Is sleep duration associated with overweight/obesity in Indigenous Australian adults?

Background: Associations between high BMI and sleep duration and chronic illness are recognised. Short sleep is an accepted predictor of high BMI for children, including Indigenous Australian children. Short sleep has also been associated with high BMI in Australian adults, although not specifically in Indigenous Australian adults. This study aims to determine whether the relationship between sleep duration and BMI observed in non-Indigenous adults holds for Indigenous adults. Methods: Data collected from 5204 non-Indigenous and 646 Indigenous participants aged over 18 years in a nationally representative Australian Health Survey 2011-2013 were analysed. Sleep duration was self-reported as the time between going to bed and time waking up; BMI was derived from measurement and categorised into normal weight (BMI = 18.5-24.9) and overweight/obese (BMI ≥ 25). Logistic regression was performed for the non-Indigenous and Indigenous groups separately to examine the association between sleep duration and BMI in each group. Results: Proportionally more Indigenous people were classified as overweight/obese than non-Indigenous (χ2 = 21.81, p < 0.001). Short sleep was reported by similar proportions in both groups (Indigenous 15% vs non-Indigenous 17%) whereas long sleep of > 9 h was reported by proportionally more Indigenous than non-Indigenous people (41% vs 26%). Without accounting for possible confounders, the association between sleep duration and BMI for the Indigenous group was not significant but a possible dose-response relationship was evident, with the odds of overweight/obesity being greatest for those who typically slept < 7 h (OR = 1.77, 95% CI 0.38-3.94) and < 6 h (OR = 1.55, 95%CI = 0.58-4.14). The same model for the non-Indigenous group was significant, with the odds of overweight/obesity being greatest for those who typically slept < 6 h (OR = 1.67, 95%CI 1.25-2.25). The risk of overweight/obesity diminished for both groups with sleep > 7 h. Accounting for a range of socioeconomic and personal confounders attenuated the strength of these relationships marginally. Conclusion: Adding to reports relating sleep duration and BMI for Australian adults, this study provides evidence for an inverse relationship in non-Indigenous adults and suggests a similar trend for Indigenous adults. This trend was non-significant but is consistent with previous results for Indigenous children.