The antecedents and outcomes of persistent and remitting adolescent depressive symptom trajectories: a longitudinal, population-based English study

Background Depression often first emerges in adolescence and, for many, is a lifelong disorder. The long-term clinical course of depression is highly variable. We aimed to examine the adult outcomes of adolescent-onset trajectories of clinically significant depressive symptoms and to identify factors differentiating trajectories that persist and desist in adulthood. Methods We included participants from the English population-based Avon Longitudinal Study of Parents and Children with data on depressive symptoms. Self-reported depression symptoms were assessed on ten occasions when participants were age 10·5–25 years using the short Mood and Feelings Questionnaire, and major depressive disorder episodes were assessed at age 13·0 years, 15·0 years, 17·5 years, and 25·0 years. We characterised trajectories of depression symptoms using latent class growth analysis, for which we required depression data at least once from each of three key phases: ages 10·5–13·5 years; 16·5–18·5 years; and 21–25 years. We examined adult outcomes by assessing lifetime suicidal self-harm and functional impairment at age 24·0 years, and employment, education, and the self-reported Strengths and Difficulties Questionnaire at age 25·0 years. Findings We studied 4234 participants: 2651 (63%) female, 1582 (37%) male, and one individual with missing sex data. The mean age was 10·6 years (SD 0·2) at baseline and 25·8 years (SD 0·5) at the final timepoint. Data on ethnicity were not available in our data set. We identified four depression trajectory classes: adolescent-persistent depression with onset early in adolescence (7%, n≈279), adolescent-limited depression with onset later in adolescence and remittance by adult life (14%, n≈592), adult-increasing depression (25%, n≈1056), and stable-low levels of depression (54%, n≈2307). The adolescent-persistent class was associated with poor adult outcomes for functional impairment (62%), suicidal self-harm (27%), mental health difficulties (25%), and not being in education, employment, or training (16%). Adolescent-limited depression was associated with transient adolescent stress, but by early adulthood functional impairment and mental health difficulties were similar to the stable-low group. Major depressive disorder polygenic score (odds ratio [OR] 1·36, 95% CI 1·04–1·79), adolescent educational attainment (OR 0·47, 0·30–0·74), and any early childhood adversity (OR 2·60, 1·42–4·78), that persisted into adulthood (OR 1·60, 1·38–1·87) distinguished the adolescent-persistent and adolescent-limited groups. Interpretation The future course of adolescent depression can be differentiated by age at onset during adolescence, adolescent academic attainment, early and persistent adversity, and genetic loading. A detailed social and educational history could be helpful in making clinical decisions about the intensity of interventions for young people with clinically elevated depressive symptoms who seek help