Temporal patterns in accelerometer-measured physical activity (PA) among U.S. youth: Differences by weight status

Purpose: To describe accelerometer-measured PA by time of day, sex, & weight status in a nationally representative sample of youth.

Methods: Included youth (6-19 years) from the 2003-6 National Health & Nutrition Examination Surveys had at least one 10+ hour weekday of accelerometer data (N=4852; 50% male). Temporal patterns were represented by segments that broadly reflect before (6-9am), during (9am-3pm), after school (3-6pm), & evening (6-9pm) time. Weight status groups were: Normal (5th-84th %ile), overweight (85th-94th %ile), & obese (≥95th %ile). Statistics for mean counts per minute (cpm) were expressed for each time segment across all valid weekdays. Covariates were age, race/ethnicity, & sex. Sample weights accounted for the complex survey design.

Results: From 6-9am, the count rate for overweight youth was higher than obese youth [mean cpm(SE)=212.7(15.1) vs. 186.5(11.5); p<.05]. From 9am-3pm, the count rate for normal weight youth was higher than obese youth [mean cpm(SE)=453.1(6.3) vs. 402.4(9.8); p<.05]. The largest significant rate differences between normal weight & obese youth occurred from 3-6pm [mean cpm(SE)=627.7(11.6) vs. 504.9(13.4); p<.05] & from 6-9pm [mean cpm(SE)=506.7(15.2) vs. 414.7(15.4); p<.05]. Patterns & relationships were similar by sex.

Conclusions: Accelerometer count rate is highest in normal weight vs. overweight & obese youth throughout the day. Describing temporal PA patterns provide insight into times when interventions might be most effective. For example, between 3-6pm when there are the greatest differences between normal weight & overweight/obese groups. Such analyses also may help inform school policy decisions regarding physical education & recess time.