A Comparison of Gait Characteristics Between Older Women with [625708]

A Comparison of Gait Characteristics Between Older Women with
and Without Peripheral Neuropathy in Standard and Challenging
Environments
James K. Richardson, MD, Sibylle B. Thies, MS, Trina K. DeMott, MS, PT,
and James A. Ashton-Miller, PhD
OBJECTIVES: To compare gait patterns in older women
with and without peripheral neuropathy (PN) in standard
(smooth surface, normal lighting) and challenging environ-
ments (CE) (irregular surface, low lighting).
DESIGN: Observational, controlled study of 24 subjects.
SETTING: Biomechanical research laboratory.
PARTICIPANTS: Twenty-four older women, 12 with PN
and 12 without PN (mean age /C6standard deviation 567.1/C6
7.9 and 70.2 /C64.3, respectively).
MEASUREMENTS: Gait parameters and, in the 12 PN
subjects, neuropathy severity.
RESULTS: The CE was associated with increases in step
width, step-width variability, step-width range, step width–
to–step length ratio, step time and step-time variability, and
decreases in step length and speed. The PN subjects dem-onstrated a greater step width–to–step length ratio and step
time and shorter step length and slower speed than the
control subjects. In adapting to the CE, the PN subjectsdemonstrated greater increases in step width–to–step length
ratio and step-time variability and a greater decrease in step
length than did the control subjects. In the standard envi-ronment, only one gait parameter correlated with PN se-
verity, whereas in the CE, four gait parameters did so.
CONCLUSION: The subjects demonstrated a gait that
was slower, less efficient, and more variable temporally and
in the frontal plane in the CE. Control and PN subjects
demonstrated similar variability in medial-lateral stepplacement in the CE but at the cost of speed and efficien-
cy for the PN subjects. Because the CE magnified gait dif-
ferences between the two groups of subjects and caused gaitchanges in the PN subjects that correlated with PN severity,the CE may offer improved resolution for detecting gait
abnormalities. J Am Geriatr Soc 52:1532–1537, 2004.
Key words: gait; balance; neuropathyBecause longer axons are usually more vulnerable to
disease, a diffuse peripheral neuropathy (PN) typically
reduces sensorimotor function in a distal-to-proximal pat-
tern, leading to reduced sensation and motor impairments
in the feet and ankles.1Given the importance of somato-
sensory input to the maintenance of balance,2it is not sur-
prising that older persons with PN are at a markedly
increased risk for injurious falls.3,4Furthermore the prev-
alence of PN is high, estimated to affect slightly more than20% of persons aged 65 to 75 in the United States.
5This
prevalence appears to be increasing, likely driven by in-
creasing rates of type II diabetes mellitus and glucose in-tolerance.
6
The study of gait is of interest because a large propor-
tion of falls in older persons, with and without PN, occurduring ambulation.
3,7Differences between neuropathic and
normal gait identified under ideal circumstances (level sur-
face with good lighting) include decreased walking speed,increased frontal plane motion at the ankle,
8increased time
in double support,9and substitution of hip flexion for plan-
tar flexion.10Despite these differences, PN subjects have
not been found to be unstable while walking on a levelsurface with good lighting. Underscoring this point, one
study found that PN patients achieved better dynamic up-
per body stability than did control subjects and concludedthat falls by patients with PN were ‘‘likely due to their in-
ability to develop and execute appropriate avoidance and/
or response strategies when faced with unexpected obsta-cles or large-scale perturbations during locomotion.’’
11This
conclusion agrees with the authors’ clinical experience.
Therefore, gait patterns of older persons with PN in an
environment that causes unexpected perturbations are ofinterest. In this study, two groups of older women, one
group with and the other without PN, were observed walk-
ing on a flat surface with high lighting (standard environ-ment (SE)) and an irregularly contoured surface with low
lighting (challenging environment (CE)). Lower extremity
and trunk positions were continuously recorded so that gaitparameters could be determined. Gait parameters and their
clinical relevance included step-width variability and step-
width range because of their association with frontal planecontrol;
12step width–to–step length ratio because of itsAddress correspondence to Dr. James K. Richardson, Associate Professor,
University of Michigan, Department of Physical Medicine and
Rehabilitation, Medical Professional Bldg. D5200, Ann Arbor, MI 48109.
E-mail: jkrich@umich.eduFrom the Departments of Physical Medicine and Rehabilitation andMechanical Engineering, University of Michigan, Ann Arbor, Michigan.
JAGS 52:1532–1537, 2004
r2004 by the American Geriatrics Society 0002-8614/04/$15.00

relationship to gait efficiency;13step length and step time,
which determine speed and are of functional importance in
time-contingent activities such as crossing a street with alight;
14and step-time variability because of its relationship
to falls.15It was hypothesized that the CE and the presence
of PN would lead to increasing gait variability, both tem-porally and in the frontal plane, and decreased speed and
efficiency.
METHODS
Subjects
PN subjects were recruited from the University of Michigan
Electrodiagnostic Laboratory and Physical Medicine and
Rehabilitation Outpatient Orthotics and Prosthetics Clinic.All patients underwent history, physical examination, andelectrodiagnostic testing. History and physical examination
focused on the identification of central neurological or
musculoskeletal abnormalities. The University of Michiganinstitutional review board approved the project, and all
subjects gave written informed consent.
Inclusion criteria for PN subjects were: age between 50
and 85, symptoms consistent with PN, ability to speak and
understand English, ability to ambulate household distanc-
es without an assistive device, physical examination con-sistent with PN (absent or decreased Achilles reflexes,
decreased distal lower extremity sensation (vibration, pin-
prick and light touch) which improved proximally) andelectrodiagnostic evidence of a diffuse, primarily axonalpolyneuropathy as evidenced by an abnormal sural re-
sponse (absent or amplitude o6mV) and peroneal or tibial
motor responses (absent or amplitude o2.0 or 3.0 mV, re-
spectively). The subjects were also evaluated using the
Michigan Diabetes Neuropathy Score (MDNS), a 0- to 46-
point scale (higher score reflecting more severe PN) thatcorrelates well with more extensive neuropathy staging
scales.
16
Exclusion criteria were weight greater than 300 lbs
(135.7 kg), history of balance disorder unrelated to PN, ev-
idence on physical examination of vestibular or central
neurological dysfunction, or a musculoskeletal abnormalitysuch as severe scoliosis or amputation.
The control subjects were recruited through the Uni-
versity of Michigan Older Americans Independence Center
Research Participant Program. The inclusion/exclusion cri-teria for the control subjects were the same as those for the
PN subjects, with the exception that the control subjects
had no symptoms or signs of PN. The control subjectsunderwent the same screening history and physical exam-
ination as the PN subjects but did not undergo electro-
diagnostic testing.
Subject Preparation and Experimental Apparatus
The subjects were placed in a safety harness that was at-tached by climbing rope to an overhead track (Figure 1).The cords were adjusted to prevent the knees from coming
into contact with the floor when the subject hung unsup-
ported. Subjects wore standard, flat-soled athletic shoessupplied by the laboratory. For all trials, subjects were in-
structed to walk at their own pace, as if they were ‘‘walking
to mail a letter.’’ The SE included a flat, linoleum tile walkingsurface and lighting maintained at more than 900 lux by
overhead light. The CE included an irregularly contoured
walking surface and dimmed room lights ( o50 lux). The
walkway (1.5 /C210 m) was created by randomly arranging
prism-shaped pieces of wood (1.5 cm high, 3.5 cm wide, 6–
16 cm long) beneath dark industrial carpet (Figure 1). Two
optoelectronic markers (infrared-emitting diodes) wereplaced 5 cm apart on a malleable aluminum strip (10 cm /C2
1.5 cm) inserted under the tongue of each shoe. The top
marker was located anterior to the center of the malleoli(estimated to be the center of the tibiotalar joint). A waist
marker was placed on a belt in the midline at the level of the
umbilicus. Two foot switches, each a force-sensing resistor,
Figure 1. Subject preparation and experimental apparatus. Top
illustration: Schematic of irregular walkway used during the ex-
periment. Wooden prisms are oriented randomly beneath an in-
dustrial carpeted surface. Waist and ankle electronic markers areshown. Lower illustration: Sample data for one trial are shown.
Optoelectronic position data for the right (R) and left (L) ankle
markers were used for calculation of step width (SW) and steplength (SL). The waist marker (W) trajectory is also shown.
Kinematic data were scaled according to the 10-cm bars in the
lower left corner.GAIT IN WOMEN WITH AND WITHOUT NEUROPATHY 1533 JAGS SEPTEMBER 2004–VOL. 52, NO. 9

were placed underneath the insole of each shoe. One switch
was placed under the first metatarsophalangeal joint and
the other beneath the calcaneus. Double support was de-fined as the period of time in the gait cycle during which at
least one switch inside each shoe was activated. Step width,
step length, and average speed were measured at 100 Hzusing an optoelectronic camera system (Optotrak 3020,
Northern Digital Corp., Waterloo, Ontario) toward which
the subject walked within the boundaries of the walkway.
Gait/Data Analysis
The kinematic and force data were processed using a cus-tom algorithm written in MATLAB to quantify step width,
step length, and walking speed. Speed was calculated by
taking the time derivative of the waist marker during whatwas defined as the ‘‘comfortable gait speed’’ interval. This
interval was found by excluding data taken when the waist
velocity was less then 85% of the maximum velocity forthat trial. This was done to eliminate steps taken while thesubject accelerated to and decelerated from comfortable
gait speed. Similarly, the other gait parameters were only
included in the analysis during this interval. Step time wasdetermined by calculating the time elapsed between closure
of the right and left metatarsal foot switch during comfort-
able gait speed. Step width and step length were defined asshown in Figure 1.
Statistical Analysis of Gait Parameters
SPSS version 11.0 (SPSS Inc. Chicago, IL) was used for allanalyses. Repeated measures analysis of variance (ANOVA)was used to identify subject group and environment effects,and subject group-by-environment interactions. Age and
body mass index (BMI) (weight in kg/height in m
2) were
used as covariates. Environment-by-age and environment-by-BMI interactions were also explored. Descriptive statis-
tics of gait parameters for both subject groups under each
environmental condition were generated. Step-width rangewas determined by subtracting the smallest from the largest
step width for each subject in each environment. The stand-
ard deviations of step width and step time were used asmeasures of step-width and step-time variability. Two-
tailed paired ttests were used to compare differences in
mean gait parameters within each subject group in the twoenvironmental conditions. Standard two-tailed ttests were
used to compare group differences in demographic data and
mean gait parameters when walking under the same envi-
ronmental condition. Pearson correlations were used to ex-plore relationships between clinical PN severity, as
determined by MDNS score, and gait parameters within
the 12 PN subjects. To adjust for multiple comparisons ofmeans a P-valueo.006 was considered significant and a P-
value/C21.006 and o.0125 was considered to be a trend for
thettests and repeated measures ANOVA.
RESULTS
Subjects
The PN and control subjects were of similar age (mean /C6
standard deviation 67.1 /C67.9 vs 70.2 /C64.3,P5.238) and
BMI (30.3 /C69.5 vs 30.3 /C612.5 kg/m2,P5.993). As antic-
ipated, the PN subjects demonstrated higher MDNS scores(17.7/C65.7 vs 1.4 /C61.5,Po.001). The causes of PN in the
PN subjects were diabetes mellitus (n 55), connective tissue
disease (n 54), past exposure to chemotherapy (n 51) and
idiopathic (n 52).
The Effect of Environment
Repeated measures ANOVA showed that environment had
a significant effect on all gait parameters. The CE was as-
sociated with increases in step width, step-width variability,step-width range, step width–to–step length ratio, step
time, step-time variability, and decreases in step length and
speed (Figure 2). Therefore, the subjects demonstrated aslower, wider-based, and more variable gait in the CE.
Direct comparisons of gait parameters under the two
environmental conditions, along with accompanying pairedttest results, are shown in Table 1. The effect of the CE was
to make the gait of the PN subjects more temporally var-
iable, wider-based, and slower Fthe last due to decreased
step length and increased step time. In general, the controlsubjects demonstrated similar, but less marked, changes in
gait parameters in the CE, but unlike the PN subjects, the
control subjects did not decrease their step length or in-crease step width in the CE. As a result, the control subjects
did not demonstrate an increase in step width–to–step
length ratio or decrease in speed in the CE.
The Effect of Subject Group
Multivariate analysis showed that subject group affected allgait parameters except step width, step-width range, and
variability (Figure 2). The PN subjects demonstrated sig-
nificantly increased step width–to–step length ratio, de-creased step length and speed, and increased step time and
step-time variability (trend). Therefore the PN subjects
demonstrated a gait that was slower due to decreased steplength and more temporally variable. No significant age or
BMI effects were identified.
Direct comparisons of gait parameters between subject
groups, and the results of the corresponding ttests, are
presented in Table 1. Group differences identified are sim-
ilar to those noted with multivariate analysis, with the ex-ception that, in the SE, PN subjects did not show differencesfrom the controls in step-time variability. With regard to
step-width variability and range, there were no group dif-
ferences in either environment, and in the CE, the step-width variability values were nearly equal (40.7 /C610.1 vs
39.8/C610.9 mm).
Subject Group-by-Environment Interactions
Significant subject group-by-environment interactions for
step width–to–step length ratio, step length, and step-timevariability (trend) were identified (Figure 2). More specif-
ically, in adapting to the CE the PN subjects demonstrated a
greater increase in step width–to–step length ratio, a greaterdecrease in step length, and a greater increase in step-time
variability than did the control subjects. No subject group-
by-environment interactions were identified for step-widthvariability, step-width range, or step time. Finally, no sig-
nificant BMI-by-environment or age-by-environment inter-
actions were identified for any of the gait parameters.1534 RICHARDSON ET AL. SEPTEMBER 2004–VOL. 52, NO. 9 JAGS

Effect of PN Severity
In the SE, only one gait parameter (mean step width,
correlation coefficient ( r)5/C00.578; P5.049) correlated
with a clinical measure of PN severity, whereas in the CE,four gait parameters (mean step width, r5/C00.671; P5
.017; step-width variability, r50.577; P5.049; step-width
range, r50.675; P5.020; step-time variability, r5/C00.619;
P5.032) did so. Therefore with increasing PN severity, the
PN subjects’ gait changed minimally in the SE, but in the CE,
their gait became narrower and more variable in terms oftime and width and steps were taken more quickly.DISCUSSION
The Effect of Environment
The fact that the CE induced significant changes in all gait
parameters suggests that its effect on gait was profound. In
the CE, steps were slower, wider-based, and more variable in
the frontal plane and temporally. These gait changes mayhave been advantageous to the subjects given that none fell.The slower pace may have allowed better control by de-
creasing the momentum of the subject’s center of mass. The
wide-based gait decreases the chance of a collision betweenswing and stance limbs, a hazard that has been demonstrated130140150160170180190200Mean Step Width (mm)PN
Controls
Subject Group: P = .017
Environment: P = .003
PN by Environment: P =
.069
202530354045Step Width Variability (mm)PN
Controls
Subject Group: P = .227
Environment: P < .001
PN by Environment: P =
.421
120130140150160170180190200Step Width Range (mm)PN
Controls
Subject Group: P = .494
Environment: P = .004
PN by Environment: P =
.390 130140150160170180190200Mean Step Width (mm)PN
Controls
Subject Group: P = .017
Environment: P = .003
PN by Environment: P =
.069
202530354045Step Width Variability (mm)PN
Controls
Subject Group: P = .227
Environment: P < .001
PN by Environment: P =
120130140150160170180190200Step Width Range (mm)PN
Controls
Subject Group: P = .494
Environment: P = .004
PN by Environment: P =
.390 130140150160170180190200
Standard Challenging
Environment EnvironmentStandard Challenging
Environment Environment
Standard Challenging
Environment EnvironmentStandard Challenging
Environment Environment
Standard Challenging
Environment Environment
Standard Challenging
Environment EnvironmentStandard Challenging
Environment EnvironmentStandard Challenging
Environment EnvironmentMean Step Width (mm)PN
Controls
Subject Group: P = .017
Environment: P = .003
PN by Environment: P =
.069
202530354045Step Width Variability (mm)PN
Controls
Subject Group: P = .227
Environment: P < .001
Environment: P =
120130140150160170180190200Step Width Range (mm)PN
Controls
Subject Group: P = .494
Environment: P = .004
PN by Environment: P =
.390
0.20.250.30.350.40.450.50.55Width:Length RatioPN
Controls
Subject Group: P = .003
Environment: P = .001
PN by Environment: P = .003 350400450500550Mean Step Length (mm)PN
Controls
Subject Group: P < .001
Environment: P < .001
PN by Environment: P < .001
0.40.50.60.70.80.911.11.2Mean Speed (m/sec)PN
Controls
Subject Group: P < .001
Environment: P < .001
PN by Environment: P =
.024
0.40.450.50.550.60.650.7Mean Step Time (sec)PN
Controls
Subject Group: P = .001
Environment: P < .001
PN by Environment: P =
.167
00.020.040.060.080.1Step Time Variability (sec)PN
Controls
Subject Group: P = .013
Environment: P < .001
PN by Environment: P =
.007
Figure 2. Plots demonstrate interactions between subject group and environment for each gait parameter. Adjacent text box lists main
effect significance of subject group and environment and their interaction as determined by repeated measures analysis of variance,
adjusting for body mass index and age. PN 5peripheral neuropathy.GAIT IN WOMEN WITH AND WITHOUT NEUROPATHY 1535 JAGS SEPTEMBER 2004–VOL. 52, NO. 9

in older adults when given challenges to their mediolateral
stability.17
The Effect of Subject Group
The PN subjects demonstrated slower speed due to shorter
step length and longer step time in the SE than the controlsubjects but no differences in step-width range, step-width
variability, or step-time variability. These latter findings sug-
gest that the PN subjects were not less stable than the controlsubjects in the SE, a finding consistent with that of another
study.
11PN subjects demonstrated larger step width–to–step
length ratios in the SE than the controls, suggesting a lessefficient gait for the former.
13Overall, the data suggest that
older female PN patients under ideal conditions are able to
regulate their gait to a similar degree as healthy women of
similar age, but at the cost of speed and efficiency.
In general, the two groups of subjects adapted to the CE
similarly, although the PN subjects often made more ex-
treme adjustments. The most notable exception to this sim-ilarity of adaptation was that the PN subjects markedly
shortened their step length, but the control subjects did not.
The PN subjects simultaneously showed a tendency to in-crease step width. As a result the PN subjects markedly
increased their step width–to–step length ratio and de-
creased speed, whereas the control subjects changed neitherparameter. The PN gait patterns represent a marked con-trast to healthy young men and women who adapted to an
irregular surface by increasing step time and stride length
while maintaining speed.
18
Why did the PN patients make these gait changes in the
CE? Widening the base of support during dual stance may
have increased stability. Alternatively, or additionally, in-creasing step width may have loaded the thigh adductors so
as to improve frontal plane proprioception and control at
the hip. This could compensate for known frontal planeimpairments in ankle proprioception and motor functionassociated with PN.
19,20Increasing step width may also
have decreased the risk of collision between the swing and
stance limbs. Shortening step length may have been neces-sary to accommodate the step-width increase or to mini-
mize time spent in single limb support in the CE. In support
of the latter, previous studies found that PN patients haveless ability to maintain one-legged balance on flat and var-
iable surfaces.
21,22
In the CE, there were no group differences in step-
width range, and the two groups demonstrated nearly iden-
tical step-width variabilities (Table 1 and Figure 2). One
way to interpret these data is that these similarly agedgroups of women tolerated similar degrees of foot-place-ment irregularity in the frontal plane in the CE, but the PN
subjects needed to slow their speed and increase step width–
to–step length ratio, thereby reducing efficiency, to a muchgreater degree than the control subjects to achieve similar
control. With previous evidence that closely links lateral
stability and mediolateral foot placement,
12,23subject con-
cern for maintenance of dynamic stability in the frontal
plane may underlie this strategy.
The Effect of PN Severity
In the CE, more severe PN was positively associated with
step-width variability and range and negatively associatedTable 1. A Comparison of Gait Parameters for Peripheral Neuropathy (PN) and Control Subjects in Standard and Challenging Environments, Using Unadju sted tTests
PN Subjects Control SubjectsEnvironment
ComparisonsSubject Group
Comparisons
Standard
EnvironmentChallenging
EnvironmentStandard
EnvironmentChallenging
Environment PN Group Control Standard Challenging
Gait Parameter Mean /C6Standard Deviation P-value/C3P-valuew
Mean step width, mm 172.0 /C635.3 198.6 /C650.2 145.6 /C629.8 154.3 /C626.1 .012 .086 .060 .013
Step-width variability, mm 34.5 /C67.5 40.7 /C610.1 29.9 /C66.9 39.8 /C610.9 .020 .003 .136 .851
Step width range, mm 157.4 /C627.6 178.7 /C640.3 142.1 /C637.0 186.8 /C660.1 .110 .015 .262 .699
Step width:length 0.37 /C60.08 0.51 /C60.17 0.27 /C60.07 0.29 /C60.06 .006 .063 .003 .001
Mean speed, m/s 0.83 /C60.15 0.66 /C6.17 1.15 /C60.22 1.08 /C60.20 o.001 .014 o.001 o.001
Mean step time, seconds 0.63 /C60.08 0.68 /C60.10 0.53 /C60.06 0.56 /C60.06 o.001 .001 .002 .001
Step time-variability, seconds 0.042 /C60.03 0.092 /C60.03 0.030 /C60.02 0.051 /C60.03 o.001 .001 .157 .002
/C3Paired ttest.
wStandard ttest.1536 RICHARDSON ET AL. SEPTEMBER 2004–VOL. 52, NO. 9 JAGS

with step-time variability and (a trend) step time. One pos-
sible explanation for this is that the subjects with more severe
PN had a progressive decrease in stability during single limbsupport in the CE. Consistent with this, previous study iden-
tified impairments in neuropathic patients in maintaining
balance transferring from bipedal to unipedal stance
22and
during termination of gait.24Given these limitations, sub-
jects with more severe PN may have needed to take a series of
quick successive steps, a common pattern in older personsconfronted with a postural perturbation,
25leading to re-
duced step time and step-time variability. The resultant de-
creased step time gave the subjects with more severe PN less
time to modify the trajectory of the swing limb, leading to anincreased variability of frontal plane foot placement.
Only one gait parameter demonstrated a significant
correlation with PN severity in the SE. In contrast, four gaitparameters significantly correlated with PN in the CE.
These findings suggest that studying PN patients under
challenging circumstances is more likely to yield insight intogait abnormalities than study under ideal conditions. Theabsence of a correlation between PN severity and speed
suggests that the changes in gait are due to factors other
than slowing down.
Implications
The most effective gait allows a person to move over avariety of terrains safely and with minimal effort. The datasuggest that older women with PN are reasonably stable onflat surfaces with good lighting but that their gait speed
under these ideal circumstances is somewhat slower than
needed in the community for crossing streets with walkingsignals (0.83 m/s, or 68% of the 1.22 m/s needed).
14When
the environment is challenging, the speed of the PN subjects
is severely reduced (0.66 m/s or 54% of the 1.22 m/s nec-essary for crossing streets). Moreover, the accompanying
reduction in step length compromises gait efficiency so that
PN patients would likely have reduced endurance in thecommunity. Therefore, although PN subjects appear able to
navigate irregular surfaces under low light conditions safe-
ly, there is a significant cost in terms of speed and efficiency.These patients will likely benefit from a gait aid
26and
planning when moving in the community.
REFERENCES
1. Amato AA, Dumitru D. Approach to Peripheral Neuropathy. In: Dumitru D,
Amato AA, Zwarts MJ, eds. Electrodiagnostic Medicine. Philadelphia: Hanley
& Belfus, 2002, pp 885–898.
2. Fitzpatrick R, McCloskey DI. Proprioceptive, visual and vestibular thresholds
for the perception of sway during standing in humans. J Physiol 1994;478:173–186.
3. Cavanagh PR, Derr JA, Ulbrecht JS et al. Problems with gait and posture in
neuropathic patients with insulin-dependent diabetes mellitus. Diabet Med
1992;9:469–474.4. Richardson JK, Ching C, Hurvitz EA. The relationship between electromyo-
graphically documented peripheral neuropathy and falls. J Am Geriatr Soc1992;40:1008–1012.
5. Franklin GM, Kahn LB, Baxter J et al. Sensory neuropathy in non-insulin-
dependent diabetes mellitus. The San Luis Valley Diabetes Study. Am JEpidemiol 1990;131:633–643.
6. Harris MI, Flegal KM, Cowie CC et al. Prevalence of diabetes, impaired fasting
glucose, and impaired glucose tolerance in U.S. adults: The Third NationalHealth and Nutrition Examination Survey, 1988–94. Diabetes Care 1998;21:518–524.
7. Berg WP, Alessio HM, Mills EM et al. Circumstances and consequences of
falls in independent community-dwelling older adults. Age Ageing 1997;26:261–268.
8. Katoulis EC, Ebdon-Parry M, Lanshammar H et al. Gait abnormalities in
diabetic neuropathy. Diabetes Care 1997;20:1904–1907.
9. Courtemanche R, Teasdale N, Boucher P et al. Gait problems in diabetic ne-
uropathic patients. Arch Phys Med Rehab 1996;77:849–855.
10. Mueller MJ, Minor SD, Sahrmann SA et al. Differences in the gait charac-
teristic of patients with diabetes and peripheral neuropathy compared with
age-matched controls. Phys Ther 1994;74:299–308.
11. Dingwell JB, Cusumano JP, Sternad D et al. Slower speeds in patients with
diabetic neuropathy lead to improved local dynamic stability of continuousoverground walking. J Biomech 2000;33:1269–1277.
12. Bauby CE, Kuo AD. Active control of lateral balance in human walking.
J Biomech 2000;33:1433–1440.
13. Donelan JM, Kram R, Kuo AD. Mechanical and metabolic determinants of the
preferred step width in human walking. Proc R Soc Lond B Biol Sci 2001;268:1985–1992.
14. Langlois JA, Keyl PM, Guralnik JM et al. Characteristics of older pedes-
trians who have difficulty crossing the street. Am J Public Health 1997;87:393–397.
15. Hausdorff JM, Rios DA, Edelberg HK. Gait variability and fall risk in com-
munity-living older adults: A 1-year prospective study. Arch Phys Med Rehabil
2001;82:1050–1056.
16. Feldman EL, Stevens MJ, Thomas PK et al. A practical two-step quantitative
clinical and electrophysiological assessment for the diagnosis and staging ofdiabetic neuropathy. Diabetes Care 1994;17:1281–1289.
17. Maki BE, Edmondstone MA, McIlroy WE. Age-related differences in laterally
directed compensatory stepping behavior. J Gerontol A Biol Sci Med Sci
2000;55A:M270–M277.
18. Menz HB, Lord SR, Fitzpatrick RC. Acceleration patterns of the head and
pelvis when walking on level and irregular surfaces. Gait Posture 2003;18:35–46.
19. Van den Bosch CG, Gilsing MG, Lee SG et al. Peripheral neuropathy effect on
ankle inversion and eversion detection thresholds. Arch Phys Med Rehabil
1995;76:850–856.
20. Gutierrez EM, Helber MD, Dealva D et al. Mild diabetic neuropathy affects
ankle motor function. Clin Biomech 2001;16:522–528.
21. Ashton-Miller JA, Yeh MW, Richardson JK et al. A cane reduces loss of bal-
ance in patients with peripheral neuropathy: Results from a challenging uni-pedal balance test. Arch Phys Med Rehabil 1996;77:446–452.
22. Richardson JK, Ashton-Miller JA, Lee SG et al. Moderate peripheral ne-
uropathy impairs weight transfer and unipedal balance in the elderly. ArchPhys Med Rehabil 1996;77:1152–1156.
23. MacKinnon CD, Winter DA. Control of whole body balance in the frontal
plane during human walking. J Biomech 1993;26:633–644.
24. Meier MR, Desrosiers J, Bourassa P et al. Effect of type II diabetic periph-
eral neuropathy on gait termination in the elderly. Diabetologia 2001;
44:585–592.
25. Jensen JL, Brown LA, Woollacott MH. Compensatory stepping: The biome-
chanics of a preferred response among older adults. Exp Aging Res 2001;27:361–376.
26. Richardson JK, Thies S, DeMott T et al. Interventions improve gait regularity
in patients with peripheral neuropathy while walking on an irregular surface
under low light. J Am Geriatr Soc 2004;52:510–515.GAIT IN WOMEN WITH AND WITHOUT NEUROPATHY 1537 JAGS SEPTEMBER 2004–VOL. 52, NO. 9