An 8-Year Prospective Study of Depressive Symptoms and Change in Insulin From Adolescence to Young Adulthood

To evaluate whether depressive symptoms predict change in fasting insulin among adolescents followed into young adulthood. We hypothesized that higher depressive symptoms would predict increased insulin and that puberty and race/ethnicity would moderate this relationship.

Data came from the Princeton School District Study, a school-based longitudinal cohort of non-Hispanic black and white adolescents (2001-2011). Depressive symptoms, fasting insulin, and body mass index were measured at baseline (adolescence) and 8 years later (young adulthood) in 685 participants. Puberty was assessed using a validated protocol measuring sex steroids and physical changes. The primary outcome was change in fasting insulin. Analyses accounted for age, sex, race, parental education, baseline insulin, body mass index z score, puberty, and time to follow-up.

At baseline, depressive symptoms were correlated with insulin (ρ = 0.13, p = .001). High baseline insulin predicted insulin change (B = -11.50, standard error [SE] = 2.30, p < .001). Depressive symptoms also predicted insulin change, but only for pubertal adolescents (B = -0.23, SE = 0.11, p = .038). This relationship was moderated by race (p = .047); depressive symptoms predicted insulin change only among pubertal black adolescents (p = .030), not white (p = .49), and in the direction opposite that hypothesized (Bblacks = -0.51, SE = 0.23). Post hoc analyses revealed that pubertal black adolescents with high depressive symptoms had the highest baseline insulin, which stayed high across the follow-up period.

Among pubertal black adolescents, elevated depressive symptoms are associated with increased risk for sustained hyperinsulinemia from adolescence into adulthood. These youths may be particularly vulnerable for Type 2 diabetes.