Count- versus MAD-based accelerometry-assessed movement behaviors and associations with child adiposity and fitness.

Abstract

Estimations of time spent sedentary and in various physical activity intensities may vary according to data reduction methods applied. This study compared associations between children's accelerometer data and adiposity and fitness markers using open source (mean amplitude deviation, MAD) and proprietary (counts) data reduction methods. Complete-case accelerometer, adiposity (Body Mass Index z-score, waist circumference), and fitness (cardiorespiratory, musculoskeletal) data from 118 children (10.4±0.6years, 49% girls) were analyzed. Estimates of sedentary behavior, light-, moderate-, vigorous- (VPA), and moderate- to vigorous-intensity (MVPA) physical activity were calculated using count- and MAD-based data reduction methods. Linear regression models between time in movement behaviours and fitness and adiposity markers were conducted. Significant differences in estimates of time spent in all intensities were observed between MAD-based and count-based methods. Both methods produced evidence to suggest that sedentary behavior was detrimentally, and physical activity (any intensity) was beneficially, associated with waist circumference. MVPA and VPA were beneficially associated with fitness markers using both data reduction measures. Overall, findings suggest that estimates of sedentary time and physical activity were not comparable. However, the strength and direction of the associations obtained between the different data reduction methods and adiposity and fitness outcomes were fairly comparable, with both methods finding stronger associations for VPA compared to MVPA. This suggests that future studies may be able to pool data using different data reduction approaches when examining associations between activity and health risk factors, albeit with caution.