Background Wealth inequalities are increasing in many countries but their impact on health, and particularly mortality, has been little studied. Existing studies have focused on specific age groups using survey data, with no studies investigating mortality inequalities by wealth across adulthood. We investigated the relationship between individual wealth and mortality across the adult life course in the Swedish population.

Methods We studied the entire Swedish adult population in 1990 using national registries, with follow-up for 19 years. Our main exposure of interest was the amount of tax paid in a historical tax on wealth, calculated using a comprehensive assessment including real estate (value in prior year minus outstanding mortgages), high value personal items (e.g. cars, jewellery and boats), bank accounts, shares, government bonds and property annuities. Inequalities by wealth were contrasted with individual- and household measures of income and the extent to which associations were attenuated by four other measures of socioeconomic position investigated.

Relative indices of inequality (RII) were calculated for all-cause and cause-specific mortality for six different age groups (25–39, 40–54, 55–64, 65–74, 75–84, 85+ years) using Poisson regression with robust standard errors. All analyses were stratified by sex and adjusted for age (five year bands) and follow-up year.

Results Large inequalities in mortality by wealth were observed and their importance as a predictor of mortality remained more stable throughout the adult life course than income-based measures. The RII for wealth was 2.28 (95% CI 2.17 to 2.40) in men aged 25–34 years compared to 5.41 (95% CI 5.17 to 5.66) by individual income. Men experienced greater inequalities across all ages (e.g. the wealth-based RII was 1.82 (95% CI 1.72 to 1.93) for women aged 25–34 years), except amongst the over 85s. Compared to individual and household income, wealth became more important over the working age period. Adjustment for education, social class, individual- and household-income led to only modest reduction in the association between wealth and mortality. Wealth was associated with nearly all causes of death and the highest relative risk increases were seen for alcohol, drugs, infections and external causes of death.

Discussion Wealth is strongly predictive of mortality throughout the adult life course. Study limitations include reliance on a historical dataset and potential under-ascertainment of off-shore assets (although using a rank-based measure limits its likely impact). Redistribution of income may not be sufficient to narrow health inequalities – policymakers may need to focus on the increasingly unequal distribution of wealth in high-income countries.