School performance and later diagnoses of nonaffective psychoses, bipolar disorder, and depression

Objective It is unclear whether there are differences between specific school subject performance and later psychiatric disorders. We examined whether mean grade point average (MGPA) and specific school subjects associated with diagnoses of nonaffective psychoses, bipolar disorder and depression. Methods In this register-based study, we studied the Finnish population born in 1987 who had available MGPA and six specific school grades (age = 15.4-16.4 years; n = 50,508). Grades were analyzed with smoothing splines. Covariates included sex, urbanicity, parental education level and parental diagnosed psychiatric disorders. Outcomes were incident nonaffective psychosis, bipolar disorder and depression diagnosed in specialized services until year 2015 (age = 28.0-28.9 years). Results During the follow-up, 727 individuals were diagnosed with nonaffective psychoses, 489 with bipolar disorder and 3492 with depression. MGPA was inversely associated with all outcomes. In multivariate models including specific school subjects and covariates, the school subject with largest risk ratios (RR) was Physical Education (RR and Bonferroni-corrected confidence interval [CI] at -1.5 SD: nonaffective psychoses 1.63, 1.36-1.95; bipolar disorder 1.64, 1.30-2.05; depression 1.72, 1.53-1.93). Higher grades in Art were associated with nonaffective psychoses and depression (RR and Bonferroni-corrected CI at +1.5 SD: nonaffective psychoses 1.48, 1.11-1.96; depression 1.22, 1.07-1.38). Conclusion There was a robust association between poorer scores on Physical Education and risk for psychosis, bipolar disorder and depression. Higher grades in Art were also associated with risk for later disorders. Subject specific school performance may be more informative about mental disorder risk than overall school performance.