Gait and Posture 10 (1999) 10 – 20

www.elsevier.com/locate/gaitpos

Quadriceps muscle strength and dynamic stability in elderly

persons
Donna Moxley Scarborough a,*, David E. Krebs a,b, Bette Ann Harris b
a
Biomotion Laboratory, Massachusetts General Hospital, Ruth Sleeper Hall, Room 010, 40 Parkman Street, Boston, MA 02114, USA
b
MGH Institute of Health Professions, Boston, MA, USA

Accepted 14 April 1999

Abstract

Several measures of dynamic stability during two functional activities correlated to quadriceps femoris muscle strength. A total
of 34 disabled elders (aged 60–88) living in the Boston area consented to maximum isometric quadriceps muscle strength testing,
chair rise and gait analysis. During chair rise, quadriceps strength significantly correlated with maximum upper body vertical
linear momentum, r=0.53, PB0.005, anterior posterior linear momentum, r= 0.38, PB 0.05, and the time to complete the chair
rise, r= −0.48, PB0.05, n= 29. Stride length and gait velocity correlated (r= 0.56, P B 0.001 and r= 0.51, P B 0.002, n=34)
with quadriceps muscle strength. The maximum range of whole body anteroposterior (A/P) linear momentum during gait also
correlated with quadriceps strength (r=0.47, P =0.004, n = 31). Dynamic stability during chair rise and gait, at preferred speed,
correlates directly with quadriceps femoris muscle strength in functionally limited elderly individuals. In our sample, elders
performed one of three movement strategies to arise from a chair, and quadriceps strength did not statistically differ between the
chair rise strategy groups. However, persons with the greatest quadriceps strength values were more stable regardless of which
chair rise strategy they performed. Our data indicate that clinicians should not suggest that patients use compensatory momentum
inducing locomotor strategies unless the patient has sufficient strength to control these induced forces. © 1999 Elsevier Science
B.V. All rights reserved.

Keywords: Quadriceps muscle strength; Dynamic stability; Chair rise; Strategies; Gait; Elderly

1. Introduction detailed analysis of dynamic stability, for example

Falls among the elderly are common and often result
in injuries requiring hospitalization. Falls occur most
frequently during walking, followed by locomotor ac-
tivities of daily living, such as getting up from a chair
[1]. A critical component in the safe performance of
daily activities is dynamic stability, the ability to con-
trol the center of gravity (CoG) within and outside the
base of support (BoS) [2,3]. The poor dynamic stability
observed in many elderly individuals may commonly
result from a combination of chronic deterioration of
musculoskeletal, neuromuscular and somatosensory
systems, caused by genetic traits, lifestyle and other
factors [4,5]. The relationship of muscle strength to
* Corresponding author.
E-mail address: dscarborough@partners.org (D. Moxley Scarbor-
ough)
trunk and CoG movement and the momentum genera-
tion during chair rise and walking in elderly individuals
is essential to determine which muscles exert the most
influence on dynamic stability. Based upon studies
which identify those muscles most critical for dynamic
stability during routine activities of daily living, such as
arising from a chair, care providers can develop more
rational approaches to treatment for reducing the likeli-
hood of falls in at-risk individuals.
Several investigations stress the importance of
quadriceps muscle strength in older persons for the
successful performance of chair rise and gait [5–12].
Timed sit to stand tests can estimate quadriceps
strength, but do not quantify dynamic stability during
the task [10,11]. Several investigators have contributed
a variety of detailed kinematic and kinetic data during
chair rise with self selected feet and arm positioning
and with constrained arm and feet positions [9,13–

0966-6362/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved.
PII: S 0 9 6 6 - 6 3 6 2 ( 9 9 ) 0 0 0 1 8 - 1
 D. Moxley Scarborough et al. / Gait and Posture 10 (1999) 10–20
19].One investigation has shown that quadriceps muscle
strength is critically important to develop the moment
across the knee that is necessary to enable functionally
limited elders to arise from a chair [9]. Recently,
Schenkman et al. confirmed the relationship of required
lower extremity strength to performance of successful
chair rise with challenging the subjects with lowering
chair heights [20]. In this particular study the investiga-
tors found a significant correlation between combined
lower extremity peak torque strength values and func-
tional reach, r =0.47, P =0.001 [20]. However, to our
knowledge there has been no report on the relationship
of kinematic or kinetic data, measuring dynamic stabil-
ity, during chair rise to specific muscle strength mea-
surements. Correlations to dynamic stability measures
such as time-distance parameters including stride
length, gait velocity and double support time to lower
extremity muscle strength have been investigated during
gait [2,10–12,21–24]. However, we have not found
reports on the relationship of leg muscle strength and
whole body CoG momentum as a measure of dynamic
stability, during gait.
The ability to generate and control segmental and
whole body momentum is a prerequisite for movement.
Therefore, we believe momentum plays an intricate role
in dynamic stability as it influences ‘‘the ability to
control the center of gravity (CoG) within and outside
the base of support (BoS)’’ [2]. The coordination of
internal forces, such as muscle and ligament forces, and
the influence of external forces, such as gravity, impacts
the ability to control momentum. Including momentum
as a measure of dynamic stability should provide infor-
mation about locomotor dynamics. Past studies of mo-
mentum generation have primarily focused on
describing movement patterns during chair rise [14–16].
Recently, Kaya and colleagues investigated momentum
control during chair rise and gait among healthy elders
and elders with bilateral vestibular hypofunction [25].
These investigators suggested healthy elders limit the
speed of rising from a chair and gait velocity to lessen
momentum generation because of limits in balance
control or strength which are needed to dissipate mo-
mentum generated by faster gait velocity [25]. The
previous studies have not correlated dynamic stability
measures such as momentum generation to specific
muscle strength measures. The influence of specific
muscles’ strength on dynamic stability, and on safe
performance of these activities, deserves further investi-
gation. Clinicians frequently suggest different move-
ments to compensate for weakness. For example,
during chair rise clinicians often advise the use of
repeated forward trunk flexion, referred to as ‘rocking’,
to build momentum to assist with lift-off during chair
rise. A weaker individual may not be capable of safely
controlling the momentum generation creating loss of
dynamic stability and resulting in a failed chair rise,
11
stumbling forward or falling backward into the chair
[19,25].
The continuous firing pattern of the quadriceps dur-
ing chair rise has been reported [17]. Resisted exercise
training can significantly improve quadriceps strength
among elderly persons [11]. Therefore, we investigated
the role quadriceps muscle strength plays in dynamic
stability during chair rise and gait. We have focused on
the influence of the upper body (encompassing the
pelvis and all segments superior) during chair rise to
include the most dynamic movements of body mass
being translated and rotated over the fixed feet. The
upper body demonstrates the greatest variation in
movement patterns during chair rise [15,20]. Therefore,
the use of upper body measurements could provide
more comprehensive information on dynamic stability
during chair rise. During gait we expected to see per-
sons with weaker quadriceps strength compensate by
implementing a larger range of momentum generation,
to assist in the development of speed and movement.
We hypothesized that quadriceps strength correlated:
(1) inversely with chair rise cycle time, maximum trunk
flexion, maximum range of lateral trunk flexion, upper
body anteroposterior (A/P) linear momentum and
whole body CoG to ankle A/P difference at lift-off
during chair rise; (2) directly with chair rise maximum
upper body vertical linear momentum; (3) directly with
gait stride length, average velocity, and center of pres-
sure (CoP)-CoG moment arm (defined in Section 2);
and (4) inversely with gait double support time and
range of whole body CoG lateral and A/P linear
momentum.
2. Methods
2.1. Subjects
Of 50 possible subjects, 34 participants met the inclu-
sion/exclusion criteria. The subjects ranging in age from
60 to 88 years, were involved in the Strong-for-Life
in-home, resistance strength training study [25]. All
subjects provided written informed consent and Table 1
provides characteristic data of all the participants from
our sample of convenience. To be selected each partici-
pant had to indicate that their health status limited
their ability to perform one or more functional activi-
ties (Table 1), but be able to both follow commands
and ambulate independently for 25 feet [26]. Each
participant received medical approval from their pri-
mary care physician and each passed an in-home safety-
exercise test which monitored heart rate, blood
pressure, and respiratory rate during a series of exer-
cises. Excluded were subjects with terminal illness, neu-
rologic disease, diabetes mellitus, major loss of vision
(legally blind), and painful or gross musculoskeletal
12 D. Moxley Scarborough et al. / Gait and Posture 10 (1999) 10–20

structural abnormalities (joint deformity such as in
rheumatoid arthritis).
2.2. Instrumentation
Strength measurements were performed at each sub-
ject’s home with a hand-held Nicholas MMT Dy-
namometer (Model c 01160; Lafayette Instrument).
The total body biomechanical analysis was performed
at the Massachusetts General Biomotion Laboratory.
Kinematic and kinetic data acquisition is described in
detail elsewhere [8,27,28]. The instrumentation in-
cluded: two Kistler piezoelectric force plates, two com-
puter terminals, four Selspot II™ optoelectronic
cameras; and 64 infrared light emitting diodes
(irLEDs). The irLEDs are mounted on 11 plastic arrays
attached securely to the following respective body seg-
ments of each participant: right and left feet, shanks,
thighs and arms, pelvis trunk and head (Fig. 1).
TRACK™, PVWAVE and SuperPlot II software were
used for data collection, processing and analysis. Sub-
jects performed the chair rise activity on an armless,
backless, height adjustable chair [27]. The array posi-
tion and orientation data were converted into body
segment positions yielding 6 df for each of the 11 body
segments and a three-dimensional android model (Fig.
4) [27]. The instrumentation and processing of raw
kinematic data yields resolutions of B 1° and B1 mm

Table 1
List of subject characteristicsa

Subject ID no. Age (years) Gender Functional Height (m) Weight (kg) Quadriceps MVIC Leg tested
limitationb strength (kg-Output/
kg-BW)

1 65.42 F c,f 1.52 52.27 8.70 (0.00) R

2 80.00 M b,c,d,e 1.66 87.73 8.29 (0.15) R
3 75.75 F a,b,e 1.56 61.36 8.37 (0.52) L
4 75.66 F a,b,c 1.57 59.09 10.19 (0.16) R
5 68.92 M a,b 1.63 65.91 11.76 (0.05) R
6 82.83 M c 1.65 85.91 7.91 (0.26) R
7 64.83 F a,b,c,e 1.65 66.36 17.23 (1.21) R
8 70.08 F c,e 1.50 68.64 5.73 (0.14) R
9 81.66 F a–i 1.55 72.73 9.81 (0.35) R
10 66.16 F b–d,f,g,i 1.60 68.18 10.27 (0.38) R
11 75.42 M c 1.83 86.36 6.82 (0.63) R
12 88.08 F a–d,f 1.57 57.73 11.06 (0.72) R
13 67.92 F a–e 1.51 58.18 10.51 (0.17) R
14 68.75 F a–f 1.57 79.55 10.71 (0.20) R
15 85.92 M c 1.83 59.09 11.92 (1.41) L
16 65.58 M a–d,f,g 1.70 90.91 5.8 (0.21) R
17 67.16 F a–f 1.73 92.27 4.66 (0.38) R
18 80.33 F a,e 1.65 75.00 13.97 (1.93) R
19 72.50 M a,c,f,g 1.73 59.09 15.15 (1.41) R
20 68.16 F a,b 1.57 47.73 8.33 (0.34) R
21 77.83 M e 1.91 79.55 13.31 (0.65) R
22 66.00 F e 1.57 77.27 4.12 (0.17) R
23 88.42 F a–f 1.70 69.09 5.53 (0.56) R
24 83.50 F a–g,i 1.50 54.09 5.88 (0.48) L
25 70.25 F a–i 1.73 67.27 4.90 (0.14) R
26 79.00 F a–g,i 1.56 63.64 7.14 (0.00) R
27 65.41 F a–e 1.60 55.91 13.86 (0.40) R
28 67.42 F a–c,e–h 1.57 102.73 7.10 (0.09) R
29 81.50 F b,c,e,f 1.63 63.64 6.5 (0.51) R
30 79.20 F a–g 1.56 56.36 12.46 (0.74) R
31 73.66 F a–g 1.70 93.18 3.68 (0.60) R
32 73.58 F a,be–h 1.50 77.27 0.54 (0.07) L
33 60.25 F c,d 1.66 84.09 5.35 (0.08) R
34 72.17 M f 1.84 84.09 11.38 (0.88) R
n= 34 x̄ = 73.80 (7.48) F= 25, x̄ = 1.64 (0.11) x̄ =26.67 (4.99) x̄ =8.79 (3.71) R =30, L=
M= 9 4

a
Standard deviations in parentheses.
b
List of functional limitations, based on health status, include subject responding as either ‘limited a little’ or ‘limited a lot’ to one or more of
the following: (a) moderate activities, such as moving a table, pushing a vacuum cleaner, bowling or playing golf; (b) lifting or carrying groceries;
(c) climbing several flights of stairs; (d) climbing one flight of stairs; (e) bending, kneeling or stooping; (f) walking more than a mile; (g) walking
several blocks; (h) walking one block; (i) bathing or dressing yourself.
 D. Moxley Scarborough et al. / Gait and Posture 10 (1999) 10–20
Fig. 1. Subject with testing arrays in chair rise test position.
[27,28]. Upper body (head, arms, trunk and pelvis) and
whole body linear momentum values were calculated as
the product of each segment’s mass times the segment’s
CoG velocity (momentum=mass times velocity, kgM/
s)] [8]. Segmental mass estimations and kinematic data
are used to calculate the CoG of each segment and all
segments are used to calculate whole body CoG [27].
Kinetic data including vertical reaction forces, and
center of pressure (CoP) were measured from the force
plates [27].
2.3. Protocol
Subjects’ muscle strength was assessed at each partic-
ipant’s home within 2 weeks prior to the Biomotion
Laboratory tests. Quadriceps muscle strength was mea-
sured by a single tester. Strength was defined as the
force generated during a maximum voluntary isometric
contraction (MVIC) measured in kilograms (kg) with a
hand held dynamometer during a 3-s isometric hold at
60° knee flexion; the participant was seated in a chair
with feet on the floor [26]. The dynamometer was
calibrated at 90.50 kg prior to each test and was
placed against the right shin two finger widths (about 5
cm) above the malleoli. The subject was asked to ‘‘push
13
as hard as you can into me’’. The right knee was chosen
for consistency across subjects; however, left knee mea-
surements were collected from four subjects, who were
fearful of potential right knee pain during quadriceps
muscle strength testing. An initial practice measurement
was performed followed by two recorded strength mea-
surements. The two recorded MVIC values were each
divided by the individual’s weight [29]. The average of
the two values was used in the statistical analysis.
All participants were barefoot during chair rise and
gait tests. For each chair rise trial the participant was
seated with the greater trochanters approximately 4 cm
from the front edge of a backless chair, adjusted to
100% knee height, measured as the distance from the
right medial tibial plateau to the floor. Each subject
performed one practice and two recorded trials. Each
participant placed their feet as requested in a self
constrained position, 10 cm apart, with 18° of ankle
dorsi flexion. Trials were considered successful when
participants kept their arms folded in a constrained
position in front of their abdomen and their feet re-
mained still during the chair rise. Each participant was
asked to arise from the chair as they ‘normally do’
beginning on ‘go’ after the cue ‘one, two, ready, go’.
Data collection began on ‘ready’ and once the subject
stood upright, s/he looked straight ahead and stood as
still as possible during the remaining portion of the
total 7-s data collection period. Arm and foot position
constraints were used to improve the consistency of the
body position during chair rise testings [14,18]. A total
of 29 subjects were able to arise independently, without
taking a step or unfolding their arms, at 100% knee
height, but five subjects were unable to arise from the
100% knee height chair without using their arms to
push up or assistance from the tester; they repeated the
same protocol with a 120% knee height chair. The
variables chosen to measure dynamic postural stability
during chair rise were analyzed only for subjects able to
perform chair rise at 100% knee height.
From start of movement (SOM) to end of rise (EOR)
we analyzed maximum upper body vertical and antero-
posterior linear momentum, maximum trunk flexion
(measured relative to room coordinates) and maximum
range of lateral trunk flexion. These measures were
collected between the SOM time and EOR time of chair
rise as defined in Fig. 2. The chair rise cycle time was
also calculated for all subjects as the time from SOM
through EOR. The A/P difference between the CoG
and the ankle joint (the average of the left and right
ankle joint centers’ positions) in the sagittal plane at the
time of lift-off from the chair was normalized to the
A/P difference of CoG to ankle joint at the end of rise
for each individual. This value was then normalized to
each individual’s height in meters.
For each gait trial the subject walked approximately
8 m. Data collection began when each subject entered
14 D. Moxley Scarborough et al. / Gait and Posture 10 (1999) 10–20
Fig. 2. The determination of chair rise events. (A) Start of movement
(SOM), defined as the time at which forward upper body momentum
begins (vertical line labeled ‘S’). (B) Lift-off (LO), defined as the time
at which the thigh segment moves upwards from its original position
by 2° (vertical line labeled ‘L’). (C) End of rise (EOR), defined as the
time at which the upper body CoG reaches its highest vertical
position (vertical line labeled ‘E’).
the 2 × 2 ×2-m viewing volume, approximately 3 m
beyond the starting point. Each subject was asked to
‘‘move forward in as straight a line as possible, walk-
ing at your normal pace, as if you were taking a brisk
walk in the park’’. The following gait variables were
determined: stride length (distance traveled between
consecutive ipsilateral heel strikes), gait velocity (aver-
age whole body CoG velocity during one gait cycle),
double support time (one of the two periods of time
during which the body is supported by both limbs,
expressed as % of the gait cycle), maximum moment
arm (the distance between the whole body CoG and
CoP, in both the anteroposterior and medial lateral
planes during single limb stance phase) [17], range of
Table 2
Normalized quadriceps muscle strengtha values
Subjects n Mean S.D. Range
Performed CR at 100% knee 29 9.58 3.33 4.12–17.23
height
Performed CR at 120% knee 4 5.18 1.26 3.68–5.88
height
Unable to perform CR 1 0.54
a
Strength equals the total output of knee extension strength in kg
divided by total body weight in kg (kg-output/kg-BW). CR, chair
rise.
whole body CoG anteroposterior (A/P) linear momen-
tum (the absolute difference between the maximum
and minimum anteroposterior linear momentum values
during a gait cycle), and range of whole body CoG
lateral linear momentum (the absolute difference of the
maximum right and left lateral momentum values, dur-
ing a gait cycle).
2.4. Data analysis
The average value for all data from two trials was
analyzed. Means, standard deviations, and linear corre-
lation analyses were performed for all data. Pearson
product-moment correlations between quadriceps
strength and all the dynamic stability variables for
chair rise and gait were performed to test each hypoth-
esis. For analysis of group differences between chair
rise strategy groups to be more conservative an
ANOVA (two-tailed) was performed. Statistical signifi-
cance of correlations and group differences was deter-
mined at PB 0.05.
3. Results
3.1. Quadriceps muscle strength
The weight adjusted quadriceps strength for all sub-
jects ranged from 0.54 to 17.23 kg per kg of body
weight (Table 2).
3.2. Relationship of quadriceps muscle strength to
measures of dynamic stability during chair rise
Quadriceps strength had a significant, moderately
strong correlation with maximum upper body vertical
linear momentum, (r=0.53, PB 0.005) and there also
was a significant direct correlation with maximum up-
per body anteroposterior (A/P) linear momentum, (r=
0.38, PB0.05), (Fig. 3). Fig. 3C demonstrates that as
the strength of the quadriceps muscle increased, the
faster each subject could arise from the chair. There
was a significant inverse correlation between quadri-
ceps muscle strength and chair rise cycle time (r= −
0.48 and PB 0.05). The participant with the least
quadriceps muscle strength (0.54 kg, S.D.= 0.07) was
unable to perform chair rise at either 100% or the
increased seat height of 120% of knee height. There
were four other elders with low values of knee extensor
strength who were unable to perform chair rise with
the seat of the chair at 100% of knee height but could
arise from the chair with the seat at 120% of knee
height. The whole body CoG to ankle A/P difference
at lift-off did not correlate with quadriceps strength
(r= − 0.07, P= 0.37).
 D. Moxley Scarborough et al. / Gait and Posture 10 (1999) 10–20
Fig. 3. Relationships between relative quadriceps muscle strength
(kg-Output/kg-BW) and dynamic stability during chair rise. (A) The
linear relationship between relative quadriceps muscle strength and
maximum upper body vertical linear momentum during chair rise
(r = 0.53, P B 0.005). (B) The linear relationship between relative
quadriceps muscle strength and maximum upper body A/P linear
momentum (r= 0.38, P B0.05). (C) The linear relationship between
quadriceps muscle strength and chair rise cycle time (r = −0.48,
P B0.01). The bold lines represent the regression lines and the dashed
lines show the 95% confidence interval ranges (n= 29).
Trunk motion during chair rise was not significantly
related to knee extensor strength. Quadriceps muscle
strength was not correlated with either maximum trunk
flexion (r= − 0.13, P = 0.4874) or range of lateral
trunk flexion, (r= − 0.097, P =0.6165). Table 3 sum-
marizes the correlations of the measures of dynamic
stability during chair rise.
15
3.3. Mo6ement strategies of chair rise
During chair rise subjects used three major move-
ment patterns, or strategies. The most frequently ob-
served was the momentum-transfer chair rise strategy
(Fig. 4) first described by Schenkman and colleagues
[15,16]. Elders in our study also performed the exagger-
ated trunk flexion chair rise strategy (Fig. 5), described
previously and referred to as the stabilization strategy
by Hughes and Schenkman [15]. The third strategy we
observed has not previously been reported. The domi-
nant vertical chair rise strategy is described in detail in
Fig. 6. The subjects in this sample consistently per-
formed the same strategy for each of the consecutive
two chair rise trials. Table 4 presents descriptive find-
ings of average quadriceps strength values and subject
characteristics for each of the three chair rise strategy
groups. There was no statistically significant differences
in quadriceps strength across groups, P= 0.99.
3.4. Relationship of quadriceps strength to dynamic
stability during gait
As hypothesized, quadriceps strength had a signifi-
cant, moderately high correlation with stride length
(r=0.558, PB 0.001) and preferred gait velocity (r=
0.51, PB0.002) (Fig. 7). Double support time corre-
lated inversely with quadriceps strength (r= −0.41,
PB 0.02). Maximum moment arm had a lower but
positive correlation with quadriceps muscle strength
(r=0.35, PB 0.05). The range of the whole body CoG
A/P linear momentum values correlated directly with
quadriceps strength (r= 0.47, P= 0.004, n= 31). There
was no statistical correlation between quadriceps
strength and maximum range of whole body lateral
linear momentum (r= 0.25, P=0.08, n= 33). Table 5
summarizes the correlations among the measures of
dynamic stability during gait.
3.5. Effect of age on quadriceps strength and measures
of dynamic stability during chair rise and gait
Age did not predict quadriceps muscle strength (r=
0.0843, P= 0.6638). Of the 34 subjects examined, 19
were young elderly (60–75 years) and 15 were old
elderly ( \ 75.1 years). The subject who was unable to
perform chair rise at 100 or 120% knee height was in
the young elderly group. Of the persons requiring the
increased 120% knee height chair, half were in the
young elderly group and the other half in the old
elderly group. Of the persons who were able to success-
fully arise from a chair at 100% knee height, 16 were in
the young elderly and 13 were in the old elderly age
groups. Age did not correlate significantly with mea-
sures of dynamic stability during gait or chair rise.
16 D. Moxley Scarborough et al. / Gait and Posture 10 (1999) 10–20

Table 3
Linear correlations of dynamic stability measures during chair rise

Upper body vertical Chair rise Maximum trunk Range of lateral CoG to ankle A/P
linear momentum cycle time flexion trunk flexion difference

Upper body A/P linear
momentum
Upper body vertical linear
momentum
Chair rise cycle time
Maximum trunk flexion
r=0.67, P= 0.0001* r = −0.28, P= 0.074 r = 0.37, P =0.03* r=−0.01, P =0.48 r =0.08, P =0.34
r= −0.71, P =0.0001* r=−0.04, P=0.42 r =−0.14, P = 0.22 r=−0.02, P =0.46
r = 0.42, P =0.01* r= 0.25, P = 0.09 r =−0.02, P = 0.45
r =0.29, P= 0.13 r=−0.14, P =0.23

* Correlation reaches level of significance.

4. Discussion functionally impaired elderly subjects [14,15,18]. Inves-

tigators established the importance of quadriceps
Development of effective strengthening treatment strength for successful chair rise, especially from low
programs for prevention of falls and maintaining safe, chair heights, among healthy adults and functionally
independent function among elderly persons depends impaired elders [9,22].
on determination of how specific muscle strength influ-
ences dynamic stability during the common activities of
arising from a chair and walking. We focused on the
quadriceps femoris muscle strength because of its large
size, ease of testing and known association of power
generation during chair rise [10,11,15,17,21]. The de-
tailed biomechanical analysis of several dynamic stabil-
ity measures allowed us to discover the important role
quadriceps muscle strength plays in dynamic stability
during chair rise and during gait among elderly per-
sons. Contrary to clinical wisdom and our hypothesis, Fig. 4. Momentum-transfer chair rise strategy. At time of lift-off, the
upper body anterior momentum transfers to total body vertical
during chair rise our elderly subjects used momentum
momentum with continued anterior momentum [16,17]. The follow-
differently than expected, demonstrating three strategies ing is our kinematic description of the momentum-transfer strategy,
to accomplish chair rise, independent of quadriceps to assist the clinician when observing persons: at the time of lift-off
muscle strength (Table 4). During chair rise some elders upper body flexion continues with initiation of knee extension which
avoid the potentially destabilizing, but more functional, smoothly transitions to simultaneous back and knee extension, until
erect posture is attained. (The vertically directed line originating from
momentum transfer strategy, apparently because they the force plates within the floor represents the ground reaction force.
lack the coordination and/or strength in muscle groups The cross seen at the level of the pelvis represents the whole body
other than the quadriceps needed to dampen and con- CoG.)
trol the increased forces induced by the rapid move-
ment resulting from momentum generation. However,
persons with the greatest quadriceps strength values
typically performed chair rise faster and generated
greater vertical and A/P linear momentum, demonstrat-
ing better stability and ability to control movement,
regardless of which chair rise strategy they performed.
These findings suggest that clinicians caring for elderly
persons should monitor quadriceps strength and recom-
mend quadriceps strengthening exercises for maintain-
ing and improving dynamic stability.
Fig. 5. Exaggerated trunk flexion chair rise strategy. The exaggerated
Having normal quadriceps strength is important in trunk flexion chair rise strategy has been described previously and
maintaining dynamic stability during the common ac- referred to as the stabilization strategy [8,17]. We have expanded the
tivity of arising from a chair. Schenkman and col- description as: exaggerated trunk flexion prior to lift-off, frequently
leagues described four phases of chair rise during two followed by further trunk flexion, placing the CoG over the feet
resulting in delayed trunk extension during the final transition to the
strategies of chair rise [14,15,18]. The momentum-trans- erect posture. (The vertically directed line originating from the force
fer chair rise strategy was observed in healthy/normal plates within the floor represents the ground reaction force. The cross
subjects and the stabilization strategy was observed in seen at the level of the pelvis represents the whole body CoG.)
 D. Moxley Scarborough et al. / Gait and Posture 10 (1999) 10–20
Fig. 6. Dominant vertical rise chair rise strategy. The dominant
vertical chair rise strategy is described as follows: at time of lift-off,
upper body anterior momentum lessens and cessation of forward
trunk flexion immediately follows with subsequent dominance of
vertical momentum and knee extension. (The vertically directed line
originating from the force plates within the floor represents the
ground reaction force. The cross seen at the level of the pelvis
represents the whole body CoG.)
We found a strong correlation between quadriceps
muscle strength and maximum upper body vertical
linear momentum during chair rise. Quadriceps muscle
contraction creates the necessary vertical velocity,
which in turn determines the vertical linear momentum
of each person as they arise from a chair. We also
found quadriceps muscle strength correlated inversely
with chair rise cycle time. These observations confirm
and extend previous conclusions, that the time required
to arise from a chair provides a good indirect measure-
ment of knee extensor strength [10,11].
In contrast to our original hypothesis, lateral trunk
flexion during chair rise does not correlate with quadri-
ceps muscle strength. We hypothesized that persons
with stronger quadriceps would demonstrate greater
dynamic stability and thus have less lateral instability.
Persons with weak quadriceps muscles may also have
weakness in lateral and other proximal stabilizing mus-
cles, potentially leading to lateral instability. However,
the predominant anteroposterior (A/P) direction of
movement that occurs during chair rise leads to insta-
bility primarily in the same A/P direction, not to lateral
Table 4
Chair rise strategy groups with respective quadriceps strength values and subject characteristics
n Quadriceps strengtha
Momentum transfer strategy Mean 17 9.62
(S.D.) (3.59)
Range 5.35–17.23
Exaggerated flexion strategy Mean 4 9.71
(S.D.) (3.74)
Range 4.90–13.31
Dominant vertical rise strategy Mean 9.42
(S.D.) (2.99)
Range 8 4.12–13.97
a
Quadriceps muscle strength values [(kg-output/kg-BW)100]. There was no statistical difference in quadriceps strength between chair rise
strategy groups (ANOVA, P=0.958).
17
instability [19]. Future investigations such as the rela-
tionships of lateral trunk flexion and lateral center of
gravity displacement during chair rise to the strength of
the hip abductors and ankle everters, may help to
establish if these muscles, in addition to the quadriceps
femoris, have a primary role in dynamic stability during
chair rise.
We also hypothesized that the weaker elderly partici-
pants would generate greater A/P momentum to assist
with lift-off, to compensate for their decreased strength.
Instead, quadriceps muscle strength correlated directly
with maximum upper body A/P linear momentum.
There are at least two explanations for this unexpected
finding. First, stronger persons are better able to gener-
ate a greater velocity on arising, thus creating higher
momentum. Second, and more interestingly, the weaker
elders, in general, do not attempt to generate greater
A/P momentum as a compensatory strategy. This is
probably because they lack the strength to control
increased forces. We had expected, as suggested by
previous investigators, that weaker participants would
increase trunk flexion prior to lift-off in order to posi-
tion their CoG over their ankles and to generate mo-
mentum [14,18]. The exaggerated trunk flexion strategy
was performed by several participants in our investiga-
tion, however, their quadriceps muscle strength ranged
from low to high values.
Some of the elderly participants in our study used the
dominant vertical chair rise strategy, and these elderly
individuals also have values of quadriceps muscle
strength ranging from low to high values. Surprisingly,
certain weaker elderly participants use this strategy,
despite the increased leg muscle effort required and the
apparent energetic inefficiency of this movement strat-
egy to arise from a chair. Several investigators have
suggested that the whole body CoG to ankle position
A/P difference at lift-off varies with the movement
strategies of chair rise [3,15]. We did not find a statisti-
cally significant correlation between quadriceps strength
and whole body CoG to ankle position A/P difference
Age (years) Height (m) Weight (kg)
72.63 1.61 69.89
(8.58) (0.067) (15.08)
60.25–88.42 1.50–1.73 47.73–102.73
74.86 1.75 64.55
(8.97) (0.165) (11.73)
65.42–85.92 1.52–1.91 52.27–79.55
74.47 1.64 74.43
(5.34) (0.121) (9.85)
66.00–81.66 1.55–1.84 59.09–86.36
18 D. Moxley Scarborough et al. / Gait and Posture 10 (1999) 10–20
Fig. 7. Relationships between relative quadriceps muscle strength
(kg-Output/kg-BW) and dynamic stability during gait. (A) Relation-
ship between relative quadriceps muscle strength and stride length
(r= 0.56, P = 0.001, n = 34). (B) Relationship between relative
quadriceps muscle strength and gait velocity (r = 0.51, P= 0.002,
n =34). (C) Relationship between relative quadriceps muscle strength
and the maximum range of whole body A/P linear momentum
(r= 0.47, P =0.004, n = 31). The bold lines represent the linear
regression lines and the dashed lines show the 95% confidence interval
ranges.
at lift-off. These data suggest that quadriceps femoris
muscle strength alone does not determine the strategy
for arising from a chair in functionally limited elderly
persons.
The reasons accounting for the different strategies
used to arise from a chair in our study group require
further investigation. Participants who are unable to
arise from a chair at 100% of their knee height have
failed in their attempt while using one of the above
three strategies. Stepping forward, backwards or falling
back onto the seat (sit-back) are body reaction move-
ments, that we have observed during failed attempts at
arising from the chair [19]. These adjustments for posi-
tioning the body can sometimes lead to falls and in-
juries [1,19]. Continued investigations to determine
which strategy provides the least likelihood of failure in
arising from a chair among elderly persons with all
ranges of quadriceps strength may be valuable to the
clinician for designing preventative health programs for
functionally limited, weaker, at-risk individuals. At
present clinicians may instruct elderly persons to use
greater momentum, or ‘rocking’, to assist in their lift-
off during chair rise. The biomechanical analysis per-
formed in our study suggests using great caution in
offering this advise as it may inadvertently induce a
person with weak quadriceps muscles to fall forward if
the elder cannot control the increased momentum asso-
ciated with this pattern of movement [19]. Muscle
strength and motor control are also needed to dampen
the effects of this additional generated momentum. Our
data suggest that coordinated anteroposterior, then ver-
tical momentum peaks are needed to successfully arise
from a chair (Figs. 4–6).
Our investigation further supports the importance of
quadriceps strength to dynamic stability during gait. As
hypothesized, stride length, gait velocity, and maximum
moment arm correlated directly and double support
time correlated inversely with quadriceps muscle
strength. Similar findings from a previous investigation
of 26 elderly persons demonstrated that quadriceps
strength correlated significantly with step length (r=
0.69 PB 0.001) [21]. Our data also supported previous
findings from larger sample sizes that stride length
correlates directly with gait velocity [12,24]. Longer
stride length obviates shuffling; shuffling is observed
frequently in older persons and may also decrease falls
which often occur during hurried gait or when turning
in older individuals [1,30]. Creating longer strides al-
lows a person to increase gait speed, which is important
for independence. The ability of an elderly person to
increase their walking speed to avoid a moving obstacle
on the sidewalk (such as a person on roller skates, a
dog, or a bicyclist etc.) or to cross a road quickly is an
important safety issue among independent, community
dwelling, elderly, persons.
Our data indicate that elderly persons with greater
quadriceps muscle strength had shorter double support
time than weaker individuals. Double support time has
been used as a parameter of dynamic stability by many
investigators [2,7,23,24]. Elders with greater quadriceps
strength values had larger maximum moment arm val-
ues during gait. Therefore, stronger elderly participants
move their CoG farther away from their BoS, than the
weaker participants. This demonstrates confidence in
controlling the CoG beyond the BoS ‘comfort zone’
 D. Moxley Scarborough et al. / Gait and Posture 10 (1999) 10–20
Table 5
Linear correlations of dynamic stability measures during gait
Gait velocity Double support time
Stride length r = 0.94, (n=34) r = −0.69, (n= 34)
P =0.0001* P= 0.0001*
Gait velocity r= −0.70, (n= 34)
P = 0.0001*
Double support
time
Maximum Moment
arm
Range of A/P mo-
mentum
* Correlation reaches level of significance.
which is a primary skill required in dynamic stability.
The relationship of quadriceps strength to maximum
range of A/P linear momentum values during gait
suggests greater quadriceps strength allows elderly per-
sons the ability to generate greater forward momentum
likely contributing to increased gait velocity. Persons
with weaker quadriceps strength values did not develop
high momentum values as a strategy to increase gait
velocity, suggesting inability to control the additional
forces that result from greater momentum generation.
In activities such as walking, it appears that greater
quadriceps muscle strength enhances dynamic stability
and may play an important role in reducing the risk of
falls and related injuries in functionally limited elderly
individuals.
Limitations of the present investigation include a
modest sample size and a predominance of female
participants. Strength measured in Newtons of force, as
opposed to our kilogram units, might allow more direct
comparison to results of other investigations that use
force units. This study implemented a standardized,
scientifically accepted isometric strength testing proto-
col, which has been used when correlating strength to
functional tasks [26,29,31]. Further investigations of the
correlations between quadriceps muscle strength and
the strength of other muscles, such as the trunk exten-
sors, the abdominal muscles, hip flexors, abductors and
extensors, ankle plantar flexors, dorsi flexors, inverters
and everters and measurements of dynamic stability
during chair rise and gait should be carried out in a
larger sample of functionally limited elders. The influ-
ence of exercise training, specifically quadriceps femoris
muscle strengthening, on different measurements of
dynamic stability among elderly individuals and the
effects of such training on the frequency of falls and
injuries, needs to be assessed. Nonetheless, our data
represent the largest sample and most sophisticated
analysis reported to date evaluating the relationship
between strength and a variety of dynamic stability
measurements during chair rise and gait.
19
Moment arm Range of A/P Range of lateral
momentum momentum
r = 0.82, (n =34) P = r = 0.14, (n = 34) P= r= 0.36, (n =34)
0.0001* 0.25 P= 0.03*
r= 0.80, (n =34) P= r =−0.05, (n =27) r =0.17, (n =29)
0.0001* P= 0.41 P=0.19
r=−0.44, (n = 34) r =−0.13, (n =27) r =−0.09, (n =29)
P= 0.0099* P =0.26 P=0.32
r =−0.21, (n = 31) r =0.24, (n = 33)
P= 0.13 P =0.09
r =0.20, (n =31)
P=0.14
5. Conclusions
With these elders, increased generation of A/P mo-
mentum was not used as a compensatory strategy for
weak quadriceps strength during chair rise or gait. The
results of our investigation introduces clinical perspec-
tives on three different movement strategies to accom-
plish chair rise among elders and the relationship of
quadriceps strength to dynamic stability measures.
Knee extensor strength correlates with most measures
of dynamic stability during chair rise and gait. Our data
indicate that clinicians should not suggest that patients
use compensatory momentum inducing locomotor
strategies unless the patient has sufficient strength to
control these induced forces. Adequate quadriceps mus-
cle strength is essential for the ability to successfully
perform functional activities of arising from a chair and
walking, as importantly, quadriceps strength enhances
the generation and control of momentum and mainte-
nance of good dynamic stability, ultimately influencing
safety and the likelihood of falls, during these two
locomotor activities of daily living.
Acknowledgements
This research was supported by grants from the
National Institutes of Health (R01AG12561 and P50
AG11669-03). We are grateful for the cooperation and
assistance of New England Research Institute’s staff
from its Roybal Center’s Strong for Life exercise pro-
ject, and for the assistance from Massachusetts General
Hospital Biomotion Laboratory staff, in particular Jose
V. Ramirez, MD, for his edits and locomotor data
collection. Acknowledgment and appreciation is ex-
tended to Patrick O. Riley, Ph.D. for his assistance
during the initial phases of this research. An extended
appreciation, from the first author, goes to my father,
Richard T. Moxley III, MD for his dedication, guid-
ance, research skills and encouragement.
20 D. Moxley Scarborough et al. / Gait and Posture 10 (1999) 10–20
References
[1] Tinetti ME, Speechley M, Ginter SF. Risk factors for falls
among elderly persons living in the community. N Engl J Med
1988;319:1701– 7.
[2] Krebs DE, Lockert J. Vestibulopathy and gait. In: Spivach BS,
editor. Evaluation and management of gait disorders. New
York: Marcel Dekker, 1995:93–115.
[3] Pai Y-C, Patton J. Center of mass velocity—position predictions
for balance control. J Biomech 1997;30:347–54.
[4] Horak FB, Shupert CL, Mirka A. Components of postural
dyscontrol in the elderly: a review. Neurobiol Aging
1989;10:727 – 38.
[5] Lord SR, Clark RD, Webster IW. Postural stability and associ-
ated physiological factors in a population of aged persons. J
Gerontol Med Sci 1991;46:69–76.
[6] Buchner DM, de Lateur BJ. The importance of skeletal muscle
strength to physical function in older adults. J Behav Med
1991;13:91 – 7.
[7] Wolfson L, Whipple R, Amerman P, Tobin J. Gait assessment in
the elderly: a gait abnormality rating scale and its relation to
falls. J Gerontol Med Sci 1990;45:M12–9.
[8] Hutchinson EB, Riley PO, Krebs DE. A dynamic analysis of the
joint forces and torques during rising from a chair. IEEE Trans
Rehab Engin 1994;2:49–56.
[9] Hughes MA, Myers BS, Schenkman ML. The role of strength in
rising from a chair in the functionally impaired elderly. J
Biomech 1996;29:1509–13.
[10] Bohannon RW, Smith J, Hull D, Palmeri D, et al. Deficits in
lower extremity muscle and gait performance among renal trans-
plant candidates. Arch Phys Med Rehabil 1995;76:547–51.
[11] Fiatarone MA, Marks EC, Ryan ND, Meredith CN, et al.
High-intensity strength training in nonagenarians: effects on
skeletal muscle. J Am Med Assoc 1990;263:3029–34.
[12] Perry J, Mulroy SJ, Renwick SE. The relationship of lower
extremity strength and gait parameters in patients with post-po-
lio syndrome. Arch Phys Med Rehabil 1993;74:165–9.
[13] Schultz AB, Alexander NB, Ashton-Miller JA. Biomechanical
analyses of rising from a chair. J Biomech 1992;25:1383– 91.
[14] Schenkman M, Berger RA, Riley PO, Mann RW, et al. Whole-
body movements during rising to standing from sitting. Phys
Ther 1990;70:638 – 48.
[15] Hughes MA, Schenkman ML. Chair rise strategy in the func-
tionally impaired elderly. J Rehabil Res Dev 1996;33:409–12.
[16] Pai Y-C, Naughton BJ, Chang RW, Rogers MW. Control of
body centre of mass momentum during sit-to-stand among
young and elderly adults. Gait Posture 1994;2:109 – 16.
[17] Millington PJ, Myklebust BM, Shambes GM. Biomechanical
analysis of the sit-to-stand motion in elderly persons. Arch Phys
Med Rehabil 1992;73:609 – 17.
[18] Ikeda ER, Schenkman ML, Riley PO, Hodge WA. Influence of
age on dynamics of rising from a chair. Phys Ther 1991;71:473–
81.
[19] Riley PO, Krebs DE, Popat R. Biomechanical analysis of failed
sit to stand. IEEE Trans Rehab Engin 1997;5:353 – 9.
[20] Schenkman M, Hughes MA, Samsa G, Studenski S. The relative
importance of strength and balance in chair rise by functionally
impaired older individuals. J Am Geriatr Soc 1996;44:1441–6.
[21] Judge JO, Davis RB, Ounpuu S. Step length reductions in
advanced age: the role of ankle and hip kinetics. J Gerontol Med
Sci 1996;51A:303 – 12.
[22] Imms FJ, Edholm OG. Studies of gait and mobility in the
elderly. Age Ageing 1981;10:147 – 56.
[23] Gill-Body KM, Krebs DE. Locomotor stability problems associ-
ated with vestibulopathy: assessment and treatment. Phys Ther
Practice 1994;13:232 – 45.
[24] Perry J. Gait analysis: normal and pathological function. Thoro-
fare, NJ: SLACK, 1992:19 – 47 also pp. 287 – 97.
[25] Kaya BK, Krebs DE, Riley PO. Dynamic stability in elders:
momentum control in locomotor ADL. J Gerontol Med Sci
1998;53A:M126– 34.
[26] Jette AM, Lachman M, Giorgetti MM, Assmann SF, et al.
Exercise: it’s never too late!. Am J Public Health 1998;89:66–72.
[27] Riley PO, Mann RW, Hodge WA. Modelling of the biomechan-
ics of posture and balance. J Biomech 1990;23:503 – 6.
[28] Antonsson EK, Mann RW. Automatic 6-D.O.F. kinematic tra-
jectory acquisition and analysis. J Dyn Syst Meas Control
1989;111:31 – 9.
[29] Andrews AW, Bohannon RW. Normative values for isometric
muscle force measurements obtained with hand-held dynamome-
ters. Phys Ther 1996;76:248 – 59.
[30] Alexander NB. Gait disorders in older adults. J Am Geriatr Soc
1996;44:434 – 51.
[31] Krebs DE, Jette AM, Assmann SF. Moderate exercise improves
gait stability in disabled elders. Arch Phys Med Rehabil
1998;79:1489 – 95.