The heterogeneous depression trajectory and its predictors in coronary heart disease patients undergoing home-based cardiac rehabilitation: a cohort study Authors: Zhen Yang, Xutong Zheng, Liyu Xu, Yu Gao, Chunqi Zhang, Aiping Wang Published: 2024 ABSTRACT AbstractBackgroundPsychological management, particularly addressing depression, is crucial for the effectiveness of home-based cardiac rehabilitation. This study aimed to explore the depression trajectories of coronary heart disease patients during home-based cardiac rehabilitation, identify trajectories associated with cardiovascular readmission, and integrate them into a heterogeneous depression trajectory while examining its predictors.MethodsA prospective cohort study was conducted at a large cardiac rehabilitation center in mainland China. Participants completed the Patient Health Questionnaire-9 to assess depression levels during the 6-month home-based cardiac rehabilitation, with monthly follow-ups. Using latent class growth models to explore depression trajectories. The relationship between different trajectories and cardiovascular readmission was determined using Cox proportional hazards regression, identifying heterogeneous depression trajectory. Logistic regression analysis was employed to explore the influencing factors of heterogeneous depression trajectory.ResultsA total of 346 eligible patients with coronary heart disease participated in the study. Four distinct depression trajectories were identified: sustained no depression (48.0%), delayed onset (15.9%), low U-shaped depression (25.1%), and sustained depression (11.0%). Depression trajectories significantly impacted cardiovascular readmission rates, with higher risks observed in the delayed onset (HR: 4.707, 95% CI: 1.766–12.544) and sustained depression (HR: 8.832, 95% CI: 3.281–23.773) groups. These two groups were combined and termed heterogeneous depression trajectory. Importantly, education level, number of chronic diseases, resilience, social support, and anxiety were independent predictors of heterogeneous depression trajectory.ConclusionsDepression trajectories during home-based cardiac rehabilitation are significantly heterogeneous and influence cardiovascular outcomes. Early identification and management of high-risk factors can enhance psychological health and reduce readmission rates.Keywords: Cardiac rehabilitation, Home-based, Depression, Coronary artery disease, Depression trajectories Introduction Cardiac rehabilitation is a comprehensive program designed to improve the cardiac and overall functional status affected by cardiovascular diseases, prevent recurrent cardiovascular events, enhance quality of life, and facilitate reintegration into normal social life [1]. As a priority recommendation, exercise-based cardiac rehabilitation significantly improves cardiac function and prognosis in patients with coronary artery disease [2, 3]. However, the persistent nature of heart diseases, combined with logistical challenges such as transportation barriers, means that institution-based CR may be financially taxing and time-consuming for patients [4]. As a solution, Home-Based Cardiac Rehabilitation (HBCR) provides a cost-effective and accessible alternative model that brings similar improvements in reducting cardiovascular risk, promoting mental and spiritual well-being, enhancing cardiac function and prognosis, and improving the quality of life for patients [5–9]. HBCR management includes five key components: medication adherence, physical activity, healthy eating, psychological support, and smoking cessation [10]. Psychological management, particularly addressing depression, is crucial for the effectiveness of HBCR [10]. Studies indicate that depression affects approximately 20–30% of cardiac patients, with even higher rates observed in those undergoing cardiac rehabilitation [11, 12]. Depression symptoms (e.g., low mood, lack of motivation, and fatigue) can reduce adherence to rehabilitation exercises and medication, increasing the risk of cardiovascular readmissions [13, 14]. Moreover, depression can severely impact mental health and quality of life, exacerbating cardiac symptoms such as chest pain and dyspnea [15, 16]. Therefore, managing depression in coronary artery disease patients is a critical focus in HBCR. Effective identification and management of depression not only improve psychological health but also enhance rehabilitation outcomes and reduce the risk of adverse cardiovascular events. Depression in coronary artery disease patients undergoing HBCR is dynamic, evolving over time and influenced by various factors, unfolding through different patterns or trajectories, and closely related to adverse cardiovascular outcomes [17]. The heterogeneity of depression trajectories underscores the urgent need to identify key influencing factors to facilitate targeted interventions. In terms of patient care, understanding the significance of depression trajectories not only aids in comprehending the psychological health status of patients but also provides a scientific basis for developing personalized intervention strategies, thereby improving overall health management and rehabilitation outcomes. Additionally, the implications of this research for healthcare systems are equally important; by optimizing depression management strategies, it can enhance resource allocation efficiency, reduce cardiovascular readmission rates, and ultimately improve patients’ quality of life and satisfaction. Current methods for assessing depression in coronary artery disease patients during HBCR primarily rely on one-time cross-sectional evaluations. While these methods capture a snapshot of depressive symptoms, they fail to account for the dynamic nature of depression over time. This limitation highlights a critical gap in our ability to assess and identify the prevalence and key predictors of depression in coronary artery disease patients. To bridge this gap, this study aimed to explore the depression trajectories of coronary heart disease patients during HBCR, identify trajectories associated with cardiovascular readmission, and integrate them into a heterogeneous depression trajectory while examining its predictors. The study hypothesizes that different depression trajectories are significantly associated with the risk of cardiovascular readmission, and that factors such as education level, number of chronic diseases, resilience, social support, and anxiety are independent predictors of depression trajectories. The research questions include: (1) What types of depression trajectories are present in coronary artery disease patients during HBCR? (2) How do these depression trajectories impact the risk of cardiovascular readmission? (3) What factors play a critical role in different depression trajectories? By highlighting the diversity in depression trajectories and their predictive value for adverse outcomes, this study aims to provide a foundation for personalized psychological interventions and theoretical support for future research on depression management within cardiovascular rehabilitation. Methods Study design We conducted a prospective cohort study from July 2023 to April 2024 at a large cardiac rehabilitation center in mainland China. The cohort included eligible coronary artery disease patients based on baseline characteristics such as age, gender, and prior medical history to ensure representativeness of the results. Participants received institutional cardiac rehabilitation prior to enrollment, with baseline data collected one day before or on the day of discharge, serving as the initial reference point for their psychological and physiological status. During the six-month HBCR period, eligible patients completed monthly outpatient follow-ups, which included a questionnaire assessing depressive mood. The study also monitored clinical outcomes, such as cardiovascular readmissions within six months. The follow-up schedule was designed to account for the dynamic nature of depressive mood changes, allowing timely assessments across different stages of HBCR. All participants provided informed consent, and the study protocol was approved by the Ethics Review Committee of the First Affiliated Hospital of China Medical University (No. 2023.66). Participants and sampling Inclusion criteria were: (1) patients aged ≥ 18 years, (2) patients undergoing HBCR, and (3) patients who gave informed consent and volunteered to participate in this study. Exclusion criteria included: (1) patients who is unable to communicate normally due to a hearing or speech impairment. (2) patients who have received psychological counseling and treatment. Sample size was determined based on the 10-events-per-variable (EPV) rule in logistic regression, ensuring at least 10 outcome events for each predictor in the regression equation [18]. We anticipated up to 8 predictors in the model. Preliminary findings indicated a depression prevalence of 0.23 among coronary artery disease patients in HBCR. Accounting for a 0.2 attrition rate, the required final sample size was calculated to be 421. Measurement tools General demographic questionnaire Developed based on literature review and group discussions, this questionnaire included age, gender, marital status, education level, monthly income, comorbidities, and residence. Patient health questionnaire-9 (PHQ-9) The PHQ-9 was used to assess depressive symptoms in coronary artery disease patients [19]. It comprises 9 items with a 4-point Likert scale, ranging from 0 (not at all) to 3 (nearly every day), with a total score of 0 to 27. Higher scores indicate more severe depression, with thresholds of 5, 10, 15, and 20 for mild, moderate, moderately severe, and severe depression, respectively. In this study, the Cronbach’s alpha coefficients of the scale at six time points were 0.768, 0.779, 0.803, 0.732, 0.801, and 0.793, respectively. Perceived social support scale (PSSS) The PSSS assessed perceived social support levels in coronary artery disease patients [20], covering 12 items across three dimensions: family support, friend support, and other support. Responses were collected using a 7-point Likert scale, with total scores ranging from 12 to 84. Higher scores indicate higher perceived social support, with scores of 12–36 indicating low support, 37–60 moderate support, and 61–84 high support. In this study, the Cronbach’s alpha coefficients of the scale was 0.748. 10-item connor-davidson resilience scale (CD-RISC-10) The CD-RISC-10 evaluated psychological resilience in coronary artery disease patients [21], consisting of 10 items on a single dimension. Responses were collected using a 5-point Likert scale, ranging from 0 (not true at all) to 4 (true nearly all the time), with a total score of 0 to 40. Higher scores indicate higher resilience, with 0–13 indicating low resilience, 14–27 moderate resilience, and 28–40 high resilience. In this study, the Cronbach’s alpha coefficients of the scale was 0.812. Self-rating anxiety scale (SAS) The SAS assessed anxiety levels in coronary artery disease patients [22], comprising 20 items. Responses were collected using a 4-point Likert scale. The raw score range was 20–80, which was then multiplied by 1.25 to obtain the standard score. Higher scores indicate higher anxiety levels, with thresholds of 50–59 for mild anxiety, 60–69 for moderate anxiety, and 70 + for severe anxiety. In this study, the Cronbach’s alpha coefficients of the scale was 0.758. Follow-up information questionnaire Cardiovascular readmission events were identified through telephone follow-ups and recorded by follow-up nurses in the questionnaire. Patients were contacted monthly over the six-month follow-up period to collect readmission information. The items covered in the questionnaire included the timing of readmissions, reasons for readmission, changes in symptoms, and an assessment of the patients’ psychological state during rehabilitation. For non-responders, follow-up nurses made two additional phone attempts within one week to ensure data completeness. Cardiovascular readmission referred to any unplanned readmission due to cardiovascular disease during the follow-up period, providing valuable insights into the potential impact of depressive symptoms on cardiovascular health outcomes in coronary artery disease patients. Data analysis Data were analyzed using SPSS 27.0 software. Categorical variables were expressed as frequencies and percentages and analyzed using chi-square tests, as these tests are appropriate for assessing differences in proportions between groups. Continuous variables were expressed as means with standard deviations, and t-tests or Mann-Whitney U tests were chosen based on the distribution characteristics: t-tests for normally distributed data, and the non-parametric Mann-Whitney U test when the normality assumption was not met. Latent class growth models were established using Mplus 8.3. Latent class growth models is a statistical method based on longitudinal data that can identify latent trajectory patterns over time within a study population. Given the dynamic nature of depressive symptoms, Latent class growth models enables us to classify patients into distinct trajectory groups, facilitating a deeper exploration of depression heterogeneity. Linear models were tested for 1 to 7 trajectory groups, with progressively more complex models assessed using fit indices such as Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and adjusted BIC, with lower values indicating better model fit [23]. Entropy values ranged from 0 to 1, with values ≥ 0.8 indicating > 90% classification accuracy. Model comparisons used the bootstrap likelihood ratio and Lo-Mendell-Rubin likelihood ratio test (LMR), with p < 0.05 indicating the K model fits better than the K-1 model. We employed the Cox proportional hazards regression model to explore the impact of different depression trajectories on cardiovascular readmissions within six months in coronary artery disease patients, identifying heterogeneous depression trajectory. Adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated. To reduce the impact of potential confounding variables, we controlled for general sociodemographic factors such as age, gender, marital status, education level, and medical history in the multivariate regression analyses. Univariate and multivariate logistic regression analyses were then used to explore influencing factors for heterogeneous depression trajectory. Results General demographic characteristics of participants A total of 346 coronary artery disease patients participated in this study, with 42.8% being male and 57.2% female. Over half of the patients (52.9%) were between the ages of 40 and 60. The majority (58.1%) were married. Monthly incomes of 42.2% of the patients ranged between 3000 and 6000 RMB. Nearly half of the patients (47.1%) had a moderate level of education. Most patients (53.5%) had 3 to 4 chronic diseases. A majority (53.5%) resided in urban areas. Detailed information is presented in Table 1. Table 1The general demographic information of the participantsVariables n %Age<407020.240–6018352.9>609326.9GenderMale14842.8Female19857.2SpouseNo14541.9Yes20158.1Educational LevelsPrimary education10329.8Secondary education16347.1Higher education8023.1Monthly Income (RMB)<30008123.43000–600014642.2>600011934.4Number of chronic diseases<38123.43–418553.5>48023.1Place of ResidenceVillage16146.5City18553.5 The general demographic information of the participants Prevalence and trajectories of depression in coronary artery disease patients During the HBCR period, the prevalence of depressive symptoms in coronary artery disease patients peaked at 48.77% in the fifth month and was lowest at 19.36% in the second month. As shown in Table 2, the four-group trajectory model with free estimation was identified as the optimal model, with an entropy value of 1.000, AIC of 5712.074, BIC of 5792.849, and aBIC of 5726.231, all of which were relatively low. Both BLRT and LMR tests were statistically significant (p < 0.05). Based on the levels and trends of depression (Fig. 1), these four trajectories were named accordingly. A total of 166 patients (48.0%) had no depression during the HBCR period, remaining below the decision threshold with a stable trend (Group 1: Sustained No Depression). 55 patients (15.9%) initially exhibited no depressive symptoms, but their symptoms gradually increased over time (Group 2: Delayed Onset). 87 patients (25.1%) had depressive symptoms below the decision threshold and showed a declining trend in the first three months, followed by a steady increase after the third month (Group 3: Low U-shaped Depression). 38 patients (11.0%) consistently exhibited depressive symptoms above the decision threshold, with a stable trend (Group 4: Sustained Depression). Table 2Model selection criteria of the Latent Class Growth Model (LCGM) analysisClassAICBICaBICEntropyLMRBLRTClass probabilityLinear19298.6329329.4049304.026127791.0787833.3897798.4940.9990.00000.00000.73121/0.2687336436.9936490.8436446.4310.9990.00000.00000.10983/0.47977/0.4104045957.9266023.3155969.3860.9980.00000.00000.47977/0.10983/0.15896/0.2514555961.1476038.0765974.6300.8280.18240.07500.25145/0.23988/0.23988/0.115896/0.1098365964.4706052.9385979.9750.8250.43300.47740.08382/0.25145/0.23699/0.24277/0.15896/0.02601Quadratic18418.1158456.5798424.856126910.6886964.5396920.1271.0000.00000.00000.10983/0.8901736076.5596145.7956088.6931.0000.00000.00000.47977/0.41040/ 0.1098345299.0895383.7115313.9201.0000.00000.00000.10983/0.25145/0.15896/0.4797755266.4215366.4295283.9490.9180.02190.00000.15896/0.22543/0.25145/0.025434/0.1098365239.2085354.6015259.4320.9290.00730.00000.15896/0.22543/0.08092/0.025145/0.02890/0.25434Free estimation19292.8299338.9869300.919127781.1207838.8177791.2320.9990.00000.00000.73121/0.2687936176.7086245.9436188.8421.0000.00000.00000.47977/0.41040/0.1098345712.0745792.8495726.2311.0000.00000.00000.10983/0.25145/0.47977/0.1589655703.1905795.5055719.3700.8820.06680.00530.28035/0.19942/0.25145/0.10983/0.1589665707.9175811.7715726.1190.8930.10230.00000.15607/0.00289/0.019942/0.28035/0.10983/0.25145 Model selection criteria of the Latent Class Growth Model (LCGM) analysis Trajectory of depression in home-based cardiac rehabilitation for coronary heart disease patients within 6 months post-discharge Heterogeneous depression trajectory in coronary artery disease patients During the follow-up of 346 coronary artery disease patients, 42 experienced cardiovascular readmissions, accounting for 12.14% of the total population. This readmission rate highlights the challenges faced by patients during home-based cardiac rehabilitation, underscoring the need for clinical management to address depressive symptoms and their potential impact on readmission risk. Kaplan-Meier survival analysis indicated that Group 4 had a significantly higher cardiovascular readmissions rate compared to other groups, followed by Group 2. The log-rank test showed a statistically significant difference in the cumulative incidence of cardiovascular readmissions among the four groups (X2 = 29.566, p < 0.001). Adjusted Cox proportional hazards regression results revealed that after controlling for general sociodemographic factors, compared to Group 1, both Group 2 (HR: 4.707, 95% CI: 1.766–12.544, p = 0.002) and Group 4 (HR: 8.832, 95% CI: 3.281–23.773, p < 0.001) had significantly increased risks of cardiovascular readmissions within six months post-discharge (Table 3). Consequently, Groups 2 and 4 were identified and combined as the “Heterogeneous Adherence Trajectory,” encompassing 93 patients, which is 26.88% of the cohort. Table 3Heterogeneity depression trajectory analysis based on Cox proportional hazards modelTrajectory typeUnadjusted modelAdjusted modelHR95% CIp valuesHR95% CIp valuesGroup 1Reference——Reference——Group 25.3862.232–12.999< 0.0014.7071.766–12.5440.002Group 32.4130.952–6.1140.0631.9610.744–5.1640.173Group 47.4552.996–18.550< 0.0018.8323.281–23.773< 0.001 Heterogeneity depression trajectory analysis based on Cox proportional hazards model Predictors of heterogeneous depression trajectory Univariate analysis indicated that six variables were significantly associated with heterogeneous depression trajectory during HBCR in coronary artery disease patients (p < 0.05): age, education level, number of chronic diseases, resilience, social support, and anxiety, as detailed in Table 4. Multivariate logistic regression analysis (Table 5) revealed that education level, number of chronic diseases, resilience, social support, and anxiety were independent predictors of heterogeneous depression trajectory during HBCR in coronary artery disease patients (p < 0.05). Table 4Univariate analysis of predictive factors for heterogeneous depression trajectories (n = 428)VariablesPatients do not belong to heterogeneous depression trajectories (n = 253)Patients who belong to heterogeneous depression trajectories(n = 93)χ2 valuesp-valuesAge6.0600.048 <4058 (22.9)12 (12.9) 40–60134 (53.0)49 (52.7) >6061 (24.1)32 (34.4)Gender0.0370.848 Male109 (43.1)39 (41.9) Female144 (56.9)54 (58.1)Spouse0.2380.626 No119 (47.0)41 (44.1) Yes134 (53.0)52 (55.9)Educational Levels22.459<0.001 Primary education58 (22.9)45 (48.4) Secondary education127 (50.2)36 (38.7) Higher education68 (26.9)12 (12.9)Monthly Income (RMB)5.2710.072 <300053 (20.9)28 (30.1) 3000–6000105 (41.5)41 (44.1) >600095 (37.6)24 (25.8)Number of chronic diseases12.0230.002 <370 (27.7)11 (11.8) 3–4133 (52.6)52 (55.9) >450 (19.8)30 (32.3)Place of Residence0.0310.860 Village117 (46.2)44 (47.3) City136 (53.8)49 (52.7)Resilience14.835<0.001 Low56 (22.1)40 (43.0) Medium119 (47.1)33 (35.5) High78 (30.8)20 (21.5)Social Spport30.912<0.001 Low68 (26.9)55 (59.2) Medium129 (51.0)27 (29.0) High56 (22.1)11 (11.8)Anxiety11.8260.008 None65 (25.7)9 (9.7) Mild76 (30.0)30 (32.3) Moderate84 (33.2)37 (39.8) Severe28 (11.1)17 (18.3) Univariate analysis of predictive factors for heterogeneous depression trajectories (n = 428) Table 5Multivariate analysis of predictors for heterogeneous depression trajectoriesCharacteristicβS.E.ORp-value95%CILower limitUpper limitEducational Levels Primary educationReference Secondary education-1.1360.3100.321<0.0010.1750.590 Higher education-1.4610.4050.232<0.0010.1050.513Number of chronic diseases <3Reference 3–40.8210.4022.2720.0411.0334.996 >41.1800.4413.2530.0081.3697.726Resilience LowReference Medium-0.8550.3230.4250.0080.2260.802 High-0.9450.3630.3890.0090.1910.791Social Spport LowReference Medium-1.3160.3060.268<0.0010.1470.489 High-1.1670.4050.3110.0040.1410.689Anxiety NoneReference Mild0.9090.4592.4810.0481.0086.104 Moderate1.0540.4482.8690.0191.1926.905 Severe1.4670.5434.3370.0071.49612.576Constant-0.5090.5650.6010.367 Multivariate analysis of predictors for heterogeneous depression trajectories Discussion In this study, depression in coronary artery disease patients exhibited significant heterogeneity during HBCR, which could be categorized into four distinct trajectory types: sustained no depression, delayed onset, low U-shaped depression, and sustained depression. Understanding these depression trajectories is crucial for clinical practice and patient outcomes, as they not only reveal the dynamic changes and heterogeneity of depressive symptoms but also closely associate with the risk of cardiovascular readmission, particularly in the delayed onset and sustained depression groups, collectively termed the heterogeneous depression trajectory. Notably, education level, number of chronic diseases, resilience, social support, and anxiety were independent predictors of the heterogeneous depression trajectory. These findings not only reveal the dynamic changes, heterogeneity, and potential influencing factors of depression during HBCR in coronary artery disease patients but also underscore the importance of early identification and personalized management of depression. By better understanding these trajectories, we can provide a scientific basis for developing more effective intervention strategies, which can improve patients’ psychological health, rehabilitation outcomes, and reduce the risk of cardiovascular readmission. This study revealed significant trajectory heterogeneity in depression among coronary artery disease patients during HBCR, with different trajectory types exhibiting distinct characteristics in psychological states and disease progression. These trajectory patterns reflect the dynamic changes in depression, emphasizing the importance of continuous monitoring of depression during rehabilitation [24]. Particularly, the peak in depression at the fifth month may be associated with various challenges and pressures faced by patients in the later stages of rehabilitation, such as the complexity of rehabilitation plans, difficulty in long-term adherence, and potential lack of social support [25, 26]. The findings of the delayed onset and low U-shaped depression groups further highlight the potential volatility and delayed effects of depression. These patients may not exhibit significant depression in the early stages, but over time, symptoms gradually emerge or recur. This delayed effect might be related to patients’ initial adaptation to rehabilitation, gradually increasing physical burden, and the accumulation of long-term psychological stress [27, 28]. Patients in the sustained depression group consistently exhibited high levels of depression throughout the rehabilitation period, indicating that these patients might require more intensive and personalized psychological interventions. The persistent presence of depression not only reduces adherence to rehabilitation plans but also increases the risk of cardiovascular readmission, further impacting overall rehabilitation outcomes [29, 30]. Our study found that the depression trajectories significantly influenced the risk of cardiovascular readmission in coronary artery disease patients during HBCR, particularly in the sustained depression and delayed onset groups. These findings highlight the potential negative impact of depression on cardiac rehabilitation outcomes and emphasize the importance of screening and personalized psychological interventions [31]. Kaplan-Meier survival curves and Cox proportional hazards regression analysis both indicated that patients in the sustained depression group had a significantly higher risk of cardiovascular readmission compared to other groups. This suggests that persistent depression may lead to greater psychological and physiological stress during rehabilitation, thereby increasing the incidence of cardiovascular events [32]. The high readmission risk in the delayed onset group is also noteworthy, indicating that depression not promptly identified and managed in the early stages of rehabilitation may gradually worsen and adversely affect cardiovascular health [33]. Combining Groups 2 and 4 into the “Heterogeneous Depression Trajectory” further underscores the heterogeneous impact of different depression trajectory types on patient prognosis. Approximately 27% of patients belong to this category, suggesting a need for special attention to this group with significant changes in depression in clinical practice. Early identification of these high-risk patients and providing appropriate psychological support and interventions can not only improve their psychological health but also potentially reduce the incidence of cardiovascular readmissions and enhance overall rehabilitation outcomes. The study identified education level, number of chronic diseases, resilience, social support, and anxiety as independent predictors of heterogeneous depression trajectory in coronary artery disease patients during HBCR. These results provide new perspectives for understanding the formation mechanisms of depression trajectories and offer a scientific basis for developing clinical intervention strategies. Firstly, compared to primary education, patients with secondary and higher education levels had a significantly lower risk of heterogeneous depression trajectory. This finding may reflect the advantages of higher education levels in health knowledge, self-management skills, and resource acquisition, enabling patients to cope more effectively with the challenges and pressures of cardiac rehabilitation [34, 35]. Secondly, the relationship between the number of chronic diseases and depression trajectories suggests that patients with more chronic diseases face a higher risk of depression. This result implies that patients with multiple chronic diseases may experience greater disease burden, leading to increased psychological stress and a higher incidence of depression [36]. Therefore, during rehabilitation, special attention should be given to these high-risk patients, providing comprehensive management and support through multidisciplinary collaboration. The findings on resilience and social support further emphasize the crucial role of psychological and social factors in managing depression. Patients with high resilience and social support levels had significantly lower risks of depression during rehabilitation, indicating that enhancing patients’ psychological resilience and social support networks is an effective intervention strategy [37, 38]. Psychological resilience helps patients cope better with adversity, while strong social support provides emotional and practical assistance, alleviating feelings of loneliness and stress [37, 39]. Finally, anxiety, whether mild, moderate, or severe, significantly increased the risk of heterogeneous depression trajectory. This finding indicates a close link between anxiety and depression, where the presence of anxiety symptoms exacerbates depression. Therefore, early identification and management of anxiety symptoms during rehabilitation are crucial and can be addressed through psychological counseling, medication, and other interventions to alleviate patients’ anxiety and depression. Limitations Despite the important insights provided by this study, several limitations need to be considered. First, this study was conducted at a single center in mainland China, which may restrict the generalizability of the findings. The specific demographics and healthcare context of this center may not reflect the diversity of other populations or healthcare settings. Future research should aim to replicate this study in multiple centers and varied populations to enhance the external validity and applicability of the findings. Second, the assessment of depression relied on self-reported questionnaires, which may introduce reporting bias, with patients potentially underestimating or overestimating their depression levels for various reasons. To reduce this bias, future studies should consider incorporating objective psychological assessment tools and clinical interviews. Additionally, we would like to emphasize that it is not possible to infer causal relationships due to the observational nature of the study. Lastly, although this study explored multiple potential predictors, it did not include all possible factors influencing depression trajectories, such as socioeconomic status, family support structure, and patients’ lifestyle habits. Future research should further investigate the impact of these factors to comprehensively understand the dynamic changes and predictors of depression in coronary artery disease patients. Conclusion This study revealed significant heterogeneity and independent predictors of depression trajectories in coronary artery disease patients during HBCR. These findings emphasize the importance of continuous psychological assessment and personalized management during rehabilitation. Specifically, education level, number of chronic diseases, resilience, social support, and anxiety are crucial factors influencing depression trajectories. Therefore, in future research, early identification and management of these high-risk factors could lead to more effective intervention strategies, improving patients’ psychological health, reducing the risk of cardiovascular readmission, and enhancing overall rehabilitation outcomes. Acknowledgements The authors are grateful to patients with coronary heart disease who participated in this study, and also to the health providers for their strong support in sampling. Author contributions Z.Y. contributed to the conceptualization, methodology, investigation, formal analysis and writing – original draft. X.T.Z. and X.L.Y. contributed to the conceptualization, methodology, and formal analysis. C.Q.Z. and Y.G. contributed to the investigation and formal analysis. A.P.W. contributed to the conceptualization, methodology, supervision, and writing – review & editing. All gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy. Funding None. Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate All procedures were conducted in accordance to the Declaration of Helsinki of 1964 and its further modifications. All participants gave informed consent forms. The research proposal was approved by the Ethics Review Committee of the First Affiliated Hospital of China Medical University (No. 2023. 66). Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.