Association between preoperative prognostic nutritional index and postoperative delirium in patients undergoing coronary artery bypass grafting: a two-center retrospective cohort study
Background Evidence regarding the association between preoperative prognostic nutritional index (PNI) and postoperative delirium (POD) specifically in patients undergoing CABG remains limited. This study aims to examine the relationship. Methods CABG patients extracted from the Medical Information Mart for Intensive Care IV (MIMIC-IV) and Fuwai Central China Cardiovascular Hospital (FWCC) were retrospectively included in the study. Participants were categorized according to the PNI score: low-PNI group (PNI ≤ 48) and normal-PNI group (PNI > 48). The POD was examined through Confusion Assessment Method for the ICU (CAM-ICU). The association between them was investigated by logistic regression, restricted cubic splines, as well as subgroup analysis. The area under the curves (AUCs), net reclassification index (NRI) and integrated discrimination improvement (IDI) were utilized to examine the predictive performance. Results The median PNI was 49.03 (44.96, 53.36) in the MIMIC-IV cohort and 49.75 (46.35, 52.60) in the FWCC cohort. The incidence of POD was 18.07% (215/1190) and 18.03% (156/865), respectively. Lower PNI was significantly associated with an increased risk of POD, and patients with low PNI had a markedly higher risk of POD than those with normal PNI (OR, 1.570; 95% CI, 1.113–2.214; P = 0.01) in the MIMIC cohort. Restricted cubic spline analysis further demonstrated an inverse association between PNI and POD. Adding PNI to the baseline prediction models improved predictive performance, with the most consistent improvement observed in IDI, whereas improvements in AUC and NRI varied according to model complexity. Similar findings were observed in the FWCC cohort, supporting the robustness and incremental predictive value of PNI. Conclusion Lower preoperative PNI was independently associated with higher odds of POD among patients undergoing CABG. Incorporating PNI into baseline prediction models provided modest and inconsistent incremental discrimination, most evident in IDI.
Authors
- 程兆云
- X Li
- Sheng Yue
- Peng Bao
- Xingwei Wang
- Yuzhen Sun
Institutions
- Zhengzhou University (CN)
- Zhengzhou Central Hospital (CN)
Publication Details
- Journal
- Frontiers in Medicine
- Published
- 2026-11-18
- DOI
- https://doi.org/10.3389/fmed.2026.2006346
- Primary Topic
- Intensive Care Unit Cognitive Disorders
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Science and Technology Department of Henan Province