Figure 1 – Scatter-plots illustrating selected variability-related predictors of 6MW/TUG test performance. Linear best-fit lines have been added to indicate significance (Control Group only).
Brad
Manor, MS. and Li Li, PhD.
Louisiana
State University, Louisiana, United States
E-mail: lli3@lsu.edu
Gait variability is often quantified to evaluate the integrity of the human locomotor system. While the magnitude of variability is most often examined, the application of non-linear analysis to human walking has recently been employed to examine the structure of variability.
It is currently unknown if these properties of gait variability predict locomotion-based physical performance across different populations. Therefore, the purpose of this study was to determine the ability of measures related to the magnitude and structure of variability to predict locomotion-based performance in healthy older adults and those suffering from movement disorder associated with peripheral neuropathy (PN).
METHODS
Twelve individuals with diagnosed PN
(mean ± SD age = 67.0 ± 10.9 yr) and 12
aged-matched healthy controls (age = 69.1 ± 10.0 yr) were recruited.
Locomotion-based physical performance was assessed by two common field tests. The 6-Minute Walk (6MW) test was completed with cones placed 30m apart along a lighted hallway. Participants walked as far as possible within 6 minutes back and forth around the cones, and distance (m) was recorded. The Timed Up-and-Go (TUG) test was conducted with the participant seated and a cone placed 3m in front of the chair. The time (sec) taken to stand up, walk around the cone, and sit back down was recorded.
On a separate day, gait variability was assessed during treadmill
walking. 6MW distance was used to determine treadmill walking speed.
Participants completed 3min trials during which motion analysis (60 Hz) was
used to collect sagittal plane hip, knee, and ankle joint angle kinematics.
Stride duration variability (CoVar) was calculated by the ratio of
stride duration variability and the mean stride time from 30 strides. Joint
angle variability (JTvar) was calculated as the average standard deviation for
each joint from the ensemble curves of the same 30 strides (Li et al 2005).
The structure of gait variability was assessed by short- (λST) and long-term (λLT) Finite-Time Lyapunov Exponents associated with lower-extremity joint angles. λST and λLT were computed for each joint from ensemble curves of 100 consecutive strides (Dingwell & Cusumano 2000). JTvar, λST, and λLT values were averaged across lower-extremity joint to produce mean scores.
Potential group differences in locomotion-based physical performance (6MW, TUG), and gait variability (CoVar, JTvar, λST, λLT) were assessed with independent-samples T-tests. Pearson’s R used to examine the relationships between measures of locomotion-based physical performance and gait dynamics.
Table
1: Group means ± Standard deviation (SE)
for Locomotion-Based Physical Performance (i.e., Field Tests) and Gait Dynamics
|
|
Field Tests |
Magnitude/Structure of Stride-to-Stride Variability |
||||
|
6MW (m) |
TUG (sec) |
CoVar (sec) |
JTvar (deg) |
λST† |
λLT† |
|
|
CON |
530 ± 25 |
7.0 ± 0.5 |
2.1 ± 0.1 |
1.7 ± 0.1 |
3.0 ± 0.1 |
0.17 ± 0.01 |
|
PN |
391 ± 27* |
9.5 ± 0.6* |
2.9 ± 0.3* |
2.6 ± 0.2* |
3.1 ± 0.1 |
0.17 ± 0.01 |
Note:
* p < .05, † units = [<in[dj(i)]>/Stride]*100).
RESULTS AND DISCUSSION
The PN group exhibited reduced
locomotion-based physical performance as evidenced by a 26% decrease in 6MW (p <
.05) and a 36% increase in TUG (p < .05) (Table 1). Compared to controls, the
PN group exhibited increased magnitude of gait variability (CoVar, JTvar)
yet no difference in its structure (λST,λLT)
(Table 1).
In
the Control group, λST significantly predicted performance in
the 6MW (R = -.71, p < .01) and TUG (R = .77, p < .01) tests. Individuals
with decreased λST values tended to perform better. Controls
with less CoVar also tended to complete the TUG test in less time (R = .67, p =
.02) (Figure 1).
Within
the PN group, no measure of the magnitude or structure of variability predicted
locomotion-based physical performance.
In healthy older adults, the
magnitude and structure of variability can predict performance in
locomotion-based function. The high correlation between λST and both the 6MW and TUG suggests that this
property may more closely reflect locomotor system integrity.
The breakdown of correlation between measures of gait variability physical performance in the PN group points to the heterogeneity of this population. While gait variability may differentiate between individuals with PN and healthy older adults, additional research is needed to examine other PN-related alterations that lead to decreased locomotor performance.
Dingwell, J.B., Cusumano, J.P. (2000). Chaos,10, 848-863.
Li, L., al. (2005). Hum Mov Sci. 24, 257-67.
Figure 1 – Scatter-plots illustrating selected variability-related predictors of 6MW/TUG test performance. Linear best-fit lines have been added to indicate significance (Control Group only).