Differences in lower-extremity muscular activation during walking between healthy older and young adults.

Abstract

Previous studies have identified differences in gait kinetics between healthy older and young adults. However, the underlying factors that cause these changes are not well understood. The objective of this study was to assess the effects of age and speed on the activation of lower-extremity muscles during human walking. We recorded electromyography (EMG) signals of the soleus, gastrocnemius, biceps femoris, medial hamstrings, tibialis anterior, vastus lateralis, and rectus femoris as healthy young and older adults walked over ground at slow, preferred and fast walking speeds. Nineteen healthy older adults (age, 73+/-5 years) and 18 healthy young adults (age, 26+/-3 years) participated. Rectified EMG signals were normalized to mean activities over a gait cycle at the preferred speed, allowing for an assessment of how the activity was distributed over the gait cycle and modulated with speed. Compared to the young adults, the older adults exhibited greater activation of the tibialis anterior and soleus during mid-stance at all walking speeds and greater activation of the vastus lateralis and medial hamstrings during loading and mid-stance at the fast walking speed, suggesting increased coactivation across the ankle and knee. In addition, older adults depend less on soleus muscle activation to push off at faster walking speeds. We conclude that age-related changes in neuromuscular activity reflect a strategy of stiffening the limb during single support and likely contribute to reduced push off power at fast walking speeds.

Full text

Page 1

Differences in lower-extremity muscular activation during walking
between healthy older and young adults
Anne Schmitza, Amy Sildera, Bryan Heiderscheita,b, Jane Mahoneyc, Darryl G. Thelena,b,d,*
aDepartment of Biomedical Engineering, University of Wisconsin-Madison, 3039 Mechanical Engineering Building, 1513 University Ave., Madison, WI 53706, USA
bDepartment of Orthopedics and Rehabilitation, University of Wisconsin-Madison, 3039 Mechanical Engineering Building, 1513 University Ave., Madison, WI 53706, USA
cDepartment of Medicine, University of Wisconsin-Madison, 2870 University Ave., Suite 106, Madison, WI 53705, USA
dDepartment of Mechanical Engineering, University of Wisconsin-Madison, 3039 Mechanical Engineering Building, 1513 University Ave., Madison, WI 53706, USA
a r t i c l e i n f o
Article history:
Received 16 May 2008
Received in revised form 5 September 2008
Accepted 7 October 2008
Keywords:
Aging
Gait
Speed
Lower extremity
Muscles
EMG
a b s t r a c t
Previous studies have identified differences in gait kinetics between healthy older and young adults.
However, the underlying factors that cause these changes are not well understood. The objective of this
study was to assess the effects of age and speed on the activation of lower-extremity muscles during
human walking. We recorded electromyography (EMG) signals of the soleus, gastrocnemius, biceps
femoris, medial hamstrings, tibialis anterior, vastus lateralis, and rectus femoris as healthy young and
older adults walked over ground at slow, preferred and fast walking speeds. Nineteen healthy older
adults (age, 73 ± 5 years) and 18 healthy young adults (age, 26 ± 3 years) participated. Rectified EMG sig-
nals were normalized to mean activities over a gait cycle at the preferred speed, allowing for an assess-
ment of how the activity was distributed over the gait cycle and modulated with speed. Compared to the
young adults, the older adults exhibited greater activation of the tibialis anterior and soleus during mid-
stance at all walking speeds and greater activation of the vastus lateralis and medial hamstrings during
loading and mid-stance at the fast walking speed, suggesting increased coactivation across the ankle and
knee. In addition, older adults depend less on soleus muscle activation to push off at faster walking
speeds. We conclude that age-related changes in neuromuscular activity reflect a strategy of stiffening
the limb during single support and likely contribute to reduced push off power at fast walking speeds.
? 2008 Elsevier Ltd. All rights reserved.
1. Introduction
Joint kinetic measures have proven effective in distinguishing
the changes in gait mechanics associated with aging. Notably, older
adults exhibit decreased peak ankle plantar flexor power during
push off (McGibbon and Krebs, 2004; Winter, 1991; Winter et al.,
1990), accompanied by either increased peak hip extensor power
during early-stance (McGibbon and Krebs, 2004) and/or increased
peak hip flexor power generation during late-stance (McGibbon
and Krebs, 2004; Judge et al., 1996). This distal to proximal shift
in power production (DeVita and Hortobagyi, 2000) exists even
among active, healthy older adults, and seems to be more pro-
nounced at faster walking speeds (Silder et al., 2008).
While age-related changes in joint kinetics are well docu-
mented, the underlying factors that drive these changes are not
well understood. In particular, it is unclear whether biomechanical
changes in muscle or adaptations in neuromuscular activity play a
more prominent role. For example, it is feasible that muscle activa-
tion patterns are unchanged with age and that the changes in joint
kinetics are simply a result of age-related muscle remodeling. Sar-
copenia is a well documented effect of aging (Frontera et al., 2000,
1991), and a decrease in muscle strength and power is generally
associated with this loss of muscle mass (Frontera et al., 2000,
1991; Visser et al., 2002). Such structural changes alone could
diminish power output when the plantar flexors contract during
the push off phase of walking. Sarcopenia is also characterized by
an overall decrease in the number of muscle fibers and cross-sec-
tional area due to fatty and connective tissue replacement of the
muscle fibers (Huang et al., 1999; Lexell, 1995; Luff, 1998; Proctor
et al., 1998). This non-contractile tissue replacement may cause
increased joint stiffness. Since hip flexor tightness is known to con-
tribute substantially to the hip power generated during pre-swing
(Whittington et al., 2008), it is feasible that an increase in passive
hip joint stiffness could allow for enhanced energy storage in the
passive hip flexors during mid to late stance. The release of this
energy during pre-swing could then contribute to the increased
hip flexor power output observed experimentally in older adult
gait.
1050-6411/$ - see front matter ? 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jelekin.2008.10.008
* Corresponding author. Address: Department of Mechanical Engineering, Uni-
versity of Wisconsin-Madison, 3039 Mechanical Engineering Building, 1513 Uni-
versity Ave., Madison, WI 53706, USA. Tel.: +1 608 262 1902; fax: +1 608 265 2316.
E-mail address: thelen@engr.wisc.edu (D.G. Thelen).
Journal of Electromyography and Kinesiology 19 (2009) 1085–1091
Contents lists available at ScienceDirect
Journal of Electromyography and Kinesiology
journal homepage: www.elsevier.com/locate/jelekin

Page 2

Alternatively, the changes in gait mechanics may arise via neu-
rally mediated changes in muscle excitation patterns. It has previ-
ously been shown that older adults exhibit increased coactivation
of knee muscles during walking and that this may contribute to in-
creased metabolic costs for locomotion (Mian et al., 2006). In-
creased levels of cocontraction have also been observed about
the ankle in older adults during isometric exertions (Patten and Ka-
men, 2000) and challenging postural tasks (Benjuya et al., 2004;
Manchester et al., 1989). However, it is not known whether similar
ankle muscle cocontraction strategies are employed about the an-
kle during the single support phase of walking. If present, this
could increase active ankle joint stiffness while contributing to a
decrease in net power output about the ankle. In this scenario,
the greater hip power output in older adults may reflect a compen-
satory mechanism, whereby an increased neural drive of the hip
muscles is used to compensate for decreased power output at
the ankle (McGibbon, 2003).
An assessment of muscle activities in the lower limb muscles is
important to gain additional insights into the relative influence of
biomechanical and neural factors on older adult gait. In this study,
we investigated the modulation of lower-extremity electromyog-
raphy (EMG) signals with walking speed in healthy young and
healthy older adults. We hypothesized that older adults would ex-
hibit increased activation of ankle dorsi- and plantar flexors during
the single support phase of gait, consistent with ankle muscle coac-
tivation that is observed during challenging postural tasks. We also
hypothesized that the distal to proximal shift in power output
would be reflected by changes in neuromuscular activity at faster
walking speeds, such that older adults would exhibit decreased
plantar flexor activity during push off accompanied by increased
hip muscle activity during early and late stance.
2. Materials and methods
Nineteen healthy older adults (7 males, 12 females; age, 73 ± 5
years) and 18 healthy young adults (8 males, 10 females; age,
26 ± 3 years) participated in this study (Table 1). Subjects were ex-
cluded based on current or history of orthopedic diagnosis, muscu-
loskeletal trauma, persistent joint pain, and any known cardiac,
neural, gait impairment, or balance issues. Older adults were
screened by a geriatrician to determine additional exclusion crite-
ria based on cognitive impairment (score <24 on the Mini-Mental
State Exam, MMSE) (Folstein et al., 1975), plantar sensory impair-
ment (unable to perceive 5.07 Semmes–Weinstein (10-g) monofil-
ament), or loss of vibratory sensation at the great toe (unable to
perceive vibration at the great metatarsal–phalangeal joint with
a 128 Hz tuning fork). Physical activity levels of the older adults
were assessed using the CHAMPS questionnaire (Stewart et al.,
2001). The Dynamic Gait Index was used as a clinical assessment
of gait and dynamic balance with a score below 19 (maximum pos-
sible score of 24) correlated to an increased risk of falling (Whitney
et al., 2000). Each subject gave informed consent in agreement
with a protocol approved by the University of Wisconsin’s Health
Sciences Institutional Review Board.
Subjects were first asked to perform repeated walking trials at
their preferred speed over a 12 m walkway. Sacral marker kine-
matics were recorded in these trials and used to determine the
average preferred speed over a gait cycle. Subjects then performed
a total of fifteen walking trials, five trials each at 80% (slow), 100%
(preferred), and 120% (fast) of their preferred speed, with the
ordering of speed randomized. Trials were accepted if the mea-
sured speed, as determined from the sacral marker kinematics,
was within 5% of the target speed. If not, subjects were asked to re-
peat the trial and instructed whether to slow down or speed up rel-
ative to the most recent trial.
Whole body motion was tracked using 42 motion capture mark-
ers, 23 of which were placed on palpable anatomical landmarks.
The additional 19 markers were used to enhance segment tracking
and reduce skin motion artifact (Cappozzo et al., 1995). Kinematic
data were collected at 100 Hz using an eight-camera passive mo-
tion capture system (Motion Analysis Corporation, Santa Rosa,
CA) and processed with motion capture software (EVaRT v5.0).
Ground reaction forces were synchronously recorded at 2000 Hz
for two successive foot strikes using three force plates (Model
BP400600, AMTI, Watertown, MA) imbedded in the middle of the
walkway. EMG signals for the soleus, gastrocnemius, biceps femo-
ris, medial hamstrings, tibialis anterior, vastus lateralis, and rectus
femoris were also synchronously recorded at 2000 Hz using pre-
amplified single differential electrodes with 10 mm inter-electrode
distance (DE-2.1, DelSys, Inc, Boston, MA). The electrodes were
coated with conducting gel prior to application and interfaced with
an amplifier/processor unit (CMRR > 85 dB at 60 Hz; input imped-
ance > 100 MX). The electrode locations were determined by the
same investigator for each subject using standard EMG electrode
locations that placed the electrodes at the center of the muscle bel-
ly (Basmajian, 1989).
The EMG signals were processed using MATLAB (MATLAB
R2006a, MathWorks, Inc., Natick, MA). The data was first passed
through a 10–500 Hz sixth order Butterworth bandpass filter and
full wave rectified (Konrad, 2005). To create a linear envelope,
the data was subsequently passed through a sixth order Butter-
worth low-pass filter with a 6 Hz cutoff frequency. Heel contact
times were determined from the vertical ground reaction force
traces using a 10 N threshold. Rectified EMG data were interpo-
lated to obtain a data series of 1001 points over the gait cycle, from
heel contact to the subsequent heel contact of the same limb. The
EMG signals for each subject were normalized to the mean signal
for each muscle over the entire gait cycle of that subject’s preferred
speed walking trials.
For each of the seven muscles recorded, an average EMG trace
for each subject at each speed was obtained by ensemble averaging
across all five trials. We then computed the average normalized
EMG activity within selected phases of the gait cycle (Perry,
1992): loading (0–10% of the gait cycle), mid-stance (10–30%), ter-
minal stance and pre-swing (30–60%), initial swing (60–73%), and
terminal swing (87–100%). For each muscle, average EMG activities
were computed for the phases needed to test our hypotheses. To as-
sess the potential presence of cocontraction during stance, we
quantified and compared the average muscle activities from all
musclesduring loading, mid-stanceand terminal stance/pre-swing.
To assess the potential role that neural factors play in the distal to
proximalshift in powerproduction,we also quantified rectus femo-
ris activity during initial swing and the biceps femoris and medial
hamstring EMG activities during terminal swing. A repeated mea-
sures ANOVA was used to assess the effects of age (young subjects,
older subjects) and speed (slow, preferred, and fast) on the normal-
ized muscle activities in these phases (STATISTICA 6, StatSoft, Inc.,
Tulsa, OK). Post hoc analyses were subsequently performed using
Tukey’s HSD test to evaluate significant speed–age interactions.
Statistical significance was defined at p < 0.05.
Table 1
Average (SD) characteristics of the adults who participated in the study.
Young adults (n = 18) Older adults (n = 19)
8 Men 10 Women 7 Men12 Women
Age (years)
Body mass (kg)
Height (m)
27 (4)
81 (7)
1.84 (0.09)
25 (2)
61 (7)
1.66 (0.06)
73 (4)
74 (8)
1.75 (0.07)
72 (6)
67 (12)
1.66 (0.09)
1086
A. Schmitz et al./Journal of Electromyography and Kinesiology 19 (2009) 1085–1091

Page 3

3. Results
There were no significant differences between the young and
older adults with respect to preferred walking speed, cadence, or
the percent of the gait cycle spent in swing (Table 2). The older
adults tested in this study (CHAMPS score = 11,772) were substan-
tially more active than typical older adults (CHAMPS score = 3386,
Stewart et al., 2001). In addition, all older adults displayed normal
gait and dynamic balance (Dynamic Gait Index score = 23.8 (0.7)
Whitney et al., 2000).
Table 3 summarizes the mean EMG values and significant fac-
tors for each gait phase tested, and Fig. 1 displays the EMG curves
for muscles that showed significant differences between older and
young adults.
3.1. Loading phase
Tibialis anterior, soleus, biceps femoris, and rectus femoris
activities showed a speed effect, with activity increasing as speed
increased. The older adults used significantly more gastrocnemius
activity than the young adults across all three speeds (p = 0.013),
with activity significantly increasing for both age groups as speed
increased. The vastus lateralis showed a significant speed–age
interaction (Fig. 2). Tukey’s post hoc test revealed age was only a
factor at the faster walking speed, with the older adults displaying
22% less activity than the younger adults. The medial hamstrings
also showed a speed–age interaction (Fig. 2). Age was only signif-
icant at the faster walking speed, where the older adults showed
36% more activity than the young adults.
3.2. Mid-stance
Gastrocnemius and biceps femoris activities showed significant
speed effects, whereby activity increased as speed increased. With
the older adults showing significantly more activity across all three
speeds, soleus (p = 0.001), vastus lateralis (p = 0.001), and rectus
femoris (p = 0.006) showed significant age effects, along with a
speed effect where activity increased as speed also increased. The
tibialis anterior only showed an age effect where the older adults
exhibited more activity than the younger adults across all three
speeds (p = 0.000). Medial hamstrings activity showed a speed–
age interaction (Fig. 2). Tukey’s post hoc analysis showed age
was only significant at the faster walking speed, with older adults
displaying 52% more medial hamstring activity than the young
adults.
3.3. Terminal stance and pre-swing
Speed was the only significant effect in tibialis anterior, gastroc-
nemius, biceps femoris, vastus lateralis, medial hamstring, and rec-
tus femoris activities. These activities significantly increased as
speed increased. Soleus activity exhibited a speed–age interaction
(Fig. 2). At the fast walking speed, older adults used 20% less soleus
activity than the young adults.
Table 2
Mean (SD) walking speed, cadence and percent of gait cycle spent in swing phase for
both groups at each of the walking speeds tested.
Young adultsOlder adults
Slow Preferred FastSlow PreferredFast
Walking speed
(m/s)
Cadence
(steps/min)
Swing phase
(% gait cycle)
1.07
(0.10)
98 (11)
1.33
(0.13)
112 (10)
1.60
(0.13)
122 (10)
1.06
(0.10)
100 (6)
1.32
(0.13)
115 (7)
1.58
(0.16)
125 (9)
32 (1) 34 (1)35 (1) 32 (2)34 (2) 35 (2)
Table 3
Mean (SD) normalized electromyography activities for each group during select phases of the gait cycle. Each muscle’s activity level was normalized to the mean rectified activity
it exhibited at the preferred walking speed. A two-way, repeated measures ANOVA was used to determine the influence of age and walking speed on the muscle activities
observed.
Phase of gait cycle MuscleSlowPreferredFast
p Values
Young adultsOlder adultsYoung adultsOlder adults Young adults Older adultsSpeedAgeSpeed?Age
0.801
0.553
0.514
0.416
0.011
0.009
0.351
Loading response (0–10%)TA
SOL
GAS
BF
VL
MH
RF
2.14 (0.40)
0.48 (0.15)
0.29 (0.11)
1.70 (0.77)
2.06 (0.47)
2.11 (0.81)
1.31 (0.25)
1.97 (0.31)
0.44 (0.15)
0.41 (0.18)
1.81 (0.65)
2.15 (0.47)
1.97 (0.62)
1.43 (0.34)
2.61 (0.48)
0.52 (0.19)
0.32 (0.15)
1.73 (0.84)
2.90 (0.63)
1.66 (0.51)
1.69 (0.45)
2.30 (0.36)
0.57 (0.18)
0.47 (0.17)
2.12 (0.80)
2.86 (0.47)
1.90 (0.61)
1.93 (0.50)
2.82 (0.80)
0.73 (0.38)
0.51 (0.39)
1.89 (0.97)
3.96 (1.17)
1.46 (0.51)
2.11 (0.56)
2.54 (0.74)
0.80 (0.42)
0.77 (0.55)
2.35 (0.93)
3.24 (0.98)
1.98 (0.80)
2.16 (0.60)
0.000
0.000
0.000
0.038
0.000
0.006
0.000
0.068
0.697
0.013
0.166
0.234
0.263
0.328
Mid-stance (10–30%) TA
SOL
GAS
BF
VL
MH
RF
0.44 (0.22)
0.86 (0.26)
1.14 (0.49)
0.71 (0.34)
1.13 (0.26)
1.06 (0.56)
0.92 (0.19)
0.83 (0.41)
1.07 (0.23)
1.08 (0.36)
0.88 (0.56)
1.44 (0.40)
0.90 (0.51)
1.09 (0.26)
0.49 (0.21)
0.99 (0.28)
1.20 (0.41)
0.76 (0.36)
1.31 (0.25)
0.87 (0.35)
1.06 (0.20)
0.92 (0.35)
1.21 (0.34)
1.17 (0.45)
0.99 (0.59)
1.55 (0.30)
0.98 (0.59)
1.19 (0.26)
0.55 (0.29)
1.08 (0.35)
1.27 (0.53)
0.89 (0.49)
1.39 (0.31)
0.82 (0.34)
1.11 (0.36)
0.96 (0.41)
1.57 (0.47)
1.50 (0.62)
1.30 (0.69)
1.76 (0.49)
1.25 (0.61)
1.44 (0.39)
0.092
0.000
0.001
0.000
0.000
0.219
0.000
0.000
0.001
0.755
0.100
0.001
0.395
0.006
0.941
0.054
0.101
0.086
0.577
0.000
0.170
Terminal stance and pre-swing
(30–60%)
TA
SOL
GAS
BF
VL
MH
RF
0.29 (0.14)
1.71 (0.35)
1.96 (0.33)
0.50 (0.23)
0.47 (0.21)
0.49 (0.18)
0.65 (0.16)
0.32 (0.17)
1.69 (0.24)
1.69 (0.31)
0.48 (0.23)
0.54 (0.23)
0.42 (0.21)
0.64 (0.16)
0.39 (0.15)
2.13 (0.29)
2.06 (0.32)
0.61 (0.26)
0.55 (0.18)
0.49 (0.22)
0.83 (0.18)
0.38 (0.16)
1.98 (0.22)
1.96 (0.29)
0.45 (0.24)
0.55 (0.19)
0.39 (0.17)
0.83 (0.26)
0.58 (0.36)
2.67 (0.53)
2.31 (0.44)
0.84 (0.70)
0.66 (0.35)
0.55 (0.30)
1.20 (0.42)
0.57 (0.25)
2.23 (0.52)
2.22 (0.68)
0.58 (0.30)
0.76 (0.21)
0.52 (0.25)
1.23 (0.49)
0.000
0.000
0.000
0.002
0.000
0.007
0.000
0.992
0.048
0.172
0.137
0.352
0.338
0.983
0.846
0.007
0.381
0.180
0.438
0.507
0.949
Initial swing (60–73%)RF0.77 (0.27)0.62 (0.22)1.18 (0.33)1.05 (0.39)1.93 (0.88) 1.50 (0.66)
0.000
0.0840.168
Terminal swing (87–100%)BF
MH
1.90 (0.66)
2.14 (0.73)
1.79 (0.71)
2.06 (0.54)
2.63 (0.62)
2.95 (0.61)
2.46 (0.83)
2.73 (0.79)
3.06 (0.89)
3.40 (0.81)
3.28 (1.32)
3.27 (1.19)
0.000
0.000
0.928
0.541
0.194
0.822
Note: TA = tibialis anterior; SOL = soleus; GAS = gastrocnemius; BF = biceps femoris; VL = vastus lateralis; MH = medial hamstrings; RF = rectus femoris.
A. Schmitz et al./Journal of Electromyography and Kinesiology 19 (2009) 1085–1091
1087

Page 4

1
2
3
4
5

2
4
6
So leus
*
*
*
*
2
4
6
2
4
6
0 20 40 60 80 100
Older Adults
Young Adults
Gastrocnemius
Slow
Preferred
Fast
2
4
6
2
4
6
2
4
6
0 20 40
% Gait Cycle
60 80 100
Slow
SlowSlow
Slow
Preferred
Preferred
Preferred
Preferred
Fast
Fast
Fast
Fast
Rectus Femoris
0 20 40
% Gait Cycle
60 80 100
% Gait Cycle
% Gait Cycle
% Gait Cycle
1
2
3
4
5
1
2
3
4
5
Vastus Lateralis
2
4
6
2
4
6
2
4
6
0 20 40 60 80 100
*
*
*
*



Medial Hamstrings
0 20 40
% Gait Cycle
60 80 100
2
4
6
2
4
6
2
4
6
Slow
Preferred
Fast
*
*

Tibialis Anterior
2
4
6
2
4
6
2
4
6
0 20 40 60 80 100
*
*
*
*
*
*
*
*
*
Normalized EMG Signals
Normalized EMG Signals
Normalized EMG Signals
Normalized EMG Signals
Normalized EMG Signals
Normalized EMG Signals
Fig. 1. Ensemble averaged electromyographic activities for the young and older adults at the slow, preferred and fast walking speeds. The shaded portion represents plus and
minus one standard deviation of the young adults’ data. The horizontal lines above the curves indicate phases of the gait cycle where the average normalized activities over a
phase of the gait cycle differed significantly (?p < 0.05) between the age groups, as determined by Tukey’s post hoc tests. The phases of the gait cycle that were statistically
compared for each muscle are given in Table 3.
1088
A. Schmitz et al./Journal of Electromyography and Kinesiology 19 (2009) 1085–1091

Page 5

3.4. Initial swing
Only rectus femoris activity showed a speed effect, with activity
significantly increasing as speed increased.
3.5. Terminal swing
Biceps femoris and medial hamstring activity did not show any
age effects but did show speed effects, with activity levels increas-
ing with speed.
4. Discussion
We quantified lower-extremity EMG patterns in healthy young
and older adults at three walking speeds. Our results show older
adults adopt a neuromuscular activation pattern that seems to uti-
lize greater coactivation of muscles about the ankle at all speeds
andaboutthekneeatfastspeedsduringmid-stanceanddependless
on soleus muscle activation to push off at faster walking speeds. In-
creased levels of hamstring activity in the older adults were also
present during loading and mid-stance at faster walking speeds,
which may contribute to the increased hip extensor power during
early-stance commonly observed in older adult gait (DeVita and
Hortobagyi, 2000; McGibbon and Krebs, 2004; Silder et al., 2008).
The differences in muscle activity between groups observed in
this study cannot be attributed to gait speed differences, since both
young and older subjects exhibited similar walking speeds. Prior
studies have reported a slower walking speed in older adults of
?70 years of age (Kerrigan et al., 1998; McGibbon, 2003; McGib-
bon and Krebs, 1999; Winter, 1991). In this study, the older adults’
average preferred walking speed (1.32 m/s) was near the upper
end of the range measured for this age group (1.03–1.30 m/s, Judge
et al., 1996; Kang and Dingwell, 2008; Kerrigan et al., 1998; McGib-
bon and Krebs, 1999, 2004). The faster speeds seen in this study
may, in part, reflect the high activity levels of the older adults
tested (CHAMPS score = 11,772). Another recent cross-sectional
study also found that active healthy older adults exhibited similar
walking speeds to height and weight matched younger adults
(Kang and Dingwell, 2008).
We observed a greater prevalence of age effects on uniarticular
muscle activities (soleus, vastus lateralis), than on biarticular mus-
cles, that cross the same joint (gastrocnemius, rectus femoris). For
example, older adults exhibited comparable gastrocnemius activa-
tion to the young adults during push off at all speeds but showed
20% less soleus activity during push off at the faster walking speed.
This difference may relate to the different biomechanical roles that
these muscles play during normal gait. The results of a forward dy-
namic model suggest that the uniarticular soleus is more impor-
tant in generating forward propulsion of the trunk, while the
biarticular gastrocnemius plays a bigger role in initiating swing
limb motion (Neptune et al., 2001). A recent empirical study has
confirmed differential biomechanical function of the soleus and
gastrocnemius during normal walking (Stewart et al., 2007). Thus,
our results would suggest that in older adults, the gastrocnemius
retains its role in swing initiation. Since older adults show less so-
leus activity during push off at the fast walking speed but retain
similar walking speeds, the soleus seems to diminish its role in
providing forward propulsion. In addition to the changes at the an-
kle, older adults showed 22% less activity in the uniarticular knee
extensor (vastus lateralis) during the loading phase of fast walking
while the biarticular rectus femoris exhibited no change in activa-
tion during loading. It has previously been observed that there is a
0
1
2
3
4
5
6
SlowPreferred
Speed
Fast
Mean Normalized EMG Signals
Vastus Lateralis During Loading
Young
Old
Young
Old
Young
Old
Young
Old
0
0.5
1
1.5
2
2.5
3
3.5
SlowPreferred
Speed
Fast
Mean Normalized EMG Signals
Medial Hamstrings During Loading
SlowPreferred
Speed
Fast
SlowPreferred
Speed
Fast
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
Mean Normalized EMG Signals
Medial Hamstrings During Mid-Stance
0
0.5
1
1.5
2
2.5
3
3.5
Mean Normalized EMG Signals
Soleus DuringTerminal Stance and Pre-Swing
*
*
*
*
Fig. 2. Mean normalized electromyographic activities for the young and older
adults during phases of the gait cycle where muscle activity showed a significant
speed–age interaction. The asterisk (?) denote where age was a significant factor in
post-hoc tests.
A. Schmitz et al./Journal of Electromyography and Kinesiology 19 (2009) 1085–1091
1089

Page 6

slight reduction in the negative work done by the knee extensors
during loading (DeVita and Hortobagyi, 2000), which this study
would attribute to reduced vastus activation.
Prior studies have reported that older adults often exhibit in-
creased hip extensor power and/or increased hip flexor power
generation in conjunction with reduced ankle plantar flexor
power, particularly at faster walking speeds (DeVita and Hort-
obagyi, 2000; Judge et al., 1996; McGibbon, 2003; McGibbon
and Krebs, 2004; Winter et al., 1990). Similarly, we previously
showed that the older adults analyzed in this study generated
more peak hip extensor power and did significantly more positive
work by both the hip extensors and flexors, when compared to
the younger subjects (Silder et al., 2008). We did not record sur-
face EMG activity from the uniarticular hip flexors and extensors
and thus are unable to definitively assess how these muscles may
have contributed to the biomechanical differences observed. How-
ever, at the fast walking speed the older adults showed an average
of 36% more medial hamstring activity during loading and 52%
more activity during mid-stance than the young adults. These re-
sults suggest that the hamstrings may contribute to the increased
hip extensor power observed biomechanically and thus may serve
to compensate for reduced soleus power output during plantar
flexion (McGibbon, 2003). The older adults activated the rectus
femoris more than the young adults during mid-stance across
all three speeds but showed no difference in activity during termi-
nal stance and pre-swing. Therefore, an increase in hip flexor
work, if generated actively, likely must arise from modulation of
uniarticular hip flexors such as the iliacus and psoas muscles.
Cocontraction has been observed as a task independent strategy
that is believed to be used to stiffen the joint and enhance stability
(Benjuya et al., 2004; Hortobagyi and DeVita, 2006; Manchester
et al., 1989). Our results suggest that older adults may employ
greater coactivation of the uniarticular soleus and tibialis anterior
muscles during mid-stance at all three speeds, which may increase
ankle stiffness and accommodate potential balance concerns. In-
creased activity levels were also present in the vastus lateralis
(at all speeds) and hamstrings (fast speed) during mid-stance, sug-
gesting that the joint stiffening strategy may extend to the knee
when older adults increase their walking speed. However, it is
interesting to note that none of the older adults exhibited lower-
extremity sensory deficits as measured by the plantar monofila-
ment test or great toe vibratory sensation testing. This suggests
that sensory deficits are either not present or are not detectable
via standard clinical means. Alternatively, cortical mechanisms
may be responsible for the increased cocontraction commonly seen
in older adults, with the scaling of flexion, extension, and coactivity
commands in the cortex becoming more inaccurate with age
(Hortobagyi and DeVita, 2006). A potential side-effect of increased
cocontraction of opposing muscles would be higher metabolic
costs, as discussed by Mian et al. (2006).
In this study, we normalized EMG signals to average activities
measured at the preferred walking speed. Such a normalization
approach was beneficial in assessing how EMG activity was dis-
tributed across the gait cycle and how activity levels were modu-
lated with increasing and decreasing speed. However in using this
approach, we did not attain a measure of the activity level with
respect to maximum. Furthermore, while the young and older
adults walked at the same preferred speeds, it is possible that they
were operating at differing percentages of their aerobic capacities.
As a result, it is not possible to determine whether the reduced so-
leus activity in older adults during terminal stance and pre-swing
at the faster speed represented a saturation effect (i.e. near max-
imal activity) or alternatively represented a deliberate attempt
to rely less on the soleus to push off. However, in either case, it
is clear that reduced push off power must, in part, have a neural
basis and thus, does not strictly represent a reduction in strength
due to age-related sarcopenia (Frontera et al., 2000, 1991; Visser
et al., 2002).
In summary, age-related changes in the neuromuscular activa-
tion of walking are clearly present, even among active healthy old-
er adults. The changes in activity observed in this study would
actively induce the distal-to-proximal shift in power production
observed in the gait of older adults (DeVita and Hortobagyi,
2000) and would also act to enhance joint stiffness during single
limb support.
Acknowledgements
This publication was made possible by Grant Number AG24276
from the National Institutes of Health and a National Science Foun-
dation pre-doctoral fellowship (AS). We gratefully acknowledge
the contribution of Ben Whittington, MS.
References
Basmajian JVe. Biofeedback: principles and practice for clinicians. 3rd ed.; 1989.
Benjuya N, Melzer I, Kaplanski J. Aging-induced shifts from a reliance on sensory
input to muscle cocontraction during balanced standing. J Gerontol A Biol Sci
Med Sci 2004;59A(2):166–71.
Cappozzo A, Catani F, Croce UD, Leardini A. Position and orientation in space of
bones during movement: anatomical frame definition and determination. Clin
Biomech (Bristol, Avon) 1995;10(4):171–8.
DeVita P, Hortobagyi T. Age causes a redistribution of joint torques and powers
during gait. J Appl Physiol 2000;88(5):1804–11.
Folstein M, Folstein S, McHugh P. Mini-mental state: a practical method for grading
the cognitive state of patients for the clinician. J Psychiat Res 1975;12(3):
189–98.
Frontera WR, Hughes VA, Lutz KJ, Evans WJ. A cross-sectional study of muscle
strength and mass in 45- to 78-yr-old men and women. J Appl Physiol
1991;71(2):644–50.
Frontera WR, Hughes VA, Fielding RA, Fiatarone MA, Evans WJ, Roubenoff R. Aging
of skeletal muscle: a 12-yr longitudinal study. J Appl Physiol 2000;88(4):
1321–6.
Hortobagyi T, DeVita P. Mechanisms responsible for the age-associated increase in
coactivation of antagonist muscles. Exercise Sport Sci Rev 2006;34(1):29–35.
Huang RP, Rubin CT, McLeod KJ. Changes in postural muscle dynamics as a function
of age. J Gerontol 1999;54A(8):B352–7.
Judge JO, Davis R, Ounpuu S. Step length reductions in advanced age: the role of
ankle and hip kinetics. J Gerontol Med Sci 1996;51(6):M303–12.
Kang H, Dingwell J. Separating the effects of age and walking speed on gait
variability. Gait Posture 2008;27(4):572–7.
Kerrigan DC, Todd MK, Della Croce U, Lipsitz LA, Collins JJ. Biomechanical gait
alterations independent of speed in the healthy elderly: evidence for specific
limiting impairments. Arch Phys Med Rehab 1998;79(3):317–22.
KonradP.TheABC ofEMG:apractical
electromyography. Scottsdale, AZ; 2005.
Lexell J. Human aging, muscle mass, and fiber type composition. J Gerontol A Biol
Sci Med Sci 1995;50A(Spec):11–7.
Luff AR. Age-associated changes in the innervation of muscle fibers and changes in
the mechanical properties of motor unites. Ann NY Acad Sci 1998;854:92–101.
Manchester D, Woollacott MH, Zederbauer-Hylton N, Marin O. Visual, vestibular,
and somatosensory contributions to balance control in the older adult. J
Gerontol 1989;44(4):M118–27.
McGibbon CA. Toward a better understanding of gait changes with age and
disablement: neuromuscular adaptation. Exercise Sport Sci Rev 2003;31(2):
102–8.
McGibbon CA, Krebs DE. Effects of age and functional limitation on leg joint power
and work during stance phase of gait. J Rehab Res Dev 1999;36(3):173–82.
McGibbon CA, Krebs DE. Discriminating age and disability effects in locomotion:
neuromuscular adaptations in musculoskeletal pathology. J Appl Physiol
2004;96(1):149–60.
Mian OS, Thom JM, Ardigo LP, Narici MV, Minetti AE. Metabolic cost, mechanical
work, and efficiency during walking in young and older men. Acta Physiol (Oxf)
2006;186(2):127–39.
Neptune RR, Kautz SA, Zajac FE. Contributions of the individual ankle plantar flexors
to support, forward progression and swing initiation during walking. J Biomech
2001;34(11):1387–98.
Patten C, Kamen G. Adaptations in motor unit discharge activity with force control
training in young and older human adults. Eur J Appl Physiol 2000;83(2):
128–43.
Perry J. Gait analysis: normal and pathological function; 1992.
Proctor DN, Balagopal P, Nair KS. Age-related sarcopenia in humans is associated
with reduced synthetic rates of specific muscle proteins. J Nutr 1998;128(2):
351S–5S.
Silder A, Heiderscheit B, Thelen DG. Active and passive contributions to joint
kinetics during walking in older adults. J Biomech 2008;41(7):1520–7.
introduction tokinesiological
1090
A. Schmitz et al./Journal of Electromyography and Kinesiology 19 (2009) 1085–1091

Page 7

Stewart AL, Mills KM, King AC, Haskell WL, Gillis D, Ritter PL. CHAMPS physical
activity questionnaire for older adults: outcomes for interventions. Med Sci
Sports Exercise 2001;33(7):1126–41.
Stewart C, Postans N, Schwartz MH, Rozumalski A, Roberts A. An exploration of the
function of the triceps surae during normal gait using functional electrical
stimulation. Gait Posture 2007;26(4):482–8.
Visser M, Kritchevsky SB, Goodpaster BH, Newman AB, Nevitt M, Stamm E, et al. Leg
muscle mass and composition in relation to lower extremity performance in
men and women aged 70–79: the health, aging and body composition study. J
Am Geriatr Soc 2002;50(5):897–904.
Whitney SL, Hudak MT, Marchetti GF. The dynamic gait index relates to self-
reported fall history in individuals with vestibular dysfunction. J Vestibul Res
2000;10(2):99–105.
Whittington B, Silder A, Heiderscheit B, Thelen DG. The contribution of passive-
elastic mechanisms to lower extremity joint kinetics during human walking.
Gait Posture 2008;27(4):628–34.
Winter DA. The biomechanics and motor control of human gait: normal, elderly and
pathological. 2nd ed.; 1991.
Winter D, Patla A, Frank J, Walt S. Biomechanical walking pattern changes in the fit
and healthy elderly. Phys Ther 1990;70(6):340–7.
Anne Schmitz received the BS degree in Mechanical
Engineering at the University of Wisconsin-Madison
in 2006. Currently, she is a Master’s candidate in
BiomedicalEngineering
research interests include computational biome-
chanics, simulation of human gait, and finite element
analysis. Clinical applications include knee surgery
simulation and cerebral palsy gait analysis.
atUW-Madison.Her
Amy Silder received the BS degree (2003) in Bio-
systems Engineering from Michigan State University
and the MSE degree (2005) from the University of
Wisconsin-Madison, where she is currently a doc-
toral student in Biomedical Engineering. Her research
interests include imaging musculotendon mechanics
via magnetic resonance imaging, running and walk-
ingmechanics, and
rehabilitation.
injurymechanics and
Bryan Heiderscheit, PT, PhD is an Assistant Professor
in the Departments of Orthopedics & Rehabilitation
and Biomedical Engineering at the University of
Wisconsin-Madison. He
therapy training at the University of Wisconsin-La
Crosse (1994) and his doctorate in biomechanics
from the University of Massachusetts (2000). He is
co-director of the Neuromuscular Biomechanics
Laboratory at the University of Wisconsin-Madison
and director of the Runners’ Clinic through the Uni-
versity of Wisconsin Sports Medicine Clinic. His
research is aimed at understanding and enhancing
movement coordination as it relates to injury and
completed his physical
aging, with recent projects focused on the mechanisms of running-related injuries
and falls-risk detection in older adults. He has received research support from the
NIH, NFL Medical Charities and American Physical Therapy Association.
Jane Mahoney, MD, is an Associate Professor in the
Geriatics Section of the Department of Medicine at
the University of Wisconsin-Madison School of
Medicine and Public Health, and the Medical Director
of Care Wisconsin. She received her MD degree
(1986) from the University of California – San Fran-
cisco. Her internship and residency (1986–1989)
were done in Internal Medicine at the University of
Wisconsin Hospitals and Clinics in Madison, WI;
followed by a fellowship in Geriatrics (1989–1991) at
the William S. Middleton Memorial Veterans Hospi-
tal, also in Madison, WI. Her research interests
include epidemiologic, randomized trial, and dis-
semination research examining risk of falls in the elderly and interventions to
decrease falls and improve mobility. She has received funding from the NIH, CDC,
and American Physical Therapy Foundation.
Darryl G. Thelen received the BS degree in Mechan-
ical Engineering from Michigan State University, in
1987 and the MSE and PhD degrees in Mechanical
Engineering from the University of Michigan, Ann
Arbor, MI, in 1988 and 1992, respectively. He is
currently an Associate Professor in the Department of
Mechanical Engineering at the University of Wis-
consin-Madison. His research interests include com-
putational biomechanics, simulation and control of
human locomotion, and dynamic imaging of mus-
culotendon mechanics.
A. Schmitz et al./Journal of Electromyography and Kinesiology 19 (2009) 1085–1091
1091