Purpose This study aims to investigate depression, its significant association and risk factors with demographic and psychosocial characteristics of public university students in Bangladesh. Design/methodology/approach This study used a cross-sectional quantitative research design using the Beck Depression Inventory Scale (BDI-II). A total of 689 valid responses from four public universities in Bangladesh were collected using a stratified random sampling technique. The bivariate chi-squared test was performed for association, and a binary logistic regression was used to isolate psychosocial risk factors. Findings The prevalence of depression among tertiary-level students was 65.6%. The logistic regression analysis identified female gender (OR = 2.236, 95%CI = 1.528, 3.271), lower maternal education (OR = 2.215, 95%CI = 1.204, 4.074), underweight BMI (OR = 4.172, 95%CI = 1.116, 14.921), below normal sleep duration (OR = 2.292, 95%CI = 1.271, 4.133), lack of extracurricular activities (OR = 1.598, 95%CI = 1.112, 2.298), history of COVID-19 infection (OR = 1.558, 95%CI = 1.025, 2.370), dissatisfaction with government initiatives for COVID-19 control (OR = 1.625, 95%CI = 1.061, 2.489) and youth development (OR = 1.830, 95%CI = 1.212, 2.765), and perceiving a lack of a professional environment (OR = 1.975, 95%CI = 1.363, 2.863) as significant predictors of higher odds of depression. Interestingly, being married (OR = 0.366, 95%CI = 0.157, 0.855) was associated with lower odds of depression, which warrants further exploration within the specific cultural context of Bangladesh. Social implications These findings emphasize the importance of targeted mental health interventions, policy reforms and academic support programs to address depression among students in higher education institutions. Originality/value This study addressed a significant knowledge gap, as previous studies primarily focused on specific subgroups of students in Bangladesh while predicting depression among university students.
Muyeed et al. (Sat,) studied this question.