Journal of Psychiatric Research 47 (2013) 391e400

Contents lists available at SciVerse ScienceDirect

Journal of Psychiatric Research

journal homepage: www.elsevier.com/locate/psychires

Review

A systematic review of studies of depression prevalence in university students

Ahmed K. Ibrahim a, b, *, Shona J. Kelly c, Clive E. Adams d, Cris Glazebrook d
a
Community Health School, Faculty of Medicine, Assiut University, Asyut, Egypt
b
Division of Epidemiology, Community Health Sciences School, D Floor, West Block, Queens Medical Centre, University of Nottingham, Nottingham, UK
c
Social Epidemiology and Evaluation Research Unit, Division of Health Sciences, University of South Australia, Adelaide, Australia
d
Institute of Mental Health, University of Nottingham Innovation Park, Triumph Road, Nottingham NG7 2TU, UK

a r t i c l e i n f o a b s t r a c t

Article history: Background: Depression is a common health problem, ranking third after cardiac and respiratory diseases
Received 3 June 2012 as a major cause of disability. There is evidence to suggest that university students are at higher risk of
Received in revised form depression, despite being a socially advantaged population, but the reported rates have shown wide
28 November 2012
variability across settings.
Accepted 28 November 2012
Purpose: To explore the prevalence of depression in university students.
Method: PubMed, PsycINFO, BioMed Central and Medline were searched to identify studies published
Keywords:
between 1990 and 2010 reporting on depression prevalence among university students. Searches used
Systematic review
Depression
a combination of the terms depression, depressive symptoms, depressive disorders, prevalence,
Prevalence university students, college students, undergraduate students, adolescents and/or young adults. Studies
Students were evaluated with a quality rating.
University Results: Twenty-four articles were identified that met the inclusion and exclusion criteria. Reported
prevalence rates ranged from 10% to 85% with a weighted mean prevalence of 30.6%.
Conclusions: The results suggest that university students experience rates of depression that are substan-
tially higher than those found in the general population. Study quality has not improved since 1990.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Background diagnostic tool and sampling used (Weissman et al., 1996; Marsella,
1978). Although there has been an increasing concern about
Depression is one of the most common health problems for depression in specific groups such as adolescents or the elderly
university students (Lyubomirsky et al., 2003; Vredenburg et al., (Winter et al., 2011; Springer et al., 2011; Lim et al., 2011; Gladstone
1988). Depression is considered as a multi-problematic disorder et al., 2011; McKenzie et al., 2010), the problem of university
that leads to impairment in inter-personal, social, and occupational students’ depression has received relatively little attention, despite
functioning (Sadock and Kaplan, 2007). The basic characteristic of evidence of a steady rise in the number of depressed university
depression is a loss of positive affect which manifests itself in a range students (Ceyhan et al., 2009). Studies have reported wide variations
of symptoms, including sleep disturbance, lack of self-care, poor in the proportion of students identified as depressed, from relatively
concentration, anxiety and lack of interest in everyday low rates around 10% (Goebert et al., 2009; Vazquez and Blanco, 2006;
experiences (NICE, 2009). Level of impairment can be classified clin- Vazquez and Blanco, 2008) to high rates of between 40% and 84%
ically by standardized diagnostic interview but in prevalence studies (Bayati et al., 2009; Garlow et al., 2008; Khan et al., 2006). This wide
depression is typically identified through a validated, self-report variation appears to be influenced by many factors including methods
screening instrument. The prevalence of depression seems to be of assessment (Weissman et al., 1996; Marsella, 1978), geographical
affected by many factors including; population studied, location (Steptoe et al., 2007; Weissman et al.,1996) and demographic
socio-demographic factors (e.g. sex, age) (Steptoe et al., 2007; Kaplan factors such as SES (Kaplan et al., 2008; Steptoe et al., 2007).
et al., 2008), place of study (Weissman et al.,1996; Steptoe et al., 2007) The cost of affective disorders can be particularly high in young
people because they represent the future of any community, its
hope and potential leaders (El-Gendawy et al., 2005). Depression
* Corresponding author. Department of Public Health and Community Medicine, in this early life stage can lead to an accumulation of negative
Medical Faculty, Assiut University, Asyut, Egypt. Tel.: þ20 1127533610; fax: þ20
consequences through adult life through its impact on career pros-
8823254633.
E-mail addresses: [email protected], [email protected] pects and social relationships (Denise et al., 1996; Aalto-Setälä et al.,
(A.K. Ibrahim). 2001). Depression has been linked to poorer academic achievements

0022-3956/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jpsychires.2012.11.015
392 A.K. Ibrahim et al. / Journal of Psychiatric Research 47 (2013) 391e400

(Hysenbegasi et al., 2005), relationship instability (Whitton and reviewed studies, published between January 1990 and October
Whisman, 2010), suicidal thoughts and attempts (Jeon, 2011) and 2010, reporting on depression among undergraduate university
poorer work performance (Harvey et al., 2011). Although arguably students. Searches used the keywords depression, depressive
university students are more likely to be advantaged in socio- symptoms, depressive disorders, prevalence, university students,
economic terms which is considered protective against depression college students, undergraduate students, adolescents and/or
(Lowe et al., 2009), there are many factors that might increase young adults were used in the searches. Additional articles were
students’ vulnerability to depression. These factors include changes identified through the reference lists of the retrieved articles and
in life style resulting in sleep and eating disturbances, financial previous review studies.
stressors, family relationship alterations, academic worries and Inclusion criteria were that: 1) the study sample included
preoccupation with post-graduation life (NIMH, 2003). exclusively undergraduate students in higher education; 2) the
There is a strong perception, both in the US and in the UK, that study included an aim to establish prevalence of depression and;
demands for psychological services by university students have grown 3) the study reported prevalence rates. The exclusion criteria
and that university counseling services are also dealing with more were 1) the study did not report response rate; 2) clinical trials
severe mental illness (Hunt and Eisenberg, 2010). Despite this, a recent studies and; 3) failure to report a separate prevalence rate for
literature review of studies on depression and treatment outcomes depression. Demographic data, sample size, diagnostic instrument
among US College students carried out from 1990 to 2009 identified used and prevalence data on students’ depression were abstracted.
only four studies and concluded that research on depression and Searches were limited to articles published in the last two
treatment outcomes among US college students are present but scarce decades yielding a total of 2303 citations. After examining the titles,
and inconclusive. They also found wide variability in inclusion and abstracts (if abstract was unavailable, the article was nevertheless
exclusion criteria and tools for diagnosis of depression and determi- counted) and the reference lists for related articles, 94 articles were
nation of its severity (Miller and Chung, 2009). Another systematic retrieved, including five Non-English articles (French 1, Japanese 1,
review of research published between January 1980 and May 2005 Mexican 1, Korean 2) and 89 English language studies were
looking at the prevalence of depression, anxiety, and other indicators examined thoroughly. Non-English articles were translated with
of psychological distress among US and Canadian medical students the help of PhD students from Japan, Spain, and Korea, studying at
found higher rates of depression than is seen in the general pop- the University of Nottingham, who were expert in both languages;
ulation. The review also pointed to a lack of research into the causes of English and the other language.
students’ depression and its impact on academic performance, After careful reading of these articles, an additional 70 articles
dropout rates and professional development (Dyrbye et al., 2006). were excluded as a result of the following justifications: the study
To our knowledge, no systematic review of studies examining population was non-university adolescents or young adults (13),
the prevalence of depression in undergraduate university students studies evaluating treatment of depression and/or clinical trials and
has been published. In the light of this research gap, this review has either not reporting prevalence rate and/or response rate (14), studies
two main objectives: (I) to identify studies reporting on rates of not reporting response rate and/or prevalence (23), no separate
depression among university students (II) to examine the hypoth- prevalence rate for depression (8), studies did not aim to establish
esis that there is an increase in the rates of depression among prevalence (12). The remaining 24 articles were included and were
undergraduate university students. evaluated for quality (Fig. 1). Prevalence rates across studies were
calculated as weighted means using RevMan software which takes
2. Method into account variation in cut-off used (RevMan, 2011). The prevalence
rate per study was multiplied by the corresponding sample size and
A systematic literature review of PubMed, PsycINFO, BioMed divided by the total sample size to give a weighted prevalence of
Central and Medline databases was carried out to identify peer- depression and 95% CIs were calculated (IBM-SPSS, 2009).

2303 were retrieved

After reviewing the titles

167 were available for examination

After careful reading of

the online abstracts

94 studies were eligible for examination

70 studies were excluded:

13 studies on Non - university adolescents & young adults
14 studies on treatment of depression and/or clinical trials
11 studies not reporting response rate
12 studies not reporting prevalence
8 studies examined anxiety and depression
12 studies not aimed to establish prevalence

24 were eligible for inclusion

Fig. 1. The study flow chart.

 A.K. Ibrahim et al. / Journal of Psychiatric Research 47 (2013) 391e400 393

3. Quality evaluation 8

7
The 24 articles were read extensively and, as there is no agreed
quality assessment instrument for epidemiological prevalence
6
studies, we adapted one developed by Parker and colleagues
(Parker et al., 2008). Articles scored one point for each of the 5
following quality markers: (1) the target population was defined
clearly, (2) complete, random or consecutive recruitment, (3) the 4

targeted sample is representative or the report presents evidence

3
that the results can be generalized to the general undergraduate
population (4) the response rate was equal or greater than 70%, (5) 2
the scale used is a validated measure of depression with valid cut-
offs for classification of depression, (6) the sample size is adequate 1
with a minimum sample size of 300 (Loney et al., 1998), (7) the
confidence intervals (CI) or standard error (SE) are reported. The 0
1985 1990 1995 2000 2005 2010 2015
last two quality criteria were added because the larger the sample,
the more precise the results are (Strachan, 1997). Additionally, CI Fig. 2. Change of the studies quality scores over time.
and SE are important for the reliability assessment of the outcome
of prevalence studies. In the study results either CI or SE should be
computed and always reported (Loney et al., 1998). A full descrip- Hong Kong (Song et al., 2008; Wang et al., 2010), one from China
tion of the quality assessments for the examined studies is included (Zong et al., 2010), and two from South Korea (Choi, 2003; Roh et al.,
in Table 1. In Fig. 2 the quality scores of the included studies from 2010)). Only two studies were carried out in Arabic countries (Egypt
1990 were plotted against the year of study. The regression line and Lebanon) (El-Gendawy et al., 2005; Mehanna and Richa, 2006).
indicates average quality scores over time. The remaining study was international in scope and deliberately
sampled university students from high, middle and low-income
countries (Steptoe et al., 2007).
4. Results Medical students were targeted in 12 studies (Arslan et al., 2009;
Dahlin and Runeson, 2005; Dion and Giordano, 1990; Goebert et al.,
Out of a total of 2303 publications, only 24 studies satisfied all 2009; Hendryx et al., 1991; Kaya et al., 2007; Mancevska et al.,
the inclusion and exclusion criteria Fig. 1. The majority of the 2008; Roh et al., 2010; Rosal et al., 1997; Schwenk et al., 2010;
included studies (n ¼ 15) had been carried out in Western coun- Thompson et al., 2010; Tjia et al., 2005), while eleven studies
tries. Nine had been carried out in the USA (Eisenberg et al., 2007; collected data from a sample of different faculties (Choi, 2003;
Garlow et al., 2008; Goebert et al., 2009; Hendryx et al., 1991; Curran et al., 2009; Eisenberg et al., 2007; El-Gendawy et al., 2005;
Roberts et al., 2010; Rosal et al., 1997; Schwenk et al., 2010; Garlow et al., 2008; Mehanna and Richa, 2006; Mikolajczyk et al.,
Thompson et al., 2010; Tjia et al., 2005), one in Canada (Dion and 2008; Roberts et al., 2010; Song et al., 2008; Wong et al., 2006;
Giordano, 1990), one in Sweden (Dahlin and Runeson, 2005), one Zong et al., 2010), and only one study excluded medical students
in Ireland (Curran et al., 2009), two in Turkey (Arslan et al., 2009; (Steptoe et al., 2007). The majority of studies (n ¼ 18) used
Kaya et al., 2007) and one in Macedonia (Mancevska et al., 2008). In a convenience sample (Choi, 2003; Curran et al., 2009; Dahlin and
addition, one study used data from four EU countries (Mikolajczyk Runeson, 2005; Dion and Giordano, 1990; Garlow et al., 2008;
et al., 2008). Five studies sampled East Asian students (two from Goebert et al., 2009; Hendryx et al., 1991; Kaya et al., 2007;

Table 1
Quality assessments of the studies.

SN Source Quality score Sample definition Recruitment Representative sample Response rate Scale Sample size CI or SE
1 Dion et al. 4 1 0 0 1 1 1 0
2 Hendryx et al. 3 1 0 0 1 1 0 0
3 Rosal et al. 3 0 0 0 0 1 1 1
4 Choi, M. 3 0 0 1 0 1 0 1
5 El-Gendawy et al. 6 1 1 1 1 1 1 0
6 Tjia et al. 3 1 0 0 0 1 1 0
7 Dahlin et al. 4 1 0 0 1 1 1 0
8 Mehanna et al. 5 1 1 0 1 1 1 0
9 Wong et al. 3 1 0 0 0 1 1 0
10 Kaya et al. 4 1 0 0 1 1 1 0
11 Steptoe et al. 6 1 0 1 1 1 1 1
12 Eisenberg et al. 6 1 1 1 0 1 1 1
13 Song et al. 4 1 0 0 0 1 1 1
14 Mikolajczyk et al. 6 1 1 1 0 1 1 1
15 Garlow et al. 4 1 0 0 0 1 1 1
16 Mancevska et al. 4 1 0 0 1 1 1 0
17 Goebert et al. 4 0 0 0 1 1 1 1
18 Curran et al. 2 0 0 0 0 1 1 0
19 Arslan et al. 7 1 1 1 1 1 1 1
20 Roh et al. 4 1 0 0 0 1 1 1
21 Thompsom et al. 2 0 0 0 1 1 0 0
22 Roberts et al. 5 1 1 1 0 1 1 0
23 Zong et al. 2 1 0 0 0 1 0 0
24 Schwenk et al. 3 0 0 0 0 1 1 1
394 A.K. Ibrahim et al. / Journal of Psychiatric Research 47 (2013) 391e400

Mancevska et al., 2008; Roh et al., 2010; Rosal et al., 1997; Schwenk 2006; Mikolajczyk et al., 2008; Roberts et al., 2010) was 14e71%
et al., 2010; Song et al., 2008; Steptoe et al., 2007; Thompson et al., with a weighted mean of 35.3% (95% CI, 34.3e36.6). This was
2010; Tjia et al., 2005; Wong et al., 2006; Zong et al., 2010), whereas higher than the mean rate observed in studies using convenience
random sampling was the strategy in six studies (Arslan et al., 2009; sampling (Choi, 2003; Curran et al., 2009; Dahlin and Runeson,
Eisenberg et al., 2007; El-Gendawy et al., 2005; Mehanna and Richa, 2005; Dion and Giordano, 1990; Garlow et al., 2008; Goebert
2006; Mikolajczyk et al., 2008; Roberts et al., 2010). Moreover, all et al., 2009; Hendryx et al., 1991; Kaya et al., 2007; Mancevska
studies adopted a cross-sectional design except for one longitudinal et al., 2008; Roh et al., 2010; Rosal et al., 1997; Schwenk et al.,
design (Rosal et al., 1997). A range of measures were used to identify 2010; Song et al., 2008; Steptoe et al., 2007; Thompson et al.,
depression in the articles included in this review. Twenty three 2010; Tjia et al., 2005; Wong et al., 2006; Zong et al., 2010),
studies used a cut-off score on a depression rating scale to classify where the prevalence ranged between 10.3% and 84.5% with
depression status and only one study using a (semi) structured a weighted mean at 29% (95% CI, 28.3e29.7). Comparison of studies
interview (the Mini International Neuropsychiatric Interview with small sample sizes (less than 300) with those with larger
(MINI)) to establish DSM-IV criteria (Roh et al., 2010). sample sizes found no obvious effect. Additionally, there was
Quality was evaluated for all the 24 studies according to the a modest but significant inverse relationship between the depres-
criteria demonstrated in Table 1. According to these criteria the sion prevalence rate and the response rate of the study (r ¼ 0.3,
maximum possible score for quality is 7. Actual scores ranged from p < 0.05) with poorer response rates associated with higher prev-
2 to 7, with a mean of 4.04 (SD: 1.4). The number of studies alence rates.
assessing the prevalence of depression in undergraduate students Sixteen articles reported gender difference, the majority of them
increased over time but no substantial increase in the quality of (n ¼ 9) (Dahlin and Runeson, 2005; Dion and Giordano, 1990;
studies over time was observed, as shown in Fig. 2. The overall Goebert et al., 2009; Roberts et al., 2010; Roh et al., 2010; Rosal
sample size in the current review was 48,650, with a minimum of et al., 1997; Schwenk et al., 2010; Song et al., 2008; Steptoe et al.,
102 and a maximum of 17,348 participants. The mean age ranged 2007) found higher prevalence among female compared to male
from 15 to 26 years. Gender of the participants was reported in all students, six articles could not detect any statistically significant
studies except two (Curran et al., 2009; Thompson et al., 2010). gender differences (Arslan et al., 2009; Eisenberg et al., 2007;
Percentages of males in the 22 studies reporting on sex ranged from El-Gendawy et al., 2005; Kaya et al., 2007; Tjia et al., 2005; Zong
28% to 64%. The cut-off was defined from the way depression was et al., 2010) and one found that males had a higher rate of
defined in each study (Table 2). depression (Wong et al., 2006). For the 16 studies reporting on
The prevalence of depression is shown in Table 2. Overall, gender female participants reported higher rates of depression
depression was present in nearly one-third of the total students with a weighted mean average of 29.6% (95% CI, 29.2e30.1)
studied with a weighted mean prevalence of 30.6% (95% CI, compared to 24.9% (95% CI, 24.4e25.4) in males.
30.2e31.1). Prevalence rates ranged between 10% (95% CI, The influence of student age on depression prevalence was
7.7e14.3) and 84.5% (95% CI, 80.3e86.7). Reported rates of depres- discussed in seven studies. Three found higher prevalence among
sion in undergraduate students fluctuated over the publication time younger students (Arslan et al., 2009; Eisenberg et al., 2007; Roh
period with no discernible trend (r ¼ 0.03, p > 0.05) (Figs. 3 and 4). et al., 2010), two articles stated that older students have higher
Eight different scales were used in the 24 articles included in rates (El-Gendawy et al., 2005; Schwenk et al., 2010), and no
the review. The Beck Depression Inventory (BDI) was the most difference by age was found in two articles (Kaya et al., 2007; Tjia
common tool used (n ¼ 12) (Arslan et al., 2009; Curran et al., 2009; et al., 2005). As regards the year of study, higher prevalence rates
Dion and Giordano, 1990; Hendryx et al., 1991; Kaya et al., 2007; were observed in earlier years of study (which is consistent with
Mancevska et al., 2008; Mehanna and Richa, 2006; Mikolajczyk higher rates among younger students) in six articles (Arslan et al.,
et al., 2008; Roberts et al., 2010; Steptoe et al., 2007; Tjia et al., 2009; El-Gendawy et al., 2005; Goebert et al., 2009; Mancevska
2005; Zong et al., 2010) with a weighted depression prevalence et al., 2008; Mehanna and Richa, 2006; Roh et al., 2010), while
mean of 24% (95% CI, 23.1e24.9), followed by the Center for equal rates over the university study years were observed in two
Epidemiological Studies Depression Scale (CES-D) in four studies studies (Roberts et al., 2010; Tjia et al., 2005). Socio-economic
(Goebert et al., 2009; Rosal et al., 1997; Song et al., 2008; determinants of prevalence were recorded in seven publications
Thompson et al., 2010) showed a weighted mean of 36.8 (95% CI, which, concluded that the greater the family income the lower the
35.2e38.4) and 47.7% (95% CI, 46.2e49.2) was the weighted mean prevalence of depression (Eisenberg et al., 2007; El-Gendawy et al.,
in three articles using PHQ-9 (Eisenberg et al., 2007; Garlow et al., 2005; Kaya et al., 2007; Mancevska et al., 2008; Mikolajczyk et al.,
2008; Schwenk et al., 2010). 2008; Roh et al., 2010; Steptoe et al., 2007), however two of these
Regarding the nature of the studied population, the prevalence seven studies reported higher prevalence rates among students
of depression found in studies carried out in medical student whose parents had higher education (Kaya et al., 2007; Steptoe
samples (Arslan et al., 2009; Dahlin and Runeson, 2005; Dion and et al., 2007).
Giordano, 1990; Goebert et al., 2009; Hendryx et al., 1991; Kaya
et al., 2007; Mancevska et al., 2008; Roh et al., 2010; Rosal et al., 5. Discussion
1997; Schwenk et al., 2010; Thompson et al., 2010; Tjia et al.,
2005) ranged from 10.3% to 59%, with a weighted mean of 25.6% The current review included studies published between January
(95% CI, 23.2e26.6). However, research on prevalence of depression 1990 and October 2010 and reporting on depression among
among a greater range of university students (Choi, 2003; Curran undergraduate university students including medical students.
et al., 2009; Eisenberg et al., 2007; El-Gendawy et al., 2005; According to this current review the average depression prevalence
Garlow et al., 2008; Mehanna and Richa, 2006; Mikolajczyk et al., is 30.6%, a higher rate than the 9% found in the general population
2008; Roberts et al., 2010; Song et al., 2008; Wong et al., 2006; rates of the US (range 6e12%) (Gonzalez et al., 2010). Moreover,
Zong et al., 2010) shows wider variability (range, 14e85%), with a community-based cross-national survey of depression prevalence
a higher weighted mean of 35.6% (95% CI, 34.9e37.8). For the carried out in 10 countries in North America, Latin America, Europe,
sampling methodology; the range of prevalence rates reported for and Asia and using the Composite International Diagnostic (CIDI),
studies using random sampling technique (Arslan et al., 2009; reported a mean prevalence of 9.8%, again much lower than the
Eisenberg et al., 2007; El-Gendawy et al., 2005; Mehanna and Richa, weighted mean in this systematic review of studies confined to
 A.K. Ibrahim et al. / Journal of Psychiatric Research 47 (2013) 391e400 395

Table 2
Demographics and methodologies employed in 22 studies examining depression among university students from 1990 to 2010.

SN Source Year Country Period of Study Samplea Scaleb Cut-off Quality score
1 Dion et al. 1990 Canada 1988 1a 21-BDI Normal 5
2 Hendryx et al. 1991 USA NR 1a 21-BDI-I Normal 4
3 Rosal et al. 1997 USA 1987e1989 1a 20-CES-D Normal 4
4 Choi, M. 2003 S. Korea 2002 2a 20-ZSRDS Normal 3
5 El-Gendawy et al. 2005 Egypt 2004 2b 52-ZDS Normal 7
6 Tjia et al. 2005 USA 2001e2002 1a 13-BDI-II 7 mild 4
7 Dahlin et al. 2005 Sweden 2001e2002 1a 12-MDI Normal 5
8 Mehanna et al. 2006 Lebanon 2003e2004 2b 13-BDI-II Normal 5
9 Wong et al. 2006 Hong Kong 2003 2a 42-DASS Normal 5
10 Kaya et al. 2007 Turkey NR 1a 21-BDI-II 17 5
11 Steptoe et al. 2007 23 EU 1999e2001 3a 13-BDI-II 8 8
12 Eisenberg et al. 2007 USA 2005 2b PHQ-9 Normal 8
13 Song et al. 2008 Hong Kong 2006 2b 20-CES-D Normal 6
14 Mikolajczyk et al. 2008 4 EU 2005 2b 20-M-BDI Normal 8
15 Garlow et al. 2008 USA 2002e2005 2b PHQ-9 Normal 5
16 Mancevska et al. 2008 Macedonia 2007e2008 1a 21-BDI-II 17 5
17 Goebert et al. 2009 USA 2003e2004 1a 20-CES-D 16 Mild 6
18 Curran et al. 2009 Ireland NR 2a 21-BDI-I Normal 2
19 Arslan et al. 2009 Turkey 2007e2008 1b 21-BDI-I 19 8
20 Roh et al. 2010 S. Korea 2006e2007 1a 9-MINI-RR Normal 6
21 Thompsom et al. 2010 USA 2002e2003 1a 20-CES-D Normal 2
22 Roberts et al. 2010 USA NR 2b 21-BDI-II 20 5
23 Zong et al. 2010 China NR 2a 21-BDI-II 14 3
24 Schwenk et al. 2010 USA 2009 1a PHQ-9 Normal 4

SN Sample size Response rate Sex male% Mean age Prevalence Covariates measuredc
1 432 82% 33% 20.3 y 34% Sex-women[, ethnicity[
2 110 74.8% 64% 24 y 19% Alexithymia[
3 300 48e88% 53% NR 18e39% Sex-women[, perceived stress[
4 298 45.1% 56% NR 26.8% Coping flexibilityY, number of stressful life events[, perceived controlY,
psychological sym. > somatic symptoms
5 1000 82.4% 52% 19.3 y 71% Sex¼, age[, family structure[, SES[, Residence-rural[
6 564 57.1% 54.4% 25 y 15.2% Sex¼, age¼, year of study¼
7 342 90.4% 40.9% 26.1 y 12.9% Sex-women[, stress-A
8 677 74.9% 62.2% 21.7 y 52.7% Study subject¼, study yearY
9 7915 27.5% 37% 20 y 35.1% Sex-males[, psychiatric problems history[
10 754 80.5% 42.6% 21.9 y 26.9% Sex¼, age¼, family structure[, father educ.¼, mother educ.[, family incomeY,
History of general health or psychiatric problems[
11 17,348 90% 43% 23.5 y 21% Sex-women[, parent educ.[, family wealthY, sense of controlY
12 2843 56.6% 50% 20 y 13.8% Sex¼, ageY, financial struggle[, race; white[, family setting[
13 1677 55.7% 50.3% 18.5 y HK ¼ 43.9% Sex; HK¼, B-women[, neuroticism level[, self-esteemY, perfectionism[
B ¼ 24.6%
14 2146 60e95% 36.8% Y23 y 29.5% Sex-A, income-A, country, depression; Poland > Bulgaria > Germany > Denmark
15 729 8.1% 28.3% 14.7 y 84.5% Suicidal ideation-A, stress-A, anxiety-A
16 354 75e92% 33.9% 19.3 y 10.4% Study yearY, family incomeY, stress-A substance use-A
17 1343 88% 48% NR 25% Sex-women[, ethnicity¼, study yearY, psychiatric problems history[, suicidal ideation[
18 338 62.7% NR NR 13.9% Faculty-medicals[, social supportY, suicidal ideation[, stress[, alcohol abuse¼
19 822 80.8% 45.9% 20.8 y 21.8% Sex¼, ageY, study yearY, BMI¼
20 7357 52.2% 36% 21 y 10.3% Sex-women[, ageY, living alone[, financial difficulty[, academic performanceY
21 102 85% NR NR 59.1% Before; 59% depression, 30% suicidal ideation.
After; 24% depression and 3% suicidal ideation
22 428 18% 35% 19 y 22% Sex-women[, year of study¼, aggression (scale)[, cigarette smoking[, alcohol abuse¼
23 266 62% 55% 18.5 y 21% Sex¼, coping flexibilityY, number of stressful life events[, perceived controlY
24 505 65.7% 41.6% NR 44% Sex-women[, age[, suicidal ideation[ quitting school ideas[, depression history[

NR ¼ not reported, þCI ¼ 95% confidence interval, SE ¼ standard error.

a
1 ¼ Medical sample, 2 ¼ university sample, 3 ¼ university sample except medicals. a ¼ Convenience sample, b ¼ random.
b
See the study coding in Appendix 1.
c
[ e Increase, Y e decrease, ¼ e equal, A ¼ adjusted (details in Appendix 1).

student populations (Andrade et al., 2003). Another community- However, a large cross-sectional study of a representative
based study carried out in Australia to track the changes in sample carried out in the USA as part of the National Epidemiologic
depression prevalence over 10 years period found that the preva- Survey on Alcohol and Related Conditions (NESARC) could not
lence was 10.3% in 2008 (Goldney et al., 2010). Previous studies on detect any significant difference in the prevalence of depression
young adult populations also found a lower prevalence compared between college students (7.85%, 95% CI 6.33e9.82) and their
with the current results (10.8e22%) (Denise et al., 1996; Aalto- matched non-college attendants (7.79, 95% CI 6.37e9.60) using the
Setälä et al., 2001). This might be due to the fact that students DSM-IV diagnostic criteria (Blanco et al., 2008).
experienced more stresses concerning their futures and employ- It has been suggested that rates of depression in undergraduate
ment or that they were less satisfied with their studies. It might also student have increased over time (Ceyhan et al., 2009; Denise et al.,
indicate that being a student is one of the factors that predispose to 1996), but the current review could not detect this trend. This could
depression (separation from home and lack of family support) be explained by differences in the study methods, tools used, or the
(NIMH, 2009). cultural differences of the studied population. Still a growing
396 A.K. Ibrahim et al. / Journal of Psychiatric Research 47 (2013) 391e400

90 of the general population (n ¼ 8764) in five European countries

80 (UK, Spain, Netherlands, Norway, and Greece). The overall preva-
lence of depression was estimated at 8.6% (95% CI 7.05e10.37), with
70
10.05% (95% CI 7.80e12.85) of females affected and 6.61% of males
60 (95% CI 4.92e8.83) (Ayuso-Mateos et al., 2001), markedly lower
50 than weighted mean prevalence rates reported for students in the
present review. Similarly community studies adopting the PHQ-9 to
40
screen for depression also found lower rates (4.2e9.2%) (Martin
30 et al., 2006; Yu et al., 2011) in comparison to our results where
20 the weighted prevalence mean was 47.7%. Furthermore, a compar-
ison of two studies of the prevalence of depression in Egyptian
10
samples using the Zagazig Depression Scale (ZDS) (Fawzy et al.,
0 1982; Ibrahim et al., 2010, 2012) found a much lower rate in the
1985 1990 1995 2000 2005 2010 2015
general population (26%), (Fawzy et al., 1982) compared to
Fig. 3. Change of the depression prevalence rate over time. university students 71% (Ibrahim et al., 2011). This was supported
by another using a comparable scale (CDS), which found only a 9%
prevalence of depression in the general population (Carroll et al.,
concern has been expressed about university students’ mental 1981).
health (Ceyhan et al., 2009), this was evident in our findings as the Gender difference in vulnerability to depression was evident in
quantity of publications show an increase over time. However, the the current study, consistent with other studies carried out both in
quality of study was more or less stable across time with a mean general populations (Velde et al., 2010; Van de Velde et al., 2010;
quality score of 4/7. Angst et al., 2002) and in university students (Young et al., 2010;
The BDI was most frequently reported outcome measure and Ovuga et al., 2006; Ghodasara et al., 2011). Although the difference
although rates based on BDI were lower than those based on the was statistically significant, it was not large. This was supported by
CES-D or the PHQ, the weighted depression rate in the current a previous study which concluded that gender differences are
review for studies used the BDI was high (24%) compared to studies markedly evident in the prevalence rate for major depression but
carried out in community-based samples (ranged between 5 and less so for minor depression, and this relation persisted across all
15%) (Yeung et al., 2002; Kuan-Pin et al., 2007; Katon and age groups (Van de Velde et al., 2010).
Schulberg, 1992; Poole et al., 2009). In addition, the European Many could argue that these high figures reflect an extreme
Outcome of Depression International Network (ODIN) used the BDI dose of normality as the majority of university student are
to explore the prevalence of depression in representative samples emerging from the hormonal and psychosocial chaos of

Fig. 4. Forest plot of studies on depression among college students.

 A.K. Ibrahim et al. / Journal of Psychiatric Research 47 (2013) 391e400 397

adolescence into adulthood and that there is an inflation of figures, from depression are less likely to volunteer in studies using
but in the current review we included studies that that used well- a convenience sample. This suggests that many studies may
validated tools. The use of screening tools such as the PHQ may pick underestimate the prevalence of depression in university samples.
up psychological distress rather than clinical depression and so It was also concluded that there was an inverse relationship
may inflate rates of disorder. They may also miss young people with between prevalence on one hand and sample size and response rate
an atypical presentation of depression. In view of the high mean on the other. As sample size and response rate are crucial for any
rate of depression found in this systematic review of studies which prevalence study, special attention should be drawn to their
have used well established depression scales it is important to determination and reporting (Bonita et al., 2006).
validate these measures in student populations. Also, as a screening This review encountered several limitations. The major limita-
tool there is the possibility of fallacies (positive or negative) and tion was the possibility of missing studies not directly reporting on
this should be considered when interpreting the results. The above depressive prevalence (i.e. studies examining the prevalence of
indicated that each measure should be tested for validity and reli- general distress and using measures that screen for depression as
ability in this vulnerable group before its implementation for one of the elements of general distress e.g. the General Health
depression screening. Alternatively, a well-validated and reliable Questionnaire of Symptom Checklist (SCL-90)). Additionally the co-
tool for depression screening among university students, as morbidity of anxiety and depression may lead to over-estimation of
a distinguish group in the community, should be developed and the prevalence rates in the studied papers. Publication bias is the
validated cross-culturally to avoid any diagnostic bias and to enable main drawback in any systematic review where it is proposed that
the researcher to identify the depression probability among the extreme results are more likely to be published especially in highly
studied group accurately. respected journals so conclusions exclusively based on published
An earlier review of depression among US and Canadian Medical studies (Dubben and Beck-Bornholdt, 2005), therefore, can be
students has been published (Dyrbye et al., 2006). It was part of misleading. Secondly, the average prevalence of depression in the
a more extensive review that also investigated the other sources of current review (30.6%) may have been attenuated by including
psychiatric distress such as anxiety. A similar search technique to some studies that reporting only rates of major depressive disorder
this review was used but the older review did not include a quality rather than minor depressive states. Another limitation is that
assessment of the selected articles and had a more limited scope, a limited number of studies were included in this review as many
including only students in medical faculties. Of the 40 included studies reported the prevalence of depression prevalence but did
studies, 23 articles evaluated depression among medical students, not report a response rate. This is important because the lower the
of those only 10 studies reported depression prevalence, in which response rate, the less valid (for both external and internal validity)
a slightly lower overall prevalence rate of 22.3% was reported the study as differences between non-respondents and respon-
compared to the 25.6% found in our review. This difference is dents may exist (non-response bias) in other perspectives than just
probably due to the fact that most of the studies included in the their willingness to take part in a survey (Denscombe, 2008, 2009).
review were excluded from our review due to failure to meet one or Excluding articles reanalyzing data from the same database could
more of the more stringent inclusion criteria for example studies be considered a strength since it avoids including the same data
were published before 1990 or response rates were not reported many times. At the same time researcher may have not been aware
(Dyrbye et al., 2006). of this double counting. Finally, any systematic review is affected by
The majority of studies identified were carried out in the West the weaknesses and limitations of the included studies themselves
(68%), and only two studies of those included used data from such as small sizes and poor response rates. Although all studies
developing Arabic countries. This may reflect both a publishing bias used validated measures only one used a clinical interview. Relying
and a general lack of research in developing countries which is on self-report of symptoms is likely to impact on the sensitivity and
unfortunate given the potentially higher vulnerability to depres- specificity of the classification of depression.
sion in people in less economically developed countries due to
financial struggles and the poorer quality of health care (Ben-Ezra 6. Conclusion and recommendations
and Essar, 2004; Patel et al., 2001; Licinio et al., 2008). Mental
health studies have suggested that medical education may have an Although there is a need for more in-depth research to confirm
inevitable negative effect on mental health and increase the risk of the findings of this review, there is accumulating evidence to
depression (Dyrbye et al., 2006; Nguyen, 2011). As a result, many suggest that depression represents a significant health concern in
Medical Schools adopt screening programs for depression for all 1st university populations with, on average, nearly a third of students
year students, which is not the case in other faculties. In the current affected. Furthermore the weighted mean prevalence of depressive
review we could not find any evidence of increased risk of disorders in students of 30.6% is considerably higher than rates
depression in studies recruiting medical students only, but reported in general populations. This systematic review empha-
controversially we found that studies with more heterogenous sizes that depression is a common mental health problem in
student samples had a higher weighted mean (36%) compared to university students. Although females are more at risk, the high
medical student studies (26%). This may be due to the fact that rates for male students are particularly concerning since they are
medical students are well-acquainted with mental disorders and typically less willing to access support. The results of this review
they are exposed to mental cases and learn how to deal with these suggest that more attention should be given to the identification
disorders. We also feel that the frequent recruitment of medical and management of depression in university settings. With current
students in psychological studies was due to the accessibility of economic pressures, vulnerability may increase further unless
students and good response rates. research is conducted to establish effective interventions for
It is well-known that probability sampling strengthen the management of depression in students.
external validity (generalizability) of the study results, conclusions In light of the results of this review, a proposed design for future
and inferences, however it is time-consuming, costly and requires research on the prevalence of depression among Higher Education
a level of skill (Nutbrown, 2007). In this review, it was noticed that students could consider the following; a longitudinal study design
articles adopting probability sampling reported a substantially with a considerable sample size (>300) and a randomly-selected
higher prevalence of depression compared to studies using less representative sample of students from different study disciplines
rigorous sampling (35% vs. 29%), perhaps because those suffering and from variable socio-economic backgrounds. This could
398 A.K. Ibrahim et al. / Journal of Psychiatric Research 47 (2013) 391e400

encompass several nested cross-sectional studies that include References

comparative general population samples. A well-validated and
reliable tool for depression screening designed for university Aalto-Setälä T, Marttunen M, Tuulio-Henriksson A, Poikolainen K, Lönnqvist J. One-
month prevalence of depression and other DSM-IV disorders among young
students should be implemented. Additionally, more research adults. Psychological Medicine 2001;31:791e801.
should investigate the validity and reliability of well-known Andrade L, Caraveo-Anduaga J, Berglund P, Bijl R, Vollebergh W, Dragomirecka E,
depression diagnostic tools in university student samples. et al. The epidemiology of major depressive episodes: results from the Inter-
national Consortium of Psychiatric Epidemiology (ICPE) surveys. International
Journal of Methods in Psychiatric Research 2003;12:3e21.
Conflict of interest Angst J, Gamma A, Gastpar M, Lépine P, Mendlewicz J, Tylee A. Gender differences in
depression. Epidemiological findings from the European DEPRES I and II
studies. European Archives of Psychiatry and Clinical Neuroscience 2002;252:
None. 201e9.
Arslan G, Ayranci U, Unsal A, Arslantas D. Prevalence of depression, its correlates
Contributors among students, and its effect on health-related quality of life in a Turkish
university. Upsala Journal of Medical Sciences 2009;114:170e7.
Ayuso-Mateos J, Vazquez-Barquero J, Dowrick C, Lehtinen V, Dalgard O, Casey P,
Authors contributed equally to the work. et al. Depressive disorders in Europe: prevalence figures from the ODIN study.
British Journal of Psychiatry 2001;179:308e16.
Bayati A, Beigi M, Salehi M. Depression prevalence and related factors in Iranian
Role of funding source
students. Pakistan Journal of Biological Sciences 2009;12:1371e5.
Ben-Ezra M, Essar N. Depression and anxiety in developing countries. The Lancet
None. 2004;364:1488e523.
Blanco C, Okuda M, Wright C, Hasin D, Grant B, Liu S, et al. Mental health of college
students and their non-college-attending peers: results from the national
Acknowledgments epidemiologic study on alcohol and related conditions. Archives of General
Psychiatry 2008;65:1429e37.
I am very grateful for the Ministry of Higher Education, Egyptian Bonita R, Beaglehole R, Kjellstrom T. Basic epidemiology. 2nd ed. Geneva: WHO;
2006.
Government specially Assiut University for sponsoring my whole Carroll B, Feinberg M, Smouse P, Rawson S, Greden J. The Carroll rating scale for
studies. It is a pleasure to express my deepest gratitude and grateful depression, development, reliability and validation. British Journal of Psychiatry
appreciation to the University of Nottingham for supporting this 1981;138:194e200.
Ceyhan A, Ceyhan E, Kurty Y. Investigation of university students’ depression.
study. Eurasian Journal of Educational Research 2009;36:75e90.
Choi M. Symptoms, depression, and coping behaviors of university students. Taehan
Appendix 1 Kanho Hakhoe Chi 2003;33:433e9 (in Japan).
Curran T, Gawley E, Gill M, Crumlish N. Depression, suicidality and alcohol abuse
among medical and business students. Irish Medical Journal 2009;102:249e52.
Coding of the systematic review Dahlin M, Runeson B. Stress and depression among medical students: a cross-
sectional study. Medical Education 2005;39:594e604.
Denise L, Terrie M, Avshalom C, Lynn M, Phil S, Warren S. Psychiatric disorder in
Association
a birth cohort of young adults: prevalence, comorbidity, clinical significance,
and new case incidence from ages 11e21. Journal of Consulting and Clinical
Psychology 1996;64:552e62.
[ Increase factor is associated with increased risk of depression Denscombe M. The length of responses to open-ended questions: a comparison of
Y Increase factor is associated with decreased risk of depression online and paper questionnaires in terms of a mode effect. Social Science
¼ No difference Computer Review 2008;26:359e68.
Denscombe M. Item non-response rates: a comparison of online and paper ques-
A Adjusted
tionnaires. International Journal of Social Research Methodology 2009;12:281e
91.
Dion K, Giordano C. Ethnicity and sex as correlates of depression symptoms in
a Canadian university sample. International Journal of Social Psychiatry 1990;
Tool 36:30e41.

Code Tool Cut-off used

1 a 21-BDI (Beck Depression Inventory) (0e9) Minimal depression or no, (10e18) mild depression
b 21-BDI-I (Beck Depression Inventory)-1st revision (19e29) Moderate depression, (30e63) severe depression
c 21-BDI-II (Beck Depression Inventory)-2nd revision (0e13) Minimal depression or no, (14e19) mild depression
(20e28) Moderate depression, (29e63) severe depression
d 13-BDI-II (Beck Depression Inventory)-2nd revision shortened (0e7) Minimal depression or no, (8e11) mild depression
(11e15) Moderate depression, (16e39) severe depression
e 20-M-BDI (Beck Depression Inventory) German modificationa (0e35) Minimal depression or no, (36e100) high depression
2 20-CES-D (Center for Epidemiological Studies Depression) (0e15) Minimal depression or no, (16e60) high depression
3 52-ZDS (Zagazig Depression Scale) (0e9) Minimal depression or no, (10e19) mild depression
(20e29) Moderate depression, (30e52) severe depression
4 12-MDI (Major Depression Inventory) (0e25) Minimal depression or no, (26e60) high depression
5 42-DASS (Depression Anxiety Stress Scale) (0e9) minimal depression or no, (10e13) mild depression
(14e20) Moderate depression, (21e27) severe depression
(28e42) Extremely severe
6 PHQ-9 (Patient Health Questionnaire) (0e4) minimal depression or no, (5e9) mild depression
(10e14) Moderate depression, (15e19) moderately severe
(20e27) Severe depression
7 ZSRDS (Zung Self Rating Depression Scale) (20e49) Normal range, (50e59) mildly depressed
(60e69) Moderately depressed, (70) severely depressed
8 9-MINI-RR (Mini International Neuropsychiatric Interview) (0e4) Minimal depression or no, (5e9) high depression
a
The modification of the original BDI included two approaches: (a) the four items per symptom which assessed the specific symptom’s intensity in the original BDI, were
replaced by a single statement per symptom with a six point Likert scale measuring its frequency in the last 4 weeks (with the two extreme categories labeled as 0 ¼ ‘Never’,
5 ¼ ‘Almost Always’), (b) one symptom, which had low specificity (loss of weight) was excluded. The reduction in the number of items per symptom is consistent with another
recent modification of BDI (BDI-II).
 A.K. Ibrahim et al. / Journal of Psychiatric Research 47 (2013) 391e400 399

Dubben H, Beck-Bornholdt H. Systematic review of publication bias in studies on Marsella A. Thoughts on cross-cultural studies on the epidemiology of depression.
publication bias. BMJ 2005;331:433e4. Culture, Medicine and Psychiatry 1978;2:343e57.
Dyrbye L, Thomas M, Shanafelt T. Systematic review of depression, anxiety, and Martin A, Rief W, Klaiberg A, Braehler E. Validity of the brief patient health ques-
other indicators of psychological distress among U.S. and Canadian medical tionnaire mood scale (PHQ-9) in the general population. General Hospital
students. Academic Medicine 2006;81:354e73. Psychiatry 2006;28:71e7.
Eisenberg D, Gollust S, Golberstein E, Hefner J. Prevalence and correlates of McKenzie M, Olsson C, Jorm A, Romaniuk H, Patton G. Association of adolescent
depression, anxiety, and suicidality among university students. American symptoms of depression and anxiety with daily smoking and nicotine depen-
Journal of Orthopsychiatry 2007;77:534e42. dence in young adulthood: findings from a 10-year longitudinal study. Addic-
El-Gendawy S, Hadhood M, Shams R, Ibrahim A. Epidemiological aspects of depres- tion 2010;105:1652e9.
sion among Assiut University students. Assiut Medical Journal 2005;29:81e9. Mehanna Z, Richa S. Prevalence of anxiety and depressive disorders in medical
Fawzy M, El-Maghraby Z, El-Amin H, Sahloul M. The Zagazig depression scale students. Transversal study in medical students in the Saint-Joseph University
manual. Cairo: El-Nahda El-Massriya; 1982 (in Arabic). of Beirut. Encephale 2006;32:976e82 (in French).
Garlow S, Rosenberg J, Moore J, Haas A, Koestner B, Hendin H, et al. Depression, Mikolajczyk R, Maxwell A, El Ansari W, Naydenova V, Stock C. Prevalence of
desperation, and suicidal ideation in college students: results from the Amer- depressive symptoms in university students from Germany, Denmark, Poland
ican Foundation for Suicide Prevention College screening project at Emory and Bulgaria. Social Psychiatry and Psychiatric Epidemiology 2008;43:105e12.
University. Depress and Anxiety 2008;25:482e8. Miller E, Chung H. A literature review of studies of depression and treatment outcomes
Ghodasara S, Davidson M, Reich M, Savoie C, Rodgers S. Assessing student mental among U.S. college students since 1990. Psychiatric Services 2009;60:1257e60.
health at the Vanderbilt University School of Medicine. Academic Medicine Nguyen M. Why medical school is depressing and what we should be doing about
2011;86:116e21. it? AMSJ 2011;2:65e8.
Gladstone T, Beardslee W, O’Connor E. The prevention of adolescent depression. NICE. Depression: the treatment and management of depression in adults, NICE
Psychiatric Clinics of North America 2011;34:35e52. clinical guideline 90. NICE clinical guideline 23. London: National Institute for
Goebert D, Thompson D, Takeshita J, Bryson P. Depressive symptoms in medical Health and Clinical Excellence; 2009.
students and residents: a multischool study. Academic Medicine 2009;84: NIMH. National Institute of Mental Health: depression and college students. New
236e41. York: NIMH (NIH Publication); 2003.
Goldney R, Eckert K, Hawthorne G, Taylor A. Changes in the prevalence of major NIMH. National Institute of Mental Health: depression & suicide among college
depression in an Australian community sample between 1998 and 2008. students: a fact sheet for physicians. In: N. P. (ed.); 2009. p. 700e44.
Australian and New Zealand Journal of Psychiatry 2010;44:901e10. Nutbrown C. A student’s guide to methodology. 2nd ed. London: Sage Publications
Gonalez O, Berry J, Mcknighty-Eliy I, Strine T, Edwards V, Lu H, et al. Current depression Ltd; 2007.
among adults e United States, 2006 and 2008. MMWR 2010;59:1229e35. Ovuga E, Boardman J, Wasserman D. Undergraduate student mental health at
Harvey S, Glozier N, Henderson M, Allaway S, Litchfield P, Holland-Elliott K, et al. Makerere University, Uganda. World Psychiatry 2006;5:51e2.
Depression and work performance: an ecological study using web-based Parker G, Beresford B, Clarke S, Gridley K, Pitman R, Spiers G, et al. Technical report
screening. Occupational Medicine 2011;61:209e11. for SCIE research review on the prevalence and incidence of parental mental
Hendryx M, Haviland M, Shaw D. Dimensions of alexithymia and their relationships health problems and the detection, screening and reporting of parental mental
to anxiety and depression. Journal of Personality Assessment 1991;56:227e37. health problems. York Social Policy Research Unit, University of York; 2008.
Hunt J, Eisenberg D. Mental health problems and help-seeking behavior among Patel V, Abas M, Todd C, Reeler A. Depression in developing countries: lessons from
college students. Journal of Adolescent Health 2010;46:3e10. Zimbabwe. BMJ 2001;322:482e4.
Hysenbegasi A, Hass S, Rowland C. The impact of depression on the academic Poole P, Bramwell R, Murphy P. The utility of the Beck depression inventory fast
productivity of university students. Journal of Mental Health Policy and screen (BDI-FS) in a pain clinic population. European Journal of Pain 2009;13:
Economics 2005;8:145e51. 865e9.
IBM-SPSS. Statistical package for social science. Ver.19. Standard version. SPSS Inc.; RevMan. (Review manager). Version 5.1: The Nordic Cochrane Centre. 5.1 ed.
2009. Copenhagen: The Cochrane Collaboration; 2011.
Ibrahim A, Kelly S, Challenor C, Glazebrook C. Establishing the reliability and val- Roberts S, Carol A, Kim R, Hounchell J. Relationships between aggression,
idity of the Zagazig depression scale in a UK student population: an online pilot depression, and alcohol, tobacco: implications for healthcare providers in
study. BMC Psychiatry 2010;10. http://dx.doi.org/10.1186/1471-244X-10-107. student health. Journal of the American Academy of Nurse Practitioners 2010;
Ibrahim A, Kelly S, Glazebrook C. Analysis of an Egyptian study on the socioeco- 22:369e75.
nomic distribution of depressive symptoms among undergraduates. Social Roh M, Jeon HMH, Han S, Bong-Jin H. The prevalence and impact of depression
Psychiatry and Psychiatric Epidemiology 2011. http://dx.doi.org/10.1007/ among medical students: a nationwide cross-sectional study in South Korea.
s00127-011-0400-x. Academic Medicine 2010;85:1384e90.
Ibrahim A, Kelly S, Glazebrook C. Reliability and validity of an Arabic version of Rosal M, Ockene I, Ockene J, Barrett S, Ma Y, Hebert J. A longitudinal study of students’
Hamilton depression scale in an Egyptian University student sample. depression at one medical school. Academic Medicine 1997;72:542e6.
Comprehensive Psychiatry 2012;53:638e47. Sadock B, Kaplan H. Kaplan and Sadock’s synopsis of psychiatry: behavioral
Jeon H. Depression and suicide. Journal of the Korean Medical Association 2011;54: sciences/clinical psychiatry. Baltimore: Lippincott Williams & Wilkins; 2007.
370e5. Schwenk T, Davis L, Wimsatt L. Depression, stigma, and suicidal ideation in medical
Kaplan G, Shema S, Leite C. Socioeconomic determinants of psychological well- students. JAMA 2010;304:1181e90.
being: the role of income, income change, and income sources during the Song Y, Huang Y, Liu D, Kwan J, Zhang F, Sham P, et al. Depression in college:
course of 29 years. Annals of Epidemiology 2008;18:531e7. depressive symptoms and personality factors in Beijing and Hong Kong college
Katon W, Schulberg H. Epidemiology of depression in primary care. Special section: freshmen. Comprehensive Psychiatry 2008;49:496e502.
developing guidelines for treating depressive disorders in the primary care Springer D, Rubin A, Beevers C. Treatment of depression in adolescents and adults:
setting. General Hospital Psychiatry 1992;14:237e47. clinician’s guide to evidence-based practice. In: Clinician’s guide to evidence-
Kaya M, Genç M, Kaya B, Pehlivan E. Prevalence of depressive symptoms, ways of based practice series. West Sussex: John Wiley & Sons; 2011.
coping, and related factors among medical school and health services higher Steptoe A, Tsuda A, Tanaka Y, Wardle J. Depressive symptoms, socio-economic
education students. Turk Psikiyatri Dergisi 2007;18:137e46. background, sense of control, and cultural factors in university students from
Khan M, Mahmood S, Badshah A, Ali S, Jamal Y. Prevalence of depression, anxiety 23 countries. International Journal of Behavioral Medicine 2007;14:97e107.
and their associated factors among medical students in Karachi, Pakistan. Strachan D. The nature of epidemiological studies. In: Williams H, Strachan D,
Journal of the Pakistan Medical Association 2006;56:583e6. editors. The challenge of dermato-epidemiology. London: Informa Healthcare;
Kuan-Pin S, Tsan-Hung C, Chieh-Liang H, Ming H, Chin-Chih L, Wei-Che C, et al. 1997.
Different cutoff points for different trimesters? The use of Edinburgh postnatal Thompson D, Goebert D, Takeshita J. A program for reducing depressive symptoms
depression scale and Beck depression inventory to screen for depression in and suicidal ideation in medical students. Academic Medicine 2010;85:1635e9.
pregnant Taiwanese women. General Hospital Psychiatry 2007;29:436e41. Tjia J, Givens J, Shea J. Factors associated with undertreatment of medical student
Licinio J, Wong M, Silva De Lima M, Soares B. Depression in developing countries. depression. Journal of American College Health 2005;53:219e24.
Biology of depression: from novel insights to therapeutic strategies. Wiley-VCH Van de Velde S, Bracke P, Meuleman B. Gender differences in depression in 25
Verlag GmbH & Co; 2008. European countries after eliminating measurement bias in the CES-D. Social
Lim M, Chang W, Yu X, Chiu H, Chong M, Kua E. Depression in Chinese elderly Science Research 2010;39:396e404.
populations. Asia-Pacific Psychiatry 2011;3:46e53. Vazquez F, Blanco V. Symptoms of depression and related factors among Spanish
Loney P, Chambers L, Bennett K, Roberts J, Stratford P. Critical appraisal of the health university students. Psychological Reports 2006;99:583e90.
research literature: prevalence or incidence of a health problem. Chronic Vazquez F, Blanco V. Prevalence of DSM-IV major depression among Spanish
Diseases in Canada 1998;19:170e6. university students. Journal of American College Health 2008;57:165e71.
Lowe G, Lipps G, Young R. Factors associated with depression in students at the Velde S, Brackea P, Levecque K. Gender differences in depression in 23 European
University of the West Indies, Mona, Jamaica. West Indian Medical Journal countries. Cross-national variation in the gender gap in depression. Social
2009;58:21e7. Science & Medicine 2010;71:305e13.
Lyubomirsky S, Kasri F, Zehm K. Dysphoric rumination impairs concentration on Vredenburg K, O’Brien E, Krames L. Depression in college students: personality and
academic tasks. Cognitive Therapy and Research 2003;27:309e30. experiential factors. Journal of Counseling Psychology 1988;35:419e25.
Mancevska S, Bonzinovska I, Tecce J, Pluncevik-Gligoroska J, Sivevska-Smilevska E. Wang J, Schmitz N, Dewa C. Socioeconomic status and the risk of major depression:
Depression, anxiety and substance use in medical students in the Republic of the Canadian national population health survey. Journal of Epidemiology and
Macedonia. Bratislavske Lekarske Listy 2008;109:568e72. Community Health 2010;64:447e52.
400 A.K. Ibrahim et al. / Journal of Psychiatric Research 47 (2013) 391e400

Weissman M, Bland R, Canino G, Faravelli CEA. Cross-national epidemiology of Yeung A, Howarth S, Raymond C, Nierenberg A, Fava M. Use of the Chinese version
major depression and bipolar disorder. JAMA 1996;276:293e9. of the Beck depression inventory for screening depression in primary care.
Whitton S, Whisman M. Relationship satisfaction instability and depression. Journal Journal of Nervous and Mental Disease 2002;190:94e9.
of Family Psychology 2010;24:791e4. Young C, Fang D, Zisook S. Depression in AsianeAmerican and Caucasian under-
Winter Y, Korchounov A, Zhukova T, Bertschi N. Depression in elderly graduate students. Journal of Affective Disorders 2010;125:379e82.
patients with Alzheimer dementia or vascular dementia and its influence Yu X, Stewart S, Wong P, Lam T. Screening for depression with the patient
on their quality of life. Journal of Neurosciences in Rural Practice 2011;2: health questionnaire-2 (PHQ-2) among the general population in Hong
27e32. Kong. Journal of Affective Disorders 2011. http://dx.doi.org/10.1016/j.jad.2011.05.007.
Wong J, Cheung E, Chan K, Ma K, Tang S. Web-based survey of depression, anxiety Zong J, Cao Y, Shi Y, Wang Y, Yan C, Abela J, et al. Coping flexibility in college
and stress in first-year tertiary education students in Hong Kong. Australian and students with depressive symptoms. Health and Quality of Life Outcomes
New Zealand Journal of Psychiatry 2006;40:777e82. [Online] 2010;8.