| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 85 No. 5 May 2001, pp. 1923-1931
Copyright ©2001 by the American Physiological Society
Biomotion Laboratory, Massachusetts General Hospital Department of Orthopaedics, and MGH Institute of Health Professions, Boston, Massachusetts 02114
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
McGibbon, Chris A. and
David E. Krebs.
Age-Related Changes in Lower Trunk Coordination and Energy
Transfer During Gait.
J. Neurophysiol. 85: 1923-1931, 2001.
The effects of aging on lower trunk
(trunk-low-back joint-pelvis) coordination and energy
transfer during locomotion has received little attention; consequently,
there are scant biomechanical data available for comparison with
patient populations whose upper body movements may be impaired by
orthopaedic or neurologic disorders. To address this problem, we
analyzed gait data from a cross-sectional sample of healthy adults
(n = 93) between 20 and 90 yr old (n = 44 elderly, >50 yr old; n = 49 young, <50 yr old).
Gait characteristics of elders were mostly typical: gait speed of
elders (1.13 ± 0.20 m/s) was significantly (P = 0.007) lower than gait speed of young subjects (1.20 ± 0.18 m/s).
Although elders had less low-back (trunk relative to pelvis) range of
motion (ROM; P = 0.013) during gait than young
subjects, no age-related differences were detected in absolute trunk
and pelvis ROM or peak pitch angles during gait. Despite similar upper
body postures, there was a strong association between age and
pelvis-trunk angular velocity phase angle (r = 0.48, P < 0.001) with zero phase occurring at approximately
55 yr of age; young subjects lead with the pelvis while elderly
subjects lead with the trunk. Age related changes in gait speed and
low-back ROM were unable to explain the above findings. The
trunk-leading strategy used by elders resulted in a sense reversal of
the low-back joint power curve and increased (P = 0.013) the mechanical energy expenditure required for eccentric control
of the lower trunk musculature during stance phase of gait. These data
suggest an age-related change in the control of lower trunk movements
during gait that preserves upper body posture and walking speed but
requires a leading trunk and higher mechanical energy demands of lower trunk musculature
two factors that may reduce the ability to recover from dynamic instabilities. The behavioral and motor control aspects of
these findings may be important for understanding locomotor impairment
compensations in aging humans and in quantifying falls risk.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Locomotion in
bipedal primates, humans in particular, is unique in the sense that
control of the upright trunk is especially critical for maintaining
control of balance (Krebs et al. 1992
; MacKinnon
and Winter 1993; Nashner and McCollum 1985;
Winter 1987). The effects of natural aging may impact
locomotor function due to age-related degeneration of the
musculo-skeletal system (Frontera et al. 1991;
Grimby 1995; Tinetti and Ginter 1988;
Wolfson et al. 1995), sensory feedback systems
(Allum et al. 1997; Peterka et al. 1990;
Stelmach and Worringham 1985) and the brain's motor control structures (Rogers and Bloom 1985). Because the
upper body constitutes two-third body weight, subtle uncoordinated
movements of the upper body may affect the ability to recover from
dynamic instabilities.
Studies of coordination in human walking have focused primarily
on the lower limbs (Bianchi et al. 1998; Grasso
et al. 1998, 2000; Grillner 1981)
of young, healthy persons. Few studies have examined the upper body
during gait and have focused primarily on trunk kinematics in the young
and elderly (Krebs et al. 1992; Murray
1967; Opila-Correia 1990; Thorstensson et
al. 1984) or pelvis and trunk kinematics in young adults
(Stokes et al. 1989; van Emmerik and Wagenaar
1996). It is unknown what changes in pelvis-trunk (lower trunk)
interactions, if any, occur with aging; consequently, there are scant
biomechanical data available for comparison with patient populations
whose upper body movements may be impaired by orthopaedic or neurologic
disorders. We took as our primary hypothesis that aging in healthy
(asymptomatic) adults affects the kinematic coordination of the lower
trunk during gait, and we tested whether such alterations are modulated
by age-related changes in kinematic gait parameters, such as walking speed and range of movement (Himann et al. 1988;
Larish et al. 1988; Oberg et al. 1993;
Ostrosky et al. 1994; Sullivan et al. 1994).
Although examining the kinematics of the lower trunk can yield
information about how humans coordinate trunk and pelvis movements during gait, the kinematics alone cannot provide a mechanistic explanation for the observed behaviors. Information on muscle involvement in controlling the observed movements is thus required. Past studies have shown that mechanical energy analysis can provide useful information about muscle coordination and motor control (Aleshinsky 1986a-e; Bianchi et al.
1998; Prilutsky and Zatsiorsky 1994;
Prilutsky et al. 1996; Winter 1990;
Winter et al. 1990). Therefore we also examined the
energy transfers across the lower trunk via inverse dynamic analysis of
low-back joint moments during gait. We took as our secondary hypothesis
that alterations in sagittal plane lower trunk coordination with aging
is a result of alterations in flexor/extensor muscle coordination, as
reflected by changes in the mechanical energy transfer patterns. The
kinematic and kinetic results were then used to discuss behavioral and
motor control aspects of lower trunk coordination with aging.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Subjects and gait analysis procedures
The sample consisted of 93 healthy adults (33 males and 60 females) ranging in age from 20.2 to 88.8 yr old (49.1 ± 21.9 yr old, mean ± SD). All subjects were considered healthy at the time of data collection, having no orthopaedic or neurologic disorders affecting locomotion or balance as determined by a staff physician and physical therapist. All subjects provided informed consent prior to inclusion in the study according to institutional policy on human research.
Subjects performed several (range, 1-5; mode, 3) level walking trials
at their self-selected speed on a 10-m carpeted walkway in their bare
feet. The 10-m walkway provided ample distance for subjects to reach
their steady-state gait pattern, normally attained before three steps
(~2 m) from a stand still (Miller and Verstraete 1996). Four Selspot II (Selective Electronics, Partille,
Sweden) optoelectric cameras were used to track body segment fixed
arrays of infrared light-emitting diodes (irLEDs), and two Kistler
(Kistler Instruments, Winterthur, Switzerland) piezoelectric force
plates were used to record ground reaction forces. Each body segment array consisted of three to five irLEDs imbedded in a flat plastic disk
and were securely strapped to the mid-sections of 11 body segments
(both feet, shanks, thighs and arms, and pelvis, trunk, and head) using
neoprene and Velcro straps. The camera configuration gave an
approximate viewing volume of 2 × 2 × 2 m3 in the center of the 10-m walkway. Force
plates were unobtrusively (carpet covered) situated in the center of
the viewing volume.
Data were sampled at 150 Hz, and raw kinematic data were filtered using
a low-pass Butterworth filter (4th order, 6-Hz cutoff). irLED data were
processed using TRACK (Massachusetts Institute of Technology,
Cambridge, MA) software to give three-dimensional rotations and
translations of each body segment as previously described (Riley
et al. 1990). Body segment mass, center of mass, and mass
moment of inertia properties were computed from regression equations
(McConville et al. 1980) using subject-specific
anatomical measurements. For this analysis, only upper body segment
data (pelvis, trunk, arms, and head) were required to test our
hypotheses. We defined the trunk as a rigid segment between
the neck joint (C1-C2) and
low-back joint
(L4-L5), and the
pelvis as a rigid segment between the low-back joint and the
hips. We use the term lower trunk to refer to the
trunk-low-back joint-pelvis as a system.
Upper body segments were defined using a static standing pointing
procedure developed by Riley et al. (1990). Briefly,
subjects donned the irLED arrays and stood with feet 30 cm apart.
During these "static" pointing trials, the tester used two
hand-held arrays to point to landmarks that were subsequently used to
define each segment's anthropometrics and embedded coordinate system. The low-back joint was defined at the
L4-L5 level as the
mid-point of a line joining the left and right superior iliac crests,
and the neck joint was defined at the
C1-C2 level as the center
of a circle circumscribing the neck below the tragus of the ear
(Hutchinson et al. 1994; Riley et al.
1990). The segment coordinate system for each body segment,
defined from the static pointing trials, were then transformed into the
segment's irLED array coordinate system; measurement of the irLED
arrays' position and orientation during arbitrary movements enabled
estimation of the segments' skeletal kinematics. Precision of array
measurements is assessed at 1 mm in translation and <0.1° in
orientation (Antonsson and Mann 1989).
Data analysis and key variables
INDEPENDENT VARIABLES: AGE AND GENDER.
Two groups of subjects were created from the sample according to age,
thus creating two independent variables (age and gender), each with two
levels. The majority of young subjects were in their 20s and 30s, and
the majority of elderly subjects were in their 60s and 70s: thus a
cut-point of 50 yr in age was selected (n = 49 < 50 yr old, and n = 44 
50 yr old) to ensure the
mean age of the groups was adequately separated.
DEPENDENT VARIABLES: PHASE SHIFT AND MECHANICAL ENERGY TRANSFER. Consistent with our hypotheses, primary dependent variables for quantifying lower trunk coordination consisted of 1) the angular velocity phase angle between the pelvis and trunk (lower trunk phase shift) and 2) mechanical energy transfer between the pelvis and trunk.
Lower trunk phase shift was determined from peak-to-peak analysis of pelvis and trunk angular velocities within the subject's gait cycle. Because there was only one set of force plates, successive heel strike times for the same foot could not be registered; therefore the gait cycle was determined from peak knee flexion to peak knee flexion of the same limb, depending on which leg (left or right) was visible for both events (Fig. 1). Phase shift was expressed in degrees relative to the gait cycle (1 cycle = 2
radians = 360°). There were typically three peaks in pelvis and
trunk angular velocity (Fig. 1): in the early gait cycle
(
1, negative peaks approximately around heel
strike), mid gait cycle (2, positive peaks
approximately around mid-late stance), and late gait cycle (3, negative peaks approximately around heel
strike of the opposite foot). These three values were averaged to
arrive at a mean phase shift () during the cycle. Phase shift angles
were computed such that a positive value indicated that the trunk was
leading the pelvis.
|
|
(1) |
M
t and p were
angular velocities of the trunk and pelvis, respectively. The amount of
mechanical energy expended was denoted by
Ub, and was calculated by integrating the
net joint power curve over specific intervals of time
(t1 to
t2, described in more detail below).
|
(2) |
and others (Eng and Winter
1995; Judge et al. 1996; Prilutsky et al.
1996), a positive net power
(Pb > 0) means that joint
muscles are doing a net amount of positive work; therefore concentric muscle contractions control the observed movements. Conversely, a
negative net power (Pb < 0) means the
joint muscles are doing a net amount of negative work; therefore
eccentric muscle contractions control the observed movements. Thus our
analysis can separate movements into concentric and eccentric control
modes. Because the pelvis-trunk phase shift is a function of trunk and
pelvis angular velocities, the direction of phase shift (a leading
trunk or a leading pelvis) will affect the sign of the relative angular velocity term (t p) in Eq. 1, and hence
affect the control mode.
However, the energy possessed by a segment is only partially explained
by the muscle energy that controls it; energy can also be transferred
between segments without the use of muscles (for example, the shank
dynamics of an above knee prosthesis during swing phase of gait). The
transfer of energy (or flow of power) is determined from the proximal
and distal segment powers in Eq. 1. As described in detail
elsewhere (Aleshinsky 1986a-e), only when both segments
rotate in the same direction can there be a transfer of mechanical
energy between segments, thereby reducing the burden placed on the
muscles for controlling movements. However, no inter-segmental transfer
occurs if segments rotate in opposite directions (the muscles do all
the work). Thus our analysis also separates movements into transfer and
nontransfer modes. Partitioning joint energy expenditures into
concentric/eccentric and transfer/nontransfer modes (defined by the
time intervals t1 and
t2 in Eq. 2) can therefore be used to quantify the effects of alterations in lower trunk motor
control strategies during gait. We chose to analyze three energy
expenditure modes: U
|
COVARIATES: UPPER BODY POSTURE, RANGE OF MOTION, AND GAIT SPEED.
It has been documented in prior reports that upper body posture and
range of motion (Sullivan et al. 1994) and walking speed (Himann et al. 1988; Larish et al. 1988)
change with age and that walking speed in particular may influence
upper body posture (Thorstensson et al. 1984) and the
relative phase of the trunk and pelvis (van Emmerik and Wagenaar
1996). Therefore the following variables were documented for
use as covariates when testing the main effects of age and gender on
lower trunk coordination variables: 1) peak trunk angle,
2) peak pelvis angle, 3) peak low-back angle,
4) trunk range of motion (ROM), 5) pelvis ROM,
6) low-back ROM, and 7) gait speed. Peak and
range were documented for the trunk and pelvis pitch angles (in room
coordinates), as well as for the low-back flexion angle (trunk relative
to pelvis) during the gait cycle. Gait speed was calculated as the
anterior-posterior center of gravity displacement during the gait cycle
divided by cycle time and normalized to height.
STATISTICAL ANALYSIS AND HYPOTHESIS TESTING. All variables were evaluated within a single gait cycle and averaged over repeated trials for each subject (73 of the 93 subjects had data for 2 or more trials: 31 had data for 2 trials, 34 had data for 3 trials, 4 had data for 4 trials, and 4 had data for 5 trials). Trial-to-trial repeatability was assessed for dependent variables and covariates using a two-way mixed effects interclass correlation coefficient (ICC, where ICC > 0.80 was considered high repeatability). Subjects who had only one trial were included in the hypothesis testing, but excluded from repeatability tests. Associations between variables were evaluated with Pearson's product moment correlations, and between-groups differences were evaluated with two-way ANOVA using age and gender as independent variables. Covariates influences were assessed with partial correlation analysis and two-way analysis of covariance (ANCOVA). Interaction effects and homogeneity of variance were assessed for each main effects test. An alpha level of 0.05 was chosen for all statistical tests. SPSS for Windows (Version 8.0, SPSS, Chicago, IL) was used for all statistical analyses.
Procedures for testing hypotheses were as follows. First, main effects analysis was performed for covariates (treating them as dependent variables), as well as correlation analysis among covariates and between the covariates and age. Hypothesis one was then tested by two-way ANOVA (phase shift as dependent variable, and age and gender groups as the independent variables), and then by ANCOVA (gait speed or upper body angles as covariates) to assess whether the covariates explained the ANOVA main effects differences. Pearson correlations and partial correlations were also performed (with age as a continuous variable) as a secondary approach for testing hypotheses. Hypothesis two was tested similar to hypothesis one (except mechanical energy expenditures were the dependent variables).| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Subject characteristics and covariates
Young subjects were not significantly different in height (P = 0.095) and weight (P = 0.948) compared with old subjects (Table 1). There were nonsignificant age differences (P = 0.098) between males and females; however, males were taller (P < 0.001) and heavier (P < 0.001) than females (Table 1).
|
Young subjects walked faster per unit height (P = 0.007) compared with old subjects, and females walked faster per unit height (P = 0.007) compared with males (Table 2). Low-back flexion ROM was significantly greater in young subjects compared with old subjects (P = 0.021) but not significantly different (P = 0.893) between males and females (Table 2). There were no significant differences between old and young subjects or between males and females in peak trunk pitch angle (P = 0.470 and P = 0.230), peak pelvis pitch angle (P = 0.370 and P = 0.639), peak low-back flexion angle (P = 0.953 and P = 0.912), trunk ROM (P = 0.096 and P = 0.242), and pelvis ROM (P = 0.453 and P = 0.190).
|
Correlation analysis indicated weak-to-moderate significant
relationships between gait speed and age (r = 0.366,
P < 0.001), low-back ROM and age (r = 0.256, P = 0.013), and between low-back ROM and gait
speed (r = 0.267, P = 0.010). When
controlling for gait speed, ANOVA indicated no significant difference
between young and old subjects in low-back ROM (P = 0.095). Because gait speed and low-back ROM were statistically
different between gender and/or age groups, and had statistically
significant associations with age, they were used as covariates in the
main effects analysis of the primary dependent variables.
Trial-to-trial repeatability for these variables was found to be high (ICC = 0.970 for gait speed and ICC = 0.853 for low-back flexion ROM). The homogeneity of variance test was not violated (Levene's test, P > 0.05) for any of the above main effects tests, and there were no significant interactions (P > 0.05) in the two-way ANOVA tests.
Lower trunk phase shift
Lower trunk phase shift for age and gender groups are summarized
in Table 3. Average phase shift () was
significantly different between young and old subjects
(P < 0.001) but not different between males and
females (P = 0.452). On average, young subjects lead trunk angular motion with their pelvis, while old subjects lead pelvis
angular motion with their trunk. Phase shift was significantly associated with age (r = 0.480, P < 0.001), gait speed (r = 0.287, P = 0.005) and low-back ROM (r = 0.227, P = 0.029). Figure 3 shows the relationship
between phase shift and age.
|
|
To determine whether this relationship was controlled by gait speed and/or low-back ROM, partial correlation and analysis of covariance tests were conducted between phase shift and age controlling for gait speed and low-back ROM. Gait speed accounted for only a small proportion of the association between lower trunk phase shift and age; when controlling for gait speed the correlation remained significant (phase vs. age partial r = 0.420, P < 0.001), and the difference between young and old subjects remained significant (P < 0.001). Furthermore, controlling for low-back ROM was even less explanatory (phase vs. age partial r = 0.448, P < 0.001), which is not surprising due to the apparent dependence of low-back ROM on gait speed.
Repeatability was assessed for the three (1,
2, and 3)
peak-to-peak phase delay measures and was found to be high (ICC = 0.854). Trial-to-trial repeatability for average pelvis-trunk phase
shift angle was also found to be high (ICC = 0.911). The homogeneity of variance test was not violated (Levene's test, P > 0.05) for the main effects test, and there were no
significant interactions (P > 0.05) in the two-way
ANOVA or ANCOVA tests.
Low-back mechanical energy transfer
Mechanical energy expenditures of the low-back are summarized in
Table 3. Analysis of variance indicated that older subjects transferred
more eccentric muscle energy,
U
|
Trial-to-trial repeatability of the mechanical energy data were found
to be high (ICC = 0.915 for
U
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Our data suggest that a reversal in the angular velocity phase relationship of the trunk and pelvis occurs with aging in asymptomatic adults that is independent of walking speed and range of motion of the low back. Furthermore, this reversal in phase relationship parallels an apparent reversal of the muscle lengthening and shortening (eccentric and concentric) sequences of low-back flexor-extensor musculature, as indicated by the mechanical energy transfer patterns of the low back. Thus there appears to be two distinctive lower trunk coordination strategies used by individuals during preferred-speed gait (trunk-leading and pelvis-leading), where elderly subjects appear more likely to use a trunk-leading strategy than young subjects. Our data also suggest that elderly subjects increase the mechanical energy demands of their low-back musculature, and that this increase is a consequence of using the trunk-leading strategy.
Effects of aging on lower trunk coordination during gait
We found that the correlation between age and pelvis-trunk angular
velocity phase angle (phase shift) was moderately strong (r = 0.480, P < 0.001) with the
regression line crossing the zero phase line at approximately 55 yr of
age. The predominant lower trunk coordination strategy for younger
subjects was to lead their trunk with the pelvis (mean of 10°),
while the predominant strategy for elderly subjects was to lead their
pelvis with the trunk (mean of +7°). Of particular interest was that
the observed age-related differences in lower trunk coordination did
not appear to be mediated by forward velocity of the center of mass or
upper body posture: lower trunk phase shift correlated with gait speed
and low-back ROM, but controlling for those variables did not have a
large effect on the relationship between lower trunk phase shift and age.
Although we did detect an age- (but not gender) related difference in
low-back ROM, the difference between young and old subjects was small
(~1°) and could be statistically explained by differences in
walking speed. These data indicate that the upright posture of the
trunk was similar for young and old, and males and females, in this
study. Furthermore, magnitudes and range of trunk ROM were similar to
data reported by others (Krebs et al. 1992;
Murray 1967; Opila-Correia 1990;
Stokes et al. 1989; Thorstensson et al.
1984). Age and gender differences were found in gait speed, and
there was a significant association between age and gait speed. Van Emmerik and Wagenaar (1996) showed that within
healthy young individuals, the relative phase between trunk and pelvis
increased with faster walking. It should be noted that van
Emmerik and Wagenaar (1996) computed continous phase and
discrete phase relationships between the pelvis and trunk based on
angular displacements in the transverse plane. Nonetheless, in our
study, the relationship between gait speed and phase shift was
consistent with van Emmerik and Wagenaar (1996),
although the variation in gait speed across our subjects did not
account for much of the variance in phase shift (~8%), probably
because subjects in our study walked at only their preferred speed.
Although walking speed is a commonly used measure of gait performance
in elderly subjects (Himann et al. 1988; Larish
et al. 1988; Oberg et al. 1993; Ostrosky
et al. 1994), our results suggest that gait speed is not
appreciably affected by different strategies of lower trunk
coordination. This may be important if one considers the potential
consequences of the lower trunk coordination strategy used.
A gait style in which the pelvis leads the trunk would suggest a
predominant lower extremity controller schema; the CNS is using the
lower body to direct movements of the upper body. Conversely, a gait
style in which the trunk leads the pelvis would suggest a predominant
upper body controller schema; the CNS is using the massive upper body
to direct movements of the lower body. It stands to reason that the
trunk-leading coordination strategy may impact the ability to recover
from a trip or slip. Adequate lagging of the trunk behind the pelvis
would allow more time for the motor cortex to process sensory
information and issue corrective commands to stabilize the upper body
when the path of the lower extremities is disturbed. Leading with the
massive trunk may not allow a sufficient upper body stabilizing
recovery response. Because gait speed explained only a small portion of
the age-phase shift relationship, gait speed may not always be the best
predictor of falls risk. Pavol et al. (1999) showed that
elders who walk faster are less likely to recover from a trip and that
trunk orientation was not a good predictor of falls recovery. They
(Pavol et al. 1999) suggested that available power
reserves do not affect the ability to recover from fall. Our data offer
an explanation; the ability to generate rapid torque (power burst) may
not be sufficient for trip recovery response when the trunk leads the
pelvis by a substantial margin. Future studies should attempt to
quantify the influence of lower trunk coordination on falls recovery,
and determine specifically what (if any) magnitude of trunk lead
constitutes a threat to successful trip recovery.
The low-back joint power profiles in Fig. 4 suggest that elders
contract low-back muscles (flexor-extensor) eccentrically in the double
support phases and early single support phase and concentrically in
late single support phase, while younger subjects do the opposite. The
trunk-leading strategy used by elders resulted in an increase in the
mechanical energy expenditure of the low-back musculature; elderly
subjects expended more energy in eccentric control than younger
subjects, a difference that was explained by differences in lower trunk
phase shift. This finding suggests that elderly subjects rely more on
eccentric control of low-back muscles to regulate energy transfer
during the double limb support phase of gait (Fig. 4), which implies an
alteration in activation strategies of the nervous system (Enoka
1996). When accounting for the phase shift differences among
subjects, the younger subjects used more concentric muscle power
compared with elderly subjects, suggesting that elders reduce the
concentric energy output of trunk muscles to minimize the energy
transferred proximally to the trunk during the less dynamically stable
single support phase of gait.
Behavioral and motor control aspects of lower trunk coordination strategies
These data raise some important questions: Why would elders select an upper body, trunk-leading, control strategy, particularly given its potentially negative consequences? What is the mechanism that enables the trunk-leading strategy to occur without appreciably affecting upper body posture or walking speed in healthy adults?
Eccentric muscle activity is not as metabolically demanding as
concentric muscle activity (Willems et al. 1995), and
hence may be a compensatory pattern adopted by elders to minimize
low-back muscle fatigue. However, the mechanisms of muscle fatigue
under low contraction force conditions are not well understood, and recent evidence suggests that neural adaptations following
immobilization of muscles (which might translate into increased
sedentary lifestyle of the elderly) may actually increase muscle
endurance (Semmler et al. 2000), particularly in women
(Semmler et al. 1999). It is known, however, that
prolonged eccentric activity generates a higher concentration of plasma
creatine kinase that can induce significantly prolonged muscle pain
compared with that caused by concentric activity (Lund et al.
1998), and therefore eccentric efficiency is not a satisfactory
explanation. It is also likely that strength deficits of the lower
extremities are, at least, a contributing factor. Previous gait
analysis studies have shown increased hip power in healthy elders
compared with young subjects (Judge et al. 1996), and
increased hip and low-back power in disabled elders compared with
healthy elders (McGibbon et al. 2001), suggesting that a
trunk-leading strategy may be a response to lower extremity impairment.
In essence, these past studies suggest that pelvis and trunk
musculature is used to advance, or pull, the leg into swing phase when
the lower extremity muscles (ankle plantar flexors and knee extensors)
are weakened from aging and/or disability. Although strength was not
measured in our sample, past studies support our expectation that our
elderly subjects were weaker than our younger subjects (Frontera
et al. 1991).
What facilitates this compensatory response? The firing sequences for
leg muscles in automated movements such as gait are, at least in
animals (Golubitsky et al. 1999), issued via central pattern generators (CPGs). It has been proposed that the functional organization of CPGs constitutes a network of coupled oscillators that
produce coordinated multi-joint movements (Grillner et al. 1995; Pearson 1993). Experimental data have
shown that the phase coupling of kinematic patterns of the lower
extremity segments are highly invariant at different walking speeds
(Bianchi et al. 1998) and with different trunk postures
(Grasso et al. 2000). Walking backward has been shown to
preserve the kinematic profiles of segmental angle elevations but with
a reversal in the phase coupling of limb segments and patterns of
muscle activity (Grasso et al. 1998). Grasso et
al. (1998) and Grillner (1981) hypothesized a
reorganization of motor programming such that the CPGs recruited for
backward movements were the same series of coupled oscillators but with
a sign reversal of the phase coupling.
Although theories surrounding the role of CPGs and their interaction
with high- and mid-level control centers of the brain remain
controversial (Mitz and Winstein 1993), the reversal in lower trunk phase without a statistically significant alteration in
gait speed and upper body posture (and in the presence of normal vestibulocerebellar function) suggests that reorganization of lumbo-sacral muscle phasic oscillators may occur in response to lower
extremity weakness (or other impairments) in otherwise asymptomatic elderly adults, to maintain proper attitude control of the trunk and
overall gait function. More detailed studies that use electromyography would be required to test the above hypothesis.
Limitations and conclusions
Studies of locomotion in aging individuals are confounded by aging effects of all body systems including neurologic, musculo-skeletal, cardiac, and respiratory systems. Although the subjects in our study were healthy and free of orthopaedic, cardiac-respiratory, and neurologic disorders, we cannot preclude that subtle age-related changes in multiple body systems influenced their locomotor abilities. We only examined the mechanics of the sagittal plane, and therefore it is unknown what age-related changes, if any, occur in the frontal and transverse planes. Our model of the upper body consisted of rigid segment models of the pelvis, trunk, arms, and head; treatment of the trunk and perhaps the arms as multi-jointed segments may also yield interesting information. It should also be understood that mechanical energy analysis, based on inverse dynamic models, provides only indirect muscle action data. Energy transfer calculations for the low back are a function of low-back net moments, which can be calculated top-down or bottom-up. Although there are no published studies that suggest one method is better, preliminary comparisons from our lab show very close agreement between the two methods (unpublished observations); a top-down approach was chosen here to enable calculation of lower trunk energy transfers without the use of force plates.
In conclusion, our data suggest that lower trunk coordination strategy during gait may be altered in humans of advanced age and is likely a response to subtle age-related changes in locomotor function. Future studies comparing lower trunk coordination between healthy subjects and patients with neurologic (e.g., cerebellar disease, Parkinsons, and spinal cord injury) and musculoskeletal (e.g., arthritis, low-back pain, and muscle wasting) dysfunction are warranted and may shed light on this previously unreported characteristic of gait with aging. The behavioral and motor control aspects of these findings may be important for understanding locomotor impairment in aging humans and in quantifying falls risk.
| |
ACKNOWLEDGMENTS |
|---|
This work was supported by National Institute on Aging Grant R01-AG-11255.
| |
FOOTNOTES |
|---|
Address for reprint requests: C. A. McGibbon, Massachusetts General Hospital, Biomotion Laboratory, Boston, MA 02114 (E-mail: cmcgibbon{at}partners.org).
Received 18 September 2000; accepted in final form 5 February 2001.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
This article has been cited by other articles:
|
|
 |
|
  D. E Krebs Invited Commentary Physical Therapy, December 1, 2007; 87(12): 1667 - 1667. [Full Text] [PDF]
|
|
|
|
 |
|
 B. K. Barry and R. G. Carson The Consequences of Resistance Training for Movement Control in Older Adults J. Gerontol. A Biol. Sci. Med. Sci., July 1, 2004; 59(7): M730 - M754. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||
