TRACKING OF PHYSICAL ACTIVITY FROM AGES 4 TO 17 IN AN ANGLOAND MEXICAN-AMERICAN COHORT

PURPOSE We assessed physical activity (PA) tracking in a Euro- and Mexican-American cohort from ages 4 to 17 using both objective and subjective measures. METHODS 1 Objective PA data were obtained from direct observation in children's homes using BEACHES (12 occasions; ages 4–7), CALTRAC accelerometer (4 occasions; ages 11–12), and CSA accelerometer (4 occasions; ages 15–17). Listwise analyses for objective measures included 147 children (N=76 boys, 71 girls). Log transformations of CALTRAC and CSA measures corrected for skewed distributions. Best Linear Unbiased Predictors (BLUPs) were computed for each measure to accommodate subjects with missing data. Partial correlations were conducted separately for boys and girls, controlling for ethnicity. RESULTS 1 Significant correlations (i.e., tracking) among the three PA measures (i.e., across the three time periods) were not found for either gender. METHODS 2 Subjective PA data were obtained via Physical Activity Recall (PAR) interviews in children's homes on 4 occasions (2 interviews, 6-months apart at ages 11–12, and 2 interviews, 6-months apart at ages 16–17). Listwise analyses for PAR included 177 children (N=88 boys, 89 girls). BLUPs were computed for each measurement period. Partial correlations were conducted separately for boys and girls, controlling for ethnicity. RESULTS 2 Significant relationships in self-reported PA across time periods for both boys (r=.372, p = .000) and girls (r=.311, p = .000) were found. CONCLUSIONS Similar to other studies, PA tracking was higher using subjective than objective measures. Studying the tracking of PA as children move from early childhood through adolescence faces numerous challenges, including (1) rapid changes in the type of PA and its contexts (e.g., location) as children age, and (2) the need for technological advancements that will provide a valid samples of PA at different developmental levels. Supported by HL52449