A cross-sectional study was conducted over a 3-month period between February 3, 2023, and April 30, 2023, among students at the University of Buea in Cameroon. The University of Buea is geographically positioned in Buea, the capital of the Southwest Region, a part of Cameroon’s Anglophone zone that currently faces acute socio-political crises. The university has an average population of over 33,000 students distributed across 11 faculties and schools.

The study population comprised all enrolled students at the University of Buea, regardless of their specific study program.

Students enrolled at the Higher Technical Teachers Training College in Kumba, an institution of the University of Buea located outside Buea, were excluded. We also excluded students who refused to provide consent and those with incomplete questionnaires.

To calculate sample size, we used Cochran’s formula for cross-sectional studies.

$n_i=\frac{Z^2\cdot P\left(1-P\right)}{d^2}$

where n is the sample size, P is the expected proportion in the population, q is $1-P$, d is the margin of error set at 0.05, and Z is the standard normal variant at a 95% confidence interval (standard value 1.96). We tolerated an absolute error (d) of 5%.

The maximum sample size was obtained using an expected prevalence of depression of 31.9%, which was obtained from research carried out on students of the Faculty of Letters and Social Sciences of the University of Dschang ( Mboua et al., 2021 ).

Therefore,

$n_i=\frac{{\left(1.96\right)}^2\times 0.319\left(1-0.319\right)}{{0.05}^2}$

$n_i=\frac{3.8416\times 0.319\times 0.681}{0.0025}$

$n_i=\frac{0.834545}{0.0025}$

$n_i=334$

We took a non-response rate of 10%, 10/100 (333) = 33.

This resulted in a minimal sample size (n = 333 + 33 = 367 students).

We performed a multi-staged cluster sampling method using four different steps. A convenient random sampling method was used to select university faculty (FLPS, FSMS, FE, FS, FAVM, COT, and FHS). A probability proportional to size was used to select the number of participants from each faculty member. Proportionate to size was used to select the number of students for each cycle of study (undergraduate, master’s, and PHD). Lastly, a convenience sampling method was used to select the departments and the different years of study (1^st to 7^th year).

Data were collected on sociodemographic, academic, socioeconomic, and lifestyle characteristics. The DASS-21 scale () was used to screen for depression, anxiety, and stress among students, whereas the University of Buea GPA grading system was used to evaluate academic performance. The questionnaires were pre-tested on 35 students at a private university. Furthermore, to ensure most students participated, and to increase the sample size, a snowball sampling method was used to recruit participants online using a well-structured Google form shared with students via social media platforms. The Google form carried same information as printed questionnaires, including sociodemographic, academic, socio-economic, lifestyle factors and the DASS-21 scale.

Students who satisfied the inclusion criteria were deemed eligible to participate in the study and were subsequently recruited. 1150 participants were recruited, of which 164 were online responses and 986 were on-site responses. Fifty-five incomplete questionnaires were excluded. All participant information was coded, and confidentiality was maintained throughout the study. Depression, anxiety, and stress were evaluated using the DASS-21, a free, self-report tool that is quick to administer and score ( Henry & Crawford, 2005 ). This tool is reliable and valid for measuring individual scores for depression, anxiety, and stress, and provides an overall score for general mental health (; Medvedev, 2023 ). This scale categorizes depression, anxiety, and stress into different levels of severity, including normal, mild, moderate, severe, and extremely severe. ( Gomez, 2016 ).

Academic performance was evaluated using self-reported Cumulative Grade Point Averages (CGPAs) from the preceding academic semester. These CGPAs were subsequently categorized into distinct performance levels: First Class (3.6 - 4.0), second class Upper (3.0 - 3.5), second class Lower (2.5 - 2.9), and Third Class (below 2.5).

This study further investigated the potential associations between several variable categories and the prevalence of depression, anxiety, and stress among university students. The examined variables included socio-demographic characteristics (age, sex, relationship status), academic parameters (level of education encompassing undergraduate, master’s, and doctoral studies, specific program of study, faculty affiliation, and academic year), socioeconomic indicators (employment status, income level), and lifestyle factors (alcohol consumption, smoking habits, and history of traumatic experiences).

The raw dataset was initially managed and curated using Epidata v. 6.0. The data were subsequently transferred to the Statistical Package for the Social Sciences (SPSS) version 25 for statistical analysis. Descriptive statistical methods were used to analyze the data. Frequencies and percentages were calculated for categorical variables and are presented graphically as bar and pie charts. Continuous variables were summarized using measures of central tendency (mean and median) and measures of dispersion (standard deviation and interquartile range), which were selected based on the distributional characteristics of each variable. To delineate potential associations with depression, anxiety, and stress, Chi-square and Fisher’s exact tests were used. Variables exhibiting a significance level of P ≤ 0.2 in these bivariate analyses were subsequently incorporated into multivariate logistic regression models. This analytical method was used to determine the independent predictors of the conditions. The results are expressed as odds ratios (ORs) and their respective 95% confidence intervals (CIs). The threshold for statistical significance was set a priori at P < 0.05.

Ethical review and formal approval for this study were obtained from the Institutional Review Board of the Faculty of Health Sciences, University of Buea (IRB Number: 2023/1917-01/UB/SG/IRB/FHS), complemented by administrative authorization from the Regional Delegation of Public Health in the Southwest Region.

This study adhered to the principles outlined in the Declaration of Helsinki regarding research involving human subjects. Comprehensive informed written consent was a prerequisite for participation, obtained directly from adult subjects and from the legal guardians of minor participants who also provided their explicit assent. The voluntary nature of participation was underscored, guaranteeing the unencumbered right to withdraw without prejudice

Baseline sociodemographic characteristics of the participants are presented in Table 1 .

Of the 1095 participants, 702 (64.1%) were female, and 393/1095 (35.9%) were male. The average age of the participants was 22.1 ± 3.99 (SD). Most of the students were single (78.6%). Of the seven faculties of the University of Buea selected for this study, the FLPS group was the most populated (26.7%). Most students were undergraduates (79.7%). Most of the participants from the first and second years of the study 38.4% and 38.5%, respectively, (57.1%) were involved in extracurricular activities, (35.0%) had a GPA range between 2.5 - 2.9.

A significant proportion of the student population has experienced financial constraints, with 50% reporting insufficient income. The student body was largely defined by familial financial support (86.8%) and full-time academic dedication (90.2%), and was further characterized by a notable prevalence of self-employment among over half of the students (56.8%).

According to this study, 41.3% drank alcohol, and 1.8% smoked cigarettes. Many students (62.6%) had traumatic experiences, with the three most common traumatic experiences being gunshots (25%), the death of a loved one (19.5%), and robbery (8.4%). Up to 77.7% of students lived in armed conflict areas.

Based on the findings, 28.8% of the students had depression, 40.3% had anxiety, and 9.2% had stress. Regarding the severity of symptoms associated with depression, anxiety, and stress, 18.3%, 15.1%, and 7.5% of the participants exhibited mild symptoms, respectively. Moderate symptoms were observed in 9.7%, 19.1%, and 0% of participants, respectively. Severe symptoms were present in 0.7%, 5.8%, and 1.7% of participants, respectively. Finally, extremely severe symptoms were reported in 0%, 0.4%, and 0% of participants with depression, anxiety, and stress, respectively ( Figure 2 ).

Regarding the co-occurrence of depression, anxiety, and stress, 18.2% of the participants had anxiety symptoms alone, 6.5% had only depressive symptoms, 0.9% had only stress symptoms, 15.3% had anxiety and depression symptoms, 0.2% had both depressive and stress symptoms, 1.4% had anxiety, and 6.8% had anxiety, depression, and stress symptoms ( Figure 3 ).

A logistic regression model was used to identify the factors independently associated with depression. We found that students at the PhD level were 2.79 times more likely to be depressed than undergraduates (AOR = 2.79, P = 0.018), while students in their sixth year of medical school were more than 8.25 times (AOR = 8.25, P = 0.025) more likely to be depressed than those in their final year. Students who did not participate in extracurricular activities were 0.63 times less likely to experience depression (AOR = 0.63, P = 0.032). In contrast, students who received financial support from other sources, family, or scholarships were (AOR = 11.21, P = 0.001; AOR = 2.55, P = 0.001; AOR = 3.75, P = 0.023) times more likely to experience depression than those who were self-sponsored. Additionally, the likelihood of full-time students experiencing depression was 0.57 times less than that of part-time students (AOR = 0.94, P = 0.027) ( Table 2 ).

Table 3 presents the factors that contribute to anxiety.

CT = College of Technology, FHS = Faculty of Health Sciences, FE = Faculty of Education, FS = Faculty of Sciences, FAVM = Faculty of agriculture and Veterinary medicine, FLPS = Faculty of Law and Political sciences, FSMS = Faculty of Social and management sciences.

Compared to students from FSMS, those attending CT and FHS were 0.46 (AOR = 0.46, P = 0.038) and 0.51 (AOR = 0.51, P = 0.003) times less likely to experience anxiety, respectively. In contrast, students from FE were 1.64 (AOR = 1.64, P = 0.033) times more prone to anxiety than their FSMS counterparts. Students who relied on a family member as their sponsor were 1.67 times more likely to experience anxiety than those who were self-sponsored (AOR = 1.67, P = 0.019). Females were 1.55 (AOR = 1.55, P = 0.002) times more anxious than males, while students who reported their income as sufficient were 1.51 (AOR = 1.51, P = 0.023) times more likely to have anxiety than those who reported that their income was sufficient. Students who did not smoke cigarettes and those who had never resided in a conflict area were 0.3 times (AOR = 0.3, P = 0.022) and 0.63 times (AOR = 0.63, P = 0.004) less likely to have anxiety, respectively, than those who smoked cigarettes and had lived in a conflict area.

The factors associated with stress among study participants are listed in Table 4 .

Female students were found to have a 1.87 times higher likelihood of experiencing stress than male students (AOR = 1.87, P = 0.012). Additionally, students with a GPA below 2 were 2.95 times more likely to experience stress than those with a GPA ranging from 3.5 to 4.0 (2.95, P = 0.016). Moreover, students who had not undergone a traumatic experience were 0.43 times less likely to experience stress (AOR = 0.43, P = 0.001) than those who had faced such an experience.

The relationships between depression, anxiety, stress, and academic outcomes were examined to assess their impact on academic performance. As shown in Table 5 , a significant correlation was identified between depression (χ² = 15.74, P = 0.003), stress (χ² = 15.67, P = 0.004), and poor academic performance. Figure 4 illustrates that most participants experiencing depression, anxiety, and stress had a Grade Point Average (GPA) ranging from 2.5 to 2.9.

However, it is crucial to consider certain limitations when interpreting the findings of this study. Depression, anxiety, and stress were assessed using the DASS-21 questionnaire, a concise and efficient screening instrument. The scale is valid and reliable in diverse settings in the post-COVID era ( Ali et al., 2021 ) and has high-quality as a screening tool, but is not a clinical diagnostic tool ( Lee et al., 2019 ). Participants are susceptible to information recall and response biases, which can affect the accuracy of the results. Secondly, the cross-sectional design of this study limits our ability to infer causal relationships. Moreover, this study was conducted within a single university context, where the dual impact of the COVID-19 pandemic and armed conflict is evident. Several other factors could have influenced our findings, although none of the academic factors were explored. Furthermore, our study does not adequately address potential confounding variables such as study habits, sleep quality, social support, and socioeconomic background. These limitations may limit the generalizability of our findings to other universities in the country.

Notwithstanding these limitations, this study significantly enhances our understanding of the current prevalence of depression, anxiety, and stress among university students in Cameroon, with a particular focus on the University of Buea. A large sample of students, stratified by subject area in different faculties and levels of study, was used. The findings of this study will be useful for policymakers in planning interventions and the development of guidelines and policies for mental health and wellbeing for university students of Buea and the broader Cameroonian education system.

Further research could improve the current study’s limitations by using a wider national sample and a longitudinal design, more rigorous clinical diagnostic tools, and evaluating other variables, such as study habits, sleep quality, and biological factors that influence mental health and academic performance.