Harmful alcohol habits did not explain the social gradient of sickness absence in Swedish women and men

Abstract

Background: the aim of this study was to examine the prevalence and socioeconomic distribution of harmful alcohol habits in sick-listed women and men, and whether the social gradient in sickness absence could be explained by the socioeconomic distribution of harmful alcohol habits. Methods: this cross-sectional questionnaire study included newly sick-listed individuals (n=2 798, 19-64 years, 66% women) from Sweden. The outcome variable, self-reported harmful alcohol habits, was measured with the Alcohol Use Disorder Identification Test. Registered socioeconomic variables (education, income, occupational class) were explanatory variables with age as confounder and selfreported health, symptoms, mental wellbeing, and self-efficacy as mediators. Chi2-tests and logistic regression models were applied. Results: 9% of sick-listed women and 22% of men had harmful alcohol habits. Women with a low annual income (≤149 000 SEK) had higher odds ratios (OR=2.47; 95% CI=1.43-4.27) of harmful alcohol habits than those with ≥300 000 SEK/year. The significance of low income remained when mediators were introduced into the logistic regression model (OR=2.03, 95% CI=1.13-3.65). In the model including age, income was no longer significant. Men with low income were more likely to have harmful alcohol habits than men with high income (OR=2.59; 95% CI=1.45-4.62). When mediators were included low income remained significant (OR=2.88; 95% CI=1.56-5.31). Income was no longer significant when age was introduced. Education and occupational status were not significant. Conclusions: harmful alcohol habits were common among sick-listed women and men. The socioeconomic differences in harmful alcohol habits did not explain the social gradient in sickness absence.