Stress among medical students is a recognized global health concern driven by intensive academic demands and clinical responsibilities. Prior regional studies indicate high rates of depression, anxiety, and stress among students in South and Southeast Asia. Bangladesh shows elevated burdens of mental health conditions and suicidality relative to global rankings. This study aimed to quantify perceived stress among clinical-year medical students in Dhaka and to identify specific academic and psychological determinants using the Perceived Stress Scale-10 (PSS-10) and multinomial logistic regression.
A descriptive cross-sectional study was conducted covering the period January 1 to December 31, 2022, with data collection from September 3 to September 21, 2022. Two medical colleges in Dhaka were selected by stratified random sampling to represent one public (Shaheed Suhrawardy Medical College & Hospital) and one private (Shaheed Monsur Ali Medical College and Hospital) institution. The target population comprised 3rd, 4th, and 5th year MBBS students. The minimum calculated sample size was 345; allowing for 10% non-response, the final target was 380 and 380 valid responses were obtained. Participants provided written informed consent.
A pre-tested, structured, self-administered questionnaire (English and Bengali versions) captured sociodemographic data, a broad set of academic activities–related items, psychological factors, and the PSS-10. The PSS-10 yields scores from 0 to 40, classified as low (0–13), moderate (14–26), and high (27–40) stress. The questionnaire covered academic domains (for example: academic performance, study and exam score anxiety, curriculum difficulty, administration and lecture satisfaction, teaching strategy, availability of learning materials, recreation time) and psychological variables (for example: time management skills, parental support, social life perception, loneliness, interpersonal conflicts).
Completed questionnaires were reviewed daily and entered into IBM SPSS Statistics (version 23). Data preprocessing included editing, coding, recoding, and handling a low proportion of missing values (<5%) via mode imputation. Bivariate associations were assessed with chi-square tests; predictors with p<0.10 in bivariate analysis were advanced to regression. Multinomial logistic regression models were fitted to identify factors associated with moderate and high stress relative to low stress. Three regression models were specified: Model 1 (academic variables), Model 2 (psychological variables), and Model 3 (combined academic and psychological variables). Results were reported as Relative Risk Ratios (RRRs) with p-values. Analyses were conducted with two-sided tests and a significance level set at 10% using SPSS v23 and Stata v14.2.
Among 380 respondents, 68.7% were female and 51.8% were from Shaheed Suhrawardy Medical College. The mean PSS-10 score was 21.64 (SD 6.64). Using PSS-10 cutoffs, 67.9% of students were classified as having moderate stress and 22.6% as having high stress. Bivariate analysis identified 29 academic factors and 9 psychological factors that were significantly associated with stress at the selected threshold and were thus considered in regression modeling.
Model 1 (academic factors): Significant academic correlates of increased likelihood of moderate or high stress included poor academic performance indicators, study and exam score anxiety, fear of future uncertainty and changing living conditions, lower administration satisfaction, and lack of entertainment facilities.
Model 2 (psychological factors): Significant psychological determinants included inadequate time management skills, negative social life perception, and the feeling of loneliness.
Model 3 (combined academic and psychological factors): When academic and psychological variables were entered together, significant predictors of stress included study and exam score worry, difficulty understanding context, fear of new living situations, administration and lecture satisfaction, teaching strategy satisfaction, time management, parental support for study, and feeling lonely.
The models produced RRRs and p-values to indicate the magnitude and statistical significance of associations for moderate and high stress compared with the low-stress reference group. The study used a 10% significance threshold for inclusion and interpretation.
The Institutional Review Board of the National Institute of Preventive and Social Medicine (NIPSOM) approved the protocol. Institutional permissions were granted, and informed written consent was obtained from each participant. Data were anonymized; the minimal dataset was archived in a public repository (DOI provided in the source) and managed for long-term access by a designated institutional contact. Hard copies and the SPSS dataset are retained by the corresponding author in accordance with ethical approvals.
The study found a high prevalence of perceived stress among clinical-year medical students in Dhaka, with most students classified as moderately stressed and a substantial minority highly stressed. Both academic pressures—particularly exam-related anxiety, perceived administrative and teaching shortcomings, and fears about future living—and psychological factors such as poor time management and loneliness were associated with higher stress levels. The authors recommend targeted interventions addressing identified academic and psychological stressors, further assessment of students with moderate and high stress, and rehabilitation strategies to reduce the burden of mental stress among medical students.
The source reports that the study was cross-sectional and limited to two medical colleges in Dhaka, which may affect generalizability across all Bangladeshi medical students. The report also notes that mode imputation was used for a small proportion of missing values (<5%). Other limitations referenced in the source text were not detailed in this summary because specific additional limitations were not reported in the provided material.