The effect of refractive blur on postural stability

$Page 1: The effect of refractive blur on postural stability$
The effect of refractive blur on posturalstability

Vijay Anand1, John Buckley1, Andy Scally2 and David B. Elliott1

1Department of Optometry, University of Bradford, Bradford, and 2The Institute for Health Research,

School of Health Studies, University of Bradford, Bradford, UK

Abstract

The effect of refractive blur upon postural stability was investigated under three conditions: normal

standing, standing with input from the somatosensory system disrupted and standing with input from

the somatosensory and vestibular systems disrupted. Standing stability was assessed using the

centre of pressure (COP) signal from force plate data in four young subjects (mean 23.9 ± 3.1 years)

and five repeated sets of measurements were taken. The subjects looked straight ahead at a

horizontal and vertical square wave pattern of 2.5 cycles (degree))1. Under each of the three test

conditions, standing stability was measured with the optimal refractive correction and under binocular

blur levels of 0, + 1, + 2, + 4, and + 8 D and with eyes closed. In the normal standing condition,

dioptric blur had only a mild effect on postural stability. However refractive blur produced large

increases in postural instability when input from one or both of the other two sensory systems were

disrupted. We hypothesized that dioptric blur would have an even great effect on postural stability if

the visual target used was of higher spatial frequency. This was confirmed by repeated

measurements on one subject using a target of 8 cycles (degree))1. The study highlights the

possible importance of an optimal correction to postural stability, particular in situations (or people)

where input from the somatosensory and/or vestibular systems are disrupted, and where the visual

surrounds are of high spatial frequency.

Keywords: postural stability, refractive blur, somatosensory system, spatial frequency, vestibular

system

Introduction

Balance control in the human body is regulated bythree different yet integrated sensory systems: thesomatosensory, vestibular and visual. The somatosen-sory system includes various receptors that provideinformation about pressure distribution (cutaneous),muscle tension (Golgi tendon organs), joint anglechanges (joint receptors) and muscle length changes(spindles). The vestibular system has two main func-tions, first, concerned with rotational movements of thehead, and secondly providing information regarding

the body’s position with respect to gravity and move-ment. The role of visual information to posturalstability has been studied for many years. For example,Edwards (1946) demonstrated that the availability ofvisual information can reduce postural instability by asmuch as 50%. Furthermore, postural stability has beenshown to be an important risk factor for falls or fear offalling, with the visual contribution to postural stabilitybeing greater in fallers compared with non-fallers(Turano et al., 1994). Vision is particularly importantin stabilising posture under more challenging condi-tions, when input from the somatosensory system isdisrupted (Lord et al., 1991; Elliott et al., 1994; Turanoet al., 1994; Lord and Menz, 2000). Under conditionsof reduced somatosensory input, the amount ofpostural instability has been shown to correlate withmeasures of visual acuity (VA) (Lord et al., 1991),contrast sensitivity (CS) (Lord et al., 1991; Elliottet al., 1994; Turano et al., 1994; Lord and Menz,2000) and stereoacuity (Lord and Menz, 2000).

Received: 19 March 2002

Revised form: 16 June 2002

Accepted: 10 June 2002

Correspondence and reprint requests to: Dr David B. Elliot, Depart-

ment of Optometry, University of Bradford, Bradford BD7 IDP, UK.

E-mail address: [email protected]

Ophthal. Physiol. Opt. 2002 22: 528–534

ª 2002 The College of Optometrists528

In this study, we investigated the effect of refractiveblur upon postural stability under normal standingconditions and under challenging conditions whenthe input from the somatosensory and/or vestibularsystems was disrupted. Understanding the influence ofrefractive blur on postural stability is important giventhe widespread availability of optimal refractivecorrection via spectacles, contact lenses or refractivesurgery. At present, refractive blur is a major cause ofvisual impairment, particularly in the elderly. Forexample, surveys suggest that about one-quarter ofelderly people in the UK could have improved visionwith updated spectacles (Wormald et al., 1992; vander Pols et al., 1999). Furthermore, Jack et al. (1995)indicated that of 200 elderly patients admitted to anacute geriatric clinic, 101 (50.5%) had impaired vision(best eye acuity worse than 6/18 Snellen). They founda particularly high prevalence (76%) of visual impair-ment in the patients admitted following a fall andthat in 79% of these patients, visual impairmentwas reversible, mainly by correcting refractive errors(40%).The previous literature on postural stability changes

with refractive error is limited and somewhat conflicting.Edwards (1946) reported an increase in median bodyinstability of about 51% (mean increase 28%) with theaddition of a + 5 D lens in 50 young subjects. Morerecently, Paulus and colleagues (Paulus et al., 1984,1989; Straube et al., 1990), reported a series of studiesthat examined the effect of dioptric blur on posturalstability with input from the somatosensory systemdisrupted. In their early study (Paulus et al., 1984)they found a 25% increase in postural instability whenfive myopic subjects removed their spectacles (foursubjects had myopic errors between )3 and )5 D andone had a myopic correction of )11 D). They alsofound about a 25% increase in postural instability to+ 4 and + 6 D blur in a group of 10 young subjects ina later study (Straube et al., 1990). This increased to50% and nearly 100%, respectively, for+ 8 and+ 10 Dblur. However, in their other study, they found amuch smaller (� 10%) and not statistically significantincrease in postural instability when 16 myopes andhyperopes between 2 and 5 D removed their spectacles(the ages of the subjects was not reported; Paulus et al.,1989). In addition, two high myopes (more than )15 D)and five aphakics showed no significant differencein postural stability with or without their spectacles(average 2% increase without spectacles), which theysuggested was caused by distortions produced by thehigh-powered spectacles.The aim of this study was to determine the effect of

refractive blur upon postural stability during quietstance with and without input from the somatosensoryand/or vestibular systems disrupted.

Methods

Four subjects, two male and two female (mean age 23.9,SD 3.1 years) were recruited from the student popula-tion at the University of Bradford after screening fortheir suitability for participation in the study. Subjectscreening included a medical history, optimal VAmeasurement and ophthalmoscopic ocular screening.Exclusion criteria included any neuromuscular, skeletalor cardiovascular disorders that could interfere withbalance control, any medication other than that forbirth control, a history of falls, hypotension, amblyopia,strabismus, eye disease or ocular surgery, binocularvisual acuity worse than )0.1 logMAR (Snellen equiv-alent � 6/5) and any visible ocular disease. The tenets ofthe Declaration of Helsinki were followed and the studygained approval from the University ethical committee.Informed consent was obtained after the nature of thestudy had been fully explained.The data were collected on six visits. At the first visit,

subjects were screened for inclusion into the study,informed consent was gained and a �familiarisationsession� was provided. A subjective over-refraction oftheir spectacle prescription was used to obtain thesubjects optimal refractive correction for 4 m. Binocularvisual function was subsequently assessed using VA andCSmeasurements. Binocular VAwas measured using theoptimal refractive correction and an ETDRS logMARchart, using a by-letter scoring system, a chart luminanceof 160 cd m)2 and a 4-m working distance. Binocular CSwas measured using the Pelli–Robson chart at 1 m usingthe optimal refractive correction plus a + 0.75 DSworking distance lens, a by-letter scoring system and achart luminance of 200 cd m)2. Binocular VA and CSwere subsequently remeasured using additional binocularblur lenses of + 1, + 2,+ 4 and + 8 DS.In the following five visits, repeated postural stability

measurements were taken. Standing postural stabilitywas determined while subjects stood stationary on aforce plate (AMTI OR6-7, Advanced MechanicalTechnology Inc., Boston, MA, USA) mounted flushwith the floor. Efficient postural stability is maintainedby keeping the centre of mass within a limited areaover the base of support. Displacements of the centre ofpressure (COP) in the anterior–posterior (A–P) andmedial–lateral (M–L) directions were derived from theforce and moment profiles measured by the forceplate. Fluctuations in the displacement of the COPsignal were quantified using the root mean square(RMS) of the amplitude, sampled over a 25-s period(Winter et al., 1990). These fluctuations reflect theresponse of the CNS to changes of the centre of mass(Horak et al., 1989; Winter et al., 1990). Subjects wereasked to stand still on the force plate for 30-s periodswith their arms by their sides and their feet placed so

Refractive blur and postural stability: V. Anand et al. 529

ª 2002 The College of Optometrists

that the inner edges of both feet were one foot length(their own) apart. They were asked to keep looking atthe middle of a visual target that consisted of a hori-zontal and vertical square wave pattern of 2.5 cycles(degree))1 (Simoneau et al., 1992; Kunkel et al., 1998).This consists of a fundamental sine-wave grating of2.5 cycles (degree))1 plus higher spatial frequency edgeinformation. Intermediate spatial frequencies have beenshown to provide better visual stabilisation of posturethan lower or higher frequencies (Kunkel et al., 1998).Throughout the pattern, theWeber contrastwas 25%andthis was chosen to represent contrast levels typicallyfound in the home, and the target covered an area of1.1 m2 with a viewing distance of 1 m. The target wasadjusted for height for each subject so that its centre wasat eye level. Viewing was binocular with the subject’soptimal correction and the incorporation of a workingdistance lens (+ 0.75 DS, to correct for the differencebetween refraction distance of 4 m and the target distanceof 1 m), using full aperture lenses in a trial frame at adistance of 1 m. Standing postural stability was evaluatedin this control condition and also when the input from thesomatosensory and/or vestibular sensory systems weredisrupted. The somatosensory system inputwas disruptedby asking the subjects to stand on a foam mat (16 cmthick) over the force plate (Lord et al., 1991; Lord andMenz, 2000). The compliant nature of the foam makes itdifficult for the kinaesthetic system to accurately providebody orientation information with respect to the ground.The vestibular system input to balance stability wasdisrupted by asking the subjects to extend their headbackward at 45�. To ensure that there was no change inthe visual input for this condition, the visual target wasraised and orientated to a 45� position to ensure the samevisual target and test distance as in the control condition.Under each of the three test conditions, standing posturalstability was measured with the optimal refractivecorrection for 1 m and under binocular blur levels of+ 1, + 2, + 4 and + 8 D and with eyes closed. Theorder of the 18 postural stability measurements wasrandomized. These 18 measurements were subsequentlyrepeated five times during subsequent visits.The data were analysed using a generalized estimating

equation (GEE) population-averaged model, with �sub-ject� as the grouping variable, using the Stata version 7.0statistical program (Stata Corp., College Station, TX,USA). An exchangeable correlation structure wasjudged to be appropriate, given the experimental design.The terms in the model were:(1) APML: A fixed factor with two levels – A–P and M–L directions of stability.(2) Sensory disruption: A fixed factor with three levels –normal standing, disruption of the somatosensory sys-tem input by asking the subjects to stand on foam anddisruption of both the somatosensory and vestibular

system by asking the subjects to stand on foam and tilttheir head back by 45�.(3) Blur: A fixed factor with six levels – eyes open with noblur, 1, 2, 4 and 8 D blur, and eyes closed. The eyes closedcondition was included in this factor as it was measuredfor each sensory disruption condition and allowed sub-sequent comparison of eye closed measurements to thosewith eyes open and various amounts of blur.The interactions of blur and sensory disruption, and

blur and APML were also included in the model.

Results

The effect of refractive blur on logMAR VA and Pelli–Robson CS for the four subjects is shown in Figure 1. Boxand whisker plots of the RMS COP data in the A–P andM–L planes are shown in Figures 2 and 3, respectively,and show a general increase in COP displacements withincreasing levels of refractive blur and further increaseswhen information from the somatosensory and vestibularsystems were disrupted. We had assumed that the within-subject correlation would, on average, be similar for alltime points. This assumption was checked by includingthe measurement occasion, first as a covariate (checkingfor a linear trend) and secondly as a factor, in the model.In both cases, occasion was not significant (p ¼ 0.40 andp ¼ 0.53, respectively) indicating that there was notraining effect present in the data.Findings from the GEE population-averaged model

are given in Table 1. Results were consistent across allsubjects. Subject, APML, sensory disruption and blurwere all highly significant factors in the model, as wasthe sensory disruption/blur interaction term. TheAPML/blur interaction was shown to be significant,although at a much lower level compared with the otherfactors. The model was checked by plotting the predict-ed values of stability against the actual values andagainst the studentised residuals. There was generallyclose agreement between the actual and predicted values

Figure 1. Subject logMAR visual acuity and Pelli–Robson contrast

sensitivity scores as a function of refractive blur.

530 Ophthal. Physiol. Opt. 2002 22: No. 6


of stability and departures from model assumptionswere not severe, which suggested that the model was agood approximation of the data.

Discussion

Blur and visual function

Blur had a significantly greater effect on logMAR VAthan on Pelli–Robson CS (Figure 1). The Pelli–Robson

chart measures CS at or slightly below the peak of theCS function at about 1.5 cycles (degree))1, and isunaffected by small amounts of refractive blur (Bradleyet al., 1991). The results indicate little or no change inPelli–Robson CS for both 1.0 and 2.0 D blur, similar toprevious findings (Bradley et al., 1991) (Figure 1).

Medial–lateral vs. anterior–posterior postural stability

The APML term in the model was highly significant(p < 0.001), indicating that there were highly significantdifferences between postural stability in the fore-aft andlateral directions in all conditions. In side-by-sidestance, AP stability is under the control of the ankle,whereas ML stability is controlled by the hips (Winter

Figure 2. Box-and-whisker plots of centre of pressure RMS mea-

surements in the (a) anterior–posterior direction and (b) medial–

lateral direction. Data are shown with eyes open (EO) and eyes open

with varying amounts of blur under normal standing condi-

tions(EO*D), and when standing on foam (suffixed by F) and when

standing on foam with their head tilted back 45� (suffixed by FH).

Figure 3. Centre of pressure RMS measurements in the anterior–

posterior direction as a function of the amount of refractive blur.

Linear regression equations were fit to the data for (a) normal

standing, COP ¼ 2.95 + 0.23 blur, r ¼ 0.963; (b) with somatosen-

sory system input disrupted, COP ¼ 5.48 + 0.47 blur, r ¼ 0.982; (c)

with somatosensory and vestibular system inputs disrupted,

COP ¼ 5.93 + 0.56 blur, r ¼ 0.998. The COP with eyes closed in

each of the three conditions are shown for comparison.

Table 1. Summary table for the generalised estimating equation

(GEE) population-averaged model

Factor v2 (d.f.) p-Value

APML 11.90 (1) 0.0006

Sensory input disruption 80.72 (2) <0.0001

Blur 26.82 (5) 0.0001

Blur–sensory input disruption interaction 305.76 (10) <0.0001

Blur–APML interaction 14.18 (5) 0.0145



et al., 1990). The relatively small perturbations topostural stability evoked in this study would tend tobe controlled by an ankle strategy rather than hipmovement (Horak et al., 1989; Winter et al., 1990). Themean COP RMS displacement in the A–P direction wasalways greater than the associated M–L measure (seeFigures 1 and 2). Romberg coefficients (ratio of posturalstability with eyes open with no blur to eyes closed) inthe A–P and M–L directions were similar (Table 2),which indicates that the visual system had a consistentinfluence. However, the APML/blur interaction term inthe model was significant (p < 0.05 with and withoutthe eyes closed data in the model) indicating that blurhad a slightly different effect on A–P and M–L stability.Typically, blur increased A–P instability more, withpercentage increases of 52.0 and 74.2% with 8 D blur onthe normal and foam surface compared with increases of19.7 and 49.9% in the M–L direction. This agrees withprevious findings (Paulus et al., 1984) and suggests thatrefractive blur may have a greater effect on the visualstimuli that provide information to control A–P stabilitythan on the stimuli providing information for M–Lstability. Visual input to A–P stabilization is providedby changes in disparity and target size (Paulus et al.,1984).

The effect of blur

The blur term in the model was highly significant(v2 ¼ 26.82; p ¼ 0.0001, Table 1), indicating posturalinstability increased with refractive blur (if the eyesclosed data were removed from the model little changewas found; v2 ¼ 24.51, d.f. ¼ 4, p ¼ 0.0001). The blur/sensory disruption interaction term in the model wasalso highly significant (p < 0.0001), indicating that blurhad significantly different effects under the differentconditions of sensory system disruption. This termremained highly significant if the eyes closed data wereremoved from the model (v2 ¼ 35.62, d.f. ¼ 8,p < 0.0001). These results highlight the fact that theinput from the visual system became increasingly

important as the input from the other two systems wasdisrupted, which is in agreement with previous literature(Lord et al., 1991; Teasdale et al., 1991; Elliott et al.,1994; Turano et al., 1994; Lord and Menz, 2000). In thenormal standing condition, dioptric blur had only a mildeffect on postural stability. The COP RMS mediansincreased by 19.7% (M–L) and 52.0% (A–P) with theaddition of 8.0 D blur (v2 ¼ 18.10, p < 0.0001). Theinput from the visual system tends to be ignored in thissituation, as indicated by the similarity of measurementsin the eyes closed and 8 D blur conditions (13%difference for M–L, no difference for A–P; v2 ¼ 0.68,p ¼ 0.41). However, when input from one or both of theother sensory systems were disrupted, the visual systemplayed an increasingly important role, i.e. with thesomatosensory system disrupted, changes in posturalstability with dioptric blur were larger, with COP RMSmedians increasing by 49.9% (M–L) and 74.2% (A–P)with the addition of 8.0 D blur (v2 ¼ 64.79, p � 0.0001).In addition the COP RMS with eyes closed in thiscondition was also larger than the COP RMS with the8 D blur (28.5% reduction with the blurred visualinformation for M–L, 16.0% reduction for A–P;v2 ¼ 16.11, p < 0.0001). Furthermore, when the inputfrom both the somatosensory and vestibular systemswas disrupted, the increases in postural instability withdioptric blur increased further. By moving the visualstimulus to the 45-degree position, the effect of the headtilt was to disrupt input from the vestibular systemalone, with theoretically no effect on input from thevisual system. With both the somatosensory andvestibular systems disrupted mean COP RMS increasedby 78.1% (M–L) and 77.6% (A–P) with the addition of8.0 D blur (v2 ¼ 92.23, p � 0.0001) and the reductionin mean COP RMS between the eyes closed and 8 Dblur condition was again generally larger (27.3%reduction with the blurred visual information forM–L, 35.3% reduction for A–P; v2 ¼ 116.29,p � 0.0001). The increasing importance of the visualsystem in controlling stability can also be illustratedusing the Romberg coefficients in the three measurementconditions (Table 2).

Comparison with previous studies

Our results of increases in mean COP RMS in normalstanding of 15.8% (M–L) and 40.6% (A–P) with 4.0 Dblur compare well with those of Edwards (1946), whofound an increase of 51% (mean increase 28%) with theaddition of a + 5 D lens in 50 young subjects. However,our results suggest a much larger effect of refractive bluron balance than the studies of Paulus and colleagues(Paulus et al., 1984, 1989; Straube et al., 1990); but whythis was so is unclear. Our findings indicate that theeffect of dioptric blur is influenced by the usefulness of

Table 2. Group mean Romberg coefficients (COP RMS in eyes

open/COP RMS in eyes closed) in three standing conditions and two

directions of postural stability. Values closer to zero indicate a

greater importance of vision

Medial–lateral

direction

Anterior–posterior

direction

Normal standing condition 0.73 0.66

With disrupted

somatosensory input

0.48 0.48

With disrupted

somatosensory and

vestibular input

0.41 0.36



the input from the somatosensory and vestibularsystems. The differences in our findings and those fromPaulus et al. may have resulted from methodologicaldifferences in how the experimental set-up disrupted theinput from these systems. The report that suggested theleast blur effect (Paulus et al., 1989) reported stability interms of sway path which may be a less sensitivemeasure than RMS COP. In addition, they measuredpostural instability when 16 myopes and hyperopesbetween 2 and 5 D removed their spectacles (the ages ofthe subjects was not reported and the hyperopes mayhave been able to accommodate clearly on the targetduring the test period; Paulus et al., 1989). The visualtarget used in the experimental set-up will also be animportant factor. Paulus et al. (1984) used a screenrandomly covered with different coloured dots ofdifferent sizes [0.02–0.57 degrees in diameter, funda-mental frequencies of 1–17 cycles (degree))1]. We used ahorizontal and vertical square wave pattern of 2.5cycles (degree))1 with 25% contrast. Because dioptricblur has been shown to have a greater effect on visionwith targets of lower contrast and higher spatialfrequency (Campbell and Green 1965), we hypothesisedthat using a visual target with higher spatial frequencywould result in blur having a greater effect on posturalstability. To test this hypothesis, we repeated allmeasurements on one subject using a target of 8cycles (degree))1. The results indicate that blur had anincreased effect on standing stability with the 8cycles (degree))1 target compared with the 2.5cycles (degree))1 target (Figure 4). An analysis ofvariance (ANOVAANOVA) indicated that postural instabilitywas significantly greater with the 8 cycles (degree))1

target (F1,17 ¼ 33.6, p < 0.001). Given that dioptricblur has also been shown to have a greater effect onvision with targets of lower contrast (Campbell andGreen 1965), it is likely that using a visual target withlower contrast would also result in blur having a greatereffect on postural stability. It is possible that if the visualtarget used by Paulus et al. did not provide enoughvisual information to aid postural stability, then anydioptric blur of the target would have less chance todisrupt stability. Further research is required to deter-mine the effects of monocular blur on postural stability,whether the effects of dioptric blur are different forelderly subjects and whether the effects on posturalstability of adding positive lenses is similar to that ofsubjects removing their spectacles. It is possible that themagnification effects of positive lenses could also havesome influence of postural stability. We also wish toinvestigate the effects of dioptric blur on subjects whosequality of input from the somatosensory and/or vesti-bular systems is affected by diseases such as diabetes(Oppenheim et al., 1999).

Summary

Findings indicate that increasing levels of refractive blurcan significantly increase postural instability, particularlywhen the visual surround is of high spatial frequency andwhen the information provided by the somatosensoryand/or vestibular systems is disrupted. As it is likely thatpoor quality input from these sensory systems occurswiththick carpeting or shoes (Lord and Bashford 1996;Redfern et al., 1997), when looking or reaching to acupboard above eye level (Simoneau et al., 1992), or withvarious systemic diseases (Oppenheim et al., 1999), thesefindings highlight that individuals requiring refractivecorrection may be at a greater risk of falling than thosewho have optimal refractive correction.

Acknowledgements

This work was supported by VPPP Foundation.

References

Bradley, A., Hook, J. and Haeseker, J. (1991) A comparison ofclinical acuity and contrast sensitivity charts: effect ofuncorrected myopia. Ophthal. Physiol. Opt. 11, 218–226.

Campbell, F. W. and Green, D. G. (1965) Optical and retinalfactors affecting visual resolution. J. Physiol. 181, 576–593.

Edwards, A. S. (1946) Body instability and vision. J. Exp.

Psycho. 36, 526–535.Elliott, D. B., Patla, A. E., Flannagan, J. G. et al. (1994) TheWaterloo vision and mobility study: postural control strat-

egies in subjects with ARM. Ophthal. Physiol. Opt. 15, 553–559.

Figure 4. Comparison of centre of pressure RMS measurements in

the anterior–posterior direction as a function of the amount of

refractive blur using visual targets of 2.5 and 8 cycles (degree))1.



Horak, F. B., Shupert, C. L. and Mirka, A. (1989) Compo-

nents of postural dyscontrol in the elderly: a review.Neurobiol. Ageing 10, 727–738.

Jack, C. I. A., Smith, T., Neoh, C., Lye, M. and McGalliard,

J. N. (1995) Prevalence of how vision in elderly patientadmitted to an acute geriatric unit in Liverpool – elderlypeople who fall are more likely to have low vision.

Gerontology 41, 280–285.Kunkel, M., Freudenthaler, N., Steinhoff, B. J., Baudewif, J.and Paulus, W. (1998) Spatial frequency related efficacyof visual stabilisation of posture. Exp Brain Res. 121,

471–477.Lord, S. R. and Bashford, G. M. (1996) Shoe characteristicsand balance in older women. J. Am. Ger. Soc. 44, 429–433.

Lord, S. R., Clark, R. D. and Webster, I. W. (1991) Visualacuity and contrast sensitivity in relation to falls in anelderly population. Age Ageing 20, 175–181.

Lord, S. R. and Menz, H. B. (2000) Visual contribution topostural stability in older adults. Gerontology 46, 306–310.

Oppenheim, U., Kohen-Raz, R., Kohen-Raz, A., Alex, D. andAzarya, M. (1999) Postural characteristics of diabetic

neuropathy. Diabetes Care 22, 328–332.Paulus, W. M., Straube, A. and Brandt, Th (1984) Visualstabilization of posture-physiological stimulus characteris-

tics and clinical aspects. Brain 107, 1143–1163.Paulus, W. M., Straube, A., Quintern, J. and Brandt, T. (1989)Visual performance in ametropia and with optical distortion

produced by bifocals and multifocals. Acta Otolaryngologia486, 243–246.

Redfern, M. S., Moore, P. L. and Yarskey, C. M. (1997) The

influence of flooring on standing balance among olderpersons. Human Factors 39, 445–455.

Simoneau, G. G., Leibowitz, H. W., Ulbrecht, J. S., Tyrell, R.

A. and Cavanagh, P. R. (1992) The effects of visual factorsand head orientation on postural steadiness in women 55–70years of age. J. Gerontol. 47, M151–M158.

Straube, A., Paulus, W. M. and Brandt, Th (1990) Influence ofvisual blur on object motion detection, self motion detectionand postural balance. Behav. Brain Res. 40, 1–6.

Teasdale, N., Stelmach, G. E. and Breunig, A. (1991) Postural

instability characteristics of the elderly under normal andaltered visual and support conditions. J. Gerontol. 46, B238–B244.

Turano, K., Rubin, G. S., Herdman, S. J., Chee, E. and Fried,L. P. (1994) Visual stabilization of posture in the elderly:fallers vs. nonfallers. Optom Vision Sci. 71, 761–769.

Van der Pols, J. C., Thompson, J. R., Bates, C. J. et al.(1999) Is the frequency of having an eye test associatedwith socio-economic factors? A national cross sectionalstudy in British elderly. J. Epidemiol. Commun. H. 53, 737–

738.Winter, D. A., Patla, A. E. and Frank, J. S. (1990) Assessmentof balance control in humans. Med. Prog. Tech. 16, 31–51.

Wormald, R. P. L., Wright, L. A., Courtney, P. et al. (1992)Visual problems in the elderly population and implicationsfor services. Br. Med. J. 304, 226–1229.



Documents

The effect of refractive blur on postural stability