Lorem, Ipsum

Mental Health & Prevention (2015) 3, 48–56

http://dx.doi.org/1 2212-6570/& 2015 E

nCorresponding au E-mail addresses

[email protected] frank-hagen.hofman michael.sperth@stw rainer.holm-hadulla@

1Tel.: +49 622154

Mental health of students and its development between 1994 and 2012

Hanna Bergera,n, Gabriele Helga Frankeb, Frank-Hagen Hofmanna,c,1, Michael Spertha,c,1, Rainer Matthias Holm-Hadullaa,c,d,1

aCounseling Service for Students, Heidelberg University, Gartenstrasse 2, D-69115 Heidelberg, Germany bUniversity of Magdeburg-Stendal, Psychology of Rehabilitation, Osterburger Straße 25, 39576 Hansestadt Stendal, Germany cStudentenwerk Heidelberg, Psychosoziale Beratung für Studierende, Gartenstr. 2, 69115 Heidelberg, Germany dUniversidad Diego Portales, Santiago de Chile, Chile

Received 26 July 2014; received in revised form 28 October 2014; accepted 14 January 2015 Available online 23 January 2015

KEYWORDS Medical students; Psychology students; Mental health pro- blems; Psychological disorder; Development over time; Health care

0.1016/j.mhp.201 lsevier GmbH. All

thor. Tel.: +49 17 : [email protected] e (G.H. Franke), [email protected] .uni-heidelberg.de stw.uni-heidelber

3750.

Abstract The present study investigated the prevalence of psychological distress in medical and psychology students (n=293) at a German University. Sampling structure, questionnaire and process of this study were a reproduction of a prior study carried out at the same university in 1994 (n=346) which allowed for assessing changes over time. Results show that 12% of students suffered from clinically relevant psychological complaints. Compared to 1994, self-reported mental health problems have significantly decreased. Possible explanations such as an improvement of health care, economic conditions and study conditions are discussed. & 2015 Elsevier GmbH. All rights reserved.

1. Introduction

Stimulated in part by the implementation of the Bologna- Process that led to the introduction of the Bachelor-/ Master-degrees to the German educational system and

5.01.001 rights reserved.

51135146. ni-heidelberg.de (H. Berger),

erg.de (F.-H. Hofmann), (M. Sperth), g.de (R.M. Holm-Hadulla).

entailed significant changes in most study paths, concerns were voiced that many students might have difficulties adapting to the new system. Consequently, the mental health of students has received growing attention in the last few years. Mass media coverage focused on seemingly increasing demands that students have to face and reported that many would either drop out completely or risk mental health issues, also pointing at the increasing incidence of mental illness in Germany. Health insurance companies have indeed recorded an increase of psychological disorders in the German population since the 1990s (Grobe & Dörning, 2011; TK Stress Studie, 2012). Whether this development is due to an actual increase of morbidity or rather to

49Mental health of students and its development between 1994 and 2012

methodical distortions, e.g. through modified diagnostic criteria, remains unclear (Jacobi, 2013).

A recent study has found a 12-month prevalence of 27.7% for any kind of psychological disorder (Jacobi, Höfler, & Strehle, 2014) in the German general population.

Studies on the prevalence of psychological disorders in university students remain scarce, however. In the United States of America, research on student's mental health was strongly stimulated by the fact that suicide constitutes the second leading death cause in American college students (Garlow et al., 2008; Hunt & Eisenberg, 2010; Van Orden et al., 2008; Wilcox et al., 2010), hinting at severe mental problems in this population. In their most recent survey based on data from over 80,000 college students, the American College Health Association found that a total of 14.9% of students reported having been diagnosed with a depressive disorder at some point during their studies while 32% (n=3746) of those had been diagnosed within the past 12 months (ACHA, 2009).

In Germany, a representative survey carried out by the German Student Services (“Deutsches Studentenwerk”, DSW) found that 27% of students reported to have been impaired in their studies by psychological distress, while 10% of all students felt severely impaired. The prevalence of depressive symptoms varied between 14% and 22% depend- ing on age and gender. Participants mostly complained about difficulties with academic performance, lacking self- esteem, depressiveness, test anxiety and general anxiety (Hahne, 1999). A more recent survey initiated by the DSW showed that 7% of the participating students rated them- selves as significantly impaired in their studies due to health problems, 42% of which mentioning mental disorders (Middendorff, Apolinarski, Poskowsky, Kandulla, & Netz, 2013). In another study that focused solely on disabled students and students suffering from chronic diseases, 45% of the participants indicated that mental disorders had a debilitating effect on their studies with depression, anxiety disorders, eating disorders, and personality disorders being the most prevalent (Unger et al., 2012).

In a screening survey conducted at a German university, about 22% of the students met the criteria for at least one psychological disorder as defined by the PHQ-D (Bailer et al., 2008).

Compared to the general population students seem to be rather more affected by psychological complaints. In a meta-analysis including 40 studies on psychological com- plaints of medical students in the USA and Canada, pre- valence rates for depression and anxiety were consistently higher than in the general population between 1980 and 2005 (Dyrbye, Thomas, & Shanafelt, 2006). The DEGS1- survey on the health of German adults showed that psycho- logical disorders in the German population are most frequent in the age group of 18–35, which includes students as well (Jacobi et al., 2014). European studies largely confirm these results (Seliger & Brähler, 2007; Bíro et al., 2010).

Studies finding no difference between students and non- student peers are only sporadic (Hunt & Eisenberg, 2010).

Scientific literature on the development of mental health problems in students over time is scarce and results are heterogeneous. According to an investigation on the devel- opment of psychological distress in samples taken of a campus counselling center in Kansas between 1988 and

2001 (n=13,257 in total) most complaints increased, includ- ing anxiety, depression, suicidal ideas and personality dis- orders (Benton et al., 2003).

In contrast to that, both type and extent of psychological distress in clients consulting the psychosocial counselling center of the University of Heidelberg remained rather stable between 1993 and 2008 (Holm-Hadulla et al., 2009). Only some problems such as somatoform disorders and test anxiety showed an increase.

The evaluation of two field samples that had been investigated in the scope of that survey as well (n=471 in total) resulted in stable prevalence rates over time (Holm- Hadulla et al., 2009). An up-to-date study analyzing data of the psychotherapeutic ambulance for students at the Uni- versity of Goettingen even found a slight decrease of mental health problems between 2006 and 2010 (Klug, Strack, & Reich, 2013).

On the other hand studies focusing on the impact of the Bologna reform on students self-reported distress levels present rather high distress levels in Bachelor students after the implementation of the reform (e.g. Middendorff, Poskowsky, & Isserstedt, 2012; Ortenburg, 2013). According to a large survey on 4037 German Bachelor students 59% reported to have currently experienced nervous and anxious feelings and about half of the students felt overstrained.

The present study was motivated by the deficient data situation concerning the development of mental health problems in field study populations of students and the contradicting results from different psychotherapeutic counselling centers. A detailed survey at the University of Heidelberg contributes new data to student's mental health problems, accounting for the satisfaction with life and studies and socio-demographic factors as well. A comparison with data of a very similar study carried out in Heidelberg in 1994 (Soeder, 1995) allowed for assessing changes over time.

2. Methods

2.1. Process of the study

A sample of medical and psychology students (n=293) was collected at the University of Heidelberg in the 2012/2013 winter semester following the model of a historical control study carried out at the University of Heidelberg in 1994 (n=346).

The current cross-sectional study was supposed to match the historical control as well as possible, including sample structure and data collection. In both samples, psychologi- cal distress was measured with the Psychosocial Complaints List (PCL; Holm-Hadulla & Soeder, 1997) and the Symptom- Checklist 90 (SCL-90; Franke, 2002). Additionally, the Satisfaction with Life and Studies Scale (SLSS; Holm-Hadulla & Hofmann, 2007) served for assessing the student's level of general satisfaction in the investigation of 2012.

Expecting a response rate of 46% (Soeder, 1995), 800 questionnaires were distributed in a total of 24 courses to medical and psychology students of different stages of studies (first year to final year). Courses were not selected randomly but were limited to well attended, compulsory

H. Berger et al.50

courses, thus providing best possible representativeness for each subject. Students were supposed to return completed questionnaires via the post boxes of student offices in the Department of Psychology, the Department of Anatomy and the clinic for internal medicine.

2.2. Measuring instruments

2.2.1. Psychosocial complaints list (PCL) 22 items comprise subjectively perceived psychological and psychosocial strains referring to the past 12 months. A six level scale grades the severity of impairment. All items are listed in greater detail in Table 4. The responses to all single items add up to a total score. Based on previous investiga- tions a score 43 concerning single items and a total score 427.34 can be classified as clinically relevant (Holm-Hadulla & Soeder, 1997; Sperth, Hofmann, & Holm-Hadulla, 2014).

The retest-reliability is r=0.72 and the internal consis- tency α=0.76.

2.2.2. Symptom-checklist SCL-90 The SCL-90 (Franke, 2002) is a standardized instrument for the self-evaluation of psychological and psychosomatic complaints. A total of 90 questions refer to the past 7 days and are assigned to 9 categories (see Table 5). A five point Likert scale specifies the extent of the subjective suffering. Global scores give information about the overall psycholo- gical distress level. The Global Severity Index (GSI) is the mean score of all responses and measures the psychological distress in relation to all 90 questions. The Positive Symp- tom Distress Index (PSDI) measures the overall intensity of positive responses and the Positive Symptom Total (PST) comprises the number of items with positive responses. The manual states a retest-reliabilty between r=0.69 and r=0.93. The internal consistency for the subscales range from α=0.75 to α=0.97 (Franke, 2002).

2.2.3. Satisfaction with life and studies scale (SLSS) The SLSS (Holm-Hadulla & Hofmann, 2007) comprises 7 ques- tions on the overall satisfaction of students. Among others,

Table 1 Sociodemographic variables of the 2012 and 1994 sam

2012

Total (n=293)

Medicine (n=164)

Psyc (n=

Gender Female N 214 101 113

Percentage (%) 73 61.6 87. Male N 79 63 16

Percentage (%) 27 38.4 12.

Age Mean 22.89 22.9 22. SD 3.705 3.199 4. Median 22 22.0 22.

Semester Mean 4.39 5.05 4. SD 2.835 2.420 3. Median 5.0 7 5

the questions include the global satisfaction with life, satisfaction with personal accomplishments or with social surroundings. For details see Table 2.

The responses are ranked on a scale ranging from 1 (=not at all) to 5 (=very much).

The internal consistency is α=0.84.

3. Statistical methods

Data analysis was supported by SPSS 21. In addition, the online tool GraphPad was used for comparing means with the historical control for which no raw data were available. Mean, median and SD were calculated for continuous data. The Kolmogorov–Smirnov-test showed a left skewed distribution which was to be expected as the sample did not consist of patients. Provided the Levene-test resulted in homogeneity of variance t-tests for independent samples and variance ana- lyses were applied for assessing differences between the samples of 2012 and 1994. The comparison of global scores between the norm sample of the general population and the student sample of 2012 were carried out analogously.

Uni- and multivariate variance analyses described the influ- ence of socio-demographic variables on psychological distress levels.

All p-values should only be interpreted with regard to the exploratory and not confirmatory study design. The sample size was geared to the historical control (n=346), leading to a Cohen's effect size of ca. 0.09 (given an α-error=0.05 and a power (1�β) of 0.8). The level of significance was defined as pr0.05.

4. Results

A return rate of 38% resulted in a sample size of n=293. Table 1 shows the sample's most important socio- demographic features.

Compared to the University of Heidelberg's student statistics female students were overrepresented by 10% both in the 2012 and 1994 student samples. In both samples the average age was lower than the one of the entire student body in the corresponding years. In the 2012/2013

ples.

1994

ho-logy 129)

Total (n=346)

Medicine (n=190)

Psycho-logy (n=156)

221 98 123 6 63.9 51.6 78.8

125 92 33 4 36.1 48.4 21.2

88 24.3 22.7 26.1 276 4.0 – – 0 23 22.0 25.0

77 4.64 4.1 5.3 292 – – –

– – –

51Mental health of students and its development between 1994 and 2012

winter semester the average age of all psychology students in Heidelberg plus all medical students at the faculties of medicine in Heidelberg and Mannheim was 26 with a median of 25 (Universität Heidelberg, 2014).

5. Results of the SLSS

Table 2 lists the mean values relating to the total 2012 student sample.

Pearson correlation coefficients display the interrelation between the satisfaction with life and studies and psycho- logical distress measured by using the PCL and the SCL-90.

All of the SLSS's three total scores correlated significantly and negatively with all total scores of the PCL and SCL-90 (in each case pr0.0001).

6. Results of the PCL

Tables 3 and 4 demonstrate the recent project results, the main results are presented next to the data of 1994. The portion of students excessing the cut-off point for clinically relevant impairment was 15.7% in the 2012 sample. In the univariate variance analysis, the subject of studies showed a significant effect both on the PCL total score and the score for psychological problems (in each case p=0.032), but missed statistical significance when analyzing the whole model. Neither age nor gender had any significant effects. There were no interaction effects. Mild difficulties to work- and concentrate occurred most frequently. More severe impairments concerned test anxiety most frequently, fol- lowed by general anxiety, lacking self-esteem, psychoso- matic symptoms and depressiveness (see Table 4).

T-tests showed that both the PCL total score and it's two subscores were significantly reduced in the 2012 sample compared to historical control. The difference between the two samples was higher for psychological than for psycho- social problems (see Table 3).

The effect sizes were middle range altogether (Cohen's d between d=0.3 for the subscore “psychosocial problems” and d=0.5 for the PCL total score) and highest for psycho- logical problems (d=0.6).

Table 2 SLSS: mean values of the 2012 student sample.

Mean

Healthiness/productivity 2.73 (0.88) Being content with oneself 2.89 (0.90) Getting along with others 3.11 (0.76) Satisfaction with life 2.89 (0.90) Satisfaction with academic performance 2.63 (0.98) Satisfaction with current study situation 2.81 (0.93) Satisfaction with general study conditions 2.62 (0.98)

Total score 19.68 (4.34) Subscore satisfaction with life 11.62 (2.71) Subscore satisfaction with studies 8.06 (2.31)

SD in brackets.

7. Results of the SCL-90

After normalizing transformations to T-scores on the basis of a recent norm sample (n=1061; Franke, 2014), 11.9% (n=35) of students in the 2012 sample suffered from clinically relevant psychological impairment (T-values on two subscales and/or T-GSIZ63). Here, depressiveness constituted the major problem area. The overall psycholo- gical impairment of the 2012 students sample was in accordance with the one of the norm sample.

The current SCL-90 student norm sample had been collected in the years 2011/2012 and consists of 51.6% females and 48.4% males. Various subjects of study and different universities in former eastern and western Germany were involved and ensure good representativeness for German university students (Franke, 2014).

Uni- and multivariate analyses testing the effect of socio- demographic variables with the GSI as dependent variable did not yield any significant effects including interaction effects.

When comparing the different subscales of the SCL-90 simultaneously, females scored significantly higher on the subscales phobic anxiety (p=0.040) and somatization (p= 0.005).

Psychology students scored significantly higher than medical students on the subscales anxiety (p=0.011) and hostility (p=0.022). Age and stage of studies did not show any significant effect.

Overall, psychological distress levels were significantly reduced in the 2012 students sample compared to 1994 including all subscales and the global scores GSI and PSDI (see Table 5).

Effect sizes were middle range or small, for the GSI a Cohen's d=0.38, and for the PSDI d=0.21.

The decrease was most evident on the subscales inter- personal sensitivity and paranoid ideation, followed by obsessive–compulsive, anxiety and depression which are problems typical of students (see below). In this context, compulsiveness should rather be interpreted in the meaning of difficulties to work and concentrate as these are, next to compulsiveness in a narrower sense, equally covered by this subscale (Soeder, 1995). When regarding the mean values of the subscale's single items, those describing rather difficul- ties to work and concentrate were highest and showed the largest decrease compared to 1994.

When comparing the single items belonging to the sub- scale paranoid ideation, the mean value for “ideas and perceptions others don't share” showed the largest decrease since 1994.

8. Comparison with the general population

In Table 6, the results of the SCL-90 for the 2012 student sample are contrasted with those of a recent norm sample representative for the German population (n=2025; 2014). Details to the norm sample have been published elsewhere (Franke, 2014).

Table 7 contains the p-values belonging to the comparison of means with the 2012 student sample. At first glance psychological distress levels seem to be lower in students than in the general population. However, the difference of

Table 3 PCL: Total score and subscores for the 2012 and 1994 student samples.

2012 (n=293) 1994 (n=346) p Mean difference Confidence interval

Total score 17.3 (10.8)n 23.3 (13.1) o0.0001 �6.000 �7.887 to �4.113 Psychosocial problems 6.9 (4.6)n 8.1 (5.1) o0.002 �1.200 �1.961 to �0.439 Psychological problems 10.5 (7.7)n 15.2 (9.8) o0.0001 �4.700 �6.088 to �3.312

Mean values (SD in brackets). p-values resulting from t-tests. Asterisks mark significant results.

Table 4 PCL: Results of the 2012 student sample.

Percentage

Psychosocial problems Score 40

Score 43

Partnership problems 66.1 15.1 Conflicts with parents 55.6 7.8 Disease or death of close

friends/ relatives 51.9 13.3

Physical diseases 57 5.5 Housing problems 46.4 6.5 Financial problems 48.1 7.2 Considerable strain due to own child 3.1 0.3

Psychological problems Score 40

Score 43

Difficulties to work and concentrate 77.1 6.1 Test anxiety/feelings of insecurity

Towards authorities 66.2 9.2

Contact problems 47.9 2.7 Low self-esteem 70 7.5 Feelings of anxiety that I cannot explain 58.7 7.8 Feelings of aggression that are

hard to control 23.2 0.7

Compulsive thoughts. compulsive acts 20.5 0.7 Depressiveness 64.2 6.5 Suicidal thoughts 9.9 1.4 Sexual problems 27.6 2.4 Psychosomatic complaints 59.5 6.9 Problems with alcohol or drugs 12.6 0.7 Medicince-dependency 1 0 Extreme psychological distress 3.4 0.7 Other problems/complaints 32.4 4.4

Frequency of complaints in percent for all scores 40 and for clinically relevant impairment with scores 43.

H. Berger et al.52

mean GSI values misses statistical significance using a t-test. Conversely, the mean PSDI was significantly higher in the 2012 student sample than in the general population (p=0.0023).

Die PSDI indicates the average extent of impairment of those items with a value 40 whereas the GSI includes the average impairment of all 90 items.

With a Cohens' d=0.2 the effect size was small, though. Moreover, students scored higher on the subscales depres-

sion, interpersonal sensitivity and obsessive–compulsive compared to the general population. Only in the case of

depression this difference of mean scores reached statistical significance.

9. Discussion

9.1. Prevalence of mental health problems in students

In the 2012 student sample, a proportion of 11.9–15.7% of medical and psychology students reported clinically rele- vant mental health problems. These figures are in accor- dance with data of sample of students at the University of Heidelberg obtained between 2007 and 2008 (Holm-Hadulla et al., 2009) and with the large DSW-survey from 1999 (Hahne, 1999).

However, it should be mentioned that German medical and psychology students distinguish themselves from most other subjects by the strictly limited access (numerous clausus). From that point of view, one has to expect that typical student-problems such as lack of motivation and ambivalence towards the professional perspective will not occur as much as in an average student population. For this reason the data of the present survey were compared to the current SCL-90 student norm sample (Franke, 2014). As regards the latter, generalizability to German university students can be assumed due to the involvement of various subjects of study and institutions from different parts of Germany (Franke, 2014). Normalizing transformations to T-values displayed average psychological distress levels in the 2012 student sample measured with the SCL-90. Conse- quently, the current survey at the University of Heidelberg did not feature any subject-specific or regional deviances of psychological distress levels in students. Still, generaliza- bilty of the study results to students of other subjects is not given due to the limited sample coming along with the focus on replicating a historical control study.

Severely impaired students were probably underrepre- sented because they are less likely to attend courses regularly and might be less willing to take the extra effort of participating in a survey.

Owing to this selection bias, the proportion of 12% of clinically relevant mental health problems as found in the 2012 survey probably underestimates the actual rate con- cerning medical and psychology students.

9.2. Frequent mental health problems in students

The data analysis of both the PCL and the SCL-90 featured the same characteristic problem areas which prevail in national and international studies of the past 20 years as well (Bundesministerium für Bildung und Wissenschaft,

Table 5 SCL-90: Mean values of the 2012 and 1994 student samples next to the recent norm sample.

2013 Norm sample (students) (n=1061)

2012 Student sample, Heidelberg (n=293)

1994 Student sample, Heidelberg (n=346)

P Mean difference

Confidence interval

Hostility 0.38 (0.52) 0.36 (0.41)n 0.46 (0.50) o0.0065 �0.1000 �0.1719 to �0.0281

Anxiety 0.35 (0.49) 0.29 (0.31)n 0.42 (0.44) o0.0001 �0.1300 �0.1902 to �0.0698

Depression 0.54 (0.59) 0.57 (0.53)n 0.69 (0.59) o0.0075 �0.1200 �0.2079 to �0.0321

Paranoid Ideation 0.36 (0.50) 0.29 (0.40)n 0.50 (0.56) o0.0001 �0.2100 �0.2869 to �0.1331

Phobic Anxiety 0.17 (0.39) 0.10 (0.21)n 0.19 (0.29) o0.0001 �0.0900 �0.1300 to �0.0500

Psychoticism 0.24 (0.40) 0.19 (0.24)n 0.31 (0.38) o0.0001 �0.1200 �0.1705 to �0.0695

Somatization 0.43 (0.47) 0.36 (0.35)n 0.42 (0.40) o0.0459 �0.0600 �0.1189 to �0.0011

Interpersonal Sensitivity

0.50 (0.54) 0.46 (0.44)n 0.69 (0.60) o0.0001 �0.2300 �0.3131 to �0.1469

Obsessive�compulsive 0.62 (0.59) 0.59 (0.48)n 0.72 (0.55) o0.0017 �0.1300 �0.2110 to �0.0490

GSI 0.42 (0.43) 0.38 (0.29)n 0.51 (0.38) o0.0001 �0.1300 �0.1833 to �0.0767

PST 23.69 (17.52) 23.19 (13.21) – – – – PSDI 1.42 (0.52) 1.38 (0.34)n 1.46 (0.41) o0.0081 �0.0800 �0.1392 to

�0.0208

SD in brackets. P-values, mean differences and confidence intervals resulting from t-tests comparing mean values of the 2012 and 1994 student samples. Asterisks mark significant results.

Table 6 SCL-90: Mean values of the 2012 student sample compared to the norm sample of the general population.

2012 Student sample, Heidelberg 2013 Norm sample of adults

Total (n=293)

Males (n=79)

Females (n=214)

Total (n=2025)

Males (n=1.012)

Females (n=1013)

Hostility 0.36 (0.41) 0.35 (0.45) 0.36 (0.40) 0.37 (0.49) 0.37 (0.48) 0.38 (0.50) Anxiety 0.29 (0.31) 0.26 (0.31) 0.31 (0.31) 0.32 (0.47) 0.29 (0.42) 0.35 (0.51) Depression 0.57 (0.53)n 0.50 (0.54) 0.60 (0.52) 0.49 (0.59) 0.47 (0.57) 0.52 (0.61) Paranoid Ideation 0.29 (0.40)n 0.32 (0.42)n 0.28 (0.39)n 0.48 (0.59) 0.49 (0.56) 0.46 (0.62) Phobic anxiety 0.10 (0.21)n 0.06 (0.14)n 0.11 (0.23)n 0.17 (0.36) 0.15 (0.32) 0.18 (0.40) Psychoticism 0.19 (0.24) 0.22 (0.26) 0.17 (0.23) 0.23 (0.42) 0.23 (0.39) 0.23 (0.45) Somatization 0.36 (0.35)n 0.27 (0.23)n 0.39 (0.38)n 0.50 (0.50) 0.46 (0.50) 0.54 (0.50) Interpersonal

sensitvity 0.46 (0.44) 0.39 (0.39) 0.48 (0.46) 0.44 (0.54) 0.42 (0.50) 0.45 (0.57)

Obessive–compulsive 0.59 (0.48) 0.59 (0.51) 0.59 (0.47) 0.55 (0.58) 0.56 (0.57) 0.54 (0.58) GSI 0.38 (0.29) 0.34 (0.28) 0.40 (0.30) 0.41 (0.45) 0.40 (0.42) 0.43 (0.47) PST 23.19 (13.21) 21.54 (14.28) 23.80 (12.78) 24.99 (18.04) 24.32 (18.07) 25.65 (18.00) PSDI 1.38 (0.34)n 1.35 (0.32) 1.40 (0.35)n 1.30 (0.43) 1.28 (0.43) 1.32 (0.42)

SD in brackets. Those values that were higher in students than in the general population are written in bold. Asterisks mark significant results.

53Mental health of students and its development between 1994 and 2012

1986, Hahne, 1999; Holm-Hadulla et al., 2009; Knigge-Illner, 2002; Wynaden, Wichmann, & Murray, 2013). These are difficulties to work and concentrate, depressiveness,

lacking self-esteem and accordingly interpersonal sensitivity as well as test anxiety, depending on the measuring instru- ment. Apart from the latter, the first three positions are

Table 7 P-values. mean differences and confidence intervals for the comparison between the 2012 student sample and the total 2013 norm sample of the general population.

P total P males P females Mean difference Confidence interval

Hostility o0.7393 o0.7202 o0.5830 �0.0100 �0.0690 to 0.0490 Anxiety o0.2894 o0.5343 o0.2695 �0.0300 �0.0856 to 0.0256 Depression o0.0282 o0.6512 o0.0743 0.0800 0.0084 to 0.1516 Paranoid ideation o0.0001 o0.0084 o0.0001 �0.1900 �0.2599 to �0.1201 Phobic anxiety o0.0012 o0.0133 o0.0135 �0.0700 �0.1123 to �0.0277 Psychoticism o0.1113 o0.8228 o0.0579 �0.0400 �0.0893 to 0.0093 Somatization o0.0001 o0.0008 o0.0001 �0.1400 �0.1994 to �0.0806 Interpersonal sensitivity o0.5449 o0.6025 o0.4706 0.0200 �0.0460 to 0.0860 Obsessive–compulsive o0.2603 o0.6501 o0.2376 0.0400 �0.0298 to 0.1098 GSI o0.2679 o0.2986 o0.3705 �0.0300 �0.0832 to 0.0232 PSDI o0.0023 o0.1570 o0.0094 0.0800 0.0285 to 0.1315

H. Berger et al.54

consistent with the survey of 1994. Eventually, the results of the recent SCL-90 norm sample also confirm the findings of the present study (Franke, 2002, 2014).

Not only are these major problem areas relevant due to their frequent occurrence in students, but also due to their negative impact on the general satisfaction with life.

The SCL-90 subscales reflecting these problems evidenced the greatest negative correlations with the SLSS-score “Satisfaction with life”.

9.3. Mental health problems in students compared to the general population

Compared to the recent SCL-90 norm sample of the general population (Franke, 2014), the overall psychological distress levels in the 2012 student sample were largely the same.

However, medical and psychology students from Heidel- berg were slightly more affected by depressiveness, com- pulsiveness and social insecurity. Remarkable is the significantly increased mean PSDI in the 2012 student sample compared to the norm sample of the general population in combination with an approximately equal mean GSI. This could imply that psychological complaints in medical and psychology students are less diverse than in the general population, but more frequent and more severe in some typical problem areas. This seems plausible as students form a social subgroup featuring a more homo- geneous distribution of specific risk factors, external con- ditions and biographical features such as age and level of education.

9.4. Development over time

The prevalence of mental health problems was significantly lower in the 2012 student sample than in the one of 1994. The difference was biggest for problem areas typical of students. The considerable decrease on the SCL-90 subscale paranoid ideation could be explained by an increasing cultural diversity and variety of ideologies at universities as the item “ideas and opinions others do not share” showed the largest difference of means.

As to the interpretation of the PCL it should be noted that the cut-off point defining clinically relevant impairment had been slightly modified since 1994. In the present study a cut-off point of 27.34 (Sperth et al., 2014) was applied, whereas a cut-off point of 34 had been valid at the time the historical control study was carried out (Soeder, 1995). Consequently, the effect size of the decrease of psycholo- gical distress levels as calculated in the present study thereby slightly underestimates the actual effect size.

There is little scientific literature confirming decreasing mental health problems in students so far (Klug et al., 2013). Sporadic European studies present at least consistent prevalence rates of psychological distress in students over time (Holm-Hadulla et al., 2009; Knigge-Illner, 2002; Quince et al., 2012).

According to a recently published paper prevalence rates of psychological disorders have been rather constant since 1989 in the general population in Germany (Jacobi et al., 2014).

On account of limited reference data from previous investigations sampling errors and selection bias have to be considered as well. However, a descriptive comparison with large norm samples of 2002 and 2013 displays similar tendencies supporting the findings of the present study. Here too, the decrease of psychological distress levels is most evident for typical problems reported particularly frequently by students (Franke, 2002, 2014). The decline of psychological complaints does not extend to the repre- sentative norm samples of the general population (Franke, 1995, 2002, 2014) possibly implying a gradual approximation of distress levels in both groups.

A quantitative and qualitative improvement of the gen- eral health care situation or either greater utilization of counselling services and therapy are two of many possible explanations for the observed decline in prevalence rates of mental health problems in students.

The universities' psychological counselling centers for instance present increasing numbers of clients (Knigge- Illner, 2002; PBS, 2010) and also the demand for ambulatory psychotherapeutic care by students has increased during the last few years (Holm-Hadulla et al., 2009; Knigge-Illner, 2002; TK Stress Studie, 2012). Furthermore, student coun- selling services have become more effective (Skopinceva

55Mental health of students and its development between 1994 and 2012

et al., 2013; Sperth et al., 2014). Besides, growing eco- nomic prosperity and more favorable study conditions might be significant in this context as well. The employment market is generally more promising than 10 or 20 years before, not least for medical students who face a growing demand for physicians in Germany.

Modernized university programs with greater diversity of training schemes and forms of learning and flatter hierar- chies offer more opportunities to students possibly prevent- ing study-related psychological distress to a certain degree. This might also extend to reforms leading to more struc- tured framework conditions for students, possibly facilitat- ing organization and self-management throughout the course of studies. Examples of such reforms are the new medical course at the University of Heidelberg implemented in 2001 or even the controversially discussed Bologna- reform. The latter is quite often cited to increase academic pressure and subjectively perceived stress levels in students (Gusy, Lohmann, & Drewes, 2010; Gusy, Lohmann, & Marcus, 2012; Thees et al., 2012; Sieverding, Schmidt, Obergfell, & Scheiter, 2013) but actually there is no study showing an increase of mental disorders by reason of the Bologna- process.

Indeed, self-reported stress levels of Bachelor students seem to be high (Ortenburg, 2013) but maybe this does not necessarily imply an increasing susceptibility to mental health problems. It is remarkable that a large survey on German Bachelor students did not only find high distress levels but also a proportion of 90% of students feeling able to cope with difficult study-related problems.

The observed decline of psychological distress according to the present study and data of the norm samples discussed (Franke, 2002, 2014; Klug et al., 2013) is based on a decline of self-reported complaints. In view of our performance- oriented society it is also conceivable that not the actual complaints have declined but only the willingness of admit- ting to them in a non-therapeutic setting.

A certain stigmatization of people impaired in their productivity seems plausible and can be found in the general population as well (Jacobi et al., 2014) where psychological disorders attract negative attention in a working environment with high demands to social skills. In this context this leads to a higher demand for psychother- apeutic counselling and psychotropic drugs (Jacobi et al., 2014) which can also be found in case of students (Holm- Hadulla et al., 2009; Grobe & Dörning, 2011; Knigge-Illner, 2002; Middendorff et al., 2012; Quince et al., 2012; PBS, 2010; TK Stress Studie, 2012).

9.5. Sample and methodical limits

Neither the 2012 nor the 1994 sample are representative random samples which entails possible distortions. The lower average age of the recent sample compared to the student statistics of the University of Heidelberg can partly be ascribed to an overrepresentation of psychology students in the Bachelor's program compared to the Master's pro- gram. Furthermore, advanced students not attending courses while preparing for their final exams, earning a doctorate or, in case of medical students, completing their final and exclusively practical year, could not be included.

As regards sample size and – structure, the 2012 sample matches the historical control very well, ensuring a valid comparison between the two. All in all, the decline of mental health problems in students as observed in the present study seems plausible when recent data of other German universities are taken into account (Franke, 2014; Klug et al., 2013).

10. Conclusions

Constituting the first study of its' kind in Europe the present paper compares psychological distress levels found in two analogously conducted cross-sectional studies on student samples of a German university collected at different points in time featuring very similar sample sizes and – structures. The observed decrease of psychological complaints in students represents the central finding of this study and hints to a new epidemiological development. Although sample-related distortions have to be taken into account and other recent surveys hint to increasing complaints about study-related stress and excessive demands at uni- versities, the results are in accordance with some new data existing on the topic in Germany. The diversity in the discussed results may be caused by the fact that the compared findings nearly always have their origin in self report data of students. These reports are influenced by many different factors and trends. To increase the validity and reliability of research-outcomes in the topic of mental health problems of students we need a representative study that is based on psychodiagnostic instruments and inter- views applied by psychological experts and that would refer to the diagnostic categories of international classification systems.

Possibly, the observed development is due to an improve- ment of health care utilization, economic conditions and study conditions or to an unwillingness of communicating mental health problems in the face of a growing pressure to perform.

Conflict of interest

The present study has been carried out with the approval of the ethics committee of the faculty of medicine at the University of Heidelberg.There has been no funding of the study and no conflict of interest.

References

American College Health Association (2009). American College Health Association-National College Health Assessment Spring 2008 Reference Group Data Report (abridged): the American College Health Association. Journal of American College Health, 57(5), 477.

Bailer, J., Schwarz, D., & Witthöft, M., et al. (2008).Prävalenz psychischer Syndrome bei Studierenden einer deutschen Uni- versität. Psychotherapie, Psychosomatik, Medizinische Psycho- logie, 58(11), 423–429.

Benton, S. A., Robertson, J. M., & Tseng, W., et al. (2003).Changes in counseling center client problems across 13 years. Professional Psychology: Research and Practice, 34(1), 66–72.

H. Berger et al.56

Bíro, E., Balajti, I., & Andány, R., et al. (2010).Determinants of mental well-being in medical students. Social Psychiatry and Psychiatric Epidemiology, 45(2), 253–258.

Bundesministerium für Bildung und Wissenschaft (Eds.) (1986). Das soziale Bild der Studierendenschaft der Bundesrepublik Deutschland. 11. Sozialerhebung des Deutschen Studenten- werks. Bonn.

Dyrbye, L. N., Thomas, M. R., & Shanafelt, T. D. (2006). Systematic review of depression, anxiety, and other indicators of psycholo- gical distress among U.S. and Canadian medical students. Academic Medicine, 81(4), 354–373.

Franke, G. (1995). SCL-90-R. Die Symptom-Checkliste von Deroga- tis–Deutsche Version–Manual. Göttingen: Beltz Test.

Franke, G. H. (2002). SCL-90-R. Symptomcheckliste von L. R. Derogatis. Zweite, vollständig überarbeitete und neu normierte Auflage. Göttingen: Beltz Test.

Franke, G. H. (2014). SCL-90s-S. Symptom-Checklist-90s-Standard – Manual. Göttingen: Hogrefe.

Garlow, S. J., Rosenberg, J., & Moore, J. D., et al. (2008). Depression, desperation, and suicidal ideation in college stu- dents: results from the American Foundation for Suicide Pre- vention College Screening Project at Emory University. Depression and Anxiety, 25, 482–488.

Grobe, T. & Dörning, H. (2011). Gesundheits report 2011 – Veröffentlichungen zum Betrieblichen Gesundheitsmanagement der TK, Band 26. Herausgegeben von der Techniker Kranken- kasse. Lübeck: Schipplick Winkler PrintMedien.

Gusy, B., Lohmann, K., & Marcus, K. (2012). Sind Bachelor-/ Masterstudierende stärker ausgebrannt? Prävention und Gesund- heitsförderung, 7(4), 237–245.

Gusy, B., Lohmann, K., & Drewes, J. (2010). Burnout bei Studier- enden, die einen Bachelor-Abschluss anstreben. Prävention und Gesundheitsförderung, 5(3), 271–275.

Hahne, R. (1999). Studium und psychische Probleme. Sonderaus- wertung zur 15. Sozialerhebung des Deutschen Studentenwerks. Dt. Studentenwerk: Bonn.

Holm-Hadulla, R. M., & Soeder, U. (1997). Psychische Beschwerden und Störungen von Studierenden. Psychotherapie, Psychosoma- tik, Medizinische Psychologie, 47(12), 419–425.

Holm-Hadulla, R. M., & Hofmann, F.-H. (2007). Lebens- und Studienzufriedenheitsskala. Berlin: Deutsches Studentenwerk.

Holm-Hadulla, R., Hofmann, F., & Sperth, , et al. (2009).Psychische Beschwerden und Störungen von Studierenden. Psychothera- peut, 54, 346–356.

Hunt, J., & Eisenberg, D. (2010). Mental health problems and help- seeking behavior among college students. Journal of Adolescent Health, 46(1), 3–10.

Jacobi, F. (2013). Epidemiologie psychischer Störungen. Häufigkeit und Krankheitslast in Deutschland. Psychotherapeut, 58, 191.

Jacobi, F., Höfler, M., & Strehle, J., et al. (2014).Psychische Störungen in der Allgemeinbevölkerung. Der Nervenarzt, 85(1), 77–87.

Klug, C., Strack, M., & Reich, G. (2013). Belastungen von Bachelor- und Diplom-Studierenden. Psychotherapeut, 58(2), 159–164.

Knigge-Illner, H. (2002). Psychosoziale Probleme Studierender im Wandel der Zeiten aus der Sicht Psychologischer Beratung. Gruppendynamik und Organisationsberatung, 33(1), 43–56.

Middendorff, E., Poskowsky, J., & Isserstedt, W. (2012). Formen der Stresskompensation und Leistungssteigerung bei Studierenden. Hannover: HIS Hochschul-Informations-System.

Middendorff, E., Apolinarski, B., Poskowsky, J., Kandulla, M., & Netz, N. (2013). Die wirtschaftliche und soziale Lage der

Studierenden in Deutschland 2012. Berlin: Bundesministeriums für Bildung und Forschung.

Ortenburger, A. (2013). Beratung von Bachelorstudierenden in Studium und Alltag. Ergebnisse einer HISBUS-Befragung zu Schwierigkeiten und Problemlagen von Studierenden und zur Wahrnehmung, Nutzung und Bewertung von Beratungsangebo- ten. HIS:Forum Hochschule 3/2013. Hannover: HIS.

Psychotherapeutische Beratungsstelle für Studierende des Studen- tenwerks Heidelberg (PBS) (2010). Geschäftsbericht 2009. Inter- nes Dokument. Heidelberg: Studentenwerk.

Quince, T. A., Wood, D. F., & Parker, R. A., et al. (2012).Prevalence and persistence of depression among undergraduate medical students: a longitudinal study at one UK medical school. BMJ Open, 2, 4.

Seliger, K., & Brähler, E. (2007). Psychische Gesundheit von Studierenden der Medizin. Eine empirische Untersuchung. Psy- chotherapeut, 52(4), 220–286.

Sieverding, M., Schmidt, L. I., Obergfell, J., & Scheiter, F. (2013). Stress und Studienzufriedenheit bei Bachelor-und Diplom-Psy- chologiestudierenden im Vergleich. Psychologische Rundschau, 64(2), 94–100.

Skopinceva, M., Ackermann, E., & Frommer, J., et al. (2013). Psychische Belastung von ratsuchenden Studierenden. In: A. Fischer (Ed.), Nachwuchswissenschaftlerkonferenz Ost- und Mitteldeutscher Fachhochschulen (pp. 455–456). Glückstadt: Hülsbusch.

Soeder, U. (1995). Die Prävalenz von psychosozialen Beschwerden bei Heidelberger Studierenden. Heidelberg: Psychologisches Institut der Universität Heidelberg.

Sperth, M., Hofmann, F. H., & Holm-Hadulla, R. M. (2014). Effektivität integrativer psychotherapeutischer Beratung für Studierende. Psychotherapie Psychosomatik � Medizinische Psy- chologie, 64(06), 224–231.

Techniker Krankenkasse (2012). TK-Stress-Studie NRW-Studenten, 2012. Ergebnisse einer repräsentativen Forsa-Umfrage aus Mai. ⟨https://www.tk.de/centaurus/servlet/contentblob/456454/ Datei/80754/Forsa-Studie%20Studentenalltag%20in%20NRW.pdf⟩. Stand: 11.10.2014.

Thees, S., Gobel, J., & Jose, G., et al. (2012).Die Gesundheit von Studierenden im Bologna-Prozess. Prävention und Gesund- heitsförderung, 7(3), 196–202.

Universität Heidelberg. Studierendenstatistik. ⟨http://www.uni-hei delberg.de/studium/interesse/faecher/statistik.html⟩. Stand: 2.2.2014.

Unger, G., et al. (2012).Studying with impairments in Germany. A 2011 survey of the situation for students with a disability or chronic illness. Deutsches Studentenwerk: BMBF.

Van Orden, K. A., Witte, T. K., & James, L. M., et al. (2008).Suicidal ideation in college students varies across semesters: The mediating role of belongingness. Suicide and Life-Threatening Behaviour, 38, 427–435.

Wilcox, H. C., Arria, A. M., & Caldeira, K. M., et al. (2010). Prevalence and predictors of persistent suicide ideation, plans, and attempts during college. Journal of Affective Disorders, 27 (1–3), 287–294.

Wynaden, D., Wichmann, H., & Murray, S. (2013). A synopsis of the mental health concerns of university students: Results of a text- based online survey from one Australian university. Higher Education Research & Development, 32(5), 846–860.