U-shaped relationship between serum phosphate levels and mortality in critically ill patients with atrial fibrillation: Insights from the MIMIC-IV database

Abstract

Introduction

Atrial fibrillation (AF), a common arrhythmia in critically ill patients, significantly affects prognosis. While phosphate disturbances are strongly associated with cardiovascular risks, the impact of serum phosphate on critically ill AF patients' prognosis remains uncertain.

Methods

Utilizing data from the MIMIC-IV database, we retrospectively analyzed 11,744 critically ill AF patients. The primary outcome was 30-day all-cause mortality, with secondary outcomes at 90 days and 1 year. Cox regression models quantified the association between serum phosphate and mortality. Nonlinear associations were evaluated using restricted cubic splines (RCS), with inflection points further characterized through segmented Cox proportional hazards model.

Results

Mortality rates at 30 days, 90 days, and 1 year were 12.7 %, 14.4 %, and 15.8 %, respectively. Kaplan-Meier analysis showed higher mortality in patients with high phosphate levels. Adjusted Cox regression demonstrated that hyperphosphatemia independently predicted increased mortality at 30 days (HR: 1.55, 95 % CI: 1.36–1.78, p < 0.001), 90 days (HR: 1.55, 95 % CI: 1.37–1.76, p < 0.001), and 1 year (HR: 1.57, 95 % CI: 1.39–1.77, p < 0.001). RCS and two-piece Cox regression revealed a U-shaped nonlinear relationship between serum phosphate and mortality, with risk decreasing below a threshold and increasing above it. Similar patterns were observed across all time points.

Conclusion

Our findings demonstrated a U-shaped relationship between serum phosphate levels and mortality in critically ill AF patients, highlighting the importance of maintaining optimal phosphate levels in managing this population.