Abstract

Objectives. To assess the relationship between habitual sleep disturbances and the incidence of type 2 diabetes, and to obtain an estimate of the risk.

Research Design and Methods. Systematic search of publications using MEDLINE (1955-April 2009), EMBASE, the Cochrane Library and manual searches without language restrictions. We included studies if they were prospective, with follow-up >3 years, had an assessment of sleep disturbances at baseline and incidence of type 2 diabetes. We recorded several characteristics for each study. We extracted quantity and quality of sleep, how they were assessed, and incident cases defined with different validated methods. We extracted relative risks (RR) and 95% C.I. and pooled them using random effects models. We performed sensitivity analysis, and assessed heterogeneity and publication bias.

Results. We included ten studies (13 independent cohort samples; 107,756 male and female participants, follow-up range 4.2 to 32 years, 3,586 incident cases of type 2 diabetes). In pooled analyses, quantity and quality of sleep predicted the risk of developing type-2 diabetes. For short duration of sleep (≤5-to-6h per night) the RR was 1.28; 95% CI 1.03 to 1.60; p=0.024 (heterogeneity p=0.015); for long duration of sleep (>8-to-9h per night) 1.48 (1.13 to 1.96; p=0.005); for difficulty in initiating sleep 1.57 (1.25 to 1.97; p<0.0001) and for difficulty in maintaining sleep 1.84 (1.39 to 2.43; p<0.0001).

Conclusions. Quantity and quality of sleep consistently and significantly predict the risk of the development of type 2 diabetes. The mechanisms underlying such relation may differ between short and long sleepers.