Association of red blood cell distribution width-platelet ratio with mortality after coronary artery bypass grafting

Introduction

Coronary heart disease is common worldwide and the long-term therapeutic effect of coronary artery bypass grafting (CABG) has been confirmed. However, myocardial injury may occur in nearly a quarter of patients after CABG because of diverse mechanisms (Weidenmann et al., 2020). Postoperative adverse events following CABG still need concern and attention. Some research has reported the correlation between inflammation and immune-related indicators and adverse prognosis in cardiovascular diseases (Leiherer et al., 2016; Lin et al., 2024; Wasilewski et al., 2018). The red blood cell distribution width (RDW) has traditionally been utilized to diagnose anemia (Arkew et al., 2022). Previous studies have revealed that RDW has a strong association with inflammation, which is a vital contributor to coronary heart disease and can lead to plaque instability (Reed, Rossi & Cannon, 2017). The activation of the inflammatory response may affect the normal maturation of red blood cells by interfering with the erythrocyte membrane, leading to increased RDW (Lin et al., 2021; Peng, Sasmita & Luo, 2024). RDW has been identified as an independent prognostic indicator for individuals with myocardial infarction (MI) (Huang et al., 2021; Peng, Sasmita & Luo, 2024). Besides, platelets are involved in thrombosis and play a crucial role in the regulation of hemostasis and coagulation (Pezeshkpoor, Oldenburg & Pavlova, 2022). Abnormal platelet levels are associated with a higher occurrence of major adverse cardiovascular events in patients with acute coronary syndrome (Dahlen et al., 2021; Ito et al., 2018; Roh et al., 2020). Platelets are excessively activated and consumed under inflammatory conditions (Setarehaseman, Mohammadi & Maitta, 2025). Ischemia-reperfusion injury, swelling and hypoxia of microvascular endothelial cells and myocardial cells, microvascular spasm and embolism, the activation and release of inflammatory factors due to oxidative stress, and other factors are potential mechanisms for postoperative adverse events following CABG surgery (Chi et al., 2017; Tang et al., 2024). Elevated RDW and diminished platelet levels imply the concurrent presence of an inflammatory response and immune system dysfunction (Setarehaseman, Mohammadi & Maitta, 2025). Furthermore, recent research indicates that employing the ratio of RDW to platelet as a prognostic indicator can markedly enhance prognostic value when compared to utilizing each indicator in isolation (Li et al., 2019). The red blood cell distribution width-platelets ratio (RPR), as a combination parameter, has been reported as a new indicator to reflect the severity of inflammation (Tang et al., 2024). It has been shown to have a strong association with the prognosis of cancer, ischemic stroke, heart failure, and hepatic diseases (Dallio et al., 2024; Lin et al., 2024; Takeuchi et al., 2017; Xu & Peng, 2023). The role of RPR on the prognosis of patients after CABG still remains unclear gaps to elucidate. Thus, this study aims to explore the association between RPR with mortality in patients after CABG and provide a theoretical basis for predicting clinical outcomes after CABG.

Materials & Methods

Results

Association between RPR and in-hospital mortality

Univariate logistic regression analyses showed that platelet (odds ratio [OR] 0.361, 95% CI [0.236–0.553] per standard deviation [SD] increase, p < 0.001), RDW (OR 1.539, 95% CI [1.217–1.945] per SD increase, p < 0.001), RPR (OR 1.741, 95% CI [1.485–2.041] per SD increase, p < 0.001), BNP (OR 1.290, 95% CI [1.006–1.655] per SD increase, p = 0.045), LVEF (OR 0.708, 95% CI [0.519–0.966] per SD increase, p = 0.029), previous stroke (OR 2.520, 95% CI [1.248–5.088], p = 0.010), and EuroSCORE (OR 1.360, 95% CI [1.019–1.815] per SD increase, p = 0.037) were accountable for in-hospital mortality. The details of factors selected by univariable logistic regression analyses were presented in the Supplemental Information 1. Moreover, RDW (OR = 1.354, 95% CI [1.010–1.815] per SD increase, p = 0.042) and RPR (OR 1.493, 95% CI [1.119–1.992] per SD increase, p = 0.006) were independent risk predictors by multivariate logistic regression analysis (Table 2). As shown in Fig. 1, the Lowess curve showed an approximate positive linear relationship between RPR and in-hospital mortality after CABG.

Table 2:

Logistic regression assessing risk factors for in-hospital mortality.

Variables	Univariate analysis		Multivariate analysis
	OR (95% CI)	p value	OR (95% CI)	p value
Platelet	0.361 (0.236, 0.553)	<0.001	0.724 (0.409, 1.281)	0.267
RDW	1.539 (1.217, 1.945)	<0.001	1.354 (1.010, 1.815)	0.042
RPR	1.741 (1.485, 2.041)	<0.001	1.493 (1.119, 1.992)	0.006
BNP	1.290 (1.006, 1.655)	0.045	1.228 (0.939, 1.607)	0.134
LVEF	0.708 (0.519, 0.966)	0.029	0.724 (0.520, 1.009)	0.057
EuroSCORE	1.360 (1.019, 1.815)	0.037	1.280 (0.937, 1.748)	0.121
Previous stroke	2.520 (1.248, 5.088)	0.010	2.141 (0.987, 4.643)	0.054

DOI: 10.7717/peerj.19472/table-2

Notes:

Abbreviations

BNP

B-type natriuretic peptide

CI

confidence interval

LVEF

left ventricular ejection fraction

OR

odds ratio

RDW

red blood cell distribution width

RPR

red blood cell distribution width-platelet ratio

WBC

white blood cell

ROC analysis

ROC analyses were performed according to the selected predictors by multivariable logistic regression analysis. RPR (AUC 0.716, 95% CI [0.617–0.814]) showed better discrimination than RDW (AUC 0.578, 95% CI [0.477–0.680], p = 0.002) for the in-hospital mortality among patients after CABG (Fig. 2). The cut-off value (0.107) of RPR was calculated for further analysis, and groups were further divided into the high RPR group (≥ 0.107) and the low RPR group (< 0.107). The details were depicted in Table 3. The incidence of in-hospital and one-year mortality were both significantly higher in the high RPR group (13.02%, p < 0.001; 18.60%, p < 0.001), as compared to the low RPR group (1.63%; 1.82%).

Table 3:

Baseline characteristics between high and low RPR groups.

Variables	High RPR (n = 215)	Low RPR (n = 1043)	χ2/Z	p value
Age (years)	63 (58, 66)	62 (57, 66)	−1.056	0.291
Gender, male, n (%)	136 (63.26)	580 (55.61)	4.251	0.039
BMI (kg/m2)	26.13 (21.91, 30.31)	25.84 (21.90, 29.58)	−0.719	0.472
MAP (mm Hg)	98 (86.67, 108.33)	97.33 (86.67, 108)	−0.436	0.663
NYHA, n (%)			3.462	0.326
I	1 (0.47)	12 (1.15)
II	146 (67.91)	759 (72.77)
III	61 (28.37)	242 (23.20)
IV	7 (3.26)	30 (2.88)
Previous MI, n (%)	108 (50.23)	552 (52.92)	0.518	0.472
Previous PCI, n (%)	50 (23.26)	278 (26.65)	1.068	0.301
Previous stroke, n (%)	28 (13.02)	121 (11.60)	0.345	0.557
Number of stenosed arteries	3 (3, 3)	3 (3, 3)	−0.125	0.901
Number of grafts	4 (3, 4)	4 (3, 4)	−0.188	0.851
SYNTAX score	31 (26, 35)	31 (25, 35)	−0.067	0.946
EuroSCORE	4 (3, 6)	4 (3, 5)	−0.225	0.822
Cardiopulmonary bypass, n (%)	87 (40.47)	449 (43.05)	0.487	0.485
WBC (K/μl)	6.9 (5.8, 8.1)	7 (5.8, 8.1)	−0.372	0.710
Platelet (K/μl)	112 (82, 126)	211 (179, 266)	−22.232	<0.001
RDW (%)	15.2 (14.1, 17.1)	14.3 (13.5, 15.7)	−5.899	<0.001
Hemoglobin (g/L)	138 (133, 141)	138 (134, 142)	−1.599	0.110
RPR	0.14 (0.12, 0.18)	0.07 (0.05, 0.08)	−23.116	<0.001
Creatinine (umol/L)	65 (59, 71)	63 (58, 69)	−2.131	0.033
BUN (mmol/L)	7 (6.1, 7.8)	6.6 (5.8, 7.6)	−2.758	0.006
Albumin (g/L)	43 (39, 46)	42 (38, 47)	−0.637	0.524
CK-MB (U/L)	12 (10, 15)	13 (10, 15)	−0.781	0.435
Troponin I (ng/ml)	0.37 (0.06, 1.98)	0.27 (0.06, 0.98)	−0.741	0.459
BNP (pg/ml)	295 (112, 754)	262 (97, 706)	−1.094	0.274
LVEF (%)	45 (39, 51)	46 (40, 51)	−0.825	0.409
Mild to moderate MR, n (%)	39 (18.14)	157 (15.05)	1.291	0.256
Mild to moderate AR, n (%)	14 (6.51)	83 (7.96)	0.524	0.469
Mild to moderate TR, n (%)	9 (4.19)	40 (3.84)	0.059	0.809
Mild to moderate PAH, n (%)	19 (8.84)	113 (10.83)	0.757	0.384
Peripheral artery disease, n (%)	44 (20.47)	191 (18.31)	0.544	0.461
Hyperlipemia, n (%)	20 (9.30)	108 (10.35)	0.216	0.642
COPD, n (%)	29 (13.49)	142 (13.61)	0.002	0.961
Diabetes, n (%)	100 (46.51)	498 (47.75)	0.109	0.741
In-hospital mortality, n (%)	28 (13.02)	17 (1.63)	67.086	<0.001
One-year mortality, n (%)	40 (18.60)	19 (1.82)	112.323	<0.001

DOI: 10.7717/peerj.19472/table-3

Notes:

Abbreviations

AR

aortic regurgitation

BMI

body mass index

BNP

B-type natriuretic peptide

BUN

blood urea nitrogen

CK-MB

creatine kinase-myocardial band

COPD

chronic obstructive pulmonary disease

LVEF

left ventricular ejection fraction

MAP

mean arterial pressure

MI

myocardial infarction

MR

mitral regurgitation

NYHA

New York Heart Association

PAH

pulmonary arterial hypertension

PCI

percutaneous coronary intervention

RDW

red blood cell distribution width

RPR

red blood cell distribution width-platelet ratio

TR

tricuspid regurgitation

Survival and subgroup analysis

In the multivariable logistic regression model, high RPR correlated with elevated risks of in-hospital mortality (OR 6.097, 95% CI [2.308–16.104], p < 0.001) and one-year mortality (OR 6.395, 95% CI [2.610–15.666], p < 0.001) after adjusting for all included covariates. As depicted in the Kaplan–Meier survival curve, patients with low RPR show better one-year survival than those with high RPR (Fig. 3). As depicted in Figs. 4 and 5, subgroup analyses revealed that high RPR consistently had increased risks of in-hospital mortality and one-year mortality in each subgroup.

Discussion

Currently, few studies have discussed the relationship between RPR and in-hospital mortality in patients following CABG. ROC curves demonstrated better discrimination of RPR than RDW for in-hospital mortality when RPR was analyzed as a continuous variable. For patients undergoing CABG, we have calculated the cut-off value of RPR and revealed that high RPR was an independent risk factor for in-hospital mortality and one-year mortality. Moreover, Kaplan–Meier survival curves showed the better prognosis of the low RPR group, which verified that the cut-off value of RPR can be used to stratify patients’ short-term and mid-term clinical outcomes. Univariable and multivariable logistic regression analyses and subgroup analyses were performed in order to provide a reliable conclusion for the association of RPR and the prognosis of patients with CABG.

RDW has historically been used to diagnose anemia (Arkew et al., 2022). Recent evidence indicates that elevated RDW is significantly linked to poor prognosis in various diseases (Arkew et al., 2022). In our study, high RDW is also an independent risk factor for mortality, which is consistent with previous studies (Bozorgi et al., 2016a; Bozorgi et al., 2016b; Wasilewski et al., 2018). RDW has been confirmed to be significantly associated with a variety of inflammatory biomarkers and cytokines (Wang et al., 2021). The inflammatory response has negative impacts on bone marrow function, as well as disturbing erythropoietin generation and inhibiting iron metabolism (Wang et al., 2021; Wang & Hsu, 2021). Oxidative stress alters the half-life of red blood cells and releases more premature red blood cells into the peripheral circulation, subsequently leading to an increase in red blood cell heterogeneity (Bujak et al., 2015; Wang & Hsu, 2021). In addition, surgical revascularization can cause an inflammatory response and activation of the adrenergic system and thus affect the maturation of red blood cells, leading to increased RDW (Yao et al., 2023). A meta-analysis clarified that increased RDW could indicate an adverse prognosis in heart failure patients (Shao et al., 2015). Likewise, in an adult congenital heart disease cohort, RDW has been shown to be remarkably associated with cardiovascular events, independently of N-terminal pro-BNP and other prognostic biomarkers (Baggen et al., 2018). In addition, Hou et al. (2018) found that RDW has a positive correlation with frailty in elderly patients. Frailty also serves as a risk factor for mortality and hospital readmissions in heart failure (Zhang et al., 2018).

Platelet-dependent thrombosis is a key factor in the progression of coronary heart disease. Decreased platelet count is associated with increased risk of death, infarct size, and reinfarction rates (Dahlen et al., 2021; Ito et al., 2018; Roh et al., 2020). Low preoperative platelet levels may potentially increase the likelihood of postoperative bleeding (Gunertem et al., 2021). Consequently, this elevated risk of postoperative bleeding can lead to a higher mortality rate after surgery (Ezelsoy et al., 2020). In clinical practice, aggressive antiplatelet or anticoagulant treatment may increase the risk of hemorrhagic stroke (Xu & Peng, 2023). On the contrary, several studies have also shown that high platelet levels are also associated with adverse clinical outcomes (Malyszczak et al., 2020). Platelets mediate and initiate thrombotic occlusion of the entire coronary artery and aggregate in the circulation, resulting in impaired microcirculation and inducing myocardial ischemia during reperfusion (Roh et al., 2020). In this study, a univariate logistic regression analysis indicated that platelet was a risk factor. However, after using multivariate logistic regression analysis to correct other confounding factors, no significant association was found between platelet levels and in-hospital mortality for patients after CABG. We considered that RPR, a factor with more predictive value, was included in the multivariate logistic regression model, which was the main reason for the insignificant difference in the relationship between platelets and in-hospital mortality. Besides, there are still no unified and consistent criteria for the threshold of platelets in predicting clinical prognosis among published articles. Large cohort studies are required to further detect its cut-off value on prognosis.

Based on the above factors, this study mainly focuses on the ratio of RDW and platelet-RPR. RDW and platelets as readily gained indicators from routine laboratory examinations, are commonly used in clinical practice, which can be conveniently and dynamically monitored on the advantage of the rapidity and low cost. Besides, RDW is also an independent risk predictor of in-hospital mortality after CABG in the multivariate regression model. ROC analysis revealed the discrimination of RPR was significantly better than that of RDW, confirming that RPR is a better predictive indicator. On the basis of current research combined with our results, the cut-off value of RPR can clarify the risk stratification on the short-term and mid-term prognosis, and high RPR implies a poor prognosis for patients undergoing CABG. The cut-off value of preoperative RPR, instead of RDW or platelet, may provide better guidance for the assessment of patients’ states and clinical outcomes, as well as timely adjust the treatment strategy. The predicting value of RPR on early prognosis needs to be paid more attention to. The cut-off value of RPR needs external validation by using multicenter data to ensure its generalizability and clinical applicability. This study identified the prognostic value of preoperative RPR on mortality after CABG, and the underlying biological mechanism by which RPR affects patients’ prognosis remains to be elucidated in subsequent studies.

There exist several limitations in this study. This is a single-center retrospective study. Data from other medical centers are needed to further ensure the robustness and generalizability of our results. Potential bias and undetected factors might exist although univariable and multivariable logistic regression, and subgroup analyses were used to adjust the confounding factors. Only the relationship between RPR and mortality can be inferred on account of the nature of the retrospective study design. Randomized controlled trials or prospective cohort studies are needed to further explore the causal association between RPR and mortality for patients after CABG. Besides, the relationship between RPR levels measured at various time intervals and clinical outcomes also deserves additional investigation. Future studies should explore whether changes in RPR measurements over time are associated with disease progression or the effectiveness of treatment.

Conclusions

In conclusion, preoperative high RPR could serve as an independent risk predictor for in-hospital mortality and one-year mortality, which can be utilized to assess the prognosis of patients and can further provide guidance for the treatment in patients following CABG. Further studies are necessary to discover the mechanism of RPR on prognosis for patients undergoing CABG.

Supplemental Information