Abstract
BACKGROUND: Suicide is a major public health concern in India, with rising rates among adolescents. School-going adolescents are particularly vulnerable due to academic pressure and related mental health issues, especially depression. OBJECTIVES: While prior studies have examined the prevalence and individual predictors of suicidal ideation (SI), limited research has explored the mechanism linking academic stress (AS) to SI through major depressive symptoms (MDS), particularly in under-researched regions like Manipur, Northeast India. This study examines the associations between AS, MDS, and SI, and whether MDS mediates the relationship between AS and SI, adjusting for a range of socio-demographic factors. METHODS: A cross-sectional study was conducted among 861 school-going adolescents (Grades 9-12) from six randomly selected schools in Imphal, Manipur, between August and October 2024. Data were collected using self-administered questionnaires, assessing AS (Educational Stress Scale for Adolescents [ESSA]), MDS (Patient Health Questionnaire-8 [PHQ-8]), and SI (Item 9 of the Patient Health Questionnaire-Adolescents [PHQ-A]) and socio-demographics. Analyses included descriptive statistics, chi-square tests, logistic regression, and mediation analysis. RESULTS: High AS (88.3%), positive for MDS (41.7%), and SI (43.9%) were prevalent. SI prevalence was significantly higher among students with high AS (47.5%) and MDS (69.4%). Logistic regression identified both AS (OR = 2.71) and MDS (OR = 5.49) as significant predictors of SI. Mediation analysis confirmed that AS predicted MDS (OR = 12.02), which in turn predicted SI, with a significant partial mediation effect. Causal mediation analysis revealed that 51% of AS's total effect on SI was mediated through MDS (ACME = 0.125, p < 0.001). CONCLUSION: AS predicts SI directly and via MDS, indicating partial mediation. Findings advocate the need for early, school-based interventions targeting both AS and MDS, especially in high-risk regions like Manipur, Northeast India.