Download Muscle Coordination During Rapid Force Production by Young and

Journal of Gerontology: MEDICAL SCIENCES
2005, Vol. 60A, No. 2, 232–240
Copyright 2005 by The Gerontological Society of America
Muscle Coordination During Rapid Force
Production by Young and Older Adults
Benjamin K. Barry, Stephan Riek, and Richard G. Carson
Perception and Motor Systems Laboratory, School of Human Movement Studies,
The University of Queensland, Brisbane, Australia.
Background. Older adults typically exhibit dramatic reductions in the rate of force development and deficits in the
execution of rapid coordinated movements. The purpose of the current study was to investigate the association between
the reduced rate of force development exhibited by older adults and the ability to coordinate groups of muscles.
Methods. The performance of a visually guided aiming task that required the generation of isometric torque about the
elbow joint was compared in 10 young adults (age range, 19 to 29 years) and 10 older adults (age range, 65 to 80 years).
Participants were required to exert isometric torque in flexion, extension, pronation, or supination, or in combinations of
these directions, to reach a target in minimum time. Surface electromyograms were obtained from the biceps brachii,
triceps brachii, brachioradialis, and flexor carpi radialis.
Results. Older participants exhibited slower target acquisition times compared with young participants (p , .05), with
the extent of the differences between the groups varying markedly between target locations.
Conclusions. The impairment in performance, although partially attributable to a general decline in the ability to
produce force rapidly, was also affected by the requirements for muscular coordination. At the neuromuscular level,
differences between the young and the elderly were expressed most prominently in the bifunctional muscle biceps brachii
and in certain temporal aspects of muscular coordination.
T
HE ability to rapidly develop force declines markedly
in older age (1,2) and is closely associated with decrements in the physical capabilities of older adults (3–5).
To design rehabilitative strategies intended to enhance the
movement capabilities of older adults, the particular
mechanisms that contribute to the deterioration of the
ability to rapidly develop force must be identified. Beyond
the reduction in rapid force development that arises from
muscle atrophy and the slowing of muscle contractile
properties (6,7), there also exists the potential for alterations
in the neural control of muscles to contribute to the lower
rate of force development that occurs with increasing age
(8–10). The marked slowing of movement in older adults
(11) has been linked to reductions in the ability to
coordinate muscles effectively (8,10).
Christou and Carlton (12) provided evidence that older
adults experience particular difficulties in coordination
during rapid actions. Older adults exhibited considerably
greater inter-trial variability compared with young adults
when they repeatedly performed a rapid discrete force
trajectory matching task. However, the 2 groups were not
distinguished when they produced slow continuous contractions (12). Darling and coworkers (9) reported that older
adults varied more in the production of aiming trajectories,
particularly when they were required to move rapidly. Less
effective timing of agonist–antagonist muscle activation
patterns accompanied these performance deficits. In addition, older adults are less accurate and less smooth when
they make movements to targets using the elbow joint in
conjunction with the shoulder joint, compared with
conditions in which the elbow joint is moved alone (10).
Progressively greater decreases in performance with an
increasing number of joints indicate a coordination deficit
232
in older adults. These findings support the view that older
adults exhibit deficiencies in muscle coordination that are
exacerbated by increases in the speed and complexity of
movement (10,13–15).
To produce rapid, goal-directed movements, muscle activation patterns are initially specified in a feed-forward manner and
subsequently modified on the basis of visual and proprioceptive
feedback (16). Muscle activation patterns are currently believed
to be formulated on the basis of an ‘‘internal model’’ that
provides an abstract representation of the transformation
between the command signal to the motor effectors and
consequent movement trajectories (17,18). Necessarily, the
ultimate output of an internal model is the specific motor
commands to initiate the patterns of muscle activation that are
suitable to accomplish a desired motor act (17,18). It is
generally accepted that these motor commands do not specify
the activity of muscles independently, but rather they control
groups of muscles and motor units through a network of
common neural inputs (19,20). Clearly, there is sufficient
flexibility within these neural circuits to allow muscle activation
patterns to be modulated to meet the demands imposed by
specific movement tasks (21). At the same time, the possible
combinations of muscle activation patterns are constrained by
the structure and plasticity of these neural circuits.
Degenerative changes in the aging neuromuscular system,
such as a decrease in the number of corticospinal fibers (22);
a decrease in intracortical inhibition (23); neuronal degradation in other higher nervous centers, such as the cerebellum
and the basal ganglia (24–26); and the reduced number of
motoneurons in the spinal cord (27) would be expected to
limit the flexibility with which muscle coordination patterns
are formulated and executed. Potentially this may impede the
ability of older adults to coordinate muscles optimally when
RAPID FORCE GENERATION BY OLDER ADULTS
233
affecting the elbow joint. The participants were all right handed
according to their preferential use of that particular hand during
sports (e.g., lawn bowls) and other daily activities. The
University of Queensland Medical Research Ethics Committee
approved all experimental procedures.
Force–Torque Recordings
All experimental tasks were performed using only the
dominant limb with participants seated and grasping
a manipulandum (Figure 1). The manipulandum was instrumented with a multiple-degree-of-freedom force–torque
transducer (Delta; ATI Industrial Automation, Apex, NC).
Force (flexion–extension, 0.25 N resolution) and torque about
the long axis of the forearm (pronation–supination, 0.015 Nm
resolution) were sampled at 1000 Hz at an analog-to-digital
interface (AT-mio-16E-10; National Instruments, Austin,
Texas) and the data stored on a personal computer. Labview
(version 5.0; National Instruments) was used to write the
custom experimental control and data acquisition routines.
Subsequent to collection, flexion and extension force recordings were converted to torque data for all analyses.
Figure 1. The experimental setup incorporates a computer display and
a manipulandum instrumented with a multidimensional force–torque transducer.
Participants were seated in an adjustable-height chair 65 cm from a computer
display (display size, 42.5 cm) positioned at eye level. All experimental tasks
were performed using only the dominant limb, which was placed in the pendent
position with the elbow flexed at 908 and the forearm in a neutral position. The
elbow joint was secured in a padded brace with a Velcro strap, and the hand
grasped the manipulandum. Padded clamps were located above and below the
hand to minimize the force required to grip the manipulandum. The outer edges
of the screen display were scaled to 100% MVC in flexion, extension, pronation,
and supination (top, bottom, left, and right screen edges, respectively). Targets
were displayed at the appropriate percentage of these screen boundaries (i.e.,
30% or 50%). The combination targets required that either 30% or 50% of the
individual direction MVCs be generated simultaneously (e.g., 50% flexion MVC
plus 50% supination MVC).
they try to develop force rapidly, especially in circumstances
that require the formation of novel patterns of muscle
activation. Any deficits exhibited by older adults are
therefore likely to be exacerbated by movements that impose
greater demands on muscular coordination.
The purpose of this experiment was to determine whether
the deficiencies in the rapid development of force exhibited
by older adults are contingent on inefficiencies in intermuscular coordination. Using the elbow joint as a model,
the experimental task required young and older adults to
produce isometric flexion, extension, pronation, or supination torques, in isolation or combination, to rapidly acquire
visually displayed targets. The task assessed, therefore, the
flexibility with which muscle synergies were composed and
recomposed to meet the demands of a variety of movement
tasks, each of which imposed unique coordination demands.
We evaluated age-related performance decrements in terms
of the formation of torque trajectories and the associated
patterns of muscle activation.
METHODS
Participants
From healthy populations, we recruited 10 older volunteers
(9 men, 1 woman; mean age 6 SD, 71.7 6 4.7; age range, 65 to
80 years) and 10 young volunteers (6 men, 4 women; mean age,
22.7 6 2.8; age range, 19 to 29 years) who had no disorders
Experimental Procedures
Experimental sessions began with adjustment of the
equipment dimensions for each participant and then a
familiarization program. Participants practiced exerting torque against the manipulandum using only the elbow joint
while maintaining wrist and shoulder posture. Once they had
become comfortable with this, they performed some practice
trials of the aiming task, after which the experimental tests
were begun. First we assessed maximal voluntary contractions (MVCs) and then the visually guided aiming task.
Maximal voluntary contractions.—Maximal voluntary
contractions were obtained for isometric flexion, extension,
pronation, and supination to determine separately for each
participant the torque levels that would be used in the
aiming task. After several practice attempts, the volunteers
performed MVCs three times for flexion, extension,
pronation, and supination. Successive attempts were separated by 2 to 3 minutes’ recovery (28). The maximum force
achieved for the three attempts was taken to be the MVC. If,
however, at least two of the attempts were not within 10% of
this value, extra attempts were completed to ensure that
a maximum had been achieved (28). Visual feedback of the
torque produced was provided to participants via a bar scale
display on the computer monitor.
Visually guided aiming.—The aiming task required the
generation of isometric torque about the elbow joint to
acquire targets located in positions that corresponded to the
application of flexion, extension, pronation, and supination
torques, either in isolation or in combination (i.e., flexion and
supination, flexion and pronation, extension and supination,
extension and pronation). When the target was presented, the
participants had to exert torque in the appropriate direction
and to reach the target in minimum time. The location of the
targets corresponded to isometric force levels that were either
30% or 50% of MVC. The 8 target positions were presented
in pseudorandom order, with each target location presented
twice in each of 8 blocks of 16 trials (for a total of 128 trials).
234
BARRY ET AL.
Figure 2. Sample data are from the torque trajectories produced during the acquisition of a combined extension and supination target by a young adult and an older
adult. Torque time-series data are included for the sample trials to depict the information contained in the trajectory descriptors. Dashed lines denote the movement
onset and target acquisition. (No trajectory history was displayed during the aiming trials.) Additional analyses of the torque trajectory data required that the X axis
(pronation/supination) and Y axis (flexion/extension) of the visually guided aiming torque recordings be combined to determine a resultant torque time series. The peak
RTD was obtained from the differentiated time series, and the temporal location of this peak was expressed as a proportion of the target acquisition time and is referred
to as the percentage time to peak RTD. To quantify the contribution of the peak RTD burst to the total target acquisition movement, the area under curve of the
differentiated time series was calculated for the segment between zero crossings that contained the peak positive value of the time series. This was expressed as
a proportion of the area under the curve for the velocity time series from movement onset to target acquisition (% peak RTD/total area). Effectively, this peak RTD
burst corresponded to the primary ‘‘submovement’’ of the torque produced in acquiring a target. The advantage of quantifying its contribution to the total aiming
movement in terms of the total change in torque, instead of the total movement duration, is that this particular measure also accounts for (by virtue of the normalization)
the magnitude of adjustments in the rate of torque production that follow the primary submovement. A temporal index that describes the proportion of the total target
acquisition time at which the primary submovement ended does not incorporate consideration of the magnitude of subsequent adjustments in torque, but rather only
their duration. With reference to the sample data, it is clear for the young adult that the burst of contraction containing the peak RTD occupies most of the target
acquisition period. In contrast, for the older adult, the initial burst only represented a small proportion of the trial length and many peaks and troughs occurred on the
‘‘velocity’’ trace. Trajectory error (in newton-meters) was defined as the RMS of the departure of the resultant trajectory from a straight line between the home position
and the target. This measure indicated the extent to which torque was produced purely in the required direction. The number of discrete modifications to the torque
trajectory was quantified as the instances in which a switch in the sign of the acceleration occurred (i.e., zero crossings of the acceleration time series).
For the first 64 trials, the targets were defined at 30%
MVC. In the second block of 64 trials, they were defined at
50% MVC. At the start of each trial, the computer screen
displayed a yellow dot in the middle, corresponding to the
zero torque (resting) position of the arm. When a cursor
indicating the torque level currently generated by the participant fell within this ‘‘home zone,’’ the color of the dot
changed from yellow to red. The participant had to maintain
zero net torque at the start of every trial. After a random
foreperiod (1 to 3 seconds), a target was presented, accompanied by an auditory cue (370 milliseconds, 1.2 kHz square
wave). The target was also displayed as a yellow dot, which
changed to red whenever the applied torque fell within a
target zone (defined for all directions as 65% of the smallest
of the MVCs for flexion, extension, pronation, or supination
relative to the target force level [e.g., 65% supination
torque]; the home zone was defined by the same criteria).
Once the level of applied torque was maintained in the
target zone for a continuous period of 0.1 second, a second
auditory cue was generated to signify the end of a trial.
Torque Data Processing and Analysis
The torque time series data recorded in the visually
guided aiming task were digitally low-pass filtered by using
a second-order, dual-pass Butterworth filter with a 15 Hz
cutoff frequency. Custom software (Labview) was used to
identify the movement onset and target contact. Movement
onset was defined as the point in time at which the torque in
either direction exceeded the boundary of the home zone.
Trials in which the target was not acquired within the
designated period (10 s) or the starting torque was not
within the required limits when the target was presented
were discarded. Reaction time was defined as the elapsed
time from target presentation to movement onset. Target
acquisition time was calculated as the elapsed time from
movement onset to the time when the target was first acquired, provided that contact with the target was maintained
for the subsequent 100 milliseconds. Subsequent analyses
(Figure 2) were completed using custom software written in
Matlab (Mathworks, Natick, MA).
Electromyographic Data
After standard skin preparation procedures, miniature preamplified bipolar surface electrodes (stainless steel) were
placed over the long head of the biceps brachii, the lateral head
of triceps brachii, brachioradialis, and flexor carpi radialis.
The electrodes (separation 13 mm) were positioned over the
midbelly of each muscle. One self-adhesive gel electrode
(Conmed, Utica, NY) was placed on the posterior aspect of
the wrist as a common grounding electrode. The electromyographic (EMG) signals were amplified (2000 times) and
band-pass filtered (30–500 Hz). These data were sampled at
1000 Hz and stored in the manner described previously.
The EMG recordings for each muscle were full-wave
RAPID FORCE GENERATION BY OLDER ADULTS
235
rectified and digitally low-pass filtered at 40 Hz (Butterworth dual pass). The root mean square (RMS) of the EMG
recorded during the visually guided aiming task was
calculated for the period between target presentation and
acquisition. The values thus obtained were then expressed as
a percentage of the maximum RMS recorded for that muscle
during a corresponding MVC (e.g., the maximum RMS
value obtained for the triceps during an extension MVC).
Custom software (Labview) was used to identify distinct
bursts of activity recorded from each muscle that occurred
during the target acquisition movements. The primary burst
detection algorithm identified burst onsets when the restinglevel EMG was exceeded by 2.5 standard deviations, and
offsets when the EMG returned to below this level. Additional temporal and threshold criteria were also imposed to
enhance the reliable detection of EMG bursts.
Statistical Analyses
Comparison of the young and older groups, for values
obtained at both the 30% and 50% target force levels, were
conducted based on simple main effects analysis of variance
designs fage (between) 3 target (within)g. The main effect
of age was determined across all target positions and within
each target position. Because the preliminary analysis failed
to reveal the presence of reliable age-related differences between the 30% and 50% MVC conditions, the values were
combined for the purposes of the current analyses. Separate
analyses of variance were conducted for each dependent
measure. Effect sizes were calculated according to the
method of Cohen (29). By convention, a small effect size is
indicated by an f value of .1, a medium effect size by an f
value of .25, and a large effect size by an f value of .4 (29).
RESULTS
Visually Guided Aiming Performance
Reaction time was consistently slower for the older
volunteers by an average of 28% (young adults: 402.6
milliseconds; older adults: 514.5 milliseconds; F(1,18) ¼
7.45, p , .05, f ¼ .15) (Figure 3A). The difference in
reaction time between young and older adults was very
similar across the different target directions. Target
acquisition times were longer overall for older volunteers,
by an average of 79% (young adults, 791.7 milliseconds;
older adults, 1417.5 milliseconds; F(1,18) ¼ 7.04, p , .05,
f ¼ .15) (Figure 3B). The extent of the age-related difference
varied markedly among the different target locations,
ranging from essentially equivalent target acquisition times
for combined extension and pronation (F(1,18) ¼ 0.01, p ¼
.942, f ¼ .01), to the marked differences observed for
extension, supination, and extension plus supination. In the
latter conditions, the target acquisition times recorded for
the older adults were approximately 1.5 to nearly 2.5 times
longer than those exhibited by the young adults.
The peak rate of torque development (RTD) was also
lower by 45% for the older volunteers across the 8 target
Figure 3. (A) Reaction time (s), (B) target acquisition time (s), and (C) peak
rate of torque development (newton-meters s1) are shown for young and older
participants in the visually guided aiming task. *p , .05; ysmall effect size;
z
moderate effect size; §large effect size. Shaded regions indicate the 95%
confidence intervals. Flex, flexion; flesup, flexion and supination; supn,
supination; extsup, extension and supination; extd, extension; extpro, extension
and pronation; pron, pronation; flepro, flexion and pronation.
BARRY ET AL.
236
Table 1. Trajectory Descriptors for Young and Older Participants During the Visually Guided Aiming Task
Flex
Flesup
Supn
Extsup
Extd
Extpro
Pron
Flepro
%ttpk_RTD
Young
Older
11.9 6 4.8
13.8 6 13.8
11.4 6 7.1
16.1 6 8.1
13.8 6 9.7
30.6 6 24.9*y
11.9 6 6.2
17.1 6 13.0
17.4 6 13.4
16.2 6 16.0
13.7 6 9.6
19.7 6 25.6
10.2 6 6.9
16.1 6 16.0
11.2 6 4.6
14.6 6 11.2
76.7 6 14.4
58.3 6 18.1*y
59.8 6 10.9
52.3 6 20.7
66.0 6 16.0
37.4 6 20.9*y
64.5 6 11.8
36.4 6 20.2*y
72.6 6 16.6
43.1 6 21.7*y
49.5 6 19.9
38.8 6 15.7
59.1 6 21.9
41.9 6 20.9
72.3 6 9.9
43.7 6 20.0*y
0.30 6 0.15
0.42 6 0.30
0.58 6 0.26
1.53 6 1.29*y
1.74 6 0.97
0.51 6 0.24*y
0.52 6 0.25
1.14 6 0.90*y
0.31 6 0.19
0.46 6 0.27
1.12 6 1.37
1.62 6 1.56
2.20 6 1.78
0.61 6 0.29*y
0.44 6 0.23
1.31 6 1.11*y
8.7 6 3.0
12.3 6 6.5*
12.7 6 4.5
21.4 6 21.7
7.7 6 5.1
22.3 6 20.0*
13.0 6 5.9
36.0 6 25.0*y
9.7 6 4.0
22.6 6 17.2*y
27.6 6 30.6
24.6 6 15.3
10 6 5.8
17.5 6 9.9*y
10.3 6 4.2
18.4 6 10.0*y
pkRTD: area, %
Young
Older
Traj Error, Nm
Young
Older
No. submvts
Young
Older
Mean 6 standard deviation for young (top) and (bottom) participants. The percentage of the target acquisition time to peak rate of torque development (RTD)
(%ttpk_RTD), impulse in the peak RTD burst as a percentage of the total impulse (pkRTD: area), root mean square of the trajectory ‘error’ (Traj. Error), the number of
discrete modifications to the torque trajectory or number of ‘sub-movements’ (No. submvts).
*p , .05 young vs older adults. yEffect sizes 0.40.
Flex ¼ flexion; Flesup ¼ flexion and supination; Supn ¼ supination; Extsup ¼ extension and supination; Extd ¼ extension; Extpro ¼ extension and supination; Pron
¼ pronation; Flepro ¼ flexion and pronation.
locations (young adults, 155.4 Nm1; older adults, 85.7
Nm1; F(1,18) ¼ 5.35, p , .05, f ¼ .13) (Figure 3C). This
difference was statistically significant for 4 of the 8 target
locations: flexion (44%; F(1,18) ¼ 5.34, p , .05, f ¼ .37),
extension (48%; F(1,18) ¼ 4.99, p , .05, f ¼.35), extension
plus pronation (54%; F(1,18) ¼ 7.57, p , .05, f ¼ .44), and
flexion plus pronation (51%; F(1,18) ¼8.37, p , .01, f¼.46);
and it was expressed prominently for 3 of the remaining
target positions: flexion plus supination (38%; F(1,18) ¼ 3.77,
p ¼ .068, f ¼.31), pronation (46%; F(1,18) ¼ 3.49, p ¼ .078, f
¼.30), and extension plus supination (43%; F(1,18) ¼ 3.27, p
¼ .087, f ¼ .29). For movements directed to supination
targets, however, the young and older adults did not differ
appreciably (11%; F(1,18) ¼ 0.09, p ¼ .772, f ¼.05).
In addition to the lower peak RTD exhibited by the older
group, the peak RTD tended to occur later (Table 1). This
trend was expressed across the 8 targets (young adults,
12.7%; older adults, 18%; F(1,18) ¼ 2.87, p ¼ .108, f ¼ .09),
but only for the supination target was this difference
statistically significant (F(1,18) ¼ 9.10, p , .01, f ¼ .48).
The proportion of the total trial impulse achieved during the
peak RTD burst was clearly lower for the older volunteers
(young adults, 65.1%; older adults, 44%; F(1,18) ¼ 5.26, p ,
.01, f ¼ .22). The older adults also displayed a larger
number of switches in the sign of the acceleration time
series. The overall difference was such that approximately
75% more zero crossings were exhibited by the older
volunteers (young adults, 12.5; older adults, 22, F(1,18) ¼
6.15, p , .05, f ¼ .14) and we observed statistically
significant differences for 6 of the 8 targets. The RMS of the
trajectory error quantified the departure from a straight line
trajectory toward the specified target. For flexion plus
supination (F(1,18) ¼ 11.33, p , .01, f ¼ .53), extension
plus supination (F(1,18) ¼ 18.89, p , .01, f ¼ .69), and flexion
plus pronation (F(1,18) ¼ 22.48, p , .01, f ¼ .75), older adults
exhibited greater trajectory deviations than did the young
adults. These data indicate that older adults were typically
less able than the young adults to coordinate the production
of torque in 2 degrees of freedom. There was no overall age-
related difference in trajectory error because the young group
exhibited much larger trajectory deviations for the supination
and pronation targets compared with the older group.
Investigations conducted in our laboratory indicate that
minimization of the trajectory error when acquiring targets in
these particular directions does not necessarily enhance
performance. Instead it may indicate that older adults adopt
a more ‘‘conservative’’ strategy in these instances by focusing
on torque production in the target axis only and avoiding the
coordination of torque in both degrees of freedom.
Electromyographic Variables
The overall patterns of muscular activation were very
similar between the young and older adults for the triceps
brachii, brachioradialis, and flexor carpi radialis (Figure 4).
This is apparent in the pattern of variation for each muscle in
the normalized amplitude (RMS) of EMG across the 8 target
locations. We observed statistically significant differences
only for the biceps brachii between the two groups, whereby
the older adults exhibited greater activation of this muscle
for extension targets (young adults, 3.16%; older adults,
5.81%; F(1,18) ¼ 6.87, p , .05, f ¼ .41) and pronation targets
(young adults, 2.24%; older adults, 4.86%; F(1,18) ¼ 7.62,
p , .05, f ¼ .44). In contrast, for the combined flexion and
supination target, the older group exhibited a tendency to
activate biceps brachii to a lesser degree than did the young
group (young adults, 51.8%; older adults, 34.6%; F(1,18) ¼
4.24, p ¼ .054, f ¼ .33). The pattern of results for the
normalized rate of burst onset (Figure 5) very closely
resembled that of the normalized EMG RMS. For the biceps
brachii, we observed a tendency for the older adults to exhibit faster onset rates during extension (F(1,18) ¼ 3.14, p ¼
.094, f ¼ .28) and pronation (F(1,18) ¼ 3.24, p ¼ .089, f ¼ .28)
while producing a significantly slower onset rate during
combined flexion and supination (F(1,18) ¼ 5.37, p , .05,
f ¼ .37), and a tendency to also do so when pure flexion
was required (F(1,18) ¼ 3.39, p ¼ .082, f ¼ .29).
A clear difference in the muscle recruitment patterns
adopted by the young and older adults was evident in the
RAPID FORCE GENERATION BY OLDER ADULTS
237
Figure 4. Normalized RMS of EMG were recorded from (A) biceps brachii, (B) triceps brachii, (C) brachioradialis, and (D) flexor carpi radialis during the visually
guided aiming task.*p , .05; ylarge effect size. Error bars display 95% confidence intervals. Flex ¼ flexion; Flesup ¼ flexion and supination; Supn ¼ supination; Extsup
¼ extension and supination; Extd ¼ extension; Extpro ¼ extension and supination; Pron ¼ pronation; Flepro ¼ flexion and pronation.
timing of muscle burst onsets (Figure 6). Across all 8 targets,
the average EMG burst onset time, relative to the time of
movement initiation, occurred later in the older adults. When
the action of a particular muscle opposed the required
direction of torque, the older adults tended to activate the
muscle later than did the young adults. In some instances, and
more frequently for the older group, the muscle burst onsets
occurred after the initiation of movement. Clearly, the
initiation of bursts at this latency was preceded by activity
in other muscles, primarily the agonists.
DISCUSSION
In a simple two-degree-of-freedom isometric aiming task,
we found that the ability of older adults to rapidly produce
torque to acquire a visually displayed target varied according
to the coordination demands of the required movement. The
dependence of the age-related difference in rapid torque
production on the direction of isometric torque application is
indicative of dysfunctions in muscular coordination in the
elderly. Older adults are apparently less able than young
adults to modulate the pattern of activity of groups of
muscles to match the demands posed by specific movement
tasks. This contention is supported by the differences we
observed between the young and older adults in both the
timing and magnitude of muscle activation.
Differences in Rapid Torque Production
The torque trajectories generated during the aiming task
displayed many of the primary characteristics observed in
other visually guided aiming tasks (16). This was an initial
impulse that preceded peak ‘‘velocity’’ and was typically
sustained for most of the duration of the ‘‘movement.’’ This
initial phase was frequently followed by one or several
relatively brief additional impulses. For the older adults, not
only was the peak rate of torque development lower but the
initial impulse during which the peak rate of torque
development was achieved represented a smaller proportion
of the total impulse required to reach the target. Consequently, the older adults exhibited a much greater number
of additional impulses or ‘‘submovements’’ in acquiring
a target. This may reflect a reduction in the quality of feedforward movement programming and an increased use of
on-line control to achieve targets, as has been frequently
reported for older adults when they perform aiming tasks
(13,15,30). A reduced ability to plan movements was also
implied by the increased reaction times displayed by older
adults in general, and in particular by the trend for the time
taken to initiate movements to combined extension and
supination to be most prolonged (10,13).
Combined extension and supination, for which the
performance of the older adults was most impaired, imposes
significant demands with respect to the coordination of
muscles acting about the elbow joint. In producing
combined extension and supination torques, triceps brachii
generates extension torque (31,32). Biceps brachii, in
producing a component of the supination torque, also
generates a flexion torque, and hence acts in opposition to
triceps brachii (31,32). This constraint presents a particularly
challenging problem for central nervous system control
238
BARRY ET AL.
Figure 5. The normalized mean rate of EMG burst onset were recorded from (A) biceps brachii, (B) triceps brachii, (C) brachioradialis, and (D) flexor carpi radialis
during the visually guided aiming task. *p , .05; ymoderate effect size. Error bars show 95% confidence intervals. The rate at which the EMG increased during the onset
of each burst was calculated on a second copy of the EMG channel low-pass filtered at 6 Hz. From the derivative of this time series, the peak rate of increase of the EMG
was identified for the first burst of activity in each muscle. The mean burst onset rate was defined as the mean slope of the EMG in the region about the peak rate of increase
of the EMG for which the magnitude of the signal was increasing (i.e., the average of the segment between zero crossings of the differentiated time series that contained the
peak positive value of this time series). The values thus obtained were normalized with respect to the peak amplitude of EMG achieved during an MVC. Flex ¼ flexion;
Flesup ¼ flexion and supination; Supn ¼ supination; Extsup ¼ extension and supination; Extd ¼ extension; Extpro ¼ extension and supination; Pron ¼ pronation; Flepro ¼
flexion and pronation.
when a task requires that extension and supination torque be
generated either rapidly or at near-maximal levels. The
particular difficulties exhibited by the older adults when
faced with conflicting task demands suggest that the aged
neuromuscular system may no longer retain the flexibility
necessary to rapidly recompose muscle synergies in
a functionally adaptive manner.
Deficiencies in motor planning and reliance on feedbackbased control pose crucial limitations on the ability of older
adults to execute rapid movements (13,33,34). Feed-forward
control of movements and the associated internal models are
frequently ascribed to the cerebellum (17). Potentially, the
deficits in aged motor function may be explained in part by the
cerebellar degeneration that is reported to occur in older adults
(24). Both animal (25) and human (26) studies have shown
that cerebellar degeneration has a close association with the
deterioration of motor skills. The findings of this study
suggest that in older adults the muscle activation patterns
devised based on feed-forward control are either inappropriate for the specific movement task, or that there is a decrease in
the flexibility with which muscle synergies may be modified
to implement the desired patterns of muscle activation.
Differences in Muscle Activation
Particular problems with the biceps brachii were implicated because the target location for which older adults
performed mostly poorly posed substantial coordination
demands on that muscle. In addition, the age-related differences in the EMG data were most prominently expressed
in the biceps brachii muscle. We observed increased levels of
activation for targets for which the biceps acted as an
antagonist, and the rate of biceps activation was lower for
older adults during movements for which it was the primary
agonist. During MVC, older adults exhibit less consistency in
the activation of the biceps brachii muscle than do young
adults, but such differences have not been observed for the
triceps brachii (35). Correspondingly, although it has been
reported that the elbow extensors and flexors undergo
a similar degree of atrophy as part of the aging process, the
decrease in strength is markedly greater in the elbow flexors
(36). When they evaluated decreases in motor performance in
older adults, Shinohara and colleagues (37) recently
suggested that when there is a disproportionate change in
the force-generating capabilities of muscles involved in
a particular synergy, previously suitable combinations of
neural commands become less appropriate (37). In this
regard, the accentuated degeneration of biceps brachii may
act to disrupt the utility of the muscle activation patterns that
mediate the generation of torque about the elbow joint.
It is important to acknowledge, however, that although
a disproportionate change in the force-generating capabilities
of muscles involved in a particular synergy may disturb the
RAPID FORCE GENERATION BY OLDER ADULTS
239
Figure 6. Burst onset time of EMG were recorded from (A) biceps brachii, (B) triceps brachii, (C) brachioradialis, and (D) flexor carpi radialis during the visually
guided aiming task. *p , .05; ylarge effect size. Error bars display 95% confidence intervals. Burst onset time was quantified as the occurrence of the first burst for each
muscle expressed relative to the movement onset time. As noted in Methods, some burst detection criteria in addition to the 2.5 standard deviation threshold were also
imposed to enhance the reliable detection of EMG bursts. To ensure that all burst activity was identified, including that of low amplitude, the peak EMG value was
identified for each trial, in the period between target presentation and target acquisition, and a threshold was established at 20% of this value. This threshold was
substituted in place of the mean plus 2.5 standard deviations criterion, and the burst detection procedure was reapplied to those data in which a burst was not identified
by the primary algorithm. Any burst shorter than 20 ms yielded by either algorithm was discarded. The remaining bursts, for which the gap between multiple bursts of
a given muscle were less than 60 ms, were combined. A final test was imposed to ensure that the bursts thus identified had lasted at least 100 ms. Flex ¼ flexion; Flesup
¼ flexion and supination; Supn ¼ supination; Extsup ¼ extension and supination; Extd ¼ extension; Extpro ¼ extension and supination; Pron ¼ pronation; Flepro ¼
flexion and pronation.
formation of muscle activation patterns, it is also possible that
such changes may be well compensated for by the nervous
system but would still result in different patterns of EMG. The
possibility also exists that significant differences in the
magnitude of muscle activation were only identified in the
biceps brachii because this muscle may have been activated in
the most reproducible way across the participants and the trials.
Although not reported in the results, the coefficient of variation
of EMG activity across trials was very similar for young and
older adults and for all four muscles from which recordings
were made. Another alternative explanation for the observed
differences in biceps brachii EMG is the possibility of agerelated differences in the normalized EMG–force relationship
arising from changes in the muscle properties of older adults.
Ng and Kent-Braun (38) reported that older adults exhibit
increased levels of normalized EMG for the same relative force
at low percentages of MVC. It is very likely, however, that their
findings (38) are attributable to increased levels of agonistantagonist activation that are typical of the elderly. Therefore,
the observed differences in normalized EMG in the current
study may be expected to indicate genuine differences in the
degree of biceps brachii activation (38,39).
Implications
The ability to rapidly generate torque has been identified
as an important contributor to the decrements in the physical
capabilities of older adults (3–5) and may be a crucial factor
in the recovery of balance and the production of protective
responses when they fall (2,40). In addition to the loss
of muscle mass and the slowing of muscle contractile
properties, deterioration of the capacity to coordinate
muscles also appears to contribute to the reduced ability
to develop torque rapidly. Indirectly, therefore, deficiencies
in the coordination of muscles will impair the functional
capabilities of older adults by contributing to the age-related
reduction in the rate of torque development. Deficiencies in
muscle coordination may also be expected to directly impair
functional capabilities by rendering older adults less able to
produce the specific muscle actions required to perform
effectively the wide variety of movement tasks that are
encountered in everyday living.
Resistance training interventions tailored to enhance the
rapid development of force have been advocated as a key
strategy to improve the movement capabilities of older adults
(41). The observed deficits in the ability of older adults to
240
BARRY ET AL.
appropriately coordinate muscles when producing rapid
contractions suggests that a significant benefit of resistance
training interventions may be derived from adaptations in
muscular coordination. Adaptations that benefit the facility
for muscular coordination are likely to be a particularly
important consideration for resistance training interventions
prescribed to older adults, because there is an expectation that
the adaptations accrued through the repeated performance of
a small number of exercises will transfer to a broad range of
movement tasks. The capacity of older adults to modulate
patterns of muscle activation according to the requirements of
different functional tasks will determine the effectiveness of
the transfer of these adaptations.
ACKNOWLEDGMENTS
Supported by the Australian Research Council.
Presented in part at the 21st Annual Meeting of the Australian
Neuroscience Society, Brisbane, Australia, January 2001.
The authors thank Dave Perkins for constructing the testing device;
Robert Bryant for his assistance with the electronic equipment; Kirsten
Willms for her help in data processing; Jonathon Shemmell, Dr. James
Tresilian, and Dr. Guy Wallis for their advice and assistance in data
analysis; and the volunteers who so generously donated their time.
Address correspondence to Benjamin K. Barry, Department of Integrative Physiology, University of Colorado, Boulder, CO 80309-0354.
E-mail: [email protected]
REFERENCES
1. Hakkinen K, Hakkinen A. Muscle cross-sectional area, force production and relaxation characteristics in women at different ages. Eur J
Appl Physiol Occup Physiol. 1991;62:410–414.
2. Izquierdo M, Aguado X, Gonzalez R, Lopez JL, Hakkinen K. Maximal
and explosive force production capacity and balance performance in men
of different ages. Eur J Appl Physiol Occup Physiol. 1999;79:260–267.
3. Suzuki T, Bean JF, Fielding RA. Muscle power of the ankle flexors
predicts functional performance in community-dwelling older women.
J Am Geriatr Soc. 2001;49:1161–1167.
4. Foldvari M, Clark M, Laviolette LC, et al. Association of muscle power
with functional status in community-dwelling elderly women. J
Gerontol A Biol Sci Med Sci. 2000;55:M192–M199.
5. Bean JF, Kiely DK, Herman S, et al. The relationship between leg
power and physical performance in mobility-limited older people.
J Am Geriatr Soc. 2002;50:461–467.
6. Larsson L, Li X, Frontera WR. Effects of aging on shortening velocity
and myosin isoform composition in single human skeletal muscle cells.
Am J Physiol. 1997;272:C638–C649.
7. Hook P, Sriramoju V, Larsson L. Effects of aging on actin sliding speed
on myosin from single skeletal muscle cells of mice, rats, and humans.
Am J Physiol Cell Physiol. 2001;280:C782–C788.
8. Morgan M, Phillips JG, Bradshaw JL, Mattingley JB, Iansek R,
Bradshaw JA. Age-related motor slowness: simply strategic? J
Gerontol A Biol Sci Med Sci. 1994;49:M133–M139.
9. Darling WG, Cooke JD, Brown SH. Control of simple arm movements
in elderly humans. Neurobiol Aging. 1989;10:149–157.
10. Seidler RD, Alberts JL, Stelmach GE. Changes in multi-joint
performance with age. Motor Control. 2002;6:19–31.
11. Seidler RD, Stelmach GE. Reduction in sensorimotor control with age.
Quest. 1995;47:386–394.
12. Christou EA, Carlton LG. Old adults exhibit greater motor output
variability than young adults only during rapid discrete isometric
contractions. J Gerontol A Biol Sci Med Sci. 2001;56A:B524–B532.
13. Yan, JH, Thomas JR, Stelmach GE. Aging and rapid aiming arm
movement control. Exp Aging Res. 1998;24:155–168.
14. Light KE, Spirduso WW. Effects of adult aging on the movement
complexity factor of response programming. J Gerontol A Biol Sci Med
Sci. 1990;45:P107–P109.
15. Cooke JD, Brown SH, Cunningham DA. Kinematics of arm movements in elderly humans. Neurobiol Aging. 1989;10:159–165.
16. Meyer DE, Abrams RA, Kornblum S, Wright CE, Smith JE. Optimality
in human motor performance: Ideal control of rapid aimed movements.
Psychol Rev. 1988;95:340–370.
17. Kawato M. Internal models for motor control and trajectory planning.
Curr Opin Neurobiol. 1999;9:718–727.
18. Mussa-Ivaldi FA. Modular features of motor control and learning. Curr
Opin Neurobiol. 1999;9:713–717.
19. Holdefer RN, Miller LE. Primary motor cortical neurons encode
functional muscle synergies. Exp Brain Res. 2002;146:233–243.
20. Schneider C, Devanne H, Lavoie BA, Capaday C. Neural mechanisms
involved in the functional linking of motor cortical points. Exp Brain
Res. 2002;146:86–94.
21. Windhorst U, Burke RE, Dieringer N, et al. What are the outputs of
motor behaviour and how are they controlled? In: Humphrey DR,
Freund H-J, eds. Motor Control: Concepts and Issues. New York:
Wiley; 1991:101–119.
22. Eisen A, Entezari-Taher M, Stewart H. Cortical projections to spinal
motoneurons: changes with aging and amyotrophic lateral sclerosis.
Neurology. 1996;46:1396–1404.
23. Peinemann A, Lehner C, Conrad B, Siebner HR. Age-related decrease
in paired-pulse intracortical inhibition in the human primary motor
cortex. Neurosci Lett. 2001;313:33–36.
24. Sjobeck M, Dahlen S, Englund E. Neuronal loss in the brainstem and
cerebellum—part of the normal aging process? A morphometric study
of the vermis cerebelli and inferior olivary nucleus. J Gerontol A Biol
Sci Med Sci. 1999;54:B363–B368.
25. Forster MJ, Dubey A, Dawson KM, Stutts WA, Lal H, Sohal RS. Agerelated losses of cognitive function and motor skills in mice are
associated with oxidative protein damage in the brain. Proc Natl Acad
Sci U S A. 1996;93:4765–4769.
26. Raz N, Williamson A, Gunning-Dixon F, Head D, Acker JD.
Neuroanatomical and cognitive correlates of adult age differences in
acquisition of a perceptual-motor skill. Microsc Res Tech. 2000;51:85–93.
27. Tomlinson BE, Irving D. The numbers of limb motor neurons in the
human lumbosacral cord throughout life. J Neurol Sci. 1977;34:213–219.
28. Hunter SK, White M, Thompson M. Techniques to evaluate elderly
human muscle function: a physiological basis. J Gerontol A Biol Sci
Med Sci. 1998;53:B204–B216.
29. Cohen J. Statistical Power Analysis for the Behavioural Sciences. New
York: Academic Press; 1969.
30. Pohl PS, Winstein CJ, Fisher BE. The locus of age-related movement
slowing: sensory processing in continuous goal-directed aiming. J
Gerontol B Psychol Sci Soc Sci. 1996;51:P94–P102.
31. Buchanan TS, Rovai GP, Rymer WZ. Strategies for muscle activation
during isometric torque generation at the human elbow. J Neurophysiol.
1989;62:1201–1212.
32. Jamison JC, Caldwell GE. Muscle synergies and isometric torque
production: influence of supination and pronation level on elbow
flexion. J Neurophysiol. 1993;70:947–960.
33. Haaland KY, Harrington DL, Grice JW. Effects of aging on planningand
implementing arm movements. Psychol Aging. 1993;8:617–632.
34. Seidler-Dobrin RD, Stelmach GE. Persistence in visual feedback
control by the elderly. Exp Brain Res. 1998;119:467–474.
35. Jakobi JM, Rice CL. Voluntary muscle activation varies with age and
muscle group. J Appl Physiol. 2002;93:457–462.
36. Klein CS, Rice CL, Marsh GD. Normalized force, activation, and
coactivation in the arm muscles of young and old men. J Appl Physiol.
2001;91:1341–1349.
37. Shinohara M, Latash ML, Zatsiorsky VM. Age effects on force
produced by intrinsic and extrinsic hand muscles and finger interaction
during mvc tasks. J Appl Physiol. 2003;95:1361–1369.
38. Ng AV, Kent-Braun JA. Slowed muscle contractile properties are not
associated with a decreased emg/force relationship in older humans.
J Gerontol A Biol Sci Med Sci. 1999;54:B452–B458.
39. Galganski ME, Fuglevand AJ, Enoka RM. Reduced control of motor
output in a human hand muscle of elderly subjects during submaximal
contractions. J Neurophysiol. 1993;69:2108–2115.
40. DeGoede KM, Ashton-Miller JA. Biomechanical simulations of forward
fall arrests: effects of upper extremity arrest strategy, gender and agingrelated declines in muscle strength. J Biomech. 2003;36:413–420.
41. Evans WJ. Exercise strategies should be designed to increase muscle
power. J Gerontol Biol Sci Med Sci. 2000;55A:M309–M310.
Received June 16, 2003
Accepted October 22, 2003
Decision Editor: John E. Morley, MB, BCh