Recent U.S. surveillance studies document rising depressive symptoms among young adults, but less is known about how these symptoms develop across emerging adulthood in different birth cohorts and whether delays in adult social role acquisition contribute to cohort differences.

We analyzed 19 cohorts of Monitoring the Future panel data (born 1971–1989) followed biennially from ages 18 to 30 (1989–2019; N = 7,861). Depressive symptoms were measured using a four-item index (range: 4 – 20). Quadratic latent growth curve models estimated symptom trajectories and cohort effects, including age-by-cohort interactions. Five binary time-varying social roles were evaluated as mediators: full-time college attendance, full-time employment, residential independence from parents, marriage, and parenthood. Models adjusted for sex and race/ethnicity with mediation models further adjusting for lagged depressive symptoms, urbanicity, and parental education.

Results indicated a developmental reversal in cohort effects. Each five-year increase in birth cohort was associated with lower depressive symptoms at age 20 (−0.129; 95% CI: −0.209, −0.048) but higher symptoms at age 30 (0.154; 95% CI: 0.048, 0.259). Social roles partially mediated cohort effects, with modest contributions concentrated after the reversal emerged. Declining marriage explained 11.73% of the total effect at age 30 and declining full-time employment explained 9.8% at age 28. No moderation by sex was observed.

More recent cohorts entered emerging adulthood with lower depressive symptoms than earlier cohorts, but this advantage reversed by age 30. Declining marriage and full time work accounted for a modest portion of cohort differences. Substantial direct cohort effects persisted after accounting for roles, suggesting other unmeasured factors also influence depressive symptom trajectories during emerging adulthood.