Psychological and contextual risk factors for first-onset depression among adolescents and young people around the globe: A systematic review and meta-analysis

Abstract

Aim: Identifying predictors for future onset of depression is crucial to effectively developing preventive interventions. We conducted a systematic review and meta-analysis to identify risk factors for first-onset depression among adolescents and young people.

Methods: We searched MEDLINE (Ovid), PsycINFO, Cochrane Database, Web of Science, Lilacs, African Journals Online and Global Health (July 2009 to December 2020) for longitudinal studies assessing risk factors for first-onset depression among adolescents and young people aged 10-25 years. Meta-analyses generated summary odds ratio (OR) estimates.

Registration: PROSPERO CRD42018103973.

Results: Nineteen studies representing 21 unique populations were included in the meta-analysis. Among studies reporting race/ethnicity, 79% of participants were of White/European descent. Seventeen studies were from high-income countries, with only two from an upper-middle-income country (China). Odds for first-onset depression were significantly greater for girls compared to boys (n = 13; OR = 1.78 [1.78, 2.28], p < 0.001) and for youth with other mental health problems at baseline (n = 4; OR = 3.20 [1.95, 5.23], p < 0.001). There were non-significant associations for negative family environment (n = 8; OR = 1.60 [0.82, 3.10], p = 0.16) and parental depression (n = 3; OR = 2.30 [0.73, 7.24], p = 0.16).

Conclusions: Most longitudinal studies do not report risk factors specifically for first-onset depression. Moreover, predictive data are limited to predominantly White populations in high-income countries. Future research must be more ethnically and geographically representative. Recommendations are provided for consistent and comprehensive reporting of study designs and analyses of risk factors for first-onset depression.

Keywords: adolescent; developing countries; early diagnosis; risk assessment; risk factors.

FIGURE 1

PRISMA flow diagram of study selection process. PRISMA flow diagram showing the study selection process for the systematic review and meta‐analysis. Study selection stages include (1) Identification: Number of records identified through the search across databases, additional sources identified through hand searches, and number of duplicates removed before screening. (2) Screening: Included number of titles and abstracts screened (12 years, 2009–2020) and records excluded; and (3) Eligibility: full text articles found eligible for full‐text screening, records excluded with reasons, and final number of studies included in the meta‐analysis. *Inclusive of first round screening of 12 753 titles and abstracts with four reviewers, and second round of re‐screening for 6394 titles and abstracts with three reviewers

FIGURE 2

Forest plot of risk factor categories predicting first‐onset depression outcome among study populations (n = 19) †Forest plot shows four risk factor categories (gender, family environment, other mental health problems, and parental depression) in which three or more studies were compiled per category. We conducted random‐effects meta‐analyses in Cochrane Review Manager version 5.3 and calculated odds ratios (OR) for binary outcomes with 95% confidence intervals and two‐sided p‐values for each risk factor category. Heterogeneity was assessed using the I‐squared statistic within Cochrane Review Manager 5.3. * Forest Plot Abbreviations: Chi‐square (Chi²); Confidence interval (CI); Degrees of freedom (df); Heterogeneity (I ²); Interval variable (IV); Standard error (SE)