Complaints of difficulty to fall asleep increase the risk of depression in later life: The health in men study

Objectives: To determine if complaints of poor sleep are associated with incident depression in older men. Methods: Cohort study with an average follow up period of 6 years (range 3 months to 8.5 years). Participants were 5127 community-dwelling Western Australian older men aged 70-90 years who provided information about sleep problems. The primary outcome of interest of the study was a recorded diagnosis of depressive episode, recurrent depressive disorder or dysthymia in the Western Australian Data Linkage System. Participants completed a health questionnaire that included questions assessing difficulty falling asleep, remaining awake, as well as early morning awakening. Other measured factors included age, education, country of birth, living arrangements, social support, smoking, body mass index, and prevalent diabetes, hypertension, arthritis, chronic respiratory diseases, coronary artery disease, stroke, and cancer. Biochemical measurement of C-reactive protein, testosterone and plasma homocysteine were available for 3800 men. Results: We found that 60% of men reported at least one sleep problem and that the unadjusted hazard ratio (HR) of depression was higher in men who complained of difficulties to initiate sleep (HR = 2.19, 95% confidence interval - 95% CI = 1.47-3.27) or who remained awake most of the night (HR = 1.94, 95% CI = 1.15-3.27). There was no association between early morning awakening and incident depression. The association between incident depression and subjective difficulty falling asleep remained after the analyses were adjusted for other measured factors (HR = 1.83, 95% CI = 1.20-2.79). The association between depression and remaining awake was no longer significant once the analyses were adjusted for confounding (HR = 1.43, 95% CI = 0.812.53). A sensitivity analysis confirmed these results. Limitations: The evaluation of the exposure (sleep disturbance) was limited to self-rating questions that were not externally validated. The diagnosis of depression was based on administrative record linkage rather than structure clinical interviews. The observational nature of the study limits our ability to ascribe a causal relationship between complaints of poor sleep and incident depression. Conclusions: Complaints of difficulty falling asleep increase the risk of incident depression in older men. Clarifying the mechanisms that underlie this association should become an international research priority, as they may contribute to guide interventions designed to decrease the burden of depression in later life. (C) 2011 Elsevier B.V. All rights reserved.