The associations between carbohydrate and protein intakes with habitual sleep duration among adults living in urban and rural areas

Background & aims: Macronutrient intake was associated with habitual sleep duration, a recognized metabolic risk factor. However, results are not clear for populations with large diversities in dietary habits and urbanization levels. In this study, we aimed to evaluate the associations between relative macronutrient intakes and sleep duration, and the potential modification effect of urban–rural residence among adult Chinese population. Methods: We analyzed data from the China Health and Nutrition Survey 2009. In total, 9239 men and women aged 18 years or older were included. Dietary intakes were measured using a combination of a 3-day recall and household-based food weighing approach. Sleep duration was self-reported as hours of sleeping every 24 h. Results: Our participants have a mean sleep duration of 7.9 ± 1.2 h. After multivariate adjustment, higher intake of carbohydrate as percentage of total energy was associated with longer sleep duration (β ± SE = 0.43 ± 0.12, P < 0.001); and an inverse association was observed for the relative protein intake (β ± SE = −2.12 ± 0.47, P < 0.001). After stratification, the inverse association for protein was only detected in rural residents (β ± SE = −3.30 ± 0.66, P < 0.001), but not in urban participants (β ± SE = −0.44 ± 0.74, P = 0.55) (P for interaction = 0.002). Similarly, a marginal interaction with residence was observed for carbohydrate–sleep association. Conclusions: Our study suggested that the relative intakes of protein/carbohydrate were associated with sleep duration among adult Chinese, especially those in rural areas.