The geriatric nutritional risk index as a prognostic risk factor for critically ill patients with atrial fibrillation: a retrospective study based on MIMIC-IV and local hospital cohort external validation

Background Malnutrition is known to worsen outcomes in many diseases. This study explored its effect on prognosis in critically ill patients with atrial fibrillation (AF). Methods A total of 4,564 patients were studied. Their nutritional status was assessed using the Geriatric Nutritional Risk Index (GNRI), and they were divided into four groups based on GNRI quartiles, with a median of 100.23 and interquartile range of 90.31–111.27. The primary outcome was 28-day mortality. Kaplan–Meier analysis evaluated survival differences among the GNRI groups, and a multivariate Cox model assessed GNRI’s independent effect on mortality risk. A restricted cubic spline (RCS) analysis explored potential non-linear associations between GNRI and 28-day mortality. Results There is a negative correlation between GNRI and 28-day mortality, a finding supported by a log-rank test result with p < 0.001. According to the multivariate analysis, higher GNRI values were independently associated with a lower risk of 28-day mortality (quartile 4 compared to quartile 1, the hazard ratio (HR) was 0.78 with a 95% confidence interval (CI) of 0.62–0.97, and p = 0.029). Subgroup analysis showed no association in those with obesity, or in cardiovascular ICUs or post-cardiac surgery settings. No link was found in patients with a history of AF or on CRRT. However, the association was consistently seen in septic patients with new-onset AF. Additionally, the RCS analysis revealed a non-linear relationship between GNRI and 28-day mortality, showing that as GNRI increased, the risk of 28-day mortality decreased among patients. Findings from external validation cohorts also confirm this relationship. Conclusion Among critically ill patients, elevated GNRI scores are associated with improved outcomes in patients with new-onset AF, but this association is not present in patients with preexisting AF. In the subgroup of septic patients with new-onset AF, GNRI may not reliably predict clinical prognosis, emphasizing the need for further validation prior to clinical use. The GNRI provides a straightforward, unbiased assessment of both nutritional status and systemic inflammation burden.