Psychology

Case Report

The effect of knee resizing illusions on pain and swelling

in symptomatic knee osteoarthritis: a case report
Erin MacIntyrea, Maja Sigersethb,c, Brian W. Pullingc, Roger Newportd, Tasha R. Stantonc,e,*

Abstract
Introduction: Resizing illusions that manipulate perceived body size are analgesic in some chronic pain conditions. Little is known
whether such illusions may also alter other physiological features, such as swelling.
Objectives: To determine the effects of a knee resizing illusion on knee pain and swelling in symptomatic osteoarthritis.
Methods: This case study was extracted from a larger study evaluating the analgesic effects of resizing illusions in people with knee
osteoarthritis. A mediated reality system (alters real-time video) was used to provide resizing “stretch” and “shrink” illusions of the
knee. Knee pain intensity (0–100 numerical rating scale) was measured before and after illusion and after sustained (3 minutes) and
repeated (n 5 10) illusions. In this case study, knee swelling (leg circumference below, at, and above the knee) was also measured.
Results: The 55-year-old male participant reported a long history of episodic knee pain and swelling that was subsequently
diagnosed as severe osteoarthritis in 2013. In the first testing session, the participant experienced an increase in pain with the shrink
illusion and a decrease in pain with stretch illusion. A noticeable increase in knee swelling was also observed. Thus, in sessions 2/3,
swelling was also assessed. The stretch illusion decreased pain to the largest extent, but resulted in increased knee swelling.
Repeated and sustained stretch illusions had cumulative analgesic effects but resulted in cumulative increases in swelling. While the
shrink illusion increased pain, sustained (;10 minutes) visual minification of the entire knee and leg reduced both pain and swelling.
Conclusion: Our case report suggests that both pain and swelling may be modifiable by altering body-relevant sensory input in
symptomatic knee osteoarthritis.
Keywords: Osteoarthritis, Body illusion, Mediated reality, Mental representation, Pain, Swelling

1. Introduction matrix dynamically adapts to changes in (perceived) body struc-

ture.12 Indeed, visual illusions that distort the size of the affected limb
There is growing evidence that body resizing illusions are analgesic in
in people with complex regional pain syndrome (CRPS) modulate
some pain conditions,4 including knee osteoarthritis.21 Recent
both movement-evoked pain and swelling.14
theoretical models, namely the cortical body matrix model,12
The effect of body illusions on physiological regulation is
postulate that both pain and physiological regulation (eg, swelling)
particularly relevant in knee osteoarthritis, where fluctuating or
of the body are intertwined with body perception. That is,
persistent swelling is common.3 However, CRPS is an atypical pain
a multisensory representation of the body is involved in maintaining
condition (eg, swelling of the affected limb can be induced by merely
integrity and homeostatic regulation of the body, and this body
imagining movement11), making it uncertain whether illusion-
induced swelling changes would also occur in osteoarthritis.
Sponsorships or competing interests that may be relevant to content are disclosed Therefore, in this case report, we examine the effect of illusory
at the end of this article.
a
resizing of the painful knee (using mediated reality) on pain intensity
Sports and Spinal Physiotherapy, Sunshine Coast, Queensland, Australia,
b and swelling in a patient with symptomatic knee osteoarthritis.
Department of Global Public Health and Primary Care, Member of Physiotherapy
Research Group, University of Bergen, Bergen, Norway, c IIMPACT in Health, The
University of South Australia, Adelaide, South Australia, Australia, d School of Sport,
Exercise and Health Sciences, Loughborough University, Loughborough, United
2. Methods
Kingdom, e Neuroscience Research Australia, Sydney, New South Wales, Australia This case was taken from our previous knee osteoarthritis illusion
*Corresponding author. Address: The University of South Australia, IIMPACT in study.21 The initial protocol and case study were approved by the
Health, GPO Box 2471, Adelaide, SA 5001, Australia. Tel.: 1618 8302 2090. E-mail
UniSA Human Research Ethics Board (ID:28496); the participant
address: [email protected] (T.R. Stanton).
provided written informed consent for both.
Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf
of The International Association for the Study of Pain. This is an open access article
The participant attended 3 sessions, with resizing illusions
distributed under the Creative Commons Attribution License 4.0 (CCBY), which performed on his symptomatic knee using the MIRAGE illusion
permits unrestricted use, distribution, and reproduction in any medium, provided the system (tested in standing; Fig. 1). The participant viewed video
original work is properly cited. of his own knee from the front (as if looking in a mirror) through
PR9 4 (2019) e795 a head mounted display. To create the resizing illusions, video
http://dx.doi.org/10.1097/PR9.0000000000000795 images of the knee were altered using customised LabVIEW

4 (2019) e795 www.painreportsonline.com 1

2
·
E. MacIntyre et al. 4 (2019) e795 PAIN Reports®

software. The knee was visually stretched or shrunk (;50% of the entire limb (50% visual minification) was also tested (last in
normal size), accompanied by the experimenter applying each session) and was sustained for 10 minutes (pain/swelling
congruent tactile input on the calf (gentle longitudinal traction or measured before/after illusion).
compression, respectively). One-sample t tests evaluated if baseline pain scores differed
Session 1 evaluated whether body illusions produced more significantly from the postillusion pain scores (for all repeated/
analgesia than control conditions. The participant underwent 2 sustained illusions).
illusion conditions (stretch/shrink: congruent vision-touch) and
6 stretch/shrink control conditions (touch-only, vision-only,
3. Results
incongruent vision-touch), in a randomised order with pain
intensity (0–100 numerical rating scale9) assessed before and A 55-year-old Caucasian man (P.C.) reported a 30-year history of
after condition. Each illusion/condition lasted ;4 to 5 seconds, episodic knee pain after a horse-riding accident, where the horse
with the knee returned to normal size for the 2-minute break fell and rolled over his left leg. He reported having a sore knee and
between each condition. back, but that the pain subsided over the following weeks. No
The illusion that produced the greatest pain reduction was medical attention was sought. He reported that his knee was
used in sessions 2 and 3. To determine cumulative analgesic without incident until his forties when he began experiencing
benefits, this illusion was sustained for 3 minutes (participant periods of knee pain and swelling, which was diagnosed as a torn
viewed his visually altered knee, rating pain intensity every 30 meniscus. He underwent 3 knee arthroscopies for the meniscal
seconds) in session 2. The illusion was also repeated 10 times injury and reported good outcome (periods without pain/swelling
(pain intensity rated before/after illusion; 30 seconds break after each procedure). P.C. reported that his knee was
between illusions) in both sessions. asymptomatic until 2013 (;53 years old), at which time his knee
During sessions 2/3, additional procedures were undertaken in pain and swelling returned after an increase in activity level. His
this participant. Knee swelling was assessed (before/after symptoms were managed pharmaceutically with prednisone and
illusion), through tape measure, at 3 locations (skin marked for methotrexate to good effect, although medication side effects
consistency): 1 cm below patella inferior pole; at patella midpoint; prevented long-term use. He also reported undertaking physio-
and 1 cm above patella superior border. A unisensory illusion of therapy (exercise/education), acupuncture, and osteopathy. In

Figure 1. The MIRAGE mediated reality set-up and and experimental conditions.
4 (2019) e795 www.painreportsonline.com 3

2015, he underwent repeat knee radiographs which showed Upon follow-up (September 2018, ;2 years after study), P.C.
severe osteoarthritis. See Table 1 for participant demographics reported that he underwent a total knee replacement (October
and session-relevant details. 2015), followed by extensive physiotherapy which improved, but
In session 1, P.C. had an increase in knee pain (110) with the did not entirely relieve, his knee pain and swelling (Table 1).
shrink illusion and a pain decrease (25) with the stretch illusion
(no change in pain with control conditions). P.C. had a visually
4. Discussion
noticeable increase in knee swelling post-stretch illusion.
In sessions 2/3, repeated and sustained application of the This case study presents the first evidence that body resizing
stretch illusion both had cumulative effects resulting in increased illusions may modulate both pain and swelling in knee osteoar-
analgesia (Figs. 2A/B and 3A), but also increased swelling, thritis. That the effects on pain and swelling were conflicting
primarily superior to the patella (Figs. 2C and 3C). P.C. reported during the knee stretch illusion (congruent vision-touch; visually
transient, but consistent, feelings of nausea during the stretch altering the knee), but consistent during a unisensory minimisa-
illusion. In contrast, when P.C. underwent the sustained tion illusion (vision-only; visually altering the entire limb), suggests
unisensory minification illusion (reduced overall size of the knee/ differential pathways of effect based on the type of sensory
leg), knee swelling reduced (Figs. 2C and 3C), and pain manipulation.
decreased (measured only in session 3; Fig. 3B). Statistically Past work in CRPS has shown that the perceived anthropometric
significant pain reductions were present for all repeated/ characteristics of the body (size of the painful limb) are linked to
sustained illusions (Table 1). pain10,14 and to its physiological regulation, namely swelling.14 Such

Table 1
Demographics and pain measures for the case study participant (P.C.) and the original cohort.
Demographics/outcomes P.C. Original cohort
Age (y) 55 67.3 (9.9)
Height (cm) 189 167.2 (11.2)
Weight (kg) 99 82.7 (16.3)
Sex Male 3 male, 9 female
Fremantle Knee Awareness Questionnaire 22 14.0 (8.4)
Oxford Knee Score 12 24.1 (8.1)
Perceived knee size (% of true size) 102 104.1 (0.05)
Session 1
Average baseline knee pain (past 48 h) 75 48.0 (24.3)
Maximum knee pain (past 48 h) 90 66.3 (28.6)
Minimum knee pain (past 48 h) 0 6.3 (10.9)
Current knee pain (beginning of session) 25 43.3 (22.2)
Session 2
Current knee pain (beginning of session) 25 26.7 (28.6)
Sustained stretch illusion
Baseline pain rating 25 26.3 (13.1)
Pain reduction 219* (276%) 22.7 (3.2) (210%)
One-sample t test results t1,5 5 4.52, P 5 0.006 N/A
Repeated stretch illusion
Baseline pain rating 25 31.7 (12.9)
Pain reduction 25 (220%) 26 (7.7) (219%)
One-sample t test results t1,18 5 3.95, P 5 0.0009 N/A
Session 3
Current knee pain (beginning of session) 75 45.6 (36.6)
Repeated stretch illusion
Baseline pain rating 60 50.4 (24.6)
Pain reduction 220* (233%) 220 (13.1) (240%)
One-sample t test results t1,18 5 3.72, P 5 0.002 N/A
Sustained minified leg illusion
Baseline pain rating 40 N/A
Pain reduction 210 (225%) N/A
One-sample t test results t1,4 5 3.26, P 5 0.03 N/A
Follow-up (September 2018)
Current knee pain 20 N/A
Average knee pain (previous 24 h) 30 N/A
Current subjective rating of knee swelling 0 N/A
Average subjective rating of knee swelling 0 N/A
(previous 24 h)
All data are mean (SD) unless otherwise indicated. Pain and subjective rating of swelling measured using a 101-point NRS, where 0 5 no pain/swelling and 100 5 most pain/swelling imaginable. The Freemantle Knee
Awareness Questionnaire ranges from 0 to 36, where higher scores indicate less knee awareness,16 and the Oxford knee Score ranges from 0 to 48 where higher values indicate less disability.6 Perceived knee size evaluated
using established methodology.8 Pain reduction scores are calculated from taking the final pain rating (after illusion) minus the baseline rating. For sustained illusions, the baseline pain rating referred to either the first rating
taken directly after the illusion was applied or the rating taken before the illusion being applied (whichever was most conservative). The percentage of pain reduction is expressed as the change in pain induced by the illusion
divided by the baseline pain rating. One-sample t test results compared the baseline pain ratings to the postillusion pain ratings; bolded p-values represent statistically significant results. N/A 5 not applicable because this test
was not performed in the full study population or the study population was not followed up.
* Denotes a pain decrease greater than minimum clinically important difference for pain.7,18
4
·
E. MacIntyre et al. 4 (2019) e795 PAIN Reports®

Figure 2. Session 2 results. (A) Pain intensity ratings during repeated stretch illusions; (B) pain intensity ratings during sustained stretch illusion; (C) knee swelling
measurements (leg circumference in centimeters) at baseline, after repeated stretch illusions, after sustained stretch illusion, and after a sustained minimised knee/
leg illusion. NRS, Numerical Rating Scale.

findings, combined with evidence of bidirectional links between body a nonaffine transformation (site-specific manipulation—the knee
ownership and physiological regulation,2,13 have led to the proposal itself “stretches”), while a minimised limb illusion is an affine
of the cortical body matrix theory.12 We extend this theory by transformation (rigid body alteration of the entire limb). Past work
showing that altered sensory input can have differential influences on shows alterations in physiological regulation as a function of body
pain and swelling, potentially suggesting unique drivers of each. ownership13 and that loss of body ownership can be analge-
Why might the stretch illusion reduce pain but increase sic,17,19,20 raising the possibility that either illusion may have
swelling (with effects heightened for repeated application) induced a loss of knee ownership. Nausea was reported only
whereas a minimised view of the entire limb reduces both pain during the stretch illusion suggesting that the knee was not
and swelling? First, that a sustained stretch illusion (visual change disowned for that condition. Rather, nausea typically occurs
not repeated) also reduced pain and increased swelling (Fig. 2B/ when what you see does not match what you feel (sensory-
2C) suggests against such effects being solely driven by neural mismatch), with cybersickness heightened as a function of
processing initiated with viewing the real-time change in body size immersion.23 The stretch illusion alters the knee itself, inducing an
(eg, via increased visual attention1). Second, differences in the incongruence between the seen and the felt knee which may be
type of illusion may be relevant. Illusions differed on the presence/ sufficient to elicit protective responses such as nausea and
absence of a tactile component, but tactile-only control increased swelling or, in past work, feelings of disgust (which
conditions did not influence pain/swelling suggesting against have not been reported for full limb minimisation).15 However,
a simple effect of touch. In addition, the stretch illusion is such incongruence is typically algesic5 or has no effect on pain,22

Figure 3. Session 3 results. (A) Pain intensity ratings during repeated stretch illusions; (B) pain intensity ratings during sustained minimised knee/leg illusion; (C)
knee swelling measurements (leg circumference in centimeters) at baseline, after repeated stretch illusions, and after a sustained minimised knee/leg illusion. NRS,
Numerical Rating Scale.
4 (2019) e795 www.painreportsonline.com 5

neither of which was seen here. Instead, perceived body size may [2] Barnsley N, McAuley JH, Mohan R, Dey A, Thomas P, Moseley GL. The
also play a role in analgesia. People with knee osteoarthritis have rubber hand illusion increases histamine reactivity in the real arm. Curr
Biol 2011;21:R945–6.
an altered perception of their painful knee16—P.C. perceived his [3] Bijlsma JW, Berenbaum F, Lafeber FP. Osteoarthritis: an update with
knee to be larger than it actually was (2% larger). Research in relevance for clinical practice. Lancet 2011;377:2115–26.
hand osteoarthritis has shown that regardless of the illusion [4] Boesch E, Bellan V, Moseley GL, Stanton TR. The effect of bodily illusions
(stretch or shrink), illusory resizing normalises mental hand on clinical pain: a systematic review and meta-analysis. PAIN 2016;157:
516–29.
representation.8 Such findings may support a role of mental
[5] Daenen L, Nijs J, Roussel N, Wouters K, Van Loo M, Cras P.
representation in the analgesic effects during both enlarging and Sensorimotor incongruence exacerbates symptoms in patients with
reducing illusions seen here. chronic whiplash associated disorders: an experimental study.
Rheumatology (Oxford) 2012;51:1492–9.
[6] Dawson J, Fitzpatrick R, Murray D, Carr A. Questionnaire on the
4.1. Study limitations perceptions of patients about total knee replacement. J Bone Joint Surg
Br 1998;80:63–9.
Limb ownership was not evaluated during illusions (purposefully [7] Farrar JT, Young JP, LaMoreaux L, Werth JL, Poole RM. Clinical
to ensure participant blinding). Future work is warranted to importance of changes in chronic pain intensity when measured on an
explore the effects of these illusions on ownership and mental 11-point numerical pain rating scale. PAIN 2001;94:149–58.
[8] Gilpin HR, Moseley GL, Stanton TR, Newport R. Evidence for distorted
representation of the body and to explore phenomenological
mental representation of the hand in osteoarthritis. Rheumatology
accounts (does a stretched knee appear swollen? ie, perceived (Oxford) 2015;54:678–82.
swelling influences actual swelling). In addition, knee swelling [9] Jensen MP, Karoly P, Braver S. The measurement of clinical pain
measurement over time (without any illusion) was not evaluated, intensity: a comparison of six methods. PAIN 1986;27:117–26.
making it possible that swelling would have reduced without [10] Matamala-Gomez M, Diaz Gonzalez AM, Slater M, Sanchez-Vives MV.
Decreasing pain ratings in chronic arm pain through changing a virtual
a minimised knee illusion. However, that there was no visual body: different strategies for different pain types. J Pain 2019;20:
change in knee swelling for ;15 minutes after illusion in session 1 685–97.
supports the present findings. [11] Moseley GL. Imagined movements cause pain and swelling in a patient
with complex regional pain syndrome. Neurology 2004;62:1644.
[12] Moseley GL, Gallace A, Spence C. Bodily illusions in health and disease:
5. Conclusions physiological and clinical perspectives and the concept of a cortical “body
matrix”. Neurosci Biobehav Rev 2012;36:34–46.
We present the first detailed case study in which body resizing [13] Moseley GL, Olthof N, Venema A, Don S, Wijers M, Gallace A, Spence C.
illusions modulate osteoarthritic pain and swelling. Our findings add Psychologically induced cooling of a specific body part caused by the
to the growing evidence that body-relevant sensory input can have illusory ownership of an artificial counterpart. Proc Natl Acad Sci USA
2008;105:13169–73.
profound influences on pain and physiological regulation of the body. [14] Moseley GL, Parsons TJ, Spence C. Visual distortion of a limb
modulates the pain and swelling evoked by movement. Curr Biol
2008;18:R1047–8.
Disclosures [15] Newport R, Auty K, Carey M, Greenfield K, Howard EM, Ratcliffe N, Thair
H, Themelis K. Give it a tug and feel it grow: extending body perception
T.R. Stanton received travel and accommodation support from Eli Lilly
through the univeral nature of illusory finger stretching. Iperception 2015;
Ltd for speaking engagements (2014; unrelated to the present topic). 6:1–4.
R. Newport is the creator the MIRAGE mediated-reality systems. The [16] Nishigami T, Mibu A, Tanaka K, Yamashita Y, Yamada E, Wand BM,
University of Nottingham (United Kingdom) received equipment fees Catley MJ, Stanton TR, Moseley GL. Development and psychometric
for creation of mediated-reality systems for external laboratories. The properties of knee-specific body-perception questionnaire in people with
knee osteoarthritis: the Fremantle Knee Awareness Questionnaire. PLoS
remaining authors have no conflicts of interest to declare. One 2017;12:e0179225.
[17] Pamment J, Aspell JE. Putting pain out of mind with an “out of body”
illusion. Eur J Pain 2017;21:334–42.
Acknowledgements [18] Salaffi F, Stancati A, Silvestri CA, Ciapetti A, Grassi W. Minimal clinically
important changes in chronic musculoskeletal pain intensity measured on
T.R. Stanton supported by a National Health and Medical a numerical rating scale. Eur J Pain 2004;8:283–91.
Research Council Career Development Fellowship (ID1141735). [19] Siedlecka M, Klimza A, Łukowska M, Wierzchoń M. Rubber hand illusion
This work was supported by a University of South Australia internal reduces discomfort caused by cold stimulus. PLoS One 2014;9:
research grant. The funders had no role in study design, data e109909.
[20] Solc à M, Ronchi R, Bello-Ruiz J, Schmidlin T, Herbelin B, Luthi F,
collection and analyses, decision to publish, or preparation of the
Konzelmann M, Beaulieu JY, Delaquaize F, Schnider A, Guggisberg
manuscript. AG, Serino A, Blanke O. Heartbeat-enhanced immersive virtual
reality to treat complex regional pain syndrome. Neurology 2018;91:
Article history: e479–89.
Received 4 June 2019 [21] Stanton TR, Gilpin HR, Edwards L, Moseley GL, Newport R. Illusory
resizing of the painful knee is analgesic in symptomatic knee
Received in revised form 9 October 2019 osteoarthritis. PeerJ 2018;6:e5206.
Accepted 13 October 2019 [22] Wand BM, Szpak L, George PJ, Bulsara MK, O’Connell NE, Moseley GL.
Moving in an environment of induced sensorimotor incongruence does
not influence pain sensitivity in healthy volunteers: a randomised within-
References subject experiment. PLoS One 2014;9:e93701.
[1] Abrams RA, Christ SE. Motion onset captures attention. Psychol Sci [23] Weech S, Kenny S, Barnett-Cowan M. Presence and cybersickness in
2003;14:427–32. virtual reality are negatively related: a review. Front Psychol 2019;10:158.