Prevalence And Factors Associated With Depression Among Medical Students In Africa; A Systematic Review Study

Background: Depression in its severe form can lead to suicide. In Africa, a couple of studies have been carried to evaluate depression in medical students. These studies have provided variable findings on the prevalence of depression in medical students and associated factors. Given the considerable variability and somewhat controversial findings in the different studies, there is a need for a systematic review study to critically appraise the existing literature. Methods: This was a systematic review of studies published between January 1, 2000, and December 31, 2019 in the selected databases. We used Medical Subject Headings (MeSH) combining relevant keywords to search the databases. Results: Data was extracted from 4 studies included in this review involving a total number of 1848 medical students. The prevalence of depressive symptoms range from 23.3 to 76.5% and major depression was reported in 15.6% and 30.6% of participants in 2 studies. Female gender and being a student in higher level of studies were significantly associated with depression in all included studies. Other factors associated with higher levels of depression included smoking, alcohol consumption, being unmarried, major life events, presence of other mental health condition and the presence of chronic medical condition. Conclusion: Despite paucity of research on depression among medical students in Africa, this review shows that depressive symptoms are common and associated with a number of socioeconomic factors. Being a female student and/or higher level of medical education is associated with depression. There is need for more robust studies to evaluate the impact of depression on medical students.


3
Depression is a chronic medical condition characterised by lack of interest, decrease energy and absence of pleasure, feeling of guilt, low self-worth, insomnia and poor concentration 1 . This condition affects more than 300 million people worldwide and thus a major contributor to global disease burden 2 . Depressive symptoms are common among individuals of all walks of life and thus health professionals are not exempted. In university medical students, previous systematic reviews have shown that the prevalence of depression is 6-66.5% in North America 3 , while a more global systematic review and meta-analysis study reported a prevalence of 27.2% 4 . In Africa, there is still a paucity of data on depression among medical students and worse still; no systematic review study has been done to culminate evidence from different studies done in Africa. Many factors have been found to be associated with depression worldwide including stress. The prevalence of stress among medical student is 25-90% and has been demonstrated to be an important determinant of depression 5,6 . Stress contributes to a higher prevalence of depression among medical students than the general population. Other factors demonstrated to be associated with depression include the presence of chronic medical conditions, alcohol consumption, level of study and smoking 4 .
Depression in medical students can lead to a number of adverse outcomes. These outcomes range from impairment of functioning in the classroom to stress-induced disorders and deteriorating clinical performance. In the worst-case scenario, depression can lead suicide. Literature has reported higher rates of suicide in medical professionals than the general population 7 . This is especially true among female medical students and professionals 8 . Due to the effect of depression on the later professional life of medical students, it is important it is identified during medical training and appropriately managed. A systematic review will provide more evidence of the diverse factors associated with depression.
In Africa, a couple of studies have been carried to evaluate depression in medical students. These studies were carried out among medical students from diverse sociodemographic backgrounds. Given the considerable variability and somewhat controversial findings in the different studies, there is a need for a systematic review study to critically appraise existing literature to determine the prevalence of depression in medical students in Africa, its associated factors in different sub-regions in Africa.

Review questions
This 19 year (2000-2019) systematic review study was done to answer the following questions: 1. What is the prevalence of depression in medical students in Africa?

Methods
This study is reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols (PRISMA-P) 2015 Guidelines 9 .

Inclusion criteria
Study designs and participants: observational studies (cross-sectional, case-control and cohort studies) on the prevalence of depression and its associated factors in medical students in Africa.
We included studies published between January 1, 2000, and December 31, 2019 in the selected databases. No language restrictions were applied and study with duplicate results of the same study, only the most comprehensive and up-to-date version was considered.

Exclusion criteria
We excluded all interventional studies carried on depression in medical students. We also excluded unpublished manuscripts and conference abstracts as well as editorials and reviews letters to editors.

Primary database search
We used Medical Subject Headings (MeSH) combining relevant keywords with names of countries in Africa to search PubMed/MEDLINE for eligible studies. If the name of a country has changed, both names will be included in the search. Key words included in the search included: depression, depressive symptoms, medical students and Africa. Other databases included in the searches include: Excerpta Medica Database (EMBASE) and PsycINFO.
Articles returned by the search will be saved to the Mendeley software. Duplicates were highlighted and taken out from the list of potential studies. The retained articles were reviewed for eligibility by going through the abstracts. The full articles were then accessed via PubMed or the corresponding journal's site. The reference lists of all eligible research articles were scrutinised for additional potentially relevant articles.

Data management and Screening
After de-duplication of records by Mendeley software, a standardised pretested questionnaire containing inclusion and exclusion criteria was drafted to orient the screening of titles and abstracts. The full texts of all potentially eligible studies were rigorously reviewed using the pretested questionnaire. A PRISMA flowchart was generated to reflect the entire review process and also detailing potentially eligible studies which were excluded. Six articles were assessed for eligibility, 4 studies were retained. The two excluded studies were experimental studies which were Randomised control trials.

Data extraction and items
A data extraction sheet was produced on Microsoft Excel spread sheet. Data from the different selected studies included were then be extracted. The following items were extracted from the selected studies: authors, year of study, year of publication and journal, country where study was conducted, geographical region, language of publication, study design and setting (country), duration of study, sample size, objectives, mean/median age of participants, age range of participants, level of student (clinical and preclinical), gender distribution. Other factors of social/personal life were collected including: gender, the presence of a chronic disease, level of alcohol consumption, smoking, marital status, monthly income. Finally information on associations (X 2 , ORs, RR, CIs and p values) between depression and these factors was collected when available. The STROBE checklist 10 for observational studies was used to evaluate the quality of reporting each paper reviewed. All studies were reported with following the STROBE guidelines except for the earliest study carried out in Nigeria in 2008 11 . The risk of bias in each study was evaluated using the Risk of Bias Tool for Prevalence Studies developed by Hoy and collaborators 12 . GRADE score was not used for assessment of quality of evidence for this particular review study 13 .  (Table 1).

Results
We excluded 2 studies which were interventional in nature. A variety of methods were used to diagnose depression and assess the degree of depressive symptoms. All studies used self-rating questionnaires for the diagnosis of depression. Two (02) studies used the Zung Self Rating Depression Scale, one study uses the DASS 21 self-reporting questionnaire and a final study used the PHQ-9 questionnaire for diagnosis of depression and depressive symptoms. The prevalence of depressive symptoms range from 23.3 to 76.5% and major depression was reported in 15.6% and 30.6% of participants in 2 studies of the included studies (Table 2).
Female gender and being a student in higher level of studies were significantly associated with depression in all included studies. Other factors associated with higher levels of depression included smoking, alcohol consumption (2 of 4), being unmarried, major life events, presence of other mental condition, and visit to any traditional, complementary or alternate medical practitioner and the presence of chronic medical condition. One study reported a positive effect of spirituality on depression while 2 studies revealed that depression was not associated with academic performance.  21 . One particular point of interest in many studies in the past is to find the impact of depression on academic performance. In this review 2 studies did not find any association between depression and academic performance 15,16 . This finding is in total contrast with previous studies where depressed students had higher risk of academic problems 22,23 .
Generally the statistical analyses carried out in the included studies were satisfactory however only three studies gave details of the approach of the multivariate analysis used.
The different studies included different factors (variables) in the multivariate analysis models 14-16 . A few basic problems were however identified in the statistical methods used to produce the models these studies. The study by Ngasa and collaborators did not mention which variable they controlled for in their logistic model for the relationship between depression and academic performance 16 . The studies Mohammed and colleagues controlled only for age in multivariate analysis 15 . All included studies used a p value cut off of 0.05 for statistical significance. This is no longer recommended, and it's better to present p value as a spectrum of evidence rather than a binary significant or nonsignificant metric 24 .
The main limitation of this study stems from the fact that the relevant studies included in the review used a wide variety of instruments to assess depression. This made it difficult to carry out an efficient comparison between the different included studies.

Conclusion
Irrespective of the tool used to assess depression or the geographical location, our analysis demonstrated that depression was generally high among medical students in Africa. This review shows that depressive symptoms are common and associated with a number of socioeconomic factors. Being a female student and/or higher level of medical education is associated with higher degree of depression. We therefore make the following recommendations: We recommend large scale RCTs to generate more evidence of the link between depression and its associated factors as there was generally high risk of bias in the included studies.
Particular attention should be paid to female medical students as almost all studies revealed higher prevalence than in the male counterparts.
Emphasis should be placed on preventing medical students from smoking cigarettes or consuming alcohol. These factors were reported in most studies to be associated with depression.