Contribution of psychosocial factors to socioeconomic inequalities in mortality among older Australian men: a population-based cohort study

Background Among older people, the extent to which psychosocial factors explain socioeconomic inequalities in mortality is debated. We aimed to investigate the potential mediating effect of psychosocial factors on socioeconomic inequalities in mortality. Methods We used data from a prospective population-based cohort (the Concord Health and Ageing in Men Project; baseline recruitment in 2005-2007), in Sydney, Australia. The main outcomes were all-cause and cause-specific mortality. Socioeconomic status (SES; educational attainment, occupational position, source of income, housing tenure, and a cumulative SES score) was assessed at baseline. Measures of structural and functional social support, as well as depressive and anxiety symptoms were assessed three times during follow-ups. Associations were quantified using Cox regression. Mediation was calculated using change-in-estimate method. Results 1522 men (mean age at baseline: 77 center dot 4 +/- 5 center dot 5 years) were included in the analyses with a mean (SD) follow-up time of 9 center dot 0 (3 center dot 6) years for all-cause and 8 center dot 0 (2 center dot 8) years for cause-specific mortality. At baseline, psychosocial measures displayed marked social patterning. Being unmarried, living alone, low social interactions, and elevated depressive symptoms were associated with higher risk of all-cause and cardiovascular disease (CVD) mortality. Psychosocial factors explained 35% of SES inequalities in all-cause mortality, 29% in CVD mortality, 12% in cancer mortality, and 39% in non-CVD, non-cancer mortality. Conclusion Psychosocial factors may account for up to one-third of SES inequalities in deaths from all and specific causes (except cancer mortality). Our findings suggest that interventional studies targeting social relationships and/or psychological distress in older men aiming to reduce socioeconomic inequalities in mortality are warranted.