Depression and anxiety symptoms in cardiac patients: a cross-sectional hospital-based study in a Palestinian population

This cross-sectional hospital-based study was conducted on patients consecutively admitted to the cardiology and cardiac surgery departments of An-Najah National University Hospital, Arab-Specialized Hospital, Watani Hospital and Nablus Specialized Hospital in the Northern West Bank city of Nablus, Palestine. Patients were eligible for the study if they were between 30 and 80 years of age and had an existing or newly confirmed cardiac diagnosis warranting hospitalization during the period between March and November 2017. In the present study, cardiac diagnoses considered included CAD, ST elevation or non-ST elevation MI, angina, HF, cardiac arrhythmia, valve disease or any other cardiac disease. Diagnoses were confirmed using hospital medical charts. Patients were excluded if they had a normal cardiac catheterization (CATH), an acute or past stroke, end-stage kidney disease (including dialysis patients), peripheral vascular disease, major co-morbidities affecting mental health (alcohol abuse, drug abuse), neurological disorders (dementia, Alzheimer’s disease, epilepsy, Parkinson’s disease), cognitive impairment, a severe clinically diagnosed psychiatric condition or any other condition affecting the quality of their responses.

Eligible patients were recruited for in-person interviews by trained medical research assistants. Patients were identified using hospital registries and medical records. Interviews were conducted while patients were awaiting treatment (after CATH) or after their treatment, within 1 week of their admission to the hospital. Eligible patients were informed about the study objectives and benefits and written informed consent was obtained from those who agreed to participate. The study was approved by the Ethics Committee of Nordwest-und Zentral Schweiz (EKNZ) in Basel, Switzerland, and by the Institutional Review Board (IRB) committee at An-Najah National University in Nablus, Palestine.

Data was collected using a structured questionnaire consisting of two parts (Tables 1 and 2; see Additional file 1 for more detailed description). The first part included detailed socio-demographic and clinical information obtained from patients’ administrative and medical charts.

The second part was administered during a private interview and consisted of a sequence of screening instruments. These validated tools have demonstrated to be suitable for clinical populations to assess for depression and anxiety symptoms, QoL, post-traumatic stress disorder (PTSD), social support, resilience, self-esteem, somatic symptoms and lifestyle behaviors. The questionnaire, including the instruments, was translated from English to Arabic and back-translated from Arabic to English by two bilingual experts.

The primary endpoints in the present study were depression and anxiety. Descriptive statistics for stress as an endpoint are presented in an Additional file 2: Table S2, with no further description. The four predictor blocks investigated for association with depression and anxiety were socio-demographic, clinical, psychosocial and lifestyle factors (Table 1). Endpoints and predictors were described as means and standard deviations (SD) for quantitative variables and as absolute values and percentages for categorical variables. Differences in predictor variables according to presence or absence of depression and anxiety symptoms were described using chi-squared test and the Wilcoxon rank sum test, as appropriate. Fisher’s exact test was used for results presenting a frequency below five. Multivariate ordered logistic regression analyses were performed to examine the independent association between predictor variables and depression or anxiety symptoms. All variables were entered in each of the models at once. Results are presented in separate models for depression and anxiety and are expressed as odds ratio (OR) and 95% confidence intervals (95% CI). The cut-offs for the respective outcome variables were normal, mild/moderate, and severe/very severe according to standardized cut-offs. Correlations between the outcome variables and other instruments used in the study were assessed using Spearman’s rank correlation coefficient. Cronbach alpha was used to assess the internal consistency of the different scales. Statistical significance was defined as a two-sided p-value <0.05. All data was analyzed using the STATA Data Analysis and Statistical Software, version 14.

Characteristics of the participants are presented in Table 3.

Table 4 shows the proportion of patients with depression (CDS), depression (DASS-depression), anxiety (DASS-anxiety) and stress (DASS-stress) symptoms at different severity levels. Cutoffs for the levels of CDS and DASS subscales are also presented in Table 4. Based on our findings, the mean ± SD depression score on the CDS was 101.3 ± 15.6 and the overall proportion of patients with depression was 78.7%. According to the recommended cutoffs for the CDS, 21.3, 25.2 and 53.5% of the sample had no, mild-to-moderate and severe-to-very severe depression symptoms, respectively. The means ± SDs on the DASS-42 were 9.4 ± 8.6, 9.4 ± 6.8 and 15.2 ± 9 for depression, anxiety, and stress, respectively. It was found that the overall proportion was 52.9, 53.1 and 37.4% for the presence of depressive, anxiety and stress symptoms according to the DASS-42. Based on recommended cutoffs, 47.1, 33.4 and 19.5% of patients reported normal, mild-to-moderate and severe-to-very severe depressive symptoms (DASS-depression). In addition, 46.9% patients did not report any anxiety, while 33.9% reported mild-to-moderate anxiety and 19.2% reported severe-to-very severe anxiety symptoms. According to the stress scale, 62.6% reported having no stress, 30% reported having mild-moderate stress symptoms and 7.4% severe-very severe stress symptoms. Patients that had mild-moderate and severe-very severe symptoms of depression or anxiety according to the DASS-42 were more likely to have severe depressive symptoms on the CDS. Nevertheless, 8 (4.0%) and 9 (4.6%) patients without any signs of depressive symptoms on the CDS, showed symptoms of depression and anxiety, respectively, on the DASS-42.

Spearman correlation coefficients were calculated between the CDS and other instruments used in the study (see Fig. 1). The correlation of the CDS with the DASS-depression was =0.57 (p = < 0.001), with DASS-stress =0.51 (p < 0.001), and with DASS-anxiety =0.50 (p < 0.001). The correlations between the DASS subscales were as follows: DASS-depression and DASS-anxiety (0.65, p < 0.001); DASS-depression and DASS-stress (0.63, p < 0.001); DASS-anxiety and DASS-stress (0.61, p < 0.001). CDS was positively correlated with the PHQ-15 (0.44, p < .001) and the PTSD-PCL-S (0.37, p < .001)and the DASS-depression was also positively correlated with the PTSD-PCL-S (0.57, p < .001).

The bivariate distribution of socio-demographic, clinical, psychological, and lifestyle characteristics according to the presence or absence of depression symptoms (CDS), anxiety symptoms (DASS-anxiety) and stress symptoms (DASS-stress) is presented in Additional file 3: Table S1 (depression (CDS) and anxiety) and Additional file 2: Table S2 (stress). Briefly, with regard to socio-demographic factors, bivariate analysis revealed depressive and anxiety symptoms were more frequent among females than males (depression: 84.9% vs. 76.4%, p = 0.003; anxiety: 72.4% vs. 55.6%, p = < 0.001) and among those with lower educational level (depression: 81.0% [no high school diploma] vs. 78.0% [high school diploma] and 71.1% [college degree], p = 0.02; anxiety: 63.2% [no high school diploma] vs. 59.8% (high school diploma) and 52.8% (college degree), p < 0.05). Both, depressive and anxiety symptoms were most prevalent among those unemployed and housewives (depression: p < 0.001; anxiety: p < 0.001).

Multivariable ordered logistic regression was performed to determine the independent association of factors in the four predictor blocks with a) depressive symptoms, categorized as no depressive symptoms, moderate depressive symptoms, and severe depressive symptoms according to the CDS and b) anxiety symptoms, categorized as minimal anxiety symptoms, mild-moderate anxiety symptoms, and severe-severe anxiety symptoms according to the DASS-anxiety subscale. The results for the four blocks of variables are presented in Table 5.

Overall, most of the psychosocial factors were consistently associated with both, depression and anxiety. Participants with depression or anxiety were more likely to exhibit at least some symptoms of PTSD. Odds ratios tended to be higher for anxiety than for depression (moderate symptoms vs. minimal symptoms: OR_depresssion 1.87 (95% CI 1.29–2.71) vs. OR_anxiety 3.01 (95% CI 2.12–4.27). Patients with depression or anxiety had a lower score for the mental component of QoL (SF-12-MCS) [OR_depression 0.96 (95% CI 0.95–0.97); OR_anxiety 0.98 (95% CI 0.97–0.99)]. High resilience and high social support were inversely associated with depression and anxiety [high vs. low social support: OR_depression 0.71 (95% CI 0.52–0.97); OR_anxiety 0.74 (95% CI 0.54–1.00)]. The inverse association with resilience tended to be stronger in the presence of anxiety [high vs. very low resilience: OR_depression 0.42 (95% CI 0.18–0.94); OR_anxiety 0.22 (95% CI 0.11–0.48)].

In addition to psychosocial factors, the physical component of QoL (SF-12-PCS) and somatic symptoms (PHQ-15) also showed consistent associations with both depression and anxiety [high vs. minimal somatic symptoms: OR_depression 3.00 (95% CI 1.73–5.18); OR_anxiety 7.64 (95% CI 3.96–14.78)]; [SF-12-PCS: OR_depression 0.98 (95% CI 0.97–1.00); OR_anxiety 0.97 (95% CI 0.96–0.99)]. Current smoking was associated with depression with the strongest odds ratio observed for depression and former smoking: OR 1.87 (95% CI 1.17–2.97). Finally, patients with depression and anxiety were more likely to be physically inactive compared to patients without the respective psychological problems.

A few factors exhibited associations with only one of the two mental health outcomes. For example, patients residing in villages or camps were less likely to show symptoms of anxiety compared to patients living in the city [(OR_village 0.74 (0.54–1.00); OR_camp 0.56 (0.31–1.01)]. Unemployment was positively associated with depression, but not anxiety. The presence of previous cardiac diagnoses was positively associated with anxiety, but inversely associated with depression [(OR_depression 0.67 (95% 0.47–0.96); OR_anxiety 1.52 (95% 1.05–2.21)]. Symptoms of depression were more frequent among patients with a cardiac diagnosis for more than 10 years compared to patients with a diagnosis for a year or less [years since first diagnosis ≥10 vs. ≤ 1 year OR 1.71 (95% 1.06–2.75)]. Patients who were diagnosed with angina were more likely to have anxiety symptoms than those diagnosed with an MI. Underweight and obese participants were more likely to exhibit symptoms of anxiety than those of normal weight and overweight.

Cronbach α for the primary outcome variable, CDS was 0.86 and 0.92 (DASS-depression), 0.82 (DASS-anxiety), and 0.89 (DASS-stress) for the DASS subscales, respectively, indicating high consistency for the relevant psychometric scales. This indicates all scales exhibit acceptable internal consistency with little likelihood of item redundancy. Inter-item correlations between the 26 items of the CDS and the total CDS scores ranged from 0.08 to 0.58, and all correlations were statistically significant at the 0.01 level. Cronbach α for the other scales used in the study was: 0.78 for PHQ-15, 0.86 for PTSD-PCL-S, 0.82 for ESSI and 0.88 for RS-14.