Sleep duration and chronic conditions

Abstract

Noncommunicable diseases (NCDs), including obesity, diabetes, hypertension, and heart disease, are the leading cause of mortality globally. NCD etiology remains unclear and can only be partly explained by known risk factors including diet and physical activity. Sleep, as an important physiologic process, is associated with neurobiological functions. Previous observational studies have shown relationships of sleep duration with NCDs including obesity, diabetes, hypertension, and anemia. However, whether the associations are causal, and should be translated into public health action remains unclear.

To assess the role of sleep duration in factors potentially driving NCDs, i.e., obesity, diabetes, hypertension, and anemia, I used two different methods, observational and Mendelian randomization (MR) studies. In the population-representative Hong Kong birth cohort “Children of 1997”, I assessed the associations of sleep duration with body mass index (BMI) z-score and obesity/overweight at ~11 years of age cross-sectionally, and examined longitudinal associations of sleep duration with subsequent BMI z-score and obesity/overweight from 11~16 years. Then I assessed the associations of sleep duration at ~17.5 years with fasting plasma glucose (FG)/ glycated hemoglobin (HbA1c) and hemoglobin (Hgb)/ hematocrit (Hct). Given puberty is a key developmental phase, the association of age at puberty with sleep duration was assessed observationally. As validation, two-sample MR was used to assess the causal associations of sleep duration with obesity, diabetes, hypertension and Hgb/Hct. To further clarify, I investigated the effect of age at puberty on sleep duration to further elucidate whether sleep duration is a cause of chronic disease.

I found longer sleep duration was related to lower BMI in both the observational and MR study in children, but such an association was not validated in the MR study in adults, especially in women. Similarly, sleep duration was not related to diabetes or FG/HbA1c. Earlier age at puberty increased sleep duration. Shorter and short sleep duration caused hypertension, but hypertension increased sleep duration in the bi-directional MR study. Lastly, sleep increased both Hgb and Hct, particularly in men.

Sleep duration had a protective effect on hypertension and anemia but had no effect on diabetes and obesity in adults. Earlier age at puberty decreased sleep duration, taken together with previous evidence that earlier puberty has a negative effect on chronic diseases, age at puberty was a potential confounder of the associations of sleep duration with obesity, diabetes, and hypertension. Potential sex differences and the mechanisms of these findings warrant further investigation.