Data for the present study was drawn from the ongoing BOOM (Bochum Optimism and Mental Health) project, a large-scale, cross-cultural, longitudinal investigation of risk and protective factors in mental health (Maercker et al., 2015; Schönfeld, Brailovskaia, Bieda, Zhang, & Margraf, 2016; Schönfeld, Brailovskaia, & Margraf, 2017). The Ethics Committee of the Faculty of Psychology of the Ruhr-Universität Bochum approved the study. All national regulations and laws regarding research on human subjects were followed, and required permission was obtained. Participants were provided information about the purpose of the study and an assurance of their anonymity in participation, and gave written informed consent in order to participate. Furthermore, all participants were provided with information where to receive help in case of psychological distress.

In Germany, each year (since 2011) an invitation e-mail including a link to the BOOM-baseline survey is sent to all students enrolled at the Ruhr-Universität Bochum, a large state university. After completing the baseline survey including questionnaires on different aspects of positive mental health (e.g., life satisfaction, self-efficacy, social support) as well as mental health problems (e.g., depression, anxiety, and stress), participants are asked whether they allow the BOOM-research-team to contact them for further assessment waves.

Data for the present study was collected between October/December 2015 (T1) and again between October/December 2016 (T2). A collective invitation e-mail including a link to an anonymous online survey using the Unipark-server (www.unipark.de) was sent to a randomly selected sample of 300 students of the Ruhr-Universität Bochum who had already taken part in an earlier recruitment wave within the BOOM-Project. In October/December 2016, a further collective invitation e-mail including a link to the second online self-report survey was sent to those students who had participated in the survey in 2015. Both surveys included the same questionnaires.

In total 207 German students (n=147, 70.3% female; age: M=26.04, SD=5.33) at the Ruhr-Universität Bochum took part at the first measurement time point (T1) in 2015 and at the follow-up assessment (T2) in 2016. At T1, most students were either in their third (18.8%) or eleventh (15.9%) semester (range: 1-16 semesters). Students were registered in 18 different faculties. Most German students either lived together with a partner (n=124, 61.4%) or were singles (n=80, 38.6%).

Suicidal Ideation and Behavior. Twelve-month suicide ideation (“How often have you thought about killing yourself in the past year?”) was assessed using the respective item of the Suicidal Behaviors Questionnaire – Revised (SBQ-R; Osman et al., 2001). The suicide ideation item is presented with a 5-point [1=never; 5=very often (5 or more times)] Likert scale. Lifetime suicide attempts were assessed using the respective item-response from the SBQ-R: “Have you ever attempted suicide, and really hoped to die?” If participants affirmed this question, they were also asked how often they attempted suicide in their lifetime. Though the original SBQ-R consists of four items, only the aforementioned items were used in the current study due to space constraints. The SBQ-R has been recommended for screening purposes (Batterham et al., 2015) and has repeatedly been used in clinical and non-clinical samples (Osman et al., 2001).

Depression. Depression was measured by a subscale of the Depression-Anxiety-Stress Scales 21 (DASS-21; Henry & Crawford, 2005). The seven depression items are presented with a 4-point (0=did not apply to me at all; 3=applied to me very much or most of the time) Likert scale. Psychometric properties of the DASS are well established in both clinical and non-clinical studies (Henry & Crawford, 2005). A score ≥ 10 points to a mild depressive symptomatology (Lovibond & Lovibond, 1995). Internal consistency in the current sample was α=.89.

Positive mental health. The nine-item Positive Mental Health Scale (PMH-scale; Lukat et al., 2016) assesses mainly emotional aspects of well-being. It was developed to assess a single holistic concept of positive emotionality related to positive mental health. The PMH-scale's items derived from the Trierer Personality Inventory (Becker, 1989), the Freiburg Personality Inventory (Fahrenberg, Selg, & Hampel, 1989), the Mental Health Scale (Tönnies, Plöhn, & Krippendorf, 1996), the Berner Well-Being Inventory (Grob et al., 1991), and one item formulated by Lukat et al. (2016). Participants respond to statements such as “In general, I am confident”, “All in all, I am satisfied with my life”, “I feel that I am actually well equipped to deal with life and its difficulties” on a scale ranging from 0 (I disagree) to 3 (I agree). Unidimensional structure and good convergent and discriminant validity are demonstrated in samples comprised of students, patients and the general population (Lukat et al., 2016). Cronbach's alpha was α=.92 in the current study.

All statistical analyses were conducted using the statistical analysis program R 3.3.2. Initially correlation analyses were carried out to explore associations between key variables. Changes in study variables from baseline (T1) to follow-up (T2) were analyzed using t-tests for dependent samples. To evaluate the relationship between depression, positive mental health and suicide ideation, a 4-step hierarchical linear regression analysis was conducted with T2 suicide ideation (SBQ-R) as dependent variable. Age and gender – both factors being differently associated with suicide ideation (Nock et al., 2014) – were entered as control variables in Step 1; depression severity (DASS) at T1 was entered in Step 2; positive mental health (PMH) at T1 was entered in Step 3, and the interaction of T1 depression severity and positive mental health was entered in Step 4. All metric predictors were z-standardized before entering into the regression analyses. If the interaction term adds significant predictive variance to the regression model, it indicates a moderating effect of positive mental health on the association between depression and suicide ideation (Hayes, 2013).

Descriptive statistics for each measure are presented in Table 1. At baseline, 43 participants (21.5%) indicated some degree of suicidal ideation (i.e. SBQ-R Suicide Ideation Item > 0) within the past 12 months. Lifetime suicide attempts were reported by seven participants (3.4%). At the follow-up assessment 12 month later, 49 participants (23.6%) indicated some degree of suicidal ideation within the past 12 months. No participant attempted suicide between the two assessments. T-tests for dependent samples revealed that suicide ideation declined from the baseline to the follow-up assessment, t(206)=5.47, p=.000, whereas depression increased, t(206)=−3.38, p=.001, and positive mental health did not change, t(206)=0.64, p=.522.

Correlations indicated that depressive symptoms correlate positively with suicide ideation, and both depressive symptoms as well as suicide ideation correlate negatively with positive mental health at both assessments (see Table 1).

Table 2 shows the results of the hierarchical linear regression analysis with suicide ideation at T2 as criterion variable.

In Model 1, neither age nor gender was associated with suicide ideation. In Model 2, depression emerged as a significant predictor of suicide ideation and in Model 3 positive mental health was a significant predictor of suicide ideation. In Model 4, the two-way-interaction of depression and positive mental health was a significant predictor. Although significant, the additional variance explained in Model 4 was rather low.

Nonetheless, positive mental health was found to moderate the effect of depression severity on suicide ideation. As can be seen in Figure 1, for those who reported higher levels of positive mental health (M+1SD), depression severity showed no association with suicide ideation over time.

Figure 1.

Results of linear regression analysis with depression at T1 (depression_T1) predicting suicide ideation at T2 (SI_T2) for mean values (M) values M+1SD and M-1SD of the moderator PMH at T1. Light lines represent 95% confidence intervals of each regression line and shaded areas the range between regression line and corresponding 95% confidence interval.

(0.08MB).