Prevalence of sleep disturbance and associated risk factors in UK Biobank participants with alcohol use disorders and major depression

Introduction Current understanding of the differences in sleep disturbance (SD) and associated risk factors in patients with alcohol use disorders (AUD), major depressive disorders (MDD), and comorbid AUD+MDD is limited. Methods Data from the UK Biobank (UKB) (n=47,825) were utilized to categorize subjects into those with MDD (n=5,991), AUD (n=12,952), both (MDD+AUD)(n=3,219), and controls (n=25,663). We used generalized linear models (GLMs) to test whether rates of SD and sleep duration differed among the groups and determine the clinical predictors of SD. Rates of SD and sleep duration were compared using regression analyses accounting for demographic (age, sex, ethnicity, and Townsend deprivation index) and clinical (body mass index, neuroticism score, and alcohol consumption) factors. Results After accounting for diagnostic category, SD was associated with age, female sex, white ethnicity, and higher BMI, neuroticism and alcohol consumption scores (all p<0.0001). The unadjusted prevalence of SD was 25.6%, 25.9%, 39.2%, and 41.1% in control, AUD, MDD, and MDD+AUD categories respectively. Rates of SD in controls and AUD group as well as MDD alone and MDD+AUD did not differ in unadjusted models (p=0.45 and 0.075, respectively). Prevalence of SD differed in the four groups (p<0.0001 for all pairwise comparisons) after adjusting for demographic confounders. After further adjustment for clinical factors, effect sizes were reduced, but pairwise comparisons remained significant, except in the AUD versus MDD group (all p<0.05). After adjusting for demographic and clinical factors, sleep duration did not differ among the groups. Conclusion Demographic and clinical characteristics associated with SD were similar in patients with MDD, AUD, and MDD+AUD. The differences in rates of SD between the diagnostic groups were attenuated but persisted after accounting for these confounders. Genetic and other factors capable of influencing SD in patients with MDD, AUD, and comorbid MDD+AUD merit future investigation.