Association between the geriatric nutritional risk index and postoperative delirium in gastric surgery patients: an analysis of the MIMIC-IV database

Abstract

Background: This study explores the correlation between nutritional status, as determined by the Geriatric Nutritional Risk Index (GNRI), and the incidence of postoperative delirium (POD) in patients undergoing gastric surgery.

Methods: Data were obtained from the MIMIC-IV 2.2 database for patients aged 18 years or older who underwent gastric surgery. Patients were categorized into the malnourished group (GNRI < 98) and the non-malnourished group (GNRI ≥ 98). Multivariable logistic regression was performed to assess the association between GNRI and POD, and various potential confounders were adjusted to ensure the robustness of the results. Non-linear relationships between GNRI and POD risk were evaluated through restricted cubic spline (RCS) analysis. Subgroup analyses were conducted to examine the effect of GNRI on POD across different patient categories, and interactions were calculated. Propensity score matching (PSM) was employed to reduce confounding bias.

Results: The study included a total of 4,818 patients, of whom 1,133 (23.5%) developed POD. Patients with a GNRI < 98 had a significantly higher risk of POD compared with those with a GNRI ≥ 98 (odds ratio (OR): 2.21, 95% confidence interval (CI): 1.93-2.53, p < 0.001). Even after adjustment for relevant confounders, GNRI remained significantly associated with POD (OR:1.24, 95% CI: 1.04-1.48, p < 0.001). This association was further supported by the results from PSM (OR:1.23, 95% CI: 1.01-1.51, p = 0.045). RCS analysis demonstrated a non-linear relationship between GNRI and POD risk (p < 0.05). Subgroup analyses revealed significant interactions within the cardiovascular disease, renal replacement therapy, benzodiazepine medication, and vasoactive drug subgroups (p for interaction < 0.05). After the patient population was adjusted to individuals aged 65 and older, this correlation remained significant (p for interaction < 0.05).

Conclusions: This study identifies a significant association between GNRI and the incidence of POD in patients undergoing gastric surgery. Improving nutritional status before surgery may lower the risk of developing POD.

Keywords: GNRI; Gastric surgery; MIMIC-IV database; Nutritional status; Postoperative delirium.

Fig. 1

Flowchart of Patient Selection. Flowchart for selecting patients from the MIMIC-IV database. Abbreviation: MIMIC-IV, Intensive Care Medical Information Market IV

Fig. 2

Restricted cubic spline plot between GNRI and delirium, with the red bold line indicating the odds ratio and the shaded area representing the 95% confidence interval

Fig. 3

Forest plot of subgroup analysis. Notes: Binary logistic regression is applied to evaluate the connection between POD and nutritional status in various subgroups. The results are shown as OR with 95%CI. Interaction p-values are computed through binary logistic regression to assess interactions between subgroups and nutritional status. Ca: Serum calcium; Na: Serum sodium; K: Serum kalium; Creatinine: Serum Creatinine; BUN: Blood Urea Nitrogen; Glu: Glucose; MI: Myocardial Infarction; CHF: Congestive Heart Failure; Ventilation status: Mechanical Ventilation; Dialysis type: Renal Replacement Therapy; Vas: Vasoactive Drugs; Dex: Dexmedetomidine; BZDs: Benzodiazepines; Prop: Propofol