Objective: The aim of this study was to investigate the possible relationship among sleep complaints, sleep duration, and the development of diabetes prospectively over a 12-year period in a middle-aged Swedish population.
Research design and methods: A random sample of 2,663 subjects aged 45-65 years living in mid-Sweden were sent a postal questionnaire including questions about sleep complaints, sleep duration, sociodemographic characteristics, behavioral risk factors, medical conditions, and depression (response rate 70.2%). Twelve years later, a new questionnaire with almost identical questions was sent to all the survivors (n = 1,604) in 1995, and the questionnaire was answered by 1,244 subjects (77.6%).
Results: Men reporting new diabetes at follow-up more often reported short sleep duration (< or =5 h per night) (16.0 vs. 5.9%, P < 0.01), difficulties initiating sleep (16.0 vs. 3.1%, P < 0.001), and difficulties maintaining sleep (28.0 vs. 6.3%, P < 0.001) at baseline than men who did not develop diabetes. Women reporting new diabetes at follow-up reported long sleep duration (> or =9 h per night) more often at baseline than women not developing diabetes (7.9 vs. 2.4%, P < 0.05). In multiple logistic regression models, the relative risk (95% CI) for development of diabetes was higher in men with short sleep duration (2.8 [1.1-7.3]) or difficulties maintaining sleep (4.8 [1.9-12.5]) after adjustment for age and other relevant risk factors. Short or long sleep duration or sleep complaints did not influence the risk of new diabetes in women.
Conclusions: Difficulties maintaining sleep or short sleep duration (< or =5 h) are associated with an increased incidence of diabetes in men.