Low sleep and low socioeconomic status predict high body mass index: a 4-year longitudinal study of Australian schoolchildren.

OBJECTIVE: The objective of this study was to examine the longitudinal relationships between body mass index (BMI), sleep duration and socioeconomic status (SES) in a 4-year cohort of 939 children aged 7-12 years.
METHODS: Children and their mothers completed an annual questionnaire to assess usual weekday sleep and wake times, amount of sleep, physical activity, parental education and school SES. 93% of children were enrolled (939/1010) and retention was 88%, 83% and 81% in consecutive years. Height and weight were measured annually.
RESULTS: BMI increased with decreasing amount of sleep and less sleep predicted greater International Obesity Task Force measures of obesity and overweight. In all 4 years, after controlling for baseline BMI, low SES was a significant predictor of high BMI. Children in the upper tertile of sleep in year 1 had a 2.3 kg lower weight gain (standard error [SE]: 0.5) between years 1 and 4 (P < 0.0001) than children in the lower tertile of sleep and a 0.45 kg m(-2) lower increase in BMI (SE: 0.15) (P = 0.004). The difference between children with consistently low and high sleep duration over 4 years was 1 BMI point. Those with the lowest BMI were the children with both high SES and high sleep duration. PA was not associated with BMI.
CONCLUSIONS: Both low SES and short sleep duration predict obesity risk in children after controlling for baseline BMI and this trend becomes stronger as children enter adolescence. Obesity prevention should include a sleep promotion component and this may be more beneficial to children of low SES and/or socially disadvantaged backgrounds.