Prognostic nutritional index is useful for predicting the prognosis of patients with infective endocarditis undergoing surgery: a retrospective study

Objective: To investigate the predictive efficacy of preoperative prognostic nutritional index (PNI) for clinical outcomes in patients with infective endocarditis (IE) undergoing surgical treatment.

Methods: A retrospective analysis was conducted on 373 IE patients who underwent cardiac valve surgery in the Department of Cardiac Surgery, the First Affiliated Hospital of Xinjiang Medical University from January 1, 2009 to December 31, 2023. According to the preoperative PNI scores, patients were divided into the malnourished group (132 cases) and the well-nourished group (241 cases). Univariate and multivariate Logistic regression analyses were used to explore the association between preoperative PNI and postoperative in-hospital mortality as well as 1-year all-cause mortality.

Results: This study analyzed 373 patients who underwent surgery for IE to assess the impact of preoperative nutritional status on prognosis. According to the preoperative PNI, 35.4% (132 patients) were malnourished. Compared with the well-nourished group, the malnourished group had a lower body mass index (BMI) (20.94 vs. 22.84, P < 0.001) and a higher proportion of NYHA class III–IV heart function (53.79% vs. 31.95%, P < 0.001). Multivariate analysis revealed that a lower preoperative PNI score (OR = 0.91, 95% CI: 0.83∼0.99, P = 0.033) and longer cardiopulmonary bypass (CPB) time (per 10-min increase: OR = 1.15, 95% CI: 1.07∼1.24, P < 0.001) were independent risk factors for postoperative in-hospital mortality. Similarly, preoperative PNI score (OR = 0.95, 95% CI: 0.90–0.99, P = 0.040) and CPB time (per 10-min increase: OR = 1.10, 95% CI: 1.04–1.16, P = 0.001) were also independent predictors of 1-year all-cause mortality. ROC curve analysis showed that the predictive ability of PNI for postoperative in-hospital mortality (AUC = 0.74) was better than that for 1-year mortality (AUC = 0.61).

Conclusion: Preoperative PNI score can effectively reflect the nutritional-immune status of patients with IE. It is not only an independent risk factor for predicting postoperative in-hospital mortality and 1-year all-cause mortality, but also has certain short-term predictive ability and identification value for long-term outcomes, which can provide a reference for clinical early identification of high-risk patients and formulation of nutritional intervention strategies.

1 Introduction

Infective endocarditis (IE) is a severe infectious disease with concerning morbidity and mortality, posing a significant threat to human health. The latest statistics indicate that the global incidence of IE is approximately 13.8 cases per 100,000 population, resulting in around 66,300 deaths annually (1). IE can lead to severe complications such as valvular damage, sepsis, septic embolism, stroke, and acute heart failure, which not only endanger patients’ lives but also increase the difficulty of treatment and nutritional consumption, thereby elevating the risk of malnutrition (2).

In recent years, studies have found that malnutrition is associated with increased mortality, higher infection risk, and prolonged hospital stays in IE patients, significantly impairing their prognosis (3, 4). Preoperative nutritional status has a notable impact on postoperative outcomes in patients undergoing cardiac surgery (5). Research by Juliana et al. (6) demonstrated that patients with heart disease who are malnourished preoperatively have a higher mortality rate than those with good nutritional status. Additionally, patients undergoing cardiac surgery are more prone to postoperative complications such as cardiac cachexia, sarcopenia, and cognitive dysfunction, accompanied by longer intensive care unit (ICU) stays, prolonged postoperative mechanical ventilation, and higher incidences of postoperative cardiovascular and infectious complications. Therefore, preoperative nutritional status is a key driver of clinical outcomes in cardiac surgery patients.

As a serology-based tool for assessing immunonutritional status, the prognostic nutritional index (PNI) has been confirmed to be significantly associated with in-hospital and long-term mortality in various cardiovascular diseases, including acute coronary syndrome and heart failure (7–10). However, research on its role in prognostic assessment of IE remains limited. Against this backdrop, this study focuses on exploring the predictive efficacy of preoperative PNI scores for clinical outcomes in IE patients undergoing surgical treatment.

2 Materials and methods

2.1 Study population

This was a retrospective observational study. The study population included patients who were hospitalized in the Department of Cardiac Surgery, the First Affiliated Hospital of Xinjiang Medical University, from January 1, 2009, to December 31, 2023. This study has been approved by the Hospital Ethics Committee (No. K202411-32) and supported by the Open Projects of the National Key Laboratory for the Etiology and Prevention of High-Incidence Diseases in Central Asia, co-built by the Province and the Ministry Fund (No: SKLHIDCA-2023-24). These patients were clearly diagnosed with IE according to the modified Duke diagnostic criteria (11) and underwent valve surgery under cardiopulmonary bypass (CPB). Inclusion criteria: (1) Diagnosis of IE based on the modified Duke diagnostic criteria; (2) Eligibility for and receipt of cardiac surgery; (3) Adults aged ≥18 years; (4) Complete clinical data. (5) Complete 1-year follow-up data (confirmed via outpatient records and telephone interviews, with no missing information on all-cause mortality). Exclusion criteria: (1) Complicated with other cardiac surgeries (e.g., congenital heart disease, coronary atherosclerotic heart disease, Marfan syndrome, ascending aortic aneurysm, etc.); (2) Previous or in-hospital diagnosis of other consumptive diseases (e.g., malignant tumors, immunodeficiency-related diseases, hyperthyroidism, etc.); (3) Pregnant or lactating women; (4) Patients with incomplete clinical data. Additionally, to ensure that all patients ultimately included in the study cohort have complete follow-up information, those with incomplete 1-year follow-up data were excluded during the initial enrollment phase. The detailed screening process is shown in Figure 1.

FIGURE 1

Figure 1. Study flow diagram.

2.2 Treatment methods

In accordance with the 2023 ESC guidelines (1), the surgery aimed to completely remove the infected foci to prevent persistent or recurrent infection, and to repair or replace the damaged valves based on the extent of valvular lesions. All patients underwent open-heart surgery with CPB to remove intracardiac vegetations and perform valve surgery. After the operation, all patients were transferred to the cardiac surgery ICU for close monitoring and treatment.

2.3 Data collection

Clinical data of patients were collected from the hospital’s electronic medical record system, and baseline data were entered into Excel. Baseline data included: (1) Demographic data: age, gender, BMI, etc.; (2) Symptoms and signs at admission: fever, shortness of breath, chest tightness or pain, lower extremity edema, etc.; (3) Comorbidities: diabetes, hypertension, etc.; (4) Preoperative laboratory tests: blood routine, coagulation function, liver and kidney function, etc.; (5) Surgery-related clinical data; (6) Follow-up data.

2.4 Malnutrition scoring systems

2.4.1 Body mass index (BMI) scoring

Based on the World Health Organization (WHO) criteria, patients were classified into malnutrition (<18.5 kg/m²), normal weight (18.5–24.9 kg/m²), overweight (25.0–29.9 kg/m²), and obese (≥30 kg/m²) according to their body mass index (BMI).

2.4.2 Prognostic nutritional index scoring

It mainly includes serum albumin and total peripheral blood lymphocyte count. The formula for PNI is: PNI = serum albumin (g/L) + 5 × total peripheral blood lymphocyte count (×10⁹/L) – with the following nutritional status cut-offs immediately defined: a score >38 indicates normal nutrition, scores of 35–38 indicate moderate malnutrition, and a score <35 indicates severe malnutrition (8).

2.5 Follow-up and grouping

All patients with infective endocarditis were followed up for 1 year through outpatient visits and telephone calls. Given that patients with incomplete follow-up data were excluded during enrollment, all 373 finally included patients completed the 1-year follow-up, achieving a 100% follow-up rate. The endpoint of follow-up was all-cause mortality. A self-designed follow-up form was used to record all-cause mortality events during follow-up. Patients were divided into the malnourished group (132 cases) and the well-nourished group (241 cases) according to preoperative PNI scores.

2.6 Statistical analysis

Statistical analyses were performed using SPSS version 26.0 and R version 4.4.2. Continuous variables were tested for normality using the Shapiro-Wilk test. Those conforming to a normal distribution were expressed as mean ± standard deviation and compared between groups using the independent samples t-test; non-normally distributed variables were expressed as median (interquartile range) and compared using the Mann-Whitney U test. Categorical variables were described as frequency (percentage) and compared using the χ² test or Fisher’s exact test. Multivariable logistic regression analysis was employed to identify independent risk factors, incorporating variables with a P-value < 0.05 in univariate analysis and clinically relevant variables. Multicollinearity was assessed using the variance inflation factor. Receiver operating characteristic (ROC) curve analysis was conducted to evaluate predictive performance, calculating the area under the curve (AUC) with its 95% confidence interval. Differences between AUCs of different ROC curves were compared using DeLong’s test. All statistical tests were two-tailed, and a P-value < 0.05 was considered statistically significant.

3 Results

3.1 Baseline preoperative characteristics of patients

Among the 373 enrolled patients, the mean age was 41.9 years (range: 18–73 years), with 292 males (78.3%) and 81 females (21.7%). Regarding clinical features, weight loss of more than 5% within 3 months was the most common presentation (96 cases, 25.7%). Comorbidities and relevant medical history included 12 patients with prosthetic valves (PV, 3.2%), 20 cases of diabetes mellitus (5.4%), 53 cases of hypertension (14.2%), and 60 patients with a history of cerebrovascular accident (CVA, 16.1%). In addition, there were 16 cases of liver dysfunction (4.3%), 32 cases of renal dysfunction (8.6%), and 12 cases of electrolyte disorders (3.2%). A total of 196 patients (52.5%) had positive blood cultures. The pathogen distribution was as follows: 100 cases of Streptococcus infection (26.8%), 29 cases of Staphylococcus infection (7.8%), and notably, 10 cases of Brucellosis (2.7%). Echocardiographic findings revealed 364 patients with cardiac vegetations (97.6%), 342 patients with left-sided infective endocarditis (91.7%), 73 patients with aortic valve perforation (19.6%), and 13 patients with aortic perivalvular abscess (3.5%). Additionally, 8 cases (2.1%) had a left ventricular ejection fraction ≤40%. The distribution of surgical procedures was as follows: 148 cases (39.7%) underwent aortic valve replacement/repair, 114 cases (30.6%) underwent mitral valve replacement/repair, 3 cases (0.8%) underwent tricuspid valve replacement/repair, and 108 cases (28.9%) underwent combined valve replacement/repair.

The mean preoperative BMI of the patients was 22.2 kg/m². Based on BMI classification: 58 cases (15.6%) were malnourished (<18.5 kg/m²), 234 cases (62.7%) had normal weight (18.5–24.9 kg/m²), 69 cases (18.5%) were overweight (25.0–29.9 kg/m²), and 12 cases (3.2%) were obese (>30 kg/m²). The mean preoperative PNI was 40.7. According to PNI classification, 132 cases (35.4%) were malnourished, including 54 cases of moderate malnutrition and 78 cases of severe malnutrition. Further analysis revealed that among the 132 malnourished patients identified by PNI, 87 cases (65.9%) had a normal preoperative BMI, and only 30 cases (22.7%) met the BMI criteria for malnutrition. Among the 78 cases with severe malnutrition, 52 cases (66.7%) still had a normal preoperative BMI (Figure 2 and Table 1).

FIGURE 2

Figure 2. Multidimensional analysis of preoperative malnutrition in patients based on BMI and PNI.

TABLE 1

Table 1. Cross-distribution of preoperative BMI categories and PNI status in patients with infective endocarditis (column %).

3.2 Comparison of preoperative clinical characteristics

Between IE Malnourished Group and Well-Nourished Group According to the preoperative PNI scoring criteria, patients were divided into the malnourished group (132 cases, 35.4%) and the well-nourished group (241 cases, 64.6%). The age distribution of the two groups was relatively similar. Comparison of clinical data (Table 2) showed that, compared with the well-nourished group, the malnourished group was significantly associated with lower BMI (20.94 ± 3.38 vs. 22.84 ± 3.80, P < 0.001) and had a higher proportion of cardiac insufficiency with New York Heart Association (NYHA) Classification III-IV (53.79% vs. 31.95%, P < 0.001). Comparison of laboratory test results (Table 3) revealed that the levels of B-Type Natriuretic Peptide, procalcitonin, and Neutrophil-to-Lymphocyte Ratio (NLR) in the malnourished group were significantly higher than those in the well-nourished group (all P < 0.01); while the levels of creatine kinase, total cholesterol, serum albumin, hemoglobin, and platelet count were significantly lower (all P < 0.05).

TABLE 2

Table 2. Comparison of clinical data between the well-nourished group and the malnourished group.

TABLE 3

Table 3. Comparison of laboratory test results between the well-nourished group and the malnourished group.

3.3 Model diagnostics: collinearity (VIF)

To assess multicollinearity, variance inflation factors [VIF = 1/(1−R²)] were computed using ordinary least squares. All covariates in both multivariable models had VIFs close to 1 and <2 (approximate range 1.02∼1.09), indicating no meaningful multicollinearity (Supplementary Tables 1, 2). Because serum albumin is a component of PNI, albumin was not entered alongside PNI to avoid structural collinearity.

3.4 Analysis of risk factors for postoperative in-hospital mortality in IE patients

In this study, 19 cases (5.1%) of postoperative in-hospital mortality occurred among IE patients. Univariate logistic regression analysis (Table 4) showed that NYHA class III–IV (OR = 3.51, 95% CI: 1.35∼10.21, P = 0.013), preoperative PNI score (OR = 0.87, 95% CI: 0.80∼0.94, P < 0.001), and CPB time modeled per 10-min increment were significantly associated with postoperative in-hospital mortality (OR = 1.10, 95% CI: 1.01∼1.34, P < 0.001). After including variables with P < 0.05 from the above univariate analysis into the multivariable logistic regression (Table 5) and modeling CPB time per 10 min, preoperative PNI score (OR = 0.91, 95% CI: 0.83∼0.99, P = 0.033) and CPB time (per 10 min: OR = 1.15, 95% CI: 1.07∼1.24, P < 0.001) remained independently associated with postoperative in-hospital mortality in IE patients.

TABLE 4

Table 4. Univariate analysis of risk factors for in-hospital mortality in IE patients undergoing surgery.

TABLE 5

Table 5. Multivariate analysis of risk factors for in-hospital mortality in IE patients undergoing surgery.

3.5 Analysis of risk factors for postoperative 1-year all-cause mortality in IE patients

A total of 47 cases (12.6%) of 1-year all-cause mortality occurred after surgery in IE patients. Univariate logistic regression (Table 6) showed that age (OR = 1.03, 95% CI: 1.01∼1.06, P = 0.029), preoperative PNI score (OR = 0.93, 95% CI: 0.89∼0.98, P = 0.005), and CPB time modeled per 10-min increment were significantly associated with 1-year all-cause mortality (OR = 1.11, 95% CI: 1.10∼1.18, P < 0.001). In the multivariable logistic regression (Table 7) with CPB time modeled per 10 min, preoperative PNI score (OR = 0.95, 95% CI: 0.90∼0.99, P = 0.040) and CPB time (per 10 min: OR = 1.10, 95% CI: 1.04∼1.16, P = 0.001) remained independent predictors of 1-year all-cause mortality, whereas age was no longer statistically significant (P = 0.118).

TABLE 6

Table 6. Univariate analysis of risk factors for 1-year all-cause mortality in IE patients undergoing surgery.

TABLE 7

Table 7. Multivariate analysis of risk factors for 1-year all-cause mortality in IE patients undergoing surgery.

3.6 Functional form and linearity

For in-hospital mortality, PNI showed significant non-linearity (RCS, P_{non–linearity}−0.0277), while CPB time did not (P = 0.087). For 1-year mortality, CPB time exhibited clear non-linearity (P = 0.001), whereas PNI did not (P = 0.152). Given event counts and parsimony, we retained linear terms in primary models; spline curves are presented as sensitivity analyses (Supplementary Figures 1, 2). For clinical interpretability, we further report OR per 5-point decrease in PNI from the multivariable models: In-hospital mortality: OR per 1-point increase = 0.91 (95% CI: 0.83∼0.99), OR per 5-point decrease = 1.60 (95% CI: 1.05∼2.54). 1-year mortality: OR per 1-point increase = 0.95 (95% CI: 0.90–0.99), OR per 5-point decrease = 1.29 (95% CI: 1.05∼1.69).

3.7 Absolute risks and adjusted ORs by PNI strata/quartiles

To enhance clinical interpretability, the PNI was categorized into three strata: severe malnutrition (PNI < 35), moderate malnutrition (PNI 35–38), and normal nutrition (PNI > 38). Absolute risks with 95% confidence intervals (CIs), calculated using the Clopper–Pearson method, are reported for each stratum.

For in-hospital mortality (Table 8), after adjusting for CPB time (per 10-min increment) and NYHA class III–IV, the absolute risk was 2.5% (95% CI: 0.9–5.3) in the normal PNI group, 7.4% (95% CI: 2.1–17.9) in the moderate malnutrition group, and 11.5% (95% CI: 5.4–20.8) in the severe malnutrition group. Compared to the normal PNI group, the adjusted odds ratios were 2.54 (95% CI: 0.66–9.81, P = 0.178) for moderate malnutrition and 2.14 (95% CI: 0.61–7.53, P = 0.236) for severe malnutrition.

TABLE 8

Table 8. In-hospital mortality: absolute risks and adjusted ORs by PNI strata.

For 1-year all-cause mortality (Table 9), after adjusting for age and CPB time (per 10-min increment), the absolute risk was 9.5% (95% CI: 6.1–14.0) in the normal PNI group, 13.0% (95% CI: 5.4–24.9) in the moderate malnutrition group, and 21.8% (95% CI: 13.2–32.6) in the severe malnutrition group. The adjusted odds ratios were 1.33 (95% CI: 0.53–3.34, P = 0.540) for moderate malnutrition and 1.98 (95% CI: 0.93–4.18, P = 0.075) for severe malnutrition, using the normal PNI group as reference. Although the severe malnutrition group showed a trend toward higher mortality risk, the adjusted OR were not statistically significant (P > 0.05).

TABLE 9

Table 9. 1-year all-cause mortality: absolute risks and adjusted ORs by PNI strata.

3.8 Predictive efficacy of PNI for prognosis in patients with infective endocarditis by ROC curve

Results of ROC curve analysis showed that the area under the curve (AUC) of PNI for predicting postoperative in-hospital death in patients with infective endocarditis was 0.74 (95% CI: 0.64–0.83), and the AUC for predicting 1-year all-cause death was 0.61 (95% CI: 0.53∼0.70) (Figure 3).

FIGURE 3

Figure 3. Predictive efficacy of PNI for in-hospital death and 1-year all-cause death in patients with infective endocarditis.

4 Discussion

Despite the proliferation of research on IE, little is known about the relationship between the PNI and IE. Based on our findings, three key observations can be made. First, IE patients classified as malnourished or well-nourished by PNI criteria exhibit distinct clinical characteristics. The malnourished group shows lower BMI and a higher proportion of NYHA Classification III-IV (53.79% vs. 31.95%). They also have significantly elevated inflammatory markers (e.g., C-reactive protein, procalcitonin) and markedly reduced nutrition-related indicators (e.g., serum albumin, lymphocytes). These findings suggest that preoperative PNI can effectively distinguish the nutritional-inflammatory status and degree of cardiac function impairment in IE patients. Second, the preoperative PNI score is an independent risk factor for in-hospital mortality and 1-year all-cause mortality in IE patients undergoing surgery. A lower score indicates higher risk. This score demonstrates predictive efficacy for both short-term and long-term prognosis. Prolonged CPB time is also an independent risk factor, suggesting that surgical trauma and nutritional-immune status jointly influence prognosis.

These results further supplement existing research. Several surgical risk scores (e.g., EuroSCORE II, ACC/AHA risk models) include age, heart failure, renal failure, microbial type, and local complications (12, 13). However, their application is limited by heterogeneity in IE manifestations and variable disease progression. Kahraman et al. (14) first reported that PNI is associated with in-hospital mortality in IE inpatients, providing a basis for this study. Our analysis confirmed that preoperative PNI score is an independent risk factor for 1-year all-cause mortality in surgically treated IE patients (OR = 0.95, 95% CI: 0.90∼0.99, P = 0.040). Limited evidence has addressed the long-term prognostic value of PNI in this population. Kahraman et al. reported an association between PNI and in-hospital mortality, but did not explore 1-year outcomes. Our findings extend their work by confirming PNI’s independent predictive role for long-term mortality. Meanwhile, its independent predictive value for in-hospital mortality was also verified (OR = 0.91, 95% CI: 0.83∼0.99, P = 0.03). This result is consistent with Cheng et al. (15) In very elderly patients with non-valvular atrial fibrillation, moderate to severe malnutrition assessed by PNI was an independent predictor of composite events (thromboembolism + all-cause death) (HR = 3.37, 95% CI: 1.90∼6.00, P < 0.001). This confirms the universal significance of PNI as a prognostic indicator in cardiovascular diseases. PNI directly reflects the patient’s underlying status rather than acting through complications. Patients with abnormal preoperative PNI are more likely to develop uncontrolled infection and heart failure due to systemic imbalance. This effect is evident in both short-term in-hospital outcomes and long-term prognosis. Thus, PNI serves as a new independent predictor for long-term risk assessment in IE patients. Similar conclusions were reported by Chen et al. (16, 17), who showed that PNI predicts both short-term and long-term prognosis in heart disease patients.

Notably, in the multivariate analysis of in-hospital mortality, NYHA class lost significance after adjustment (P = 0.335). PNI, however, remained independently associated with prognosis. This suggests that PNI’s effect is not dependent on NYHA classification and highlights its unique role as a supplementary risk indicator. ROC curve analysis confirmed its predictive value. The AUC for in-hospital mortality was 0.74 (95% CI: 0.64∼0.83), indicating moderate predictive efficacy. The AUC for 1-year mortality was 0.61 (95% CI: 0.53∼0.70), showing modest identification value. This aligns with the trend of PNI’s C-index (0.77, 95% CI: 0.71∼0.82) in cardiac valve surgery patients reported by Cho et al. (18) The weaker performance for long-term outcomes reflects the complexity of prognosis, influenced by postoperative interventions and disease progression. Nonetheless, PNI retains unique value as a preoperative marker. It is based on quantitative serological data and does not require complex indicators. PNI provides objective evidence of the link between nutritional-immune status and prognosis. It is a simple, independent tool that supports clinical decision-making, risk stratification, and outcome prediction in cardiovascular disease (19–21).

A key phenomenon emerged in our study. Among the 132 malnourished patients identified by PNI, the majority (65.9%, 87 cases) had normal BMI. In addition, 11.4% (15 cases) were overweight or obese. These data show that normal BMI does not always equal good nutrition, and obesity may coexist with malnutrition. Compared with BMI, which only categorizes body shape, PNI better captures the essence of nutritional-immune status. It helps avoid delays in clinical decisions caused by misjudging nutrition based on body weight.

The predictive advantage of PNI comes from its integration of albumin and lymphocytes. Albumin and lymphocytes are both crucial in IE progression. In this study, serum albumin in the malnourished group was markedly lower than in the well-nourished group (27.15 vs. 35.44 g/L, P < 0.001). This is consistent with findings that hypoalbuminemia is an independent risk factor for early postoperative mortality in IE (22). Hypoalbuminemia not only reflects malnutrition but also reduces immunoglobulin synthesis, impairs cellular immunity, and weakens pathogen clearance. This creates a vicious cycle of “malnutrition–immunosuppression–disease progression” (23). Similarly, lymphocyte counts were significantly lower in the malnourished group (1.26 vs. 1.82 × 10⁹/L, P < 0.001). Low lymphocyte count is closely linked to severity and mortality of infectious diseases (24, 25). Zhang et al. (26) also showed that a total lymphocyte count <1500/mm³ is strongly associated with malnutrition. Thus, low lymphocytes in IE patients reflect both immune dysfunction and nutritional deficiency, increasing the risk of poor outcomes. By integrating these two markers, PNI captures the “nutrition–inflammation–immunity” interaction, which explains its superior predictive value.

Prognostic nutritional index has direct clinical implications. Patients with low PNI should receive individualized nutritional intervention via multidisciplinary collaboration (cardiac surgery and nutrition teams). Albumin and high-quality protein should be supplemented to correct hypoalbuminemia, along with trace elements such as zinc and selenium to enhance lymphocyte function (27). At the same time, careful preoperative preparation can shorten CPB time, an independent risk factor identified in this study, thereby reducing risks of organ hypoperfusion and bleeding (28). Dynamic monitoring of PNI with cardiac and inflammatory indicators may also help detect uncontrolled infection or heart failure early. The mean age of our surgical IE cohort was 42 years, lower than the 60–70 years typically reported in Western IE cohorts (29–31). This likely reflects regional differences in case mix and etiology. In high-income countries, population aging and prosthetic/device-related IE shift the age distribution upward. In contrast, native-valve disease and streptococcal infections remain more common in some regions (32). Therefore, our findings are most generalizable to younger surgical IE populations. Extrapolation to older or device-related cases should be done with caution and ideally validated externally.

To enhance comparability, we reported key descriptors: native vs. prosthetic valve IE, left- vs. right-sided involvement, blood-culture positivity, pathogen spectrum, embolic events, and vegetation/abscess characteristics. Age was adjusted in multivariable models (33, 34). For clinical communication, we prespecified PNI categories: >38 (normal), 35–38 (moderate malnutrition), and <35 (severe malnutrition). There is no universally accepted cut-off. Kahraman et al. reported a cut-off of ≈35.6 for IE (AUC 0.691), supporting our use of the lower range. In other cardiac surgical cohorts, higher thresholds have been reported–for example, ≈46.1 in adult cardiac surgery with CPB (35) and ≈43.4–48.3 in TAVR cohorts (36, 37). This reflects population- and procedure-dependence. We therefore analyzed PNI both categorically and continuously.

Our 1-year mortality model’s AUC of 0.61 indicates modest discrimination. PNI alone has limited utility for individual risk stratification and is better as an auxiliary signal. This is consistent with recent IE studies, where single markers show only moderate accuracy, while multivariable models integrating clinical, microbiological, imaging, and perioperative factors perform better (38, 39). From a nutritional assessment perspective, PNI is complementary to indices such as GNRI and CONUT. Their relative performance varies across populations. In cardiac surgery and structural-heart cohorts, malnutrition is consistently linked with poor outcomes. Recent meta-analyses suggest GNRI may perform as well as, or even better than, PNI for short-term mortality. PNI, however, provides a simple immune-nutrition signal. Combined or comparative use of metrics is preferable to reliance on one index (40, 41). Dynamic assessment may also improve prediction. Perioperative changes in PNI have been associated with outcomes, suggesting it should be considered a time-varying risk indicator. In practice, integrating PNI with clinical factors (age, NYHA class, CPB time), microbiological/echocardiographic data, and other nutrition scores is more realistic. Based on evidence, we position PNI as an adjunct to established risk models rather than a stand-alone tool.

5 Limitations

This study has several limitations. First, its single-center, retrospective design, combined with a limited sample size and a highly homogeneous population, may introduce selection bias. Second, it should be noted that the mean age of the IE population in this cardiac surgery cohort is relatively young. Therefore, we explicitly caution against overgeneralizing the conclusions to elderly IE populations or those with a high incidence of device-related IE, to avoid misapplication of the findings due to differences in population characteristics. Furthermore, the PNI, as a composite parameter derived from serum albumin and lymphocyte count–both of which are influenced by nutritional status and systemic inflammation–cannot fully distinguish the independent effects of nutritional deficits from those of inflammatory responses. Thus, the PNI should be interpreted primarily as a comprehensive inflammatory-nutritional risk marker rather than a pure nutritional indicator. Third, nutritional status was assessed only at admission without capturing dynamic changes during hospitalization. Fourth, the validation of the PNI-based predictive model focused primarily on discrimination, while comprehensive assessments of calibration and clinical utility remain lacking. Future prospective studies with larger sample sizes and external validation cohorts are warranted. These should incorporate more specific nutritional indicators and employ tools such as decision curve analysis to evaluate clinical net benefit, thereby facilitating the translation of such models into clinical practice.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

The studies involving humans were approved by the Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University (No. K202411-32). The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

Author contributions

KA: Writing – review & editing, Methodology, Formal analysis, Writing – original draft. ZL: Data curation, Writing – review & editing, Project administration, Writing – original draft. MD: Investigation, Writing – original draft, Data curation. AW: Investigation, Data curation, Writing – original draft. LS: Writing – original draft, Investigation. WZ: Writing – original draft, Project administration, Conceptualization, Supervision.

Funding

The author(s) declare that financial support was received for the research and/or publication of this article. This study was funded in part by Open Projects of the National Key Laboratory for the Etiology and Prevention of High-Incidence Diseases in Central Asia, co-built by the Province and the Ministry Fund (grant number: SKLHIDCA-2023-24).

Acknowledgments

We would like to thank all authors for their substantial work on data collecting and follow-up.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Correction note

A correction has been made to this article. Details can be found at: 10.3389/fnut.2025.1736985.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnut.2025.1685875/full#supplementary-material

References

1. Delgado V, Ajmone Marsan N, de Waha S, Bonaros N, Brida M, Burri H, et al. 2023 ESC guidelines for the management of endocarditis. Eur Heart J. (2023) 44:3948–4042. doi: 10.1093/eurheartj/ehad193

PubMed Abstract | Crossref Full Text | Google Scholar

2. Rajani R, Klein J. Infective endocarditis: a contemporary update. Clin Med. (2020) 20:31–5. doi: 10.7861/clinmed.cme.20.1.1

PubMed Abstract | Crossref Full Text | Google Scholar

3. Bullock A, Greenley S, McKenzie G, Paton L, Johnson M. Relationship between markers of malnutrition and clinical outcomes in older adults with cancer: systematic review, narrative synthesis and meta-analysis. Eur J Clin Nutr. (2020) 74:1519–35. doi: 10.1038/s41430-020-0629-0

PubMed Abstract | Crossref Full Text | Google Scholar

4. Felder S, Lechtenboehmer C, Bally M, Fehr R, Deiss M, Faessler L, et al. Association of nutritional risk and adverse medical outcomes across different medical inpatient populations. Nutrition. (2015) 31:1385–93. doi: 10.1016/j.nut.2015.06.007

PubMed Abstract | Crossref Full Text | Google Scholar

5. Lopez-Delgado J, Muñoz-Del Rio G, Flordelís-Lasierra J, Putzu A. Nutrition in adult cardiac surgery: preoperative evaluation, management in the postoperative period, and clinical implications for outcomes. J Cardiothorac Vasc Anesth. (2019) 33:3143–62. doi: 10.1053/j.jvca.2019.04.002

PubMed Abstract | Crossref Full Text | Google Scholar

6. Juliana N, Abd Aziz N, Maluin S, Abu Yazit N, Azmani S, Kadiman S, et al. Nutritional status and post-cardiac surgery outcomes: an updated review with emphasis on cognitive function. J Clin Med. (2024) 13:4015. doi: 10.3390/jcm13144015

PubMed Abstract | Crossref Full Text | Google Scholar

7. Buzby G, Mullen J, Matthews D, Hobbs C, Rosato E. Prognostic nutritional index in gastrointestinal surgery. Am J Surg. (1980) 139:160–7. doi: 10.1016/0002-961090246-9

Crossref Full Text | Google Scholar

8. Raposeiras Roubín S, Abu Assi E, Cespón Fernandez M, Barreiro Pardal C, Lizancos Castro A, Parada J, et al. Prevalence and prognostic significance of malnutrition in patients with acute coronary syndrome. J Am Coll Cardiol. (2020) 76:828–40. doi: 10.1016/j.jacc.2020.06.058

PubMed Abstract | Crossref Full Text | Google Scholar

9. Narumi T, Arimoto T, Funayama A, Kadowaki S, Otaki Y, Nishiyama S, et al. Prognostic importance of objective nutritional indexes in patients with chronic heart failure. J Cardiol. (2013) 62:307–13. doi: 10.1016/j.jjcc.2013.05.007

PubMed Abstract | Crossref Full Text | Google Scholar

10. Yoshihisa A, Kanno Y, Watanabe S, Yokokawa T, Abe S, Miyata M, et al. Impact of nutritional indices on mortality in patients with heart failure. Open Heart. (2018) 5:e000730. doi: 10.1136/openhrt-2017-000730

PubMed Abstract | Crossref Full Text | Google Scholar

11. Fowler V, Durack D, Selton-Suty C, Athan E, Bayer A, Chamis A, et al. The 2023 duke-international society for cardiovascular infectious diseases criteria for infective endocarditis: updating the modified duke criteria. Clin Infect Dis. (2023) 77:518–26. doi: 10.1093/cid/ciad271

PubMed Abstract | Crossref Full Text | Google Scholar

12. Nashef S, Roques F, Sharples L, Nilsson J, Smith C, Goldstone A, et al. EuroSCORE II. Eur J Cardiothorac Surg. (2012) 41:734–44; discussion 744–5. doi: 10.1093/ejcts/ezs043

PubMed Abstract | Crossref Full Text | Google Scholar

13. Goff D Jr., Lloyd-Jones D, Bennett G, Coady S, D’Agostino R, Gibbons R, et al. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American college of cardiology/American heart association task force on practice guidelines. Circulation. (2014) 129:S49–73. doi: 10.1161/01.cir.0000437741.48606.98

PubMed Abstract | Crossref Full Text | Google Scholar

14. Kahraman S, Zencirkıran Aguş H, Kalkan A, Uzun F, Ertürk M, Kalkan M, et al. Prognostic nutritional index predicts mortality in infective endocarditis. Turk Kardiyol Dern Ars. (2020) 48:392–402. doi: 10.5543/tkda.2020.25899

PubMed Abstract | Crossref Full Text | Google Scholar

15. Cheng N, Dang A, Lv N, He Y, Wang X. Malnutrition status in patients of very advanced age with nonvalvular atrial fibrillation and its impact on clinical outcomes. Nutr Metab Cardiovasc Dis. (2019) 29:1101–9. doi: 10.1016/j.numecd.2019.06.021

PubMed Abstract | Crossref Full Text | Google Scholar

16. Wada H, Dohi T, Miyauchi K, Jun S, Endo H, Doi S, et al. Relationship between the prognostic nutritional index and long-term clinical outcomes in patients with stable coronary artery disease. J Cardiol. (2018) 72:155–61. doi: 10.1016/j.jjcc.2018.01.012

PubMed Abstract | Crossref Full Text | Google Scholar

17. Chen X, Wei X, Huang J, Ke Z, Tan N, Chen J, et al. The prognostic nutritional index might predict clinical outcomes in patients with idiopathic dilated cardiomyopathy. Nutr Metab Cardiovasc Dis. (2020) 30:393–9. doi: 10.1016/j.numecd.2019.10.007

PubMed Abstract | Crossref Full Text | Google Scholar

18. Cho J, Shim J, Kim K, Lee S, Kwak Y. Impact of preoperative nutritional scores on 1-year postoperative mortality in patients undergoing valvular heart surgery. J Thorac Cardiovasc Surg. (2022) 164:1140–1149.e3. doi: 10.1016/j.jtcvs.2020.12.099

PubMed Abstract | Crossref Full Text | Google Scholar

19. Tang J, Yang L, Yang G, Li Y, Zhu Y, Li H, et al. Prognostic nutritional index as a predictor of cardiovascular and all-cause mortality in American adults with hypertension: results from the NHANES database. Front Cardiovasc Med. (2025) 11:1465379. doi: 10.3389/fcvm.2024.1465379

PubMed Abstract | Crossref Full Text | Google Scholar

20. Ding H, Zou X. Association of prognostic nutritional index with all-cause and cardiovascular mortality in adults with depression: NHANES 2005-2018. Front Nutr. (2025) 12:1599830. doi: 10.3389/fnut.2025.1599830

PubMed Abstract | Crossref Full Text | Google Scholar

21. Xu W, Zhang L, Yang Q, Cao Y, Rao R, Lv L, et al. Associations of prognostic nutritional index with cardiovascular all-cause mortality among CVD patients with diabetes or prediabetes: evidence from the NHANES 2005-2018. Front Immunol. (2025) 16:1518295. doi: 10.3389/fimmu.2025.1518295

PubMed Abstract | Crossref Full Text | Google Scholar

22. Huang S, Chen J, Chu T, Luo L, Liu Q, Feng K, et al. Pathogenic spectrum of infective endocarditis and analysis of prognostic risk factors following surgical treatment in a tertiary hospital in China. BMC Infect Dis. (2024) 24:1440. doi: 10.1186/s12879-024-10350-y

PubMed Abstract | Crossref Full Text | Google Scholar

23. Arques S. Human serum albumin in cardiovascular diseases. Eur J Intern Med. (2018) 52:8–12. doi: 10.1016/j.ejim.2018.04.014

PubMed Abstract | Crossref Full Text | Google Scholar

24. Wang Z, Zhang W, Chen L, Lu X, Tu Y. Lymphopenia in sepsis: a narrative review. Crit Care. (2024) 28:315. doi: 10.1186/s13054-024-05099-4

PubMed Abstract | Crossref Full Text | Google Scholar

25. Martínez-Urbistondo M, Gutiérrez-Rojas Á, Andrés A, Gutiérrez I, Escudero G, García S, et al. Severe lymphopenia as a predictor of COVID-19 mortality in immunosuppressed patients. J Clin Med. (2021) 10:3595. doi: 10.3390/jcm10163595

PubMed Abstract | Crossref Full Text | Google Scholar

26. Zhang K, Qin W, Zheng Y, Pang J, Zhong N, Fei J. Malnutrition contributes to low lymphocyte count in early-stage coronavirus disease-2019. Front Nutr. (2022) 8:739216. doi: 10.3389/fnut.2021.739216

PubMed Abstract | Crossref Full Text | Google Scholar

27. An P, Wan S, Luo Y, Luo J, Zhang X, Zhou S, et al. Micronutrient supplementation to reduce cardiovascular risk. J Am Coll Cardiol. (2022) 80:2269–85. doi: 10.1016/j.jacc.2022.09.048

PubMed Abstract | Crossref Full Text | Google Scholar

28. Li H, Li Y, Jia Z, Ren L, Nappi F, Ma X, et al. Postoperative follow-up of 221 patients with infective endocarditis from Gaoligong mountain area of Yunnan in China: a retrospective, single-center, observational cohort study. J Thorac Dis. (2024) 16:3325–37. doi: 10.21037/jtd-24-470

PubMed Abstract | Crossref Full Text | Google Scholar

29. Li M, Kim J, Sastry B, Chen M. Infective endocarditis. Lancet. (2024) 404:377–92. doi: 10.1016/S0140-673601098-5

Crossref Full Text | Google Scholar

30. Cresti A, Chiavarelli M, Scalese M, Nencioni C, Valentini S, Guerrini F, et al. Epidemiological and mortality trends in infective endocarditis, a 17-year population-based prospective study. Cardiovasc Diagn Ther. (2017) 7:27–35. doi: 10.21037/cdt.2016.08.09

PubMed Abstract | Crossref Full Text | Google Scholar

31. Jordal S, Kommedal Ø, Haaverstad R, Saeed S, Davidsen ES, Salminen PR, et al. Epidemiological and microbial trends of infective endocarditis in western Norway: a 7-year prospective observational study. BMC Infect Dis. (2024) 24:702. doi: 10.1186/s12879-024-09596-3

PubMed Abstract | Crossref Full Text | Google Scholar

32. Bielick C, Arnold C, Chu V. Cardiovascular implantable electronic device infections: a contemporary review. Infect Dis Clin North Am. (2024) 38:673–91. doi: 10.1016/j.idc.2024.07.004

PubMed Abstract | Crossref Full Text | Google Scholar

33. Okamoto H, Nishi T, Kamisaka K, Sasahira Y, Kanaoka K, Sumita Y, et al. Trends in the clinical characteristics and outcome of infective endocarditis: a nationwide study from 2016 to 2021. J Am Heart Assoc. (2025) 14:e037188. doi: 10.1161/JAHA.124.037188

PubMed Abstract | Crossref Full Text | Google Scholar

34. Weber C, Petrov G, Luehr M, Aubin H, Tugtekin S, Borger M, et al. Surgical results for prosthetic versus native valve endocarditis: a multicenter analysis. J Thorac Cardiovasc Surg. (2021) 161:609–619.e10. doi: 10.1016/j.jtcvs.2019.09.186

PubMed Abstract | Crossref Full Text | Google Scholar

35. Lee S, Ko K, Choi C, Park C, Park K, Son K. Does the prognostic nutritional index have a predictive role in the outcomes of adult cardiac surgery? J Thorac Cardiovasc Surg. (2020) 160:145–153.e3. doi: 10.1016/j.jtcvs.2019.08.069

PubMed Abstract | Crossref Full Text | Google Scholar

36. Panç C, Yılmaz E, Gürbak I, Uzun F, Ertürk M. Effect of prognostic nutritional index on short-term survival after transcatheter aortic valve implantation. Turk Kardiyol Dern Ars. (2020) 48:585–93. doi: 10.5543/tkda.2020.97709

PubMed Abstract | Crossref Full Text | Google Scholar

37. Ozilhan M, Açıkgöz S. The prognostic nutritional index in prediction of two-year mortality in patients undergoing transcatheter aortic valve replacement. Med Sci Discov. (2023) 10:684–9. doi: 10.36472/msd.v10i9.1029

Crossref Full Text | Google Scholar

38. Cho C, Lim K, Siu I, Ho J, Chow S, Fujikawa T, et al. Infective endocarditis risk scores: a narrative review. J Thorac Dis. (2025) 17:2662–78. doi: 10.21037/jtd-2024-2041

PubMed Abstract | Crossref Full Text | Google Scholar

39. Zi-Yang Y, Qi W, Liu X, Li H, Wang S, Yu D, et al. Mortality predicting models for patients with infective endocarditis: a machine learning approach. BMC Med Inform Decis Mak. (2025) 25:229. doi: 10.1186/s12911-025-03025-4

PubMed Abstract | Crossref Full Text | Google Scholar

40. Luo P, Shi K, Luo Y, Ren H. Prognostic value of the geriatric nutritional risk index in patients undergoing cardiac surgery: a systematic review and meta-analysis. Front Nutr. (2025) 12:1628671. doi: 10.3389/fnut.2025.1628671

PubMed Abstract | Crossref Full Text | Google Scholar

41. Lai A, Warrier M, Ng E, Lin C, Chin Y, Kong G, et al. Cardiovascular outcomes in acute coronary syndrome and malnutrition: a meta-analysis of nutritional assessment tools. JACC Adv. (2023) 2:100635. doi: 10.1016/j.jacadv.2023.100635

PubMed Abstract | Crossref Full Text | Google Scholar