Welcome to poster #578 and winner of the ASB President’s Award for the best poster!!

We hope this webpage looks like a traditional printed poster when viewed on a computer, but it also has interactive elements like flip boxes, videos, and filterable galleries that allow you to see subsets of plots. So, hover over items and click/tap/swipe on them. FYI, default features are the quick main takeaways, while additional details involve a click/tap/swipe.

Self-pace treadmill controllers influence gait variability but not average walking speed

Cesar R. Castano

Department of Mechanical and Aerospace Engineering
University of Central Florida
castanocesar@knights.ucf.edu

Helen J. Huang

Department of Mechanical and Aerospace Engineering
Disability, Aging, and Technology (DAT) cluster
University of Central Florida
hjhuang@ucf.edu

Purpose: To investigate how self-pace treadmill controller sensitivity affects walking speed, step length, and step width on different different slopes (decline, level, and incline).

Hypothesis: More sensitive self-pace controllers would have greater walking speed fluctuations, step length variability, and step width variability on all slopes.

Background

Self-pace treadmills allow subjects to walk at their chosen walking speed while preserving natural gait fluctuations [1,2].

Equipment availability, laboratory space, and experiment design have led to a variety of custom self-pace controllers with different parameters, which may influence a subject’s gait.

controller sensitivity did not affect average gait parameters but did affect gait variability

All
speed
step length
step width
average
total variance

average speed

 Average speed was nearly identical regardless of controller sensitivity on each slope.

speed total variance

Speed variance increased as controller sensitivity increased (black asterisks).

average step length

Average step length was similar regardless of sensitivity within each slope.

step length total variance

Step length variance increased as self-pace controller sensitivity increased in each of the slopes (black asterisks).

average step width

Average step width was similar within each slope regardless of controller sensitivity.

step width total variance

Step width variance was not affected by controller sensitivity within each slope.

Curly brackets indicate significant differences between slopes within a given sensitivity shown as color coded asterisks, * low, * medium, * high (Tukey HSD, p<0.05). Square brackets with black asterisks indicate significant differences between sensitivities within a slope (Tukey HSD, p<0.05).

detrended variances were similar regardless of self-pace controller sensitivity, but speed-trend step length variances increased with higher sensitivity self-pace controllers

All
step length
step width
sum variances
stats

step length sum variances

Speed-trend step length variances were 34%-58% of the total step length variances. Within each slope, both speed-trend and total step length variances increased with increasing self-pace controller sensitivity. We used methods described in [3] for detrending variances.

step width sum variances

The speed-trend step width variance was negligible (<0.05%) and was also unaffected by controller sensitivity. We used methods described in [3] for detrending variances.
&nbsp

step length stats

Within a slope, detrended step length variance was not affected by controller sensitivity, whereas speed-trend step length variance increased with higher self-pace controller sensitivity (square brackets with *, Tukey HSD p<0.05). Detrended variances were smallest for level walking, regardless of which sensitivity level was used (curly brackets with * low, * medium, and * high sensitivities, Tukey HSD, p<0.05).

step width stats

Controller sensitivity had no significant effects on detrended or speed-trend step width variances.

final thoughts

Increasing self-pace controller sensitivity:

did not affect: 1) average walking speeds or gait parameters; 2) detrended variances; and 3) speed-trend step width variance
resulted in increased 1) speed variance, 2) total step length variance, and 3) speed-trend step length variance.

Gait variability during self-pace treadmill walking included additional speed-related components that increased with more sensitive controllers, which may affect the interpretation of gait variability metrics.

references

[1] Choi, J.-S., et al., 2017. J. Biomech. 65, 154–160
[2] Sloot, L.H., et al., 2014. Gait Posture 39, 478–484.
[3] Collins S.H. and Kuo A.D., 2013. PLOS ONE 8, e73597.

acknowledgements

Supported in part by NIH R01AG054621 to HJH, Gerontology Research Award from the Learning Institute for Elders (LIFE) at UCF to HJH, and a UCF ORC Fellowship to CRC.