Authors: Qiuge Zhao, Xiaofei Sun, Yanting Zhang, Yuzhen Zhang, Cancan Chen
Categories: Research, Heart failure, Anxiety symptoms, Depressive symptoms, Network analysis
Source: BMC Psychiatry
Authors: Qiuge Zhao, Xiaofei Sun, Yanting Zhang, Yuzhen Zhang, Cancan Chen
Anxiety and depressive symptoms are common among patients with heart failure (HF). Physical limitations, lifestyle changes, and uncertainties related to HF can result in the development or exacerbating of anxiety and depressive symptoms. However, the central and bridge symptoms of anxiety and depressive symptoms network among patients with HF remain unclear. Network analysis is a statistical method that can discover and visualize complex relationships between multiple variables. This study aimed to establish a network of anxiety and depressive symptoms and identify the central and bridge symptoms in this network among patients with HF.
This study employed a cross-sectional study design and convenience sampling to recruit patients with HF. This study followed the Helsinki Declaration and was approved by the Research Ethics Committee of Hospital. The Generalized Anxiety Disorder-7 (GAD-7) and Patient Health Questionnaire (PHQ-9) were administered to evaluate anxiety and depressive symptoms among patients with HF, respectively. Network analysis of anxiety and depressive symptoms was performed using R.
In the anxiety and depressive symptoms network, PHQ2 (feeling down, depressed, or hopeless), PHQ7 (inability to concentrate), and GAD4 (difficulty relaxing) were the most central symptoms. Anxiety and depressive symptoms were linked by PHQ2 (feeling down, depressed, or hopeless), GAD6 (becoming easily annoyed or impatient), GAD5 (unable to sit still because of anxiety), GAD7 (feeling afraid that something terrible is about to happen), and PHQ6 (feeling bad or like a failure, or disappointing oneself or family).
This study identified the central and bridge symptoms in a network of anxiety and depressive symptoms. Targeting these symptoms can contribute to interventions for patients with HF at risk of—or suffering from—anxiety and depressive symptoms, which can be effective in reducing the comorbidity of anxiety and depression.
Heart failure (HF)—a complex clinical syndrome—affects approximately 64.3 million people worldwide [1]. In China, about 8.9 million patients suffer from HF [2]. Owing to the aging population and improved survival rates of coronary artery disease, HF incidence is expected to double in the next 40 years, and healthcare costs will continue increasing [3]. Despite the progress in HF treatment, it is still associated with a high incidence, and the prognosis of patients with HF is generally poor, such as decreased functional status and quality of life, repeated hospitalization, and high mortality [4, 5]. Physical symptoms—including difficulty breathing, edema, and fatigue —may occur among patients with HF owing to insufficient cardiac output caused by abnormal cardiac structure and/or function [6]. Patients with HF, in addition to HF’s well-known physical symptoms, also exhibit psychological symptoms, particularly anxiety and depressive symptoms [7, 8].
Numerous patients with HF feel anxious owing to fear and concern about their condition. Anxiety can manifest as excessive concern for one’s own health, fear of symptoms or complications, or feeling uneasy and restless [7]. Owing to HF’s prolonged and recurrent nature, patients experience under both physical and psychological pressure, which can precipitate anxiety symptoms. Anxiety symptoms are common among patients with HF, with a prevalence rate of 20–50% [9]. Severe anxiety symptoms are the primary factors preventing changes in lifestyle, diet, and medication adherence, increasing readmission frequency and decreasing quality of life [10, 11]. Moreover, anxiety may precipitate breathing difficulties, panic, and chest pain, which can further exacerbate HF symptoms, result in hospitalization, and increase preventable healthcare utilization [12]. Considering anxiety’s adverse effects, paying attention to the anxiety symptoms of patients with HF is necessary.
Depressive symptoms are another common psychological problem among patients with HF. The characteristics of depressive symptoms are sadness, loss of interest or happiness, low self-worth, guilt, sleep or appetite disorders, fatigue, and lack of concentration [13]. Previous studies have demonstrated that the incidence of depression in patients with HF was five times higher than that among the general population [14], with 41.9% of patients with HF suffering from depression and 28.1% suffering from moderate to severe depression [15]. Depression can trigger neurohormonal activation by increasing the activity of the hypothalamic-pituitary-adrenal axis, result in hypercortisolemia, hypertension, arrhythmia, hypercoagulable state, and cytokine release, which may negatively impact the prognosis of HF [16, 17]. Accumulating evidence has suggested that depression is an important risk factor for recurrent hospitalization and increased mortality among patients with HF [18]. Compared to patients with HF with—than those without—depressive symptoms are 1.45 times more likely to be hospitalized [19] and have a 1.40 times higher mortality risk [20]. Additionally, depressive symptoms may result in reduced treatment compliance, poor functional status [21], difficulties in self-care, and decreased quality of life among patients with HF [22].
Anxiety and depression are common among patients with HF and are related to disease progression [7]. Suffering from both anxiety and depressive symptoms can negatively impact daily activities and physical condition among patients with HF [23]. Moreover, a close relationship exists between anxiety and depression, implying that people with one disease are at a higher risk of developing another [24]. A previous study found that the activation of one symptom can quickly spread to other symptoms, precipitating more chronic symptoms over time [25]. Gerymski et al. [26] found that anxiety and depression are associated with reduced sexual satisfaction in female patients with HF using network analysis. Wang et al. [27] found significant correlations between anxiety and depression, subjective and objective social support in the symptom network of patients with chronic HF. However, the key symptoms of depression and anxiety among patients with HF—and whether these central symptoms affect other symptoms—remain unclear.
Network analysis is a statistical method that can discover and visualize complex relationships between multiple variables and identify potential influencing factors [28]. An important goal of network analysis methods is to identify central symptoms and bridge symptoms in a network and analyze the mechanisms affecting their connectivity [29]. A network diagram is a visual representation of a system comprising nodes and edges. Nodes represent variables, and edges represent connections between nodes [30]. Centrality indices are utilized to identify the most important nodes in a network, including strength, closeness, and betweenness. Strength centrality determines the strength of the relationship between a node and its neighboring nodes. Closeness centrality identifies the average distance between a node and other nodes. Betweenness centrality identifies the frequency at which a node is connected between two nodes [31]. Network analysis has been widely used to explore depression and anxiety symptoms in different populations, such as adolescents [32], chronic disease patients [33], and older adults [34]. Owing to age [32], gender [35], and cultural differences [36], differences prevail in the strength, closeness, and betweenness of anxiety and depressive symptoms; central symptoms also vary among different groups. Accurately identifying the central anxiety and depressive symptoms among patients with HF and their associations can provide greater information for psychological care.
Therefore, this study used network analysis to explore the relationship between anxiety and depressive symptoms and identify central and bridge symptoms among patients with HF. We assume that the network model of anxiety and depressive symptoms has some key symptoms affecting other symptoms. The network model of anxiety and depressive symptoms is closely linked by several central symptoms.
This multicenter cross-sectional study was conducted from March to December 2023 in four tertiary hospitals in China. The inclusion criteria were as Participants must have an HF diagnosis according to the international Guidelines for the Diagnosis and Treatment of HF, be 18 years old and above, be able to complete the questionnaire in writing or orally, and provide informed consent and voluntarily participate in this study. The exclusion criteria were as Participants have an acute myocardial infarction attack within the past month, and exhibit other serious complications, such as malignant tumors, liver failure, kidney failure, lung failure, and mental disorders.
This study followed the Helsinki Declaration and was approved by the Research Ethics Committee of Hospital. The researchers who received training before the study’s commencement utilized a unified guiding language face-to-face to introduce this study’s purpose and content to the participants. Per the sample size requirements in network analysis,^37^ the sample size should be greater than the total number of parameters (including threshold parameters and paired related parameters). The threshold parameter equals to the number of nodes, and the paired related parameters are equal to (total number of nodes) × (total number of nodes − 1)/2. In this study, 16 nodes need to be constructed; hence the threshold parameter is 16, and the paired related parameters are (16 × 15) /2 = 120. Therefore, the minimum sample size required is 136. This study’s sample size is 277, which fulfills the requirements.
The patients’ demographic characteristics included age, gender, educational level, and marital and work status. The clinical characteristics included left ventricular ejection fraction (LVEF), body mass index (BMI), HF duration, New York Heart Association (NYHA) class, and comorbidities. Demographic characteristics were self-reported by patients, whereas clinical characteristics were obtained from their medical records.
This study utilized the Generalized Anxiety Disorder-7 (GAD-7) scale to measure anxiety levels among patients with HF over the past two weeks. This scale— developed by Spitzer et al. [38]—comprise seven items. Each item is scored on a scale of 0 to 3 (0 = completely absent, 1 = some days, 2 = more than seven days, 3 = almost every day), and the total score is obtained by adding all items. The higher the total score, the more severe the anxiety. The GAD-7 has acceptable reliability and validity among patients with cardiovascular disease in China, with a score greater than 10 indicating that patients have anxiety symptoms. In this study, Cronbach’s α of the GAD-7 was 0.898.
The Patient Health Questionnaire (PHQ-9) [39] was used to measure depressive symptoms among patients with HF in the past two weeks. This scale is based on the nine criteria for diagnosing depression in the Diagnostic and Statistical Manual of Mental Disorders IV (DSM-IV). Each item is scored on a scale of 0–3 (0 = completely absent, 1 = some days, 2 = more than seven days, 3 = almost every day), with a total score of 0–27. Higher scores indicate more severe depression. A total score greater than 10 indicates depressive symptoms. The Chinese version of PHQ-9 has been found to have good psychometric characteristics. [40] In this study, Cronbach’s α of the PHQ-9 was 0.807.
Descriptive analysis of continuous and categorical variables and the calculation of Cronbach’s α coefficients of the scales were conducted using SPSS 25.0. Network analysis of anxiety and depressive symptoms was performed using the R software (version 4.02).
The data used for network analysis were all continuous variables; therefore, a Gaussian graph model (GGM) was employed. Using the graphical least absolute shrinkage and selection operator (GLASSO) [41], weak edges were removed to reduce false correlations, resulting in a concise network. Sixteen nodes were present in the drawn network. The thicker and darker the edges between nodes, the stronger the correlation between the two nodes. The green and red edges represented positive and negative correlations, respectively. This study conducted three main analyses —namely, network estimation, centrality indices analysis, and stability estimation. To evaluate the importance of each node (symptom), three key centrality indices were calculated—namely, strength, closeness, and betweenness. The bridge centrality index was calculated to identify bridge symptoms of anxiety and depressive symptoms; its accuracy was evaluated by calculating the marginal weights’ 95% confidence interval (CI). Additionally, a stability assessment was conducted by calculating the stability coefficients related to strength centrality using the case study’s discarded subset bootstrap (2,500 bootstrap samples). The stability coefficient should be greater than 0.5, and at least greater than 0.25. If it is 0.7 or greater, it is optimal [37].
Overall, this study included 277 patients with HF, comprising 165 men (59.6%) and 112 women (40.4%); their mean age was 63.60 ± 14.09 years. Most patients had a junior high school education or lower (70.4%), were married (90.6%), and were unemployed (75.8%). Their mean LVEF was 43.35 ± 14.36%. More than half of the patients had an HF duration of over six months (70.0%), were classified as NYHA class III (54.1%), and had three or more comorbidities (66.4%). Table 1 presents the sample characteristics.
The mean scores of anxiety and depressive symptoms were 4.01 ± 3.94 and 5.18 ± 3.72, respectively. Furthermore, 30 patients (10.8%) scored ≥ 10 on the GAD-7 and were identified as having anxiety symptoms, while 39 patients (14.1%) score ≥ 10 on the PHQ-9 and were identified as having depressive symptoms.
Table 1Sample characteristics (n = 277)Characteristicsn (%)/mean ± SDAge (years)63.60 ± 14.09Gender Men Women165 (59.6%)112 (40.4%)Education Lower than high school195 (70.4%) High school or above82 (29.6%)Marital status Single/divorced/widow26 (9.4%) Married251 (90.6%)Work status Unemployed210 (75.8%) Employed67 (24.2%)LVEF (%)43.35 ± 14.36BMI (kg/m^2^)23.76 ± 4.56HF Duration < 6 months83 (30.0%) ≥ 6 months194 (70.0%)NYHA class II72 (26.0%) III150 (54.1%) IV55 (19.9%)Comorbidity < 389 (32.1%) ≥ 3184 (66.4%)Unreported4 (1.5%)Note LVEF: left ventricular ejection fraction; BMI: body mass index; HF: heart failure; NYHA: New York Heart Association; SD: standard deviation
Figure 1 presents the network structure of anxiety and depressive symptoms among patients with HF. In the network of anxiety symptoms, the strongest edges (in descending order) were between GAD2 and GAD3 (edge weight = 0.453), between GAD4 and GAD5 (edge weight = 0.424), between GAD1 and GAD2 (edge weight = 0.267), between GAD3 and GAD7 (edge weight = 0.228), between GAD1 and GAD4 (edge weight = 0.155), between GAD2 and GAD7 (edge weight = 0.143), and between GAD5 and GAD6 (edge weight = 0.100). In the network of depressive symptoms, the strongest edges (in descending order) were between PHQ7 and PHQ8 (edge weight = 0.328), between PHQ1 and PHQ2 (edge weight = 0.265), between PHQ4 and PHQ5 (edge weight = 0.241), between PHQ3 and PHQ4 (edge weight = 0.206), between PHQ6 and PHQ9 (edge weight = 0.185), between PHQ1 and PHQ7 (edge weight = 0.184), between PHQ6 and PHQ7 (edge weight = 0.177), between PHQ3 and PHQ5 (edge weight = 0.158), between PHQ8 and PHQ9 (edge weight = 0.147), and between PHQ1 and PHQ4 (edge weight = 0.138). In the network of anxiety and depressive symptoms, the strongest edges (in descending order) were between GAD6 and PHQ2 (edge weight = 0.208), between GAD7 and PHQ2 (edge weight = 0.172), between GAD7 and PHQ6 (edge weight = 0.143), between GAD5 and PHQ7 (edge weight = 0.131), between GAD2 and PHQ6 (edge weight = 0.130), between GAD4 and PHQ5 (edge weight = 0.115), between GAD1 and PHQ2 (edge weight = 0.102), and between GAD5 and PHQ8 (edge weight = 0.101).
In the constructed network structure, the highest strength was observed for GAD2 (strength = 1.17), GAD4 (strength = 1.16), GAD3 (strength = 1.06), and PHQ2 (strength = 1.00; Fig. 2a). The higher expected influence was GAD2 (value = 1.17), GAD4 (value = 1.16), GAD3 (value = 1.06), and PHQ2 (value = 1.00; Table 2). For closeness and betweenness (Fig. 2a), the highest (arranged in descending order) were PHQ1, PHQ7, PHQ2, and GAD4, suggesting that they are closest to other nodes and are usually located between the other two nodes in the network. Based on strength, closeness, betweenness, and expected influence, PHQ2 (feeling down, depressed, or hopeless), PHQ7 (inability to concentrate), and GAD4 (difficulty relaxing) were the most central symptoms. The nodes with the strongest bridge strength were PHQ2 (feeling down, depressed, or hopeless), GAD6 (becoming easily annoyed or impatient), GAD5 (unable to sit still because of anxiety), GAD7(feeling afraid that something terrible is about to happen), and PHQ6 (feeling bad or like a failure, or disappointing oneself or family; Fig. 2b).
The network structure’s stability was assessed by estimating the edge weights’ 95% CI. In this study, the central stability coefficient was 0.60 > 0.50, indicating that the network was stable (Fig. 3). The difference test for the edge weights indicated that some edges were statistically significance (Fig. 4).
Table 2Descriptive statistics of the GAD-7 and PHQ-9 itemsItemMeanSDExpected InfluenceGAD1 Feeling nervous, anxious, or eager0.780.730.70GAD2 Unable to stop or control worries0.710.791.17GAD3 Excessively worried about various things0.680.811.06GAD4 Difficulty relaxing0.420.621.16GAD5 Unable to sit still because of anxiety0.290.520.98GAD6 Becoming easily annoyed or impatient0.650.750.87GAD7 Feeling afraid that something terrible is about to happen0.470.730.86PHQ1 Lack of motivation or interest in doing things0.780.710.96PHQ2 Feeling down, depressed, or hopeless0.700.741.00PHQ3 Difficulty falling asleep, restless sleep, or excessive sleeping0.850.810.56PHQ4 Feeling tired or lacking vitality1.230.790.79PHQ5 Loss of appetite or overeating0.770.800.62PHQ6 Feeling bad or like a failure, or disappointing oneself or family0.330.590.87PHQ7 Inability to concentrate0.290.580.93PHQ8 Slowness of movement or speech detectable by others, or walking around constantly because ofrestlessness0.190.440.89PHQ9 Thoughts of dying or hurting oneself in some way0.040.240.42SD: standard deviation; GAD-7: Generalized Anxiety Disorder scale-7; PHQ-9: Patient Health Questionnaire-9
Fig. 1Network structure of anxiety and depressive symptoms
Fig. 2Centrality and bridge centrality (a) Centrality (b) Bridge centrality
Fig. 3Stability of centrality
Fig. 4Bootstrap difference test for edge weights
This study constructed a network of anxiety and depressive symptoms among patients with HF. In the network, PHQ2 (feeling down, depressed, or hopeless), PHQ7 (inability to concentrate), and GAD4 (difficulty relaxing) were the most central symptoms. Additionally, anxiety and depressive symptoms were linked by the symptoms of PHQ2 (feeling down, depressed, or hopeless), GAD6 (becoming easily annoyed or impatient), GAD5 (inability to sit still because of anxiety), GAD7 (feeling afraid that something terrible is about to happen), and PHQ6 (feeling bad or like a failure, or disappointing oneself or family). Interventions targeting core and bridging symptoms in the network of anxiety and depressive symptoms are potentially valuable in improving mental health in patients with HF.
This study revealed that 10.8%, 14.1%, and 7.58% of patients with HF exhibited anxiety, depressive symptoms, and combined anxiety and depressive symptoms, respectively. The prevalence of anxiety and depressive symptoms in this study was lower than that in a previous study. Hamatani et al. [42] found that 16% and 28% of hospitalized patients with HF exhibited significant anxiety and depressive symptoms, respectively. This difference may be attributable to the following (a) different approaches used to measure anxiety and depressive symptoms, (b) different standards and cut-off values of measurement tools, (c) different timing of measuring anxiety and depressive symptoms during the research process, and (d) different exclusion and inclusion criteria. [43]
Notably, PHQ2 (feeling down, depressed, or hopeless) was a central symptom in the network of anxiety and depressive symptoms among patients with HF. Depressed mood or hopelessness is one of the core symptoms of depressive symptoms, in the DSM-IV A-criteria for major depression [44]. Holzapfel et al. [45] found that CHF patients had lower levels of depressed mood or hopelessness compared to non-CHF patients. Further, Yang et al. [46] found that PHQ2 (feeling down, depressed, or hopeless) was a central symptom among older adults who are functionally impaired. Moreover, PHQ2 (feeling down, depressed, or hopeless) was the main central symptom of anxiety and depression networks in various populations, which indubitably emphasizes the importance of feeling down, depressed, or hopeless and its commonality in most populations [47]. In the network of anxiety and depressive symptoms, feeling down, depressed, or hopeless was a predictor of not only depressive but also anxiety symptoms. As HF is a progressive and unpredictable disease, patients may experience multiple exacerbation during the course of the disease, frequently accompanied by readmission and additional visits [48]. Patients with HF may experience depressed mood because of their inability to predict or control threatening situations [49]. Furthermore, PHQ7 (inability to concentrate) was a central symptom in the network of anxiety and depressive symptoms among patients with HF—inconsistent with a previous study. Chen et al. [50] found that PHQ7 (inability to concentrate) was a bridge symptom of anxiety and depression among patients with tinnitus. This inconsistency may be related to the poor prognosis patients with HF. Owing to the high hospitalization rate and mortality of HF, most patients would focus on treatment and daily care since its diagnosis, while their attention when doing other things may be affected. Additionally, GAD4 (difficulty relaxing) is a central symptom in the network of anxiety and depressive symptoms among patients with HF— consistent with a previous study on older adults with hypertension [51]. Ma et al. [51]. investigated 4,993 patients with hypertension and found that GAD-4 was a central symptom in the network of anxiety and depressive symptoms. Patients with HF may worry excessively about their health, symptoms, or complications and have trouble relaxing [7].
Notably, PHQ2 (feeling down, depressed, or hopeless) and PHQ6 (feeling bad or like a failure, or disappointing oneself or family) were bridge symptoms of anxiety and depression among patients with HF. In the network, PHQ2 (feeling down, depressed, or hopeless) was both a central and bridge symptom—consistent with Li et al.’s findings [52]. Sadness (feeling down, depressed, or hopeless) was more central than most other symptoms [53], and an increase in sadness may trigger concerns about catastrophic consequences. In this study, PHQ6 (feeling bad or like a failure, or disappointing oneself or family) was an important channel for the interaction between anxiety and depressive symptoms—consistent with previous research on adults with disabilities [54]. Further, HF cause a decline in patients’ physical functioning, affects their social roles, makes them dependent on family or friends, and makes them feel worthless [55]. Moreover, GAD6 (becoming easily annoyed or impatient), GAD5 (unable to sit still because of anxiety), and GAD7 (feeling afraid that something terrible is about to happen) were bridge symptoms of anxiety and depression among patients with HF. Bian et al. [47] found that GAD6 (becoming easily annoyed or impatient) was a main bridge symptom connecting anxiety symptoms, depressive symptoms, and personal mastery among community-dwelling older adults. Further, GAD5 (unable to sit still because of anxiety) may play a prominent role in activating and maintaining the psychopathological network of anxiety and depression among adults with disabilities [54]. Additionally, GAD7 (feeling afraid that something terrible is about to happen) was a bridge symptom of anxiety and depression among patients with tinnitus [50]. Noteworthily, HF is a chronic progressive disease with long-term recurrent attacks [56]. Patients with HF are under both physical and psychological pressure [56], rendering them prone to irritability and worrying about their prognosis.
Identifying central and bridge symptoms in the network of anxiety and depressive symptoms among patients with HF has potential clinical significance. Per the network theory of psychopathology, interventions targeting important central symptoms may exert the greatest effect in disrupting the entire network and reducing its severity, thus aiding in intervention and treatment [57, 58]. Therefore, interventions targeting symptoms such as “feeling down, depressed or hopeless,” “inability to concentrate,” and “difficulty relaxing” may help prevent and treat anxiety and depressive symptoms among patients with HF. Additionally, discontinuing important bridge symptoms can disrupt the connection between—and thereby reduce —comorbidities [59]. Thus, this study recommends using bridge symptoms such as “feeling down, depressed, or hopeless,” “becoming easily annoyed or impatient,” “unable to sit still because of anxiety,” and “feeling bad or like a failure, or disappointing oneself or family” when framing intervention goals to prevent and reduce comorbidity anxiety and depressive symptoms. Mindfulness and social support are important protective factors against anxiety and depression [60]. Healthcare providers should focus on strengthening social support, mindfulness, and coping strategies among patients with HF to alleviate their anxiety and depressive symptoms.
This study exhibited the following limitations. First, a cross-sectional study cannot determine causal relationships. Longitudinal studies can be conducted to explore the dynamic network of anxiety and depressive symptoms among patients with HF. Second, anxiety and depressive symptoms were measured using self-report scales, which may be have caused recall bias, thus necessitating a careful interpretation of the results. Third, as this study focused on the Chinese population with HF, the findings’ generalizability to other populations remains unclear. The sample range should be expanded, especially among different regions and ethnicities, to verify the findings’ applicability and generalizability. Fourth, when conducting the network analysis, covariates or confounding factors were not included. Future studies should consider the impact of comorbidities, disease course, status of medication and personality traits on the network of anxiety and depressive symptoms.
In summary, we determined that PHQ2 (feeling down, depressed, or hopeless), PHQ7 (inability to concentrate), and GAD4 (difficulty relaxing) are central symptoms in the network of anxiety and depressive symptoms among patients with HF, while PHQ2 (feeling down, depressed, or hopeless), GAD6 (becoming easily anxious or impatient), GAD5 (unable to sit still because of anxiety), GAD7 (feeling afraid that something terrible is about to happen), and PHQ6 (feeling bad or like a failure, or disappointing oneself or family) are key nodes connecting anxiety and depressive symptoms. This study provides a theoretical basis for targeted interventions to improve the mental health of—and promote disease recovery among patients with HF.