MULTIVARIATE LOGISTIC REGRESSION ANALYSIS OF DEPRESSION RISK FACTORS AMONG NIGERIAN UNIVERSITY STUDENTS

Abstract

Depression is a growing public health issue among university students, yet limited studies in Nigeria have applied multivariate techniques to identify its key predictors across varying institutional contexts. This study employed a cross-sectional design involving 1,200 undergraduate students from three Federal Universities in South-West Nigeria— Federal University of Agriculture, Abeokuta (FUNAAB); the Federal University of Technology, Akure (FUTA); and the University of Lagos (UNILAG). Data were gathered through a structured questionnaire integrating validated instruments such as the Beck Depression Inventory-II (BDI-II), Sexual Risk Survey, Bullying Victimization Scale, Alcohol Use Disorders Identification Test (AUDIT), and self-developed measures of academic stress and performance. Depression, treated as a binary outcome, was analyzed using multivariate logistic regression. Variables with a p-value < 0.20 from univariate analysis were included in the final model. About 30% of students reported moderate-to-severe depression. Significant predictors included female gender (Adjusted Odds Ratio (AOR) = 1.62), age 21–23 (AOR = 1.41), high academic stress (AOR = 2.18), self-sponsorship (AOR = 1.72), high-risk alcohol use (AOR = 1.94), bullying victimization (AOR = 3.27), and risky sexual behavior (AOR = 1.68). The model demonstrated good fit (Hosmer-Lemeshow p = 0.48) and strong explanatory power (Nagelkerke’s R² = 0.31). The findings highlight the multifaceted nature of depression and the importance of targeted interventions within university settings.