Prognostic value of red cell distribution width in acute myocardial infarction

Full Text

Introduction

Myocardial infarction (MI) is widely recognized as the predominant clinical manifestation of coronary artery disease (CAD), a leading cause of cardiovascular mortality globally with far-reaching health implications [1]. Despite the well-defined diagnostic criteria and established treatment strategies for acute MI [2], elevated mortality rates persist, emphasizing the crucial importance of identifying high-risk patients for optimal survival outcomes. Accurate risk stratification is thus imperative in the context of acute MI [3, 4].

In the pathogenesis of CAD, inflammation assumes a key role, contributing to plaque instability and rupture [5, 6]. The red blood cell distribution width (RDW) is a metric utilized in routine complete blood counts (CBCs) to measure the size variability among red blood cells [7]. Recent investigations have proposed that inflammatory processes, neurohormonal activity, and activation of the adrenergic system may influence erythrocyte maturation by disrupting the erythrocyte membrane, resulting in elevated RDW [8, 9]. While traditionally utilized for the differential diagnosis of anemia and hematological disorders, RDW has attracted attention for its potential prognostic relevance in individuals with acute MI [7, 10]. Recent studies indicate an adverse prognostic correlation between elevated RDW and various cardiovascular conditions, including stable CAD, heart failure, general population, acute MI, and stroke [6, 7, 10–14]. Despite the diversity in RDW research related to the clinical prognosis of patients with cardiovascular diseases [14–17], the predictive power of RDW in mortality for acute MI patients remains uncertain. The aim of the present study was to investigate the relationship between RDW on admission and long-term mortality within 18 months in patients with acute MI.

Materials and methods

The study was designed as a single-­center prospective observational cohort investigation, conducted at the Vinogradov municipal clinical hospital (Moscow, Russia). All patients aged >18 years admitting with acute MI and undergoing coronary angiography (CAG) < 24 hours after symptom onset from January 1, 2020, to December 31, 2021 were included. We excluded men or women who were with type 3, 4 and type 5 MI as well as those who developed MI during hospitalization. MI was diagnosed by using the Third universal definition of MI [18].

The baseline demographic and clinical characteristics, cardiovascular risk factors and comorbidities, data on physical examination, blood tests and imaging methods (electrocardiography, echocardiography, CAG), and medications during hospitalization were collected. Access 2 Immunoassay System (Beckman Coulter. USA) was used for the measurement of cardiac Troponin I with 99^th percentile upper reference limit (URL) being 0.02 ng/L. Patients with incomplete medical history were not originally included in the dataset. The CBCs, thus including the measurement of RDW and hemoglobin, was performed in all patients at admission using a Siemens ADVIA 2120i hematology analyzer (Siemens Healthcare Diagnostics, Erlangen, Germany). The Global Registry of Acute Coronary Events (GRACE) 2.0 score was used to assess risk stratification of MI patients [19].

The primary outcome was mortality occurred within 18 months after discharge. Mortality was defined as death from any causes that was recorded in patients’ electronic medical records and death registers. In cases where patients were not followed up at our facility, the endpoint was monitored through telephone communication with patients’ relatives. At the study closing date all of follow-up information was available. The study complies with the guidelines of the Declaration of Helsinki and was independently approved by the local Ethics Committee of the Institute of Medicine, Peoples’ Friendship University of Russia. All patients provided written informed consent.

Statistical analysis. The baseline characteristics of all patients were stratified according to the RDW tertiles. Categorical variables were described as frequencies and percentages, while continuous variables were presented using mean, median (Me), and interquartile range (IQR) values. Chi-square test or Fisher’s exact test was employed to compare categorical variables, and the Kruskal–Wallis test was used for continuous variables to compare groups. Correlations between RDW and other parameters were assessed using Spearman’s rank correlation test. Logistic binomial regression was employed to evaluate the independent effects of RDW on clinical outcomes. Univariate logistic regression analysis was utilized to identify associations with mortality, generating odds ratios (OR) and their 95% confidence intervals (CIs). All variables found to be significantly associated with RDW were entered into a multivariate model using a stepwise method. The predictive accuracy of RDW for mortality within 18 months after acute MI was identified by receiver operating characteristics (ROC) curve analysis to measure the sensitivity and specificity of RDW, and the area under the curve (AUC) was calculated. Statistical analysis was performed using SPSS 25.0 (SPSS Inc., Chicago, IL, USA). All analyses with P values  <  0.05 were considered statistically significant, and all reported P values were 2‑sided.

Results and discussion

We identified 577 patients with MI undergoing CAG. The median age of patients was 65 (IQR: 56—74) years, 60.7% were male (n = 350). The median and IQR of RDW were 14.2% and 13.5—15.0%, respectively. The baseline characteristics of patients stratified according to quartiles of RDW are shown in Table 1. Patients in the highest quartile of RDW, compared to the lowest quartile, displayed a higher frequency of cardiovascular risk factors such as arterial hypertension, CAD, previous MI, and more concomitant comorbidities such as previous cerebrovascular accidents, atrial fibrillation, peripheral artery disease, and anemia. The median value of the GRACE score also increased in parallel with RDW values.

Table 1
Baseline characteristics and laboratory findings of the study population stratified according to quartiles of red blood cell distribution width

Characteristics	Quartiles of red blood cell distribution width (RDW), %				P
	< 13.5	13.5—14.2	14.3—15.0	>15.0
Number of patients	127	167	149	137	–
Age, years, Me (IQR)	64 (55; 70)	64 (55; 74)	67 (56; 75)	67 (56.7; 77)	0.119
Men, n (%)	82 (64.6)	104 (62.3)	90 (60.4)	74 (55.2)	0.447
ST elevation, n (%)	58 (45.7)	85 (50.9)	73(49)	56 (41.8)	0.423
Arterial hypertension, n (%)	110 (86.6)	143 (85.6)	141 (94.6)	122 (91)	0.041
CAD, n (%)	45 (35.4)	63 (37.7)	78 (52.3)	76 (56.7)	< 0.001
Previous MI, n (%)	19 (15)	27 (16.2)	38 (25.5)	40 (29.9)	0.005
Prior revascularization, n (%)	15 (11.8)	15 (9.0)	19 (12.8)	23 (17.2)	0.201
Chronic HF, n (%)	7 (5.5)	10 (6.0)	13 (8.7)	10 (7.5)	0.699
Diabetes mellitus, n (%)	22 (17.3)	36 (21.6)	38 (25.5)	30 (22.4)	0.437
Previous cerebrovascular accident, n (%)	2 (1.6)	11 (6.6)	17 (11.4)	11 (8.2)	0.016
Atrial fibrillation, n (%)	8 (6.3)	9 (5.4)	25 (16.8)	20 (14.9)	0.001
CKD, n (%)	7 (5.5)	11 (6.6)	10 (6.7)	14 (10.4)	0.428
Peripheral artery disease, n (%)	1 (0.8)	1 (0.6)	7 (4.7)	9 (6.7)	0.005
Chronic obstructive pulmonary disease, n (%)	12 (9.4)	22 (13.2)	27 (18.1)	22 (16.4)	0.183
Anemia, n (%)	20 (15.7)	26 (15.6)	41 (27.5)	69 (51.5)	< 0.001
Systolic BP, mm Hg, Me (IQR)	140 (120; 160)	135 (120; 150)	140 (120; 158)	137 (118; 160)	0.911
Heart rate, b. p.m, Me (IQR)	74 (66; 88)	76 (68; 86)	78 (68; 90)	77 (69.5; 96)	0.094
Troponin, ng/mL, Me (IQR)	0.38 (0.08; 2.70)	0.39 (0.09; 2.94)	0.37 (0.10; 3.45)	0.34 (0.09; 2.61)	0.917
Hemoglobin, g/L, Me (IQR)	140 (128; 148)	141 (129; 149)	137 (123.5; 146)	124 (101; 138.2)	< 0.001
Creatinine, µmol/Le Me (IQR)	93 (83; 108)	92 (77; 106)	95 (79; 112)	91 (79; 113)	0.504
Non-obstructive CAD, n (%)	16 (12.6)	17 (10.2)	17 (11.4)	16 (11.9)	0.927
LV EF,%, Me (IQR)	46 (42;55)	46 (40.7; 55)	44 (40; 54.7)	44 (38; 50)	0.001
1 vessel CAD, n (%)	20 (15.7)	33 (19.8)	16 (10.7)	17 (12.7)	0.125
2 vessels CAD, n (%)	27 (21.3)	34 (20.4)	41 (27.5)	22 (16.4)	0.146
3 vessels CAD, n (%)	64 (50.4)	83 (49.7)	75 (50.3)	79 (59)	0.356
Percutaneous coronary intervention, n (%)	105 (82.7)	140 (83.8)	115 (77.2)	99 (73.9)	0.121
GRACE score, points, Me (IQR)	113 (95; 132)	116 (96; 139)	124 (100.5; 147)	124 (100; 149.2)	0.009
Medical treatment:
Beta-blockers, n (%)	118 (92.9)	156 (93.4)	138 (92.6)	114 (85.1)	0.044
ACEi/ARBs, n (%)	115 (90.6)	145 (86.8)	133 (89.3)	115 (85.8)	0.608
Aspirin, n (%)	123 (96.9)	162 (97)	140 (94)	123 (91.8)	0.136
P2Y12 inhibitors, n (%)	123 (96.9)	163 (97.6)	146 (98)	134 (100)	0.626
Statins, n (%)	123 (96.9)	161 (96.4)	141 (94.6)	134 (100)	0.617
Anticoagulants, n (%)	27 (21.3)	42 (25.1)	44 (29.5)	39 (29.1)	0.373

Note: ACEi: angiotensin converting enzyme inhibitor; ARB: angiotensin receptor blocker; BP: blood pressure, CAD: coronary artery disease; CKD: chronic kidney disease; GRACE: Global Registry of Acute Coronary Events; HF: heart failure; IQR: interquartile range; LV EF: left ventricular ejection fraction; Me: median, MI: myocardial infarction; RDW: red blood cell distribution width.

RDW levels were correlated well with age, history of CAD, previous MI, previous cerebrovascular accidents, atrial fibrillation, peripheral artery disease and the GRACE risk score. There was a negative correlation between RDW and the left ventricular ejection fraction (LV EF) and hemoglobin. We did not observe any significant correlation between RDW and other parameters (Table 2).

Table 2
Spearman’s correlations analysis between red blood cell distribution width and other parameters

Variables	rho-value	p
Age	0.116	0.005
CAD	0.179	< 0.001
Previous MI	0.159	< 0.001
Previous cerebrovascular accident	0.099	0.018
Atrial fibrillation	0.132	0.001
Peripheral artery disease	0.136	0.001
Hemoglobin	-0.286	< 0.001
LV EF	-0.162	< 0.001
GRACE score	0.16	< 0.001

Note: CAD: coronary artery disease; GRACE: Global Registry of Acute Coronary Events; LV EF: left ventricular ejection fraction; MI: myocardial infarction.

As regards the follow-up, the median RDW value in patients who died during hospitalization and 18‑month follow up was significantly higher than that of survived patients (14.8% vs. 14.2%; p = 0.005, respectively and 15.0% vs. 14.1%; p < 0.001, respectively). After stratifying the entire study population into quartiles of RDW, the incidence of in-hospital death and death during 18 months increased in parallel with RDW quartiles (p  =  0.008 and p < 0.001, respectively) (Table 3).

Univariate logistic analysis demonstrated that RDW was associated with mortality within 18 months in patients with acute MI (OR 1.38; 95% CI, 1.20—1.58; p < 0.001). This association was then confirmed in multivariate analysis (Table 4), which showed that RDW remained independently associated with 18‑month death (adjusted OR, 1.33; 95% CI, 1.12—1.58; p < 0.001). A higher GRACE score and three-­vessel CAD were found to be additional independent predictors of 18‑month death.

Table 3
Incidence of in-hospital and 18‑month death in patients with an acute myocardial infarction, stratified according to quartiles of red blood cell distribution width

Outcome	Quartiles of red blood cell distribution width (RDW),%				P
	< 13.5	13.5—14.2	14.3—15.0	>15.0
In-hospital death, n (%)	4 (3.1)	6 (3.6)	6 (4.0)	15 (11.2)	0.008
18‑month death, n (%)	5 (3.9)	9 (5.4)	20 (13.4)	32 (23.9)	< 0.001

Table 4
Multivariate logistic regression analysis to assess predictors of 18‑month mortality in patients with an acute myocardial infarction

Variables	Univariate Analysis		Multivariate Analysis
	OR (95% CI)	P	OR (95% CI)	P
RDW, %	1.38 (1.20—1.58)	< 0.001	1.33 (1.12—1.58)	0.001
Age, years	1.10 (1.07—1.13)	< 0.001	1.03 (0.99—1.08)	0.167
Female gender	2.87 (1.69—4.86)	< 0.001	1.71 (0.83—3.51)	0.143
CAD history	3.14 (1.81—5.46)	< 0.001	1.27 (0.62—2.58)	0.511
Previous cerebrovascular accident	3.69 (1.78—7.66)	< 0.001	1.64 (0.65—4.15)	0.293
Diabetes mellitus	2.12 (1.22—3.67)	0.008	1.45 (0.69—3.04)	0.327
Atrial fibrillation	2.60 (1.34—5.03)	0.005	1.41 (0.63—3.14)	0.400
CKD	3.10 (1.47—6.50)	0.003	1.01 (0.37—2.77)	0.983
Anemia	3.91 (3.31—6.61)	< 0.001	1.22 (0.60—2.48)	0.577
Killip class ≥2	4.87 (2.86—8.28)	< 0.001	1.03 (0.48—2.17)	0.948
LV EF ≤40%	2.56 (1.43—4.60)	0.002	1.61 (0.79—3.27)	0.188
GRACE score ≥140	9.83 (5.50—17.59)	< 0.001	3.26 (1.22—8.74)	0.019
Three-vessel CAD	4.32 (2.30—8.12)	< 0.001	3.42 (1.57—7.44)	0.002

Note: CAD: coronary artery disease; CI: confidence interval; CKD: chronic kidney disease; GRACE: Global Registry of Acute Coronary Events; LV EF: left ventricular ejection fraction; OR: odds ratio, RDW: red blood cell distribution width.Top of Form

In ROC curves analysis, the AUC value was 0.708 in the evaluation of RDW as a predictor of 18‑month mortality (Figure 1). The optimal cut-off RDW for estimating 18‑month mortality was 14.2, with 78.8% sensitivity and 54.8% specificity (adjusted OR, 4.19; 95% CI, 1.90—9.25; p < 0.001).

Fig. 1. The receiver-­operating characteristic (ROC) curve for red blood cell distribution width for predicting 18‑month mortality (area under curve = 0.708, 95% confidence interval: 0.642—0.775, p < 0.001)

In this study we aimed to establish the relationship between RDW and mortality in patients with acute MI patients. The main findings of this study are as follows: 1) patients with acute MI who died within 18 months had higher RDW on admission; 2) the rate of in-hospital and 18‑month mortality was significantly higher in patients with high RDW than in those with low RDW; 3) RDW on admission was an independent predictor of 18‑month mortality with acute MI.

Several studies previously investigated the role of RDW in predicting adverse outcomes after an acute MI. A high RDW value was found to be a significant predictor of adverse outcomes in patients with acute coronary syndrome (ACS), and especially of both in-hospital and long-term cardiovascular mortality. Uyarel et al. studied 2,506 patients undergoing primary percutaneous coronary intervention (PCI) for ST-segment elevation MI, and showed that patients with elevated RDW (i. e., 16.1%) at admission had higher in-hospital mortality rate compared to those with normal RDW (7.6% vs. 3.6%; P < 0.001) [19, 20]. In a cross-­sectional study, included 3101 patients with acute MI, RDW was a significant risk predictor of in-hospital mortality after adjusting for age, sex, clinical and laboratory variables (tertile 3 (≥14.2%) vs tertile 1 (≤13.2%): hazard ratio [HR] 2.3; 95% CI 1.39–4.01; p for trend < 0.05) predictor of in-hospital mortality [13]. Khaki et al followed 649 patients with acute MI for 6 months and found that the 6‑month mortality rate was significantly higher in patients with high RDW (≥14.6%) than in those with a low RDW ( < 14.6%) (24.3% vs. 7.9%; p < 0.001) [21]. Gul et al. studied 310 patients with non-­ST elevation MI and explored the association between RDW at admission and 3‑year outcome [22]. The overall mortality rate was significantly higher in the high RDW group (>14.0%) comparted to the low RDW group (≤14.0%) (19% vs. 6%, p < 0.001). A significant association was also found between high admission RDW and adjusted risk of cardiovascular mortality (HR, 3.2; 95% CI, 1.3-7.78). Similar to these results, our study also showed a positive correlation between RDW, and 18‑month mortality of acute MI. Survival rates were highest when the RDW was 13.5% and lowest when the RDW was ≥15.0% after adjusting for age, sex, comorbidity, Killip class, LV EF and GRACE score. More specifically, a RDW higher than 14.2% was found to be associated with a more than 4‑fold enhanced risk of 18‑month death after an acute MI. Notably, the association between RDW and 18‑month mortality was confirmed to be independent from other known risk factors of adverse outcome after cardiac ischemia, thus highlighting the valuable role of measuring RDW in this clinical setting.

While recent research has established RDW as a prognostic indicator for cardiovascular disease, the precise mechanistic pathways underlying the association between elevated RDW and adverse clinical outcomes in cardiovascular conditions remain incompletely elucidated. It is plausible that inflammatory and neurohormonal activation play a crucial role in establishing these mechanistic links, as suggested by recent investigations [23, 24]. Inflammatory factors are implicated in the desensitization of bone marrow erythroid progenitor cells, hindering the antiapoptotic and promaturation effects of erythropoietin. This process leads to an influx of reticulocytes into the peripheral blood, contributing to an elevation in RDW [25]. Additionally, it induces alterations in the lipid structure of red blood cells and compromises their degeneration capacity. Consequently, there is a reduction in the oxygen-­carrying capacity of red blood cells and an increase in whole blood viscosity, thereby heightening the mortality risk in patients with MI [11, 16].

Our findings revealed a noteworthy positive correlation between RDW, age, and cardiovascular disease. Elevated RDW values have consistently been associated with adverse outcomes, including mortality and complications, in various chronic and prevalent conditions such as coronary artery disease [14], ischemic cerebrovascular disease [15], atrial fibrillation [26], peripheral artery disease [27], and anemia [28]. The higher RDW levels observed in older patients can be attributed to the increased prevalence of anemia, inflammatory burden, comorbidities, poor nutritional status, and age-associated diseases in this demographic [11, 22]. This observation contributes to a better understanding of the potential linkage between RDW and 18‑month mortality among patients with acute MI.

The GRACE score’s prognostic significance in ACS aligns with current guidelines [2, 29]. Our study establishes a significant correlation between RDW and the GRACE score, particularly for in-hospital and 18‑month mortality. When combined with three-­vessel CAD, this association enhances prognostic insights, especially in emergency outpatient settings. This dual assessment holds potential for guiding decisions on patient transfer and treatment strategies.

Our study has several limitations. The single-­center design and a relatively small sample size may limit the generalizability of our findings. We did not concurrently assess factors that influence RDW, such as inflammatory markers, malnutrition indicators, and natriuretic peptides, potentially introducing confounding variables. Additionally, RDW was measured only once, preventing an exploration of temporal variations during hospitalization. Future studies with serial measurements are needed to assess the incremental prognostic value of such an approach. Furthermore, our study’s exclusive focus on hemoglobin levels overlooks potential influences on RDW from deficiencies in iron, vitamin B12, or folate. Comprehensive assessments of these factors are essential for a more nuanced interpretation of RDW values.

Conclusion

RDW showed a correlation with traditional cardiovascular factors. Higher RDW at admission was independently associated with an elevated risk of 18‑month mortality in acute MI patients, underscoring its potential as a prognostic marker in this population.

About the authors

Truong Huy Hoang

Pham Ngoc Thach University of Medicine; Tam Duc Heart Hospital

Author for correspondence.
Email: truonghh@pnt.edu.vn
ORCID iD: 0000-0002-2013-2647
SPIN-code: 5684-0275
Ho Chi Minh, Vietnam

Victor V. Maiskov

RUDN University; Vinogradov Municipal Clinical Hospital

Email: truonghh@pnt.edu.vn
ORCID iD: 0009-0002-2135-2606
SPIN-code: 7370-7545
Moscow, Russian Federation

Imad A. Merai

RUDN University; Vinogradov Municipal Clinical Hospital

Email: truonghh@pnt.edu.vn
ORCID iD: 0000-0001-6818-8845
SPIN-code: 4477-7559
Moscow, Russian Federation

Zhanna D. Kobalava

RUDN University; Vinogradov Municipal Clinical Hospital

Email: truonghh@pnt.edu.vn
ORCID iD: 0000-0002-5873-1768
SPIN-code: 9828-5409
Moscow, Russian Federation

References

Supplementary files