Too much or not enough sleep may raise an individual's risk of developing type 2 diabetes, independent of age, blood pressure, smoking status, and waist circumference. For the study, researchers assessed the impact of sleep duration on the development of diabetes in >1100 middle-aged and elderly men who were diabetes-free in 1987 to 1989. The participants were followed for 15 years. During the long-term study, the researchers found that men getting no more than 6 hours of sleep a night, as well as those getting more than 8 hours, faced a considerable increased risk of developing the disease, compared with men getting 7 to 8 hours of sleep a night.

Specifically, the risk of diabetes was 2-fold higher in men with short sleep duration and >3-fold higher in men reporting long sleep duration, compared with men sleeping 7 to 8 hours a night. The researchers noted, however, that relative risks for diabetes were lowered "considerably" when adjusted for testosterone. The observation suggested that the effects of sleep duration could be mediated through changes in the body's levels of this hormone. (The findings were reported in Diabetes Care, March 2006.)