How common depressive symptoms are and their social and demographic links among university students in Cyprus

The aim of the present study is to investigate a) the prevalence of clinical depressive symptoms, and b) possible associations between clinical depressive symptoms and i) individual, ii) parental, iii) academic characteristics, as well as iv) health status and v) substance abuse related variables.

A descriptive cross-sectional study with internal comparisons was performed between November 2010 and May 2011 amongst all undergraduate students of the Cyprus University of Technology (CUT), one of three public universities in the Republic of Cyprus. CUT is the second largest university with 2,452 active students across 10 Departments in 5 Faculties, offering access to a free education via national examinations.

The study was approved by the National Bioethics Committee as well as the Ethics Committee of the University. Prior to data collection, the Heads of all Departments were informed about the purpose of the study and the data collection procedures, securing their consent. All active undergraduate students (N 1,783 at the time of the study) were eligible to participate, independent of age, gender and nationality. Students enrolled in postgraduate programs of study, doctoral programs or students enrolled in other short-term educational programs were excluded.

The questionnaire pack along with an information sheet explaining the purpose of the study was distributed to the students during class time (either in lecture theatres, classrooms, or labs) and verbal consent was obtained. Participation in the study was voluntary and anonymous in order to guarantee confidentiality. After a short briefing, any students who did not wish to participate had the opportunity to leave the classroom. The final sample consisted of 1.500 undergraduate students (response rate of 85%). With regard to the 283 non-participants, 240 were absent on the day of the survey, 20 declined participation and 23 were excluded from the analysis due to missing/incomplete data. The questionnaires were returned in a collection box in sealed envelopes. The research team coordinated the data collection with the Studies and Student Affairs Office of the University in order to ensure that it would not coincide with mid-terms, final examinations or any other potentially stressful study-related activities, such as hospital or industry placements, internships etc.

The Statistical Package for Social Sciences Software (SPSS - version 17) was used to analyse the data. Descriptive statistics for all socio-demographic and other variables were calculated, expressed as appropriately in frequencies, mean values and standard deviation. The CES-D item scores were summed up to provide an overall score, theoretically ranging from 0 to 60, after reversing positively phrased items 4, 8, 12, and 16. Unless more than 5 items on the scale were missing (in which case a score is generally not calculated), in all other cases, the overall score was the sum of the items for which a response was provided divided by the number of answered items. The associations between the occurrence of depressive symptoms and each of the socio-demographic characteristics under study were investigated using chi-square tests. For all statistical tests, p values of 0.05 or lower were considered statistically significant. In parallel, odds ratio (and 95% confidence intervals) of depressive symptoms for each of the socio-demographics characteristics were estimated in logistic regression models. Backward stepwise multivariable logistic regression models were used in order to select among the large number of variables considered the final set associated with depressive symptoms controlling for the potential confounding effect of the rest of the variables in the final model. Unlike a previous analysis of the data focusing on the association between stressful life events and depressive symptoms [19], the aim here was not to test certain variables based on a priori defined hypothesis. Since the intention was to select the set of variables among the over 30 (some of which are highly intercorrelated) that best describe an increased odds of depressive symptoms in University student population, use of a backward technique was deemed appropriate. For all statistical tests in multivariable logistic regression model, p values of 0.10 or lower were considered statistically significant.

With regard to individual characteristics, statistically significant differences in the prevalence of clinical depressive symptoms were noted by gender, with higher frequency among female students (29.5% vs. 24, 3%, p 0.042). Similarly, those residing in rural areas presented higher prevalence of clinical depressive symptoms compared to students living in sub-urban or metropolitan areas (40.3% vs. 34.1% and 25.6%, p 0.034, respectively). In addition, students who were living alone reported higher prevalence of clinical depressive symptoms (31.5% vs. 26.9%, p 0.030) (Table 1).

A statistically significant difference in the prevalence of depressive symptoms was noted with regard to the subject of studies. In particular, students from the faculty of Engineering and Technology reported the higher occurrence of clinical depressive symptoms (35.1%, p 0.001). In addition, associations were observed with the level of academic performance. Students who failed to pass or complete a course reported higher prevalence of clinical depressive symptoms compared to those students who successfully completed the requirements within each semester (35.1% vs. 26.3%, p 0.003). Students who achieved significantly lower grades than their peers presented the highest frequency of clinical depressive symptoms (50%, p 0.045). Additionally, higher prevalence of symptoms was observed among students with learning disabilities (43.9% vs. 23.6%, p < 0.001), those whose subject of studies was not their first choice (30.4% vs. 25.5%, p < 0.036), those reporting low satisfaction with their studies (49.7% vs. 22.2%, p < 0.001) or the quality of education received (43.7% vs. 21.1%, p < 0.001).

The primary scope of the present study was to provide an estimate for the prevalence of clinical depressive symptoms among university students in Cyprus. Furthermore, the study explored clinical depressive symptoms and their association with individual, parental, academic and health-related characteristics. The study revealed that one third of students reported clinical depressive symptoms (27.9%). Moreover, the evidence identified strong links between these symptoms and specific socio-demographic, academic and health behavior characteristics.

The frequency of clinical depressive symptoms was found to be consistent with most of the studies reported by other European countries and studies from the USA. Remarkably, the frequency of clinical depressive symptoms among female students ranges from 25% to 46%, whereas lower frequency is commonly observed among male students; ranging from 12% to 34% [13,16,24,25]. The consistency of research findings could support the idea that universality does not only apply on consumed goods, but also on sentiment and general mental health status in specific populations [26].

In contrast, one might argue that this consistency might seem paradoxical due to concrete cultural differences between the aforementioned countries and Cyprus [26]. However, despite the Cypriot culture being similar to and resembling the Greek culture Papazisis et al. [27] reported a higher prevalence of depressive symptoms (52.4%) in a sample of Greek nursing students. This difference however, may stem from the fact the Papazisis et al. [27] applied a different methodological approach and instrument to assess the level of depressive symptoms. In addition the sample only comprised of nursing students. Moreover, one may argue that their results support the fact that the professional and social status of Greek nurses is in fact low [28], a factor which has been found to be associated with psychological distress [29].

In contrast, the prevalence of depressive symptoms in the present study was much higher than the study of nursing students in Sweden (10.2%) [5]. What is more, nursing students in Sweden reported the lowest prevalence of depressive symptoms, universally [26]. A possible explanation for this is that Swedish nurses perceive their professional work as respectful and fulfilling [30], which may function as a buffer system against psychological distress [28,29]. However, this research was performed over a decade ago (2002) and may not represent the current Swedish university students.

In this study there were significant differences in clinically significant depressive symptoms between genders. Female students reported a higher frequency (OR 1.70 95% CI: 1.17 – 2.33) and this is supported by several studies. Nevertheless, there are also a number of studies which contradict this finding or conclude no discrepancies between the two genders [9,14,31-34]. Generally, a higher prevalence of depressive symptoms amongst females has been attributed to socio-cultural factors, including factors related to gender role, as well as biological and psychological explanations [34-36]. In addition, it has also been proposed that women are generally more likely to accept and report such symptoms [4,36]. Overall, this finding is in agreement with the W.H.O’s reports that stated higher frequency of psychopathology in women [37].

Moreover, our study shows that the Cypriot students with permanent residency in rural and sub rural areas had a significantly higher risk of depression when compared to those living in metropolitan areas. Students in rural and sub rural areas are 1.60 times more likely to manifest clinically significant depressive symptoms compared with residing in metropolitan areas (OR 1.60 95% CI: 1.16-2.08). This may relate to both economic and social factors. For instance, a large proportion (23.5%) of students residing in rural areas reported poor economic family circumstances. In addition, young people living in rural areas may feel more “isolated”, receiving limited social stimuli compared to their colleagues living in metropolitan areas [26]. All the above might have an impact on the available social support, the development of coping mechanisms or to spirituality issues, and cause emotional pressure [26], which may lead to increased risk of mental disturbances [26,38].

In addition, our research reports that there is a significant connection between students with children and levels of depression. The study showed that students with children are 3 times more likely to develop depressive symptoms than those without (OR 3.06 95% CI: 0.96 – 9.74). Other research evidence supports this finding [5,26]. It is evident that being a mother, bearing the responsibilities of motherhood along with those of academic pursuits may lead to exposure to prolonged stress [5,26]. Such demands, in conjunction with other stressful life events e.g. a positive family history of mental disorder, poor physical health, drug abuse and disapproval of academic pursuits, may put students with the aforementioned characteristics at a high risk of disturbed mental status [19].

An additional risk factor revealed was the family status of the student’s parents, and in particular the case of separated or divorced parents. Specifically, students whose parents were divorced were 1.60 times more at risk of showing clinically significant depressive symptoms than students whose parents were living together (OR 1.60 95% CI: 2.87 – 2.91). This could be explained through findings which support that the availability of effective support systems functions as a buffer system against mental health problems [39]. Divorce or separation of parents can lead to a decrease in family support and may be a traumatic stress or for a child. There are studies showing that individuals who have experienced a significant loss during childhood (such as death of a loved one or parental divorce) are more likely to develop depressive symptoms during adulthood [39-45].

Interestingly it was found that students whose fathers’ have higher educational attainment (i.e. university degree) were more likely to report clinical depressive symptoms. (OR 1.75 95% CI: 1.13 – 2.61). This could derive from increased parental pressure from parents with a higher educational background [46] as the latter may have higher expectations of their children, i.e. to follow their academic footsteps and go on to occupations with higher social acknowledgement, and subsequent higher socio-economical power and status [26]. This might also explain why students from the Engineering and Technology School (commonly higher pass marks are required in the national entry exams) were found to be at a higher risk of depressive symptoms compared to students from other departments, e.g. the Nursing Department [26].

As mentioned above, evidence from this study suggests Engineering and Technology Cypriot students, manifest higher levels of clinically significant depressive symptoms than students studying in other faculties. Specifically in the multivariate model (Table 6), in relation to the faculty of Geotechnical Sciences and Environmental Management, only the faculty of Engineering and Technology showed a marginally statistically significant effect in comparison with the other faculties (OR 1.80, 95% CI: 1.00 - 3.08, p = 0.047). Although these differences in depression levels were not due to demographic mediator factors, such as age, gender, residency or the students’ year of study [26]; such differences might due to students varied experiences and variations in the organizational structure of each faculty [26]; and/or this most likely related to academic pressure (students in this faculty showed high level of academic performance) and social demands that college environment place on those students [19]; and/or may due to increase parental pressure from those parents with a higher educational background [46]. All the above may result in high levels of stress that eventually lead to depressive symptomatology.

Analysis of our results showed a significantly higher occurrence of clinically significant depressive symptoms amongst Cypriot students with learning difficulties. Cypriot students with learning difficulties have almost twice (OR 1.85, 95% CI: 1.30 - 2.87) to manifastate clinical depressive symptoms compared with others students without learning difficulties. In addition these students demonstrated a reduction in learning, may caused by a decrease in the level of information absorbed, and/or a decrease in their learning ability [47]. This special group seems to have particular psychological and practical difficulties, due to the peculiarities of their learning difficulties [48]; this with combination with anxiety, stress and academic failure, may probably led to academic and social isolation [49].

Moreover, remarkable findings of this study is that 1 in 3 Cypriot students whom reported with no or low level of satisfaction with their course/specialism those with no or low level of satisfaction with the quality of the education system those with no or low level of satisfaction with living arrangement, their social life and the quality of available university facilities, was up to 2.73 (no or low level of satisfaction with their course/specialism , OR 1.95 95% CI: 1.37 – 2.77, no or low level of satisfaction with the quality of the education system, OR 1.66 95% CI: 1.21 – 2.27 and no or low level of satisfaction with living arrangement, their social life and the quality of available university facilities, OR 2.73 95% CI: 2.00 – 3.72) times more at risk of manifestating clinically significant depressive symptoms compared with those whose reported with high level of satisfaction with their course/specialism, the quality of the education system and living arrangement, their social life and the quality of available university facilities. It is obvious that the quality of the relationship between students and university, affects their general, social, academic and emotional adjustment at the university [49], as well as their intention to remain and study until graduation. The interaction between the students and the university and their relation to the above, is somehow an indicator of general adjustment [50] and thus an indicator of mental health status [26].

The present study demonstrated that any physical health problems and physical disabilities were associated with depressive symptoms. Of course, physical illness is a typical symptom of depressive symptoms in students [51-53]. However, little attention has been given to this remark, because students are considered as an age group with very good physical health. One possible explanation for this could lie in the fact that students who experience physical illness have to deal both with the difficulties of daily activities and, possibly, with the social isolation coming from the combination of an introvert character and poor acceptance by fellow students.

Similarly, our results revealed a significant difference between students who self-assessed as having physical and mental health problems and those who did not. Participants who assessed their physical (OR 1.45, 95% CI: 0.79 – 2.52) or mental health (OR 11.30, 95% CI: 7.05 – 18.08) during last month as poor or very poor presented higher prevalence of clinically significant depressive symptoms compared to those students who assessed their physical health during last month as very good or excellent. In addition, an important finding in our research was that all Cypriot students (N 4) who had received treatment regarding mental health in a hospital or a clinic still showed symptoms of depression. An interesting characteristic of depression in young people is that young people who recover from a major depressive episode continue to manifest concomitant psychopathology, depressive thinking, low self-esteem, negative expectations about the future and ineffective problem solving strategies [53]. Students have a high probability of depression recurrence and of developing a new episode. In addition to these problems, students are also insensitively teased by peers because of their mental health problem, which negatively impacts on the students’ social behavior and can lead to increased introvercy.

In relation to family history of depression, in this study, the prevalence of clinically significant depressive symptoms in Cypriot students was significantly higher compared with those who had a family history depression (OR 2.85 95% CI: 1.77 – 4.60), or mental health problem (OR 1.88 95% CI: 1.80 – 4.62). Undoubtedly, the existence of a history of depression in the family is a key factor that increases the incidence of depression among Cypriot university students. It is worth noting, however, that the most common symptoms of depression in people with a family history appear in both the student and the general population. Many researches pointed out that depression was more frequent in students who had a family history of it compared with those who did not [6,25,54], and identified results similar to the present study supporting that students with a family history of depression are up to 2.5 times more likely to have depression compared with those with no family history [55]. Our results may be explained with the genetic epidemiology data suggesting that younger age of onset is associated with family history of depression [56].

With regards to drug abuse, the prevalence rates of substance use behaviors have been well documented among undergraduate students [57-60]. In regards to drug and substance use, the current research shows that smoking (OR 1.40 95% CI: 1.00 - 1.89) and drug addiction show an increased risk of manifestating clinically significant depressive symptoms (up to 5.45 times experiencing symptoms of depression for drug users, OR 5.44 95% CI: 2.95 -8.84). One explanation for this could be that since the level of stress and anxiety in Cypriot university students is rather high due to a stressful environment, a low allowance, and an intensive study tempo, they may use substances more. This is also in agreement with a study showing that smokers with anxiety disorders reported greater anxiety sensitivity, anxiety symptoms, agoraphobic avoidance, depressed mood, negative effect, stress, and life interference, when compared to non-smokers [60]. There is also some evidence that nicotine has antidepressant properties, which may explain the relationship between smokers and high occurrence of depression as people may self-medicate through smoking [61]. However, based on the cross-sectional design of the majority of such studies, including the present one, we cannot conclude whether drug abuse is a mean of self- medication against depressive symptoms, or if drug abuse leads to depressive symptoms. In any case, drug abuse seems to co-occur with depressive symptoms [60].

The above findings need to be viewed in the context of certain methodological limitations. The data collection took place in the students’ classrooms, hence, students that were absent on that day or students who declined to participate were not included in the analysis. While the response rate was particularly high, it is not clear whether the observed prevalence of depressive symptoms may in fact be an underestimate of the true prevalence, since it is likely that those who suffer from psychological distress or mental problems skip classes more often. More importantly, this was a descriptive cross-sectional study, precluding any inference on causality. At least with regards to some characteristics of the students reverse causality may be at play. For instance, behaviour characteristics (smoking, drugs addiction), or academic characteristics (e.g. failing a unit/course), may be the result rather than the cause of depression. Furthermore, Nowadays, depression research has given increasing consideration to the possibility of complex and reciprocal relations between stress and depression. Not only does stress increase the risk for depression (i.e. a stress exposure model of depression), but depression, or depressogenic vulnerabilities, in turn, may also increase susceptibility to stressful events that are at least in part influenced by the individual [39,62-64]. Longitudinal studies should aim to explore particular life-related factors that may lead to depressive symptoms. Finally, cross-national comparisons are difficult and there is a need for collaborative international studies to investigate the prevalence of depressive symptoms among student populations across different settings and cultures, by employing common psychometric tools and standard methodology. Nevertheless, the large sample in the present study and the use of the most appropriate and robust scale (the CES-D) to measure students’ depressive symptoms, allowed for a more accurate measurement of the manifestation of depressive symptoms and its correlation with socio-demographic characteristics. More importantly, in contrast to previous studies, the present study did not focus on the effect of certain demographic variables on depressive symptoms, but examine a large number of socio-demographic and educational variables. Additionally, it should be noted that backward techniques tend to capitalize on chance, while variables excluded from the final model are no more controlled for their possible confounding effect. Nevertheless, unlike a previous analysis of the same data focusing on the association between stressful life events and depressive symptoms [19] the aim here was not to test specific variables but to reduce the set of variables from the over 30 variables considered (many of which highly intercorrelated) that best describe increased odds of depressive symptoms among a University student population.