Background

Acute coronary syndrome (ACS) is still among the leading causes of global mortality and morbidity [1]. Studies in recent years have raised a particular concern that malnutrition is not insubstantial in ACS patients, even among those obese [2, 3]. Malnutrition fosters an increased burden through the vicious cycle of sarcopenia, cardiac dysfunction, and cardiogenic cachexia in cardiovascular diseases (CVD) [4, 5], and it’s also a modifiable risk factor on which physicians could act [6]. The nutritional risk index (NRI) has been commonly used as a nutritional screening tool in these years’ studies for its simplicity, accessibility, and discrimination abilities [7,8,9,10]. As a systemic and dynamic pathological process, inflammation plays multiple maladaptive roles contributing to the progression and destabilization of ACS, including atherogenesis, plaque evolution, thrombogenesis, as well as myocardial damage and repair [11, 12]. However, there is still an unmet need to take cognizance of the double burden of malnutrition and inflammation on ACS patients and identify those who would benefit more from proactive anti-inflammatory therapy.

Dysregulated inflammation with superimposed undernutrition may constitute a significant yet relatively concealed threat in ACS patients, which needs additional attention and integrated responses. Here, we aimed to explore the combined effects and interplay of inflammation and malnutrition on outcomes in ACS patients to provide valuable information as we move toward more individualized CVD care.

Methods

Study design and population

In this single-center, retrospective observational cohort, patients presenting with ACS undergoing percutaneous coronary intervention (PCI) between June 2016 and November 2017 at Beijing Anzhen Hospital were consecutively included. Patients with missing data on height, weight, serum albumin, hs-CRP, and a known history of coronary artery bypass grafting were excluded. Baseline demographic data, previous medical history, clinical presentations, laboratory and echocardiographic results, angiographic and procedural characteristics, and medications were collected from the electronic medical records. This study followed the Helsinki Declaration of Human Rights and was approved by the institutional review board (IRB) of Beijing Anzhen Hospital (IRB number: 2016034x). The IRB waived the need for written patient consent as this study involved a retrospective analysis of clinically acquired data.

Definitions

We dichotomized the inflammatory status of ACS populations based on enrollment hs-CRP measurements of 2 mg/L or greater (high-inflamed) vs. less than 2 mg/L (uninflamed), consistent with the current consensus [13, 14]. Nutritional status was graded using the NRI, which was calculated using [1.519 × serum albumin (g/L)] + 41.7 × (current weight/ideal weight)] for patients under the age of 65 and [1. 489×serum albumin (g/L)] + 41.7 × (current weight/ideal weight)] for those over 65, the latter formula is also known as the geriatric NRI (GNRI) [15]; the ideal body weight was considered using the Lorenz formula, that is, [height (cm) -100 - ([height (cm) − 150]/4)] for men, and [height (cm)-100-([height (cm)-150]/2.5)] for women [16]; malnutrition was considered as a dichotomous variable as scores of NRI or GNRI lower than 100 considered malnutrition [15]. Acute coronary syndrome and its components were determined by clinical diagnosis at discharge extracted from the electronic medical record. Anemia was hemoglobin lower than 120 g/L for men and lower than 110 g/L for women [17]. Complete revascularization was defined as treating all coronary artery segments > 1.5 mm in diameter and ≥ 50% diameter stenosis in the same procedure [18], which was determined according to the final coronary angiography reports.

Outcomes

The primary outcome was major adverse cardiovascular events (MACEs), defined as the composite proportion of cardiac mortality, non-fatal MI, non-fatal stroke, and unplanned revascularization. Cardiac mortality includes deaths that result from an AMI, sudden cardiac death, heart failure, and cardiac procedures. An unplanned revascularization was defined as a revascularization procedure that was not planned at the time of the index AMI admission and that was not performed in the setting of a recurrent AMI (to avoid examining the clinical impact of recurrent AMIs). The study index date was the admission date, and complete follow-up was until any MACE occurred or December 2020, whichever came first. Patients were followed up every six months by telephone or through available medical records.

Statistical analysis

Categorical variables were summarized as frequencies and percentages; numerical variables were summarized as the mean (standard deviation, SD) or median (interquartile range, IQR), depending on the data distribution. Categorical variables were compared with the χ² test, and continuous variables with Mann-Whitney non-parametric tests. Receiver-operating characteristic curve (ROC) analyses were performed to estimate the discriminative abilities of the nutritional scores and systemic inflammatory indicators on MACEs. The associations between the variables and the binary study outcomes were assessed using multivariable Cox proportional hazards regression models with a hazard ratio (HR) and 95% confidence interval (CI) in the total population and by stratified groups. Besides, the interactions between malnutrition and inflammation were tested by conducting likelihood ratio tests.

Potential covariates were identified by reviewing existing literature and clinician consensus. The following variables were extracted from the baseline assessment data: demographic factors (age and sex); clinical characteristics (heart rate, systolic blood pressure, diastolic blood pressure, hypertension, diabetes, peripheral artery disease, old myocardial infarction, prior PCI, prior stroke, anemia, total cholesterol [TC], low-density lipoprotein cholesterol [LDL-C], creatinine clearance [CrCl], fast plasma glucose [FBG], glycated hemoglobin A1c [HbA1C], left ventricular systolic dysfunction [LVSD], multivessel coronary artery disease, SYNTAX [Synergy Between PCI With Taxus and CABG] score, and complete revascularization); and discharge pharmacotherapy (dual antiplatelet therapy [DAPT], statin, and beta-blockers). Kaplan-Meier survival probability estimates were calculated from the date of PCI up to the total available follow-up at three years, stratified according to nutritional and inflammatory status, and assessed with the log-rank test. The concordance index, net reclassification improvement, and integrated discrimination improvement statistical analyses were performed to evaluate the incremental value of the hs-CRP or NRI in addition to the GRACE (Global Registry of Acute Coronary Events) risk model. A two-sided P-value of less than 0.05 was considered to indicate statistical significance. All analyses were done with R software (version 3.5.0, Austria) and Prism software (version 5.0, USA).

Results

A total of 1,743 ACS patients treated with PCI who met the enrollment criteria were included (251 patients without complete informative data were excluded), with an average age of 60 years, of whom 76.5% were male. Patients were stratified into four phenotypes according to their NRI and hs-CRP categories: nourished and uninflamed group (n = 1051), nourished and high-inflamed group (n = 573), malnourished and uninflamed group (n = 46), as well as malnourished and high-inflamed group (n = 73) (Fig. 1). Over a median follow-up of 30 months (IQR, 30–36months), 351 (20.1%) MACEs occurred, including 37 (2.1%) cardiac deaths, 43 (2.5%) non-fatal MI, 22 (1.3%) non-fatal stroke, and 249 (14.3%) unplanned coronary revascularization. Compared with the other groups, malnourished and highly inflamed patients were more prone to complicated diabetes, anemia, and renal dysfunction, presenting with larger proportions of acute myocardial infarction (AMI), displaying worse cardiac function and more complex coronary lesions; they were also more often to be prescribed with furosemide on discharge. After 30 months of follow-up, the malnourished and high-inflamed group showed the highest rates of MACEs (41.1%), cardiac death (12.3%), non-fatal MI (4.1%), and unplanned revascularization (24.7%) among the four phenotypes (Table 1).

Fig. 1
figure 1

Flow chart of the study population enrollment and stratification Abbreviations ACS, acute coronary syndrome; CABG, coronary angioplasty bypass grafting; CRP, C-reactive protein; PCI, percutaneous coronary intervention

Full size image
Table 1 Baseline clinical characteristics and outcomes of patients stratified by the nutritional and inflammatory statuses
Full size table

The adjusted hazard ratio and 95%CI of inflammation or malnutrition on clinical outcomes are shown in Table 2; the association between the covariates included in the multivariable model and MACEs was displayed in Supplementary Table 1. Malnutrition showed a significant correlation to the subsequent cardiac death, unplanned revascularization, and overall MACEs; the high-inflamed status was related to the increased MACEs, mainly driven by the incidence of unplanned revascularization.

Table 2 Predictive value of malnutrition or inflammation for outcomes in univariate and multivariate proportional hazards regression analyses
Full size table

Kaplan-Meier survival analyses showed that malnourished patients with elevated hs-CRP levels exhibited the most unfavorable prognosis in terms of MACEs, cardiac death, and unplanned revascularization. However, no difference among the four groups was found in non-fatal MI or non-fatal stroke (Fig. 2). Using nourished without the elevated hs-CRP group as the reference, malnourished patients with elevated hs-CRP showed significantly higher MACEs, cardiac death, and unplanned revascularization in multivariate analyses (all P < 0.05); the nourished with elevated hs-CRP group presented with increased unplanned revascularization (P = 0.005). Interestingly, patients in the malnourished group without elevated hs-CRP exhibited no significant difference in affecting MACEs and each component compared with the reference group (Table 3). We further conducted the subgroup analyses evaluating the predictive value of nutrition-inflammation status in patients stratified by ACS type, and no evidence of subgroup interactions was observed between patients diagnosed with AMI or unstable angina (Supplementary Table 3).

Fig. 2
figure 2

Kaplan-Meier survival curves for MACEs and each component up to 3 years Abbreviations hs-CRP, high-sensitivity C-reactive protein; MACEs, major adverse cardiac events; MI, myocardial infarction; PCI, percutaneous coronary intervention

Full size image
Table 3 Predictive value of each group on outcomes in multivariate proportional hazards regression analyses a
Full size table

We further conducted stratified analyses within subgroups and calculated interaction statistics to estimate the interplay between baseline malnutrition and inflammation regarding MACEs. As shown in Table 4, the P-value for interaction between nutritional status (nourished vs. malnourished) and hs-CRP levels (≤ 2 mg/L vs. >2 mg/L) on MACEs was 0.248; when including hs-CRP in the model as a continuous variable, it remained a risk enhancer among the malnourished (P = 0.003) but not the nourished (P = 0.620), with evidence for interaction (P for interaction = 0.001). Simultaneously, as it was included in the model as a continuous variable, the NRI score shared a significant correlation with MACEs in high-inflamed patients (P = 0.003) while failing to present it in those without (P = 0.869), with evidence for interaction (P for interaction = 0.005) (Table 5). When hs-CRP and malnutrition were both included in the model as continuous variables, there was evidence for effect modification between each other on the association of MACEs (P for interaction = 0.046).

Table 4 Interplay of inflammation or malnutrition on MACEs in multivariable analyses a. Examining the effect of inflammation on MACEs in patients stratified by nutritional status
Full size table
Table 5 Examining the effect of malnutrition on MACEs in patients stratified by inflammatory status
Full size table

Finally, the incremental predictive effect of hs-CRP or NRI beyond the established GRACE risk score was evaluated. As adding each variate into the GRACE model significantly increased the original predictive value on MACEs or cardiac death, a synergistic effect showed when combining both hs-CRP and NRI with the GRACE model (Table 6 and Supplementary Table 4).

Table 6 Incremental effect of inflammation or malnutrition beyond the GRACE risk model to predict MACEs and cardiac death
Full size table

Discussion

This study of retrospective data is the first to highlight the importance of an easily overlooked population of ACS patients, who suffered a double burden of malnutrition and high-inflamed status. We validated that among ACS patients, malnutrition combined with inflammation is associated with a significantly increased risk of long-term cardiac outcomes and may interplay by amplifying the harmful effects of each other, adding both of which into the GRACE risk score would provide an increment predictive value.

Statistics about the prevalence of malnutrition among ACS patients varied across these years’ studies. A large cohort involving 10,161 AMI patients who underwent PCI showed that 11% were at risk of malnutrition, as determined by GNRI [19]. Komici et al. evaluated the nutritional status of 174 elderly AMI patients by Mini Nutritional Assessment (MNA), and 12% were considered malnourished [20]; Tonet et al. reported 4% of elderly ACS patients undernourished using the MNA–Short Form [21]. Data from 5,062 ACS patients suggested that malnutrition prevailed in over 50% of cases based on the controlling nutritional status (CONUT) score, NRI, or prognostic nutritional index (PNI) [22]. Recently, Kong et al. observed 1,829 AMI patients, 76.5%, 61.8%, and 39.3% of whom were malnourished according to the CONUT score, NRI, and PNI, respectively [3]. These variations reflect the necessity and importance of setting universal, standardized, and pragmatic nutritional evaluation criteria to facilitate risk stratification and management. In 2018, the American Society for Parenteral and Enteral Nutrition (ASPEN) led the launching of the GLIM criteria as a consensus-based criterion of malnutrition [23]; however, the demand for body composition measuring techniques may hinder its widely applied in a general setting, especially a coronary care unit. Current frequently used screening tools have exhibited various discrimination abilities. However, some may not be suitable, considering the particularity of the ACS population. To be specific, one of the items of the criterion of the CONUT score is total cholesterol, which could be strongly perturbed by patients who have received lipid-lowering therapies. Thus, there’s an unmet need to establish an ACS-suitable nutritional evaluation method to facilitate clinical evaluation and prevention systems.

In our cohort, older adults, females, and leaner people were found more frequently in the malnourished group, and they also displayed lower cholesterol levels; these clinical characteristics are consistent with the former studies [3, 22]. One of the notable observations of our study is that a heavy burden of systemic inflammation marked by elevated hs-CRP prevailed in malnourished ACS patients by over 60%. Inflammation is thought to be a key driver for sarcopenia elicited by an underlying illness, leading to disease-related malnutrition and increased mortality [24, 25]. CAD has been recognized as not only a lipid metabolic disorder but also an inflammation-mediated systemic disease, with the pro- and anti-inflammatory reactions interacting to exacerbate or ameliorate the atherosclerotic process and causing catabolic activity, energy expenditure, and function loss. One of the primary underlying mechanisms is the circulating cytokines released from the systemic inflammatory response (i.e., interleukin 6, interleukin 1β, and tumor necrosis factors), triggering mechanisms that contribute to the pathogenesis of malnutrition, including brain circuits disorder, protein degradation, and neuroendocrine disturbance, ultimately causing disease-related anorexia, sarcopenia, cognitive decline, and frailty [11, 12, 26, 27]. In 2012, ASPEN and the Academy of Nutrition and Dietetics launched a consensus statement emphasizing the need to identify inflammation early in the diagnostic procedure of malnutrition to determine its etiology and guide treatment [28]. In a traumatic and acute stress-causing event like ACS, albumin serves as a practical plasma biomarker that can be used to indicate and monitor catabolic metabolism and systemic inflammatory response to reflect disease severity; hypoalbuminemia was also one of the most striking features of patients with MACEs in our cohort.

Nutritional status influences the body’s inflammatory response, and consistent nutritional management is crucial for patients with critical illness to recover from inflammatory injury [27, 29]. Although the prevalence of malnutrition exhibited significant discrepancies, its association with poorer outcomes remains clear and sound [19,20,21,22, 30, 31]. However, the importance of screening and monitoring for malnutrition in patients with ACS hasn’t drawn enough concern from clinical guidelines and practice [32, 33]. The subgroup analysis of the present study suggested the bidirectional association between inflammation and malnutrition; however, the interaction differed when handling inflammation and malnutrition as binary or continuous values, which indicates that the stratification standards to define high inflammation or malnutrition seemingly may affect the results. In this scenario, studies with larger scale and ACS-suitable criteria for stratification seemed indispensable to facilitate the understanding of the interplay between inflammation and malnutrition. Given the existing results, we could speculate that patients with worse nutritional conditions may urgently need deactivating inflammation, and patients with higher hs-CRP would benefit more from nutritional support. Previous studies among hospitalized patients showed that baseline inflammatory status was associated with their response to nutritional therapy and that nutritional support alone is inadequate to prevent muscle loss during severe, sustained, or repeated bouts of inflammation [34, 35], which emphasizes the importance of considering the double burden of inflammation and malnutrition in disease assessment, intervention, and monitoring. Regarding ACS patients, there is still a lack of evidence investigating the treatment response to nutritional support or anti-inflammatory therapy among subgroups with different degrees of inflammation or malnutrition.

To the best of our knowledge, the present study is the first analysis to investigate the interplay between inflammation and malnutrition in affecting cardiac outcomes among ACS patients, which would provide preliminary implications for future research. We are also aware of several limitations. First, as a retrospective study, we couldn’t conclude direct causality, account for other potential confounding factors, and were more susceptible to selection, information, and recall bias. Second, the proportion of AMI patients (26%) in our ACS cohort was lower than the statistics from large-scale Chinese registries [36, 37], which may be attributed to the limited sample size; thus, the study results do not necessarily reflect the outcome of ACS patients treated according to a modern standard. Third, the diagnostic accuracy of NRI among ACS patients was yet to be confirmed, and the GLIM criteria were regrettably not used due to a lack of indispensable information; these existing problems need to be addressed in future research works as we move towards a global definition of malnutrition. Fourth, the serum hs-CRP, albumin, and body weight were measured only once at admission and are subject to measurement error and physiologic variability; the dynamic and long-term monitoring of inflammatory and nutritional status changes would provide valuable and indispensable information. Fifth, as it was a single-center study based on Chinese people, the results should be carefully interpreted and extended to other ethnicities.

Conclusions

Among patients with ACS undergoing PCI, the double burden of inflammation and malnutrition was not uncommon and strongly associated with long-term cardiac outcomes in a synergistic manner. Identifying patients’ inflammatory and nutritional status will better facilitate the modern individualized care of ACS. Future prospective studies are needed to evaluate whether the heterogeneity in the therapeutic effect of nutritional support or anti-inflammatory interventions exists in patients with different inflammatory or nutritional burdens.