Background Sophisticated measures of inequality of opportunity for health require the empirically challenging and ethically controversial task of disentangling causal networks between circumstances and choices throughout an individual’s life. ‘Life course’ gradients – bivariate associations between childhood social circumstances and adult health – may provide a simple indicator of societal lifetime inequality of opportunity for health, since there is widespread ethical agreement children cannot be held responsible for their social circumstances. We compare conventional ‘current’ social gradients in adult health with ‘life course’ gradients, representing a simple inequality of opportunity metric.

Methods We use data from the UK 1958 National Child Development Study, applying multiple imputation methods to account for attrition. The primary health variable is SF-36 general health score (0–100) measured at age 50, with dichotomised self-assessed general health included as a robustness check. The primary SES variables are income, social class, and education measured at age 50 and age 16. SES variables at age 16 are measured by parental proxy and several alternatives are explored in robustness checks, including parental SES earlier in childhood. Gradients are calculated using the slope and relative indices of inequality (SII, RII) for SF-36 general scores and risk differences and ratios for self-assessed health.

Results Life course income gradients in SF-36 scores are 4.53 and 5.22 for women and men, respectively, using the SII (p<0.001) and 1.07 and 1.08 for women and men, respectively, using the RII (p<0.001). Life course gradients in SF-36 scores are also found using mother’s education (SII: 11.32/10.84 for women/men, p<0.001; RII: 1.19/1.18 for women/men, p<0.001), father’s education (SII: 9.97/7.75 for women/men, p<0.001; RII: 1.16/1.13 for women/men, p<0.001) and household social class (SII: 11.60/7.11 for women/men, p<0.001; RII: 1.19/1.12 for women/men, p<0.001). Life course gradients are approximately half the size of current gradients for parental education and women’s parental social class and one third the size for parental income and men’s parental social class. Results are robust across inequality measures, adult health variables, childhood SES variables, and missing data assumptions.

Discussion Children in the most socioeconomically disadvantaged households at age 16 tend to experience substantially worse general health at age 50 than children in the most advantaged households. This trend is more pronounced when using static SES markers such as education. Lifecourse social gradients provide a simple indicator of inequality of opportunity for health, which are 31–56% the magnitude of health inequality measured using conventional current social gradients.