Does household income mediate the association between education and health in Canada?

The study aim was to investigate whether household income mediates the association between education and health in a nationally representative sample of Canadian adults. The data came from the Longitudinal and International Study of Adults linked to income data from the Canada Revenue Agency. Odds ratios and predicted probabilities from binary logistic regression models were used to describe associations between education and (a) self-rated health, (b) longstanding illness or health problem, (c) emotional, psychological or mental health problem and (d) symptoms of psychological distress. The Karlson-Holm-Breen decomposition method was used to investigate the potentially mediating role of household income in these associations. The analyses were conducted separately for women and men. Education was significantly associated with all four health indicators for both women and men. Of the four health indicators, education was most strongly associated with self-rated health for both women and men. Education was more strongly associated with self-rated health and the presence of an emotional, psychological or mental health problem for women than for men. Curiously, men with a postgraduate degree were significantly more likely than men with a bachelor degree to report symptoms of psychological distress. Only modest proportions of the associations between education and health could be attributed to differences in household income. Education and household income manifested independent associations with all four health indicators among women and with three of four health indicators among men. Education and household income are joint independent predictors of health in Canada. Accordingly, both should be included in research on socioeconomic health inequalities in this context.