SenseCam-coded body positions associated with accelerometer-determined cadence
- Presented on May 21, 2014
Purpose: To quantify the proportion (%) of time spent in different body positions within incremental bands of accelerometer determined cadence (steps/min).
Methods: This was a secondary analysis of SenseCam camera images and ActiGraph GT3X+ accelerometer data concurrently worn during waking hours by 40 adults (70% male; 36±12 years-old; BMI=23±3 kg/m2) for 3-5 days. Body positions coded from images included: 1) sedentary, 2) standing still, 3) standing moving, 4) walking/running, 5) biking, or 6) changing position. For each participant, we calculated the proportion of time spent in each body position within previously published incremental cadence bands: 0 (non-movement during wearing time), 1-19 (incidental movement), 20-39 (sporadic movement), 40-59 (purposeful steps), 60-79 (slow walking), 80-99 (medium walking), 100-119 (brisk walking), and 120+ steps/min (all faster locomotion). Means and 95% confidence intervals (bootstrapped) were computed to describe the sample-level proportion of time spent in each body position within each cadence band.
Results: Sedentary behavior was more frequent at lower cadences, M=78% (CI: 72-81%) at 0 steps/min and M=43% (CI: 39-49%) at 1-19 steps/min, and less frequent at higher cadences, M=1% (CI: 0-4%) at 100-119 steps/min and M=1% (CI: 0-3%) at 120+ steps/min. Standing (still or moving), walking/running, biking, and changing position were more frequent at higher cadences, with walking/running most frequently identified, M=71% (CI: 63-79%) at 80-99 steps/min, M=90% (CI: 83-95%) at 100-119 steps/min, and M=87% (CI: 70-95%) at 120+ steps/min.
Conclusions: Further research is needed to evaluate the validity of cadence as a simple form of pattern analysis for identifying body positions.
- Mahara Proenca
- John Schuna
- Jacqueline Kerr
- Simon Marshall
- Catrine Tudor-Locke
ISBNPA 2014 Annual Conference