The relationship between obstructive sleep apnea and insomnia: a population-based cross-sectional polysomnographic study

Background: The relationship between insomnia and objectively measured obstructive sleep apnea (OSA) severity has not previously been investigated in both genders in the general population. The main aim of this population-based polysomnography (PSG) study was to evaluate the cross-sectional association between severity of OSA and DSM-V insomnia and insomnia severity. Methods: A random sample of 1200 participants in the third Nord-Trondelag Health Study (HUNT3) was invited and 213 (18%) aged between 21 and 82 years underwent an ambulatory PSG, a semi-structured interview, and a sleep-specific questionnaire. A proxy DSM-V insomnia diagnosis as well as an Insomnia Symptom Score (ISS, range 0-12) were calculated from three insomnia questions and one daytime sleepiness symptom question. Participants were then divided into three groups according to their apnea-hypopnea index (AHI): AHI < 5 (without OSA), AHI 5-14.9 (mild OSA), and AHI >= 15 (moderate-to-severe OSA). Associations between prevalence of insomnia and OSA groups were assessed by logistic regression models adjusted for age and gender. Associations between ISS and OSA were assessed in a general linear model with contrasts. Results: A total of 25.2% (29.1% women, 12.5% men) had insomnia. Insomnia prevalence did not differ between subjects with and without OSA, but ISS differed significantly between OSA categories (ANCOVA df 2, F = 6.73, p = 0.001). ISS was lower in the moderate-to-severe OSA-group compared to those without OSA (mean difference -2.68; 95% [CI -4.33, -1.04]; p = 0.002). In subjects with moderate-to-severe OSA, ISS correlated negatively with age (Pearson r = -0.66, p = 0.015). Conclusion: In this population-based PSG study, no overall statistical association between OSA and insomnia prevalence was found. However, participants with moderate-to-severe OSA reported less insomnia symptoms than subjects without OSA, in particular in older individuals. (c) 2018 Elsevier B.V. All rights reserved.