Abstract
Objective
This study sought to examine the associations between four inflammatory biomarkers and 1-year adverse cardiovascular outcomes among hospitalized patients with heart failure with reduced ejection fraction (HFrEF) and to compare their prognostic performance.
Methods
A multicenter retrospective cohort study was conducted involving 1814 hospitalized patients with HFrEF (LVEF <40%). Four inflammation-derived biomarkers were derived from standard blood test results. The relationships between these biomarkers and the 1-year incidence of adverse cardiovascular outcomes were analyzed using Cox proportional hazards regression and restricted cubic spline (RCS) modeling. Event-free survival across biomarker tertiles was compared via Kaplan–Meier survival analysis. The discriminative ability of each biomarker was assessed through receiver operating characteristic (ROC) curves, time-dependent ROC analyses, and C-statistics. Additional sensitivity analyses were performed to verify the stability and reliability of the results.
Results
Elevated levels of all four biomarkers were significantly associated with an increased risk of adverse cardiovascular outcomes in patients with HFrEF. Each 1-SD increase in systemic inflammation response index (SIRI), systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR), and monocyte-to-lymphocyte ratio (MLR) was associated with a 32.5%, 46.1%, 54.0%, and 98.3% higher hazard of adverse cardiovascular outcomes, respectively. Tertile analyses showed that the increased risk of adverse cardiovascular outcomes was mainly observed in the highest biomarker tertiles. Among the four markers, SIRI showed better discriminative performance than the other indices in this cohort, with an AUC of 0.705, which was supported by time-dependent ROC and C-statistic analyses.
Conclusion
Higher SIRI, SII, NLR, and MLR levels were associated with an increased risk of adverse cardiovascular outcomes in patients with HFrEF. SIRI showed numerically better discrimination among hospitalized patients with HFrEF in this cohort, but further external validation is required before clinical application.
Keywords: multicenter retrospective cohort study, heart failure with reduced ejection fraction, inflammatory biomarkers, adverse cardiovascular outcomes
Introduction
Heart failure (HF) remains a major cause of morbidity, mortality, and repeated hospitalization worldwide.1,2 More than 64 million people are estimated to be living with HF globally, and the burden is expected to increase further with population aging and improved survival from cardiovascular disease.3,4 Despite advances in medical therapy, clinical outcomes remain heterogeneous, making early identification of patients at high risk of adverse events an important issue in HF management.
Inflammatory activation is closely involved in the progression of HF. Persistent immune and inflammatory responses may contribute to cardiomyocyte injury, ventricular remodeling, endothelial dysfunction, and worsening cardiac function.5,6 Beyond its direct effects on the myocardium, inflammation is closely implicated in a range of cardiovascular pathologies, including the onset and perpetuation of hypertension, atherosclerotic plaque formation, and plaque rupture.7–10 Traditional methods of assessing systemic inflammation have predominantly relied on individual biomarkers like white blood cell count, C-reactive protein (CRP), and procalcitonin, which frequently prove insufficient to capture the full complexity of the inflammatory state. In recent years, composite inflammatory indices derived from routine complete blood count (CBC) parameters have gained increasing attention owing to their low cost, ease of measurement, and strong reproducibility.11–13 The neutrophil-to-lymphocyte ratio (NLR) and monocyte-to-lymphocyte ratio (MLR) reflect relative changes in circulating leukocyte populations, whereas the systemic immune-inflammation index (SII) and systemic inflammation response index (SIRI) incorporate several blood cell components. Associations between these indices and cardiovascular or systemic diseases have been reported in previous studies.8,11–15 These indices, however, are not specific to HF-related inflammation. In hospitalized patients with HFrEF, these indices may also be influenced by concurrent inflammatory conditions, including infection, which can alter circulating leukocyte profiles and independently affect clinical outcomes.1,16,17 Therefore, elevated CBC-derived inflammatory indices may reflect both HF-associated inflammatory activation and systemic inflammation arising from other clinical conditions. These biomarkers should consequently be interpreted as indicators of overall inflammatory burden rather than direct measures of HF-specific inflammation. Previous studies have linked NLR, SII, and SIRI to adverse outcomes in patients with HF, including acute HF and HFrEF.18,19 However, direct comparisons of multiple CBC-derived inflammatory indices within the same HFrEF cohort remain limited, and their relative prognostic performance is unclear. In addition, whether these indices provide incremental prognostic discrimination beyond conventional clinical characteristics remains unclear. Therefore, this multicenter retrospective cohort study examined the associations of SIRI, SII, NLR, and MLR with 1-year adverse cardiovascular outcomes in hospitalized patients with HFrEF.
Methods
Study Design and Population
This multicenter retrospective cohort study included patients hospitalized with heart failure with reduced ejection fraction (HFrEF) at the Second Hospital of Shanxi Medical University and Changzhi People’s Hospital between January 2021 and December 2024. HFrEF was defined as LVEF <40%. Patients with severe comorbidities, malignant tumors, a hospital stay of <3 days, or incomplete follow-up information were excluded. Among the eligible patients, 79 patients (4.2%) were excluded due to incomplete follow-up information or loss to follow-up. The detailed selection process and numbers of excluded patients are shown in Figure 1. A total of 1814 patients were included in the final analysis.
Figure 1.
Flowchart of patient selection and study population.
Data Collection
A standardized protocol was used to collect baseline clinical data, including age, sex, systolic and diastolic blood pressure, smoking and drinking status, and major comorbidities such as diabetes mellitus (DM), coronary heart disease (CHD), and hyperlipidemia. Detailed definitions are provided in the Supplementary Materials. Laboratory variables included red blood cell, neutrophil, lymphocyte, monocyte, and platelet counts; fasting glucose; alanine aminotransferase (ALT); aspartate aminotransferase (AST); albumin; total cholesterol (TC); triglycerides (TG); high-density lipoprotein cholesterol (HDL-C); low-density lipoprotein cholesterol (LDL-C); blood urea nitrogen (BUN); and B-type natriuretic peptide (BNP). Medication data included antiplatelet agents, β-blockers, diuretics, angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin II receptor blockers (ARBs), calcium channel blockers (CCBs), and statins during hospitalization and at discharge. Echocardiography was performed within 48 hours after admission by experienced sonographers according to standardized institutional protocols. For details, see the Supplementary Materials.
Calculation of Inflammatory Biomarkers
Novel inflammatory biomarkers were calculated based on CBC parameters in the following:16 (1) SIRI = neutrophil count×monocyte count/lymphocyte count; (2) SII = platelet count×neutrophil count/lymphocyte count; (3) NLR = neutrophil count/lymphocyte count; (4) MLR = monocyte count/lymphocyte count. All hematologic parameters are reported in standardized units (×109/L).8,11,13,20
Study Outcomes
The primary endpoint was 1-year adverse cardiovascular outcomes, defined as a composite of HF rehospitalization and cardiovascular death. HF rehospitalization was defined as an unplanned hospitalization due to worsening HF, whereas cardiovascular death was defined as death attributable to a documented cardiovascular cause. Follow-up began at hospital discharge and continued for 12 months. Outcome information was obtained through telephone follow-up and available medical records. HF rehospitalizations were verified against hospital records whenever possible, whereas cardiovascular deaths were identified based on documented cardiovascular causes from medical records and follow-up information; no independent endpoint adjudication committee was used. If both endpoint components occurred, only the first event was counted, and the date of the first event was used as the event date. Patients without adverse cardiovascular outcomes were censored at the date of the last available follow-up or at 365 days after discharge, whichever occurred first.
Statistical Analysis
Continuous variables were expressed as mean ± SD or median (IQR), as appropriate, and categorical variables as n (%). Baseline characteristics were compared between patients with and without adverse cardiovascular outcomes using Student’s t-test or the Mann–Whitney U-test for continuous variables and the χ2-test or Fisher’s exact test for categorical variables, as appropriate. Univariable and multivariable Cox proportional hazards models were used to evaluate the associations between inflammatory biomarkers and adverse cardiovascular outcomes. The proportional hazards assumption was assessed using Schoenfeld residual tests, and no substantial violations were identified (Figure S1). Four models were constructed: Model 1 was unadjusted; Model 2 adjusted for sex, age, BMI, smoking status, and alcohol consumption; Model 3 further adjusted for SBP, DBP, ALT, AST, TC, TG, LDL-C, HDL-C, glucose, and BNP; and Model 4 additionally adjusted for CHD, DM, and hyperlipidemia. Kaplan–Meier curves with Log rank tests and restricted cubic spline analyses were used to assess event-free survival and potential nonlinear associations, respectively. Prognostic discrimination was evaluated using ROC curves, time-dependent ROC analyses, and C-statistics. Time-dependent ROC analyses were performed at 2, 3, 4, 6, 8, 10, and 12 months to assess the dynamic discriminative ability of inflammatory biomarkers during follow-up, whereas conventional ROC curves were generated for the 1-year outcome. Further details are provided in the Supplementary Materials. All analyses were conducted using R version 4.2.2, with a two-sided P < 0.05 considered statistically significant.
Results
Baseline Characteristics of the Study Population
A total of 1814 patients hospitalized with HFrEF (LVEF <40%) at the two centers between January 2021 and December 2024 were included. Patients were stratified according to the occurrence of adverse cardiovascular outcomes within 1 year. During follow-up, 588 patients (32.4%) experienced adverse cardiovascular outcomes. HF rehospitalization occurred in 586 patients (32.3%), and cardiovascular death occurred in 171 patients (9.4%); some patients experienced both endpoint components. As shown in Table 1, the mean age was 60.27 ± 8.50 years, and 73.59% were female. Patients with adverse cardiovascular outcomes had elevated BMI and DBP and were more prone to being current smokers and drinkers. They also exhibited higher levels of ALT, AST, albumin, TC, and BNP, and lower HDL-C. The prevalence of CHD and hyperlipidemia was higher in the adverse cardiovascular outcomes group, who were also more frequently treated with lipid-lowering agents, antiplatelet drugs, diuretics, and ventricular remodeling–targeted therapies. Notably, all novel inflammatory biomarkers were significantly elevated in patients with adverse cardiovascular outcomes, while no significant differences were observed in the remaining variables.
Table 1.
Baseline Characteristics of the Study Population
| Characteristics | Overall | Non-Adverse Cardiovascular Outcomes | Adverse Cardiovascular Outcomes | P-value |
|---|---|---|---|---|
| N | 1814 | 1226 | 588 | |
| Age, (years) | 60.27 ± 8.50 | 60.29 ± 8.43 | 60.22 ± 8.66 | 0.883 |
| Female, n (%) | 1335 (73.59) | 854 (69.66) | 481 (81.80) | <0.001 |
| BMI, kg/m2) | 25.97 ± 4.00 | 25.63 ± 3.97 | 26.66 ± 3.99 | <0.001 |
| SBP, mmHg | 144.74 ± 18.59 | 144.37 ± 18.37 | 145.50 ± 19.04 | 0.228 |
| DBP, mmHg | 88.33 ± 13.87 | 87.81 ± 13.52 | 89.43 ± 14.52 | 0.020 |
| Smoking, n (%) | 274 (15.10) | 173 (14.11) | 101 (17.18) | 0.047 |
| Drinking, n (%) | 216 (11.91) | 125 (10.20) | 91 (15.48) | 0.002 |
| Laboratory data | ||||
| ALT, U/L | 18.19 (13.00–29.83) | 17.00 (12.00–27.00) | 22.00 (14.91–33.00) | <0.001 |
| AST, U/L | 19.00 (15.00–24.21) | 18.05 (15.00–24.00) | 20.00 (16.00–26.02) | <0.001 |
| Albumin, g/L | 40.61 ± 3.55 | 40.33 ± 3.48 | 41.22 ± 3.61 | <0.001 |
| TC, mmol/L | 4.14 ± 0.94 | 4.10 ± 0.92 | 4.20 ± 0.98 | 0.036 |
| TG, mmol/L | 0.67 (0.56–1.65) | 0.65 (0.55–1.37) | 0.76 (0.59–2.13) | <0.001 |
| HDL-C, mmol/L | 1.13 ± 0.29 | 1.16 ± 0.28 | 1.09 ± 0.28 | <0.001 |
| LDL-C, mmol/L | 2.61 (2.12–3.19) | 2.62 (2.11–3.19) | 2.61 (2.14–3.19) | 0.921 |
| BUN, mg/dL | 4.84 (3.98–5.74) | 4.77 (3.90–5.70) | 5.00 (4.10–5.88) | 0.012 |
| Glucose, mmol/L | 5.73 ± 0.91 | 5.75 ± 0.95 | 5.68 ± 0.81 | 0.118 |
| BNP, pg/mL | 363.40 (178.20–789.36) | 324.00 (171.00–723.10) | 398.00 (205.20–867.72) | <0.001 |
| SIRI | 0.82 (0.56–1.20) | 0.75 (0.53–1.03) | 1.16 (0.71–1.82) | <0.001 |
| SII | 469.08 (338.18–659.85) | 431.98 (318.10–583.66) | 598.05 (406.61–849.58) | <0.001 |
| NLR | 1.95 (1.52–2.59) | 1.82 (1.45–2.31) | 2.39 (1.71–3.27) | <0.001 |
| MLR | 0.22 (0.17–0.29) | 0.21 (0.17–0.26) | 0.27 (0.20–0.36) | <0.001 |
| Medical history, n (%) | ||||
| DM | 161 (8.88) | 114 (9.30) | 47 (7.99) | 0.360 |
| Hyperlipidemia | 934 (51.49) | 521 (42.50) | 413 (70.24) | <0.001 |
| CHD | 721 (39.75) | 432 (35.24) | 289 (49.15) | <0.001 |
| Therapeutic drugs, n (%) | ||||
| Statin | 1281 (70.62) | 775 (63.21) | 506 (86.05) | <0.001 |
| Antiplatelet agents | 721 (39.75) | 432 (35.24) | 289 (49.15) | <0.001 |
| Diuretics | 1775 (97.85) | 1196 (97.55) | 579 (98.47) | 0.208 |
| β-blockers | 1569 (86.49) | 1040 (84.83) | 529 (89.97) | 0.003 |
| CCBs | 1043 (57.50) | 578 (47.15) | 465 (79.08) | <0.001 |
| ACEIs/ARBs | 1666 (91.84) | 1122 (91.52) | 544 (92.52) | 0.230 |
Note: Continuous variables are expressed as mean ± SD or median (IQR); categorical variables are presented as n (%).
Abbreviations: BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; ALT, alanine aminotransferase; AST, aspartate aminotransferase; TC, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; BUN, blood urea nitrogen; SIRI, systemic inflammation response index; SII, systemic immune-inflammation index; NLR, neutrophil-to-lymphocyte ratio; MLR, monocyte-to-lymphocyte ratio; BNP, B-type natriuretic peptide; DM, diabetes mellitus; CHD, coronary heart disease; CCBs, calcium channel blockers; ACEIs, angiotensin-converting enzyme inhibitors; ARBs, angiotensin II receptor blockers.
Incidence of Adverse Cardiovascular Outcomes Across Tertiles of Inflammatory Biomarkers
Patients were stratified into tertiles according to the distribution of each inflammatory biomarker, with the following cutoff values: SIRI: T1 < 0.649, T2 = 0.649–1.061, T3 > 1.061; SII: T1 < 380.554, T2 = 380.554–583.762, T3 > 583.762; NLR: T1 < 1.659, T2 = 1.659–2.330, T3 > 2.330; MLR: T1 < 0.191, T2 = 0.191–0.262, T3 > 0.262. As shown in Figure 2, the incidence of adverse cardiovascular outcomes progressively increased across tertiles of all inflammatory biomarkers. As shown in Figure 2, the observed incidence of adverse cardiovascular outcomes increased across tertiles of the four inflammatory biomarkers, with significant overall trends (P for trend <0.001).
Figure 2.
Incidence of adverse cardiovascular outcomes across tertiles of four inflammatory biomarkers in patients with HFrEF. (A) SIRI; (B) SII; (C) NLR; (D) MLR.
Association Between Inflammatory Biomarkers and Adverse Cardiovascular Outcomes
A Multivariable Cox proportional hazards regression analysis was conducted to examine the associations between inflammatory biomarkers and adverse cardiovascular outcomes in patients with HFrEF. Higher levels of all inflammatory biomarkers showed significant positive associations with a greater likelihood of adverse cardiovascular outcomes. When analyzed as categorical variables, the associations were most pronounced in the T3, whereas the estimates for the T2 were weaker and generally did not reach statistical significance (Table 2). Kaplan–Meier analyses similarly showed a higher cumulative incidence of adverse cardiovascular outcomes in the T3 group than in the lower tertiles (Figure 3). Furthermore, RCS analyses showed that the risk of adverse cardiovascular outcomes became more evident at higher biomarker concentrations (Figure 4).
Table 2.
Multivariable Cox Regression Analysis of Inflammatory Biomarkers Associated with Adverse Cardiovascular Outcomes Among Patients with HFrEF
| Adverse Cardiovascular Outcomes | Model 1 | Model 2 | Model 3 | Model 4 |
|---|---|---|---|---|
| HR (95% CI), P-value | HR (95% CI), P-value | HR (95% CI), P-value | HR (95% CI), P-value | |
| SIRI | ||||
| SIRI (per 1-SD increase) | 1.260 (1.220, 1.300), <0.001 | 1.320 (1.270, 1.370), <0.001 | 1.330 (1.270, 1.380), <0.001 | 1.325 (1.270, 1.380), <0.001 |
| SIRI (tertiles) | ||||
| T1 | Reference | Reference | Reference | Reference |
| T2 | 1.211 (0.95, 1.543), 0.122 | 1.240 (0.973, 1.581), 0.081 | 1.225 (0.962, 1.562), 0.102 | 1.223 (0.959, 1.558), 0.104 |
| T3 | 2.709 (2.205, 3.325), <0.001 | 2.834 (2.306, 3.482), <0.001 | 2.868 (2.334, 3.524), <0.001 | 2.862 (2.327, 3.516), <0.001 |
| P for trend | <0.001 | <0.001 | <0.001 | <0.001 |
| SII | ||||
| SII (per 1-SD increase) | 1.427 (1.363, 1.494), <0.001 | 1.431 (1.366, 1.501), <0.001 | 1.462 (1.395, 1.534), <0.001 | 1.461 (1.391, 1.533), <0.001 |
| SII (tertiles) | ||||
| T1 | Reference | Reference | Reference | Reference |
| T2 | 1.166 (0.925, 1.471), 0.186 | 1.166 (0.924, 1.472), 0.189 | 1.193 (0.946, 1.506), 0.135 | 1.195 (0.946, 1.509), 0.134 |
| T3 | 2.443 (1.989, 3.001), <0.001 | 2.457 (2.001, 3.018), <0.001 | 2.553 (2.075, 3.141), <0.001 | 2.551 (2.073, 3.139), <0.001 |
| P for trend | <0.001 | <0.001 | <0.001 | <0.001 |
| NLR | ||||
| NLR (per 1-SD increase) | 1.453 (1.378, 1.532), <0.001 | 1.535 (1.445, 1.631), <0.001 | 1.541 (1.447, 1.639), <0.001 | 1.540 (1.447, 1.639), <0.001 |
| NLR (tertiles) | ||||
| T1 | Reference | Reference | Reference | Reference |
| T2 | 0.994 (0.786, 1.254), 0.950 | 1.030 (0.815, 1.302), 0.802 | 1.033 (0.818, 1.306), 0.784 | 1.031 (0.816, 1.303), 0.799 |
| T3 | 2.306 (1.885, 2.818), <0.001 | 2.398 (1.961, 2.933), <0.001 | 2.417 (1.976, 2.959), <0.001 | 2.414 (1.973, 2.956), <0.001 |
| P for trend | <0.001 | <0.001 | <0.001 | <0.001 |
| MLR | ||||
| MLR (per 1-SD increase) | 1.784 (1.649, 1.933), <0.001 | 1.968 (1.789, 2.166), <0.001 | 1.983 (1.791, 2.195), <0.001 | 1.983 (1.791, 2.194), <0.001 |
| MLR (tertiles) | ||||
| T1 | Reference | Reference | Reference | Reference |
| T2 | 1.131 (0.895, 1.433), 0.273 | 1.146 (0.907, 1.452), 0.105 | 1.110 (0.877, 1.409), 0.284 | 1.110 (0.877, 1.408), 0.283 |
| T3 | 2.494 (2.030, 3.065), <0.001 | 2.585 (2.102, 3.179), <0.001 | 2.573 (2.092, 3.166), <0.001 | 2.572 (2.094, 3.167), <0.001 |
| P for trend | <0.001 | <0.001 | <0.001 | <0.001 |
Notes: Model 1: no covariates were adjusted. Model 2: adjusted for sex, age, BMI, smoking status, and alcohol consumption. Model 3: Model 2 further adjusted for SBP, DBP, ALT, AST, TC, TG, LDL-C, HDL-C, glucose, and BNP. Model 4: further adjusted for CHD, DM, and hyperlipidemia based on Model 3.
Abbreviations: HR, hazard ratio; CI, confidence interval; SD, standard deviation. Abbreviations as in Table 1.
Figure 3.
Kaplan–Meier survival curves for adverse cardiovascular outcomes across tertiles of inflammatory biomarkers in patients with HFrEF. Survival differences were assessed using Log rank tests. (A) SIRI; (B) SII; (C) NLR; (D) MLR.
Figure 4.
Restricted cubic spline analyses of inflammatory biomarkers and adverse cardiovascular outcomes in patients with HFrEF. (A) SIRI; (B) SII; (C) NLR; (D) MLR.
Comparative Evaluation of the Predictive Performance of Inflammatory Biomarkers
ROC curve analyses were performed to assess and compare the discriminative performance of the inflammatory biomarkers for adverse cardiovascular outcomes and its individual components. As shown in Table 3, for the overall adverse cardiovascular outcomes endpoint, SIRI had the highest AUC (0.705), followed by MLR (0.680), NLR (0.668), and SII (0.667). Pairwise DeLong tests showed that SIRI showed significantly higher AUC values than SII, NLR, and MLR (P < 0.01), and the differences remained significant after Holm correction. For rehospitalization, time-dependent ROC analyses showed that the discriminative performance of the four biomarkers varied during follow-up (Figure 5A). In the conventional ROC analysis, SIRI had the highest AUC of 0.665 (95% CI: 0.624–0.705), followed by MLR (AUC: 0.663, 95% CI: 0.622–0.705), NLR (AUC: 0.643, 95% CI: 0.603–0.684), and SII (AUC: 0.637, 95% CI: 0.596–0.678) (Figure 5B). Pairwise DeLong test results are presented in Table S1. For cardiovascular death, time-dependent ROC analyses similarly showed changes in discriminative performance during follow-up (Figure 6A). In the ROC analysis, SIRI showed the highest AUC of 0.775 (95% CI: 0.703–0.846), followed closely by MLR (AUC: 0.772, 95% CI: 0.701–0.842), NLR (AUC: 0.760, 95% CI: 0.690–0.831), and SII (AUC: 0.747, 95% CI: 0.675–0.818) (Figure 6B). Pairwise DeLong test results are presented in Table S2. In the C-statistic analysis based on the fully adjusted Model 4, the addition of SIRI, SII, NLR, and MLR yielded C-statistics of 0.747 (95% CI: 0.726, 0.768), 0.730 (95% CI: 0.711, 0.752), 0.734 (95% CI: 0.716, 0.756), and 0.728 (95% CI: 0.708, 0.746), respectively (Table 4). Overall, SIRI showed the highest discriminative performance for adverse cardiovascular outcomes as well as for HF rehospitalization and cardiovascular death.
Table 3.
ROC Curve Analysis and Pairwise DeLong Comparisons of the Discriminative Performance of Inflammatory Biomarkers for Adverse Cardiovascular Outcomes Among Patients with HFrEF
| Index | Optimal Cut-Off | AUC (95% CI) | Sensitivity | Specificity | PPV | NPV | ΔAUC Compared with SIRI | Z | P value | Holm-Adjusted P value |
|---|---|---|---|---|---|---|---|---|---|---|
| SIRI | 1.284 | 0.705 (0.677, 0.733) | 0.469 | 0.904 | 0.701 | 0.779 | – | – | – | – |
| SII | 621.759 | 0.667 (0.639, 0.694) | 0.476 | 0.796 | 0.529 | 0.759 | 0.038 | 3.750 | <0.001 | <0.001 |
| NLR | 2.28 | 0.668 (0.640, 0.696) | 0.549 | 0.735 | 0.5 | 0.772 | 0.037 | 3.959 | <0.001 | <0.001 |
| MLR | 0.289 | 0.680 (0.652, 0.708) | 0.453 | 0.847 | 0.588 | 0.762 | 0.025 | 2.898 | 0.004 | 0.004 |
Note: ΔAUC was calculated as the AUC of SIRI minus the AUC of the corresponding comparator; positive values indicate better discriminative performance of SIRI.
Abbreviations: AUC, area under the curve; CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value. Abbreviations as in Table 1.
Figure 5.
Time-dependent ROC and conventional ROC curves of four inflammatory biomarkers for predicting HF rehospitalization in patients with HFrEF. (A) Time-dependent ROC analysis; (B) ROC curve analysis.
Figure 6.
Time-dependent ROC and conventional ROC curves of four inflammatory biomarkers for predicting cardiovascular deaths in patients with HFrEF. (A) Time-dependent ROC analysis; (B) ROC curve analysis.
Table 4.
C-Statistic Analysis of Inflammatory Biomarkers for Predicting Adverse Cardiovascular Outcomes in Patients with HFrEF
| Model | C-Statistic (95% CI) | ΔC-Statistic Compared with Model 4 | P value |
|---|---|---|---|
| Model 4 | 0.682 (0.660, 0.704) | Reference | – |
| Model 4+SIRI | 0.747 (0.726, 0.768) | 0.065 | <0.001 |
| Model 4+SII | 0.730 (0.711, 0.752) | 0.048 | <0.001 |
| Model 4+NLR | 0.734 (0.716, 0.756) | 0.052 | <0.001 |
| Model 4+MLR | 0.728 (0.708, 0.746) | 0.046 | <0.001 |
Notes: Model 4: adjusted for Sex, Age, BMI, Smoking, Drinking, SBP, DBP, ALT, AST, TC, TG, LDL-C, HDL-C, Glucose, BNP, CHD, DM, Hyperlipidemia. ΔC-statistic was calculated as the C-statistic of Model 4 plus the corresponding inflammatory biomarker minus that of Model 4 alone; positive values indicate improvement in discrimination after addition of the biomarker.
Abbreviations: Abbreviations as in Table 1.
Sensitivity Analyses
To ensure the robustness of our findings, several sensitivity analyses were performed to account for potential confounding effects of baseline characteristics and comorbid conditions. First, since smoking can exacerbate vascular injury and worsen HF prognosis, all current smokers were excluded, and the associations between inflammatory biomarkers and adverse cardiovascular outcomes were reassessed. The results remained consistent, with elevated levels of all biomarkers still significantly associated with higher cardiovascular risk (Table S3). Second, as alcohol consumption may aggravate disease progression, we repeated the analyses after excluding all drinkers. The associations persisted after excluding drinker (Table S4). Furthermore, as DM may increase susceptibility to infection and other complications, all diabetic patients were excluded. Similar associations were observed after excluding patients with DM (Table S5). Additional adjustment for study center yielded materially unchanged associations (Table S6). E-values indicated that any such hidden confounder would require a strong association with both the exposure and outcome to fully account for the observed associations, suggesting that unmeasured confounding is unlikely to overturn our main conclusions and further supporting the robustness of our findings (Table S7). Sensitivity analyses yielded broadly consistent results, supporting the stability of the observed associations between higher inflammatory biomarker levels and adverse cardiovascular outcomes in patients with HFrEF.
Discussion
The present study assessed the associations of SIRI, SII, NLR, and MLR with 1-year adverse cardiovascular outcomes in hospitalized patients with HFrEF. Higher levels of all indices were associated with increased risk after multivariable adjustment, particularly among patients in the highest tertile. Although MLR showed the largest HR per 1-SD increase, SIRI showed relatively better discrimination in this cohort and improved model performance when added to clinical predictors. Time-dependent ROC, Kaplan–Meier, and RCS analyses consistently supported the prognostic relevance of these inflammatory indices. The stronger association observed for MLR should be interpreted cautiously, as regression estimates reflect relative risk rather than discrimination, which may depend on the integrated contribution of multiple inflammatory components and clinical characteristics.
HF is a cardiovascular disorder arising from compromised cardiac pumping capacity and is characterized by substantial morbidity and mortality, posing a major global public health challenge.1,21 With population aging, its prevalence and hospitalization rates continue to rise, especially in developing and resource-limited regions.4,22,23 HF patients often have multiple comorbidities-such as hypertension, diabetes, COPD, and chronic kidney disease-which increase disease complexity and worsen outcomes.22,24–26 Therefore, identifying reliable biomarkers for early detection of high-risk patients is crucial for improving prognosis and guiding precision treatment.
Inflammation represents an important biological response to tissue injury and pathological stress and plays a critical role in the development and progression of cardiovascular diseases. Chronic inflammatory activation can impair vascular endothelial function and promote cardiovascular complications through sustained immune dysregulation and tissue remodeling.5,27 In addition, prolonged inflammation may disrupt immune homeostasis through persistent activation of cellular and humoral immune responses, contributing to chronic inflammatory state.28–30 Beyond immune dysregulation, persistent inflammation contributes to pathological tissue remodeling, particularly myocardial fibrosis, which plays an important role in HF progression.5,8,30 CBC-derived inflammatory indices, including SIRI, SII, NLR, and MLR, have attracted increasing attention because of their accessibility, computational simplicity, and potential prognostic value.8,11,12,31 Previous studies have suggested that SIRI may serve as a potential prognostic marker in various clinical settings, including osteoporosis, stroke, and sepsis.11,13,20 In the context of HF, chronic inflammatory activation contributes to disease progression through multiple interconnected mechanisms. It promotes cardiomyocyte apoptosis, fibrosis, and adverse ventricular remodeling, leading to impaired cardiac function and accelerated disease progression.32–34 Elevated inflammatory mediators further aggravate myocardial damage and necrosis.35–38 Ciculating inflammatory mediators, including TNF-α, IL-6, IL-1β, and CRP, are elevated in HF and may impair cardiomyocyte survival and function, thereby amplifying myocardial injury.36–38 Inflammatory activation additionally promotes oxidative stress and endothelial dysfunction, worsening coronary microvascular impairment, myocardial ischemia, and metabolic dysregulation.39–41 In acute decompensated HF, elevated inflammatory biomarkers such as CRP and galectin-3 have been associated with disease severity and adverse clinical outcomes, reflecting systemic inflammatory activation and impaired organ perfusion.39,42–44 Sustained inflammatory activation has been associated with HF rehospitalization and mortality, highlighting the prognostic relevance of inflammation in HF.45,46 Beyond direct myocardial injury, inflammation may exert systemic effects through persistent immune activation, further contributing to disease progression in HF.46,47, In addition to inflammatory biomarkers, parameters reflecting perfusion and metabolic status may provide complementary information for HF risk stratification.48,49 Collectively, these inflammatory mechanisms provide a biological basis linking systemic inflammatory burden with HF progression and adverse cardiovascular outcomes.
A strength of this study is the direct comparison of four readily available CBC-derived inflammatory indices in a multicenter cohort of hospitalized patients with HFrEF. Nevertheless, several limitations should be acknowledged. First, the retrospective design may have introduced selection bias and residual confounding. In addition, CBC-derived inflammatory indices may be influenced by concurrent infection or other unrecognized inflammatory conditions, which could not be completely excluded. Second, several established HF prognostic factors, including continuous LVEF, renal function, serum sodium, hemoglobin, atrial fibrillation, HF etiology, NYHA class, and complete guideline-directed medical therapy at discharge, were not consistently available and therefore could not be incorporated into the multivariable models. Third, only baseline CBC measurements were available, preventing assessment of longitudinal changes in inflammatory indices during follow-up. Fourth, outcome information was partly obtained through telephone follow-up and available medical records, which may have introduced potential misclassification. Although the first occurrence of the composite endpoint was used for time-to-event analysis, detailed temporal relationships between changes in inflammatory status and clinical events could not be evaluated. Finally, external validation was not performed, and formal calibration assessment was unavailable, limiting evaluation of generalizability. The relatively high proportion of women and treatment patterns in this cohort may also affect applicability to other populations. Further prospective studies with more comprehensive clinical information and serial biomarker measurements are warranted.
Conclusions
This study found that higher CBC-derived inflammatory indices were associated with an increased risk of 1-year adverse cardiovascular outcomes in hospitalized patients with HFrEF. SIRI showed numerically better discriminative performance and may help identify patients at higher risk, although further external validation is needed before clinical application.
Funding Statement
No funding was received for this article.
Ethics Statement
The study protocol was reviewed and approved by the Ethics Committees of the Second Hospital of Shanxi Medical University (Approval No. 2024-YX-202) and Changzhi People’s Hospital (Approval No. 2024035). Informed consent was exempted given the retrospective nature of the study. Patient data were anonymized and handled confidentially throughout the study. The study was conducted in accordance with the principles of the Declaration of Helsinki.
Disclosure
The authors report no conflicts of interest in this work.
References
- 1.Boorsma EM, Ter Maaten JM, Damman K, et al. Congestion in heart failure: a contemporary look at physiology, diagnosis and treatment. Nat Rev Cardiol. 2020;17(10):641–13. doi: 10.1038/s41569-020-0379-7 [DOI] [PubMed] [Google Scholar]
- 2.Greene SJ, Fonarow GC, Butler J. Risk profiles in heart failure: baseline, residual, worsening, and advanced heart failure risk. Circ Heart Fail. 2020;13(6):e007132. doi: 10.1161/CIRCHEARTFAILURE.120.007132 [DOI] [PubMed] [Google Scholar]
- 3.Mbanze I, Spracklen TF, Jessen N, Damasceno A, Sliwa K. Heart failure in low-income and middle-income countries. Heart. 2025;111(8):341–351. doi: 10.1136/heartjnl-2024-324176 [DOI] [PubMed] [Google Scholar]
- 4.Savarese G, Becher PM, Lund LH, et al. Global burden of heart failure: a comprehensive and updated review of epidemiology. Cardiovasc Res. 2023;118(17):3272–3287. doi: 10.1093/cvr/cvac013 [DOI] [PubMed] [Google Scholar]
- 5.Adamo L, Rocha-Resende C, Prabhu SD, Mann DL. Reappraising the role of inflammation in heart failure. Nat Rev Cardiol. 2020;17(5):269–285. doi: 10.1038/s41569-019-0315-x [DOI] [PubMed] [Google Scholar]
- 6.Thorp EB, Filipp M. Contributions of inflammation to cardiometabolic heart failure with preserved ejection fraction. Annu Rev Pathol. 2025;20(1):143–167. doi: 10.1146/annurev-pathmechdis-111523-023405 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sinha G. Doubt cast on inflammation’s stop signals. Science. 2022;376(6593):565–566. doi: 10.1126/science.abq8310 [DOI] [PubMed] [Google Scholar]
- 8.Shen D, Cai X, Hu J, et al. Inflammatory indices and MAFLD prevalence in hypertensive patients: a large-scale cross-sectional analysis from China. J Inflamm Res. 2025;18:1623–1638. doi: 10.2147/JIR.S503648 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Chen R, Zhang H, Tang B, et al. Macrophages in cardiovascular diseases: molecular mechanisms and therapeutic targets. Signal Transduct Target Ther. 2024;9(1):130. doi: 10.1038/s41392-024-01840-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Libby P, Kobold S. Inflammation: a common contributor to cancer, aging, and cardiovascular diseases-expanding the concept of cardio-oncology. Cardiovasc Res. 2019;115(5):824–829. doi: 10.1093/cvr/cvz058 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ma H, Cai X, Hu J, et al. Association of systemic inflammatory response index with bone mineral density, osteoporosis, and future fracture risk in elderly hypertensive patients. Postgrad Med. 2024;136(4):406–416. doi: 10.1080/00325481.2024.2354158 [DOI] [PubMed] [Google Scholar]
- 12.Tuzimek A, Dziedzic EA, Beck J, Kochman W. Correlations between acute coronary syndrome and novel inflammatory markers (systemic immune-inflammation index, systemic inflammation response index, and aggregate index of systemic inflammation) in patients with and without diabetes or prediabetes. J Inflamm Res. 2024;17:2623–2632. doi: 10.2147/JIR.S454117 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Cai X, Song S, Hu J, et al. Systemic inflammation response index as a predictor of stroke risk in elderly patients with hypertension: a cohort study. J Inflamm Res. 2023;16:4821–4832. doi: 10.2147/JIR.S433190 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Liu X, Li X, Chen Y, et al. Systemic immune-inflammation Index is associated with chronic kidney disease in the US population: insights from NHANES 2007-2018. Front Immunol. 2024;15:1331610. doi: 10.3389/fimmu.2024.1331610 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Huertas-Campos LF, Torres-Pesantes L, Rodriguez-Sagastegui JO, et al. Prognostic performance of the neutrophil-lymphocyte ratio for early complications following sleeve gastrectomy. Metab Target Organ Damage. 2025;5(2):21. doi: 10.20517/mtod.2024.109 [DOI] [Google Scholar]
- 16.Arrigo M, Gayat E, Parenica J, et al. Precipitating factors and 90-day outcome of acute heart failure: a report from the intercontinental GREAT registry. Eur J Heart Fail. 2017;19(2):201–208. doi: 10.1002/ejhf.682 [DOI] [PubMed] [Google Scholar]
- 17.Cho JH, Cho HJ, Lee HY, et al. Neutrophil-lymphocyte ratio in patients with acute heart failure predicts in-hospital and long-term mortality. J Clin Med. 2020;9(2):557. doi: 10.3390/jcm9020557 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Özmen M, Altınkaya O, Saraç İ, et al. Prognostic impact of systemic immune inflammatory index in patients with reduced ejection fraction heart failure. BMC Cardiovasc Disord. 2025;25(1):559. doi: 10.1186/s12872-025-05053-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Xie L, Wang Q, Lu H, et al. The systemic inflammation response index as a significant predictor of short-term adverse outcomes in acute decompensated heart failure patients: a cohort study from Southern China. Front Endocrinol. 2024;15:1444663. doi: 10.3389/fendo.2024.1444663 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Xu T, Song S, Zhu K, et al. Systemic inflammatory response index improves prognostic predictive value in intensive care unit patients with sepsis. Sci Rep. 2025;15(1):1908. doi: 10.1038/s41598-024-81860-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Baman JR, Ahmad FS. Heart Failure. Jama. 2020;324(10):1015. doi: 10.1001/jama.2020.13310 [DOI] [PubMed] [Google Scholar]
- 22.Osmanska J, Jhund PS. Contemporary management of heart failure in the elderly. Drugs Aging. 2019;36(2):137–146. doi: 10.1007/s40266-018-0625-4 [DOI] [PubMed] [Google Scholar]
- 23.Khan MS, Shahid I, Bennis A, Rakisheva A, Metra M, Butler J. Global epidemiology of heart failure. Nat Rev Cardiol. 2024;21(10):717–734. doi: 10.1038/s41569-024-01046-6 [DOI] [PubMed] [Google Scholar]
- 24.Sudevan R. Socioeconomic determinants of heart failure mortality: lessons from Denmark. Eur J Prev Cardiol. 2021;28(1):76–77. doi: 10.1093/eurjpc/zwaa042 [DOI] [PubMed] [Google Scholar]
- 25.Böhm M, Butler J, Coats A, et al. Empagliflozin in resistant hypertension and heart failure with preserved ejection fraction: the EMPEROR-Preserved trial. Eur Heart J. 2025;46(14):1304–1317. doi: 10.1093/eurheartj/ehae938 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Vivodtzev I, Lacasse Y. Stimulating advances in chronic heart failure and COPD. Chest. 2009;136(1):5–6. doi: 10.1378/chest.09-0408 [DOI] [PubMed] [Google Scholar]
- 27.Hall CJ, Wicker SM, Chien AT, et al. Repositioning drugs for inflammatory disease - fishing for new anti-inflammatory agents. Dis Model Mech. 2014;7(9):1069–1081. doi: 10.1242/dmm.016873 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Zarrin AA, Bao K, Lupardus P, Vucic D. Kinase inhibition in autoimmunity and inflammation. Nat Rev Drug Discov. 2021;20(1):39–63. doi: 10.1038/s41573-020-0082-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Krainer J, Siebenhandl S, Weinhäusel A. Systemic autoinflammatory diseases. J Autoimmun. 2020;109:102421. doi: 10.1016/j.jaut.2020.102421 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Behr J, Salisbury ML, Walsh S, et al. The role of inflammation and fibrosis in interstitial lung disease treatment decisions. Am J Respir Crit Care Med. 2024;210(4):392–400. doi: 10.1164/rccm.202401-0048PP [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Fan W, Wei C, Liu Y, et al. The prognostic value of hematologic inflammatory markers in patients with acute coronary syndrome undergoing percutaneous coronary intervention. Clin Appl Thromb Hemost. 2022;28:10760296221146183. doi: 10.1177/10760296221146183 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Mouton AJ, Li X, Hall ME, Hall JE. Obesity, hypertension, and cardiac dysfunction: novel roles of immunometabolism in macrophage activation and inflammation. Circ Res. 2020;126(6):789–806. doi: 10.1161/CIRCRESAHA.119.312321 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Glinton KE, Ma W, Lantz C, et al. Macrophage-produced VEGFC is induced by efferocytosis to ameliorate cardiac injury and inflammation. J Clin Invest. 2022;132(9):e140685. doi: 10.1172/JCI140685 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Hanna A, Frangogiannis NG. Inflammatory cytokines and chemokines as therapeutic targets in heart failure. Cardiovasc Drugs Ther. 2020;34(6):849–863. doi: 10.1007/s10557-020-07071-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Zhao M, Cao N, Gu H, et al. AMPK attenuation of β-adrenergic receptor-induced cardiac injury via phosphorylation of β-arrestin-1-ser330. Circ Res. 2024;135(6):651–667. doi: 10.1161/CIRCRESAHA.124.324762 [DOI] [PubMed] [Google Scholar]
- 36.Li Z, Ding Y, Peng Y, et al. Effects of IL-38 on macrophages and myocardial ischemic injury. Front Immunol. 2022;13:894002. doi: 10.3389/fimmu.2022.894002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Zhang L, Zhao S, Wang Y. Diannexin alleviates myocardial ischemia-reperfusion injury by orchestrating cardiomyocyte oxidative damage, macrophage polarization and fibrotic process by TLR4-NF-kB-mediated inactivation of NLRP3 inflammasome. Int Immunopharmacol. 2024;130:111668. doi: 10.1016/j.intimp.2024.111668 [DOI] [PubMed] [Google Scholar]
- 38.Zhuang L, Zong X, Yang Q, Fan Q, Tao R. Interleukin-34-NF-κB signaling aggravates myocardial ischemic/reperfusion injury by facilitating macrophage recruitment and polarization. EBioMedicine. 2023;95:104744. doi: 10.1016/j.ebiom.2023.104744 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Kinoshita D, Suzuki K, Yuki H, et al. Sex-specific association between perivascular inflammation and plaque vulnerability. Circ Cardiovasc Imag. 2024;17(2):e016178. doi: 10.1161/CIRCIMAGING.123.016178 [DOI] [PubMed] [Google Scholar]
- 40.Nomura CH, Assuncao-Jr AN, Guimarães PO, et al. Association between perivascular inflammation and downstream myocardial perfusion in patients with suspected coronary artery disease. Eur Heart J Cardiovasc Imag. 2020;21(6):599–605. doi: 10.1093/ehjci/jeaa023 [DOI] [PubMed] [Google Scholar]
- 41.Chandrasekaran P, Weiskirchen R. Diabetes mellitus and heart disease. Metab Target Organ Damage. 2024;4(2):18. doi: 10.20517/mtod.2024.15 [DOI] [Google Scholar]
- 42.Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med. 2005;352(16):1685–1695. doi: 10.1056/NEJMra043430 [DOI] [PubMed] [Google Scholar]
- 43.Libby P, Loscalzo J, Ridker PM, et al. Inflammation, immunity, and infection in atherothrombosis: JACC review topic of the week. J Am Coll Cardiol. 2018;72(17):2071–2081. doi: 10.1016/j.jacc.2018.08.1043 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Dai C, Zheng J, Qi L, et al. Chronic stress boosts systemic inflammation and compromises antiviral innate immunity in Carassius gibel. Front Immunol. 2023;14:1105156. doi: 10.3389/fimmu.2023.1105156 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Iwasaki M, Saito J, Zhao H, Sakamoto A, Hirota K, Ma D. Inflammation triggered by SARS-CoV-2 and ACE2 augment drives multiple organ failure of severe COVID-19: molecular mechanisms and implications. Inflammation. 2021;44(1):13–34. doi: 10.1007/s10753-020-01337-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Beal AL, Cerra FB. Multiple organ failure syndrome in the 1990s. Systemic inflammatory response and organ dysfunction. Jama. 1994;271(3):226–233. doi: 10.1001/jama.1994.03510270072043 [DOI] [PubMed] [Google Scholar]
- 47.Dharap SB, Ekhande SV. An observational study of incidence, risk factors & outcome of systemic inflammatory response & organ dysfunction following major trauma. Indian J Med Res. 2017;146(3):346–353. doi: 10.4103/ijmr.IJMR_1538_15 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Tetaj N, Segreti A, Piccirillo F, et al. Diagnostic and prognostic value of arterial blood gas and electrolyte analyses in heart failure. Rev Cardiovasc Med. 2026;27(3):47958. doi: 10.31083/RCM47958 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Ferro A, Segreti A, Tetaj N, et al. Resting and exercise lactate dynamics in heart failure: guiding therapeutic optimization. Biomedicines. 2026;14(4):884. doi: 10.3390/biomedicines14040884 [DOI] [PMC free article] [PubMed] [Google Scholar]