble to reduce distortion and eliminate either
refracted pencil becomes astigmatic. Instead Distortion oblique astigmatism or mean oblique error
of the rays re-uniting in a single image point, Distortion affects the shape of the image by supplying lenses of very steeply curved
they form two-line foci at right angles to one rather than its sharpness and is caused by form, but such lenses are expensive to pro-
another and between them, where the the fact that the power of a spherical duce and appear very bulbous.
refracted pencil has its least cross-sectional surface increases toward its periphery, The two aberrations that remain – over
area – a disc of least confusion. The plane with the change in shape increasing as the which the designer can exert some influence
containing the optical axis of the surface is eye uses wider and wider zones of a – are oblique astigmatism and curvature of
referred to as the tangential plane and the spherical lens. field. A useful guide to the effects of oblique
plane at right angles to this is referred to as astigmatism and curvature of field in a given
the sagittal plane. The tangential and sagit- PINCUSHION DISTORTION spectacle lens is obtained by studying a field
tal oblique vertex sphere powers of a spec- Plus lenses produce pincushion distortion, diagram for the lens form. A field diagram
tacle lens are determined by accurate which is the type of distortion typically seen (Figure 1A) is a plot of the tangential and sag-
trigonometric ray tracing through the lens. when a strong plus lens is used as a magni- ittal oblique vertex sphere powers against
The chief ray passes through the centre of fier. The characteristic pincushion-shape the ocular rotation of the eye viewing
rotation of the eye which is supposed to lie image also gives the impression that the through the lens. In the case of a perfect lens
on the optical axis of the lens. It is also object being viewed is concave, whereby the – such as the +4.00 D design that has an
assumed that the primary line of sight coin- centre of the object seems further from the ideal field diagram illustrated in Figure 1A –
cides with the optical axis.5 When it does eye than the edges. the tangential and sagittal oblique vertex
not, as would be the case when the lens is sphere powers remain +4.00 D for all zones
tilted or decentred before the eye – for BARREL DISTORTION of the lens. Unfortunately, this performance
example, when a pantoscopic or face form Minus lenses produce barrel distortion and is impossible to obtain in a single lens with
tilt is applied to the lens – more complicated is often reported by myopes who view just two surfaces, at least for this power.
ray tracing techniques must be applied, through peripheral zones of their lenses. An The performance of a +4.00 D design
which invariably requires the use of com- object afflicted with barrel distortion gives made in plano-convex form is shown in
puter software to trace skew rays through rise to a convex appearance of the target; Figure 1B. When the eye views along the
the lens.5 the centre of the target appears to be closer optical axis of the lens, the power of the lens
Before the arrival of the computer, third- than the edges. is indeed +4.00 D. However, when the eye
order equations were employed to deter- When the form of a lens is changed, the rotates through 35 from the optical axis, the
mine the best form of spectacle lenses.8 amount of distortion that is exhibited by the real effect of the lens is +4.25 D in the sagit-
More recently, fifth-order equations have lens also changes and is probably the chief tal meridian and +5.75 D in the tangential
been published to determine the optimum cause of the perceptual problems that occur meridian. This effect can be expressed as
4
A Ocular rotation B Ocular rotation the back vertex power of the lens. Such a
form is described as a minimum tangential
40˚
T&S
40˚
S error form11 and is seen to suffer from an
ever-increasing amount of aberrational
+4.00 30˚ +4.25/+1.50 30˚ T astigmatism, albeit small, as the eye rotates
20˚ 20˚ away from the optical axis. The oblique
10˚ 10˚ astigmatic error amounts to about +0.25 D
35˚ 35˚
0˚ 0˚ at 35 and the blurring effect of this small
+3.0 +4.0 +5.0 +3.0 +4.0 +5.0
cylinder is certain to be less than the 0.25 D
Field diagram Field diagram
sphere blur found in the point-focal form
For 30˚ rotation the depicted in Figure 2A.
+4.00 Vertex sphere +4.00 Vertex sphere effect is +4.25/+1.50
In Figure 2C, the bending of the lens has
S = plot of sagittal powers been reduced still further to a +7.62 D base
T = plot of tangential powers
curve. It can be seen in the field diagram
that the tangential and sagittal oblique ver-
Figure 1. A: Field diagram for an ideal +4.00 D lens. The graph indicates that the tan-
tex sphere powers have increased to just
gential (T) and sagittal (S) oblique vertex sphere powers remain +4.00 D for all direc-
the point where the focal lines within the
tions of gaze. B: Field diagram for +4.00 D lens made in plano-convex form. The graph
eye would lie either side of, and equidistant
shows that the lens is afflicted with aberrational astigmatism which increases as the
from, the retina. At 35 the off-axis power of
eye rotates away from the optical axis of the lens.
the lens is +3.85 DS/+0.30 DC, the tangential
being equivalent to a power +4.25 D sphere Figure 2, which shows how the off-axis per- power is +4.15 D and the sagittal power
with a +1.50 D cylinder, and is so different formance of +4.00 D lenses varies for three +3.85 D, compared with the paraxial power
from the paraxial power that it cannot be curved forms with front curves +9.62 D, which is +4.00 DS. The mean oblique power
ignored. Clearly, a plano-convex design for a +8.12 D and +7.62 D. of the lens is +4.00 D. This form of lens is
lens of power +4.00 D is a poor choice. In Figure 2A, the lens has been bent into a known as a Percival lens design and is free
In general, the surface powers which are form where the oblique astigmatic error has from mean oblique error for the zone in
chosen for any given lens are those which been entirely eliminated. Such a form is question.
make the power obtained during oblique described as a point-focal lens form, from Today, most best-form lens series are
gaze as close as possible to the power the German word Punktal, which means designed to be free from tangential error
obtained when the eye looks along the opti- ‘point-forming’. This of course is the name when fitted at an average vertex distance.
cal axis of the lens.10 Although a +4.00 D still used by Carl Zeiss to describe their clas- When minimum T-error-form lenses are
lens for which the power remains the same sic series of point-focal lenses. fitted at a longer vertex distance than nor-
for all directions of gaze cannot be made, At 35 , the power of the lens has dropped mal, they tend to perform like point-focal
the performance shown in Figure 1B can to +3.70 D; that is, when the astigmatism is lenses. However, when fitted at a shorter
certainly be improved upon. fully corrected, the mean oblique power of vertex distance, that is, closer to the eye,
the lens changes by −0.30 D. The lens has a they tend to perform like Percival-form
mean oblique error at 35 of −0.30 D. lenses.
Best-form spectacle lenses If the form of the lens is flattened from The same principles are involved in the
the point-focal bending, the tangential design of minus spectacle lenses. The tan-
The control which the designer can exercise power increases. For the +8.12 D bending gential (F 0 T) and sagittal (F 0 S) oblique vertex
over these two aberrations is illustrated in depicted in Figure 2B, it is now the same as sphere powers are given in Tables 1–3 for
−4.00 D lenses made in point-focal form
(+4.70 D base curve), minimum tangential
A Ocular rotation B Ocular rotation C Ocular rotation error form (+3.70 D base curve) and
Percival-form (+3.25 D base curve) for ocular
T&S ST S T rotations out to 35 .
40˚ 40˚ 40˚
30˚ 30˚ 30˚

Aspherical surfaces
20˚ 20˚ 20˚
10˚ 10˚ 10˚ The term ‘aspherical surface’ usually refers

0˚ to a surface that is rotationally symmetrical
0˚ 0˚
+3.0 +4.0 +5.0 +3.0 +4.0 +5.0 +3.0 +4.0 +5.0 but at the same time, not spherical. The
Field diagram Field diagram Field diagram simplest aspherical surfaces are the con-
icoids, obtained by rotating a conic
Figure 2. Field diagrams for +4.00 D lens designs made in various best forms. A: Point- section about the z-axis. The conic sections
focal form, where oblique astigmatic error = 0. B: Minimum tangential error form, are illustrated in Figure 3A. They differ from
where F0 T = F0 V. C: Percival-form, where mean oblique power = F0 V and hence mean one another by their degree of asphericity,
oblique error = 0. which is denoted in the diagram by the
5
from the rotation of a conic section about

Ocular rotation F0 T F0 S OAE MOP MOE its z-axis (Figure 3A). The spherical surface is
35 −3.75 −3.78 +0.03 −3.77 +0.23 also a member of the family. An ellipse
30 −3.85 −3.85 0.00 −3.85 +0.15 rotated about its z-axis produces an ellip-
25 −3.92 −3.90 −0.02 −3.91 +0.09 soid. If the major axis of the ellipse is hori-
20 −3.96 −3.94 −0.02 −3.95 +0.05 zontal, the solid is referred to as a prolate
15 −3.98 −3.99 −0.01 −3.99 +0.01 ellipsoid. If the minor axis is horizontal, the
10 −3.99 −3.99 0.00 −3.99 +0.01 solid is referred to as an oblate ellipsoid.
A parabola rotated about the z-axis gener-
5 −4.00 −4.00 0.00 −4.00 0.00
ates a paraboloid and a hyperbola gener-
0 −4.00 −4.00 0.00 −4.00 0.00
ates a hyperboloid. These conicoidal
MOE: mean oblique error, MOP: mean oblique power, OAE: oblique astigmatic error.
surfaces are astigmatic at every point on the
surface except at the vertex, and their use
Table 1. Aberrational data for −4.00 D lens made in point-focal form, F1 = +4.70, enables the designer to make a lens of any
tC = 1.5 mm, n = 1.50, centre of rotation distance = 27 mm. All aberrations expressed power, in any bending, free from aberra-
in dioptres. tional astigmatism by choosing an aspheri-
cal surface the asphericity of which
neutralises the astigmatism of oblique
Ocular rotation F0 T F0 S OAE MOP MOE
incidence.
35 −3.95 −3.84 −0.11 −3.89 +0.11
The designer is not limited to conicoidal
30 −4.00 −3.90 −0.10 −3.95 +0.05
surfaces. If the equation for a conic
25 −4.02 −3.93 −0.09 −3.98 +0.02 section is written in the form:
20 −4.02 −3.96 −0.06 −3.99 +0.01
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
15 −4.02 −3.98 −0.04 −4.00 0.00
z = y 2 = r 0 þ ðr 0 2 −py 2 Þ [2]
10 −4.01 −3.99 −0.02 −4.00 0.00

5 −4.00 −4.00 0.00 −4.00 0.00
0 −4.00 −4.00 0.00 −4.00 0.00 where r0 represents the radius of curvature
MOE: mean oblique error, MOP: mean oblique power, OAE: oblique astigmatic error. of the surface at the vertex, and expanded
by the binomial theorem, the following
series is obtained:
Table 2. Aberrational data for −4.00 D lens made in Min. T-Error form, F1 = +3.70,
tC = 1.5 mm, n = 1.50, centre of rotation distance = 27 mm. All aberrations expressed z = Ay 2 þ By 4 þ Cy 6 þ Dy 8 þ … [3]
in dioptres.
where: A = 1/2r0, B = p/(22.2!r03), C = 3p2/23.3!
0 0 r05, D = 15p3/24.4!r07.
Ocular rotation FT FS OAE MOP MOE
35 −4.05 −3.88 −0.17 −3.97 +0.03 By altering the co-efficients A, B, C, D and
30 −4.08 −3.92 −0.16 −4.00 0.00 so on, the designer can configure the sur-
25 −4.07 −3.95 −0.12 −4.02 −0.02 faces to deform them from their pure con-
icoidal forms. These deformed conicoids are
20 −4.06 −3.97 −0.09 −4.02 −0.02
often referred to as polynomial or higher-
15 −4.04 −3.99 −0.05 −4.02 −0.02
order aspherical surfaces.
10 −4.02 −3.99 −0.03 −4.01 −0.01
5 −4.00 −4.00 0.00 −4.00 0.00
0 −4.00 −4.00 0.00 −4.00 0.00 Aspheric lenses for the
MOE: mean oblique error, MOP: mean oblique power, OAE: oblique astigmatic error. correction of aphakia
The field diagrams illustrated in Figure 4

Table 3. Aberrational data for −4.00 D lens made in Percival-form, F1 = +3.25,
show the off-axis performances of +12.00 D
tC = 1.5 mm, n = 1.50, centre of rotation distance = 27 mm. All aberrations expressed
lenses made in various forms. Figure 4A
in dioptres.
illustrates the field diagram for a form that
employs spherical surfaces. The increase in
p-value for each member of the family. K ðor QÞ = ðp – 1Þ: tangential power of the lens and the large
Other methods for expressing the amount of aberrational astigmatism are
asphericity include the eccentricity, e, which A typical aspherical surface is the ellipsoid seen in the diagram. The sagittal power
is related to p by the equation: illustrated in Figure 3B, which would be gen- remains about +12.00 D, but the tangential
erated by an ellipse rotating about its major power increases, reaching about +14.00 D at
e = √ ð1 – pÞ diameter. The ellipsoid is one member of a 35 from the optical axis. At 35 , the real
family of aspherical surfaces known collec- effect of this form with spherical surfaces is
or by K (or Q), where: tively as the conicoids, since they result +11.91 D with a +2.00 D cylinder – not the
6
A y B both the tangential and the sagittal radii of

Hyperbola p < 0 curvature for the surface increase, with the
Parabola p = 0 B tangential radius changing at a faster rate
Prolate ellipse P than the sagittal radius. Inspection of the
p>0<1 CSP field diagram in Figure 4B confirms that this
is just what is required to combat the aber-
Oblate
ellipse A rational astigmatism for this form of lens –
z a greater decrease in the tangential power
p>1 C0 CSB
Circle of the lens.
p=1 r0 By careful choice of eccentricity for the
ellipsoid it is possible to eliminate oblique
astigmatism for wide zones of the lens.
CTP
Aspheric lenses of the type needed for the
correction of aphakia usually employ a con-
vex prolate ellipsoidal surface to eliminate
CTB aberrational astigmatism in the post-
cataract range of prescriptions.
Figure 3. A: The conic sections. B: How an ellipsoidal surface corrects aberrational The improvement in off-axis performance
astigmatism. A is the vertex of the curve. C0 is the centre of curvature of the surface can be judged from the field diagram shown in
at the vertex. AC0 is the radius of curvature of the surface at the vertex, r0. P is a Figure 4B, which illustrates the zonal variation
point on the curve. PCTP is the radius of curvature of the surface at point P in the tan- in oblique vertex sphere powers for a point-
gential meridian, which is the plane of the diagram. CTP lies on the evolute, C0CTB, focal +12.00 D lens made with a −3.00 D back
which is the locus of the tangential centres of curvature of the surface between curve and a suitably chosen ellipsoidal front
points A and B. PCSP is the radius of curvature of the surface at point P in the sagittal surface, whose p-value is +0.65. It can be seen
meridian, which lies at right angles to the plane of the diagram. CSP lies on the evo- for this design that the tangential and sagittal
lute, C0CSB, which is the locus of the sagittal centres of curvature of the surface oblique vertex sphere powers remain the
between points A and B. same for all zones out to 40 , but the lens per-
formance is by no means perfect. The mean
oblique power, which now is the same as the
+12.00 D sphere intended. At this point, the One of the simplest surfaces to provide
tangential and sagittal oblique vertex sphere
lens exhibits 2.00 D of unwanted the correct variation in neutralising astigma-
powers, drops off rapidly as the eye rotates
astigmatism. tism is the ellipsoid. It is easy to see how
away from the optical axis of the lens.
When the designer is not limited to the this surface introduces neutralising astigma-
This loss in power, the mean oblique error,
use of spherical surfaces, oblique astigmatism by considering how the surface alters
amounts to almost 1.00 D at 35 from the
tism can be eliminated to provide an in shape as the eye rotates away from the
optical axis, but at least the error in off-axis
increase in the field of useful vision. This is pole of the curve. Figure 3B illustrates the
performance is a spherical one. It goes with-
achieved by employing a surface which itself instantaneous centres of curvature for the
out saying that, ideally, the designer would
is astigmatic, with the surface astigmatism point, P, on the surface of a convex prolate
like the marginal power of the aspheric
varying in just the right way to counteract ellipsoidal surface. The evolutes for the
design to increase in order to provide a con-
the astigmatism of oblique incidence. section AB are also shown and it is seen that
stant correction for all zones of the lens.
The large drop in tangential power does
provide the advantage for lens powers in
A Ocular rotation B Ocular rotation C Ocular rotation
this range, in that there is a reduction in dis-
tortion compared with the spherical design.
S T&S S
40˚ 40˚ 40˚ Figure 4C illustrates the zonal variation in
T T
30˚ 30˚ 30˚ oblique vertex sphere powers for a +12.00 D
lens made with a convex polynomial surface.
20˚ 20˚ 20˚ The astigmatism is zero out to about 20 ,
10˚ 10˚ 10˚ then increases slightly before dropping to
zero at about 32 . Beyond 35 , there is a
0˚ 0˚ 0˚
+10.0 +12.0 +14.0 +10.0 +12.0 +14.0 +10.0 +12.0 +14.0 rapid drop in tangential power as the surface
Field diagram Field diagram Field diagram becomes concave in this zone of the lens.
Figure 4. Field diagrams for +12.00 D lenses made in various forms. A: +12.00 D lens
made with spherical surfaces. Note that for a 20 rotation of the eye, the effective Aspheric lenses for the normal
prescription is +12.00/+0.56 and at 30 , the effective prescription is +11.93/+1.37. B: power range
+12.00 D lens made with convex prolate ellipsoidal surface. At 30 the effective pre-
scription is +11.33 DS. C: +12.00 D lens made with convex polynomial surface. At 30 Aspherical surfaces are now employed for
the effective prescription is +11.55/+0.09. lenses of low power, as required for the
7
This expression enables the correct

Refractive Front Form of front Back Centre asphericity to be found to produce a best-
index curve curve curve thickness
form lens for any chosen bending.
1.50 +8.25 D Sphere −4.45 D 4.4 mm Table 5 gives the off-axis performance of
1.50 +5.42 D Hyperboloid −1.50 D 3.9 mm the second +4.00 D lens described in
p = −1.4
Table 4, made with a −1.50 D back surface
1.50 +3.96 D Hyperboloid 0.00 3.7 mm power and a convex aspherical surface
p = −10.5
whose p-value has been chosen to eliminate
tangential error. This form has a very small
Table 4. +4.00 D lenses made in CR 39 in various forms at 60 mm diameter with degree of aberrational astigmatism. The sur-
0.5 mm edge thickness face is a convex hyperboloid whose p-value
is −1.4.
An even greater saving in thickness is
obtained when a higher refractive index
Ocular rotation F0 T F0 S OAE MOP MOE
material is employed. If the same base
35 +3.94 +3.79 +0.15 +3.86 −0.14 curve is used the saving is two-fold. First,
30 +4.00 +3.86 +0.14 +3.93 −0.07 there is the obvious reduction in the sags of
25 +4.02 +3.91 +0.11 +3.97 −0.03 the curves, since longer radii of curvature
20 +4.02 +3.95 +0.07 +3.98 −0.02 are employed. Second, since the use of the
15 +4.02 +3.97 +0.05 +3.99 −0.01 same power base curve on a higher refrac-
10 +4.01 +3.99 +0.02 +4.00 0.00 tive index material requires a longer radius
5 +4.00 +4.00 0.00 +4.00 0.00 of curvature at the vertex, the lens is flatter
0 +4.00 +4.00 0.00 +4.00 0.00 still and requires greater asphericity for the
surface to restore the off-axis performance.
MOE: mean oblique error, MOP: mean oblique power, OAE: oblique astigmatic error.
This is illustrated in Table 6, which shows
how the centre thickness of 60 mm diame-
Table 5. Aberrational data for +4.00 D lens made in Min. T-Error form, F2 = −1.50, tC = ter +4.00 D lenses would reduce when made
3.9 mm, n = 1.50, centre of rotation distance = 27 mm. All aberrations expressed in in 1.60, 1.67 and 1.74 index materials. The
dioptres. asphericity of the convex surfaces indicated
in the figure has been chosen to provide the
same off-axis performance for each lens.
Another important advantage of these
usual range of prescriptions.12 The use of astigmatism, an aspherical surface can be low-power aspheric designs for hypermetro-
aspheric forms for the low to medium employed, the form of which is such that it pia can be gleaned from Table 6. They lend
power range allows the production of thin- introduces negative surface astigmatism to themselves far better to a system of supply
ner and lighter lenses for this range of pre- neutralise the increase in tangential power, of large diameter plus uncut lenses, which
scriptions. The reduction in thickness is the or if required, produce a point-focal lens. are subsequently edged to smaller diame-
result of a two-stage process. First, the lens If the error in tangential power, δT, is ters depending upon the choice of shape
is made much flatter in form by employing known for a given incidence height, y1, at and size of the lens.
a shallower base curve. the front surface of radius, r1 and power F1,
The mechanical details of a series of +4.00 then the necessary asphericity, p, to elimi-
D uncut lenses made in CR 39 material at nate the tangential error can be found Aspheric lenses for myopia
60 mm diameter, all with an edge thickness from:13
of 0.5 mm, is shown in Table 4. Note that The principle of flattening a curved lens
the first form is a standard CR 39 lens made p = 1 þ ðr 1 =y 1 Þ2 :f1 − ½F 1 =ðF 1 − δT Þ2=3 g [4] form to make it thinner and then
with spherical surfaces. The second form is
a typical CR 39 aspheric lens with a shallow
inside curve and the third form is plano-
convex in form with the tangential error Refractive Front Form of front Back Centre
index curve curve curve thickness
neutralised by the hyperboloidal front sur-
face. It can be seen that by flattening and 1.60 +5.45 D Hyperboloid −1.50 D 3.2 mm
p = −3.35
aspherising the lens form, a saving in centre
1.67 +5.45 D Hyperboloid −1.50 D 2.9 mm
thickness is obtained. The flatter the lens, p = −5.0
the thinner it becomes.
1.74 +5.45 D Hyperboloid −1.50 D 2.7 mm
When spherical surfaces are used for p = −6.9
flatter-form lenses, aberrational astigma-
tism arises when the eye views through off-
axis portions of the lens. To eliminate the Table 6. +4.00 D lenses made in various materials at 60 mm diameter with 0.5 mm
tangential error and greatly reduce the edge thickness
8
aspherising one surface to restore the off- This ideal is difficult to achieve with the The equation to a toroidal surface is:
axis performance of the flatter-form lens usual form of toroidal surface where it is
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
can be applied equally to minus lenses. impossible to optimise each principal merid-
x 2 þ y 2 þ z2 −2r E z þ 2ðr E – r T Þ r T − r T 2 −y 2
In the days when the concave surface of a ian and maintain the correct cylindrical
spectacle lens incorporated any cylindrical effect for all zones of the lens. Most [6]
correction, aspheric minus lens series origi- designers agree that a best-form astigmatic
nally employed a convex aspherical surface, lens is one where the cylinder power is the where rE is the equatorial radius of curva-
the purpose of which was to increase the same for each principal meridian even ture and rT is the transverse radius of curva-
convexity of the front surface toward the though the power is not maintained along ture of the surface. Again, solving for z for
edge of the lens. With the advent of free- these meridians. In practice, this is best any given (x,y) co-ordinates enables the on-
form production techniques, concave achieved for plus astigmatic lenses when board computer to produce the surface.
aspherical surfaces designed to flatten the the convex surface is toroidal and of barrel The form in which the information should
surface toward its edge have become the form. For minus astigmatic lenses, the cylin- be presented to the generator is described
norm. In the case of astigmatic prescrip- der should be included on the concave sur- in the Vision Council Data communication
tions, concave atoroidal surfaces may be face. Once again, the barrel form is standard, entry 5.6.3.1, SURFMT.
employed. preferred.
In cases of very high myopia, the principle Free-form production techniques allow
of blending has been applied to the the asphericity to be varied between the Progressive lenses
workshop-flattened lenticular, to produce a principal meridians of the lens, resulting in
blended concave lenticular with a truly invis- an atoroidal surface and atoric lenses (one The first commercially successful progressive
ible dividing line. surface being atoroidal), which offer the power lens was developed by Bernard Mai-
These blended lenticulars for myopia, best solution for astigmatic prescriptions.14 tenaz of Essel, one of the founding members of
such as the Wrobel Super-lenti and the An isometric view of a concave atoroidal Essilor International, and introduced in Europe
Rodenstock Lentilux designs, enjoy excellent surface is illustrated in Figure 5. When used under the trade name Varilux in 1959.15 It was
cosmetic properties and allow very high with the flatter forms which are employed pointed out in 1963 by the German scientist
minus prescriptions, even in excess of for low-power aspheric lenses, a concave Minkwitz16 that a surface in which power
−20.00 D, to be dispensed in relatively thin atoroidal surface may be combined with an increased across its face must also have an
and lightweight form. Several manufac- aspherical front surface when the design astigmatic effect, whereby the astigmatism
turers now offer these designs under vari- becomes bi-aspheric in form. In addition to increases at twice the rate of change in power
ous trade names. reducing the cylinder error, this form of con- and lies at right angles to the direction of the
struction produces a small saving in thick- change in power. Since that time, the goal of
ness. However, there is little advantage to progressive lens design has been to minimise
Astigmatic lenses be gained by using a bi-aspheric construc- the effects of so-called Minkwitz astigmatism.
tion for purely spherical prescriptions. This In 1995 a patent was taken out by Kelch
A best-form astigmatic lens would have the is because – in the case of point-focal lenses et al.17 describing a progressive lens which
correct principal powers and an equal cylin- – although there is a slight decrease in had a convex progressive power surface and
drical effect along each meridian of the lens. mean oblique error, this is accompanied by an aspherical (for spherical prescriptions) or
a slight increase in distortion and the saving atoroidal (for astigmatic prescriptions) con-
in thickness is negligible. cave surface. This strategy optimised the
design for the required prescription in that
the design criterion is restored to the lens,
Free-form surface description no matter what prescription or cylinder axis
Elliptical direction is prescribed.
section p = +5 When a surface is to be produced by means Before the advent of free-form production
of computer numerical control machining, techniques, one convex base curve with its
the surface curves must be translated to given near addition was expected to be
three-dimensional co-ordinates (x,y,z). This used for powers covering a range of per-
Elliptical is a simple matter for the surfaces consid- haps four dioptres or so in spherical power
section p = +30 ered so far, for which equations would be and also cylinders in 0.25 intervals up to
known to the designer. For example, an 4.00 D with any prescribed axis direction.
aspherical surface whose p-value is known, The progressive surface would have been
is described by the equation: optimised for a single spherical power in
the middle of this range. If the prescription
deviated from the power for which the sur-
x 2 þ y 2 þ pz2 − 2r 0 z = 0 [5] face was designed, the symmetry of the iso-
cylinder lines would be destroyed and they
Figure 5. Concave atoroidal surface in and solving for z for any given (x,y) co- would then encroach upon the areas which
which principal meridians are oblate ordinates enables the on-board computer were supposed to provide clear vision. The
ellipses to generate the z-heights for the surface. Kelch patent described a method whereby
9
the combination of the progressive surface the practitioner stipulates which area of the 10. Fry GA. Choosing the base curve for an ophthalmic
lens. Am J Optom Physiol Opt 1978; 55: 238–248.
and the aspherical concave prescription sur- lens should be optimised for various occu- 11. Davis JK, Fernald HG, Rayner AW. An analysis of oph-
face restored the iso-cylinder lines to their pations.19 Also, a progressive design with thalmic lens design. Am J Optom Arch Am Acad Optom
previous ideal design position. the upper portion devoted to distance vision 1964; 41: 400–421.
12. Jalie M, inventor, Ophthalmic Spectacle Lenses Having
The modern method used by most major and the lower portion to intermediate a Hyperbolic Surface. United States patent US
lens manufacturers to produce free-form vision20 is claimed to be ideal for use at a 4289387. 1981 Sep 15.
13. Davis JK, Fernald HG, Inventors; American Optical Cor-
progressive surfaces is to provide a data- computer workstation.
poration, assignee. Ophthalmic Lens Series. United
base containing the (x,y,z) co-ordinates for Prism-controlled designs have been States patent US 3960442. 1976 Jun 1.
the initial design of a progressive surface. described which incorporate – in addition to 14. Jalie M, Inventor; Opticorp Inc., Assignee. Aspheric
Lenses. United States patent US 5083859. 1992
Then, the z-co-ordinates for the prescription an increase in power from distance to near Jan 28.
surface are computed for the same values – an increasing horizontal prism as the eye 15. Cretin-Maitenaz B, Inventor; Societe Industrielle et
of x and y and simply added to the z-co- rotates down from distance to near.21 With commerciale des Ouvriers Lunetiers, assignee.
Multifocal Lens Having a Locally Variable Power.
ordinates for the progressive surface. For the advent of free-form production tech- United States patent US 2869422. 1959 Jan 20.
their flagship designs, an optimisation ray niques, it is possible to supply progressive 16. Minkwitz G. On the surface astigmatism of a fixed
symmetrical aspheric surface. Opt Acta (Lond) 1963;
tracing routine may be run to ensure that designs with different cylinder powers
10: 223–227.
the finished lens performs in the manner and/or axes in the distance and near por- 17. Kelch G, Lahres H, Wietschorke H, Inventors; Carl-
which the designer intended.18 tions.22 Another interesting progressive Zeiss-Stiftung, assignee. Spectacle Lens. United States
patent US 544503. 1995 Aug 22.
In recent years, a series of low-addition pro- design has an additional intermediate por- 18. Hof A, Hanssen A, Inventors; Carl-Zeiss-Stiftung,
gressive power lenses has been introduced, tion below the near portion to enable sub- Assignee. A Spectacle Lens with Spherical Front Side
mainly for intermediate and near vision use, jects to clearly see steps and the ground at and Multifocal Back Side and Process for its Produc-
tion. United States patent US 6089713. 2000 Jul 18.
when only some 50 per cent of the full near their feet.23 19. Baumbach P, Esser G, Mueller W, et al., Inventors;
addition is required.16 These so-called degres- Optische Werke G Rodenstock (DE) Assignee. Method
of Manufacturing Ophthalmic Lenses. United States
sive lenses are especially useful when worn at
REFERENCES patent US 6685316 B2, 2004 Feb 3.
the computer where, mainly, only mid- 1. Henker O. Introduction to the Theory of Spectacles. 20. Dorsch R, Haimerle W, Inventors; Rodenstock GmbH
distance and near vision is required. Some Jena: Jena Optikerschule, 1924. Assignee. Progressive Spectacle Lens For Seeing
2. Emsley HH, Swaine W. Ophthalmic Lenses, 6th Objects at a Large or Average Distance. United States
manufacturers promote very low-addition ed. London: Hatton Press, 1961. patent US 7033022 B2. 2006 Apr 25.
progressive power lenses for use by pre-pres- 3. Emsley HH. Aberrations of Thin Lenses. London: Con- 21. Poulain I, Drobe B, Haro C, et al., Inventors; Essilor
byopes, suggesting that they may help to stable, 1956. International (Companie Generale d’Optique)
4. Kingslake R. Lens Design Fundamentals. London: Aca- Assignee. Ophthalmic Lens with Progressive Addition
relieve computer vision syndrome at the demic Press Inc, 1978. of Power and Prism. United States patent US 7216977
workstation. Needless to say, because the 5. Jalie M. The Principles of Ophthalmic Lenses, 5th B2. 2007 May 15.
ed. London: ABDO, 2016. 22. Donetti B, Petignaud C, Hernandez M, Inventors;
addition is very small, and the corridor length
6. Fannin TE, Grosvenor T. Clinical Optics. Boston: But- Essilor International (Companie Generale d’Optique)
so long, the Minkwitz astigmatism exhibited terworths, 1987. Assignee. Method for Determination of an Ophthal-
by these lenses is very small, with the designs 7. Atchison DA, Smith G. Optics of the Human Eye. mic Lens Using an Astigmatic Prescription for Far
Edinburgh: Butterworths, 2000. Sight and for Near Sight. United States Patent US
offering wide fields of clear vision. 8. Atchison DA. Third-order theory and aspheric specta- 7249850 B2. 2007 Jul 31.
Several patents have also appeared for cle lens design. Ophthalmic Physiol Opt 1984; 4: 23. Giraudet G, Poulain I, Inventors; Essilor International
vocational progressive designs which have 179–186. (Companie Generale d’Optique) Assignee. Progressive
9. Miks A, Novak J. Fifth order theory of the astigma- Lens for Ophthalmic Spectacles having an Additional
not yet reached the market. For example, a tism of thin spectacle lenses. Optom Vis Sci 2011; 88: Zone for Intermediate Vision. United States Patent US
progressive design can be employed where 1369–1374. 8061838 B2. 2011 Nov 22.
10
INVITED REVIEW
Stand magnifiers for low vision: description, prescription,

assessment
Andrew Carkeet BAppSci (Optom) MSc Stand magnifiers are still one of the most commonly prescribed classes of low vision
BA PhD devices. Their performance can be difficult to understand because stand magnifiers usually
School of Optometry and Vision Science, and Institute do not give an image at infinity. This review summarises the methods of describing image
of Health and Biomedical Innovation, Queensland enlargement for stand magnifiers, emphasising their relationship to equivalent viewing dis-
University of Technology, Brisbane, Queensland, tance (EVD). This is done in terms of the underlying optical equations, and measurement
Australia methods, and methods of prescribing. In the past, methods of determining EVD have been
E-mail: [email protected] somewhat indirect, requiring accurate measurement of lens power, and image position.
The use of digital photography provides an alternative, more direct, simpler method of
determining EVD, which can be accomplished in-office. This method is described and it is
demonstrated how it gives comparable results to older methods with small, clinically non-
meaningful differences, that may be due to differences in image distance reference planes.
Describing the performance of stand magnifiers in terms of their dioptric power, or in terms
of ‘nominal magnification’ or ‘trade magnification’, is imprecise and misleading. It is better
to use indices such as equivalent viewing power and EVD, which take into account the mag-
nifier dioptric power, the image position of the magnifier and the distance a patient is from
the magnifier. While EVD is a useful index for prescribing stand magnifiers, manufacturers
do not always provide sufficient technical details to determine EVD for their stand magni-
Submitted: 11 April 2019 fiers, and available tables of EVDs are more than a decade old and are likely to need
Revised: 3 July 2019 updating. Photographic comparison provides a method for determining EVD, and this
Accepted for publication: 3 July 2019 method can also be applied to other low vision devices.
Key words: equivalent viewing distance, low vision, magnification, stand magnifiers
Stand magnifiers are still commonly used speed with stand magnifiers in age-related Also, for a patient, the relative image
for low vision management. Although there macular degeneration patients.6 enlargement provided by the system will
is an increasing use of optoelectronic However, nearly all stand magnifiers form depend on the dioptric power of the magni-
devices, which may sometimes yield better images which are not located at infinity. If a fier, the exit vergence of the magnifier, and
reading performance,1,2 many patients still patient is emmetropic, or is wearing a dis- the distance from the eye of the patient to
find a use for stand magnifiers, because of tance spectacle correction, then additional the magnifier.
their low cost, portability, and ease of use. positive focusing power must be used with The optical principles behind stand mag-
In one tertiary hospital low vision clinic, the magnifier to form a clear image on the nifiers might seem complicated to apply in
stand magnifier prescription was relatively retina of the patient. This is either in the low vision practice, but there are concepts
stable over three decades until 2003, with form of accommodation, or if the patient is which make their prescription easier. This
the use of illuminated magnifiers becoming presbyopic (as most are), in the form of a paper will review the optical principles
more common. In a study of outreach low spectacle addition. underpinning stand magnifiers; the various
vision clinics in 2004, they were the most Practitioners, then, must consider stand clinical descriptions of image enlargement –
frequently prescribed near low vision device magnifier prescription in terms of providing in particular, equivalent viewing distance
for adult patients.3 A 2016 study in a paedi- an optical system comprised of the magni- (EVD), the accuracy of the specifications of
atric low vision population found that stand fier itself and the potential spectacle addi- the manufacturer for stand magnifiers, in-
magnifiers were the most commonly pre- tion along with the distance prescription of office methods of assessing clinical param-
scribed near low vision device.4 They are a patient. The power of the addition will be eters of stand magnifiers, and required
easy for patients to learn to use although determined by how far the image is formed near additions. Where possible, these
patients may sometimes have difficulty behind the magnifier, the distance from the topics will be distilled into clinically useful
scanning text with a stand magnifier5 and eye of the patient to the magnifier, and the prescribing principles for low vision stand
in-office practice may improve reading residual accommodation of the patient. magnifiers.
11
Stand magnifiers for low vision Carkeet
Optics of stand magnifiers If the patient is presbyopic and has no EID G −ℓ0
accommodation, then they will need a spec- EVD = = [4]
; ;
tacle addition to see the image which is a dis-
Figure 1 illustrates the optical principles
tance of G − ℓ0 from the eye, that is a For stand magnifiers, and the image dis-
behind stand magnifiers. The object is tance ℓ0 is often known with appropriate
spectacle addition of power AddMax (the
placed at the base of the stand magnifier, at precision, as is the power of the magnifier
maximum addition required if no accommo-
the object distance ℓ from the magnifier FSM. Lateral magnification can be calculated
dation is used) is indicated where:
lens. Its image will be formed at the image from those two values:
distance ℓ0 from the magnifier lens. Most
stand magnifiers are constructed so that the 1 1 L0
AddMax = = =− [3]
EID G− ℓ0 1−GL0 ℓ’ ℓ’ ℓ’ 1
image is closer than infinity and ℓ0 is nega- ;= = 1 = 1 = ℓ’ L’ − F SM = ℓ’ ’ −F SM
ℓ L ’ ℓ
tive. Useful relationships are given by stan- L −F SM
To summarise, a stand magnifier will pro-
dard paraxial refraction equations. Object [5]
duce an enlarged image, and that image will
vergence L = 1ℓ and image vergence will be be at a distance from the patient. Both those So for a given combination of ℓ , FSM and G 0
negative (divergent) with L0 = ℓ10 . The image factors, lateral magnification and image posi-
size 70 and object size 7 are related by the tion, need to be considered when assessing
G −ℓ0
the potential value to a patient of a stand EVD = [6]
lateral magnification equations: ℓ0 1
−F SM
magnifier. There are a number of clinical ℓ0
metrics that have been used to do this.

7 0 ℓ0 L EVD might appear difficult to calculate, but
;= = = [1]
7 ℓ L0 as seen below it is easy to determine with
EVD photographic methods, and provides a com-
(Note: for this, Fincham’s notation m for lateral
mon metric that can be applied to other
magnification7 has been used. Bailey et al.8
near magnifier types such as hand magni-
use ER, enlargement ratio, to describe lateral Probably the most useful of these metrics is
fiers, and electronic devices.
magnification, that is ; = ER, as do others.)9 EVD, described by Bailey et al.8,10 The EVD is
The stand magnifier dioptric power is defined as the distance at which an object
denoted as FSM, then would need to be from the unaided eye, to Other metrics for stand
subtend the same angle that it would when magnifiers
L0 = F SM + L [2] viewed through the magnifier. For example,
from the tables of Bailey et al.,8 if a COIL
Equivalent viewing power
The distance from the magnifier lens to #4210 (specified 10 x 36 D) stand magnifier
Equivalent viewing power (EVP)11–13 can be
the eye is G and the eye-image distance EID= is used with eye-lens distance G of 10 cm,
thought of as the spectacle addition
G − ℓ0 . the EVD is 3.8 cm. In other words when (or accommodation required) for an object
(Note: the subscript SM is used to denote viewing through the magnifier, objects will actually positioned at the EVD. It is the recip-
variables pertaining to the specific case of subtend the same angle as if the patient rocal of EVD, that is
stand magnifiers throughout. When variables had viewed, unaided, the object at 3.8 cm.
apply to magnifiers in general, for example EVD is the EID divided by the lateral magnifi-
1
hand magnifiers, the subscript M is used.)9 cation of the image: EVP = [7]
EVD
EVP can also be considered and calculated

as the equivalent power of a lens system
made up of the stand magnifier lens, and
the maximum addition (or required accom-
modation) separated by the eye-lens dis-
tance G. Thus, based on the equivalent
power equation:
EVP = F SM + AddMax −GF SM AddMax [8]
For the COIL #4210 used at z = 10 cm,

AddMax would be 2.08 D and EVP would
be 26.4 D.
If the lateral magnification m for the stand
magnifier is known then EVP can be calculated.
EVP = m × Add Max [9]
Figure 1. Object image relationship stand magnifier of power FSM. Object distance is Given the EVD and EVP are reciprocals of
ℓ, image distance ℓ0 , object height i, image height i0 and eye-lens distance z. each other, the two indices should be
12
similarly useful from a clinical point of view. Third, the use of power alone ignores M calculated by this method is sometimes
Tables using one method can be easily parameters which will affect the functional referred to as ‘trade magnification’.9,15 This
converted to the other. performance of a stand magnifier. For a equation is based on an arbitrary reference
stand magnifier, the lateral magnification distance of 25 cm. It assumes the exit ver-
Dioptric power will be constant, but as can be seen from gence from the magnifier is −4.00 D and
Most manufacturers will specify a dioptric Equation 6, EVD depends also on the image that eye-lens distance G = 0. These assump-
power for stand magnifiers. That power is distance ℓ0 from the lens and the eye to lens tions are almost never appropriate for stand
nearly always a misleading indicator of the distance, G. In the example above – the COIL magnifiers and so the equations are inap-
image enlarging properties of a stand #4210 – ℓ0 was measured by Bailey et al.8 as propriate for stand magnifiers.
magnifier. −38 cm. If a patient had positioned their eye For a given lens power, trade magnification
First, it is often inaccurate. For the example close to the lens (at G = 2.5 cm) the EVD gives a higher value, by one, than does nomi-
above – a COIL #4210 stand magnifier – the would be 3.2 cm. If a patient used the stand nal magnification. Perhaps for this reason
specified dioptric power was +36.00 D, but magnifier from further away (at G = 25 cm) trade magnification is more often used by
previous authors8 measured equivalent the EVD would be 5.0 cm. That represents a manufacturers to designate the enlargement
power of the magnifier as actually +30.70 D. change in performance of about 0.2 log properties of stand magnifiers. However,
This issue with inaccurate power specification units. If the reading threshold of a patient whether trade or nominal magnification is
given by the manufacturer is well-known and was N6 in the first position, it would be used to label a stand magnifier, they are
has been reported by a number of authors. expected to be N10 in the second eye-lens nearly always misleading and usually over-
Bailey et al.8 measured the equivalent power distance. This is something that cannot be state the magnification of the device.
of a series of 82 stand magnifiers of which predicted from stand magnifier power Brown et al.9 surveyed 66 stand magni-
40 had labelled dioptric powers designated alone. fiers, 48 of which were labelled with trade
by the manufacturer; of those, 27 had pow- magnification and 18 were labelled with
ers that differed by 0.50 D from the labelled Near ‘relative magnification’ nominal magnification. With G = 0, that is
powers. Of those there were three magnifiers There is a long-standing convention in which the eye coincident with the magnifier plane,
with labelled power underestimating actual near magnification is specified in terms of all of those labelled with nominal magnifica-
power, and for 24 magnifiers the manufac- the ratio of the angular subtense α0 of an tion underestimated the enlargement per-
turer overestimated the power of the magni- image viewed through a magnifier, to the formance of the device, because none had
fier. Of those, 21 were from the one angular subtense α25 of an object viewed at an image located at infinity. However, the
manufacturer – COIL – although it should be a standard distance, usually 25 cm image would need additional accommoda-
noted that 25 per cent of COIL magnifiers (or 10 inches).7 Bailey refers to this as ‘relation or a spectacle addition to be in focus.
(Equation 6) had labelled powers that mat- tive magnification’ (M), and this paper Of the 48 labelled with trade magnification,
ched actual powers to within 0.50 D. This adopts that use. Relative magnification is only nine had images acceptably close to
issue with COIL magnifiers had been usually denoted by an uppercase M to dis- 25 cm from the lens, 31 had images closer
observed by others. Chung and Johnston14 tinguish near relative magnification M,7 than 25 cm from the lens (and trade magni-
also noted differences between measured from lateral magnification m. fication would therefore underestimate per-
and manufacturer-specified powers for COIL For a magnifier (for example, a hand mag-
formance), and eight had images further
stand magnifiers, as did Bullimore and nifier) of power FM and a focal length f 0M than 25 cm from the lens (and trade magni-
Bailey,10 who found that manufacturer- with the object at its primary focal point,
specified powers were overstated in nearly fication would therefore overestimate per-
and the image at infinity, this equation can
all COIL stand magnifiers they measured, by formance). If a reasonable working distance
be written as:
between zero and 28 per cent. G of 25 cm was used for any device, any
Second, that power can be difficult to mea- trade or nominal magnification greater than
α0 0:25 F M
sure. Stand magnifier equations assume thin M= = 0 = [10] one overestimated performance.
α25 fM 4
lenses, but the stand magnifier lenses are However, one can accurately use EVD to
often thick enough so that assumption will calculate M for a stand magnifier image
This definition is sometimes referred to as
be problematic. While it is more appropriate (that is, MSM), compared with an object at
‘nominal magnification’.9,15
to specify equivalent lens power, Fe, for FSM, 25 cm. Because EVD is the distance that the
The equation is sometimes modified to
that power is referenced to principal planes, object would subtend the same angle that it
allow for some accommodation by the
which are imaginary planes associated with would when viewed through the magni-
patient, traditionally an amount of 4.00 D,
the lens. Fe needs to be determined indi- fier, then:
and to assume that the eye of the observer
rectly, for example using methods that mea-
is so close to the magnifier that working dis- 0:25
sure image magnification for known object MSM = [12]
tance G is approximately zero. Under those EVD
distances. Sometimes it is easier to measure
circumstances the magnifier can be consid-
back vertex power of the lens (F0 v or BVP), 0:25
ered as having a power of FM + 4.00 D. If that MSM = [13]
which is referenced to the back surface of G −ℓ0
is the case, then: ℓ ðℓ10 − F SM Þ
0
the lens, a real physical structure. Some
manufacturers may designate power in
ℓ0 1
ℓ0
− F SM
terms of BVP14 instead of the more appropri- FM + 4 FM MSM = [14]
M= = +1 [11] 4ðG −ℓ0 Þ
ate Fe. 4 4
13
Smith et al.16 have expressed this in terms hand magnifiers (which assumes an infinite
Using EVD for prescribing stand
of vergence exiting the magnifier lens image for the magnifier), as the equation for
magnifiers
(instead of image position): a stand magnifier. For a hand magnifier
held with the object at its primary focal
g ðF SM −L0 Þ As shown above, EVD is a useful measure-
point and an image at infinity12
MSM = [15] ment for describing image enlargement in
ð1−GL0 Þ
A low vision. It is a common metric that can
FoV M = [18] be calculated for stand magnifiers, near
where g = 0.25 m. FM × z
By way of example, the stand magnifier additions, hand magnifiers, and near elec-
There can be significant errors in estimates tronic magnifiers. Bailey et al.8 suggested the
previously discussed – the COIL #4210 – was
of field size if Equation 18 is used for a following steps for determining the appropri-
described by the manufacturer as having a
stand magnifier. ate EVD for a patient based on near tasks and
power of +36.00 D and a magnification M of
For the COIL #4210 mentioned above acuity.
10X, which comes from the definition
(FSM = +30.70 D, ℓ0 = −38 cm, with A = 36 mm), 1. Decide on a goal print size. This will be
M = F/4 + 1. In fact, the stand magnifier was
used at an eye-lens distance of G = 25 cm, then task-related. For example, a patient may
measured as having a power of 30.70 D, so
Equation 18 gives a field size of 4.7 mm. wish to read regular print books of N12.
applying the trade magnification equation However, the correct equation – Equation 17 Bailey et al.8 initially set this as the goal
M = F/4 + 1 would give an M of 34.7/4 = 8.7X. – (EVD of 5.0 cm) gives a field size of 7.2 mm, print size, because it was believed that
But actual MSM = 0.25/EVD. For an eye-lens so using the wrong equation significantly setting a goal close to threshold was opti-
distance G of 2.5 cm, the EVD was 3.2 cm underestimates field of view. If a close work- mal because it maximised the amount of
(0.032 m), so MSM = 0.25/0.032 = 7.8X. For ing distance, G = 2.5 cm, is used, then using text that could be fitted into the field.22
G = 10 cm, EVD was 3.8 cm (0.038 m) and the incorrect equation (Equation 18) gives a More recently it has been demonstrated
MSM = 6.6X. For G = 25 cm, EVD was 5.0 cm, field size of 47 mm, which slightly overesti- that increasing print size above threshold
so MSM = 5.0X. This is half the M specified mates the value given by the correct equa- improves reading rates in normal and
by the manufacturer (trade magnification). tion (Equation 17) of 46 mm. low vision patients.23–27 It is considered
Although 25 cm is the most commonly It should be noted that the calculations appropriate to modify this goal print size
used reference distance for industry calcula- for field of view (Equations 16–18) depend to allow for an acuity reserve so that the
tions of M, some authors use additional ref- patient is not working at threshold.
on the assumptions that thin lenses are
erence distances7 including 40 cm.7,17 Researchers28,29 have suggested that an
used, that pupil size is negligibly small, and
acuity reserve of two-fold would give flu-
Bailey17 differentiates near relative magnifi- that the full extent of the lens can be
ent reading, with larger acuity reserves
cation specified by different reference dis- used.8,14,19 The usable field of the magnifier,
giving higher reading rates, although
tances by using subscripts (that is, M0.25 or that can be seen clearly through the magni-
Legge et al. have suggested individual
M0.40) and shows magnification equations fier, may be different. This is because mag-
testing at different print sizes to establish
for the later reference distance. nifier lenses probably do not have negligible
fluency rates.30,31 Using an acuity reserve
In summary, manufacturers nearly always thickness, and are likely to be aspheric. Also,
of 2X goal print size would be N6 (N12/2).
describe their stand magnifiers in terms of for very close working distances, the base of
2. Test reading threshold at a known working
trade magnification. But irrespective of the magnifier may act as the field stop and
distance with an appropriate near correc-
whether trade magnification or nominal mag- limit how much text can be seen. In addi-
tion (the addition is to ensure optimal
nification is used, it almost always overesti- tion, there are at least two possible refer-
performance). For example, a reading
mates performance, especially for larger eye- ence planes from which the eye position
threshold was obtained of N18 at a work-
lens distances. can be judged when calculating field size.
ing distance of 40 cm with a 2.5 D addition.
Stationary field of view should be judged
3. Determine a required EVD as:
(and G calculated) with respect to the pupil
Stand magnifier field of view of the patient, approximately 3 mm20 Goal reading print size
EVD = Test distance
behind the corneal plane. For field of fixa- Threshold print size
The linear extent of a field that is visible tion, the field which can be viewed with a [19]
through a stand magnifier (FoVSM) with an preferred locus of fixation, G, of the patient
aperture A can be shown to be:18,19 should be calculated from the centre of
In this case EVD would be N6/N18 x
rotation of the eye, approximately 15 mm
40 cm = 13.3 cm.
A behind the cornea.21
FoV SM = [16] 4. Choose a device which gives that EVD at
EVP × G Chung and Johnston14,19 noted discrepancies
appropriate eye-lens distance. Tables of
between measured and calculated field sizes
appropriate optical devices with their
for very close observation distances where the EVDs can be found in Bailey et al.,8 Bul-
which can be written in terms of EVD:8
base of the magnifier limits the field, and mea- limore et al.,10 and Lovie-Kitchin and
A × EVD sured field could be as low as 17 per cent of Whittaker.28 Tables provided by Chung
FoV SM = [17] the calculated field. However, even for larger
G and Johnston14,19 use similar tables for
distances, it was not uncommon for calculated EVP. Table 1 is taken from Lovie-Kitchin
Some authors (for example, Wolffsohn13) field to overestimate and underestimate mea- and Whittaker28 (their table 5). Suppose
have erroneously reported the equation for sured field by 10–20 per cent. the patient wanted a device to use at
14
Manufacturer Description of manufacturer Measured Predict performance (z = eye-lens distance)

and ID# FSM z = 2.5 cm z = 10 cm z = 25 cm
D EVD Eye- EVD Eye- EVD Eye-
image (cm) image (cm) image (cm)
COIL 5213 8D/3X Hi-power aspheric tilt 7.9 9.0 24.9 11.7 32.4 17.1 47.4
Eschenbach 2050 10D/2.5X hand mag, converts to 9.0 7.1 12.9 11.2 20.4 19.4 35.4
stand
Eschenbach 2630 10D/3.5X aspheric, tilt, 9.4 7.6 20.8 10.4 28.3 15.9 43.3
75 x 50 mm
Schweizer 320/90 12D/3X, Visolette, paperweight, 10.8 5.4 9.4 9.7 16.9 18.3 31.9
90 mm
COIL 5214 12D/4X Hi-power aspheric tilt 10.8 8.4 69.5 9.3 77.0 11.2 92.0
Eschenbach 1420 Bright field, paperweight, 65 mm 15.6 4.1 7.1 8.5 14.6 17.3 29.6
65 mm
Eschenbach 1421 Bright field, paperweight, 65 mm 17.2 4.0 7.1 8.2 14.6 16.5 29.6
40 mm
COIL 4206 20D/6X Hi-power aspheric 18.7 5.1 51.6 5.8 59.1 7.3 74.1
Eschenbach 2626 23D/6X, aspheric 23.0 4.1 32.8 5.0 40.3 6.9 55.3
COIL 4208 28D/6X Hi-power aspheric 23.4 4.2 73.0 4.6 80.5 5.5 95.5
Peak 1961 40D/10X Loup 29.2 3.3 23.3 4.4 30.8 6.5 45.8
Eschenbach 1153 32D/8X aplanatic 28.9 3.4 44.5 4.0 52.0 5.1 67.0
COIL 4212 44D/12X Hi-power 35.1 3.0 22.5 4.0 30.0 6.1 45.0
COIL 4210 36D/10X Hi-power 30.7 3.2 40.5 3.8 48.0 5.0 63.0
COIL 4215 56D/15X Hi-power focusable 40.3 2.5 18.1 3.5 25.6 5.6 40.6
Eschenbach 2628 38D/10X, aspheric 37.1 2.7 44.2 3.1 51.7 4.0 66.7
Peak 1962 15X Loup 45.3 2.2 20.9 3.0 28.4 4.7 43.4
Peak 1964 22X Loup 64.9 1.7 10.1 3.0 17.6 5.5 32.6
COIL 4220 76D/20X Hi-power 53.6 1.9 15.9 2.9 23.4 4.7 38.4
†
For most of the higher-powered devices (for example EVD < 6 cm), an eye-lens distance of 25 cm is not practical because the field of
view is too small.
Table 1. Table of equivalent viewing distances (EVDs) for stand magnifiers taken from Lovie-Kitchin and Whittaker28
an eye-lens distance G of 10 cm. COIL 5213 to determine EVD and associated spectacle set-up for testing the stand magnifier). At
has an EVD of 11.7 (slightly better than additions from parameters provided by that large object distance, the small uncer-
the required EVD) and an EID of 32.4 cm. some manufacturers, or by careful mea- tainty in the location of the first principal
5. Trial the selected magnifier with the appro- surements of the stand magnifiers them- plane of the lens is insignificant
priate spectacle addition at the appropri- selves. At a minimum, a good estimate of (Figure 2A). The author uses two bright
ate eye-lens distance. In this case the EID FSM and image position ℓ0 can be used to LED pen-torches shone through pinholes
is 32.4 cm, which would correspond to an
determine EVD, but there are other (Figure 2A), positioned at an object distance
addition power of +3.00 D. If fluency is not
methods to determine EVD. Eschenbach −4 m. This is imaged on a suitable translu-
achieved at the appropriate print size
(N12) then other devices can be tried,
provides good detail about their stand mag- cent object. Bailey34 suggested using translu-
based on the patient’s actual performance. nifier products on their websites,33 including cent tape on a graticule for a loupe. The
FSM, L0 and lateral magnification m, which image size can be read off the loupe grati-
Prescribing magnification using EVD in make calculating EVD easy. But for other cule. The author uses tape on a ruler as the
this way is one of a number of similar companies, the data are more difficult to image screen and measures image size from
methods which can be based on the obtain; however, the EVD can be relatively pixels on a photograph (Figure 2C). The
patient’s near acuity.32
easily measured. image size 7t0 is carefully measured. From
Bailey34 described simple in-office mea- Equation 1 mt can be calculated and there-
surements which can be used to deter- fore l t0 can be calculated:
Measurement of EVD and image mine magnifier lens power, and the image
position position for the lens. To determine FSM, a 70t
= ;t and ℓ0t = ;t ℓt [1b] [1c]
;t
bright object of known size 7t is positioned
More recent tables of EVDs for low vision at a long-distance ℓt from the lens. (The The paraxial refraction equation can be
devices are difficult to obtain. It is possible subscript t is used to indicate that this is a used to determine FSM:
15
A spectacle addition at different working dis-

tances, but it is unlikely to give a precise
Object Magnifier Image enough estimate of image position ℓ0 to esti-
mate EVD using Equation 6.
Bailey37 later suggested an alternative,
more precise, method using a travelling
near-focusing telescope. The telescope is
first focused on the exit surface of the mag-
nifier, then moved until it is focused on the
image formed by the magnifier, and the dis-
tance through which the telescope is moved
provides an estimate of ℓ0 . This method has
been used to measure image position in
B
most subsequent studies of stand magnifier
parameters.8,10,14,19,35
Measuring EVD using

photographic comparison
Using information about stand magnifier

power FSM and image position ℓ0 is an indi-
C rect way to determine EVD. Recently Carkeet
et al.38 used a photographic technique to
measure EVD. This technique is illustrated in
Figure 3, showing measurements of an
Eschenbach System Vario Plus 155339 used
at eye-lens distance z of 25 cm. A digital
camera was used, but as an in-office tech-
nique, the resolution of a modern smart
phone camera or tablet camera will usually
be adequate. The camera should be
focused, set at infinity, because allowing the
camera to autofocus may change image size
in unknown ways.
For some smart phones and tablets, this
may require the use of an additional appli-
cation. The focus is adjusted for near
objects by placing different lenses of differ-
Figure 2. Bailey34 method for determining power of a stand magnifier. A: Optical set ent powers in front of the camera lens,
up with -ℓ > > ℓ0 . B: Object, two lights shone through pinhole. C: Image, shown by two simulating a spectacle addition. Although
arrows, formed on translucent tape on a ruler. the camera is placed close to the trial lens,
that trial lens is used as the reference
plane for the measurements, on the
F SM = L0t −Lt = ℓ10 − ℓ1t [2b] magnifier will provide an estimate of its exit
assumption that vergence exiting that lens
t
vergence. To improve this judgement, the
will be zero, and the distance from the trial
Variants of this method have been used in a observer can look through the stand magni- lens to the camera does not affect image
number of studies for determining stand fier and trial lens using a telescope focused size. Pilot work showed this to be true. The
magnifier dioptric power.8,10,11,14,19,35 for distance.36 trial lens power is adjusted until the image
To determine EVP and to determine the Alternatively, light from a distant object on the screen is clear. This can be done
appropriate spectacle addition at different can be refracted through the lens in reverse, reliably to the nearest 0.25 D, giving an
eye-lens positions, an estimate of the image to form an image onto a translucent screen estimate of AddMax for a given set of condi-
position ℓ0 is required. Bailey34 suggested at the base of the magnifier. The retinos- tions. ‘z’ is defined as the distance from the
using a retinoscopy rack or trial lenses at copy rack can be adjusted until an image is stand magnifier exit surface to the centre
the magnifier lens face, to focus light from formed on the screen, and the exit vergence of the trial lens (that is, the ‘spectacle
the base of the magnifier for an emmetropic will be equal and opposite in power to the plane’).
eye viewing with relaxed accommodation retinoscopy rack lens. These methods will The technique is simple and intuitive.
through the magnifier. The power of the provide a sufficiently precise estimate of exit Objects of known scale are used – graph
lens neutralising the exit vergence of the vergence to determine the appropriate paper in this case – and a reference
16
Trial lens
19.4 pixels/mm
Camera
dO = 25 cm
Trial lens
Camera
Image Object 51 pixels/mm
EVD = Pixels/mm in reference image x dO

Pixels/mm in magnified image
EVD = 19.143/51.0 x 25 cm
EVD = 9.38 cm
Figure 3. Arrangement for photographic determination of equivalent viewing distance (EVD). In the example an image is taken of
the object, at a reference distance of do (in this case 25 cm) (19.413 pixels/mm). An image is also taken at a distance z from the
stand magnifier (51 pixels/mm). From Equation 20, EVD = 19.413/51 × 25 cm = 9.38 cm.
photograph is taken at a known distance d0. reference distance do = 25 cm and cent. The photo comparison method and the
In the example, d0 was set at 25 cm, but it 51 pixels/mm was measured in the magni- Bailey method EVDs agree to close to five per
could be at another distance which gives suf- fier image. This gives an EVD value of cent, which would be acceptable from a clinical
ficient image quality. The trial lens is 9.38 cm. point of view, being a difference of only 0.02 log
adjusted to give a clear image on the cam- This ‘photo comparison’ method with units.
era. As a check it should be equal to the Equation 20 was used to determine EVD The data in the upper plot are very close
reciprocal of d0. An image is taken. The values for the Eschenbach System Vario to the five per cent difference line and show
appropriate stand magnifier is positioned Plus series of stand magnifiers, for that the photo comparison method yields
over the object and the camera is set up at z = 2.5 cm, 10 cm, and 25 cm, along with the slightly higher estimates of EVD (that is,
the desired distance z. The trial lens which appropriate spectacle additions. For com- less image enlargement) than the Bailey
gives the clearest screen image is inserted. parison, EVDs were determined based on method. Some of that difference between
The image through the magnifier is photo- FSM measured using Bailey’s magnification the two measurements is expected because
graphed. An appropriate section of the method8,10,14,19,34 and ℓ0 measured using the the reference plane used to calculate exit
image can be selected to be analysed. telescope method10,14,19,34,37 referenced to vergence in the Bailey method is the surface
For high-powered magnifiers with periph- the exit surface of the magnifier, and Equa- of the lens and a more appropriate refer-
eral distortion, a clear section of the image tion 6, along with AddMax calculated using ence plane is the second principal point of
at the centre of the magnifier is chosen. The Equation 3. Eschenbach provides considerable the magnifier lens, which lies closer to the
number of pixels per millimetre in the image technical information about these magnifiers image and object.
seen through the stand magnifier is deter- in their catalogue,33 including FSM and L0 From the data of Chung and Johnston19 on
mined (the cursor functions on Microsoft which was used to derive ℓ0 and these COIL stand magnifiers, the second principal
Paint are used for this) as is the number were used to calculate EVDs, again using plane lay, on average, 1 cm from the lens sur-
of pixels per millimetre in the reference Equation 6, along with AddMax calculated face. If the second principal plane of the
image. This allows comparison of the relative using Equation 3. Eschenbach magnifiers in Table 2 lies at a
scales of the images. Then EVD can be calcu- These results are shown in Table 2. similar position then it can be calculated that
lated as Figure 4 shows difference-versus-mean lateral magnification m for the Bailey method
plots to compare the photo comparison EVDs would be overestimated by 3.5 per cent,
pixels=mm in reference image × d o with the Bailey method EVDs (upper frame) and resulting in EVD measurements being 3.5 per
EVD = [20]
pixels=mm in magnifier image the photo comparison EVDs with the EVDs cent too low. Thus, the position of the refer-
taken from catalogue specifications (lower ence plane for exit vergence is likely to
In the example, 19.143 pixels/mm was frame). The dashed straight line on both plots account for a considerable amount of the dif-
measured in the reference image at a represents inter-method differences of five per ference between the two methods.
17
z = 2.5 cm z = 10 cm z = 25 cm
Catalogue number EVD (cm) AddMax (D) EVD (cm) AddMax (D) EVD (cm) AddMax (D)
Eschenbach System Vario Plus
15806 − 3 × 7.6D 8.80 +5.50 12.43 +3.75 19.78 +2.25
15817 − 3.9 × 11.4D 7.79 +4.50 10.56 +3.50 15.85 +2.25
15826 − 2.8 × 7D 8.79 +6.00 12.86 +4.00 20.63 +2.25
15599 − 3 × 12D 6.70 +6.00 9.83 +4.00 15.79 +2.50
15549 − 4 × 16D 5.95 +4.25 7.85 +3.25 11.90 +2.00
15539 − 5 × 20D 4.99 +3.75 6.48 +2.75 9.38 +1.75
15527 − 6 × 24 D 4.39 +3.25 5.49 +2.75 7.98 +1.75
15516 − 7 × 28D 3.71 +2.75 4.54 +2.25 6.14 +1.50
15507 − 10 × 38D 2.67 +3.25 3.44 +2.50 4.93 +1.75
15577 − 12.5 × 50D 2.18 +3.25 2.55 +2.50 3.60 +1.75
Bailey method
15806 − 3 × 7.6D 8.35 +5.47 11.78 +3.88 18.63 +2.45
15817 − 3.9 × 11.4D 7.33 +4.64 9.89 +3.44 14.99 +2.27
15826 − 2.8 × 7D 8.37 +6.09 12.20 +4.18 19.85 +2.57
15599 − 3 × 12D 6.43 +6.07 9.36 +4.17 15.21 +2.57
15549 − 4 × 16D 5.66 +4.33 7.50 +3.27 11.17 +2.19
15539 − 5 × 20D 4.78 +3.87 6.17 +3.00 8.95 +2.07
15527 − 6 × 24 D 4.14 +3.70 5.29 +2.90 7.59 +2.02
15516 − 7 × 28D 3.58 +2.86 4.34 +2.35 5.87 +1.74
15507 − 10 × 38D 2.65 +3.53 3.35 +2.79 4.75 +1.97
15577 − 12.5 × 50D 2.06 +3.39 2.59 +2.70 3.64 +1.92
Catalogue
15806 − 3 × 7.6D 8.18 +5.71 11.68 +4.00 18.69 +2.5
15817 − 3.9 × 11.4D 6.86 +4.44 9.15 +3.33 13.72 +2.22
15826 − 2.8 × 7D 8.54 +5.71 12.20 +4.00 19.51 +2.50
15599 − 3 × 12D 6.25 +5.71 8.93 +4.00 14.29 +2.50
15549 − 4 × 16D 5.50 +3.64 7.00 +2.86 10.00 +2.00
15539 − 5 × 20D 4.60 +3.51 5.81 +2.78 8.23 +1.96
15527 − 6 × 24 D 3.96 +3.08 4.88 +2.5 6.71 +1.82
15516 − 7 × 28D 3.47 +2.74 4.18 +2.27 5.61 +1.69
15507 − 10 × 38D 2.62 +2.60 3.13 +2.17 4.16 +1.64
15577 − 12.5 × 50D 2.03 +2.60 2.42 +2.17 3.21 +1.64
Table 2. Eschenbach stand magnifier equivalent viewing distances (EVDs) and AddMax for different eye-lens differences z deter-
mined using the photo comparison method, Bailey method34,37 and from parameters in the Eschenbach catalogue33
Estimates of spectacle additions using the single patient with 6/150 acuity obtaining depth be taken with a smart phone with autofocus.
Bailey telescope method match those from of focus levels between 1.00 D and 3.00 Field of view does influence reading perfor-
the photographic method to within 0.50 D for D. However, there is no current study of how mance23,24 with optimum recommended field
the different techniques. Catalogue estimates defocus with real-world use of stand magnifiers of view for reading being 16–20 characters.25
match less precisely, to within 0.65 D. This is affects parameters such as reading speed. Fine et al.42 have shown that reading perfor-
probably sufficient for low vision purposes. Empirically, assessing field of view has been mance increases with stand magnifiers even
Legge et al.39 observed tolerance to defocus usually accomplished by an observer posi- when there are already 13 characters in the
increased with defocus, patients whose acuity tioned at the appropriate distance, and there field. Some patients show a reduction in read-
was 6/12 having depth of focus of 0.50 D and can be a mismatch between empirical field of ing rate with magnifiers when print size gets
those with acuity of 6/120 having depth of focus view and calculated field of view.19 Photo- too large.27 Field of view may therefore be a
between 1.50 D and 5.00 D. Tucker and graphic methods modified from those used in useful addition to tables of EVDs as incorpo-
Charman40 observed a similar relationship. Figure 3 could also be used. The appropriate rated by other authors,14,19,43 allowing practi-
Jacobs and Johnston41 trialled different stand reference plane for this is the entrance pupil tioners to distinguish between devices of
magnifiers with different spectacle additions on a of the camera, and the measurement could similar EVDs.
18
2.0 required for viewing clearly through the

magnifier. It must be stressed that there has
been no research on whether using a specta-
Photo comparison EVD–
1.0 cle addition has an effect on reading perfor-

Bailey EVD (cm)
mance with stand magnifiers and this may

be an area for further research. From a clini-
0.0 cal perspective, practitioners can trial appro-
priate additions to see if patients report
z = 2.5 cm subjective improvement with a stand magni-
z = 10.0 cm
fier. Presbyopic practitioners may find it
-1.0 z = 25.0 cm
quicker and easier to measure the addition
on themselves, looking through the magni-
fier at the appropriate working distance.
-2.0
Table 2 does not contain estimates of field
0.0 5.0 10.0 15.0 20.0
of view. This is likely to be a useful parame-
Mean photo comparison EVD and Bailey EVD (cm) ter as well, and should be included if avail-
able. Where possible, such field of view
2.0 measurements should be based on empiri-
cal measurements. Again, this could be done
with photographic methods, but using the
Photo comparison EVD–
magnifier at the appropriate distance, with a

catalogue EVD (cm)
1.0
ruler as an object, which will usually be suffi-
ciently accurate for in-office measurements.
The photo comparison method for deter-
0.0
mining EVD for stand magnifiers works well. It
z = 2.5 cm could be adapted to measuring EVD (and field
z = 10.0 cm of view) in other near devices such as hand
-1.0 z = 25.0 cm
magnifiers, near telescopes, prism spectacles,
and electronic magnifiers, as well as measur-
ing the magnification in distance telescopes.
-2.0
0.0 5.0 10.0 15.0 20.0
REFERENCES
Mean photo comparison EVD and catalogue EVD (cm) 1. Virgili G, Acosta R, Grover LL et al. Reading aids for
adults with low vision. Cochrane Database Syst Rev
2013: CD003303.
Figure 4. Difference-versus-mean plots comparing equivalent viewing distance (EVD) 2. Peterson RC, Wolffsohn JS, Rubinstein M et al. Benefits
calculated by photo comparison with Bailey34,37 method (upper frame) and with EVD of electronic vision enhancement systems (EVES) for
the visually impaired. Am J Ophthalmol 2003; 136:
from catalogue measurements (lower frame), for a series of Eschenbach stand magni- 1129–1135.
fiers. The dashed line denotes a five per cent difference. 3. Lindsay J, Bickerstaff D, McGlade A et al. Low vision
service delivery: an audit of newly developed outreach
clinics in Northern Ireland. Ophthalmic Physiol Opt
products become available. There is cur- 2004; 24: 360–368.
Conclusions 4. Gao G, Yu M, Dai J et al. Demographic and clinical
rently a need for an updated set of measure- characteristics of a paediatric low vision population in
ments for commonly used stand magnifiers, a low vision clinic in China. Clin Exp Optom 2016; 99:
The labels of manufacturers on stand magni- and indeed other forms of near devices. If 274–279.
fiers are nearly always misleading, in terms of practitioners need to characterise the EVD of 5. Bowers A, Cheong AM, Lovie-Kitchin JE. Reading with
optical magnifiers: page navigation strategies and dif-
their enlarging properties. Measures such as a stand magnifier that is not catalogued then ficulties. Optom Vis Sci 2007; 84: 9–20.
EVD or EVP will be more useful for prescribing the photographic method described above is 6. Cheong AM, Lovie-Kitchin JE, Bowers AR et al. Short-
term in-office practice improves reading performance
devices. Some manufacturers provide sufficient sufficiently simple to be used in-office.
with stand magnifiers for people with AMD. Optom Vis
information about their stand magnifiers to Most tables of EVDs for stand magnifiers Sci 2005; 82: 114–127.
estimate EVD, but many do not. EVD tables are specify EID for different z values.8,10,28 7. Fincham WHA, Freeman MH. The principles of optical
instruments. In: Optics, 8th ed. London: Butterworths,
therefore likely to be useful for the practitioner. Table 2 expresses this parameter as its recip- 1980. pp. 155–174.
It should be noted that Table 2 is the first rocal, AddMax. This is likely to be easier for 8. Bailey IL, Bullimore MA, Greer RB et al. Low vision
practitioners to use than EID, as many magnifiers--their optical parameters and methods for
table of EVDs for stand magnifiers published prescribing. Optom Vis Sci 1994; 71: 689–698.
in reviewed literature in 20 years. More vision-impaired patients will have presbyo- 9. Brown WL, Siemsen DW. Magnification labels for
recent tables are available online,43 updated pia. For them, it may be simplest to use stand magnifiers: always misleading and usually unac-
hievable. Optometry 2008; 79: 9–17.
in 2006, containing EVDs for stand magni- AddMax to prescribe a spectacle addition or
10. Bullimore MA, Bailey IL. Stand magnifiers: an evalua-
fiers, prism spectacles, hand magnifiers and to compare with an existing addition. For tion of new optical aids from COIL. Optom Vis Sci
electronic magnifiers. For EVD to remain non-presbyopic patients, for example chil- 1989; 66: 766–773.
11. Johnston AW. Understanding how simple magnifiers
useful, such tables must be made readily dren using stand magnifiers,44,45 AddMax will provide image enlargement. Clin Exp Optom 2003; 86:
available and regularly updated as new be an estimate of accommodative demand 403–408.
19
12. Rumney NJ. Hand magnifiers. In: Jackson JA, Wolffsohn JS, 23. Legge GE, Pelli DG, Rubin GS et al. Psychophysics of 35. Rumney NJ. New range of stand magnifiers based on
Bailey IL, eds. Low Vision Manual. Edinburgh: Butterworth reading--I. Normal vision. Vision Res 1985; 25: standardized image position. Ophthalmic Physiol Opt
Heinemann Elsevier, 2007. pp. 198–209. 239–252. 1994; 14: 419–422.
13. Wolffsohn JS. Stand magnifiers. In: Jackson JA, 24. Legge GE, Rubin GS, Pelli DG et al. Psychophysics of 36. Dillehay SM, Pensyl CD. Low vision aids and the pres-
Wolffsohn JS, Bailey IL, eds. Low Vision Manual. Edinburgh: reading--II. Low vision. Vision Res 1985; 25: 253–265. byope. J Am Optom Assoc 1991; 62: 704–710.
Butterworth Heinemann Elsevier, 2007. pp. 210–222. 25. Whittaker SG, Lovie-Kitchin J. Visual requirements for 37. Bailey IL. Low vision- locating the image in stand mag-
14. Chung ST, Johnston AW. New stand magnifiers do not reading. Optom Vis Sci 1993; 70: 54–65. nifiers - an alternative method. Optom Wkly 1983; 74:
meet rated levels of performance. Clin Exp Optom 26. Bullimore MA, Bailey IL. Reading and eye movements 487–488.
1990; 73: 194–199. in age-related maculopathy. Optom Vis Sci 1995; 72: 38. Carkeet A, Gambrill B, Harding M et al. Measurement of
15. International Standards Organization.15253. Opthalmic 125–138. stand magnifier equivalent viewing distance,
Optics and Instruments- Optical Devices for Enhancing 27. Cheong AC, Lovie-Kitchin JE, Bowers AR. Determining using digital photography. Poster. Annual Meeting of the
Low Vision. Geneva, Switzerland, 2000. magnification for reading with low vision. Clin Exp American Academy of Optometry Seattle, 2013.
16. Smith G, Jacobs RJ, Johnston AW. Procedures for the Optom 2002; 85: 229–237. 39. Legge GE, Mullen KT, Woo GC et al. Tolerance to
optical assessment of low vision aids. Aust J Optom 28. Lovie-Kitchin JE, Whittaker SG. Prescribing near magni- visual defocus. J Opt Soc Am A 1987; 4: 851–863.
1979; 62: 195–200. fication for low vision patients. Clin Exp Optom 1999; 40. Tucker J, Charman WN. The depth-of-focus of the
17. Bailey IL. Magnification for near vision. Optom Mon 82: 214–224. human eye for Snellen letters. Am J Optom Physiol Opt
1980; 71: 73–76. 29. Lovie-Kitchin J. Reading with low vision: the impact of 1975; 52: 3–21.
18. Johnston AW. Technical note: the relationship research on clinical management. Clin Exp Optom 41. Jacobs RJ, Johnston AW. Evaluation of low vision aids.
between magnification and field of view for simple 2011; 94: 121–132. Aust J Optom 1982; 65: 213–216.
magnifiers. Aust J Optom 1982; 65: 74–77. 30. Ahn SJ, Legge GE, Luebker A. Printed cards for mea- 42. Fine EM, Kirschen MP, Peli E. The necessary field of
19. Chung ST, Johnston AW. Practical options for magnifi- suring low-vision reading speed. Vision Res 1995; 35: view to read with an optimal stand magnifier. J Am
cation and field of view of stand magnifiers. Clin Exp 1939–1944. Optom Assoc 1996; 67: 382–389.
Optom 1989; 72: 140–147. 31. Legge GE, Ross JA, Isenberg LM et al. Psychophysics of 43. Bailey IL. Berkeley Yellow Pages- Bailey, Catalog of Optical
20. Smith G, Atchison DA. A3 schematic eyes. In: The Eye reading. Clinical predictors of low-vision reading Parameters of Low Vision Aids 2018. Available at: https://
and Visual Optical Instruments. Cambridge: Cambridge speed. Invest Ophthalmol Vis Sci 1992; 33: 677–687. www.researchgate.net/publication/329105246_Berkeley_
University Press, 1997. pp. 777–792. 32. Wolffsohn JS, Eperjesi F. Predicting prescribed Yellow_Pages-_Bailey. Accessed 19/07/2019.
21. Smith G, Atchison DA. The eye. In: The Eye and Visual magnification. Ophthalmic Physiol Opt 2004; 24: 334–338. 44. Liebrand-Schurink J, Cox RF, van Rens GH et al. Effec-
Optical Instruments. Cambridge: Cambridge University 33. Eschenbach. Eschenbach Products Stand Magnifiers tive and efficient stand magnifier use in visually
Press, 1997. pp. 291–316. 2013. Available at: https://eschenbach.com/products/ impaired children. Front Psych 2016; 7: 944.
22. Lovie-Kitchin J, Bowman KJ. Senile Macular illuminated-stand-magnifiers.asp. Accessed 19/07/2019. 45. Liebrand-Schurink J, Boonstra FN, van Rens GH et al.
Degeneration-Management and Rehabilitation, 1st 34. Bailey IL. Verifying near vision magnifiers- part 2. Shape of magnifiers affects controllability in children with
ed. Boston: Butterworths, 1985. p. 96. Optom Mon 1981; 72: 34–38. visual impairment. Acta Ophthalmol 2016; 94: 761–767.
20
INVITED REVIEW
Correction of presbyopia: old problems with old

(and new) solutions
Pete S Kollbaum OD PhD FAAO We live in a three-dimensional world and the human eye can focus images from a wide
Arthur Bradley PhD range of distances by adjusting the power of the eye’s lens (accommodation). Progressive
School of Optometry, Indiana University, Bloomington, senescent changes in the lens ultimately lead to a complete loss of this ability by about age
Indiana, USA 50, which then requires alternative strategies to generate high-quality retinal images for far
E-mail: [email protected] and close viewing distances. This review paper highlights the biomimetic properties and
underlying optical mechanisms of induced anisometropia, small apertures, dynamic lenses,
and multi-optic lenses in ameliorating the visual consequences of presbyopia. Specifically,
the advantages and consequences of non-liner neural summation leveraged in monovision
treatments are reviewed. Additionally, the value of a small pupil is quantified, and the
impact of pinhole pupil location and their effects on neural sensitivity are examined. Differ-
ent strategies of generating multifocal optics are also examined, and specifically the interac-
Submitted: 13 July 2019 tion between ocular and contact or intraocular lens aberrations and their effect on resulting
Revised: 5 September 2019 image quality are simulated. Interestingly, most of the novel strategies for aiding presbyopic
Accepted for publication: 20 September and pseudophakic eyes (for example, monovision, multifocality, pinhole pupils) have
2019 emerged naturally via evolution in a range of species.
Key words: accommodation, anisometropia, aspheric, multifocal, presbyopia, pupil, spherical aberration
We live in a three-dimensional world, and with other mammals.18,19 This paper builds upon visual acuity.26,29 The improvement in near
monofocal optics, only one object distance can recent comprehensive reviews20,21 to exam- vision created by correcting one eye can be
generate a focused image at any given time.1,2 ine the optical implications of current and dramatic, but monovision does not replicate
Therefore, most of the retinal image is likely to potential strategies to focus the retinal image the range of high-quality vision observed in
be out of focus most of the time. However, despite the loss of accommodative ability. young adults (Figure 1). There is a small reduc-
optical defocus can be catastrophic for human tion of peak acuity and contrast sensitivity
vision3 and may also be a stimulus for failed down to levels observed with monocular
emmetropisation and the continued eye Induced anisometropia vision.30,31 Additionally, intermediate distance
growth that underlies myopia.4–6 vision (where neither eye has a focused
Most vertebrate eyes focus images on the Although different distances can be focused image) is significantly degraded, an effect that
retina by adjusting the optical power of the by employing different optical powers in each is exacerbated by high add powers23,26
eye as stimulus distance changes by either eye (for example, chameleons16), evolution (Figure 1). Also, because monocular defocus is
moving the eye’s lens (for example, fish7–9 has not developed this approach for humans especially detrimental to stereopsis,32,33
and cats10), adjusting the power of the eye’s for whom anisometropia is rare and a possi- patients fit with monovision corrections have
lens (for example, humans), or adjusting the ble cause of visual cortical disruption if pre- compromised stereo acuity34,35 and discrimi-
power of the cornea (for example, sent in early life (amblyopia, loss of nation of suprathreshold horizontal dispar-
chameleons).11–13 However, less common stereopsis).22 However, inducing anisometro- ities.36 Furthermore, in binocularly vulnerable
strategies are employed that use a pinhole pia in older presbyopic or pseudophakic patients monovision can induce strabismus.37
pupil to expand the depth of field (for exam- patients (‘monovision’) has been shown to Unlike a bifocal lens which simultaneously
ple, nautilus14), multifocal optics (for example, expand the binocular depth of field by gener- produces focused and defocused images
fish15), and different powers in different eyes ating clear bifocality of the binocular visual that add optically (and therefore linearly) at
(for example, chameleons16). Variants of system,23–26 and high-quality vision at two dis- the retina, a monovision patient must
these less common strategies are now being tances (Figure 1).27 The success of monovision employ binocular neural summation to add
utilised by older humans who have lost the hinges on the observation that vision with one focused and defocused images in the cortex
ability to adjust power of their lenses (presby- focused and one defocused eye is dominated where non-linear summation38 or cross-eye
opes and pseudphakes17,18), a senescent trait by the focused image,28 as shown quantita- neural inhibition39 is thought to reduce the
also observed in non-human primates and tively in studies of contrast sensitivity23–25 and visibility of the defocused image.28
21
Multifocal optics: old and new solutions Kollbaum and Bradley
Clinical strategies that employ pinhole

pupils to expand the eye’s depth of field
have a long history (for example, Daza de
Valdes, 1591, cited in a recent review by
Charman21). Three key parameters must be
selected: pupil axial location, pupil size, and
Normalised log CS
whether to employ a fixed pupil or one that

varies with light level.
Small pupils have been introduced into
the spectacle plane,52–54 corneal plane55,56
and iris plane.57 Placing a small pupil into
the spectacle plane will expand the eye’s
depth of field (as proposed by Donders), but
will also drastically reduce the field of view
(for example, a 1 mm pinhole placed 17 mm
in front of the eye’s entrance pupil will start
vignetting the retinal image for field eccen-
tricities of about +/−3 degrees),52,53 which
limits the clinical utility of such an approach.
One strategy to deal with this field restric-
Target vergence (D) tion has been to employ multiple pinholes
in the spectacle plane, obtaining some
Figure 1. Contrast sensitivity plotted as a function of target vergence (dioptres) for depth of field expansion and each individual
four different subjects: A young adult (24 years of age; circle symbols and black pinhole allowing light to reach the retina
dashed line), a pre-presbyope not yet requiring a reading add (44 years of age; square from different field locations producing a
symbols and black solid line), a full presbyope who had been using a reading add for complex patchwork of images and scoto-
several years (44 years of age; red circle symbols and red solid line), and a full presby- mas.53 Placing the pinhole into the corneal
ope wearing a +3.00 D add in front of one eye (blue triangle symbols and blue solid plane either surgically58–60 or with a contact
line). Y-axis is log normalised CS (log CS/best corrected binocular single vision log con- lens61 reduces the field constriction effect.
trast sensitivity). However, light from the mid-periphery will
still be blocked from entering the eye,62–66
and total peripheral field occlusion can only
observers, who typically require some opti- be avoided by employing an annular ‘iris’,
Small aperture optics which for large field angles allows light to
cal correction to read at near.48
The impact of pupil size on depth of field enter the eye around the outside of the
The need for multiple optical powers can be can be seen in Figure 2 where image quality inlay.63 These field constriction problems
avoided by employing pinhole optics, which (visual Strehl ratio on optical transfer can be avoided by placing the pinhole pupil
can expand the depth of field of the function)49–51 is plotted as a function of target into or very close to the iris plane (for exam-
eye21,40,41 because the blur resulting from vergence (adapted from Xu et al.41) assuming ple, by employing pharmacological
defocus is approximately proportional to high photopic light levels and a typically aber- miotics),67–70 or including a small aperture
pupil diameter.42 The invertebrate nautilus rated older eye. The depth of field versus dif- into the anterior surface of an intraocular
employs this type of pinhole eye design, fraction blur trade-off can be seen. The impact lens (IOL),55,57,71 or piggy-back implantation
achieving imaging and depth of field with a of diffraction is striking once the pupil size is of a pinhole aperture.72
small pupil,14 and to a lesser extent, there is reduced to under 2 mm, and peak image Selection of pinhole pupil size is affected
some argument that human vision also quality for the 0.5 mm pupil is less than by many factors, but a balance must be
employs this small pupil strategy. Pupil size 10 per cent of that achieved with mid-sized struck between improving the defocused
generally decreases over the age of 50,43 pupils. However, for near distances between near vision without compromising focused
and the retention of near vision pupil miosis 100 and 50 cm, the 1 mm pupil provides the distance vision.73 The value of pinhole optics
in older eyes44,45 can further reduce pupil best image quality, and it outperforms the for presbyopic vision occurs because blur
size for the near viewing conditions where 0.5 mm pupil over more than a 3.5 D range. resulting from defocus will approximately
defocus is more likely. The smaller pupils in The degradation in image quality due to the scale in proportion to the amount of
older eyes effectively correct for the added spherical aberration reduces peak defocus and pupil diameter,42 for example,
increased aberrations found in these image quality for the largest pupil (7 mm) to the blur present in the retinal image when a
eyes.46,47 However, the senescent decrease that observed with a 1 mm pupil. Maximum distance-corrected presbyope views a stimu-
in photopic and mesopic pupil diameters in peak image quality is observed with pupil lus at 50 cm through a 6 mm pupil can be
65 year old eyes (for example, ranging from diameters between 2.5 and 4.5 mm. For this halved by adding a +1.00 D correcting lens,
4.7 to 3.8 photopically and 6.8 to 4.9 mm aberrated eye that is paraxially emmetropic, or reducing pupil diameter to 3 mm. How-
mesopically43) fails to provide sufficient there is a small myopic shift as the pupil is ever, this proportionality rule ignores the
expansion of the depth of field for older dilated due to positive spherical aberration. impact of diffraction and optical aberrations.
22
1 and Drug Administration approved

Image quality (normalised VSOFT)
Crystalens (Bausch and Lomb, Rochester,

NY, USA) employs a hinged haptic to try to
0.5 mm
utilise existing ciliary muscle activity to move
0.8 1 mm
1.5 mm the lens axially in pseudophakic eyes.81
2 mm However, despite the addition of aspheric
2.5 mm optics to the most recent lens design
0.6 3 mm (Crystalens AO) aimed at creating spherical
3.5 mm aberration-free surfaces,82 inadequate axial
4 mm movement has prevented this device from
0.4 5 mm being widely successful.83
6 mm
7 mm
0.2 Lens power variations
Multifocality can also be achieved by non-

0 dynamic methods. Similar to some fish,15
-3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 optical corrections in which optical power
varies with spatial location across the pupil
Target vergence (dioptres)
have been implemented in modern
multifocal IOLs84–86 and contact lenses.87–90
Figure 2. Image quality for different pupil diameters is plotted as a function of target
Radially symmetric or concentric design
vergence (dioptres) for a model eye with 0.2 microns of primary spherical aberration
multifocal lenses are most common,20 but
(C40). Image quality is quantified by the visual Strehl ratio on optical transfer function
early multifocal contact lenses employed a
metric49–51 normalised to the diffraction limited value observed with a 3 mm diame-
segmented geometry inherited from spec-
ter pupil.
tacle lenses,91–93 as did the earliest bifocal
IOL (a variant of which is still produced).94
It also fails to capture the significant impact focused vision in the central field at phot- Segmented bifocal contact lenses, how-
of pupil constriction on retinal illuminance opic light levels.53 For example, the KAMRA ever, suffered from rotational
and vision.73 As pupil size decreases, the corneal inlays employ a 1.6 mm artificial instabilities,92 and are now typically only
role of diffraction blur in degrading image pupil, which has proven effective at provid- implemented as gas-permeable lens
quality is amplified,41 but the impact of ing reading at near in photopic condi- designs utilised for a select patient
higher-order aberrations is reduced, which tions.60,79 Pharmacological pupil miotics group.20 Optical modelling reveals, how-
produces a peak image quality in a focused have reduced pupil diameters to between ever, that meridionally segmented optical
eye for pupil diameters between 2 and 1–2 mm at photopic light levels, providing designs may offer superior optical
3 mm.73,74 At high light levels, where significant improved near vision and effec- performance.95,96
Weber’s law is effective, the accompanying tive near reading.68 However, laboratory The core principle of a concentric
drop in retinal illuminance produced by studies reveal that such small pupils will sig- multifocal design is that it will be rotationally
pupil constriction has no effect on vision.75 nificantly degrade focused (distance) vision insensitive and include two or more powers
However, at low photopic and mesopic light when lighting drops to low photopic and contained in geometrically separate zones
levels, reducing retinal illuminance lowers mesopic levels.73,74 Balancing the need for located at different distances from the lens
contrast sensitivity due to photon noise expanded depth of field with the need to centre97 (Figure 3). Designs that incorporate
effects,73,76 and the combined effects of dif- avoid loss of distance vision may require a two powers in alternating annular zones
fraction and reductions in retinal illumi- small pupil approach that adjusts with light often also include significant regions of the
nance can significantly impair distance level, and this behaviour is generally lens in which there is a gradual power change
vision when pupil diameters are decreased observed with miotics drugs.80 As pointed with increasing radial distance.98,99 For exam-
below 2.5 mm.74 These studies emphasise out by Charman,21 effectiveness of expan- ple, in Figure 3, which shows high-resolution
that, as is the case for normal human ding the eye’s depth of field may also be Shack-Hartmann two-dimensional power
vision,77,78 the optimum pupil diameter for enhanced if the eye is slightly myopic. maps and radially averaged power profile
presbyopia treatment is larger at low envi- maps, this can be seen in both the centre-
ronmental light levels.74 Interestingly, in distance (top left) and centre-near (lower left)
highly aberrated eyes, depth of focus can Lens mobility designs, and to a lesser extent in the concen-
be expanded by either constricting the tric ring design (top right). Manufacturing limi-
pupil (pinhole pupil effect) or dilating the Similar to the approach taken by some tations may necessitate a spatial spread of
pupil (multifocal optics effect).41 fish7–9 and cats,10 multifocality can also be the transition in optical power, thereby
Pinhole spectacle devices have employed achieved by adjusting the position of the adding multiple powers, and thus authenti-
pupils less than 1 mm (for example, eye’s lens. This approach lies at the heart of cating ‘multifocal’ labelling despite the fact
0.9 mm52,53), with a resulting diffraction strategies to re-establish accommodation in that the lens design may have been strictly
limit of < 30 cpd, which predictably impairs pseudophakic eyes. In particular, the Food bifocal. This manufacturing by-product is
23
Figure 3. This figure shows power profiles of four multifocal contact lenses, two centre-distance designs (Biofinity CD and MiSight,
both CooperVision, Pleasanton, CA, USA), and two centre-near designs (Biofinity CN and AirOptix, CooperVision and Alcon, Ltd,
Fort Worth, TX, USA, respectively). Radial power was computed from wavefront slope data collected over a 7 mm analysis diame-
ter (Power = slope/radius178,179) measured with a validated129 Clear Wave (Lumetrics, Inc., Rochester, NY, USA) single pass Shack-
Hartmann aberrometer (sample spacing of 0.104 mm). To avoid exaggerated computational noise errors near to the lens centre,
power computations are not made in the central 0.6 mm, so this very centre data should be ignored. Colour maps show the radial
symmetry of these four designs. In each colour map, the mean power is assigned green colour, and positive and negative devia-
tions are coded by warm (yellow, orange) and cold (cyan, blue) colours, respectively.
quite different from a multifocal design con- designs.104–106 These diffractive designs can design to generate a pupil size dependency
taining a spatially extensive power gradient be distinguished from the previously in which the add optical power produced by
(similar to adding spherical aberration) from described concentric refractive lenses the diffractive optical element contributes a
the lens centre to its periphery (Figure 3, because every location across the lens con- higher proportion of the light in the image
lower right). tributes to each power, and thus these with small pupils and a much smaller pro-
Another type of multifocal lens employs lenses are often referred to as ‘full aperture’ portion for large pupils. This ‘centre-near’
many concentric zones that each introduce designs,107–110 in which the relative amount approach biases the optics toward the near
half wavelength (or other fractional wave- of energy in the distance and near images add for near viewing (due to near viewing
length) optical path length steps between does not necessarily vary with pupil size, as it pupil miosis, which is present in older
zones. These are usually referred to as does with concentric refracting designs.97–99 eyes116,117) but biases the optics to distance
diffractive lenses because their imaging Diffractive optics have been incorporated vision at night.101,112 For example, a centre-
properties rely on specific optical path into several IOLs,26,101,103,106,111,112 but have near two-zone design with a central 3 mm
length changes between zones100,101 and had limited commercial success as contact diameter zone may become effectively a
result in constructive interference (images) lenses (for example, Echelon113,114 and monofocal near correction at high light
at focal distances determined by the ring Diffrax115). levels97,98,118 resulting in all of the image
geometry.102 In these lenses, a half wave- The pupil size dependency of concentric energy being defocused when viewing dis-
length (pi) phase shift between each zone refractive designs can be considered either tant objects. Similar to the diffractive lenses,
produces a lens with two powers, a bifo- a liability or an attractive feature of the an alternative design approach that varies
cal26,101,103 and other phase step combina- design,97 and indeed, several IOLs have spe- power across meridians and not as a func-
tions can be used to generate trifocal cifically modified the standard diffractive tion of radius95 may also have optical
24
characteristics that are approximately invari- contact lens or IOL.97,122 Therefore, the sig- Therefore, to achieve a desired level of mul-
ant as pupil size changes. Also, such designs nificant positive spherical aberration tifocality in the corrected eye requires larger
avoid the inherently poor optical quality exhibited by older eyes,123–125 and the cor- radially varying power changes (spherical
available from thin annular apertures97 neas of pseudophakes126 may augment any aberration) in the correcting lens of a
resulting in generally superior optical quality centre-distance multifocal that also contrib- centre-near design than of a centre-
of the meridionally varying designs.95 Such utes more positive power with increasing distance design.27,97,122 Although it is a
meridionally varying lenses were first intro- radial distance from the lens centre (positive simple matter to add the extra radially
duced into IOLs by cementing together half spherical aberration). However, in the case varying power needed for the centre-near
lenses with differing powers.94 of centre-near designs which inherently con- designs, high levels of spherical aberration
tain negative spherical aberration,27,98,99,122 in a contact lens will introduce more coma
ocular positive spherical aberration will sub- as the lens decentres,128,129 a problem well
Eye’s inherent multifocality tract from the add power provided by the documented in the contact lens44,130–133
multifocal lens containing negative spherical and IOL134–140 literature. Because many
Although the presbyopic eye is often consid- aberration.97 Importantly, ocular spherical commercially available multifocal contact
ered as a monofocal optical system to be aberration may be either a help or hin- lenses and IOLs employ centre-near
augmented with multifocal lenses in presby- drance depending on the type of design designs this problem is likely to be
opia and pseudophakia, the eye’s optics are being fit, and may likely contribute to the commonplace.
actually multifocal. For example, ocular lon- variable patient responses often experi- Figure 4 quantifies this issue by examining
gitudinal chromatic aberration (LCA) of the enced with these multifocal designs.127 the impact of lens decentration when the
human eye generates approximately a 2.5 D
difference in refractive state at the two
extremes of the visible spectrum. Whitefoot
and Charman119 examined the impact on
depth of focus by doubling or correcting
LCA. Depth of focus increased by 0.50 D
when doubling LCA and decreased by
0.30 D when correcting LCA. Some
diffractive lens designs have the opposite
sign of longitudinal chromatic aberration to 120
refractive optics (more power at long wave-
length versus short).102,103,115 Therefore, the
100
reverse LCA in the first diffractive order
Strehl ratio x 1,000
used to contribute bifocal add power can

mostly correct the ocular LCA,103 providing 80
improved polychromatic image quality.
These studies reveal that the multifocality of
human eyes due to the wavelength depen- 60
dence of refractive index offers little true
multifocality for polychromatic stimuli, as
40
expected due to the large visual attenuation
of the spectral margins.120
The human eye also exhibits significant 20
multifocality at each wavelength, primarily
due to spherical aberration and coma. The
former produces optical power that varies 0
with radius squared, typically resulting in a -2 -1 0 1 2
myopic shift of about 2.00 D at the edge of
an 8 mm diameter pupil.121 Coma, on the
Target vergence (dioptres)
other hand, can create a meridional linear
ramp of power across the pupil. Therefore, Figure 4. Impact on image quality of adding spherical aberration to achieve multi-
the typical eye has both inherent radial and focality. The graph plots the Strehl Ratio image quality metric ×1,000 as a function of
meridional multifocality, but as most pres- target vergence in dioptres for an aberrated presbyopic eye with a 6 mm pupil and
byopes can attest, these levels of inherent +0.2 microns of C40 (black curve), and for the same eye after spherical aberration has
multifocality remain insufficient to provide been added to give the eye+contact lens either +0.4 microns of C40 spherical aberra-
the depth of focus required to navigate our tion (centre-distance model, red) or −0.4 microns of C40 spherical aberration (centre-
three-dimensional world. However, the nat- near model, blue). Further, 0.5 mm or 1.0 mm of lens decentration was introduced
urally occurring changes in refractive state into the same centre-distance and centre-near models (dashed and dotted lines,
across the pupil will add to (or subtract respectively). Images of logMAR letter charts show the peak image quality achievable
from) any multifocality provided by a with and without 1 mm contact lens decentration.
25
added lens is designed to achieve a net

spherical aberration level of 0.4 microns A 100
across a 6 mm pupil. The model included
the reported higher-order aberrations of
older eyes,141 with a baseline spherical aber- 80
ration level (C40) of +0.2 microns. To achieve
matching eye+lens resultant positive and
60
negative multifocality levels of positive or
negative (C40) of 0.4 microns, the centre-
distance lens must add +0.2 microns to the 40
eye’s +0.2 microns, but the centre-near lens
must add −0.6 microns. The resulting peak 20
image quality is slightly higher for the
centre-distance model, but the impact of
lens decentration is now much more dra-
0
matic for the centre-near model because 0 1 2 3 4 5 6
the coma introduced by decentration scales
with the magnitude of the spherical aberra-
B 100
tion levels within the contact lens.128 Peak
Percentage of pupil covered
image quality (Figure 4, see inserts) and

overall image quality ultimately is affected
80
more by lens decentration in the centre-
near design because of the higher levels of
lens spherical aberration required to 60
achieve multifocality.
40
Problem of the simultaneously
present defocused images 20
Although zonal, diffractive bifocal, and 0

multifocal lenses provide the different pow- 0 1 2 3 4 5 6
ers required to focus distance and near
stimuli on the retina, the focused image is C 100
always coupled with a defocused image pro-
duced by the optical power inappropriate
for the stimulus distance (for example, near 80
optic and distant target). The amount of
defocused light in a simple two-zone con-
60
centric design will vary systematically with
pupil size97,98,122 (Figure 5A). When the
outer zone of a two-zone concentric design 40
is defocused, the defocused point spread
function will be an annulus.97,142,143 In this 20
case, the resulting annular ‘halo’ will
increase in size as the pupil dilates and be
most visible when the stimulus contrast is
0
highest, as it is when viewing lights at night 0 1 2 3 4 5 6
(a common source of visual disturbance
clinically reported with multifocal optics). In
Pupil diameter (mm)
addition to reducing the size of these haloes
by reducing the add power (described as Figure 5. Three examples showing the relationship between pupil diameter (x-axis, mm)
either low add multifocal lenses or and the proportion of light imaged by each optic (y-axis) of radially varying zonal lenses.
‘extended depth of focus’ lenses),144 two In each plot, the proportion of pupil covered by the distance (blue), near (red) and transi-
alternative strategies have been proposed tion (black dashed) optics are plotted as a function of pupil diameter in mm for three
to minimise the visibility of these haloes. In centre-distance designs. Specifically, A: a lens with a 3 mm centre zone and an abrupt
one, the size of the defocused halo is change to the add power, B: a centre-distance design with a 3 mm centre zone sur-
reduced by coupling positive defocus with rounded by 1.2 mm annulus across which power gradually changed to the near add
negative spherical aberration (that is, power, and C: a five zone design with alternating distance (zones one, three, and five) and
near (zones two and four) powers. Adapted from Bradley et al.97 and Altoaimi et al.98
26
entering the pupil from the peripheral field

A B must pass through the peripheral optics of
the contact lens.63,155–158 Of course, a simi-
lar variation in refractive state across the
visual field may also potentially be created
by aspheric lenses containing significant
positive spherical aberration.159 Likewise,
orthokeratology also produces an eye with
myopia in the peripheral optics or equiva-
lently an eye with positive corneal plane
spherical aberration.27,160–162 Consistent
with the hypothesis of removing periph-
eral hyperopia from the retinal image,
these multifocal concentric ring163 and
annular164–166 contact lens designs and
the multifocal orthokeratology corneas
have generally shown promise at slowing
myopia development.163,167–173
Lessons from evolution
Figure 6. Simulated retinal images generated by a radially varying zonal bifocal lens Evolution developed some of the multifocal
when the centre zone is focused, and the add zone is nominally +2.00 D defocused. strategies now used to correct presbyopia, and
Centre zone diameter is 3 mm, pupil diameter is 6 mm. In A: 0.15 microns of positive the need for improved surrogates for accom-
spherical aberration was added to the add zone, and in B: negative spherical aberra- modation grows every day as the presbyopic
tion was added. population grows toward an estimated 1.8 bil-
lion by 2050.174,175 Additionally, the value of
including negative spherical aberration in multifocal optics has expanded due to its ability
Adopting multifocal optical
the add zone) and vice versa. This produces to slow myopia progression.163,167–173 As work
designs for the control of
a smaller but higher contrast halo. Alterna- continues to optimise these designs, there may
myopia
tively, contrast of ghost images can be be as yet untried evolutionary strategies that
reduced by coupling positive spherical aber- can be adapted. The true long-term solution,
ration with positive defocus (including posi- Although multifocal optical designs were
and where much of our future research likely
tive spherical aberration in the add zone) originally developed to provide increased
may need to occur may be in preventing the
depth of focus for eyes lacking autofocus
and vice versa, which generates larger but onset of presbyopia in the first place.176 Or,
lower contrast haloes with now clear capability (presbyopic and pseudophakic
better yet, maybe even preventing ageing from
edges.145 The results of these two eyes) as described above, recent experi-
occurring in the first place?177 However, until
aberration-based strategies are simulated in ments on infant monkeys have shown that then, a key understanding of the strengths and
Figure 6 for two designs of centre-distance adding some extra plus power to an other- limitations of accommodation surrogates and
optics with distance targets. On the left wise hyperopic eye will slow the growth of how they might be applied in our clinical prac-
(Figure 6A), positive spherical aberration has the vitreal chamber resulting in reduced tices to aid our patients is critical.
been added to the near add zone increasing myopia.146–148 Significantly, Smith et al. also
the size of the blur, but reducing its con- showed that added plus power could be
restricted to the peripheral retina and have REFERENCES
trast, whereas on the right (Figure 6B), nega- 1. Flitcroft DI. The complex interactions of retinal, opti-
tive spherical aberration has been added to a similar impact.149 These two results have cal and environmental factors in myopia aetiology.
led to a rapid proliferation of optical inter- Prog Retin Eye Res 2012; 31: 622–660.
the add zone reducing the blur size but
2. Hoffman DM, Banks MS. Focus information is used
increasing its contrast. When positive ventions to control myopia development to interpret binocular images. J Vis 2010; 10: 13.
defocus is generated by the add zone for that employ added plus power, notably into 3. Thorn F, Schwartz F. Effects of dioptric blur on
Snellen and grating acuity. Optom Vis Sci 1990;
distance stimuli, adding opposite sign nega- the peripheral retina.150–154 Plus power can
67: 3–7.
tive spherical aberration creates high con- be added to the peripheral retinal image 4. Raviola E, Wiesel TN. Effect of dark-rearing on experi-
trast phase reversed regions of the spatial using some of the same approaches mental myopia in monkeys. Invest Ophthalmol Vis Sci
1978; 17: 485–488.
frequency spectrum103 which can disrupt described above, such as by several types of 5. Smith EL 3rd. Spectacle lenses and emmetropization:
the spatial structure of the focused image. centre-distance concentric ring designs (for the role of optical defocus in regulating ocular devel-
opment. Optom Vis Sci 1998; 75: 388–398.
This phase effect can be easily seen by com- example, MiSight; CooperVision, Pleasanton,
6. Smith EL 3rd, Hung LF. The role of optical defocus in
paring the phase correct but low contrast CA, USA) or annular (Biofinity Multifocal, regulating refractive development in infant monkeys.
images in Figure 6A with the higher contrast CooperVision). The axial separation of the Vision Res 1999; 39: 1415–1435.
7. Schwab IR, Dubielzig RR, Schobert C. Evolution’s Wit-
but phase altered image in Figure 6B. Which cornea and contact lens from the eye’s ness: How Eyes Evolved. New York: Oxford University
would be more disruptive to a patient? entrance pupil necessitates that rays Press, 2012.
27
8. Schwab IR, Hart N. Cover illustration. More than 36. Smith CE, Allison RS, Wilkinson F et al. Monovision: 61. Freeman E. Pinhole contact lenses. Am J Optom Arch
black and white. Br J Ophthalmol 2006; 90: 406. consequences for depth perception from large dis- Am Acad Optom 1952; 29: 347–352.
9. Khorramshahia O, Schartaua JM. Krogera RHH. A parities. Exp Eye Res 2019; 183: 62–67. 62. Albarran Diego C, Montes-Mico R, Pons AM et al.
complex system of ligaments and a muscle keep the 37. Pollard ZF, Greenberg MF, Bordenca M et al. Strabis- Influence of the luminance level on visual perfor-
crystalline lens in place in the eyes of boy fishes (tel- mus precipitated by monovision. Am J Ophthalmol mance with a disposable soft cosmetic tinted contact
eosts). Vision Res 2008; 48: 1503–1508. 2011; 152: 479–482 e471. lens. Ophthalmic Physiol Opt 2001; 21: 411–419.
10. Hughes A. Observing accommodation in the cat. 38. Legge GE, Rubin GS. Binocular interactions in sup- 63. Atchison DA, Blazaki S, Suheimat M et al. Do small-
Vision Res 1973; 13: 481–482. rathreshold contrast perception. Percept Psychophys aperture presbyopic corrections influence the visual
11. Alevogt R, Alevogt R. Studien zur Kinematik der Cha- 1981; 30: 49–61. field? Ophthalmic Physiol Opt 2016; 36: 51–59.
maÈ leonzunge. Z Vgl Physiol 1954; 36: 66–77. 39. Ding J, Sperling G. A gain-control theory of binocular 64. Langenbucher A, Goebels S, Szentmary N et al.
12. Flanders M. Visually guided head movement in the combination. Proc Natl Acad Sci U S A 2006; 103: Vignetting and field of view with the KAMRA corneal
African chameleon. Vision Res 1985; 25: 935–942. 1141–1146. inlay. Biomed Res Int 2013; 2013: 154593.
13. Kirmse W, Kirmse R, Milev E. Visuomotor operation 40. Hickenbotham A, Tiruveedhula P, Roorda A. Compar- 65. Nau A. A contact lens model to produce reversible
in transition from object fixation to prey shooting in ison of spherical aberration and small-pupil profiles visual field loss in healthy subjects. J Am Optom Assoc
chameleons. Biol Cybern 1994; 71: 209–214. in improving depth of focus for presbyopic correc- 2012; 83: 279–284.
14. Muntz WRA, Raj U. On the visual system of Nautilius tions. J Cataract Refract Surg 2012; 38: 2071–2079. 66. Carkeet A. Field restriction and vignetting in contact
Pompilius. J Exp Biol 1984; 109: 253–263. 41. Xu R, Wang H, Jaskulski M et al. Small-pupil versus lenses with opaque peripheries. Clin Exp Optom
15. Gagnon Y, Soderberg B, Kroger R. Optical advan- multifocal strategies for expanding depth of focus of 1998; 81: 151–158.
tages and function of multifocal spherical fish lenses. presbyopic eyes. J Cataract Refract Surg 2019; 45: 67. Renna A, Alio JL, Vejarano LF. Pharmacological treat-
J Opt Soc Am A Opt Image Sci Vis 2012; 29: 1786–1793. 647–655. ments of presbyopia: a review of modern perspec-
16. Ott M, Schaeffel F, Kirmse W. Binocular vision and 42. Smith G. Angular diameter of defocus blur discs. tives. Eye Vis (Lond) 2017; 4: 3.
accommodation in prey-catching chameleons. Am J Optometry and Physiol Optics 1982; 59: 885–889. 68. Abdelkader A. Improved presbyopic vision with
J Comp Physiol A 1998; 182: 319–330. 43. Winn B, Whitaker D, Elliott DB et al. Factors affecting miotics. Eye Contact Lens 2015; 41: 323–327.
17. Duane A. An attempt to determine the normal range light-adapted pupil size in normal human subjects. 69. Abdelkader A, Kaufman HE. Clinical outcomes of
of accommodation at various ages, being a revision Invest Ophthalmol Vis Sci 1994; 35: 1132–1137. combined versus separate carbachol and
of Donder’s experiments. Trans Am Ophthalmol Soc 44. Chateau N, De Brabander J, Bouchard F et al. Infra- brimonidine drops in correcting presbyopia. Eye Vis
1908; 11: 634–641. red pupillometry in presbyopes fitted with soft con- (Lond) 2016; 3: 31.
18. Howland HC, Sivak JG. Penguin vision in air and tact lenses. Optom Vis Sci 1996; 73: 733–741. 70. Benozzi J, Benozzi G, Orman B. Presbyopia: a new
water. Vision Res 1984; 24: 1905–1909. 45. Xu R, Gil D, Dibas M et al. The effect of light level and potential pharmacological treatment. Med Hypothesis
19. Bito LZ, Kaufman PL, DeRousseau CJ et al. Presbyo- small pupils on presbyopic reading performance. Discov Innov Ophthalmol 2012; 1: 3–5.
pia: an animal model and experimental approaches Invest Ophthalmol Vis Sci 2016; 57: 5656–5664. 71. Dick HB, Piovella M, Vukich J et al. Prospective multi-
for the study of the mechanism of accommodation 46. Elliott SL, Choi SS, Doble N et al. Role of high-order center trial of a small-aperture intraocular lens in
and ocular ageing. Eye (Lond) 1987; 1: 222–230. aberrations in senescent changes in spatial vision. cataract surgery. J Cataract Refract Surg 2017; 43:
20. Wolffsohn JS, Davies LN. Presbyopia: effectiveness of cor- J Vis 2009; 9: 24.1–16. 956–968.
rection strategies. Prog Retin Eye Res 2019; 68: 124–143. 47. Guirao A, Gonzalez C, Redondo M et al. Average opti- 72. Trindade C. Small aperture (pinhole) intraocular
21. Charman WN. Pinholes and presbyopia: solution or cal performance of the human eye as a function of implant to increase depth of focus. In: Application UP
sideshow? Ophthalmic Physiol Opt 2019; 39: 1–10. age in a normal population. Invest Ophthalmol Vis Sci ed. US, 2014.
22. Barrett BT, Bradley A, Candy TR. The relationship 1999; 40: 203–213. 73. Xu R, Wang H, Thibos LN et al. Interaction of aberra-
between anisometropia and amblyopia. Prog Retin 48. Pointer JS. The presbyopic add. I. Magnitude and dis- tions, diffraction, and quantal fluctuations determine
Eye Res 2013; 36: 120–158. tribution in a historical context. Ophthalmic Physiol the impact of pupil size on visual quality. J Opt Soc
23. Legras R, Hornain V, Monot A et al. Effect of induced Opt 1995; 15: 235–240. Am A Opt Image Sci Vis 2017; 34: 481–492.
anisometropia on binocular through-focus contrast 49. Cheng X, Thibos LN, Bradley A. Estimating visual 74. Xu R, Thibos L, Bradley A. Effect of target luminance
sensitivity. Optom Vis Sci 2001; 78: 503–509. quality from wavefront aberration measurements. on optimum pupil diameter for presbyopic eyes.
24. Vandermeer G, Legras R, Gicquel JJ et al. Quality of J Refract Surg 2003; 19: S579–S584. Optom Vis Sci 2016; 93: 1409–1419.
vision with traditional monovision versus modified 50. Cheng X, Bradley A, Thibos LN. Predicting subjective 75. Van Nes F, Baouman M. Spatial modulation transfer
monovision. Acta Ophthalmol Suppl 2013; 91: S085. judgment of best focus with objective image quality in the human eye. JOSA 1967; 57: 401–406.
25. Vandermeer G, Rio D, Gicquel JJ et al. Subjective metrics. J Vis 2004; 4: 310–321. 76. Banks MS, Geisler WS, Bennett PJ. The physical limits
through-focus quality of vision with various versions 51. Thibos LN, Hong X, Bradley A et al. Accuracy and pre- of grating visibility. Vision Res 1987; 27: 1915–1924.
of modified monovision. Br J Ophthalmol 2015; 99: cision of objective refraction from wavefront aberra- 77. Campbell FW, Gregory AH. Effect of size of pupil on
997–1003. tions. J Vis 2004; 4: 329–351. visual acuity. Nature 1960; 187: 1121–1123.
26. Ravikumar S, Bradley A, Bharadwaj S et al. Expan- 52. Kim WS, Park IK, Chun YS. Quantitative analysis of 78. Woodhouse JM. The effect of pupil size on grating
ding binocular depth of focus by combining mono- functional changes caused by pinhole glasses. Invest detection at various contrast levels. Vision Res 1975;
vision with diffractive bifocal intraocular lenses. Ophthalmol Vis Sci 2014; 55: 6679–6685. 15: 645–648.
J Cataract Refract Surg 2016; 42: 1288–1296. 53. Kim WS, Park IK, Park YK et al. Comparison of objec- 79. Seyeddain O, Riha W, Hohensinn M et al. Refractive
27. Kollbaum PS. Optical aberrations of contact lenses and tive and subjective changes induced by multiple- surgical correction of presbyopia with the AcuFocus
eyes corrected with contact lenses. In: Optometry: Indiana pinhole glasses and single-pinhole glasses. J Korean small aperture corneal inlay: two-year follow-up.
University. Bloomington, Indiana, 2007. pp. 196–233. Med Sci 2017; 32: 850–857. J Refract Surg 2010; 26: 707–715.
28. Schor C, Landsman L, Erickson P. Ocular dominance 54. Wittenberg S. Pinhole eyewear systems: a special 80. McDonald JE 2nd, El-Moatassem Kotb AM,
and the interocular suppression of blur in mono- report. J Am Optom Assoc 1993; 64: 112–116. Decker BB. Effect of brimonidine tartrate ophthalmic
vision. Am J Optom Physiol Opt 1987; 64: 723–730. 55. Dick HB. Small-aperture strategies for the correction solution 0.2% on pupil size in normal eyes under dif-
29. Fernandez EJ, Schwarz C, Prieto PM et al. Impact on of presbyopia. Curr Opin Ophthalmol 2019; 30: ferent luminance conditions. J Cataract Refract Surg
stereo-acuity of two presbyopia correction approaches: 236–242. 2001; 27: 560–564.
monovision and small aperture inlay. Biomed Opt 56. Seyeddain O, Hohensinn M, Riha W et al. Small- 81. Marchini G, Pedrotti E, Sartori P et al. Ultrasound bio-
Express 2013; 4: 822–830. aperture corneal inlay for the correction of presbyo- microscopic changes during accommodation in eyes
30. Loshin DS. Binocular summation with monovision pia: 3-year follow-up. J Cataract Refract Surg 2012; 38: with accommodating intraocular lenses: pilot study
contact lens correction for presbyopia. Int Contact 35–45. and hypothesis for the mechanism of accommoda-
Lens Clinic 1982; 9: 161–173. 57. Srinivasan S. Small aperture intraocular lenses: the tion. J Cataract Refract Surg 2004; 30: 2476–2482.
31. Collins M, Goode A, Brown B. Distance visual acuity new kids on the block. J Cataract Refract Surg 2018; 82. Incorporated BaL. Bausch and Lomb Crystalens
and monovision. Optom Vis Sci 1993; 70: 723–728. 44: 927–928. Accommodating Posterior Chamber Intraocular Lens
32. Lit A. Presentation of experimental data. J Am Optom 58. Campos M, Beer S, Nakano EM et al. Complete Device Description. [Cited 07 Jan 2019] Available at:
Assoc 1968; 39: 1098–1099. depigmentation of a small aperture corneal inlay https://www.bausch.com/ecp/our-products/cataract-
33. Westheimer G, McKee SP. Stereoscopic acuity with implanted for compensation of presbyopia. Arq Bras surgery/lens-systems/crystalens-ao, 2016.
defocused and spatially filtered retinal images. JOSA Oftalmol 2017; 80: 52–56. 83. Marcos S, Ortiz S, Perez-Merino P et al. Three-
1980; 70: 772–778. 59. GOt W. Correction of presbyopia with a small aper- dimensional evaluation of accommodating intraocu-
34. Back A, Grant T, Hine N. Comparative visual perfor- ture corneal inlay. J Refract Surg 2011; 27: 842–845. lar lens shift and alignment in vivo. Ophthalmology
mance of three presbyopic contact lens corrections. 60. Dexl AK, Seyeddain O, Riha W et al. Reading perfor- 2014; 121: 45–55.
Optom Vis Sci 1992; 69: 474–480. mance after implantation of a small-aperture corneal 84. McNeely RN, Pazo E, Spence A et al. Visual quality
35. Freeman MH, Charman WN. An exploration of modi- inlay for the surgical correction of presbyopia: two- and performance comparison between 2 refractive
fied monovision with diffractive bifocal contact year follow-up. J Cataract Refract Surg 2011; 37: rotationally asymmetric multifocal intraocular lenses.
lenses. Cont Lens Anterior Eye 2007; 30: 189–196. 525–531. J Cataract Refract Surg 2017; 43: 1020–1026.
28
85. Tan N, Zheng D, Ye J. Comparison of visual perfor- 109. Choi J, Schwiegerling J. Optical performance mea- 133. Mackie IA, Mason D, Perry BJ. Factors influencing
mance after implantation of 3 types of intraocular surement and night driving simulation of ReSTOR, corneal contact lens centration. Br J Physiol Opt 1970;
lenses: accommodative, multifocal, and monofocal. ReZoom, and Tecnis multifocal intraocular lenses in 25: 87–103.
Eur J Ophthalmol 2014; 24: 693–698. a model eye. J Refract Surg 2008; 24: 218–222. 134. Holladay JT, Calogero D, Hilmantel G et al. Special
86. Lane SS, Morris M, Nordan L et al. Multifocal intraoc- 110. Zheleznyak L, Kim MJ, MacRae S et al. Impact of cor- report: American academy of ophthalmology task
ular lenses. Ophthalmol Clin North Am 2006; 19: neal aberrations on through-focus image quality of force summary statement for measurement of tilt,
89–105 vi. presbyopia-correcting intraocular lenses using an decentration, and chord length. Ophthalmology 2017;
87. Legras R, Rio D. Simulation of commercial vs theoret- adaptive optics bench system. J Cataract Refract Surg 124: 144–146.
ically optimised contact lenses for presbyopia. Oph- 2012; 38: 1724–1733. 135. Monestam E. Frequency of intraocular lens disloca-
thalmic Physiol Opt 2017; 37: 297–304. 111. Simpson MJ. Re: assessing the optical performance tion and pseudophacodonesis, 20 years after cata-
88. Monsalvez-Romin D, Dominguez-Vicent A, Garcia- of multifocal (diffractive) intraocular lenses. Ophthal- ract surgery - a prospective study. Am J Ophthalmol
Lazaro S et al. Power profiles in multifocal contact mic Physiol Opt 2009; 29: 207. 2019; 198: 215–222.
lenses with variable multifocal zone. Clin Exp Optom 112. Simpson MJ. Diffractive multifocal intraocular lens 136. Perez-Merino P, Marcos S. Effect of intraocular lens
2018; 101: 57–63. image quality. Appl Optics 1992; 31: 3621–3626. decentration on image quality tested in a custom
89. Kim E, Bakaraju RC, Ehrmann K. Power profiles of 113. Atchison DA, Ye M, Bradley A et al. Chromatic aber- model eye. J Cataract Refract Surg 2018; 44: 889–896.
commercial multifocal soft contact lenses. Optom Vis ration and optical power of a diffractive bifocal con- 137. Schroder S, Schrecker J, Daas L et al. Impact of intra-
Sci 2017; 94: 183–196. tact lens. Optom Vis Sci 1992; 69: 797–804. ocular lens displacement on the fixation axis. J Opt
90. Wagner S, Conrad F, Bakaraju RC et al. Power pro- 114. Bradley A, Abdul Rahman H, Soni PS et al. Effects of Soc Am A Opt Image Sci Vis 2018; 35: 561–566.
files of single vision and multifocal soft contact target distance and pupil size on letter contrast sen- 138. Uzel MM, Ozates S, Koc M et al. Decentration and tilt
lenses. Cont Lens Anterior Eye 2015; 38: 2–14. sitivity with simultaneous vision bifocal contact of intraocular lens after posterior capsulotomy.
91. Robboy M, Erickson P. Performance comparison of lenses. Optom Vis Sci 1993; 70: 476–481. Semin Ophthalmol 2018; 33: 766–771.
current hydrophilic alternating vision bifocal contact 115. Freeman M, Stone J. A new diffractive bifocal contact 139. Zhu X, He W, Zhang Y et al. Inferior decentration of
lenses. Int Contact Lens Clinic 1987; 14: 237–243. lens. Trans BCLA 1987; 10: 15–22. multifocal intraocular lenses in myopic eyes.
92. Borish IM, Soni S. Bifocal contact lenses. J Am Optom 116. Radhakrishnan H, Charman WN. Age-related Am J Ophthalmol 2018; 188: 1–8.
Assoc 1982; 53: 219–229. changes in static accommodation and accommoda- 140. Zhu X, Zhang Y, He W et al. Tilt, decentration, and
93. Toshida H, Takahashi K, Sado K et al. Bifocal contact tive miosis. Ophthalmic Physiol Opt 2007; 27: internal higher-order aberrations of sutured
lenses: history, types, characteristics, and actual 342–352. posterior-chamber intraocular lenses in patients
state and problems. Clin Ophthalmol 2008; 2: 117. Altoaimi BH, Almutairi MS, Kollbaum P et al. Accom- with open globe injuries. J Ophthalmol 2017; 2017:
869–877. modative behavior of eyes wearing aspheric single 3517461.
94. Hoffer KJ, Savini G. Multifocal intraocular lenses: his- vision contact lenses. Optom Vis Sci 2017; 94: 141. Wang L, Koch DD. Ocular higher-order aberrations in
torical perspective. In: Alió JL, Joseph P, eds. 971–980. individuals screened for refractive surgery. J Cataract
Multifocal Intraocular Lenses: The Art and the Practice, 118. Charman WN, Radhakrishnan H. Accommodation, Refract Surg 2003; 29: 1896–1903.
Essentials in Ophthalmology. Switzerland: Springer, pupil diameter and myopia. Ophthalmic Physiol Opt 142. Charman WN. Theoretical aspects of concentric
2014. pp. 5–28. 2009; 29: 72–79. varifocal lenses. Ophthalmic Physiol Opt 1982; 2:
95. de Gracia P, Hartwig A. Optimal orientation for angu- 119. Whitefoot HD, Charman WN. Hyperchromatic lenses 75–86.
larly segmented multifocal corrections. Ophthalmic as potential aids for the presbyope. Ophthalmic Phy- 143. Charman WN, Murray IJ, Nacer M et al. Theoretical and
Physiol Opt 2017; 37: 610–623. siol Opt 1995; 15: 13–22. practical performance of a concentric bifocal intraocu-
96. de Gracia P, Dorronsoro C, Marcos S. Multiple zone 120. Bradley A, Glenn A. Fry Award Lecture 1991: percep- lar implant lens. Vision Res 1998; 38: 2841–2853.
multifocal phase designs. Opt Lett 2013; 38: tual manifestations of imperfect optics in the human 144. Gatinel D, Loicq J. Clinically relevant optical proper-
3526–3529. eye: attempts to correct for ocular chromatic aberra- ties of bifocal, trifocal, and extended depth of focus
97. Bradley A, Nam J, Xu R et al. Impact of contact lens tion. Optom Vis Sci 1992; 69: 515–521. intraocular lenses. J Refract Surg 2016; 32: 273–280.
zone geometry and ocular optics on bifocal retinal 121. Thibos LN, Ye M, Zhang X et al. Spherical aberration 145. Bradley A, Kollbaum PS, Thibos LN. Multifocal correc-
image quality. Ophthalmic Physiol Opt 2014; 34: of the reduced schematic eye with elliptical tion providing improved quality of vision. In: USPTO
331–345. refracting surface. Optom Vis Sci 1997; 74: 548–556. ed. United States, 2012.
98. Altoaimi BH, Kollbaum P, Meyer D et al. Experimen- 122. Kollbaum PS, Bradley A, Thibos LN. Comparing the 146. Arumugam B, Hung LF, To CH et al. The effects of
tal investigation of accommodation in eyes fit with optical properties of soft contact lenses on and off simultaneous dual focus lenses on refractive devel-
multifocal contact lenses using a clinical auto-refrac- the eye. Optom Vis Sci 2013; 90: 924–936. opment in infant monkeys. Invest Ophthalmol Vis Sci
tor. Ophthalmic Physiol Opt 2018; 38: 152–163. 123. Guirao A, Redondo M, Artal P. Optical aberrations of 2014; 55: 7423–7432.
99. Plainis S, Atchison DA, Charman WN. Power profiles the human cornea as a function of age. J Opt Soc Am 147. Arumugam B, Hung LF, To CH et al. The effects of the
of multifocal contact lenses and their interpretation. A Opt Image Sci Vis 2000; 17: 1697–1702. relative strength of simultaneous competing defocus
Optom Vis Sci 2013; 90: 1066–1077. 124. Artal P, Berrio E, Guirao A et al. Contribution of the signals on emmetropization in infant rhesus mon-
100. Klein SA. Understanding the diffractive bifocal con- cornea and internal surfaces to the change of ocular keys. Invest Ophthalmol Vis Sci 2016; 57: 3949–3960.
tact lens. Optom Vis Sci 1993; 70: 439–460. aberrations with age. J Opt Soc Am A Opt Image Sci Vis 148. Smith EL 3rd, Hung LF, Huang J et al. Effects of local
101. Davison JA, Simpson MJ. History and development of 2002; 19: 137–143. myopic defocus on refractive development in mon-
the apodized diffractive intraocular lens. J Cataract 125. Brunette I, Bueno JM, Harissi-Dagher M et al. Optical keys. Optom Vis Sci 2013; 90: 1176–1186.
Refract Surg 2006; 32: 849–858. quality of the eye with the Artisan phakic lens for the cor- 149. Smith EL 3rd, Ramamirtham R, Qiao-Grider Y et al.
102. Buralli D, Morris G, Rogers J. Optical performance of rection of high myopia. Optom Vis Sci 2003; 80: 167–174. Effects of foveal ablation on emmetropization and
holographic kinoforms. Appl Optics 1989; 28: 976–983. 126. Wang L, Santaella RM, Booth M et al. Higher-order form-deprivation myopia. Invest Ophthalmol Vis Sci
103. Ravikumar S, Bradley A, Thibos LN. Chromatic aber- aberrations from the internal optics of the eye. 2007; 48: 3914–3922.
ration and polychromatic image quality with J Cataract Refract Surg 2005; 31: 1512–1519. 150. Mutti DO, Sinnott LT, Reuter KS et al. Peripheral
diffractive multifocal intraocular lenses. J Cataract 127. Bakaraju RC, Ehrmann K, Ho A et al. Inherent ocular refraction and eye lengths in myopic children in the
Refract Surg 2014; 40: 1192–1204. spherical aberration and multifocal contact lens opti- bifocal lenses in nearsighted kids (BLINK) study.
104. Lee S, Choi M, Xu Z et al. Optical bench performance cal performance. Optom Vis Sci 2010; 87: 1009–1022. Transl Vis Sci Technol 2019; 8: 17.
of a novel trifocal intraocular lens compared with a 128. Guirao A, Williams DR, Cox IG. Effect of rotation and 151. Liu Y, Wildsoet C. The effect of two-zone concentric
multifocal intraocular lens. Clin Ophthalmol 2016; 10: translation on the expected benefit of an ideal bifocal spectacle lenses on refractive error develop-
1031–1038. method to correct the eye’s higher-order aberra- ment and eye growth in young chicks. Invest
105. Alio JL, Plaza-Puche AB, Alio Del Barrio JL et al. Clini- tions. J Opt Soc Am A Opt Image Sci Vis 2001; 18: Ophthalmol Vis Sci 2011; 52: 1078–1086.
cal outcomes with a diffractive trifocal intraocular 1003–1015. 152. Charman WN, Mountford J, Atchison DA et al.
lens. Eur J Ophthalmol 2018; 28: 419–424. 129. Kollbaum P, Jansen M, Thibos L et al. Validation of Peripheral refraction in orthokeratology patients.
106. Cochener B, Boutillier G, Lamard M et al. A compara- an off-eye contact lens Shack-Hartmann wavefront Optom Vis Sci 2006; 83: 641–648.
tive evaluation of a new generation of diffractive tri- aberrometer. Optom Vis Sci 2008; 85: E817–E828. 153. Smith EL 3rd, Hung LF, Huang J. Relative peripheral
focal and extended depth of focus intraocular 130. Belda-Salmeron L, Drew T, Hall L et al. Objective hyperopic defocus alters central refractive development
lenses. J Refract Surg 2018; 34: 507–514. analysis of contact lens fit. Cont Lens Anterior Eye in infant monkeys. Vision Res 2009; 49: 2386–2392.
107. Barton K, Freeman MH, Woodward EG et al. 2015; 38: 163–167. 154. Shen J, Clark CA, Soni PS et al. Peripheral refraction
Diffractive bifocal contact lenses in aphakia and 131. Woods RL, Saunders JE, Port MJ. Optical perfor- with and without contact lens correction. Optom Vis
pseudophakia. A pilot study Eye (Lond) 1991; 5: mance of decentered bifocal contact lenses. Optom Sci 2010; 87: 642–655.
344–347. Vis Sci 1993; 70: 171–184. 155. Almutleb ES, Bradley A, Jedlicka J et al. Simulation of
108. Schwiegerling J. Analysis of the optical performance 132. Wake E, Tienda JB, Uyekawa PM et al. Centration and a central scotoma using contact lenses with an
of presbyopia treatments with the defocus transfer coverage of hydrogel contact lenses. Am J Optom opaque centre. Ophthalmic Physiol Opt 2018; 38:
function. J Refract Surg 2007; 23: 965–971. Physiol Opt 1981; 58: 302–308. 76–87.
29
156. Fedtke C, Manns F, Ho A. The entrance pupil of the 164. Smith MJ, Walline JJ. Controlling myopia progression 172. Hiraoka T, Kakita T, Okamoto F et al. Long-term
human eye: a three-dimensional model as a function in children and adolescents. Adolesc Health Med Ther effect of overnight orthokeratology on axial length
of viewing angle. Opt Express 2010; 18: 22364–22376. 2015; 6: 133–140. elongation in childhood myopia: a 5-year follow-up
157. Mathur A, Gehrmann J, Atchison DA. Pupil shape as 165. Walline JJ, Greiner KL, McVey ME et al. Multifocal con- study. Invest Ophthalmol Vis Sci 2012; 53: 3913–3919.
viewed along the horizontal visual field. J Vis 2013; tact lens myopia control. Optom Vis Sci 2013; 90: 173. Charm J, Cho P. High myopia-partial reduction ortho-
13: 3. 1207–1214. k: a 2-year randomized study. Optom Vis Sci 2013;
158. Spring KH, Apparent Shape SWS. Size of the pupil 166. Walline JJ, Lindsley K, Vedula SS et al. Interventions 90: 530–539.
viewed obliquely. Br J Ophthalmol 1948; 32: 347–354. to slow progression of myopia in children. Cochrane 174. Holden BA, Fricke TR, Ho SM et al. Global vision
159. Cheng X, Xu J, Chehab K et al. Soft contact lenses Database Syst Rev 2011: CD004916. impairment due to uncorrected presbyopia. Arch
with positive spherical aberration for myopia con- 167. Cho P, Cheung SW, Edwards M. The longitudinal Ophthalmol 2008; 126: 1731–1739.
trol. Optom Vis Sci 2016; 93: 353–366. orthokeratology research in children (LORIC) in Hong 175. Frick KD, Joy SM, Wilson DA et al. The global burden
160. Hiraoka T, Matsumoto Y, Okamoto F et al. Corneal Kong: a pilot study on refractive changes and myopic of potential productivity loss from uncorrected pres-
higher-order aberrations induced by overnight control. Curr Eye Res 2005; 30: 71–80. byopia. Ophthalmology 2015; 122: 1706–1710.
orthokeratology. Am J Ophthalmol 2005; 139: 168. Walline JJ, Jones LA, Sinnott LT. Corneal reshaping and 176. Tsuneyoshi Y, Higuchi A, Negishi K et al. Suppression
429–436. myopia progression. Br J Ophthalmol 2009; 93: of presbyopia progression with pirenoxine eye drops:
161. Hiraoka T, Okamoto C, Ishii Y et al. Contrast sensitiv- 1181–1185. experiments on rats and non-blinded, randomized
ity function and ocular higher-order aberrations fol- 169. Kakita T, Hiraoka T, Oshika T. Influence of overnight clinical trial of efficacy. Sci Rep 2017; 7: 6819.
lowing overnight orthokeratology. Invest Ophthalmol orthokeratology on axial elongation in childhood myo- 177. Lai RW, Lu R, Danthi PS et al. Multi-level remodeling
Vis Sci 2007; 48: 550–556. pia. Invest Ophthalmol Vis Sci 2011; 52: 2170–2174. of transcriptional landscapes in aging and longevity.
162. Gifford P, Li M, Lu H et al. Corneal versus ocular 170. Santodomingo-Rubido J, Villa-Collar C, Gilmartin B BMB Rep 2019; 52: 86–108.
aberrations after overnight orthokeratology. Optom et al. Myopia control with orthokeratology contact 178. Southwell WH. Wave-front estimation from wave-
Vis Sci 2013; 90: 439–447. lenses in Spain: refractive and biometric changes. front slope measurements. J Opt Soc Am 1980; 70:
163. Chamberlain P, Peixoto-de-Matos P, Logan NS et al. Invest Ophthalmol Vis Sci 2012; 53: 5060–5065. 998–1006.
A three-year randomized clinical trial of misight 171. Cho P, Cheung SW. Retardation of myopia in Ortho- 179. Himebaugh NL, Nam J, Bradley A et al. Scale and
lenses for myopia control. Optom Vis Sci 2019; 96: keratology (ROMIO) study: a 2-year randomized clinical spatial distribution of aberrations associated with
556–567. trial. Invest Ophthalmol Vis Sci 2012; 53: 7077–7085. tear breakup. Optom Vis Sci 2012; 89: 1590–1600.
30
INVITED REVIEW
Customised aberration-controlling corrections for keratoconic

patients using contact lenses
Amit Navin Jinabhai PhD BSc (Hons) Technological advancements in the design of soft and scleral contact lenses have led to the
MCOptom FBCLA FEAAO FHEA development of customised, aberration-controlling corrections for patients with
Division of Pharmacy and Optometry, School of Health keratoconus. As the number of contact lens manufacturers producing wavefront-guided
Sciences, Faculty of Biology Medicine and Health, The corrections continues to expand, clinical interest in this customisable technology is also
University of Manchester, Manchester, UK increasing among both patients and practitioners. This review outlines key issues surround-
E-mail: [email protected] ing the measurement of ocular aberrations for patients with keratoconus, with a particular
focus on the possible factors affecting the repeatability of Hartmann-Shack aberrometry
measurements. This review also discusses and compares the relative successes of studies
investigating the design and fitting of soft and scleral customised contact lenses for patients
with keratoconus. A series of key limitations that should be considered before designing
customised contact lens corrections is also described. Despite the challenges of producing
and fitting customised lenses, improvements in visual performance and comfortable wear-
ing times, as provided by these lenses, could help to reduce the rate of keratoplasty in
keratoconic patients, thereby significantly reducing clinical issues related to corneal graft
Submitted: 8 April 2019 surgery. Furthermore, enhancements in optical correction, provided by customised lenses,
Revised: 20 May 2019 could lead to increased independence, particularly among young adult keratoconic patients,
Accepted for publication: 23 May 2019 therefore leading to improvements in quality of life.
Key words: aberration-controlling lenses, customised scleral lenses, customised soft lenses, higher-order aberrations, keratoconus, verti-
cal coma
Keratoconus is an ectatic disease of the cor- where present, corneal scarring, which be significantly elevated in keratoconic
nea, typically characterised by stromal tissue induces unwanted light scatter.9 eyes,8,21,23–26 as the maximal stromal thin-
thinning causing the cornea to take on a Although keratoconus is most usually ning classically occurs at either the inferior5
steepened, conical shape.1 Such alterations bilateral, inter-ocular asymmetry is com- or inferior-temporal position.17,18 Light
occur due to significant changes in the bio- mon, with Nichols et al.10 reporting that the waves arriving at the keratoconic eye, from a
mechanical properties of the cornea,2 degree of asymmetry is usually largest in distant source, will be distorted by compara-
resulting in the stromal lamellar matrix no patients with more advanced disease. tively differing amounts at the (flatter) supe-
longer following a highly regularised, orthog- Unlike other ectatic conditions, such as pel- rior and (steeper) inferior cornea.20,27 The
onal pattern. Instead, there are distinct areas lucid marginal corneal degeneration,11 keratoconic cone apex also distorts incoming
of poorly aligned collagen intermixed with keratoconus characteristically affects the light waves by ‘rotating’ them,27 thereby
collagen that is arranged in the conventional inferior-central two-thirds of the cornea;12 inducing trefoil (or triangular astigmatic)
quasiregular fashion.3 Subsequently, the however, reports of centrally,13 inferiorly,14 aberrations.28,29 Furthermore, the steep-
keratoconic corneal shape becomes more inferior-nasally15 and superiorly positioned ened cone also induces spherical aberra-
easily distorted and typically shows a high cone apices16 have also been published. tion.20,28,29 These notable differences in HOA
degree of protrusion. The keratoconic cor- Other studies indicate that the cone apex is terms, compared to normal eyes, have
nea can also develop apical scarring, which most commonly displaced inferior–temporally supported the use of aberrometry measure-
may typically be attributed to rigid corneal in keratoconus.17,18 Overall, the nature and ments as a useful tool to detect subclinical
contact lens wear and/or disease progres- exact location of the corneal steepening is keratoconus25,28 as well as to grade its
sion over time.4 As the retina usually unique for each keratoconic eye. severity.30
remains unaffected in keratoconus, the Alterations in the profile of the keratoconic Despite these uses, a debatable issue, in
reduced visual performance found, com- cornea (Figure 1) induce large magnitudes of relation to the measurement of HOAs in
pared to normal eyes, is directly attributable HOAs,6,19 which differ significantly from keratoconic patients, is their repeatability,
to a combination of irregular astigmatism,5 those measured in healthy eyes.20–22 Vertical particularly when compared to repeated
higher-order aberrations (HOAs)6–8 and, coma (Z (3,-1)) is most commonly found to aberration measurements made in normal
31
Aberration-controlling lenses for keratoconus Jinabhai
Mild Moderate Severe
Pachymetry
(in µm)
Topography
(in D)
Wavefront
higher-order
aberrations
(in µm)
Figure 1. A collection of Pentacam data images highlighting the progression of keratoconus between three different keratoconic
patients; specifically, an eye with ‘mild’ keratoconus (left-hand column), an eye with ‘moderate’ keratoconus (central column) and
an eye with ‘severe’ keratoconus (right-hand column). The upper row presents pachymetry maps (numerical data are shown in
microns [μm]), the middle row presents topography maps (numerical data are shown in dioptres [D]) and the lower row presents
corneal wavefront higher-order aberration maps (numerical data are shown in microns [μm]).
eyes. Using a Scheimpflug-based topogra- comparatively poorer in the same group of of ocular aberrations would be of greater
pher, both Shankar et al.29 and Sideroudi keratoconic patients. Interestingly, Ortiz- importance than just anterior corneal sur-
et al.31 have previously reported poor Toquero et al.34 have reported that anterior face aberrations alone. This is due to the
repeatability of anterior and posterior cor- corneal HOA measurements, made using a fact that the eye’s internal optics (the poste-
neal surface aberrations, respectively. This Placido-based topographer, were actually rior corneal surface and the crystalline lens)
finding was further supported by Jinabhai more repeatable in 36 keratoconic eyes than are known to partly compensate for the
et al.,32 who reported poor repeatability of measurements made in 36 normal eyes. aberrations of the anterior cornea in both
ocular HOA measurements made using the Correspondingly, Shetty et al.35 suggested normal36,37 and keratoconic eyes.19,22,38,39 In
Hartmann–Shack technique. In contrast, that using a programmable, liquid-crystal- fact, Chen and Yoon38 proposed that in
Bayhan et al.33 reported comparable levels on-silicon phase modulating adaptive optics keratoconus, some level of compensation
of intra-examiner repeatability between set-up, to evaluate ocular HOAs, yielded a exists between the coma root-mean-square
anterior corneal aberrations measured in high intra-session repeatability for eyes with (RMS) error aberrations of the anterior and
41 keratoconic eyes and 31 normal eyes mild to moderate keratoconus. posterior corneal surfaces. Their results
using a combined Scheimpflug-Placido As the broad aim of this review is to con- indicated that the level of compensation
topographer. However, the authors’ data sider HOA measurements with respect to seemed to vary with the severity of disease;
indicate that the repeatability of their poste- their potential optical correction, discussion on average 22, 24 and 14 per cent of the
rior corneal aberration measurements was of the issues surrounding the measurement anterior surface’s coma RMS error
32
aberrations were compensated for by the Inward pathway

posterior surface in severe, moderate and
mild keratoconic eyes, respectively. In con-
trast, no such compensatory effects for
coma RMS error were found in their normal
subjects.
The majority of the studies that have
explored the correction of ocular HOAs in
keratoconic patients, using either soft, rigid
corneal or rigid scleral contact lenses, have Outward pathway
Relay lenses
used the Hartmann–Shack principle to mea-
sure their patient’s aberrations,23,26,40–56
whereas comparatively fewer studies have
used either skiascopic methods7,57–59 or the Retinal
laser ray-tracing method.60 point source
Due to its comparative popularity, some
key contributors that are likely to impact
on the repeatability of ocular aberration CCD sensor Reflected, deformed
measurements made using the Hartmann– wavefront exiting the eye
Shack technique, in keratoconic patients, Micro-lenslet array,
include: positioned conjugate
• spot-imaging errors at the wavefront to the eye’s pupil
sensor
• computational limitations Figure 2. Upper: a diagram depicting the optical path that light takes through the
• small (fixational) eye movements during Hartmann–Shack aberrometer as it enters the eye under investigation. Lower: a dia-
measurements gram depicting the optical path that light waves, reflecting off the retina, take on
• changes in aberrations due to micro- their way outward, toward the micro-lenslet array, before being imaged onto the
fluctuations in accommodation and/or charge-coupled device (CCD) sensor.
changes in the tear film during measure-
ments.
Tilted wavefront
spot image position Tilted wavefront
Spot-imaging issues
For the Hartmann–Shack technique, the
aberrated wavefront emerging from the eye Magnitude
is relayed onto a micro-lenslet array of spot image
(Figure 2, lower image) thereby generating a displacement
pattern of multiple spot images, which is
then analysed by computerised software. By
measuring the displacement of the spot
image, with respect to a fixed ‘reference’ Section of the
charge-coupled
point (Figure 3), the software then attempts
device (CCD) sensor
to reconstruct the original aberrated that corresponds to this
wavefront falling onto the lenslet array particular micro-lenslet
(Figure 4, lower image).
When attempting to evaluate the optical An individual
quality of the keratoconic eye using the micro-lenslet
Hartmann–Shack technique, the fundamental Planar wavefront
Planar wavefront spot image formed
problem lies in acquiring the spot images at at the lenslet’s reference position
the wavefront sensor, as the cornea may which is denoted by the black cross (X)
often be very distorted or even scarred (partic-
ularly in severe cases of keratoconus). The Figure 3. A schematic diagram depicting the local wavefront slope formed at the
measurement performance of the wavefront Hartmann–Shack wavefront sensor. The planar wavefront (shown in red) passing
sensor directly depends on how accurately through the micro-lenslet forms a spot image along its optical axis. The centroid posi-
the centre of each spot can be detected by the tion of the spot image is defined as the ‘reference’ position for the micro-lenslet (den-
sensor’s centroiding algorithm.61 In general, oted by the black cross, X). The tilted wavefront (shown in green) passing through the
data derived from a Hartmann–Shack sensor micro-lenslet forms a spot image which is displaced away from the reference position
does not consider the ‘optical quality’ of the (X) and corresponds to the local slope (shown in green) formed in front of the micro-
individual spots formed by the lenslet array. lenslet.
33
Only the degree of their ‘displacement’ is

Face-on view Side-on view
needed to compute the local wavefront slope
over each lenslet aperture (Figure 3). How-
ever, it is important to note that the optical
quality of these spot images can vary greatly
between normal and keratoconic eyes.62,63
A fundamental limitation of the Hartmann–
Shack sensor is the requirement that each spot
image, generated by any given micro-lenslet,
must land within the ‘virtual sub-aperture’ of a
certain photon detector at the charge-coupled
device (CCD). To this end, the ‘dynamic range’
A ‘perfect’ of a Hartmann–Shack aberrometer is some-
CCD wavefront shape what restricted by the diameter of each indi-
A ‘regular’ lattice-like sensor Micro-lenslet vidual micro-lenslet, which typically ranges
array of spot images array Local slope of the wavefront somewhere between 0.3 to 0.5 mm,64 but can
captured from an formed over each lenslet’s be as large as 0.75 mm.65
‘ideal’ eye aperture The computerised software typically used
in commercially available Hartmann–Shack
aberrometers is not usually capable of cor-
rectly identifying the following spot image
registration issues:56,61,66
Face-on view Side-on view
• when two (or more) separate spot images
are formed at the same CCD photon
detector sub-aperture (Figure 5A)
• a spot image which perfectly overlaps
with another, formed by an adjacent
micro-lenslet (Figure 5B)
• a spot image that ‘crosses over’ the allo-
cated path of another spot (Figure 5C)
• missing spot images – when one or more
spot images are formed in an area that
falls entirely outside of the CCD sensor
An ‘aberrated’ (Figure 5D).
CCD wavefront shape While a ‘displaced’ spot image is obviously
An ‘irregular’ pattern sensor aberrant from the ‘chief’ or ‘reference’ ray,
of spot images captured Micro-lenslet
Local slope of the wavefront the magnitude of this displacement pro-
from an ‘aberrated’ eye array vides no indication of the image’s quality.
formed over each lenslet’s
On the other hand, a blurry spot image may
aperture
actually contain more aberration, optical
scatter and refractive blur compared to a
Figure 4. The imaging principles underpinning the Hartmann–Shack aberrometry
sharper spot image.62,63 Figure 6 shows a
technique for a perfect (upper) and an aberrated eye (lower). Upper: the Hartmann–
typical example of the appearance of the
Shack device’s micro-lenslet array (where individual lenslets are typically between 0.3
Hartmann–Shack spot images captured
to 0.5 mm in diameter) effectively subdivides the wavefront into multiple beams. The
from an eye with severe keratoconus, while
‘local slope’ of the wavefront, over each individual lenslet’s aperture, will determine
Figure 7 depicts the reconstruction of a
the location of each individual spot image on the charge-coupled device (CCD) sensor.
highly aberrated wavefront derived from a
The upper illustration depicts the results for an ‘ideal’ eye, where the solid red line
grossly distorted spot image array formed
shows a ‘perfect’ wavefront. Lower: an aberrated wavefront produces an irregular
at the CCD array, as is typically found in
pattern of spots on the CCD sensor. Displacement of each spot from the
patients with keratoconus.
corresponding lenslet’s axis gives a measure of the ‘local slope’ of the wavefront. The
With the Hartmann–Shack technique, the
lower illustration depicts the results of an ‘aberrated’ eye where the wavy purple line
CCD sensor assumes that all of the spot
shows an ‘aberrated’ wavefront.
images will lie ‘flat’ over the finite diameter
of the lenslet in question.67 This assumption
begins to break down even for coarse,
lower-order aberrations when the magni-
tude of those aberrations is large. In such
cases, the wavefront is significantly curved
over the lenslet’s aperture, resulting in a
34
Charge-coupled device (CCD) sensor simply by using a micro-array of lenslets with

a shorter second focal length.61,66 However,
this concept is somewhat flawed, as higher-
powered micro-lenses would result in a
Virtual sub-aperture for ‘centroiding’ algorithm
reduction in sensitivity, that is, a decrease in
the sensor’s ability to detect differences
A. ‘Multiple’ spot images formed at the same between differing magnitudes of aberration.
section of the CCD sensor Conversely, increasing the micro-lenses’ sec-
ond focal length could provide permissible
B. ‘Overlapping’ spot images formed at the sensitivity, which may still be clinically useful;
Incident
same section of the CCD sensor however, this adjustment limits the sensor’s
aberrated
dynamic range.61 Another option would be
wavefront
C. ‘Crossed-over’ spot images formed at the to rearrange the physical spacing of the
CCD sensor lenslets to allow maximum resolution;68
however, this is not easily achieved with
most commercially available Hartmann–
Shack aberrometers and would likely impact
D. ‘Missing’ spot image, essentially formed
on the spatial resolution of the wavefront.
away from the CCD sensor entirely
Rarer and more complex methods have
Micro-lenslet array been suggested to help reduce the loss of
data for the Hartmann–Shack technique,
Figure 5. Examples of spot-imaging issues impacting on the Hartmann–Shack tech- which include using:
nique, including: A: ‘multiple’ spot images, B: ‘overlapping’ spot images, C: ‘crossed- • complex ‘unwrapping’ algorithms in the
over’ spot images and D: a ‘missing’ spot image computational spot detection process;
mathematical modifications to assign
spots to their corresponding lenslet, even
blurry spot, which is difficult to localise inducement of a ‘hazy’ image of the spot, if deviated outside their sub-aperture69
because its centre cannot always be accu- rather than a true geographical deviation of • astigmatic lenses in the micro-lenslet
rately located.67 If the aberrations are large the spot image.9,63 Although these blurry array, which produce ‘line’ rather than
enough, these blurry spots can even over- spot images are problematic, they contain ‘spot’ images, all the cylindrical lenses are
lap, which considerably complicates the useful information about the degree and orientated at a multitude of different
analysis. location of optical scatter sources within the angles allowing simpler recognition of the
It is also possible that ‘micro-aberrations’ eye.63 line image from a given cylindrical aper-
may exist within the spot images, which are Fundamentally, a simple solution to help ture lens70
too small to be detected by the wavefront decrease the amount of ‘crossover’ or ‘over- • a spatial modulator array; this device is
sensor’s detector.67 These micro-aberrations lapping’ of the aberrated spot images would placed in front of the lenslet array and
could still be detrimental to the retinal image be to reduce the physical amount of spot dis- allows the selective switching ‘on’ and ‘off’
quality as they may contribute to the placement possible. This may be achieved of certain sub-apertures, allowing a defi-
nite assignment of the spots to their
corresponding sub-apertures onto the
CCD sensor71
• an image-processing algorithm, along-
side an astigmatic lenslet array, that is
capable of tracing line foci which fall
outside the bounds of the conventional
search72
• a ‘spot searching’ method, which involves
fitting an astigmatic micro-lenslet to the
centre of every group of 2-×-2 spherical
micro-lenslets within the overall array
(also known as a dual micro-lenslet array).
This optical set-up allows the generation
of both spot and line images, which cre-
Figure 6. A comparison of the raw Hartmann–Shack spot images (formed at the ates a unique discernible pattern which is
charge-coupled device) between a healthy normal eye (left) and an eye with severe then processed using binary computa-
keratoconus (right). While the left-hand image shows a regular series of spot images tions and mask processing.64
arranged into a lattice-like array, the right-hand image shows substantial spot However, it is worth noting that most
image irregularities, such as missing spot images as well as fainter, blurrier spot commercially available Hartmann–Shack
images. aberrometers do not easily allow any of the
35
Face-on view Side-on view analysis, primarily due to their orthogonality

over a circular pupil and their representation
of ‘classic’ Seidel aberration terms such as
coma and spherical aberration.76 Computa-
tion of the Zernike co-efficients requires a set
of discrete orthogonal polynomials to be con-
structed using the Gram-Schmidt method,
where the co-efficients are calculated by
‘fitting’ the wavefront slope gradients and
orthogonal polynomials using a ‘least-
squares’ method.77 This methodology essen-
tially describes how each aberration coeffi-
cient makes up a proportion of the total
A ‘grossly-distorted’ CCD A ‘highly aberrated’ wavefront and aims to minimise the absolute
pattern of spot images sensor wavefront shape error between the measured sampled points
Micro-lenslet
captured from a and the Zernike terms which are fitted to the
array Local slope of the wavefront
‘highly aberrated’ data. However, investigators should be wary
over each lenslet’s aperture
keratoconic eye that the number of sampling points available
from the Hartmann–Shack aberrometer will
Figure 7. A highly aberrated wavefront produces a grossly distorted pattern of spots typically be far greater than the number of
on the charge-coupled device (CCD) sensor (left-hand image). Displacement of each Zernike polynomial terms that can be fitted
spot from the corresponding lenslet’s axis gives a measure of the ‘local slope’ of the to the wavefront to describe its shape.67 This
wavefront. The illustration above depicts the results from a ‘highly aberrated’ has led to other research exploring newer
keratoconic eye, where the wavy green line shows a highly aberrated wavefront methods to reconstruct and compute
(right-hand image). HOAs.78
Unfortunately, the direct subtraction or
comparison of ‘corneal’ and ‘total ocular’
above modifications to be made to the be moved several times during data cap- aberrations is not usually possible and can
devices’ original set-up; as a result, these ture. Additionally, neither study discusses give rise to major inaccuracies, purely
modifications are not routinely used in clini- that some spots, due to gross corneal dis- because different reference axes may have
cal investigations. tortion/scarring, will become deviated away been used during the topography and aber-
Compared to the aforementioned stud- from the aberrometer’s CCD sensor alto- rometry measurements. Most commercially
ies, the use of a ‘translatable grid’ has been gether, nor do they clarify if their available wavefront aberrometers measur-
implemented as a more practical approach translating-grid method could possibly ing the total ocular aberrations tend to use
for increasing the Hartmann–Shack account for this. Finally, it is entirely plausi- the patient’s line of sight as the reference
method’s ‘dynamic range’ – the maximum ble that some spot images, no matter how axis. Aberrations measured with respect to
measurable wavefront slope of an aber- many times the grid is moved, are likely to this axis therefore have the pupil centre as
rated ray.73,74 The translatable grid fall onto the grid itself, causing ‘lost’ data the Cartesian origin.79 On the other hand,
increases the width of the area that any (as no image will be registered at the CCD corneal topographers generally align the
given aberrated spot can fall onto, by sensor). Nevertheless, the results from Pan- videokeratographic axis with the corneal
blocking every other neighbouring lenslet, tanelli et al.73 provided a possible solution sighting centre (the intersection of the line of
for any given lenslet. Yoon et al.74 used this for evaluating highly aberrated eyes. Trans- sight with the corneal surface).80 Such inac-
translatable grid device on four normal latable grids can only usually be used in curacies in comparing corneal versus ocular
human eyes, a keratoconic eye and an custom-built aberrometers, as commercially aberrations may be accounted for (both
aberrated phase plate, and reported no dif- available instruments will not permit such mathematically and geometrically) and may
ference in sensitivity, repeatability or accu- adaptations of their micro-lenslet array. be minimised by using an instrument that
racy compared to measurements made can take simultaneous corneal and ocular
without the grid. Pantanelli et al.73 success- Computational limitations aberration measurements.20,22,24
fully used this method on 190 normal con- The aberration co-efficients recorded by
trol subjects, as well as 19 eyes with an aberrometer are mathematically
keratoconus. They found higher levels of reconstructed from a discrete number of Small (fixational) eye
negative vertical coma, third-order trefoil sampling points and are therefore not the movements
and fourth-order spherical aberration in perfect mirror image of the true aberrant From the results of their reliability study of
the keratoconic eyes than in the normal wavefront.61,74,75 The mathematical approach normal eyes, Cheng et al.81 suggested that
eyes. Nonetheless, there are limitations to classically used in aberrometry employs the small fixational eye movements, which may
using a translatable grid, in that inaccura- Zernike polynomial expansion series, which is have occurred during the measurement
cies and errors could arise due to the lon- based on the expansion of the derivatives of process, could induce variability in aberra-
ger data acquisition times required when the wavefront aberration. Zernike polyno- tion measurements. Therefore, it is highly
using this technique, as the grid needs to mials are extremely popular in wavefront likely, for eyes with keratoconus, that any
36
such small eye movements during the mea- (in white light), rather than just defocus and when interpreting the data presented by
surement stage would induce larger varia- astigmatism alone. Scores of > 1.0 represent Marsack et al.,75 it is important to note that
tions in aberrations. This point is of a positive VB, indicating a gain in visual per- the authors assumed a ‘perfect’ on-eye
particular importance as the magnitude of formance through aberration correction, alignment of their ‘customised lens’ correc-
fixational eye movements are likely to be whereas eyes with a score of 1.0 would gain tion (that is, a contact lens that was custom-
larger in keratoconic patients, than in nor- no VB from correction. Williams et al.86 found designed to correct lower-order aberrations,
mal subjects, due to poorer levels of acuity. that the VB in 109 normal eyes ranged from while also simultaneously reducing some of
While this suggestion currently remains 1.5 to 8.0, whereas the VB for their four the HOAs associated with keratoconus).
unexplored, it is worth noting that both keratoconic eyes varied between 2.5 to 25.0. However, this is an unrealistic assumption
Mihaltz et al.82 and Tan et al.6 have demon- Similar magnitudes of positive VB scores for even a prism-ballasted toric soft lens,
strated that the magnitude of the lower- were also found for four keratoconic patients which typically moves vertically by 0.3 to
order and HOAs measured in keratoconic by Guiaro et al.,87 at MTF spatial frequencies 1.0 mm upon blink,93,94 with approximately
eyes will be greatly influenced by the loca- of 16 and 32 cycles/degree. Furthermore, 2–15 degrees of rotation.95 The results of
tion of the cone apex. using their translatable grid-integrated modelling simulation by Marsack et al.75
Hartmann–Shack wavefront aberrometer to showed that mild and moderate cases of
Changes in aberrations due to increase dynamic range,74 Pantanelli et al.73 keratoconus would theoretically benefit
micro-fluctuations in also reported encouraging VB scores (ranging more than those with severe keratoconus if
accommodation or changes in from 2.5 to 10.5) for 15 keratoconic patients. the number of Zernike radial orders that are
the tear film The authors calculated their VB scores using corrected are truncated. This finding was
In normal eyes, other possible sources of the metric of volume under the MTF across perhaps to be expected, as severe to
variance in HOA measurements include spatial frequencies of zero to 60 cycles/degree. advanced keratoconic patients are more
changes in aberrations due to micro- Overall, these three studies each highlighted likely to show larger magnitudes of aberra-
fluctuations in accommodation83 or varia- that the theoretical benefit of using tions at the higher radial orders. Therefore,
tions in the tear film84 during the measure- customised correction methods for in order to successfully correct aberrations
ment process. However, in patients with keratoconic eyes was far superior to that of in keratoconus, with aberration-controlling
moderate keratoconus, Radhakrishnan normal eyes. customised lenses, it may be necessary to
et al.85 found that while HOAs did alter with modify the strategy of correction depending
accommodation and tear film changes DO ALL ABERRATION TERMS NEED on each individual patient’s disease severity.
immediately post-blink; the magnitude of CORRECTING?
these changes were relatively small com- The literature indicates that correcting every The use of non-customised
pared to their patients’ manifest aberra- single aberration term may not be benefi- contact lenses to correct
tions. In support of these results, Chen cial, as lens decentrations are likely to hin- aberrations
et al.48 proposed that tear film-induced vari- der the VB yielded for both normal87,88 and To date, a number of studies have investigated
ations in aberrations could be somewhat keratoconic eyes.40,89,90 Furthermore, due to the use of non-customised soft lenses,43,57,58
countered by standardising aberration mea- changes with accommodation and varia- rigid corneal lenses7,23,40,42,46,47,54,57 and rigid
surements, by ensuring that all readings are tions in the tear film, it is also widely scleral lenses51,55,59 to reduce the magnitude
taken two seconds post-blink. accepted that very few HOA terms are of manifest HOAs in keratoconic patients, with
completely stable in either normal81,83,84,91 each demonstrating varying degrees of suc-
Correcting ocular aberrations or keratoconic eyes.32,51,85 Nonetheless, cess, as well as revealing some important
in keratoconic patients López-Gil et al.92 reported that the aberra- findings.
VISUAL BENEFIT tions created through decentration of Perhaps rather predictably, both Jinabhai
Despite potential issues regarding their customised lenses were likely to be smaller et al.43 and Abdu et al.57 agreed that non-
repeatability, Williams et al.86 have proposed than the difference between the total RMS customised corneal lenses provided better
there is usually an identifiable and significant error measured with and without visual performance and superior aberration
‘visual benefit’ (VB) to correcting the ocular customised lenses in place for two normal correction than non-customised soft contact
HOAs of the keratoconic eye measured using eyes. Thus López-Gil et al.92 hypothesised lenses. Compared to soft lenses, corneal lenses
the Hartmann–Shack method. The authors that, “wearing a customised contact lens can mask the manifest corneal aberrations,
calculated their VB scores as the ratio of the over a course of time will show a clear induced through keratoconus, by effectively
modulation transfer function (MTF) measured benefit… especially for patients with moder- ‘replacing’ the irregular corneal surface with
with a ‘customised aberration correction’ in ate to high amounts of aberration”. On the smooth and regular refractive surfaces.96,97
place (that is, a correction that specifically other hand, Marsack et al.75 proposed that However, both Jinabhai et al.43 and Abdu
corrected for all lower-order and HOA terms), correction of all the HOA terms of the et al.57 confirmed that even with corneal
to the MTF found with just the second-order keratoconic eye are not worthwhile. The lenses in situ, there were still some residual
aberrations corrected – all of their MTFs were authors suggested that only correcting HOAs present, which were typically larger in
computed for a pupil diameter of 5.7 mm between the third and up to the fifth magnitude than the aberrations measured
and a spatial frequency of 16 cycles/degree.86 Zernike orders would give most keratoconic in normal, healthy eyes.91,98 These findings
Accordingly, their VB scores indicated the eyes better visual performance and lessen corroborated the results of other previous
potential increase in retinal image contrast by the likelihood of inducing superfluous aber- studies of keratoconic patients habitually
correcting all of the monochromatic HOAs rations due to lens decentrations. However, wearing corneal lenses7,23,46,47 or scleral
37
lenses.59 Other authors have proposed that visual performance, would be appealing to lens translation/rotation observed with the
these residual aberrations are most likely many keratoconic patients and practitioners ‘trial’ lens. Their customised lens reduced
attributable to irregularities of the posterior alike. Ideally, this hydrogel contact lens would uncorrected higher-order RMS (HORMS)
corneal surface,22,38,39 which is also known be silicone-based, of a regular thickness and error by 67 per cent; however, the authors
to become significantly distorted in would have the capabilities to reduce the did not measure visual acuity as part of
keratoconus.99 manifest ocular HOAs associated with their experiment. The authors proposed
To investigate the potential impact of cor- keratoconus. that their residual aberrations could have
recting such residual aberrations on been induced by variations in the tear film,
visual performance, Yang et al.54 used a SIMULATIONS or even on-eye vertical translation and/or
customised adaptive optics visual simulator de Brabander et al.90 simulated the visual rotation of the customised lens with blinks.
(a 37-actuator deformable mirror) to mea- performance achieved by using a Sabesan et al.50 conducted a comparison
sure contrast sensitivity function (using sine- customised soft lens to correct HOAs in nine study for three keratoconic eyes (two severe
wave gratings, presented at five different moderate keratoconic eyes and reported and one moderate) investigating the effec-
spatial frequencies) with corneal contact large improvements in the MTF with the tiveness of front-surface, customised soft
lenses in situ for 20 eyes of 19 keratoconic ‘perfect’ alignment of a customised lens. lenses versus conventional soft and corneal
patients. Compared to without it, the authors Like the results of Guirao et al.,88,105 de lenses. The authors accounted for possible
reported improved contrast sensitivity func- Brabander et al.90 reported that decen- soft lens decentrations by fitting ‘trial’ lenses
tion with their simulator, particularly at low trations of their customised soft lens led to with three different base curves and
(two cycles/degree) and intermediate spatial a partial loss in the VB gained for assessed their fittings to ascertain which
frequencies (four, eight and 16 cycles/degree). keratoconic patients. However, it should be was the most stable for each eye, ahead of
Overall, the results from this study highlighted noted that the authors calculated the effects manufacturing their customised soft lenses.
that better correction of residual aberrations of rotation and translation separately from In the most successful case, the uncorrected
could likely improve visual performance in each other. Clinically, it is widely accepted HORMS error was reduced by 75 per cent
keratoconic patients.54 that soft lenses will translate and rotate with the customised lens, but only by 17 per
upon blinking, simultaneously,93,94 and that cent with a conventional soft lens. Com-
these movements are not mutually exclu- pared to the conventional lenses, the
The use of customised sive. Nonetheless, the results of de customised lenses gave improved low-
corrections for keratoconic Brabander et al.90 showed that rotations up contrast (20 per cent) logarithm of the mini-
patients to a maximum of five degrees and transla- mum angle of resolution (logMAR) acuities,
Ahead of discussing individual studies, it is tions up to a maximum of 1 mm, upon by an average of 2.1 lines. For one of their
important to acknowledge that the majority blinking would be permissible to still yield a severe cases, the customised lens gave an
of the literature regarding the design and use benefit from a customised lens. improvement of three lines of low-contrast
of customised lenses,26,41,44,45,48–50 phase Yoon and Jeong106 simulated the logMAR acuity compared to the patient’s
plates49 or adaptive optics (typically in the decentration of customised contact lenses habitual corneal lens. For high-contrast
form of a deformable mirror)54,100–103 is for two post-penetrating keratoplasty and (100 per cent) acuity there was very little dif-
largely limited to ‘non-surgical’ keratoconic two keratoconic eyes. They found that com- ference between the subject’s corneal lens
corneas only, which typically show no apical pared to normal eyes (VB = 3), a customised and the customised lens; however, the
scarring. While this is not representative of correction gave their highly aberrated eyes a customised lens still performed best. In
the full spectrum of keratoconic patients, such threefold improvement in visual perfor- agreement with Jeong and Yoon49, Sabesan
studies provide key information about the mance (VB = 9). The authors’ results also et al.50 also noted that some small residual
impact of correcting optical aberrations with- suggested that highly aberrated eyes were errors persisted even with their customised
out the confounding factor of ‘optical scatter- more tolerant to decentrations than normal lenses in situ.
ing’ due to the presence of corneal scarring. eyes; for a 0.2 mm translation vertically, the Marsack et al.45 produced a front-surface,
Contact lenses are discrete, simple to use VB gained was reduced by only a third for customised aberration-controlling soft lens
and relatively inexpensive to manufacture, highly aberrated eyes, but by half in for a patient with moderate keratoconus
and therefore represent a suitable device with normal eyes. and compared this to the patient’s habitual,
which to correct HOAs. This idea was first pro- conventional soft contact lens. The results
posed by Smirnov104 who acknowledged that showed that both high-contrast (87 per cent)
“it is possible to manufacture a lens compen- CUSTOMISED SOFT LENSES and low-contrast (four per cent) logMAR
sating for the wave(front) aberrations of the Jeong and Yoon49 manufactured a front-sur- acuity were improved with the customised
eye” and that “these lenses must obviously be face, customised aberration-controlling soft lens compared to with the habitual soft lens.
contact ones”. As corneal lenses will typically lens for a patient with advanced In contrast to the results of Sabesan et al.,50
show significant magnitudes of on-eye keratoconus. On-eye lens decentration was Marsack et al.45 found that high-contrast
decentration with blinking, a number of stud- accounted for by first fitting a conventional logMAR acuity was improved (by 1.5 lines;
ies have investigated the use of soft contact soft ‘trial’ lens and monitoring its centration p = 0.03) more than low-contrast logMAR
lenses for providing a customised correction using an infrared pupil camera linked to acuity (which only improved by one line of
of aberrations.26,41,44,45,48–50 A soft lens their aberrometer. The aberration correc- letters; p = 0.11); however, such differences
design, which could achieve maximum on-eye tion was transferred onto the final may be due to the differing contrast levels
lens comfort as well as providing optimal customised lens, accounting for any on-eye used between these two studies.
38
Interestingly, although Marsack et al.45 customised lenses for three keratoconic was from −0.56 μm to −0.15 μm. These
noted that the habitual lenses typically eyes (one severe and two moderate cases) reductions in HOAs were believed to have
translated on-eye with blinks, they did not and reported that all three customised contributed to the improvements in high-
incorporate this factor into the design of lenses provided better logMAR acuities com- contrast (100 per cent) logMAR visual acu-
their final customised lenses. Nonetheless, pared to the patients’ habitual corneal ities measured with the customised lenses,
their customised lens successfully reduced lenses. For their patient with severe compared to the patient’s best-corrected
the uncorrected HORMS error from 0.99 μm keratoconus, their habitual corneal lens pro- spectacle acuities (the largest improvement
to 0.37 μm over a 5 mm pupil (compared to vided a high-contrast (91 per cent) acuity of reported was from 0.52 to 0.06 log units). In
0.77 μm found with the habitual conven- 0.04 0.09 log units, whereas their broad agreement with the results of
tional soft lens). customised lens gave an improved acuity of Sabesan et al.,50 Katsoulos et al.26 also
Chen et al.48 proposed a method of −0.05 0.05 log units. Conversely, the found greater improvements in low-contrast
enhancing the fitting of aberration- patient’s habitual corneal lens provided a (50 per cent) logMAR visual acuities com-
controlling soft lenses by custom-designing low-contrast (52 per cent) acuity of pared to best-corrected spectacle acuities
the lens’ back surface (using topographical 0.58 0.04 log units, whereas the (the largest improvement was from 1.00 to
data) to help reduce residual aberrations customised lens yielded an acuity of 0.10 log units). The authors’ rationale for
induced through lens translations/rotations. 0.61 0.04 log units. Encouragingly, the using a 75 per cent correction was based on
The authors reported that compared to con- mean uncorrected HORMS error was previous studies which had shown that
ventional lenses, their customised lenses reduced from 1.57 0.03 μm to 0.76 decentrations of a partial wavefront aberra-
reduced both horizontal and vertical transla- 0.03 μm with the customised lens, and to tion correction, rather than the full correc-
tions by a factor of two and reduced rota- 0.50 0.15 μm with the habitual corneal tion, would still yield a helpful VB compared
tions by a factor of five. However, the lens. Similarly, for one of their moderate to conventional contact lenses.90,105 Acknowl-
authors’ customised lenses only successfully cases, the uncorrected HORMS error was edging that not all keratoconic cones are
reduced uncorrected HORMS error for one reduced from 0.61 0.02 μm to 0.39 always decentred in the same position away
of their three patients with moderate 0.02 μm using their habitual corneal lens, from the individual eye’s line of sight,6,82
keratoconus, from 1.66 0.06 μm to and to 0.38 0.07 μm with the customised Katsoulos et al.26 also proposed that more
0.61 0.04 μm, whereas for one patient, the lens. The high-contrast (91 per cent) acuity centrally located cones could require the
author’s customised lens actually induced a for this particular patient was 0.20 0.02 correction of spherical aberration in order to
significant increase in HORMS error, from log units with their habitual corneal lens, achieve optimal visual performance.
1.17 0.04 μm (when uncorrected) to which reduced to 0.14 0.02 log units with Building on the approach used by Katsoulos
1.30 0.10 μm, which was largely attributed the customised lens. However, the patient’s et al.,26 Jinabhai et al.41 explored the effective-
to overcorrection of a majority of Zernike habitual corneal lens provided a low- ness of aberration correction provided by
terms with their customised lens in situ. For contrast (37 per cent) acuity of 0.58 0.04 customised lenses that gave either a 50 per
the remaining patient, there was no signifi- log units, whereas the customised lens cent or a 100 per cent correction of both verti-
cant change in HORMS error (uncorrected: yielded an acuity of 0.59 0.04 log units. cal and horizontal third-order coma, over a
0.70 0.03 μm, versus with the customised Nonetheless, these two cases highlighted natural 4 mm pupil. The authors’ rationale for
lens: 0.69 0.08 μm). that customised soft lenses have the poten- using a ‘partial’ correction was based on previ-
Unfortunately, Chen et al.48 did not mea- tial to provide comparable results to corneal ous studies which confirmed that
sure either high- or low-contrast visual acu- lenses in terms of low-contrast acuity, yet decentration of a ‘full’ wavefront-guided cor-
ity in their study. However, because they superior results in terms of high-contrast rection (through either rotation and/or trans-
measured both corneal surfaces’ aberra- acuity. lation) induces superfluous residual
tions as well as the total ocular aberrations, Katsoulos et al.26 used a rather different aberrations,40,60,89,90 thereby diminishing
they were able to partly model the magni- approach to producing customised soft visual performance. Jinabhai et al.41 com-
tude of the aberrations of the eye’s internal lenses for eight mild to moderate pared their two customised lenses to non-
optics. Their modelling results indicated that keratoconic eyes; their lenses were designed customised, conventional toric soft lenses and
the posterior corneal surface and crystalline to correct for around 75 per cent of the the patient’s habitual corneal lenses. Unlike in
lens were also responsible for some of the eye’s manifest third-order negative vertical previous studies, the authors used a subjec-
residual aberrations measured with their coma aberration, as well the second-order tive over-refraction result to determine the
customised lenses on-eye for their three Zernike terms extracted directly from their lower-order powers of both their customised
keratoconic patients. Chen et al.48 also aberrometry data for a 4 mm pupil diame- and non-customised soft toric lenses. This
acknowledged that their customised lenses ter. In all eight cases, a reduction in was because both Katsoulos et al.26 and
only had a central 5 mm optical zone of uncorrected HORMS error was seen (the Jinabhai et al.8 had previously demonstrated
aberration correction, which would likely largest reduction was from 0.86 μm to that the lower-order sphere and cylinder
cause problems with glare if the lenses were 0.42 μm); however, the authors did not terms, measured objectively using Hartmann–
worn in scotopic conditions. explain if the mean differences were signifi- Shack aberrometry, did not readily corre-
Marsack et al.44 compared visual perfor- cant. On the other hand, Katsoulos et al.26 spond with the sphere and cylinder powers
mance and ocular aberrations using reported a significant reduction in the mag- measured during a subjective refraction for
bespoke wavefront-guided soft contact nitude of uncorrected vertical coma aberra- keratoconic patients. Such variability between
lenses versus the subject’s own habitual cor- tion with their customised lenses these methods may be attributable to errors
neal lenses. The authors produced (p < 0.005). The largest reduction reported at the wavefront sensor.61
39
Jinabhai et al.41 reported significant Jinabhai et al.40 demonstrated that, on-eye rotation and less than 200 μm of on-
changes in mean third-order vertical coma depending on their magnitude as well as the eye translations with blinks over a 20-second
aberration for 12 keratoconic eyes, from eye’s inherent wavefront error, superfluous period. Sabesan et al.52 found that, com-
−0.93 0.34 μm when uncorrected, to lower-order and HOAs, induced through pared with the ‘best-fitting’ spherical optic
+0.18 0.39 μm with the ‘100 per cent unwanted customised lens decentrations, can lenses (1.17 0.57 μm), the customised
lenses’ (p = 0.002); to −0.17 0.30 μm with reduce the effectiveness of the wavefront cor- scleral lenses significantly reduced the mean
the ‘50 per cent lenses’ (p = 0.002) and to rection. These induced, residual aberrations HORMS error (to 0.37 0.19 μm; p < 0.05)
+0.39 0.14 μm with the patient’s habitual are typically proportional to the amount of for their keratoconic patients. Equally, the
corneal lenses (p = 0.002). In contrast, displacement, as well the magnitude of the authors also reported a significant improve-
their non-customised toric lenses did not displaced aberration.88,108 Of particular ment in mean monocular, distance high-
significantly reduce vertical coma importance, Jinabhai et al.40 reported that ver- contrast logMAR visual acuity between the
(−0.66 0.43 μm). While both the ‘100 per tical translations typically induced larger resid- study lenses, by an average of 1.9 lines
cent lenses’ and the habitual corneal ual spherical and cylindrical errors than (p < 0.05), in addition to significant improve-
lenses produced a ‘positive shift’ in vertical horizontal translations. The authors’ results ments in sinusoidal grating-based contrast
coma, the differences between these two suggested that vertical translations of a full, sensitivities at four (increased by a factor of
modes of correction were not statistically customised HOA correction might be limited 2.4), eight (increased by a factor of 1.8) and
significant. The authors also found no sig- to no more than 0.1 mm. 12 cycles/degree (increased by a factor of
nificant differences in horizontal third- 1.4), respectively. While these results indi-
order coma measurements between these CUSTOMISED SCLERAL LENSES cate that the reduction of aberrations was
five measurement conditions. Compared to customised soft lenses, which fairly successful, it is worth noting that
In spite of the apparent improvements in typically induce superfluous aberrations due residual aberrations still remained even
vertical coma aberration with their two to their unavoidable degree of ‘on-eye’ with the customised scleral lenses in situ;
customised lenses, Jinabhai et al.41 reported movement and their variable conformity to these were most likely due to the small
that the patient’s habitual corneal lenses the keratoconic corneal profile, customised lens movements observed immediately
provided significantly better distance high- scleral contact lenses are likely to offer a after blinking, or even due to keratoconus-
contrast (95 per cent) logMAR acuity, dis- greater degree of on-eye stability for induced distortions at the posterior cor-
tance low-contrast (15 per cent) logMAR acu- keratoconic patients.52,53,109 Moreover, neal surface and/or the crystalline lens.38
ity and near vision SKILL card107 scores scleral lenses also have the added benefit of Another contributing factor could have
compared to the ‘100 per cent lenses’ improving optical performance, by providing been the slight mismatch between the
(p ≤ 0.002). However, the authors found no a ‘regular’, rigid first optical surface, while pupil size at which the HOAs were
significant differences in high-contrast, low- also simultaneously increasing lens wear measured with the Hartmann–Shack
contrast or SKILL card scores measured with comfort by allowing a majority of the lens’ aberrometer (6 mm) and the actual size of
the ‘50 per cent lenses’, versus either the weight to bear onto the conjunctiva.110 the zone of customisation within the scleral
habitual corneal lenses or the ‘100 per cent Sabesan et al.52 designed customised lenses (which varied between patients,
lenses’. While it was clear that the corneal scleral lenses, for six keratoconic patients from 7.0 to 8.5 mm).
lenses provided the best visual performance (11 eyes), by first identifying their ‘best- In accordance with Sabesan et al.,52
scores of all the possible lens options that fitting’ conventional lenses. These ‘best- Marsack et al.53 also utilised a posterior sur-
were investigated, the authors also noted fitting’ lenses each had a ‘central optic’, the face scleral toric landing zone in their
that the ‘50 per cent lenses’ generally pro- corrective properties of which were purely peripheral lens designs to help provide on-
vided better visual performance scores com- spherical, and a ‘customisable periphery’ eye rotational stability. However, in contrast
pared to the ‘100 per cent lenses’. which had toric properties, and also allowed to the approach used by Sabesan et al.,52
The authors acknowledged that their quadrant-specific adjustments to be made Marsack et al.53 designed their ‘best-fitting’
customised lenses’ visual performance results (where necessary) to stabilise the lens while conventional, or ‘intermediate’, scleral
were likely to have been affected by small on- simultaneously minimising com- lenses to incorporate a ‘spherical equivalent’
eye lens translations (despite on-eye rotations pression/impingement on the conjunctiva. defocus power within the central optic,
being accounted for during the manufacturing With these ‘best-fitting’ lenses in situ, the which was derived from a subjective refrac-
process), as well as differences between the authors carefully evaluated and accounted tion routine; this was to produce a starting
patient’s natural pupil size (during the visual for both on-eye rotation and translations lens the weight of which would be more
performance measurements) and the size of after measuring ocular lower-order and closely matched to that of the ‘final’
the zone of customisation of both their ‘50 HOAs (via Hartmann–Shack aberrometry) customised lens. This allowed for more
per cent’ and ‘100 per cent’ lenses. These clini- through the lenses. Using the ‘best-fitting’ accurate measurements of ‘on-eye’ lens
cal findings corroborated the results of the lens parameters as a starting point, the rotation and/or translations of the ‘interme-
authors’ previous study,40 which modelled the authors used their aberration measure- diate’ lens ahead of designing and
effects of customised lens translations and ments to create front-surface, customised manufacturing the customisation zone of
rotations on the correction of aberrations in scleral lenses, manufactured using a sub- their ‘final’ lenses. Marsack et al.53 used this
keratoconic patients, where the theoretical micron-precision lathe. Once verified and approach to design lenses for seven
customised lens was designed to fully correct fitted on-eye, the customised scleral lenses keratoconic patients (14 eyes). Once the
all high-order Zernike terms, up to the fifth were found to provide good temporal stabil- ‘intermediate’ lens had settled, a Hartmann–
order. Using computerised simulations, ity, showing no more than two degrees of Shack aberrometer was used to evaluate
40
aberrations through the lens, over a 7 mm such deficits could be overcome through reg- lenses need to be redesigned/modified
pupil diameter, from the second to the fifth ular perceptual learning/training.111 At pre- and refitted on a regular basis, which
order (inclusive); these measurements were sent, only one study has explored the impact could prove to be financially inconvenient
then used to design the final customised of allowing keratoconic patients to habituate to the patient, while also requiring sub-
lens optic. Specifically, Marsack et al.53 to a wavefront-guided, customised scleral stantial clinical chair time
explained that the ‘final’ customised scleral lens correction for a substantial time • poor customised lens care by the patient
lenses contained the baseline level of period.109 Using the customisation methodol- (for example damage to the contact lenses
defocus correction that was previously ogy outlined in their previous investigation,53 during cleaning, storing or handling)
incorporated into the ‘intermediate’ lens, Hastings et al.109 conducted a randomised • customised contact lenses can only opti-
plus compensation for the residual objective study, which included a crossover design, all- mally correct HOAs at one given pupil
lower-order and HOAs measured via aber- owing comparisons between ‘conventional’ size. When the patient’s pupil increases
rometry. Measurements of the ‘intermedi- scleral lenses (acting as a control) and beyond this size, the effectiveness of the
ate’ lens’ on-eye decentrations were made ‘customised’ scleral lenses, over two eight- aberration control will begin to reduce;
using a customised camera system; how- week (approximate) periods, for eight however, the patient might still experi-
ever, unlike Sabesan et al.,52 Marsack patients with keratoconus. Although the ence some degree of benefit.
et al.53 only measured their lens’ decen- authors reported that both sets of lenses Another issue to overcome in manufactur-
trations over a period of 10 seconds. were worn on a ‘daily’ basis, for each eight- ing customised scleral/soft contact lenses is
Overall, Marsack et al.53 reported that, week period, specific data on how many the need for highly specialised equipment.
compared to the ‘intermediate’ lenses, the hours of lens wear per day were not pres- The cost of buying and maintaining micron-
‘final’ customised lenses provided signifi- ented in their paper. Expectedly, Hastings precision lathe machines is typically high,
cantly lower magnitudes of residual mean et al.109 found that the mean HORMS error which therefore significantly increases the
lower-order RMS error (p < 0.001) and resid- reduced from +0.46 0.24 μm with the con- cost of producing customised lenses. A fur-
ual mean HORMS error (p < 0.02), over a ventional lenses, to +0.26 0.08 μm with the ther challenge is the requirement of specialist
6 mm pupil, for all 14 eyes. However, the customised lenses (p = 0.004). However, training for practitioners, to enable them to
authors’ individual patient data revealed although improved, the difference in the carry out the assessment, fitting and subse-
that three (designed for two patients) of mean area under the log contrast sensitivity quent modification of these complex lens
their 14 ‘final’ customised lenses actually function did not reach statistical significance designs.
induced more residual aberrations than the between the two lens types (conventional In order to gain the maximum benefit
‘intermediate’ lenses. The most likely reason lenses = 13.91 2.20 log units, customised from HOA customised corrections, another
for these anomalies was proposed to be on- lenses 15.82 2.34 log units [p = 0.09]). Simi- difficulty to overcome is the need to
eye lens decentration, resulting in the larly, a non-significant improvement was also improve the efficacy of correcting lower-
wavefront-compensating optical zone reported for mean high-contrast logMAR acu- order terms with customised lenses, as
becoming misaligned with respect to the ity (conventional lenses = −0.03 0.09 log any under- or overcorrection of these
patient’s pupil. Nonetheless, the authors units, customised lenses −0.09 0.10 log terms may potentially diminish, or even
reported that, on average, 10 of the 14 eyes units [p = 0.07]). eliminate, any benefits gained by cor-
achieved a mean improvement of 1.5 lines recting the comparatively smaller-
of high-contrast monocular logMAR acuity, FURTHER LIMITATIONS TO THE USE OF magnitude HOA terms.
compared to their habitual mode of correc- CUSTOMISED LENSES Nonetheless, further technological advances
tion, which was similar to the findings of While some limitations of using either soft in the field of scleral/soft lens design are likely
Sabesan et al.52 or scleral customised contact lenses have to promote the emergence of customised
Even with their customised scleral lenses already been discussed in this review (for aberration-controlling lenses for keratoconic
in situ, both Sabesan et al.52 and Marsack example, on-eye lens decentration), there patients, particularly as improvements in visual
et al.53 noted that a significant reduction in are other more general limitations that also performance and comfortable wearing times,
HOAs did not always yield a significant need to be considered, with respect to the provided by such lenses, could reduce the rate
improvement in visual performance. This key correction of HOAs, which include: of keratoplasty in keratoconic patients, thereby
finding suggests there might be a degree of • optical limits that are set by diffraction; significantly reducing clinical issues related to
post-receptoral neural deficit present in these are related to the patient’s pupil size corneal graft surgery. Additionally, enhance-
keratoconic patients, which limits the degree • the limit of the photoreceptors’ ‘packing ments in optical correction, provided by
of visual improvement possible, even with density’ at the foveola centralis customised lenses, could lead to increased
the near-normal or better than normal levels • any errors in the manufacturing process of independence, particularly among young adult
of optical quality, as is provided by wearing a incorporating the required aberration mag- keratoconic patients, thereby contributing to
well-aligned customised wavefront correc- nitudes onto the customised lenses48,52 improvements in quality of life.
tion. Such a deficit may be attributable to • potential light scattering and/or glare
long-term exposure to an asymmetrically effects from the boundaries of the lens’
blurry retinal image.103 Further studies are customisation zone (more likely to be
REFERENCES
needed to explore this concept; for example, noticeable in scotopic conditions) 1. Rabinowitz YS. Keratoconus. Surv Ophthalmol 1998;
it is currently unclear if the degree of neural • changes in the patient’s habitual lower- 42: 297–319.
2. Shah S, Laiquzzaman M, Bhojwani R et al. Assess-
deficit is related to the severity of the order and/or HOAs due to disease pro- ment of the biomechanical properties of the cornea
patient’s disease and/or age, or whether gression would mean that the customised with the ocular response analyzer in normal and
41
keratoconic eyes. Invest Ophthalmol Vis Sci 2007; 48: incorporating correction for vertical coma aberra- 47. Choi J, Wee W, Lee J et al. Changes of ocular higher
3026–3031. tion. Ophthalmic Physiol Opt 2009; 29: 321–329. order aberration in on-and off-eye of rigid gas per-
3. Meek KM, Tuft SJ, Huang Y et al. Changes in collagen 27. Jinabhai A, Radhakrishnan H, O’Donnell C. Higher- meable contact lenses. Optom Vis Sci 2007; 84:
orientation and distribution in Keratoconus corneas. order aberrations in keratoconus: a review. Optom 42–51.
Invest Ophthalmol Vis Sci 2005; 46: 1948–1956. Pract 2009; 10: 141–160. 48. Chen M, Sabesan R, Ahmad K et al. Correcting ante-
4. Barr JT, Wilson BS, Gordon MO et al. Estimation of 28. Gobbe M, Guillon M. Corneal wavefront aberration rior corneal aberration and variability of lens move-
the incidence and factors predictive of corneal scar- measurements to detect keratoconus patients. Cont ments in keratoconic eyes with back-surface
ring in the Collaborative Longitudinal Evaluation of Lens Anterior Eye 2005; 28: 57–66. customized soft contact lenses. Opt Lett 2007; 32:
Keratoconus (CLEK) study. Cornea 2006; 25: 16–25. 29. Shankar H, Taranath D, Santhirathelagan CT et al. 3203–3205.
5. Zadnik K, Barr JT, Gordon MO et al. Biomicroscopic Repeatability of corneal first-surface wavefront aber- 49. Jeong TM, Yoon G. Customized correction of
signs and disease severity in Keratoconus. Cornea rations measured with Pentacam corneal topogra- wavefront aberrations in abnormal human eyes by
1996; 15: 139–146. phy. J Cataract Refract Surg 2008; 34: 727–734. using a phase plate and a customized contact lens.
6. Tan B, Baker K, Chen YL et al. How keratoconus influ- 30. Alio J, Shabayek M. Corneal higher order aberrations: J Korean Physical Soc 2006; 49: 121–125.
ences optical performance of the eye. J Vis 2008; a method to grade keratoconus. J Refract Surg 2006; 50. Sabesan R, Jeong TM, Carvalho LA et al. Vision
8: 1–10. 22: 539–545. improvement by correcting higher-order aberrations
7. Negishi K, Kumanomido T, Utsumi Y et al. Effect of 31. Sideroudi H, Labiris G, Giarmoulakis A et al. Repeat- with customized soft contact lenses in keratoconic
higher-order aberrations on visual function in ability, reliability and reproducibility of posterior cur- eyes. Opt Lett 2007; 32: 1000–1002.
keratoconic eyes with a rigid gas permeable contact vature and wavefront aberrations in keratoconic and 51. Yildiz E, Toklu MT, Vural ET et al. Change in accom-
lens. Am J Ophthalmol 2007; 144: 924–929. cross-linked corneas. Clin Exp Optom 2013; 96: modation and ocular aberrations in keratoconus
8. Jinabhai A, O’Donnell C, Radhakrishnan H. A compari- 547–556. patients fitted with scleral lenses. Eye Contact Lens
son between subjective refraction and aberrometry- 32. Jinabhai A, Radhakrishnan H, O’Donnell C. Repeat- 2018; 44: S50–S53.
derived refraction in keratoconus patients and con- ability of ocular aberration measurements in 52. Sabesan R, Johns L, Tomashevskaya O et al.
trol subjects. Curr Eye Res 2010; 35: 703–714. patients with keratoconus. Ophthalmic Physiol Opt Wavefront-guided scleral lens prosthetic device for
9. Jinabhai A, O’Donnell C, Radhakrishnan H et al. Forward 2011; 31: 588–594. keratoconus. Optom Vis Sci 2013; 90: 314–323.
light scatter and contrast sensitivity in keratoconic 33. Bayhan HA, Aslan Bayhan S, Muhafiz E et al. Repeat- 53. Marsack JD, Ravikumar A, Nguyen C et al. Wavefront-
patients. Cont Lens Anterior Eye 2012; 35: 22–27. ability of aberrometric measurements in normal and guided scleral lens correction in keratoconus. Optom
10. Nichols JJ, Steger-May K, Edrington TB et al. The rela- keratoconus eyes using a new Scheimpflug-Placido Vis Sci 2014; 91: 1221–1230.
tion between disease asymmetry and severity in topographer. J Cataract Refract Surg 2014; 40: 54. Yang B, Liang B, Liu L et al. Contrast sensitivity func-
keratoconus. Br J Ophthalmol 2004; 88: 788–791. 269–275. tion after correcting residual wavefront aberrations
11. Jinabhai A, Radhakrishnan H, O’Donnell C. Pellucid 34. Ortiz-Toquero S, Rodriguez G, de Juan V et al. repeat- during RGP lens wear. Optom Vis Sci 2014; 91:
corneal marginal degeneration: a review. Cont Lens ability of wavefront aberration measurements with a 1271–1277.
Anterior Eye 2011; 34: 56–63. placido-based topographer in normal and 55. Montalt JC, Porcar E, Espana-Gregori E et al. Visual
12. Romero-Jiménez M, Santodomingo-Rubido J, keratoconic eyes. J Refract Surg 2016; 32: 338–344. quality with corneo-scleral contact lenses for
Wolffsohn JS. Keratoconus: a review. Cont Lens Ante- 35. Shetty R, Kochar S, Grover T et al. Repeatability of a keratoconus management. Cont Lens Anterior Eye
rior Eye 2010; 33: 157–166. commercially available adaptive optics visual simula- 2018; 41: 351–356.
13. Jafri B, Lichter H, Stulting RD. Asymmetric keratoconus tor and aberrometer in normal and keratoconic 56. Shi Y, Queener HM, Marsack JD et al. Optimizing
attributed to eye rubbing. Cornea 2004; 23: 560–564. eyes. J Refract Surg 2017; 33: 769–772. wavefront-guided corrections for highly aberrated
14. Lafond G, Bazin R, Lajoie C. Bilateral severe 36. Artal P, Benito A, Tabernero J. The human eye is an eyes in the presence of registration uncertainty. J Vis
keratoconus after laser in situ keratomileusis in a example of robust optical design. J Vis 2006; 6: 1–7. 2013; 13: 1–15.
patient with forme fruste keratoconus. J Cataract 37. Artal P, Guirao A, Berrio E et al. Compensation of 57. Abdu M, Mohidin N, Mohd-Ali B. Visual performance
Refract Surg 2001; 27: 1115–1118. corneal aberrations by the internal optics in the and aberration associated with contact lens wear in
15. Maguire LJ, Lowry JC. Identifying progression of sub- human eye. J Vis 2001; 1: 1–8. patients with keratoconus: a pilot study. Clin Optom
clinical keratoconus by serial topography analysis. 38. Chen M, Yoon G. Posterior corneal aberrations and 2014; 6: 47–57.
Am J Ophthalmol 1991; 112: 41–45. their compensation effects on anterior corneal aber- 58. Gumus K, Kahraman N. A new fitting approach for
16. Weed KH, McGhee CN, MacEwen CJ. Atypical unilat- rations in keratoconic eyes. Invest Ophthalmol Vis Sci providing adequate comfort and visual performance
eral superior keratoconus in young males. Cont Lens 2008; 49: 5645–5652. in keratoconus: soft HydroCone (Toris K) lenses. Eye
Anterior Eye 2005; 28: 177–179. 39. Nakagawa T, Maeda N, Kosaki R et al. Higher-order Contact Lens 2016; 42: 225–230.
17. Auffarth GU, Wang L, Völcker HE. Keratoconus evalu- aberrations due to the posterior corneal surface in 59. Gumus K, Gire A, Pflugfelder SC. The impact of the
ation using the Orbscan topography system. patients with keratoconus. Invest Ophthalmol Vis Sci Boston ocular surface prosthesis on wavefront
J Cataract Refract Surg 2000; 26: 222–228. 2009; 50: 2660–2665. higher-order aberrations. Am J Ophthalmol 2011;
18. Demirbas N, Pflugfelder S. Topographic pattern and 40. Jinabhai A, Charman WN, O’Donnell C et al. Optical 151: 682–690.e2.
apex location of keratoconus on elevation topogra- quality for keratoconic eyes with conventional RGP 60. Marsack JD, Rozema JJ, Koppen C et al. Template-
phy maps. Cornea 1998; 17: 476–484. lens and simulated, customised contact lens correc- based correction of high-order aberration in
19. Barbero S, Marcos S, Merayo-Lloves J et al. Validation tions: a comparison. Ophthalmic Physiol Opt 2012; keratoconus. Optom Vis Sci 2013; 90: 324–334.
of the estimation of corneal aberrations from 32: 200–212. 61. Yoon G, Pantanelli S, MacRae S. Optimising the
videokeratography in keratoconus. J Refract Surg 41. Jinabhai A, O’Donnell C, Tromans C et al. Optical Shack-Hartmann wavefront sensor. In: Krueger RR,
2002; 18: 263–270. quality and visual performance with customised soft Applegate RA, MacRae S, eds. Wavefront Customized
20. Maeda N, Fujikado T, Kuroda T et al. Wavefront aber- contact lenses for keratoconus. Ophthalmic Physiol Visual Correction - The Quest for SuperVision 2, Chap-
rations measured with Hartmann–Shack sensor in Opt 2014; 34: 528–539. ter 16. Thorofare, New Jersey: Slack Inc., 2004.
patients with keratoconus. Ophthalmol 2002; 109: 42. Jinabhai A, Radhakrishnan H, O’Donnell C. Visual acu- pp. 131–136.
1996–2003. ity and ocular aberrations with different rigid gas 62. Lombardo M, Lombardo G. New methods and tech-
21. Lim L, Wei R, Tan D et al. Evaluation of higher order permeable lens fittings in keratoconus. Eye Contact niques for sensing the wave aberrations of human
ocular aberrations in patients with keratoconus. Lens 2010; 36: 233–237. eyes. Clin Exp Optom 2009; 92: 176–186.
J Refract Surg 2007; 23: 825–828. 43. Jinabhai A, Radhakrishnan H, Tromans C et al. Visual 63. Mihashi T, Hirohara Y, Bessho K et al. Intensity analy-
22. Schlegel Z, Lteif Y, Bains HS et al. Total, corneal, and performance and optical quality with soft lenses in sis of Hartmann–Shack images in cataractous,
internal ocular optical aberrations in patients with keratoconus patients. Ophthalmic Physiol Opt 2012; keratoconic, and normal eyes to investigate light
keratoconus. J Refract Surg 2009; 25: S951–S957. 32: 100–116. scattering. Jpn J Ophthalmol 2006; 50: 323–333.
23. Kosaki R, Maeda N, Bessho K et al. Magnitude and 44. Marsack JD, Parker KE, Applegate RA. Performance 64. Shinto H, Saita Y, Nomura T. Shack-Hartmann wavefront
orientation of Zernike terms in patients with of wavefront-guided soft lenses in three keratoconus sensor with large dynamic range by adaptive spot sea-
keratoconus. Invest Ophthalmol Vis Sci 2007; 48: subjects. Optom Vis Sci 2008; 85: 1172–1178. rch method. Appl Optics 2016; 55: 5413–5418.
3062–3068. 45. Marsack JD, Parker KE, Niu Y et al. On-eye 65. de Oliveira OG, de Lima Monteiro DW. Optimization
24. Shah S, Naroo S, Hosking S et al. Nidek OPD-scan performance of custom wavefront-guided soft con- of the Hartmann–Shack microlens array. Opt Lasers
analysis of normal, keratoconic, and penetrating ker- tact lenses in a habitual soft lens-wearing Eng 2011; 49: 521–525.
atoplasty eyes. J Refract Surg 2003; 19: S255–S259. keratoconic patient. J Cataract Refract Surg 2007; 66. Thibos LN, Hong X. Clinical applications of the Shack-
25. Bühren J, Kühne C, Kohnen T. Defining subclinical 23: 960–964. Hartmann Aberrometer. Optom Vis Sci 1999; 76:
keratoconus using corneal first-surface higher-order 46. Marsack JD, Parker KE, Pesudovs K et al. Uncorrected 817–825.
aberrations. Am J Ophthalmol 2007; 143: 381–389. wavefront error and visual performance during RGP 67. Thibos LN, Applegate RA. Assessment of optical qual-
26. Katsoulos C, Karageoriadis L, Vasileiou N et al. Cus- wear in keratoconus. Optom Vis Sci 2007; 84: ity. In: Krueger RR, Applegate RA, MacRae S, eds. Cus-
tomized hydrogel contact lenses for keratoconus, 463–470. tomized Corneal Ablation - The Quest for SuperVision,
42
Chapter 6. Thorofare, New Jersey: Slack Inc., 2001. 82. Mihaltz K, Kranitz K, Kovacs I et al. Shifting of the line 97. Charman WN. Rigid lens optics. In: Efron N,
pp. 55–63. of sight in keratoconus measured by a Hartmann– ed. Contact Lens Practice, Chapter 14. Oxford, London:
68. Llorente L, Marcos S, Dorronsoro C et al. Effect of Shack sensor. Ophthalmol 2010; 117: 41–48. Butterworth-Heinemann, 2002. pp. 171–176.
sampling on real ocular aberration measurements. 83. Ninomiya S, Fujikado T, Kuroda T et al. Changes of 98. Hartwig A, Atchison DA. Analysis of higher-order
J Opt Soc Am A Opt Image Sci Vis 2007; 24: 2783–2796. ocular aberration with accommodation. aberrations in a large clinical population. Invest
69. Pfund J, Lindlein N, Schwider J. Dynamic range Am J Ophthalmol 2002; 134: 924–926. Ophthalmol Vis Sci 2012; 53: 7862–7870.
expansion of a Shack--Hartmann sensor by use of a 84. Montés-Micó R, Alio JL, Munoz G et al. Temporal 99. de Sanctis U, Loiacono C, Richiardi L et al. Sensitivity
modified unwrapping algorithm. Opt Lett 1998; 23: changes in optical quality of air-tear film Interface at and specificity of posterior corneal elevation measured
995–997. anterior cornea after blink. Invest Ophthalmol Vis Sci by Pentacam in discriminating keratoconus/subclinical
70. Lindlein N, Pfund J, Schwider J. Expansion of the 2004; 45: 1752–1757. keratoconus. Ophthalmol 2008; 115: 1534–1539.
dynamic range of a Shack-Hartmann sensor by using 85. Radhakrishnan H, Jinabhai A, O’Donnell C. Dynamics 100. Rocha KM, Vabre L, Chateau N et al. Enhanced visual
astigmatic microlenses. Opt Eng 2000; 39: 2220–2225. of ocular aberrations in keratoconus. Clin Exp Optom acuity and image perception following correction of
71. Lindlein N, Pfund J, Schwider J. Algorithm for expan- 2010; 93: 164–174. highly aberrated eyes using an adaptive optics visual
ding the dynamic range of a Shack-Hartmann sensor 86. Williams DR, Yoon GY, Porter J et al. Visual benefit of simulator. J Refract Surg 2010; 26: 52–56.
by using a spatial light modulator array. Opt Eng correcting higher order aberrations of the eye. 101. Sabesan R, Ahmad K, Yoon G. Correcting highly aber-
2001; 40: 837–840. J Refract Surg 2000; 16: S554–S559. rated eyes using large-stroke adaptive optics.
72. Ares M, Royo S, Caum J. Shack-Hartmann sensor 87. Guirao A, Porter J, Williams DR et al. Calculated J Refract Surg 2007; 23: 947–952.
based on a cylindrical microlens array. Opt Lett 2007; impact of higher-order monochromatic aberrations 102. Sabesan R, Yoon G. Visual performance after cor-
32: 769–771. on retinal image quality in a population of human recting higher order aberrations in keratoconic eyes.
73. Pantanelli S, MacRae S, Jeong TM et al. Characteriz- eyes: erratum. J Opt Soc Am A 2002; 19: 620–628. J Vis 2009; 9: 1–10.
ing the wave aberration in eyes with keratoconus or 88. Guirao A, Williams DR, Cox I. Effect of rotation and 103. Sabesan R, Yoon G. Neural compensation for long-
penetrating keratoplasty using a high–dynamic range translation on the expected benefit of an ideal term asymmetric optical blur to improve visual per-
wavefront sensor. Ophthalmol 2007; 114: 2013–2021. method to correct the eye’s higher-order aberrations. formance in keratoconic eyes. Invest Ophthalmol Vis
74. Yoon G, Pantanelli S, Nagy LJ. Large-dynamic-range J Opt Soc Am A Opt Image Sci Vis 2001; 18: 1003–1015. Sci 2010; 51: 3835–3839.
Shack-Hartmann wavefront sensor for highly aber- 89. López-Gil N, Castejón-Mochón JF, Fernández- 104. Smirnov MS. Measurement of the wave aberration
rated eyes. J Biomed Opt 2006; 11: 0305021- 0305023. Sánchez V. Limitations of the ocular wavefront cor- of the human eye. Biophysics 1962; 7: 766–795.
75. Marsack JD, Pesudovs K, Sarver E et al. Impact of rection with contact lenses. Vision Res 2009; 49: 105. Guirao A, Cox I, Williams DR. Method for optimizing
Zernike-fit error on simulated high-and low-contrast 1729–1737. the correction of the eye’s higher-order aberrations
acuity in keratoconus: implications for using Zernike- 90. de Brabander J, Chateau N, Marin G et al. Simulated in the presence of decentrations. J Opt Soc Am A Opt
based corrections. J Opt Soc Am A Opt Image Sci Vis optical performance of custom Wavefront soft con- Image Sci Vis 2002; 19: 126–128.
2006; 23: 769–776. tact lenses for keratoconus. Optom Vis Sci 2003; 80: 106. Yoon G, Jeong TM. Effect of the movement of cus-
76. Born M, Wolf E. Geometrical theory of aberrations. 637–643. tomized contact lens on visual benefit in abnormal
In: Principles of Optics, Chapter 5, 7th 91. Radhakrishnan H, Charman WN. Age-related eyes. J Vis 2003; 3: 38.
ed. Cambridge, UK: Cambridge University Press, changes in ocular aberrations with accommodation. 107. Haegerstrom-Portnoy G, Brabyn J, Schneck ME
1999. pp. 228–260. J Vis 2007; 7: 1–21. et al. The SKILL card. An acuity test of reduced lumi-
77. Ríos S, Acosta E, Bará S. Hartmann sensing with 92. López-Gil N, Castejón-Mochón JF, Benito A et al. nance and contrast. Smith- Kettlewell Institute low
Albrecht grids. Opt Commun 1997; 133: 443–453. Aberration generation by contact lenses with luminance. Invest Ophthalmol Vis Sci 1997; 38:
78. Gatinel D, Malet J, Dumas L. Polynomial decomposi- aspheric and asymmetric surfaces. J Refract Surg 207–218.
tion method for ocular wavefront analysis. J Opt Soc 2002; 18: S603–S609. 108. Kollbaum PS, Jansen M, Thibos LN et al. Validation of
Am A Opt Image Sci Vis 2018; 35: 2035–2045. 93. Tomlinson A, Ridder WH, Watanabe R. Blink-induced an off-eye contact lens Shack-Hartmann wavefront
79. Thibos LN, Applegate RA, Schwiegerling JT et al. VSIA- variations in visual performance with Toric soft con- aberrometer. Optom Vis Sci 2008; 85: E817–E828.
taskforce-members. Standards for reporting the tact lenses. Optom Vis Sci 1994; 71: 545–549. 109. Hastings GD, Applegate RA, Nguyen LC et al. Com-
optical aberrations of eyes. In: Lakshminarayan V, 94. Young G, McIlraith R, Hunt C. Clinical evaluation of parison of wavefront-guided and best conventional
ed. Vision Sciences and It’s Applications. Vol 35 of OSA factors affecting soft Toric lens orientation. Optom scleral lenses after habituation in eyes with corneal
Trends in Optics and Photonics Series. Washington, Vis Sci 2009; 86: E1259–E1266. ectasia. Optom Vis Sci 2019; 96: 238–247.
DC: Optical Society of America, 2000. pp. 232–244. 95. Edrington TB. A literature review: the impact of rota- 110. Ticak A, Marsack JD, Koenig DE et al. A comparision
80. Mandell RB, Chiang CS, Klein SA. Location of the tional stabilization methods on toric soft contact lens of three methods to increase scleral contact lens on-
major corneal reference points. Optom Vis Sci 1995; performance. Cont Lens Anterior Eye 2011; 34: eye stability. Eye Contact Lens 2015; 41: 386–390.
72: 776–784. 104–110. 111. Sabesan R, Barbot A, Yoon G. Enhanced neural func-
81. Cheng X, Himebaugh N, Kollbaum PS et al. Test- 96. Griffiths M, Zahner K, Collins MJ et al. Masking of tion in highly aberrated eyes following perceptual
retest reliability of clinical Shack-Hartmann measure- irregular corneal topography with contact lenses. learning with adaptive optics. Vision Res 2017; 132:
ments. Invest Ophthalmol Vis Sci 2004; 45: 351–360. CLAO J 1998; 24: 76–81. 78–84.
43
INVITED REVIEW
Optical changes and visual performance with orthokeratology
Augustine N Nti OD Orthokeratology has undergone drastic changes since first described in the early 1960s. The
David A Berntsen OD PhD original orthokeratology procedure involved a series of lenses to flatten the central cornea
The Ocular Surface Institute, College of Optometry, and was plagued by variable results. The introduction of highly oxygen-permeable lens
University of Houston, Houston, Texas, USA materials that can be worn overnight, corneal topography, and reverse-geometry lens
E-mail: [email protected] designs revolutionised this procedure. Modern overnight orthokeratology causes rapid, reli-
able, and reversible reductions in refractive error. With modern designs, patients can wear
lenses overnight, remove them in the morning, and see clearly throughout the day without
the need for daytime refractive correction. Modern reverse-geometry lens designs cause
central corneal flattening and mid-peripheral corneal steepening that provides clear foveal
vision while simultaneously causing a myopic shift in peripheral retinal defocus. The periph-
eral myopic retinal defocus caused by orthokeratology is hypothesised to be responsible for
reductions in myopia progression in children fitted with these lenses. This paper reviews
the changes in orthokeratology lens design that led to the reverse-geometry ortho-
keratology lenses that are used today and the optical changes these lenses produce. The
optical changes reviewed include changes in refractive error and their time course, high-
and low-contrast visual acuity changes, changes in higher-order aberrations and visual qual-
ity metrics, changes in accommodation, and changes in peripheral defocus caused by ortho-
keratology. The use of orthokeratology for myopia control in children is also reviewed, as
are hypothesised connections between orthokeratology-induced myopic peripheral defocus
and slowed myopia progression in children, and safety and complications associated with
lens wear. A better understanding of the ocular and optical changes that occur with ortho-
Submitted: 2 April 2019 keratology will be beneficial to both clinicians and patients in making informed decisions
Revised: 17 June 2019 regarding the utilisation of orthokeratology. Future research directions with this lens modal-
Accepted for publication: 18 June 2019 ity are also discussed.
Key words: aberrations, myopia, myopia control, orthokeratology, visual acuity
Uncorrected refractive errors were the lead- myopia, the procedure can also be used to Traditional orthokeratology involved
ing global cause of moderate or severe temporarily reduce hyperopia using wearing a series of flatter-fitting rigid
vision impairment in 2015.1 Flaxman et al.1 steeper-fitting contact lenses that increase lenses during the day that gradually flat-
noted that although the prevalence of central corneal curvature. Toric ortho- tened the central cornea further with
uncorrected refractive error is declining, the keratology can also be used to temporarily each lens until myopia was eliminated.
total number of people with vision impair- correct moderate-to-high levels of astigma- After the targeted reduction in myopia
ment due to uncorrected refractive error tism (greater than 1.25 D). was achieved, lens wear was gradually
increased from 6.2 million in 1990 to 7.4 Attempts to manipulate the cornea to reduced. With time, this wearing schedule
million in 2015. Options to correct refractive reduce or correct refractive error goes far allowed for periods in the day during
errors include spectacles, contact lenses, back to the early Japanese who placed small which no lenses were worn. These early
and refractive surgery. bags of shot on the eyelids overnight to attempts at orthokeratology unfortunately
Orthokeratology, a specialty contact lens reduce myopia.3 However, it was not until produced variable results, unpredictable
option, is the temporary ‘reduction, modifi- the 1950s that clinicians realised that flat vision, and did not gain widespread
cation or elimination of refractive anomalies fitted rigid contact lenses produced corneal acceptance and usage.
by the programmed application of contact changes that could eliminate myopia. Early In 1989, Wlodyga and Stoyan produced
lenses or other related procedures’.2 Myopic work on traditional orthokeratology was the first reverse-geometry orthokeratology
orthokeratology involves fitting lenses flatter pioneered by Jessen in 1962 when he wrote lenses.4 These lenses had a secondary curve
than the flat meridian to reduce the curva- a paper on ‘orthofocus techniques’, which that was steeper than the base curve. This
ture of the central cornea. While ortho- described how rigid contact lenses can be design allowed for better centration of the
keratology is predominantly used to correct used to correct myopia. lens and more rapid reduction in refractive
44
Optical changes with orthokeratology Nti and Berntsen
error, leading to the term ‘accelerated

orthokeratology’.
This paper reviews the optical changes
that occur with orthokeratology, the effect of
orthokeratology on vision, and future direc-
tions involving the use of orthokeratology.
Modern orthokeratology
Four developments significantly influenced

modern orthokeratology: reverse-geometry
designs, highly oxygen-permeable lens
materials, corneal topography, and
computer-controlled lathing, which allowed
for consistent manufacturing of highly spe-
cialised lenses.
Since the introduction of reverse-
geometry lens designs in 1989, all modern
orthokeratology lenses use some modifica-
tion of this design. ‘Reverse geometry’ refers
to the first back peripheral curve radius
being steeper than the base curve
(Figure 1).
Newer lens materials with high oxygen
permeability have enabled modern ortho-
keratology lenses to be worn safely over-
night, thus the term ‘overnight
orthokeratology’. These lenses mould or
reshape the cornea overnight and provide
patients with clear unaided vision during the
day after lens removal. These corneal
changes are reversible, and patients wear
their lenses each night to maintain clear
unaided vision during waking hours. The
widespread availability of corneal topogra-
phers allows practitioners to measure the
centration and effects of orthokeratology
lenses on the cornea and appropriately
modify treatment.
Although orthokeratology can be used to Figure 1. Representative cross-section and fluorescein pattern of a well-fitted ortho-
correct astigmatism, hyperopia and keratology reverse-geometry lens (not drawn to scale). Note that some lens designs
myopia, orthokeratology is predominantly may have additional back-surface curves beyond what is shown. The back-surface
used to reduce myopia. Much of the rest curves are labelled as follows. A: Base curve. B: Reverse curve. C: Alignment curve.
of this review will focus on myopic D: Peripheral curve.
orthokeratology.
Corneal changes cornea happens mostly in the anterior por- annulus, there was greater steepening of the
tion of the cornea.5–10 The central corneal temporal sector compared with the nasal sec-
Myopic orthokeratology epithelium thins, while the mid-peripheral tor. They found no asymmetries in the verti-
The goal of myopic orthokeratology is to cornea thickens.5,9,11,12 cal sectors.
flatten the central cornea to reduce myopia, Orthokeratology can also lead to asym- Based on histological sectioning of cat
and the mid-peripheral cornea is steepened metric changes in the cornea.13 In a retro- corneas after being fitted with ortho-
in the process. Figure 2 shows corneal spective study, Maseedupally et al.13 reported keratology, Choo et al.14 reported that
topography for a patient before and after that after 14 days of orthokeratology, there mid-peripheral corneal thickening occurs
orthokeratology and the resulting corneal was greater flattening in the temporal sector in the epithelium; however, other studies
power changes. Based on studies measuring of the central cornea compared with the suggest this thickening is stromal in ori-
corneal changes, the flattening of the nasal sector. In the paracentral corneal gin.11,15,16 Although some studies reported
45
Figure 2. Corneal topography tangential maps of an eye A: before being fitted with orthokeratology and B: after orthokeratology
fitting. C: A power difference map shows the changes in corneal power after being fitted with orthokeratology. Warmer colours
represent higher corneal power. Topography maps courtesy of Maria Walker, OD MS.
flattening of the posterior cornea,15,16 it is Orthokeratology generally does not induce

Optical changes with myopic
generally accepted that orthokeratology astigmatism in patients with spherical
orthokeratology
primarily causes flattening of the anterior refractive errors.42,43 In myopic patients
cornea without bending the whole with astigmatism, standard spherical ortho-
cornea.5,7,10,17 Refractive error changes and keratology does not consistently change the
Significant corneal changes occur as early stability of changes amount of astigmatism. Some studies have
as 10 minutes after an orthokeratology lens Multiple studies have evaluated refractive reported that orthokeratology causes signifi-
is worn.6 Thus, just one night of ortho- error changes and their stability using over- cant changes in astigmatism43–45 while
keratology improves visual acuity and night orthokeratology.18,26–39 Nichols et al.39 others have found no significant change in
reduces myopia,9,17,18 providing a rapid and conducted a prospective study of refractive astigmatism.29,31
reversible means of correcting myopia. error changes in patients undergoing over- For patients with moderate-to-high levels
A meta-analysis by Li et al.19 found that the night orthokeratology and found that most of astigmatism (more than −1.50 D), toric
reduction in corneal thickness occurs within of the reduction in myopia occurred during orthokeratology is effective at reducing the
the first week of orthokeratology and the first seven days of overnight lens wear. amount of astigmatism. One of the earliest
remains stable thereafter. There was no significant reduction in myo- case reports utilising toric orthokeratology
Studies suggest that the rate of corneal pia after the seventh day. Other studies reported reductions in astigmatism of up to
changes with orthokeratology depends on have similarly reported that most of the −2.50 D three weeks after the start of treat-
the amount of targeted refractive change; refractive error reduction occurs within the ment, although there was variability in resid-
orthokeratology lenses targeted to correct first week of orthokeratology.18,30 ual astigmatism over time.46 Subsequent
higher amounts of myopia cause faster cor- Orthokeratology lenses generally have studies have shown the effectiveness of
neal changes than lenses to correct lower approval to correct myopia up to −5.00 D to toric orthokeratology. Chen et al.47 reported
amounts of myopia.8,20 However, how long −6.00 D, although this varies by country and a 79 per cent reduction in refractive astig-
it takes the cornea to return to its normal lens design. Eye-care providers may choose matism after one month of toric ortho-
shape depends on the duration of treat- to fit lenses off-label to correct higher keratology in myopic children with
ment. The longer orthokeratology lenses amounts of myopia depending on the laws moderate-to-high astigmatism (−1.25 up to
are used, the longer it takes the cornea to of the country in which they practise. For −3.50 D of astigmatism).
fully return to its normal shape,21 and patients with myopia higher than the Since myopic orthokeratology temporarily
reversibility may be slower in some approved limits of a particular lens design, flattens the central cornea, there is a grad-
children.22 an option is partial-reduction ortho- ual myopic regression over the course of
keratology, where a portion of the myopia is the day once orthokeratology lenses are
Hyperopic orthokeratology corrected by orthokeratology with the resid- removed and the cornea slightly steepens;
Hyperopic orthokeratology induces a myo- ual myopia corrected by spectacles during however, this myopic regression is minimal
pic shift in refractive error by steepening the day to provide clear vision.40 and accounted for by targeting a refractive
the central cornea.23,24 Hyperopic ortho- Overnight orthokeratology using spherical error of roughly +0.50 D in the morning.
keratology flattens the mid-peripheral cor- lens designs is generally indicated for After 28 days of orthokeratology in a group
nea23,24 by thickening the central corneal patients with low-to-moderate (−1.50 D or of myopic subjects, Sorbara et al.33 found
epithelium and thinning the mid-peripheral less) astigmatism. This indication is because only a −0.32 D change in refractive error
corneal epithelium.25 Similar to myopic moderate-to-high astigmatism can cause over a 14-hour period after removing lenses
orthokeratology, corneal changes are rapid decentration of orthokeratology lenses in the morning. Comparable results
(significant changes after one night of lens which can lead to decentred treatment, were reported by Soni et al.,18 who found a
wear) and reversible.23 poor vision, and may induce astigmatism.41 −0.37 D change in refractive error over a
46
12-hour period after lens removal in myopic successful orthokeratology wear have been acuity after orthokeratology is associated
subjects who had been wearing ortho- reported to be relatively stable. Although with increases in spherical aberration. These
keratology lenses for 30 days. Similarly, two studies reported minor fluctuations in reductions in low-contrast visual acuity are
Stillitano et al.48 found a −0.33 D change in spherical aberration over the course of the also greater as pupil size increases.31
refractive error over a 10-hour period after day, the other higher-order aberration As one would expect based on increased
six months of orthokeratology in myopic terms were stable.31,48 While some studies higher-order aberrations and reduced low-
subjects. have reported increases in horizontal coma contrast visual acuity reported after ortho-
Importantly, the refractive changes are after orthokeratology, others have found no keratology, retinal image quality is affected.
reversible. Kobayashi et al.36 reported that increase in coma.28,31,32 This discrepancy in Berntsen55 reported retinal image quality in
after one year of orthokeratology, the whether horizontal coma is increased by 20 myopic subjects before and one month
changes in refractive error were reversible. orthokeratology may be attributable to dif- after orthokeratology using six single-valued
In their study, 15 myopic subjects wore ferences in lens centration between studies. retinal image quality metrics previously
orthokeratology lenses for 52 weeks and reported to be highly correlated with low-
then discontinued lens wear. Spherical Visual acuity and visual quality contrast visual acuity. All six metrics showed
equivalent refractive error returned to base- metrics a significant reduction in retinal image qual-
line one month after discontinuation of lens As expected, orthokeratology leads to signif- ity after orthokeratology. Five of the six
wear, demonstrating that the corneal icant improvements in uncorrected high- image quality metrics were correlated with
changes are not permanent. and low-contrast visual acuity due to the increases in positive spherical aberration
elimination of uncorrected myopia.27,33,39 and reductions in low-contrast visual acuity
Higher-order aberration changes These improvements in uncorrected high- after orthokeratology. An example of a point
While orthokeratology causes a significant contrast visual acuity are relatively stable spread function before and after ortho-
reduction in myopia, the combination of throughout the day, with very little regres- keratology is shown in Figure 3. Although
central corneal flattening and mid- sion. After 30 days of orthokeratology, high-contrast visual acuity is not affected by
peripheral corneal steepening leads to a sig- Nichols et al.39 observed less than a 0.02 orthokeratology, these results demonstrate
nificant increase in higher-order logMAR (one letter) reduction in uncorrected that increased higher-order aberrations due
aberrations.31,32,34,48–51 Joslin et al.28 con- high-contrast visual acuity over eight hours. to orthokeratology influence visual
ducted one of the earliest studies that mea- Other researchers have reported similar performance.
sured the effect of reverse-geometry results.18,33 Orthokeratology can also pro-
orthokeratology lenses on higher-order vide unaided high-contrast visual acuity that Accommodation changes
aberrations. They found a significant is comparable to spectacles. In a study by Several studies have examined accommoda-
increase in higher-order aberrations after Sorbara et al.,33 89 per cent of subjects tive changes with orthokeratology. Felipe-
one month of wear for both a 3-mm and wearing spectacles had 6/6 or better visual Marquez et al.56 conducted a prospective
6-mm pupil diameter. For the 6-mm pupil acuity while 83 per cent of the ortho- study of 21 young adults wearing spectacles
diameter, mean higher-order root-mean- keratology subjects achieved the same level and 51 similar adults wearing ortho-
square (RMS) wavefront error increased by of vision without correction. Similarly, after keratology lenses. After three months of
0.425 μm, with the highest individual 28–60 days of successful orthokeratology, orthokeratology treatment, there were no
Zernike term increase of 0.306 μm being for uncorrected low-contrast visual acuity is significant changes in negative relative
spherical aberration (Z40). also stable throughout the day with accommodation, positive relative accommo-
Similarly, Berntsen et al.31 reported an decreases of between two to three dation, amplitude of accommodation,
average increase in higher-order RMS for a letters.33,39 accommodative lag, or monocular accommo-
5-mm pupil diameter after one month of As described earlier, patients wearing dative facility, and no differences in these
orthokeratology lens wear of 0.180 μm and orthokeratology lenses for one year had accommodative measurements between
an increase in spherical aberration of up to their refractive error return to baseline levels
0.186 μm. They also reported that while within one month of discontinuing ortho-
orthokeratology caused significant increases keratology. As expected, the same study also
in third- through sixth-order RMS for a reported that uncorrected visual acuity ret-
5-mm pupil, the increase in positive spheri- urned to baseline values roughly one month
cal aberration was a major contributor to after discontinuing orthokeratology.36
the increase in higher-order aberrations. As Best-corrected high-contrast visual acuity
well, Stillitano et al.52 later reported a before and during orthokeratology are also
0.39 μm increase in higher-order aberra- not significantly different, demonstrating
tions for a 6.5 mm pupil diameter, also that increased higher-order aberrations due
reporting a significant increase in spherical to orthokeratology do not have a clinically Figure 3. Higher-order aberration point
aberration after orthokeratology. Most meaningful effect on high-contrast acu- spread functions (PSFs) from an eye with
higher-order aberrations have been ity.31,35,54 However, best-corrected low- −3.00 D of spherical myopia at baseline
reported to stabilise within one week of contrast visual acuity is reduced by up to (left) and one month after ortho-
beginning orthokeratology treatment.53 0.12 logMAR (six letters) due to orthokeratology (right). Change in spherical
Higher-order aberration levels throughout keratology when compared to spectacle cor- aberration = 0.24 microns (5 mm pupil
the day after one month or more of rection. This reduction in low-contrast visual diameter).55
47
spectacle and orthokeratology-wearing sub- retina) compared with central refraction relative peripheral myopic defocus after
jects. Kang et al.57 also examined the effect along the line of sight.62 This relative periph- orthokeratology.37,38,64–67 The simultaneous
of orthokeratology on accommodative facility eral hyperopia has been hypothesised to creation of peripheral myopic defocus while
in 15 myopic young adults and found no sig- promote myopia progression.63 Studies still providing clear foveal vision has been
nificant change in facility after 28 nights of measuring peripheral refraction of myopic hypothesised to be the reason why ortho-
orthokeratology. eyes before and after orthokeratology have keratology lenses have been reported to
However, Gifford et al.58 found a lower found that while orthokeratology corrects slow the progression of myopia. Changes in
accommodative lag in orthokeratology sub- central myopia, mid-peripheral steepening relative peripheral refraction reported in
jects compared with control subjects wear- of the cornea yields a myopic shift in periph- some of these studies are shown in
ing single-vision contact lenses in a eral refractive error that results in periph- Figure 4.
retrospective study. In a subsequent eral myopic defocus after orthokeratology These changes in peripheral refraction
12-month prospective study, Gifford et al.59 (light focused in front of the retina). occur rapidly after the start of ortho-
reported improved accommodative Queiros et al.37 measured peripheral keratology. Kang and Swarbrick65 measured
responses with orthokeratology. They mea- refraction along the horizontal meridian of the time course of changes in peripheral
sured negative relative accommodation, 28 myopic subjects before and after one refraction with orthokeratology over two
positive relative accommodation, and month of orthokeratology. They reported a weeks, and reported that the most signifi-
accommodative lag in myopic children and hyperopic change in central refractive error cant change in mean peripheral refraction
young adults. Baseline measurements were (elimination of uncorrected myopia) within occurred after the first night of lens wear.
performed while wearing single-vision con- the central 20 of retinal eccentricity, no They also found no significant differences
tact lenses, and measurements were change in mean spherical equivalent at 25 between the mean spherical equivalent
repeated after one month and 12 months of eccentricity, and a myopic shift in spherical refraction measured after seven nights of
orthokeratology lens wear. There was no equivalent beyond 25 . They also found that orthokeratology lens wear and those mea-
significant change in negative relative treating greater amounts of myopia with sured after 14 nights of lens wear.
accommodation from baseline; however, orthokeratology resulted in greater myopic Kang et al.64 also conducted a study to
there was a significant increase in positive shifts in peripheral refractive error at eccen- determine whether changing the optic zone
relative accommodation in both children tricities of 20 or more. diameter or peripheral tangent curve of an
and adults beginning one month after Multiple studies have also reported that orthokeratology lens alters changes
starting orthokeratology. Accommodative the reshaping of the cornea with ortho- observed in peripheral refraction. They
lag decreased after one month of orthokeratology results in a conversion of relative measured peripheral refraction of 17 myopic
keratology in children, and after 12 months peripheral hyperopic defocus in the horizon- subjects after wearing orthokeratology
in adults. tal meridian before orthokeratology to lenses for two weeks. After a washout
The strongest evidence of changes in
accommodation with orthokeratology
comes from a randomised study of 240 myo-
pic children conducted by Han et al.60 Chil-
dren were randomly assigned to wear either
single-vision spectacles (n = 90), ortho-
keratology (n = 90), or multifocal spectacles
(n = 60) with concentric rings designed to
reduce paracentral defocus. In children
wearing orthokeratology, accommodative
lag was significantly lower after one year
when compared to baseline and to the
single-vision control group. Accommodative
facility also improved in the orthokeratology
group. These results demonstrate that long-
term orthokeratology improves accommo-
dative accuracy and facility. Gifford et al.61
reported that increased accommodation in
myopic young adults undergoing ortho-
keratology could be an adaptation to coun-
teract the increase in positive spherical
aberration caused by orthokeratology.
Figure 4. Spherical equivalent relative peripheral refraction (in dioptres) measured
Peripheral optics by autorefraction across the horizontal meridian of the eye from several studies
Myopic eyes generally have relative periph- before orthokeratology (solid lines with filled symbols) and after orthokeratology
eral hyperopia in the horizontal meridian, (dashed lines with open symbols). Square symbols = Kang and Swarbrick38 (n = 16),
meaning refractive error in the periphery is triangle symbols = Kang and Swarbrick67 (n = 19), circle symbols = Gonzalez-Meijome
more hyperopic (light focused behind the et al.66 (n = 34).
48
period, subjects wore a similar ortho- A similar result was reported by Ritchey retina. Several results from animal studies
keratology lens with a smaller optic zone et al.,72 in which orthokeratology was have shown that peripheral hyperopic
diameter on one eye and a steeper periph- reported to reduce dependence on correc- defocus can lead to myopia progression,
eral tangent curve on the other eye. They tion versus extended-wear soft contact while peripheral myopic defocus can slow
reported that the change in optic zone lenses. Berntsen et al.73 found that com- myopia progression.79 Based on these stud-
diameter and peripheral tangent curve did pared to their habitual correction of either ies, it is thought that the peripheral myopic
not lead to significant changes in peripheral spectacles or soft contact lenses, depen- defocus created by orthokeratology serves
refraction when compared with the pattern dence on correction and visual symptoms as a signal to slow the growth of the eye,
of peripheral refraction produced by the of participants were reduced after one reducing the progression of myopia. There
original lenses. Kang and Swarbrick68 also month of orthokeratology, although glare has also been a suggestion that visual sig-
found no evidence of differences in periph- increased. nals from different meridians of the retina
eral refraction when comparing three differ- Santodomingo-Rubido et al.74 adminis- may be differentially effective in influencing
ent orthokeratology lens designs. The lenses tered a paediatric refractive error profile central refractive error changes,80 but it is
differed in the total diameter, sphericity of questionnaire to children assigned to wear still unclear whether myopic defocus in spe-
the base curve, the number of reverse cur- orthokeratology or single-vision spectacles cific meridians of the retina are more effec-
ves, and the design of both the reverse and and found they rated orthokeratology as tive at slowing myopia progression. Work in
alignment curves. being better than spectacles in terms of this area over the last two decades has
Although most studies have only evalu- overall vision, far distance vision, symptoms, progressed from case reports81 and pilot
ated changes in peripheral refraction in the appearance, satisfaction, effect on activities, studies to randomised, controlled clinical
horizontal meridian, myopic shifts in refrac- academic performance, handling, and peer trials evaluating the efficacy of overnight
tion with orthokeratology have also been perceptions. Zhao et al.75 also reported orthokeratology for slowing the progression
reported in the vertical meridian.67,69 Kang higher vision-related quality of life with of myopia in children. Figure 5 summarises
and Swarbrick67 measured peripheral orthokeratology compared to spectacles in the myopia control effects reported.
refraction along the vertical meridian in myopic children with 75 per cent of the par- Cho et al.82 conducted the first pilot study
19 myopic subjects before and after ticipants preferring orthokeratology to spec- to evaluate whether orthokeratology could
14 nights of orthokeratology. Subjects had tacles. Additional studies have reported potentially slow eye growth in myopic chil-
peripheral myopia in the vertical meridian subjective satisfaction with orthokeratology dren. Their study found that after two years
at baseline, and orthokeratology caused a on a visual analogue scale, with participants of orthokeratology, axial elongation was
myopic shift in peripheral refraction that rating vision as 7.8/10 in one study76 and reduced by 46 per cent in myopic children
increased the amount of peripheral myopic 9.1/10 in another.77 These study results compared with spectacle-wearing historical
defocus. demonstrate that most patients using ortho- controls. Although this pilot study used his-
keratology lenses are satisfied with the torical controls, their work provided initial
vision it provides. evidence supporting further research to
Effects on visual quality of life determine whether orthokeratology was
effective in slowing myopia progression.
Another aspect of orthokeratology that is Myopia control Walline et al.83 also reported a 55 per cent
important to understand is its effect on reduction in axial growth over two years in
vision-related quality of life and on vision- Various interventions have been shown to myopic children wearing orthokeratology
related tasks. One of the early studies be effective at slowing the progression of when compared with soft contact lens-
reporting vision-related quality of life myopia. These include pharmaceuticals wearing historical controls. Studies were
between orthokeratology and soft contact such as atropine and pirenzepine, and opti- subsequently published in which myopic
lenses was conducted by Lipson et al.70 cal interventions such as bifocal and pro- children were followed for two years wear-
Study participants were randomly assigned gressive addition spectacles, bifocal and ing either orthokeratology lenses or specta-
to wear orthokeratology lenses or soft con- multifocal contact lenses and ortho- cles lenses. Although children and parents
tact lenses for eight weeks. After a washout keratology.78 Overnight orthokeratology could self-select the modality their child
period, subjects wore the other lens type for lenses were initially produced to flatten the would receive, these studies also reported
another eight weeks. After each lens-wear central cornea overnight and provide clear two-year reductions in eye growth of
period, participants completed the National unaided vision throughout the day. between 32 to 36 per cent in children wear-
Eye Institute Refractive Error Quality of Life Although the approved indication for ortho- ing orthokeratology lenses.84,85
Instrument – 42 questionnaire.71 Partici- keratology lenses is to temporarily correct The first randomised, controlled clinical
pants, on average, rated orthokeratology as myopia, research evaluating the efficacy of trial evaluating the efficacy of ortho-
having fewer activity limitations, fewer this lens modality for slowing the progres- keratology in slowing myopia progression in
symptoms, and less dependence on correc- sion of myopia in children has gained con- children was the Retardation of Myopia in
tion than soft contact lenses. Compared to siderable traction. Orthokeratology (ROMIO) study. Cho and
orthokeratology, soft contact lenses As described above, the central corneal Cheung86 randomly assigned 102 myopic
resulted in less glare. Additionally, 68 per flattening and mid-peripheral corneal steep- children into orthokeratology or single-
cent of the participants preferred ortho- ening created by modern reverse-geometry vision spectacles and followed them for two
keratology to soft contact lenses as a form orthokeratology lens designs produce a years. Of the 78 subjects who completed
of correction. myopic shift in refraction in the peripheral the study, those wearing orthokeratology
49
modality similar to their original assignment

(orthokeratology [n = 14] versus either specta-
cles or soft contact lenses [n = 16]) returned.
Seven years after starting initial treatment,
orthokeratology had slowed axial elongation
by 33 per cent compared with single-vision
corrections.
The longest prospective study of ortho-
keratology for myopia control was con-
ducted by Hiraoka et al.87 over a five-year
period. They reported statistically significant
accrual of a treatment effect over the first
three years. There was no statistically signifi-
cant difference in axial elongation between
the orthokeratology and control groups in
years four and five, but there was also no
Figure 5. Percent reduction in axial eye growth in subjects fitted with ortho- evidence of loss of the previously accrued
keratology versus control subjects wearing spectacles or soft contact lenses from sev- treatment effect. This prospective study sug-
eral studies.40,47,82–88 Subjects were followed for two years (blue solid bars) or five gests that the myopia control effect of
years (red striped bar). orthokeratology lenses is greatest in the ini-
tial years of wear with reduced efficacy over
time. This finding may be because as the
myopic children become older during the
lenses had 43 per cent less axial elongation elongation between myopic children with later stages of these long-term studies, their
compared with the single-vision spectacle moderate-to-high astigmatism (−1.25 to myopia progression naturally decreases,
control group. Axial length was measured −3.50 D) wearing toric orthokeratology making it more difficult to find a significant
every six months. For each six-month period lenses and single-vision spectacles. Of the difference in the rate of myopia progression
of the study, there was a significant reduc- 35 myopic children who completed the between the orthokeratology and control
tion in axial elongation in children wearing study, Chen et al.47 found that the children groups.
orthokeratology lenses compared with wearing toric orthokeratology lenses had The treatment effects reported in these
spectacle-wearing controls, demonstrating a 52 per cent slower axial eye elongation com- studies are an average across all children.
continued treatment effect over the two- pared with spectacle-wearing controls over Individual progression results vary, with
year period. a two-year period. some children having greater treatment
The studies described above evaluated Longer retrospective studies have been effects than others. One explanation that
the effect of orthokeratology on myopic chil- published reporting myopia progression in has been proposed for these differences in
dren with low-to-moderate myopia. Charm children wearing orthokeratology lenses. myopia control is differences in pupil size.
and Cho40 also investigated the myopia con- A study by Downie and Lowe89 reported Chen et al.92 conducted a non-randomised
trol effect of partial-reduction ortho- that orthokeratology slowed myopia pro- study where myopic children wore either
keratology in highly myopic children gression for up to eight years, while another orthokeratology lenses or single-vision spec-
(−5.75 D or more spherical equivalent myo- study by Lee et al.90 reported significantly tacles. They evaluated the effect of pupil
pia) over two years. In their study, ortho- slower myopia progression in children wear- size on myopia progression and found a sig-
keratology lenses were used to treat 4.00 D ing orthokeratology lenses up to six years nificant association between larger pupils
of myopia, and the residual myopia was after the start of treatment. One limitation and slower myopia progression in children
corrected with single-vision spectacles worn of both studies, apart from the fact that they wearing orthokeratology lenses, proposing
during the day. They found that in this were retrospective record reviews, is that that larger pupils allowed a larger area of
group of highly myopic subjects, partial- both studies measured myopia progression the peripheral retina to experience myopic
reduction orthokeratology reduced myopia by over-refraction over the orthokeratology defocus. Children wearing orthokeratology
progression by 63 per cent over two years lenses as opposed to measuring changes in lenses who had larger pupils (greater than
compared with children in the control group axial length. 6.43 mm) experienced significantly slower
wearing single-vision spectacles to correct Long-term data from a prospective study axial elongation over two years (0.36
all of their myopia. However, this study had was reported by Santodomingo-Rubido 0.22 mm) compared to orthokeratology-
a relatively high dropout rate (37 per cent in et al.91 They recruited subjects for a two-year wearing children with smaller pupils
the orthokeratology group and 16 per cent study comparing the effect of orthokeratology (0.74 0.32 mm). However, it is perplexing
in the control group) which could potentially (n = 29) versus single-vision spectacles that children wearing orthokeratology
introduce bias in their results. (n = 24) on axial growth. Orthokeratology lenses with smaller pupils had faster axial
Toric orthokeratology is indicated for slowed axial elongation by 32 per cent over elongation than children wearing single-
patients with high amounts of corneal astig- two years compared to spectacles. Five years vision spectacle lenses with small pupils
matism. A non-randomised prospective after their initial study ended, roughly half of who progressed only 0.47 0.21 mm over
study by Chen et al.47 evaluated axial the subjects who had continued with a two years. Given this discrepancy, the
50
influence of pupil size should be studied orthokeratology was effective at slowing the late 1990s and early 2000s in China and
further. axial eye growth by 0.26 mm (95% CI 0.21 to have been attributed to the use of non-gas-
Apart from pupil size, Santodomingo- 0.31) over a two-year period. Sun et al.,97 in permeable lens materials, improperly
Rubido et al.93 proposed other factors that another meta-analysis, also reported a trained practitioners, and the use of tap
were correlated with the inhibitory effects of 0.27 mm (95% CI 0.22 to 0.32) reduction in water to clean and store lenses.107 When
orthokeratology on myopia progression. axial elongation compared to controls over good clinical practice guidelines are
They included gender, age, age at myopia two years, representing approximately followed, the incidence of adverse events in
onset, rate of myopia progression, baseline 45 per cent reduction in axial elongation. clinical practice is the same for ortho-
amount of myopia, anterior chamber depth, A network meta-analysis of published keratology lenses as it is for other overnight
corneal power and shape, iris diameter, and randomised controlled trials on interven- contact lens modalities.103,105 In the USA,
refractive error of the parents. However, tions to slow myopia progression by Huang the overall estimated incidence of microbial
subsequent studies found no correlation et al.98 found that compared with controls, keratitis with orthokeratology is 7.7 per
between reduction in myopia progression orthokeratology reduced axial elongation by 10,000 years of wear (95% CI 0.9 to 27.8).108
with orthokeratology and baseline amount 0.15 mm per year (95% CI 0.08 to 0.22). This It is important that practitioners discuss
of myopia, gender and corneal toricity.40,86 myopia control effect of orthokeratology possible complications with their patients
Further work is needed to better understand was similar to low-dose (0.01%) atropine and emphasise the importance of strict
the factors that are necessary to identify (0.15 mm per year; 95% CI 0.05 to 0.25) and adherence to appropriate lens hygiene
myopic children who may benefit most from multifocal contact lenses (0.11 mm per year; practices.
orthokeratology and those who may be bet- 95% CI 0.03 to 0.20), but lower than high
ter suited for other myopia control methods. dose (1.0 or 0.5%) atropine (0.21 mm per Iron deposition
There is evidence suggesting the potential year; 95% CI 0.16 to 0.28). Corneal iron deposition has been reported
for a rebound increase in axial elongation While orthokeratology and soft multifocal during orthokeratology. These iron rings are
when myopic children discontinue ortho- contact lenses for myopia control had similar benign and have no effect on vision but
keratology treatment. In a contralateral eye effects on slowing eye growth, fitting ortho- may represent some change in corneal
crossover study conducted by Swarbrick keratology typically requires more chair time. physiology. They occur in both myopic109,110
et al.,94 myopic children were randomly In a case series of 110 myopia control and hyperopic111 orthokeratology, as early
fitted with a rigid gas-permeable lens in one patients, Turnbull et al.99 reported that chil- as one week109 or as late as three and a half
eye for daytime wear and overnight ortho- dren undergoing orthokeratology had a signif- years111 after starting orthokeratology. Cho
keratology in the other eye. After six icantly higher number of clinic visits and chair et al.109 reported that the rings resolved two
months, the lens allocation was swapped time compared with children wearing months after discontinuing orthokeratology.
between eyes after a two-week recovery multifocal contact lenses. The greater number
period, and the children were followed for of visits and potential need for morning visits
another six months. The rate of axial elon- during school time could influence a parent’s Future directions
gation after eyes that had been wearing decision regarding a myopia control modality.
orthokeratology were switched to rigid gas- With the current interest in orthokeratology
permeable lenses was greater than the rate by researchers, clinicians, and patients, the
of axial elongation in fellow eyes originally Safety and complications future of orthokeratology looks very promis-
fitted with rigid gas-permeable lenses. ing. There have been suggestions that
In another study by Cho and Cheung,95 a With the growing use of orthokeratology for multifocal orthokeratology designs could be
subset of orthokeratology-wearing children myopia correction and control, there is great developed to further increase the benefit of
from two previous studies agreed to be ran- interest in the safety of orthokeratology. this lens modality for myopia control,112
domly assigned to either continue wearing Complications from orthokeratology range although no trials have been published on
orthokeratology for 14 months or to discon- from easily treatable, low-grade corneal such a lens design. Additional studies are
tinue orthokeratology for seven months and staining to potentially sight-threatening needed to determine if optimisation of the
then resume orthokeratology for another microbial keratitis. Other complications such orthokeratology lens design could lead to a
seven months. The authors found an as central corneal epitheliopathy,100 recur- better myopia control effect.
increase in axial elongation in children who rent binding of the contact lens to the Depending on the refractive error of a
discontinued orthokeratology lens wear cornea,101 and corneal bubble and dimple patient, both myopic and hyperopic ortho-
compared to both children still in ortho- formation102 have also been reported. The keratology can be used to create mono-
keratology and children who had never majority of these complications are minor, vision correction in presbyopes. In a small
worn orthokeratology. Fortunately, eye elon- easily managed, and do not lead to a reduc- cohort of 16 emmetropic presbyopes,
gation slowed again once children resumed tion in visual acuity.103 Gifford and Swarbrick113 used hyperopic
orthokeratology. Further controlled studies orthokeratology to cause a 1.11 D myopic
are needed to explore the potential for an Infectious keratitis change in refractive error, essentially provid-
axial growth rebound after discontinuing Although cases of microbial keratitis have ing these presbyopes a roughly 1.00 D add.
orthokeratology. also been reported,104–106 with Pseudomo- Although this study was published about
Finally, a meta-analysis of the myopia con- nas aeruginosa and acanthamoeba being the five years ago, there has been little else
trol effect of orthokeratology by Si et al.96 most prevalent causative organisms,104 the published on this topic. The field would ben-
showed that compared to controls, highest number of infections occurred in efit from future work determining whether
51
a greater add can be created in emmetropic progression. There is continued research into 21. Kang SY, Kim BK, Byun YJ. Sustainability of ortho-
keratology as demonstrated by corneal topography.
patients or if this is a viable option in optimising the design of orthokeratology Korean J Ophthalmol 2007; 21: 74–78.
patients with low hyperopia. lenses in attempts to potentially improve 22. Wu R, Stapleton F, Swarbrick HA. Residual corneal
Another future application involves com- visual quality and myopia control effects, and flattening after discontinuation of long-term ortho-
keratology lens wear in Asian children. Eye Contact
bining orthokeratology with atropine for to find new applications for these lenses. Lens 2009; 35: 333–337.
myopia control. Since atropine and ortho- 23. Lu F, Sorbara L, Simpson T et al. Corneal shape and
optical performance after one night of corneal
keratology are thought to slow myopia pro-
REFERENCES refractive therapy for hyperopia. Optom Vis Sci 2007;
gression through different mechanisms, 1. Flaxman SR, Bourne RRA, Resnikoff S et al. Global 84: 357–364.
combining the two treatments could lead to causes of blindness and distance vision impairment 24. Gifford P, Swarbrick HA. Time course of corneal
1990-2020: a systematic review and meta-analysis. topographic changes in the first week of overnight
a greater reduction in myopia progression Lancet Glob Health 2017; 5: e1221–e1234. hyperopic orthokeratology. Optom Vis Sci 2008; 85:
than orthokeratology alone. The greater 2. Brungardt TF. Orthokeratology: an analysis. Int Con- 1165–1171.
effect of the combined treatment could also tact Lens Clin 1976; 3: 56–58. 25. Gifford P, Alharbi A, Swarbrick HA. Corneal thickness
3. Kerns RL. Research in orthokeratology. Part I: intro- changes in hyperopic orthokeratology measured by
be from low-dose atropine causing slight duction and background. J Am Optom Assoc 1976; 47: optical pachometry. Invest Ophthalmol Vis Sci 2011;
increases in pupil size,114 exposing a greater 1047–1051. 52: 3648–3653.
4. Wlodyga RJ, Bryla C. Corneal molding: the easy way. 26. Mountford J. An analysis of the changes in corneal
area of the retina to the myopic defocus.
Contact Lens Spectrum 1989; 4: 58–65. shape and refractive error induced by accelerated
Two published reports provide evidence 5. Swarbrick HA, Wong G, O’Leary DJ. Corneal orthokeratology. Int Contact Lens Clin 1997; 24:
supporting this synergistic effect.115,116 Chil- response to orthokeratology. Optom Vis Sci 1998; 128–144.
75: 791–799. 27. Rah MJ, Jackson JM, Jones LA et al. Overnight ortho-
dren in both studies who received ortho- 6. Sridharan R, Swarbrick H. Corneal response to short- keratology: preliminary results of the lenses and
keratology and low-dose atropine progressed term orthokeratology lens wear. Optom Vis Sci 2003; overnight orthokeratology (LOOK) study. Optom Vis
more slowly than children in orthokeratology 80: 200–206. Sci 2002; 79: 598–605.
7. Tsukiyama J, Miyamoto Y, Higaki S et al. Changes in 28. Joslin CE, Wu SM, McMahon TT et al. Higher-order
alone; however, one study was only three the anterior and posterior radii of the corneal curva- wavefront aberrations in corneal refractive therapy.
months in length,115 and the other was retro- ture and anterior chamber depth by ortho- Optom Vis Sci 2003; 80: 805–811.
keratology. Eye Contact Lens 2008; 34: 17–20. 29. Koffler BH, Smith VM. Myopia reduction using cor-
spective.116 There is currently an ongoing
8. Villa-Collar C, Gonzalez-Meijome JM, Queiros A et al. neal refractive therapy contact lenses. Eye Contact
two-year randomised clinical trial in which Short-term corneal response to corneal refractive Lens 2004; 30: 223–226.
children were randomly assigned to either therapy for different refractive targets. Cornea 2009; 30. Soni PS, Nguyen TT, Bonanno JA. Overnight ortho-
28: 311–316. keratology: refractive and corneal recovery after
orthokeratology alone or orthokeratology 9. Zhong X, Chen X, Xie RZ et al. Differences between discontinuation of reverse-geometry lenses. Eye
with 0.01% atropine that should shed more overnight and long-term wear of orthokeratology Contact Lens 2004; 30: 254–262.
light on the efficacy of this combined therapy contact lenses in corneal contour, thickness, and cell 31. Berntsen DA, Barr JT, Mitchell GL. The effect of over-
density. Cornea 2009; 28: 271–279. night contact lens corneal reshaping on higher-order
for myopia control.117 10. Yoon JH, Swarbrick HA. Posterior corneal shape aberrations and best-corrected visual acuity. Optom
Finally, future research is needed to under- changes in myopic overnight orthokeratology. Optom Vis Sci 2005; 82: 490–497.
Vis Sci 2013; 90: 196–204. 32. Hiraoka T, Matsumoto Y, Okamoto F et al. Corneal
stand how axial eye growth changes in myopic
11. Alharbi A, Swarbrick HA. The effects of over- higher-order aberrations induced by overnight
children after discontinuing orthokeratology. night orthokeratology lens wear on corneal orthokeratology. Am J Ophthalmol 2005; 139:
Overall, one should expect to continue hearing thickness. Invest Ophthalmol Vis Sci 2003; 44: 429–436.
2518–2523. 33. Sorbara L, Fonn D, Simpson T et al. Reduction of
of new developments regarding the use of this 12. Wang J, Fonn D, Simpson TL et al. Topographical myopia from corneal refractive therapy. Optom Vis
contact lens modality. thickness of the epithelium and total cornea after Sci 2005; 82: 512–518.
overnight wear of reverse-geometry rigid contact 34. Hiraoka T, Okamoto C, Ishii Y et al. Time course of
lenses for myopia reduction. Invest Ophthalmol Vis changes in ocular higher-order aberrations and con-
Sci 2003; 44: 4742–4746. trast sensitivity after overnight orthokeratology.
Conclusions 13. Maseedupally V, Gifford P, Lum E et al. Central and Invest Ophthalmol Vis Sci 2008; 49: 4314–4320.
paracentral corneal curvature changes during 35. Hiraoka T, Okamoto C, Ishii Y et al. Mesopic contrast
orthokeratology. Optom Vis Sci 2013; 90: sensitivity and ocular higher-order aberrations after
Overnight orthokeratology can temporarily 1249–1258. overnight orthokeratology. Am J Ophthalmol 2008;
14. Choo JD, Caroline PJ, Harlin DD et al. Morphologic 145: 645–655.
reduce myopia, hyperopia, and astigmatism. changes in cat epithelium following continuous wear 36. Kobayashi Y, Yanai R, Chikamoto N et al. Reversibility
This reduction in refractive error leads to of orthokeratology lenses: a pilot study. Cont Lens of effects of orthokeratology on visual acuity, refrac-
Anterior Eye 2008; 31: 29–37. tive error, corneal topography, and contrast sensitiv-
improvements in uncorrected high- and low-
15. Owens H, Garner LF, Craig JP et al. Posterior corneal ity. Eye Contact Lens 2008; 34: 224–228.
contrast visual acuity with very little regres- changes with orthokeratology. Optom Vis Sci 2004; 37. Queiros A, Gonzalez-Meijome JM, Jorge J et al.
sion throughout the day. Orthokeratology 81: 421–426. Peripheral refraction in myopic patients after ortho-
16. Gonzalez-Mesa A, Villa-Collar C, Lorente-Velazquez A keratology. Optom Vis Sci 2010; 87: 323–329.
also increases higher-order aberrations and
et al. Anterior segment changes produced in 38. Kang P, Swarbrick H. Peripheral refraction in myo-
can improve accommodative accuracy. Spe- response to long-term overnight orthokeratology. pic children wearing orthokeratology and gas-
cifically, myopic orthokeratology increases Curr Eye Res 2013; 38: 862–870. permeable lenses. Optom Vis Sci 2011; 88:
17. Chen D, Lam AK, Cho P. Posterior corneal curvature 476–482.
positive spherical aberration. change and recovery after 6 months of overnight 39. Nichols JJ, Marsich MM, Nguyen M et al. Overnight
Although high-contrast visual acuity with orthokeratology treatment. Ophthalmic Physiol Opt orthokeratology. Optom Vis Sci 2000; 77: 252–259.
2010; 30: 274–280. 40. Charm J, Cho P. High myopia-partial reduction ortho-
orthokeratology is similar to visual acuity with
18. Soni PS, Nguyen TT, Bonanno JA. Overnight ortho- k: a 2-year randomized study. Optom Vis Sci 2013; 90:
spectacles, increased higher-order aberrations keratology: visual and corneal changes. Eye Contact 530–539.
reduce best-corrected low-contrast visual acu- Lens 2003; 29: 137–145. 41. Chan B, Cho P, Cheung SW. Orthokeratology practice
19. Li F, Jiang ZX, Hao P et al. A meta-analysis of central in children in a university clinic in Hong Kong. Clin
ity and retinal image quality. Mid-peripheral corneal thickness changes with overnight ortho- Exp Optom 2008; 91: 453–460.
corneal steepening with myopic orthokeratology. Eye Contact Lens 2016; 42: 141–146. 42. Walline JJ, Rah MJ, Jones LA. The children’s overnight
keratology also causes a myopic shift in 20. Kim WK, Kim BJ, Ryu IH et al. Corneal epithelial and orthokeratology investigation (COOKI) pilot study.
stromal thickness changes in myopic ortho- Optom Vis Sci 2004; 81: 407–413.
peripheral refraction and is hypothesised to keratology and their relationship with refractive 43. Cheung SW, Cho P, Chan B. Astigmatic changes in
contribute to observed reductions in myopia change. PLoS One 2018; 13: e0203652. orthokeratology. Optom Vis Sci 2009; 86: 1352–1358.
52
44. Mountford J, Pesudovs K. An analysis of the astig- profile after orthokeratology for different degrees of study. Invest Ophthalmol Vis Sci 2012; 53:
matic changes induced by accelerated ortho- myopia. Curr Eye Res 2016; 41: 199–207. 3913–3919.
keratology. Clin Exp Optom 2002; 85: 284–293. 67. Kang P, Swarbrick H. New perspective on myopia 88. Paune J, Morales H, Armengol J et al. Myopia control
45. Hiraoka T, Furuya A, Matsumoto Y et al. Quantitative control with orthokeratology. Optom Vis Sci 2016; 93: with a novel peripheral gradient soft lens and ortho-
evaluation of regular and irregular corneal astigma- 497–503. keratology: a 2-year clinical trial. Biomed Res Int 2015;
tism in patients having overnight orthokeratology. 68. Kang P, Swarbrick H. The influence of different OK 2015: 507572.
J Cataract Refract Surg 2004; 30: 1425–1429. lens designs on peripheral refraction. Optom Vis Sci 89. Downie LE, Lowe R. Corneal reshaping influences
46. Chan B, Cho P, de Vecht A. Toric orthokeratology: a 2016; 93: 1112–1119. myopic prescription stability (CRIMPS): an analysis of
case report. Clin Exp Optom 2009; 92: 387–391. 69. Queiros A, Amorim-de-Sousa A, Lopes-Ferreira D the effect of orthokeratology on childhood myopic
47. Chen C, Cheung SW, Cho P. Myopia control using et al. Relative peripheral refraction across 4 merid- refractive stability. Eye Contact Lens 2013; 39:
toric orthokeratology (TO-SEE study). Invest ians after orthokeratology and LASIK surgery. Eye Vis 303–310.
Ophthalmol Vis Sci 2013; 54: 6510–6517. (Lond) 2018; 5: 12. 90. Lee YC, Wang JH, Chiu CJ. Effect of orthokeratology
48. Stillitano I, Schor P, Lipener C et al. Stability of 70. Lipson MJ, Sugar A, Musch DC. Overnight corneal on myopia progression: twelve-year results of a ret-
wavefront aberrations during the daytime after reshaping versus soft disposable contact lenses: rospective cohort study. BMC Ophthalmol 2017;
6 months of overnight orthokeratology corneal res- vision-related quality-of-life differences from a ran- 17: 243.
haping. J Refract Surg 2007; 23: 978–983. domized clinical trial. Optom Vis Sci 2005; 82: 91. Santodomingo-Rubido J, Villa-Collar C, Gilmartin B
49. Lian Y, Shen M, Huang S et al. Corneal reshaping and 886–891. et al. Long-term efficacy of orthokeratology contact
wavefront aberrations during overnight ortho- 71. National Eye Institute. NEI Refractive Error Quality of lens wear in controlling the progression of childhood
keratology. Eye Contact Lens 2014; 40: 161–168. Life Instrument-42 (NEI RQL-42) (Internet). [Cited myopia. Curr Eye Res 2017; 42: 713–720.
50. Santolaria-Sanz E, Cervino A, Gonzalez-Meijome JM. 17 Jun 2013.] Available at: https://nei.nih.gov/ 92. Chen Z, Niu L, Xue F et al. Impact of pupil diameter
Corneal aberrations, contrast sensitivity, and light catalog/nei-refractive-error-quality-life-instrument- on axial growth in orthokeratology. Optom Vis Sci
distortion in orthokeratology patients: 1-year results. 42-nei-rql-42 2012; 89: 1636–1640.
J Ophthalmol 2016; 2016: 8453462. 72. Ritchey ER, Barr JT, Mitchell GL. The comparison of 93. Santodomingo-Rubido J, Villa-Collar C, Gilmartin B
51. Santodomingo-Rubido J, Villa-Collar C, Gilmartin B overnight lens modalities (COLM) study. Eye Contact et al. Factors preventing myopia progression with
et al. Short- and long-term changes in corneal aber- Lens 2005; 31: 70–75. orthokeratology correction. Optom Vis Sci 2013; 90:
rations and axial length induced by orthokeratology 73. Berntsen DA, Mitchell GL, Barr JT. The effect of over- 1225–1236.
in children are not correlated. Eye Contact Lens 2017; night contact lens corneal reshaping on refractive 94. Swarbrick HA, Alharbi A, Watt K et al. Myopia control
43: 358–363. error-specific quality of life. Optom Vis Sci 2006; 83: during orthokeratology lens wear in children using a
52. Stillitano IG, Chalita MR, Schor P et al. Corneal 354–359. novel study design. Ophthalmology 2015; 122:
changes and wavefront analysis after ortho- 74. Santodomingo-Rubido J, Villa-Collar C, Gilmartin B 620–630.
keratology fitting test. Am J Ophthalmol 2007; 144: et al. Myopia control with orthokeratology contact 95. Cho P, Cheung SW. Discontinuation of ortho-
378–386. lenses in Spain: a comparison of vision-related keratology on eyeball elongation (DOEE). Cont Lens
53. Stillitano I, Schor P, Lipener C et al. Long-term follow- quality-of-life measures between orthokeratology Anterior Eye 2017; 40: 82–87.
up of orthokeratology corneal reshaping using contact lenses and single-vision spectacles. Eye Con- 96. Si JK, Tang K, Bi HS et al. Orthokeratology for myopia
wavefront aberrometry and contrast sensitivity. Eye tact Lens 2013; 39: 153–157. control: a meta-analysis. Optom Vis Sci 2015; 92:
Contact Lens 2008; 34: 140–145. 75. Zhao F, Zhao G, Zhao Z. Investigation of the effect 252–257.
54. Johnson KL, Carney LG, Mountford JA et al. Visual of orthokeratology lenses on quality of life and 97. Sun Y, Xu F, Zhang T et al. Orthokeratology to control
performance after overnight orthokeratology. Cont behaviors of children. Eye Contact Lens 2018; 44: myopia progression: a meta-analysis. PLoS One 2015;
Lens Anterior Eye 2007; 30: 29–36. 335–338. 10: e0124535.
55. Berntsen DA. Retinal image quality after ortho- 76. Hiraoka T, Okamoto C, Ishii Y et al. Patient satisfac- 98. Huang J, Wen D, Wang Q et al. Efficacy comparison
keratology. Invest Ophthalmol Vis Sci 2018; 59: 1748. tion and clinical outcomes after overnight ortho- of 16 interventions for myopia control in children: a
56. Felipe-Marquez G, Nombela-Palomo M, Cacho I et al. keratology. Optom Vis Sci 2009; 86: 875–882. network meta-analysis. Ophthalmology 2016; 123:
Accommodative changes produced in response to 77. Santolaria E, Cervino A, Queiros A et al. Subjective 697–708.
overnight orthokeratology. Graefes Arch Clin Exp satisfaction in long-term orthokeratology patients. 99. Turnbull PR, Munro OJ, Phillips JR. Contact lens
Ophthalmol 2015; 253: 619–626. Eye Contact Lens 2013; 39: 388–393. methods for clinical myopia control. Optom Vis Sci
57. Kang P, Watt K, Chau T et al. The impact of ortho- 78. Wildsoet CF, Chia A, Cho P et al. IMI - interventions 2016; 93: 1120–1126.
keratology lens wear on binocular vision and accom- for controlling myopia onset and progression report. 100. Ng LH. Central corneal epitheliopathy in a long-term,
modation: a short-term prospective study. Cont Lens Invest Ophthalmol Vis Sci 2019; 60: M106–M131. overnight orthokeratology lens wearer: a case
Anterior Eye 2018; 41: 501–506. 79. Smith EL 3rd. Prentice award lecture 2010: a case for report. Optom Vis Sci 2006; 83: 709–714.
58. Gifford K, Gifford P, Hendicott PL et al. Near binocu- peripheral optical treatment strategies for myopia. 101. Chui W, Cho P. Recurrent lens binding and central
lar visual function in young adult orthokeratology Optom Vis Sci 2011; 88: 1029–1044. Island formations in a fast-responding ortho-
versus soft contact lens wearers. Cont Lens Anterior 80. Schmid GF. Association between retinal steepness keratology lens wearer. Optom Vis Sci 2003; 80:
Eye 2017; 40: 184–189. and central myopic shift in children. Optom Vis Sci 490–494.
59. Gifford KL, Gifford P, Hendicott PL et al. Zone of clear 2011; 88: 684–690. 102. Stillitano I, Maidana E, Lui M et al. Bubble and cor-
single binocular vision in myopic orthokeratology. 81. Cheung SW, Cho P, Fan D. Asymmetrical increase in neal dimple formation after the first overnight wear
Eye Contact Lens 2019. https://doi.org/10.1097/ICL. axial length in the two eyes of a monocular ortho- of an orthokeratology lens: a case series. Eye Contact
0000000000000614. keratology patient. Optom Vis Sci 2004; 81: 653–656. Lens 2007; 33: 253–258.
60. Han X, Xu D, Ge W et al. A comparison of the effects 82. Cho P, Cheung SW, Edwards M. The longitudinal 103. Liu YM, Xie P. The safety of orthokeratology - a
of orthokeratology lens, medcall lens, and ordinary orthokeratology research in children (LORIC) in Hong systematic review. Eye Contact Lens 2016; 42:
frame glasses on the accommodative response in Kong: a pilot study on refractive changes and myopic 35–42.
myopic children. Eye Contact Lens 2018; 44: 268–271. control. Curr Eye Res 2005; 30: 71–80. 104. Kam KW, Yung W, Li GKH et al. Infectious keratitis
61. Gifford P, Li M, Lu H et al. Corneal versus ocular 83. Walline JJ, Jones LA, Sinnott LT. Corneal reshaping and orthokeratology lens use: a systematic review.
aberrations after overnight orthokeratology. Optom and myopia progression. Br J Ophthalmol 2009; 93: Infection 2017; 45: 727–735.
Vis Sci 2013; 90: 439–447. 1181–1185. 105. Hiraoka T, Sekine Y, Okamoto F et al. Safety and effi-
62. Mutti DO, Sholtz RI, Friedman NE et al. Peripheral 84. Kakita T, Hiraoka T, Oshika T. Influence of overnight cacy following 10-years of overnight orthokeratology
refraction and ocular shape in children. Invest orthokeratology on axial elongation in childhood for myopia control. Ophthalmic Physiol Opt 2018; 38:
Ophthalmol Vis Sci 2000; 41: 1022–1030. myopia. Invest Ophthalmol Vis Sci 2011; 52: 281–289.
63. Smith EL 3rd, Hung LF, Huang J. Relative peripheral 2170–2174. 106. Li W, Sun X, Wang Z et al. A survey of contact lens-
hyperopic defocus alters central refractive develop- 85. Santodomingo-Rubido J, Villa-Collar C, Gilmartin B related complications in a tertiary hospital in China.
ment in infant monkeys. Vision Res 2009; 49: et al. Myopia control with orthokeratology con- Cont Lens Anterior Eye 2018; 41: 201–204.
2386–2392. tact lenses in Spain: refractive and biometric 107. Watt KG, Swarbrick HA. Trends in microbial keratitis
64. Kang P, Gifford P, Swarbrick H. Can manipulation of changes. Invest Ophthalmol Vis Sci 2012; 53: associated with orthokeratology. Eye Contact Lens
orthokeratology lens parameters modify peripheral 5060–5065. 2007; 33: 373–377.
refraction? Optom Vis Sci 2013; 90: 1237–1248. 86. Cho P, Cheung SW. Retardation of myopia in ortho- 108. Bullimore MA, Sinnott LT, Jones-Jordan LA. The
65. Kang P, Swarbrick H. Time course of the effects of keratology (ROMIO) study: a 2-year randomized clini- risk of microbial keratitis with overnight corneal
orthokeratology on peripheral refraction and corneal cal trial. Invest Ophthalmol Vis Sci 2012; 53: reshaping lenses. Optom Vis Sci 2013; 90:
topography. Ophthalmic Physiol Opt 2013; 33: 7077–7085. 937–944.
277–282. 87. Hiraoka T, Kakita T, Okamoto F et al. Long-term 109. Cho P, Chui WS, Cheung SW. Reversibility of corneal
66. Gonzalez-Meijome JM, Faria-Ribeiro MA, Lopes- effect of overnight orthokeratology on axial length pigmented arc associated with orthokeratology.
Ferreira DP et al. Changes in peripheral refractive elongation in childhood myopia: a 5-year follow-up Optom Vis Sci 2003; 80: 791–795.
53
110. Rah MJ, Barr JT, Bailey MD. Corneal pigmentation in 113. Gifford P, Swarbrick HA. Refractive changes from children with myopia: first year results. Jpn J
overnight orthokeratology: a case series. Optometry hyperopic orthokeratology monovision in presby- Ophthalmol 2018; 62: 544–553.
2002; 73: 425–434. opes. Optom Vis Sci 2013; 90: 306–313. 116. Wan L, Wei CC, Chen CS et al. The synergistic
111. Kirkwood BJ, Rees IH. Central corneal iron line aris- 114. Kaymak H, Fricke A, Mauritz Y et al. Short-term effects effects of orthokeratology and atropine in
ing from hyperopic orthokeratology. Clin Exp Optom of low-concentration atropine eye drops on pupil size slowing the progression of myopia. J Clin Med
2011; 94: 376–379. and accommodation in young adult subjects. Graefes 2018; 7: E259.
112. Loertscher M. Multifocal orthokeratology associated Arch Clin Exp Ophthalmol 2018; 256: 2211–2217. 117. Tan Q, Ng AL, Cheng GP et al. Combined atropine
with rapid shortening of vitreous chamber depth in 115. Kinoshita N, Konno Y, Hamada N et al. Additive with orthokeratology for myopia control: study
eyes of myopic children. Cont Lens Anterior Eye 2013; effects of orthokeratology and atropine 0.01% oph- design and preliminary results. Curr Eye Res 2019;
36: e2. thalmic solution in slowing axial elongation in 44: 671–678.
54
INVITED REVIEW
Optical mechanisms regulating emmetropisation and refractive

errors: evidence from animal models
Ranjay Chakraborty* PhD BS Our current understanding of emmetropisation and myopia development has evolved from
Optometry decades of work in various animal models, including chicks, non-human primates, tree
Lisa A Ostrin† PhD OD FAAO shrews, guinea pigs, and mice. Extensive research on optical, biochemical, and environmen-
Alexandra Benavente-Perez‡ PhD MS BS tal mechanisms contributing to refractive error development in animal models has provided
Optometry insights into eye growth in humans. Importantly, animal models have taught us that eye
Pavan Kumar Verkicharla§ PhD BS growth is locally controlled within the eye, and can be influenced by the visual environment.
Optometry This review will focus on information gained from animal studies regarding the role of opti-
*College of Nursing and Health Sciences, Optometry cal mechanisms in guiding eye growth, and how these investigations have inspired studies
and Vision Science, Flinders University, Adelaide, in humans. We will first discuss how researchers came to understand that emmetropisation
Australia is guided by visual feedback, and how this can be manipulated by form-deprivation and
†
University of Houston College of Optometry, lens-induced defocus to induce refractive errors in animal models. We will then discuss vari-
Houston, Texas, USA ous aspects of accommodation that have been implicated in refractive error development,
‡
State University of New York College of Optometry,
including accommodative microfluctuations and accommodative lag. Next, the impact of
New York, USA
§
higher order aberrations and peripheral defocus will be discussed. Lastly, recent evidence
Myopia Research Lab, Prof. Brien Holden Eye
suggesting that the spectral and temporal properties of light influence eye growth, and how
Research Centre, LV Prasad Eye Institute, Hyderabad,
India
this might be leveraged to treat myopia in children, will be presented. Taken together, these
Email: ranjay.chakraborty@flinders.edu.au findings from animal models have significantly advanced our knowledge about the optical
mechanisms contributing to eye growth in humans, and will continue to contribute to the
development of novel and effective treatment options for slowing myopia progression in
Submitted: 2 July 2019
children.
Revised: 24 September 2019
Accepted for publication: 25 September
2019
Key words: accommodation, emmetropisation, form-deprivation, longitudinal chromatic aberration, myopia, peripheral defocus
Both the optical power in the anterior seg- optics, and images of distant objects focus elongation, also known as form-deprivation
ment of the eye and axial length determine behind the photoreceptor plane. myopia (FDM).4–8 Studies also use lenses to
refractive state.1–3 Emmetropisation is an This review focuses on optical mecha- alter the image plane with respect to the
active, visually guided mechanism whereby nisms of eye growth and refractive error retina, resulting in image defocus that
the axial length and the combined optical development. We will discuss how extensive induces compensatory alterations in ocular
powers of the cornea and lens precisely investigations on animal models have growth, known as lens-induced myopia or
match with each other to eliminate neonatal formed our current understanding of optical hyperopia.9–15 Both form-deprivation and
refractive errors, and bring the eye to per- mechanisms of emmetropisation, and hel- lens-induced defocus result in abnormal eye
fect focus (also known as emmetropia). In ped in developing improved optical inter- growth and refractive error development,
non-accommodating emmetropic eyes, ventions for refractive error management. with associated anatomical, optical, and bio-
visual images of distant objects are clearly chemical changes in the anterior and poste-
focused at the retinal photoreceptors. Any rior segments of the eye (see inclusive
disruption to this homeostatic mechanism Optical defocus and visual reviews).16–18 In this section, we will define
of ocular growth results in the development regulation of ocular growth the process of emmetropisation, summarise
of refractive errors. In myopia, or near- different optical aspects of FDM and lens-
sightedness, the eye is too long for the opti- The visual environment plays an important induced ametropias, including their similari-
cal power of the cornea and lens, and role in the regulation of ocular growth and ties and differences, and describe how these
images of distance objects focus in front of emmetropisation. Experimental studies of experimental models have informed us
the photoreceptor plane. In hyperopia, or myopia employ diffusers to blur the image about refractive error development in
far-sightedness, the eye is too short for the on the retina, which induces axial humans.
55
Optical regulation of eye growth Chakraborty, Ostrin, Benavente-Perez et al.
Emmetropisation
At birth, most animal species typically
exhibit variable degrees of hyperopia.19–26
During the early period of postnatal devel-
opment there is a systematic reduction in
both the degree and variability of hyperopia,
bringing the eye closer to emmetropia, or in
some cases, low myopia.27 Similar to animal
models, the majority of newborn infants are
born moderately hyperopic, typically in the
range of +2 to +4 D, which reduces signifi-
cantly during the first 18 months of life.28–30
There is a concurrent rapid increase of
~ 2–3 mm in axial length during the first
one–two years of life, primarily due to an
expansion in the vitreous chamber.31–33 The
rapid reduction in hyperopia and the
changes in axial length during the early
phase of emmetropisation are strongly cor-
related.32,33 More importantly, based on evi-
dence from animal models, the increase in
axial length during the postnatal period in
infant human eyes is believed to be modu-
lated by active visual feedback from the
hyperopic refractive error.32 While axial
length is the primary biometric component
of emmetropisation in humans, there is also
a passive contribution from reductions in
corneal and crystalline lens power during
postnatal eye growth.31,32,34
FDM
Form-deprivation as an experimental model Figure 1. Ocular compensation for form-deprivation. A: A diffuser causes non-
of myopia was first described by Wiesel and directional blur and a reduction in contrast of the retinal image. B: The absence of
Raviola35 in neonatal monkeys with lid visual feedback related to the effective refractive state of the eye causes a thinning
fusion. Soon after, FDM was successfully of the posterior choroid and an increase in ocular growth, resulting in myopia, known
induced in tree shrews,36 chicks,6 and cats37 as form-deprivation myopia (FDM). The blurred eye represents the original shape of
by suturing their eyelids, and in macaque the eye prior to form-deprivation.
monkeys by opacifying the cornea soon
after birth.38 Subsequent studies imposed
form-deprivation by securing translucent
diffusers over the eye using a mask,2,13,39 has been reported in a wide range of animal FDM is a graded phenomenon; the degree
glue,40,41 Velcro,4,42,43 or a head-mounted species, including birds,7,10,55 rodents,39,56 of axial myopia is positively correlated with
pedestal.44–47 These studies have consis- non-human primates,35,54 and even fish,57 the degree of reduction in retinal image
tently shown that depriving the retina of the magnitude of myopia and the rate of contrast.54,59 Therefore, even mild distor-
form or patterned vision produces axial ocular elongation varies among species. For tions in the quality of the retinal image may
myopia compared to untreated eyes, example, chick eyes can develop myopia of potentially lead to some degree of myopia.
suggesting that a sharp, high-contrast reti- up to 17 D after 10 days of form In any given animal model, there are signifi-
nal image is essential for normal eye growth deprivation,7 whereas primates develop cant individual (or between-subject) differ-
(Figure 1). FDM is believed to be an ‘open- approximately 5–6 D of myopia after ences in the myopic response to form-
loop’ condition, in which myopia occurs as a 17 weeks of form-deprivation.8,58 Despite deprivation. This suggests that both visual
result of unrestricted eye growth due to quantitative differences between species, environment and individual genetic factors
absence of visual feedback from the form- potentially due to differences in experimen- contribute to FDM,48 which is consistent
deprived retina and absence of a defined tal paradigms and/or inherent ocular ana- with our current understanding of the
refractive endpoint.35,48 tomical variations between animal models, aetiology of myopia in humans. Further-
FDM is primarily a result of increased these results importantly point toward a more, the ability of the eye to respond to
axial length, mainly an elongated vitreous ubiquitous visual mechanism of ocular form-deprivation declines with age in
chamber, and is accompanied by thinning of growth modulation that is conserved across chicks,7,55 monkeys,60 tree shrews,5 and
the choroid and sclera.4,7,10,49–54 While FDM species. marmosets.50 However, older chickens55,61
56
and monkeys60 do continue to respond to

form-deprivation, albeit to a lesser magni-
tude, suggesting that the visual system
retains some degree of plasticity even at the
end of the ‘initial infantile phase’ of
emmetropisation. Constant darkness also
deprives the eye of form vision, causing ocu-
lar elongation and significant corneal flat-
tening in chicks and monkeys.51,62,63 The
reduction in corneal power related to the
loss of circadian cues in darkness strongly
influence the refractive development of the
eye, leading to more hyperopic refraction in
young animals. Finally, in all animal models,
removing the diffusers at the end of the
form-deprivation period leads to recovery,
defined by a rapid and systematic reduction
in the magnitude of experimentally induced
myopia.4,7,10,55 The rapid deceleration in eye
growth during recovery from FDM largely
occurs as a result of changes in the vitreous
chamber and choroidal thickness.7,10 Recov-
ery from FDM depends on the magnitude of
myopia and the age at which diffusers are
removed.64 In animals, the ability of the eye
to recover from myopia declines with age in
optically mature eyes, with stable corneal
and lens powers.1 Interestingly, in humans,
visual deprivation from ptosis,65 congenital
cataract,66 corneal opacity,67 and vitreous
haemorrhage68 are associated with myopia,
which likely results from mechanisms simi-
lar to FDM observed in animals.
Lens-induced refractive errors

As shown in Figure 2, a large part of our cur-
rent knowledge of visual regulation of ocular Figure 2. Schematic of lens-induced refractive errors. A: A normal eye with no
growth has originated from experimental imposed lens defocus exhibits normal ocular growth and choroidal thickness. B:
studies describing the ability of the eye to Hyperopic defocus induced with negative lenses (blue) results in choroidal thinning
compensate for myopic or hyperopic which moves the retina backward, an increase in axial elongation, and a myopic shift
defocus, imposed with positive or negative in refraction. C: Myopic defocus induced with positive lenses (green) leads to a thick-
lenses, respectively. When defocus is ening of the choroid which brings the retina forward, a slowing of axial elongation,
induced, the eye undergoes compensatory and a hyperopic shift in refraction. The blurred eye in B and C represents the original
changes in ocular growth to match its axial shape of the eye prior to the introduction of lens defocus. Adapted from Wallman
length to the altered focal plane, thereby and Winawer, 2004.1
decreasing or eliminating the imposed refrac-
tive error.9,16 Myopic defocus induced with
positive lenses leads to a thickening of the ceases when the imposed defocus is appro- powers ranging from −10 to +20 D,70 while
choroid which brings the retina forward, a priately compensated.69 The effects of lens Old and New World monkeys’ eyes exhibit a
slowing of axial elongation, and a hyperopic defocus on ocular growth have been demon- relatively smaller operating range of −5 to
shift in refraction. Conversely, hyperopic strated in chickens,7,9–12,70,71 tree shrews,72 +8 D.13,75 These differences in the magnitude
defocus induced with negative lenses results monkeys,13,14 marmosets,73 guinea pigs,15 of the response may reflect interspecies dif-
in choroidal thinning which moves the retina and mice,49,74 suggesting that this vision- ferences in refractive error, ocular anatomy,
backward, an increase in axial elongation, dependent regulatory mechanism of eye or other physiological processes, such as
and a myopic shift in refraction. For either growth is conserved across species. accommodation and vergence mechanisms.
defocus condition, the change in axial length Similar to FDM, there are significant inter- Growing eyes in young animals can fully
is largely attributed to the changes in the vit- species differences in ocular responses to compensate for defocus imposed by specta-
reous chamber.10 Lens-induced defocus is a lens-induced defocus. Chick eyes can com- cle lenses as long as they are in the linear
‘closed-loop’ condition, in which axial growth pensate for a large range of spectacle lens response range of their emmetropisation
57
mechanism.70,73 Imposing higher degrees of lenses of varying powers is generally skewed differences between the two experimental
defocus beyond the operating limits of lens toward the more positive powered lens com- conditions.
compensation results in little or no change in ponent, leading to hyperopic refractive This body of work in animal models has
refractive error in mice,49 chicks76 and pri- errors. In infant macaques, the refractive laid a robust scientific foundation for devel-
mates.13 When the imposed defocus is development with dual-focus lenses of con- oping optical treatments that alter retinal
removed, all animal models show rapid centric annular zones and alternating powers image quality to reduce myopia progression
recovery by reversing the changes in both of −3 D/0 D and +3 D/0 D is largely domi- in young human eyes, including ortho-
choroidal thickness and axial eye growth to nated by the more anterior (or relatively keratology and bifocal contact lenses.99,100
restore normal vision.7,10 Interestingly, con- more myopic/less hyperopic) image plane.92
sistent with observations in animal models, Spatial and temporal integration of these
recent studies in young adult humans have competing visual signals determine the over- Accommodation
documented small, short-term bidirectional all nature and direction of refractive develop-
changes in axial length and choroidal thick- ment. Finally, spatial integration of visual Accommodation is the dioptric power change
ness in response to one–two hours of signals across the central and peripheral ret- of the eye to focus diverging rays on the ret-
imposed hyperopic and myopic defocus.77–80 ina may also modulate the eye’s response to ina, and has been implicated in myopia
During lens-imposed defocus, the sign, fre- lens-induced defocus (see section on periph- development. Accommodation is initiated by
quency, duration, and magnitude of defocus eral defocus for details). several cues, including retinal defocus, chro-
experienced by the eye change constantly While both hyperopic defocus and form- matic aberrations,101 and optical vergence.102
depending on the visual scene. Therefore, deprivation induce axial myopia (discussed Evidence from animal studies suggests that
the nature of vision-dependent eye growth above), the mechanisms underlying the two emmetropisation is guided by retinal
depends on the temporal integration of experimental conditions may be different. defocus. Speculation exists whether
visual signals over time (see reviews).16,81 For instance, blocking the parasympathetic accommodation-related defocus plays a role
Previous studies have shown that the tempo- innervation to the eye through ciliary in emmetropisation. Several animal models
ral integration of visual signals for ocular ganglionectomy inhibited FDM in chicks,93 of myopia are known to show active accom-
growth regulation is non-linear. For instance, but had no effect on the compensatory modation, including the chick,9,103
exposing the eye to successive periods of responses to lens-induced defocus.12 Previ- marmoset, 104
and rhesus monkey, 105,106
and
hyperopic and myopic defocus of equal dura- ous chick studies have reported significant have been utilised to examine the influence
tions lead to reduced axial elongation and differences in the inner retinal function of accommodation in eye growth. Non-
hyperopic refractive error in chicks.82,83 Fur- between form-deprivation and negative lens human primates exhibit lenticular accommo-
thermore, studies in chicks have shown that wear.94 In another study, Choh et al.95 dation, similar to humans. On the other
myopic defocus has a greater effect on showed that in chicks with optic nerve sec- hand, chicks demonstrate both corneal and
refractive development compared to hyper- tion, the change in axial length with diffusers lenticular accommodation.107–110 Characteris-
opic defocus, suggesting that the visual sys- was about 50 per cent greater compared to tics of accommodation that have been linked
tem may be using distinct visual mechanisms the eyes experiencing lens defocus, despite to emmetropisation and myopia include
for ocular compensation to hyperopic and similar degrees of imposed ‘spatial blur’ on accommodative microfluctuations,111 accom-
myopic defocus.81,84,85 The nature of lens the retina in absence of accommodation modative lag,112 tonic accommodation,113,114
compensation, as well as FDM, depend on (optic nerve section eliminates active accom- and blur interpretation.115
the frequency and duration of exposure, not modation). In addition, environmental light- Early studies that utilised minus lens-
just the total duration of exposure in a ing manipulations have varying effects on induced hyperopic defocus to induce experi-
day.81,85–87 In chick eyes, several brief periods FDM and lens-induced defocus. For instance, mental myopia were based on the belief
of defocus throughout the day produce a six days of rearing in bright lighting that this affected eye growth via
larger ocular response than a single or a few completely inhibited FDM in chicks, but had accommodation-related mechanisms, that
longer (and less frequent) daily episodes of no effect on the refractive endpoint of nega- is, animals wearing minus lenses could
defocus of the same total duration.87 Recent tive lens-induced myopia.96 Similarly, high accommodate to compensate for hyperopic
studies in chicks88 and humans77 have intensity light levels eliminated the develop- defocus, which would be a signal for the eye
reported that ocular response to lens- ment of FDM, but only had modest effect on to grow.9 Additionally, monocular topical
induced defocus also depends on the time of compensation to hyperopic defocus in chicks application of atropine, a nonspecific mus-
day of exposure to defocus. A number of cur- and macaques.97,98 In another interesting carinic antagonist, effectively reduces exper-
rent optical treatment strategies for myopia study by Nickla and Totonelly, the D2 antago- imental myopia of the treated eye in animal
control, such as multifocal contact lenses and nist spiperone prevented the ocular growth models. The protective effects of atropine
orthokeratology, produce simultaneous com- inhibition induced by brief periods of clear were originally attributed to cycloplegic
peting hyperopic and myopic defocus signals vision in form-deprived eyes, but had no effects on the smooth muscle of the ciliary
across a large portion of the retina. In guinea effect on eyes wearing negative lenses, body, thereby eliminating accommoda-
pigs reared with dual-focus lenses of alter- suggesting that the dopaminergic mecha- tion.116 However, several studies in animal
nating −5 D/0 D, +5 D/0 D or −5 D/+5 D nisms mediating the protective effects of models have shown that lens-induced
power zones, the refractive change is equiva- brief periods of unrestricted vision may be defocus compensation and the mechanism
lent to the average of the two constituent different for form-deprivation versus lens- of atropine in myopia control are indepen-
powers.89 In chicks90 and marmosets,91 the induced defocus conditions. These results dent of accommodation. Accommodation
refractive compensation with dual-focus warrant further investigation into the cues originate in the retina, and afferent
58
signals are carried by the optic nerve to not a cause. Similarly, a recent study in Peripheral defocus
higher brain centres. The efferent pathway chicks showed that accommodative lag does
travels from the Edinger-Westphal nucleus not predict the magnitude of lens-induced Emmetropisation is largely an active process
of the midbrain to the ciliary ganglion, and myopia.126
guided by visual feedback that can be
is ultimately carried by the ciliary nerves to On the other hand, Diether and
achieved without input from the central
the ciliary muscle of both eyes.117 Accom- Wildsoet127 investigated whether accommo-
retina137–140 or the brain,62,141,142 although
modation signals from the Edinger-Westphal dation influenced the chick eye’s ability to
an intact optic nerve is required for a fine-
nucleus will initiate a binocular and consen- decode focusing errors and the relationship
tuned response.143,144 The ability of the
sual accommodation response. Lesions in with spatial frequency and contrast. They
chick retina to selectively guide eye growth
either the afferent or efferent components found that when accommodation was elimi-
of the accommodation pathway do not pre- nated through ciliary nerve section, decoding in localised areas experiencing partial retinal
vent lens-induced myopia.118 For example, of, and compensation for, imposed defocus deprivation was a key finding by Hodos and
optic nerve section in rhesus monkeys,119 was impaired. Specifically, ciliary nerve Kuenzel,40 and Wallman et al.,137 later con-
and Edinger-Westphal nucleus lesioning and section biased the eye growth response firmed in rodents and non-human pri-
ciliary nerve section in chicks,10,12,118 do not toward more myopia when competing hyper- mates.139,140,145 In addition to local form-
prevent defocus-induced myopia; eye opic and myopic signals were present. The deprivation, the retina can respond to
growth can still be regulated via visual cues authors concluded that accommodation regionally imposed defocus.121,146–148 Chicks
even without active accommodation. Addi- plays a role in decoding defocus during and non-human primates exposed to nega-
tionally, it was shown in chicks that atropine emmetropisation. Taken together, these tive and positive hemi-field defocus can
reduces experimental myopia through non- results have led researchers to suggest that a develop myopia and hyperopia in the
accommodative mechanisms, as atropine complex relationship exists between accom- corresponding retinal area.121,149 Not only
has no effect on the striated muscle of the modation and emmetropisation, involving hemi-retinal, but also peripheral defocus
chick ciliary body.120 These findings provide multiple neural pathways, feedback loops, can modify eye growth.91,147,148,150 The com-
support that eye growth is controlled by and interactions between temporal and spa-
pensation to negative peripheral defocus is
local mechanisms within the eye. Further tial patterns of defocus (see section on
in the same direction, but of lesser degree,
support that argues against a role of accom- peripheral defocus for details).
than the compensation to full-field negative
modation in defocus-induced eye growth Studies in animal models regarding the
defocus, whereas the compensation to posi-
comes from studies showing that growth in role of accommodation in myopia have both
tive defocus is in the same direction and
local regions of the eye can be modulated informed and complemented studies in
degree,146 in some cases greater,147,148 than
by defocus to only part of the visual field,121 humans, and vice versa. Early evidence in
whereas accommodation changes focus uni- humans linking near work to increased myo- the compensation to full-field positive
formly across the visual field. pia prevalence128,129 spurred much of the defocus. In chicks, marmosets, and
Findings that active accommodation is not work regarding interactions between accom- macaques, small treatment zones of periph-
necessary for experimental myopia do not modation and experimental myopia in ani- eral hyperopic defocus effectively stimulate
preclude the presence of a role of accom- mal models. With the finding that hyperopic axial eye growth,91,148,151 but larger treat-
modation in the development of myopia, defocus in animal models produces myopia, ment zones of peripheral myopic defocus
due to its association with retinal researchers investigated whether accommo- are required to slow growth and signifi-
defocus.122 Studies in chicks, tree shrews, dative lags in children promote myopia. cantly alter axial refraction (Figure 3).146–148
and rhesus monkeys have shown that brief However, results from various studies have In another study, Schippert and Schaeffel
periods of clear vision during a hyperopic shown conflicting results, such that some compared the central and peripheral (+450
defocus period inhibit experimental myopia report increased lags exist before the onset and −450) refractive development in chicks
development.12,123,124 These findings sug- of myopia130–132 and others report that wearing either full-field spectacle lenses of
gest that if an animal can eliminate defocus increased lags appear only after myopia +6.9 D and −7 D or lenses with central holes
induced from a minus lens by accommodat- onset.112 Findings in animals showing that
of 4, 6, and 8 mm diameter for four days.
ing, a similar myopia inhibitory effect should myopic defocus slows or prevents experi-
The study found that there was almost com-
be achieved. mental myopia have provided rationale for
plete ocular compensation for full-field
Consequently, high lags of accommoda- investigating whether the use of bifocal or
lenses, but no significant change in central
tion resulting in hyperopic retinal defocus progressive addition lenses slow myopia
refraction with holes in the centre of the
would be a stimulus for the eye to grow. To progression in children; however, results
lenses, suggesting that peripheral defocus
test this hypothesis, Troilo et al. examined have also been equivocal, showing a range
accommodative behaviour before and after of efficacy between studies from none133,134 does not necessarily affect central refractive
the induction of experimental myopia in to modest levels.135,136 While more recent development.152 Overall, these findings pro-
awake marmosets,125 and found that an studies show that bifocal or multifocal con- vide strong evidence supporting the conser-
increased accommodative lag was present tact lenses more effectively slow the pro- vation of emmetropisation mechanisms
after defocus-induced myopia. However, gression of myopia compared to spectacle across species, and highlight the importance
accommodative performance before lens lenses, the underlying mechanisms are not of studying animal models to understand
treatment did not predict the amount of well understood, and may be influenced by emmetropisation in humans.
myopia induced, suggesting that the both accommodation and peripheral For many years, the high foveal sensitivity
increased lag was a consequence of myopia, defocus. to defocus was thought to be essential for
59
Figure 3. Left panel shows scatter plot showing how the inter-ocular vitreous growth differences change as a function of contact
lens treatment zone area (mm2) multiplied by the power of the treatment zone (D). Right panel shows box plots describing inter-
ocular differences in ocular growth rate during treatment in untreated controls (white) and marmosets treated single-vision −5 D
(dark red), single-vision +5 D (dark blue), multizone +5/−5 D (black), −5 D/3 mm (light red), +5 D/3 mm (blue) and +5 D/1.5 mm
(light blue). The data shown are means SE. Adapted from Benavente-Perez et al., 2014.148 The Association for Research in Vision
and Ophthalmology is the copyright holder of this figure.
defocus detection and emmetropisation. retina would be expected to drive eye occur during both normal emmetropisation
However, non-foveated species like the growth. In humans, the pattern of periph- and lens-induced myopia.161 Not only does
fish,57 and species with comparatively low eral refraction is known to vary with central relative peripheral refraction change toward
spatial resolution like chickens,7 tree refraction; myopes tend to exhibit relative relative hyperopia during periods of
shrews,36 and guinea pigs,4 respond to form- peripheral hyperopia along the horizontal increased eye growth, there is evidence that
deprivation, suggesting that the foveal contri- axis, and hyperopes tend to exhibit relative peripheral refraction also changes when eye
bution may not be essential. Work by Smith peripheral myopia.154–156 Whether the growth decelerates. In marmoset eyes, rela-
et al.138 confirmed this hypothesis after peripheral profile may be a cause or a con- tive peripheral refraction changes toward
describing how rhesus monkeys with foveal sequence of central refractive development relative myopia during emmetropisation
ablation emmetropised normally and com- continues to be controversial.157 In rhesus periods of slower growth or recovery from
pensated for form-deprivation. The periph- monkeys, FDM causes a shift in peripheral visual compensation.162
eral retina could function alone, and an refraction toward relative hyperopia, which The role of peripheral refraction and its
intact fovea was not essential for increases with the degree of central myo- interaction with the temporal properties of
emmetropisation. Foveal information is also pia.158 Imposing full-field defocus on the ret- visually guided eye growth has also been
not essential for defocus compensation. The ina of marmosets triggers compensatory evaluated in marmosets.163 Peripheral
eyes of chicks and non-human primates can changes in eye growth and refractive state refraction at baseline can predict the com-
recover from induced refractions in the that lead to asymmetries in the refraction of pensatory changes in eye growth only in
absence of foveal signals.140,153 The ability of the peripheral retina.159 In both marmosets combination with on-axis refraction, or after
the eye to respond to localised visual manip- and rhesus monkeys, the strength of the the eyes have begun to compensate for the
ulations confirmed that emmetropisation relationship between central and peripheral imposed negative defocus. Therefore,
does not depend on a central neural mecha- refraction varies with eccentricity, as does peripheral refraction changes as a conse-
nism, but on a local and regionally selective the degree of peripheral refraction quence of myopia development and can
retinal mechanism located within the eye, assymetry.159–161 Marmosets exhibit nasal- predict myopia progression when eyes have
which opened new avenues for eye growth temporal asymmetries in peripheral refrac- started to develop myopia.163 These results,
manipulation in humans. tion that change with age toward relative combined with previous results from Ben-
Since the mechanism of emmetropisation nasal hyperopia. The changes are similar, avente and Troilo’s group,160–163 provide evi-
is contained within the eye and peripheral but greater, in animals treated with full-field dence of an interaction between the
defocus can alter refractive development, negative defocus, suggesting that asymme- refractive asymmetry of the peripheral ret-
the defocus experienced by the peripheral try changes in peripheral refraction can ina and the visual experience of the central
60
retina, such that peripheral refraction

appears to be both cause and effect of axial
growth, becoming a possible factor in the
progression of myopia and offering a means
to control it.
Ocular shape and peripheral

defocus
The nature of ocular growth in myopia has
been described as either global, equatorial
or axial, depending on the region of the eye
undergoing stretching.164–166 Given the
magnitude of defocus depends on the loca-
tion of the image with respect to the retinal
plane, the shape of the retina has been
investigated in association with peripheral
refraction and refractive error (see review
by Verkicharla et al.).166 The evidence from
research studies involving primates and
human participants are in agreement with
the hypothesis that peripheral defocus influ-
ences peripheral eye shape and myopia at
large. Previous human studies have mea-
sured the shape of the retina (or the poste-
rior eye) using magnetic resonance
imaging167,168 and other indirect optical
methods169–173 and correlated it with
peripheral refraction. As shown in Figure 4,
myopic eyes with steeper or prolate retinal
shape exhibit relative peripheral hyperopia,
whereas hyperopic or emmetropic eyes with
flatter or oblate retinas exhibit relative
peripheral myopia.166–168,171,174,175
In infant rhesus monkeys158,176 and
marmosets,177 the change in peripheral
refraction with form-deprivation or lens-
induced defocus applied to a specific part of Figure 4. Possible uncorrected image shells relative to the retina. A: Steeper or pro-
the visual field correlated with the inter- late myopic retina where the central refraction (c) is focused in front of the retina
ocular differences in vitreous chamber and the peripheral refraction at extreme point p is focused behind the retina, causing
shape, as observed with magnetic reso- a relative peripheral hyperopia. B: Flatter or oblate emmetropic retina where the cen-
nance imaging. A large number of studies tral refraction (c) is focused at the retina and the peripheral refraction at extreme
involving humans have investigated periph- point p is focused in front of the retina, causing a relative peripheral myopia. Note:
eral refraction, with only a few looking at this is an exaggerated view and might represent refraction beyond central 70 degrees.
peripheral refraction and retinal shape in Adapted from Verkicharla et al., 2012, with permission.166
combination. Verkicharla et al.178,179
reported that peripheral refraction, periph-
eral eye lengths, and retinal shapes mea- assessed the posterior pole shape in highly is now growing evidence based on the longi-
sured from partial coherence interferometry myopic eyes by measuring the curvature of tudinal studies conducted in Caucasian and
were significantly affected by race, as well Bruch’s membrane using optical coherence Chinese populations indicating that periph-
as by meridian and refraction. East Asians tomography, and reported significant pro- eral hyperopic defocus may not predict the
were found to have steeper retinas and truding and undulating changes in the development or the progression of
greater relative peripheral hyperopia than shape of the posterior eye with time in myopia.182–189 These recent findings suggest
Caucasians, with steepness being greater highly myopic eyes, indicating a potential there is a complex interaction between
along the horizontal meridian than the verti- role of the posterior eye shape in progres- peripheral hyperopic defocus and myopia,
cal meridian. The differences among races sion of high myopia. and there may be a combination of optical
and meridians were attributed to structural Although various anti-myopia strategies factors that influence ocular growth associ-
variations of the eye between races, and have been designed to counteract the periph- ated with peripheral blur. Because retinal
more space in the orbit around the eye ver- eral hyperopic defocus that are also effective shape can alter the magnitude of defocus in
tically than horizontally. Wakazono et al.180 in controlling myopia progression,99,181 there the peripheral retina, it may play an
61
important role in mediating the development myopia and other refractive disorders. Stud- aberration in chicks. Interestingly, human
and progression of myopia in humans. ies in chicks,192 marmosets,193 and myopic eyes also exhibit higher levels of posi-
194
Future studies are warranted to fully eluci- monkeys, as well as humans,195 have tive spherical aberration.199 The optical
date the influence of retinal shape in the reported a systemic reduction in HOAs with changes associated with FDM or lens-induced
pathogenesis of myopia. age due to changes in the curvature and ametropias are believed to be a combination
thickness of the cornea and the crystalline of changes in the curvatures and refractive
lens, as well as refractive index of the lens. index of the eye’s optical components, as well
Higher order monochromatic Despite some interspecies differences, most as dynamic changes in the relative position of
aberrations animal studies have noted a relatively small the crystalline lens with respect to the cor-
influence of HOAs in age-dependent improve- nea.198 Finally, a study by Ramamirtham
While the optical characteristics of the eye ments in spatial vision and contrast sensitiv- et al.198 found increased HOAs in both myo-
are largely dominated by lower order aber- ity. It is hypothesised that a large part of the pic and hyperopic monkey eyes that were
rations (vertical/oblique astigmatism and reduction in HOAs during the early postnatal strongly correlated with the degree of lower
defocus), presence of higher order mono- period occurs passively, without any major order aberrations and axial ametropias,
chromatic aberrations (HOAs) can degrade input from the visual environment.196 Animal suggesting that the changes in HOAs may
retinal image quality, and therefore, may studies have also examined the relationship occur as a consequence, not a cause, of
play a role in the development of refractive between experimentally induced ametropias refractive errors.
errors. HOAs may also interact with lower and HOAs. In this respect, form-deprivation Previous human studies have noted that
order aberrations (or spherical refractive and hyperopic defocus induced experimental spherical aberration, coma, and trefoil are
error) and/or change the eye’s depth of myopias have been found to be associated the largest contributors of HOAs in normal
focus to alter the optics, and hence refrac- with greater levels of HOAs in different ani- healthy eyes; however, there are significant
tive development of the eye.190,191 mal models, albeit with minor interspecies inter-subject variations in the type and mag-
Animal studies have provided important differences.192,197,198 For instance, experimen- nitude of HOAs in the population.190,200,201
insights into the changes in HOAs during tal ametropias are associated with more posi- Consistent with the observation in animals,
emmetropisation and how these aberrations tive spherical aberration in monkeys, but the influence of HOAs in the development
might be involved in the development of greater amounts of negative spherical of myopia is unclear, with some studies
Figure 5. Spherical-equivalent, spectacle-plane refractive corrections plotted as a function of age for the treated (filled symbols)
and fellow eyes (open symbols) of representative lens-reared controls (top row) and red-light-reared monkeys (bottom row). Ani-
mals were reared with −3.0 D lenses in front of their treated eyes and plano lenses in front of their fellow eyes. The thin grey
lines in each plot represent data for the right eyes of the 39 normal control monkeys. Overall, red light-rearing prevented lens-
induced defocus in treated eyes, and resulted in hyperopia in both treated and fellow control eyes. Adapted from Hung et al.,
2018, with permission.223
62
suggesting an increase in HOAs are associ- influence of spectral composition of light on Unlike results in fish, chicks, and guinea
ated with myopia (particularly spherical eye growth. In fish, chicks, and guinea pigs, pigs, long wavelength light rearing in non-
aberration and coma),202,203 while others eyes were less myopic when raised under human primates results in a decrease in eye
reporting no significant change in ocular short wavelength (violet and blue) light growth.223 Recent studies in rhesus mon-
aberrations with myopia.204–206 It is impor- compared to those raised under longer keys have shown that animals raised with
tant to note that accommodation also wavelength (green or red) light.209,212,218–221 red filters over one or both eyes,224 or in
induces changes in HOA, as well as lower In these studies, short-term eye growth ambient light dominated by long wave-
order aberrations.207,208 Overall, evidence matched the direction and magnitude lengths (produced by red light-emitting
from animal and human studies suggest a predicted by LCA.212,222 However, longer- diodes),223 demonstrate choroidal thicken-
possible role of HOAs in visual regulation of term eye growth under these same condi- ing and slowed eye growth, resulting in less
ocular growth; however, longitudinal clinical tions surpassed what would be predicted myopic refractive errors (Figure 5). Animals
studies during childhood are needed to con- by LCA, indicating a more complex interac- raised under long wavelength light with no
firm their role in human emmetropisation. tion between chromatic cues and other lens treatment demonstrate choroidal thick-
signals for eye growth. Jiang et al. hypo- ening in both eyes, as do animals with
thesised that blue light might preferentially a +3 D or a −3 D lens over one eye. Animals,
Spectral and temporal stimulate the ON pathway to inhibit myopic on average, maintained slightly hyperopic
characteristics of light eye growth.218 refractive errors. Similar effects of red light
The spectral and temporal characteristics of

light represent another optical mechanism
implicated in eye growth, and includes such
factors as longitudinal chromatic aberration
(LCA), wavelength, intensity, and exposure
time. Experiments in a laboratory setting in ani-
mal models allow these characteristics of the
visual environment to be manipulated to better
understand their influence on eye growth.
Evidence suggests that the eye utilises
LCA to guide axial eye growth and the
refractive state.209,210 LCA causes long wave-
lengths to be focused in a more hyperopic
plane than shorter wavelengths. Therefore,
the eye’s total refraction varies inversely
with wavelength, rendering the eye rela-
tively more hyperopic (less myopic) for long
wavelength light. Broadband light results in
colour fringes on the retinal image that pro-
vide a signal as to whether defocus is hyper-
opic or myopic.211 Studies have shown that
the human eye can utilise LCA as a direc-
tional cue for accommodation;101,212 how-
ever, the role of LCA in emmetropisation is
not fully understood. The ability of the ret-
ina to detect LCA depends on the distribu-
tion of short, medium, and long wavelength
sensitive cones.213 In humans, the short
wavelength sensitive cones are fewest in
number.214 Early reports in humans suggest
that there is a reduction in sensitivity of
short wavelength cones in myopia when
tested electrophysiologically.215,216 Another
mechanism by which the retina might dis-
criminate between long and short wave-
lengths may be through spectral tuning of
the intrinsically photosensitive retinal gan-
glion cells, which are most sensitive to short Figure 6. Broadband illumination in lux (yellow trace) and power in μw/cm2 of the
wavelength light.217 blue (400–500 nm), green (500–600 nm), and red (600–700 nm) components across
Animal studies have shown conflicting 24 hours of A: typical indoor lighting and B: outdoors in full sun in Houston, TX, USA.
results across species with respect to the Ostrin, unpublished data.
63
have been shown in tree shrews.225,226 Addi- a powerful influence on refractive state by 12. Schmid KL, Wildsoet CF. Effects on the compensatory
responses to positive and negative lenses of inter-
tionally, tree shrews reared in short wave- controlling the axial length and overall mittent lens wear and ciliary nerve section in chicks.
length flickering light demonstrated growth of the eye during the postnatal Vision Res 1996; 36: 1023–1036.
myopia.226 The chromatic cues dominated developmental period. These studies have 13. Smith EL 3rd, Hung LF. The role of optical defocus in
regulating refractive development in infant monkeys.
control of eye growth such that defocus been instrumental to our current under- Vision Res 1999; 39: 1415–1435.
cues were largely ignored. Specifically, even standing of emmetropisation and develop- 14. Hung LF, Crawford ML, Smith EL. Spectacle lenses
alter eye growth and the refractive status of young
animals that wore negative lenses, which ment of refractive errors in humans.
monkeys. Nat Med 1995; 1: 761–765.
would normally induce myopia, remained Research on birds, mammals, rodents, and 15. Howlett MH, McFadden SA. Spectacle lens compen-
relatively hyperopic when reared in long non-human primates suggests that the eye sation in the pigmented guinea pig. Vision Res 2009;
49: 219–227.
wavelength light. The mechanisms of the uses several optical cues to modulate its 16. Troilo D, Smith EL 3rd, Nickla DL et al. IMI - report on
protective effects of narrowband wave- growth during emmetropisation. Visual sig- experimental models of emmetropization and myo-
length light on myopia remain elusive. nals from active accommodation, HOAs, pia. Invest Ophthalmol Vis Sci 2019; 60: M31–M88.
17. Schaeffel F, Feldkaemper M. Animal models in myo-
Numerous studies have observed that peripheral defocus, and chromatic aberra- pia research. Clin Exp Optom 2015; 98: 507–517.
increased ambient illumination is protective tions could modulate the sign and magni- 18. Smith EL 3rd, Hung LF, Arumugam B. Visual regula-
tion of refractive development: insights from animal
for FDM in animal models,98,227,228 and time tude of defocus on the retina, and hence
studies. Eye (Lond) 2014; 28: 180–188.
outdoors is protective for myopia in chil- refractive development of the eye. In addi- 19. Bradley DV, Fernandes A, Lynn M et al.
dren.229 The mechanisms of the protective tion, these optical mechanisms may alter Emmetropization in the rhesus monkey (Macaca mul-
atta): birth to young adulthood. Invest Ophthalmol Vis
effects of light are unknown, and might the temporal and spatial integration of Sci 1999; 40: 214–229.
include both neurochemical factors, such as defocus signals across the retina. Further- 20. Graham B, Judge SJ. Normal development of refrac-
alterations in melanopsin and dopamine more, features of ambient lighting, such as tive state and ocular component dimensions in the
marmoset (Callithrix jacchus). Vision Res 1999; 39:
cascades,97,230 and optical factors, such as the duration and intensity of lighting, may 177–187.
miosis-induced increased depth of field, a flat- also affect visually guided ocular growth. 21. Norton TT. Animal models of myopia: learning how
vision controls the size of the eye. ILAR J 1999; 40: 59–77.
ter dioptric scene, and spectral distribution of Several of these findings have been success-
22. Zhou X, Qu J, Xie R et al. Normal development of
sunlight and chromatic cues. Ambient illumi- fully translated into effective optical treat- refractive state and ocular dimensions in guinea
nation outdoors ranges from 1,000 lux on a ment strategies for refractive error pigs. Vision Res 2006; 46: 2815–2823.
23. Norton TT, McBrien NA. Normal development of
cloudy day in the shade to greater than correction in humans, and will continue to refractive state and ocular component dimensions in
150,000 lux in direct sun,231 and the spectral aid in the development of novel and effec- the tree shrew (Tupaia belangeri). Vision Res 1992; 32:
composition is comprised of a broad range of tive treatment options in the future. 833–842.
24. Li T, Troilo D, Glasser A et al. Constant light produces
wavelengths, including ultraviolet, visible, and severe corneal flattening and hyperopia in chickens.
infrared wavelengths. Because the intensity of ACKNOWLEDGEMENT Vision Res 1995; 35: 1203–1209.
25. Schmid KL, Hills T, Abbott M et al. Relationship
the entire visible spectrum exponentially Special thanks to Daniel Moderiano (Flinders
between intraocular pressure and eye growth in
increases outdoors compared to indoors University) for help generating figures. chick. Ophthalmic Physiol Opt 2003; 23: 25–33.
(Figure 6), it is difficult to determine if the pro- 26. Pardue MT, Stone RA, Iuvone PM. Investigating
mechanisms of myopia in mice. Exp Eye Res 2013;
tective effects of outdoor light are attributed
REFERENCES 114: 96–105.
to a specific narrowband region of the spec- 1. Wallman J, Winawer J. Homeostasis of eye growth 27. Troilo D, Judge SJ. Ocular development and visual
trum. Another possible mechanism of the and the question of myopia. Neuron 2004; 43: deprivation myopia in the common marmoset
447–468. (Callithrix jacchus). Vision Res 1993; 33: 1311–1324.
protective effect in children of time outdoors 28. Ingram RM, Barr A. Changes in refraction between
2. Smith EL 3rd. Spectacle lenses and emmetropization:
is a corresponding decrease in time spent the role of optical defocus in regulating ocular devel- the ages of 1 and 3 1/2 years. Br J Ophthalmol 1979;
opment. Optom Vis Sci 1998; 75: 388–398. 63: 339–342.
indoors performing near tasks and being
3. van Alphen G. On emmetropia and ametropia. Opt 29. Wood IC, Hodi S, Morgan L. Longitudinal change of
exposed to accommodation-related defocus. Acta (Lond) 1961; 142: 1–92. refractive error in infants during the first year of life.
Findings in animal models demonstrating 4. Howlett MH, McFadden SA. Form-deprivation myopia Eye (Lond) 1995; 9: 551–557.
in the guinea pig (Cavia porcellus). Vision Res 2006; 30. Cook RC, Glasscock RE. Refractive and ocular find-
that spectral and temporal properties of ings in the newborn. Am J Ophthalmol 1951; 34:
46: 267–283.
light affect emmetropisation have led to 5. Siegwart JT Jr, Norton TT. The susceptible period for 1407–1413.
speculation that modifying the characteris- deprivation-induced myopia in tree shrew. Vision Res 31. Gordon RA, Donzis PB. Refractive development of
1998; 38: 3505–3515. the human eye. Arch Ophthalmol 1985; 103: 785–789.
tics of indoor ambient illumination could be 6. Wallman J, Turkel J, Trachtman J. Extreme myopia 32. Mutti DO, Mitchell GL, Jones LA et al. Axial growth
used as a method to prevent slow progres- produced by modest change in early visual experi- and changes in lenticular and corneal power during
ence. Science 1978; 201: 1249–1251. emmetropization in infants. Invest Ophthalmol Vis Sci
sion of myopia in children. However, con-
7. Wallman J, Wildsoet C, Xu A et al. Moving the retina: 2005; 46: 3074–3080.
flicting results between animal models and choroidal modulation of refractive state. Vision Res 33. Pennie FC, Wood IC, Olsen C et al. A longitudinal
a rudimentary understanding of mecha- 1995; 35: 37–50. study of the biometric and refractive changes in full-
8. Smith EL 3rd, Hung LF, Kee CS et al. Effects of brief term infants during the first year of life. Vision Res
nisms involved suggest that further studies 2001; 41: 2799–2810.
periods of unrestricted vision on the development of
are warranted before manipulation of form-deprivation myopia in monkeys. Invest 34. Sorsby A, Leary GA, Richards MJ. Correlation ametro-
indoor illumination is utilised as a treatment Ophthalmol Vis Sci 2002; 43: 291–299. pia and component ametropia. Vision Res 1962; 2:
9. Schaeffel F, Glasser A, Howland HC. Accommodation, 309–313.
strategy in children. refractive error and eye growth in chickens. Vision 35. Wiesel TN, Raviola E. Myopia and eye enlargement
Res 1988; 28: 639–657. after neonatal lid fusion in monkeys. Nature 1977;
10. Wildsoet C, Wallman J. Choroidal and scleral mecha- 266: 66–68.
Conclusion nisms of compensation for spectacle lenses in 36. Sherman SM, Norton TT, Casagrande VA. Myopia in
chicks. Vision Res 1995; 35: 1175–1194. the lid-sutured tree shrew (Tupaia glis). Brain Res
11. Irving EL, Callender MG, Sivak JG. Inducing myopia, 1977; 124: 154–157.
A large volume of work on animal models hyperopia, and astigmatism in chicks. Optom Vis Sci 37. Wilson J, Sherman S. Differential effects of early
suggests that the visual environment exerts 1991; 68: 364–368. monocular deprivation on binocular and monocular
64
segments of cat striate cortex. J Neurophysiol 1977; 61. Papastergiou GI, Schmid GF, Laties AM et al. Induc- 85. Zhu X, Wallman J. Temporal properties of compensa-
40: 891–903. tion of axial eye elongation and myopic refractive tion for positive and negative spectacle lenses in
38. Wiesel TN, Raviola E. Increase in axial length of the shift in one-year-old chickens. Vision Res 1998; 38: chicks. Invest Ophthalmol Vis Sci 2009; 50: 37–46.
macaque monkey eye after corneal opacification. 1883–1888. 86. Nickla DL, Sharda V, Troilo D. Temporal integration
Invest Ophthalmol Vis Sci 1979; 18: 1232–1236. 62. Troilo D, Wallman J. The regulation of eye growth characteristics of the axial and choroidal responses
39. Lu F, Zhou X, Zhao H et al. Axial myopia induced by a and refractive state: an experimental study of to myopic defocus induced by prior form deprivation
monocularly-deprived facemask in guinea pigs: a emmetropization. Vision Res 1991; 31: 1237–1250. versus positive spectacle lens wear in chickens.
non-invasive and effective model. Exp Eye Res 2006; 63. Guyton DL, Greene PR, Scholz RT. Dark-rearing inter- Optom Vis Sci 2005; 82: 318–327.
82: 628–636. ference with emmetropization in the rhesus monkey. 87. Winawer J, Wallman J. Temporal constraints on lens
40. Hodos W, Kuenzel WJ. Retinal-image degradation Invest Ophthalmol Vis Sci 1989; 30: 761–764. compensation in chicks. Vision Res 2002; 42:
produces ocular enlargement in chicks. Invest 64. Qiao-Grider Y, Hung LF, Kee CS et al. Recovery from 2651–2668.
Ophthalmol Vis Sci 1984; 25: 652–659. form-deprivation myopia in rhesus monkeys. Invest 88. Nickla DL, Thai P, Zanzerkia Trahan R et al. Myopic
41. Wallman J, Ledoux C, Friedman MB. Simple devices Ophthalmol Vis Sci 2004; 45: 3361–3372. defocus in the evening is more effective at inhibiting
for restricting the visual fields of birds. Behav Res 65. O’Leary DJ, Millodot M. Eyelid closure causes myopia eye growth than defocus in the morning: effects on
Methods Instrum 1978; 10: 401–403. in humans. Experientia 1979; 35: 1478–1479. rhythms in axial length and choroid thickness in
42. Ashby R, Ohlendorf A, Schaeffel F. The effect of 66. von Noorden GK, Lewis RA. Ocular axial length in chicks. Exp Eye Res 2017; 154: 104–115.
ambient illuminance on the development of depriva- unilateral congenital cataracts and blepharoptosis. 89. McFadden SA, Tse DY, Bowrey HE et al. Integration
tion myopia in chicks. Invest Ophthalmol Vis Sci 2009; Invest Ophthalmol Vis Sci 1987; 28: 750–752. of defocus by dual power Fresnel lenses inhibits
50: 5348–5354. 67. Gee SS, Tabbara KF. Increase in ocular axial length in myopia in the mammalian eye. Invest Ophthalmol Vis
43. Beresford JA, Crewther SG, Crewther DP. Anatomical patients with corneal opacification. Ophthalmology Sci 2014; 55: 908–917.
correlates of experimentally induced myopia. Aust N 1988; 95: 1276–1278. 90. Tse DY, Lam CS, Guggenheim JA et al. Simultaneous
Z J Ophthalmol 1998; 26: S84–S87. 68. Miller-Meeks MJ, Bennett SR, Keech RV et al. Myopia defocus integration during refractive development.
44. Siegwart JT Jr, Norton TT. Goggles for controlling the induced by vitreous hemorrhage. Am J Ophthalmol Invest Ophthalmol Vis Sci 2007; 48: 5352–5359.
visual environment of small animals. Lab Anim Sci 1990; 109: 199–203. 91. Benavente-Perez A, Nour A, Troilo D. The effect of
1994; 44: 292–294. 69. Morgan IG, Ashby RS, Nickla DL. Form deprivation simultaneous negative and positive defocus on eye
45. Faulkner AE, Kim MK, Iuvone PM et al. Head- and lens-induced myopia: are they different? Oph- growth and development of refractive state in mar-
mounted goggles for murine form deprivation myo- thalmic Physiol Opt 2013; 33: 355–361. mosets. Invest Ophthalmol Vis Sci 2012; 53:
pia. J Neurosci Methods 2007; 161: 96–100. 70. Irving EL, Sivak JG, Callender MG. Refractive plasticity 6479–6487.
46. Chakraborty R, Park H, Aung MH et al. Comparison of the developing chick eye. Ophthalmic Physiol Opt 92. Arumugam B, Hung LF, To CH et al. The effects of
of refractive development and retinal dopamine in 1992; 12: 448–456. simultaneous dual focus lenses on refractive devel-
OFF pathway mutant and C57BL/6J wild-type mice. 71. Nickla DL, Wildsoet C, Wallman J. Visual influences opment in infant monkeys. Invest Ophthalmol Vis Sci
Mol Vis 2014; 20: 1318–1327. on diurnal rhythms in ocular length and choroidal 2014; 55: 7423–7432.
47. Chakraborty R, Yang V, Park HN et al. Lack of cone thickness in chick eyes. Exp Eye Res 1998; 66: 93. Schmid GF, Papastergiou GI, Lin T et al. Autonomic
mediated retinal function increases susceptibility to 163–181. denervations influence ocular dimensions and intra-
form-deprivation myopia in mice. Exp Eye Res 2019; 72. Norton TT, Siegwart JT Jr, Amedo AO. Effectiveness of ocular pressure in chicks. Exp Eye Res 1999; 68:
180: 226–230. hyperopic defocus, minimal defocus, or myopic 573–581.
48. Schaeffel F, Howland HC. Properties of the feedback defocus in competition with a myopiagenic stimulus 94. Fujikado T, Kawasaki Y, Suzuki A et al. Retinal func-
loops controlling eye growth and refractive state in in tree shrew eyes. Invest Ophthalmol Vis Sci 2006; tion with lens-induced myopia compared with form-
the chicken. Vision Res 1991; 31: 717–734. 47: 4687–4699. deprivation myopia in chicks. Graefes Arch Clin Exp
49. Tkatchenko TV, Shen Y, Tkatchenko AV. Mouse 73. Graham B, Judge SJ. The effects of spectacle wear in Ophthalmol 1997; 235: 320–324.
experimental myopia has features of primate myo- infancy on eye growth and refractive error in the 95. Choh V, Lew MY, Nadel MW et al. Effects of inter-
pia. Invest Ophthalmol Vis Sci 2010; 51: 1297–1303. marmoset (Callithrix jacchus). Vision Res 1999; 39: changing hyperopic defocus and form deprivation
50. Troilo D, Nickla DL, Wildsoet CF. Form deprivation 189–206. stimuli in normal and optic nerve-sectioned chicks.
myopia in mature common marmosets (Callithrix 74. Barathi VA, Boopathi VG, Yap EP et al. Two models of Vision Res 2006; 46: 1070–1079.
jacchus). Invest Ophthalmol Vis Sci 2000; 41: experimental myopia in the mouse. Vision Res 2008; 96. Bartmann M, Schaeffel F, Hagel G et al. Constant
2043–2049. 48: 904–916. light affects retinal dopamine levels and blocks dep-
51. Gottlieb MD, Fugate-Wentzek LA, Wallman J. Differ- 75. Troilo D, Totonelly K, Harb EN. Imposed anisometro- rivation myopia but not lens-induced refractive
ent visual deprivations produce different ametropias pia, accommodation, and regulation of refractive errors in chickens. Vis Neurosci 1994; 11: 199–208.
and different eye shapes. Invest Ophthalmol Vis Sci state. Optom Vis Sci 2009; 86: 31–39. 97. Ashby RS, Schaeffel F. The effect of bright light on
1987; 28: 1225–1235. 76. Irving EL, Callender MG, Sivak JG. Inducing ametro- lens compensation in chicks. Invest Ophthalmol Vis
52. McBrien NA, Lawlor P, Gentle A. Scleral remodeling pias in hatchling chicks by defocus–aperture effects Sci 2010; 51: 5247–5253.
during the development of and recovery from axial and cylindrical lenses. Vision Res 1995; 35: 98. Smith EL 3rd, Hung LF, Huang J. Protective effects of
myopia in the tree shrew. Invest Ophthalmol Vis Sci 1165–1174. high ambient lighting on the development of form-
2000; 41: 3713–3719. 77. Moderiano D, Do M, Hobbs S et al. Influence of the deprivation myopia in rhesus monkeys. Invest
53. Gottlieb MD, Joshi HB, Nickla DL. Scleral changes in time of day on axial length and choroidal thickness Ophthalmol Vis Sci 2012; 53: 421–428.
chicks with form-deprivation myopia. Curr Eye Res changes to hyperopic and myopic defocus in human 99. Walline JJ, Lindsley K, Vedula SS et al. Interventions
1990; 9: 1157–1165. eyes. Exp Eye Res 2019; 182: 125–136. to slow progression of myopia in children. Cochrane
54. Smith EL 3rd, Hung LF. Form-deprivation myopia in 78. Chakraborty R, Read SA, Collins MJ. Monocular myo- Database Syst Rev 2011; 12: CD004916.
monkeys is a graded phenomenon. Vision Res 2000; pic defocus and daily changes in axial length and 100. Cho P, Cheung SW. Retardation of myopia in ortho-
40: 371–381. choroidal thickness of human eyes. Exp Eye Res 2012; keratology (ROMIO) study: a 2-year randomized clini-
55. Wallman J, Adams JI. Developmental aspects of 103: 47–54. cal trial. Invest Ophthalmol Vis Sci 2012; 53:
experimental myopia in chicks: susceptibility, recov- 79. Chakraborty R, Read SA, Collins MJ. Hyperopic 7077–7085.
ery and relation to emmetropization. Vision Res defocus and diurnal changes in human choroid and 101. Kruger PB, Mathews S, Aggarwala KR et al. Chro-
1987; 27: 1139–1163. axial length. Optom Vis Sci 2013; 90: 1187–1198. matic aberration and ocular focus: Fincham
56. Chakraborty R, Park HN, Hanif AM et al. ON pathway 80. Wang D, Chun RK, Liu M et al. Optical defocus rapidly revisited. Vision Res 1993; 33: 1397–1411.
mutations increase susceptibility to form-deprivation changes choroidal thickness in schoolchildren. PLoS
102. Del Aguila-Carrasco AJ, Marín-Franch I, Bernal-
myopia. Exp Eye Res 2015; 137: 79–83. One 2016; 11: e0161535. Molina P et al. Accommodation responds to optical
57. Shen W, Vijayan M, Sivak JG. Inducing form- 81. Zhu X. Temporal integration of visual signals in lens vergence and not defocus blur alone. Invest
deprivation myopia in fish. Invest Ophthalmol Vis Sci compensation (a review). Exp Eye Res 2013; 114: Ophthalmol Vis Sci 2017; 58: 1758–1763.
2005; 46: 1797–1803. 69–76. 103. Ostrin LA, Liu Y, Choh V et al. The role of the iris in
58. Smith EL 3rd, Hung LF, Harwerth RS. Effects of opti- 82. Zhu X, Winawer JA, Wallman J. Potency of myopic chick accommodation. Invest Ophthalmol Vis Sci
cally induced blur on the refractive status of young defocus in spectacle lens compensation. Invest 2011; 52: 4710–4716.
monkeys. Vision Res 1994; 34: 293–301. Ophthalmol Vis Sci 2003; 44: 2818–2827. 104. Clarke RJ, Coimbra CJ, Alessio ML. Oculomotor areas
59. Bartmann M, Schaeffel F. A simple mechanism for 83. Winawer J, Zhu X, Choi J et al. Ocular compensation involved in the parasympathetic control of accom-
emmetropization without cues from accommodation for alternating myopic and hyperopic defocus. Vision modation and pupil size in the marmoset (Callithrix
or colour. Vision Res 1994; 34: 873–876. Res 2005; 45: 1667–1677. jacchus). Braz J Med Biol Res 1985; 18: 373–379.
60. Smith EL 3rd, Bradley DV, Fernandes A et al. Form 84. Zhu X, Park TW, Winawer J et al. In a matter of 105. Ostrin LA, Glasser A. Autonomic drugs and the
deprivation myopia in adolescent monkeys. Optom minutes, the eye can know which way to grow. Invest accommodative system in rhesus monkeys. Exp Eye
Vis Sci 1999; 76: 428–432. Ophthalmol Vis Sci 2005; 46: 2238–2241. Res 2010; 90: 104–112.
65
106. Croft MA, Kaufman PL, Crawford KS et al. Accommo- before and at the onset of myopia in children. 152. Schippert R, Schaeffel F. Peripheral defocus does not
dation dynamics in aging rhesus monkeys. Optom Vis Sci 2005; 82: 273–278. necessarily affect central refractive development.
Am J Physiol 1998; 275: R1885–R1897. 131. Drobe B, de Saint-André R. The pre-myopic syn- Vision Res 2006; 46: 3935–3940.
107. Glasser A, Troilo D, Howland HC. The mechanism of drome. Ophthalmic Physiol Opt 1995; 15: 375–378. 153. Wildsoet CF, Schmid KL. Optical correction of form
corneal accommodation in chicks. Vision Res 1994; 132. Goss DA. Clinical accommodation and heterophoria deprivation myopia inhibits refractive recovery in
34: 1549–1566. findings preceding juvenile onset of myopia. Optom chick eyes with intact or sectioned optic nerves.
108. Glasser A, Murphy CJ, Troilo D et al. The mechanism Vis Sci 1991; 68: 110–116. Vision Res 2000; 40: 3273–3282.
of lenticular accommodation in chicks. Vision Res 133. Shih YF, Hsiao CK, Chen CJ et al. An intervention trial 154. Ferree GR, Rand G. Interpretation of refractive condi-
1995; 35: 1525–1540. on efficacy of atropine and multi-focal glasses in tions in the peripheral field of vision: a further study.
109. Troilo D, Wallman J. Changes in corneal curvature controlling myopic progression. Acta Ophthalmol Arch Ophthalmol 1933; 9: 925–938.
during accommodation in chicks. Vision Res 1987; 27: Scand 2001; 79: 233–236. 155. Millodot M. Effect of ametropia on peripheral refrac-
241–247. 134. Edwards MH, Li RW, Lam CS et al. The Hong Kong tion. Am J Optom Physiol Opt 1981; 58: 691–695.
110. Schaeffel F, Howland HC, Farkas L. Natural accom- progressive lens myopia control study: study design 156. Mutti DO, Sholtz RI, Friedman NE et al. Peripheral
modation in the growing chicken. Vision Res 1986; and main findings. Invest Ophthalmol Vis Sci 2002; 43: refraction and ocular shape in children. Invest
26: 1977–1993. 2852–2858. Ophthalmol Vis Sci 2000; 41: 1022–1030.
111. Day M, Gray LS, Seidel D et al. The relationship 135. Leung JT, Brown B. Progression of myopia in Hong 157. Seidemann A, Schaeffel F, Guirao A et al. Peripheral
between object spatial profile and accommodation Kong Chinese schoolchildren is slowed by wearing refractive errors in myopic, emmetropic, and hyper-
microfluctuations in emmetropes and myopes. J Vis progressive lenses. Optom Vis Sci 1999; 76: opic young subjects. J Opt Soc Am A Opt Image Sci Vis
2009; 9: 1–13. 346–354. 2002; 19: 2363–2373.
112. Mutti DO, Mitchell GL, Hayes JR et al. Accommoda- 136. Gwiazda J, Hyman L, Hussein M et al. A randomized 158. Huang J, Hung LF, Ramamirtham R et al. Effects of
tive lag before and after the onset of myopia. Invest clinical trial of progressive addition lenses versus sin- form deprivation on peripheral refractions and ocu-
Ophthalmol Vis Sci 2006; 47: 837–846. gle vision lenses on the progression of myopia in lar shape in infant rhesus monkeys (Macaca mul-
113. Gwiazda J, Bauer J, Thorn F et al. Shifts in tonic children. Invest Ophthalmol Vis Sci 2003; 44: atta). Invest Ophthalmol Vis Sci 2009; 50: 4033–4044.
accommodation after near work are related to 1492–1500. 159. Totonelly KC, Coletta NJ, Troilo D. Lens–induced
refractive errors in children. Ophthalmic Physiol Opt 137. Wallman J, Gottlieb MD, Rajaram V et al. Local retinal refractive errors alter the pattern of peripheral
1995; 15: 93–97. regions control local eye growth and myopia. Science refractive state in marmosets. Invest Ophthalmol Vis
114. McBrien NA, Millodot M. The relationship between 1987; 237: 73–77. Sci 2006; 47: 3323.
tonic accommodation and refractive error. Invest 138. Smith EL 3rd, Ramamirtham R, Qiao-Grider Y et al. 160. Benavente-Perez A, Nour A, Troilo D. Peripheral
Ophthalmol Vis Sci 1987; 28: 997–1004. Effects of foveal ablation on emmetropization and refraction as a predictor for induced changes in vit-
115. Gwiazda J, Thorn F, Bauer J et al. Myopic children form-deprivation myopia. Invest Ophthalmol Vis Sci reous chamber growth rates in marmosets. Invest
show insufficient accommodative response to blur. 2007; 48: 3914–3922. Ophthalmol Vis Sci 2012; 53: 4662.
Invest Ophthalmol Vis Sci 1993; 34: 690–694. 139. Smith EL 3rd, Huang J, Hung LF et al. Hemiretinal 161. Benavente-Perez A, Nour A, Troilo D. Asymmetries in
116. Young FA. The effect of atropine on the development form deprivation: evidence for local control of eye peripheral refraction in marmosets change with
of myopia in monkeys. Am J Optom Arch Am Acad growth and refractive development in infant emmetropization and induced eye growth. Invest
Optom 1965; 42: 439–449. monkeys. Invest Ophthalmol Vis Sci 2009; 50: Ophthalmol Vis Sci 2014; 55: 2731.
117. Crawford K, Terasawa E, Kaufman PL. Reproducible 5057–5069. 162. Benavente-Perez A, Nour A, Troilo D. The role of
stimulation of ciliary muscle contraction in the 140. Smith EL 3rd, Kee CS, Ramamirtham R et al. Periph- peripheral refraction in emmetropization: evidence
cynomolgus monkey via a permanent indwelling eral vision can influence eye growth and refractive from marmoset studies. Eye Vision 2016; 3: 1–11.
midbrain electrode. Brain Res 1989; 503: 265–272. development in infant monkeys. Invest Ophthalmol 163. Benavente-Perez A, Nour A, Yan L et al. The role of
118. Schaeffel F, Troilo D, Wallman J et al. Developing Vis Sci 2005; 46: 3965–3972. peripheral refraction in the temporal integration of
eyes that lack accommodation grow to compensate 141. Raviola E, Wiesel TN. An animal model of myopia. N induced eye growth in marmosets. Ophthalmic Phy-
for imposed defocus. Vis Neurosci 1990; 4: 177–183. Engl J Med 1985; 312: 1609–1615. siol Opt 2013; 33: 661–674.
119. Raviola E, Wiesel TN. Neural control of eye growth 142. Norton TT, Essinger JA, McBrien NA. Lid-suture myo- 164. Atchison DA, Jones CE, Schmid KL et al. Eye shape in
and experimental myopia in primates. Ciba Found pia in tree shrews with retinal ganglion cell blockade. emmetropia and myopia. Invest Ophthalmol Vis Sci
Symp 1990; 155: 22–38. Vis Neurosci 1994; 11: 143–153. 2004; 45: 3380–3386.
120. McBrien NA, Moghaddam HO, Reeder AP. Atropine 143. Wildsoet CF. Neural pathways subserving negative 165. Stone RA, Flitcroft DI. Ocular shape and myopia. Ann
reduces experimental myopia and eye enlargement lens-induced emmetropization in chicks-insights Acad Med Singapore 2004; 33: 7–15.
via a nonaccommodative mechanism. Invest from selective lesions of the optic nerve and ciliary 166. Verkicharla PK, Mathur A, Mallen EAH et al. Eye
Ophthalmol Vis Sci 1993; 34: 205–215. nerve. Curr Eye Res 2003; 27: 371–385. shape and retinal shape, and their relation to
121. Diether S, Schaeffel F. Local changes in eye growth 144. Troilo D, Gottlieb MD, Wallman J. Visual deprivation peripheral refraction. Ophthalmic Physiol Opt 2012;
induced by imposed local refractive error despite causes myopia in chicks with optic nerve section. 32: 184–199.
active accommodation. Vision Res 1997; 37: 659–668. Curr Eye Res 1987; 6: 993–999. 167. Atchison DA, Pritchard N, Schmid KL et al. Shape of
122. Charman WN. Near vision, lags of accommodation 145. McFadden SA. Partial occlusion produces local form the retinal surface in emmetropia and myopia. Invest
and myopia. Ophthalmic Physiol Opt 1999; 19: deprivation myopia in the guinea pig eye. Invest Ophthalmol Vis Sci 2005; 46: 2698–2707.
126–133. Ophthalmol Vis Sci 2002; 43: 189. 168. Gilmartin B, Nagra M, Logan NS. Shape of the poste-
123. Shaikh AW, Siegwart JT Jr, Norton TT. Effect of inter- 146. Morgan IG, Ambadeniya MP. Imposed peripheral rior vitreous chamber in human emmetropia and
rupted lens wear on compensation for a minus lens myopic defocus can prevent the development of myopia. Invest Ophthalmol Vis Sci 2013; 54:
in tree shrews. Optom Vis Sci 1999; 76: 308–315. lens–induced myopia. Invest Ophthalmol Vis Sci 2006; 7240–7251.
124. Kee CS, Hung LF, Qiao-Grider Y et al. Temporal con- 47: 3328. 169. Schmid GF. Axial and peripheral eye length mea-
straints on experimental emmetropization in infant 147. Liu Y, Wildsoet C. The effect of two-zone concentric sured with optical low coherence interferometry.
monkeys. Invest Ophthalmol Vis Sci 2007; 48: bifocal spectacle lenses on refractive error develop- J Biomed Opt 2003; 8: 655–662.
957–962. ment and eye growth in young chicks. Invest 170. Schmid GF. Variability of retinal steepness at the
125. Troilo D, Quinn N, Baker K. Accommodation and Ophthalmol Vis Sci 2011; 52: 1078–1086. posterior pole in children 7–15 years of age. Curr Eye
induced myopia in marmosets. Vision Res 2007; 47: 148. Benavente-Pérez A, Nour A, Troilo D. Axial eye Res 2003; 27: 61–68.
1228–1244. growth and refractive error development can be 171. Atchison DA, Charman WN. Can partial coherence
126. Aleman A, Schaeffel F. Lag of accommodation does modified by exposing the peripheral retina to rela- interferometry be used to determine retinal shape?
not predict changes in eye growth in chickens. Vision tive myopic or hyperopic defocus. Invest Ophthalmol Optom Vis Sci 2011; 88: 601–607.
Res 2018; 149: 77–85. Vis Sci 2014; 55: 6765–6773. 172. Risk factors for idiopathic rhegmatogenous retinal
127. Diether S, Wildsoet CF. Stimulus requirements for 149. Smith EL 3rd, Hung LF, Huang J et al. Effects of opti- detachment. The Eye Disease Case-Control Study
the decoding of myopic and hyperopic defocus cal defocus on refractive development in monkeys: Group. Am J Epidemiol 1993; 137: 749–757.
under single and competing defocus conditions in evidence for local, regionally selective mechanisms. 173. Mallen EAH, Kashyap P. Technical note: measure-
the chicken. Invest Ophthalmol Vis Sci 2005; 46: Invest Ophthalmol Vis Sci 2010; 51: 3864–3873. ment of retinal contour and supine axial length
2242–2252. 150. Liu Y, Wildsoet C. The effective add inherent in using the Zeiss IOLMaster. Ophthalmic Physiol Opt
128. Angle J, Wissmann DA. The epidemiology of myopia. 2-zone negative lenses inhibits eye growth in myopic 2007; 27: 404–411.
Am J Epidemiol 1980; 111: 220–228. young chicks. Invest Ophthalmol Vis Sci 2012; 53: 174. Charman WN, Radhakrishnan H. Peripheral refrac-
129. Curtin BJ. The Myopias: Basic Science and Clinical Man- 5085–5093. tion and the development of refractive error: a
agement, 1st ed. Philadelphia, PA: Harper & Row, 151. Smith EL 3rd, Hung LF, Huang J. Relative peripheral review. Ophthalmic Physiol Opt 2010; 30: 321–338.
1985. hyperopic defocus alters central refractive develop- 175. Schmid GF. Association between retinal steepness
130. Gwiazda J, Thorn F, Held R. Accommodation, accom- ment in infant monkeys. Vision Res 2009; 49: and central myopic shift in children. Optom Vis Sci
modative convergence, and response AC/A ratios 2386–2392. 2011; 88: 684–690.
66
176. Smith EL 3rd, Hung LF, Huang J et al. Effects of local 195. Brunette I, Bueno JM, Parent M et al. Monochromatic 215. Yamamoto S, Nitta K, Kamiyama M. Cone electroreti-
myopic defocus on refractive development in mon- aberrations as a function of age, from childhood to nogram to chromatic stimuli in myopic eyes. Vision
keys. Optom Vis Sci 2013; 90: 1176–1186. advanced age. Invest Ophthalmol Vis Sci 2003; 44: Res 1997; 37: 2157–2159.
177. Totonelly K, Troilo D. Eye shape and off-axis refrac- 5438–5446. 216. Kawabata H, Murayama K, Adachi-Usami E. Sensitiv-
tive state following induced axial refractive errors in 196. Artal P, Guirao A, Berrio E et al. Compensation of ity loss in short wavelength sensitive cones in myo-
marmoset monkeys. Invest Ophthalmol Vis Sci 2008; corneal aberrations by the internal optics in the pic eyes. Nippon Ganka Gakkai Zasshi 1996; 100:
49: 3589. human eye. J Vis 2001; 1: 1–8. 868–876.
178. Verkicharla PK, Suheimat M, Schmid KL et al. Differ- 197. Kisilak ML, Campbell MC, Hunter JJ et al. Aberrations 217. Gamlin PD, McDougal DH, Pokorny J et al. Human
ences in retinal shape between East Asian and Cau- of chick eyes during normal growth and lens induc- and macaque pupil responses driven by
casian eyes. Ophthalmic Physiol Opt 2017; 37: tion of myopia. J Comp Physiol A Neuroethol Sens Neu- melanopsin-containing retinal ganglion cells. Vision
275–283. ral Behav Physiol 2006; 192: 845–855. Res 2007; 47: 946–954.
179. Verkicharla PK, Suheimat M, Schmid KL et al. Periph- 198. Ramamirtham R, Kee CS, Hung LF et al. Wave aber- 218. Jiang L, Zhang S, Schaeffel F et al. Interactions of
eral refraction, peripheral eye length and retinal rations in rhesus monkeys with vision-induced ame- chromatic and lens-induced defocus during visual
shape in myopia. Optom Vis Sci 2016; 93: 1072–1078. tropias. Vision Res 2007; 47: 2751–2766. control of eye growth in guinea pigs (Cavia porcellus).
180. Wakazono T, Yamashiro K, Miyake M et al. Time- 199. Llorente L, Barbero S, Cano D et al. Myopic versus Vision Res 2014; 94: 24–32.
course change in eye shape and development of hyperopic eyes: axial length, corneal shape and opti- 219. Foulds WS, Barathi VA, Luu CD. Progressive myopia
staphyloma in highly myopic eyes. Invest Ophthalmol cal aberrations. J Vis 2004; 4: 288–298. or hyperopia can be induced in chicks and reversed
Vis Sci 2018; 59: 5455–5461. 200. Thibos LN, Bradley A, Hong X. A statistical model of by manipulation of the chromaticity of ambient light.
181. Huang J, Wen D, Wang Q et al. Efficacy comparison the aberration structure of normal, well-corrected Invest Ophthalmol Vis Sci 2013; 54: 8004–8012.
of 16 interventions for myopia control in children: a eyes. Ophthalmic Physiol Opt 2002; 22: 427–433. 220. Kröger RH, Wagner HJ. The eye of the blue acara
network meta-analysis. Ophthalmology 2016; 123: 201. Castejón-Mochón JF, López-Gil N, Benito A et al. Ocu- (Aequidens pulcher, Cichlidae) grows to compensate
697–708. lar wave-front aberration statistics in a normal for defocus due to chromatic aberration. J Comp Phy-
182. Sng CC, Lin XY, Gazzard G et al. Peripheral refraction young population. Vision Res 2002; 42: 1611–1617. siol A 1996; 179: 837–842.
and refractive error in Singapore Chinese children. 202. He JC, Sun P, Held R et al. Wavefront aberrations in 221. Torii H, Kurihara T, Seko Y et al. Violet light exposure
Invest Ophthalmol Vis Sci 2011; 52: 1181–1190. eyes of emmetropic and moderately myopic school can be a preventive strategy against myopia pro-
183. Sng CC, Lin XY, Gazzard G et al. Change in peripheral children and young adults. Vision Res 2002; 42: gression. EBioMedicine 2017; 15: 210–219.
refraction over time in Singapore Chinese children. 1063–1070. 222. Rucker FJ, Osorio D. The effects of longitudinal chro-
Invest Ophthalmol Vis Sci 2011; 52: 7880–7887. 203. Paquin MP, Hamam H, Simonet P. Objective mea- matic aberration and a shift in the peak of the
184. Mutti DO, Sinnott LT, Mitchell GL et al. Relative surement of optical aberrations in myopic eyes. middle-wavelength sensitive cone fundamental on
peripheral refractive error and the risk of onset and Optom Vis Sci 2002; 79: 285–291. cone contrast. Vision Res 2008; 48: 1929–1939.
progression of myopia in children. Invest Ophthalmol 204. Atchison DA, Schmid KL, Pritchard N. Neural and 223. Hung LF, Arumugam B, She Z et al. Narrow-band,
Vis Sci 2011; 52: 199–205. optical limits to visual performance in myopia. Vision long-wavelength lighting promotes hyperopia and
185. Lee TT, Cho P. Relative peripheral refraction in chil- Res 2006; 46: 3707–3722. retards vision-induced myopia in infant rhesus mon-
dren: twelve-month changes in eyes with different 205. Carkeet A, Luo HD, Tong L et al. Refractive error and keys. Exp Eye Res 2018; 176: 147–160.
ametropias. Ophthalmic Physiol Opt 2013; 33: monochromatic aberrations in Singaporean children. 224. Smith EL 3rd, Hung LF, Arumugam B et al. Effects of
283–293. Vision Res 2002; 42: 1809–1824. long-wavelength lighting on refractive development
186. Atchison DA, Li SM, Li H et al. Relative peripheral 206. Cheng X, Bradley A, Hong X et al. Relationship in infant rhesus monkeys. Invest Ophthalmol Vis Sci
hyperopia does not predict development and pro- between refractive error and monochromatic aber- 2015; 56: 6490–6500.
gression of myopia in children. Invest Ophthalmol Vis rations of the eye. Optom Vis Sci 2003; 80: 43–49. 225. Gawne TJ, Ward AH, Norton TT. Long-wavelength
Sci 2015; 56: 6162–6170. 207. Lopez-Gil N, Iglesias I, Artal P. Retinal image quality (red) light produces hyperopia in juvenile and ado-
187. Atchison DA, Rosén R. The possible role of periph- in the human eye as a function of the accommoda- lescent tree shrews. Vision Res 2017; 140: 55–65.
eral refraction in development of myopia. Optom Vis tion. Vision Res 1998; 38: 2897–2907. 226. Gawne TJ, Siegwart JT Jr, Ward AH et al. The wave-
Sci 2016; 93: 1042–1044. 208. Vilupuru AS, Roorda A, Glasser A. Spatially variant length composition and temporal modulation of
188. Rotolo M, Montani G, Martin R. Myopia onset and changes in lens power during ocular accommoda- ambient lighting strongly affect refractive develop-
role of peripheral refraction. Clin Optom (Auckl) 2017; tion in a rhesus monkey eye. J Vis 2004; 4: 299–309. ment in young tree shrews. Exp Eye Res 2017; 155:
9: 105–111. 209. Liu R, Qian YF, He JC et al. Effects of different mono- 75–84.
189. Mutti DO, Sinnott LT, Reuter KS et al. Peripheral chromatic lights on refractive development and eye 227. Cohen Y, Belkin M, Yehezkel O et al. Dependency
refraction and eye lengths in myopic children in the growth in Guinea pigs. Exp Eye Res 2011; 92: 447–453. between light intensity and refractive development
bifocal lenses in nearsighted kids (BLINK) study. 210. Rucker FJ, Wallman J. Chick eyes compensate for under light-dark cycles. Exp Eye Res 2011; 92:
Transl Vis Sci Technol 2019; 8: 17. chromatic simulations of hyperopic and myopic 40–46.
190. Charman WN. Wavefront aberration of the eye: a defocus: evidence that the eye uses longitudinal 228. Chen S, Zhi Z, Ruan Q et al. Bright light suppresses
review. Optom Vis Sci 1991; 68: 574–583. chromatic aberration to guide eye-growth. Vision Res form-deprivation myopia development with activa-
191. Charman WN. Aberrations and myopia. Ophthalmic 2009; 49: 1775–1783. tion of dopamine D1 receptor signaling in the ON
Physiol Opt 2005; 25: 285–301. 211. Rucker FJ. The role of luminance and chromatic cues pathway in retina. Invest Ophthalmol Vis Sci 2017; 58:
192. García de la Cera E, Rodríguez G, Marcos S. Longitu- in emmetropisation. Ophthalmic Physiol Opt 2013; 2306–2316.
dinal changes of optical aberrations in normal and 33: 196–214. 229. Xiong S, Sankaridurg P, Naduvilath T et al. Time
form-deprived myopic chick eyes. Vision Res 2006; 212. Seidemann A, Schaeffel F. Effects of longitudinal spent in outdoor activities in relation to myopia pre-
46: 579–589. chromatic aberration on accommodation and vention and control: a meta-analysis and systematic
193. Coletta NJ, Marcos S, Troilo D. Ocular wavefront emmetropization. Vision Res 2002; 42: 2409–2417. review. Acta Ophthalmol 2017; 95: 551–566.
aberrations in the common marmoset Callithrix 213. Gisbert S, Schaeffel F. M to L cone ratios determine 230. Ostrin LA. The ipRGC-driven pupil response with light
jacchus: effects of age and refractive error. Vision Res eye sizes and baseline refractions in chickens. Exp exposure and refractive error in children. Ophthalmic
2010; 50: 2515–2529. Eye Res 2018; 172: 104–111. Physiol Opt 2018; 38: 503–515.
194. Ramamirtham R, Kee CS, Hung LF et al. Monochro- 214. Roorda A, Williams DR. The arrangement of the 231. Ostrin LA. Objectively measured light exposure in
matic ocular wave aberrations in young monkeys. three cone classes in the living human eye. Nature emmetropic and myopic adults. Optom Vis Sci 2017;
Vision Res 2006; 46: 3616–3633. 1999; 397: 520–522. 94: 229–238.
67
INVITED REVIEW
Higher order aberrations, refractive error development and

myopia control: a review
Rohan PJ Hughes MOptom BVisSc Evidence from animal and human studies suggests that ocular growth is influenced by
Stephen J Vincent PhD BAppSc (Optom) visual experience. Reduced retinal image quality and imposed optical defocus result in pre-
(Hons) dictable changes in axial eye growth. Higher order aberrations are optical imperfections of
Scott A Read PhD BAppSc (Optom) the eye that alter retinal image quality despite optimal correction of spherical defocus and
(Hons) astigmatism. Since higher order aberrations reduce retinal image quality and produce varia-
Michael J Collins PhD MAppSc DipAppSc tions in optical vergence across the entrance pupil of the eye, they may provide optical sig-
(Optom) nals that contribute to the regulation and modulation of eye growth and refractive error
Contact Lens and Visual Optics Laboratory, School of development. The magnitude and type of higher order aberrations vary with age, refractive
Optometry and Vision Science, Queensland University error, and during near work and accommodation. Furthermore, distinctive changes in
of Technology, Brisbane, Australia higher order aberrations occur with various myopia control treatments, including atropine,
E-mail: [email protected] near addition spectacle lenses, orthokeratology and soft multifocal and dual-focus contact
lenses. Several plausible mechanisms have been proposed by which higher order aberra-
tions may influence axial eye growth, the development of refractive error, and the treat-
ment effect of myopia control interventions. Future studies of higher order aberrations,
Submitted: 31 March 2019 particularly during childhood, accommodation, and treatment with myopia control interven-
Revised: 1 July 2019 tions are required to further our understanding of their potential role in refractive error
Accepted for publication: 28 July 2019 development and eye growth.
Key words: eye growth, higher order aberrations, myopia control, refractive error development, visual experience
The prevalence of myopia has dramatically astigmatism with conventional sphero- proportionately to a decrease in the dioptric
risen over the past 60 years1 with significant cylindrical lenses, can significantly influence power of the optical components of the eye,
regional variations in myopia prevalence retinal image quality,16 the accommodation which suggests biological, passive regulation
across the world, from approximately response of the eye,17 and the relative focal of eye growth,19 a process termed
15 per cent of adults in Australia,2 to 70–90 plane of different regions of the entrance emmetropisation.20 Refractive errors are pri-
per cent in South East Asian countries such pupil.18 Therefore, there are various mecha- marily determined by axial length changes21
as China,3 South Korea,4 Singapore,5 and nisms through which they may play a role in that are disproportionate to the change in
Taiwan.6 By 2050, it is estimated that 50 per guiding eye growth and the development of the ocular refractive power, where a slowed
cent of the global population will be myopic refractive errors. This review summarises the and increased rate of axial eye growth
(> −0.50 D), with one-fifth of these being literature examining HOAs in animal models of results in hyperopia and myopia, respec-
highly myopic (> −5.00 D).7 The numerous refractive error development and changes in tively, due to a failure in emmetropisation.22
sight-threatening ocular conditions that are the HOA profile in humans with age, refractive Exposure of the eye to different visual expe-
associated with myopia, including retinal error, abnormal visual development and vari- riences can disrupt emmetropisation, which
detachment,8 myopic maculopathy,9 ous myopia control interventions. Additionally, suggests that the eye also uses visual input
10 11
glaucoma, and cataract, represent a sig- possible mechanisms linking HOAs with refrac- to actively influence eye growth in humans.23
nificant public health concern both in terms tive error development and the treatment A range of animal models have demon-
of the global economy12 and the visual con- effect of myopia control interventions are dis- strated that complete visual obscuration by lid
sequences of these ocular pathologies.13 cussed in detail. suture (in chicks,24 mice,25 rabbits,26 tree
While the aetiology of refractive error is shrews,27 marmosets28 and rhesus monkeys29)
multifactorial,14 evidence from animal stud- or the deprivation of form vision using translu-
ies suggest that visual experience is an Visual regulation of eye growth cent filters (diffusers) (in fish,30 mice,31 guinea
important factor in eye growth regulation.15 pigs32 and rhesus monkeys33) typically results
Higher order aberrations (HOAs), defined as During infancy and childhood, structural in excessive axial elongation and myopia. Simi-
optical aberrations that remain following changes occur within the eye to minimise larly, humans with unilateral visual obstruction
the optimal correction of defocus and refractive error. Axial length increases from congenital ptosis,34,35 cataract,35 corneal
68
Aberrations and refractive error development Hughes, Vincent, Read et al.
opacity36 or vitreous haemorrhage37 also typi- myopigenic stimulus such as imposed hyper- acquired refractive errors prior to treatment
cally develop axial myopia due to form depriva- opic defocus55,59 or form deprivation56,57,59 (Figure 1C).
tion. First reported by Schaeffel et al.38 in the results in significantly greater ocular HOAs The findings of experimentally induced
chick model, imposed defocus also results in associated with the development of significant ametropia in animal models suggest that
predictable bidirectional changes in eye growth ametropia compared to untreated eyes the changes in HOAs associated with refrac-
in a variety of species.39 Exposure to hyperopic (Figure 1A); however, both the treated and tive error development are predominated
defocus leads to an increased ocular growth untreated eyes show a reduction in HOAs over by an increase in the asymmetric aberra-
rate to minimise the imposed refractive error, time (Figure 1B). The increase in ocular HOAs tions of the third radial order.55–57,59 Eyes
while the opposite occurs in response to myo- observed in chicks reared with monocularly with experimentally acquired ametropia
pic defocus, as demonstrated in chicks,38,40 imposed negative lenses55 and diffusers56 are show increased magnitudes of coma and
mice,25 guinea pigs,41 fish,42 tree shrews,43,44 predominantly due to changes in third order trefoil, therefore such asymmetric HOAs
marmosets45,46 and rhesus monkeys.47 RMS (root mean square wavefront error), may provide a signal that influences ocular
Recently, short-term, transient, bidirectional while fourth order55 and spherical aberra- growth, which has been hypothesised based
axial length48 and choroidal thickness48–50 tion56 RMS were minimally affected. Similarly, on longitudinal data from human
changes in response to defocus have also been the magnitude of coma and trefoil RMS (both studies.60–62 Additionally, Wildsoet and
reported in adult humans, but to a much third order terms) increased in monkeys who Schmid63 demonstrated that the chick eye is
smaller degree than in animal models. Insights developed refractive errors from imposed able to modulate ocular growth on the basis
from the chick model have shown that the defocus and form deprivation.59 Coletta of optical vergence, hence it may be possi-
response to imposed defocus occurs rapidly, et al.57 also showed strong interocular correla- ble that the eye uses vergence cues from
within minutes of the visual stimuli being intro- tions for each radial order of HOAs, except these asymmetric HOAs to influence eye
duced.51 Additionally, the sign-dependent third order RMS, in monocularly form- growth. Furthermore, monkey eyes that
responses to imposed defocus appear to be deprived marmosets. developed experimentally induced hyper-
locally mediated,52,53 which indicates that the Following removal of the visual stimuli in opia or myopia, both exhibited an increase
eye can detect odd-error cues for eye growth lens-treated and form-deprived eyes, the in magnitude and inter-subject variability of
within the retinal image. Temporal integration increase in HOAs generally reduced; how- HOAs compared to emmetropic eyes.59
of these cues from the retinal image are ever, the HOAs remained higher in the While it remains possible that an increase in
thought to modulate scleral remodelling and treated eyes than in the fellow untreated HOAs provides a form deprivation-like stim-
axial eye growth.54 eyes.57,59 Additionally, Ramamirtham et al.59 ulus due to a reduction in retinal image
found that some eyes showed no recovery quality, or that individual HOAs produce a
from their experimentally induced ametro- visual signal that promotes or inhibits ocular
Evidence from animal studies pia, and in these eyes, an increase in total growth, the overall trends observed in ani-
ocular HOAs during the recovery phase was mal studies across various species suggest
During normal visual development, chick,55,56 observed, rather than a decrease. Interest- that an increase in HOAs occurs coinciden-
marmoset57 and rhesus monkey58 eyes dis- ingly, there was no difference in the HOA tally with refractive error development.
play a decrease in HOAs over time, similar to profile between the eyes that recovered and The reduction of HOAs55–58 and the time
the reduction in neonatal refractive error. A those that failed to recover from their course of the increase in HOAs during the
A 0.18 B 0.30 Untreated chick eyes C 0.70

Untreated marmoset eyes Untreated rhesus monkey eyes
0.16 Form deprived marmoset eyes Form deprived chick eyes Eyes with no ametropia recovery
0.60
0.25 Eyes with ametropia recovery
0.14
0.50
HOA RMS (µm)
HOA RMS (µm)
0.12 0.20
RMS (µm)
0.10 0.40
0.15
0.08 0.30
0.06 0.10
0.20
0.04
0.05 0.10
0.02
0.00 0.00 0.00

3rd 4th 5th 6th 7th Z(3,-3) 0 2 4 6 8 10 12 14 0 50 100 150 200 250 300
Radial order/Zernike coefficient Age (days) Age (days)
Figure 1. Higher order aberrations (HOAs) associated with animal models of experimental myopia showing A: the greater level of
HOAs during or immediately following form deprivation compared to untreated fellow eyes in marmosets (reproduced from Col-
etta et al.57), B: the change in HOAs in chick eyes during treatment with form deprivation compared to untreated control eyes
(reproduced from Garcia de la Cera et al.56), and C: the change in HOAs in treated rhesus monkey eyes that developed form depri-
vation or lens-induced ametropia compared to an untreated control group (reproduced from Ramamirtham et al.59), where the
three time points represent pre-treatment, immediately post-treatment and following a period of recovery in rhesus monkeys.
Note that the eyes that did not recover from their experimental ametropia showed increased HOAs compared to untreated eyes
and treated eyes that exhibited recovery from induced ametropia. In A, B, and C, asterisks indicate statistically significant group
differences. In A and B, error bars represent the standard deviation, and in C, the standard error of the mean.
69
0.5 sectional study of Chinese children (n = 1,634)

HOA 3rd order 4th order measured HOAs under cycloplegia and
reported a trend of increasing HOA RMS from
0.4 three to 17 years (Figure 3), primarily due to a
negative shift in primary vertical coma (Z 3−1 ),
RMS aberrations (µm)
0.3 primary spherical aberration (Z 04 ), and sec-

ondary trefoil (Z 3− 5 ), and positive shifts in pri-
mary trefoil (Z 3− 3 ) and secondary
0.2 astigmatism (Z 24 ).77
Conversely, Brunette
et al.69 examined Canadian children under
cycloplegia and demonstrated a reduction in
0.1
HOA RMS during childhood and suggested
that HOAs are regulated similarly to lower
0.0 order aberrations (spherical and astigmatic
0 20 40 60 80 refractive errors) during emmetropisation.
Age (years) While this finding is consistent with various
animal models,55–58 their sample included
Figure 2. The change in higher order aberrations (HOAs), third order and fourth order only 29 subjects under the age of 20 years
root mean square wave front error (RMS) with age over a 5 mm pupil (polynomial whereas Zhang et al.77 examined over 1,600
regression functions adapted from Brunette et al.69). Dotted lines represent the 95% subjects in this age group. Significant differ-
confidence intervals for the regression functions. HOA, third order and fourth order ences in the refractive error range of the
RMS varied in an approximate quadratic association with age, decreasing during examined populations of Brunette et al.69
childhood to a minimum between 30–40 years and subsequently increasing with age. and Zhang et al.77 may explain the inconsis-
tency in their results, with −3.50 to +3.50 D
(across all included ages from six to 82 years)
development of ametropia55–57,59 suggests a across a lifetime (six to 82 years), the change and −10.00 to +8.25 D, respectively, although
passive scaling effect due to growth in ocu- in HOA RMS was best described by a second Zhang et al.77 reported no significant differ-
lar structures,55–58 and that any increase in order polynomial, where the elderly (over ences in HOAs between the myopes,
the level of HOAs is likely a consequence, 60 years) displayed greater HOA RMS values emmetropes and hyperopes within each age
rather than a cause, of refractive error than those 20 to 60 years old, with a mini- group. Additionally, Caucasian and Asian
development. However, modelling demon- mum at approximately 40 years (Figure 2). adults,78,79 and Chinese and Malay
strates that simple scaling of the optical Similarly, other studies have found that HOA children,80 have been reported to exhibit
components in chicks55 and rhesus mon- RMS remains stable between the ages of HOA profile differences, particularly for pri-
keys58 cannot account for the total changes approximately 20 and 55 years.70,71 Brunette mary spherical aberration (Z 04 ); therefore,
observed in HOAs, which may provide some et al.69 also showed that coma and spherical ethnic variation may also exist between the
evidence for visual, rather than entirely pas- aberration RMS vary with age in an approxi- HOA profiles of Canadian and Chinese chil-
sive, regulation of HOAs. mate quadratic association, reaching a mini- dren. Given the cross-sectional designs of
mum between 20 to 30 years, and increasing both studies, longitudinal studies are
with older age.71,72 Primary horizontal coma required to further the current understand-
HOAs and age (Z 13 ) and spherical aberration (Z 04 ) have ing of the temporal variations in HOAs dur-
shown negative65 and positive associations ing childhood.
On-axis HOAs with age,71,72 respectively, between approxi- The studies of HOAs and age in adults
Total ocular HOAs are influenced by the mately 20 and 70 years. show consistent trends in HOAs as a func-
refractive elements within the eye, specifi- Cataract development typically causes tion of age; however, a factor which is typi-
cally the curvature, alignment, refractive internal ocular HOAs to increase,73–75 pre- cally neglected is the effect of natural pupil
index, and axial separation of the anterior dominantly positive shifts in coma RMS and size. In adults, it is well-established that
and posterior surfaces of the cornea and primary spherical aberration (Z 04 ) for cortical pupil size decreases with age.81,82 Winn
crystalline lens. In humans, a partial com- and nuclear cataracts, respectively.73,75 et al.81 demonstrated that average pupil size
pensatory balance exists between the ante- Since anterior corneal HOAs exhibit negligi- at 20 years was ~4.5 mm and decreased by
rior corneal and internal HOAs (the ble variation throughout adulthood68,72,76 ~0.02 mm per year to ~3.2 mm at 85 years,
combination of the posterior cornea and and with cataract formation,75 it is likely that under typical indoor room lighting (~263
crystalline lens), whereby the internal HOAs these age-related lenticular changes result lux). Contrary to adults, pupil size in indoor
are of reduced magnitude and opposite in in a breakdown of the partial internal com- room lighting during childhood has been
sign to the anterior corneal HOAs.64 pensation of HOAs and account for the shown to increase from ~5 mm at birth to
Several studies have reported a linear reported changes in HOAs with age,68 par- ~6.1 mm by late adolescence; however, the
increase in HOA RMS between the ages of ticularly over 60 years.69 exact luminance during measurement was
20 and 70 years.65–68 In a cross-sectional The changes in HOAs observed during child- not reported.83 Each of the studies examin-
analysis, Brunette et al.69 demonstrated that hood are not consistent. A recent large cross- ing age-related changes in HOAs report the
70
Pre-school age School age Adolescent

3 0.5
Refractive power (D)

1 0
y (mm)
-1 -0.5
-2
-3 -1
-3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3
x (mm) x (mm) x (mm)
Figure 3. Refractive power maps generated from the higher order aberrations (HOAs) (third to fifth order) in pre-school age chil-
dren, school age children, and adolescents (data from Zhang et al.77). The observed changes between age groups result from posi-
tive shifts in primary trefoil (Z 3−3 ) and secondary astigmatism (Z 24 ), and negative shifts in primary vertical coma (Z 3− 1 ), primary
spherical aberration (Z 04 ) and secondary trefoil (Z 5−3 ), which produce the increase in positive and negative power at the centre and
margin of the 6 mm pupil diameter, respectively.
HOAs over fixed pupil diameters of greater eccentricity interaction for all third and
HOAs and refractive error
than 4.5 mm; however, the HOA profile fourth order Zernike terms, except primary
through the natural pupil may have differed trefoil (Z 3−3 ) and quadrafoil (Z 4− 4 ), which sug-
since pupil size varies with age. Further- Cross-sectional studies
gests a difference between age groups in
more, it is possible that the consistently Numerous cross-sectional studies have
the off-axis variation of these HOAs. How-
observed increase in HOAs reported in older compared HOAs between subjects with
ever, on average, the magnitude of these
adults may be offset by the natural age- established refractive errors; however, the
HOAs across the visual field was reported to
related pupillary miosis, since HOAs decrease results have not been consistent (Table 1).
be minimal except for primary vertical (Z 3− 1 )
with decreasing pupil size.16 Several studies of adults have found that
and horizontal (Z 13 ) coma, and spherical myopic eyes show significantly higher levels
aberration (Z 04 ).85 Most significantly, the of ocular HOA RMS than emmetropic
Off-axis HOAs combination of the coma terms increased eyes,89–91 but others have found no differ-
Off-axis HOAs are typically of greater magni- approximately linearly across the visual ences.92,93 Llorente et al.94 showed that
tude than on-axis HOAs, particularly for coma field, and older eyes exhibited a greater rate hyperopic eyes exhibit greater HOA RMS
terms, likely as a result of the change in align- of change than younger eyes,85 where the than myopic eyes; however, this finding has
ment and shape differences of the ocular orientation of the off-axis variation of the not been duplicated.92 Spherical aberration
refractive surfaces from off-axis incident light combined coma terms aligned with the axis and coma RMS have been reported to
rays.84–87 Changes in off-axis HOAs also occur of the term, as expected due to the change in increase with increasing levels of myopia.90
with age. Emmetropic adolescents (11 to alignment and shape of the cornea and crys- Similarly, most studies have shown that
14 years) show greater levels of off-axis HOA talline lens. For example, vertical coma varied third order, fourth order, and coma-like
RMS compared to on-axis measurements,88 across the vertical meridian, horizontal coma RMS values are higher in myopes than
with a magnitude similar to young adult varied across the horizontal meridian, and the emmetropes and hyperopes,89,91 but this is
emmetropes.85 Primary vertical coma (Z 3− 1 ) combined terms varied along oblique visual also not a universal finding.94 While some
and primary horizontal coma (Z 13 ) also field meridians being measured. Primary studies have shown no trend,92,94 a positive
increase off-axis, while primary spherical spherical aberration (Z 04 ) was stable across correlation between primary spherical aber-
aberration (Z 04 )
remains stable across the the visual field in each group; however, the ration (Z 04 ) and refractive error has been
88
visual field. older subjects displayed more positive observed, whereby spherical aberration
In young (20 to 30 years) and old (50 to values on average.85 Studies of peripheral becomes more negative with increasing
71 years) emmetropes, HOA RMS also varies HOAs in children, in addition to longitudinal myopia.91,93
with eccentricity in an approximate quadratic studies of off-axis HOAs, ocular biometry Cross-sectional studies of off-axis HOAs in
association along the horizontal and vertical and refractive error are required to further young adults have shown that HOA RMS
meridian; however, the rate of change with examine changes in on- and off-axis HOAs increases more rapidly with visual field
eccentricity is greater in older eyes.85 Addi- with age and their potential role in eye eccentricity in myopes than in
tionally, Mathur et al.85 found an age by growth and refractive error development. emmetropes;86 however, Osuagwu et al.87
71
72
Authors Year N Ages N by refractive Sex Location Measurement Accommodation PD HOA 3rd 4th SA Coma
(years) group and range (D) (ethnicity) technique control (mm)
Carkeet 2002 273 7.9–12.7 138 M 36 HM 147 female, Singapore Hartman-Shack 1% Cyclopentolate 5 M=E=H M = E = H HM < LM = E = H HM < LM = E = H M = E = H
et al.80 (≤ −3.00) 126 male (Chinese, Malay, (3 drops) (same for Z04 )
Indian)
102 LM
(−3.00 ≤ −0.5)
123 E (−0.5 ≤ +1.00)

12 H (≥ +1.00)
He et al.89 2002 170 10–17 87 M (≤ −0.75) NR China, America Psychophysical Badal ≥6 M>E NR NR NR NR
ray tracing
83 E (0.50)
146 18–29 92 M (≤ −0.75) M>E M>E M>E NR NR
54 E (0.50)
Paquin 2002 33 18–32 26 M 5 HM (≤ −6.00) NR Canada Hartman-Shack 1% Tropicamide; 5; 9 M > E NR NR " M / " SA "M/"
Clinical and Experimental Optometry 103.1 January 2020

et al.90 2.5% Phenylephrine coma
13 MM
(−6.00 ≤ −3.00)
9 LM
(−3.00 ≤ −1.00)
7 E (≥ −1.00)
Cheng et al.92 2003 162 26.1 5.6 124 M (≤ −0.75) NR America Hartman-Shack 0.5% Cyclopentolate 6 M=E=H NR NR M=E=H NR
(mean SD)
19 E (−0.75 ≤ +0.75)
19 H (≥ +0.75)
Llorente 2004 46 23–40 24 M (≤ −0.8) NR Spain Laser ray 1% Tropicamide 6.5 M<H H>M NR H > M (Z04 ) H>M
et al.94 tracing
22 H (≥ +0.5)
Kirwan 2006 82 (162 4–14 25 M (≤ −0.70) 42 female, Ireland Hartman-Shack 1% Cyclopentolate 6 M>H M=H M>H M=H M>H
et al.95 eyes) 40 male (Z3−1 )
137 H (≥ 0.06)
93 th
Kwan et al. 2009 116 19–29 86 M 30 HM 62 female, Hong Kong Hartman-Shack Internal instrument 5 M=E E=M E > M; " M / # 4 E > M; "M / #SA; E=M
(≤ −5.00) 54 male (Chinese) fixation target "M / # Z04
56 MM
(−5.00 ≤ −0.50)
30 E (≥ −0.50)
Martinez 2009 771 5.7–7.9 53 E (−0.50 ≤ +0.50) 368 female, Australia Hartman-Shack 1% Cyclopentolate; 5 E<H E=H E<H E < H; #H / # Z04 E=H
et al.101 403 male (Caucasian) 1% Tropicamide
718 H (≥ +0.50)
643 11.2–13.9 150 E (−0.50 ≤ +0.50) 318 female,
325 male
493 H (≥ +0.50)
Li et al.97 2012 86 7–13 64 M 21 MM 45 female, China (Chinese) Hartman-Shack 0.5% Tropicamide 5 M=E M=E M=E M=E Z13 varied
(−6.00 ≤ −3.00) 41 male (5 drops); 1.50 D
target fogging
43 LM
(−3.00 ≤ −0.50)
22 E (−0.50 ≤ +0.50)
Table 1. Summary of cross-sectional cohort studies examining on-axis higher order aberrations between refractive error groups
© 2019 Optometry Australia

Authors Year N Ages N by refractive Sex Location Measurement Accommodation PD HOA 3rd 4th SA Coma
(years) group and range (D) (ethnicity) technique control (mm)
Philip et al.99 2012 675 16–19 125 M 25 MM 339 female, Australia Hartman-Shack 1% Cyclopentolate; 5 MM = LM = E < H M = E = H MM = LM = E < H MM = LM = E < H M = E = H
(< −3.00) 336 male 1% Tropicamide (same for Z04 )
100 LM
(−3.00 < −0.50)
© 2019 Optometry Australia

197 E (−0.50 < −0.50)
353 H (≥ +0.50)
Zhang et al.96 2013 148 6–16 99 PM (−4.25 1.58) NR China (Chinese) Hartman-Shack 1% Tropicamide 6 PM > SM PM > SM PM = SM PM = SM PM > SM
(more -ve)
49 SM (−3.79 1.92)
98
Little et al. 2014 317 9–16 33 M (≤ −0.50) 162 female, Northern Ireland Hartman-Shack 1% Cyclopentolate 5 M=E=H M=E=H M=E=H M=E=H M=E=H
156 male
85 E (−0.13 < +0.50)
199 H (≥ +0.50)
Yazar et al.91 2014 1,034 18.3–22.1 217 M (≤ −0.50) 477 female, Australia (85% Hartman-Shack 1% Tropicamide; 10% 6 M > E; " M / " M>E>H M>E>H " M / # Z04 "M/"
530 male Caucasian) Phenylephrine HOA coma
476 E (−0.50 < +0.50)
314 H (≥ +0.50)
Papmastorakis 2015 557 10–15 320 M (≤ −0.50) 266 female, Greece Hartman-Shack Chart fixation (0.25 D 5 NR NR NR " M / # Z04 NR
et al.100 291 male demand)
201 E (−0.50 < +0.50)
36 H (≥ +0.50)
Philip et al.88 2018 618 11.09–13.9 91 M (≤ −0.50) 52 female, Australia Hartman-Shack 1% Cyclopentolate; 5 M=E<H M=E=H M=EM<HE<H M<E<H NR
39 male (Caucasian, 1% Tropicamide
Asian, Middle
166 E (−0.50 < +0.50) 77 female,
Eastern)
89 male
361 H (≥ +0.50) 179 female,
182 male
Coma: coma RMS, E: emmetropia, H: hyperopia, HM: high myopia, HOA: higher order aberrations RMS, LM: low myopia, M: myopia, MM: moderate myopia, N: number of participants in sample/refractive group, NR: not
reported, PD: pupil diameter for analysis, PM: progressive myopia, 3rd: third order RMS, 4th: fourth order RMS, SM: stable myopia, SA: spherical aberration RMS, Z 3−1 : primary vertical coma, Z 04 : primary spherical aberration,
Z 13 : primary horizontal coma, ": increase, #: decrease.
Table 1. Continued
73
found no significant differences between third order and coma RMS than stable aberration (Z 04 ) in young adults,86,87 hyper-
refractive error groups. Consistent with the myopes. While several studies have found no opic eyes showed more positive values than
findings of on-axis HOA studies, myopes difference in HOAs between refractive error emmetropic and myopic eyes at all
exhibit more negative primary spherical groups in children,80,97,98 Philip et al.99 observed eccentricities.88
aberration (Z04 ) than emmetropes,86 and reported that low hyperopes and emmetropes The lack of consistency concerning on-axis
hyperopes display more positive primary exhibit increased HOAs compared to HOA profiles between refractive error groups
spherical aberration (Z04 ) than emmetropes emmetropes and low myopes in a group of demonstrates the potential variability in the
and myopes on average across the visual older adolescents aged 16 to 19 years. measurement of HOAs across individuals, dif-
field.87 Coma varies with visual field eccen- Hyperopic adolescents exhibit more posi- ferent ethnicities and ages. Cross-sectional
tricity, with primary vertical coma (Z3− 1 ) tive primary spherical aberration (Z 04 )99 and cohort comparison studies do not control for
increasing from the superior to inferior field greater fourth order and spherical aberra- this individual variation and therefore longitudi-
tion RMS than myopic and emmetropic ado- nal assessments of HOAs associated with
and primary horizontal coma (Z13 ) increasing
lescents.88 Additionally, and in agreement changes in refractive error during childhood
from the nasal to temporal field.86,87 While
with the findings of adult studies, primary (repeated measures of the same children over
Mathur et al.86 reported that the rate of off-
axis change in coma is double in myopes spherical aberration (Z 04 ) tends to become time) may provide further insights into the rela-
more negative with increasing myopia100 or tionship between HOAs and refractive error.
than in emmetropes, this finding was not
confirmed by Osuagwu et al.87 across the decreasing hyperopia;101 however, not all
studies agree.80,98 Myopic and hyperopic Longitudinal studies
same visual field range. This may be due to
adolescent eyes also exhibit more positive Few studies have longitudinally examined
study differences in the level of myopia
and negative levels of secondary spherical HOAs and refraction (Table 2). Philip et al.88,102
between the two cohorts, since axial length
aberration (Z 06 ), respectively,99 inversely tracked refractive and HOA changes over
and corneal and retinal shape are
associated with primary spherical aberration approximately five years in Australian adoles-
influenced by refractive error and may
cents of mixed ethnicity. Emmetropic subjects
affect the rates of change of off-axis HOAs. (Z 04 ). This inverse association, where nega-
who underwent a myopic shift of at least
While each of these studies measured tive secondary spherical aberration (Z 06 ) and 0.50 D during the study, exhibited a reduction
HOAs either under cycloplegia, or using a positive primary spherical aberration (Z 04 ) in third order and coma RMS,102 while an
fixed distance target or the internal fixation exist, produces greater relative positive increase was reported in subjects with stable
target of the instrument (presumably refractive power in the periphery of the refractions.88 The emmetropes,102 myopes
focused for relaxed accommodation), vari- pupil and vice versa, which suggests that and hyperopes88 who underwent a myopic
ous aberration measurement techniques myopic and hyperopic eyes experience shift also showed a significant negative shift in
and instruments have been utilised which more relative negative and positive refractive primary spherical aberration (Z 04 ) and a reduc-
may account for the broad inconsistencies power in the periphery of the pupil, respec- tion in fourth order and spherical aberration
between these studies. Given that myopia tively (Figure 4). Some researchers have found RMS, while the opposite was found in sub-
typically develops during childhood and that myopic eyes exhibit higher RMS values jects with stable refractions.88,102 Addition-
adolescence, it is difficult to draw conclu-
for vertical coma (Z 3−1 ), horizontal coma (Z 13 ), ally, a moderate, statistically significant
sions from these cross-sectional cohort
and third order aberrations than relationship (r = 0.49, p < 0.001) was
studies of adult subjects with established
emmetropic89 and hyperopic95 eyes; how- observed between the changes in spherical
refractive errors and it is therefore also
ever, the majority of studies report minimal equivalent refraction and primary spherical
valuable to examine the association
differences between refractive error aberration (Z 04 ), after adjusting for age, gen-
between refractive error and HOAs in chil-
groups.80,98,99,101 Interestingly, Zhang et al.96 der and ethnicity, whereby a myopic shift
dren and adolescents.
reported that myopes with faster progres- was associated with a shift toward negative
Fewer cross-sectional studies have exam-
sion rates exhibited more negative primary primary spherical aberration (Z 04 ).102 Philip
ined HOAs in children (mostly under
vertical coma (Z 3− 1 ) than stable myopes. et al.88 also reported small changes in off-
cycloplegia), and like studies of adults, there
Differences in off-axis HOAs have been axis primary horizontal coma (Z 13 ) in the
is disagreement regarding the relationship
reported between refractive error groups in nasal and temporal fields of myopes,
between HOAs and refractive error. Kirwan
adolescents (11 to 14 years).88 Myopic eyes
et al.95 examined children aged four to emmetropes and hyperopes who under-
tend to display greater levels of HOA, third went a myopic shift, becoming more nega-
14 years and found that myopic children
order and coma RMS than hyperopic eyes in tive and positive, respectively. This
exhibited greater HOA RMS than hyperopes.
the temporal visual field, while hyperopes
He et al.89 similarly found higher levels of corresponded with increases in third order,
exhibit higher amounts of fourth order and coma and HOA RMS in both horizontal
HOA RMS in myopes than emmetropes in
spherical aberration RMS than myopes. peripheral locations in these subjects.
children aged 10 to 17 years, measured with
Philip et al.88 also found that primary verti- Lau et al.61 reported higher spherical
a natural pupil and relaxed accommodation.
Further supporting a role for HOA in myopia cal coma (Z 3−1 ) was more negative in myopic aberration and HOA RMS values in Hong
development, Zhang et al.96 examined eyes than hyperopic eyes in the temporal Kong children who underwent slower axial
myopes aged between six and 16 years and field, while myopes displayed more positive eye growth, after controlling for factors
observed that those with a higher rate of primary horizontal coma (Z 13 ) than hyper- known to affect axial elongation such as
progression (greater than 0.50 D per year) opes in the inferior visual field. Like the find- age, gender, and baseline refractive error.
exhibited significantly higher levels of HOA, ings for off-axis primary spherical Reduced axial elongation was also
74
Positive Z40 Negative Z60 Positive Z40/Negative Z60

3
1
y (mm)
0
0.4
-1

-2 0.2
-3
Negative Z40 Positive Z60 Negative Z40/Positive Z60 0
3
2 -0.2
1
-0.4
y (mm)
-1
-2
-3
-3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3
Figure 4. Refractive power maps generated for a 6 mm pupil demonstrating the oppositely signed combinations of 0.08 μm pri-
mary spherical aberration (Z 04 ) and 0.02 μm secondary spherical aberration (Z 06 ). Note that the combination of positive primary
spherical aberration (Z 04 ) and negative secondary spherical aberration (Z 06 ) results in more relative positive refractive power
toward the pupil margin, whereas the opposite occurs for the counter-scenario.
associated with less positive oblique trefoil spherical aberration (Z04 ). Primary vertical et al.18 reported greater dispersion of the
(Z 33 ), more positive primary trefoil (Z 3−3 ) and coma (Z3−1 ) and spherical aberration (Z04 ) point spread function, and a decreased mod-
more positive spherical aberration with (Z 04 ), ulation transfer function and Visual Strehl
exhibited a positive and negative correlation
ratio in progressing myopes compared with
each 0.1 μm increment of each term associ- with the change in refractive error and axial
emmetropes, at both far and near distances.
ated with ~0.13, 0.11 and 0.11 mm differ- length, respectively, while the opposite
Conversely, a cross-sectional study of chil-
ence in axial eye growth per year, trends were observed for corneal primary
dren (nine to 10 years) and adolescents
respectively.61 Interestingly, given the partial horizontal coma (Z13 ). Both ocular coma (15 to 16 years) found minimal differences in
compensatory effect of anterior corneal terms followed the same correlations as the the Visual Strehl ratio between myopes,
HOAs by internal HOAs,68 Hiraoka et al.103 corneal coma terms, but the positive corre- hyperopes and emmetropes; however, sub-
found that myopia progression and axial lation between ocular primary spherical ject numbers varied considerably between
elongation correlate independently with
aberration (Z04 ) and refractive change was the refractive groups.98 In a cohort of
many corneal HOAs, and more strongly than
not significant. emmetropic adolescents (16 to 19 years),
ocular HOAs in Japanese children. Corneal
Greater levels of HOAs and reduced reti- Philip et al.102 showed that the Visual Strehl
HOAs exhibited strong positive and negative
nal image quality in myopic eyes is not a ratio reduced significantly during the five-
correlations with refractive error shift and
universal finding. McLellan et al.104 showed year study period, and this reduction was
change in axial length, respectively, indicat-
that HOAs measured in myopic adults larger in subjects who became myopic; how-
ing that increased corneal HOAs (baseline
(mean age 41 years) consistently degraded ever, there was no difference in the Visual
measurement or averaged across the study) Strehl ratio between refractive groups at the
the modulation transfer function less than
were associated with reduced myopia pro- initial visit. The inconsistent findings of these
randomly generated HOA profiles, which
gression and axial elongation. The strongest suggests that HOA terms are likely to be studies suggest that any differences in retinal
correlations for individual corneal HOA interdependent and interact to minimise the image quality observed between refractive
terms were observed for primary vertical overall effect on image quality in myopic error groups may be related to individual
coma (Z3−1 ), horizontal coma (Z13 ) and eyes. In young adults (19 to 28 years), Collins variability or methodological differences
75
rather than a cause of refractive error devel-
Table 2. Summary of longitudinal studies examining temporal associations between on-axis higher order aberrations, myopia progression (MP) and axial elonga-
opment, therefore further longitudinal stud-
Coma: coma RMS, E: emmetropia, H: hyperopia, HOA: higher order aberrations RMS, M: myopia, NR: not reported, PD: pupil diameter used for analysis, SA: spherical
" corneal HOA
# HOA; # SA; # Z3−3 " HOA; " SA; "

0
(+ve); # Z04 (+ve); # Z3 (+ve); " Z4
" corneal and
# corneal and
" corneal and

ies are required.
ocular Z3− 1
Cross-sectional cohort and longitudinal
Z3−3 (−ve)
ocular Z13
ocular Z04
(+ve); "
studies have demonstrated that HOAs such
aberration RMS, Z 04 : primary spherical aberration, Z 3−1 : primary vertical coma, Z 13 : primary horizontal coma, +ve: positive, −ve: negative, ": increase, #: decrease.
# AE
as coma (Z 3− 1 and Z 13 ), vertical trefoil (Z 3−3 )
−3
NR
NR
and primary spherical aberration (Z 04 ) show
relatively consistent trends between refrac-
# corneal HOA; #
tive error groups, and temporal variations

ocular Z3−1 ; "
corneal and
corneal and
corneal and
ocular Z13 ; #
ocular Z04 of these terms are associated with changes
Z33 (−ve)
in refraction and axial length, respectively. This
indicates that the composition of the HOA pro-
" AE
NR
NR
file (the combination or interaction of the
terms) may play a more significant role in the
" 4th order (E, H);
" initial 3rd order
" corneal HOA; "
modulation of eye growth and refractive error

" initial coma
development rather than the magnitude of

ocular Z3−1 ; #
corneal and
corneal and
ocular Z13 ; "
" Z3−1 (−ve)
" 4th order
corneal Z04
" SA (E, H)
individual Zernike term co-efficients. These
" Z04 (+ve)
HOAs may provide image cues to the retina to

# MP
" SA
enable the eye to rapidly and accurately

NR
respond to different visual stimuli. Wilson

# initial coma; # Z04
et al.105 reported that only even, not odd,

# initial 3rd order;
(+ve); " 3rd order;
order (M, E, H); #

# corneal HOA; #
" 3rd order (E); "
radial order HOA terms (such as fourth and

coma (E); # 4th
# 4th order; "
ocular Z3−1 ; "

corneal and
corneal and
sixth order), provide odd-error cues within the

SA (M, E, H)
ocular Z13 ; #
coma; # SA
corneal Z04
retinal image which may enable the correct

identification of the sign of defocus.
" MP
NR
Additionally, the accommodative response

of the eye appears to be driven by the detec-
(mm)
tion of optical vergence to guide direction and

5
5
Duration Measurement Accommodation PD
magnitude;106 therefore, it is likely that the ret-

ina can detect relative optical vergence varia-
1% Tropicamide
1% Tropicamide
1% Tropicamide
Cyclopentolate;
Cyclopentolate;
Cyclopentolate;
Cyclopentolate
tions across the pupil. For example, in an eye

with no HOAs, light rays from the centre and
(3 drops)
the periphery of the pupil would focus per-

control
fectly at the same retinal location; however, in

Hartman-Shack 1%
Hartman-Shack 1%
Hartman-Shack 1%
Hartman-Shack 1%
the same eye with the addition of positive or

negative primary spherical aberration (Z 04 ), the
peripheral light rays would be relatively con-
Placido disc
technique
vergent or divergent to the central light ray,

(corneal)
(ocular);
respectively. Horizontal and vertical meridio-

nal retinal shape are non-identical107 and
therefore, asymmetric terms such as coma
or trefoil may also produce relatively con-
(years)
vergent or divergent light rays through dif-

5
ferent pupil positions which may be

detectable by the retina.
11.24–13.77
6–12
6.1–12.6
11.09–13.9
Initial age
(years)
HOAs and abnormal visual

development: anisometropia
and amblyopia
2018 113 M;
166 E;
361 H
2018 91 M;
2014 176 E
Hiraoka 2017 64 M
24 H
Authors Year N
Non-amblyopic anisometropia
Given that the fellow eyes of an individual
typically display a high degree of interocular
et al.103
et al.102
symmetry for both spherical and astigmatic

et al.61
et al.88
tion (AE)
Philip
Philip
refractive error (isometropia),108 non-

Lau
amblyopic anisometropia is a unique ocular

condition in which the two eyes experience
76
a similar environment but develop markedly amblyopic eye of strabismic amblyopes, negative.124–126,129–132 Secondary spherical
different refractive errors in the absence of which were typically more hyperopic, aberration (Z 06 ) also undergoes a relatively
ocular pathology or an amblyogenic factor, exhibited a greater amount of trefoil (Z 33 ) small change,124,132 but is less consistent
typically due to asymmetric axial eye than the fellow non-amblyopic eyes. A weak with respect to the direction of change. Con-
growth.109 Interestingly, the majority of correlation was observed between the inter- sistent with HOA differences between
HOAs are highly correlated between the ocular difference in primary spherical aber- myopes and emmetropes, myopes exhibit
eyes of both isometropes16,70,89,110,111 and ration (Z 04 ) and the magnitudes of less positive, or more negative, fourth order
anisometropes.110,112,113 Tian et al.112 found anisometropia and amblyopia, where the aberrations than emmetropes, and show
that the more myopic eye of non-amblyopic more hyperopic, or more amblyopic eye, larger fourth order aberration changes than
myopic anisometropes exhibited more posi- had more positive primary spherical aberra- emmetropes during accommodation.131
tive primary spherical aberration (Z 04 ) than tion (Z 04 ). The latter finding supports the typi- In a study of off-axis HOAs during accom-
the fellow, less myopic eye, and suggested cal trend of primary spherical aberration modation, a significant interaction was
that this may simply be a consequence of reported between accommodation and
(Z 04 ) to be less positive (or more negative)
the eye being more myopic, rather than an eccentricity for all third and fourth order
with increasing levels of myopia. However,
underlying cause of excessive eye growth.
these cross-sectional contralateral studies aberrations except primary trefoil (Z 3−3 ), sec-
Osuagwu et al.111 examined off-axis HOAs ondary astigmatism (Z 4−2 ) and primary
do not provide convincing evidence that
and conversely found that the less myopic
HOAs underpin the development of refrac- spherical aberration (Z 04 ).84 This suggests
eye of non-amblyopic myopic
tive error, anisometropia or amblyopia and that accommodation produces a change in
anisometropes exhibited more positive pri-
suggest that additional factors are involved. the variation of these HOAs across the
mary spherical aberration (Z 04 ) on average visual field; however, Mathur et al.84
across the visual field; however, there was reported these off-axis variations were of
negligible interocular difference in the rate HOAs during near work and minimal magnitude except for primary hori-
of change with increasing eccentricity. Pri- accommodation zontal coma (Z 13 ) and spherical aberration
mary vertical coma (Z 3−1 ) was found to
(Z 04 ) which became more positive and nega-
increase more rapidly from the superior to Near work has long been considered an tive with accommodation, respectively, aver-
inferior visual field in the more myopic eye; environmental risk factor for myopia devel- aged across the visual field.
however, the rate of change for primary opment;119,120 however, this association
The on-axis HOA profile also varies during
horizontal coma (Z 13 ) across the horizontal remains contested.121,122 Changes in the
downgaze, with most of the change arising
meridian exhibited no significant interocular magnitude of HOAs have been reported to
from a negative shift in primary trefoil (Z 3−3 )
difference.111 These cross-sectional findings occur during near work and accommoda-
and positive shifts in secondary spherical
do not provide clear and consistent evi- tion, which provides a potential mechanism
dence of a solitary role for on- or off-axis for the reported link between myopia devel- aberration (Z 06 ) and primary (Z 3−1 ) and sec-
HOAs in the development of non-amblyopic opment and near work. ondary vertical coma (Z 5−1 ), although signifi-
anisometropia and longitudinal studies are Buehren et al.123 demonstrated that a cant changes also occur in secondary
required to elucidate potential underlying two-hour reading task increased HOA RMS astigmatism (Z 4−2 and Z 24 ), tetrafoil (Z 44 ) and
mechanisms. in both emmetropes and myopes. Myopes pentafoil (Z 5−5 ).132 Ghosh et al.132 also
exhibited greater HOA RMS at both distance showed that accommodation during down-
Amblyopia and near, and a larger increase in HOA RMS gaze produced a greater negative shift in
Unilateral amblyopia results from a signifi- from distance to near fixation. Correspond- primary spherical aberration (Z 04 ) and pri-
cant interocular difference in image quality ingly, the increase in HOA RMS associated
mary vertical coma (Z 3−1 ), and a greater posi-
or visual experience during early life, typi- with the near task resulted in a reduction in
tive shift in secondary spherical aberration
cally hyperopic anisometropia,114 strabis- retinal image quality, with myopic eyes
(Z 06 ) than during accommodation alone.
mus or form deprivation.115 In children, exhibiting poorer retinal image quality at
significant differences in HOAs have gener- Given these findings, and that anterior cor-
distance and near than emmetropes, and
ally not been observed between the ambly- neal HOAs133 and elevation134 remain stable
undergoing a greater reduction at near than
during accommodation in primary gaze, the
opic and the non-amblyopic fellow eye in emmetropes.18 Given that near work typi-
monocular amblyopes, whether the cause changes in terms such as primary vertical
cally involves accommodation, downgaze
of their amblyopia is strabismic95 or refrac- and convergence, it is of interest to under- coma (Z 3−1 ) and trefoil (Z 3−3 ) during down-
tive.116 In ‘idiopathic’ amblyopia (reduced stand the changes in HOAs that occur inde- gaze (and near work) are likely associated
visual acuity with no amblyogenic factor), pendently with each aspect of near work. with lid-induced corneal deformation at the
while there was no interocular difference in For a fixed pupil diameter, HOA RMS con- superior pupil margin,123,135,136 while the
the means of individual terms, interocular sistently increases with greater accommoda- variations in primary (Z 04 ) and secondary (Z 06 )
differences were observed in the composition demands,124,125 although some have spherical aberration are likely the result of
tion of the HOA profile and the interaction found this occurs only with demands above accommodation.
between individual terms.117 Vincent 3 D.126–129 The major consistent change in Subsequent studies have confirmed that
118
et al. reported no difference in total ocu- HOAs that occurs during accommodation is typical accommodation demands (2–3 D) pro-
lar HOA RMS between the fellow eyes of a decrease in primary spherical aberration duce poorer retinal image quality than a 0 D
adult refractive amblyopes; however, the (Z 04 ), becoming less positive or more accommodation demand for a fixed pupil
77
diameter124 and Buehren et al.137 also retinal image, changes in the profile or 7.9 mm, respectively. Based on the polyno-
reported similar results during accommoda- magnitude of HOAs associated with these mial regression reported by Salmon and
tion for a natural pupil. These findings indicate interventions may influence axial eye van de Pol,153 HOA RMS would have
that with normal levels of accommodation growth and myopia progression. increased by ~0.14, 0.43 and 0.54 μm due
during near work, even with natural to the pupil dilation associated with 0.01%,
accommodation-induced pupil miosis, HOAs Anti-muscarinic agents 0.1% and 0.5% atropine, respectively.
increase and retinal image quality is reduced. Arguably the myopia intervention that has Therefore, the change in HOA RMS
Long periods of exposure to reduced retinal shown the greatest efficacy in animals143 resulting from pupil mydriasis is signifi-
image quality as a result of increased HOAs or and humans144,145 is the non-selective anti- cantly greater than the change in HOAs
altered HOA profile during prolonged near muscarinic pharmacological agent, atropine. resulting from the cycloplegic hyperopic
work may therefore provide a stimulus within Atropine reduces myopia progression in a shift. This suggests that if the effect of
the retinal image to which the eye responds dose-dependent manner,144–146 although, atropine on eye growth is mediated via a
by increasing its axial growth. questions remain about its efficacy in mechanism involving HOAs, it may be the
The combination of different HOA terms slowing axial elongation, particularly for result of the increased pupil size rather
can cause different effects on the quality of lower concentrations.146,147 Given the cyclo- than the changes in HOAs associated with
the retinal image. Thibos et al.138 demon- plegic effect of atropine in humans,148 its the hyperopic shift from cycloplegia.
strated that the combination of hyperopic mechanism of myopia control was originally The effect of other concentrations of
defocus and negative primary spherical thought to be related to changes in the atropine on HOAs have not been examined,
aberration (Z 04 ) produces a retinal image of accommodative system.149 However, animal although similar findings have been dem-
poorer quality than if positive primary studies suggest that anti-muscarinic agents onstrated following the instillation of other
spherical aberration (Z 04 ) was present with influence growth via an alternative, non- topical anti-muscarinic agents. In similarly
hyperopic defocus. Given that myopes have accommodative mechanism.143,149 While yet aged myopic children, an increase in HOA
to be confirmed in humans, evidence from and spherical-like RMS of 0.025 and
been shown to exhibit higher accommoda-
the chick suggests that atropine binds to 0.014 μm for a 6 mm pupil, respectively,
tive lags139,140 (producing hyperopic
receptors within the retina (possibly a combi- was observed following the instillation of
defocus), and primary spherical aberration
nation of muscarinic and non-muscarinic)149 1% cyclopentolate eye drops.154 Addition-
(Z 04 ) typically becomes negative with accom-
and triggers a signalling cascade to the sclera ally, a positive shift in ocular primary spher-
modation, this combination of optical
via the retinal pigment epithelium and cho- ical aberration (Z 04 ) occurred coincidentally
changes may result in reduced retinal image
roid, mediated by nitric oxide.150 However, with a 0.50 D hyperopic shift, but these
quality and provide a stimulus for ocular
the cycloplegic and mydriatic effects of atro- changes were of smaller magnitude com-
growth. Buehren et al.141 also modelled dif-
pine, which change the pupil diameter and pared to atropine and there was no change
ferent combinations of the terms that most
the crystalline lens shape and thickness, alter observed in coma-like RMS or primary hori-
consistently vary with near work; positive
the ocular HOA profile and may provide an zontal coma (Z 13 ).154 Interestingly, 0.5%
vertical trefoil (Z 3−3 ), negative primary verti-
optical signal which influences eye growth. tropicamide eye drops produced a small
cal coma (Z 3−1 ) and negative primary spheri- The twice-daily instillation of 1% atropine increase of 0.017 μm in total coma but negli-
cal aberration (Z 04 ). The sphero-cylindrical eye drops for one week in hyperopic Japa- gible changes in HOA and spherical aberra-
correction that minimised the wavefront nese children (three to 12 years) produced a tion RMS;155 however, given the post-
error and maximised retinal image quality small but significant increase of 0.044, 0.032 instillation interval prior to measurement
produced by the typical combination of and 0.023 μm in ocular HOA, coma-like and was only five minutes in this study, the full
these terms was a low hyperopic, against- spherical-like RMS, respectively, with no manifestation of optical changes may not
the-rule astigmatic correction. This indicates demonstrable change in corneal HOAs.151 have been observed since maximal
that the change in the wavefront generated Both primary horizontal coma (Z 13 ) and cycloplegia due to tropicamide occurs
during near work may mimic hyperopic approximately 20 minutes post-instilla-
spherical aberration (Z 04 ) approximately dou-
defocus and provide a stimulus to the retina tion.148 One per cent cyclopentolate and
bled following the use of atropine, becom-
that promotes myopic eye growth (Figure 5). 0.5% tropicamide have been shown to have
ing more positive; however, given that the some156 and negligible157 effect in slowing
HOAs were analysed over the same fixed myopia progression, respectively. The mag-
pupil size (6 mm) before and after atropine, nitude of the changes in HOAs with
HOAs and myopia control the authors suggested that these changes cyclopentolate and tropicamide are smaller
interventions were likely the result of the 1.18 D hyper- than with atropine; therefore, this may con-
opic shift.151 tribute to the differences observed in their
The use of optical and pharmacological Although Hiraoka et al.151 did not report myopia control efficacy. Like atropine, the
interventions in clinical practice to prevent on the changes in pupil size due to 1% atro- mydriasis from these pharmacological
or slow the progression of myopia have pine, post-hoc analysis indicated that if pupil agents are likely to influence HOAs more
become more widespread in recent size increased from 4 to 6 mm, HOA RMS significantly than the hyperopic shift associ-
years;142 however, the underlying mecha- would have increased by ~0.28 μm. Chia ated with cycloplegia. Longitudinal studies
nisms of these treatments are not fully et al.152 reported a change in photopic pupil are required to examine a potential link
understood. Given that most myopia con- size with 0.01%, 0.1% and 0.5% atropine between the changes in HOAs from anti-
trol treatments alter the quality of the from a baseline of ~4.7 mm to 5.8, 7.4 and muscarinic agents and myopia control.
78
Negative spherical aberration Optimal image plane

during accommodation (axial elongation
Superior corneal distortions
resulting from lid forces required to optimise
during near work/downgaze hyperopic defocus-like
stimulation)
Ideal wavefront entering eye

from near distance
Cornea Total
2.5 0.5

0
y (mm)
-0.5
-2.5 -1
-2.5 0 2.5 -2.5 0 2.5
x (mm) x (mm)
Figure 5. Schematic of a potential mechanism between near work and myopia development involving higher order aberrations
(HOAs). The increase in positive trefoil (Z 3−3 ) and negative primary vertical coma (Z 3−1 ) from lid-induced superior corneal distortion
during downgaze, and increase in accommodation-induced negative primary spherical aberration (Z 04 ) from the change in the crys-
talline lens results in rays from the edge of the entrance pupil exhibiting negative vergence relative to paraxial pupil rays. These
rays produce a plane of best focus (optimal retinal image) posterior to the retina, emulating hyperopic defocus which may
encourage axial eye growth. Note the increase in negative refractive power, particularly in the superior third of the pupil in the
included corneal and total ocular refractive power maps (generated from third and fourth order HOAs for a 5 mm pupil).
Figure and data adapted from Buehren et al.123,135,141
Spectacle lenses: bifocal and reduced myopia progression over three Near addition lenses were originally
progressive addition lenses years by approximately 14 per cent,158 or thought to act by reducing the near accom-
Several optical interventions, including spec- 25 per cent in those with near esophoria modation demand.158–161 Since the near
tacles and contact lenses, have been devel- and a lag of accommodation at near.159 addition zones induce localised superior rel-
oped and studied for their potential ability Bifocal lenses, with and without base-in ative peripheral retinal myopic
to slow the progression of myopia. Of the prism in the near segment, have demon- defocus,162,163 and the relative superior
spectacle lens designs, the most promising strated a greater level of myopia control peripheral refractive shift was found to be
have been progressive addition and bifocal over three years, with a reduction in myopia associated with a reduction in the rate of
spectacle lenses.145 In comparison to single- progression of approximately 40 per cent central myopic refractive progression of a
vision designs, progressive addition lenses and 50 per cent, respectively.160 one-year period in progressive addition lens
79
wearers, this may be a possible alternative in primary spherical aberration (Z 04 ) is the myopia control effect of ortho-
mechanism by which these lenses slow thought to be the result of mid-peripheral keratology; however, further longitudinal
myopia progression.163 Progressive addition corneal steepening,173 and comatic changes studies that examine the changes in on- and
and bifocal spectacle lenses will vary the are likely the result of lens, and therefore off-axis corneal and total ocular HOAs, pupil
optics of the eye as a result of a change in treatment zone, decentration.173,177 While size and their association with eye growth
accommodation response from the near both the corneal and ocular HOAs increase, before and during orthokeratology are
addition161,164 as well as the variable optics the corneal changes are substantially required to provide further insights into a
in the periphery of the lenses. greater172,173 which suggests some internal potential role for HOAs in the myopia con-
It is currently unknown what changes optical adaptation in response to the cor- trol effect of orthokeratology.
occur to the HOA profile when looking neal modifications over time, perhaps due
through the different segments of a bifocal to an altered accommodative response.178 Soft contact lenses: multifocal
lens; however, the intermediate and near Significant changes in off-axis HOAs also and dual-focus
zones of a progressive lens produce an occur as a result of orthokeratology treat- Soft contact lenses with modified refractive
increase in HOA RMS of 0.119 and ment. On average, orthokeratology pro- profiles have also shown significant efficacy
0.071 μm, respectively, relative to the dis- duces a significant increase in the in reducing myopia progression and axial
tance area (for a 5 mm pupil).165 Predomi- magnitude and peripheral rate of change of eye growth. On average, multifocal and dual-
nantly, vertical coma (Z 3−1 ) and trefoil (Z 3− 3 ) HOA RMS across the visual field.179 Typi- focus lenses reduce myopia progression by
exhibit changes, particularly in the lens cally, minimal variation in primary spherical approximately 30–50 per cent; however,
periphery.166,167 While the optics vary sub- aberration (Z 04 ) is observed across the visual there is significant inter-study variation as a
stantially across a progressive addition lens, field; however, Mathur et al.179 demon- result of lens design, study duration, and par-
it is unknown what effect this may have on strated a significant positive shift on average ticipant characteristics.182 These lenses can
the HOAs of the eye or how this may affect following orthokeratology, with one subject be broadly categorised as multifocal or dual-
eye growth and refractive error develop- exhibiting a quadratic variation (more posi- focus lenses, according to how the optical
ment, since pupil size and accommodation tive at the centre of the visual field) and profile of the lens varies across the optic
during lens wear will also vary and may another subject showing an overall positive zone. A multifocal or aspheric lens design
influence retinal image quality. shift along the horizontal meridian. Most provides a central zone of distance power
notably, vertical coma increased from supe- with a progressive increase in positive power
rior to inferior eccentricities and horizontal toward the edge of the optic zone. A dual-
Orthokeratology coma increased from nasal to temporal focus lens similarly has a central distance
Some contact lens designs have shown sig- prior to treatment, which reversed post- zone, surrounded by multiple concentric
nificantly greater myopia control efficacy orthokeratology.179 alternating zones of relative positive power
than spectacle lenses. Overnight wear of Hiraoka et al.60 reported that the change and the central distance refraction. Given
rigid reverse-geometry lens designs, or in corneal coma-like aberrations following that single-vision spherical rigid183 and soft
orthokeratology, produces central corneal one year of orthokeratology in Japanese contact lenses184 have minimal effect on
flattening and mid-peripheral corneal steep- children exhibited a moderate negative lin- myopia progression, the modified optics
ening. Numerous studies168–170 have dem- ear correlation with axial elongation must contribute to the myopia control effect
onstrated a significant and repeatable (r = −0.46, p = 0.0003), whereby less axial of multifocal and dual-focus lenses.
slowing of axial elongation in children by eye growth was associated with a larger The measured effect of different soft con-
approximately 45 per cent on average with change in coma-like aberrations. Conversely, tact lens designs on HOAs are similar; how-
orthokeratology.145 Santodomingo-Rubido et al.175 found no sig- ever, there are some notable differences
Orthokeratology produces a significant nificant correlation between corneal HOA (Figure 6). Distance-centre multifocal contact
increase in on-axis HOAs, even after one changes and axial elongation after three lenses185,186 produce significant positive
night of wear.171 Following seven nights of and 24 months of orthokeratology in shifts in primary spherical aberration (Z 04 )
treatment, the increase in corneal HOAs European children. Chen et al.180 found that ranging from 0.125 μm with a low (+1.50 D)
ranges from 0.199 μm over a 5 mm pupil,172 a larger pupil size during orthokeratology to 0.245 μm with a high add (+2.50 D) for a
to 0.71 μm over a 6 mm pupil.173 Addition- treatment in Chinese children was associ- 5 mm pupil.185 In addition, Fedtke et al.186
ally, the increase in ocular HOAs has been ated with slower axial eye growth than a demonstrated that primary horizontal coma
reported to be 0.175 μm over a 5 mm smaller pupil, and suggested that this is due (Z 13 ) increases with multifocal contact lenses
pupil172 to 0.63 μm over a 6.5 mm pupil,174 to a greater relative peripheral myopic shift. due to lens decentration. On-eye modelling
with the changes in ocular and corneal This change in peripheral refraction was of distance-centre dual-focus lenses through
HOAs typically stabilising after 30 nights of confirmed with modelling by Faria-Ribeiro a schematic eye185 showed that these lenses
lens wear.174 The predominant changes in et al.,181 who also demonstrated that on- also shift primary spherical aberration (Z 04 )
HOAs are positive shifts in corneal172,173 and off-axis HOAs, particularly primary hori- more positively when measured across a
and ocular171–174 primary spherical aberra- zontal coma (Z13 ) and primary spherical 3 mm pupil; however, primary spherical
tion (Z 04 ), and corneal173 and ocular171,173,174 aberration (Z 04 ), also increase with greater aberration (Z 04 ) became negative when
primary horizontal coma (Z 13 ). In addition, pupil size as a result of corneal topographi- analysed over a 4 mm pupil and more so
changes in corneal primary vertical coma cal changes during orthokeratology. This over a 5 mm pupil. This suggests that the
(Z 3− 1 ) have been reported.175,176 The change finding may indicate that HOAs influence concentric, alternating power profile of the
80
A Single vision B Distance-centre multifocal C Dual-focus

1
3

2 0.5
1 0
y (mm)
0 -0.5
-1 -1
-2 -1.5
-3 -2
-3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3
2
Normalised frequency
of occurence
1.5
0.5
0
-2 -1 0 1 -2 -1 0 1 -2 -1 0 1
Refractive power (D) Refractive power (D) Refractive power (D)
Figure 6. Refractive power maps and associated histograms generated from the difference in ocular higher order aberrations
(HOAs) (third to eighth order), measured using a commercial Hartmann-Shack wavefront sensor (COAS-HD, Wavefront Sciences)
during soft contact lens wear compared to a bare eye condition in a moderately myopic young adult with normal vision for
A: single-vision, B: distance-centre multifocal, and C: dual-focus soft contact lenses over a 6 mm pupil. Each lens had the same dis-
tance zone refractive power (−4.00 D), with a +2.00 D addition power in the distance-centre multifocal and dual-focus lens. Note
that although the overall refractive power distribution is similar between the three lenses, the location of the refractive powers
within the pupil plane varies between the three lenses, with an increase in positive refractive power in the mid-periphery and
periphery of the dual-focus and distance-centre multifocal contact lenses compared to the single-vision lens, consistent with the
differences in primary spherical aberration (Z 04 ) between these lenses.
lenses causes changes in HOAs that are over one year of wear in Chinese children. reduction in axial elongation and myopia
markedly pupil-dependent. Multifocal con- Fujikado et al.190 reported on the myopia progression in American children after six
tact lenses have been shown to affect the control effect of a multifocal soft contact months, respectively; however, the efficacy
accommodation response;187,188 however, lens which included +0.50 D at 4 mm from reduced by 12 months to 39 per cent and
the change in HOAs during accommodation the lens centre with a unique 0.5 mm nasal 20 per cent, respectively.191 A crossover
with multifocal contact lens wear is yet to be decentration of the optic zone. The study in New Zealand children examining
examined. Future studies examining the decentration was designed to better align the effect of a dual-focus lens design with
change in HOAs during accommodation with, and produce more symmetrical optics +2.00 D zones reported a reduction in myo-
while wearing dual-focus and multifocal con- across, the pupil, and produced a reduction pia progression and axial elongation of
tact lenses may provide valuable insights in axial elongation of 47 per cent over 36 per cent and 50 per cent, respectively,
into accommodative function during lens 12 months in Japanese children but did not over 10 months.192 A recent two-year
wear and potential mechanisms for myopia demonstrate a significant effect on refrac- randomised, controlled trial in Chinese chil-
control. tive myopia progression. Another novel soft dren193 examined four novel contact lens
A multifocal soft lens design by contact lens design incorporated positive designs, two of which incorporated periph-
Sankaridurg et al.,189 which incorporated spherical aberration (+0.175 μm for a 5 mm eral positive refractive shifts of +1.50 and
+2.00 D at 4.5 mm from the optical centre pupil, an amount purported by the authors +2.50 D at 3 mm from the lens centre (simi-
(9 mm optic zone diameter), resulted in a to negate the accommodation-induced neg- lar to the commercial distance-centre
reduction in myopia progression and axial ative shift of spherical aberration) and multifocal), and two which manipulated the
elongation of approximately 34 per cent resulted in a 65 per cent and 54 per cent HOAs to improve retinal image quality at
81
and anterior to the retina, but diminished combinations of illumination and accommo- 7. Holden BA, Fricke TR, Wilson DA et al. Global preva-
lence of myopia and high myopia and temporal
image quality for planes posterior to the ret- dation, and how these factors may tempo- trends from 2000 through 2050. Ophthalmology
ina. These lenses reduced myopia progres- rally interact to influence eye growth and 2016; 123: 1036–1042.
sion and axial elongation by ~30 per cent refractive development. 8. Lam DS, Fan DS, Chan WM et al. Prevalence and
characteristics of peripheral retinal degeneration in
and 22 per cent, respectively, with negligible Chinese adults with high myopia: a cross-sectional
differences between the lens designs.193 prevalence survey. Optom Vis Sci 2005; 82: 235–238.
Based on the difference in refractive Conclusion 9. Chen SJ, Cheng CY, Li AF et al. Prevalence and associ-
ated risk factors of myopic maculopathy in elderly
power between the centre and periphery of Chinese: the Shihpai eye study. Invest Ophthalmol Vis
the lenses, and calculations described by Several plausible theories based on retinal Sci 2012; 53: 4868–4873.
10. Shen L, Melles RB, Metlapally R et al. The association
Carkeet et al.,80 the lens designs by Fujikado image quality and vergence cues suggest of refractive error with glaucoma in a multiethnic
et al.190 and Sankaridurg et al.189 would that HOAs may play a role in the develop- population. Ophthalmology 2016; 123: 92–101.
have produced approximately +0.055 μm of ment and control of refractive error and eye 11. Pan CW, Cheng CY, Saw SM et al. Myopia and
age-related cataract: a systematic review and
primary spherical aberration (Z04 )
over a growth. However, animal studies suggest meta-analysis. Am J Ophthalmol 2013; 156:
5 mm pupil, while the lens designs recently that changes in the HOA profile may simply 1021–1033.
12. Zheng YF, Pan CW, Chay J et al. The economic cost of
reported by Sankaridurg et al.193 would be a consequence of experimentally myopia in adults aged over 40 years in Singapore.
have produced approximately +0.102 μm induced refractive error rather than a cause. Invest Ophthalmol Vis Sci 2013; 54: 7532–7537.
Likewise, variations in the HOA profile 13. Ohno-Matsui K, Lai TY, Lai CC et al. Updates of path-
over a 5 mm pupil. The Anstice and Phil-
ologic myopia. Prog Ret Eye Res 2016; 52: 156–187.
lips192 lens would have produced a greater between myopic, hyperopic and 14. Flitcroft DI. The complex interactions of retinal, opti-
level of positive spherical aberration emmetropic eyes in both children and cal and environmental factors in myopia aetiology.
Prog Ret Eye Res 2012; 31: 622–660.
(approximately +0.213 μm over a 4.78 mm adults have not produced clear, consistent,
15. Troilo D. Neonatal eye growth and emmetropisation
pupil) and showed a greater reduction in and reliable evidence to substantiate these - a literature review. Eye 1992; 6: 154–160.
myopia progression and axial elongation. theories. Characteristic temporal variations 16. Liang J, Williams DR. Aberrations and retinal image
quality of the normal human eye. J Opt Soc Am A Opt
This, together with the reports that commer- and changes during near work and accom- Image Sci Vis 1997; 14: 2873–2883.
cially available soft multifocal and dual-focus modation provide some evidence of a role 17. Plainis S, Ginis HS, Pallikaris A. The effect of ocular
for certain HOAs, such as primary spherical aberrations on steady-state errors of accommoda-
lenses also produce positive spherical aber-
tive response. J Vis 2005; 5: 466–477.
ration, support a role for spherical aberra- aberration (Z 04 ) coma (Z 3−1 and Z 13 ) and trefoil 18. Collins MJ, Buehren T, Iskander DR. Retinal image
tion in the myopia control effect of these (Z 3− 3 and Z 33 ); however, longitudinal studies quality, reading and myopia. Vision Res 2006; 46:
196–215.
lenses, in what may be a magnitude- examining the changes in HOAs during 19. Mutti DO, Mitchell GL, Jones LA et al. Axial growth
dependent manner. infancy, childhood and adolescence, and and changes in lenticular and corneal power during
emmetropization in infants. Invest Ophthalmol Vis Sci
their association with refractive error and 2005; 46: 3074–3080.
ocular structural development are required 20. Brown NP, Koretz JF, Bron AJ. The development
Consideration of pupil size and maintenance of emmetropia. Eye 1999; 13:
to comprehensively investigate these theo-
83–92.
ries. Additionally, further longitudinal stud- 21. Saw S-M, Gazzard G, Shin-Yen EC et al. Myopia and
Although HOAs may influence refractive ies of the changes in HOAs before and after associated pathological complications. Ophthalmic
error development and the treatment effect the introduction of various pharmacological Physiol Opt 2005; 25: 381–391.
of myopia control interventions via several 22. Flitcroft DI. Emmetropisation and the aetiology of
and optical interventions for the control of refractive errors. Eye 2014; 28: 169–179.
potential mechanisms, an important consid- myopia progression are necessary to estab- 23. Wildsoet CF. Active emmetropization - evidence for
eration is the effect of pupil size. Most stud- lish a clearer link between the optical
its existence and ramifications for clinical practice.
Ophthalmic Physiol Opt 1997; 17: 279–290.
ies typically report HOAs over a fixed pupil changes associated with these treatments 24. Wallman J, Turkel J, Trachtman J. Extreme myopia
size of 5 mm or greater; however, this may and their demonstrated efficacy. produced by modest change in early visual experi-
not be a realistic pupil size for children ence. Science 1978; 201: 1249–1251.
25. Barathi VA, Boopathi VG, Yap EP et al. Two models of
across a range of visual tasks and environ- experimental myopia in the mouse. Vision Res 2008;
ments, and accommodation demands.194 REFERENCES 48: 904–916.
Pupil size is dynamic and influenced by sev- 1. Dolgin E. The myopia boom. Nature 2015; 519: 26. Verolino M, Nastri G, Sellitti L et al. Axial length
276–278. increase in lid-sutured rabbits. Surv Ophthalmol
eral factors, including age,81,82 ambient 2. Attebo K, Ivers R, Mitchell P. Refractive errors in an 1999; 44: 103–108.
illumination,81,194 and accommoda- older population - the Blue Mountains eye study. 27. Sherman SM, Norton TT, Casagrande VA. Myopia in
Ophthalmology 1999; 106: 1066–1072. the lid-sutured tree shrew (Tupaia glis). Brain Res
tion.137,194 Exposure to photopic illumina-
3. Chen M, Wu A, Zhang L et al. The increasing preva- 1977; 124: 154–157.
tion conditions, such as bright outdoor lence of myopia and high myopia among high school 28. Troilo D, Judge SJ. Ocular development and visual
lighting, is likely to reduce the pupil diame- students in Fenghua City, eastern China: a 15-year deprivation myopia in the common marmoset
population-based survey. BMC Ophthalmol 2018; (Callithrix jacchus). Vision Res 1993; 33: 1311–1324.
ter to under 3–4 mm and may result in 18: 1–10. 29. Wiesel TN, Raviola E. Myopia and eye enlargement
HOAs of negligible magnitude (diffraction- 4. Han SB, Jang J, Yang HK et al. Prevalence and risk fac- after neonatal lid fusion in monkeys. Nature 1977;
limited pupil size, or Marechal criterion). It is tors of myopia in adult Korean population: Korea 266: 66–68.
National Health and Nutrition Examination Survey 30. Shen W, Vijayan M, Sivak JG. Inducing form-
possible that the reduced HOAs in bright 2013-2014 (KNHANES VI). PLoS ONE 2019; 14: deprivation myopia in fish. Invest Ophthalmol Vis Sci
outdoor lighting may explain the reduced e0211204. 2005; 46: 1797–1803.
5. Koh V, Yang A, Saw SM et al. Differences in preva- 31. Schaeffel F, Burkhardt EV, Howland HC. Measure-
axial elongation in non-myopic and myopic
lence of refractive errors in young Asian males in Sin- ment of refractive state and deprivation myopia in
children195 as a result of the improved reti- gapore between 1996-1997 and 2009-2010. two strains of mice. Optom Vis Sci 2004; 81:
nal image quality. Further research is Ophthalmic Epidemiol 2014; 21: 247–255. 99–110.
6. Lin LL, Shih YF, Hsiao CK et al. Prevalence of myopia 32. Howlett MH, McFadden SA. Form-deprivation myopia
required to examine pupil size dynamics in Taiwanese schoolchildren: 1983 to 2000. Ann Acad in the Guinea pig (Cavia porcellus). Vision Res 2006;
and HOAs under varying levels and Med Singapore 2004; 33: 27–33. 46: 267–283.
82
33. Smith EL III, Hung LF. Form-deprivation myopia in 59. Ramamirtham R, Kee CS, Hung LF et al. Wave aberra- 83. Silbert J, Matta N, Tian J et al. Pupil size and
monkeys is a graded phenomenon. Vision Res 2000; tions in rhesus monkeys with vision-induced ametro- anisocoria in children measured by the plusoptiX
40: 371–381. pias. Vision Res 2007; 47: 2751–2766. photoscreener. J AAPOS 2013; 17: 609–611.
34. O’Leary DJ, Millodot M. Eyelid closure causes myopia 60. Hiraoka T, Kakita T, Okamoto F et al. Influence of 84. Mathur A, Atchison DA, Charman WN. Effect of
in humans. Experientia 1979; 35: 1478–1479. ocular wavefront aberrations on axial length elonga- accommodation on peripheral ocular aberrations.
35. von Noorden GK, Lewis RA. Ocular axial length in tion in myopic children treated with overnight ortho- J Vis 2009; 9: 1–11.
unilateral congenital cataracts and blepharoptosis. keratology. Ophthalmology 2015; 122: 93–100. 85. Mathur A, Atchison DA, Charman WN. Effects of age
Invest Ophthalmol Vis Sci 1987; 28: 750–752. 61. Lau JK, Vincent SJ, Collins MJ et al. Ocular higher- on peripheral ocular aberrations. Opt Express 2010;
36. Gee KS, Tabbara KF. Increase in ocular axial length in order aberrations and axial eye growth in young 18: 5840–5853.
patients with corneal opacification. Ophthalmology Hong Kong children. Sci Rep 2018; 8: 6726. 86. Mathur A, Atchison DA, Charman WN. Myopia and
1988; 95: 1276–1278. 62. Kim J, Lim DH, Han SH et al. Predictive factors associ- peripheral ocular aberrations. J Vis 2009; 9: 1–12.
37. Miller-Meeks MJ, Bennett SR, Keech RV et al. Myopia ated with axial length growth and myopia progres- 87. Osuagwu UL, Suheimat M, Atchison DA. Peripheral
induced by vitreous hemorrhage. Am J Ophthalmol sion in orthokeratology. PLoS ONE 2019; 14: aberrations in adult hyperopes, emmetropes and
1990; 102: 199–203. e0218140. myopes. Ophthalmic Physiol Opt 2017; 37: 151–159.
38. Schaeffel F, Glasser A, Howland HC. Accommodation, 63. Wildsoet CF, Schmid KL. Emmetropization in chicks 88. Philip K, Sankaridurg PR, Ale JB et al. Profile of off-
refractive error and eye growth in chickens. Vision uses optical vergence and relative distance cues to axis higher order aberrations and its variation with
Res 1988; 28: 639–657. decode defocus. Vision Res 2001; 41: 3197–3204. time among various refractive error groups. Vision
39. Wallman J, Winawer J. Homeostasis of eye growth and 64. Artal P, Guirao A, Berrio E et al. Compensation of Res 2018; 153: 111–123.
the question of myopia. Neuron 2004; 43: 447–468. corneal aberrations by the internal optics in the 89. He JC, Sun P, Held R et al. Wavefront aberrations in
40. Diether S, Schaeffel F. Local changes in eye growth human eye. J Vis 2001; 1: 1–8. eyes of emmetropic and moderately myopic school
induced by imposed local refractive error despite 65. Atchison DA, Markwell EL. Aberrations of children and young adults. Vision Res 2002; 42:
active accommodation. Vision Res 1997; 37: 659–668. emmetropic subjects at different ages. Vision Res 1063–1070.
41. Howlett MH, McFadden SA. Spectacle lens compen- 2008; 48: 2224–2231. 90. Paquin MP, Hamam H, Simonet P. Objective mea-
sation in the pigmented Guinea pig. Vision Res 2009; 66. McLellan JS, Marcos S, Burns SA. Age-related surement of optical aberrations in myopic eyes.
49: 219–227. changes in monochromatic wave aberrations of the Optom Vis Sci 2002; 79: 285–291.
42. Shen W, Sivak JG. Eyes of a lower vertebrate are sus- human eye. Invest Ophthalmol Vis Sci 2001; 42: 91. Yazar S, Hewitt AW, Forward H et al. Comparison of
ceptible to the visual environment. Invest Ophthalmol 1390–1395. monochromatic aberrations in young adults with dif-
Vis Sci 2007; 48: 4829–4837. 67. Wang L, Koch DD. Ocular higher-order aberrations in ferent visual acuity and refractive errors. J Cataract
43. Shaikh AW, Siegwart JT, Norton TT. Effect of inter- individuals screened for refractive surgery. J Cataract Refract Surg 2014; 40: 441–449.
rupted lens wear on compensation for a minus lens Refract Surg 2003; 29: 1896–1903. 92. Cheng X, Bradley A, Hong X et al. Relationship
in tree shrews. Optom Vis Sci 1999; 76: 308–315. 68. Artal P, Berrio E, Guirao A et al. Contribution of the between refractive error and monochromatic aber-
44. Norton TT, Siegwart JT, Amedo AO. Effectiveness of cornea and internal surfaces to the change of ocular rations of the eye. Optom Vis Sci 2003; 80: 43–49.
hyperopic defocus, minimal defocus, or myopic aberrations with age. J Opt Soc Am A Opt Image Sci Vis 93. Kwan WCK, Yip SP, Yap MKH. Monochromatic aber-
defocus in competition with a myopiagenic stimulus 2002; 19: 137–143. rations of the human eye and myopia. Clin Exp
in tree shrew eyes. Invest Ophthalmol Vis Sci 2006; 69. Brunette I, Bueno JM, Parent M et al. Monochromatic Optom 2009; 92: 304–312.
47: 4687–4699. aberrations as a function of age, from childhood to 94. Llorente L, Barbero S, Cano D et al. Myopic versus
45. Graham B, Judge SJ. The effects of spectacle wear in advanced age. Invest Ophthalmol Vis Sci 2003; 44: hyperopic eyes: axial length, corneal shape and opti-
infancy on eye growth and refractive error in the mar- 5438–5446. cal aberrations. J Vis 2004; 4: 288–298.
moset (Callithrix jacchus). Vision Res 1999; 39: 189–206. 70. Levy Y, Segal O, Avni I et al. Ocular higher-order 95. Kirwan C, O’Keefe M, Soeldner H. Higher-order aber-
46. Troilo D, Totonelly K, Harb E. Imposed anisometro- aberrations in eyes with supernormal vision. rations in children. Am J Ophthalmol 2006; 141:
pia, accommodation, and regulation of refractive Am J Ophthalmol 2005; 139: 225–228. 67–70.
state. Optom Vis Sci 2009; 86: E31–E39. 71. Radhakrishnan H, Charman WN. Age-related 96. Zhang N, Yang XB, Zhang WQ et al. Relationship
47. Smith EL III, Hung LF. The role of optical defocus in changes in ocular aberrations with accommodation. between higher-order aberrations and myopia pro-
regulating refractive development in infant monkeys. J Vis 2007; 7: 1–21. gression in schoolchildren: a retrospective study. Int
Vision Res 1999; 39: 1415–1435. 72. Amano S, Amano Y, Yamagami S et al. Age-related J Ophthalmol 2013; 6: 295–299.
48. Read SA, Collins MJ, Sander BP. Human optical axial changes in corneal and ocular higher-order wavefront 97. Li T, Zhou X, Chen Z et al. Relationship between ocu-
length and defocus. Invest Ophthalmol Vis Sci 2010; aberrations. Am J Ophthalmol 2004; 137: 988–992. lar wavefront aberrations and refractive error in Chi-
51: 6262–6269. 73. Rocha KM, Nosé W, Bottós K et al. Higher-order nese school children. Clin Exp Optom 2012; 95:
49. Chiang ST-H, Phillips JR, Backhouse S. Effect of retinal aberrations of age-related cataract. J Cataract Refract 399–403.
image defocus on the thickness of the human cho- Surg 2007; 33: 1442–1446. 98. Little JA, McCullough SJ, Breslin KM et al. Higher
roid. Ophthalmic Physiol Opt 2015; 35: 405–413. 74. Fujikado T, Kuroda T, Maeda N et al. Light scattering order ocular aberrations and their relation to refrac-
50. Wang D, Chun RK, Liu M et al. Optical defocus rapidly and optical aberrations as objective parameters to tive error and ocular biometry in children. Invest
changes choroidal thickness in schoolchildren. PLoS predict visual deterioration in eyes with cataracts. Ophthalmol Vis Sci 2014; 55: 4791–4800.
ONE 2016; 11: e0161535. J Cataract Refract Surg 2004; 30: 1198–1208. 99. Philip K, Martinez AA, Ho A et al. Total ocular, ante-
51. Zhu X, Park TW, Winawer J et al. In a matter of 75. Kuroda T, Fujikado T, Maeda N et al. Wavefront anal- rior corneal and lenticular higher order aberrations
minutes, the eye can know which way to grow. Invest ysis in eyes with nuclear or cortical cataract. in hyperopic, myopic and emmetropic eyes. Vision
Ophthalmol Vis Sci 2005; 46: 2238–2241. Am J Ophthalmol 2002; 134: 1–9. Res 2012; 52: 31–37.
52. Wallman J, Gottlieb MD, Rajaram V et al. Local retinal 76. Wang L, Dai E, Koch DD et al. Optical aberrations of 100. Papamastorakis G, Panagopoulou S, Tsilimbaris MK
regions control local eye growth and myopia. Science the human anterior cornea. J Cataract Refract Surg et al. Ocular higher-order aberrations in a school
1987; 237: 73–77. 2003; 29: 1514–1521. children population. J Optom 2015; 8: 93–100.
53. Miles FA, Wallman J. Local ocular compensation for 77. Zhang N, Liu L, Yang B et al. Higher-order aberra- 101. Martinez AA, Sankaridurg PR, Naduvilath TJ et al.
imposed local refractive error. Vision Res 1990; 30: tions in children and adolescents of Southwest Monochromatic aberrations in hyperopic and
339–349. China. Optom Vis Sci 2018; 95: 53–59. emmetropic children. J Vis 2009; 9: 1–14.
54. Rada JA, Shelton S, Norton TT. The sclera and myo- 78. Cerviño A, Hosking SL, Ferrer-Blasco T et al. A pilot 102. Philip K, Sankaridurg P, Holden B et al. Influence of
pia. Exp Eye Res 2006; 82: 185–200. study on the differences in wavefront aberrations higher order aberrations and retinal image quality in
55. Kisilak ML, Campbell MC, Hunter JJ et al. Aberrations between two ethnic groups of young generally myo- myopisation of emmetropic eyes. Vision Res 2014;
of chick eyes during normal growth and lens induc- pic subjects. Ophthalmic Physiol Opt 2008; 28: 105: 233–243.
tion of myopia. J Comp Physiol A Neuroethol Sens Neu- 532–537. 103. Hiraoka T, Kotsuka J, Kakita T et al. Relationship
ral Behav Physiol 2006; 192: 845–855. 79. Lim KL, Fam HB. Ethnic differences in higher-order between higher-order wavefront aberrations and
56. García de la Cera E, Rodríguez G, Marcos S. Longitu- aberrations: spherical aberration in the south east natural progression of myopia in schoolchildren. Sci
dinal changes of optical aberrations in normal and Asian Chinese eye. J Cataract Refract Surg 2009; 35: Rep 2017; 7: 7876.
form-deprived myopic chick eyes. Vision Res 2006; 2144–2148. 104. McLellan JS, Prieto PM, Marcos S et al. Effects of
46: 579–589. 80. Carkeet A, Luo HD, Tong L et al. Refractive error and interactions among wave aberrations on optical
57. Coletta NJ, Marcos S, Troilo D. Ocular wavefront monochromatic aberrations in Singaporean children. image quality. Vision Res 2006; 46: 3009–3016.
aberrations in the common marmoset Callithrix Vision Res 2002; 42: 1809–1824. 105. Wilson BJ, Decker KE, Roorda A. Monochromatic
jacchus: effects of age and refractive error. Vision Res 81. Winn B, Whitaker D, Elliott DB et al. Factors affecting aberrations provide an odd-error cue to focus direc-
2010; 50: 2515–2529. light-adapted pupil size in normal human subjects. tion. J Opt Soc Am A Opt Image Sci Vis 2002; 19:
58. Ramamirtham R, Kee CS, Hung LF et al. Monochro- Invest Ophthalmol Vis Sci 1994; 35: 1132–1137. 833–839.
matic ocular wave aberrations in young monkeys. 82. Birren JE, Casperson RC, Botwinick J. Age changes in 106.
Del Aguila-Carrasco AJ, Marín-Franch I, Bernal-
Vision Res 2006; 46: 3616–3633. pupil size. J Gerontol 1950; 5: 216–221. Molina P et al. Accommodation responds to optical
83
vergence and not defocus blur alone. Invest 131. Collins MJ, Wildsoet CF, Atchison DA. Monochromatic tropicamide, and artificial tear drops application in
Ophthalmol Vis Sci 2017; 58: 1758–1763. aberrations and myopia. Vision Res 1995; 35: normal eyes. Eye Contact Lens 2018; 44: 109–112.
107. Atchison DA, Pritchard N, Schmid KL et al. Shape of 1157–1163. 156. Yen MY, Liu JH, Kao SC et al. Comparison of the
the retinal surface in emmetropia and myopia. Invest 132. Ghosh A, Collins MJ, Read SA et al. The influence of effect of atropine and cyclopentolate on myopia. Ann
Ophthalmol Vis Sci 2005; 46: 2698–2707. downward gaze and accommodation on ocular aber- Ophthalmol 1989; 21: 180–187.
108. Huynh SC, Wang XY, Ip J et al. Prevalence and associ- rations over time. J Vis 2011; 11: 1–13. 157. Shih YF, Chen CH, Chou AC et al. Effects of different
ations of anisometropia and aniso-astigmatism in a 133. Read SA, Buehren T, Collins MJ. Influence of accom- concentrations of atropine on controlling myopia in
population based sample of 6 year old children. Br J modation on the anterior and posterior cornea. myopic children. J Ocul Pharmacol Ther 1999; 15:
Ophthalmol 2006; 90: 597–601. J Cataract Refract Surg 2007; 33: 1877–1885. 85–90.
109. Vincent SJ, Collins MJ, Read SA et al. Myopic anisome- 134. He JC, Gwiazda J, Thorn F et al. Change in corneal 158. Gwiazda J, Hyman L, Hussein M et al. A randomized
tropia: ocular characteristics and aetiological consid- shape and corneal wave-front aberrations with clinical trial of progressive addition lenses versus sin-
erations. Clin Exp Optom 2014; 97: 291–307. accommodation. J Vis 2003; 3: 456–463. gle vision lenses on the progression of myopia in
110. Hartwig A, Atchison DA, Radhakrishnan H. Higher- 135. Buehren T, Collins MJ, Carney L. Corneal aberrations children. Invest Ophthalmol Vis Sci 2003; 44:
order aberrations and anisometropia. Curr Eye Res and reading. Optom Vis Sci 2003; 80: 159–166. 1492–1500.
2013; 38: 215–219. 136. Vincent SJ, Read SA, Collins MJ et al. Corneal changes 159. Gwiazda J, Chandler DL, Cotter SA et al. Progressive-
111. Osuagwu UL, Suheimat M, Atchison DA. Mirror sym- following near work in myopic anisometropia. Oph- addition lenses versus single-vision lenses for
metry of peripheral monochromatic aberrations in thalmic Physiol Opt 2013; 33: 15–25. slowing progression of myopia in children with high
fellow eyes of isomyopes and anisomyopes. Invest 137. Buehren T, Collins MJ. Accommodation stimulus– accommodative lag and near esophoria. Invest
Ophthalmol Vis Sci 2016; 57: 3422–3428. response function and retinal image quality. Vision Ophthalmol Vis Sci 2011; 52: 2749–2757.
112. Tian Y, Tarrant J, Wildsoet CF. Optical and biometric Res 2006; 46: 1633–1645. 160. Cheng D, Woo GC, Drobe B et al. Effect of bifocal
characteristics of anisomyopia in human adults. 138. Thibos LN, Bradley A, Liu T et al. Spherical aberration and prismatic bifocal spectacles on myopia progres-
Ophthalmic Physiol Opt 2011; 31: 540–549. and the sign of defocus. Optom Vis Sci 2013; 90: sion in children. JAMA Ophthalmol 2014; 132:
113. Vincent SJ, Collins MJ, Read SA et al. Interocular sym- 1284–1291. 258–264.
metry in myopic anisometropia. Optom Vis Sci 2011; 139. Gwiazda J, Thorn F, Bauer J et al. Myopic children 161. Cheng D, Schmid KL, Woo GC. The effect of positive-
88: 1454–1462. show insufficient accommodative response to blur. lens addition and base-in prism on accommodation
114. Weakley DR Jr. The association between non- Invest Ophthalmol Vis Sci 1993; 34: 690–694. accuracy and near horizontal phoria in Chinese myo-
strabismic anisometropia, amblyopia, and subnor- 140. Gwiazda J, Bauer J, Thorn F et al. A dynamic relation- pic children. Ophthalmic Physiol Opt 2008; 28:
mal binocularity. Ophthalmology 2001; 108: 163–171. ship between myopia and blur-driven accommoda- 225–237.
115. Smith EL III, Hung LF, Arumugam B et al. Observations tion in school-aged children. Vision Res 1995; 35: 162. Smith EL III, Campbell MC, Irving E. Point-counter-
on the relationship between anisometropia, amblyopia 1299–1304. point. Does peripheral retinal input explain the
and strabismus. Vision Res 2017; 134: 26–42. 141. Buehren T, Iskander DR, Collins MJ et al. Potential promising myopia control effects of corneal res-
116. Kirwan C, O’Keefe M. Higher order aberrations in higher-order aberration cues for sphero-cylindrical haping therapy (CRT or ortho-K) & multifocal soft
children with amblyopia. J Pediatr Ophthalmol refractive error development. Optom Vis Sci 2007; 84: contact lenses? Ophthalmic Physiol Opt 2013; 33:
Strabmismus 2008; 45: 92–96. 163–174. 379–384.
117. Prakash G, Sharma N, Saxena R et al. Comparison of 142. Walline JJ. Myopia control: a review. Eye Contact Lens 163. Berntsen DA, Barr CD, Mutti DO et al. Peripheral
higher order aberration profiles between normal 2016; 42: 3–8. defocus and myopia progression in myopic children
and amblyopic eyes in children with idiopathic 143. McBrien NA, Moghaddam HO, Reeder AP. Atropine randomly assigned to wear single vision and pro-
amblyopia. Acta Ophthalmol 2011; 89: e257–e262. reduces experimental myopia and eye enlargement gressive addition lenses. Invest Ophthalmol Vis Sci
118. Vincent SJ, Collins MJ, Read SA et al. Monocular via a nonaccommodative mechanism. Invest 2013; 54: 5761–5770.
amblyopia and higher order aberrations. Vision Res Ophthalmol Vis Sci 1993; 34: 205–215. 164. Schilling T, Ohlendorf A, Varnas SR et al. Peripheral
2012; 66: 39–48. 144. Chia A, Lu QS, Tan D. Five-year clinical trial on atro- design of progressive addition lenses and the lag of
119. Goss DA. Nearwork and myopia. Lancet 2000; 356: pine for the treatment of myopia 2: myopia control accommodation in myopes. Invest Ophthalmol Vis Sci
1456–1457. with atropine 0.01% eyedrops. Ophthalmology 2016; 2017; 58: 3319–3324.
120. Sorsby A. School myopia. Br J Ophthalmol 1932; 16: 123: 391–399. 165. Ghosh A, Collins MJ, Read SA et al. Measurement of
217–222. 145. Huang J, Wen D, Wang Q et al. Efficacy comparison ocular aberrations in downward gaze using a modi-
121. Jones-Jordan LA, Sinnott LT, Graham ND et al. The of 16 interventions for myopia control in children: a fied clinical aberrometer. Biomed Opt Express 2011;
contributions of near work and outdoor activity to network meta-analysis. Ophthalmology 2016; 123: 2: 452–463.
the correlation between siblings in the collaborative 697–708. 166. Villegas EA, Artal P. Spatially resolved wavefront aberra-
longitudinal evaluation of ethnicity and refractive 146. Yam JC, Jiang Y, Tang SM et al. Low-concentration tions of ophthalmic progressive-power lenses in normal
error (CLEERE) study. Invest Ophthalmol Vis Sci 2014; atropine for myopia progression (LAMP) study myo- viewing conditions. Optom Vis Sci 2003; 80: 106–114.
55: 6333–6339. pia control. Ophthalmology 2018; 126: 113–124. 167. Blendowske R, Villegas EA, Artal P. An analytical
122. Mutti DO, Zadnik K. Has near work’s star fallen? 147. Bullimore MA, Berntsen DA. Low-dose atropine for model describing aberrations in the progression cor-
Optom Vis Sci 2009; 86: 76–78. myopia control: considering all the data. JAMA ridor of progressive addition lenses. Optom Vis Sci
123. Buehren T, Collins MJ, Carney LG. Near work induced Ophthalmol 2018; 136: 303. 2006; 83: 666–671.
wavefront aberrations in myopia. Vision Res 2005; 148. Gettes BC. Drugs in refraction. Int Ophthalmol Clin 168. Cho P, Cheung SW, Edwards M. The longitudinal
45: 1297–1312. 1961; 1: 237–248. orthokeratology research in children (LORIC) in Hong
124. Li YJ, Choi JA, Kim H et al. Changes in ocular 149. McBrien NA, Stell WK, Carr B. How does atropine Kong: a pilot study on refractive changes and myopic
wavefront aberrations and retinal image quality with exert its anti-myopia effects? Ophthalmic Physiol Opt control. Curr Eye Res 2005; 30: 71–80.
objective accommodation. J Cataract Refract Surg 2013; 33: 373–378. 169. Cho P, Cheung SW. Protective role of ortho-
2011; 37: 835–841. 150. Carr BJ, Stell WK. Nitric oxide (NO) mediates the inhi- keratology in reducing risk of rapid axial elongation:
125. Iida Y, Shimizu K, Ito M et al. Influence of age on ocu- bition of form-deprivation myopia by atropine in a reanalysis of data from the ROMIO and TO-SEE
lar wavefront aberration changes with accommoda- chicks. Sci Rep 2016; 6: 9. studies. Invest Ophthalmol Vis Sci 2017; 58:
tion. J Refract Surg 2008; 24: 696–701. 151. Hiraoka T, Miyata K, Nakamura Y et al. Influences 1411–1416.
126. Cheng H, Barnett JK, Vilupuru AS et al. A population of cycloplegia with topical atropine on ocular 170. Lin HJ, Wan L, Tsai FJ et al. Overnight orthokeratology
study on changes in wave aberrations with accom- higher-order aberrations. Ophthalmology 2013; is comparable with atropine in controlling myopia.
modation. J Vis 2004; 4: 272–280. 120: 8–13. BMC Ophthalmol 2014; 14: 1–8.
127. Zhou XY, Wang L, Zhou XT et al. Wavefront aberra- 152. Chia A, Chua WH, Cheung YB et al. Atropine for the 171. Stillitano I, Chalita MR, Schor P et al. Corneal
tion changes caused by a gradient of increasing treatment of childhood myopia: safety and efficacy changes and wavefront analysis after ortho-
accommodation stimuli. Eye 2015; 29: 115–121. of 0.5%, 0.1%, and 0.01% doses (atropine for the keratology fitting test. Am J Ophthalmol 2007; 144:
128. He JC, Burns SA, Marcos S. Monochromatic aberra- treatment of myopia 2). Ophthalmology 2012; 119: 378–386.
tions in the accommodated human eye. Vision Res 347–354. 172. Gifford P, Li M, Lu H et al. Corneal versus ocular
2000; 40: 41–48. 153. Salmon TO, van de Pol C. Normal-eye zernike coeffi- aberrations after overnight orthokeratology. Optom
129. Ninomiya S, Fujikado T, Kuroda T et al. Changes of cients and root-mean-square wavefront errors. Vis Sci 2013; 90: 439–447.
ocular aberration with accommodation. J Cataract Refract Surg 2006; 32: 2064–2074. 173. Lian Y, Shen M, Huang S et al. Corneal reshaping
Am J Ophthalmol 2002; 134: 924–926. 154. Hiraoka T, Miyata K, Nakamura Y et al. Influence of and wavefront aberrations during overnight ortho-
130. Atchison DA, Collins MJ, Wildsoet CF et al. Measure- cycloplegia with topical cyclopentolate on higher- keratology. Eye Contact Lens 2014; 40: 161–168.
ment of monochromatic ocular aberrations of order aberrations in myopic children. Eye 2014; 28: 174. Stillitano I, Schor P, Lipener C et al. Long-term
human eyes as a function of accommodation by the 581–586. follow-up of orthokeratology corneal reshaping
Howland aberroscope technique. Vision Res 1995; 35: 155. Amirshekarizadeh N, Hashemi H, Jafarzadehpur E using wavefront aberrometry and contrast sensitiv-
313–323. et al. Higher-order aberrations after cyclopentolate, ity. Eye Contact Lens 2008; 34: 140–145.
84
175. Santodomingo-Rubido J, Villa-Collar C, Gilmartin B 182. Li SM, Kang MT, Wu SS et al. Studies using concentric one-year results. Invest Ophthalmol Vis Sci 2011; 52:
et al. Short- and long-term changes in corneal aber- ring bifocal and peripheral add multifocal contact 9362–9367.
rations and axial length induced by orthokeratology lenses to slow myopia progression in school-aged 190. Fujikado T, Ninomiya S, Kobayashi T et al. Effect
in children are not correlated. Eye Contact Lens 2016; children: a meta-analysis. Ophthalmic Physiol Opt of low-addition soft contact lenses with
43: 358–363. 2017; 37: 51–59. decentered optical design on myopia progression
176. Santodomingo-Rubido J, Villa-Collar C, Gilmartin B 183. Walline JJ, Jones LA, Mutti DO et al. A randomized in children: a pilot study. Clin Ophthalmol 2014; 8:
et al. The effects of entrance pupil centration and trial of the effects of rigid contact lenses on myopia 1947–1956.
coma aberrations on myopic progression follow- progression. Arch Ophthalmol 2004; 122: 1760–1766. 191. Cheng X, Xu J, Chehab K et al. Soft contact lenses
ing orthokeratology. Clin Exp Optom 2015; 98: 184. Walline JJ, Jones LA, Sinnott L et al. A randomized with positive spherical aberration for myopia con-
534–540. trial of the effect of soft contact lenses on myopia trol. Optom Vis Sci 2016; 93: 353–366.
177. Hiraoka T, Mihashi T, Okamoto C et al. Influence of progression in children. Invest Ophthalmol Vis Sci 192. Anstice NS, Phillips JR. Effect of dual-focus
induced decentered orthokeratology lens on ocular 2008; 49: 4702–4706. soft contact lens wear on axial myopia progres-
higher-order wavefront aberrations and contrast 185. Bakaraju RC, Ehrmann K, Ho A et al. Inherent ocular sion in children. Ophthalmology 2011; 118:
sensitivity function. J Cataract Refract Surg 2009; 35: spherical aberration and multifocal contact lens opti- 1152–1161.
1918–1926. cal performance. Optom Vis Sci 2010; 87: 1009–1022. 193. Sankaridurg P, Bakaraju RC, Naduvilath T et al.
178. Tarrant J, Liu Y, Wildsoet CF. Orthokeratology can 186. Fedtke C, Ehrmann K, Thomas V et al. Peripheral Myopia control with novel central and peripheral
decrease the accommodative lag in myopes. Invest refraction and aberration profiles with multifocal plus contact lenses and extended depth of focus
Ophthalmol Vis Sci 2009; 50: 4294. lenses. Optom Vis Sci 2017; 94: 876–885. contact lenses: 2 year results from a randomised
179. Mathur A, Atchison DA. Effect of orthokeratology on 187. Gong CR, Troilo D, Richdale K. Accommodation and clinical trial. Ophthalmic Physiol Opt 2019; 39:
peripheral aberrations of the eye. Optom Vis Sci phoria in children wearing multifocal contact lenses. 294–307.
2009; 86: E476–E484. Optom Vis Sci 2017; 94: 353–360. 194. Gislén A, Gustafsson J, Kröger RH. The accommoda-
180. Chen Z, Niu L, Xue F et al. Impact of pupil diameter 188. Kang P, Wildsoet CF. Acute and short-term changes in tive pupil responses of children and young adults at
on axial growth in orthokeratology. Optom Vis Sci visual function with multifocal soft contact lens wear in low and intermediate levels of ambient illumination.
2012; 89: 1636–1640. young adults. Cont Lens Anterior Eye 2016; 39: 133–140. Vision Res 2008; 48: 989–993.
181. Faria-Ribeiro M, Navarro R, González-Méijome JM. 189. Sankaridurg P, Holden B, Smith EL III et al. Decrease 195. Read SA, Collins MJ, Vincent SJ. Light exposure and
Effect of pupil size on wavefront refraction during in rate of myopia progression with a contact lens eye growth in childhood. Invest Ophthalmol Vis Sci
orthokeratology. Optom Vis Sci 2016; 93: 1399–1408. designed to reduce relative peripheral hyperopia: 2015; 56: 6779–6787.
85
INVITED REVIEW
Peripheral refraction and higher order aberrations
Dmitry Romashchenko* MSc Peripheral image quality influences several aspects of human vision. Apart from off-axis
Robert Rosén† PhD visual functions, the manipulation of peripheral optical errors is widely used in myopia con-
Linda Lundström* PhD trol interventions. This, together with recent technological advancements enabling the mea-
*Department of Applied Physics, Royal Institute of surement of peripheral errors, has inspired many studies concerning off-axis optical
Technology, Stockholm, Sweden aberrations. However, direct comparison between these studies is often not straightfor-
†
R&D, Johnson & Johnson Vision, Groningen, ward. To enable between-study comparisons and to summarise the current state of knowl-
The Netherlands edge, this review presents population data analysed using a consistent approach from
E-mail: [email protected] 16 studies on peripheral ocular optical quality (in total over 2,400 eyes). The presented data
include refractive errors and higher order monochromatic aberrations expressed as Zernike
This is an open access article under the terms of the
co-efficients (reported in a subset of the studies) over the horizontal visual field. Addition-
Creative Commons Attribution-NonCommercial
ally, modulation transfer functions, describing the monochromatic image quality, are calcu-
License, which permits use, distribution and
lated using individual wavefront data from three studies. The analysed data show that
reproduction in any medium, provided the original
work is properly cited and is not used for commercial optical errors increase with increasing eccentricity as expected from theoretical modelling.
purposes. Compared to emmetropes, myopes tend to have more hypermetropic relative peripheral
refraction over the horizontal field and worse image quality in the near-periphery of the
nasal visual field. The modulation transfer functions depend considerably on pupil shape
(for angles larger than 30 ) and to some extent, the number of Zernike terms included.
Moreover, modulation transfer functions calculated from the average Zernike co-efficients
Submitted: 28 March 2019 of a cohort are artificially inflated compared to the average of individual modulation trans-
Revised: 18 June 2019 fer functions from the same cohort. The data collated in this review are important for the
Accepted for publication: 18 June 2019 design of ocular corrections and the development and assessment of optical eye models.
Key words: myopia, ocular modulation transfer function, peripheral higher order aberrations, peripheral refraction, retinal image quality
This review summarises the results of ear- reported that mobility, including the risk of It has also been suggested that manipu-
lier studies on the peripheral optical errors falling, is also highly dependent on vision lating peripheral image quality might pre-
of the human eye. Knowledge of the periph- beyond the fovea.12,13 Further, limiting off- vent myopia onset or slow down its
eral optical quality is of importance to axis vision can affect the performance of progression.19–23 In recent years the preva-
several fields within optometry and ophthal- search tasks, where well-controlled saccadic lence of myopia has continued to increase
mology:1 development of technical aids with eye movements are required.14 and currently affects approximately 30 per
intact or improved perception and mobility; Knowledge of the peripheral retinal image cent of the population worldwide.19,21,24
correction of peripheral optical errors to quality can be useful for the development of This is of serious concern, because high
improve vision for various ocular diseases; optical aids for patients with reduced retinal myopia is a risk factor for severe ocular
and manipulation of peripheral image qual- functionality (for example, due to age-related pathologies (such as myopic macular degen-
ity to halt progressing myopia. macular degeneration), retinitis pigmentosa, eration25) and therefore, many research
Many activities in everyday life require and glaucoma. For instance, patients with studies have been dedicated to myopia con-
sufficient image quality on the peripheral central visual field loss have shown improved trol. Studies in chickens,26,27 monkeys28–31
retina. Unlike central vision, designed pri- visual performance with optical corrections and guinea pigs32 have shown that periph-
marily for resolution tasks, peripheral vision that enhance the image contrast on the eral image quality has the potential to drive
is responsible for various forms of detec- peripheral retina.15–17 Treating pseudophakic myopia development; but the entire mecha-
tion. Even though peripheral high-contrast patients can also be challenging since intra- nism as yet is not completely understood.
resolution is limited by the sampling density ocular lenses, currently available on the mar- Nevertheless, specific peripheral aberration
of the retina, both detection and low- ket, can decrease peripheral retinal image patterns are already implemented in myo-
contrast resolution depend on the quality of quality.18 Thus, explicit knowledge of the pia control methods through different types
the peripheral image.2–8 It has been demon- peripheral ocular aberrations and image of multifocality. However, all of the available
strated that peripheral vision is essential quality may be highly beneficial from a clini- optical treatments (including multifocal soft
for driving,9–11 and several studies have cal and research perspective. contact lenses, spectacles that alter peripheral
Clinical and Experimental Optometry 103.1 January 2020 © 2019 The Authors. Clinical and Experimental Optometry published by John Wiley & Sons Australia, Ltd
on behalf of Optometry Australia
86
Peripheral refraction and aberrations Romashchenko, Rosén and Lundström
Study Measurements Subjects Horizontal VF Used data Comments

technique
Lotmar and Retinoscopy 363, all emmetropes 20 , 40 , 60 in J0 J0 calculated from
Lotmar38 nasal and interval Sturm
temporal VF
Millodot39 Topcon refractor 62 subjects (13 (−60; +60) in 10 RPR, J0 J0 calculated from
emmetropes, 30 steps interval Sturm
myopes, 19
hypermetropes)
Mutti et al.40 Canon R-1 822 children aged 5 to Foveal and 30 RPR, J0 J0 calculated from
autorefractor 14 years nasal VF of the cylinder power
right eye
Gustafsson Double-pass 20 emmetropes, either (−60; +60) in 10 RPR, J0
et al.41 technique left or right eye steps
measured per subject
Seidemann PowerRefractor and 31 young adult subjects: 0 , 15 , (20 ), 30 , RPR, J0 J0 calculated from
et al.42 double-pass 8 emmetropic, 18 (40 ), 45 nasal VF interval Sturm
technique myopic, 5 hypermetorpic
Atchison Shin-Nippon 116 subjects, (−35; +35) in 5 RPR, J0 Polynomial fit to
et al.43 SRW-5000 emmetropes and steps the graphs
myopes
Shen et al.44 COAS 34 adult subjects: 8 (−30; +30) M, J0 Polynomial fit to
emmetropes, 26 myopes horizontal VF the graphs
*Lundström Laboratory 43 subjects 0 , 20 and 30 Zernike co-efficients J0 calculated from
et al.45 Hartmann-Schack nasal VF up to 9th order C22
wavefront sensor
*Mathur COAS-HD 19 subjects: 10 (−21; +21) Zernike co-efficients Polynomial fit to
et al.46 emmetropes, 10 myopes (colour map) up to 6th order colour maps
(raw data available for
20 subjects)
Baskaran COAS-HD VR 30 younger and 30 older (−40; +40) in 10 RPR, J0, C(4,0), C(1,3) J0 calculated from
et al.47 emmetropes steps C22
*Jaeken and Scanning wavefront 202 eyes of 101 subjects: (−40; +40) at 1 Zernike co-efficients J0 calculated from
Artal48 sensor 64 non-myopes and 37 intervals up to 3rd order C22
myopes
Bakaraju BHVI-EyeMapper 26 participants, (−50; 50) M, J0, C(1,3), C(3), M and J0 as
et al.53 emmetropes and horizontal VF C(4,0) polynomial fit to
myopes the graphs
Osuagwu COAS-HD 29 subjects, 19 (−20; +20) 2nd and 3rd order 1. Polynomial fit
et al.49 isomyopic (colour maps) Zernike co-efficients, to colour maps
(anisometropia < 1 D) C(4,0) 2. Only right eyes
data used
Osuagwu COAS-HD 49 young adults: 9 (−21; +21) RPR, J0, C(3,−3), C(3, 1. Polynomial fit
et al.50 hypermetropes, 20 (colour maps) −1), C(1,3), C(4,0) to colour maps
emmetropes, 20 myopes 2. Emmetropes:
SE (−0.5; +0.75) D
Philip COAS 678 adolescents: 176 Foveal and 30 M, J0, 3rd and 4th
et al.51 emmetropic, 96 myopic nasal and orders of Zernike
and 375 hypermetropic temporal VF co-efficients
Osuagwu COAS-HD 37 eyes: 18 Caucasians, (−21; +21) RPR, J0, C(3,−1), C Polynomial fit to
et al.52 19 East Asians (colour maps) (1,3), C(3), C(4,0) colour maps
The first group contains studies with only J0 and relative peripheral refraction available, the second group shows those containing
Zernike co-efficients, and the asterisk (*) marks studies for which raw wavefront data were available. The table contains only the
details relevant to the analysis over the horizontal visual field (VF) of this review. See Table S1 for full details of the studies.
M: mean sphere, RPR: relative peripheral refraction.
Table 1. List of studies from which the data were extracted
defocus, and orthokeratology) are only par- and progression of myopia need to be fur- retinal image quality is limited and the com-
tially effective and subject-dependent.33,34 ther investigated. parison between different studies is often not
This suggests that peripheral aberrations Despite the importance of peripheral aber- straightforward. Even though there are guide-
as well as their effect on the development rations, direct access to population data on lines for reporting ocular aberrations,35–37
© 2019 The Authors. Clinical and Experimental Optometry published by John Wiley & Sons Australia, Ltd Clinical and Experimental Optometry 103.1 January 2020
87
they do allow some freedom for data repre- • If the relative peripheral refraction over Relative Peripheral Refraction = MðθÞ−Mðθ = 0Þ,
sentation. One of the possible discrepancies the horizontal VF was not readily available where θ is the angle in horizontal VF (nega-
between studies is the difference in pupil size in the article, it was calculated from tive for temporal VF).
and shape (spherical or elliptical) over which Zernike co-efficients using the following • If astigmatism was not readily available as
peripheral Zernike co-efficients are calculated. formulas: the horizontal Jackson cross cylinder ( J0),
Confusion can also arise from different data
pffiffiffi pffiffiffi it was calculated using one of the follow-
types (for instance, J0/Cylinder/C(2,2) for astig- 4 3 0 12 5 0 ing methods (also see ‘Comments’ column
M= − c + c , [1]
matism), visualisation styles (table/chart/col- r2pupil 2 r2pupil 4 in Table 1):
our map) and the sign convention used to
encode the angles of the visual field.
This paper is therefore intended to pro-
vide a comprehensive overview of results Emmetropes Myopes
from previously published studies of periph-
eral ocular aberrations. An analysis of ocular 2 2
Relative peripheral refraction, D

modulation transfer functions (MTF) is also
presented, which is essential to estimate the
effect of the optical aberrations on central 0 0
and peripheral vision. The data presented in
this review have potential use in both
research and clinical applications, including -2 -2
the design of optical eye models and the
development of optical corrections.
-4 -4
Peripheral ocular
-6 -6
aberrations data
-40 -20 0 20 40 -40 -20 0 20 40
Peripheral ocular aberrations and their
effect on retinal image quality were Horizontal visual field angle, degrees
assessed using data from 16 articles, listed
in Table 1.38–53 For three studies, marked Hypermetropes
with an asterisk, wavefront data for each Millodot M, 1981
individual subject were generously shared 2 Mutti D et al., 2000
by the authors.45,46,48 The full list of articles Gustafsson J et al., 2001
considered for this review is provided in Seidemann A et al., 2002

Table S1. The final decision to include an Atchison D et al, 2006
0
article was governed by the following Lundström L et al., 2009
criteria: (1) available data in multiple eccen- Mathur A et al., 2009
tricities over the horizontal visual field (VF); Baskaran K et al., 2011
(2) wavefront data represented as a set of -2 Jaeken B & Artal P, 2011
Zernike co-efficients; and (3) the number of Osuagwu UL et al., 2016
participants in the study (at least 20). Osuagwu UL et al., 2017
All processing and analyses presented were -4 Shen J et al., 2018
conducted using the following guidelines. Philip K et al., 2018
• If the angular steps in data representations Average value
were denser than in the original publication, -6
available data was linearly interpolated.
Myopes
• No additional recalculations for the wave-
-40 -20 0 20 40 Shen J et al., 2018.
length were made. Defocus data were
unchanged from the original publications, Horizontal visual field angle, degrees Low myopia
assuming these measurements already Shen J et al., 2018.
compensated for any differences between Moderate and high myopia
the measurement wavelength and the vis-
ible spectrum wavelengths.
• The ocular wavefront measurements are Figure 1. Relative peripheral refraction in dioptres for emmetropes (top, left; 1,098
represented as standard Zernike co- subjects), myopes (top, right; 427 subjects; weighted average spherical equiva-
efficients for a 4 mm circular pupil so that lent = −3.17 0.98 D) and hypermetropes (bottom; 482 subjects; weighted average
comparison between different studies can spherical equivalent = +1.25 0.49 D). Negative visual angles correspond to the tem-
be made.1,35–37,45 poral visual field (nasal retina).
88
• from Zernike co-efficients: −2.90 1.10 D), and 20 per cent hyper- subjects. For more specific information, refer
metropes (weighted average spherical to the ‘Subjects’ and ‘Used data’ columns of
pffiffiffi pffiffiffiffiffiffi
2 6 2 6 10 2 equivalent +1.35 0.69 D) Table 1.
J0 = − c + c ; [2]
r 2pupil 2 r2pupil 4 • age range five to 58 years Figures 1–3 show the population average
• except one study (Bakaraju et al.53) no defocus (relative peripheral refraction), hori-
• using Sturm interval (taking half of the pupil dilation, cycloplegia or fogging zontal astigmatism ( J0), primary spherical
dioptric difference between the two line • ethnicity not reported, but the studies aberration, and horizontal coma across the
foci assuming J45 = 0); have been conducted in Europe, Northern horizontal VF. The weighted average curves,
• J0 from cylinder power assuming J45 = 0: America and Australia. represented by the thick lines, were calcu-
The number of subjects and the amount of lated for the areas where data from more
Cylinder available data vary among the included stud- than one study were available. The shaded
J0 = : [3] ies (refer to the figures captions). Therefore, areas show one standard deviation for
2
the sample size for each individual type of regions with data from at least three stud-
• Relative peripheral refraction data were analysis may differ from the total number of ies. In Figure 1, relative peripheral refraction
divided into three refractive groups:
myopes, emmetropes and hyper-
metropes. If not specified in the original
article, the classification was made using Lotmar W & Lotmar T, 1974
0 Millodot M, 1981
these refractive error intervals: foveal
Mutti D et al., 2000
refractive error ≤ −0.50 D for myopes;
Gustafsson J et al., 2001
−0.50 D < foveal refractive error < +0.50 D
Seidemann A et al., 2002
for emmetropes; and foveal refractive
-2 Atchison D et al, 2006
error ≥ +0.50 D for hypermetropes. Lundström L et al., 2009
• The population average optical errors,
J0, D
Mathur A et al., 2009

both in tables and figures, were calculated Baskaran K et al., 2011
taking the number of subjects into consid- Jaeken B & Artal P, 2012
eration (that is, weighted average). -4 Bakaraju R et al., 2016
The combined effect of ocular aberra- Osuagwu UL et al., 2016
tions on retinal image quality was assessed Osuagwu UL et al., 2017
by calculating the monochromatic MTFs Shen J et al., 2018
using all available Zernike co-efficients for -6 Philip K et al., 2018
each individual subject from three studies, Osuagwu UL et al., 2018
-60 -40 -20 0 20 40 60 Average value
marked with an asterisk in Table 1.45,46,48
The average MTF curves presented in this Horizontal visual field angle, degrees
review refer to the average of individual
MTFs (not MTFs derived from Zernike co- Figure 2. J0 in dioptres for all subjects. Sample size: 2,493 subjects. Negative visual
efficients averaged across a cohort of indi- angles correspond to the temporal visual field (nasal retina).
viduals). For the off-axis horizontal VF MTF
calculations, the elliptical shape of the pupil
was taken into account by scaling the hori- Horizontal coma
zontal radius of the pupil by cos(θ), where θ 0.4 Lundström L et al., 2009
Mathur A et al., 2009
C(3,1), μm
is the angle in the horizontal VF.54 The 0 Baskaran K et al., 2011

MTFs were represented and analysed as Jaeken B & Artal P, 2012
2-D functions obtained as an average of -0.4 Bakaraju R et al., 2016
the original MTF curves over all pupil Osuagwu UL et al., 2016
-0.8
-40 -20 0 20 40 Osuagwu UL et al., 2017
meridians. Philip K et al., 2018
Osuagwu UL et al., 2018
Spherical aberration Average value
Results 0.04
C(4,0), μm
0
The subject group in this review is the combi- -0.04
nation of those for the studies listed in -0.08
Table 1. Overall, it can be described as follows:
-0.12
• 2,492 phakic subjects, both male and -40 -20 0 20 40
female Horizontal visual field angle, degrees
• no reported ocular conditions or
surgeries Figure 3. Horizontal coma C(3,1) and primary spherical aberration C(4,0) in μm for all
• 60 per cent emmetropes, 20 per cent myopes subjects (for a 4 mm pupil diameter). Sample size: 1,045 subjects. Negative visual
(weighted average spherical equivalent angles correspond to the temporal visual field (nasal retina).
89
is represented separately for emmetropes 1

(top, left), myopes (top, right) and hyper- Fovea
metropes (bottom). Horizontal astigmatism 10° nasal visual field
(Figure 2) was not divided into subgroups 0.8 20° nasal visual field
because there was no correlation between 30° nasal visual field
off-axis astigmatism and central refractive
error. Figure 3 represents Zernike co- 0.6
MTF
efficients for primary horizontal coma (top)
and primary spherical aberration (bottom).
0.4
Population weighted average Zernike co-
efficients for the horizontal VF are listed in
Table 2. The values were obtained using all 0.2
studies from Table 1 with available
wavefront data.45–53 Calculations for each
angular position were made for a 4 mm 0
pupil diameter using the full extent of avail- 0 10 20 30
able wavefront data, that is all Zernike co- Spatial frequency, cycles/degrees
efficients and all angles reported. However,
it is important to mention that all of these
Angle, Spatial frequency, cycles/degrees
studies contained measurements for differ-
ent angular extents. degrees 5 10 15 20 25 35
The MTF curves for the emmetropic sub- 0 0.74 ± 0.18 0.50 ± 0.22 0.35 ± 0.20 0.25 ± 0.16 0.20 ± 0.14 0.13 ± 0.10
jects, obtained using the three studies mar- 10 0.63 ± 0.18 0.36 ± 0.19 0.25 ± 0.16 0.18 ± 0.13 0.14 ± 0.11 0.09 ± 0.08
ked in Table 1 with an asterisk,45,46,48 are
20 0.37 ± 0.15 0.18 ± 0.10 0.11 ± 0.07 0.08 ± 0.05 0.06 ± 0.04 0.04 ± 0.03
plotted in Figure 4 for four angles in the hor-
izontal VF. The calculations were carried out 30 0.24 ± 0.11 0.11 ± 0.07 0.07 ± 0.05 0.05 ± 0.04 0.04 ± 0.03 0.02 ± 0.02
for the following sample sizes: 84 subjects
for fovea, 71 subjects for 10 , 84 subjects for Figure 4. Average modulation transfer function (MTF), calculated from available Zernike
20 , and 74 subjects for 30 . The table below co-efficients,45,46,48 for emmetropes in four angles of the nasal visual field (shown as solid
the figure shows the average MTF value lines). The shaded areas represent the standard deviation at each eccentricity. Sample
standard deviation for six different spatial sizes: 84 subjects for fovea, 71 subjects for 10 , 84 subjects for 20 , and 74 subjects for 30
frequencies. As can be seen, the MTF mono- of the nasal visual field. The table at the bottom shows average standard deviation for
tonically decreases with the off-axis angle. each curve at spatial frequencies up to 35 cycles/degree.
Discussion peripheral refraction (hypermetropic, with the for the horizontal VF the refractive error
peripheral image behind the retina) due to the in the horizontal (tangential) meridian is
elongated shape of the eye (Figure 1). noticeably more negative than in the verti-
This analysis pools peripheral ocular aberra-
cal (sagittal) meridian. Thus, for the major-
tion data from a number of published stud-
Astigmatism ity of the horizontal VF, the vertical line
ies to summarise the current understanding
As predicted by Coddington’s equations,55 astig- focus is located more anterior to the
of optics and image quality across the hori-
matism ( J0) increases with increasing horizontal peripheral retina, whereas the horizontal
zontal VF in the human eye. All reviewed
off-axis angle (Figure 2), best described by a line focus is closer to the retina.
studies clearly show an increase in ocular
quadratic function. Thus, second order polyno-
optical errors with increasing off-axis angle,
mials can be fitted to the average curve in the Spherical aberration
consistent with optical theory.
figure (equation [4]; θ in degrees will give J0 in Both primary and higher order spherical
dioptres). This nature of peripheral astig- aberrations are present for on-axis as well
Defocus
matism also dictates that the vertical as for the off-axis object points (Table 2).
To be able to compare the peripheral spher-
astigmatism is rather small in the horizon- However, for most of the VF primary
ical equivalent between different refractive
tal VF (for 20 nasal VF: J0 = spherical aberration C(4,0) is dominant.
error groups, relative peripheral refraction
[−0.57 0.13] D, J45 = [0.06 0.07] D). Figure 3 (top) shows that, on average, pri-
is often used. The relative peripheral refrac-
With this in mind, Figure 2 illustrates that mary spherical aberration does not change
tion not only depends on the optical aberra-
tion field curvature (due to the oblique
incidence of light), but also on the ocular 8
shape. Therefore, both hypermetropes and >
< J0 = − 5:23 10 −4 θ2 + 5:05 10 −3 θ, θ ≤ 0 ðfitting error RMS = 0:037DÞ,
emmetropes on average have a negative rela- [4]
>
: J0 = − 3:17 10 −4 θ2 − 5:05 10 −3 θ, θ > 0, ðfitting error RMS = 0:059DÞ
tive peripheral refraction (myopic, with the
peripheral image in front of the retina),
whereas myopes tend to have positive relative
90
Zernike term Off-axis angle (negative angles correspond to temporal VF and positive to nasal VF)
−30 −20 −10 0 10 20 30
C(2,−2) +0.116 0.227 +0.035 0.136 0 0.116 −0.038 0.125 −0.057 0.128 −0.047 0.154 −0.126 0.251
C (2,2) +0.341 0.366 +0.030 0.239 −0.059 0.214 −0.030 0.223 +0.070 0.217 +0.296 0.349 +0.823 0.547
C(3,−3) +0.009 0.040 −0.009 0.040 −0.014 0.031 −0.019 0.044 −0.013 0.033 −0.007 0.037 +0.008 0.065
C(3,−1) +0.011 0.069 +0.015 0.046 0.010 0.039 +0.007 0.044 −0.003 0.036 +0.001 0.038 −0.001 0.050
C(3,1) +0.204 0.107 +0.108 0.066 +0.048 0.042 +0.006 0.041 −0.059 0.041 −0.125 0.076 −0.252 0.136
C(3,3) +0.037 0.042 +0.012 0.039 +0.009 0.031 0 0.036 −0.007 0.030 −0.018 0.405 −0.054 0.084
C(4,−4) +0.004 0.011 +0.002 0.016 0.003 0.009 +0.002 0.015 +0.001 0.008 +0.002 0.011 +0.007 0.022
C(4,−2) −0.002 0.012 −0.001 0.008 −0.001 0.005 0 0.009 −0.001 0.006 0 0.009 0 0.016
C(4,0) +0.014 0.026 +0.009 0.020 +0.010 0.018 +0.015 0.020 +0.017 0.018 +0.016 0.019 +0.013 0.029
C(4,2) +0.005 0.018 0 0.011 0 0.010 0 0.012 +0.001 0.010 +0.002 0.013 −0.001 0.027
C(4,4) +0.004 0.011 +0.001 0.010 +0.002 0.012 +0.001 0.013 +0.002 0.010 +0.003 0.012 −0.003 0.263
C(5,−5) 0 0.002 0 0.001 0 0.004 0 0.002 −0.001 0.005 −0.002 0.009
C(5,−3) 0 0.001 0 0.001 +0.001 0.004 0 0.001 −0.001 0.004 0 0.006
C(5,−1) 0 0.001 0 0.001 0 0.003 0 0.001 −0.002 0.004 −0.004 0.007
C(5,1) −0.001 0.003 −0.001 0.002 0 0.003 0 0.002 +0.001 0.005 +0.004 0.010
© 2019 The Authors. Clinical and Experimental Optometry published by John Wiley & Sons Australia, Ltd
C(5,3) 0 0.001 0 0.001 0 0.003 0 0.001 +0.001 0.003 +0.002 0.081
C(5,5) 0 0.002 0 0.002 −0.001 0.005 0 0.001 +0.001 0.005 +0.003 0.012
C(6,−6) 0 0.001 0 0.001 0 0.002 0 0.001 0 0.002 −0.002 0.004
C(6,−4) 0 0.001 0 0.001 0 0.001 0 0.001 0 0.001 0 0.004
C(6,−2) 0 0 0 0.001 0 0.001 0 0.001 0 0.003
C(6,0) 0 0.001 0 0.001 −0.001 0.003 0 0.001 0 0.003 0 0.006
C(6,2) 0 0.001 0 0.001 0 0.001 0 0.001 0 0.002 +0.001 0.004
C(6,4) 0 0.001 0 0.001 0 0.001 0 0.001 0 0.002 +0.001 0.004
C(6,6) 0 0.001 0 0.001 0 0.002 0 0.001 0 0.003 +0.001 0.004
Table 2. Population weighted average standard deviation of Zernike co-efficients (in micrometres for a 4 mm circular pupil) over the horizontal visual field
(VF) from the studies marked in Table 1 with an asterisk45,46,48
91
much over the horizontal VF; however, EFFECT OF PUPIL SHAPE ON MTF average peripheral optical errors are large
there is some variation throughout the For large off-axis angles the elliptical shape of compared to their intrasubject variation.
population. the pupil affects the appearance of the ocular
MTF curve. The difference in MTFs calculated Retinal image quality and
Horizontal coma using a 4 mm cosine-scaled elliptical pupil myopia
Foveal coma is usually small, and horizontal and a circular pupil becomes considerable for The connection between myopia development
coma is dominant for purely horizontal off- eccentricities of 30 and higher (standard and peripheral image quality in the human eye
axis angles. Primary horizontal coma shows deviation of difference in MTFs = 0.05 @ is not straightforward. Hoogerheide et al.56
a clear increase with increasing eccentricity two cycles/degree, for studies marked with once suggested that relative peripheral hyper-
(Figure 3, bottom) while the standard devia- asterisk in Table 1, 30 horizontal VF). metropia is a risk factor for myopia develop-
tion for the average curve remains relatively ment, but this conclusion was made without
low. A linear function can be fitted to the EFFECT OF AVERAGING METHODS ON MTF considering the change in ocular shape with
average curve in the figure (θ in degrees will In Figure 5 the average MTF and the MTF ocular growth.57 More recent studies show that
give the horizontal coma in μm for a 4 mm from average Zernike co-efficients are based relative peripheral hypermetropia is most likely
pupil diameter): on different mathematical approaches. The a consequence of myopia and not its precur-
calculation of the average MTF consists of sor.58,59 That is, relative peripheral refraction
C ð3,1ÞðθÞ = − 7:80 10 −3 θ −1:420 10 −2 : obtaining individual MTF curves calculated depends on the degree of myopia.43 This is
[5] separately for each set of Zernike co- also observed in the available raw data for
efficients of each subject, and then averag- 62 myopic subjects45,48 (Figure 6). However,
Using a third order polynomial would only ing these MTF curves. In contrast, the MTF substantial differences in relative peripheral
improve the fitting root-mean-square-error from average Zernike co-efficients implies refraction for different degrees of myopia start
by 0.016 μm (from 0.040 to 0.0249 μm). calculation of only one MTF curve from the appearing only at rather high eccentricities of
set of already-averaged Zernike co-effi- the VF (20 and higher). It should further be
Calculation of the ocular MTF cients. Figure 5 contains MTFs for four VF mentioned that, as suggested earlier, compari-
The central and peripheral MTFs in Figure 4 angles calculated with both described rou- son of relative peripheral refraction in the hori-
are calculated by averaging curves for all tines using available raw data for zontal VF is most representative beyond 40 of
pupil meridians. Because of off-axis astig- emmetropic subjects.45,46,48 For each angle, eccentricity.60
matism across the majority of the horizontal the MTF from average Zernike co-efficients Nevertheless, many reasonably effective
VF, objects with horizontal lines are associ- shows unrealistically high values. It is also myopia control interventions rely on manipu-
ated with better image quality than those worth noting that this difference is largest in lating the peripheral retinal image quality; the
with vertical lines. Therefore, the calculated the central VF and gradually decreases optical treatments with the highest efficacy
MTFs represent the average retinal image towards the periphery. This is because the are orthokeratology and multifocal soft
quality for a stimulus containing details with
all possible orientations. Apart from that,
the shape of the ocular MTF itself depends 1
Fovea. Average MTF
on several parameters as well as the
Fovea. Average co-efficients
method of calculation.
0.8 10° nasal visual field. Average MTF
10° nasal visual field. Average co-efficients
EFFECT OF NUMBER OF ZERNIKE TERMS
20° nasal visual field. Average MTF
ON MTF
0.6 20° nasal visual field. Average co-efficients
By definition, Zernike series have an infi-
30° nasal visual field. Average MTF
MTF
nite number of elements; in practice the

30° nasal visual field. Average co-efficients
decomposition of a wavefront is more lim-
0.4
ited. Although the residual error is gener-
ally small, in some cases it can have a
noticeable effect on the shape of the ocu- 0.2
lar MTF. In particular, accurate individual
MTFs require more Zernike terms than
population average curves. This can be 0
illustrated by comparing the MTFs (1) for 0 10 20 30
the full available extent of Zernike terms, Spatial frequency, cycles/degrees
and (2) for Zernike terms up to the third
order and primary spherical aberration. Figure 5. Average modulation transfer function (MTF) and MTF from average Zernike
While the average difference between co-efficients for four angular positions in the nasal visual field (VF) for the
these MTFs is close to zero, in individual emmetropic cohort with available data.45,46,48 Average MTF is calculated by averaging
cases it can be rather high (standard curves from individual Zernike data sets; MTF from average Zernike co-efficients is
deviation of difference in MTFs = 0.05 @ obtained by averaging individual Zernike data sets and subsequent MTF calculation.
20 cycles/degree, for studies marked with Sample sizes: 84 subjects for fovea, 71 subjects for 10 , 84 subjects for 20 , and 74 sub-
an asterisk in Table 1, fovea). jects for 30 of the nasal VF.
92
contact lenses.33,35 In orthokeratology, a Relative peripheral refraction as a function of foveal refractive error
reverse geometry rigid contact lens worn over-
Spherical equivalent [-0.50; -1.49] D (average -0.96 D)
night flattens the central cornea, which
0.6 Spherical equivalent [-1.50; -2.49] D (average -1.89 D)

decreases the overall optical power of the eye.
Spherical equivalent [-2.50; -3.49] D (average -2.85 D)
However, this flattening in conjunction with 0.4
Spherical equivalent ≤ -3.50 D (average -4.07 D)
relative steepening of the mid-peripheral cor-
nea also results in increased off-axis astigma- 0.2
tism and inverted coma.61,62 Currently
0
available multifocal soft contact lens correc-
tions impose a large depth of focus in the -0.2
periphery.63,64
In order to further develop these optical -0.4
myopia control interventions, it is important
to compare peripheral image quality -0.6
between myopes and emmetropes. Figure 7 -0.8
shows the average MTFs for emmetropes 0 5 10 15 20 25 30
and myopes with simulated central refrac- Nasal visual field, degrees
tive error correction: foveal defocus and pri-
mary astigmatisms were subtracted from Figure 6. Relative peripheral refraction as a function of foveal refractive error across
every individual set of Zernike co-efficients the nasal visual field. The curves are obtained using data from the three studies, mar-
(central and peripheral). As can be seen, the ked in Table 1 with an asterisk.45,46,48 Sample sizes: 19 subjects with spherical equiva-
average MTF for myopic subjects is gener- lent (−0.50; −1.49 D), 17 subjects with spherical equivalent (−1.50; −2.49 D),
ally lower than that of emmetropic subjects 17 subjects with spherical equivalent (−2.50; −3.49 D), nine subjects with spherical
over the horizontal VF. However, this differ- equivalent ≤ −3.50 D.
ence becomes less prominent with increas-
ing eccentricity and disappears at about 20
off-axis angle. Nonetheless, the results of
1
Figure 7 need to be interpreted with cau- Fovea. Emmetropes
tion, because there were cases with myopic Fovea. Myopes
subjects having better image quality than 10° nasal visual field. Emmetropes
emmetropic ones (standard deviation not 0.8
10° nasal visual field. Myopes
shown on the figure). 20° nasal visual field. Emmetropes
0.6 30° nasal visual field. Emmetropes
Conclusion
MTF
This paper summarises the findings of 0.4

16 recent publications on the peripheral
refractive errors and higher order aberra-
tions in the horizontal VF. The presented 0.2
data demonstrate an increase in aberrations
with off-axis angle, well predicted by the
optical aberrations theory.
0
Increasing amounts of the peripheral hori- 0 10 20 30
zontal astigmatism and coma lead to an
Spatial frequency, cycles/degrees
asymmetric profile of peripheral retinal
image quality. Furthermore, the horizontal
Figure 7. Average modulation transfer function (MTF) for emmetropes and myopes
meridian (vertical line focus) is more myopic
(with ideal central refractive correction) from the three studies marked in Table 1
than the vertical meridian (horizontal line
with an asterisk45,46,48 for four angular positions in the nasal visual field (0 , 10 , 20 ,
focus). Comparison between the different
and 30 ). Sample sizes: 84 emmetropes and 72 myopes for fovea; 71 emmetropes and
refractive groups shows that relative periph-
47 myopes for 10 ; 84 emmetropes and 72 myopes for 20 ; 74 emmetropes and
eral refraction is positive for myopic sub-
62 myopes for 30 . Weighted average spherical equivalent of the whole myopic group
jects while being negative for emmetropic
−2.41 0.55 D.
and hypermetropic subjects.
Additionally, with ideal foveal refractive
correction, myopes tend to have worse of Zernike terms used (mostly in the fovea) from average Zernike co-efficients with
MTFs than emmetropes; this effect is less and the shape of the pupil (at angles ≥ 30 ). the latter demonstrating artificially high
prominent for high eccentricities. The shape A considerable difference was also found retinal image quality, especially in the
of MTF curve itself depends on the number between the average MTF and the MTF central VF.
93
ACKNOWLEDGEMENTS 21. Holden B, Sankaridurg P, Smith E et al. Myopia, an 45. Lundström L, Gustafsson J, Unsbo P. Population distri-
underrated global challenge to vision: where the current bution of wavefront aberrations in the peripheral
The authors would like to gratefully acknowl- data takes us on myopia control. Eye 2014; 28: 142–146. human eye. J Opt Soc Am A 2009; 26: 2192–2198.
edge Professor David Atchison, Doctor 22. Goldschmidt E, Jacobsen N. Genetic and environmen- 46. Mathur A, Atchison DA, Charman WN. Myopia and
Karthikeyan Baskaran, Professor Pablo Artal, tal effects on myopia development and progression. peripheral ocular aberrations. J Vis 2009; 9: 1–12.
Eye 2014; 28: 126–133. 47. Baskaran K, Unsbo P, Gustafsson J. Influence of age
and Doctor Bart Jaeken for providing data essen- 23. Wallman J, Winawer J. Homeostasis of eye growth and on peripheral ocular aberrations. Optom Vis Sci 2011;
tial for this article. This review was supported by the question of myopia. Neuron 2004; 43: 447–468. 88: 1088–1098.
24. Williams KM, Bertelsen G, Cumberland P et al. Increas- 48. Jaeken B, Artal P. Optical quality of emmetropic and
the MyFUN project, that receives funding from
ing prevalence of myopia in Europe and the impact of myopic eyes in the periphery measured with high-
the European Union’s Horizon 2020 Research education. Ophthalmology 2015; 122: 1489–1497. angular resolution. Invest Ophthalmol Vis Sci 2012; 53:
and Innovation Programme under the Marie 25. Verkicharla PK, Ohno-Matsui K, Saw SM. Current and 3405–3413.
predicted demographics of high myopia and an 49. Osuagwu UL, Suheimat M, Atchison DA. Mirror sym-
Sklodowska-Curie grant agreement No 675137. update of its associated pathological changes. Oph- metry of peripheralmonochromatic aberrations in fel-
thalmic Physiol Opt 2015; 35: 465–475. low eyes of isomyopes and anisomyopes. Invest
REFERENCES 26. Schaeffel F, Glasser A, Howland HC. Accommodation, Ophthalmol Vis Sci 2016; 57: 3422–3428.
1. Lundström L, Rosén R. Peripheral aberrations. In: refractive error and eye growth in chickens. Vision Res 50. Osuagwu UL, Suheimat M, Atchison DA. Peripheral
Artal P, ed. Handbook of Visual Optics, Vol. I, 1st 1988; 28: 639–657. aberrations in adult hyperopes, emmetropes and
ed. Boca Raton, FL: CRC Press, 2017. pp. 313–335. 27. Irving EL, Sivak JG, Callender MG. Refractive plasticity myopes. Ophthalmic Physiol Opt 2017; 37: 151–159.
2. Lundström L, Manzanera S, Prieto PM et al. Effect of of the developing chick eye: a summary and update. 51. Philip K, Sankaridurg PR, Ale JB et al. Profile of off-axis
optical correction and remaining aberrations on Ophthalmic Physiol Opt 2015; 35: 600–606. higher order aberrations and its variation with time
peripheral resolution acuity in the human eye. Opt 28. Hung L-F, Crawford MLJ, Smith EL. Spectacle lenses among various refractive error groups. Vision Res
Express 2007; 15: 12654–12661. alter eye growth and the refractive status of young 2018; 153: 111–123.
3. Thibos LN, Cheney FE, Walsh DJ. Retinal limits to the monkeys. Nat Med 1995; 1: 761–765. 52. Osuagwu UL, Verkicharla P, Suheimat M et al. Peripheral
detection and resolution of gratings. J Opt Soc Am A 29. Smith EL 3rd, Kee C-S, Ramamirtham R et al. Periph- monochromatic aberrations in young adult Caucasian
1987; 4: 1524–1529. eral vision can influence eye growth and refractive and East Asian eyes. Optom Vis Sci 2018; 95: 234–238.
4. Thibos LN, Still DL, Bradley A. Characterization of spa- development in infant monkeys. Invest Ophthalmol Vis 53. Bakaraju RC, Fedtke C, Ehrmann K et al. Peripheral
tial aliasing and contrast sensitivity in peripheral Sci 2005; 46: 3965–3972. refraction and higher-order aberrations with
vision. Vision Res 1996; 36: 249–258. 30. Benavente-Pérez A, Nour A, Troilo D. Axial eye growth cycloplegia and fogging lenses using the BHVI-
5. Williams DR, Coletta NJ. Cone spacing and the visual and refractive error development can be modified by EyeMapper. J Optom 2016; 9: 5–12.
resolution limit. J Opt Soc Am A 1987; 4: 1514–1523. exposing the peripheral retina to relative myopic or 54. Mathur A, Gehrmann J, Atchison DA. Pupil shape as
6. Wang YZ, Thibos LN, Bradley A. Effects of refractive error hyperopic defocus. Invest Ophthalmol Vis Sci 2014; 55: viewed along the horizontal visual field. J Vis 2013;
on detection acuity and resolution acuity in peripheral 6765–6773. 13: 1–8.
vision. Invest Ophthalmol Vis Sci 1997; 38: 2134–2143. 31. Whatham AR, Judge SJ. Compensatory changes in eye 55. Freeman MH, Hull CC. Optics, 11th ed. Oxford: But-
7. Rosén R, Lundström L, Unsbo P. Influence of optical growth and refraction induced by daily wear of soft contact terworth-Heinemann, 2003. pp. 250–251.
defocus on peripheral vision. Invest Ophthalmol Vis Sci lenses in young marmosets. Vision Res 2001; 41: 267–273. 56. Hoogerheide J, Rempt F, Hoogenboom WP. Acquired
2011; 52: 318–323. 32. McFadden SA, Howlett MHC, Mertz JR. Retinoic acid myopia in young pilots. Ophthalmologica 1971; 163:
8. Venkataraman AP, Papadogiannis P, Romashchenko D signals the direction of ocular elongation in the 209–215.
et al. Peripheral resolution and contrast sensitivity: Guinea pig eye. Vision Res 2004; 44: 643–653. 57. Rosén R, Lundström L, Unsbo P et al. Have we mis-
effects of monochromatic and chromatic aberrations. 33. Huang J, Wen D, Wang Q et al. Efficacy comparison of interpreted the study of Hoogerheide et al. (1971)?
J Opt Soc Am A 2019; 36: B52–B57. 16 interventions for myopia control in children: a net- Optom Vis Sci 2012; 89: 1235–1237.
9. Owsley C, McGwin G. Vision impairment and driving. work meta-analysis. Ophthalmology 2016; 123: 58. Atchison DA, Li SM, Li H et al. Relative peripheral
Surv Ophthalmol 1999; 43: 535–550. 697–708. hyperopia does not predict development and pro-
10. Wetton MA, Horswill MS, Hatherly C et al. The devel- 34. Smith EL III, Campbell MCW, Irving E. Does peripheral gression of myopia in children. Invest Ophthalmol Vis
opment and validation of two complementary mea- retinal input explain the promising myopia control Sci 2015; 56: 6162–6170.
sures of drivers’ hazard perception ability. Accid Anal effects of corneal reshaping therapy (CRT or ortho- 59. Mutti DO, Sinnott LT, Mitchell GL et al. Relative periph-
Prev 2010; 42: 1232–1239. K) & multifocal soft contact lenses? Ophthalmic Physiol eral refractive error and the risk of onset and pro-
11. Wood JM, Troutbeck R. Effect of restriction of the bin- Opt 2013; 33: 379–384. gression of myopia in children. Invest Ophthalmol Vis
ocular visual field on driving performance. Ophthalmic 35. American National Standard for Ophthalmics: Methods Sci 2011; 52: 199–205.
Physiol Opt 1992; 12: 291–298. for Reporting Optical Aberrations of Eyes; ANSI 60. Mathur A, Atchison DA. Peripheral refraction patterns
12. Patino CM, mcKean-Cowdin R, Azen SP et al. Central and Z80.28–2004. Merrifield, VA: Optical Laboratory Asso- out to large field angles. Optom Vis Sci 2013; 90: 140–147.
peripheral visual impairment and the risk of falls and ciation, 2004. 61. Mathur A, Atchison D. Effect of orthokeratology on
falls with injury. Ophthalmology 2010; 117: 199–206.e1. 36. Ophthalmic Optics and Instruments – Reporting Aberra- peripheral aberrations of the eye. Optom Vis Sci 2009;
13. Berencsi A, Ishihara M, Imanaka K. The functional role tions of the Human Eye, ISO 24157:2008, 1st ed. Case 86: E476–E484.
of central and peripheral vision in the control of pos- postale 56 CH-1211 Geneva 20: International Organi- 62. Gonzalez-Meijome JM, Faria-Ribeiro MA, Lopes-
ture. Hum Mov Sci 2005; 24: 689–709. zation for Standardization, 2008. Ferreira DP et al. Changes in peripheral refractive pro-
14. Dakin S, Fiser J, Pasquale LR et al. Effects of peripheral 37. Thibos L, Applegate RA, Schwiegerling JT et al. Stan- file after orthokeratology for different degrees of
visual field loss on eye movements during visual sea- dards for reporting the optical aberrations of eyes. myopia. Curr Eye Res 2016; 41: 199–207.
rch. Front Psychol 2012; 3: 1–13. J Refract Surg 2002; 18: S652–S660. 63. Rosén R, Jaeken B, Lindskoog-Pettersson A et al. Eval-
15. Lewis P, Venkataraman AP, Lundström L. Contrast 38. Lotmar W, Lotmar T. Peripheral astigmatism in the uating the peripheral optical effect of multifocal con-
sensitivity in eyes with central scotoma. Optom Vis Sci human eye: experimental data and theoretical model tact lenses. Ophthalmic Physiol Opt 2012; 32: 527–534.
2018; 95: 354–361. predictions. J Opt Soc Am 1974; 64: 510–513. 64. Ji Q, Yoo Y-S, Alam H et al. Through-focus optical char-
16. Baskaran K, Rosén R, Lewis P et al. Benefit of adaptive 39. Millodot M. Effect of ametropia on peripheral refrac- acteristics of monofocal and bifocal soft contact
optics aberration correction at preferred retinal locus. tion. Am J Optom Physiol Opt 1981; 58: 691–695. lenses across the peripheral visual field. Ophthalmic
Optom Vis Sci 2012; 89: 1417–1423. 40. Mutti DO, Sholtz RI, Friedman NE et al. Peripheral Physiol Opt 2018; 38: 326–336.
17. Lundström L, Gustafsson J, Unsbo P. Vision evaluation refraction and ocular shape in children. Invest
of eccentric refractive correction. Optom Vis Sci 2007; Ophthalmol Vis Sci 2000; 41: 1022–1030.
84: 1046–1052. 41. Gustafsson J, Terenius E, Buchheister J et al. Periph-
18. Jaeken B, Mirabet S, Marín JM et al. Comparison of eral astigmatism in emmetropic eyes. Ophthalmic Phy-
siol Opt 2001; 21: 393–400.
the optical image quality in the periphery of phakic Supporting information
and pseudophakic eyes. Invest Ophthalmol Vis Sci 42. Seidemann A, Schaeffel F, Guirao A et al. Peripheral
2013; 54: 3594–3599. refractive errors in myopic, emmetropic, and hyper-
19. Holden BA, Mariotti SP, Kocur I et al. Impact of increas- opic young subjects. J Opt Soc Am A 2002; 19: Additional supporting information may be
ing prevalence of myopia and high myopia. In: Holden 2363–2373.
43. Atchison DA, Pritchard N, Schmid KL. Peripheral found in the online version of this article at
BA, ed. Report of the Joint World Health Organization -
Brien Holden Vision Institute Global Scientific Meeting on refraction along the horizontal and vertical visual the publisher’s website:
Myopia. Sydney: University of New South Wales, 2015. fields in myopia. Optom Vis Sci 2006; 46: 1450–1458.
20. Flitcroft DI. The complex interactions of retinal, optical 44. Shen J, Spors F, Egan D et al. Peripheral refraction and
image blur in four meridians in emmetropes and Table S1. List of studies on peripheral
and environmental factors in myopia aetiology. Prog
Retin Eye Res 2012; 31: 622–660. myopes. Clin Ophthalmol 2018; 12: 345–358. refraction and higher order aberrations.
94
INVITED REVIEW

Antonio J Del Aguila-Carrasco* PhD Modern methods of measuring the refractive state of the eye include wavefront sensors
Philip B Kruger† PhD OD which make it possible to monitor both static and dynamic changes of the ocular wavefront
Francisco Lara‡ PhD while the eye observes a target positioned at different distances away from the eye. In addi-
Norberto López-Gil‡ PhD tion to monitoring the ocular aberrations, wavefront refraction methods allow measure-
*Eye and Vision Research Group, School of Health ment of the accommodative response while viewing with the eye’s habitual chromatic and
Professions, University of Plymouth, Plymouth, UK monochromatic aberrations present, with these aberrations removed, and with specific
† aberrations added or removed. A large number of experiments describing the effects of
College of Optometry, The State University of
New York, New York, NY, USA accommodation on aberrations and vice versa are reviewed, pointing out the implications
‡
Vision Science Research Group (CiViUM), Instituto for fundamental questions related to the mechanism of accommodation.
Universitario de Investigación en Envejecimiento
(IUIE), University of Murcia, Murcia, Spain
E-mail: [email protected]
Submitted: 30 March 2019

Revised: 22 May 2019
Accepted for publication: 23 May 2019
Key words: aberrations, accommodation, dynamic accommodation
Accommodation can be thought of as a possibility of changing the vergence of the Besides the changes of ocular aberrations
natural adaptive optics mechanism to target (by changing the distance between due to the change in curvature of the exter-
improve the retinal image quality of the eye and the target, or by adding lenses), nal surfaces of the crystalline lens of the
objects placed at different distances. It was to stimulate the subject’s accommodation. eye,12 the ocular wavefront may also change
Thomas Young who demonstrated at the There are several commercially available due to:
beginning of the 19th century that the devices that can measure aberrations while • displacement and tilt of the lens13
change in refractive power of the eye is stimulating accommodation (for example, • pupil changes (accommodative miosis)13
due to the crystalline lens.1,2 Currently, it is • torsions on the eye globe produced by
irx3, COAS-HD, WASCA, iTrace) as well as
well known that there are no significant binocular convergence14
custom-built systems.10
changes in corneal power during • changes of the internal iso-indicial sur-
Figure 1 shows a schematic of the meth-
accommodation,3,4 and only small changes faces of the lens.15
odology typically used to measure ocular
have been observed in the sclera.5 In addi- The study of accommodation and its rela-
aberrations during accommodation in a
tion to this, Young realised that the refrac- tionship with aberrations can be carried out
static procedure. A Badal lens (not shown) is
tive power in the periphery of his pupil usually used so the target always subtends through two time domains: static and
was greater than in the centre, and when the same visual angle regardless of its opti- dynamic. The term static accommodation
he accommodated, the refractive power cal vergence.11 After each change in ver- refers to the steady state condition of
distribution was opposite.1,2 This was the gence the target remains static for some accommodation while viewing a stationary
first observation that proved that the time before the wavefront is measured to target at a fixed distance from the eye. But
spherical aberration (SA) of the eye chan- allow time for the subject to accommodate. accommodation is never really static,
ged its sign with accommodation. Step changes in vergence (0.5 D in Figure 1), instead fluctuating continuously over a small
Two centuries after Young’s discoveries, far point (FP), maximum vergence, and tar- range. These small microfluctuations6,7 of
the measurement of spherical and other get configuration (for example, mono- accommodation are a dynamic characteris-
aberrations of the accommodated eye can chromatic/polychromatic, spatial frequency tic of accommodation even under static
be performed in vivo using wavefront sen- content) vary depending on the study. For steady state conditions. Dynamic accommoda-
sors. As accommodation dynamically dynamic studies, the target vergence is usu- tion refers to the change in ocular focus that
changes,6,7 fast wavefront sensors, such as ally continuously modified, following a pre- occurs in response to changes in accommoda-
a Hartmann-Shack need to be used.8,9 The determined vergence function such as a tive demand, including sudden step changes
experimental system should include the sinusoidal or a random step function. from one target distance to another, sinusoidal
95

Aberrations and accommodation Del Aguila-Carrasco, Kruger, Lara et al.
wavefront analysis have been applied.26,27

All of them show smaller objective AA values
than the subjective AA obtained as the diop-
tric difference between the subjective far
and NPs. Three optical reasons have been
proposed to explain such differences.
Typically, subjective AA is measured after
correcting any distance ametropia and com-
puted as the inverse of the distance to the
NP with respect to the spectacle plane. How-
ever, using this reference plane without per-
forming the corresponding mathematical
correction overestimates subjective AA, espe-
cially in young myopic subjects.28
The metric chosen to calculate the subjec-
tive AA can cause a false accommodative
error. For instance, positive SA (typical in an
unaccommodated eye) can cause the objec-
tive measurement of the FP to be more
myopic than the subjective one,29,30 and as
a consequence an accommodative lead will
be observed (Figure 2). On the other hand,
negative SA (typical in the accommodating
Figure 1. Schematic of the methodology for measuring aberrations during accommo- eye), can result in a smaller objective maxi-
dation. In this example, the stimulus (S) is initially placed 0.50 D beyond the subjec- mum accommodation than observed with
tive far point (FP) (FP +0.50 D), where a wavefront A is measured. Then, it can be the subjective method, which translates to
moved to the FP, where wavefront B is now obtained. The same procedure is an apparent accommodative lag12,29,31,32
repeated (wavefront C) until the stimulus vergence reaches the maximum vergence (Figure 2).
to be measured corresponding in this case to 10.00 D closer than the FP (FP −10.00 D), It has been demonstrated that the eye
giving the wavefront D. To cover all the intervals of accommodation it is assumed uses its depth of field both in far and near
that the largest vergence (10.00 D) is closer than the subject’s near point. vision in order to increase the subjective
AA.33 In addition to the limitation imposed by
photoreceptor sampling and photonic noise,
changes, and unpredictable sum-of-sines This review examines the relationship depth of field occurs because of the pres-
changes in target distance. Finally, dynamic between accommodation and ocular aberra- ence of HOAs when the pupil is larger during
accommodation also refers to the ongoing tions in detail. Given the differences in relaxed accommodation,24 and as a conse-
microfluctuations of accommodation.6,7 methodologies and the different types of quence of the accommodative miosis.34
Knowledge of how aberrations vary with aberrations considered by different authors,
static accommodation provides information this manuscript treats static and dynamic Monochromatic aberrations and
about the shape of the surface of the lens12 accommodation, and the effect of mono- static accommodation
as well as information about its internal chromatic and chromatic aberrations During accommodation, not only is the
structure.15 Dynamic accommodation stud- separately. defocus term modified, but other monochro-
ies usually shed light on fundamental ques- matic aberrations vary too. The change in
tions such as which cues trigger the monochromatic aberrations during accommo-
accommodation system to accurately The influence of aberrations on dation has been extensively studied.29,35–38 In
change the power of the lens and accom- the subjective and objective general, all monochromatic aberrations change
modate in the right direction,10,16–19 which amplitude of accommodation with accommodation; however, this change is
is of particular interest concerning myopia The amplitude of accommodation (AA) can generally small and subject-dependant.37
development.20–22 From an applied science be measured objectively as the dioptric The change in astigmatism is generally
perspective, knowledge of how aberrations change between the FP and the near point small,39 although there are some exceptions
change with accommodation can lead to (NP). However, the eye does not present a where the magnitude and axis vary signifi-
improved designs of multifocal and accom- constant refractive power across the whole cantly with accommodation.40,41 Changes in
modative intraocular lenses, which imitate pupil due to astigmatism and other higher- astigmatism with accommodation may be
the profile of ocular aberrations during order aberrations (HOAs), and theoretically due to an increase in lens tilt caused by the
accommodation. Knowing the effect of aber- numerous FPs and NPs exist depending on combined effects of a slacker zonular ten-
rations on accommodation can also lead to the region of interest examined within the sion and gravity.42 Astigmatism can also
new contact and intraocular lens designs pupil. Therefore, HOAs influence the AA. A change with accommodative miosis in the
with customised aberration profiles that number of objective methods (metrics) for presence of HOA, although this potential
extend the depth of field.23–25 determining accommodation or AA from explanation has not been verified to date.
96

hypothesis that the change of SA during

accommodation may play a role in myopia
development. Their explanation is based on
the fact that the combination of negative SA
(typical in the accommodated eye) with nega-
tive defocus (hyperopic image, or lag of
accommodation) increases visual detection
of the letters although it reduces image con-
trast, which may promote growth of the eye.
Chromatic aberration and static

accommodation
In a non-cycloplegic eye, even when the tar-
get vergence is kept constant, the level of
accommodation fluctuates continuously
over a small range of approximately 0.50 D
at temporal frequencies ranging up to a few
cycles per second.6,7 Chromatic dispersion of
light by the optical components of the
eye47–49 results in retinal images of polychro-
matic objects with subtle colour fringes at
the edges that reliably indicate whether the
Figure 2. Typical accommodative response. For an accommodative demand of 0 D, image is focused behind or in front of the
that is, when the stimulus is at the far point (FP) accommodation of the eye should retina.47–49 These colour fringes change sub-
be relaxed, but usually presents an accommodative lead. For vergences larger than stantially when the eye changes focus
2.00 D, the eye typically presents an accommodative lag. Objective amplitude of (Figure 4). When red light is focused on the
accommodation is found as the dioptric range between the minimum and the maxi- retina, blue light is focused in front of the
mum accommodation response. retina, and a fuzzy blue colour fringe is
formed at the image edge, so under-
accommodation (hyperopic defocus) is
Third-order aberrations (that is, coma and during accommodation; or it may have a characterised by a red colour fringe, while
trefoil) may also vary during accommodation, large positive value of SA which decreases over-accommodation (myopic defocus)
but not systematically,36,37 and in many eyes during accommodation but never becomes results in a blue colour fringe. These colour
these aberrations remain relatively stable over negative. But in any case, SA decreases with cues provide reliable directional signals for
the range of accommodation demands.35,42,43 accommodation for a fixed pupil size. accommodation.50–54
In the case of fourth-order SA, there is There are no other systematic changes in Fincham50 was the first investigator to
agreement between numerous studies about any HOA except sixth-order SA, which remove the effects of chromatic aberration
its well-defined trend, becoming less positive increases during accommodation.12,44 How- by using monochromatic light and by plac-
(or more negative) with increasing accommo- ever, the values of that aberration are usu- ing a specially designed achromatising lens
dation.29,35,36 As mentioned earlier, this was ally very small, and in many cases fall below in front of the eye. He used a coincidence
originally discovered by Young,1,2 although the experimental errors. optometer to measure accommodation
he did not give it the name of SA. After There are a few studies that have shown while trial lenses were placed in front of the
Young, many others reported this how some aberrations influence static subject’s eye and found that accommoda-
change,29,35,36 which has been proven to be accommodation. In particular, Khosravi45 tion was impaired in some subjects when
generated because of the hyperbolic shape showed that the accommodation response chromatic aberration was removed. By the mid-
of the surfaces of the crystalline lens.12 Usu- to a grating stimulus in the presence of astig- 1980s high-speed recording of accommodation
ally, in the relaxed eye corneal positive SA is matism depends on the orientation of the was available55 to test Fincham’s hypothesis that
larger than the absolute value of the crystal- grating, but for multiple orientations, the chromatic aberration provides a cue for static
line lens SA (negative value), so the total eye accommodation response usually corre- accommodation. Subjects viewed stationary tar-
has a slight positive SA. However, when the sponds with the circle of least confusion. A gets at 0 D, 2.5 D and 5 D in white and mono-
eye accommodates the crystalline lens different study used adaptive optics to study chromatic light, and in white light with
increases its SA negative value, and the total the effect of one micron of coma or fourth- chromatic aberration reversed by a specially
SA of the eye becomes negative (see order SA on the accommodation response, designed lens.52 When chromatic aberration
Figure 3). Therefore, generally speaking, the finding that those aberrations may increase was removed, some subjects had difficulty
relaxed eye has positive SA and the accom- the accommodation error, especially when accommodating and when chromatic aberra-
modated eye has negative SA. However, positive SA was induced.46 The effect of tion was reversed, so that blue light focused
there are exceptions to this rule. For fourth- and six-order SA on the accommoda- further back in the eye than red light,
instance, the eye may have negative SA when tion response has also been studied theoreti- accommodation was severely impaired, and
relaxed which becomes more negative cally by other researchers32 with the some subjects accommodated in the wrong
97

eliminate the normal blur feedback from

trial-and-error microfluctuations6,7 of
accommodation that were believed to be
essential for effective accommodation.
Accommodation responded readily to these
static simulations of LCA, and accommoda-
tion was not adversely affected by simula-
tions of LCA that included typical amounts
of TCA.
Some authors have argued that chromatic
aberration does not play a role in accommo-
dation because when an isoluminant target
is used (that is, a red target on a green back-
ground or vice versa, both with the same lumi-
nance), accommodation is not induced.56,57
However, this conclusion may not be valid58
since colour and luminance signals are mixed
in a single neural channel rather than separate
channels.59,60 Furthermore, it is well known
that many other visual functions fail under iso-
Figure 3. Example of the change of fourth-order spherical aberration with accommo- luminant target conditions, including form, col-
dation in a young subject with an amplitude of accommodation (AA) > 12.00 D. In the our, motion, and depth perception.59,61,62
relaxed eye the value is positive decreasing with accommodation and becoming nega- Further investigations are required in
tive. For large values of accommodation demand, spherical aberration tends to zero this field.
because the subject’s pupil becomes small. The magnitude of LCA depends on the
refractive index and dispersive power of the
ocular media. The crystalline lens of the eye
has a gradient refractive index structure
(GRIN) with maximum refractive index at the
centre and a minimum at the periphery.63–65
During accommodation it becomes more con-
vex, especially the anterior lens surface, and
there is also a change in the distribution of
the gradient refractive index that produces a
small increase in the equivalent refractive
index of the whole lens. The increase in the
equivalent refractive index is approximately
0.0013 per dioptre of accommodation.63–65
This is accompanied by a small increase of the
chromatic aberration of the eye amounting to
approximately three per cent per dioptre of
accommodation.49 Charman measured an
increase of approximately 0.2 dioptres of chro-
matic aberration between 422 nm and
633 nm when accommodating six dioptres.49
In another study, Jaskulski et al.66 studied
Figure 4. Ray diagrams illustrate under-accommodation (hyperopic defocus) on the the accommodation response to three target
top left side of the figure and over-accommodation (myopic defocus) on the top right vergences for three different wavelengths
side. In the presence of chromatic aberration, under-accommodation produces blur and white light, all having the same lumi-
spread-functions with a red colour fringe, whereas over-accommodation produces nance. They found a shift in refractive error
blur spread-functions with a blue colour fringe, as can be seen in the bottom row. for each colour condition corresponding to

Adapted from Del Aguila-Carrasco. 66
the defocus shift created by the LCA, but the
accommodation responses did not significantly
change. However, Kruger et al. found some
direction when chromatic aberration was chromatic aberration (LCA)53 or transverse subjects accommodated less accurately in
reversed. chromatic aberration (TCA)54 were used to monochromatic light when stationary targets
Next, computer-generated images that drive accommodation for near and far dis- were positioned significantly closer or further
simulated hyperopic and myopic defocus tances. These simulated images were away than the subject’s resting position of
with and without the effects of longitudinal viewed through small pinhole pupils to accommodation.52
98

Monochromatic aberrations and

dynamic accommodation
How does the visual system know when to
accommodate or disaccommodate and by
how much? Researchers have been trying to
answer this fundamental question for a long
time, and still there is not a completely satis-
factory answer. It is well known that the visual
system makes use of information from the
outside world, such as the intensity and wave-
length of light reflected from objects, as well
as information about the interaction of light
with the optics of the eye itself, such as the
effects of inaccurate refraction and chromatic
dispersion. This information that the visual
system uses in order to change the accom-
modation state accordingly is typically
referred to as ‘cues’ for accommodation.67 For
example, from the disparity between the two
signals, or images, formed by the two eyes,
the visual system is able to interpret depth,68
and depth perception guides accommoda-
tion.50,69 Nonetheless, most people are able
to accommodate correctly under monocular
conditions. The reason for this is that the
visual system can extract depth information
from monocular cues. Some of these monoc-
ular cues are apparent distance,70,71 changing
size,72–74 and interposition of objects.71 But
even when all these monocular cues that
allow the visual system to interpret depth are
removed, many people are still able to appro-
priately change their accommodation state. Figure 5. Ray diagrams illustrate under-accommodation (hyperopic defocus) on the top
How is this possible with the lack of external left side of the figure and over-accommodation (myopic defocus) on the top right side.
cues? In this case, the visual system uses In the presence of monochromatic aberrations, under-accommodation and over-
information extracted from the image formed accommodation produce different retinal images. Red arrows indicate some of the dif-
on the retina, or from the way light rays reach ferences between the images. The bottom row shows dynamic accommodation
the retina (optical cues for accommodation). It response for one subject while viewing a Maltese cross target in a Badal optical
is known that an out-of-focus retinal image of system moving sinusoidally toward and away from the eye at 0.2 Hz, oscillating between
a perfect eye without astigmatism and HOAs one and three dioptres (grey line) with natural aberrations present (blue line) and with
can trigger accommodation.75 However, there all aberrations corrected except for defocus (red line). Adapted from Del Aguila-
are other optical cues that are based on the Carrasco.55
fact that images formed at the retina differ if
they are focused in front (myopic defocus) or
experiment, the only way in which the eye changes in target vergence, and not changes
behind the retina (hyperopic defocus) (see
could accommodate was by trial-and-error, in blur alone. A similar experiment19 to that
upper part of Figure 5). Even-order mono-
or how Phillips and Stark referred to it, the of Phillips and Stark agreed somewhat with
chromatic aberrations, which generate differ-
eye was constantly ‘hunting’, searching for their results; nevertheless, when target blur
ent images for different signs of defocus16,76
the correct direction of accommodation. The was changed quickly, some participants’
may also play a role. Irregularly shaped
recorded responses were at times in the accommodation was worse or even absent.
pupils,16,77 and the Stiles-Crawford effect,78–80
wrong direction, and then changed rapidly An interesting conclusion of this work is that
can lead to different retinal images of the
toward the correct direction. However, their accommodation works much better when
object depending upon if they are formed in
main conclusion that blur alone drives changes in light vergence were present than
front of or behind the retina.16
One aberration that has always been accommodation seems too far-fetched from when there were only changes in target blur.
linked to accommodation has been spherical their measurement in a single subject who The majority of studies about the effect of
defocus. Phillips and Stark75 demonstrated usually responded in the wrong direction to monochromatic aberrations on dynamic
that blur alone could trigger accommodation a sudden change in target vergence. Recent accommodation have been carried out
with a remarkable experiment using a
work by Del Aguila-Carrasco et al.10 suggests recently, thanks to the development and
sophisticated system at the time. In their that accommodation responds to the actual implementation of adaptive optics (AO) in
99

aberration reduced the response for most

subjects, and reversing chromatic aberra-
tion so that red light focused further for-
ward in the eye than blue light severely
impaired the dynamic accommodative
response (Figure 6). Subjects accommo-
dated poorly to sinusoidally moving targets
in narrowband monochromatic light; their
response improved as the bandwidth of the
light increased, and the response was best
in broadband ‘white’ light.51,87,88
Using sinusoidally moving sine-wave grat-
ing targets, accommodation responded to
an intermediate band of spatial frequencies
between one and eight cycles/degree, with
peak sensitivity to the effects of chromatic
aberration between three and five
cycles/degree.89,90 Even very small amounts
of normal chromatic aberration (for exam-
Figure 6. Dynamic accommodation responses for two subjects while viewing a Mal- ple, 0.25 D) improved dynamic accommoda-
tese cross target in a Badal optical system moving sinusoidally toward and away from tion gain, while small amounts of chromatic
the eye at 0.2 Hz, oscillating between one and three dioptres (red line) with chromatic aberration in the reversed direction signifi-
aberration of the eye normal, neutralised, with monochromatic light and reversed cantly impaired the dynamic response.91 It
chromatic aberration. Accommodation (blue line) responded well with normal chro- was also established that both dynamic gain
matic aberration (first row); the response was reduced with chromatic aberration and the accuracy of static accommodation
neutralised by an achromatising lens (second row), and with monochromatic light were improved by the presence of chro-
(third row); and the response was severely impaired when chromatic aberration was matic aberration.52
reversed (fourth trace). Adapted from Kruger et al.58 All of these dynamic accommodation experi-
ments were performed under normal ‘closed-
vision.81,82 Using AO technology, some or all measuring the accommodation response in loop’ conditions where blur feedback from
the aberrations of the eye can be corrected, open-loop conditions (without feedback). small oscillations of accommodation was avail-
or different amounts of them can be induced able. But the presence of blur feedback can
Results from these simulation experiments
in real time. Since some of the ocular mask the true nature of the stimulus cue, and
suggest that the eye does not use monochro-
it was important to repeat these experiments
monochromatic aberrations change with matic aberrations to detect the sign of
under ‘open-loop’ conditions without blur feed-
accommodation,29,35–37 it is essential that defocus, since a large number of participants
back from oscillations of accommodation and
their correction is performed in real time. By did not respond to the simulations, and the
without trial-and-error changes in focus. Effec-
correcting particular monochromatic aberra- few who showed some response, could not tive dynamic accommodation responses with
tions and evaluating the accommodative properly follow the changes in blur.18 Never- high dynamic gains in the absence of blur
response of the eye, it is possible to assess theless, these studies were carried out on feedback confirmed that chromatic aberration
the effect of these aberrations on accommo- relatively small populations, thus larger sam- provides a highly reliable directional signal for
dation, if any. Recent studies manipulating
ple sizes need to be evaluated in order to dynamic accommodation.92
the eye’s natural aberrations suggest that the
draw firm conclusions. This series of dynamic accommodation
eye does not use monochromatic aberrations
experiments established that ratios of the
for accommodation,17,83–85 since no signifi-
Chromatic aberration and contrasts of the red, green and blue compo-
cant differences were found between the dynamic accommodation: the nents of the retinal image provide the optical
response with natural aberrations present, or chromatic cue signals that drive accommodation. Calcula-
corrected. In a recent experiment,17 the Fincham’s original findings50 were con- tions of the cone contrasts measured by
accommodative response of two out of eight firmed in monkeys86 and in a series of long- middle- and short-wavelength-sensitive
subjects seemed to increase slightly when experiments in humans in which the LCA of cones93 and empirical tests of this theory94
astigmatism was present while other mono- the eye was doubled, neutralised and proved that it was ratios of L-, M- and S-cone
chromatic aberrations were corrected. A dif- reversed58,73,74 while a Maltese cross target, contrasts that provide the directional signals
ferent approach has been used to elucidate viewed in a Badal optical system, moved that drive dynamic accommodation in two
whether certain monochromatic aberrations sinusoidally toward and away from the eye colour directions: red-green and blue-yellow.
do provide a cue for dynamic accommoda- at 0.2 Hz oscillating between one and three Another series of experiments showed
tion.18 The approach consists of blurring the dioptres of accommodative demand that isolated short-wavelength-sensitive
target computationally using different combi- (Figure 6). Doubling the amount of chro- cones (S-cones) drive dynamic accommoda-
nations of the subject’s own monochromatic matic aberration had no adverse effect on tion on their own, without any input from L-
aberrations together with defocus, and accommodation, neutralising chromatic cones or M-cones.95–97 In the first of these
100

experiments, accommodation was monitored more detailed studies about the optics of the image quality of objects located at different
continuously to a sine-wave grating target crystalline lens and its change during accom- distances, but also modifies its aberrations.
(three cycles/degree; 0.53 contrast) moving modation are needed, in particular, those Reciprocally, aberrations may influence the
with an unpredictable sum-of-sines motion in corresponding to the changes in its internal accommodation response, increasing, for
a Badal stimulus system under two experi- structure (iso-indicial surfaces) during accom- instance, the lag of accommodation. The most
mental conditions: a ‘blue’ condition (420 nm modation15 and their effects on the accom- significant change in HOA during accommoda-
blue grating +580 nm intense yellow homo- modation response. More detail about the tion is that experienced by fourth-order SA,
geneous adapting field) and a ‘white’ condi- shape of the back surface of the lens and its which decreases during accommodation, usu-
tion (broadband white grating). Mean change during accommodation are also ally changing its value from positive to nega-
dynamic gains for eight subjects were needed since current data are not precise tive, while chromatic aberration changes very
reduced by 50 per cent in the ‘blue’ condition enough. New imaging technology devices little during accommodation. Dynamic accom-
compared to the ‘white’ condition.95 Both S- based on optical coherence tomography modation studies have shown that monochro-
cones and LM-cones mediate static and probably combined with other wavefront matic aberrations do not seem to play a role
signed step accommodation responses to technologies will likely allow more accurate in accommodation. On the contrary, LCA pro-
changes in accommodation demand.96 S- determination of these types of lenticular vides a strong signed cue that reliably guides
cone contrast drives accommodation strongly changes in the near future. Further investiga- accommodation.
for near, resulting in significant over- tion into the change in monochromatic aber-
accommodation of more than one dioptre, rations during accommodation may lead to ACKNOWLEDGEMENTS
but the S-cone response is too slow to influ- improved designs of intraocular and contact Part of the studies presented have been per-
ence step dynamics when LM-cones partici- lenses to compensate for presbyopia. formed by some of the authors who want to
pate. The latencies and time constants for Another interesting area of research is to acknowledge their funding resources:
the accommodation response mediated by S- determine how the visual system is able to • European Research Council Starting Grant
cones alone to step changes in optical ver- detect the sign of defocus, and thus appropri- (ERC-2012-StG-309416-SACCO)
gence are two to three times longer than the ately accommodate. There are still many fun- • Universitat de València (UV-INV-PREDOC14-
latencies and time constants for accommoda- damental research studies to perform in this 179135)
tion mediated by LM-cones.96 Thus the slow regard. For example, it has not been investi- • European Research Council Marie Curie
accommodation response from S-cones actu- gated whether not having a perfectly circular ITN grant (“AGEYE” 608049)
ally reduces dynamic gain to sinusoidal target pupil is used by the visual system as a direc- • Ayudas para la realización de proyectos
motion at 0.2 Hz.97 The directional signal tional cue for accommodation. Moreover, in de investigación, Spain (grants: 05832/PI/07
from the chromatic mechanism that com- the last five years theoretical studies have and 15312/PI/10)
pares S- and LM-cone contrasts (S – [L + M]) been carried out to determine if the sign of • National Eye Institute of NIH (EYO5901).
cannot assist accommodation to sinusoidally defocus can be detected by particular struc-
moving targets.97 tures of the retinal anatomy.101,102 In particu-
REFERENCES
Finally, L-cones on their own and M-cones lar, Vohnsen et al. have carried out 1. Young T. The Bakerian Lecture on the mechanism of
on their own can mediate both static and computational simulations to show that there the eye. Philos Trans R Soc London 1801; 91: 23–88.
dynamic accommodation: L-cone contrast are different distributions of the electromag- 2. Atchison DA, Charman WN. Thomas Young’s contri-
butions to geometrical optics. Clin Exp Optom 2011;
reduces the mean accommodation level, netic field along the cone when light is focused 94: 333–340.
while M-cone contrast increases the mean either before or after the photoreceptor 3. He JC, Gwiazda J, Thorn F et al. Change in corneal
accommodation level.98 Mean accommoda- entrance plane, which may produce different shape and corneal wave-front aberrations with
accommodation. J Vis 2003; 3: 1.
tion level is decreased when L-cone contrast cone signals.101 López-Gil et al. have taken a 4. Sisó-Fuertes I, Domínguez-Vicent A, del Aguila-
is higher than M-cone contrast, and different geometric optics approach based on Carrasco A et al. Corneal changes with accommoda-
increased when M-cone contrast is higher different shadows that are cast by retinal ves- tion using dual Scheimpflug photography. J Cataract
Refract Surg 2015; 41: 981–989.
than L-cone contrast.98 In summary, L-cones sels in the peripheral retina when light is 5. Consejo A, Radhakrishnan H, Iskander DR. Scleral
reduce accommodation while both M-cones focused in front, on, or behind the blood vessel changes with accommodation. Ophthalmic Physiol
and S-cones increase accommodation.98,99 plane.102 Further experiments in humans Opt 2017; 37: 263–274.
6. Charman WN, Heron G. Fluctuations in accommoda-
The same chromatic cues, cone contrasts should be conducted to test these theoretical tion: a review. Ophthalmic Physiol Opt 1988; 8: 153–164.
and neural mechanisms that control every- hypotheses of optical vergence detection by 7. Charman WN, Heron G. Microfluctuations in accom-
day focusing of the human eye, also control the retina. The long-term goal of this funda- modation: an update on their characteristics and
possible role. Ophthalmic Physiol Opt 2015; 35:
long-term emmetropisation and developmental research is to extend what we have 476–499.
ment of myopia in animals.100 learned about cues for everyday accommoda- 8. Platt BC, Shack R. History and principles of Shack-
Hartmann wavefront sensing. J Refract Surg 2001; 17:
tion to the long-term focusing mechanism
S573–S577.
Future directions called emmetropisation, which operates to 9. Thibos LN, Hong X. Clinical applications of the Shack-
The interaction between aberrations and avoid the development of refractive errors. Hartmann aberrometer. Optom Vis Sci 1999; 76:
ocular accommodation has been studied 817–825.
10.
Del Aguila-Carrasco AJ, Marín-Franch I, Bernal-
extensively. Nevertheless, there are still a Molina P et al. Accommodation responds to optical
number of questions that need to be Conclusions vergence and not defocus blur alone. Invest
resolved and the possibilities for future Opthalmol Vis Sci 2017; 58: 1758.
11. Atchison DA, Bradley A, Thibos LN et al. Useful varia-
research on the topic are almost countless. Accommodation not only changes the refractions of the Badal optometer. Optom Vis Sci 1995; 72:
Some areas need further work. For instance, tive power of the eye to improve the retinal 279–284.
101

12. López-Gil N, Fernández-Sánchez V. The change of 35. He J, Burns S, Marcos S. Monochromatic aberrations interaction between color and luminance processing.
spherical aberration during accommodation and its in the accommodated human eye. Vision Res 2000; Vision Res 1992; 32: 1253–1262.
effect on the accommodation response. J Vis 2010; 40: 41–48. 61. Livingstone M, Hubel D. Segregation of form, color,
10: 12–12. 36. Atchison DA, Collins MJ, Wildsoet CF et al. Measure- movement, and depth: anatomy, physiology, and
13. El Hage S, Le Grand Y. Physiological optics. New York: ment of monochromatic ocular aberrations of perception. Science 1988; 240: 740–749.
Springer, 1980. human eyes as a function of accommodation by the 62. Troscianko T, Fahle M. Why do isoluminant stimuli
14. Fincham EF, Walton J. The reciprocal actions of Howland aberroscope technique. Vision Res 1995; 35: appear slower? J Opt Soc Am A 1988; 5: 871–880.
accommodation and convergence. J Physiol 1957; 313–323. 63. Dubbelman M, Van der Heijde GL, Weeber HA.
137: 488–508. 37. López-Gil N, Fernández-Sánchez V, Legras R et al. Change in shape of the aging human crystalline lens
15. Navarro R, López-Gil N. Impact of internal curvature Accommodation-related changes in monochro- with accommodation. Vision Res 2005; 45: 117–132.
gradient on the power and accommodation of the matic aberrations of the human eye as a function 64. Bahrami M, Heidari A, Pierscionek BK. Alteration in
crystalline lens. Optica 2017; 4: 334. of age. Invest Ophthalmol Vis Sci 2008; 49: refractive index profile during accommodation
16. López-Gil N, Rucker FJ, Stark LR et al. Effect of third- 1736–1743. based on mechanical modelling. Biomed Opt Express
order aberrations on dynamic accommodation. 38. Cheng H, Barnett JK, Vilupuru AS et al. A population 2016; 7: 99–110.
Vision Res 2007; 47: 755–765. study on changes in wave aberrations with 65. Khan A, Pope JM, Verkicharla PK et al. Change in
17. Bernal-Molina P, Marín-Franch I, Del Aguila- accomodation. J Vis 2004; 4: 3. human lens dimensions, lens refractive index distri-
Carrasco AJ et al. Human eyes do not need mono- 39. Liu T, Thibos LN. Variation of axial and oblique astig- bution and ciliary body ring diameter with accommo-
chromatic aberrations for dynamic accommodation. matism with accommodation across the visual field. dation. Biomed Opt Express 2018; 9: 1272–1282.
Ophthalmic Physiol Opt 2017; 37: 602–609. J Vis 2017; 17: 24. 66. Jaskulski M, Marín-Franch I, Bernal-Molina P et al.
18.
Esteve-Taboada JJ, Del Aguila-Carrasco AJ, Bernal- 40. Hofstetter HW. The correction of astigmatism for The effect of longitudinal chromatic aberration on
Molina P et al. Dynamic accommodation without near work. Optom Vis Sci 1945; 22: 121–134. the lag of accommodation and depth of field. Oph-
feedback does not respond to isolated blur cues. 41. Bannon RE. A study of astigmatism at the near point thalmic Physiol Opt 2016; 36: 657–663.
Vision Res 2017; 136: 50–56. with special reference to astigmatic accommodation. 67.
Del Aguila-Carrasco A. Light Vergence Detection in Mon-
Optom Vis Sci 1946; 23: 53–72. ocular and Monochromatic Accommodation (Monograph
19. Marín-Franch I, Del Aguila-Carrasco AJ, Bernal-
42. Radhakrishnan H, Charman WN. Changes in astigma- on the Internet). Valencia, Spain; 2017. Available from
Molina P et al. There is more to accommodation of
tism with accommodation. Ophthalmic Physiol Opt http://roderic.uv.es/handle/10550/59161.
the eye than simply minimizing retinal blur. Biomed
2007; 27: 275–280. 68. Qian N. Binocular disparity and the perception of
Opt Express 2017; 8: 4717–4728.
43. Ukai K, Ichihashi Y. Changes in ocular astigmatism depth. Neuron 1997; 18: 359–368.
20. Goss DA. Nearwork and myopia. Lancet 2000; 356:
over the whole range of accommodation. Optom Vis 69. Wheatstone CXVIII. Contributions to the physiology
1456–1457.
Sci 1991; 68: 813–818. of vision. —part the first. On some remarkable,
21. Chen JC, Schmid KL, Brown B. The autonomic control
44. Ninomiya S, Fujikado T, Kuroda T et al. Changes of and hitherto unobserved, phenomena of binocular
of accommodation and implications for human myo-
ocular aberration with accommodation. vision. Philos Trans R Soc London 1838; 128:
pia development: a review. Ophthalmic Physiol Opt
Am J Ophthalmol 2002; 134: 924–926. 371–394.
2003; 23: 401–422.
45. Khosravi B. The Influence of Ocular Astigmatism on 70. Ittelson WH, Ames A. Accommodation, convergence,
22. Weizhong L, Zhikuan Y, Wen L et al. A longitudi-
Accommodation (Monograph on the Internet). and their relation to apparent distance. J Psychol
nal study on the relationship between myopia
Manchester, UK, 1992. 1950; 30: 43–62.
development and near accommodation lag in
46. Gambra E, Sawides L, Dorronsoro C et al. Accommo- 71. Takeda T, Iida T, Fukui Y. Dynamic eye accommoda-
myopic children. Ophthalmic Physiol Opt 2008;
dative lag and fluctuations when optical aberrations tion evoked by apparent distances. Optom Vis Sci
28: 57–61.
are manipulated. J Vis 2009; 9: 1–15. 1990; 67: 450–455.
23. Benard Y, López-Gil N, Legras R. Subjective depth
47. Wald G, Griffin DR. The change in refractive power of 72. Kruger PB, Pola J. Changing target size is a stimulus
of field in presence of 4th-order and 6th-order
the human eye in dim and bright light. J Opt Soc Am for accommodation. J Opt Soc Am A 1985; 2:
Zernike spherical aberration using adaptive optics
1947; 37: 321–336. 1832–1835.
technology. J Cataract Refract Surg 2010; 36:
48. Bedford RE, Wyszecki G. Axial chromatic aberration 73. Kruger PB, Pola J. Stimuli for accommodation: blur,
2129–2138.
of the human eye. J Opt Soc Am 1957; 47: 564–565. chromatic aberration and size. Vision Res 1986; 26:
24. Benard Y, López-Gil N, Legras R. Optimizing the sub-
49. Charman WN, Tucker J. Accommodation and color. 957–971.
jective depth-of-focus with combinations of fourth-
J Opt Soc Am 1978; 68: 459–471. 74. Kruger PB, Pola J. Dioptric and non-dioptric stimuli
and sixth-order spherical aberration. Vision Res 2011;
50. Fincham EF. The accommodation reflex and its stim- for accommodation: target size alone and with blur
51: 2471–2477.
ulus. Br J Ophthalmol 1951; 35: 381–393. and chromatic aberration. Vision Res 1987; 27:
25. Yi F, Robert Iskander D, Collins M. Depth of focus
51. Aggarwala KR, Kruger ES, Mathews S et al. Spectral 555–567.
and visual acuity with primary and secondary spheri-
bandwidth and ocular accommodation. J Opt Soc Am 75. Phillips S, Stark L. Blur: a sufficient accommodative
cal aberration. Vision Res 2011; 51: 1648–1658.
A 1995; 12: 450–455. stimulus. Doc Ophthalmol 1977; 43: 65–89.
26. López-Gil N, Fernández-Sánchez V, Thibos LN et al.
52. Kruger PB, Aggarwala KR, Bean S et al. Accommoda- 76. Wilson BJ, Decker KE, Roorda A. Monochromatic
Objective amplitude of accommodation computed
tion to stationary and moving targets. Optom Vis Sci aberrations provide an odd-error cue to focus direc-
from optical quality metrics applied to wavefront
1997; 74: 505–510. tion. J Opt Soc Am A 2002; 19: 833.
outcomes. J Optom 2009; 2: 223–234.
53. Lee JH, Stark LR, Cohen S et al. Accommodation to 77. Wyatt HJ. The form of the human pupil. Vision Res
27. Tarrant J, Roorda A, Wildsoet CF. Determining the
static chromatic simulations of blurred retinal 1995; 35: 2021–2036.
accommodative response from wavefront aberra-
images. Ophthalmic Physiol Opt 1999; 19: 223–235. 78. Kruger PB, López-Gil N, Stark LR. Accommoda-
tions. J Vis 2010; 10: 4.
54. Stark LR, Lee RS, Kruger PB et al. Accommodation to tion and the Stiles-Crawford effect: theory and a
28. Bernal-Molina P, Vargas-Martín F, Thibos LN et al.
simulations of defocus and chromatic aberration in case study. Ophthalmic Physiol Opt 2001; 21:
Influence of Ametropia and its correction on mea-
the presence of chromatic misalignment. Vision Res 339–351.
surement of accommodation. Invest Opthalmol Vis
2002; 42: 1485–1498. 79. Kruger PB, Stark LR, Nguyen HN. Small foveal targets
Sci. 2016; 57: 3010.
55. Kruger PB. Infrared recording retinoscope for moni- for studies of accommodation and the Stiles–
29. Plainis S, Ginis HS, Pallikaris A. The effect of ocular
toring accomodation. Am J Optom Physiol Opt 1979; Crawford effect. Vision Res 2004; 44: 2757–2767.
aberrations on steady-state errors of accommoda-
56: 116–123. 80. Stark LR, Kruger PB, Rucker FJ et al. Potential signal
tive response. J Vis 2005; 5: 7.
56. Wolfe JM, Owens DA. Is accommodation colorblind? to accommodation from the Stiles-Crawford effect
30. Rocha KM, Vabre L, Chateau N et al. Expanding
Focusing chromatic contours. Perception 1981; 10: and ocular monochromatic aberrations. J Mod Opt
depth of focus by modifying higher-order aberra-
53–62. 2009; 56: 2203–2216.
tions induced by an adaptive optics visual simulator.
57. Switkes E, Bradley A, Schor C. Readily visible changes 81. Roorda A. Adaptive optics for studying visual func-
J Cataract Refract Surg 2009; 35: 1885–1892.
in color contrast are insufficient to stimulate accom- tion: a comprehensive review. J Vis 2011; 11: 6.
31. López-Gil N, Martin J, Liu T et al. Retinal image qual-
modation. Vision Res 1990; 30: 1367–1376. 82.
Marín-Franch I, Del Aguila-Carrasco AJ, Levecq X
ity during accommodation. Ophthalmic Physiol Opt
58. Kruger PB, Mathews S, Aggarwala KR et al. Chro- et al. Drifts in real-time partial wavefront correction
2013; 33: 497–507.
matic aberration and ocular focus: Fincham and how to avoid them. Appl Optics 2017; 56:
32. Thibos LN, Bradley A, López-Gil N. Modelling the
revisited. Vision Res 1993; 33: 1397–1411. 3989–3994.
impact of spherical aberration on accommodation.
59. Logothetis NK, Schiller PH, Charles ER et al. Percep- 83. Chen L, Kruger PB, Hofer H et al. Accommodation
Ophthalmic Physiol Opt 2013; 33: 482–496.
tual deficits and the activity of the color-opponent with higher-order monochromatic aberrations
33. Bernal-Molina P, Montés-Micó R, Legras R et al.
and broad-band pathways at isoluminance. Science corrected with adaptive optics. J Opt Soc Am A 2006;
Depth-of-field of the accommodating eye. Optom Vis
1990; 247: 214–217. 23: 1–8.
Sci 2014; 91: 1208–1214.
60. Gur M, Akri V. Isoluminant stimuli may not expose 84. Chin SS, Hampson KM, Mallen EAH. Role of ocular
34. Charman WN, Radhakrishnan H. Accommodation,
the full contribution of color to visual functioning: aberrations in dynamic accommodation control. Clin
pupil diameter and myopia. Ophthalmic Physiol Opt
spatial contrast sensitivity measurements indicate Exp Optom 2009; 92: 227–237.
2009; 29: 72–79.
102

85. Chin SS, Hampson KM, Mallen EAH. Effect of correc- 91. Kruger PB, Nowbotsing S, Aggarwala KR et al. Small 97. Kruger PB, Rucker FJ, Hu C et al. Accommodation with
tion of ocular aberration dynamics on the accommo- amounts of chromatic aberration influence dynamic and without short-wavelength-sensitive cones and
dation response to a sinusoidally moving stimulus. accommodation. Optom Vis Sci 1995; 72: 656–666. chromatic aberration. Vision Res 2005; 45: 1265–1274.
Opt Lett 2009; 34: 3274–3276. 92. Kruger PB, Mathews S, Katz M et al. Accommodation 98. Rucker FJ, Kruger PB. Accommodation responses to
86. Flitcroft DI, Judge SJ. The effect of stimulus chroma- without feedback suggests directional signals specify stimuli in cone contrast space. Vision Res 2004; 44:
ticity on ocular accommodation in the monkey. ocular focus. Vision Res 1997; 3: 2511–2526. 2931–2944.
J Physiol 1988; 398: 36. 93. Flitcroft DI. A neural and computational model for 99. Graef K, Schaeffel F. Control of accommodation by
87. Aggarwala KR, Nowbotsing S, Kruger PB. Accom- the chromatic control of accommodation. Vis Neu- longitudinal chromatic aberration and blue cones.
modation to monochromatic and white-light tar- rosci 1990; 5: 547–555. J Vis 2012; 12: 1–14.
gets. Invest Ophthalmol Vis Sci 1995; 36: 94. Kruger PB, Mathews S, Aggarwala KR et al. Accom- 100. Rucker FJ, Wallman J. Chick eyes compensate for
2695–2705. modation responds to changing contrast of long, chromatic simulations of hyperopic and myopic
88. Kotulak JC, Morse SE, Billock VA. Red-green middle and short spectral-waveband components defocus: evidence that the eye uses longitudinal
opponent channel mediation of control of of the retinal image. Vision Res 1995; 35: chromatic aberration to guide eye-growth. Vision Res
human ocular accommodation. J Physiol 1995; 2415–2429. 2009; 49: 1775–1783.
482: 697–703. 95. Rucker FJ, Kruger PB. Isolated short-wavelength sen- 101. Vohnsen B. Directional sensitivity of the retina: a lay-
89. Stone D, Mathews S, Kruger PB. Accommodation and sitive cones can mediate a reflex accommodation ered scattering model of outer-segment photorecep-
chromatic aberration: effect of spatial frequency. response. Vision Res 2001; 41: 911–922. tor pigments. Biomed Opt Express 2014; 5: 1569–1587.
Ophthalmic Physiol Opt 1993; 13: 244–252. 96. Rucker FJ, Kruger PB. The role of short-wavelength 102. López-Gil N, Jaskulski MT, Vargas-Martín F et al.
90. Mathews S, Kruger PB. Spatiotemporal transfer func- sensitive cones and chromatic aberration in the Retinal blood vessels may be used to detect the
tion of human accommodation. Vision Res 1994; 34: response to stationary and step accommodation sign of defocus. Invest Ophthalmol Vis Sci 2016; 57:
1965–1980. stimuli. Vision Res 2004; 44: 197–208. 3958.
103
INVITED REVIEW
Blur adaptation: clinical and refractive considerations
Matthew P Cufflin PhD The human visual system is amenable to a number of adaptive processes; one such pro-
Edward AH Mallen PhD cess, or collection of processes, is the adaptation to blur. Blur adaptation can be observed
School of Optometry and Vision Science, University of as an improvement in vision under degraded conditions, and these changes occur relatively
Bradford, Bradford, West Yorkshire, UK rapidly following exposure to blur. The potential important future directions of this research
E-mail: [email protected] area and the clinical implications of blur adaptation are discussed.
Submitted: 19 July 2019

Revised: 8 November 2019
Accepted for publication: 13 November 2019
Key words: adaptation, blur, defocus
Blur adaptation to navigate this area of vision research. We ADDING CONVEX LENSES
have not conducted a systematic review. Mon-Williams et al.1 added +1.00 D convex
lenses to the distance view of
Definition of blur adaptation
15 emmetropes. This allowed for a similar
Blur can be defined as a degradation of Blur adaptation using different level of myopic defocus to be applied across
image quality due to a number of influences methods of generating blur all participants, and for the adaptation
including optical aberrations, diffusion, and
effects to be investigated in non-myopes,
spatial filtering. For example, the introduc-
Refractive manipulation who had not been habitually exposed to
tion of myopic defocus diminishes the visual
REMOVING MYOPIC REFRACTIVE previous myopic defocus. While one dioptre
clarity of a scene and induces the awareness
CORRECTION of myopic defocus may initially reduce VA
of ‘blur’ in the observer. If the presence of
In response to anecdotal reports of vison by 0.35–0.40 logMAR units on an Early Treat-
blur persists, the visual system quickly
improving after periods of uncorrected ment Diabetic Retinopathy Study chart,6 the
begins to re-calibrate to these changes in an
vision, Pesudovs and Brennan2 removed the blur adaptation effect may result in a recov-
attempt to partially recover vision. Several ery of approximately 0.07–0.17 logMAR
optical correction from 10 myopic partici-
minutes later, the deleterious effects of blur pants for a period of 90 minutes. Seven of units over a 30- to 45-minute adaptation
(such as reduced high-contrast visual acuity the 10 observers displayed a small improve- period.1,7 Significantly increasing the levels
[VA]) imposed on vision will have lessened. ment in unaided acuity. The overall mean of myopic defocus up to +3.00 D yielded
This compensatory effect has been termed improvement was in the order of two let- only a moderate increase in the adaptive VA
blur adaptation, which is defined as an ters, which was statistically significant yet improvement to 0.20–0.26 logMAR units8 in
improvement in visual resolution, without clinically less significant and within the test– a group of myopes and emmetropes. The
alteration of refractive error or pupil size, retest variability (TRV) of high-contrast VA use of convex lenses to induce myopic
following exposure to a blur stimulus.1 charts.3 Rosenfield et al.4 used the same defocus in myopes and emmetropes gives
There are a number of visual adaptive pro- method but extended the adaptation period equal initial reduction in visual performance
cesses that help to improve function and to three hours, observing a much greater under conditions of cycloplegia.9
perceived image quality. Such adaptive pro- mean improvement in unaided acuity of During these experiments, participants
cesses are essential in overcoming the many 0.23 logMAR units. The disadvantage of this were instructed to view videos presented at
imperfections of the human eye. This article method of blur production is the lack of a close to optical infinity. This stimulus method
will review studies of blur adaptation in constant level of uncorrected lower-order provides an approximation of real-world
human vision with an emphasis on the clini- aberrations (myopic defocus and astigma- vision and the usual visual diet of spatial fre-
cal manifestations of this phenomenon. We tism) across all participants. In addition, nei- quencies, as well as minimising boredom.10
have used electronic databases and library ther study controlled for entrance pupil size, The importance of fully correcting any resid-
resources to identify relevant articles and which also impacts on the attenuating ability ual defocus and astigmatism prior to the
have attempted to group these elements in of retinal defocus on the modulation trans- addition of convex lenses should be noted,
a manner that we hope will help the reader fer function (MTF).5 as any residual refractive errors will lead to
104
Blur adaptation Cufflin and Mallen
inconsistencies in the level of blur experi- Marcos et al.18 for review) and has many function compared to optical blur. This is an
enced by participants. An under-corrected applications across the field of optics from important point to note when comparing
myope for example, may undergo blur adap- astronomy to retinal imaging. An AO system across studies that may use different
tation before the commencement of an comprises a device for measurement of the methods of generating blur.
experiment, thus limiting the measurable quality of an image, a method for the
effects of experimentally induced adaptation. manipulation of image quality (for example
a deformable mirror), and a control system. Quantifying the effects of blur
SPATIAL FREQUENCY MANIPULATION AO can be used to induce many different adaptation
The perception of blur may also be pro- types of blur, via the facility of, for example,
duced by manipulation of the spectral char- a deformable mirror to change individual As well as differing methods of blur produc-
acteristics of a stimulus.11 When viewing a aberration terms (for example defocus, tion, blur adaptation investigations have
natural scene the eye receives visual infor- astigmatism, spherical aberration, and coma utilised many methods to quantify this
mation coded across a range of spatial fre- aberrations) in isolation or in combination. change in blur perception both pre- and
quencies. For a natural scene there is a In the study of blur adaptation, AO can be post-adaptation. These methods include
greater low spatial frequency content, with used to present the visual system with a standard optometric acuity tests of letter-
the amplitude of each contributing fre- highly flexible array of stimuli, and has the based optotypes, grating acuity, as well as
quency believed to fall as frequency added capability of using the participant’s measures of contrast sensitivity and direct
increases.12 Consequently, a log amplitude natural aberrations to contribute to the measures of blur sensitivity.
versus log frequency plot of a natural scene stimulus. For example, the coma-type aber-
will have a gradient of approximately rations of the eye could be manipulated (for High-contrast VA
−1.00.13 Webster et al.14 have applied ampli- example rotated by some amount) and this High-contrast logMAR VA charts have been
fication factors to increase the gradient of used as a blur adapting stimulus. Thus, AO used widely in the measurement of blur
this graph, which will in turn accentuate the offers a whole range of experimental adaptation changes during and following
contribution of the lower spatial frequen- options for the study of the impact of blur exposure to defocus.1,2,4,7,22–25 VA with myo-
cies. Once the image is recombined, there on visual function. Sawides et al.19 applied pic defocus in place is measured to provide
will be an attenuation of high spatial fre- AO to study the perceived best focus of the an insight into the changes in resolution
quency information similar to that seen human visual system. It has been demon- that occur during blur adaptation. The
when myopic defocus is applied to an strated that the point of best focus across a logMAR acuity charts have an advantage
image.5 An increase in the negative slope of range of blur levels are biased toward the over Snellen charts in that the visual task
the power spectrum of −0.50 has an equiva- natural aberrations for a given individual. and crowding effects remain constant for all
lent effect on image clarity to the addition This gives evidence for the adaptation of the acuity levels.26 Table 1 summarises the
of 1.50 D of myopic defocus.15 human visual system to its own aberration blurred VA (BVA) improvements observed
pattern, thus maximising spatial vision following various durations of blur
USING SCATTER FILTERS performance. adaptation.
An alternative approach to the generation Prolonged exposure to myopic defocus
of blur is to apply a filter, as employed by INTRODUCING ASTIGMATIC BLUR has been shown to gradually improve
Vera-Diaz et al.16 This allows for the percep- The effects of blur adaptation produced by blurred VA by between 0.04 and 0.27
tion of blurred vision to be produced with- other lower-order aberrations have also logMAR units. The use of BVA to monitor
out altering the vergence of an been examined. Sawides et al.20 assessed changes in visual resolution has the advan-
accommodative target. Vera-Diaz et al.16 the impact of two minutes of adaptation to tage of speed and ease of performance and
employed a 0.2 Bangerter occlusion foil images simulating vertical or horizontal understanding for both the observer and
which reduced contrast by approximately astigmatic errors in focus. Adaptation to examiner. However, the measurement of
75 per cent and attenuated the MTF to defocus caused a shift in the perceived iso- BVA does have a major disadvantage,
50 per cent and 20 per cent of pre-blur tropic blurring of images and this aftereffect namely the variability of VA measurements
levels for spatial frequencies of ≥ 0.25 and transferred across a range of stimuli, such is known to increase in the presence of
≥ 1.25 cycles per degree (cpd) respectively. as faces and flowers. For example, adapta- blur.3,29 Carkeet et al.29 found that optical
The blur produced by Bangerter foils differs tion to vertical blurring caused an iso- defocus extended the probit interval, and
from that produced by optical defocus – for tropically blurred test image to appear thus reduced the accuracy of the endpoint
example, Bangerter foils generate monoton- horizontally blurred. This orientation-specific of high-contrast VA measurements. Simi-
ically increasing attenuation of high spatial aspect of adaptation suggests that larly, Rosser et al.3 found an increase in the
frequency content, whereas the attenuation observers adapt to uncorrected astigmatic TRV of VA measurements under conditions
effects of optical defocus on higher spatial errors, such as those present in the correc- of induced myopic defocus.
frequencies is more irregular.17 tion of low-level astigmats with spherical The time elapsed between the end of the
soft contact lenses. Ohlendorf et al.21 mea- adaptation task and the measurement of
EMPLOYING ADAPTIVE OPTICS TO sured the effect on vision of astigmatic blur visual function is small, with most studies
MANIPULATE HIGHER-ORDER generated optically (trial case lenses) versus reporting immediate measures of VA upon
ABERRATIONS blur generated by computer modelling completion of the adaptation period. In
Adaptive optics (AO) is a versatile technique (ZEMAX). The computer-generated blur had addition, these studies did not employ top-
for the manipulation of image quality (see a greater deleterious effect on visual up periods of adaptation during the
105
Study Subject refractive status Level of defocus Duration of Blurred acuity

adaptation improvement
Mon-Williams et al.1 Emmetropes +1.00 D 30 minutes RE 0.12 logMAR
LE 0.09 logMAR
BE 0.09 logMAR
Emmetropes under cycloplegia +2.00 D 30 minutes 0.26 logMAR
George and Rosenfield10 Emmetropes +2.50 D Two hours 0.13 logMAR
Myopes +2.50 D Two hours 0.27 logMAR
Wang et al.22 Myopes +2.50 D One hour 0.16 logMAR
Pesudovs and Brennan2 Uncorrected myopes 0.25 to 2.00 D 90 minutes 0.04 logMAR
Rosenfield et al.27 Uncorrected myopes 1.50 to 3.00 D Three hours 0.21 logMAR
Rosenfield et al.4 Uncorrected myopes 1.00 to 3.50 D Three hours 0.23 logMAR
Cufflin et al.7 Emmetropes 0.06 logMAR
Early-onset myopes +1.00 D 30 minutes 0.07 logMAR
Late-onset myopes 0.07 logMAR
Cufflin and Mallen23 Emmetropes 0.08 logMAR
Early-onset myopes +3.00 D 30 minutes 0.26 logMAR
Late-onset myopes 0.18 logMAR
Poulere et al.25 Emmetropes +2.00 D One hour 0.10 logMAR
0.09 (Landolt C)
Myopes 0.07 logMAR
0.13 (Landolt C)
Khan et al.28 Emmetropes +1.00 D 30 minutes 0.17 logMAR
+3.00 D 0.16 logMAR
Myopes +1.00 D 0.20 logMAR
+3.00 D 0.23 logMAR
Ghosh et al.24 Emmetropes +2.00 D One hour 0.05 logMAR
Myopes 0.07 logMAR
BE: both eyes, BVA: blurred visual acuity; LE: left eye, RE: right eye.
Table 1. Summary of the BVA improvements seen following adaptation to myopic defocus (induced by lenses over optimal correc-
tion or removal of myopic correction)
measurement of VA. However, top-up contrasts of 63 per cent and 98 per cent to myopic defocus and a 30-minute adaptation
periods have been employed in studies the testing in a group of emmetropes and time. The CSF was measured under
examining the effects of blur adaptation on myopes. The same method of acuity assess- defocused conditions for some observers and
contrast sensitivity and the subjective point ment was employed, but the level of myopic without defocus for others, and no top-up
of best focus. defocus was standardised to 2.50 D across adaptation was included. The authors
the entire subject cohort. observed a reduction in contrast sensitivity
Grating VA for all frequencies between 5 and 25 cpd fol-
Rosenfield et al.4 determined the effect of Contrast sensitivity measures lowing blur adaptation. The sensitivity of spa-
three hours of uncorrected myopic vision In addition to the significant changes in tial frequencies at 2 and 4 cpd was found to
on the grating acuity of 22 young myopes. BVA, adaptation has been shown to influ- be unaffected by blur adaptation.
Randomly orientated sine wave gratings ence contrast sensitivity. The introduction Rajeev and Metha31 employed a similar
were presented at contrasts of between 2.5 of defocus will reduce contrast levels method (30-minute adaptation to 2.00 D of
and 40 per cent in a three-alternative across a range of spatial frequencies, with myopic defocus), but CSF was measured
forced-choice paradigm to find the spatial this attenuating effect enhanced as spatial with defocus in place. The results showed
frequency threshold. The grating acuity frequency increases.30 The effects of blur an increase in sensitivity for 8 and 12 cpd,
improved at all contrasts following adaptation on the contrast sensitivity func- and a reduction in sensitivity for 0.5 cpd.
30 minutes without optical correction, and tion (CSF) can be measured with and with- This study also underlines the importance
this improvement continued for the remain- out blur in place. of top-up adaptation during CSF testing.
der of the three-hour adaptation period. The original study of refractive blur adapta- When top-up images between test stimuli
George and Rosenfield10 extended the study tion effects on contrast sensitivity was under- were removed, then adaptive effects on the
of grating VA and blur adaptation by adding taken by Mon-Williams et al.,1 with 2.00 D CSF were negligible.
106
Venkataraman et al.32 investigated the focused. This blur adaptation change was increases in accommodation response time
effects of 7.5 and 42 adapting stimuli on present in both young and old observers.35 to stepwise stimulus changes following blur
the foveal and peripheral (10 from fixation) adaptation. There was also an increase in
CSF. A myopic defocus level of 2.00 D and Blur sensitivity accommodation response phase lag when
an adaptation period of 30 minutes was Rather than using an indirect measure of tracking a sinusoidally moving accommoda-
employed. This adaptation generated an blur sensitivity, Wang et al.22 employed an tive stimulus. Adaptation to blur induced by
increase in sensitivity for 3 and 4 cpd only. ascending method of adjustment to directly Bangerter diffusing filters has also been
In this study, the CSF measures were under- determine the blur sensitivity pre- and post- shown to increase the magnitude of the
taken under clear conditions and without blur adaptation. Unlike VA, the presence of accommodative response following three
top-up of adaptation. Due to the time taken defocus reduces subjective blur sensitivity minutes of adaptation in myopic adult
to measure the CSF, it is important to con- thresholds and reduces the variability of the observers,16 although others have found lit-
sider the use of top-up periods of adapta- response.37 After positioning a target at the tle effect of prior adaptation to myopic
tion, as the decay of the effect may mask position of subjective best focus, myopic defocus on the accommodation response.8
the true effects. defocus was added at a rate of 0.1 D/sec.
These reports indicate that blur adapta- The observer was instructed to indicate the
tion influences the CSF. There is evidence to levels of defocus required to induce just Characteristics of the blur
suggest that the sensitivity to medium32 and noticeable blur, bothersome level of blur adaptation effect
high spatial frequencies31 is increased post- and non-resolvable blur. The authors
adaptation in response to their attenuation defined bothersome blur as where the ‘blur Time period of blur adaptation
by defocus. The reduction of medium and of the target became just bothersome or Significant changes in the perception of blur
higher spatial frequency contrast will bias annoying to look at’. This is a highly subjec- have been observed following as little as
the image further toward low frequency tive criterion, and requires the criteria of three minutes of blur adaptation.14
content. Webster33 has demonstrated that annoyance to remain constant throughout Ohlendorf et al.21 observed adaptation to
brief adaptation to scenes with a bias the adapting period. It was observed that astigmatic blur (produced by either lenses
toward low spatial frequency content can following blur adaptation, the subjective or simulated by image manipulation) after a
lead to a reduction in contrast sensitivity for sensitivity to all three levels of blur was 10-minute adaptation period.
low spatial frequencies. Webster’s observa- increased for a single letter target, but not Khan et al. documented significant
tions are also consistent with the preceding for an extract of text. It is suggested that the improvements in VA within four minutes of
work by Webster and Miyahara,34 where larger visual angle and reduction in periph- the introduction of 1.00 D or 3.00 D of myo-
adaptation to increasingly blurred images eral blur sensitivity are responsible for the pic defocus in young adult observers.28
reduced the contrast sensitivity at low discrepancy between a single letter and text High-contrast VA was measured at two-
frequencies only. stimulus behaviour.38 minute intervals for a period of 30 minutes,
Conversely, Cufflin et al.7 measured blur and the rate of VA improvement slowed
Subjective point of best focus sensitivity and blur discrimination thresholds considerably following the first six minutes.
A number of studies have examined the before and after blur adaptation in Figure 1 shows an example plot of the time
effect of blur adaptation on the positioning emmetropes, youth onset myopes, and young course of adaptation to 1.00 D of myopic
of the subjective point of best focus.14,35,36 adult onset myopes using similar methods of blur (from Khan et al.28).
Manipulation has been performed to accen- adjustment. Thresholds for blur detection and The longest duration of adaptation period
tuate either the low or high spatial fre- discrimination were found to be elevated (that used in studies that generated blur using
quency content present in a series of is observers were less sensitive to blur) follow- myopic defocus was three hours.4 The levels
images. Biasing image content toward the ing adaptation, with this effect being greatest of VA improvement in myopes observed
low spatial frequencies caused an image to in the youth onset myopes. here were similar to those seen by studies
appear blurred, while emphasising the high employing a significantly shorter adaptation
spatial frequency information caused the Assessing the accommodation period of 30 minutes.23,28 This suggests that
sharp transitions present in the image to response once blur adaptation has been established,
appear ‘too sharp’.14 Both pre- and post- Blur is a cue to accommodation, and any extending the adaptation period from
adaptation, the subjects viewed an assort- changes in the ability of the visual system to minutes to hours does not produce addi-
ment of images of varying degrees of spatial detect blur may have consequences for the tional improvements in VA.
frequency filtering. A staircase method was control and accuracy of this response. Le The persistence of the adaptation effect
used to determine the subject point of best et al.39 measured accommodation and pupil has also been investigated. Delshad et al.40
focus, where images were not too sharp responses before, during and after a period observed a mean improvement of 0.16
and not blurred. Adaptation to blur was of blur adaptation. An increase in accommo- logMAR units in 26 adults following
found to significantly shift this null point dation response variability of around 17 per 60 minutes exposure to 3.00 D of myopic
toward a level that was described as blurred cent was observed in both emmetropic and defocus. Following 20 further minutes of
prior to adaptation. The opposite occurred myopic participants following blur adapta- clear vision, an improvement of 0.11 logMAR
with adaptation to sharpened images, with tion. Following a wash-out period without units persisted, meaning that over two-thirds
the point of best focus translated so that blur adaptation, accommodation response of the improvements accrued during adapta-
images previously acknowledged as ‘too parameters returned to the pre-adaptation tion were retained in the short term. There is
sharp’ were now perceived as optimally levels. Cufflin and Mallen23 observed also evidence of adaptive effects persisting
107
This region experiences moderately altered

0.7 Emmetropes; 1 D blur refraction, a decrease in neural receptor den-
Myopes; 1 D blur sity, and a reduction in sensitivity to high spa-
Change in visual acuity (logMAR)
Emmetropes; exponential fit; R2 = 0.95 tial frequency stimuli. Myopic defocus of

0.6 Myopes; exponential fit; R2 = 0.97 1.00 D was added to the optimal refraction of
a group of emmetropes and myopes for an
adaptation period of 30 minutes. VA was
0.5 measured at 2 intervals from 0 to 10 into
the temporal visual field. Blur adaptation was
found to occur across all locations measured
0.4 and was found to be independent of eccen-
tricity. As the capability to detect high spatial
frequencies declines with increasing eccen-
0.3 tricity, yet the adaptation effects were unaf-
fected, this may indicate that recalibration of
the sensitivity to high spatial frequencies may
0.2 only play a limited role in blur adaptation.
Vera-Diaz et al.46 also found that eccentricity
had little impact on the ability of blur adapta-
0.1 tion to shift the point of subjective neutrality
– the adaptive effects were similar when nor-
0 5 10 15 20 25 30
mally sighted observers saw a target straight-
Time (mins) on, or when looking up to 10 past its edge.
The visual field is exposed to varying
Figure 1. Time course of blur adaptation, taken from Khan et al.28 degrees of defocus as we move away from
the optical axis. Venkataraman et al.47
assessed the effects of a 60-second adapta-
further than this. Following an adaptation aberrations, such as coma and spherical tion period to 2.00 D of non-optical defocus
period of three hours of uncorrected myopic aberration, in order to reduce their detri- on the perceived neutral focus (PNF). The
vision, refractive corrections were returned mental effects on visual resolution. Artal PNF is analogous to the subjective point of
to a group of subjects and their unaided acu- et al.43,44 used AO to recreate and manipu- best focus and was measured in 0.01 D
ities were checked six times over the subse- late the higher-order aberrations present in steps, with responses from participants indi-
quent 48 hours.27 The majority of the a subject’s vision. This allowed for a direct cating whether or not a test image was
unaided VA improvement accrued during the comparison of the visual resolution both ‘blurred’ or ‘sharp’. It was measured at the
adaptation period was still present after two with habitual aberrations in place and in the fovea, but a range of adapting stimuli com-
days of optimally focused vision. However, it presence of abnormal aberrations. The ini- binations were examined, including a
must be noted that no control groups were tial magnitude of these abnormal aberra- blurred parafovea (4 –20 ) with a clear cen-
included in this investigation. tions was identical to the habitual state, tral area (0 –4 ) and vice versa. Adaptation
although the orientation of these abnormal to a clear image at the fovea slightly
Resilience of adaptation aberrations had been altered. The observer reduced the defocus required at the PNF,
Khan et al.41 demonstrated that improve- was required to vary the magnitude of the regardless of whether the parafovea simul-
ments in defocused VA can occur even rotated aberrations until the image quality taneously had adapted to a grey, blurred or
through a disrupted period of exposure to matched that of the habitual image. The sharply focused image. Similarly, foveal
myopic blur. Equivalent improvements in VA observers consistently reduced the magni- adaptation to blurred central images
were seen in young adults when a tude of the rotated aberrations in order to increased the PNF regardless of the stimuli
15-minute adaptation period consisted of match the perception of the habitual aber- in the 4 –20 region. This indicates that the
alternating short (seven-second) periods of ration condition. This suggests that neural fovea may dominate when adapting to
blurred or clear vision was compared to a adaptation had acted to reduce the impact defocus, although the PNF was only mea-
15-minute period of sustained myopic of the habitual aberrations on the visual sured foveally, so we have no indication of
defocus. Blur adaptation effects induced by perception, and this had resulted in the effect on PNF other parts of the retina.
convex lens additions were also found to improved visual resolution under habitual Ghosh et al.24 also investigated blur adapta-
withstand intervening periods of clear vision conditions. tion foveally and at 10 in the periphery,
lasting one, five or 10 minutes in duration employing AO to correct for higher-order
with no reduction in acuity effects.42 aberrations. Following 60 minutes of adap-
Blur adaptation in peripheral tation to 2.00 D of myopic defocus, myopes
Adaptation to higher-order vision were found to experience a greater degree
aberrations of BVA improvement compared to
There is evidence to suggest that the visual Mankowska et al.45 investigated the ability of emmetropes, but only at the 10 temporal
system also adapts to higher-order the parafoveal region to adapt to defocus. retinal location – equivalent levels of
108
adaptation were seen foveally. In another frequency-specific channels in the visual cor- transfer of blur adaptation that has been
study, extending the blur stimulus to 42 tex.50 It was proposed that the gains of observed.1
from 7.5 was found to minimise the effects these channel outputs are variable and can
of blur adaptation on the CSF which were be altered in response to the attenuation of
present when the smaller target was high spatial frequencies by the human Blur adaptation and myopia
employed.32 eye.48 This attenuation also occurs in the
presence of defocus, which increasingly Myopes will be exposed to myopic defocus
attenuates content as spatial frequency at the fovea whenever they remove their
The mechanism of blur increases. Georgeson and Sullivan48 refractive correction, which is likely to give
adaptation suggested that continuous feedback from them a greater lifetime exposure to myopic
the visual system would allow the determi- defocus than emmetropes. It is perhaps rea-
Blur adaptation yields significant changes in nation of the attenuating factor for a certain sonable to expect that myopes will draw on
blur perception and improvements in BVA. channel. A correction factor could be this experience when adapting to myopic
A reduction in pupil size or hyperopic shift derived to instigate an increase in gain from defocus. There is also evidence that myopes
in refractive error during the adaptation the affected channel, thus restoring per- have reduced blur sensitivity compared to
period would account for these improve- ceived clarity.14 emmetropes.53
ments. However, evidence has shown there More recently, contrast constancy has Thorn et al.54 suggested that the myopes
is no change in pupil size, refractive error or been shown to withstand the effects of a in their study performed better (than
crystalline lens thickness following adapta- 3.00 D level of defocus. The investigators emmetropes) in the presence of defocus due
tion to blur.2 Webster et al. attributed the suggested that contrast amplification of the to their increased experience of defocused
changes in vision following blur adaptation blurred stimuli allowed for the maintenance vision. George and Rosenfield examined the
to temporary recalibrations of the neural of the contrast constancy effect.51 effect of equivalent levels of blur on
response to blur.14 Mon-Williams et al.1 At present, this deblurring mechanism defocused VA between emmetropes and
suggested that the contrast constancy the- described by Georgeson and Sullivan48 is myopes.10 They found that the improvement
ory of Georgeson and Sullivan48 may explain the most likely mechanism for blur adapta- in high-contrast BVA following blur adapta-
the adaptation-induced changes in visual tion. The gains of the spatial frequency tion was significant and equivalent for the
resolution. Georgeson and Sullivan48 channels could be modified in order to com- two refractive groups. The effect of blur
observed the ability to accurately perceive pensate for defocus and restore visual reso- adaptation on grating BVA was also investi-
contrast levels for a wide range of spatial lution back to pre-blur levels. There are gated for contrasts of between 2.5 per cent
frequencies, even those severely affected by three ways of reducing the effect of blur on and 98 per cent. The emmetropes showed
optical aberrations. vision. no change in grating acuity following adapta-
The fall-off in contrast sensitivity at spatial 1. Increase the sensitivity to high spatial tion, whereas the myopes showed significant
frequencies beyond the peak sensitivity fre- frequency content. This will partially improvements in grating acuity, but only at
quency has been attributed to the combined reverse the drop in high spatial fre- contrasts ≤ 16 per cent. This improved per-
effects of optical and neural limitations of quency sensitivity that occurs immedi- formance at low contrast suggests that there
the human visual system.49 This attenuation ately on insertion of myopic defocus. may be subtle differences in the myopic
in the MTF of the eye at higher spatial fre- This is supported by the work of Sub- adaptation response to defocus.
quencies impacted on contrast detection ramanian and Mutti.52 Poulere et al.25 found that myopes were
thresholds, yet the effect of these limitations 2. Decrease the gain of the low spatial affected less by blur, compared to
on a suprathreshold contrast matching task frequency-sensitive channels. In the pres- emmetropes, and this differential effect was
was minimal.48 Georgeson and Sullivan48 ence of blur the ability of the low spatial more obvious with letter targets compared
instructed observers to vary the contrast of frequency content to mask the high spa- to Landolt Cs. There was a correlation
a grating until it appeared to match the con- tial frequency content is increased. This between refractive error and change in VA
trast of a standard grating of 5 cpd. This is due to the relative weakness of high following imposition of blur, with
was repeated for standard grating contrasts spatial frequency content under the influ- emmetropes showing a larger reduction in
of up to 90 per cent (Michelson contrast) ence of blur. Reduction in the gain of the vision than myopes, and the change in
and spatial frequencies of 0.25 to 25 cpd. low spatial frequency channels will cause vision becoming smaller with the higher
The marked reduction in contrast sensitivity an ‘unmasking’ effect1 which will restore myopes. However, no difference in the mag-
at the higher spatial frequencies failed to the relative contributions of the low and nitude of blur adaptation effects was evi-
impact on the contrast matching ability, with high spatial frequency channels back dent between refractive groups. Ghosh
accurate contrast matching performed up to toward pre-blur levels. et al.24 also observed higher levels of blur
25 cpd. The authors proposed that a com- 3. A combination of both high spatial fre- adaptation in myopes compared to
pensation process occurred to counteract quency gain increase and unmasking emmetropes, but only at 10 in the periph-
the optical and neural attenuation of high could be employed to maximise the eral field – no difference between groups
spatial frequencies by the human eye and deblurring effect. This mechanism is was observed at the fovea.
restore the clarity of the image. This was supported by Mon-Williams et al.1 The The visual diet presented to the human
termed contrast constancy.48 deblurring process is believed to take visual system has undergone considerable
The visual system is known to process place in central binocular cells of the change in the last decade, with increasing
visual information in a series of spatial visual cortex, due to the inter-ocular reliance on handheld devices for work,
109
communication and entertainment. Such of around 1.00 to 2.00 D, may show elevated a period of adaptation, with effects being
devices can expose the eyes to stimuli that performance in terms of unaided vision. It is observed in the contrast and blur sensitivi-
differ from natural images in terms of spa- possible that the uncorrected myope could be ties of adapted observers, in addition to per-
tial frequency content and chromatic spec- in a chronic state of blur adaptation, with the formance on high-contrast acuity charts.
trum. This offers the potential of a chronic well-documented effects impacting upon clini- This effect is likely due to compensatory
blur stimulus, which may induce adaptive cal measurements. A similar effect may be changes in spatial frequency detection chan-
effects. In the area of myopia management, seen in presbyopic patients wearing mono- nels at the level of the visual cortex and
the role of relative myopic retinal blur as a vision correction, due to the chronic myopic may include both low and high spatial fre-
‘stop’ signal for eye growth is building in the blur presenting to the ‘near vision’ eye during quencies. Additionally, changes in ocular
evidence base of clinical effectiveness. distance vision. anatomy may be observed, particularly
Radhakrishnan et al.55 examined the effect The changes induced by blur adaptation since recent advances in methods of bio-
of simultaneous vision from a bifocal correc- are unlikely to be confined to the visual metric measurement, for example, axial
tion. Shifts in the perception of blur were function alone. Read et al.56 observed small, length and enhanced depth imaging ocular
observed with simultaneous vision bifocal but statistically significant changes in axial coherence tomography. Evidence suggests
designs, and the magnitude of the shifts in length following one hour of exposure to that myopes may display a slightly higher
perception were related to the proportions myopic and hyperopic defocus while viewing propensity for this blur adaptation improve-
of defocus. The impact of chronic exposure a distance target. A mean (1 SD) decrease ment. As exposure to defocus is being
to myopic defocus provides considerable in axial length of 13 14 μm was observed established as a myopia management strat-
opportunity for the study of blur adaptation. following exposure to 3.00 D of myopic egy, the impact of chronic blur on the struc-
It may be the case that blur adaptation has defocus, while similar exposure to hyper- ture and function of the visual system is
a role to play in the refinement of myopia opic defocus induced a mean increase of ripe for further investigation.
management strategies. 8 14 μm in a group of young adults. Adap-
tation to 1.50 D of myopic defocus was also REFERENCES
1. Mon-Williams M, Tresilian JR, Strang NC et al. Improv-
found to reduce the amplitude of the daily
ing vision: neural compensation for optical defocus.
Blur adaptation in disease changes in sub-foveal and parafoveal cho- Proc Biol Sci 1998; 265: 71–77.
roidal thickness, as well as shifting the 2. Pesudovs K, Brennan NA. Decreased uncorrected
vision after a period of distance fixation with specta-
Vera-Diaz et al.46 assessed blur adaptation timing of the maximum and minimum cle wear. Optom Vis Sci 1993; 70: 528–531.
(using the method described by Webster values.57 Effects of myopic and hyperopic 3. Rosser DA, Murdoch IE, Cousens SN. The effect of
et al.14 and Vera-Diaz et al.36) in a small num- defocus on choroidal thickness have also optical defocus on the test-retest variability of visual
acuity measurements. Invest Ophthalmol Vis Sci 2004;
ber of patients with central vision loss com- been observed in myopic school children.58 45: 1076–1079.
pared with normally sighted controls. Blur These changes were induced under experi- 4. Rosenfield M, Hong SE, George S. Blur adaptation in
adaptation was evident in the peripheral field mental conditions that are very similar to myopes. Optom Vis Sci 2004; 81: 657–662.
5. Charman WN, Jennings JA. The optical quality of the
of normal observers and those with central those used to study blur adaptation to monochromatic retinal image as a function of focus.
vision loss. It was found that adaptation to refractive defocus. Therefore, there are Br J Physiol Opt 1976; 31: 119–134.
6. Radhakrishnan H, Pardhan S, Calver RI et al. Effect of
short-term blur was evident in the partici- likely to be changes occurring in the ocular
positive and negative defocus on contrast sensitivity
pants with vision loss, despite their long-term structures as well as the changes in visual in myopes and non-myopes. Vision Res 2004; 44:
reduced vision; blur adaptation processes function that we see in blur adaptation. 1869–1878.
7. Cufflin MP, Mankowska A, Mallen EA. Effect of blur
appear intact and robust following pathology. Blur adaptation may have relevance to adaptation on blur sensitivity and discrimination in
There was no correlation between the myopia management procedures where a emmetropes and myopes. Invest Ophthalmol Vis Sci
2007; 48: 2932–2939.
degree of adaptation to blur and VA. lens with a relative positive addition places
8. Cufflin MP, Hazel CA, Mallen EA. Static accommodative
an image shell in front of the retina. The responses following adaptation to differential levels
likely increase in the number of patients of blur. Ophthalmic Physiol Opt 2007; 27: 353–360.
Clinical implications of blur 9. Radhakrishnan H, Pardhan S, Calver RI et al. Unequal
being prescribed myopia management
adaptation reduction in visual acuity with positive and negative
interventions will drive the need for a fuller defocusing lenses in myopes. Optom Vis Sci 2004; 81:
understanding of blur adaptation processes, 14–17.
The clinical ramifications of blur adaptation 10. George S, Rosenfield M. Blur adaptation and myopia.
and any impact on other processes, such as Optom Vis Sci 2004; 81: 543–547.
could be directly relevant to the refraction accommodation function. In the eye 11. Schmid KL, Iskander DR, Li RWH et al. Blur detection
process during an eye examination. Lengthy corrected with a conventional spectacle lens, thresholds in childhood myopia: single and dual tar-
get presentation. Vision Res 2002; 42: 239–247.
exposure to blur during refraction may induce Lin et al.59 demonstrated a significant level 12. Field DJ. Relations between the statistics of natural
blur adaptation in patients. Confirmation tests of relative hypermetropic blur away from images and the response properties of cortical cells.
of the refractive end point which rely on the visual axis. Study of any adaptation to
J Opt Soc Am A 1987; 4: 2379–2394.
13. Tolhurst DJ, Tadmor Y, Chao T. Amplitude spectra of
responses to a blurred stimulus, specifically this form of blur may provide useful insight natural images. Ophthalmic Physiol Opt 1992; 12:
the ‘+1.00 D blur test’, may yield a better than into myopia progression. 229–232.
expected level of vision. Given the relatively 14. Webster MA, Georgeson MA, Webster SM. Neural adjust-
ments to image blur. Nat Neurosci 2002; 5: 839–840.
short onset time of blur adaptation effects, 15. Tadmor Y, Tolhurst DJ. Discrimination of changes in
care should be taken to limit a patient’s expo- Summary the second-order statistics of natural and synthetic
images. Vision Res 1994; 34: 541–554.
sure to blur during the refraction process.
16. Vera-Diaz FA, Gwiazda J, Thorn F et al. Increased
Patients being examined for their first myopic In summary, prolonged exposure to blur sig- accommodation following adaptation to image blur in
correction, for example at a level of myopia nificantly improves visual performance after myopes. J Vis 2004; 4: 1111–1119.
110
17. Pérez GM, Archer SM, Artal P. Optical characterization 31. Rajeev N, Metha A. Enhanced contrast sensitivity con- 46. Vera-Diaz FA, Woods RL, Peli E. Blur adaptation to
of Bangerter foils. Invest Ophthalmol Vis Sci 2010; 51: firms active compensation in blur adaptation. Invest central retinal disease. Invest Ophthalmol Vis Sci 2017;
609–613. Ophthalmol Vis Sci 2010; 51: 1242–1246. 58: 3646–3655.
18. Marcos S, Werner JS, Burns SA et al. Vision science 32. Venkataraman AP, Winter S, Unsbo P et al. Blur adap- 47. Venkataraman AP, Radhakrishnan A, Dorronsoro C
and adaptive optics, the state of the field. Vision Res tation: contrast sensitivity changes and stimulus et al. Role of parafovea in blur perception. Vision Res
2017; 132: 3–33. extent. Vision Res 2015; 110: 100–106. 2017; 138: 59–65.
19. Sawides L, de Gracia P, Dorronsoro C et al. Vision is 33. Webster MA. Contrast sensitivity under natural states 48. Georgeson MA, Sullivan GD. Contrast constancy:
adapted to the natural level of blur present in the reti- of adaptation. Proc SPIE 1999; 3644: 58–70. deblurring in human vision by spatial frequency chan-
nal image. PLoS One 2011; 6: e27031. 34. Webster MA, Miyahara E. Contrast adaptation and the nels. J Physiol 1975; 252: 627–656.
20. Sawides L, Marcos S, Ravikumar S et al. Adaptation to spatial structure of natural images. J Opt Soc Am A Opt 49. Campbell FW, Green DG. Optical and retinal factors
astigmatic blur. J Vis 2010; 10: 22. Image Sci Vis 1997; 14: 2355–2366. affecting visual resolution. J Physiol 1965; 181:
21. Ohlendorf A, Tabernero J, Schaeffel F. Neuronal adap- 35. Elliott SL, Hardy JL, Webster MA et al. Aging and blur 576–593.
tation to simulated and optically-induced astigmatic adaptation. J Vis 2007; 7: 8. 50. Blakemore C, Campbell FW. On the existence of neu-
defocus. Vision Res 2011; 51: 529–534. 36. Vera-Diaz FA, Woods RL, Peli E. Shape and individual rones in the human visual system selectively sensitive
22. Wang B, Ciuffreda KJ, Vasudevan B. Effect of blur variability of the blur adaptation curve. Vision Res to the orientation and size of retinal images. J Physiol
adaptation on blur sensitivity in myopes. Vision Res 2010; 50: 1452–1461. 1969; 203: 237–260.
2006; 46: 3634–3641. 37. Jacobs RJ, Smith G, Chan CD. Effect of defocus on blur 51. Comerford JP, Thorn F, Chuang J. Contrast constancy
23. Cufflin MP, Mallen EA. Dynamic accommodation thresholds and on thresholds of perceived change in with refractive blur. Invest Ophthalmol Vis Sci 2002; 43:
responses following adaptation to defocus. Optom Vis blur: comparison of source and observer methods. 4719.
Sci 2008; 85: 982–991. Optom Vis Sci 1989; 66: 545–553. 52. Subramanian V, Mutti DO. The effect of blur adapta-
24. Ghosh A, Zheleznyak L, Barbot A et al. Neural adapta- 38. Wang B, Ciuffreda KJ. Blur discrimination of the tion on contrast sensitivity. Invest Ophthalmol Vis Sci
tion to peripheral blur in myopes and emmetropes. human eye in the near retinal periphery. Optom Vis 2005; 46: 5604.
Vision Res 2017; 132: 69–77. Sci 2005; 82: 52–58. 53. Rosenfield M, Abraham-Cohen JA. Blur sensitivity in
25. Poulere E, Moschandreas J, Kontadakis GA et al. Effect 39. Le R, Bao J, Chen D et al. The effect of blur adaptation myopes. Optom Vis Sci 1999; 76: 303–307.
of blur and subsequent adaptation on visual acuity on accommodative response and pupil size during 54. Thorn F, Cameron L, Arnel J et al. Myopia adults see
using letter and Landolt C charts: differences between reading. J Vis 2010; 10: 1. through defocus better than emmetropes. In:
emmetropes and myopes. Ophthalmic Physiol Opt 40. Delshad S, Collins MJ, Read SA et al. The time course Tokoro T, ed. Myopia Updates Proceedings of the 6th
2013; 33: 130–137. of blue adaptation and its persistence. In: The Associa- International Conference on Myopia. Tokyo: Springer,
26. Ferris FL 3rd, Kassoff A, Bresnick GH et al. New visual tion for Reaserch In Vision and Ophtalmology (ARVO) - 1998. pp. 368–374.
acuity charts for clinical research. Am J Ophthalmol Asia Meeting. Brisbane, Qld: Brisbane Convention & 55. Radhakrishnan A, Dorronsoro C, Sawides L et al.
1982; 94: 91–96. Entertainment Centre, 2017. Short-term neural adaptation to simultaneous bifocal
27. Rosenfield M, Portello JK, Hong SE et al. Decay of 41. Khan KA, Cufflin MP, Mallen EA. The effect of inter- images. PLoS One 2014; 9: e93089.
blur adaptation. Invest Ophthalmol Vis Sci 2003; 44: rupted defocus on blur adaptation. Ophthalmic Physiol 56. Read SA, Collins MJ, Sander BP. Human optical axial
4315. Opt 2016; 36: 649–656. length and defocus. Invest Ophthalmol Vis Sci 2010; 51:
28. Khan KA, Dawson K, Mankowska A et al. The 42. Portello J, Rosenfield M. Effect of intervening periods 6262–6269.
time course of blur adaptation in emmetropes of clear vision on blur adaptation. Optom Vis Sci 2002; 57. Chakraborty R, Read SA, Collins MJ. Monocular myopic
and myopes. Ophthalmic Physiol Opt 2013; 33: 79: 24. defocus and daily changes in axial length and choroi-
305–310. 43. Artal P, Chen L, Fernández EJ et al. Adaptive optics for dal thickness of human eyes. Exp Eye Res 2012; 103:
29. Carkeet A, Lee L, Kerr JR et al. The slope of the psy- vision: the eye’s adaptation to point spread function. 47–54.
chometric function for Bailey-Lovie letter charts: J Refract Surg 2003; 19: S585–S587. 58. Wang D, Chun RK, Liu M et al. Optical defocus rapidly
defocus effects and implications for modeling 44. Artal P, Chen L, Fernández EJ et al. Neural compensation changes choroidal thickness in schoolchildren. PLoS
letter-by-letter scores. Optom Vis Sci 2001; 78: for the eye’s optical aberrations. J Vis 2004; 4: 281–287. One 2016; 11: e0161535.
113–121. 45. Mankowska A, Aziz K, Cufflin MP et al. Effect of blur 59. Lin Z, Martinez A, Chen X et al. Peripheral defocus
30. Smith G. Ocular defocus, spurious resolution and con- adaptation on human parafoveal vision. Invest with single-vision spectacle lenses in myopic children.
trast reversal. Ophthalmic Physiol Opt 1982; 2: 5–23. Ophthalmol Vis Sci 2012; 53: 1145–1150. Optom Vis Sci 2010; 87: 4–9.
111
INVITED REVIEW
Adaptive optics imaging of the retinal microvasculature
Phillip Bedggood BOptom PhD The eye has long been recognised as the window to pathological processes occurring in the
Andrew Metha BSc BSc (Optom) brain and other organs. By imaging the vasculature of the retina we have improved the scien-
PGCertOcTher PhD tific understanding and clinical best practice for a diverse range of conditions, ranging from
Department of Optometry and Vision Sciences, The diabetes, to stroke, to dementia. Mounting evidence suggests that damage to the smallest
University of Melbourne, Melbourne, Australia and most delicate vessels in the body, the capillaries, is the first sign in many vasculopathies.
E-mail: [email protected] These are the most critical vessels involved in the exchange of metabolites with tissue. Accu-
rate assessment of retinal capillary structure and function would therefore be of great benefit
across a broad range of disciplines in medical science; however, their small size does not
make this an easy task. This has led to the development of high-resolution adaptive optics
Submitted: 3 July 2019 imaging methods to non-invasively explore retinal microvascular networks in living human
Revised: 17 September 2019 eyes. This review describes the present state of the art in the field, the scientific break-
Accepted for publication: 20 September throughs that have been made possible in the understanding of vessel structure and function
2019 in health and disease, and future directions for this emerging technology.
Key words: adaptive optics, blood flow, capillaries, diabetes, retina
Microvasculature as a window to and drainage networks. However, the critical disease, small vessels have indeed been
systemic and central nervous activities of the vascular system described observed to affect and be affected by patho-
system pathology above take place on a smaller scale, within logical processes early in diverse conditions
networks of capillaries on the order of including diabetes, hypertension, coronary
5–10 μm wide. This is so for a variety of rea- heart disease, cerebral ischaemia–reperfusion
The retina is the only tissue for which the
sons. The narrow capillary lumen forces blood injury, stroke or brain haemorrhages, demen-
body’s deep vasculature can be visualised
cells to be marshalled through in single file, tia, demyelinating disease, and sepsis.7,8 The
directly and non-invasively, offering a
deforming into elongated shapes.5,6 Resis- most well-understood example is diabetic
unique avenue for the assessment of vascu-
tance to flow is high, comprising ~1/4 to 1/3 of vasculopathy which begins with loss of peri-
lar health in major systemic conditions such
as diabetes,1 hypertension2 and coronary the vascular total; flow is accordingly slower cytes, thickening of the endothelial basement
heart disease.3 Being neural tissue, retinal which enables sufficient time for exchange of membrane9 and eventual development of
observation of blood vessels is also uniquely metabolites.5 Metabolite exchange is further microaneurysms. Dysfunction of endothelial
relevant to important neurovascular condi- facilitated by the increased cell-wall contact and/or pericyte cells and associated disruption
tions such as stroke, Alzheimer’s disease area, reduced average distance between cell of the blood–brain barrier lead to leakage and
and other dementias.3,4 contents and tissue, and much thinner capil- impaired local regulation of flow. Various com-
The vasculature can be thought of as an lary wall.5,6 The sheer abundance of capillaries binations of these processes are thought to be
interconnected network of distribution pipes provides closer average proximity to all cell compromised in a range of diseases.7,8 The
the function of which is, essentially, to deliver types in the body, particularly neurons.6,7 term ‘small vessel disease’ has been proposed
a reliable flow of blood through the entire Despite lacking a muscular coat, capillaries as a common microvasculopathy manifesting
body. Within the blood are the required nutri- are capable of active local flow regulation by in heart (leading to ischaemic heart disease),
ents, gases, hormones, and immune agents way of contractile pericyte cells distributed brain (leading to stroke and dementia), retina
which must be delivered to tissue, while at the within their walls.7 In addition to their impor- and kidneys.10
same time removing the waste products of tance to overall vascular function, being vastly
metabolism. Using standard ophthalmoscopy more numerous, they offer greater redun-
methods, the most readily visualised vessels dancy compared with large vessels and so are Challenges in study of the
in colour fundus photographs of the retina statistically more likely to reveal the earliest, smallest vessels
are ~75–150 μm wide. Vessels as small as pre-clinical hallmarks of pathology. For all
30 μm can sometimes be resolved, and from these reasons, it makes sense to pay attention Is studying individual vessels of the micro-
close observations at this scale we have learnt to the smallest vessels, as well as the large. vasculature worthwhile? Each one does not
much about the structure and some of the As technology has improved and enabled get very much of the overall flow; what flow
functional workings of the vascular supply the microvasculature to be assessed in human they do get is hard to interpret as it is
112
Adaptive optics vascular imaging Bedggood and Metha
extremely heterogenous, changing markedly interactions outside the plane of interest, • Choice of imaging wavelengths to maxi-
over time and between neighbouring capil- improving axial resolution and minimising mise back-reflection from tissue posterior
laries in seemingly unpredictable fashion. In veiling glare. Manipulation of the aperture to the vasculature, and absorption by the
some ways capillaries appear uninteresting, can reject directly returned light so that vasculature relative to other absorbers
with no muscular coat or a tunica media or the signal only shows scatter from multi- such as the lens, macular pigment or mel-
adventitia at all. They are small and accord- ple structures; such scatter requires a anin. If a narrow bandwidth is used, the
ingly difficult to study; they are typically change in refractive index, rendering oth- light is partially coherent and so contrast
about 25 per cent narrower than the blood erwise transparent objects.15 The chief may be generated by complex interfer-
cells they permit,6 forcing most to undergo limitation of confocal scanning is the ence effects including defocused phase
significant deformation which alters resis- lower frame rate as time is required to contrast.21
tance in a poorly understood manner.11 form the image; however, scanning can • Adaptive optics (AO) which, combined with
Resistance is also increased by significant be ‘frozen’ on a cross-section of a single other methods described above, compen-
interaction between blood cells and the vessel to afford extremely rapid serial sates for the distorted shape of light waves
endothelial glycocalyx; these factors reconstruction of flow.16 returning from the eye to enable resolu-
are tempered by formation of a relatively • Interferometry for example, optical coher- tion of cellular structures. The downsides
large plasma layer which effectively lubri- ence tomography (OCT), which generates are that the superior resolution is afforded
cates the flow.12 Thus the flow state is com- contrast from optical path differences as over a limited field of view, for example
plex and not easy to describe with simple small as tens of nanometres to enable 1–2 ,22 and that increased cost and com-
models of fluid mechanics. Small changes in visualisation of otherwise transparent plexity have historically limited widespread
vessel morphology, endothelial cell function, structures. Coherent light imaging ‘gates’ clinical adoption.
basement membrane thickness, local blood information to a particular range in depth,
composition, cell aggregation and other rhe- providing high axial resolution and
ological factors could all have significant minimising veiling glare. Entire tissue vol- State of the art in AO imaging of
effects on capillary flow. Thus, while study umes are acquired which provides unique the human retinal
of capillary function may provide very sensi- three-dimensional profiling of tissue. microvasculature
tive indicators of pathology, there is still However, signal acquisition is time-
much to learn. consuming relative to the speed of blood Vascular structure
The retina may be the best place in the flow. This means that, despite recent Adaptive optics was invented to compensate
body to improve our understanding of claims,17 transverse flow velocity cannot for atmospheric turbulence encountered by
microvascular flow phenomena, given the be determined – the sampling require- ground-based telescopes. The technology
transparency of the ocular media and the ments for this are described further was extended to retinal imaging over two
ability to image repeatedly and non- below. Without the ability to track flow, decades ago for study of the cone photore-
invasively in living human subjects. The first and again despite recent claims, crossings ceptor mosaic, by combining conventional
challenge is spatial resolution: as mentioned cannot be differentiated from vessel bra- flash or ‘flood’ fundus photography with
above, vessels below about 30 μm diameter nches to learn network connectivity.18 AO.23 The shadows of blood vessels were
cannot be seen by conventional or ‘flood’ Axial velocity can be measured using the noted during the first AO studies of the
illumination ophthalmoscopy (where the Doppler effect, but only a small propor- in vivo photoreceptor mosaic and used to
entire field is imaged simultaneously by tion of retinal micro-vessels are oriented aid repeated retinal alignment,24 but it was
brief flashes of light). Rather than size per axially.19 Another limitation of OCT is that not until AO was combined with confocal
se though, the main limitation is low con- acquisition is time-consuming relative to scanning methods that images directly
trast (because light is absorbed by only a the incessant motion of the eye, meaning focused on vasculature structure were dem-
thin column of haemoglobin, and veiled by that no two OCT scans look exactly the onstrated.14 Later strategies computed
other intra-ocular scatter) combined with same. The inherent distortion degrades intensity variations in time to generate
imperfect focusing of the eye (aberrations). local structural measures such as diame- label-free, high-contrast perfusion maps of
Methods to address these limitations ter and branching angle and, while useful the lumens of the smallest retinal vessel
include the following: global metrics can still be determined, networks.25,26 Figure 1 shows a perfusion
• Contrast agents introduced intravenously individual vessels or micropathology may map (B) generated by our flood AO system
or orally. Fluorescein angiography can be missed entirely due to a sudden jump with methods recently described,27 overlaid
facilitate flow measurements in some in the eye’s position.20 on a commercial OCT angiography (OCT-A)
individual capillaries without adaptive • Changes over time of recorded image data scan used to guide imaging (A).
optics,13 although the lifetime of useful occur due to movement of blood constitu- Both the blood column and vessel wall
fluorescence is limited, normal cell rheol- ents with different absorption and/or itself are composed of largely transparent
ogy may be disrupted, and intravenous refractive index characteristics. Repeated cells which nonetheless differ in refractive
injection of dyes is somewhat invasive observations of the same point on the ret- index from surrounding tissue. Collecting
which precludes routine use. ina can generate high contrast from these indirectly scattered light from these struc-
• Confocal methods,14 which image a small differences; in lower frame rate devices tures allows them to be visualised using ‘off-
point or line that is rapidly scanned this creates perfusion maps of the vascula- set aperture’ methods akin to darkfield
across the retina to construct each frame. ture; in higher frame rate devices, flow in microscopy.28,29 A related variation of this
An aperture blocks light from scattering individual vessels can be tracked. imaging modality, known as ‘split-detector’,
113
A. Commerical OCT-A
B. “Flood” AO motion contrast map
C. Velocity
map
D. Space-time
plots for
indicated
vessels
Figure 1. Mapping capillary flow in a human subject with type I diabetes. A: Background acquired with a commercial optical coherence
tomography angiography (OCT-A) device (Heidelberg Spectralis). B: Motion contrast montage acquired with our ‘flood’ adaptive optics
(AO) system at 400 fps with 593 nm light. C: Velocity mapping using pixel intensity cross-correlation, in the same region as B. Red arrows
indicate flow direction for three segments at an arterial junction, referenced in D. D: Spatiotemporal plots over a 200 ms period for ves-
sels indicated by white arrows, showing the alternating single-file passage of cells (black) and plasma (white). The right-most plot has
much lower haematocrit despite being of comparable diameter and flow velocity. Scale bar is 100 μm. Velocity colour map ranges from
0 to 4.5 mm/s.
ignores the directly scattered, confocally leukocytes, with their relative scarcity in the measurements of velocity, cell shape and
imaged rays while collecting and contrasting bloodstream allowing unambiguous tracking lumen diameter.37,38 Example data from this
light landing on either side of the confocal of each cell.33–35 Similarly, the tendency of method are shown in Figure 2, courtesy of Dr
image point.30 Such methods are sensitive cell aggregates and/or lengthy sections of Jesse Schallek’s laboratory, University of
to refractive index (phase) changes in the plasma to form in certain vessels allows Rochester.
direction of the offset or split, although at unambiguous tracking of those aggregates.36 For assessment of contemporaneous flow
the cost of some lateral resolution. A pro- An alternate strategy is to ‘freeze’ the scan- velocity across the capillary network, deviations
grammable aperture can be used to explore ning raster on a cross-section of a single ves- from the traditional point-scanning approaches
structure oriented in arbitrary directions.31 sel of interest.16 This allows extremely rapid are required. This includes high frame rate
Recently developed darkfield methods may (kilohertz) measurement of the flow profile ‘flood’ illumination imaging, which affords
be able to reveal similar details for flood- across a single vessel; smaller vessels are direct visualisation and tracking of individual
based illumination geometries.32 precluded due to difficulties in compensating cells traversing the network21,39 line scanning
for motion of the eye. More recent strategies technology which strikes a balance between
Vascular function have extended this approach to individual confocality and rapidity of acquisition,40 or
The first label-free investigations of capillary capillaries in anaesthetised rodents, allowing dual-channel point scanning with a very small
flow in the living human retina were achieved the counting of individual red cells, white temporal offset between channels, allowing
by tracking the shadows of individual cells and platelets as well as accurate pairwise cell displacements to be calculated.41
114
Figure 2. Fast cross-sectional imaging for reconstruction of capillary contents. A 1D scan line is ‘frozen’ along a cross-section per-
pendicular to the lumen of a target capillary, and rapidly sampled over time (15 kHz). The resulting space–time image is depicted
here with time horizontal and shows the ability to count different blood components and measure their velocity, shape and pack-
ing arrangement. Data were acquired in an anaesthetised rodent at 796 nm with split-detection adaptive optics scanning laser
ophthalmoscopy (AOSLO). Scale bars: horizontal = 10 ms, vertical = 5 μm. Image supplied courtesy of Dr Jesse Schallek’s labora-
tory, University of Rochester.
The resulting datasets afford detailed analysis lumen and thickening of the vessel wall, Branching
of flow patterns across the capillary network. that is an increase in the wall-to-lumen There is strong coupling expected between
Figure 1C shows a map of average velocity cal- ratio (WLR). This is hard to assess with stan- the diameter of parent and daughter bra-
culated by ‘flood’ AO at 400 fps, using pixel dard clinical retinal photography or other in nches at vascular junctions, based on
intensity cross-correlation.27 vivo methods; the gold standard is a subcu- minimisation of energy expended in
In recent years, the advent of new taneous biopsy which is invasive and may transporting and supporting the blood vol-
methods to analyse variations in amplitude not be entirely representative of neural ume. This is expressed by Murray’s law,
and phase information in OCT data (OCT-A) tissue.48 which states that the sum of the cubes of
have allowed commercial OCT systems to Using AO retinal imaging, WLR can be the daughter radii should equal that of the
provide high-contrast, noninvasive perfusion measured non-invasively and has been parent.58 The majority of vessels across vari-
maps of the retinal vasculature.42 The same shown to increase with mean blood pres- ous tissues and species conform well to this
methods have been applied to improve AO- sure, body mass index and age.49–52 The law, including the largest retinal vessels
OCT perfusion mapping,43,44 including visu- WLR in healthy individuals is highly predict- imaged with conventional retinal fundus
alisation of the choriocapillaris.45 The com- able from the size of the vessel, with a strong photography.19
bination of AO and OCT-A significantly linear relationship (R2 ~ 0.98) between lumen However, AO imaging of healthy vessels
improves both transverse and axial resolu- and total diameter for the gamut of vessels < 100 μm in diameter has demonstrated sig-
tion, with recorded vessel diameters spanning from 10 to 150 μm in size (that is, nificant departures from the expected cubic
according with histology and the distinct all except the capillaries).53 In hypertension relationship, with exponents around two or
anatomical beds appearing better sepa- the correlation remains strong but some ves- less in veins of diameter 20–100 μm and
rated. AO-OCT-A perfusion maps now rival sels show significant departures from the in arteries 20–50 μm.59 Departures from
those generated by other AO modalities. predicted relationship, with narrowing of the Murray’s law are expected where the under-
Given the recent successes of AO-OCT in lumen and concomitant thickening of lying assumptions no longer hold, that is,
visualising fine transparent structure such the wall. Evaluation of deviations from the that resistance no longer varies inversely
as ganglion cell somas and axons,46 normally tight coupling observed may pro- with the fourth power of vessel radius. Such
advances in visualisation of vascular support vide a statistically powerful biomarker for departures may occur for example where
cells with AO-OCT-A may lie in the near disease. It has recently been shown that the underlying flow is turbulent, where
future. The AO procedure itself can in princi- WLR as measured with AO is responsive to blood viscosity and/or vessel stiffness
ple be achieved in software alone with both short-term (dilation of the lumen) and change significantly across a junction, or in
swept source OCT, due to the simultaneous long-term (lumen dilation and reduction of dynamic states where flow is redistributed
acquisition of phase-stable information.47 wall thickening) pharmacological treatment across the network.58,59 Further study is
of hypertension.54 needed to develop our theories of capillary
Remodelling affecting WLR has also been flow in light of the measurements now
Scientific advances with AO noted in diabetes. While reports have been afforded by AO imaging technology; this is
imaging: vessel structure mixed on whether microvascular diameter discussed further below in the section on
increases or decreases in diabetes, likely vessel function.
The above developments have improved due to differential effects at different stages
our understanding of the normal structure of the disease,55 AO-enabled studies point Tortuosity
of the retinal microvasculature, and how to wall thickening in those vessels large AO-based investigations have demonstrated
this becomes altered in various disease enough to have a substantive wall56 that small vessel tortuosity is increased in
processes. together with narrowing of the lumen.57 diabetes, even in eyes with no clinically
Figure 3 shows images captured with both detectable retinopathy.56,60 Adaptations
Lumen and vessel wall offset-pinhole AO scanning laser ophthal- include the formation of small, sharp loops
One of the most significant indicators for moscopy (AOSLO) (A-C, courtesy of Dr Ste- and sprouts in the capillaries, analogous to
prognosis of hypertension is structural phen Burns, Indiana University) and flood the clinically familiar presentation of intra-
adaptation to sustained high blood pres- AO (D-E, using our system), revealing alter- retinal microvascular abnormalities (IRMA)
sure, which results in narrowing of the ations to the vessel wall in diabetes. in larger retinal vessels. An example of
115
Figure 4A shows a large microaneurysm (red

A B C D arrow) imaged with offset-pinhole AOSLO
(courtesy of Dr Stephen Burns, Indiana Uni-
versity), with surrounding large cysts (white
arrows) and mottled microcystic appearance
of the inner retina. Other panels show micro-
aneurysms (red arrows) imaged with our
flood AO system.
Lesions identified with AO can be repeat-
edly targeted with longitudinal follow-up,
allowing study of the natural history of
formation, progression and resolution of indi-
vidual microaneurysms and other pathol-
ogy.61,63,64 Further studies are needed to
E determine whether microaneurysms give
way to or are caused by other structural
abnormalities, such as highly tortuous
capillaries. Regardless, the rate of
microaneurysm turnover is known to be a
strong biomarker for development of
future sight-threatening complications.65
However, microaneurysms exhibit wide
variation in their size (for example, com-
pare lesion sizes in Figure 4A and 4D) and
perfusion status (for example compare
motion contrast in Figure 4E and 4G). This
means that non-AO-enabled systems with-
Figure 3. Visualisation of vessel wall and lumen in human diabetic retina. A: Adaptive
out a contrast agent are liable to miss
optics scanning laser ophthalmoscopy (AOSLO) reflectance image with horizontal off-
more than half of microaneurysms.66
set of the pinhole, revealing vessel wall fine structure in the horizontal direction.
The detailed characterisation of lesion
Focal constriction is evident in this vessel (arrows). B: AOSLO reflectance image in a
shape is critical information required for
different subject who shows thickening of the wall (arrows). C: The same data from B
computational fluid models which can pre-
with motion contrast overlay (red), revealing that the wall (white arrows) is substan-
dict the dynamics of flow through affected
tially thicker than the lumen (red arrows), especially on the right-hand side. D: Flood
regions of the retina. This allows the calcula-
AO reflectance image of an arteriole. The vessel edge is indicated (arrows). E: Motion
tion of otherwise unmeasurable parameters
contrast of the area in D, demonstrating a large dark (non-perfused) region on the
such as wall shear stress and perfusion
left side of the vessel. Extent of the functional lumen (red arrows) is much smaller
pressure. This information, if estimated
than the total vessel diameter (white arrows), especially on the left side. Scale bar is
accurately, would prove indispensable in
100 μm. AOSLO images supplied courtesy of Dr Stephen Burns, Indiana University;
deciding the likelihood that a given vessel or
flood AO images acquired with our system.
lesion will progress to lack of perfusion or
development of other abnormalities.67,68
unusual vessel malformations in diabetes exchange between red cells and tissue, lead- Such techniques can similarly be applied to
can be seen in Figure 4, with B-C showing a ing to further ischaemia and channel closure study less severe structural abnormalities
‘knot’ structure, D-E showing a hairpin loop in a vicious cycle. The result is widespread described above such as altered tortuosity
(white arrows) and F-G showing a tortuous capillary dropout and other later stage com- or branching.69
vessel (white arrows) which feeds a plications of diabetic retinopathy. In addition to microaneurysms, AO can be
microaneurysm. Microaneurysms are dis- used to image other characteristic pathologies
cussed further below. Microaneurysms and other with high resolution, such as neovascularisation,
AO imaging data has shown that tortuosity lesions in diabetes hard exudates, haemorrhages, cotton wool
changes in diabetes occur preferentially in Microaneurysms can be rendered in exqui- spots, and intraretinal cysts.70
the arteriovenous channels, which are the site detail with AO imaging, identifying
widest and most direct path between arterial unique morphological classes not previously Other focal vascular
and venous systems and are hence preferred known.61 Previous knowledge regarding the abnormalities
by the larger leukocytes.60 It has been pro- details of microaneurysm formation comes In addition to diffuse changes to the vessel
posed that, following these structural from histology, where sample sizes are lim- wall and lumen in hypertension, noted
changes, leukocytes become re-routed into ited and it is impossible to track lesions over above, microvascular remodelling underpins
the surrounding capillaries which would oth- time. Using AO, microaneurysm shapes have more familiar clinical signs such as arterio-
erwise be relied upon to traffic large num- been shown to be predictive of surrounding venous (A/V) nicking and focal arteriolar
bers of red cells. This disrupts metabolic disorganisation of the inner retina.62 narrowing. Focal narrowing has been
116
tracked rapidly over time. The method is

A B C
best suited to vessels in the 30–80 μm range
due to the significant, incessant movement
of the eye;16 however, it can in principle be
applied to individual capillaries.37 In many
of the larger vessels in the body, idealised
D E approximations of fluid mechanics hold and
a parabolic velocity profile is expected. It
has now been shown that in the small reti-
nal vessels these approximations begin to
break down; the profile becomes flatter with
decreasing vessel diameter and, in arteri-
F G oles, depends dynamically on the phase of
the cardiac cycle.38,75 These departures
from idealised models of fluid mechanics
may reflect changes in viscosity, and/or eas-
ier cell aggregation at slower flow speeds
and as lumen size becomes comparable to
Figure 4. Examples of gross structural abnormalities of the retinal microvasculature cell size.75 The method has also been used
in diabetes. A: Offset-pinhole adaptive optics scanning laser ophthalmoscopy (AOSLO) to study complex flow dynamics such as
showing an area containing a large microaneurysm (red arrow) and adjacent cysts mixing of flow downstream from a venous
(white arrows). Retina surrounding the lesion is disrupted (mottled appearance) confluence;75 this phenomenon has been
which may indicate microcysts. B: Flood AO reflectance image showing a highly tortu- replicated in a number of studies.28,37
ous ‘knot’ structure. C: Motion contrast of the area in B. D: Flood AO reflectance The tendency of blood cells to aggregate
image showing a microaneurysm (red arrow) alongside a hairpin loop (white arrow). within retinal capillaries could be a potent
E: Motion contrast of the area in D, revealing the perfused structure of the lesions. F: biomarker given the suspected impor-
Another microaneurysm (red arrow), fed by a tortuous vessel (white arrow). G: tance of aggregation in a range of disease
Motion contrast of the area in F, revealing poor perfusion of this microaneurysm processes including hypertension, diabe-
(compared to strong perfusion of the lesion in E). Scale bar is 100 μm. AOSLO images tes, metabolic syndrome, coronary dis-
courtesy of Dr Stephen Burns, Indiana University; flood AO images acquired with our ease, stroke, infectious disease and
system. haematological disease.76 Cell aggregation
is typically measured with a blood panel,
observed to form and disappear over time vessel density/connectivity. This advance by necessity a proxy for true in vivo cell
and does not appear to be associated with a was achieved with various optical, hardware, behaviour, but it can be studied directly
thicker wall, suggesting chronic over- and software improvements to AO-OCT sys- with AO. For example, long ‘tails’ form
constriction as the cause.71 Similarly, A/V tems. The choriocapillaris is of central within certain vessels in AOSLO video
nicking has been shown not to indicate importance for a number of disease sequences and are seen to elongate over
physical compression by the arteriole, processes, notably age-related macular time. These are believed to represent dynam-
rather being produced by independent con- degeneration.74 ically aggregating groups of cells. Some ves-
striction of the vein.49,72 sels show much higher tendency to form or
In cases of retinal vasculitis, focal com- traffic aggregates in a way that depends
pression of vessels, opacification of peri- Scientific advances with AO upon their particular morphology.77 The ten-
vascular tissue73 and increased scatter imaging: vessel function dency to form aggregates may therefore be
arising from infiltrates (groups of extrava- altered in disease processes that cause struc-
sating leukocytes) in post-capillary venules In addition to improving understanding of tural adaptations; this has been confirmed in
have been demonstrated and tracked over normal microvascular morphology and the diabetes, where aggregates are seen to
time with AO imaging.71 The microvascula- development of structural malformations, accrete more rapidly prior to the develop-
ture has also been studied in branch venous AO retinal imaging has facilitated significant ment of clinically detectable retinopathy; and
occlusions, where precise longitudinal track- advances in our understanding of vascular in eyes with retinopathy, the increased aggre-
ing of vessels can be used to infer the function. Vascular function can be thought gation is even more pronounced.78
ischaemic status of the retina.64 about in terms of flow at ‘rest’, as well as Individual platelets, which are low con-
‘dynamic’ changes which are brought about trast and in humans only 2.5–3 μm in diam-
Choriocapillaris in response to some stimulus. eter, have been resolved in rodents using
AO-OCT-A is currently the best modality for AOSLO by reconstruction of high-speed cap-
quantitative imaging of choriocapillaris ‘Resting’ flow state illary cross-sections.37 Data from this group
morphology,45 which is very difficult with By freezing the AOSLO scanning raster on a are shown in Figure 2 and demonstrate the
other modalities due not only to the small particular vessel as described above,16 the potential for accurate characterisation of
capillary diameter but also to the interven- cross-sectional velocity of small retinal ves- red cell shape and packing arrangement as
ing retinal pigment epithelium and tight sels can be measured and its evolution well. Platelet aggregation is a key predictor
117
for occlusive vessel disease such as myocar- attempts to infer capillary flow velocity from changes in pressure at one part of the
dial infarction;79 however, it is typically mea- the decorrelation computations employed microvascular network are propagated
sured by mixing a clotting agent with spun for OCT-A.17 downstream. In a completely rigid vessel
blood. As it has been established that the The above discussion deals primarily with the pulse wave would travel at the speed of
natural movement and shear forces that lower frame rate imaging of the microvascu- sound; real vessels have some degree
platelets are exposed to are the predomi- lature; other approaches allow imaging of of compliance, which buffers pressure
nant factor in driving aggregation,80 the typi- the capillary network full-field at high speed, changes and accordingly slows propagation
cal measurement process may not reflect in order to simultaneously determine veloc- of the pulse wave. Loss of compliance
the natural state; the potential for a non- ity in multiple micro-vessels.21,27,39–41 This (increased stiffness) in large vessels such as
invasive, in vivo blood panel afforded by AO allows investigation of the distribution of the aorta, measured as more rapid propa-
imaging may be far more sensitive than cur- flow across the microvascular network in gation of the pulse wave through these
rently available tests. space and time. In our experience, even vessels, is a major risk factor for the devel-
Just as vessels can undergo periods with the high transverse resolution of vessel opment of coronary heart disease and
whereby cell aggregates are passed with no structure afforded by AO, it is only with flow stroke.84 It is possible that the pulse wave
discernible plasma gaps, some vessels information that we are able to unambigu- in smaller vessels is also important; how-
undergo repeated periods of low haematocrit ously map network connectivity (that is, sep- ever, the high speeds and short distances
or acellularity, causing them to transiently dis- arate crosses from true branches); without involved have precluded direct measure-
appear from motion contrast maps.56,81 To this information, the task can probably only ment. In the retina, indirect methods have
illustrate variations in haematocrit, Figure 1D be done with histology.83 Figure 1C illus- produced conflicting estimates of normal
shows the evolution in vessel appearance trates the spatial flow mapping that can be pulse wave velocity which range from 20 to
over time for four segments in a diabetic sub- achieved with AO imaging. 600 mm/second.85,86 However, direct obse-
ject imaged with our flood AO system. The Using AO, it has now been established rvation of the pulse wave may be possible
right-most panel passes long sections of that the cardiac imprint is very pronounced with modern AO imaging. Extrapolating
plasma (bright) broken up sporadically by even at the capillary level, with essentially sampling requirements for measurement
cells (dark), indicating a low haematocrit; a all capillaries undergoing pulsatile variations of cell velocity,27,40 frame rates in the range
more typical haematocrit of ~50 per cent is in velocity between systole and dias- 2,000 to 60,000 fps are suggested. Such
evident in the other three panels. These varia- tole.27,37,38,40,41 However, superimposed on sampling rates are on the order of the line
tions in cellular perfusion occurred despite this pattern are complex changes which scan rate employed by current AO scanning
comparable diameter (Figure 1B) and flow occur due to variations in cell ‘traffic’, in par- systems.38,41,87
speed (Figure 1C) between the vessels. ticular with the passage of leukocytes and Understanding normal capillary flow may
Figure 2 also depicts moment-to-moment var- large aggregates of erythrocytes. Figure 5 be a particularly fruitful field of study
iations in capillary haematocrit. Persistence of demonstrates examples of two patterns of because the introduction of aberrant flow
low haematocrit states has obvious implica- flow: segment ‘1’ (red solid line) undergoes dynamics may precede (or cause) the for-
tions for metabolism and could prove a use- rhythmic flow variations in line with the mation of structural abnormalities. A partic-
ful signpost for disease processes; early expected cardiac pattern (red dotted line). ularly simple area to study would be the
intermittent drops in perfusion may presage In comparison, segment ‘2’ (solid blue line) vessels along the edge of the foveal avascu-
more pronounced capillary dropout later in appears hindered from following its lar zone (FAZ), as they comprise only a sin-
disease. AO imaging has been used to track expected cardiac pattern (dotted blue line), gle layer in depth. Since this area is also
the appearance of non-perfused capillaries being especially slowed at one of the sys- often an early casualty in disease, due per-
over time, primarily in diabetes.63,64,82 By tolic peaks (arrow). The reason for this dis- haps to the reliance of diffusion from the
combining structural and motion contrast ruption to flow is evident in the edge of the FAZ to locations within the cen-
imaging, it is possible to identify ‘ghost’ capil- corresponding single frame shown in tral fovea,67,88 it marks an excellent place to
laries which are present but not perfused by Figure 5B, where a fat, dark cell aggregate is undertake a thorough characterisation of
cells; such vessels have been observed to seen within segment 2. These examples normal network flow dynamics and changes
occur more frequently in diabetic eyes.20,56 underscore the ability of AO imaging to facil- throughout various stages of disease.
The above discussion highlights the itate accurate measurement of vessel diam-
importance of repeated observation to learn eter and other morphology, flow velocity, Stimulus-evoked changes in flow
the true patency of a vessel. This is relevant and contents of the blood column to allow The retina is the most metabolically active
to OCT-A perfusion mapping which has investigation of the ‘rules’ governing the nor- tissue in the body.89 The inner retinal blood
recently seen widespread adoption as a clin- mal trafficking of blood constituents across supply supports fundamental visual
ical and research tool for identification of the capillary network. Despite their funda- processing performed by the bipolar and
capillary dropout. Multiple OCT-A scans are mental importance, these rules have hith- ganglion cells; as the patterns of light falling
recommended to ensure that the wide vari- erto remained poorly understood due to the on the retina change in space or time, the
ations in capillary haematocrit, noted above, lack of accessibility to capillary networks need for information processing increases
do not masquerade as dropout (a similar operating in their natural state. which causes large variations in metabolic
argument can be made regarding the poten- In addition to observing pulsatility of flow activity. The autoregulation of blood flow to
tial for eye movements to skip some vessels velocity in individual vessels, it may be pos- match the ever-changing neuronal activity is
or lesions entirely). Variation in the local sible to measure propagation of the pulse termed ‘neurovascular coupling’. A major
haematocrit is also a major confounder for wave itself, that is to measure how quickly contribution to this process comes from the
118
Figure 5. Patterns of variation in capillary flow over time. A: Motion contrast image generated with flood adaptive optics (AO) at 300 fps
over three seconds, using 750 nm light. Two example segments are labelled (‘1’, ‘2’). B: One frame from the sequence, showing normal
single-file flow through segment 1, and the passage of a large cell aggregate through segment 2. Frame coincides with the systolic peak
and to the arrow in C. C: Velocity over time for segment 1 (solid red line) and segment 2 (solid blue line), alongside scaled averages for the
field (matching dotted lines) which show the pattern expected from variations in cardiac output. At the time point indicated by the
arrow, segment 2 is significantly slowed, coinciding with the presence of the fat, dark aggregate within this segment as seen in B. Scale
bar is 100 μm.
capillaries themselves by way of stimulation More recently, neurovascular coupling flow. This work is still in its infancy, with
of the contractile pericyte cells contained has been studied in healthy vessels < 30 μm many unanswered questions including the
within their walls. Pericytes respond by vari- wide, by measuring changes in vessel diame- relationship between diameter changes and
ous pathways in a feed-forward manner ter in response to locally delivered flickering flow, the manner in which the vascular
triggered by neurotransmitter release.7 light in a spot ~1 across.92 Proportional autoregulation signal is propagated along
Neurovascular coupling may be impli- changes observed were much larger in these the vascular tree,8 and whether deficits in
cated in a number of conditions. Pericyte vessels, under local stimulation, than previ- autoregulation precede the formation of
cells are among the first damaged in ously reported for the larger vessels under structural abnormalities.
diabetes,9 and impaired vascular autore- full-field stimulation (~30 per cent for capil- In contrast to the flicker-evoked changes
gulation is thought to play a major role in laries of < 10 μm diameter and ~12 per cent described above, a similar protocol expl-
dementia and stroke.7 The retina makes for vessels of 10–30 μm diameter, as opp- oring the influence of altered blood gas
particularly fertile ground for study of neu- osed to a few per cent for large retinal ves- (hyperoxia or hypercapnia) produced cha-
rovascular coupling in such conditions.4 sels90). This confirms the key role played by nges in vessel size of comparable degree to
With conventional fundus imaging, the neu- the smallest vessels in local redistribution of those obtained with flicker, but these were
rovascular response to full-field flickering flow and suggests that exploring flow regula- not focally distributed, indicating a different
light has been explored in the large retinal tion in these vessels may provide an even mechanism of action as predicted based on
vessels, eliciting changes in diameter of the more sensitive stress test for the study of current theories of the key pathways
order of a few percent for healthy vessels.90 pathological processes. involved.93
As expected, deficits in the stimulus-evoked In addition to the surprisingly large Both pericytes and glial cells are thought
flow response have been noted in a variety changes in diameter elicited, the above to form a key component of the neuro-
of disease processes.90 work showed that dilations tended to be vascular unit. The magnitude of response
Advances in AO imaging have extended focally distributed (that is, non-uniform) and location of any deficits should ideally
the study of neurovascular coupling to mea- along many vessels, which further impli- be co-localised with the presence and mor-
surement of flow changes elicited for retinal cates pericytes as the agents for change in phology of these important cells. This now
vessels in the 30–80 μm range.91 Using a accordance with their intermittent position- appears to be possible with both mural
local stimulation protocol has highlighted ing along vessels of the calibre studied.92 cells and glial cell end-feet visualised by
the local redistribution of flow to match Additional unique observations made were appropriate manipulation of AOSLO detec-
neuronal activity, with flow increasing only the ability of post-capillary venules to dilate, tion geometry.15,29,31 Pericytes in particular
when the stimulated retina lies within the which was not predicted based on previous could represent a very sensitive biomarker
feeding area of a small arteriole, and pro- ex vivo animal work, and constriction of for diabetes due to their central role in
gressively greater changes in flow occurring some vessels in response to the stimulus pathophysiology of vascular changes in that
the wider the area stimulated. which indicates targeted redistribution of condition.
119
heart disease. Full-field approaches may 5. Burton AC. Physiology and Biophysics of the Circulation:
Future directions An Introductory Text. Chicago: Year Book Medical Pub-
allow for study of the propagation of vascu- lishers, 1972.
lar autoregulation signals across the vascu- 6. Snyder GK, Sheafor BA. Red blood cells: centerpiece in
Many investigations that have been carried lar tree. Clever manipulation of detection the evolution of the vertebrate circulatory system. Am
out by AO retinal imaging research laborato- Zool 1999; 39: 189–198.
geometry in scanning modalities continues 7. Hamilton NB, Attwell D, Hall CN. Pericyte-mediated
ries are not yet amenable to routine clinical to reveal previously invisible structures such regulation of capillary diameter: a component of neu-
assessment in large numbers of patients. as pericytes and glial cells, which could be rovascular coupling in health and disease. Front Neu-
Factors limiting the broader applicability of roenergetics 2010; 2: 5.
co-localised with flow deficits and so offer 8. Bosetti F, Galis ZS, Bynoe MS et al. ‘Small blood ves-
AO imaging include: the cost, complexity greater explanatory power and more sensi- sels: big health problems?’: scientific recommenda-
and immobility of the systems; specialist tive biomarkers for disease; developments tions of the National Institutes of Health workshop.
J Am Heart Assoc 2016; 5: e004389.
expertise required in hardware deployment in AO-OCT promise to reveal still further 9. Cogan DG, Toussaint D, Kuwabara T. Retinal vascular
and operation; lack of standardised soft- details of ‘invisible’ cell classes. Multispectral patterns: IV. Diabetic retinopathy. Arch Ophthalmol
ware; small fields imaged; small fixational methods may facilitate oximetry for individ-
1961; 66: 366–378.
10. Berry C, Sidik N, Pereira AC et al. Small-vessel disease
eye movements (which are large relative to ual capillaries and even individual blood in the heart and brain: current knowledge, unmet
the objects of interest); sensitivity to scatter cells. therapeutic need, and future directions. J Am Heart
or distorted ocular media; and need for Assoc 2019; 8: e011104.
The ultimate goal is for quick, broadly 11. Dintenfass L. Inversion of the Fahraeus–Lindqvist phe-
manually assisted analysis of data including applicable, non-invasive, high spatiotempo- nomenon in blood flow through capillaries of
identification of novel lesion types. Nonethe- ral resolution observation of retinal vascula- diminishing radius. Nature 1967; 215: 1099.
less, inroads have been made in deploying 12. Fahraeus R, Lindqvist T. The viscosity of the blood in
ture structure and function to robustly narrow capillary tubes. Am J Physiol 1931; 96:
the technology in clinical paradigms. For determine an individual’s vascular health in 562–568.
example, there is a clinical device available its natural state.8 Adaptive optics extends 13. Wolf S, Arend O, Reim M. Measurement of retinal
hemodynamics with scanning laser ophthalmoscopy:
from Imagine Eyes (the ‘rxt1’), with some the vessels that can be imaged to the finest reference values and variation. Surv Ophthalmol 1994;
257 citable entries returned by Google capillary networks. Detailed measurement 38: S95–S100.
Scholar from 2010 to 2018 for the terms of structure, function and stimulus-evoked
14. Roorda A, Romero-Borja F, Donnelly WJ III et al. Adap-
tive optics scanning laser ophthalmoscopy. Opt
‘rtx1’ and ‘Imagine Eyes’. However, routine changes will allow better understanding of Express 2002; 10: 405–412.
deployment in clinical practice is unlikely to normal microvascular function as well as 15. Burns SA, Elsner AE, Sapoznik KA et al. Adaptive optics
come until there is widespread uptake in imaging of the human retina. Prog Retin Eye Res 2019;
the natural history of disease, including pre- 68: 1–30.
ophthalmic research, which may be several clinical variations which may predict the 16. Zhong Z, Petrig BL, Qi X et al. In vivo measurement of
years away as the limitations identified development of severe complications of erythrocyte velocity and retinal blood flow using
above are gradually overcome. Develop- adaptive optics scanning laser ophthalmoscopy. Opt
both retinal disease and systemic co-mor- Express 2008; 16: 12746–12756.
ments in commercial instrumentation such bidities. Such investigations, if successful, 17. Wang RK, Zhang Q, Li Y et al. Optical coherence
as OCT-A have greatly assisted research in tomography angiography-based capillary velocimetry.
would have major impacts on early diagno-
AO imaging, offering readily accessible J Biomed Opt 2017; 22: 066008.
sis and accurate prognosis for individual 18. Nesper PL, Fawzi AA. Human parafoveal capillary vas-
widefield and depth information to inform cular anatomy and connectivity revealed by optical
patients, and offer sensitive, non-invasive,
targeted study of smaller areas with AO. coherence tomography angiography. Invest
longitudinally accessible biomarkers for clin- Ophthalmol Vis Sci 2018; 59: 3858–3867.
In contrast to widespread clinical adoption
ical trials of novel therapies targeting the 19. Riva CE, Grunwald JE, Sinclair SH et al. Blood velocity
of AO technology, there are many exciting and volumetric flow rate in human retinal vessels.
vasculature to facilitate optimal capillary
avenues for further development of AO Invest Ophthalmol Vis Sci 1985; 26: 1124–1132.
perfusion. 20. Mo S, Krawitz B, Efstathiadis E et al. Imaging
imaging methods as a research tool to
foveal microvasculature: optical coherence
develop scientific understanding of retinal tomography angiography versus adaptive optics
ACKNOWLEDGEMENTS scanning light ophthalmoscope fluorescein angiog-
vascular structure and function, on the cel-
The authors would like to thank Dr Stephen raphy. Invest Ophthalmol Vis Sci 2016; 57:
lular scale, in health and disease. Precise OCT130–OCT140.
Burns (Indiana University) and Dr Jesse
and repeatable measurements facilitate lon- 21. Bedggood P, Metha A. Analysis of contrast and
Schallek (University of Rochester) for supply- motion signals generated by human blood con-
gitudinal study of various disease processes
ing sample AOSLO image data presented in stituents in capillary flow. Opt Lett 2014; 39:
as highlighted above. Recent improvements 610–613.
this review.
in technology allow faster frame rates which 22. Bedggood P, Daaboul M, Ashman RA et al. Character-
Research supported by Australian Research istics of the human isoplanatic patch and implications
have revealed elaborate flow patterns
Council Discovery Project ARC DP180103393. for adaptive optics retinal imaging. J Biomed Opt 2008;
across the microvascular network; this infor- 13: 024008.
mation could be used to seed models which 23. Liang J, Williams DR, Miller DT. Supernormal vision
REFERENCES and high-resolution retinal imaging through adaptive
identify aberrant flow profiles in individual 1. Cheung N, Wong TY. Diabetic retinopathy and sys- optics. J Opt Soc Am A 1997; 14: 2884–2892.
vessels, revealing which vessels in which temic vascular complications. Prog Retin Eye Res 2008; 24. Roorda A, Williams DR. The arrangement of the three
patients will give way to non-perfusion and 27: 161–176. cone classes in the living human eye. Nature 1999;
2. Ong Y-T, Wong TY, Klein R et al. Hypertensive retinop- 397: 520.
eventually more widespread disease compli- athy and risk of stroke. Hypertension 2013; 62: 25. Tam J, Martin JA, Roorda A. Noninvasive visualization
cations. Rapid imaging also has the potential 706–711. and analysis of parafoveal capillaries in humans.
3. Witt N, Wong TY, Hughes AD et al. Abnormalities of Invest Ophthalmol Vis Sci 2010; 51: 1691–1698.
to observe propagation of the pulse wave
retinal microvascular structure and risk of mortality 26. Chui TY, Zhong Z, Song H et al. Foveal avascular zone
in retinal micro-vessels, which would from ischemic heart disease and stroke. Hypertension and its relationship to foveal pit shape. Optom Vis Sci
provide crucial and otherwise unobtainable 2006; 47: 975–981. 2012; 89: 602.
4. Cheung CY-l, Ikram MK, Chen C et al. Imaging retina 27. Bedggood P, Metha A. Mapping flow velocity in the
information on microvascular compliance to study dementia and stroke. Prog Ret Eye Res 2017; human retinal capillary network with pixel intensity
(or stiffness) relevant to stroke and coronary 57: 89–107. cross correlation. PLoS One 2019; 14: e0218918.
120
28. Chui TY, VanNasdale DA, Burns SA. The use of for- pressure and focal vascular changes. J Hypertens adaptive optics scanning laser ophthalmoscopy.
ward scatter to improve retinal vascular imaging with 2014; 32: 890. Biomed Opt Express 2016; 7: 4958–4973.
an adaptive optics scanning laser ophthalmoscope. 50. Meixner E, Michelson G. Measurement of retinal wall- 70. Karst SG, Lammer J, Radwan SH et al. Characterization
Biomed Opt Express 2012; 3: 2537–2549. to-lumen ratio by adaptive optics retinal camera: a of in vivo retinal lesions of diabetic retinopathy using
29. Chui TY, Gast TJ, Burns SA. Imaging of vascular wall clinical research. Graefes Arch Clin Exp Ophthalmol adaptive optics scanning laser ophthalmoscopy. Int J
fine structure in the human retina using adaptive 2015; 253: 1985–1995. Endocrinol 2018; 2018: 7492946.
optics scanning laser ophthalmoscopy. Invest 51. Arichika S, Uji A, Ooto S et al. Effects of age and blood 71. Paques M, Meimon S, Rossant F et al. Adaptive optics
Ophthalmol Vis Sci 2013; 54: 7115–7124. pressure on the retinal arterial wall, analyzed using ophthalmoscopy: application to age-related macular
30. Sulai YN, Scoles D, Harvey Z et al. Visualization of reti- adaptive optics scanning laser ophthalmoscopy. Sci degeneration and vascular diseases. Prog Retin Eye
nal vascular structure and perfusion with a nonconfocal Rep 2015; 5: 12283. Res 2018; 66: 1–16.
adaptive optics scanning light ophthalmoscope. J Opt 52. Rosenbaum D, Kachenoura N, Koch E et al. Relation- 72. Paques M, Brolly A, Benesty J et al. Venous nicking
Soc Am A 2014; 31: 569–579. ships between retinal arteriole anatomy and aortic without arteriovenous contact: the role of the arterio-
31. Sapoznik KA, Luo T, De Castro A et al. Enhanced geometry and function and peripheral resistance in lar microenvironment in arteriovenous nickings. JAMA
retinal vasculature imaging with a rapidly con- hypertensives. Hypertens Res 2016; 39: 536. Ophthalmol 2015; 133: 947–950.
figurable aperture. Biomed Opt Express 2018; 9: 53. Hillard JG, Gast TJ, Chui TY et al. Retinal arterioles in 73. Errera M-H, Coisy S, Fardeau C et al. Retinal vasculitis
1323–1333. hypo-, normo-, and hypertensive subjects measured imaging by adaptive optics. Ophthalmology 2014; 121:
32. Gofas-Salas E, Mece P, Mugnier L et al. Near infrared using adaptive optics. Transl Vis Sci Technol 2016; 1311–1312.e2.
adaptive optics flood illumination retinal angiography. 5: 16. 74. McLeod DS, Grebe R, Bhutto I et al. Relationship
Biomed Opt Express 2019; 10: 2730–2743. 54. Rosenbaum D, Mattina A, Koch E et al. Effects of age, between RPE and choriocapillaris in age-related mac-
33. Martin JA, Roorda A. Pulsatility of parafoveal capillary blood pressure and antihypertensive treatments on ular degeneration. Invest Ophthalmol Vis Sci 2009; 50:
leukocytes. Exp Eye Res 2009; 88: 356–360. retinal arterioles remodeling assessed by adaptive 4982–4991.
34. Tam J, Roorda A. Speed quantification and tracking of optics. J Hypertens 2016; 34: 1115–1122. 75. Zhong Z, Song H, Chui TYP et al. Noninvasive mea-
moving objects in adaptive optics scanning laser oph- 55. Burgansky-Eliash Z, Barak A, Barash H et al. Increased surements and analysis of blood velocity profiles in
thalmoscopy. J Biomed Opt 2011; 16: 036002. retinal blood flow velocity in patients with early diabe- human retinal vessels. Invest Ophthalmol Vis Sci 2011;
35. Martin JA, Roorda A. Direct and noninvasive assess- tes mellitus. Retina 2012; 32: 112–119. 52: 4151–4157.
ment of parafoveal capillary leukocyte velocity. Oph- 56. Burns SA, Elsner AE, Chui TY et al. In vivo adaptive 76. Baskurt OK, Meiselman HJ. Erythrocyte aggregation:
thalmology 2005; 112: 2219–2224. optics microvascular imaging in diabetic patients with- basic aspects and clinical importance. Clin Hemorheol
36. Arichika S, Uji A, Hangai M et al. Noninvasive and out clinically severe diabetic retinopathy. Biomed Opt Microcirc 2013; 53: 23–37.
direct monitoring of erythrocyte aggregates in human Express 2014; 5: 961–974. 77. Arichika S, Uji A, Ooto S et al. Adaptive optics-assisted
retinal microvasculature using adaptive optics scan- 57. Lombardo M, Parravano M, Serrao S et al. Analysis of identification of preferential erythrocyte aggregate
ning laser ophthalmoscopy. Invest Ophthalmol Vis Sci retinal capillaries in patients with type 1 diabetes and pathways in the human retinal microvasculature. PLoS
2013; 54: 4394–4402. nonproliferative diabetic retinopathy using adaptive One 2014; 9: e89679.
37. Guevara-Torres A, Joseph A, Schallek J. Label free optics imaging. Retina 2013; 33: 1630–1639. 78. Arichika S, Uji A, Murakami T et al. Retinal hemorheologic
measurement of retinal blood cell flux, velocity, 58. Sherman TF. On connecting large vessels to small. characterization of early-stage diabetic retinopathy using
hematocrit and capillary width in the living mouse The meaning of Murray’s law. J Gen Physiol 1981; 78: adaptive optics scanning laser ophthalmoscopy. Invest
eye. Biomed Opt Express 2016; 7: 4228–4249. 431–453. Ophthalmol Vis Sci 2014; 55: 8513–8522.
38. Joseph A, Guevara-Torres A, Schallek J. Imaging single- 59. Luo T, Gast TJ, Vermeer TJ et al. Retinal vascular 79. Trip MD, Cats VM, van Capelle FJ et al. Platelet
cell blood flow in the smallest to largest vessels in the branching in healthy and diabetic subjects. Invest hyperreactivity and prognosis in survivors of myocar-
living retina. Elife 2019; 8: e45077. Ophthalmol Vis Sci 2017; 58: 2685–2694. dial infarction. N Engl J Med 1990; 322: 1549–1554.
39. Bedggood P, Metha A. Direct visualization and charac- 60. Tam J, Dhamdhere KP, Tiruveedhula P et al. Disrup- 80. Nesbitt WS, Westein E, Tovar-Lopez FJ et al. A shear
terization of erythrocyte flow in human retinal capil- tion of the retinal parafoveal capillary network in type gradient–dependent platelet aggregation mecha-
laries. Biomed Opt Express 2012; 3: 3264–3277. 2 diabetes before the onset of diabetic retinopathy. nism drives thrombus formation. Nat Med 2009;
40. Gu B, Wang X, Twa MD et al. Noninvasive in vivo char- Invest Ophthalmol Vis Sci 2011; 52: 9257–9266. 15: 665.
acterization of erythrocyte motion in human retinal 61. Dubow M, Pinhas A, Shah N et al. Classification of 81. Pinhas A, Razeen M, Dubow M et al. Assessment of
capillaries using high-speed adaptive optics near- human retinal microaneurysms using adaptive optics perfused foveal microvascular density and identifica-
confocal imaging. Biomed Opt Express 2018; 9: scanning light ophthalmoscope fluorescein angiogra- tion of nonperfused capillaries in healthy and
3653–3677. phy. Invest Ophthalmol Vis Sci 2014; 55: 1299–1309. vasculopathic eyes. Invest Ophthalmol Vis Sci 2014; 55:
41. de Castro A, Huang G, Sawides L et al. Rapid high res- 62. Lammer J, Karst SG, Lin MM et al. Association of 8056–8066.
olution imaging with a dual-channel scanning tech- Microaneurysms on adaptive optics scanning laser 82. Tam J, Dhamdhere KP, Tiruveedhula P et al. Subclini-
nique. Opt Lett 2016; 41: 1881–1884. ophthalmoscopy with surrounding Neuroretinal cal capillary changes in non proliferative diabetic reti-
42. Gorczynska I, Migacz JV, Zawadzki RJ et al. Comparison pathology and visual function in diabetes. Invest nopathy. Optom Vis Sci 2012; 89: E692.
of amplitude-decorrelation, speckle-variance and Ophthalmol Vis Sci 2018; 59: 5633–5640. 83. Chandrasekera E, An D, McAllister IL et al. Three-
phase-variance OCT angiography methods for imag- 63. Chui TYP, Pinhas A, Gan A et al. Longitudinal imaging of dimensional microscopy demonstrates series and
ing the human retina and choroid. Biomed Opt Express microvascular remodelling in proliferative diabetic reti- parallel Organization of Human Peripapillary Capil-
2016; 7: 911–942. nopathy using adaptive optics scanning light ophthal- lary Plexuses. Invest Ophthalmol Vis Sci 2018; 59:
43. Salas M, Augustin M, Ginner L et al. Visualization of moscopy. Ophthalmic Physiol Opt 2016; 36: 290–302. 4327–4344.
micro-capillaries using optical coherence tomography 64. Chui TY, Mo S, Krawitz B et al. Human retinal micro- 84. Mattace-Raso FU, van der Cammen TJ, Hofman A
angiography with and without adaptive optics. Biomed vascular imaging using adaptive optics scanning light et al. Arterial stiffness and risk of coronary heart
Opt Express 2017; 8: 207–222. ophthalmoscopy. Int J Retina Vitreous 2016; 2: 11. disease and stroke. Circulation 2006; 113:
44. Zawadzki RJ, Choi SS, Jones SM et al. Adaptive optics- 65. Nunes S, Pires I, Rosa A et al. Microaneurysm turn- 657–663.
optical coherence tomography: optimizing visualization over is a biomarker for diabetic retinopathy progres- 85. Li Q, Li L, Fan S et al. Retinal pulse wave velocity mea-
of microscopic retinal structures in three dimensions. sion to clinically significant macular edema: findings surement using spectral-domain optical coherence
J Opt Soc Am A 2007; 24: 1373–1383. for type 2 diabetics with nonproliferative retinopathy. tomography. J Biophoton 2018; 11: e201700163.
45. Kurokawa K, Liu Z, Miller DT. Adaptive optics optical Ophthalmologica 2009; 223: 292–297. 86. Spahr H, Hillmann D, Hain C et al. Imaging pulse wave
coherence tomography angiography for morphomet- 66. Schreur V, Domanian A, Liefers B et al. Morphological propagation in human retinal vessels using full-field
ric analysis of choriocapillaris. Biomed Opt Express and topographical appearance of microaneurysms on swept-source optical coherence tomography. Opt Lett
2017; 8: 1803–1822. optical coherence tomography angiography. Br J 2015; 40: 4771–4774.
46. Liu Z, Tam J, Saeedi O et al. Trans-retinal cellular Ophthalmol 2019; 103: 630–635. 87. Lu J, Gu B, Wang X et al. Adaptive optics parallel near-
imaging with multimodal adaptive optics. Biomed Opt 67. Fu X, Gens JS, Glazier JA et al. Progression of diabetic confocal scanning ophthalmoscopy. Opt Lett 2016; 41:
Express 2018; 9: 4246–4262. capillary occlusion: a model. PLoS Comput Biol 2016; 3852–3855.
47. Hillmann D, Spahr H, Hain C et al. Aberration-free vol- 12: e1004932. 88. Alibhai AY, Moult EM, Shahzad R et al. Quantifying
umetric high-speed imaging of in vivo retina. Sci Rep 68. Bernabeu MO, Lu Y, Abu-Qamar O et al. Estimation of microvascular changes using OCT angiography in dia-
2016; 6: 35209. diabetic retinal microaneurysm perfusion parameters betic eyes without clinical evidence of retinopathy.
48. Rizzoni D, Agabiti-Rosei E. Structural abnormalities of based on computational fluid dynamics modeling of Ophthalmol Retina 2018; 2: 418–427.
small resistance arteries in essential hypertension. adaptive optics scanning laser ophthalmoscopy. Front 89. Wangsa-Wirawan ND, Linsenmeier RA. Retinal oxygen:
Intern Emerg Med 2012; 7: 205–212. Physiol 2018; 9: 989. fundamental and clinical aspects. Arch Ophthalmol
49. Koch E, Rosenbaum D, Brolly A et al. Morphometric 69. Lu Y, Bernabeu MO, Lammer J et al. Computational 2003; 121: 547–557.
analysis of small arteries in the human retina using fluid dynamics assisted characterization of parafoveal 90. Riva CE, Logean E, Falsini B. Visually evoked
adaptive optics imaging: relationship with blood hemodynamics in normal and diabetic eyes using hemodynamical response and assessment of
121
neurovascular coupling in the optic nerve and retina. 92. Duan A, Bedggood PA, Bui BV et al. Evidence of 93. Duan A, Bedggood PA, Metha AB et al. Reactivity in
Prog Retin Eye Res 2005; 24: 183–215. flicker-induced functional hyperaemia in the smallest the human retinal microvasculature measured during
91. Zhong Z, Huang G, Chui TYP et al. Local flicker stimulation vessels of the human retinal blood supply. PLoS One acute gas breathing provocations. Sci Rep 2017; 7:
evokes local retinal blood velocity changes. J Vis 2012; 12: 3. 2016; 11: e0162621. 2113.
122
INVITED RESEARCH
The influence of orthokeratology compression factor on ocular

higher-order aberrations
Jason K Lau* BOptom Background: To investigate the influence of compression factor upon changes in ocular
Stephen J Vincent† PhD higher-order aberrations (HOAs) in young myopic children undergoing orthokeratology
Sin-Wan Cheung* PhD treatment.
Pauline Cho* PhD Methods: Subjects aged between six and < 11 years, with low myopia (0.50–4.00 D inclu-
*Centre for Myopia Research, School of Optometry, sive), low astigmatism (≤ 1.25 D), and anisometropia (≤ 1.00 D), were randomly assigned to
The Hong Kong Polytechnic University, Hong Kong wear orthokeratology lenses of different compression factors in each eye (one eye 0.75 D
Special Administrative Region, China and the fellow eye 1.75 D). HOAs were measured weekly over one month of lens wear.
†
Contact Lens and Visual Optics Laboratory, School of Wavefront analysis was conducted over a 5-mm pupil using a sixth order Zernike polyno-
Optometry and Visual Science, Queensland University mial expansion. Linear mixed models were used to examine the individual Zernike co-
of Technology, Brisbane, Queensland, Australia efficients and specific root-mean-square (RMS) error (spherical, comatic, total HOAs) metrics
E-mail: [email protected]
and their changes between the two eyes during the study period.
Results: Twenty-eight myopic (mean manifest spherical equivalent refraction: −2.10 0.58 D)
children (median [range] age: 9.3 [7.8–11.0] years) were analysed. Significant interocular dif-
ferences in HOAs at baseline were observed for Z 6−6 and Z 6−4 only (both p < 0.05). During the
lens wear period, eyes fitted with the increased compression factor showed greater changes
in primary spherical aberration (Z 04 , p = 0.04) and RMS values for spherical and total HOAs
(both p < 0.01). Considering data from both eyes together, after adjusting for the paired
nature of the data, some other Zernike terms (Z 13 and Z 06 , both p < 0.01) and the RMS value
of comatic aberrations (p < 0.001) significantly increased after one month of orthokeratology
treatment. The increase in primary spherical aberration (Z 04 ) was positively correlated with
the reduction in spherical equivalent refractive error, but only in eyes fitted with the
increased compression factor (r = 0.69, p < 0.001).
Conclusions: Increasing the orthokeratology compression factor by 1.00 D significantly
Submitted: 16 March 2019 altered some HOAs, particularly spherical aberration. Given the association between posi-
Revised: 20 May 2019 tive spherical aberration and eye growth in children, further research investigating the influ-
Accepted for publication: 21 May 2019 ence of orthokeratology compression factor on axial eye growth is warranted.
Key words: compression factor, higher-order aberrations, orthokeratology
Orthokeratology is an established treatment correcting factor, known as the compression full correction; however, currently no stud-
for paediatric myopia control1–3 and is popu- factor (that is, the Jessen factor), in addition ies have investigated the feasibility, safety,
lar among both practitioners4 and parents.5,6 to the correction for myopia, to counteract or optical outcomes of increasing the com-
It utilises reverse geometry rigid gas perme- this refractive regression to ensure good pression factor by 1.00 D in children.
able lenses worn overnight which flatten the unaided vision and patient satisfaction As the corneal shape is altered during ortho-
central cornea and steepen the mid- throughout the entire day. keratology treatment, both corneal and ocular
peripheral cornea,7 resulting in daytime myo- Despite the use of a conventional com- aberrations, primarily spherical and comatic
pia correction and provides children with an pression factor (0.75 D), under-correction of aberrations, significantly increase.15–21 Hiraoka
increased quality of life due to improved myopia (of about 0.50–0.75 D) has been et al.22 investigated the correlations between
unaided vision and convenience.5,6 However, reported10–13 and researchers have postu- ocular aberrations and axial elongation in
upon lens removal in the morning, the lated that the correcting factor was most 55 children (mean age: 10.3 1.4 years)
induced corneal changes begin to regress likely underestimated. Chan et al.14 undergoing orthokeratology treatment for one
and may result in approximately 0.50–0.75 D analysed the refractive outcome of their year and showed that children with greater
under-correction toward the end of the myopic children and suggested that an addi- increases in the root-mean-square (RMS) error
day.8–11 Therefore, most orthokeratology tional 1.00 D should be incorporated (that values for spherical, comatic, and total higher-
lens manufacturers incorporate an extra is, a 1.75 D compression factor) to achieve order aberrations (HOAs) exhibited slower
123
Orthokeratology compression factor and higher-order aberrations Lau, Vincent, Cheung et al
axial eye growth. Theoretically, a greater Lenses and solutions 125 measurements of monochromatic HOAs
change in HOAs should be induced during Four-zone orthokeratology lenses (Menicon Z were acquired (555 nm wavelength), and
orthokeratology treatment when the amount Night or Menicon Z Night Toric lenses; NKL later averaged.
of myopia corrected is higher, as a greater Contactlenzen B.V., Emmen, The Netherlands)
change in corneal shape is necessary to with a Dk of 163 (ISO unit) were used. Either Wavefront analysis
achieve the desired refractive correction (that spherical or toric lenses, based on the manu- The wavefront data acquired from COAS
is, more reduction in corneal sphericity).23 facturer’s software (Easyfit, version 2013; NKL was fitted with a sixth order polynomial and
Kang et al.24 attempted to alter corneal Contactlenzen B.V.), were fitted to each subject the Zernike co-efficients were rescaled to a
HOAs by changing the optic zone diameter (that is, the same lens design was used for 5-mm pupil through interpolation. The RMS
(from 6 mm to 5 mm) and the peripheral tan- both eyes). An extra 1.00 D was added to the of spherical (Z 04 and Z 06 combined), comatic
gent (from 1/4 to 1/2) while controlling for lens target for myopia correction in eyes (Z 3− 1 , Z 13 , Z 5− 1 , and Z 15 combined), and total
centration and refractive correction; however, randomised to wear the increased compres- HOAs (from third to sixth radial orders,
no significant difference in spherical aberration sion factor (1.75 D) lens, while the fellow eyes inclusive) were also calculated. The signs of
(Z 04 ) was found. Chen et al.25 also hypo- wore lenses with the default compression fac-
Zernike terms (Z 13 , Z 33 , Z 4− 4 , Z 4− 2 , Z 15 , Z 35 , Z 55 ,
thesised that modifying the lens diameter tor of 0.75 D.
Z 6− 6 , Z 6−4 , and Z 6−2 ) for the left eyes were
may be useful to alter the HOA profile, par- Subjects were required to wear the lenses
reversed to account for enantiomorphism
ticularly vertical coma, but to date no stud- every night, and to perform daily cleaning
(mirror symmetry) between the two
ies have systematically examined the effect and disinfection and weekly protein removal
eyes.27,28
of total lens diameter on the changes procedures (cleaning: Menicon Spray and
induced in corneal optics. Clean; rinsing: Ophtecs cleadew; disinfec-
The aim of this study was therefore to com- tion: Menicare Plus; protein removal: Sample size determination
pare the changes in ocular HOAs in young Menicon Progent). Artificial tears (Precilens The sample size was calculated with
myopic children wearing orthokeratology Aquadrop+) were also provided to avoid G*Power (version 3.1.9.2; Kiel University, Kiel,
lenses of different compression factors in the bubbles trapped underneath the lens and Germany) based on the average apical cor-
two eyes over a one-month period. facilitate lens removal when necessary. neal power difference anticipated between
the two eyes during orthokeratology treat-
Examination visits ment. A minimum interocular difference of
Lenses were delivered at the baseline visit 0.50 D was expected with a within-subject
Methods
and subsequent weekly data collection visits standard deviation of 0.70 D.18 Therefore, a
over one month were scheduled at a similar minimum of 18 subjects were required to
Study design provide 80 per cent power to detect a signifi-
time of day (two hours) to avoid the poten-
This was a double-blind, contralateral, self- cant difference with an alpha level of 0.05.
tial influence of diurnal variation on out-
controlled study investigating the effect of
come measures. The early morning visit
different orthokeratology compression fac-
(within two hours after waking) after the Statistical analyses
tors (0.75 D and 1.75 D) on the changes in
first overnight lens wear and any additional All statistical analyses were performed using
ocular HOAs. The procedures followed the
unscheduled visits were arranged when nec- SPSS version 23 (IMB Corp., Armonk, NY,
Declaration of Helsinki and the study was
essary to maintain good vision and ocular USA). The normality of the baseline demo-
approved by the Departmental Research
health throughout the study period. graphics, ocular HOA Zernike co-efficients,
Committee of the School of Optometry at
and their changes at the one-month visit
The Hong Kong Polytechnic University.
Data collection were checked with Shapiro–Wilk tests.
Informed consent of the parents was
High-contrast visual acuity (Early Treatment Paired t-tests or Wilcoxon tests, where
obtained after thorough explanation of the
Diabetic Retinopathy Study charts, 90 per appropriate, were used to compare the
nature and possible consequences of the
cent contrast; Precision Vision, Woodstock, baseline differences or changes (at the one-
study. The study was registered at
IL, USA), non-cycloplegic subjective refrac- month visit) between eyes. Linear mixed
ClinicalTrial.gov (NCT02643875).
tion, slit-lamp biomicroscopy, and Medmont models were used to assess the effect of dif-
corneal topography were conducted at each ferent compression factors on HOAs over
Subjects visit to monitor lens performance, ocular time, with restricted maximum likelihood
Chinese subjects aged between six and health, and vision. External ocular health estimation and a first-order autoregressive
< 11 years, with low myopia (0.50–4.00 D), conditions were graded according to the covariance structure.
low astigmatism (≤ 1.25 D), and anisometro- Efron grading system. Estimated marginal means, adjusted with
pia ≤ 1.00 D, were recruited. Those with high Ocular HOAs were measured from each Bonferroni corrections, are presented for
corneal toricity (≥ 2.00 D), a history of previ- eye using a Shack-Hartmann aberrometer significant between-eye differences; other-
ous myopia control treatments, any ocular (COAS; Wavefront Sciences Ltd., Albuquer- wise, the results are presented considering
or systemic diseases that may affect refrac- que, NM, USA) through natural pupils under both eyes together, after adjustment for
tive development or contact lens wear, or scotopic conditions (five lux) with the fellow paired-eye data. Changes in the Zernike co-
were non-compliant with lens wear or eye occluded. A Badal optometer,26 mounted efficients and RMS values from baseline at
related procedures, were excluded. Lenses on the COAS machine, was set with the each visit were also compared between the
were only ordered if the subjects demon- spherical equivalent refraction of the subject two eyes using paired t-tests or equivalent
strated good lens-handling skills. to control for accommodation. For each eye, (for significant terms in the linear mixed
124
models). Pearson or Spearman correlations,

when appropriate, were then used to assess Compression factor p-value
1.75 D 0.75 D
the associations between changes in these
Zernike co-efficients and changes in spheri- Visual acuities, logMAR
cal equivalent refraction. A p-value of less Unaided 0.63 0.29 0.65 0.31 0.387†
than 0.05 was considered statistically Best-corrected 0.00 0.04 0.00 0.05 0.678†
significant. Refraction, D
Myopia −2.09 0.97 −2.12 0.94 0.714†
Astigmatism −0.50 (−1.25, 0.00) 0.00 (−1.25, 0.00) 0.080‡
Results SER −2.30 1.03 −2.27 0.99 0.646†
Individual Zernike terms, μm
Thirty-six subjects were randomised and
Z 3−3 0.025 0.080 0.012 0.071 0.327†
completed the baseline data collection.
Z 3−1 −0.019 0.114 0.000 0.113 0.185†
However, eight subjects were excluded due
Z 13 0.002 (−0.094, 0.110) −0.013 (−0.104, 0.083) 0.425‡
to failure to adapt to lens wear (one), bro-
ken lens (one), poor vision (one), and loss to Z 33 −0.008 0.066 −0.009 0.059 0.938†
follow-up (two). An additional three subjects Z 4−4 0.022 0.022 0.020 0.025 0.604†
were excluded because of a poor lens fit, Z 4−2 −0.014 (−0.030, 0.026) −0.006 (−0.039, 0.027) 0.053‡
leaving 28 subjects included in the final ana- Z 04 0.057 (−0.019, 0.212) 0.067 (−0.020, 0.376) 0.633‡
lyses, consisting of 12 boys and 16 girls, with Z 24 −0.003 0.028 −0.006 0.031 0.468†
median (range) age of 9.3 (7.8–11.0) years. Z 44 0.007 0.022 0.013 0.024 0.146†
Eighteen subjects were fitted with spherical 0.964‡
Z 5−5 0.002 (−0.017, 0.038) 0.001 (−0.020, 0.031)
lenses and 10 with toric lenses. −0.003 (−0.043, 0.020) −0.004 (−0.070, 0.012) 0.585‡
Z 5−3
Table 1 displays the mean baseline visual
Z 5−1 0.009 (−0.015, 0.069) 0.007 (−0.018, 0.035) 0.076‡
acuities, refraction, and ocular HOAs for the
Z 15 0.001 (−0.014, 0.013) 0.001 (−0.046, 0.011) 0.633‡
two eyes. There were no significant baseline
Z 35 0.001 0.005 0.002 0.005 0.449†
interocular differences for refraction,
unaided visual acuities, or best-corrected Z 55 0.002 (−0.015, 0.031) 0.004 (−0.013, 0.047) 0.982‡
visual acuities (all p > 0.05). Significant inter- Z 6−6 0.002 (−0.005, 0.009) −0.001 (−0.012, 0.020) 0.031‡
ocular differences were observed at the Z 6−4 −0.003 0.002 −0.001 0.003 < 0.001†
baseline visit for some higher-order Zernike Z 6−2 0.000 (−0.008, 0.007) 0.000 (−0.006, 0.018) 0.274‡
terms (Z 6−6 and Z 6− 4 , both p < 0.05). Z 06 −0.005 (−0.016, 0.026) −0.005 (−0.016, 0.100) 0.633‡
Z 26 −0.001 (−0.026, 0.010) 0.000 (−0.021, 0.024) 0.274‡
Effect of compression factor Z 46 0.001 (−0.006, 0.015) −0.001 (−0.067, 0.023) 0.106‡
After one week of lens wear, eyes fitted with Z 66 −0.002 0.004 −0.001 0.006 0.330†
the increased compression factor showed a Root-mean-squares (RMS), μm
greater initial refractive reduction than the RMS SA 0.057 (0.004, 0.213) 0.068 (0.006, 0.389) 0.187‡
fellow eyes by 0.44 0.66 D (p = 0.001). RMS coma 0.116 0.064 0.107 0.059 0.411†
However, this difference diminished with RMS HOAs 0.194 (0.102, 0.316) 0.175 (0.098, 0.536) 0.439‡
time (mean differences: week two: Bold value indicates statistically significant value.
0.39 0.65 D, week three: 0.31 0.54 D; HOA: higher-order aberrations; RMS: root-mean-square; RMS coma: Z 3−1 , Z 13 , Z 5−1 , and
both p < 0.01), with no significant difference Z 15 combined; RMS HOAs: from third to sixth orders (inclusive); RMS SA: Z 04 and Z 06
in refractive correction observed between combined; SA: spherical; SER: spherical equivalent refraction.
the eyes at the one-month visit (mean †
Paired t-test.
spherical equivalent reduction: increased ‡
Wilcoxon test.
compression factor: 2.58 0.89 D, conven-
tional compression factor: 2.36 0.83 D;
mean difference: 0.22 0.68 D, p = 0.10). Table 1. Baseline refraction and ocular aberrations, mean SD or median (min,
Figure 1 shows the changes in Zernike co- max), of the eyes fitted with different orthokeratology compression factors
efficients and RMS values with significant dif-
ferences between the two eyes during the
study period. Eyes fitted with the increased
compression factor showed a significantly 0.147 0.176 μm, p < 0.001) after one week displayed greater increases in these terms
greater increase in primary spherical aberra- of lens wear. These terms and RMS values and RMS values than the fellow eyes by
tion (Z 04 , mean difference: 0.089 0.139 μm, stabilised and no significant changes over 0.076 0.142 μm, 0.083 0.124 μm, and
p = 0.002), the RMS of spherical aberrations time were observed for both eyes (all 0.096 0.141 μm, respectively (all p < 0.01).
(mean difference: 0.094 0.120 μm, p > 0.05). At the one-month visit, eyes fitted Despite the statistically significant differ-
p < 0.001) and total HOAs (mean difference: with the increased compression factor also ence in unaided visual acuities between
125
eyes (increased compression factor: 0.02

[−0.08 to 0.34] logMAR, conventional com-
pression factor: −0.01 [−0.10 to 0.32]
logMAR; p = 0.04) at the one-month visit,
clinically this difference was minimal (that is,
2–3 letters difference on average). There
were no significant differences in best-
corrected visual acuities between the two
eyes (increased compression factor: −0.04
[−0.14 to 0.18] logMAR, conventional com-
pression factor: −0.02 [−0.14 to 0.08]
logMAR; p = 0.87).
Changes in other HOAs

over time
Some other HOA terms and RMS values were
also altered after orthokeratology treatment,
despite no significant differences in the changes
over time between the two eyes. Adjusting for
using paired-eye data, after one week of lens
wear, primary horizontal coma (Z 13 ) and sec-
ondary spherical aberration (Z 06 ) increased
by 0.117 0.116 μm and 0.024 0.032 μm,
respectively (both p < 0.001). No further
changes in these terms or significant
between-eye differences were observed
thereafter (all p > 0.05). At the one-month visit,
tertiary horizontal astigmatism (Z 26 ) decreased,
but the change was minimal (mean change:
−0.012 0.016 μm, p = 0.006).
Correlation analysis for primary

spherical aberration (Z 04 )
Since primary spherical aberration (Z 04 ) was
the only Zernike co-efficient (unlike
unsigned RMS metrics, individual Zernike
co-efficient provides directional information)
found to be significant in both the linear
mixed model (a significant increase during
orthokeratology treatment) and higher in
eyes fitted with the increased compression
factor, correlation analyses were performed.
Examining data from each eye separately
revealed a positive association between the
change in spherical equivalent refraction
and the change in spherical aberration (Z 04 )
(Figure 2). However, this relationship was
Figure 1. The mean change in primary spherical aberration (Z 04 ) and root-mean- only significant for the eyes fitted with the
increased compression factor (r = 0.69,
square (RMS) error for spherical aberrations and total higher-order aberrations
p < 0.001).
(HOAs) in eyes fitted with orthokeratology lenses of increased compression factor
(1.75 D, black circles) and conventional compression factor (0.75 D, white circles) dur-
ing one-month orthokeratology lens wear. Each error bar represents one standard
Discussion
error of the mean. Asterisks represent significant differences between eyes (*p < 0.05;
**p < 0.01; ***p < 0.001).
This is the first prospective study to exam-
ine the influence of compression factor on
the changes in ocular HOAs during ortho-
keratology treatment. In theory, increasing
126
higher RMS values for spherical aberrations

and total HOAs after one month of lens
wear as anticipated.
Only a few researchers have attempted to
deliberately alter the ocular HOA profile to
control myopia progression. Cheng et al.31
showed that children wearing a soft contact
lens with additional positive spherical aberra-
tion had 39 per cent slower axial eye growth
compared to children fitted with conven-
tional single-vision contact lenses over one
year. The spherical aberration incorporated
in this lens design was about 0.175 μm over
a 5-mm pupil, whereas the change in spheri-
cal aberration due to orthokeratology in the
current study (both compression factors) at
the one-month visit was approximately 1.4
times greater (about 0.250 μm for a 5-mm
pupil). In addition, the magnitude of the posi-
tive shift in spherical aberration in eyes fitted
with the increased compression factor was
40 per cent more than those fitted with the
conventional compression factor.
The relationship between HOAs and axial
Figure 2. A scatter plot showing the correlation between the change in primary
eye growth has recently been investigated
spherical aberration (Z 04 ) and the change in spherical equivalent refraction for eyes
in spectacle-wearing children.32,33 Hiraoka
fitted with orthokeratology lenses of increased compression factor (1.75 D, black cir- et al.32 indicated that higher levels of verti-
cles) and conventional compression factor (0.75 D, white circles). Regression lines
cal coma (Z 3−1 ) and spherical aberration (Z 04 )
show the co-efficients of the Pearson correlation (1.75 D: solid line; 0.75 D:
were correlated with slower axial eye
dashed line).
growth, whereas Lau et al.33 demonstrated
that oblique trefoil (Z 33 ) and spherical aber-
ration (Z 04 ) were associated with slower axial
the compression factor by 1.00 D should design was applied in the current study, the
elongation, after accounting for confounding
lead to an increase in specific HOA terms intrinsic corneal properties of an individual
variables.
and RMS metrics, particularly for spherical should apply equally between eyes (for
A number of studies have also investi-
aberration, which was confirmed in this example, corneal hysteresis).30 The results of
gated the association between the
short-term study. A contralateral self- the current study suggest that a simple alter-
changes in ocular HOAs induced by ortho-
controlled design was used to provide ation of the design of the initial lens
keratology and axial eye growth during
improved control over intrinsic ocular vari- (by increasing the compression factor by
childhood. After one year of ortho-
ables of individuals, such as corneal tangent 1.00 D) did not significantly alter the central
modulus,29 refractive error, and the HOA subjective spherical equivalent outcome keratology, Hiraoka et al.22 observed sig-
profile, to minimise possible confounders. compared to the conventional compression nificant correlations between axial eye
The concept of an increased compression factor. This implies that other factors (such growth and the changes in the RMS of
factor (1.75 D) was introduced in an attempt as corneal biomechanical properties) influ- spherical, comatic aberrations, and total
to induce additional myopic correction to ence the final refractive outcomes. HOAs, but not for primary spherical aber-
negate the under-correction typically In addition, the changes in subjective ration (Z 04 ). The authors speculated that
observed with the conventional compres- refraction mainly reflect the image quality asymmetric, rather than symmetric, changes
sion factor (0.75 D) used by most manufac- subjectively perceived by the individuals at in corneal shape might be an inhibitory factor
turers. However, no clinically significant the foveal region while the changes in HOAs to axial elongation in orthokeratology-treated
effect (< 0.25 D) of compression factor on were objectively analysed over a fixed pupil
eyes. However, their study used a simple cor-
changes in corneal apical power or subjective diameter (5 mm in this study), which could
relation analysis, which did not account for
refraction following one month of lens wear be a reason for the discrepancy between
the magnitude of change observed in sub- the effect of other known contributors to eye
was found. Lam et al.29 previously analysed
jective refraction and ocular aberrations growth, such as baseline refractive error, age,
the influence of corneal biomechanical prop-
erties on orthokeratology treatment out- between the fellow eyes. and sex. If total HOAs contribute to axial elon-
comes and found that greater corneal Regarding the HOAs, in the current study, gation in children, increasing the compression
flattening was associated with lower corneal eyes fitted with orthokeratology lenses of factor by 1.00 D, which significantly increased
hysteresis and higher tangent modulus. increased compression factor demonstrated the magnitude of total HOAs and spherical
Since a randomised, self-controlled study more primary spherical aberration (Z 04 ) and aberration with minimal influence on the
127
refractive outcome and visual acuity, may be a Co. Ltd., Japan (ZG3Z) and the Research Res- 17. Lian Y, Shen M, Huang S et al. Corneal reshaping and
wavefront aberrations during overnight ortho-
potential method to increase the myopia con- idency Scheme of the School of Optometry, keratology. Eye Contact Lens 2014; 40: 161–168.
trol efficacy of orthokeratology. PolyU provided to J.K. Lau. The rinsing solu- 18. Sun Y, Wang L, Gao J et al. Influence of overnight
tions were sponsored by Ophtecs Co., Japan orthokeratology on corneal surface shape and optical
While increased levels of HOAs after quality. J Ophthalmol 2017; 2017: 3279821.
orthokeratology treatment are associated and the artificial tears were supported by 19. Berntsen DA, Barr JT, Mitchell GL. The effect of over-
with decreased visual acuity, particularly in Precilens Ltd., France, respectively. night contact lens corneal reshaping on higher-order
aberrations and best-corrected visual acuity. Optom
low-contrast environments,34 increasing the Vis Sci 2005; 82: 490–497.
compression factor by 1.00 D resulted in REFERENCES 20. Gifford P, Li M, Lu H et al. Corneal versus ocular aber-
1. Cho P, Cheung SW. Retardation of myopia in ortho- rations after overnight orthokeratology. Optom Vis Sci
comparable unaided visual acuities between
keratology (ROMIO) study: a 2-year randomized clini- 2013; 90: 439–447.
eyes wearing orthokeratology lenses of dif- cal trial. Invest Ophthalmol Vis Sci 2012; 53: 7077–7085. 21. Joslin CE, Wu SM, McMahon TT et al. Higher-order
ferent compression factors (on average a 2. Chen C, Cheung SW, Cho P. Myopia control using toric wavefront aberrations in corneal refractive therapy.
orthokeratology (TO-SEE study). Invest Ophthalmol Vis Optom Vis Sci 2003; 80: 805–811.
difference of a few letters), in spite of the Sci 2013; 54: 6510–6517. 22. Hiraoka T, Kakita T, Okamoto F et al. Influence of ocu-
significant increase in some Zernike terms 3. Charm J, Cho P. High myopia–partial reduction ortho- lar wavefront aberrations on axial length elongation
and total HOAs. However, visual perfor- k: a 2-year randomized study. Optom Vis Sci 2013; 90: in myopic children treated with overnight ortho-
530–539. keratology. Ophthalmology 2015; 122: 93–100.
mance was not measured under low lighting 4. Wolffsohn JS, Calossi A, Cho P et al. Global trends in 23. Moreno-Barriuso E, Lloves JM, Marcos S et al. Ocular
conditions in this study. myopia management attitudes and strategies in clini- aberrations before and after myopic corneal refractive
cal practice. Cont Lens Anterior Eye 2016; 39: 106–116. surgery: LASIK-induced changes measured with laser ray
5. Cheung SW, Lam C, Cho P. Parents’ knowledge and tracing. Invest Ophthalmol Vis Sci 2001; 42: 1396–1403.
perspective of optical methods for myopia control in 24. Kang P, Gifford P, Swarbrick H. Can manipulation of
Conclusions children. Optom Vis Sci 2014; 91: 634–641. orthokeratology lens parameters modify peripheral
6. Hiraoka T, Okamoto C, Ishii Y et al. Patient satisfaction refraction? Optom Vis Sci 2013; 90: 1237–1248.
and clinical outcomes after overnight orthokeratology. 25. Chen Q, Li M, Yuan Y et al. Interaction between cor-
Increasing the orthokeratology compression Optom Vis Sci 2009; 86: 875–882. neal and internal ocular aberrations induced by
factor by 1.00 D significantly increased pri- 7. Zhong Y, Chen Z, Xue F et al. Central and peripheral orthokeratology and its influential factors. Biomed Res
corneal power change in myopic orthokeratology and Int 2017; 2017: 1–8.
mary spherical aberration and the RMS its relationship with 2-year axial length change. Invest 26. Atchison DA, Bradley A, Thibos LN et al. Useful varia-
values of spherical and total HOAs, but not Ophthalmol Vis Sci 2015; 56: 4514–4519. tions of the Badal Optometer. Optom Vis Sci 1995; 72:
8. Mountford J. Retention and regression of orthokeratology 279–284.
for other Zernike co-efficients or the RMS of
with time. Int Contact Lens Clin 1998; 25: 59–64. 27. Gatinel D, Delair E, Abi-Farah H et al. Distribution and
comatic aberrations. Given the mounting 9. Johnson KL, Carney LG, Mountford JA et al. Visual per- enantiomorphism of higher-order ocular optical aber-
evidence of the potential influence of HOAs, formance after overnight orthokeratology. Cont Lens rations. J Fr Ophtalmol 2005; 28: 1041–1050.
Anterior Eye 2007; 30: 29–36. 28. Porter J, Guirao A, Cox IG et al. Monochromatic aber-
particularly spherical aberration, upon axial 10. Sorbara L, Fonn D, Simpson T et al. Reduction of myo- rations of the human eye in a large population. J Opt
eye growth in childhood, the authors hypo- pia from corneal refractive therapy. Optom Vis Sci Soc Am A 2001; 18: 1793–1803.
thesise that increasing the compression fac- 2005; 82: 512–518. 29. Lam AK, Leung SY, Hon Y et al. Influence of short-term
11. Nichols JJ, Marsich MM, Nguyen M et al. Overnight orthokeratology to corneal tangent modulus: a ran-
tor of a specific lens design may improve orthokeratology. Optom Vis Sci 2000; 77: 252–259. domized study. Curr Eye Res 2018; 43: 474–481.
the myopia control efficacy of ortho- 12. Rah MJ, Jackson JM, Jones LA et al. Overnight ortho- 30. Vincent SJ, Collins MJ, Read SA et al. Interocular sym-
keratology: preliminary results of the lenses and over- metry in myopic anisometropia. Optom Vis Sci 2011;
keratology without compromising visual
night Orthokeratology (LOOK) study. Optom Vis Sci 88: 1454–1462.
performance. However, further longitudinal 2002; 79: 598–605. 31. Cheng X, Xu J, Chehab K et al. Soft contact lenses with
investigations are required before firm con- 13. Tahhan N, Du Toit R, Papas E et al. Comparison of positive spherical aberration for myopia control.
reverse-geometry lens designs for overnight ortho- Optom Vis Sci 2016; 93: 353–366.
clusions can be drawn. keratology. Optom Vis Sci 2003; 80: 796–804. 32. Hiraoka T, Kotsuka J, Kakita T et al. Relationship
14. Chan B, Cho P, Mountford J. The validity of the between higher-order wavefront aberrations and nat-
Jessen formula in overnight orthokeratology: a ret- ural progression of myopia in schoolchildren. Sci Rep
ACKNOWLEDGEMENTS rospective study. Ophthalmic Physiol Opt 2008; 28: 2017; 7: 7876.
265–268. 33. Lau JK, Vincent SJ, Collins MJ et al. Ocular higher-order
The authors thank Mr Kin Wan for his kind 15. Hiraoka T, Matsumoto Y, Okamoto F et al. Corneal aberrations and axial eye growth in young Hong Kong
assistance in data collection. This study was higher-order aberrations induced by overnight ortho- children. Sci Rep 2018; 8: 6726.
supported by a collaborative research keratology. Am J Ophthalmol 2005; 139: 429–436. 34. Hiraoka T, Okamoto C, Ishii Y et al. Contrast sensitivity
16. Stillitano IG, Chalita MR, Schor P et al. Corneal function and ocular higher-order aberrations follow-
agreement between The Hong Kong Poly- changes and wavefront analysis after orthokeratology ing overnight orthokeratology. Invest Ophthalmol Vis
technic University (PolyU) and Menicon fitting test. Am J Ophthalmol 2007; 144: 378–386. Sci 2007; 48: 550–556.
128
PUBLISHER
Clinical and Experimental Optometry is published by John Wiley & Sons Australia, Ltd
42 McDougall Street, Milton, QLD 4064, Australia
Tel: +61 (0)7 3859 9755
Production Editor
Simon Tan (email: [email protected])
Copyright and Copying (in any format)

Copyright © 2020 Optometry Australia. All rights reserved. No part of this publication may be reproduced, stored or transmitted in any form or by
any means without the prior permission in writing from the copyright holder. Authorization to copy items for internal and personal use is granted
by the copyright holder for libraries and other users registered with their local Reproduction Rights Organisation (RRO), e.g. Copyright Clearance
Center (CCC), 222 Rosewood Drive, Danvers, MA 01923, USA (www.copyright.com), provided the appropriate fee is paid directly to the RRO. This
consent does not extend to other kinds of copying such as copying for general distribution, for advertising or promotional purposes, for
republication, for creating new collective works or for resale. Permissions for such reuse can be obtained using the RightsLink "Request
Permissions" link on Wiley Online Library. Special requests should be addressed to: [email protected].
VISUAL OPTICS: LOOKING BEYOND 2020
SPECIAL ISSUE
Vol. 100
Vol. 103 No. 1

JANUARY 2020
103 No.
No. X1 MONTH
JANUARY
2017
2020
Vol. 103 No.1 JANUARY 2020
Visual Optics: Looking Beyond 2020 issue co-ordinated by

Associate Professors Stephen Vincent and Scott Read
CLINICAL AND EXPERIMENTAL OPTOMETRY

GUEST EDITORIAL
1 Looking and seeing beyond 2020
INVITED REVIEWS
3 Modern spectacle lens design
11 Stand magnifiers for low vision: description, prescription, assessment
21 Correction of presbyopia: old problems with old (and new) solutions
31 Customised aberration-controlling corrections for keratoconic patients using contact lenses
44 Optical changes and visual performance with orthokeratology
55 Optical mechanisms regulating emmetropisation and refractive errors: evidence from animal
models
68 Higher order aberrations, refractive error development and myopia control: a review
86 Peripheral refraction and higher order aberrations
95 Aberrations and accommodation
104 Blur adaptation: clinical and refractive considerations
112 Adaptive optics imaging of the retinal microvasculature
INVITED RESEARCH Modern spectacle lens design
PAGES 1–128
123 The influence of orthokeratology compression factor on ocular higher-order aberrations Stand magnifiers for low vision
Multifocal optics
Aberration-controlling lenses for keratoconus
Optical changes with orthokeratology
Optical regulation of eye growth
Aberrations and refractive error development
Peripheral refraction and aberrations
Blur adaptation
Adaptive optics vascular imaging
Orthokeratology compression factor
iOS App Current subscribers

Download the app from iTunes or Google Play store and
Android App log in free through your association or institution
WILEY
New subscribers
Subscribe through the app
in partnership with
Website wileyonlinelibrary.com/journal/cxo The Hong Kong Society of

Cxo v103 I1 Jan

Uploaded by

Copyright:

Available Formats

Cxo v103 I1 Jan

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cxo v103 I1 Jan

Uploaded by

Copyright:

Available Formats

Vol. 103 No.

SPECIAL ISSUE VISUAL OPTICS: LOOKING BEYOND 2020

Modern spectacle lens design

EDITORIAL BOARD INTERNATIONAL ADVISORY BOARD

Looking and seeing beyond 2020

Clin Exp Optom 2020; 103: 1–2 DOI:10.1111/cxo.12993

Modern spectacle lens design

Clin Exp Optom 2020; 103: 3–10 DOI:10.1111/cxo.12930

30˚ 30˚ 30˚

10˚ 10˚ 10˚ The term ‘aspherical surface’ usually refers

from the rotation of a conic section about

The ﬁeld diagrams illustrated in Figure 4

A y B both the tangential and the sagittal radii of

This expression enables the correct

Stand magniﬁers for low vision: description, prescription,

Clin Exp Optom 2020; 103: 11–20 DOI:10.1111/cxo.12948

metrics that have been used to do this.

EVP can also be considered and calculated

EVP = F SM + AddMax −GF SM AddMax [8]

For the COIL #4210 used at z = 10 cm,

EVP = m × Add Max [9]

Manufacturer Description of manufacturer Measured Predict performance (z = eye-lens distance)

A spectacle addition at different working dis-

Measuring EVD using

Using information about stand magniﬁer

EVD = Pixels/mm in reference image x dO

2.0 required for viewing clearly through the

1.0 cle addition has an effect on reading perfor-

mance with stand magniﬁers and this may

magniﬁer at the appropriate distance, with a

Correction of presbyopia: old problems with old

Clin Exp Optom 2020; 103: 21–30 DOI:10.1111/cxo.12987

Clinical strategies that employ pinhole

whether to employ a ﬁxed pupil or one that

1 and Drug Administration approved

Crystalens (Bausch and Lomb, Rochester,

Multifocality can also be achieved by non-

used to contribute bifocal add power can

added lens is designed to achieve a net

image quality (Figure 4, see inserts) and

Although zonal, diffractive bifocal, and 0

entering the pupil from the peripheral ﬁeld

Lessons from evolution

Customised aberration-controlling corrections for keratoconic

Clin Exp Optom 2020; 103: 31–43 DOI:10.1111/cxo.12937

Mild Moderate Severe

Mild Moderate Severe

Mild Moderate Severe

aberrations were compensated for by the Inward pathway

Only the degree of their ‘displacement’ is

Charge-coupled device (CCD) sensor simply by using a micro-array of lenslets with

Face-on view Side-on view analysis, primarily due to their orthogonality

Optical changes and visual performance with orthokeratology

Clin Exp Optom 2020; 103: 44–54 DOI:10.1111/cxo.12947

Key words: aberrations, myopia, myopia control, orthokeratology, visual acuity

error, leading to the term ‘accelerated

Four developments signiﬁcantly inﬂuenced

ﬂattening of the posterior cornea,15,16 it is Orthokeratology generally does not induce

modality similar to their original assignment