Psychometric properties of the Slovenian version of the cardiac depression scale

Cardiovascular diseases (CVD) were the cause of 40% of all deaths in Slovenia in 2016, and are the seventh most common cause of visits to the general practitioner (1). The prevalence of depression in people with CVD is high (2) and is a strong predictor of mortality and additional cardiac events (3, 4). In patients with coronary artery disease (CAD), depressive symptoms contribute to a lower quality of life and to physical limitations (5, 6). In the first week after myocardial infarction, the prevalence of depression symptoms in patients is between 30 and 40%, and that of moderate to severe major depressive disorder is between 15 and 30% (6).

The European Society of Cardiology and the American Heart Association recognise depression as a risk factor for CVD and major adverse cardiac events. They recommend routine testing for depression in cardiac patients (4, 6, 7). In Slovenia, we test patients with CVD for depression at the primary care level. The Patient Health Questionnaire-2 (PHQ-2) and the Patient Health Questionnaire-9 (PHQ-9) are administered (8). However, studies have shown that depression in people with CVD differs from major depressive disorder (9). The main symptoms of major depressive disorder are depressed mood and anhedonia (10), whereas patients with CAD complain of fatigue, anxiousness, waking at night, reduced concentration, hopelessness and depressed mood (9, 11). Evidence showed that up to 75% of patients with depression and CVD go undiagnosed, as the somatic symptoms of depression are attributed to cardiac problems and not depression (12, 13). Early identification of depression symptoms is essential for providing optimum management of depression and prevent it from developing into a major depressive disorder. A specific scale for cardiac patients was therefore created in response. The authors used the most frequent responses of cardiac patients and developed the Cardiac Depression Scale (CDS) (9).

Because CVD are highly prevalent in Slovenia and depression affects approximately between 20 and 40% of all patients with cardiac disease (14), the aim of this study was to evaluate the psychometric properties of the Slovenian version of the Cardiac Depression Scale (S-CDS) in Slovenian-speaking cardiac patients.

The findings of our study suggest that the S-CDS is a valid and reliable instrument for measuring depressive symptoms in Slovenian patients with heart disease. The study population included a wide range of ages and a typical spectrum of heart disease diagnoses.

Using the recommended cut-off score of 80 for mild-to-moderate depression and 95 for severe depression (16, 18), the S-CDS demonstrated that 27.2% of the study population had mild-to-moderate depression and 33.1% had severe depression. These findings are comparable to previous studies and underscore the importance of assessing depression in cardiac patients (17, 20, 44, 45). Mean values of individual items ranged from 1.88 to 4.39 on a 7-point Likert scale. The item ‘Dying is the best solution for me’ had the lowest value. The one with the highest value was ‘I am concerned about the uncertainty of my health’. In the S-CDS, the first item operationalises hopelessness and is one of the strongest factors for depression. The second factor operationalises uncertainty. The findings are consistent with the original article (9). To increase the clarity of the construct, the factors Uncertainty and Hopelessness were later relabelled as Fear and Suicidal Ideation (44).

The S-CDS had an acceptable level of internal consistency (Cronbach’s alpha 0.92), and the item-to-total correlation coefficients, if item deleted, were positive (ranging from 0.25 to 0.7). According to the literature, we can reason that the S-CDS has good internal reliability (46). The S-CDS also had satisfactory reliability over time (ICC=0.71). All results are similar to those in the original and other CDS translations (9, 17, 19–21, 45, 47).

Factor analysis of the S-CDS recognised six factors that explained 61% of the variance. All the six new factors had acceptable reliability and had been extracted in other research studies (19, 44, 45). In the original study, seven factors were extracted (9). Factor 1–Anhedonia explains 35% of the variance and contains six items with factor coefficients greater than 0.30. A factor from the original scale (Inactivity) loaded on Factor 1, together with two items from Fear. A possible explanation for this result can be that the items from the Anhedonia and Inactivity factors measure similar content since, in the DSM-5, Anhedonia is defined by either a reduced ability to experience pleasure or diminished interest in engaging in pleasurable activities (10). The other five factors in the S-CDS were almost identical to those described originally (Sleep Disturbance, Cognition, Mood, Suicidal Ideation and Fear).

We conducted a confirmatory factor analysis to determine whether to use the S-CDS as a one- or two-dimensional scale. Based on reports from two such studies, we used six factors extracted from the exploratory factor analysis (21, 48). In the baseline model from Hare and Davis, seven factors and two dimensions were extracted (9). According to this model, we divided our six factors into two dimensions. We then re-ran the confirmatory factor analysis on six factors measuring one dimension. The goodness-of-fit measures were acceptable for both models. However, due to a very high correlation (0.93) between the two dimensions, the six-factor model with only one dimension was reasonable (Table 3). Studies conducted to validate the original and translated versions of the CDS also suggested using the scale as a one-dimensional model (17, 19, 44, 45).

We found one item not loading above 0.3 regarding the exploratory factor analysis. The item ‘I can’t be bothered doing anything much’ was removed, as it may not assess potential depression among Slovenian patients with heart disease. The confirmatory factor analysis supported its removal from the S-CDS, showing acceptable goodness-of-fit measures for a six-factor model with 25 items.

We found the high correlation between the S-CDS and the STAI-S scores (r=0.82, p<0.001) stronger than the correlation between the S-CDS and the CES-D scores. The high correlation between the STAI-S and S-CDS suggests high comorbidity between depressive symptoms and anxiety in cardiac patients. A moderate correlation between S-CDS and CES-D suggests that the CDS measures adjustment disorder with depressed mood, while the CES-D is a measurement tool for depression symptoms. Our results are in line with other studies (9, 17, 19, 45, 47, 49, 50).

In our sample, the estimated internal consistency coefficients of the CES-D and the STAI-S were 0.73 and 0.94. The values represent good internal consistency of both questionnaires (51). The questionnaires used for determining concurrent validity were not validated in Slovenian. This is because Slovenia is a small country and we have a limited source of officially translated questionnaires. However, both are widely used in research (30–34, 38, 39). Furthermore, the validation study was a part of the ANGST study, for which we used five different questionnaires (including STAI-S). The CES-D was added for validation because it is in the public domain and is a good tool for measuring depressive symptoms in the general population and in clinical settings (23, 52). Both questionnaires were previously used in cardiac patients for measuring depressive and anxiety symptoms (12, 45, 47, 53–55).

This study suggests that the S-CDS with 25 questions is a valid and reliable instrument for measuring depression symptoms in cardiac patients in Slovenia. It could be used in clinical and research settings for the early identification and more efficient management of depression.