FRONTAL PLANE HIP AND ANKLE SENSORIMOTOR FUNCTION, NOT

AGE, PREDICTS UNIPEDAL STANCE TIME

LARA ALLET, PhD,1,2,3 HOGENE KIM, MS,4 JAMES ASHTON-MILLER, PhD,4,5 TRINA DE MOTT, PT,2
and JAMES K. RICHARDSON, MD2
1
Department of Physiotherapy, University of Applied Sciences of Western Switzerland, Geneva, Switzerland
2
Department of Physical Medicine & Rehabilitation, University of Michigan, Ann Arbor, Michigan, USA
3
Health Care Directorate, University Hospitals and University of Geneva, Health Care Directorate, Geneva, Switzerland
4
Department of Biomedical Engineering, University of Michigan, Ann Arbor, Michigan, USA
5
Department of Mechanical Engineering, Biomechanics Research Laboratory, University of Michigan, Ann Arbor, Michigan, USA
Accepted 10 October 2011

ABSTRACT: Introduction: Changes occur in muscles and at the hip is of greater importance to equilibrium
nerves with aging. In this study we explore the relationship in the frontal plane than control at the ankle. For
between unipedal stance time (UST) and frontal plane hip and
ankle sensorimotor function in subjects with diabetic neuropa- example, a whole-body inverted pendulum model
thy. Methods: UST, quantitative measures of frontal plane ankle of medial–lateral control during human walking
proprioceptive thresholds, and ankle and hip motor function suggests that the hip exerts the primary influence,
were tested in 41 subjects with a spectrum of lower limb senso-
rimotor function ranging from healthy to moderately severe dia- and that minor errors in hip motion are compen-
betic neuropathy. Results: Frontal plane hip and ankle sated by adjustments at the subtalar joint.12 Simi-
sensorimotor function demonstrated significant relationships larly, a second model demonstrated that foot
with UST. Multivariate analysis identified only composite hip
strength, ankle proprioceptive threshold, and age to be signifi- placement in the frontal plane, which is regulated
cant predictors of UST (R2 ¼ 0.73), explaining 46%, 24%, and by hip abduction/adduction, was the most efficient
3% of the variance, respectively. Conclusions: Frontal plane hip method for controlling frontal plane balance while
strength was the single best predictor of UST and appeared to
compensate for less precise ankle proprioceptive thresholds. walking.13 Other studies have provided experimen-
This finding is clinically relevant given the possibility of strength- tal support for these models and demonstrated the
ening the hip, even in patients with significant peripheral neu- importance of hip frontal plane strength for bal-
ropathy.
Muscle Nerve 45: 578–585, 2012 ance control in elderly subjects when they negoti-
ate obstacles14 and for fall prevention.15,16
However, no study has described the relationship
Quantitative and qualitative changes occur in between lower limb afferent and efferent neuromus-
muscles and nerves with aging.1 These changes
cular capacities relevant to frontal plane control in
include a decrease in the number of alpha moto-
older subjects with a demonstrably significant range
neurons, reduced motoneuron excitability, and loss
of peripheral neurological function. For example,
of type II muscle fibers, leading to decreased muscle
none of the aforementioned biomechanical models
mass and slower muscle response latencies.2 Such
or experimental studies addressed the role of distal
changes, which adversely affect motor control and
balance in older persons, are even more marked afferent function (i.e., ankle proprioception). Simi-
among older persons with peripheral neuropathy larly, evaluations of lower limb neuromuscular
(PN), a common complication of diabetes mellitus. capacities associated with balance deficits in subjects
In such patients, the neuropathy is usually length-de- with PN studied either ankle proprioception5 or
pendent and results in distal sensorimotor dysfunc- ankle joint motor function,17,18 but not both, and no
tion of varied severity. As a result, diabetic patients study has evaluated hip motor function in this high-
have decreased balance,3–6 altered gait,7 and risk population.
increased fall risk,8,9 compared with healthy controls. Unipedal stance time (UST) is a convenient
Control of frontal plane stability is particularly clinical measure of balance that evaluates frontal
important given that lateral falls are associated plane postural control. It is the most challenging
with hip fractures in older adults.10,11 Biomechani- activity within the widely used Berg Balance
cal models and human studies suggest that control Scale.19 Moreover, UST is associated with
frailty,20,21 PN,6,22 activity level,23 and falls in older
Abbreviations: BMI, body mass index; COM, center of mass; MVC, persons with PN24 and without PN,25,26 and
maximum voluntary contraction; MDNS, Michigan Diabetes Neuropathy decreases markedly with age.27,28 Therefore, our
Score; PN, peripheral neuropathy; RTD, rate of torque development; UST,
unipedal stance time objective was to elucidate the relationships
Key words: age, balance, diabetic neuropathy, muscle strength, between UST and lower limb neuromuscular
proprioception
Correspondence to: L. Allet, University of Applied Sciences of Western capacities relevant to frontal plane postural control
Switzerland (HEDS) Dpt of Physiotherapy, Rue des Caroubiers 25, in older subjects with a spectrum of neuromuscu-
CH1227 Carouge, Switzerland; e-mail: [email protected]
lar function. The primary hypothesis was that hip
V
C 2011 Wiley Periodicals, Inc.

Published online 17 October 2011 in Wiley Online Library

motor function would be an independent predic-
(wileyonlinelibrary.com). DOI 10.1002/mus.22325 tor of UST. Support for this hypothesis has clinical
578 Sensorimotor Function and Balance MUSCLE & NERVE April 2012
relevance, given the fact that PN predominantly normal electrodiagnostic studies. Otherwise, they
affects distal function, which leaves the potential met the same inclusion criteria as the PN
for strengthening of hip musculature.1 subjects.

METHODS Entrance Evaluation. During the physical examina-

Forty-one subjects (16 healthy old and 25 with PN tion that focused on neurological and musculoskel-
due to diabetes) were recruited under a protocol etal findings, inclusion and exclusion criteria were
approved by our institutional review board. Written verified. Neuropathy severity was further deter-
informed consent was obtained from all partici- mined using the 46-point-scale MDNS29,30 (higher
pants. Subjects were recruited from the University score reflecting more severe neuropathy), evaluat-
of Michigan Orthotics and Prosthetics Clinic, En- ing distal sensory impairment, distal muscle
docrinology Clinic, and the Older Americans Inde- strength, and muscle stretch reflexes. Finally, all
pendence Center Human Subjects Core. Inclusion subjects underwent nerve conduction studies of
criteria for PN subjects were: the fibular nerve, as described previously.
UST. Subjects performed three trials of UST on
• Age 50–85 years. each foot.29,31 Subjects started with an intramalleo-
• Weight <136 kg. lar distance of approximately 15 cm, and then
• Known history of diabetes. transferred weight to one foot. To standardize the
• Able to walk household distances without an as- test sequence and timing of weight transfer to the
sistance/assistive device. extent possible the examiner asked, ‘‘Ready?’’ and
• Strength of ankle dorsiflexors, invertors, and upon receiving assent from the subject, gave the
evertors at least anti-gravity (grade
3 by manual cadence command, ‘‘One, two, up.’’ Subjects were
muscle testing). required to raise their non-stance limb at the ‘‘up’’
• Symptoms and signs consistent with PN: symmet- command. UST maximum was set at 30 s.
rically altered sensation in lower extremities, and
Michigan Diabetes Neuropathy Score (MDNS) of Neuromuscular Capacity Testing. Hip Abduction and

10.29 Adduction Muscle Strength. A custom whole-body
• Electrodiagnostic evidence of a diffuse PN as evi- dynamometer (BioLogic Engineering, Inc.) was
denced by bilaterally abnormal fibular motor used to measure the maximum voluntary contrac-
nerve conduction studies (absent or amplitude <2 tion (MVC) and maximum rate of torque develop-
mV and/or latency >6.2 ms and/or conduction ment (RTD) in the frontal plane at the hip.32 This
velocity <41.0 m/s), stimulating 9 cm from the re- dynamometer was found to be sensitive to the
cording site over the extensor digitorum brevis effects of age, gender, and hip angle when isomet-
distally, and distal to the fibular head proximally. ric hip strength was measured in a group of 24
young and 24 older subjects. In addition, the appa-
Exclusion criteria for PN subjects were: ratus demonstrated the ability to resolve torque
with a precision of 0.5 Newton meter (Nm). Retest
• Accidental fall 1 month prior to testing. reliability has not been evaluated; however, it was
• History or evidence of any significant central anticipated that reliability would be similar to that
nervous system dysfunction (i.e., hemiparesis, found with isometric testing in other populations
myelopathy, or cerebellar ataxia). (e.g., with a mean day-to-day difference of 10%
• Neuromuscular disorder other than PN (e.g., and a coefficient of repeatability of 11–33%).33
myopathy or myasthenia gravis). The dynamometer features a horizontal bench on
• Evidence of vestibular dysfunction. which the subject lies fully supported, allowing all
• Angina or angina-equivalent symptoms with measurements to be made in a gravity-free plane.
exercise. The pelvis and upper body were immobilized using
• Plantar skin sore or joint replacement within the adjustable harness straps at multiple points. During
previous year. maximum voluntary abduction strength tests, sub-
• Symptomatic postural hypotension. jects progressively increased their isometric effort
• Significant musculoskeletal deformity (i.e., from rest to their maximum over a count of three,
amputation or Charcot changes). held it for 2 s, and relaxed. Patients were encour-
• Lower limb or spinal arthritis or pain that limits aged verbally. To quantify rate of isometric
standing to <10 min, or walking to less than one strength development, subjects performed an
block. abduction against the lever arm as fast and as hard
• The healthy older adults were without neuro- as possible for 3 s.34 Three trials were performed
pathic symptoms, had MDNS of <10, and had with 1-min rests between trials. Subjects performed
Sensorimotor Function and Balance MUSCLE & NERVE April 2012 579
analogous maneuvers in the opposite direction for Statistics. Statistics were conducted using SPSS
hip adduction strength and rate of isometric for Windows (release 11.0.1.2001; SPSS, Inc., Chi-
strength testing. cago, Illinois). Descriptive statistics were calculated
Ankle Muscle Strength. During testing of the for all measures, including a composite score of
rate of ankle strength development, subjects stood frontal plane ‘‘hip strength,’’ calculated as the
on the test foot on a force plate (OR-6; Advanced mean of the mean peak abduction and adduction
Mechanical Technology, Inc.) and moved the cen- MVCs. Data were examined for normality and
ter of ground support reaction from the lateral screened for outliers. Pearson’s product-moment
margin of the foot to the medial margin as quickly correlation coefficients were calculated to assess
as possible, then again to the lateral margin, as relationships between neuromuscular capacities
described elswehere.17 Three trials, each with five and UST.
medial–lateral movements, were performed. Sub- A regression model determined independent
jects were allowed to touch a horizontal railing to predictors of UST. Variables were entered stepwise
keep their balance. in the order of their strength of correlation. To
During maximum voluntary strength testing, reduce the number of independent variables, only
subjects stood on the force platform touching the the best predictor variable for ankle motor func-
hand rails on both sides as needed. Subjects were tion and the best predictor variable for hip motor
then asked to lift one leg, shift their center of grav- function were retained in the final regression
ity as far lateral under their foot as they could, and model, along with the identified covariates (age
lift their hands from the rails for 3 s. The test was and body mass index).
repeated three times for the lateral, and then like- To determine whether hip strength might com-
wise repeated for the medial margin of the foot. pensate for distal afferent deficiencies (less precise
Ankle Proprioception Threshold. Subjects stood ankle proprioceptive thresholds), the residuals of
with the test foot in a 40  25 cm cradle that was the regression model using UST as the outcome
rotated by an Aerotech 1000 servomotor equipped variable and proprioceptive threshold and age as
with an 8000-line rotary encoder, as described by predictor variables were saved and ranked by mag-
Son et al.5 After an audible cue, a single ankle nitude. The hip strength of the 12 subjects with
inversion or eversion rotation of 0.1 –3 magni- the highest residuals was then compared with the
tude was randomly presented at 5 /s. The subject hip strength of the 12 subjects with the lowest
then pressed a joystick handle in the direction of residuals using a two-sided Student’s t-test. A simi-
the perceived foot rotation. Four blocks of 25 trials lar analysis was performed to determine whether
(randomly, 10 eversion, 10 inversion, and 5 more precise ankle proprioceptive thresholds
dummy trials) were presented interspersed with 2– might compensate for decreased hip strength. The
5-min rest intervals. The outcome measure was the significance level for all tests was set at P < 0.05.
ankle proprioception threshold (TH100), defined
as the smallest rotational displacement of the RESULTS
ankle that a subject could reliably detect with Of the 91 potential subjects, 21 did not pass the tel-
100% accuracy.35 ephone screening, and 18 elected to not partici-
pate. Of the 52 remaining subjects, 3 had schedul-
ing conflicts, and 5 failed the screen. Of those 44
Data Processing. Signals were amplified to volt lev- remaining subjects, 1 was lost to follow-up, and 2
els before being acquired using a 12-bit analog-to- dropped out due to medical concerns. Finally, 41
digital converter sampling at 100 HZ. The MVC subjects were enrolled. The means and standard
efforts at the hip and ankle, as well as the maximal deviations of age, body mass index (BMI), and
RTD, were normalized by individual body size, MDNS, together with the participants’ neuromuscu-
defined as the parameter body height multiplied lar capacities and UST data, are shown in Table 1.
by weight (units of Nm). Strength data were proc-
essed using a second-order least-squares polyno- Correlations. Correlations between UST and fron-
mial fit (LabVIEW) to determine the peak value. tal plane lower limb neuromuscular function were
The mean peak value obtained from the three tri- strong, and many of the functions explained more
als for each test type was used for the statistical than a third of the variability in UST (Table 2).
analyses. To determine each proprioceptive thresh- This includes all of the functions measured except
old, we calculated the mean TH100 from the four for ankle inversion and eversion MVC, and hip
blocks of 25 trials in each test direction. A sum- abduction and ankle eversion RTD. Age and BMI
mary measure of ankle proprioception was found were substantially less strongly associated with UST
from the sum of the inversion and eversion pro- than were the majority of neuromuscular RTD
prioception thresholds. variables.
580 Sensorimotor Function and Balance MUSCLE & NERVE April 2012
 Table 1. (SD) of demographic and neuromuscular function results.
Non-diabetic subjects (N ¼ 16) Diabetic patients (N ¼ 25)
Parameter All Men (N ¼ 6) Women (N ¼ 10) All Men (N ¼ 15) Women (N ¼ 10)
Age (years) 67.81 (8.97) 67.83 (11.02) 67.8 (8.16) 70.04 (8.16) 71.53 (7.17) 67.8 (9.39)
BMI (kg/m2) 28.35 (7.18) 26.24 (3.25) 29.62 (8.68) 32.41 (6.44) 30.25 (5.36) 35.66 (6.81)
Unipedal stance time (s) 22.34 (11.1) 21.87 (11.98) 22.62 (11.19) 6.9 (6.91) 15.1 (11.03) 9.52 (9.36)
MDNS (0—46 points) 1.69 (3.77) 2.5 (6.12) 1.2 (1.48) 13.56 (6.04) 14.13 (6.5) 12.7 (5.48)
Hip abduction MVC (Nm/Nm) 0.041 (0.024) 0.051 (0.028) 0.035 (0.02) 0.031 (0.01) 0.032 (0.011) 0.03 (0.009)
Hip abduction RTD (N.m/N.m.s) 0.255 (0.188) 0.312 (0.224) 0.22 (0.166) 0.154 (0.096) 0.155 (0.104) 0.154 (0.087)
Hip adduction MVC (Nm/Nm) 0.047 (0.018) 0.051 (0.018) 0.045 (0.018) 0.033 (0.012) 0.035 (0.013) 0.03 (0.011)
Hip adduction RTD (Nm/Nm/s) 0.29 (0.226) 0.4 (0.224) 0.224 (0.211) 0.199 (0.151) 0.19 (0.176) 0.213 (0.112)
Ankle eversion MVC (cm)* 1.275 (0.502) 1.501 (0.665) 1.135 (0.348) 1.017 (0.442) 1.009 (0.541) 1.028 (0.257)
Ankle inversion MVC (cm)* 2.187 (0.501) 2.544 (0.381) 1.932 (0.426) 1.596 (0.659) 1.585 (0.729) 1.614 (0.563)
Ankle inversion RTD (Nm/Nm/s)* 0.188 (0.096) 0.231 (0.086) 0.161 (0.097) 0.104 (0.064) 0.113 (0.067) 0.091 (0.06)
Ankle eversion RTD (Nm/Nm/s)* 0.243 (0.114) 0.326 (0.136) 0.191 (0.059) 0.141 (0.069) 0.15 (0.079) 0.128 (0.051)
Proprioceptive threshold ( )† 0.986 (0.757) 1.154 (1.093) 0.885 (0.511) 2.391 (1.313) 2.208 (0.832) 2.665 (1.839)

MVC, maximum voluntary contraction; RTD, rate of torque development.

*N ¼ 13 valid cases for non-diabetic subjects, N ¼ 24 valid cases for diabetic patients, and N ¼ 12 for non-diabetic subjects.

Multivariate Analyses. The final regression model DISCUSSION

included UST as the outcome variable and hip We have quantified sensory and motor lower limb
strength (as defined in Methods), ankle inversion neuromuscular capacities in a group of older sub-
RTD, ankle proprioception, and the covariates age jects over a spectrum of peripheral neurologic
and BMI as independent variables (Table 3). Maxi- health. There are three novel, clinically significant
mum hip strength was the most important predic- findings: (1) maximum voluntary hip strength in
tor of UST, explaining almost half of its variability. the frontal plane was the single best predictor of
Ankle proprioceptive thresholds and age also con- UST, a result consistent with the primary hypothe-
tributed to the model in a significant manner. The sis; (2) maximum voluntary hip strength and ankle
former explained an additional 25% of the var- proprioceptive thresholds explained the majority
iance in UST, and age explained just 3%. Overall, of the variance in UST, with age playing a trivial
the model explains nearly three fourths of the vari- role; and (3) increased hip strength appears to
ability in UST. compensate for less precise ankle proprioception.
Although frontal plane hip strength is not rou-
UST and the Ratio of a Composite Variable of Hip tinely evaluated in studies of postural control,
Strength to Ankle Proprioception. After observing there is evidence supporting its importance. For
the relationship between hip strength and UST example, during bipedal stance, anterior–posterior
and the inverse relationship between propriocep- balance is under ankle control (plantar and
tive threshold and UST, we formed a new variable,
the ratio of hip strength to proprioceptive thresh-
old. This variable was found to explain >70% of
Table 2. Bivariate correlations between unipedal stance time and
the variability of UST (Fig. 1). neuromuscular function, age, and body mass index.

Hip Strength Can Compensate for Imprecise Ankle Correlation

coefficient
Proprioception. After performing regression of
Parameter with UST P
ankle proprioceptive threshold and age on UST,
Hip strength 0.672 0.000
the residuals for all subjects were ranked, and the
Hip adduction MVC 0.664 0.000
hip strength of the upper one third (representing Hip adduction RTD 0.645 0.000
subjects who had longer USTs than would be Ankle inversion RTD 0.644 0.000
expected for prioprioceptive threshold and age) Proprioceptive threshold 0.643 0.000
was compared with that of the lower one third. The Hip abduction MVC 0.619 0.000
Age 0.492 0.001
former had significantly greater hip strength than
Ankle eversion RTD 0.490 0.001
the latter (Fig. 2a), suggesting that hip strength was Hip abduction RTD 0.481 0.001
able to compensate for less precise ankle proprio- BMI 0.392 0.009
ception. When the same analysis was performed for Ankle eversion MVC 0.351 0.018
ankle proprioceptive thresholds, subjects with Ankle inversion MVC 0.350 0.018
greater UST had significantly more precise All values calculated based on the 36 subjects who had valid results for
(smaller) proprioceptive thresholds (Fig. 2b). all variables.

Sensorimotor Function and Balance MUSCLE & NERVE April 2012 581
 Table 3. Regression model.
95% CI bound
2
Model R R Dependent variable US* Lower Upper t P
1 0.676 0.456 Hip MVC* 460.945 290.881 631.010 5.497 0.000
2 0.834 0.696 Hip MVC* 386.485 254.224 518.745 5.932 0.000
Ankle proprioception 4.179 5.794 2.564 5.254 0.000
3 0.856 0.733 Hip MVC* 339.517 205.974 473.060 5.167 0.000
Ankle proprioception 3.867 5.433 0.300 5.017 0.000
Age 0.265 0.515 .015 2.156 0.038

*US ¼ unstandardized coefficients.

dorsiflexors),36 whereas mediolateral balance is with previous work5 in which ankle inversion/ever-
controlled via frontal plane motion at the hip.36 sion proprioceptive thresholds explained approxi-
Other studies demonstrated significant correlations mately half the variance in UST (R2 ¼ 0.514) in
between hip abduction RTD and performance of
reactive and voluntary frontal plane balance in
older adults.37 A study of slips noted that older
persons used frontal plane mechanisms for recov-
ery, whereas young subjects did not.38 One way to
interpret the importance of abductor and adductor
muscles with regard to unipedal stance is to sug-
gest that a co-contraction of these muscles allows a
transient, voluntary increase in hip rotational stiff-
ness. Given that an inverted pendulum is a com-
monly used model for human standing balance,
this stiffness creates a longer pendulum, which
requires more time to fall than a shorter pendu-
lum. As a result, there is more time available for
postural adjustments, which renders the task of
one-legged balance less challenging.39 However,
once balance is disturbed, it is likely that the avail-
ability of a rapid rate of strength development
would be more important, given that balance resto-
ration occurs within fractions of a second.34
The independent contribution of ankle pro-
prioception for balancing on one leg is consistent

FIGURE 2. Hip strength and ankle proprioception in patients with

shorter and longer, respectively, USTs than anticipated. A com-
parison of (a) hip strength and (b) ankle proprioceptive thresholds
in subjects who demonstrated shorter (left) and longer (right)
USTs than would be anticipated based on their ankle propriocep-
tive threshold and age. Hip strength was calculated as the mean
of the mean peak abduction and adduction maximal voluntary
FIGURE 1. Scatterplots illustrating the relationship of the hip contractions (Nm/Nm). UST, unipedal stance time. Propriocep-
strength/proprioception ratio to UST. The equation for the curvi- tive threshold ¼ smallest rotational displacement of the ankle that
linear regression is: y ¼ 0.0098e0.067x. a subject could reliably detect with 100% accuracy.

582 Sensorimotor Function and Balance MUSCLE & NERVE April 2012
older subjects with a range of peripheral neurolog- pensate for distal sensory impairment at the ankle.
ical function. More precise ankle proprioceptive Given the fact that the majority of polyneuropa-
thresholds may reduce the lateral distance that the thies are distal, this strategy can be used in a large
center of mass (COM) can travel prior to detec- proportion of patients with lower limb neuromus-
tion. Early detection of a displaced COM would cular disease. Conversely, persons with PN and
then require only moderate strength that a major- proximal weakness that cannot be improved may
ity of older persons likely possess. In contrast, less be best served by an assistive device, appropriate
precise ankle proprioception would require greater upper limb strengthening, environmental modifi-
intensity of motor function for appropriate reposi- cation, and instruction.44,45 Finally, it should be
tioning of the COM. Supporting this explanation, noted that diminished UST need not be viewed as
healthy subjects demonstrate increased center-of- a natural consequence of aging, despite research
pressure velocities when the plantar aspect of the that has noted an inverse association between the
foot is anesthetized, which is consistent with the two and even one study suggesting age-adjusted
greater motor function requirement.40 norms for UST.27,46 Instead, a decreased UST
Ankle motor function did not show a signifi- should, in the absence of an obvious musculoskel-
cant independent influence on UST, despite the etal and/or central neurological disorder, be con-
fact that ankle inversion and eversion rates of tor- sidered a function of diminished lower limb neuro-
que generation explained approximately 40% and muscular competence.
25%, respectively, of its variance. These findings A recent study47 found that improvements in
are consistent with those of Gutierrez et al.,17 who trunk extension endurance, but not lower limb
found that ankle inversion RTD explained over strength or power, were independently associated
50% of the variance in UST (R2 ¼ 0.575). In con- with clinically meaningful change in balance in
trast, ankle maximum isometric inversion and ever- older adults. However, that protocol measured
sion strengths each explained only 12% of UST. lower limb strength by means of a seated double-
When observing subjects successfully balancing on leg press maneuver, and thus sagittal plane
one foot there are rapid postural adjustments in strength of multiple muscle groups within the
ankle inversion and eversion as the center of pres- lower limbs was simultaneously measured. This
sure is quickly manipulated to control the move- technique contrasts with that of our study, which
ments of the whole-body COM. The rapid speed measured frontal plane sensorimotor functions dis-
with which these changes occur in the subject who cretely at the hip and ankle. Therefore, although
can reliably stand on one foot is consistent with trunk extension endurance may be more impor-
ankle maximum RTD being an important motor tant to balance than to sagittal plane lower limb
function for the maintenance of unipedal stance. strength, the relative importance of trunk endur-
These findings are in line with other studies show- ance and lower limb frontal plane sensorimotor
ing that the ability of the lower limbs to create function with reference to balance has yet to be
force quickly is of greater importance than the explored.
total force a muscle group can generate.41,42 The strengths of this study include the fact that
Although highly correlated with UST, ankle RTD sensory and motor control mechanisms were quan-
had no independent influence on UST in the pres- tified simultaneously in subjects with a spectrum of
ence of ankle proprioception and hip strength. neuromuscular dysfunction. The correlations and
This is of clinical interest, given the challenge of multiple regression analyses were unusually strong.
strengthening distal musculature in PN subjects. Given the complexity of any human behavior it is
Given the established relationships between a remarkable that just two lower limb neuromuscular
diminished UST and frailty, activity level, and falls, characteristics explain nearly 75% of UST. Limita-
strategies to increase UST have clinical relevance. tions include the fact that UST is unlikely to per-
There is no clear evidence that ankle propriocep- fectly reflect a variety of relevant mobility charac-
tive thresholds can be improved by therapeutic teristics, such as gait speed and the ability to
exercise,43 and recent work has shown that an recover from a perturbation while walking. The
ankle orthosis, which decreased the temporal and lower limb sensorimotor function(s) responsible
spatial variability of neuropathic gait on an irregu- for these deserves further attention. In addition,
lar surface, did not improve ankle proprioceptive only frontal plane neuromuscular functions were
thresholds.35 Given these findings, frontal plane evaluated. It is possible that sagittal plane muscle
hip strengthening appears the best strategy for strength also influences UST. It should also be
improving UST. This strengthening should be pur- noted that the measurements of ankle motor func-
sued most aggressively in those with decreased dis- tion assumed the ankle’s center of rotation to be
tal afferent neurological function, as it appears midway between the malleoli. This was an estima-
that increased frontal plane hip strength can com- tion and therefore represents a study limitation,
Sensorimotor Function and Balance MUSCLE & NERVE April 2012 583
but an important one to note given that ankle 12. MacKinnon CD, Winter DA. Control of whole body balance in the
frontal plane during human walking. J Biomech 1993;26:633–644.
motor function was not identified as an independ- 13. Bauby CE, Kuo AD. Active control of lateral balance in human walk-
ent predictor of UST. It is possible that evaluation ing. J Biomech 2000;33:1433–1440.
14. Chou L, Kaufman K, Brey R. Correlation between muscle strength
of ankle motor function by another means, such as and balance control when negotiating obstacles. Presented at the
an open chain technique, would have led to an al- 46th annual meeting of the Orthopaedic Research Society, Orlando,
Florida; 2000; p 315.
ternative conclusion. In the past, however, we used 15. Hilliard MJ, Martinez KM, Janssen I, Edwards B, Mille ML, Zhang Y,
the closed chain technique, and its validity is sup- et al. Lateral balance factors predict future falls in community-living
older adults. Arch Phys Med Rehabil 2008;89:1708–1713.
ported by the relationship between ankle strength 16. Mille ML, Johnson ME, Martinez KM, Rogers MW. Age-dependent
determined in this fashion to the presence of neu- differences in lateral balance recovery through protective stepping.
Clin Biomech 2005;20:607–616.
ropathy and unipedal stance time.17 Due to techni- 17. Gutierrez EM, Helber MD, Dealva D, Ashton-Miller JA, Richardson
cal difficulties in the early stages of the study, a JK. Mild diabetic neuropathy affects ankle motor function. Clin Bio-
mech (Bristol, Avon) 2001;16:522–528.
portion of the ankle motor data could not be ana- 18. Giacomozzi C, D’Ambrogi E, Cesinaro S, Macellari V, Uccioli L. Mus-
lyzed for 5 subjects, so the final regression model cle performance and ankle joint mobility in long-term patients with
diabetes. BMC Musculoskelet Disord 2008;9:99.
was performed on 36 subjects. Finally, UST was 19. Wang CY, Hsieh CL, Olson SL, Wang CH, Sheu CF, Liang CC. Psy-
cut-off at 30 seconds, likely creating a ceiling effect chometric properties of the Berg balance scale in a community-
for the most able subjects. dwelling elderly resident population in Taiwan. J Formos Med Assoc
2006;105:992–1000.
In conclusion, increased frontal plane hip 20. Hurvitz EA, Richardson JK, Werner RA, Ruhl AM, Dixon MR. Unipe-
strength and/or decreased (more precise) ankle dal stance testing as an indicator of fall risk among older outpa-
tients. Arch Phys Med Rehabil 2000;81:587–591.
proprioceptive thresholds strongly influenced UST. 21. Vellas BJ, Wayne SJ, Romero L, Baumgartner RN, Rubenstein LZ,
Age, in contrast, had a trivial influence when these Garry PJ. One-leg balance is an important predictor of injurious falls
in older persons. J Am Geriatr Soc 1997;45:735–738.
neuromuscular functions were taken into account. 22. Hurvitz EA, Richardson JK, Werner RA. Unipedal stance testing in
Frontal plane hip strength was the single best pre- the assessment of peripheral neuropathy. Arch Phys Med Rehabil
2001;82:198–204.
dictor of UST and appeared to compensate for less 23. Bulbulian R, Hargan ML. The effect of activity history and current
precise ankle proprioceptive thresholds. This find- activity on static and dynamic postural balance in older adults. Phys-
iol Behav 2000;70:319–325.
ing is clinically relevant, given the possibility of 24. Richardson JK. Factors associated with falls in older patients with dif-
strengthening the hip even in the setting of signifi- fuse polyneuropathy. J Am Geriatr Soc 2002;50:1767–1773.
25. Vellas BJ, Rubenstein LZ, Ousset PJ, Faisant C, Kostek V, Nourha-
cant PN. shemi F, et al. One-leg standing balance and functional status in a
population of 512 community-living elderly persons. Aging (Milano)
This work was supported by the National Institutes of Health 1997;95–98.
(ROI AG026569-01), PHS grants (P30AG024824), and the Swiss 26. Bohannon R. Single limb stance times. A descriptive meta-analysis of
National Foundation (IZKOZ3_133925). data from individuals at least 60 years of age. Topics Geriatric Reha-
bil 2006;22:70–77.
27. Bohannon RW, Larkin PA, Cook AC, Gear J, Singer J. Decrease in
timed balance test scores with aging. Phys Ther 1984;64:1067–1070.
REFERENCES 28. Potvin AR, Syndulko K, Tourtellotte WW, Lemmon JA, Potvin JH.
Human neurologic function and the aging process. J Am Geriatr Soc
1. Narici MV, Maganaris C, Reeves N. Myotendinous alterations and 1980;28:1–9.
effects of resistive loading in old age. Scand J Med Sci Sports 2005; 29. Richardson JK. The clinical identification of peripheral neuropathy
15:392–401. among older persons. Arch Phys Med Rehabil 2002;83:1553–1558.
2. Barry BK, Carson RG. The consequences of resistance training for 30. Feldman EL, Stevens MJ, Thomas PK, Brown MB, Canal N, Greene
movement control in older adults. J Gerontol A Biol Sci Med Sci DA. A practical two-step quantitative clinical and electrophysiological
2004;59:730–754. assessment for the diagnosis and staging of diabetic neuropathy. Dia-
3. Kanade RV, van Deursen RW, Harding KG, Price PE. Investigation betes Care 1994;17:1281–1289.
of standing balance in patients with diabetic neuropathy at different 31. Richardson JK, Tang C, Nwagwu C, Nnodim J. The influence of ini-
stages of foot complications. Clin Biomech 2008;23:1183–1191. tial bipedal stance width on the clinical measurement of unipedal
4. Turcot K, Allet L, Golay A, Hoffmeyer P, Armand S. Investigation of balance time. PM R 2010;2:254–258.
standing balance in diabetic patients with and without peripheral 32. Smeesters C, Ashton-Miller JA, Cole NM. Effect of age and gender
neuropathy using accelerometers. Clin Biomech 2009;24:716–721. on maximum hip flexor power. Vancouver, BC: International Associ-
5. Son J, Ashton-Miller JA, Richardson JK. Frontal plane ankle proprio- ation of Gerontology; 2001.
ceptive thresholds and unipedal balance. Muscle Nerve 2009;39: 33. Ylinen J, Salo P, Nykanen M, Kautiainen H, Hakkinen A. Decreased
150–157. isometric neck strength in women with chronic neck pain and the
repeatability of neck strength measurements. Arch Phys Med Rehabil
6. Richardson JK, Ashton-Miller JA, Lee SG, Jacobs K. Moderate periph-
2004;85:1303–1308.
eral neuropathy impairs weight transfer and unipedal balance in the
elderly. Arch Phys Med Rehabil 1996;77:1152–1156. 34. Thelen DG, Schultz AB, Alexander NB, Ashton-Miller JA. Effects of
age on rapid ankle torque development. J Gerontol A Biol Sci Med
7. Allet L, Armand S, Golay A, Monnin D, de Bie RA, de Bruin ED.
Sci 1996;51:M226–232.
Gait characteristics of diabetic patients: a systematic review. Diabetes
35. Son J, Ashton-Miller JA, Richardson JK. Do ankle orthoses improve
Metab Res Rev 2008;24:173–191.
ankle proprioceptive thresholds or unipedal balance in older per-
8. Wallace C, Reiber GE, LeMaster J, Smith DG, Sullivan K, Hayes S, sons with peripheral neuropathy? Am J Phys Med Rehabil 2010;89:
et al. Incidence of falls, risk factors for falls, and fall-related fractures 369–375.
in individuals with diabetes and a prior foot ulcer. Diabetes Care 36. Winter DA, Patla AE, Prince F, Ishac M, Gielo-Perczak K. Stiffness
2002;25:1983–1986. control of balance in quiet standing. J Neurophysiol 1998;80:
9. Maurer MS, Burcham J, Cheng H. Diabetes mellitus is associated 1211–1221.
with an increased risk of falls in elderly residents of a long-term care 37. Chang SH, Mercer VS, Giuliani CA, Sloane PD. Relationship
facility. J Gerontol A Biol Sci Med Sci 2005;60:1157–1162. between hip abductor rate of force development and mediolateral
10. Maki BE, Holliday PJ, Topper AK. A prospective study of postural stability in older adults. Arch Phys Med Rehabil 2005;86:1843–1850.
balance and risk of falling in an ambulatory and independent elderly 38. Liu J, Lockhart TE. Age-related joint moment characteristics during
population. J Gerontol 1994;49:M72–84. normal gait and successful reactive-recovery from unexpected slip
11. Cummings SR, Nevitt MC. Non-skeletal determinants of fractures: perturbations. Gait Posture 2009;30:276–281.
the potential importance of the mechanics of falls. Osteoporos Int 39. Reeves NP, Narendra KS, Cholewicki J. Spine stability: lessons from
1994;4(suppl 1):67–70. balancing a stick. Clin Biomech 2011;26:325–330.

584 Sensorimotor Function and Balance MUSCLE & NERVE April 2012
40. Meyer PF, Oddsson LI, De Luca CJ. The role of plantar cutaneous results from a challenging unipedal balance test. Arch Phys Med
sensation in unperturbed stance. Exp Brain Res 2004;156:505–512. Rehabil 1996;77:446–452.
41. Bean JF, Kiely DK, Herman S, Leveille SG, Mizer K, Frontera WR, 45. Richardson JK, Thies SB, DeMott TK, Ashton-Miller JA. Interventions
et al. The relationship between leg power and physical performance improve gait regularity in patients with peripheral neuropathy while
in mobility-limited older people. J Am Geriatr Soc 2002;50:461–467. walking on an irregular surface under low light. J Am Geriatr Soc
42. Bento PC, Pereira G, Ugrinowitsch C, Rodacki AL. Peak torque and 2004;52:510–515.
rate of torque development in elderly with and without fall history. 46. Springer BA, Marin R, Cyhan T, Roberts H, Gill NW. Normative val-
Clin Biomech 2010;25:450–454. ues for the unipedal stance test with eyes open and closed. J Geriatr
43. Ashton-Miller JA, Wojtys EM, Huston LJ, Fry-Welch D. Can proprio- Phys Ther 2007;30:8–15.
ception really be improved by exercises? Knee Surg Sports Trauma- 47. Suri P, Kiely DK, Leveille SG, Frontera WR, Bean JF. Increased trunk
tol Arthrosc 2001;9:128–136. extension endurance is associated with meaningful improvement in
44. Ashton-Miller JA, Yeh MW, Richardson JK, Galloway T. A cane balance among older adults with mobility problems. Arch Phys Med
reduces loss of balance in patients with peripheral neuropathy: Rehabil 2011;92:1038–1043.

Sensorimotor Function and Balance MUSCLE & NERVE April 2012 585