Academic and psychological determinants of mental stress among medical students in Dhaka: A multinomial logistic regression approach Maliha Khan Majlish, Mushfiqur Rahman Khan Majlish, Sabrin Shohid, Md Samiul Islam, A N M Shamsul Islam ABSTRACT As stress among medical students is a global health concern due to comprehensive academic engagements, this study aimed to assess the educational and psychological factors related to mental stress among medical students in Dhaka city. A descriptive type of cross-sectional study was conducted in two medical colleges in Dhaka city among 380 medical students from 3rd year to 5th year, using a pre-tested, self-administered, and structured questionnaire. The data obtained were analyzed using IBM SPSS Statistics (version 23) and Stata (version 14.2) employing the chi-square association test for bivariate analysis and subsequent multinomial logistic regression for the significant variables. Among 380 respondents, 68.7% were female, with 51.8% of respondents from Shaheed Suhrawardy Medical College. The mean stress score of medical students was 21.64 (SD = 6.64), with 67.9% and 22.6% as moderately and highly stressed, respectively. In bivariate analysis, 29 academic factors and 09 psychological factors were significantly associated with stress. In multinomial logistic regression analysis, three models were conducted, containing variables related to academic stress in model 1, variables related to psychological stress in model 2, and all stress variables in model 3. From model 1, academic performance, study and exam score anxiety, fear of future uncertainty and new living, administration satisfaction, and lack of entertainment facilities are significantly associated with stress. Additionally, the factors significantly associated with stress are time management, social life perception, and feeling lonely for model 2 and all stress-related variables, including study and exam score worry, difficulty in understanding context, fear of new living, administration and lecture satisfaction, teaching strategy satisfaction, time management, parental support in study, and feeling loneliness for model 3. Proper interventions for stressors of mental stress, further assessment, and rehabilitation for high and moderate-stressed respondents could be a prime concern to lessen the burden of stress among medical students. Introduction Despite modern advancements, mental disorders persist as a major global health crisis. Stress, triggered by various situations when powerless or unable to achieve desired outcomes, is the body’s response to mental strain, releasing stress hormones and activating the defense system. However, this response can have both positive and negative effects on mental and physical health. Global statistics indicate 61% of adults under stress feel nervous, and 51% feel depressed [ 1 ]. Developed countries like the Netherlands (depression, anxiety: 20%, 17%), Saudi Arabia (stress: 33.8%), Argentina, Canada, and the USA show significant but lower ranges of mental health problems, with medical students facing higher distress than the general population [ 2 – 5 ]. Being widespread among medical students globally, anxiety and depression rates are high in the Middle East and Asia. In Southeast Asia, Pakistan (depression in 57.6%, anxiety in 74%, and stress in 57.7% of students) [ 6 ], India (depression 8.7% to 71.3%, anxiety 34.5%, and stress 51.3%) [ 7 ], Nepal (20.9% of psychological morbidity) [ 8 ], China (depression, anxiety, stress: 12.26%, 18.47%, 8.53%), Turkey (depression, anxiety, stress: 18.6%, 26.5%, 7.9%), Egypt (depression, anxiety, stress: 64.2%, 77.1%, 70.4%), and Malaysia (stress: 33.3%) reveal significant prevalence rates of anxiety, depression, and stress among students and medical professionals [ 9 – 12 , 48 ], emphasizing the global concern for mental health, regardless of cultural or socio-demographic differences [ 2 – 5 , 7 – 13 ] Moreover, Bangladesh ranks 153rd globally in suicide rates, with approximately 7 million people diagnosed with anxiety and depression, accounting for 0.84% of all fatalities [ 14 , 15 ]. Depression (57.9%), anxiety (33.7%), stress (59.7%), and sleep disorders are prevalent in Bangladesh [ 16 ], with medical students experiencing higher stress (54.6%), anxiety (65.9%), and depression (49.9%), primarily due to studies [ 17 , 18 ]. Bangladesh has 106 medical colleges, including 36 public, numerous private, and 6 under the Ministry of Defense, collectively enrolling over 10,000 students annually [ 22 ]. The MBBS program spans five years, followed by a mandatory one-year internship, with multiple phases of examinations that place substantial academic pressure on students. Stress affects 54% of medical students in Bangladesh, while the percentages of stress in medical students are comparable with respect to sex. Key stressors include workload, time constraints, family expectations, lack of practical training, academic pressures, and fear of entering the medical field, etc., which are all major contributors to mental health issues [ 18 , 23 – 25 ]. Lifestyle factors such as excessive screen time, physical inactivity, poor sleep, irregular meal frequency, and alcohol consumption are closely linked to depression and anxiety [ 20 ]. While some students develop coping strategies in their middle years, the cumulative demands of medical school continue to heighten psychological vulnerability [ 26 , 27 ]. Given that depression, stress, and anxiety among medical students are well-documented globally [ 18 – 21 , 23 – 25 ], understanding their prevalence and underlying causes in Bangladesh is imperative, as evidence for academic and psychological stressors remains limited, fragmented, and often focused on single institutions or small student groups. Few studies have systematically assessed perceived stress using a validated tool, and limited exploration of academic and psychological stressors specific to the country’s unique social context hampers policymakers and educators from designing targeted mental health interventions for medical students. This study aimed to identify the academic and psychological determinants of mental stress among medical students in Dhaka using the Perceived Stress Scale (PSS) and a multinomial logistic regression approach. Specifically, the objectives were to (i) measure the prevalence of perceived stress, (ii) examine the key academic and psychological factors associated with varying levels of stress, and (iii) generate evidence to inform policies and interventions that promote mental health and wellbeing among medical students. Methodology Study design and setting A descriptive cross-sectional study was conducted among medical students in Dhaka to evaluate their mental stress and associated factors related to academic activities from January 1 to December 31, 2022, with data collection period from September 3 to September 21, 2022. Two medical colleges, such as one government medical college as Shaheed Suhrawardy Medical College & Hospital, and one private medical college as Shaheed Monsur Ali Medical College and Hospital, around Dhaka city, were selected through stratified sampling techniques to ensure representation from both public and private institutions. The study population included students from 3 rd , 4 th , and 5 th academic years in these institutions, as students of this group are simultaneously engaged in intensive academic coursework, ward duties, and examinations. Moreover, the well-documented academic and psychological burden faced by Bangladeshi medical students has been established during their clinical years [ 24 , 28 , 29 ], creating a scenario where stress is both highly prevalent and diverse in its manifestations, making this group suitable for the study objectives. Sample size and sampling technique The required sample size was determined using the formula n = z 2 p q d 2 , where z = 1.96 (standard normal deviate at 95% confidence level), p = 0.50 (assumed prevalence of stress among medical students based on previous studies in Bangladesh to maximize sample size), q = 1–p, and d = 0.05 (margin of error). Based on this calculation, the minimum required sample size was 345. With a 10% non-response rate, the final target sample size was 380 students. At the time of data collection, the total number of 3 rd –5 th year students was 618 across the two selected medical colleges (Shaheed Suhrawardy Medical College: 319 students; Shaheed Monsur Ali Medical College: 299 students). A stratified random sampling method was employed to ensure representation from both public and private medical institutions in Dhaka. i) Stratification: All medical colleges in Dhaka were divided into two strata based on type: public medical colleges and private medical colleges. ii) Random selection: One medical college from each stratum was selected using simple random sampling, resulting in Shaheed Suhrawardy Medical College (public) and Shaheed Monsur Ali Medical College (private). iii) Census within selected colleges: All eligible 3 rd , 4 th , and 5 th -year students who provided informed consent in the selected medical colleges were invited to participate, exclusion criteria ruled out those who did not consent or were unable to participate due to severe illness or acute psychological distress. Pre-testing A pre-test was conducted at Bangladesh Medical College & Hospital using the same structured self-administered questionnaire. A total of 55 students were recruited after authorization from the college principal. Self-administered structured questionnaires were distributed and collected four days later, with an additional two days allowed if necessary. Pre-test results showed a response rate of 72.7% (n = 40). Based on the pre-test results, minor modifications were made to improve clarity and flow of the questions, including rewording ambiguous items and adjusting the sequence for better logical order. The revised questionnaire was then used at the study sites. Data collection procedures Institutional permissions were obtained, and departmental teachers facilitated student briefings. A pre-tested, structured, self-administered questionnaire and informed consent form were distributed. On average, the questionnaire comprised four sections and required 25–30 minutes to complete. The study team accounted for real-life challenges faced by medical students during data collection; for example, many students were on ward rotations or preparing for examinations. To minimize disruption, the questionnaire distribution was integrated into existing departmental briefings, with students allowed flexibility in returning responses within a week. Teachers also played a supportive role by encouraging participation while ensuring voluntary and anonymous involvement. Completed responses were collected over three weeks, yielding a total of 380 valid responses ( Fig 1 ). 10.1371/journal.pone.0354261.g001 Fig 1 Flow of study data collection procedures. Study materials Semi-structured questionnaire The questionnaire was created in both English and Bengali, with the English version distributed to respondents. It consisted of four sections covering: Socio-demographic characteristics, Academic activities-related questions, Factors associated with mental stress, and Perceived Stress Scale (PSS). Independent and dependent variables. The dependent variable comprised the mental stress score utilizing the Perceived Stress Scale-10, while the independent variables comprised the following: Sociodemographic variables : Respondent age group, sex, Medical College name, religion, nationality, academic year, marital status, family type, place of residence, accommodation, parents’ occupation, whether parents bear educational expense, whether the student has a part-time job, etc. Academic activities related variables : Academic performance, attending sessions, concentration in sessions, understanding sessions, perception of academic sessions curriculum difficulty, academic environment learning material availability, study and exam score anxiety, pressure of studies and passing exams, time management pressure, difficulty in understanding context, parent’s and teacher’s expectation pressure, fear of future uncertainty and new living, fear of dropping out of course, future choice cluelessness, administration satisfaction, lecture satisfaction, exam criteria satisfaction, career choice satisfaction, lack of entertainment facility, teaching strategy satisfaction, course activity feeling, interaction with other medical students, recreation time Psychological factors : Time management skills, parental support in study, social life perception, feeling lonely, studies hampering social interaction, illness hampering study, classmate conflict, family conflict, roommate conflict. Perceived stress scale-10 Though several scales are available to assess stress among medical students, the Perceived Stress Scale (PSS-10) is widely validated, brief, and easily adaptable to local languages, making it suitable for cross-cultural studies. In contrast, other widely validated tools measure depression, anxiety, and stress [ 30 – 32 ] simultaneously, but are comprehensive with less focus on perceived stress alone. Other self-reported questionnaires can target specific stressors but often require cultural adaptation and validation. Considering these factors, the PSS-10 was selected for this study due to its brevity, ease of administration, and suitability for assessing perceived stress in Bangladeshi medical students. The PSS-10, developed by Cohen et al. (1983), is a widely validated psychological tool used to assess perceived stress over the past month, evaluating how unpredictable, uncontrollable, and overwhelming respondents find their lives. It consists of 10 items on a 5-point Likert scale (0 = never to 4 = very often), with items 4, 5, 7, and 8 being reverse-coded (0 = 4, 1 = 3, 2 = 2, 3 = 1, 4 = 0), and the sum of all items yields a total score ranging from 0 to 40. Stress levels are classified as: Low stress (0−13 scores); Moderate stress (14−26 scores); High Stress (27−40 scores) (33). To ensure better comprehension, the questionnaire was administered in both English and Bangla, using a validated Bangla version of the PSS-10 [ 34 ]. Among 380 medical students surveyed, responses were obtained using a 10-item PSS scale, administered in both English [ 33 , 35 ] and Bangla [ 34 , 36 ] version of the English self-administered questionnaire for better comprehension. Data management Data pre-processing. At the end of each day of data collection, each questionnaire was reviewed for completeness and consistency, and data entry commenced immediately afterward. Once all data were collected, they were processed and tabulated, involving careful editing, coding, recoding, summation, and categorization to exclude irrelevant and unreliable information. The missing value count was low (less than 5%). The missing value count was low (less than 5%) and was addressed using mode imputation. Data analysis. The Chi-square association test was conducted for bivariate analysis to identify the unadjusted significance between the dependent and independent variables without controlling for other factors. Predictors with a p-value <0.10 in bivariate analysis were considered in the regression model. Multinomial logistic regression was employed to identify the factors associated with varying levels of stress (low, moderate, and high) among medical students. The study included three models: Model 1 comprised variables related to academic stress, Model 2 comprised variables related to psychological stress, and Model 3 included all stress-related variables. The results are presented as Relative Risk Ratios (RRRs) with corresponding p-values, indicating the likelihood of being in the moderate or high stress group compared to the low stress reference group, given the presence of specific predictors. All statistical tests were two-sided, and the significance level was set at 10%. IBM SPSS Statistics (version 23) and STATA version 14.2 were used to conduct statistical analysis. Ethical considerations Regarding ethical implications, the protocol was approved by the Institutional Review Board (IRB) of the National Institute of Preventive and Social Medicine (NIPSOM), and permissions from the medical colleges were obtained. Informed written consent was obtained from each respondent, and the study objectives were briefly explained to them. Privacy and confidentiality were strictly maintained, and no personal identifiers were collected. Participants had the right to withdraw from the study at any time and were assured that there would be no physical or mental harm, as no invasive procedures were involved. Results Socio-demographic factors In this study, Fig 2 shows the average PSS-10 score was 21.64 with a standard deviation of 6.64 (minimum: 3 and maximum: 39). 258 respondents (67.9%) had a moderate stress level among all 380 respondents. The perceived stress scores range between 15–27. 10.1371/journal.pone.0354261.g002 Fig 2 Histogram of PSS-10 score among the among medical students (n = 380). Among 380 respondents, 96.6% of respondents were aged between 21–25, and 68.7% of respondents were female. 51.8% of respondents were from Shaheed Suhrawardy Medical College, and the remaining from Shaheed Monsur Ali Medical College. 86.8% belong to the Islamic religion, and 15.8% of respondents were foreign nationals. Almost an equal number of respondents from different academic years participated in the study. 92.4% were unmarried respondents, and 83.9% came from nuclear families. 71.3% resided in urban residences, and 62.9% lived in hostels. In case of parents’ occupation, 38.2% of respondents’ fathers worked in government sectors, whereas 69.7% of respondents’ mothers were homemakers. 88.2% of respondents’ parents bear their educational expenses, and 74.2% of students did not have a part-time job. Academic factors Table 1 shows the academic factors that are significantly associated with perceived stress in the study sample at the 10% significance level. Almost an equal number of respondents have different perceptions of their academic performance (satisfactory or not satisfactory). 91.1% of students attended academic sessions regularly, and 69.5% gave concentration in these sessions. 66.3% understood the studies delivered in these sessions, and 78.4% found these sessions beneficial. 73.9% of students found the medical curriculum difficult. Almost an equal number of students were either satisfied or dissatisfied with the academic environment, and 68.2% found the availability of learning materials. 84.7% of respondents worry about studying, and 74.7% worry about poor exam scores. Almost an equal number of students felt the pressure of studying, passing exams, and time management in studies (93.9%, 93.4% and 91.8%). 63.7% of respondents felt difficulty in understanding context, and 95.3% felt the vastness of the academic syllabus. 62.9% felt the pressure of fulfilling parents’ expectations, and 68.4% felt the pressure of fulfilling teachers’ expectations. 79.7%, 66.7% and 56.1% of students were fearful of future uncertainty, new way of living, and dropping out of the course, respectively. 67.4% of respondents were clueless about future choices of specialty. 31.6% were satisfied with the administration, and 35.5% were satisfied with the examination criteria. An equal number of students were satisfied with class lectures and career choice (54.5% and 56.5% respectively). 41.3% were satisfied with teaching strategies, and 42.4% were satisfied with course activities. 59.7% of respondents had a satisfactory interaction with other medical students. 81.3% of students felt the lack of entertainment facilities in institutions, and 82.1% of students had no time for recreation. 10.1371/journal.pone.0354261.t001 Table 1 Academic factors associated with perceived stress (N = 380). Variable Number of Observations (%) P value* Academic Performance Satisfactory 186 (48.9%) 0.000 Not Satisfactory 194 (51.1%) Attending Sessions Yes 346 (91.1%) 0.002 No 34 (8.9%) Concentration in Sessions Yes 264 (69.5%) 0.000 No 116 (30.5%) Understanding Sessions Yes 252 (66.3%) 0.000 No 128 (33.7%) Perception of Benefit of Academic Sessions Beneficial 298 (78.4%) 0.069 Not Beneficial 82 (21.6%) Curriculum Difficult Yes 281 (73.9%) 0.000 No 99 (26.1%) Academic Environment Satisfactory 187 (49.3%) 0.001 Not Satisfactory 192 (50.5%) Learning Material Availability Available 259 (68.2%) 0.066 a Not Available 121 (31.8%) Study Worry Yes 322 (84.7%) 0.000 No 58 (15.3%) Exam Score Worry Yes 284 (74.7%) 0.000 No 96 (25.3%) Pressure of Studies Yes 357 (93.9%) 0.010 No 23 (6.1%) Pressure of Passing Exams Yes 355 (93.4%) 0.000 No 25 (6.6%) Time Management Pressure Yes 349 (91.8%) 0.001 No 31 (8.2%) Difficulty in Understanding Context Yes 242 (63.7%) 0.000 No 138 (36.3%) Parent’s Expectation Pressure Yes 239 (62.9%) 0.011 No 141 (37.1%) Teacher’s Expectation Pressure Yes 260 (68.4%) 0.003 No 120 (31.6%) Future Uncertainty Fear Yes 303 (79.7%) 0.000 No 77 (20.3%) Fear of New Living Yes 254 (66.8%) 0.000 No 126 (33.2%) Fear of Dropping Out of Course Yes 215 (56.6%) 0.000 No 165 (43.4%) Future Choice Cluelessness Yes 256 (67.4%) 0.000 No 124 (32.6%) Administration Satisfaction Yes 119 (31.3%) 0.000 No 261 (68.7%) Lecture Satisfaction Yes 207 (54.5%) 0.000 No 173 (45.5%) Exam Criteria Satisfaction Yes 135 (35.5%) 0.000 No 245 (64.5%) Career Choice Satisfaction Yes 215 (56.6%) 0.000 No 165 (43.4%) Lack of Entertainment Facility Yes 309 (81.3%) 0.003 No 71 (18.7%) Teaching Strategy Satisfaction Yes 157 (41.3%) 0.032 No 233 (58.7%) Course Activity Feeling Satisfactory 161 (42.4%) 0.000 Not Satisfactory 219 (57.6%) Medical Students Interaction Satisfactory 227 (59.7%) 0.023 Not Satisfactory 153 (40.3%) Recreation Time Yes 68 (17.9%) 0.000 No 312 (82.1%) In * column, significant when Confidence Interval (CI) 90%, p < 0.1. Among academic factors, 29 variables namely perception of academic performance, attending academic session, concentration in academic sessions, understanding academic sessions, perception of benefit of academic sessions, finding medical curriculum difficult, satisfaction with academic environment, availability of learning material, worry about study, worry about exam score, pressure of studies, pressure of passing exams, pressure of time management, difficulty in understanding context, pressure of parents’ expectations, pressure of teachers’ expectations, fear of future uncertainty, fear of new living, fear of dropping out of course, cluelessness of future choice, administration satisfaction, lecture satisfaction, exam criteria satisfaction, career choice satisfaction, lack of entertainment facility, satisfaction in teaching strategy, feeling about course activity, interaction with medical students’ and recreation time were significantly associated at the 10% significance level. The remaining academic factors are insignificant at the 10% level of significance. Psychological factors Table 2 exhibits the other factors that are significantly associated with perceived stress in the study sample at 5% significance level. Among 380 medical students, 27.6% were satisfied with their time management skills, and 86.3% had parental support in study. 35.3% of the students were satisfied with their current social life, and 66.8% were suffering from loneliness. 68.4% felt that studies hamper social interaction, and 53.2% felt that illness hampers study. 18.9% of students are engaged in conflicts with their family members, and an equal number of students are involved in conflict with roommates and classmates (23.7%). 10.1371/journal.pone.0354261.t002 Table 2 Psychological factors associated with perceived stress (N = 380). Variable Number of Observations (%) P value* Time Management Skill Satisfactory 105 (27.6%) 0.000 Not Satisfactory 275 (72.4%) Parental Support in Study Present 328 (86.3%) 0.041 Absent 52 (13.7%) Social Life Perception Satisfactory 134 (35.3%) 0.000 Not Satisfactory 246 (64.7%) Feeling Lonely Yes 254 (66.8%) 0.000 No 126 (33.2%) Studies Hamper Social Interaction Yes 260 (68.4%) 0.021 No 120 (31.6%) Illness Hampers Study Yes 202 (53.2%) 0.005 No 178 (46.8%) Classmate Conflict Yes 90 (23.7%) 0.042 No 290 (76.3%) Family Conflict No 308 (81.1%) 0.012 Yes 72 (18.9%) Roommate Conflict No 290 (76.3%) 0.008 Yes 90 (23.7%) In * column, all values are significant when Confidence Interval (CI) 95%; p < 0.10. Among psychological factors, 09 variables, namely time management skill, parental support in study, perception of social life, feeling lonely, studies hamper social interaction, illness hampers study, classmate conflict, family conflict, and roommate conflict, were significantly associated at the 10% significance level, respectively. The remaining factors are insignificant at the 10% level of significance. Multivariate analysis Academic factors. Table 3 exhibits the relative risk ratio (RRR) for various academic factors on Perceived Stress and their corresponding significance among participating medical students. RRR and significance values for each academic factor are measured for moderate and high perceived stress. From the p-value, it is observed that anxiety regarding exam scores and satisfaction with administration are significantly associated with moderate stress of the students at the 10% level of significance. The RRR value is 0.392 for anxiety related to exam scores, which means that students who do not worry about their exam scores are 60.8% less likely to experience moderate stress compared to those who do worry. Moreover, the RRR value is 3.598 for administration satisfaction, which means that students who are not satisfied with the administration system are 3.598 times more likely to experience moderate stress compared to those who are satisfied with the administration. The remaining academic factors are insignificant at 5% level of significance. 10.1371/journal.pone.0354261.t003 Table 3 Relative risk ratio of academic factors on perceived stress (N = 380). Variable Category RRR (moderate) 95% CI P value (moderate) RRR (high) P value (high) 95% CI Academic Performance Satisfactory 1 – 1 – Unsatisfactory 2.265 (0.703, 7.331) 0.172 2.979 0.099a (0.815, 10.883) Attending Sessions Yes 1 – 1 – No 0.443 (0.507, 3.403) 0.434 1.054 0.962 (0.120, 9.233) Concentration in Sessions Yes 1 – 1 – No 0.751 (0.192, 2.941) 0.682 1.051 0.947 (0.241, 4.584) Understanding Sessions Yes 1 – 1 – No 1.103 (0.308, 3.945) 0.879 1.078 0.308 (0.509, 8.490) Perception of Sessions Beneficial 1 – 1 – Not Beneficial 1.331 (0.311, 5.683) 0.699 1.106 0.901 (0.226, 5.394) Curriculum Difficulty Yes 1 – 1 – No 0.760 (0.281, 2.054) 0.589 0.786 0.712 (0.219, 2.814) Academic Environment Satisfactory 1 – 1 – Unsatisfactory 0.767 (0.221, 2.656) 0.676 0.589 0.460 (0.145, 2.397) Learning Material Availability Available 1 – 1 – Not Available 0.580 (0.174, 1.935) 0.376 0.528 0.354 (0.137, 2.037) Study Worry Yes 1 – 1 – No 0.510 (0.162, 1.603) 0.250 0.140 0.085a (0.015, 1.308) Exam Score Worry Yes 1 – 1 – No 0.392 (0.144, 1.064) 0.066a 0.216 0.028b (0.055, 0.846) Pressure of Studies Yes 1 – 1 – No 0.982 (0.169, 5.684) 0.984 1.156 0.919 (0.071, 18.891) Pressure of Passing Exams Yes 1 – 1 – No 0.328 (0.067, 1.608) 0.170 1.123 0.928 (0.091, 13.935) Time Management Pressure Yes 1 – 1 – No 1.259 (0.293, 5.396) 0.756 0.406 0.508 (0.028, 5.858) Difficulty in Understanding Context Yes 1 – 1 – No 1.669 (0.591, 4.715) 0.333 1.489 0.523 (0.438, 5.062) Parent’s Expectation Pressure Yes 1 – 1 – No 1.649 (0.518, 5.247) 0.397 1.926 0.334 (0.508, 7.298) Teacher’s Expectation Pressure Yes 1 – 1 – No 1.020 (0.337, 3.086) 0.972 1.061 0.930 (0.280, 4.020) Fear of Future Uncertainty Yes 1 – 1 – No 0.593 (0.171, 2.049) 0.409 0.133 0.050a (0.018, 0.998) Fear of New Living Yes 1 – 1 – No 0.412 (0.134, 1.266) 0.122 0.281 0.062a (0.074, 1.064) Fear of Dropping Out of Course Yes 1 – 1 – No 0.776 (0.264, 2.277) 0.645 0.537 0.318 (0.158, 1.819) Future Choice Cluelessness Yes 1 – 1 – No 0.686 (0.267, 0.176) 0.435 0.904 0.865 (0.282, 2.892) Administration Satisfaction Yes 1 – 1 – No 3.598 (1.204, 10.785) 0.022b 3.770 0.045b (1.031, 13.779) Lecture Satisfaction Yes 1 – 1 – No 1.743 (0.454, 6.695) 0.418 3.371 0.112 (0.753, 15.089) Exam Criteria Satisfaction Yes 1 – 1 – No 0.670 (0.184, 2.433) 0.543 0.901 0.892 (0.202, 4.022) Career Choice Satisfaction Yes 1 – 1 – No 1.096 (0.351, 3.417) 0.874 1.762 0.382 (0.494, 6.284) Lack of Entertainment Facility Yes 1 – 1 – No 0.490 (0.184, 1.303) 0.153 0.273 0.060a (0.071, 1.054) Teaching Strategy Satisfaction Yes 1 – 1 – No 1.059 (0.292, 3.843) 0.930 0.317 0.130 (0.071, 1.399) Course Activity Feeling Satisfactory 1 – 1 – Unsatisfactory 0.480 (0.138, 1.662) 0.247 0.740 0.685 (0.173, 3.157) Medical Students’ Interaction Satisfactory 1 – 1 – Unsatisfactory 1.924 (0.638, 5.801) 0.245 1.775 0.359 (0.521, 6.053) Recreation Time Yes 1 – 1 – No 0.858 (0.306, 2.405) 0.772 2.760 0.184 (0.618, 12.331) a significant when Confidence Interval (CI) 90%, p < 0.1. b significant when Confidence Interval (CI) 95%; p < 0.05. In case of association with high stress, perception of academic performance, anxiety related to study and exams, fear of future uncertainty and new living, administration satisfaction, and lack of entertainment facilities are significantly associated with perceived stress at the 10% level of significance. From RRR, students who are not satisfied with their academic performance are 2.979 times higher relative risk to experience high stress than low stress compared to those who are satisfied with their academic performance. Anxiety related to study and exams is also significantly associated with high stress. Students who were not worried about their studies had an 86.0% lower relative risk of being in the high stress group rather than the low stress group compared with students who were worried about their studies (RRR = 0.140). Similarly, students who do not worry about their exams have a 78.4% lower relative risk of experiencing stress than those who worry about exam scores. Students who are not fearful of future uncertainty had an 86.7% lower relative risk of experiencing stress than those who are uncertain about their future. Additionally, students who are not fearful of new living have a 71.9% lower relative risk of experiencing high stress compared to those who are fearful of new living. Similarly, moderate stress and administration satisfaction are also significantly associated with high perceived stress. Students who are not satisfied with the administration have a 3.77 times higher relative risk of experiencing high stress than those who are satisfied. Lack of entertainment facilities is also associated with high perceived stress, and students who reported no lack of entertainment facilities had a 72.7% lower relative risk of being in the high stress group rather than the low stress group compared with students who reported a lack of entertainment facilities (RRR = 0.273). The remaining variables are not significant at the 10% level of significance. Psychological factors Table 4 exhibits the relative risk ratio (RRR) for various psychological factors on Perceived Stress and their corresponding significance among participating medical students. RRR and significance values for each psychological factor are measured for moderate and high perceived stress. From the p-value, it is observed that perception of time management and feeling of loneliness are significantly associated with both moderate and high perceived stress, and perception of social life is significantly associated with high perceived stress at 5% level of significance. 10.1371/journal.pone.0354261.t004 Table 4 Relative risk ratio of psychological factors on perceived stress (N = 380). Variable Category RRR (moderate) 95% CI P value (moderate) RRR (high) 95% CI P value (high) Time Management Satisfactory 1 – 1 – Unsatisfactory 4.819 (2.0845, 11.144) 0.000 a 13.827 (4.451, 42.946) 0.000 a Parental Support in Study Present 1 – 1 – Absent 2.565 (0.316, 20.825) 0.378 3.117 (0.356, 27.235) 0.304 Social Life Perception Satisfactory 1 – 1 – Unsatisfactory 1.194 (0.517, 2.759) 0.677 4.146 (1.409, 12.201) 0.010 a Feeling Lonely Yes 1 – 1 – No 0.278 (0.113, 0.685) 0.005 a 0.105 (0.034, 0.325) 0.000 a Studies Hamper Social Interaction Yes 1 – 1 – No 0.843 (0.369, 1.926) 0.686 1.082 (0.396, 2.956) 0.876 Illness Hampers Study Yes 1 – 1 – No 1.097 (0.471, 2.556) 0.829 0.804 (0.301, 2.153) 0.666 Classmate Conflict No 1 – 1 – Yes 1.112 (0.316, 3.912) 0.868 2.774 (0.712, 10.807) 0.141 Family Conflict No 1 – 1 – Yes 1.973 (0.395, 9.839) 0.427 2.118 (0.388, 11.544) 0.386 Roommate Conflict No 1 – 1 – Yes 1.775 (0.429, 7.337) 0.427 1.770 (0.387, 8.081) 0.461 a significant when Confidence Interval (CI) 95%; p < 0.05. From the table, students who are not satisfied with their time management are 4.819 times more likely to experience moderate stress and 13.827 times more likely to experience high stress compared to those who are satisfied with their time management. Moreover, students who are not suffering from loneliness are 72.2% less likely to experience moderate stress and 89.5% less likely to experience high stress compared to those who are suffering from loneliness. Students who are not satisfied with their social life are 4.146 times more likely to experience high stress compared to those who are satisfied with their social life. The remaining variables are insignificant at the 5% level of significance. Academic and psychological factors Table 5 exhibits the relative risk ratio (RRR) for various academic and psychological factors on Perceived Stress and their corresponding significance among participating medical students. RRR and significant values for each academic and psychological factor are measured for moderate and high perceived stress. From the p-values, it is observed that students who are not worried about their exams are 66% less likely to experience moderate stress compared to those who worry. Moreover, students who face difficulty in understanding medical contexts are 4.117 times more likely to experience moderate stress compared to those who understand the context. Additionally, students who are not afraid of new living are 79.2% less likely to experience stress than those who are afraid of new living. Students who are not satisfied with their administration and lectures are 7.749 times and 5.042 times more likely to experience moderate stress compared to those who are satisfied with their administration and lectures. Furthermore, students who are not satisfied with their time management are 5.299 times more likely to experience high stress than those who are satisfied. Students who do not feel lonely are 86.7% less likely to experience moderate stress than those who are suffering from loneliness. 10.1371/journal.pone.0354261.t005 Table 5 Relative risk ratio of all factors on perceived stress (N = 380). Variable Category RRR (moderate) 95% CI P value (moderate) RRR (high) 95% CI P value (high) Academic Performance Satisfactory 1 – 1 – Unsatisfactory 1.151 (0.291, 4.539) 0.841 1.168 (0.257, 5.301) 0.840 Attending Sessions Yes 1 – 1 – No 0.168 (0.015, 1.835) 0.144 0.367 (0.031, 4.481) 0.432 Concentration in Sessions Yes 1 – 1 – No 0.451 (0.096, 2.118) 0.313 0.559 (0.105, 2.979) 0.496 Understanding Sessions Yes 1 – 1 – No 1.032 (0.235, 4.527) 0.966 2.019 (0.401, 10.166) 0.394 Perception of Sessions Beneficial 1 – 1 – Not Beneficial 0.981 (0.201, 4.804) 0.981 0.978 (0.171, 5.605) 0.980 Curriculum Difficulty Yes 1 – 1 – No 0.690 (0.196, 2.421) 0.563 0.720 (0.159, 3.259) 0.671 Academic Environment Satisfactory 1 – 1 – Unsatisfactory 0.998 (0.237, 4.209) 0.998 0.704 (0.139, 3.552) 0.671 Learning Material Availability Available 1 – 1 – Not Available 0.558 (0.147, 2.111) 0.391 0.419 (0.094, 1.869) 0.255 Study Worry Yes 1 – 1 – No 0.410 (0.104, 1.614) 0.202 0.132 (0.011, 1.531) 0.099 a Exam Score Worry Yes 1 – 1 – No 0.340 (0.095, 1.216) 0.097 a 0.125 (0.024, 0.646) 0.013 b Pressure of Studies Yes 1 – 1 – No 1.115 (0.125, 9.931) 0.922 1.491 (0.064, 34.793) 0.803 Pressure of Passing Exams Yes 1 – 1 – No 0.241 (0.033, 1.772) 0.163 1.698 (0.095, 30.245) 0.718 Time Management Pressure Yes 1 – 1 – No 1.100 (0.225, 2.365) 0.906 0.156 (0.008, 2.966) 0.217 Difficulty in Understanding Context Yes 1 – 1 – No 4.117 (1.021, 16.609) 0.047 b 3.216 (0.674, 15.341) 0.143 Parent’s Expectation Pressure Yes 1 – 1 – No 1.957 (0.441, 8.676) 0.377 2.145 0.369 Teacher’s Expectation Pressure Yes 1 – 1 (0.406, 11.330) – No 0.864 (0.235, 3.177) 0.826 0.801 0.778 Fear of Future Uncertainty Yes 1 – 1 (0.107, 3.727) – No 0.641 (0.137, 2.996) 0.572 0.157 0.112 Fear of New Living Yes 1 – 1 (0.016, 1.539( – No 0.208 (0.053, 0.809) 0.023 b 0.155 0.020 b Fear of Dropping Out of Course Yes 1 – 1 (0.032, 0.751) – No 0.756 (0.189, 3.015) 0.692 0.468 0.328 Future Choice Cluelessness Yes 1 – 1 (0.102, 2.138) – No 0.480 (0.139, 1.657) 0.246 0.625 0.524 Administration Satisfaction Yes 1 – 1 (0.148, 2.641) – No 7.749 (1.894, 31.701) 0.004 b 9.9479 0.006 b Lecture Satisfaction Yes 1 – 1 (1.921, 46.766) – No 5.042 (0.921, 27.598) 0.062 a 10.056 0.014 b Exam Criteria Satisfaction Yes 1 – 1 (1.605, 63.009) – No 0.197 (0.037, 1.044) 0.156 0.291 0.192 Career Choice Satisfaction Yes 1 – 1 (0.045, 1.855) – No 1.275 (0.327, 4.965) 0.726 1.704 0.486 Lack of Entertainment Facility Yes 1 – 1 (0.379, 7.644) – No 0.571 (0.167, 1.954) 0.373 0.338 0.180 Teaching Strategy Satisfaction Yes 1 – 1 (0.069, 1.651) – No 1.051 (0.232, 4.741) 0.949 1.209 0.075 a Course Activity Feeling Satisfactory 1 – 1 (0.037, 1.171) – Unsatisfactory 0.243 (0.056, 1.044) 0.157 0.463 0.373 Medical Students’ Interaction Satisfactory 1 – 1 (0.085, 2.515) – Unsatisfactory 1.004 (0.277, 3.631) 0.994 0.716 0.642 Recreation Time Yes 1 – 1 (0.175, 2.922) – No 0.770 (0.242, 2.453) 0.659 2.197 0.354 Time Management Satisfactory 1 – 1 (0.415, 11.626) – Unsatisfactory 5.299 (1.536, 18.287) 0.008 b 9.089 0.005 b Parental Support in Study Present 1 – 1 (1.961, 42.145) – Absent 10.045 (0.562, 179.491) 0.117 16.376 0.066 a Social Life Perception Satisfactory 1 – 1 (0.831, 32.715) – Unsatisfactory 0.488 (0.134, 1.781) 0.278 1.841 0.436 Feeling Lonely Yes 1 – 1 (0.396, 8.567) – No 0.133 (0.037, 0.472) 0.002 b 0.046 0.000 b Studies Hamper Social Interaction Yes 1 – 1 (0.011, 0.209) – No 2.149 (0.642, 7.189) 0.214 4.470 0.140 Illness Hampers Study Yes 1 – 1 (1.072, 18.642) – No 2.123 (0.621, 7.247) 0.229 2.119 0.283 Classmate Conflict No 1 – 1 (0.537, 8.352) – Yes 1.341 (0.235, 7.637) 0.741 3.382 0.199 Family Conflict No 1 – 1 (0.525, 21.751) – Yes 5.652 (0.527, 60.604) 0.152 5.642 0.170 Roommate Conflict No 1 – 1 (0.477, 66.732) – Yes 2.450 (0.435, 13.783) 0.309 2.034 0.456 a significant when Confidence Interval (CI) 90%, p < 0.1. b significant when Confidence Interval (CI) 95%; p < 0.05. In case of high stress, students who are not worried about their studies and exams are 86.8% and 87.5% less likely to experience high stress than those who are worried. Moreover, students who are not afraid of new living are 84.5% less likely to experience high stress compared to those who are afraid of new living. Administration and lecture satisfaction are also significantly associated with high stress. Students who are not satisfied with their administration and lectures are 9.479 times and 10.056 times more likely to experience high stress than those students who are satisfied with administration and lectures. Similarly, students who are not satisfied with teaching strategy are 1.209 times more likely to experience high stress compared to those who are satisfied. Moreover, students who are satisfied with their time management and do not have parental support in their study are 9.089 times and 16.376 times more likely to experience high stress. Furthermore, students who are not suffering from loneliness are 95.4% less likely to experience high stress compared to those who are suffering from loneliness. The remaining variables are insignificant at the 10% level of significance. Discussion In the study, it was identified that 258 medical respondents from Shaheed Suhrawardy Medical College and Shaheed Monsur Ali Medical College, which comprises 67.9% of the sample, reported having moderate stress by the PSS-10 score. The perceived stress prevalence among medical students is high, like studies from Pakistan [ 37 ]and Portugal [ 38 ]. 9.5% and 22.6% of the respondents were identified as low and high-stressed students, respectively. The mean PSS score was 21.64, which is higher than those of medical students in Malaysia (19.5) [ 39 ], Korea (18.6) [ 40 ] and Thailand (13.5) [ 41 ] and lower than that of Saudi Arabia (28.5 ± 3.8) [ 42 ]. Table 6 shows the frequency distribution of the responses to the perceived stress scale by the medical students. 10.1371/journal.pone.0354261.t006 Table 6 Perceived stress scale (PSS-10) response (N = 380). Variables Never n (%) Almost Never n (%) Sometimes n (%) Fairly Often n (%) Very Often n (%) Upset because something happened unexpectedly 23 (6.1%) 24 (6.3%) 186 (48.9%) 65 (17.1%) 82 (21.6%) Unable to control important things in life 41 (10.8%) 39 (10.3%) 157 (41.3%) 73 (19.2%) 70 (18.4%) Feeling nervous and stressed 20 (5.3%) 27 (7.1%) 147 (38.7%) 93 (24.5%) 93 (24.5%) Confident in the ability to handle personal problems 59 (15.5%) 102 (26.8%) 148 (38.9%) 48 (12.6%) 23 (6.1%) Felt that things were going your way 22 (5.8%) 78 (20.5%) 162 (42.6%) 74 (19.5%) 44 (11.6%) Unable to cope with all the things to do 40 (10.5%) 57 (15.0%) 157 (41.3%) 65 (17.1%) 61 (16.1%) Able to control irritations in life 42 (11.1%) 86 (22.6%) 173 (45.5%) 55 (14.5%) 24 (6.3%) Feeling that you were on top of things 13 (3.4%) 35 (9.2%) 149 (39.2%) 91 (23.9%) 92 (24.2%) Angered because of things outside of your control 35 (9.2%) 52 (13.7%) 164 (43.2%) 69 (18.2%) 60 (15.8%) Feeling of difficulties piled up so high that you could not overcome them 54 (14.2%) 66 (17.4%) 152 (40.0%) 60 (15.8%) 48 (12.6%) The perceived stress is significantly associated with several academic factors in bivariate analysis in the present study. Academic performance and anxiety regarding study and exam scores are significantly associated with the perceived stress of the medical students, which aligns with a study conducted in Germany regarding chronic stress in medical and dental students [ 43 ]. Academic curriculum and recreation time are significant academic stressors of the current study, which is in line with a study in India about perceived stress and stress sources of undergraduate medical students [ 44 ]. In Nepal, a study was conducted to identify the sources of stress among undergraduate medical students, which confirms that lecture satisfaction, lack of entertainment facilities, and exam pressure are significant sources of stress, and the current study also corroborates this finding [ 8 ]. Fear of uncertainty about the future and future choice cluelessness are important academic factors for perceived stress among medical students, which aligns with the study in Nigeria related to correlates of stress among medical students [ 45 ]. Class lecture satisfaction has a significant association with perceived stress, which aligns with a similar study conducted in Saudi Arabia [ 42 ]. Several psychological factors were also associated with the perceived stress. The most important source of stress was feeling loneliness in the current study, and it is also one of the prominent sources of stress that was identified in similar studies conducted in Pakistan [ 46 ] India [ 44 ]. Moreover, a study regarding perceived stress among undergraduate medical students was conducted in West Bengal, where it was found that peer conflict is significantly associated with perceived stress among students, and the current study is also consistent with this finding [ 47 ]. Additionally, time management skills and inability to interact with society are significantly associated with perceived stress according to the bivariate analysis. Multivariate analysis was conducted where the perceived stress category was the dependent variable and the significant variables in bivariate analysis were the independent variables. Multinomial logistic regression identified the significant variables in multivariable analysis which are namely anxiety related to study and exams, difficulty in understanding context, fear of new living, administration and lecture satisfaction, teaching strategy satisfaction, time management, parental support in study, and feeling of loneliness. Students who are not anxious about their studies and exams are less likely to experience stress compared to those who experience anxiety. Moreover, students who face difficulty in understanding medical contexts suffer from moderate stress compared to those who understand the medical topics clearly. Students who are not afraid of new living are less likely to experience moderate and high stress compared to those who are afraid of new living. Lecture and Administration satisfaction are also significant predictors for stress of medical students and those who are not satisfied with lecture and administration are more likely to experience stress. Additionally, students who are not satisfied with the teaching strategy are more likely to experience stress compared to those who are satisfied with the teaching methods. Furthermore, students who are not satisfied with their time management are more likely to experience stress compared to those who are satisfied with their time management. Absence of parental support in study and loneliness are also key predictors and students are more likely to experience stress when their parents do not support their studies and their suffering from loneliness. Limitations of the study The study sample was collected from two medical schools, and the entire analysis and results may not be generalizable to other medical schools in Bangladesh. Moreover, the data were self-reported, and hence the information was subject to reporting bias. Furthermore, the non-participation of several students might have impacted the sample size of the study. Despite these limitations, the study provides valuable insights into mental stress among medical students in Dhaka and offers an important foundation for future multi-institutional research. Conclusions The cross-sectional study established the persistence of mental stress as a burden among medical students, as it identified that about 90% of the medical students as moderately and highly stressed. A substantial proportion of medical students were suffering from mental stress, mostly due to their pressures/demands. However, several psychological factors are also responsible in this regard. Considering the vulnerability of medical students towards mental stress and their potential contribution to future health systems, medical colleges and health authorities should investigate the matter by addressing their psychological needs and formulate effective strategies to reduce the burden of mental stress among medical students.