AIMS: We examined the influence of comorbid sleep disorder on the association between type 2 diabetes (T2D) and risk of incident depression.

METHODS: The study population (N = 232,489) was based on all individuals registered aged ≥40 years with a T2D diagnosis between January 1, 2000 to December 31, 2012 in the Danish National Diabetes Register and a matched reference population. The risk of incident depression (diagnosis or anti-depressant medication) following T2D and possible effect modification of comorbid sleep disorder was estimated using adjusted Cox proportional hazards regression. Sleep disorder was defined as a diagnosis of insomnia, hypersomnia or sleep-wake schedule disorders or use of sleep medication (z-drugs or melatonin) in the Danish National Patient Registry or the Danish National Prescription Registry.

RESULTS: At study entry, 15.3 % of the participants had a sleep disorder. During follow-up, 2.6 % were diagnosed with depression and 32.1 % received antidepressant medication. The unadjusted hazard ratio (HR) for depression was 1.54 (95%CI 1.52-1.56) for patients with diabetes, which attenuated to 1.50 (1.48-1.52) after adjustment for sleep disorders, which further attenuated to 1.27 (1.26-1.29) in the model further adjusted for psychiatric and somatic comorbidities. The analyses of T2D and sleep disorder as independent and combined variables compared with none of the conditions on risk of depression, showed a HR of 1.27 (95 % CI 1.19-1.35) for T2D without sleep disorder, 1.46 (95 % CI 1.33-1.59) for sleep disorders without T2D, and 1.49 (95%CI 1.37-1.63) for both conditions.

CONCLUSIONS: T2D and sleep disorders were independently associated with subsequent risk of depression and individuals with both conditions experienced the greatest relative risk. Sleep disorders neither explained nor amplified the relation between diabetes and depression.