Fig. 2

The association between parental income across childhood and adult health separately by childhood circumstances (A–C) and adult socioeconomic attainment (D–F), Norwegian birth cohorts 1967–1973. Source: data from the Norwegian Control and Reimbursement Database, 2006–2016. Note: predicted probabilities of having had one or more consultations for any disorder (i.e., one of the disorders in A–L) from linear probability models for childhood parental income vigintiles, controlling for birth year, estimated using OLS regression. Shaded areas refer to 95% confidence intervals. Each panel shows the predicted probabilities from regressions where each childhood parental income vigintile is separately interacted with dummy variable measures of birth weight (A), mother’s marital status at child age 16 (B), parental education at child age 16 (C), educational attainment at age 35 (D), adult income quartile averaged across ages 30–36 (E), and has been or being married with own children by age 43 (F)