Comorbidity of depression and anxiety leads to a poor prognosis following angina pectoris patients: a prospective study

In this prospective study, we investigated the association between depression, anxiety, comorbidity and the occurrence of hospital readmission and recurrent CV events during a 29-month follow-up period among AP patients. Patients were admitted to Guangdong Provincial People’s Hospital from October 2017 to January 2018 for a cross-sectional study and were followed up with every year to observe their prognosis after discharge. Professional cardiologists in the hospital persistently incorporated 705 patients with suspected CHD into the study. All patients underwent examinations and therapies considered to be the most appropriate by the attending cardiologist. A total of 705 copies of the questionnaires were issued, and all the completed copies were retrieved. A well-trained psycho-cardiologist explained the psychiatric scales to each patient 1 day before the coronary angiography and assisted those with impaired vision or poor reading ability to understand the contents of the scale. The study was approved by the Medical Ethics Committee of Guangdong Provincial People’s Hospital. Written informed consent was obtained from all participants. Details about our recruitment strategy and the excluded patients were described in our previous article [21]. A total of 443 individuals with AP (according to the Braunwald standard) and diagnosed CHD were included for analysis.

Follow-up assessments were performed by telephone at 1 year and over 2 years after discharges. The average follow-up time was 26.33 ± 0.93 months. Endpoint events included all-cause mortality, cardiac and noncardiac readmission, nonfatal stroke, revascularisation and the recurrence of nonfatal myocardial infarction. Noncardiac readmission refers to noncardiac diseases that require the patient to be readmitted for further treatment. MACEs consisted of death, nonfatal stroke, cardiac readmission, revascularisation and the recurrence of nonfatal myocardial infarction. Moreover, the composite event covered all above events. Due to the insufficient samples of death, nonfatal stroke, revascularisation and recurrence of nonfatal myocardial infarction, we did not analyse these clinical events individually.

Patient health questionnaire-9 (PHQ-9) and generalised anxiety disorder-7 (GAD-7) scales were used to assess depression and anxiety at baseline, respectively. We thoroughly elaborated on the utility of these scales in our previous article [21]. In our main analysis, a score of 5 and 10 was used as the mood symptom and clinical mood disorder thresholds. PHQ-9 has been demonstrated to be a dependable measure to evaluate depression severity with mild, moderate, moderately severe to severe depression corresponding to a score of 5, 10 and 15, respectively [22]. And even mild depression symptoms, as classified by PHQ-9, are associated with relatively poor prognoses among cardiac patients [23]. The cut-off score of ≥10 for the detection of clinical depression has been shown to have 89% sensitivity and 89% specificity [24]. GAD-7 score ranges from 0 to 21 with a score of 0–4, 5–9, 10–14, 15–21 representing normal, mild, moderate and severe levels of anxiety, respectively. According to the widely use of the cut-off value of 5, we compared the differences between patients with GAD-7 scores of < 5 and ≥ 5. Moreover, a GAD-7 score of ≥10 indicated that the patient has clinical anxiety, in which the sensitivity and specificity of this diagnostic criterion attained 89 and 82%, respectively [25, 26]. Former studies have shown the underlying structure of GAD-7 to be unidimensional [27]. The Chinese versions of PHQ-9 and GAD-7 have been validated among Chinese cardiac patients [28, 29].

As depression and anxiety are two aspects of mental state of patients and the moderately strong inter correlation between them, we tried to use a more direct way to demonstrate the severity of comprehensive mood symptoms. We screened out patients with PHQ-9 scores of ≥5 and GAD-7 scores of ≥5 into the comorbidity subgroup, and patients with both scores ≥10 were placed in the clinical comorbidity subgroup. The combination of PHQ-9 and GAD-7 scales has been merged as a composite measure of depression and anxiety as PHQ-ADS scale [30]. And PHQ-ADS cutpoints of 10, 20, and 30 indicated mild, moderate, and severe levels of depression/anxiety, respectively. The reliability and validity of this combination and cutpoints have been examined and testified in latest studies [30‐32].

We compared the characteristics of patients with different depression and anxiety symptoms. The normality test was performed on continuous variables. In this study, the continuous values of PHQ-9 and GAD-7 scores were non-normal distributions. Clinical characteristics of patients were presented by questionnaire scores of < 5, 5–9, and ≥ 10.

A total of 443 patients were enrolled in the study and assessed using PHQ-9 and GAD-7 scales. The baseline characteristics of the participants are shown in Table 1. The average participant was 63.9 years old (± 9.8 years), and there were 337 (76.1%) males and 106 females (23.9%). Among all the AP patients, there were 172 (38.9%) with depression symptom (PHQ-9 scores of ≥5), of which 49 (11.1%) had clinical depression (PHQ-9 scores of ≥10). As for anxiety symptoms, 103 (23.3%), 13 (2.9%) and 11 (2.5%) patients had mild, moderate and severe anxiety symptoms, respectively. Furthermore, 71 participants suffered from the comorbidity of depression and anxiety, among them, 52 (73.2%) comprised the comorbidity subgroup and 19 (26.8%) comprised clinical comorbidity subgroup. The correlation of depression and anxiety were shown in our previous article [21].

All 443 patients received follow-up calls for readmissions, MACEs, and composite events. The censored data contained loss to follow-up and endpoint events that had not occurred at the end of the study. In this study, there were 74 cardiac readmissions, 46 noncardiac readmissions, 88 MACEs and 129 composite events during the 29-month follow-up period (Fig. 1). Five patients suffered from both noncardiac readmissions and MACEs. The incidence of follow-up events in different groups is shown in Table 1.

We initially analysed prognosis of patients with PHQ-9 and GAD-7 scores of ≥5 (see Additional file 1: Table S1). Unadjusted HRs were estimated from the Kaplan–Meier survival curve and Log-Rank tests for patients without mood symptoms and with mood symptoms. Between the nondepressed and depressed group (PHQ-9 scores of ≥5), a significant difference of noncardiac readmission ([HR] 2.16, 95% [CI] 1.19–3.93, p = 0.011) was observed. However, depression symptoms were found to be not significantly associated with cardiac readmissions ([HR] 1.03, 95% [CI] 0.64–1.65, p = 0.908) and MACEs ([HR] 1.24, 95% [CI] 0.82–1.90, p = 0.338). Moreover, we did not observe a significant relationship between anxiety symptoms (GAD-7 scores of ≥5) and prognosis. (see Additional file 1: Table S1) As for the comorbidity subgroup, we found that coexistence of depression and anxiety may lead to more noncardiac readmission ([HR] 2.28, 95% [CI] 1.05–4.94, p = 0.036). A Cox proportional regression model was established for multi-factors analyses. The multi-factors included age, gender, education and severity of coronary artery stenosis, which age, gender and education were statistically significant in univariate analysis and severity of coronary artery stenosis was an important clinical indicator. Compared to patients without mood symptoms, patients with depression symptoms (PHQ-9 scores of ≥5) didn’t show a high risk of readmission ([HR] 1.81, 95% [CI] 0.55–5.96, p = 0.333). Meanwhile, after adjustment, we did not observe significant effects of the comorbidity (both PHQ-9 and GAD-7 scores of ≥5) on the noncardiac readmission. However, there was a tendency that patients with comorbidity may had more risks than patients with individual mood symptom on the follow-up events (see Additional file 1: Table S1).

Moreover, we analysed clinical mood disorders with PHQ-9 and GAD-7 scores of ≥10 (see Table 2). Compared to patients without mood disorders, patients with clinical depression (PHQ-9 scores of ≥10) had a greater risk of noncardiac readmission ([HR] 2.81, 95% [CI] 1.30–6.10, p = 0.009). At the same time, clinical anxiety (GAD-7 scores of ≥10) were found to have a strong significance with noncardiac readmission ([HR] 3.46, 95% [CI] 1.37–8.72, p = 0.008) and composite event ([HR] 2.19, 95% [CI] 1.23–3.91, p = 0.008) while MACEs were found to have a positive correlation tendency in this group ([HR] 1.99, 95% [CI] 0.97–4.07, p = 0.061). As for the clinical comorbidity subgroup (both PHQ-9 and GAD-7 scores of ≥10), we performed Kaplan–Meier (KM) curves of endpoint events with number of subjects at risk (Fig. 2). The unadjusted analysis showed that clinical comorbidity presented great significance with occurrence of noncardiac readmission ([HR] 3.74, 95% [CI] 1.39–10.10, p = 0.009), MACEs ([HR] 2.27, 95% [CI] 1.07–4.81, p = 0.033) and composite event ([HR] 2.44, 95% [CI] 1.31–4.52, p = 0.005). After controlling for age, gender, marriage and severity of coronary artery stenosis, participants with clinical depression were found to be at a greater risk of noncardiac readmission ([HR]2.38, 95% [CI] 1.06–5.33, p = 0.035) and composite event ([HR]1.73, 95% [CI] 1.04–2.87, p = 0.036) than normal samples. Meanwhile, clinical anxiety participants reported a significant risk of more MACEs ([HR] 2.23, 95% [CI] 1.07–4.64, p = 0.032) and composite event ([HR]2.30, 95% [CI] 1.27–4.15, p = 0.006) than normal participants after multi-factors adjustment. In addition, clinical comorbidity was found to be a great predictor of noncardiac readmission ([HR] 2.91, 95% [CI] 1.03–8.18, p = 0.043), MACEs ([HR] 2.38, 95% [CI] 1.11–5.10, p = 0.025) and composite events ([HR] 2. 52, 95% [CI] 1.35–4.69, p = 0.004) (see Table 2).

In our study, patients with clinical depression symptoms (PHQ-9 score of ≥10) were found to have a 2.38 times higher risk of noncardiac readmission than those who did not suffer from depression. Pederson et al. [33] showed that depression might be an unrecognised, and possibly changeable, independent risk factor for unexpected readmission or premature death. Depression is difficult to detect in an acute care environment and is often undertreated. Patients with moderate to severe depression symptoms at discharge with no history of depression may face a twofold increase in the risk of short-term hospital readmission or death. Mitchell et al. [34] reported that the risk of hospital readmission or emergency department treatment within 30 days after discharge among patients with PHQ-9 scores of ≥5 was increased by 73%. Cancino et al. [35] showed a 49% increase in readmission within 30 days of discharge for patients with mild depression and a 96% increase for patients with moderate to severe depression. Moraska et al. [36] found that high clinical depression scores on the PHQ-9 were associated with increased risk of hospitalisation (adjusted HR: 1.79, 95% CI: 1.30–2.47). In a small study of 144 patients, Kartha et al. [37] illustrated that patients who were diagnosed with severe depression using a standardised scoring algorithm based on PHQ-9 were three times more likely than others to be rehospitalised within 90 days of discharge. We used PHQ-9 scores of < 5, 5–9 and ≥ 10 as the thresholds for non-depression, mild depression and moderate to severe depression, respectively. We also found that both mild and moderate to severe depression symptoms were associated with noncardiac readmission and composite events.

Previous literature has shown that CHD combined with depression symptoms will lead to an increase in mortality and rehospitalisation, thereby aggravating the severity of the disease and medical expenses and wasting resources. A number of studies [38] have shown that the comorbidity of mental and physical illnesses can increase the severity of disease, frequencies of readmissions and medical expenses. Even the presence of isolated mood symptoms might increase one’s risk of having medical comorbidities [39]. People with mood symptoms but no mood disorder diagnosis were found to be prone to have more lifelong diseases and health problems and longer hospital stays than those without mood diagnoses or symptoms [40]. Our study has shown that both mild depression and moderate to severe depression might be risk factors for hospital readmission.

Former studies have shown that anxiety is a significant risk factor for recurrent CV events and mortality [3, 13, 41‐43]. Although we did not analyse death, revascularisation or nonfatal myocardial infarction independently due to a lack of valid samples, we summed up cardiac readmission, nonfatal stroke and all the above as MACEs. Tully et al. [44] found that in patients with confirmed CHD, anxiety was a prognostic risk of subsequent MACEs, such as myocardial infarction, left ventricular failure and stroke. Another study showed that generalised anxiety disorder increased the MACE risk of acute coronary syndrome outpatients by nearly two times within 2 years [45]. Similar to previous researches, we identified an association between clinical anxiety and poor CVD prognoses. Although there have been few studies in which the authors analysed the relationship between anxiety and noncardiac readmission, we found a positive correlation between clinical anxiety and noncardiac readmission. We did not observe this association in the anxiety symptom group (GAD-7 score of ≥5), possibly because the anxiety symptom is transient and usually triggered by a perceived threat, and once the threat diminishes, the anxiety will diminish too. It has been proven that anxiety is prone to reduce the quality of life of patients with CHD, especially after an acute coronary syndrome event, and that anxious patients have more disabilities and somatic symptoms than others [46‐48].

The authors of many related studies analysed depression or anxiety as single risk indicators, but we studied them as a comorbid symptom. We discovered that most clinical anxiety patients also have depression symptoms. Denollet and Strik et al. [41, 49] reported that a mixed emotional state consisting of anxiety and depression symptoms, not just depression, was the most common manifestation among patients with myocardial infarction [49]. The author, thus, concluded that anxiety was a common factor of depression after myocardial infarction. A similar symptom overlap was found by Frasure-Smith [45]. Some authors [50, 51] demonstrated that this connection was produced by the same dysfunctional biology or that depression and anxiety might be derived from parallel genetic dispositions. Furthermore, Scott et al. showed that comorbid depression-anxiety disorder increased the risk of a series of physical symptoms occurring at the same time [52]. In a follow-up study of women with suspected myocardial ischemia, depression or anxiety could be used to predict CVD events and death more accurately than the independent symptoms [53]. Phillip et al. [54] showed that even after adjustment, the strongest association between CVD and death from all causes appeared in major depressive disorder and generalised anxiety disorder comorbidity. Similarly, we identified a greater risk of noncardiac readmission, MACEs, and composite events among participants with both depression and anxiety.

First, due to the small sample size and prognostic events, some statistical analyses in this study may suffer from overfitting. Second, depression and anxiety are strongly correlated and are two aspects of mental state of patients. It would be not all-sided enough to analyse them individually. While the combined PHQ-9 ≥ 5 & GAD-7 ≥ 5 and PHQ-9 ≥ 10 & GAD-7 ≥ 10 has not been strictly validated, there were some researches have already tried to testify this combination [55, 56]. Although we did not carry out professional psychiatric interviews to confirm the diagnosis of mood symptoms, we used an easy-to-deploy screening tool that has been fully validated. Moreover, our research was conducted in only one hospital specialised in cardiac care, which ensured the consistency of examination tools, methods and follow-up times and procedures but limited the generalizability of the findings.