Abstract
BACKGROUND: The extent to which income setting or rural and urban environments modify the association between sleep and obesity in young children is unclear. The aims of this cross-sectional observational study were to (i) describe and compare sleep in South African preschool children from rural low-income (RL), urban low-income (UL) and urban high-income (UH) settings; and (ii) test for associations between sleep parameters and body mass index (BMI). METHODS: Participants were preschoolers (5.2 ± 0.7y, 49.5% boys) from RL (n = 111), UL (n = 65) and UH (n = 22) settings. Height and weight were measured. Sleep, sedentary behaviour and physical activity were assessed using accelerometery. RESULTS: UL children had higher BMI z-scores (median: 0.39; interquartile range: - 0.27, 0.99) than the UH (- 0.38; - 0.88, 0.11) and RL (- 0.08; - 0.83, 0.53) children (p = 0.001). The UL children had later bedtimes (p < 0.001) and wake-up times (p < 0.001) and shorter 24 h (p < 0.001) and nocturnal (p < 0.001) sleep durations than the RL and UH children. After adjusting for age, sex, setting, SB and PA, for every hour less sleep obtained (24 h and nocturnal), children were 2.28 (95% CI: 1.28-4.35) and 2.22 (95% CI: 1.27-3.85) more likely, respectively, to belong to a higher BMI z-score quartile. CONCLUSIONS: Shorter sleep is associated with a higher BMI z-score in South African preschoolers, despite high levels of PA, with UL children appearing to be particularly vulnerable.