The review focused on adolescents aged 10 to 19 who experienced symptoms of depression or anxiety but did not meet the criteria for clinical disorders. Its authors said this group is particularly relevant in low- and middle-income countries, where 90% of the world’s adolescents live and where access to evidence-supported mental health services is limited. A 2021 State of the World’s Children report estimated that adolescent mental health conditions cost the global economy $451 billion annually in lost human capital, with the projected total rising to $620 billion by 2040.

The systematic search covered Academic Search Complete, APA PsycInfo, Global Health, MEDLINE and SocINDEX with Full Text for studies published from January 2013 through July 2024. It identified 1,710 records, removed 368 duplicates and sent 1,342 records for abstract screening. Fifty-seven full texts were assessed, resulting in the inclusion of 22 studies: 19 randomized controlled trials and three quasi-experimental studies. Inter-rater agreement during screening was high, with a Cohen’s kappa of 0.85. The protocol was registered with PROSPERO as CRD42024598603, and the review followed PRISMA 2020 and Synthesis Without Meta-analysis reporting guidance.

Eligible studies used experimental or quasi-experimental designs with a comparison group and quantitatively evaluated depression or anxiety outcomes. Medication-only interventions, studies that did not report separate scores for depression and anxiety, and studies focused on extreme trauma were excluded. Because the interventions, methods and outcome measures varied substantially, the researchers did not conduct a formal meta-analysis. Instead, they calculated standardized effect sizes, known as Cohen’s d, and synthesized the findings narratively.

Two reviewers evaluated methodological rigor with an adapted Methodological Quality Rating Scale, assessed risk of bias with version 2 of the Cochrane risk-of-bias tool for randomized trials and ROBINS-I for quasi-experimental studies, and used GRADE to assess the certainty of evidence. Among the randomized trials, six had low risk of bias, 12 raised some concerns and one had high risk. Common concerns included unblinded outcome assessors and the absence of pre-registered protocols with pre-specified analysis plans. All three quasi-experimental studies were rated at serious risk of bias.

The clearest finding involved depression interventions delivered by non-specialists. Six randomized trials had a median effect size of 0.26, with a range of 0.07 to 0.39. They showed no serious concerns across the GRADE domains, had consistent effects with overlapping confidence intervals, and included 67% adequately powered studies, producing a high certainty rating. By comparison, five specialist-delivered depression studies had a median effect size of 1.01 but a much wider range of 0.04 to 2.88. Only 20% were adequately powered, and the evidence was rated very low certainty because of risk of bias, inconsistency and imprecision.

For anxiety, only four non-specialist-delivered studies were available. Their median effect was 0.58, with a range of 0.37 to 2.00, and the evidence was rated low certainty because the effects were inconsistent. Specialist-delivered anxiety interventions were assessed in 12 studies, had a median effect of 1.35 and were rated very low certainty. The trials also addressed different forms of anxiety, including test, social and generalized anxiety, and used at least six different anxiety instruments, limiting direct comparisons. The evidence therefore could neither confirm nor refute whether specialist delivery is more effective than non-specialist delivery for anxiety.

The comparison by setting was also inconclusive. Twelve interventions were conducted in schools and 10 in non-school environments, including hospitals, community organizations and online platforms. Certainty of evidence was very low for both settings and both outcomes because of problems with bias, inconsistent effects and imprecision. Across the included studies, effect sizes ranged from 0.04 to 2.88 for depression and from 0.21 to 2.87 for anxiety. Differences in sample size, outcome measures, dosage and clinical model made it difficult to isolate the effect of implementation setting, but did not make the comparison impossible in principle.

The review also identified weaknesses in the underlying literature. Sixty-four percent of studies lacked adequate statistical power, commonly because of small samples or participant dropout; 59% had no follow-up phase; and 36% did not discuss generalizability. Geographic representation was uneven, with 10 studies from the Middle East and North Africa, including eight from Iran, compared with two from South Asia. Other limitations included a relatively small review, exclusion of gray literature, reliance on English-language publications, subjective methodological ratings, and possible differences in how adolescents across cultures perceive and report symptoms. The authors called for larger, adequately powered trials, longer follow-up, registered analysis plans, culturally validated measures, and further comparisons of provider types and settings, especially for anxiety.

Subject of Research: Effectiveness of psychological interventions for subclinical adolescent depression and anxiety in low- and middle-income countries

Article Title: Mental health interventions for adolescents with depression and anxiety in low- and middle-income countries: a systematic review of the evidence

Article References: Yin, S., Tian, T., & Auslander, W. (2026). Mental health interventions for adolescents with depression and anxiety in low- and middle-income countries: a systematic review of the evidence. Frontiers in Psychology, 17, Article 1809434. https://doi.org/10.3389/fpsyg.2026.1809434

Image Credits: AI Generated

DOI: 10.3389/fpsyg.2026.1809434

Keywords: systematic review, adolescent mental health, depression, anxiety, task shifting, non-specialist providers, low- and middle-income countries, GRADE, school-based interventions, Mental, health, interventions