Childhood social characteristics and drug-related mortality by age 41: a register-based study of Finnish birth cohorts 1982–2004

Attained socioeconomic position is associated with drug-related deaths, but whether social characteristics in childhood predict drug-related mortality remains unclear. We estimated differences in drug-related mortality between ages 16 and 41 by parental education, household income, and household type. We used administrative data on all Finnish residents born 1982–2004 (N = 1 446 548). Underlying and contributory causes of drug-related mortality (1997–2023) were drawn from Causes of Death Register records and parental education, household income and household type (age 15) were derived from administrative records. Differences in drug-related mortality by childhood social characteristics and their co-occurrence were estimated using sex-stratified Cox regression. Birth cohort differences were estimated using Kaplan-Meier failure curves. The drug-related mortality rate per 100 000 person-years for men (21.9, 95% CI: 21.1–22.8) was over 3× higher than that of women (5.7, 5.3–6.2). Among both men and women, respectively, lower parental education (basic: HR 1.86, 1.63–2.11; HR 1.78, 1.37–2.32) and non-nuclear household type (single-parent households: HR 2.10, 1.89–2.33; HR 3.06, 2.46–3.79) independently predicted higher mortality hazards net of all predictors, whereas the estimates for household income greatly attenuated. Interaction analyses between parental education and household type suggested larger relative increases in mortality hazards by lower parental education among individuals from non-nuclear families. Mortality differences were consistently observed across birth cohorts. Childhood household type and parental education independently predicted subsequent drug-related mortality among men and women. Children of lower educated parents from non-nuclear families had the highest excess risk.