LAUSR

Background Three main explanations of the relationship between income and population health have been identified: the absolute, the contextual and the relative income hypotheses. The evidence about their relevance particularly in egalitarian societies is, however, inconsistent. This study aimed to test the three hypotheses in relation to psychological distress in northern Sweden. Methods Data come from the 2014 cross-sectional survey from the four northern-most counties in Sweden, and included people aged 25-84 years (n = 21 004). Psychological distress was measured by the General Health Questionnaire-12 and income information came from population registers. Absolute income was operationalized by individual disposable income, contextual income as the municipal-level Gini coefficient and relative income by the Yitzhaki index. Prevalence ratios (PR) were calculated from log-binomial regression analyses. Results A gradient in poor mental health was observed across quintiles of individual income, with the poorest substantially more likely to report poor health compared with the highest quintile (PR = 1.56; 95% CI = 1.19, 2.04). Second, municipalities in the quintiles 2-3 of the Gini coefficient had a better mental health compared with those in the most equal municipalities. Third, a gradient in poor mental health across quintiles of relative deprivation was also found, with the most deprived quintile the most likely to report poor health (PR = 1.37; 95% CI = 1.06, 1.76). Conclusion This study suggests a strong, moderate and lack of support for the absolute, relative and contextual income effect hypotheses, respectively. Interventions targeting a reduction in the individual income gap may be necessary in order to reduce psychosocial distress differences in northern Sweden.