The association between urine ketone and new‐onset atrial fibrillation in critically ill patients

Abstract BACKGROUND AND AIMS New‐onset atrial fibrillation (NOAF) is a common manifestation in critically ill patients. There is a paucity of evidence indicating a relationship between urinary ketones and NOAF. METHODS Critically ill patients with urinary ketone measurements from the Medical Information Mart for Intensive Care (MIMIC‐IV) database were included. The primary outcome was NOAF Propensity score matching was performed following by multivariable logistic regression. RESULTS A total of 24,688 patients with available data of urine ketone were included in this study. The urine ketone of 4014 patients was tested positive. The average age of the included participants was 63.8 years old, and 54.5% of them were male. Result of the fully‐adjusted binary logistic regression model showed that patients with positive urinary ketone was associated with a significantly lower risk of NOAF (Odds ratio, 0.79, 95% CI 0.7–0.9), compared with those with negative urinary ketone. In the subgroup analysis according to diabetic status, compared with nondiabetics, patients with diabetes had lower risk of NOAF ( p ‐values for interaction < 0.05). Results of other subgroup analyses according to gender, age, infection, myocardial infarction, and congestive heart failure were consistent with the primary analysis. CONCLUSIONS Positive urinary ketone body may be associated with reduced risk of NOAF in critically ill patients during intensive care unit hospitalization. Further studies are needed to clarify the underlying mechanisms.