Related Experiment Video
Updated: Mar 17, 2026
Physical Activity Measurement in Children Accepting Table Tennis Training
Published on: July 27, 2022
Allometrically Scaled Children's Clinical and Free-Living Ambulatory Behavior
Jongil Lim1, John M Schuna, Michael A Busa
11Department of Kinesiology, University of Massachusetts, Amherst, MA; 2Department of Kinesiology, Oregon State University, Corvallis, OR; and 3Pennington Biomedical Research Center, Baton Rouge, LA.
Purpose:
This study aimed to compare clinical and free-living walking cadence in school-age children and to examine how the allometric scaling of leg length variability affects objective ambulatory activity assessment.
Methods:
A total of 375 children (154 boys and 221 girls, 9-11 yr old) completed GAITRite-determined slow, normal, and fast walks and wore accelerometers for 1 wk. Dependent variables from clinical assessment included gait speed, cadence, and step length, whereas steps per day, peak 1-min cadence, and peak 60-min cadence were assessed during free living. Analogous allometrically scaled variables were used to account for leg length differences. Free-living times above clinically determined individualized slow, normal, and fast cadence values were calculated. Differences in dependent variables between sex and sex-specific leg length tertiles were assessed.
Results:
Clinically assessed cadence (mean ± SD) was 90.9 ± 15.2 (slow), 113.8 ± 12.9 (normal), and 148.9 ± 20.9 (fast) steps per minute, respectively. During free living, participants accumulated 8651 ± 2259 steps per day. Peak 1-min cadence was 113.4 ± 12.4 steps per minute and peak 60-min cadence was 60.1 ± 11.4 steps per minute. Allometrically scaling gait variables to leg length eliminated the previously significant leg length effect observed in both clinical and free-living gait variables but did not affect the observation that girls exhibited lower levels of free-living ambulatory behavior measured by mean steps per day. On average, all groups spent <15 min·d above clinically determined slow cadence; this was unaffected by leg length.
Conclusion:
Allometrically scaling gait variables to leg length significantly affected the assessment of ambulatory behavior, such that different leg length groups appear to walk in a dynamically similar manner. Leg length effects on free-living ambulatory behavior were also eliminated by implementing estimates of time spent above individualized cadence cut points derived from clinical gait assessment.
