Association between cardiometabolic health and objectively-measured, free-living sleep parameters in a rural African setting: a pilot study.

Abstract Objectives To investigate the relationship between objectively-measured, free-living sleep quantity and quality, and cardiometabolic health, in a rural African setting in 139 adults (≥40 years, female: n=99, male: n=40). Wrist-mounted, tri-axial accelerometry data was collected over nine days. Measures of sleep quantity and quality, and physical activity were extracted from valid minute-by-minute data. Self-reported data included behavioural, health and socio-demographic variables. Biological data included body composition, resting blood pressure and fasting blood glucose, insulin and lipids. Regression models were constructed with insulin resistance (IR), Cardiometabolic (CM) risk and a metabolic z-score, as dependent variables, adjusting for socio-demographic, behavioural and biological factors. Results Nocturnal sleep time was longer in females (p=0.054) and sleep quality was better in males (p≤0.017). Few participants slept >9 hours/night (4-5%), and 46-50% slept <7 hours/night. IR and CM risk was higher in females (p≤0.006). In adjusted models, sleep variables were independently associated with IR (p<0.05). Sleep quantity was linearly (p<0.05) and non-linearly (p≤0.0196) associated with IR, and non-linearly (p≤0.0398) associated with CM risk. Sleep quality was linearly related with IR (p<0.05). A number of non-sleep behavioural variables were independently associated with CM risk (alcohol and tobacco use, p≤0.034) and IR (physical activity, sugar-sweetened beverage consumption, p<0.05).