Co-morbid OSA and insomnia increases depression prevalence and severity in men

Abstract

Background and objective: Obstructive sleep apnoea (OSA) and insomnia coexist in clinical populations but prevalence in the community and risk factors remain largely unknown. We examined the prevalence and profile of previously undiagnosed co-morbid OSA and insomnia symptoms (COMISA) in community-dwelling men. Methods: Men (n = 700, aged 58.5 ± 11.0 (mean ± SD) years) without a prior diagnosis of OSA completed full at-home unattended polysomnography, the Pittsburgh Sleep Quality Index and 36-item short form (SF-36) survey (2007–2012). Insomnia symptoms included difficulty initiating/maintaining sleep in the presence of daytime fatigue (DIMS-F). Depressive symptoms were assessed using the Beck Depression Inventory-1A, Centre for Epidemiological Studies Depression Scale and Patient Health Questionnaire-9 (PHQ-9) (2007–2010). Univariate (χ2 and analysis of variance (ANOVA)) and multiple linear regressions were used to compare data from four groups of individuals: neither disorder; previously undiagnosed OSA (apnoea–hypopnoea index ≥ 10) or DIMS-F alone; and COMISA. Results: COMISA prevalence was 6.7%. Depression prevalence (COMISA, 42.6%; DIMS-F, 21.6%; OSA, 8.4%, χ2 = 71.6, P < 0.00) and symptom scale scores (e.g. PHQ-9 mean ± SD: 16.1 ± 5.5 c.f. DIMS-F: 14.0 ± 4.9, P < 0.01 and OSA: 11.4 ± 3.0, P = 0.01) were highest in men with COMISA. In COMISA, respiratory and arousal indices were similar to those observed in OSA whilst reductions in subjective sleep and day dysfunction scores were similar to DIMS-F. After adjustment, predicted mean depression scores were all higher in DIMS-F and COMISA using linear regression (e.g. PHQ-9 β (95% CI): DIMS-F: 2.3 (1.2, 3.5); COMISA: 4.1 (3.0, 5.1)). Conclusion: Men with COMISA have a greater prevalence, and severity, of depression than men with only one disorder.