© 2019, Springer Nature Limited. Background/Objectives: Whether variation in sleep and physical activity explain marked ethnic and socioeconomic disparities in childhood obesity is unclear. As time spent in one behaviour influences time spent in other behaviours across the 24-hour day, compositional analyses are essential. The aims of this study were to determine how ethnicity and socioeconomic status influence compositional time use in children, and whether differences in compositional time use explain variation in body mass index (BMI) z-score and obesity prevalence across ethnic groups. Methods: In all, 690 children (58% European, 20% Māori, 13% Pacific, 9% Asian; 66% low-medium deprivation and 34% high deprivation) aged 6–10 years wore an ActiGraph accelerometer 24-hours a day for 5 days yielding data on sedentary time, sleep, light physical activity (LPA) and moderate-to-vigorous physical activity (MVPA). Height and weight were measured using standard techniques and BMI z-scores calculated. Twenty-four hour movement data were transformed into isometric log-ratio co-ordinates for multivariable regression analysis and effect sizes were back-transformed. Results: European children spent more time asleep (predicted difference in minutes, 95% CI: 16.1, 7.4–24.9) and in MVPA (6.6 min, 2.4–10.4), and less time sedentary (−10.2 min, −19.8 to −0.6) and in LPA (−12.2 min, −21.0 to −3.5) than non-European children. Overall, 10% more sleep was associated with a larger difference in BMI z-score (adjusted difference, 95% CI: −0.13, −0.25 to −0.01) than 10% more MVPA (−0.06, −0.09 to −0.03). Compositional time use explained 35% of the increased risk of obesity in Pacific compared with European children after adjustment for age, sex, deprivation and diet, but only 9% in Māori and 24% in Asian children. Conclusions: Ethnic differences in compositional time use explain a relatively small proportion of the ethnic differences in obesity prevalence that exist in children.
Do differences in compositional time use explain ethnic variation in the prevalence of obesity in children? Analyses using 24-hour accelerometry