THE

EFFECT

OF

POSTURE
M. A. ADAMS, W.

ON
C.

THE
HUTTON

LUMBAR

SPINE

From

the

PoIttec/inic

of Central

London

A series

of experiments

showing

how posture

affects

the lumbar

spine

is reviewed.

Postures

which

flatten

(that is, flex) the lumbar spine are compared with those that preserve the lumbar lordosis. Our review shows that flexed postures have several advantages: flexion improves the transport of metabolites in the intervertebral discs, reduces the stresses on the apophyseal joints and on the posterior half of the annulus fibrosus, and gives the spine a high compressive strength. Flexion also has disadvantages: stress on the anterior annulus and increases the hydrostatic pressure in the nucleus pulposus The disadvantages are not of much significance and we conclude that it is mechanically advantageous to flatten the lumbar spine when sitting and when lifting heavy weights. it increases the at low load levels. and nutritionally

Current rather aesthetic dont

ideas vague. and slouch.

on

what

constitutes

good

posture

are on and

The military

usual advice, possibly based traditions, is sit up straight sitting up straight not

Paradoxically,

is taken allowing Erect there is no fact, have posture

to mean sitting with a lumbar lordosis and the lumbar spine to flex and flatten its curve. As far as the reliable beneficial. evidence On lumbar spine is concerned, up straight population that sitting the contrary,

posture

is, in studies

shown that lumbar people who habitually ten the lumbar spine postures are many adults. couraged and

disc degeneration is sit or squat in postures (Fahrni and Trueman

rare among which flat1965). Such

7)
060 847
Fig. The lumbar faces ofthe for typical I

A
0:22

instinctively assumed by If these natural preferences advice given on posture,

children and by are to be disthen To such advice this end we on cadaveric affect both the the nutritional

0<20#{176}

should be founded on scientific evidence. have carried out a series of experiments lumbar spines to show how posture can mechanics of the lumbar status of the intervertebral spine discs. METHOD The normal curves of the spine help and also

curvature (0) is defined as the angle sacrum and the LI vertebral body. erect and flexed postures ( calculated 1979).

between the upper surValues ofO are shown from Andersson et al.

unstressed

spine

it averages

about

40

as measured

from and with when

the dissected spines ofcadavers Dubow 1972). In a living adult, posture, from 80 when leaning

(Farfan, Huberdeau the curvature varies backwards to 0

it to act

as a shock

absorber by reducing its longitudinal lar mechanism a metal rod can curved absent human natural. The
M. A. Adams. W. C. hutton,
and Science Polytechnic WIM SJS. Requests
(

stiffness (by a simiabsorb shock if it is

bending fully forwards (calculated and Shepherd 1984). Typical values standing and sitting postures are Three of these values are calculated
al. ( I 979).

from Pearcy, Portek within this range for shown in Figure 1. from Andersson et measurements. to their effect are those that

to form a helical spring). These spinal curves are in the child but pronounced in the adult in all populations: they therefore must be considered lumbar
PhD. MSc.

We

and have

the fourth is from our own defined postures according curvature: flexed postures

on the lumbar

curvature
Research

is defined
Fellow

in Figure

1 In the
. -

ProfessorofBiomechanics,
London. should hesent I 15 New

School Cavcndish
Hutton.

of Engineering Street. London

of Central England. for reprints

flatten the lumbar spine and produce low values of lumbar curvature, whereas erect postures are those that maintain or increase the lumbar curvature. Posture and the loading of the apophyseal joints. These joints stabilise the spine and protect the discs from both excessive flexion (Adams, Hutton and Stott 1980) and axial rotation a major role although this (Adams and Hutton 1981). They in resisting shear and compressive varies considerably with posture.
625

to Professor

also

1985 British 620X

Editorial $2.00

Society

of Bone

and

Joint

Surgery

play forces,

0301

85 4130

Vol..

67-B.

No.

4. AUGUST

1985

626

M.

A.

ADAMS,

W.

C.

HUTTON

erect posture. mitted through through 1981). sive and

the

highest compressive the posterior annulus

stresses and

are transthe lowest

the anterior annulus (Horst and Bninckmann Similarly. in flexed postures the highest compresthe anterior annulus annulus. discs, the nucleus pulthe magnitude of (Nachemson 1963). sitting pulposus
50%

/
Erect posture Flexed posture

stresses are transmitted through the lowest through the posterior In all but the most degenerate exhibits hydrostatic is dependent on disc pressure, wedging

posus which

At low load levels (appropriate in relaxed standing postures), pressure in the nucleus the same applied compressive load is about in the flexed position than in the erect

and for higher (cal1979). of the 3

position

Fig.

The effect of posture on the loading of the apophyseal joints. In the erect posture the articular surfaces resist the intervertebral shear force (S) and a proportion ofthe intervertebral compressive force (C). Extraarticular impingement can also occur (as shown by the star). In the flexed posture the articular surfaces resist the shear force but play no part in resisting the compressive force.

culated from Nachemson, Schultz and Berkson An explanation of this is given in the appendix. The effects of posture on the mechanics mature and 4. Posture intervertebral and the nutrient are the reach disc are illustrated in

Figures discs.

supply

to the intervertebral structures routes: from

The lumbar discs the body. Nutrients In the erect posture, the apophyseal joints resist the blood vessels most of the shear force acting on the spine (Hutton, Stott and Cyron 1977) as well as about 16% of the compressive force (Adams and Hutton 1980). The resulting stress between the articular surfaces is concentrated in the lower margins ofthejoint (Dunlop, Adams and Hutton 1984) row as shown and degenerate, in Figure then 2. If the the facets disc may is unusually come into narclose fluid surrounding transport processes sion-and both

largest avascular their cells by two bodies and

in from tissue

in the

vertebral

the annulus fibrosus. for each route-fluid of these are affected

There are two flow and diffu-

by posture.

apposition pressive

and may then resist up force on the spine (Adams

to 70% of the comand Hutton 1980). is due on the to extraadjacent In

Much of this abnormally articular impingement lamina or pedicle (Dunlop, such cases, the apophyseal

high resistance of the facet tips

Adams and Hutton 1984). joints show gross osteoarthni-

tic changes (Adams and Hutton 1980). In the flexed posture, the apophysealjoints shear force but now play no part in resisting tebral stress erect upper 1984) compressive between the posture parts as shown of and the force (Adams and Hutton anticular surfaces is lower is concentrated joint (Dunlop, in the Adams

resist the the interver1980). The than in the middie and and Hutton there disc. is The

Erect

posture
Fig.3

Flexed
Fig.4

posture

The effect ofposture on the mature intervertebral disc. In the erect postures (Fig. 3) the compressive force (C) produces the highest stress on the posterior annulus, while in the flexed posture (Fig. 4) the stress is highest on the anterior annulus and the hydrostatic pressure (p) in the nucleus is increased.

in Figure

2. In the flexed

posture

Fluid High while

flow

is caused

by pressure

changes

on the disc.

no extra-articular Posture and the

impingement. loading of the

intervertebral

pressure causes fluid to be expelled from the disc, low pressure (lying down, for example) allows the in the disc to suck in fluid from surroundexchange from the increase this fluid fluid to be expelled

intervertebral discs main weight-bearing ture affects the way

and vertebral bodies comprise the column of the lumbar spine. Posthis column resists the loads applied

proteoglycans

ing tissue. Flexed postures because they cause more

to it, but has little effect on the magnitude of these loads (see below: posture and muscle activity). Under load, an unwedged disc tends to behave as a hydrostatic body exerting a uniform compressive stress on the vertebral end-plates (Horst and Brinckmann 198 1 ). By wedging a disc we complicate this slightly: young mature This non-degenerate and degenerate means that when discs discs remain sustain disc hydrostatic, but pressure gradients. is wedged in the

disc than do erect postures (Adams and Hutton 1983). The effect is particularly marked in the nucleus pulposus (Fig. 5). Diffusion occurs as a result of a chemical concentration gradient: nutrients diffuse into the disc, and waste products diffuse out of it. In erect postures diffusion occurs more readily into (Fig. front
JOURNAL

the

anterior

annulus

than

into

a mature

the posterior annulus height decreases from

5), probably because to back (Adams and

the disc Hutton

THE

EFFECT

OF

POSTURE

ON

THE

LUMBAR

SPINE

627 how such postures.

riorfterior

consider
loads can

high dynamic loading ofthe spine and produce different injuries in different

a)

% Fluid

loss

in 4 hours

Erect

posture

Aexed

posture

Lifting heavy weights activity and results in a high spine. For young men lifting force can reach 8000-9000 N The effect of such forces on tested on cadaveric lumbar adjacent ments); vertebrae the outcome segment and the depends

requires high back muscle compressive force on the with maximum effort, this (Hutton and Adams 1982). the lumbar spine has been motion segments (two intervening very much without disc and ligaon posture. If being flexed,

1OIIIIIE7
b) Diffusion

the

motion

is compressed

of solute
Fig. 5

into disc

The The

percentage of fluid loss (a) from the disc depends on the posture. three values for each disc refer to the anterior annulus, the nucleus pulposus and the posterior annulus. In the erect posture solutes can diffuse more easily into the anterior halfofthe disc than into the posterior half: this imbalance is reversed in the flexed posture. The numbers are proportional to the amount of solute (b) that can diffuse during four hours.

there may be damage to the vertebral body or end-plate (Perey 1957). Ifit is wedged in a flexed posture (Fig. 7), it should not be damaged at all (Hutton and Adams 1982). However, if flexion is excessive (hyperfiexion) so that the posterior intervertebral ligaments is again be crushed of the are overstretched, vulnerable; or there the may disc then anterior be sudden and the motion segment vertebral body may posterior Hutton prolapse 1982).

intervertebral

(Adams

1984). diffusion explained (Adams Posture required ling need least

Flexed into

postures the disc

reverse

this posture

and

the overall results

rate can

of be

is increased.

These deforms

by the way flexed and Hutton 1984). and muscle to stabilise Lucas activity. the spine and

the annulus

Trunk muscle activity is and prevent it from buck1961); obviously, the Erect posture
Flexed posture
Fig. 7

(Morris,

Bresler

for stability is greatest when it is horizontal.

when the spine is vertical and This explains why the activity when the backrest of a (Andersson et al. 1975). angle, posture (as defined a minor effect on muscle

Hyperfiexion

of the back muscles is reduced chair is made to incline more However, for a given backrest by lumbar curvature) has only

The effect ofposture on spinal injury. In the erect posture, the compressive force on the spine (C) can cause end-plate fracture. In the flexed posture the spine has a high compressive strength. In hyperfiexion C can cause posterior disc prolapse or damage to the anterior vertebral body.

activity (Andersson et al. 1975). Figure 6 illustrates how trunk muscle activity is determined by the need to stabilise the spine rather than by the lumbar curvature. Posture have with fairly and considered low spinal static injury. typical loads In standing the previous and sitting In this sections postures section we we

Activities

such

as digging

or repeated

heavy

lifting

subject the lumbar spine to fatigue compressive loading. The effect of such loading has been investigated by testing cadaveric motion segments to destruction and, again, the not results flexed, depend the site on posture. If the motion of failure is in the vertebral segment is body and

on the spine.

end-plate (Hardy et al. 1958). If it is flexed, the lamellae of the annulus can become distorted and this sometimes leads to radial 1983). fissures in the disc (Adams and Hutton

Low

trunk

muscle

activity

DISCUSSION Erect posture

Flexed posture

The advantages and spine are summarised advantages. tips

disadvantages of flexing in Table I. Let us first

the lumbar consider the

Reducing the high stresses that can be found on the of the facet joints may well be significant. In a brposture, the stress between the facet surfaces cartilage 1984) can of and of and levels found in the articular (Dunlop, Adams and Hutton for the very joints (Lewin

High
Trunk

trunk

muscle
Fig. 6

activity

dotic

muscle activity is determined by the need to stabilise the spine rather than by the lumbar curvature. In the upper figures the spine is supported and there is low trunk muscle activity. whereas in the lower figures the spine is unsupported and there is high trunk muscle activity.

exceed the peak the hip and knee may be osteoarthritis

responsible in these

high incidence 1964; Adams

VOL.

67-B.

No.

4. AUGUST

985

628
Table I. Flexed posture

M.

A. ADAMS,

W.

C. HUTTON

spine is about 1000 N, or less than one-tenth required to cause failure (Hutton and Adams
annulus

ofthe 1982). to the damage

force flexed under it are a

.4dvaniugts

Reduced stresses at the apophyseal joints Reduced compressive stress Ofl the posterior Improved transport of disc metabolites High compressive strength ofthe spine
Dl.(1(IV(l?11U,gtS

There is one other posture: because flexion

possible objection can lead to disc

Increased Increased

compressive hydrostatic

stress on the anterior pressure in the nucleus

annulus at low

fatigue loads (see posture and spinal injury section), might lead to similar damage when the load levels low but persist for long periods of time. Generally,
load levels

structure provided there damage

can suffer fatigue damage enough loading cycles are is always a cut-off no value matter ofboad, how accumulates

at low applied. below long

load levels However, which load no is in that

Hutton 1980). Also, the possible extra-articular impingement between the facet tip and the adjacent lamina or pedicle in a lordotic posture could be a source ofbow back pain, especially ifthejoint capsule is trapped between the bony surfaces. Similarly. reducing posterior this stress bus into annulus causes the spinal may outward canal Yorra the compressive stress on the

applied. In the load is probably

case of the lumbar discs, the cut-off above that normally encountered

flexed sitting postures, since people who habitually sit with their lumbar spines flexed have less disc disease (Fahrni and Trueman 1965); it thus seems unlikely that the discs. In this ture on paper we have investigated spine: the effects our of poswith and spine a normal. symptom-free conclusions flexed sitting posture could be damaging to their

be of clinical significance, bulging of the posterior and intervertebral 1957). In many people, of the dural sac and cause 1981). There is also evidence

since annuforamina this pain from

(Brown. Hansen and may lead to compression (Penning and Wilmink

animal experiments that high compressive loading of the posterior annulus can eventually lead to degenerative changes in this part ofthe disc (Lindbbom 1957). Improving the metabolite transport in the disc must be of real value since its overall supply of glucose has been described as barely adequate (Maroudas ci al. 1975). Moreover, the part inner the ofthe disc with the most critical supply. the 1976), is precisely posterior part most annubus (Nachemson helped by a flexed posmetabolite changes Nachemson of a flexed in

do not necessarily apply to the spine of a person back pain. On balance, however, it is mechanically nutritionally advantageous to flatten the lumbar when sitting and when lifting heavy weights. APPENDIX Non-linear explain sure high why elastic wedging behaviour a disc pulposus, strength. of the in flexion disc. increases

In order the

to pres-

in the nucleus compressive

and why a flexed spine has a a simple model of the disc is

presented. Figure 8 shows a disc as three parallel springs representing respectively the anterior annulus, the nucleus pulposus and the posterior annulus. These springs are non-linear, that is, they become stiffer the more they are compressed or stretched. In the figure, the width of each spring on it. is drawn in proportion to the force acting

ture (Adams and Hutton 1984). Deficient transport has been linked with degenerative the disc (Nachemson ci a!. 1970; Holm and 1982). Although the high compressive strength lumbar spine is probably important, force in this area can nonetheless be intra-abdominal pressure (Bartelink flattening gin of involving compressive explained as outlined Neither (flexing) the lumbar safety, which may repeated bending

the compressive reduced in life by 1957). However, it an extra marin activities The increased

spine gives be important and lifting.

Mf

strength of a flexed by the non-linear elastic in the appendix. of the disadvantages

lumbar spine may be behaviour of the disc, listed in Table annulus (Galante the I seems is the 1967) stresses

Tfl
a) c) Mf

to be of much significance. thickest and stiffest part and applied exhibit static more has is evidently well

The anterior of the annulus capable of resisting

to it since it is usually the last part of the disc to degenerative changes (Ritchie and Fahrni 1970). As for the second disadvantage (increased hydropressure serious been in the nucleus until it is realised only the pulposus), this that the pressure at low load may seem increase where

m
b)

Erect posture
Fig. 8 elastic bending behaviour (M) and

cO
of the disc, ofcompressive

flexed
illustrating force (C).

posture
the effects of

demonstrated

levels,

there is no likelihood standing postures,

of mechanical compressive

damage: in sitting or force on the lumbar

Non-linear

THE

JOURNAL

OF BONE

AND

JOINT

SURGERY

THE

EFFECT

OF

POSTURE

ON

THE

LUMBAR

SPINE

629 At high values of of the springs explain the high (see Fig. 8d). The

nor

Figure 8a shows with any bending

the disc moment

under acting

no external forces on it. In Figure 8b force (C) but with posture). In Figure moment (M) but no lies in the so that the annulus and

the disc is subjected no bending moment 8c the disc compressive posterior bending the nucleus in tension. compressive Comparing pulposus The is subjected

to a compressive (as in the erect to a bending

cable to standing and sitting postures. C, applicable to lifting, the non-linearity becomes important and we can now compressive strength of a flexed spine anterior because spring. levels,

force: here the centre of rotation half of the disc (Rolander 1966) moment is resisted by the anterior in compression In Figure

annulus spring is considerably shortened and, of non-linearity, will be stiffer than the nuclear If compressive force (C) is increased to high the additional load will be borne primarily by the By this from porin the and the

and by the posterior annulus 8d, the bending moment and as in a flexed posture. pressure in the nucleus values of C, appli-

anterior annulus, to the advantage ofthe nucleus. mechanism the pressure in the nucleus is prevented rising to levels that tion of the end-plate. erect spine posture (b), is manifestly would fracture When high this mechanism weaker. the weak central loads are applied does not occur

force are combined, 8d with 8b we see that is higher above in flexed only applies postures.

to low

REFERENCES Adams MA, Hutton WC. The effect of posture apophysial joints in resisting intervertebral J Bone Joint Surg [Br] 1980:62 B: 358 62. on the role compressive of the forces. Holm Horst mechanical S, Nachemson disc after spinal S, Brinckmann on the end-plate A. Nutritional fusion. C/in changes in the canine intervertebral Ort/iop 1982: 169:243 58. of the distribution of axial stress body. Spine 1981 :6:217 32. spine be crushed in heavy lift-

Adams MA, Hutton WC. The relevance of torsion to the derangement ofthelumbarspine. Spine 1981:6:241-8. Adams MA, Hutton WC. Prolapsed injury. Spine 1982 :7 :I84 91. Adams MA, lumbar Hutton WC. intervertebral intervertebral disc:

P. Measurement ofthe vertebral

a hyperflexion content interverteinto lumbar of the of

Hutton WC, Adams MA. Can the lumbar ing? Spine 1982:7:586- 90. Hutton WC, Stott JRR, Cyron Spiize 1977:2:202 9. Lewin T. study. Osteoarthritis Acta Orthop BM.

Is spondylolysis

a fatigue

fracture?

The effect of posture on the fluid discs. Spine 1983:8:665 71.

Adams MA, Hutton WC. The effect of fatigue on the lumbar bral disc. J Bone Joint Surg [Br] 1983:65-B: 199-203. Adams MA, Hutton WC.
discs. intervertebral

in lumbar synovial joints: Scand 1964:Suppl 73. degeneration [An:] 1957:39 considered A:933--45.

a morphological as a pressure

The effect of posture on diffusion On/top Trans 1984:8:415 6. Stott joint. JRR. Spine The resistance I 980:5: 245-53.

Lindblom K. Intervertebral-disc atrophy. J Bone Joint Surg

Adams MA, lumbar

Hutton WC, intervertebral

to flexion

Maroudas A, Stockwell RA, Nachemson A, Urban J. Factors involved in the nutrition of the human lumbar intervertebral disc: cellularity and diffusion ofglucose in vitro. J. Anat 1975:120:113-30. Morris JM, Lucas DB, spine. J Bone Joint Bresler B. Role of the trunk in stability Surg [Am] I 961 :43 A: 327 SI.

Andersson BJ, Ortengren R, Nachemson AL, The sitting posture: an electromyographic Ortliop Cliii North Am l97S:6( I): 105-20. Andersson CBJ, Murphy RW, Ortengren influence of backrest inclination and lordosis. Spine 1979:4:S2--8. Bartelink DL. The on the lumbar B:718 25. Brown T, Hansen
sacral spine preliminary

Elfstrom G, Broman H. and discometric study. AL. The on lumbar

of the

R, Nachemson lumbar support

Nachemson AL. The influence of spinal intradiscal pressure and on the tensile sus. Aeta Orthop Stand 1963:33: 183

movements on the lumbar stresses in the annulus fibro207.

role ofabdominal intervertebraldiscs. RJ, Yorra AJ.

pressure in relieving the pressure J BoneJoint Surg[Br] 1957:39

Nachemson A, Lewin T, Maroudas A, Freeman MAR. In vitro diffusion of dye through the end-plates and the annulus fIbrosus of human inter-vertebral discs. Aeta Orthop Seand 1970:41:S89-607. Nachemson AL. The lumbar spine: an orthopaedic challenge. Spine 1976:1:S9-71. Nachemson AL, Schultz AB, Berkson MH. Mechanical properties of human lumbar spine motion segments: influences ofage, sex, disc level and degeneration. Spine 1979:4: 1 -8. Pearcy M. normal Portek 1, Shepherd J. Three-dimensional movement in the lumbar spine. Spine Biomechanics of myelography. endplate investigation. surgery:
I 970: 13:6-71.

with particular report. J Bone

Some mechanical tests on the lumboreference to the intervertebral discs: a Joint Surg [Ant] I 957 :39-A : 1 135 64. and the

Dunlop RB. lumbar Fahrni

Adams MA, Hutton WC. Disc space narrowing facetjoints. J BoneJoint Surg [Br] l984:6&-B:706--lO.

X-ray
1984:9:294

WH, Trueman GE. Comparative of a primitive population with Europeans. J BoneJoint Surg[Br]

radiological study ofthe spines North Americans and Northern 196S:47--B:S52--5. Lumbar intervertebral I972:54-A:492-5l0. lumbar annulus disc fibrosus.

analysis 7.

of a

Penning L, Wilmink JT. study of flexion-extension Perey

lumbosacral dural sac: Spine 1981 :6: 398-408.

Farfan

HF, Huberdeau RM, Dubow HI. degeneration. J Bone Joint Surg [Am]
.4(t(I

0. Fracture of the vertebral experimental biomechanical 1957: Suppl 25. WH. discs. Experimental Can J Surg

in the lumbar spine: an .4cta Ortliop Scand age changes in lumbar

Galante Hardy

JO. Tensile properties of the human Orthop Seand 1967:Suppl 100. WG,
of

Ritchie JH, Fahrni intervertebral

tests

9:690

Lissner HR, Webster JE, Gurdjian ES. Repeated loading the lumbar spine: a preliminary report. Surg Forum 1958: S.

Rolander SD. Motion of the lumbar spine with special reference to the stabilizing effect of posterior fusion: an experimental study on autopsy specimens. Acta Orthop Scand l966:Suppl 90.

VOL.

67-B.

No. 4. AUGUST

1985