Preoperative nutritional risk predicts postoperative complications but not delirium: Evidence from a multicentre...

Malnutrition is a prevalent modifiable risk factor in surgical patients. However, its specific contribution to postoperative delirium remains unclear. We hypothesised that preoperative nutritional risk would independently predict delirium and aimed to explore potential mediating pathways involving inflammation and surgical stress.

This prespecified secondary analysis of the multicentre observational study included 2422 adult patients undergoing major surgery. The primary exposure was preoperative nutritional risk. Outcomes included postoperative delirium (48 h), complications, and hospital length of stay (LOS). Associations were assessed using multivariable logistic and linear regression adjusted for confounders including baseline cognition and surgical load. Generalised additive models (GAMs) and parallel mediation analyses were employed to evaluate non-linear effects and potential mediators (haemoglobin, albumin, surgical load).

Nutritional risk was present in 117 patients (4.8% of the cohort). In fully adjusted models, nutritional risk was not associated with postoperative delirium (Odds Ratio [OR] 0.83; 95% CI 0.28-2.10; p = 0.72). Delirium was instead driven by age, baseline cognitive impairment, and surgical load. Conversely, nutritional risk was a strong independent predictor of postoperative complications (OR 2.08; 95% CI 1.31-3.29; p = 0.002) and prolonged LOS (adjusted increase ∼40%; p < 0.001). Mediation analysis revealed that while lower serum albumin predicted delirium, it did not mediate the relationship between nutritional risk and neurocognitive outcomes.

Preoperative nutritional risk is a robust marker of somatic vulnerability, predicting complications and delayed discharge, but is not an independent driver of postoperative delirium. Strategies to prevent delirium should target pathways distinct from those governing metabolic and physical recovery.