Trajectories of depressive symptoms among young people in London, UK, and Tokyo, Japan: a longitudinal cross-cohort study.

BACKGROUND

Research suggests gender inequalities in adolescent mental health are context dependent and might be preventable through social and structural change. However, variations in the size of gender inequalities in mental health across diverse cultural contexts could be due to incomparable measurement. We aimed to compare a measurement of mental health among young people in Tokyo, Japan, and London, UK, and test the hypothesis that gender inequalities in depressive symptom trajectories are larger in London than in Tokyo.

METHODS

For this longitudinal cross-cohort study, we extracted responses to the 13-item Short Mood and Feelings Questionnaire (SMFQ) by young people who participated in three consecutive waves of the Tokyo Teen Cohort (TTC) and the London-based Resilience, Ethnicity and Adolescent Mental Health (REACH) cohorts. We used multigroup and longitudinal confirmatory factor analysis to examine measurement invariance of the SMFQ by cohort, gender, and age. Latent growth curve models were used to estimate and compare mean trajectories of SMFQ from ages 11-16 years among boys and girls, overall, and in each cohort.

FINDINGS

7100 young people from TTC and REACH (3587 boys [50·5%] and 3513 girls [49·5%]) were included in the analysis. With the TTC and REACH cohorts combined, we found very strong evidence of differences in SMFQ between boys and girls, with a mean starting level of 0·71 points (95% CI 0·42-0·95) higher and mean rate of change of 0·73 points (95% CI 0·62-0·82) higher in girls versus boys. Among the 4287 participants in REACH (2097 [48·9%] boys and 2190 [51·1%] girls), a difference in SMFQ was evident between boys and girls at age 11-12 years (difference in mean intercepts: 0·75 [95% CI 0·25-1·25]). Among the 2813 participants in TCC (1490 boys [53·0%] and 1323 girls [47·0%]), differences in SMFQ between boys and girls emerged at a later age, between ages 11 years and 14 years, during which SMFQ decreased among boys and increased among girls (mean difference in slopes 0·52 [95% CI 0·40 to 0·65]). The difference in SMFQ between boys and girls widened year-on-year in both cohorts; by age 16 years, the difference in SMFQ between boys and girls in REACH (mean difference in slopes 0·98 [95% CI 0·77 to 1·20]) was around twice as large as in TTC (0·52 [0·40 to 0·65]). The annual rate of increase in SMFQ among girls in REACH (1·1 [95% CI 0·9-1·3]) was around four times greater than among girls in TTC (0·3 [0·2-0·4]). We found little evidence to suggest these differences in gender inequalities were due to incomparable measurement.

INTERPRETATION

Gender inequalities in emotional health among young people are context dependent and might be preventable through social and structural change.

FUNDING

Japanese Society for the Promotion of Science, UK Economic and Social Research Council, and European Research Council.

TRANSLATION

For the Japanese translation of the abstract see Supplementary Materials section.