This study retrospectively included patients who underwent radical gastrectomy for gastric cancer in the Department of Gastrointestinal Surgery at the First Affiliated Hospital of Guangxi Medical University from January 2019 to June 2023. All patient data were followed up by dedicated personnel for 30 days after the surgery. The inclusion criteria were as follows: (1) age ≥ 18 years; (2) confirmed diagnosis of gastric cancer and initial curative surgery; and (3) comprehensive medical history. The exclusion criteria were (1) preoperative adjuvant chemotherapy or radiotherapy; (2) concurrent primary tumors; (3) cases requiring emergency surgery because of gastric perforation and bleeding; and (4) cases undergoing concomitant resection of other organs during surgery. A total of 578 patients were included in this study, which employed a retrospective research design and obtained a waiver for informed consent. The study was approved by the ethics committee of the First Affiliated Hospital of Guangxi Medical University (2023-k293-01).

We collected several variables, including (1) general information: gender, age, BMI, smoking history, drinking history, comorbidities (hypertension, diabetes, coronary heart disease, and cerebral infarction); (2) laboratory parameters: albumin, hemoglobin, prealbumin, C-reactive protein, WBC, CA199, and AFP; (3) data related to surgical and tumor characteristics: gastrectomy, surgical approach, ASA class, TNM staging, tumor differentiation, intraoperative blood transfusion, operation time, intraoperative blood loss, preoperative nutritional intervention; (4) outcome variables: complications within 30 days after surgery. Use the following formula to calculate the NRI (11): NRI = 1.519 × albumin (g/L) + [41.7 × (weight/Wlo)]. Ideal weight (WLo) is calculated from height (H) and gender with the Lorenz formula (14). For men, Wlo (kg)=H (cm)−100−[(H−150)/4]. For women, Wlo (kg)=H(cm)–100−[(H−150)/2.5]. When the actual weight is greater than the ideal weight, the actual weight/ideal weight is taken as 1. The study population was divided into two groups according to the optimal cutoff value for the results of the ROC analysis: the low NRI group(NRI<96.7)and the high NRI group(NRI≥96.7). The Clavien–Dindo Classification system was used in categorizing postoperative complications into grades I, II, III, IV, and V (15). The presence of more than one complication in a patient was recorded as the highest level of complication. Frequent postoperative complications in patients with gastric cancer included pulmonary infection, gastrointestinal fistula, gastric atony, pleural effusion and others.

Continuous variables were presented as either mean ± standard deviation or median (interquartile range), whereas categorical variables were depicted as numbers and percentages (%). Intergroup comparisons were performed using t-tests, chi-square tests, and Fisher’s exact tests. The receiver operating characteristic (ROC) curve was used in determining the optimal cutoff value for NRI. Model 1 remained unadjusted, Model 2 was adjusted for age, sex, and BMI, Model 3 included adjustments for the variables in Model 2 and gastrectomy, surgical approach, intraoperative blood transfusion, operation time, and intraoperative blood loss, and Model 4 included adjustments for the variables in Model 3 and preoperative nutritional intervention. Subgroup analysis was conducted to confirm the reliability of the results, and stratification was based on gender, age (<60 and ≥60 years), BMI, gastrectomy, surgical approach, TNM staging, intraoperative blood transfusion and preoperative nutritional intervention. The cumulative incidence plot was employed, and a log-rank test was used to assess the difference in the cumulative risk of postoperative complications within 30 days between the low- and high-NRI groups. We excluded variables with missing values exceeding 20%, and for continuous variables with less than 5% missing values, the mean or median was used for replacement. All studies analyzed data using the statistical software packages R (http://www.R-project.org, R Foundation) and Free Statistics software version 1.8. For all analyses, a two-tailed P<0.05 was considered statistically significant.

A total of 656 patients underwent gastric cancer surgery between January 2019 and June 2023, and 578 eligible patients were finally enrolled for analysis after the exclusion of patients with other surgical procedures and missing key data ( Figure 1 ). The optimal cutoff value for NRI was 96.7 according to ROC curve analysis, and the study population was categorized into low-NRI (<96.7) and high-NRI (≥96.7) groups ( Supplementary Figure 1 ).

The baseline characteristics of the patients are summarized in Table 1 . The median age was 60.0 (52.0, 64.0), and 361 (62.5%) were male. A total of 120 (20.8%) patients suffered from postoperative complications. There were 84 (70%) pulmonary infections, 10 (8.3%) GI fistulas, 4 (3.3%) gastroparesis, 4 (3.3%) were pleural effusions, and 18 (15%) other cases in Supplementary Table 1 . The two groups were compared in terms of age, preoperative nutritional intervention, BMI, smoking, drinking, TNM staging, tumor differentiation, intraoperative blood transfusion, albumin, hemoglobin, prealbumin, C-reactive protein, and WBC. The differences were all statistically significant (P<0.05). In addition, patients in the low-NRI group presented a higher incidence of postoperative complications than those in the high-NRI group (28.2% vs. 13%), including grades II (22.4% vs. 7.7%),grade III (4.8% vs. 4.2%) and grade IV (1.0% vs. 1.1%). Additionally, the baseline characteristics of patients with and without complications are presented in Supplementary Table 2 . Patients with complications had lower NRI values and were more likely to be male compared with patients without complications. Patients with complications had a higher chance of intraoperative blood transfusion and blood loss and required longer operative times than those without.

Univariate analysis of the factors affecting 30-day postoperative complications is displayed in Supplementary Table 3 . Gender, age, NRI, albumin, hemoglobin, prealbumin, gastrectomy, preoperative nutritional intervention, intraoperative blood loss, operation time, surgical approach, and intraoperative blood transfusion may be associated with postoperative complications(P<0.05).

The results of the multivariate Cox regression analyses are presented in Table 2 , including unadjusted model 1, partially adjusted model 2, partially adjusted model 3,and fully adjusted model 4. In fully adjusted model 4, NRI was treated as a continuous variable, and a negative correlation was found between NRI and the incidence of 30-day postoperative complications. Specifically, for every 1 unit increase in NRI, a 7% decrease in the risk of 30-day postoperative complications was found (HR=0.93, 95% CI:0.90–0.96, P<0.001). This negative association was maintained in unadjusted model 1, partially adjusted model 2, partially adjusted model 3,and fully adjusted model4. When NRI was used as a categorical variable in the uncorrected model 1, the risk of complications at 30 days postoperatively was higher in the low NRI group compared with the high NRI group (HR=2.63, 95% CI: 1.71–4.03; P<0.001), and this association remained significant even after full adjustments for variables (HR=2.89, 95% CI: 1.71–4.88; P<0.001).

We further performed subgroup analysis to test whether the association between NRI and 30-day postoperative complications was stable across subgroups. As shown in Figure 2 , no significant interactions were observed in the subgroups in terms of gender, age, BMI, gastrectomy, surgical approach, TNM staging, intraoperative blood transfusion and preoperative nutritional intervention after adjusting for confounders (P for interaction > 0.05). This result indicates that the results of the negative association between NRI and the risk of 30-day postoperative complications in patients with gastric cancer remain robust regardless of baseline patient characteristics.

The cumulative incidence of complications at 30 days postoperatively showed that patients in the high NRI group (NRI ≥ 96.7) had a lower cumulative risk of 30-day postoperative complications compared with the low NRI group (NRI < 96.7) ( Figure 3 ).

A significant correlation was found between preoperative NRI levels and the incidence of 30-day postoperative complications in patients with gastric cancer. The incidence of 30-day postoperative complications in 578 gastric cancer surgery patients was 20.8%, and the incidence of postoperative complications was significantly higher in the low-NRI group than in the high-NRI group. After adjusting for confounders, the risk of 30-day postoperative complications was 2.89 times higher in the low-NRI group than in the high-NRI group. Apparently, preoperative NRI is an independent protective factor against 30-day postoperative complications in patients with gastric cancer.

Nutritional conditions have been recognized as an important factor affecting the prognoses of cancer patients, attracting considerable interest (16–18). Malnutrition is common in patients with gastric cancer, having prevalence rates ranging from approximately 29.1% to 80.4% (19–21), and is associated with poor outcomes (16–18). In some studies, NRI as an indicator of nutritional assessment is mostly used preoperatively for the prediction of the postoperative outcomes of related diseases (22, 23). Notably, it is implemented using serum albumin and weight to assess the nutritional status of patients and can be independent of subjective factors. NRI demonstrates strong prognostic value in various populations, such as in cardiac surgery (12), breast cancer (13), or head and neck cancer (24, 25). Its clinical outcomes in patients with gastric cancer are inconsistent. Kim et al. assessed preoperative NRI scores in 958 stage II and III gastric cancer and found that patients with higher NRI scores significantly outperformed those with lower scores in terms of survival (26). Song et al. (27)specifically investigated the prognostic ability of NRI in stage III gastric cancer patients, demonstrating that NRI had prognostic significance with an optimal cutoff value of 99, correlating with short survival times. Similar conclusions were drawn from other studies on gastric cancer (28–30). However, these studies examined patient survival and did not explore postoperative complications. A previous study (31) focused only on the relationship between low preoperative NRI levels and wound complications after gastrectomy (NRI≥97.5 vs. NRI<97.5: OR=0.653, 95% CI:0.326–0.974, P=0.014). Furthermore, Choi et al. (32)found through meta-analysis that those with gastric cancer identified as malnourished (NRI<97.5) had a higher risk of postoperative wound complications compared to those with good nutritional status (NRI≥97.5). However, these studies did not examine other common postoperative complications, such as pneumonia, pulmonary embolism, and heart failure. In this study, we assessed 30-day postoperative complications in patients with gastric cancer by adjusting for confounders, revealing a strong association between NRI and postoperative complications. The results in our subgroup analysis were consistent with the primary results, suggesting that the results of the independent effect of NRI on 30-day postoperative complications in patients with gastric cancer have high reliability and clinical applicability. Cumulative incidence analysis showed that patients with low NRI had a higher risk of postoperative complications compared with the high NRI group. In addition, we found no deaths (Complications: Grade V) within 30 days postoperatively in both the low and high NRI groups. This may be attributed to the relatively short duration of the study, the optimization of clinical management, and the limitations of the sample size. Previous study (27) suggest that the low NRI group may present a higher risk of death at longer follow-up (1, 3 and 5 years).

The NRI, a quantitative laboratory marker, integrates serum albumin levels and body weight, which reflect the nutritional status and degree of immune suppression in patients with gastric cancer (33, 34). Serum albumin is not only an indicator of nutritional status but is also strongly associated with systemic inflammatory status (35). Inflammatory factors released by tumors may affect liver function, inhibit albumin synthesis, and even lead to the structural degeneration of albumin, considerably reducing serum albumin levels (36, 37). In addition, body weight is an important indicator of nutritional status, which directly affects the tolerance of patients with gastric cancer patients (38). Body weight plays an independent predictive role in the prognoses of patients with gastric cancer (39, 40). A low NRI score implies malnutrition and immunosuppression in patients, thus leading to a low tolerance for surgery or chemotherapy and predisposed to poor clinical outcomes. In summary, malnutrition and immunosuppression are the likely main factors affecting poor clinical outcomes in patients with gastric cancer. In addition to these factors, NRI assessment may involve other unknown mechanisms affecting prognosis, which need to be explained by further in-depth studies.

The main significance of this study is that our findings suggest that NRI is strongly associated with 30-day postoperative complications in patients undergoing gastric cancer surgery. These findings will prompt healthcare professionals to identify high-risk patients and aid in the development of a preoperative nutritional management plan to improve the nutritional status of preoperative patients with gastric cancer and provide effective nutritional interventions. The aim was to reduce the rate of postoperative complications. In addition, our findings may contribute to future clinical studies of individualized nutritional control in gastric cancer patients to test whether the preoperative optimization of NRI can improve patients’ prognosis and adverse outcomes.

Our study has several limitations. First, it is a retrospective study, and despite our efforts to control potential confounding factors, we were unable to assess certain covariates that may have influenced the results. Second, the study only investigated the impact of NRI on early postoperative complications in patients with gastric cancer, without evaluating long-term postoperative outcomes. Third, it failed to assess the changes in NRI values after preoperative nutritional interventions and the possible impact on the cumulative incidence of postoperative complications.Despite adjusting for preoperative nutritional intervention status in multivariate models, the lack of standardized data on preoperative nutritional support limited our ability to assess its causal role. Future intervention studies should prioritize this variable to elucidate potential treatment effects. Lastly, we only collected clinical data from a single hospital, and thus the generalizability of the study findings may be limited. Future multicenter prospective studies are necessary to confirm our findings.