STRENGTHS AND LIMITATIONS OF THIS STUDY
-
This systematic review and meta-analysis used a comprehensive search to avoid missing published articles.
-
The current study was conducted using the updated Preferred Reporting Items for Systematic Reviews and Meta-Analyses-2020 checklists to compose the report.
-
All primary studies included in the current review were conducted using a cross-sectional study design, which does not establish a real cause-and-effect relationship between the covariates and outcome variables.
-
The number of studies conducted in universities is higher than studies conducted in colleges and high schools, which may result in unproportional distributions of study participants.
-
There was significant heterogeneity in both the overall meta-analysis and subgroup analysis.
Introduction
Depression is a major public health concern that manifests clinically as persistent melancholy, a lack of interest in often-joyful activities and difficulty doing daily activities, as well as problems with sleep or appetite, recurrent suicidal thoughts and trouble concentrating.1 2 The prevalence of depression is alarmingly increasing around the world.3 It significantly increases the overall global disease burden and is a leading cause of disability.4 According to the WHO, 4.4% of people worldwide experience depression at any given time.5 A depressive disorder increases a person’s risk of early mortality by 40% compared with a healthy person.6 Depression puts a significant strain on society as a multiproblematic condition that impairs social, vocational and interpersonal functioning.7 8
Depression is the greatest concerning mental health condition among students and it is the second-most common cause of death among people between the ages of 15 and 34, which includes the majority age of students.9 10 Due to developmental and academic demands, high school students are more susceptible than ever to social and psychological problems.11 12 Adolescence is seen as a stage of transition during which a person goes through physical and psychological changes and development that are limited to the time between puberty and the beginning of legal maturity.13 14 This stressful time is usually when several mental problems are most likely to develop and frequently continue into adulthood and increase the chance of developing additional mental health problems.15 Depression is a severe mental illness that affects adolescents and frequently has an impact on family connections, social functioning and academic achievement.16
When young individuals transfer to college, they are exposed to a special risk, and 53% of young adults exhibit depression symptoms to some extent while in college.17 Mental health difficulties, psychological health complaints and symptoms are all getting worse and more prevalent on college campuses.18 19 University students are a unique set of people going through a difficult transitional stage as they transition from adolescence to adulthood and as a result, make many critical life decisions.20 During this critical time, university students are under a great deal of stress, mostly due to financial strain, academic responsibilities and interpersonal interactions.8
A systematic review and meta-analysis study revealed that the pooled prevalence of depression among medical students was 30.6% with a prevalence range of 10%–85%.21 The overall prevalence of depression among Malaysian and Iranian secondary school students was 42.6% and 72.6%, respectively.22 23 Different studies around the world showed that the prevalence of depression among students was 32.8% in China college students,24 46.5% in Spanish college students,25 21.8% in Australia,26 68.5% in Hong Kong,27 43.7% in India,28 40.9% in Pakistan,29 43% Saudi Arabia30 and 23% in the USA.31 Although there were insufficient data obtained from an all-over-the-nation study to estimate the prevalence of depression in Ethiopia, a study conducted using the findings of the national health survey and a 2012 report by the Ethiopian Federal Ministry of Health revealed that the prevalence of depression in Ethiopia was 9.1% and 5%, respectively. The prevalence of depression among students in Ethiopia varied and ranged from 17.1%32 to 52.3%.33
Epidemiological studies have shown several factors that were accountable for an increased risk of student depression such as lack of free time, starting a new personal relationship, competitiveness, relocating to an unfamiliar place, examinations and workloads, fear of failing, financial problems, being female and worries about separation from family.34–38 Students who fail to perform well on a test show substantially more psychological distress than those who achieve an inadequate academic performance resulting in psychological problems.39 In Ethiopia, there are inadequate resources and teaching facilities, infrastructure gaps, shortages of teachers, and an outdated curriculum.40 41 These may contribute to the students’ depression. Depression has several significant impacts on students like greater absenteeism from class, decreased academic performance, affect social relationships and increased risk of self-injury like attempting or committing suicide.42–44
Even though different primary studies were conducted in Ethiopia among students to assess depression, their reports contained very discrepant and substantially inconsistent findings across the studies. Moreover, as far as our knowledge, there are no previous systematic reviews and meta-analyses on the pooled prevalence and determinants of depression among students in Ethiopia. Therefore, the main objective of this systematic review and meta-analysis was to estimate the pooled prevalence and associated factors of depression among high school, college and university students in Ethiopia, which has a vital role in constructing an adequate interventional programme.
Methods
Study protocol registration
This systematic review and meta-analysis was conducted to estimate the overall combined prevalence of depression among Ethiopian students and to determine its associated factors. The study protocol for the current meta-analysis was registered in the International Prospective Register of Systemic Review (PROSPERO) with registration number CRD42023424301. We also used an appropriate guideline for systematic reviews and meta-analyses reports, which is the Preferred Reporting Items for Systematic Reviews and Meta-Analyses-2020 checklists45 (online supplemental file 1).
Supplemental material
Search strategy
We have used the following electronic databases to search primary studies that describe the prevalence and factors associated with depression in Ethiopia: PubMed/MEDLINE, African Journal Online, Psychiatry Online, CINAHL, EMBASE and Science Direct. Grey literature and additional articles were searched using Google Scholar and Google. A searching strategy for this systematic review and meta-analysis was developed using combinations of free texts and restricted vocabulary such as Mesh terms and keywords that related to the outcomes of the study. The primary articles were searched by using the following keywords, “Prevalence” OR “Magnitude” OR “Epidemiology” OR “Incidence” AND “Depression” AND “Associated factors” OR “Risk factors” OR “Determinants” OR “Predictors” OR “Correlates” AND “High school” OR “College” OR “University” OR “Higher institution” AND “Students” OR “Learners” AND “Ethiopia”. The search of the primary studies was conducted from 2 April 2023 to 11 May 2023 using the English language (online supplemental file 2).
Supplemental material
Eligibility criteria
Before retrieving full-text articles, we selected the pertinent papers using their titles and abstracts. Further screening was done on the retrieved full-text articles by predetermined inclusion and exclusion criteria. Observational studies such as cross-sectional, case–control and cohort studies that reported the prevalence and associated factors of depression were included. All articles with full text, easily accessible, written in the English language and conducted in Ethiopia without restrictions on publication year were included in this review. Both published and unpublished primary studies were included. All studies conducted in high school, college and university were considered. However, case reports, reviews, case studies, conferences, qualitative findings and studies not written in English were not included in this study. Additionally, articles with insufficient information that did not provide specific results for depression were excluded from this systematic review and meta-analysis.
Data extraction
Two authors (MM and TT) independently extracted all the essential data from full-text studies using a standardised data extraction format. The extracted data contain the following components: the first author’s name, publication year, the study design, sample size, educational level, prevalence, screening tool and the region where the study was carried out. Moreover, factors associated with depression, including the 95%, CI and OR, were extracted. The agreement between the two authors (MM and TT) has been evaluated using the intrarater reliability coefficient (Cohen’s kappa). Substantial agreement was concluded between the two authors as obtained from Cohen’s kappa coefficient, which was 0.76. During the data extraction process, any disagreement between the two authors was resolved by consulting and discussing with the third and fourth authors (GR and FA). For the inconsistent data, double extraction was conducted to reach a common agreement (online supplemental file 3).
Supplemental material
Outcome measurements
The primary objective of this systematic review and meta-analysis was to assess the pooled prevalence of depression among students in Ethiopia. The overall prevalence of depression was calculated by dividing the number of students with depression by the total number of students included in this review and multiplied by 100. The second objective of this study was to identify the pooled effects of factors that are associated with depression among students in Ethiopia. The OR was calculated from the included primary studies using 2×2 tables. The primary articles included in this review were screened using the Patient Health Questionnaire (PHQ-9),46 Beck Depression Inventory (BDI-II),47 Depression, Anxiety and Stress Scale (DASS-21),48 Centre for Epidemiologic Studies Depression Scale (CES-D)49 and Hospital Anxiety and Depression Scale (HADS).50 The cut-off scores used by primary studies to assess depression were: DASS-21 and BDI-II based on the cut-off for milled depression; HADS based on the cut-off scores greater than or equal to eight; PHQ-9 based on the cut-off scores greater than or equal to five; and CES-D based on the cut-off point greater than or equal to 16. The questionnaires of primary articles that were used to assess depression were initially prepared in English, then translated into the local language, such as Amharic, Afaan Oromo and others, and finally back-translated to English by language experts. As a result, the measures for depression were administered in their local language and then translated back to English to maintain consistency.
Quality assessment
The Joanna Briggs Institute (JBI) critical appraisal checklists were used to assess the methodological quality of each included primary study. Two authors (GN and GT) independently evaluated the quality of the included studies for this review. The JBI critical evaluation checklist is a standardised assessment tool, used to assess the methodological quality of included primary articles.51 The critical appraisal checklists contain nine items with a total quality score between 0 and 9, which is classified as between 0–4, 5–7 and 8 or 9, scored as low, moderate and high quality, respectively. Studies with quality scores of 5 and above were considered in this systematic review and meta-analysis. Any disagreement between the two authors while assessing the quality of the included studies was resolved by discussing with the third author (GMT) (online supplemental file 4).
Supplemental material
Data synthesis and analysis
The data extraction was done using a Microsoft Excel spreadsheet, and the extracted data were exported it to STATA V.14.0 for further analysis. We used the inverse variance weighting approach to weight the effect sizes of the included primary studies. The findings of this review were summarised and presented using forest plots, texts and tables. To estimate the overall prevalence, a random effect meta-analysis model was employed. The I2 statistics test was used to check the presence of heterogeneity among the included articles.52 The result of an I2 test greater than 50 confirmed the presence of statistical heterogeneity among the included studies. Due to the existence of heterogeneity in this study, a random-effect meta-analysis was conducted. Subgroup analysis was carried out based on the region where the study was done, publication year, level of education and screening tools for depression to determine the source of heterogeneity. Furthermore, sensitivity analysis was also conducted using the leave-one-out technique to assess the effects of a single study on the overall prevalence. Publication bias was checked by using visual observation of the funnel plots test and Egger weighted regression test at a 5% significant level.53 54 In Egger’s test, publication bias was regarded as present when the p value was<0.05.
Patient and public involvement
Patients and the public were not involved in the review protocol, design, reporting or dissemination plans of this study.
Results
Search results
Using various electronic search methods including PubMed, Google Scholar, CINAHL and African Journal Online, a total of 2465 primary studies were retrieved for this review. Among these studies, 1942 studies were removed because of duplication. Additionally, 498 studies were excluded after we looked at their titles and abstracts because they were irrelevant to our review, were not conducted in Ethiopia, had differences in study populations and settings, and their complete texts were unavailable. Another 25 full-text articles were reviewed for eligibility using the inclusion criteria, and 8 studies were excluded for other reasons. Finally, 17 studies were eligible and included in this systematic review and meta-analysis (figure 1).
Flow chart of study selection for meta-analysis of depression among students in Ethiopia.
Characteristics of included studies
A total of 17 primary articles were included in this systematic review and meta-analysis, which assessed depression and their associated factors among students in Ethiopia. The included studies were published between 2015 and 2022 in which 11 studies were published after 2019 and the other remaining 6 studies were published before 2019. The primary studies included in this review were conducted in the five regions of Ethiopia; four studies were conducted in Addis Ababa,55–58 six studies were done in Amhara,32 59–63 four were in Oromia,33 64–66 two were from South Nation Nationality and Peoples67 68 and the other one study was conducted in Tigray region.69 A total of 8798 participants were included in this review, with sample sizes ranging from 26562 to 1040.65 Regarding the level of education, 12 studies were done among university students,32 33 57 60 61 63–66 68 70 71 2 articles were conducted from college students58 67 and the other 3 studies were conducted among high school students.59 62 69 Out of 17 included primary articles, 4 studies were conducted using PHQ-9,32 59 67 69 the other 5 studies by using BDI-II,61 63 65 66 70 4r studies were done using DASS-21,33 58 60 62 3 studies using CES-D56 64 68 and the remaining 1 study was conducted using HADS.57 Those studies included in this review were cross-sectional studies and their prevalence rate ranged from 17.1%32 to 52.3%33 (table 1).
Characteristics of studies included in this systematic review and meta-analysis on depression among students in Ethiopia
The pooled prevalence of depression among students
17 primary studies were included in this systematic review and meta-analysis to calculate the combined prevalence of depression among Ethiopian students. The pooled prevalence of depression among students in Ethiopia was 35.52% (95% CI 30.54% to 40.51%) (figure 2). Regarding heterogeneity, the statistics test (I2) was computed, and there was a significant level of heterogeneity among the included primary articles (I2=96.2, p≤0.001). Consequently, we conducted a random effect meta-analysis model to estimate the pooled prevalence of depression among students in Ethiopia. Additionally, subgroup analyses based on study region, publication year, level of education and screening tools were conducted to identify the potential sources of heterogeneity.
Forest plot showing the pooled prevalence of depression among students in Ethiopia.
Publication bias
Publication bias was checked by using both a funnel plot and an objective statistical test (Egger’s test). Accordingly, the evidence from the visual inspection of the funnel plot revealed that there was no publication bias (figure 3). Additionally, the Egger’s test confirmed the absence of publication bias in the included primary studies (p=0.056) (table 2).
Funnel plot of depression among students in Ethiopia.
Eggre’s test of depression among students in Ethiopia
Subgroup analysis
In this review, subgroup analysis was conducted based on region, screening tools, level of education and publication year. The result of this subgroup analysis revealed that the highest pooled prevalence of depression was shown in Addis Ababa (40.43%; 95% CI 28.02% to 52.83%) (Cohen’s kappa=0.75), and the lowest prevalence of depression was observed in the Amhara region (32.38%; 95% CI 22.49% to 42.27%) (Cohen’s kappa=0.67). The pooled prevalence of depression among college students 42.64% (95% CI 26.08% to 59.20%) (Cohen’s kappa=1.00) was higher than among university 35.10% (95% CI 29.60% to 40.60%) (Cohen’s kappa=0.75) and high school students 32.44% (95% CI 18.65% to 46.22%) (Cohen’s kappa=0.68). The findings of subgroup analysis by publication year indicated a higher prevalence of depression among studies that were conducted after 2019, 35.66% (95% CI 28.88% to 42.43%) (Cohen’s kappa=0.73) than studies conducted in 2019 and below 35.24% (95% CI 29.38% to 41.10%) (Cohen’s kappa=0.83). Moreover, the highest prevalence of depression was obtained from studies conducted using HADS followed by DASS-21, which was 51.30% (95% CI 45.37% to 57.23%) and 47.52% (95% CI 42.23% to 52.81%) (Cohen’s kappa=0.75), respectively (online supplemental file 5).
Supplemental material
Sensitivity analysis
In this review, sensitivity analysis was carried out by using the one-study leave-out technique to detect the impacts of a single study on the results of the overall prevalence of depression. As shown from the findings of sensitivity analysis, the result of a single study did not affect the pooled prevalence. After the omission of a single study from the analysis step by step, the pooled prevalence of depression ranges between 34.51% (95% CI 29.55% to 39.48%) and 36.68% (95% CI 31.83% to 41.52%) (online supplemental file 6).
Supplemental material
Factors associated with depression among students
In the included primary articles, there were several factors associated with depression. However, the factors that were presented in at least two primary studies were used in this systematic review and meta-analysis. The findings of the meta-analysis revealed that female sex, first-year students, alcohol drinking, family history of mental illness and parental neglect were significantly associated with depression among students. Being female was mentioned as an associated factor for depression in the six primary studies32 59 60 62 64 71 included in this systematic review and meta-analysis (Cohen’s kappa=0.68). Female participants were 2.8 times more likely to have depression compared with their counterparts (OR 2.80, 95% CI 1.73 to 4.53) (online supplemental file 7). Alcohol drinking was associated with depression in the three included articles33 65 68 (Cohen’s kappa=1.00). Participants who drink alcohol had 2.17 times greater odds of having depression compared with non-drinkers (OR 2.17, 95% CI 1.70 to 2.76) (online supplemental file 8). A family history of mental illness was mentioned as a contributing factor to depression in the included two primary studies66 72 (Cohen’s kappa=0.50). The pooled OR showed that participants with a family history of mental illness were 2.59 times more exposed to depression as compared with participants without a family history of mental illness (OR 2.59, 95% CI 1.51 to 4.46) (online supplemental file 9). The analysis of four included articles64 65 72 73 revealed that those respondents who are first-year students were 3.1 times more likely to develop depression than their counterparts (OR 3.09, 95% CI 1.84 to 5.18) (Cohen’s kappa=0.75) (online supplemental file 10). The pooled OR of the two included studies59 69 showed that those study participants who had parental neglect were 2.64 times more likely to have depression than participants who did not experience parental neglect (OR 2.64, 95% CI 1.98 to 3.53) (Cohen’s kappa=0.50) (online supplemental file 11).
Supplemental material
Supplemental material
Supplemental material
Supplemental material
Supplemental material
Discussion
This systematic review and meta-analysis was carried out to determine the pooled prevalence and associated factors of depression among students in Ethiopia. As far as we know, this systematic review and meta-analysis is the first to assess the overall prevalence of depression and its associated factors in Ethiopia. This systematic review and meta-analysis revealed that the overall pooled prevalence of depression among Ethiopian students was found to be 35.52% with a (95% CI 30.54% to 40.51%). The subgroup analysis indicated that the highest combined prevalence of depression was observed in Addis Ababa city, and college students were more depressed than high school and university students. The result of this review was in line with other studies that were conducted in Cameroon 30.6%,74 China 32.8%,24 Hong Kong 35.8%,75 Iran 33%76 and Pakistan 35.1%.77
This study was lower than the systematic review in Saudi Arabia with an estimated pooled prevalence of depression of 51.5%,78 and the systematic review and meta-analysis done in Pakistan, which was 42.66%.79 It is also lower than different studies conducted in Ethiopia 43.49%,80 Hong Kong 68.5%,27 India 43.7%,28 Pakistan 40.9%29 and Saudi Arabia 43%.30 The possible reason for the discrepancy might be due to differences in sample size, sampled study participants and sociocultural level.
The result of this systematic review and meta-analysis was higher than the findings of other studies conducted in Ethiopia 28.13%,81 Spain 18.4%,82 Australia 21.8%26 and the USA 23%.31 It was also higher than the systematic review and meta-analysis that was conducted in China, with a pooled prevalence of 28.4%.83 The observed difference might be due to the fact that other studies conducted in different countries had a single finding, whereas this study had a pooled prevalence of depression by including different primary articles. Another discrepancy might be related to variations in the accessibility and delivery of mental health facilities and psychosocial counselling services for students.
In this systematic review and meta-analysis, we also assessed the contributing factors to depression among students. The pooled effect of associated factors showed that female sex, first-year students, alcohol drinking, family history of mental illness and parental neglect were significantly associated with depression.
Accordingly, being female was 2.8 times more likely to have depression compared with male sex. This finding was consistent with other studies conducted in Ethiopia,81 Cameroon,74 Saudi Arabia78 and Brazil.84 This might be due to the fact that females are more prone to depression due to gender-related vulnerability to psychosocial stressors and they are also more sensitive to stress. The other possible reason for this association is females have maladaptive coping strategies and multiple social roles more than males.
In this review, having a family history of mental illness was significantly associated with depression among students. This was supported by other studies done in the New Zealand85 and India.86 This might be related to depression can be more likely to occur in offspring whose parents have a history of mental illness due to hereditary transmission.87 In addition to this, the health of their parents may be stressing and worrying the offspring, which may contribute to student’s depression.
The finding of the current study revealed that first-year students were more likely to develop depression as compared with those who were second-year and above students. This result was consistent with other studies carried out in India88 and Saudi Arabia.78 This could be due to the fact that students move to a new location and begin a new way of life after being separated from their family and local settings, and during their period of transition, first-year university students may be vulnerable to depression.9 89 The other possible reason is due to the absence of social relationships and unusual kinds of exams.8 20
In this systematic review and meta-analysis, alcohol drinking was a significant predictor of depression among students. The finding was supported by the results of another study that was conducted in Ethiopia81 and the USA.90 This might be due to the fact that alcohol affects the functions of neurotransmitters in the brain, which contributes to the development of depression.91 This study also revealed that students who experienced parental neglect were more likely to have depression than students who did not experience parental neglect. This was consistent with another study conducted in the Northeastern USA.92 It is due to the fact that Children who have been neglected show symptoms of attachment disorder and eventually develop fearful relationships even with their own family.93
Strengths and limitations of the study
Even though this systematic review and meta-analysis was conducted to determine the summarised estimated prevalence of depression by including 17 primary studies with a larger sample size, it has the following limitations. The first limitation of this review was that all primary articles that were included in this systematic review and meta-analysis were conducted using a cross-sectional study design, which does not establish a real cause-and-effect relationship between the covariates and outcome variables. Second, the inter-rater reliability coefficients of some of the associated factors were in the moderate or fair range, and since the result of the Egger’s test was 0.056, which does not provide confidence, publication bias has to be a possibility. The other limitation of this study was that since the number of studies conducted in universities is higher than studies conducted in colleges and high schools, this makes unproportional distributions of study participants. Furthermore, most of the questionnaires that were used to assess primary articles were translated into the local language, but it was difficult to know if the translations were reliable across studies. Finally, there was significant heterogeneity in both the overall meta-analysis and subgroup analysis.
Conclusion
In this review, the overall pooled prevalence of depression among Ethiopian students was high. Female sex, first-year educational level, family history of mental illness, alcohol drinking and parental neglect were found significantly associated with depression. Therefore, the Ministry of Health and the Ministry of Education should work together by extending early intervention programmes and mental health screening services that focus on high school, college and university students to minimise the negative effects of depression on their academic performance, health and social lives. Additionally, mental health facilities and psychosocial counselling should be provided early for students based on identified risk factors to improve their mental health problems.
Data availability statement
All data relevant to the study are included in the article or uploaded as online supplemental information.
Ethics statements
Patient consent for publication
Acknowledgments
We would like to extend our deepest thankfulness to the authors of the included primary articles as they helped as the groundwork for this systematic review and meta-analysis.
This post was originally published on https://bmjopen.bmj.com