Abstract
Purpose
Undergraduate students are known to be a high-risk group for mental health problems. The purpose of this paper is to constitute a repeated cross-sectional study on the trend of depression over the years and factors associated with depression among undergraduates.
Design/methodology/approach
Cross-sectional data from five surveys between 2013 and 2020 (N = 1,578) among the undergraduates of Universiti Tunku Abdul Rahman, a private university in Kampar Malaysia, were combined. The Depression Anxiety and Stress Scale-21 was used to screen for depression. Cochran’s Armitage test was used to detect trend in depression. Logistic regression, random forest regression and extra gradient boosting regression were used to identify risk factors and classification.
Findings
The prevalence of depressive symptoms was found to be between 26.4% and 36.8% between the years with an average of 29.9%. There was no significant time trend in the prevalence. The risk of depressive symptoms was higher among female students, those who were dependent on family for financial support and those who were stressed.
Practical implications
Periodical screening for depression is warranted for the identification of students at risk for depression. Professional cognitive-behavioral therapies, peer support and consulting services should be made available to the students in need.
Originality/value
Depression among students had been studied widely, but the trend over years remains unexplored, especially in developing countries.
Keywords
Citation
Cheong, W.S., Degeras, K., Suliman, K.R., Selvaraju, M. and Subramaniam, K. (2022), "The prevalence of depression among students in higher education institution: a repeated cross-sectional study", Journal of Public Mental Health, Vol. 21 No. 4, pp. 331-340. https://doi.org/10.1108/JPMH-12-2021-0152
Publisher
:Emerald Publishing Limited
Copyright © 2022, Emerald Publishing Limited
Introduction
The age ranges between 18 and 29 years old, or emerging adulthood, is the preparatory phase toward adulthood (Arnett et al., 2014). This stage of life typically involves enrolling in tertiary education, exploring the job markets and engaging in serious relationship and, thus, is rather volatile (Arnett et al., 2014). It also necessitates crucial decision-making such as the selection of field of study and type of career. These instabilities could be overwhelming for some, and this is evidenced by research findings which show that mental disorders are common among this age group (Solmi et al., 2022). The prevalence rates of mental disorders are higher during this stage of life than in the 30s (Gustavson, 2018). The proportion of disability-adjusted life years because of mental and substance use disorders are higher for this age group (Chadda, 2018).
Undergraduate students are usually in the emerging adulthood stage of life. The students are in the transition phase from a family-reliant lifestyle to one with independence and freedom of social life, coupled with various challenges with regards to studies, friendships, relationships, etc. Although university students may be labeled as a socially advantaged group, they are vulnerable to mental disorders as compared to the general public (Mofatteh, 2021) and peers of same age who are not in the universities (Maser et al., 2019). About one-third (30.6%) of university students were found to have depression, which is higher than the 9% prevalence of depression in the general population (Ibrahim et al., 2013). Depression is the most common mental illness associated with university students (Ibrahim et al., 2013).
Studies using screening tools reported depressive symptoms to be present among 50%–75% of the undergraduates (Asif et al., 2020; Iqbal et al., 2015; Wahed and Hassan, 2017). Prevalence of moderately severe to severe depressive symptoms were reported to be between 13.9% and 59% (Arun et al., 2022; Iqbal et al., 2015). Similar rates were reported for Malaysian university students: rate of moderate depressive symptoms was between 20% and 27.5% and severe symptoms between 4% and 10% (Islam et al., 2018; Shamsuddin et al., 2013; Teh et al., 2015).
Researchers have identified many factors that increase the risk of depression among the university student population. A narrative review classified the predictors into six broad categories: psychological, academic, biological, lifestyle, social factors and financial factors, that were intricately connected (Mofatteh, 2021). Psychological predictors for depression include loneliness (Richardson, 2017), low self-esteem and other existing mental disorders (Mofatteh, 2021). A meta-analytic study reported that the peak and median age of onset for any mental disorders were 14.5 and 18 years (Solmi et al., 2022). A study on Spanish university students reported that the age of onset for mood disorders and anxiety disorders were 15 and 16 years (Ballester et al., 2020). This shows that the students may have had episodes of mental disorders before enrolling into the universities.
Poor academic performance (Luo et al., 2021) and poor study-life integration, that is, excessive academic-related behavior and academic stress (Badri, 2020), increase the risk of depression. Students with financial difficulties or from lower socio-economic background have been shown to have a high risk for mental disorders (Islam et al., 2018; Richardson et al., 2018; Teh et al., 2015; Usher and Curran, 2019; Wahed and Hassan, 2017). The adaptation of poor lifestyle habits, such as poor sleeping habits (Luo et al., 2021), alcohol consumption (Usher and Curran, 2019), smoking (Usher and Curran, 2019), lack of physical exercise (Mofatteh, 2021) and social media addiction (Haand and Shuwang, 2020), also increases the risk of depression and mental disorders.
Mental health of university students has been an area of interest for many researchers, and consequently, there is a lot of information on epidemiology and etiology of the problem. This is necessary, given the rise in the prevalence of depression among the university students over the decades, as reported by a review (Ibrahim et al., 2013). If left untreated, then depression can have harmful effects on students, such as academic attrition, developing social disability, suicidal tendencies, volatility in relationships and poor academic performance (Mofatteh,2021; Teh et al., 2015). Previous studies conducted in specific universities or multiple institutions, with inter-institutional collaboration, generally provided cross-sectional evidences. The trend of depression over time is yet to be well explored. Consequently, the health-care professionals, research community and policymakers are aware that university students have a high rate of depression because of various reasons, but whether the incidence of depression has levelled off, increased or decreased over the years is still unclear. The long-term impact and efficacy of the youth mental health programs or mental health services are not fully understood. To fill this gap to some extent, this study was designed with two aims, first to determine the trend in depression over the years among undergraduates and, second, to identify predictors of depression among undergraduate students. Based on previous observations that depression in academic settings has increased over the decades, we hypothesize that:
The rate of depression among students increases over the years.
Socio-demographic factors increase the rate of depression.
Materials and methods
Data collection and compilation into repeated cross-sectional data
The repeated cross-sectional design was used for this study. Repeated cross-sectional data is used to collect responses from a new random sample at successive points in time. In other words, the respondents who answered the survey in the current time period will be different from the respondents in the previous and subsequent successive time periods. Because of this feature, repeated cross-sectional data are often used to analyze population or group changes over time (Rafferty et al., 2015). Unlike panel data, it does not involve repeated measures on the same individuals, but different individuals are studied over time. In this study, cross-sectional data collected on undergraduate students from Kampar over the span of five years between 2013 and 2020, in 2013, 2016, 2018, 2019 and 2020, were harmonized to create a data set with observations over time. The study in the year 2020 was conducted before the movement control orders and lockdowns because of the COVID-19 pandemic. These five studies involved with undergraduate honors research projects on various aspects of students’ mental health such as internet addiction, smartphone addiction and stress coping strategies. Similar instruments were used to screen for depression and anxiety and collect basic demographic details across the years and the information were combined to form this data set. No studies were conducted in the year 2014, 2015 and 2017.
The sampling technique used and sample size of each survey are shown in Table 1. The multistage cluster sampling was started with the selection of two faculties out of five faculties and with a center available in the Kampar campus of the University. After that, a list of programs and year of study were made for the selected faculties. For a particular program, each year of study was treated as a distinct group. A random selection was made to select a few program-year groups. Classes teaching major subjects for the program year were identified and visited by the researchers of each study. The classes teaching major subjects were selected, as those classes consist of only students in a particular program and year and will not be open to students from other majors as an elective subject. This step was carried out to avoid approaching same student twice in different classes. The responses were collected from all the students present at the time of visit to the selected classes. In the purposive sampling, all the students studying in the library on randomly selected days were invited and surveyed. Inclusion criteria for the entire study were: consenting Malaysian undergraduate students registered for the current semester. Postgraduate students and foreign students were excluded from the study.
The cross-sectional data collected over the five years were merged, cleaned and processed. Only variables that were collected in all studies were retained. The variable names in the data set and classification codes were standardized before merging. A column on year of survey was added in the merged repeated cross-sectional data set to enable study of the time trend.
Data collection tool
Participants in the study were screened for depression using the Depression, Anxiety and Stress Scale-21 items (DASS-21) instrument (Lovibond and Lovibond, 1995). The instrument consists of three self-reported scales that measure depression, anxiety and stress. The self-reported scales consist of seven items measured on a scale with four ratings which are “Did not apply to me at all,” “Applied to me to some degree, or some of the time,” “Applied to me to a considerable degree or a good part of time” and “Applied to me very much or most of the time.” The depression scale measures dysphoria, devaluation of life, hopelessness, self-deprecation, lack of interest, anhedonia and inertia. The final score can be obtained by multiplying the DASS-21 scores by two. A score between 0 and 9 indicated that a person is normal, 10 and 13 as possible mild depression, 14 and 20 as possible moderate depression, 21 and 27 as possible severe depression and 28 and above as possible extremely severe depression. The instrument was deemed suitable for depression, anxiety and stress screening among Malaysian adults (Musa et al., 2007). The normal and mild depression groups were combined to form the “no depression” group and moderate and above were combined to form “depressed” group for further analysis.
Socio-demographic details including gender, age, ethnicity, major, cumulative grade point average, source of funding for university fees and source of funding for daily expenses were collected in all the surveys and were extracted for this study. All interviews were conducted in English, which is the formal instruction medium of the university.
Ethics approval
The ethics clearances were obtained for the individual studies and the current study with the combined data from the University’s Scientific Research Ethics Committee. Written consent was obtained from the participants before the survey. Participants were aware that their participation in the study is voluntary, and no personal information such as name or student identification number was obtained, to maintain anonymity of the participants.
Statistical analysis
Analysis was performed using the IBM SPSS version 21 and sci kit learn, a free machine learning software library for Python programming language (Scikit-learn, 2022). The five cross-sectional studies were combined and cleaned for missing values. Cochran-Armitage trend test was used to determine the presence of time-wise increasing or decreasing trend in the rate of depression. Chi-square test was used to study the association between variables. Independent sample t-test was used to compare means between groups. The data was split into two random groups with a ratio 70:30. The 70% of the data or the training data was used to develop the regression model and classification algorithm including logistic regression, random forest and extra gradient boosting classifiers. The classifiers were used to develop a predictive model for depression and identify factors associated with depression. The remaining 30% of the data or the testing data set was used to test the accuracy of the prediction.
Results
A total of 1,578 students records collected in five different years between 2013 and 2020 were combined in this analysis (Table 2). The largest sample was obtained in the year 2013, n = 402 respondent, followed by the year 2019, n = 328 respondents. Majority of the participants were female and of Chinese ethnic group. The participants had mean and median age of around 21 years with a range from 17 to 30 years.
Data collected via the purposive sampling could cause selection bias and, thus, result in a non-representative sample. The samples collected via the two sampling methods (multistage cluster and purposive sampling) were compared to determine if the two groups differed in the distribution of mental disorders and demographic characteristics. There was no significant difference in the prevalence of depressive symptoms (χ2 = 0.02 and p = 0.49). Comparison of demographic characteristic showed that the purposive sample consisted more females (60% vs 47.6%) (χ2 = 24.04 and p = 0.00). The distribution of ethnicity (χ2 = 4.43 and p = 0.22) and age (t = 0.462 and p = 0.644) did not differ. This shows that the groups do not differ greatly and could be combined to form a repeated cross-sectional data.
The distribution of depressive symptoms in the study duration is shown in Figure 1. The rate of depression ranged from 26% to 37% among the participants of the study. There was an increase from 26% in year 2013 to 37% in 2018, followed by a drop to 31% and 29% in the subsequent years (Figure 1). The rates did not differ greatly from the overall prevalence of 29.9%. The Cochran-Armitage trend test confirmed this by showing an absence of trend in change in the prevalence rates over time (Z = 0.083 and p-value = 0.9339).
Female students (OR = 1.6), those receiving financial support from family (OR = 1.6) and those who were screened with moderate (OR = 5.5) and severe stress (OR = 29.4) had increased risk for depression as per Table 2. The other category for financial support included study loan, scholarship, working part time and multiple sources of support, with the depression rate for the categories: 28.9% (n = 54), 18.5% (n = 5), 30.8% (n = 4) and 24.3% (n = 43), respectively (Table 3). Students who had taken study loans and were working part time had apparently high depression rates, around 30%. The students with scholarships had lowest rates of depression. However, no further analysis was done to study the significance of the difference because of small sample sizes; for instance, only 13 students worked part time and 27 were on scholarship; the categories were combined to produce a binary category, supported by family and others.
The data was split 70:30 for creating a classification algorithm and testing the classification. The logistic regression model was able to correctly classify 83% of cases (accuracy) and 80% of positive cases were correctly identified (precision). Two other classification algorithms that were developed also had similar accuracy and precision rates. The variables that were deemed important in the classification algorithms are shown in Table 4. Stress was identified as an important contributor in all the algorithms. Age was deemed important in one algorithm as in Table 4.
Discussion
The prevalence of high depression scores among the participants ranged from 26.0% to 37.0%. On average, around 30% of the students were screened positive for depression over the study duration. This is in coherence with the review which highlighted that 30.6% of university had depression. The rate of depressive symptoms obtained in this study was within the range reported in other Malaysian universities, between 29.4% and 37.2% (Islam et al., 2018; Shamsuddin et al., 2013; Teh et al., 2015). Trend analysis showed that the fluctuation in the rate over time is insignificant. The first research hypothesis for this study was not proven. The predictors for high depressive symptoms include female gender, being financially supported by the family and stress. As hypothesized, the socio-demographic characteristics of participants increased the risk of depressive symptoms.
The lack of change in prevalence rate over the years is consistent with a Canadian study on the trend of depression among adolescents between age 12 and 19 years, which showed that there was no significant change in the depression trend of adolescents for the span of 14 years from year 2000 to year 2014 (Wiens et al., 2017). The stable depression trend suggested that current mental health resources are insufficient to tackle depression among adolescents, and more effective approaches need to be developed (Wiens et al., 2017). The findings contrasted with another study from the USA which found an increase in the rate of depression among adolescents from year 1991 to year 2018, especially girls (Keyes et al., 2019). The prevalence rate over the years could be considered as a positive finding, as the academic environment did not show extra demands over the years. However, the findings should be interpreted with caution, as an increase in the rate was noted until 2018, when multistage cluster sampling was introduced. The two years with purposive sample data had prevalence rate that were quite similar with the previous years. There are possibilities of students who were potentially depressed isolating themselves, rather than studying in a public place like a library. Purposive sampling of the students in the library could have resulted in lower rates. Future studies with a-priori hypothesis to study the fluctuation in depression rate, and using random sampling technique to obtain participants, are needed to verify if there is really a lack of trend.
A similarity between the current findings and the findings by Keyes et al. (2019) is that young women were at a higher risk for depressive symptoms. Higher risks of mental disorders among female students have been reported previously (Usher and Curran, 2019; Wahed and Hassan, 2017) According to Mojtabai et al. (2016), the risk could have emerged from cyberbullying and internet addiction issues, as a correlation was observed between heavy texting and depressive mood among girls.
Presence of stress was the strongest factor associated with depression. Academic stress caused by difficulty in mastering a subject, underperformance and being engaged in demanding courses and extreme academic activities increase the risk of depression among undergraduate students (Badri, 2020; Mofatteh, 2021). Higher rates of depression were also observed previously among final year students because of the difficulty of the subject and also worry about future career prospects (Mofatteh, 2021; Islam et al., 2018; Shamsuddin et al., 2013). Existing mental disorders affect their ability to cope with the academic stress and increase the risk of depression among the students (Mofatteh, 2021). Students who experienced mental disorders in past 12 months had higher level of role impairment, including for academic related roles (Ballester et al., 2020). In the current study, the participants were not assessed for existing mental health problems, and thus, the contribution of existing problems to development of depression could be not be studied.
In the current study, students with no loan or scholarship and funded by the family were found to have higher rate of depression. The university in which this study is conducted is a non-profit body which was built to provide affordable quality private education to all segments of the society. This feature attracts the students from middle class background. Many of them obtain scholarships or study loans to pay the university fees and parents support their expenses. Knowing that the family is channeling the hard-earned money on their education may impose a great burden on the students to perform well academically, which may lead to depression, in which case the students find difficulties in coping with their studies. This situation is reiterated further by Azim and Baig (2019) mentioning that students have called attention to finance as a notable factor in the institutions which requires students to pay high tuition fees, especially in the private universities compared to the public universities, which has led to increased and higher depression level.
There are some limitations associated with the study. The five combined studies did not have an a-priori hypothesis to detect the change in depression. Therefore, limited risk factors for depression were explored. In addition to that, different sample sizes and sampling techniques were used to obtain the responses. The non-random sampling techniques may affect the accuracy of the estimation. Different sampling methods may have resulted in a difference of 9% in prevalence between the year 2018 and 2020. Even then, a noteworthy 30% of the participants were screened positive for depression, indicating that it is a problem worth noting. The screening tool used in the study maybe a good measure of symptoms but may not be accurate in determining the presence of depression. In addition, self-rated symptoms could have been subjected to over or under reporting. The study has used reasonable overall sample size and provides a picture on trend in depression on Malaysian students, which to date has not been well explored. Future studies with clearly defined objectives to capture depression trend using diagnostic tools are warranted.
Higher educational institutions provide various psychological supports for the students such as counselling by trained professionals, peer support activities and mental health campaigns. Despite all that, the rate of depression remained stable, indicating that outreach of these programs should be improved further. Conducting mental health screening every semester, identifying students at risk and helping them will be beneficial in reducing depression among students. Previous findings reported low rates of mental health service utilization among students with mental disorders (Ballester et al., 2020). Thus, in addition to providing services such as cognitive-behavioral programs by professional, counselling services and peer support, students should also be educated on when they should get professional help and the importance of getting help for their well-being.
Conclusion
In conclusion, in the five years between 2013 and 2020, there is no change in the prevalence of depression symptoms among undergraduates. Around 30% of the students screened positive for depression symptoms. Being female, being financially supported by family and high stress levels increased the risk of depression. Routine mental health screening is warranted to identify students at risk and providing help.
For future research to determine depression trends, by using a repeated cross-sectional study, it is recommended to set the main objectives and standardize the question design, sampling method and measurement of each research period. This step not only simplifies the data merging process but also prevents information loss; the collected data will also better serve research purposes. The number of target respondents from different demographic backgrounds (such as year of study) should also be equal. By doing so, it is possible to appropriately investigate the association between various factors and depression by reducing the effects of under-sampling or over-sampling.
Figures
Figure 1
Distribution of depressive symptoms across the study duration
The compiled studies
| Year | Sampling technique | Sample size |
|---|---|---|
| 2013 | Multistage cluster | 402 |
| 2016 | Multistage cluster | 307 |
| 2018 | Multistage cluster | 289 |
| 2019 | Purposive sampling | 330 |
| 2020 | Purposive sampling | 250 |
Demographic characteristic of the respondents
| Demographic factor | 2013 | 2016 | 2018 | 2019 | 2020 | Total |
|---|---|---|---|---|---|---|
| Total sample size | 402 | 307 | 289 | 330 | 250 | 1,578 |
| Gender | ||||||
| Female | 289 (71.9%) | 204 (66.4%) | 158 (54.7%) | 174 (52.7%) | 177 (70.8%) | 826 (52.0%) |
| Male | 113 (28.1%) | 103 (33.6%) | 131 (45.3%) | 156 (47.3%) | 73 (29.2%) | 752 (48.0%) |
| Age (years) | ||||||
| Mean | 20.76 | 21.22 | 20.81 | 20.98 | 20.81 | 20.90 |
| SD | 1.315 | 1.280 | 1.150 | 1.567 | 1.375 | 1.437 |
| Median | 21 | 21 | 21 | 21 | 21 | 21 |
| IQR | 1 | 2 | 1 | 2 | 2 | 2 |
| Ethnicity | ||||||
| Chinese | 361 (89.8%) | 280 (91.2%) | 262 (90.7%) | 305 (92.4%) | 230 (92.0%) | 1,440 (91.3%) |
| Indian | 30 (7.5%) | 24 (7.8%) | 20 (6.9%) | 20 (6.1%) | 18 (7.2%) | 112 (7.1%) |
| Others | 11 (2.7%) | 3 (1.0%) | 7 (2.4%) | 5 (1.5%) | 2 (0.8%) | 26 (1.6%) |
| Depression | ||||||
| (n) | 106 | 87 | 106 | 102 | 70 | 471 |
| Prevalence (%) | 26.4 | 28.3 | 36.8 | 31.1 | 28.0 | 29.9 |
Predictors of depressive symptoms among participants
| Factor | Prevalence of depression n (%) |
Odds ratio | 95% confidence interval of odds ratio |
|---|---|---|---|
| Gender | |||
| Male⸸ | 212 (28.3) | 1.00 | |
| Female | 259 (31.4) | 1.634* | 1.190–2.245 |
| Expenses | |||
| Others⸸ | 184 (27.1) | 1.00 | |
| Family | 287 (32.1) | 1.580 | 1.150–2.171 |
| Stress | |||
| Not stressed⸸ | 166 (15.1) | 1.00 | |
| Moderate | 125 (48.6) | 5.501 | 3.834–7.891 |
| Severe | 179 (81.4) | 29.446 | 18.557–46.724 |
⸸Reference category and Expenses – resources for daily expense. The others category includes scholarship, study loan and working part time. Only 13 students worked part time and the prevalence of depression for the group was 30.8%; *p < 0.05, all other variables had p < 0.01
Classification algorithms and important variables in the classification
| Classification algorithm | Accuracy (%) | Precision (%) | Variable importance (descending order) |
|---|---|---|---|
| Logistic regression | 82.7 | 79.8 | Stress, financial support and gender |
| Random forest | 80.4 | 80.8 | Stress |
| Extreme gradient boosting | 82.3 | 78.8 | Stress and age |
References
Arnett, J.J., Žukauskienė, R. and Sugimura, K. (2014), “The new life stage of emerging adulthood at ages 18–29 years: implications for mental health”, The Lancet Psychiatry, Vol. 1 No. 7, pp. 569-576.
Arun, P., Ramamurthy, P. and Thilakan, P. (2022), “Indian medical students with depression, anxiety, and suicidal behavior: why do they not seek treatment?”, Indian Journal of Psychological Medicine, Vol. 44 No. 1, pp. 10-16.
Asif, S., Mudassar, A., Shahzad, T.Z., Raouf, M. and Pervaiz, T. (2020), “Frequency of depression, anxiety and stress among university students”, Pakistan Journal of Medical Sciences, Vol. 36 No. 5, pp. 971-976.
Azim, S.R. and Baig, M. (2019), “Frequency and perceived causes of depression, anxiety and stress among medical students of a private medical institute in Karachi: a mixed method study”, Journal of Pakistan Medical Association, Vol. 69 No. 6, pp. 840-845.
Badri, S.K.Z. (2020), “Study-life integration and young adult students mental health”, Journal of Public Mental Health, Vol. 20 No. 1, pp. 24-32.
Ballester, L., Alayo, I., Vilagut, G., Almenara, J., Cebrià, A.I., Echeburúa, E., Gabilondo, A., Gili, M., Lagares, C., Piqueras, J.A. and Roca, M. (2020), “Mental disorders in Spanish university students: prevalence, age-of-onset, severe role impairment and mental health treatment”, Journal of Affective Disorders, Vol. 273, pp. 604-613.
Chadda, R.K. (2018), “Youth & mental health: challenges ahead”, Indian Journal of Medical Research, Vol. 76 No. 11, pp. 359-361.
Gustavson, K., Knudsen, A.K., Nesvåg, R., Knudsen, G.P., Vollset, S.E. and Reichborn-Kjennerud, T. (2018), “Prevalence and stability of mental disorders among young adults: findings from a longitudinal study”, BMC Psychiatry, Vol. 18 No. 1, pp. 1-15.
Haand, R. and Shuwang, Z. (2020), “The relationship between social media addiction and depression. A quantitative study among university students in Khost, Afghanistan”, International Journal of Adolescence and Youth, Vol. 25 No. 1, pp. 780-786.
Ibrahim, A.K., Kelly, S.J., Adams, C.E. and Glazebrook, C. (2013), “A systematic review of studies of depression prevalence in university students”, Journal of Psychiatric Research, Vol. 47 No. 3, pp. 391-400.
Iqbal, S., Gupta, S. and Venkatarao, E. (2015), “Stress, anxiety and depression among medical undergraduate students and their socio-demographic correlates”, The Indian Journal of Medical Research, Vol. 141 No. 3, pp. 354-357.
Islam, M.A., Low, W.Y., Tong, W.T., Yuen, C.C.C. and Abdullah, A. (2018), “Factors associated with depression among university students in Malaysia: a cross-sectional study”, KnE Life Sciences, Vol. 4 No. 4, pp. 415-427.
Keyes, K.M., Gary, D., O’Malley, P.M., Hamilton, A. and Schulenberg, J. (2019), “Recent increases in depressive symptoms among US adolescents: trends from 1991 to 2018”, Social Psychiatry and Psychiatric Epidemiology, Vol. 54 No. 8, pp. 987-996.
Lovibond, S.H. and Lovibond, P.F. (1995), Manual for the Depression Anxiety Stress Scale, 2nd ed., The Psychological Foundation of Australia, Sydney.
Luo, W., Zhong, B.L. and Chiu, H.F.K. (2021), “Prevalence of depressive symptoms among Chinese university students amid the COVID-19 pandemic: a systematic review and meta-analysis”, Epidemiology and Psychiatric Sciences, Vol. 30 No. e31, pp. 1-21.
Maser, B., Danilewitz, M., Guérin, E., Findlay, L. and Frank, E. (2019), “Medical student psychological distress and mental illness relative to the general population: a Canadian cross-sectional survey”, Academic Medicine, Vol. 94 No. 11, pp. 1781-1791.
Mofatteh, M. (2021), “Risk factors associated with stress, anxiety, and depression among university undergraduate students”, AIMS Public Health, Vol. 8 No. 1, pp. 36-65.
Mojtabai, R., Olfson, M. and Han, B. (2016), “National trends in the prevalence and treatment of depression in adolescents and young adults”, Pediatrics, Vol. 138 No. 6, p. e20161878.
Musa, R., Fadzil, M.A. and Zain, Z. (2007), “Translation, validation and psychometric properties of Bahasa Malaysia version of the depression anxiety and stress scales (DASS)”, ASEAN Journal of Psychiatry, Vol. 8 No. 2, pp. 82-89.
Rafferty, A., Walthery, P. and King-Hele, S. (2015), “Analysing change over time: repeated cross sectional and longitudinal survey data”, UK Data Service, Vol. 1, pp. 1-25.
Richardson, T., Elliott, P. and Roberts, R. (2017), “Relationship between loneliness and mental health in students”, Journal of Public Mental Health, Vol. 16 No. 2, pp. 48-54.
Richardson, T., Yeebo, M., Jansen, M., Elliott, P. and Roberts, R. (2018), “Financial difficulties and psychosis risk in British undergraduate students: a longitudinal analysis”, Journal of Public Mental Health, Vol. 17 No. 2, pp. 61-68.
Scikit-learn (2022), “Scikit-learn machine learning in python”, available at: https://scikit-learn.org/stable/index.html/ (assessed 10 June 2022).
Shamsuddin, K., Fadzil, F., Ismail, W.S.W., Shah, S.A., Omar, K., Muhammad, N.A., Jaffar, A., Ismail, A. and Mahadevan, R. (2013), “Correlates of depression, anxiety and stress among Malaysian university students”, Asian Journal of Psychiatry, Vol. 6 No. 4, pp. 318-323.
Solmi, M., Radua, J., Olivola, M., Croce, E., Soardo, L., Salazar de Pablo, G., Il Shin, J., Kirkbride, J.B., Jones, P., Kim, J.H. and Kim, J.Y. (2022), “Age at onset of mental disorders worldwide: large-scale meta-analysis of 192 epidemiological studies”, Molecular Psychiatry, Vol. 27 No. 1, pp. 281-295.
Teh, C.K., Ngo, C.W., binti Zulkifli, R.A., Vellasamy, R. and Suresh, K. (2015), “Depression, anxiety and stress among undergraduate students: a cross sectional study”, Open Journal of Epidemiology, Vol. 5 No. 4, pp. 260-268.
Usher, W. and Curran, C. (2019), “Predicting Australia’s university students’ mental health status”, Health Promotion International, Vol. 34 No. 2, pp. 312-322.
Wahed, A.W. and Hassan, S.K. (2017), “Prevalence and associated factors of stress, anxiety and depression among medical Fayoum University students”, Alexandria Journal of Medicine, Vol. 53 No. 1, pp. 77-84.
Wiens, K., Williams, J.V.A., Lavorato, D.H., Duffy, A., Pringsheim, T.M., Sajobi, T.T. and Patten, S.B. (2017), “Is the prevalence of major depression increasing in the Canadian adolescent population? Assessing trends from 2000 to 2014”, Journal of Affective Disorders, Vol. 210, pp. 22-26.
Further reading
WHO (2021), “Adolescence health in South East Asian region”, available at: www.who.int/southeastasia/health-topics/adolescent-health/ (assessed 21 October 2021).
Acknowledgements
Acknowledgement: Sincere gratitude is conveyed to the Universiti Tunku Abdul Rahman and Faculty of Science Universiti Tunku Abdul Rahman for all the support rendered. Sincere gratitude is conveyed to Ms Chiam Shiok Shin, Ms Leow Kar Yin, Ms Chang Yee Fong, Mr Chew Yee Ming, Mr Chin Chun Kit, Mr Tan Chong Eng and Mr Teh Yong Tend for assisting the data collection process for the study.