Association between lactate-to-albumin ratio and mortality in obese patients with atrial fibrillation

The lactate-to-albumin ratio (LAR) is a composite biomarker reflecting both metabolic stress and nutritional status. This study aimed to explore its association with 28-day mortality in obese patients with atrial fibrillation (AF). We conducted a retrospective cohort study via the MIMIC-IV v3.1 database. The primary outcome was 28-day mortality. Cox proportional hazards models, LASSO regression, restricted cubic spline (RCS) modeling, and receiver operating characteristic (ROC) curve analyses were used to assess the prognostic significance of the LAR. Among 578 patients, the 28-day mortality rate was 21.5%. Nonsurvivors presented significantly greater LAR values (0.74 vs. 0.54, p < 0.001). In multivariate Cox regression adjusted for demographic, clinical, and therapeutic variables, an elevated LAR remained independently associated with 28-day mortality (HR: 1.20, 95% CI: 1.01–1.43, p < 0.05). RCS regression was utilized to flexibly model the continuous association between baseline LAR and 28-day mortality. Mortality risk increased continuously across higher LAR values, with no statistically significant departure from linearity detected ( P overall < 0.001, P non-linear = 0.212). Failure to reject the null hypothesis of linearity indicates that a linear Cox model provides a reasonable representation of the observed continuous risk continuum, rather than proving a strictly linear biological mechanism. The optimal LAR cutoff (1.185) yielded a specificity of 89.9% and a sensitivity of 33.9%. Kaplan–Meier analysis revealed significantly lower survival in patients with high LAR (HR: 3.80, 95% CI: 2.61–5.51, p < 0.001). Subgroup analyses demonstrated consistent prognostic performance across demographic and clinical strata. The LAR is independently associated with 28-day mortality in obese ICU patients with AF. Given its modest discriminative performance (AUC = 0.653), the LAR should be viewed as a modest adjunctive biomarker for risk stratification rather than a standalone clinical decision-making tool.