Depression is a prevalent comorbidity in patients with heart failure (HF) and is associated with impaired quality of life and increased risk of hospitalization and mortality (1–3). Most previous studies on depression in HF have focused on patients with chronic HF with reduced ejection fraction (HFrEF) from Western and other developed countries (3, 4). Following hospitalization for acute HF, patients are at high risk of readmission and mortality (5, 6). Approximately 30% of patients are re-hospitalized within 2–3 months of discharge, and mortality can reach 15% during this period (7). High rates of events are mainly driven by a subgroup of high-risk patients (8, 9). It is necessary to stratify patients to identify those at high risk of adverse events and potentially improve outcomes. Screening for depressive symptoms during the early post-discharge period can be easily achieved using self-reported questionnaires over the phone in minutes (10). However, the prognostic significance of depression during the early post-discharge period in patients recently hospitalized with acute HF is unknown. In addition, nearly half of patients with HF have a normal left ventricular ejection fraction (LVEF) (11). Nevertheless, data on the associations between depression and HF with preserved ejection fraction (HFpEF) are limited (4). There are substantial differences in the clinical spectrum and pathophysiological mechanisms between HFpEF and HFrEF (11). However, it is unclear whether there is an interaction between LVEF and depression on clinical outcomes.

The prevalence of depression in the general population varies significantly between countries and regions. In the World Mental Health Survey conducted by the World Health Organization, the lifetime prevalence of any mood disorder in the United States was approximately six times higher than in China (21.4 vs. 3.6%, respectively) (12). Similarly, two recent large-scale cross-sectional studies (13, 14) also showed a 5-fold difference in the lifetime prevalence of major depressive disorder between the United States and China (20.6 vs. 3.9%, respectively). A cross-sectional study (15) performed in a Chinese community-dwelling population reported that 12.0% of women and 7.7% of men with chronic HF had depression. To our knowledge, no large-scale prospective studies have been conducted on depression in patients with HF in China. Therefore, we investigated the prevalence, predictors, and prognostic value of post-discharge depression in a prospective cohort of patients hospitalized with acute HF in China.

In this large-scale prospective study of patients recently hospitalized with acute HF, 12.3% of patients had depression in the early post-discharge period. Most were undiagnosed and untreated. Female sex, older age, lower BMI, lower serum sodium concentration, worse HF severity, more comorbidities, and not using RAS inhibitors, beta-blockers, or statins at discharge were associated with depression. Post-discharge depression in patients recently hospitalized with acute HF was significantly and independently associated with an increased risk of all-cause mortality and the composite of death or HF rehospitalization. The association between post-discharge depression and adverse events was significantly stronger in relatively younger patients and in patients with HFpEF.

Previous studies in Western countries reported that the prevalence of depression in patients with HF was 20–30%, approximately three times the prevalence of major depressive disorder in the general population (2, 14). Depression was also more common in patients with a higher NYHA class, and inpatients were more likely to be depressed than outpatients (2). In this study, we assessed depressive symptoms during the early post-discharge period. This inevitably excluded patients who died before the first follow-up and those lost to the first follow-up. Discharged patients tended to have improved symptoms after in-hospital therapy. The timing of depression assessment and patient selection bias may partly explain the relatively lower prevalence in our study. Of 480 patients with depression, only seven had a history of mental illness, and only four had been treated with antidepressants, indicating an extreme lack of awareness and treatment of depression in patients with HF in China. The under-diagnosis and under-treatment of depression in patients with cardiovascular disease have been reported previously. One study showed that only 11% of patients with depression received appropriate treatment (21). The socioeconomic status (22) and cultural characteristics (e.g., more stigma toward mental illness (23)) of China may have exacerbated this problem. In the China Mental Health Survey, only 0.5% of individuals with major depressive disorder were treated adequately (24). Depression severely decreases the quality of life and increases morbidity and mortality (25). It was estimated that there were 13.7 million patients with HF in China (26), suggesting a large number of patients with undiagnosed depression that warrants attention.

The reasons why depression is prevalent in patients with HF and correlates with poor prognosis are complex and multifactorial (4). Several shared pathological mechanisms between HF and depression have been proposed, including higher levels of inflammatory markers, abnormal activation of the hypothalamic-pituitary-adrenal axis, autonomic dysfunction, and lifestyle factors (4, 27). Consistent with a previous study of chronic HF (4), our study showed that female sex, worse HF severity (higher NYHA class and NT-proBNP levels), and more comorbidities were significant predictors of post-discharge depression in univariate analysis. Depression is both a contributing factor and a consequence of the exacerbation of HF (4). The higher symptom burden, poorer quality of life, and increased activation of shared pathophysiological pathways between depression and HF may account for the higher rate of depression in patients with more severe HF (28). Previous studies reported that higher BMI and younger age were associated with depression in patients with chronic HF (29, 30) and in the general population (14, 31), whereas our findings were the opposite. These differences may partly be explained by the worse HF severity in patients with lower BMI and older age in this study, as they also had a much higher mortality rate during follow-up (Supplementary Material). Theoretically, patients with stable and chronic HF may be more similar to the general population (31). Conversely, in patients recently hospitalized with acute HF, secondary depression due to severe somatic disorders (32) may surpass the effects of a higher BMI and younger age. In this study, COPD and diabetes mellitus were independent predictors of post-discharge depression. These comorbidities may further increase the symptom burden and reduce the quality of life. Patients with COPD (33) and diabetes mellitus (34) tend to have more chronic inflammation, which may be an important mechanism for the increased risk of post-discharge depression. The use of RAS inhibitor and beta-blocker at discharge was independently associated with less post-discharge depression after adjustment. Previous studies have reported that abnormal RAS activation and increased sympathetic tone may contribute to the pathophysiological mechanism of depression in patients with HF (35, 36). Our findings support this hypothesis.

The early post-discharge period after hospitalization for acute HF carries a high risk of adverse events, and post-discharge follow-up during this period is recommended by the current European Society of Cardiology guidelines (7, 11). Our findings highlight the importance of screening for depressive symptoms in this vulnerable phase, and monitoring has significant and independent value for risk stratification. There is currently no clear consensus regarding the treatment of depression in patients with HF (11). Whether depression is a treatable target in HF or serves as a marker of HF severity remains to be determined. Potential interventions for patients with HF and post-discharge depression include treatment for HF and treatment for depression. Depressed patients tend to have lower adherence to HF treatment and poorer self-management and may require more intensive monitoring by clinicians (4). Because of the higher risk of adverse events in patients with HF and depression, a more aggressive treatment strategy for HF may provide a greater benefit. Selective serotonin reuptake inhibitors (SSRI) are widely used as antidepressants (37). However, SSRI failed to improve depression or the prognosis of HF in patients with HFrEF in two randomized controlled trials (38, 39). In the MOOD-HF trial, although escitalopram did not improve depressive symptoms more significantly than placebo, patients in both study arms showed a substantial improvement in depressive symptoms compared with baseline after nurse-coordinated disease management and optimization of HF drug therapy (38). This suggests that optimal treatment of HF may be the key to reducing depressive symptoms. Other studies have shown that exercise training (40), cognitive-behavioral therapy (41), telemedical interventional monitoring (42), and palliative care (43) improve depressive symptoms in patients with HF. These additional interventions may be beneficial for depressive symptoms in post-discharge patients with depression after the optimization of guideline-directed medical therapy for HF. However, the clinical significance of the reduction in depressive symptoms requires further research.

Previous studies on depression in patients with HFpEF are limited. This study is the first to report an interaction between post-discharge depression and in-hospital LVEF on mortality in patients recently hospitalized with acute HF. Our results were inconsistent with those of a recent post hoc analysis of the TOPCAT trial, in which no significant association was found between baseline depression and mortality in patients with chronic HFpEF after a mean follow-up of 3.3 years (44). In the TOPCAT trial, patients with a life expectancy of < 3 years were excluded. Two-thirds of patients had NYHA class I or II symptoms, indicating that the enrolled patients were more likely to have relatively stable HFpEF at baseline (45). Conversely, patients in the HERO study had NYHA class III or IV symptoms at enrolment. Worse HF severity, higher rates of adverse events, shorter time intervals between baseline evaluation of depressive symptoms and the occurrence of events, and more HF-related depression may explain the significant associations in the HERO study. In the TOPCAT trial, worsening depressive symptoms at 12 months were associated with an increased risk of all-cause mortality (HR 1.82 [95% CI: 1.13–2.93) (30), suggesting that depression may be a marker of the deterioration of chronic HFpEF. Our study extended the findings of the TOPCAT trial and indicated that, in patients who have recently experienced acute decompensated HF, depression is a stronger predictor of mortality in HFpEF than in HFrEF.

There are several limitations to this study. First, although PHQ-9 is an effective tool with high sensitivity and specificity, it remains a self-reported questionnaire, not diagnostic criteria. PHQ-9 was administered by trained nurses over the phone, not strictly self-administered. Some patients may have been afraid or unwilling to share their poor mental state when answering questions posed by nurses. Second, baseline characteristics were collected during hospitalization in the HERO study, while depressive symptoms were assessed at the first follow-up. The timing discrepancy may have introduced potential confounders and influenced our results. Because depressive symptoms were assessed at the first follow-up, only patients who (i) were discharged, (ii) did not die early after discharge, (iii) had not been rehospitalized before the first follow-up, and (iv) were prepared to answer the PHQ-9 were included. Therefore, selection bias may have influenced the results. Third, the severity of depressive symptoms in patients with HF may have fluctuated. Some studies showed that the improvement or worsening of depression over time was associated with HF outcomes (30, 40). In the HERO study, we only assessed baseline depressive symptoms; therefore, we could not evaluate the associations between dynamic changes in depressive symptoms and clinical outcomes. Finally, the enrolment of patients in the HERO registry was limited to Henan Province, China, with a population of 100 million. Further validation is required in more diverse populations.

Post-discharge depression is common in patients recently hospitalized with acute HF and is associated with adverse events. Unfortunately, depression in these patients often goes undiagnosed and untreated. Therefore, clinicians should pay more attention to the post-discharge mental health status of patients with HF. Because traditional antidepressants failed to improve depressive symptoms more significantly than placebo, further studies are needed to determine what can be done to relieve depressive symptoms and improve HF outcomes. Most previous studies on the treatment of depression in HF have focused on patients with HFrEF. Our study suggests that the association between depression and adverse events may be even stronger in patients with HFpEF than in those with HFrEF; therefore, future studies involving the treatment of depression in patients with HFpEF are necessary.

Post-discharge depression in patients recently hospitalized with acute HF is associated with an increased risk of adverse events, regardless of LVEF. Screening for depressive symptoms during the early post-discharge period may help to better identify high-risk patients and tailor patient management. Further studies are needed to determine how regular depression screening can help improve patient management and clinical outcomes.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author/s.

The studies involving human participants were reviewed and approved by the Ethics Committee on Scientific Research and Clinical Trials at the First Affiliated Hospital of Zhengzhou University (in September 2017; approval no. 2014SY-079). Written informed consent to participate in this study was provided by the patients.

JD, XD, and JL contributed to the conception and design of the study. XW, RL, XZ, LinL, and LiL participated in the collection of samples and data. RL and YL performed the statistical analysis and all the results were checked by CJ and LiL. JL wrote the first draft of the manuscript. LinL, YL, XZ, XW, and CJ wrote sections of the manuscript. CM, XD, and JD were critically revised the manuscript. All authors contributed to the article and approved the submitted version.