Relationship Between Prognostic Nutrition Index and New York Heart Association Classification in Patients with Coronary Heart Disease: A RCSCD-TCM Study

Introduction

Cardiovascular diseases (CVD) are the leading causes of death and disease burden globally.^1–3 The incidence rate increased from 271 million in 1990 to 523 million in 2019, and the death toll increased steadily from 12.1 million in 1990 to 18.6 million in 2019, making it an important public health challenge.^4,5 Atherosclerosis (AS) is the major cause of coronary heart disease (CHD) and other CVD. In 2016, approximately 2.4 million people died of atherosclerotic cardiovascular disease (ASCVD), accounting for 61% of all CVD deaths.^6,7 Malnutrition is usually caused by insufficient or unbalanced nutrient intake, leading to insufficient calories, protein, or other nutrients needed for tissue maintenance and repair. Currently, malnutrition affects every stage of life and is a global health problem affecting more than 1 billion people. However,⁸ to a large extent, clinicians understand the insufficiency and shortage of treatment.⁹ Malnutrition is common in patients with CVD, and various physiological changes and psychological pressures caused by CVD affect the nutritional status of individuals, especially in some acute diseases, which often lead to nutritional imbalance. Therefore, rapid assessment and improvement in nutritional status play a vital role in the treatment and prognosis of patients with CVD.¹⁰ Prognostic nutritional index (PNI) is a new nutritional immune inflammation index determined by analyzing the complex linear prediction model of serum albumin level, transferrin, delayed hypersensitivity, and triceps skinfold thickness.¹¹ PNI has been used to evaluate the nutritional status of patients with various diseases and is easy to calculate. Recently,^12,13 studies have shown that PNI is independently related to the long-term survival of hospitalized patients with severely decompensated acute heart failure,^14,15 and it is an important independent predictor of coronary collateral development and clinical outcome in patients with acute ST-segment elevation myocardial infarction (STEMI) undergoing primary percutaneous coronary intervention.¹⁶ Recent research has shown that PNI is inversely and significantly associated with the development of contrast-associated acute kidney injury in acute STEMI. Assessing PNI at admission may be useful for the early risk stratification of patients with STEMI.¹⁷ Another study showed that a lower PNI value is independently related to the in-hospital mortality of acute Stanford type A aortic dissection (ATAAD) and may be a useful tool for predicting the early mortality of ATAAD patients after surgical repair.^18,19 However, malnutrition assessed by PNI is manifested as being underweight or overweight. PNI combined with body mass index (BMI) may better reflect the nutritional status of the patients. However, few studies have explored the relationship between PNI and New York Heart Association (NYHA) classification, and data are particularly scarce for people with CHD. Therefore, we aimed to elucidate the relationship between the PNI of patients with CHD and NYHA classification and the complex relationship between the PNI combined with the BMI and NYHA classification.

Methods

Participants

A total of 107,301 patients with CHD diagnosed in six tertiary hospitals in Tianjin, China, from January 1, 2014, to September 30, 2020, were selected as the study participants. Patients with CHD were identified with at least one or more of the International Classification of Disease codes obtained from the hospital’s diagnosis and treatment system.

Patients below 35 or above 75 years, patients with severe liver and renal failure, infection, tumor, and cancer, and patients without clear NYHA classification and lack of PNI were included in this study. Finally, 17,413 patients were enrolled in the study. A flow chart of case screening is shown in Figure 1. This study was conducted in accordance with the principles of the Declaration of Helsinki. The registration number of the China Clinical Trials Register is ChiCTR1900024535, and the registration number of American Clinical Trials.gov is NCT04026724. This study was approved by the Ethics Committee of Tianjin University of Traditional Chinese Medicine. The approval number for the reviewed institutions is TJUTCM-EC2019008.

Figure 1 The flow chart of case screening.

Data Collection

We collected the clinical data of patients’ hospitalization, including demographics, such as age, sex, ethnicity, smoking, and drinking history, from the hospital diagnosis and treatment system. Basic demographic data were collected and recorded by the hospital doctors through standard structured questionnaires 4 h after admission. The NYHA classification was evaluated by a doctor within 24 h after admission. Fasting blood samples were collected from the study center in the morning.²⁰ Lymphocytes (LYM) were measured using an automatic hematology analyzer. Albumin (ALB), total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), fasting blood glucose (FBG), glycated hemoglobin A1c (HbA1c), and lactate dehydrogenase (LDH) levels were measured using an automatic biochemical analyzer. Fibrinogen (FIB) and D-dimer levels were measured using an automatic coagulation analyzer. Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured in the right arm artery of the subject using a standard electronic sphygmomanometer after the patient rested in a supine position for 5 min. The accuracy and reproducibility of the results were ensured by consistent training and systematic analytical processes of hospital doctors.^21,22 Hypertension was defined as SBP ≥140 mmHg and DBP ≥90 mmHg.^23,24 Hyperlipidemia was defined as TC ≥6.2 mmol/L (240 mg/dL), TG ≥2.3 mmol/L (200 mg/dL) or LDL-C ≥4.1 mmol/L (160 mg/dL), HDL-C≤1.0 mmol/L (40 mg/dL), or the use of oral lipid-lowering drugs or a self-reported doctor’s diagnosis.²⁵ Type 2 diabetes was defined as a FBG level ≥7.0 mmol/L, the use of oral hypoglycemic agents or insulin, or a self-reported doctor’s diagnosis.^26,27

The PNI was calculated as follows: albumin (ALB, g/L)+ 5×lymphocytes (LYM, 10⁹/L). According to the PNI, patients were divided into three groups: normal nutrition group (PNI≥38), moderate malnutrition (35<PNI<38), and severe malnutrition (PNI≤35).²⁸ According to the NYHA classification, cardiac function can be divided into four grades: NYHA class I: physical activity is unrestricted, and daily activities do not cause excessive fatigue, dyspnea, or palpitation; NYHA class II: slightly limited physical activity, no symptoms at rest, daily activities can cause fatigue, palpitation, dyspnea, or angina pectoris; NYHA class III: physical activity is considerably limited, and there are no symptoms at rest, which can be caused by activities lighter than daily activities; NYHA class V: cannot engage in any physical activity, and also have symptoms of congestive HF or angina pectoris at rest, which is aggravated after any physical activity.

A total of 2,052 patients with BMI data were screened from 17,413 participants. BMI was calculated for all patients and defined as the body weight (in kilograms) divided by the square of the body height (in meters).²⁹ According to the World Health Organization (WHO) Asian critical value, BMI was divided into underweight (BMI≤18.5), normal weight (18.5<BMI<25), and overweight (BMI≥25). Patients were stratified by combined subgroups of nutritional status and BMI: well-nourished-normal weight group (PNI≥38, 18.5<BMI<25), well-nourished-overweight group (PNI≥38, BMI≥25), well-nourished-underweight group (PNI≥38, BMI≤18.5), malnourished-normal weight group (PNI<38, 18.5<BMI<25), malnourished-overweight group (PNI<38, BMI≥25), and malnourished-underweight group (PNI<38, BMI≤18.5).

Statistical Analyses

Statistical analyses were performed using IBM SPSS Statistics 19.0 (IBM Corp, New York, NY, USA). Descriptive statistics were presented as median (interquartile range (IQR)) for continuous variables and number (%) for categorical variables. Baseline characteristics were compared using the Kruskal–Wallis H analysis of variance test and χ² test for continuous and categorical variables, respectively. Ordered logistic regression analysis was used to evaluate the relationship between the PNI of patients with CHD and the NYHA classification. However, ordered logistic regression analysis violated the test proportional odds assumption. Accordingly, multivariate logistic regression analysis was used to evaluate the relationship between the PNI of patients with CHD and the NYHA classification. Binary logistic regression analysis of the relationship between PNI combined with BMI and the NYHA classification was performed. Three logistic regression models were constructed: Model a: unadjusted; Model b: adjusted for age, sex, SBP, DBP, FBG, and HbA1c; Model c was based on model 2 with TC, TG, HDL-C, LDL-C, smoking, drinking, hypertension, hyperlipidemia, use of antihypertensive medication, and use of antilipidemic medication. Collinearity between independent variables was evaluated to ensure that it was appropriate to include them in the same model. Missing values were imputed using the multiple imputation method, calculated by the chain equation, and five complete datasets were created. Statistical significance was set at P<0.05.

Result

Baseline Demographics of the Study Participants

The baseline demographics of the study participants are shown in Table 1. The mean age of the patients was 64 (59–69) years, males constituted 55.5%. The highest number of patients with NYHA class II was 9,739 (55.9%), 7,071 (40.6%) patients had hypertension, and 5,739 (33.0%) patients had diabetes. FBG, LDH, FIB, D-dimer, diabetes, current antihypertensive medication, and hospitalization time were positively correlated with malnutrition, whereas HDL-C was negatively correlated with malnutrition.

Table 1 Baseline Demographics of Study Participants

Relationship Between Univariate and NYHA Classification

As shown in Table 2, sex, age, SBP, FBG, HbA1c, TC, TG, HDL-C, LDL-C, FIB, D-dimer, LDH, drinking history, current antilipidemic medication, and current antihypertensive medication were all related to NYHA classes II, III, and IV; smoking history was related to NYHA classes II and III. Hypertension history was related to NYHA classes II and IV. DBP and diabetes history were related to NYHA class III and IV.

Table 2 Relationship Between Univariate and NYHA Classification

Relationship Between PNI Index and NYHA Classification

Three logistic regression models were constructed to evaluate the relationship between the PNI and NYHA classification (Table 3). In the multivariate logistic regression analysis, when the PNI was a continuous variable, it was significantly related to NYHA class III and IV. The strongest association was observed for NYHA class IV (odd ratio [OR]: 0.863; 95% confidence interval [CI]: 0.850–0.877). After PNI was used as a classification variable, moderate and severe malnutrition were significantly related to NYHA class III and IV, and the highest association was observed in NYHA class IV (OR: 8.613; 95% CI: 5.526–13.424). In the further adjusted Model 3 (Figure 2), moderate and severe malnutrition were still significantly related to NYHA class III and IV, and the highest association was observed in NYHA class IV (OR: 6.564; 95% CI: 4.043–10.658).

Table 3 Relationship Between PNI and Classification of NYHA

Figure 2 Relationship between PNI and NYHA classification, **:P<0.01. Abbreviations: PNI, prognostic nutrition index; NYHA, New York Heart Association classification.

Relationship Between PNI Combined with BMI and NYHA Classification

Three logistic regression models were constructed to evaluate the association between PNI, BMI, and NYHA classification (Table 4). In the multivariate logistic regression analysis, we observed that malnourished-underweight patients and malnourished-overweight patients were significantly associated with poor NYHA classification, and malnourished-underweight patients (OR: 10.132; 95% CI: 2.712–37.852) were more than malnourished-overweight patients (OR: 4.825; 95% CI: 1.809–12.870). In the further adjusted Model c (Figure 3), the number of malnourished-underweight patients (OR: 8.038; 95% CI: 2.091–30.892) was still higher than that of malnourished-overweight patients (OR: 3.580; 95% CI: 1.286–9.966).

Table 4 Relationship Between PNI Combined BMI and NYHA Classification

Figure 3 Relationship between PNI combined BMI and NYHA classification, **:P<0.01.

Discussion

This study demonstrated that PNI, a new index of immune nutrition and inflammation, was significantly related to the NYHA classification of patients with CHD.^18,30 Previous studies have shown that malnutrition is highly prevalent among patients with congestive heart failure (CHF) and is strongly associated with NYHA classification, disease severity, and increased mortality.⁶ One study that examined the acute coronary syndrome (ACS) effect of malnutrition showed that malnutrition may be the trigger of ACS risk,³¹ leading to an increase in AS burden and the risk of plaque rupture, and is closely related to an increase in mortality and cardiovascular events.^32,33 Another study showed that PNI was significantly related to ventricular remodeling and diastolic function. An increase in PNI may improve cardiac function and reduce the incidence of cardiovascular events, thus reducing mortality. These results are consistent with our results.^34–36

Generally, obesity is a recognized and important predictor of morbidity and mortality of patients with CVD and other diseases (including chronic nephropathy and chronic obstructive pulmonary disease).^37,38 However, some studies have reported that obesity is related to reduced mortality in patients with HF.^39–41 A meta-analysis using BMI to relate nutritional status with HF has shown that obese patients with HF lived longer than non-obese patients; this finding was called the “obesity paradox”. Importantly, we observed that malnourished-underweight patients had a higher risk of worse NYHA classification than malnourished-overweight patients in this study, which further verified the “obesity paradox”, which is highly debated. Interestingly,³² recent studies have shown that malnourished-overweight individuals have the highest burden of complications, unsuitable cardiac remodeling, and the worst cardiac prognosis compared to malnourished-underweight individuals. We believe this inconsistency may be because BMI, waist-hip ratio, and body fat percentage were used as the definitions for being overweight and underweight in this study. In contrast, BMI was only used as the definition of weight loss and obesity in our study.^42,43 Studies have shown that the obesity paradox is observed only when BMI is used to define obesity, further verifying that our hypothesis may be reasonable.

This was a large-scale, multicenter clinical study in China, with a relatively large amount of data and more convincing results. However, this study had some limitations. First, this was a cross-sectional study, which could not explain the causal relationship between the PNI and NYHA classification in patients with CHD. In addition, we used only the NYHA classification to evaluate CHD severity. Other examination results, such as the coronary angiography Gensini score, may be more sensitive; however, these examination results were relatively few in our study. Hence, this relationship needs to be verified in future prospective studies combined with more objective indicators such as the Gensini score and coronary artery lesion index. Finally, due to the missing data on BMI in this study, the number of malnourished-overweight and malnourished-underweight patients was relatively small; thus, the results may be biased.

Nevertheless, our study was novel in comprehensively evaluating the relationship between the PNI of patients with CHD and NYHA classification and the complex relationship between the PNI combined with BMI and NYHA classification.

Conclusion

Overall, PNI was significantly related to the NYHA classification. Clinicians can highlight PNI to evaluate the nutrition of patients with CHD and perform early interventions, especially for malnourished underweight patients.