Association between serum hemoglobin and major cardiovascular adverse event in Chinese patients with ST‐segment elevation myocardial infarction after percutaneous coronary intervention

Abstract Background ST‐segment elevation myocardial infarction (STEMI) is a common clinical acute and severe disease, and it is of great significance to evaluate the prognosis of these patients. Hemoglobin levels are associated with a variety of diseases, but studies on Chinese patients with STEMI after percutaneous coronary intervention (PCI) have not been sufficient. Methods This was a secondary analysis based on a prospective cohort study of patients undergoing PCI in Taizhou, Zhejiang, China. We performed multivariable logistic regression to explore the association between the serum hemoglobin and the incidence of major cardiovascular adverse event (MACE) in patients after PCI. We also used a generalized additive model and smooth curve fitting to explain the nonlinear relationship after adjusting the potential confounders. Finally, the heterogeneity among specific groups was examined by subgroup analysis. Results Of all 462 patients enrolled in this study, 118 (25.54%) developed MACE. There was a negative correlation between serum hemoglobin and MACE in all three models (hazard ratio [HR] 0.82, 95% confidence interval [CI 0.72, 0.93], HR 0.86, 95% CI [0.76,0.98], and HR 0.87, 95% CI [0.74,0.98], respectively). In the subgroup analysis, the negative correlation existed between the patients who had myocardial infarction (MI) history (p for interaction = 0.0059) after adjusting covariates. However, no significant differences were found between age and sex groups (p for interaction = 0.1381, 0.4103, respectively). Conclusion Our results indicated that patients who received PCI with low preoperative hemoglobin were more likely to develop MACE, especially if they have already had a history of MI.


| INTRODUC TI ON
ST-segment elevation myocardial infarction (STEMI) is characterized by persistent chest pain, elevated myocardial injury markers, and ST-segment elevation in at least two consecutive leads on electrocardiogram. Worldwide, STEMI is the most common single cause of death, causing an enormous medical burden, while its incidence is still on the rise. 1 With the development of interventional technology, percutaneous coronary intervention (PCI) has brought great clinical benefits to more and more patients with STEMI. 2 At the same time, the study of its complications and risk factors has gained more attention from researchers. 3,4 Previous studies have shown that preprocedural hemoglobin levels are associated with adverse clinical events after PCI, such as in-hospital complications, 5 higher bleeding risk, 6 and poor long-term outcomes. 7 However, these studies had evident regional diversity, and the researches on the Chinese population were still insufficient. In the present study, we selected Chinese people to investigate the relationship between baseline hemoglobin level and major cardiovascular adverse event (MACE) in patients with STEMI after PCI. ship of the original data to the data-dryad website. Therefore, we utilized these data for secondary analysis on a different hypothesis without infringing the authors' rights.

| Data source and study population
This was a prospective cohort study conducted at Taizhou

| Variables
In the initial observation cohort, the clinical records of patients receiving PCI were derived from their clinical information system and stored in their data repository and analysis system. We obtained the variables from the database. In the present study, serum hemoglobin concentration was used as the independent variable, and the dependent variable was whether MACE occurred. MACE was defined as a series of events, including cardiac death, recurrent myocardial infarction (MI) of rented target vessel, cardiogenic shock, and congestive heart failure. The covariates were as follows: age, sex, systolic blood pressure (SBP), heart rate, white blood cells (WBC), platelet, fasting plasma glucose (FBG), high-density lipoproteincholesterol (HDL), low-density lipoprotein-cholesterol (LDL), albumin, creatinine, uric acid, cardiac troponin I (cTnI), creatine kinase MB (CK-MB), hypertension, diabetes mellitus, previous MI, culprit vessels, and killip grade.

| Statistical analysis
All statistical analyses were performed using package R, ver-sion3.4.3 (http://www.r-proje ct.org), and EmpowerStats software (http://empow ersta ts.com). p < 0.05 was considered statistically significant. Continuous variables were presented as mean ± standard deviation or median (maximum, minimum), while categorical variables were reported as a percentage. For continuous variables, a linear regression model was used to calculate the differences between groups. For categorical variables, the Chi-square test was used for analysis. Multivariable logistic regression models were performed to explore the relationship between the independent variable serum hemoglobin and MACE occurrence in patients after PCI. Generalized additive model and smooth curve fitting were tried to explain the nonlinear relationships after adjusting the same covariates. Then, subgroup analyses were used to find the heterogeneity between different groups stratified by age, sex, hypertension, diabetes mellitus, previous MI, albumin, creatinine, killip grade, and culprit vessels.

| Baseline characteristics of participants
The description of basic clinical characteristics, laboratory examinations, and angiographic results are shown in Table 1. Of 462 patients enrolled in the study, 118 developed MACE, and 344 did not during follow-up. 76.62% of the participants are male, and the mean age is 62.97 ± 11.92. Among different groups of age, heart rate, hemoglobin, creatinine, cTnI, and killip grade, MACE occurrence is significantly different, whereas others are not significantly different.

| The association of serum hemoglobin with MACE in patients after PCI
Three multivariate logistic regression models were constructed to investigate the relationship between serum hemoglobin and MACE ( Then, we tried to use generalized additive models and smooth curve fittings to find the visual relationship between hemoglobin and MACE stratified by age and sex, respectively (Figures 1-3).
We also performed subgroup analyses stratified by age, sex, hypertension, diabetes mellitus, previous MI, albumin, creatinine, killip grade, and culprit vessels, and results are shown in Table 3.
The test for interactions was significant for previous MI (p for interaction = 0.0059). We also observed that the negative association between hemoglobin and MACE occurrence was more significant in patients who have had a history of MI (HR = 0.32, 95% CI = 0.14, 0.73). However, the test for interactions was not statistically significant for age, sex, hypertension, diabetes mellitus, albumin, creatinine, killip grade, and culprit vessels (p values for interactions were larger than 0.05).

| DISCUSS ION
In the present study, 462 patients undergoing PCI were included.
The results revealed a significant negative association between serum hemoglobin and MACE occurrence. For every 1 g/dl increase in hemoglobin, the incidence of MACE decreased by 13%.
Additionally, with the increase of hemoglobin, the incidence further decreased in the participants who had MI history. Furthermore, we also used Figures 1-3  (<130 g/L in men and <120 g/L in women). 10 Anemia was common comorbidity in patients receiving PCI, and decreased baseline hemoglobin on admission was correlated with incrementally higher long-term risk for ischemic stroke, mortality, and major bleeding events. [11][12][13] Although the exact mechanism between anemia and MACE is not clear, patients with anemia tend to be older, have a variety of complications, and have poor basic status, which might be important reasons. 14,15 In addition, the underutilization of dual antiplatelets and statins might also be a potential cause of adverse clinical outcomes. 16

TA B L E 3 Subgroup analysis stratified
by age, sex, hypertension, diabetes mellitus, previous MI, albumin, creatinine, killip grade, and culprit vessels mortality in-hospitalization was significantly higher in patients aged ≥65, even when anemia was not present. 7 This also suggests that it is necessary to correct anemia in elderly patients who were performed PCI, which played an important role in improving their prognosis.
Admittedly, the study also has its limitations. First, the number of cases included in this cohort was relatively small, which might affect the statistical efficiency and required further analysis of large multicenter samples. Second, this study was a secondary analysis based on previous studies, and the adjusted variables were limited, so there might be bias caused by other covariables. Finally, some patients with severe cardiovascular disease and other conditions were excluded from the study, so it was not clear whether the results applied to them.

| CON CLUS IONS
In summary, for STEMI patients undergoing PCI, low preoperative hemoglobin was associated with high postoperative MACE incidence. Therefore, anemia should be corrected as far as possible especially in those with previous MI history to improve their prognosis.

DATA AVA I L A B I L I T Y S TAT E M E N T
The dataset used during the study is available at https://doi. org/10.5061/dryad.pf56m.