Association of multimorbidity burden with outcomes after percutaneous coronary intervention: a retrospective cohort study

Abstract

Background

Multimorbidity is increasingly common among patients undergoing percutaneous coronary intervention (PCI), particularly in aging populations. Although previous studies have shown that comorbidity burden influences outcomes in coronary artery disease, the association between cumulative multimorbidity burden and both clinical and functional outcomes after PCI in contemporary real-world practice remains incompletely characterized.

Methods

This single-center retrospective cohort study included 1,238 consecutive patients who underwent PCI between January 2020 and December 2023. Patients were categorized into a low-burden group (0–1 chronic disease, n = 482) and a high-burden group (≥ 2 chronic diseases, n = 756). The primary endpoint was major adverse cardiovascular events (MACE), defined as a composite of all-cause mortality, non-fatal myocardial infarction, ischemic stroke, and repeat coronary revascularization. Multivariable Cox regression and propensity score matching were performed for time-to-event outcomes. Functional outcomes, including New York Heart Association (NYHA) class, Seattle Angina Questionnaire-7 (SAQ-7), and EuroQol Five Dimensions (EQ-5D), were assessed from baseline to 12-month follow-up.

Results

Over a median follow-up of 36 months (IQR 28–42), 312 patients (25.2%) experienced MACE. The incidence of MACE was significantly higher in the high-burden group than in the low-burden group (30.4% vs. 17.0%, P < 0.001; log-rank P < 0.001). After multivariable adjustment, high multimorbidity burden remained associated with increased risk of MACE (adjusted HR = 1.52, 95% CI: 1.18–1.96, P = 0.001). All-cause mortality (13.6% vs. 7.3%, P < 0.001) and all-cause readmission (27.0% vs. 18.0%, P < 0.001) were also more frequent in the high-burden group. Functional recovery at 12 months was less favorable in patients with higher multimorbidity burden, with lower improvement in NYHA class (45.3% vs. 62.0%, P < 0.001), smaller increases in SAQ-7 score (+ 15.1 ± 6.8 vs. +22.6 ± 7.5, P < 0.001), and less improvement in EQ-5D index (+ 0.09 ± 0.11 vs. +0.14 ± 0.10, P < 0.001). Subgroup analyses showed generally consistent associations across age and sex strata, whereas disease-specific subgroup findings were interpreted cautiously because some stratification variables also contributed to the multimorbidity definition.

Conclusion

In this single-center retrospective cohort, higher multimorbidity burden was associated with increased risks of adverse clinical outcomes and less favorable functional recovery after PCI. These findings suggest that multimorbidity may provide additional prognostic information in contemporary PCI practice, although further prospective multicenter studies are needed to confirm its value in routine risk assessment.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12872-026-05975-7.

Introduction

Coronary artery disease (CAD remains a leading cause of morbidity and mortality worldwide, and percutaneous coronary intervention (PCI) has substantially improved survival in both acute and chronic coronary syndromes. The patient population undergoing PCI has changed over time, with an increasing proportion of older adults presenting with multiple chronic conditions [1, 2]. Previous registries and contemporary studies, including the MIG registry and UK/Northeast ACS cohorts, have shown that multimorbidity or comorbidity indices are associated with adverse PCI outcomes. However, the cumulative impact of multimorbidity burden on both hard and functional outcomes in contemporary real-world cohorts remains incompletely characterized [3, 4]. This study provides an updated evaluation of multimorbidity burden in a single-center PCI population, examining its association with major adverse cardiovascular events, mortality, readmission, and functional recovery, thereby offering additional insights that complement existing evidence.

In this study, we aimed to determine whether multimorbidity burden independently predicts adverse clinical and functional outcomes following PCI in a real-world cohort. We further assessed the robustness of this association using the Charlson Comorbidity Index as a validated sensitivity measure.

coronary artery disease (CAD is a leading cause of death and disability worldwide. Its clinical burden has been rising over the past decades. With the rapid development of interventional techniques, percutaneous coronary intervention (PCI) has become a major method to restore blood flow in patients with coronary artery disease (CAD [1, 2]. It has helped many acute and chronic cardiovascular patient to live longer. However, as PCI technology and therapeutic treatments continue to improve, the characteristics of the patient population have also changed significantly. In particular, the aging factor is of key concern, as it increases the overall incidence of coronary artery disease (CAD and leads to a sharp rise in the number of patients with multiple chronic diseases [3, 4]. Therefore, having multiple diseases simultaneously is becoming common among coronary artery disease (CAD patients rather than an exception.

The condition in which a patient has two or more chronic diseases simultaneously is usually defined as multimorbidity. This condition is more pronounced among elderly patients. Epidemiological data show that more than half of individuals aged ≥ 65 years have multimorbidity [5]. Among those patients who have undergone PCI, the incidence of various chronic diseases such as diabetes, hypertension, chronic kidney disease, chronic obstructive pulmonary disease (COPD) and heart failure is relatively high and continues to increase [6]. This presents growing challenges for post-PCI management. Multiple complications are not only related to the complex interactions among diseases, but also associated with multiple concurrent therapeutic treatments, chronic systemic inflammation, immune dysregulation, and reduced physiological reserves. These factors may have adverse effects on the postoperative recovery after PCI surgery and long-term cardiovascular outcomes, and such adverse effects may recur [7]. Notably, most previous studies only assessed single co-morbid disease scenarios. For instance, diabetes increases the risk of stent restenosis and adverse cardiovascular events by accelerating atherosclerosis and platelet activation [8]. In contrast, chronic kidney disease is linked to compromised drug metabolism, endothelial dysfunction, and vascular calcification, which can result in poorer outcomes following PCI [9]. Nevertheless, there are still large number of patients who experience multiple chronic diseases at the same time. Hence, focusing on only one disease condition fails to capture the overall pathophysiological burden arising from the accumulation and interaction of multiple conditions.

Adopting relatively strict inclusion criteria, frequently excluding patients with multiple chronic diseases, is a general practice in most randomized clinical trials, which limits the generalizability of the findings to the actual PCI population. Similarly, retrospective cohort studies typically examine single comorbidities rather than the cumulative burden of multimorbidity. Common risk prediction tools such as GRACE, SYNTAX, and TIMI scores also fail to consider multimorbidity as an integrated prognostic factor. Consequently, the lack of proper guidelines can significantly impact clinical decisions regarding therapeutic treatments for multimorbidity-related complex patient scenarios [1, 7].

In clinical settings, multimorbidity may influence post-PCI outcomes through multiple pathophysiological pathways. Chronic systemic inflammation and immune dysregulation are characteristic features of many chronic diseases, and are associated with the unstable condition of atherosclerotic plaques and an increased risk of thrombosis. Among this group of patients, multiple medication is also relatively common. Drug-drug interactions may weaken the efficacy of anti-platelet therapy or lipid-lowering therapy, thereby reducing the benefits of PCI surgery [10, 11]. A reduction in physiological reserves, marked by kidney damage and/or impaired lung function, can hinder postoperative recovery and increase the body’s susceptibility to complications [12]. Multimorbidity can also give way to psychological distress, reducing quality of life, affecting treatment adherence and worsening long-term cardiovascular outcomes.

Here, we aim to investigate whether the overall burden of multimorbidity can independently predict adverse outcomes after PCI. This poses a clinically meaningful question needed to be answered as the number of patients, undergoing PCI and are with multiple chronic diseases, is increasing gradually. The traditional framework that only targets a single disease may underestimate the risks faced by these patients. This study utilized real-world clinical data to assess the relationship between the burden of multiple diseases and the prognosis after PCI, and also examined whether multiple diseases could provide some additional prognostic value beyond the already identified cardiovascular risk factors.

Methods

Study design and population

This is a retrospective cohort study that utilized routinely collected clinical data. Patients who were consecutively hospitalized in the cardiology department of the hospital and underwent percutaneous coronary intervention (PCI) between January 2020 and December 2023 were included in the study. For all included cases, coronary artery disease was confirmed by angiography. The cohort included patients presenting with both stable coronary artery disease and acute coronary syndromes (unstable angina, NSTEMI, and STEMI). and all procedures were performed at the same institution. The inclusion criteria of the study were as follows [13]:

1. age ≥18 years;

2. At least one PCI procedure performed during the study period with complete procedural records;

3. Available clinical follow-up data.

Exclusion criteria included:

1. A history of coronary artery bypass grafting (CABG), to avoid confounding effects of different revascularization strategies on outcomes;

2. Presence of acute cardiogenic shock or perioperative death that precluded follow-up;

3. Incomplete clinical information or missing key variables preventing valid analysis.

The process of patient enrollment and selection is summarized in Fig. 1. Procedural characteristics collected included the number of diseased vessels, number of implanted stents, and procedural success rate as documented in operative reports. Due to the retrospective nature of the study and limitations of routinely collected data, detailed lesion morphology characteristics, stent type specifications, and specific periprocedural complications (such as no-reflow phenomenon or contrast-induced nephropathy) were not consistently available for all patients and therefore were not included in the final analysis. However, key indicators reflecting disease severity and procedural complexity, including multivessel disease, left ventricular ejection fraction, and number of implanted stents, were incorporated into the multivariable models to partially account for procedural risk differences between groups. Coexisting chronic diseases (type and number) were carefully recorded to investigate the multimorbidity burden. The study protocol was approved by the Ethics Committee of the hospital and was conducted in accordance with the principles of the Declaration of Helsinki. All patient information was anonymized prior to analysis, and no additional informed consents were obtained. To reduce confounding, propensity score matching (PSM) was performed using 1:1 nearest-neighbor matching with a caliper width of 0.02. The propensity score model included age, sex, smoking status, left ventricular ejection fraction, number of diseased vessels, number of implanted stents, hypertension, and diabetes. Covariate balance after matching was assessed using standardized mean differences (SMDs), with SMD < 0.1 considered acceptable. The matched baseline characteristics are presented in Table S1, and balance plots are shown in Figure S1.

Fig. 1.

Flow diagram of patient selection and analysis. Flowchart showing the selection of patients who underwent percutaneous coronary intervention (PCI) between January 2020 and December 2023. A total of 1,372 patients were screened, and 1,238 eligible patients were included after excluding those with prior coronary artery bypass grafting, perioperative death, or incomplete data. Patients were divided into two groups according to multimorbidity burden: low-burden (0-1 chronic diseases, n = 482) and high-burden (≥2 chronic diseases, n = 756). All patients completed follow-up with a median duration of 36 months (IQR 28 –42). Abbreviations: PCI, percutaneous coronary intervention; MACE, major adverse cardiovascular events

Due to the retrospective design and limitations of routinely collected data, detailed medication variables—including antiplatelet regimen, adherence, and statin intensity—were not consistently available for inclusion in propensity score matching. Similarly, detailed procedural characteristics, such as lesion complexity, stent type, access site, completeness of revascularization, and peri-procedural complications, could not be reliably captured. Acute presentation severity was classified by broad syndrome categories (unstable angina, NSTEMI, STEMI), while finer stratification was not consistently recorded. These limitations are acknowledged and considered in the interpretation of our finding.

Definition of multimorbidity

Multimorbidity was defined as the coexistence of two or more chronic conditions. we used the Charlson Comorbidity Index (CCI) as a validated measure of comorbidity burden in a sensitivity analysis. Pre-procedural comorbidities were obtained from clinical and discharge records, including hypertension, diabetes, chronic kidney disease, COPD, heart failure, cerebrovascular disease, peripheral vascular disease, and malignancy.

Multimorbidity was defined as the coexistence of two or more chronic conditions in the same individual, consistent with the widely accepted definition proposed in epidemiological and cardiovascular research [1, 2]. Pre-procedural comorbidities were obtained from clinical and discharge records, including hypertension, diabetes, chronic kidney disease, chronic obstructive pulmonary disease (COPD), heart failure, cerebrovascular disease, peripheral vascular disease, and malignancy.

Multimorbidity was defined as the coexistence of two or more chronic conditions. The primary analysis dichotomized patients into a low-burden group (0–1 chronic disease) and a high-burden group (≥ 2 chronic diseases), consistent with prior cardiovascular research and providing a practical framework for clinical interpretation. To provide a more granular view, we also report the full distribution of comorbidity counts across the cohort (0, 1, 2, 3, ≥ 4). In secondary analyses, we tested dose-response relationships using these finer categories to explore whether higher comorbidity counts were associated with progressively increased risk of adverse outcomes.

The chronic conditions included in this study were hypertension, diabetes, chronic kidney disease (CKD), chronic obstructive pulmonary disease (COPD), heart failure, cerebrovascular disease, peripheral vascular disease, and malignancy. These conditions were selected based on their established impact on cardiovascular prognosis and consistent documentation in clinical and discharge records [14].

To further validate our findings, we performed a sensitivity analysis using the Charlson Comorbidity Index (CCI), which accounts for disease severity and age. Patients were classified into moderate-burden (CCI ≤ 4) and high-burden (CCI ≥ 5) groups [15, 16]. Consistency between the disease-count and CCI-based analyses supports the robustness of multimorbidity as an independent predictor of post-PCI outcomes.

Data on discharge medications, including β-blockers, antiplatelet therapy, and statins, were extracted where available from electronic medical records. However, completeness of medication data varied across patients [17].

Study outcomes

The primary outcome of this study was major adverse cardiovascular events (MACE), defined as a composite of all-cause mortality, non-fatal myocardial infarction, ischemic stroke, and repeat coronary revascularization [18–21]. Repeat revascularization was defined as any subsequent coronary revascularization procedure after the index PCI, including either percutaneous coronary intervention or coronary artery bypass grafting, regardless of whether it involved the target lesion, target vessel, or a non-target vessel. These events were identified from hospital records and follow-up documentation. Because ischemia-driven classification and detailed lesion-level adjudication were not consistently available in the retrospective dataset, repeat revascularization was analyzed as any repeat coronary revascularization rather than as a target-lesion- or target-vessel-specific endpoint.

All MACE events were adjudicated by two independent cardiologists through review of medical records and follow-up data. In cases of disagreement, a senior cardiologist made the final decision.

Secondary outcomes included all-cause mortality, readmission, improvement in cardiac functional status assessed by New York Heart Association (NYHA) class, change in angina-related health status assessed by the Seattle Angina Questionnaire-7 (SAQ-7), and change in quality of life assessed by the EuroQol Five Dimensions (EQ-5D) index. NYHA class was evaluated by trained clinical staff at baseline and at 12-month follow-up using routine clinical assessment. SAQ-7 and EQ-5D data were collected at the same time points through standardized patient-reported questionnaires administered by the clinical research team as part of routine follow-up practice. Functional outcomes were analyzed as change from baseline to 12 months.

This multidimensional outcome framework is comprehensive, encompassing hard cardiovascular endpoints such as mortality and ischemic events, as well as patient-centered functional recovery indicators and quality of life indicators. In this way, a comprehensive assessment was made of the impact of multimorbidity burden on the prognosis after PCI. Readmission analysis mainly focused on the time of the first readmission. If the patient had died or the follow-up had ended, the relevant circumstances were reviewed.

Statistical analysis

All statistical analyses were conducted using SPSS version 22.0 and R software version 4.3.1. Continuous variables were assessed for normality and are presented as mean ± standard deviation or median (interquartile range), as appropriate. Between-group comparisons were performed using the independent-samples Student’s t-test or the Mann–Whitney U test. Categorical variables are expressed as counts and percentages and were compared using the chi-square test or Fisher’s exact test. Survival and event-free rates were estimated using Kaplan–Meier analysis and compared with the log-rank test.

The association between multimorbidity burden and clinical outcomes was evaluated using Cox proportional hazards regression models. Covariates for the fully adjusted primary model were prespecified based on clinical relevance and included age, sex, smoking status, left ventricular ejection fraction, Killip class, multivessel disease, number of implanted stents, and clinical presentation (acute coronary syndrome versus stable coronary artery disease). Individual comorbidities that contributed to the multimorbidity definition, including hypertension, diabetes, chronic kidney disease, chronic obstructive pulmonary disease, heart failure, cerebrovascular disease, peripheral vascular disease, and malignancy, were not simultaneously included as separate covariates in the fully adjusted primary model in order to avoid overadjustment and collinearity. Sensitivity analysis was instead performed using the Charlson Comorbidity Index as an alternative measure of comorbidity burden.

Instead, multimorbidity burden was defined using both a disease count approach and the Charlson Comorbidity Index (CCI) in sensitivity analyses. Collinearity diagnostics (e.g., Variance Inflation Factor (VIF)) were performed to assess for multicollinearity among covariates, and no significant collinearity was detected (all VIFs < 2.5). Individual comorbidities that were components of the multimorbidity definition were not simultaneously entered as independent covariates in the fully adjusted models to avoid partial collinearity and overadjustment.

A stepwise modeling strategy was applied. Model 1 adjusted for demographic variables. Model 2 additionally incorporated cardiovascular risk factors and baseline cardiac function. Model 3 further included procedural characteristics and clinical presentation, representing the fully adjusted model. Multicollinearity was assessed using variance inflation factors (all < 2.5). The proportional hazards assumption was examined using Schoenfeld residuals and log-minus-log plots, with no significant violations detected. Model discrimination and calibration were evaluated using Harrell’s C-statistic and comparison of observed versus predicted event rates across risk tertiles. Hazard ratios with 95% confidence intervals are reported. Prespecified subgroup analyses were performed according to age (< 65 vs. ≥65 years), sex, left ventricular ejection fraction (< 50% vs. ≥50%), diabetes, hypertension, chronic kidney disease, chronic obstructive pulmonary disease, cerebrovascular disease, and polyvascular disease. Interaction terms were included to assess effect modification.

Sensitivity analyses were conducted using 1:1 nearest-neighbor propensity score matching with a caliper of 0.02. Covariate balance was confirmed before reanalyzing outcomes. Missing data were < 5% and handled using complete-case analysis; multiple imputation yielded consistent results. All tests were two-sided, and P < 0.05 was considered statistically significant.

Before final manuscript preparation, all descriptive tables, matched cohort summaries, and regression outputs were regenerated from the final analytical dataset. Reported totals, subgroup counts, and supplementary materials were cross-checked for internal consistency.

Results

Baseline characteristics

During the study period, 1,238 eligible patients were included (Table 1). Among them, 482 patients (38.9%) were in the low-burden group, defined as having 0–1 chronic condition, and 756 patients (61.1%) were in the high-burden group, defined as having ≥ 2 chronic conditions. The average age of the overall cohort was 65.7 ± 10.9 years, with 72.5% of patients being male. The high-burden group was older (68.2 ± 9.8 years vs. 61.5 ± 11.2 years, P < 0.001) and had a higher proportion of females (33.9% vs. 21.4%, P < 0.001). The high-burden group also had significantly higher rates of hypertension (82.8% vs. 46.7%), diabetes (49.5% vs. 18.0%), chronic kidney disease (22.9% vs. 6.0%), COPD (17.7% vs. 4.8%), and prior stroke (12.3% vs. 3.0%) compared to the low-burden group (all P < 0.001). There were no significant differences in BMI between groups (P > 0.05). In terms of cardiac function and interventional characteristics, the high-burden group had a lower average left ventricular ejection fraction (LVEF) (49.2 ± 8.7% vs. 54.8 ± 7.9%, P < 0.001) and a higher proportion of patients with Killip class ≥ II (21.3% vs. 12.0%, P < 0.001). Additionally, the high-burden group had a higher rate of multivessel disease (64.0% vs. 39.2%, P < 0.001) and a higher number of stents implanted (2.3 ± 1.1 vs. 1.6 ± 0.9, P < 0.001). The distribution of clinical presentation (ACS vs. stable coronary artery disease) did not differ significantly between the groups (Table 1). After incorporating ACS status into the multivariable Cox regression model, the association between high multimorbidity burden and major adverse cardiovascular events (MACE) remained statistically significant (adjusted HR = 1.52, 95% CI: 1.18–1.96, P = 0.001), indicating that the observed effect was not solely driven by acute coronary syndrome (ACS) presentation.

Table 1.

Baseline demographic and clinical characteristics of patients by multimorbidity burden

Characteristics	Overall (N = 1,238)	Low-burden group (n = 482)	High-burden group (n = 756)	P value
Age, years (mean ± SD)	65.7 ± 10.9	61.5 ± 11.2	68.2 ± 9.8	< 0.001
Male, n (%)	898 (72.5)	379 (78.6)	519 (66.1)	< 0.001
Hypertension, n (%)	859 (69.4)	225 (46.7)	634 (82.8)	< 0.001
Diabetes mellitus, n (%)	442 (35.7)	87 (18.0)	355 (49.5)	< 0.001
Chronic kidney disease, n (%)	195 (15.8)	29 (6.0)	173 (22.9)	< 0.001
Chronic obstructive pulmonary disease (COPD), n (%)	151 (12.2)	23 (4.8)	134 (17.7)	< 0.001
Prior stroke, n (%)	98 (7.9)	15 (3.1)	93 (12.3)	< 0.001
Body mass index (BMI), kg/m² (mean ± SD)	25.4 ± 3.7	25.1 ± 3.5	25.6 ± 3.8	0.08
Smoking history, n (%)	523 (42.3)	215 (44.6)	308 (40.7)	0.15
Left ventricular ejection fraction (LVEF), % (mean ± SD)	51.5 ± 8.6	54.8 ± 7.9	49.2 ± 8.7	< 0.001
Killip class ≥ II, n (%)	205 (16.6)	58 (12.0)	161 (21.3)	< 0.001
Multivessel disease, n (%)	655 (52.9)	189 (39.2)	466 (64.0)	< 0.001
Number of stents (mean ± SD)	2.0 ± 1.0	1.6 ± 0.9	2.3 ± 1.1	< 0.001

The incidence of MACE was significantly higher in the high-burden group compared to the low-burden group (30.4% vs. 17.0%, P < 0.001) (Table 2). Kaplan–Meier survival analysis demonstrated clear early divergence between groups (Fig. 2). After multivariable adjustment, multimorbidity burden remained an independent predictor of MACE (adjusted HR = 1.52, 95% CI: 1.18–1.96, P = 0.001). The high-burden group also had significantly higher all-cause mortality (13.6% vs. 7.3%, P < 0.001) and all-cause readmission rates (27.0% vs. 18.0%, P < 0.001). Given the broad definition of readmission, this finding should be interpreted as reflecting greater overall illness burden after PCI rather than PCI-related prognosis alone (Table 3). Functional recovery was significantly more limited in the high-burden group, with fewer patients achieving improvement in NYHA class (45.3% vs. 62.0%, P < 0.001), smaller increases in SAQ-7 score (+ 15.1 ± 6.8 vs. +22.6 ± 7.5, P < 0.001), and less improvement in EQ-5D index (+ 0.09 ± 0.11 vs. +0.14 ± 0.10, P < 0.001).

Table 2.

Clinical outcomes after PCI according to multimorbidity burden

Outcome	Low-burden (n = 482)	High-burden (n = 756)	P value
Median follow-up, months (IQR)	35(27–41)	37(29–43)	0.21
MACE, n (%)	82 (17.0)	230 (30.4)	< 0.001
All-cause mortality, n (%)	35 (7.3)	103 (13.6)	< 0.001
Non-fatal myocardial infarction, n (%)	21 (4.4)	64 (8.5)	0.003
Ischemic stroke, n (%)	11 (2.3)	29 (3.8)	0.08
Repeat revascularization, n (%)	28 (5.8)	78 (10.3)	0.004

Data are presented as n (%) or median (IQR) unless otherwise indicated.

Fig. 2.

Kaplan–Meier curves for cumulative incidence of major adverse cardiovascular events (MACE) after PCI stratified by multimorbidity burden.Shaded areas indicate 95% confidence intervals calculated using Greenwood’s formula. The high-burden group (≥ 2 chronic diseases) exhibited significantly higher event rates compared with the low-burden group (0–1 chronic disease) (log-rank P < 0.0001). The number of patients at risk at predefined time points (0, 6, 12, 24, and 36 months) is displayed below the curves to enhance transparency and facilitate interpretation of temporal risk patterns

Table 3.

Associations between multimorbidity burden and post-PCI outcomes

Outcome	HR (95% CI)	P value	Adjusted HR (95% CI)	Adjusted P value
All-cause mortality	1.61 (1.09–2.39)	0.016	1.48 (1.02–2.22)	0.039
Readmission	1.45 (1.10–1.92)	0.007	1.39 (1.05–1.86)	0.021
Inadequate NYHA improvement (< 1 class)	1.68 (1.25–2.27)	< 0.001	1.55 (1.15–2.10)	0.004
Limited SAQ-7 improvement (< 20 points)	1.74 (1.28–2.36)	< 0.001	1.62 (1.19–2.21)	0.002
Limited EQ-5D improvement (< 0.1 index)	1.59 (1.18–2.14)	0.002	1.47 (1.08–2.01)	0.014

After propensity score matching, the final matched sample included 428 pairs. Baseline covariate balance after matching is presented in Table S1, with standardized mean differences confirming acceptable balance across matched groups. All descriptive statistics, subgroup counts, and matched-sample analyses were regenerated from the final analytical dataset to ensure internal consistency across the main text, tables, and supplementary materials.

Incidence of MACE and association with multimorbidity burden

During a median follow-up of 36 months, 312 patients (25.2%) experienced major adverse cardiovascular events (MACE). The incidence of MACE was significantly higher in the high-burden group (30.4% vs. 17.0%, P < 0.001) (Table 2). Kaplan–Meier survival analysis confirmed this difference, with early divergence in MACE rates (Fig. 2). After multivariable adjustment, high multimorbidity burden remained associated with increased risk of MACE (adjusted HR = 1.52, 95% CI 1.18–1.96, P = 0.001) (Fig. 3). Patients in the high-burden group also had higher all-cause mortality and readmission rates (P < 0.001), as well as less improvement in NYHA class, SAQ-7 scores, and EQ-5D index (P < 0.001). Subgroup analyses showed consistent associations across different clinical subgroups.

Fig. 3.

Multivariable Cox proportional hazards model for major adverse cardiovascular events (MACE) after PCI. Forest plot showing hazard ratios (HRs) and 95% confidence intervals (CIs) from the fully adjusted Cox regression model. The dashed vertical line indicates HR = 1. The model included multimorbidity burden and prespecified covariates reflecting demographic characteristics, baseline cardiac function, procedural complexity, and clinical presentation. Individual comorbidities that contributed to the multimorbidity definition were not simultaneously entered into the primary model to avoid overadjustment and collinearity. The adjusted association between high multimorbidity burden and MACE is shown using the final verified estimate reported in the main text

Because the cohort included both stable coronary artery disease and acute coronary syndrome presentations, residual confounding related to presentation severity remains possible despite adjustment for ACS status, Killip class, and left ventricular ejection fraction. The early separation of the Kaplan–Meier curves suggests that part of the observed risk difference may reflect early post-procedural vulnerability, which could not be fully disentangled in the present analysis.

Secondary outcomes and functional recovery after PCI

During follow-up, 138 patients (11.1%) died, with significantly higher mortality in the high-burden group (13.6% vs. 7.3%, P < 0.001) (Table 3). Similarly, readmission rates were higher in the high-burden group (27.0% vs. 18.0%, P < 0.001). Both mortality and readmission rates are shown in Fig. 4A-B. Functional recovery was attenuated in the high-burden group. At 12 months, fewer patients in the high-burden group showed improvement in NYHA class (45.3% vs. 62.0%, P < 0.001), and they had smaller increases in SAQ-7 score (+ 15.1 vs. +22.6, P < 0.001) and EQ-5D index (+ 0.09 vs. +0.14, P < 0.001) (Fig. 4).

Fig. 4.

Comparisons of clinical and functional outcomes between low and high multimorbidity burden groups. A-B Bar charts showing all-cause mortality, readmission, and NYHA functional improvement rates with 95% confidence intervals calculated using the Wilson score method. The horizontal bar and asterisks in panel B indicate the significance of between-group differences (*P < 0.001; χ² test). C-D Box-and-whisker plots depicting distributions of change in SAQ-7 and EQ-5D scores during 12-month follow-up period. Boxes represent the interquartile range (IQR), center lines indicate medians, whiskers denote 1.5 × IQR, and dots indicate outliers

During follow-up, patients with high multimorbidity burden had significantly higher risks of all-cause mortality (HR = 1.61, 95% CI: 1.09–2.39, P = 0.016) and readmission (HR = 1.45, 95% CI: 1.10–1.92, P = 0.007) in Cox regression analyses. Functional recovery and patient-reported outcomes were evaluated separately at 12 months as fixed-time outcomes. Compared with the low-burden group, fewer patients in the high-burden group showed improvement in NYHA class (45.3% vs. 62.0%, P < 0.001), and they had smaller mean improvements in SAQ-7 score (+ 15.1 vs. +22.6, P < 0.001) and EQ-5D index (+ 0.09 vs. +0.14, P < 0.001).

Subgroup analysis of multimorbidity-outcome associations

Stratified analyses revealed consistent associations between multimorbidity burden and increased risk of MACE across predefined clinical subgroups (Fig. 5; Table 4). However, it is important to note that in these subgroup analyses, the interpretation of multimorbidity burden as an exposure variable becomes less straightforward, as the comorbidity used for stratification (e.g., diabetes, CKD, COPD) is also a component of the multimorbidity definition. This potential overlap in defining the exposure and stratification variables must be considered when interpreting the results.

Fig. 5.

Subgroup analyses of major adverse cardiovascular events (MACE) according to multimorbidity burden. Forest plot shows adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) for high-burden versus low-burden multimorbidity groups across predefined clinical subgroups. Each point represents the HR with its 95% CI; numeric values are shown on the right. The vertical dashed line indicates HR = 1. Overall effect estimate (HR = 1.52 [95% CI: 1.28–1.81]) is summarized by dark-blue diamond at the bottom. All subgroup interactions were non-significant (P for interaction > 0.05), indicating a consistent association between multimorbidity burden and post-PCI outcomes across clinical subgroups

Table 4.

Subgroup analysis of major adverse cardiovascular events (MACE) according to multimorbidity burden

Subgroup	High-burden, % (MACE)	Low-burden, % (MACE)	Adjusted HR (95% CI)	P value	P for interaction
Age
< 65 years	24.1	13.2	1.47 (1.01–2.14)	0.042	0.66
≥65 years	34.8	21.1	1.56 (1.19–2.05)	0.001
Sex
Male	28.9	15.9	1.51 (1.12–2.04)	0.006	0.48
Female	34.5	19.8	1.63 (1.08–2.46)	0.019
Diabetes mellitus
Yes	37.5	23.4	1.68 (1.22–2.33)	0.002	0.59
No	26.1	15.5	1.42 (1.02–1.98)	0.037
Chronic kidney disease
Yes	42.0	26.7	1.82 (1.21–2.73)	0.004	0.41
No	27.3	16.2	1.48 (1.13–1.95)	0.005
Hypertension
Yes	31.5	19.4	1.46 (1.08–1.96)	0.014	0.33
No	26.8	18.0	1.32 (0.94–1.85)	0.102
COPD
Yes	40.2	25.0	1.49 (1.05–2.10)	0.028	0.47
No	27.8	16.9	1.45 (1.10–1.91)	0.008
Prior stroke
Yes	38.7	24.2	1.54 (1.03–2.31)	0.037	0.52
No	28.0	17.1	1.43 (1.10–1.86)	0.006
LVEF (%)
< 50	36.8	21.5	1.58 (1.11–2.25)	0.011	0.44
≥50	27.4	16.8	1.46 (1.08–1.97)	0.013
Coronary disease extent
Multivessel	35.7	21.0	1.61 (1.19–2.17)	0.002	0.31
Single-vessel	25.9	16.5	1.34 (0.97–1.85)	0.076

Among older patients (≥ 65 years), the high-burden group had a significantly higher incidence of MACE compared to the low-burden group (34.8% vs. 21.1%, adjusted HR = 1.56, P = 0.001). A similar pattern was observed in younger patients (< 65 years) (24.1% vs. 13.2%, adjusted HR = 1.47, P = 0.042). In sex-based analyses, both males (HR = 1.51, P = 0.006) and females (HR = 1.63, P = 0.019) exhibited significantly higher risk in the high-burden group, with slightly stronger effects in women.

Exploratory subgroup analyses suggested that the association between multimorbidity burden and MACE remained directionally similar across several clinical strata. However, findings within diabetes- and chronic kidney disease-defined subgroups should be interpreted cautiously, because these conditions also contributed to the multimorbidity burden definition and may therefore complicate interpretation of subgroup differences.

Hypertensive patients (HR = 1.46, P = 0.014) had higher MACE rates in the high-burden group, while the difference was less pronounced in normotensive patients (HR = 1.32, P = 0.102). In COPD patients, the high-burden group also had a higher risk (HR = 1.49, P = 0.028), and a similar trend was observed in non-COPD patients (HR = 1.45, P = 0.008).

For prior stroke, high-burden patients exhibited significantly greater MACE incidence (HR = 1.54, P = 0.037). Among patients with LVEF < 50%, the high-burden group had a significantly higher incidence of MACE (HR = 1.58, P = 0.011), and a similar association was seen in those with preserved LVEF (HR = 1.46, P = 0.013). In multivessel disease, the high-burden group had the highest MACE risk (HR = 1.61, P = 0.002), while single-vessel patients showed a nonsignificant trend (HR = 1.34, P = 0.076).

However, the interpretation of multimorbidity burden in these subgroup analyses should be considered with caution, as the exposure variable (multimorbidity) is partly conditioned on the disease defining the subgroup (e.g., CKD or COPD). This could introduce bias, as patients with the defining condition may have inherently different clinical characteristics compared to those without the condition, which may influence the observed associations. These analyses were performed to explore whether multimorbidity burden has a differential impact on outcomes within specific disease subgroups, but they should be interpreted as exploratory and hypothesis-generating.

Sensitivity analysis using the Charlson Comorbidity Index (CCI)

The Charlson Comorbidity Index (CCI) was used as an alternative measure of multimorbidity burden to assess the robustness of the main findings. Patients were stratified into moderate-burden (CCI ≤ 4, n = 689) and high-burden (CCI ≥ 5, n = 549) groups (Table 5). The high-CCI group had a significantly higher incidence of MACE compared to the low-CCI group (29.8% vs. 18.5%, P < 0.001) during a median follow-up of 36 months (Fig. 6). After multivariable adjustment, the association remained significant (adjusted HR = 1.37, 95% CI: 1.08–1.75, P = 0.010), confirming the independent prognostic value of CCI-based multimorbidity classification. Detailed results from the Cox regression analysis are presented in Table 6.

Table 5.

Patient classification according to the Charlson Comorbidity Index (CCI)

Group (definition)	n	% of total (N = 1,238)	Median CCI score (IQR)	3-year MACE rate (%)
Moderate burden (CCI ≤ 4)	689	55.7%	3 (3–4)	18.5%
High-burden (CCI ≥ 5)	549	44.3%	6 (5–7)	29.8%

Fig. 6.

Kaplan-Meier survival curves stratified by Charlson Comorbidity Index (CCI) classification. Kaplan-Meier curves showing the cumulative incidence of major adverse cardiovascular events (MACE) during the 36-month follow-up according to CCI-based multimorbidity burden. Patients with a high CCI burden (CCI ≥5, red line) exhibited a significantly higher cumulative incidence of MACE compared to those with a moderate CCI burden (CCI ≤4, blue line).The log-rank test demonstrated a significant difference between the two groups (P = 0.002)

Table 6.

Cox proportional hazards analysis of MACE according to Charlson Comorbidity Index (CCI) classification

Variable	HR (95% CI)	P value
Unadjusted model
High CCI burden (≥ 5) vs. Moderate CCI (≤ 4)	1.49 (1.16–1.91)	0.002
Adjusted model
High CCI burden (≥ 5) vs. Moderate CCI (≤ 4)	1.37 (1.08–1.75)	0.010
Covariates included in the adjusted model
Age (per 10 years increase)	1.12 (0.95–1.33)	0.18
Male sex	1.09 (0.87–1.36)	0.46
Current smoking	1.21 (0.95–1.54)	0.12
LVEF < 50%	1.45 (1.10–1.90)	0.008
Multivessel disease	1.27 (1.01–1.59)	0.042
Statin use at discharge	0.82 (0.65–1.03)	0.09
Dual antiplatelet therapy at discharge	0.76 (0.59–0.97)	0.028

Results from the CCI-based classification were consistent with those obtained from the disease count method, supporting the robustness of multimorbidity burden as an independent predictor of post-PCI outcomes.

Discussion

In this real-world cohort of patients undergoing PCI, we observed that higher multimorbidity burden was associated with worse clinical and functional outcomes. These findings are broadly consistent with prior studies demonstrating the prognostic importance of comorbidity burden in coronary artery disease and PCI populations. Rather than introducing a novel concept, our study provides additional evidence in a contemporary single-center setting and extends prior work by integrating both disease-count and CCI-based approaches, as well as incorporating patient-reported functional outcomes such as SAQ-7 and EQ-5D. This combined assessment offers a more comprehensive view of post-PCI prognosis in routine clinical practice. Patients with two or more chronic diseases are facing higher risks in different fields. The cumulative 3-year incidence rate of MACE in high-burden group reached 30.4%, while that in the low-burden group was only 17.0%. The all-cause mortality rate was nearly twice than that in the low-burden group (13.6% vs. 7.3%). These patients also lag behind in the functional recovery and quality of life. The improvement in NYHA classification, SAQ-7 and EQ-5D scores is significantly smaller. After multivariate adjustment and sensitivity analysis using CCI, these associations still existed, indicating that assessing multiple diseases serve as a relatively reliable prognostic marker. From these findings, it can be understood that in current PCI practices, multiple diseases exhibit a measurable and clinically significant risk gradient, but traditional cardiovascular predictive indicators cannot fully reflect this.

Another clinically relevant consideration is the coexistence of valvular heart disease in patients undergoing PCI, particularly aortic stenosis in older adults with complex cardiovascular comorbidity. In contemporary practice, some patients may undergo both coronary revascularization and transcatheter aortic valve replacement (TAVR) during the course of their disease. In this context, future coronary re-access after TAVR has become an important procedural consideration, because valve design, implantation depth, and commissural alignment can affect subsequent coronary cannulation. Recent data indicate that improved commissural alignment during TAVR is associated with lower rates of commissure-to-coronary overlap and lower rates of failed coronary access. Accordingly, in patients with combined coronary artery disease and valvular heart disease, especially those at risk of requiring future coronary intervention, the relationship between coronary anatomy, PCI strategy, and potential future coronary re-access deserves attention during longitudinal treatment planning. Although valvular heart disease was not systematically captured in the present dataset, this remains an important area for future research in multimorbid PCI populations [22].

The association between multimorbidity burden and adverse post-PCI outcomes is likely multifactorial. However, because this study was observational and did not evaluate biological mediators, medication adherence, or psychosocial pathways directly, mechanistic explanations should be interpreted cautiously. Our findings primarily support the prognostic relevance of cumulative disease burden rather than specific causal pathways. However, as our study was not designed to evaluate biological mediators directly, these explanations remain conceptual and should be interpreted cautiously. More importantly, our findings highlight the clinical implications of cumulative disease burden. Traditional cardiovascular risk models primarily focus on isolated risk factors or anatomical complexity, but they may underestimate the integrated risk carried by patients with multiple coexisting conditions. The consistent association observed across subgroup and sensitivity analyses suggests that multimorbidity burden provides incremental prognostic information beyond conventional predictors.

Incorporating multimorbidity assessment into routine PCI evaluation may therefore improve risk stratification and inform individualized management strategies. Patients with higher disease burden may benefit from closer surveillance, multidisciplinary follow-up, and personalized rehabilitation planning. In contemporary cardiovascular practice, recognizing multimorbidity as a measurable and independent risk gradient may enhance long-term outcome optimization.

Inflammation plays a critical role in the pathophysiology of atherosclerosis and can contribute to plaque instability. In patients undergoing PCI with stent placement, chronic inflammation is particularly concerning, as it may accelerate neoatherosclerosis and increase the risk of stent restenosis. This is supported by our finding that the high-comorbidity group had significantly higher revascularization rates, suggesting that inflammation may contribute to a need for repeat revascularization due to the development of in-stent restenosis. Chronic inflammatory conditions, such as diabetes, hypertension, and chronic kidney disease, exacerbate this process and are associated with worse outcomes following PCI.

This persistent low-grade inflammation often lead an increase in production of pro-inflammatory cytokines such as IL-6, TNF-α, and CRP. And these pro-inflammatory cytokines activate the NF-κB and JAK-STAT signaling pathways [23–25]. These pathways promote macrophage infiltration, matrix degradation, and lipid core expansion, destabilizing atherosclerotic plaques and increasing susceptibility to plaque rupture and thrombosis after stent implantation [26, 27]. Chronic inflammation can also promote tissue factor expression and platelet hyper-reactivity, contributing to a pro-thrombotic milieu that can diminish the protective effects of dual antiplatelet therapy [27, 28]. Comorbidities such as diabetes and chronic kidney disease promote oxidative stress, impair nitric oxide bioavailability, and upregulate endothelin-1 signaling, leading to endothelial dysfunction, that results in impaired vasodilation, microvascular dysfunction, and delayed myocardial recovery after ischemia [29]. COPD [30] or chronic kidney disease [31] exhibit sympathetic overactivation and RAAS dysregulation, promoting vascular stiffness, sodium retention, and myocardial workload. Metabolic abnormalities can induce insulin resistance, lipid metabolism dysregulations, and adipokine imbalances, all of which exacerbate vascular inflammation and contribute to adverse cardiac remodeling [32]. Concurrent use of multiple therapeutics for different underlying conditions can lead to altered drug bioavailability due to interactions mediated by CYP3A4 and P-gp. These events reduce antiplatelet activity or increase the risk of bleeding, thereby complicating the treatment strategies after PCI and related clinical outcomes [33].

From a clinical perspective, the results support the value of recognizing multimorbidity burden during post-PCI risk assessment. Patients with higher burden may require closer follow-up and more individualized management, although the specific care strategies were not evaluated in this study. Patients with a higher multiple disease burden may not benefit from a standard post-PCI approach. Therefore a specifically tailored care process may reduce risks and optimize the effectiveness of rehabilitation. Multimorbidity screening can be incorporated into routine evaluation alongside conventional indices such as SYNTAX score and LVEF In the pre-procedural stage. Patients with a high-burden profile (≥ 2 major comorbidities or CCI ≥ 5) may benefit from an enhanced prehabilitation program, which includes anemia optimization, glycemic control, renal-dose medication adjustment, and respiratory conditioning where applicable [34]. For high-risk diabetics or CKD patients, using peri-procedural hydration protocols, careful contrast dose selection, or radial access may further reduce complications [35].

After PCI, multimorbidity-stratified management should move beyond a universal DAPT-plus-statin model. For patients with chronic kidney disease or those taking multiple medications simultaneously, pharmacogenetic testing or platelet function testing may improve the selection of antiplatelet drugs, and closer renal monitoring can provide guidance for the initiation of renin-angiotensin blockers and SGLT2 inhibitors. For patients with chronic obstructive pulmonary disease or those who are physically weak, rehabilitation plans supervised by professionals, lung exercises, and nutritional support can help maintain functional abilities to some extent [36]. Multidisciplinary follow-up with cardiology, nephrology, endocrinology, and rehabilitation medicine can be operationalized through scheduled joint clinics or digital care pathways at 1, 3, and 12 months post-PCI.

Digital decision tools and machine-learning-based risk calculators may further enhance care by automatically flagging high-burden individuals, prompting task-based clinician reminders (e.g., renal function check after contrast exposure, deprescribing review, and mood screening) [37]. At the same time, a standardized multimorbidity package should be established, which includes medication coordination, frailty screening, diet and exercise plans, as well as early rehabilitation referrals, to ensure workflow consistency. Implementing multi-disease management approach may promote patient compliance and reduce unplanned readjustments, and improve the quality of life of complex PCI populations [38, 39].

Following limitations of the study need further exploration. It is a retrospective, single-center study, a design which may introduce information bias and restrict generalizability. Despite performing extensive covariate adjustments and propensity score matching, it is hard to completely rule out unmeasured residual confounding factors. The classification of multiple diseases relied on recorded diagnoses, which may not fully capture the severity or trajectory of the disease development. Despite their emerging association in multimorbidity studies, data on lifestyle factors, such as diet, physical activity and social determinants of health, was not systematically collected. Nevertheless, the findings provide meaningful real-world evidence, pinpointing the importance of recognizing multimorbidity in cardiovascular care. With the aging population and increase in multimorbidity at global level, integration of chronic disease management, frailty screening, and rehabilitation programs into routine PCI management is of urgent need. Conducting multicenter prospective studies, integrating refined disease severity indicators, and linking clinical data with public health and social care systems must be prioritized in future. It is also essential to have a national cardiovascular policy that includes a multidisciplinary care model, digital follow-up platforms, and preventive strategies for high-burden populations. Artificial intelligence can assist in risk stratification and coordinate the delivery of chronic disease care in the elderly patient populations. Continuous improvement in this regard hold promise for enhancing the quality of PCI management strategies. The mixed inclusion of stable coronary artery disease and acute coronary syndrome presentations also warrants careful interpretation. Although ACS status, Killip class, and left ventricular ejection fraction were incorporated into the adjusted models, these variables may not fully capture the heterogeneity of acute presentation severity. Moreover, the early divergence of the Kaplan–Meier curves suggests that short-term risk may have contributed importantly to the overall association. Because separate analyses by presentation type and early-versus-late events were not performed in the present study, the extent to which the observed association reflects multimorbidity burden itself versus differences in acute clinical instability cannot be fully determined.

Furthermore, early (e.g., 30-day) MACE outcomes were not separately analyzed. Therefore, the relative contribution of short-term versus long-term risk associated with multimorbidity burden cannot be fully distinguished. Future studies incorporating detailed early-phase event stratification are warranted to better characterize temporal risk patterns.

Second, detailed procedural variables and specific periprocedural complications were not comprehensively available due to the retrospective design. Although major indicators of disease severity and procedural complexity were adjusted for, residual confounding related to lesion characteristics or in-hospital procedural complications cannot be fully excluded.

Our findings suggest that routine assessment of multimorbidity burden may help improve clinical risk stratification after PCI. Future studies should determine how best to integrate multimorbidity into practical care pathways and prospective risk models. These technologies hold potential for developing more personalized treatment plans, identifying patients at highest risk for complications, and guiding post-PCI rehabilitation. However, these approaches are beyond the scope of the present study and require further empirical validation before they can be reliably integrated into clinical practice. Additionally, the use of structured multidisciplinary pathways informed by AI-driven algorithms may provide a more comprehensive approach to managing patients with multimorbidity, ultimately improving clinical outcomes and reducing healthcare costs. Again, this remains an area for future investigation, as empirical data supporting these strategies is limited.

One limitation of the study is the lack of inclusion of key medication variables, such as antiplatelet therapy and statins, in the propensity score matching or adjustment process. These medications are known to impact cardiovascular outcomes and could potentially confound the observed associations between multimorbidity burden and MACE. Future studies with more comprehensive medication data should consider including these factors to better adjust for treatment effects.

Conclusion

In this single-center retrospective cohort, higher multimorbidity burden was associated with increased risks of major adverse cardiovascular events, mortality, readmission, and less favorable functional recovery after PCI. These findings suggest that multimorbidity may provide additional prognostic information beyond conventional clinical factors. However, given the observational design and incomplete capture of some potential confounders, the results should be interpreted cautiously. Further multicenter prospective studies are needed to determine whether routine incorporation of multimorbidity assessment into PCI risk evaluation improves clinical decision-making and long-term outcomes.

Abbreviations

CABG

Coronary artery bypass grafting

CCI

Charlson Comorbidity Index

CI

Confidence interval

CKD

Chronic kidney disease

COPD

Chronic obstructive pulmonary disease

EQ

5D—EuroQol Five Dimensions questionnaire

HR

Hazard ratio

IQR

Interquartile range

KM

Kaplan—Meier

LVEF

Left ventricular ejection fraction

MACE

Major adverse cardiovascular events

MI

Myocardial infarction

NYHA

New York Heart Association

PCI

Percutaneous coronary intervention

PH

Proportional hazards

PSM

Propensity score matching

SAQ

7—Seattle Angina Questionnaire—7

SD

Standard deviation

SMD

Standardized mean difference

STROBE

Strengthening the Reporting of Observational Studies in Epidemiology

Authors’ contributions

All authors contributed to the study conception and design. XFW, XBZ and GFN performed material preparation, data collection and analysis. XFW prepared the first draft of the manuscript. WZ reviewed and edited the manuscript. All authors critically revised the manuscript. All authors read and approved the final version of the manuscript.

Funding

Not applicable.

Data availability

All data generated during the study are presented in the manuscript. Raw data files are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Ethical approval for this retrospective study was granted by the Medical Ethics Committee of Jinan No.5 People’s Hospital (24-1-07). The study was conducted in accordance with the Declaration of Helsinki. Due to the retrospective nature of the study and the use of anonymized routinely collected clinical data, the requirement for written informed consent was waived by the ethics committee.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.