School performance and hospital admission due to unipolar depression: a three-generational study of social causation and social selection

Both “social causation” and “social selection” offer plausible explanations for an association between education and mental
health. We aimed to explore these processes in unipolar depression, with a specific focus on school performance and family
tradition of education.

Grandchildren (N = 28,089, 49% female, aged 13–47 years in 2002) of a cohort born in Uppsala, Sweden, in 1915–1929 were studied in national
registers. We obtained data on final grade point average (GPA) in compulsory school, hospitalizations for unipolar depression,
grandparental/parental education and other parental social characteristics. Hospitalization in adolescence and adulthood were
studied separately, as were hospitalization for depression with or without a lifetime externalizing disorder.

Low compulsory school GPA (1–2 SD or >2 SD below average vs. average GPA) was associated with increased rate of adolescent
hospitalization for unipolar depression, both with externalizing comorbidity [hazard ratio (HR) point estimates of 66–80]
and without (HR point estimates of 4–6). By contrast, low GPA was only associated with first-time hospitalization in adulthood
for the subgroup with externalizing comorbidity (HR point estimates of 4–6). These associations were largely independent of
family education and social characteristics. Overall, low parental/grandparental education was not related to increased rates
of hospitalization.

The association between school performance and hospitalization for depression depended on adolescent hospitalization or externalizing
comorbidity, suggesting that disorders with an early onset are decisive. Contrary to the social patterning of many health
outcomes, low grandparental/parental education did not appear to increase the rate of hospitalization for unipolar depression
in the offspring.