Background Gastrointestinal tumors are highly prevalent malignant tumors worldwide, and surgery is their primary treatment modality. Although the Geriatric Nutritional Risk Index (GNRI), a simple nutritional assessment tool, has been associated with prognosis in various diseases, its impact on ICU admission risk following gastrointestinal tumor surgery remains unclear. This study aims to investigate the relationship between preoperative GNRI and postoperative ICU admission risk. Methods This retrospective cohort study analyzed 9,045 gastrointestinal tumor surgery patients from the INSPIRE database. Multivariable logistic regression with stepwise confounder adjustment was used to examine the GNRI-ICU admission association. Generalized additive models explored nonlinear relationships, while subgroup and sensitivity analyses verified result robustness. Results Multivariable regression analysis demonstrated that preoperative GNRI decrease was significantly associated with increased risk of ICU admission in patients after gastrointestinal tumor surgery. The severe nutritional risk group (GNRI 82) had a 2.01-fold higher risk of ICU admission compared to the no-risk group (GNRI 98) (95% CI: 1.09–3.69). Nonlinear analysis identified GNRI = 89 as a key threshold ( p 0.001), beyond which each unit increase in GNRI reduced the risk by 12.6%. Subgroup analysis and sensitivity analyses both confirmed the robustness of the results. Conclusion Preoperative GNRI is an independent predictor for ICU admission following gastrointestinal tumor surgery, demonstrating a significant non-linear relationship with a critical threshold at GNRI = 89. The categorical risk stratification (GNRI 82) provides clinically meaningful information for perioperative risk assessment.
Lin et al. (2026) studied this question.