24-hour Movement In Early And Late Pregnancy: A Compositional Data Approach

Physical activity (PA), sedentary behavior, and sleep during pregnancy have been separately associated with pregnancy outcomes. However, these behaviors are not independent within the 24-hour day and require a compositional data approach, which has not been applied to these behaviors during pregnancy. PURPOSE: To describe time spent in PA, sedentary behavior, and sleep within the 24-hour day during early and late pregnancy using a compositional data approach. METHODS: We used data from the control arm of the GLOW study, a randomized controlled trial of a lifestyle intervention in pregnant women with overweight and obesity. At 10 and 32 weeks gestation, waking movement was measured using wrist-worn accelerometers for 7 days and average nightly sleep duration was self-reported. Women with ≥4 days of ≥10 hours of accelerometer wear and available sleep data at both time points were included (n = 134). Waking movement was categorized as PA (light intensity and higher) or sedentary behavior (cut point: 2860 vector magnitude counts). Time spent in PA, sedentary behavior, and sleep was averaged across all valid days and normalized to 24 hours. We used compositional means and compositional variation matrices to describe the composition of the 24-hour day in early and late pregnancy. RESULTS: The mean relative amount of time spent in PA, sedentary behavior, and sleep was 19.02%, 44.04%, and 36.93% of the 24-hour day in early pregnancy, and 20.60%, 43.24%, and 36.16% in late pregnancy, respectively. The mean relative amount of time spent in PA was 1.58% (absolute) greater (23 minutes) in late pregnancy than in early pregnancy, with a corresponding lower mean relative amount of time spent in sedentary behavior (0.80% lower, 11 minutes) and sleep (0.77% lower, 11 minutes). Log-ratio variances indicated the highest proportional co-dependence was between sedentary behavior and sleep (0.054 in early pregnancy, 0.060 in late pregnancy). CONCLUSION: Pregnant women with overweight and obesity spent the largest relative portion of the 24-hour day in sedentary behavior. Our findings suggest that from early to late pregnancy relative time spent in PA within the 24-hour day increases and relative time spent in sedentary behavior and sleep within the 24-hour day decreases in this population. Funding: NIH grants K99HD100585 and R01HD073572