Abstract
BACKGROUND: To investigate the associations of serum magnesium, total calcium, and ionized calcium levels with ICU mortality, addressing conflicting evidence on electrolyte imbalances in critically ill patients. METHODS: This retrospective cross-sectional study analyzed 16,249 adult ICU patients from 208 U.S. hospitals (2014-2015) using the eICU Collaborative Database. Serum magnesium, total calcium, and ionized calcium levels were measured within 24 hours of ICU admission. ICU mortality was the primary outcome. Multivariate logistic regression, adjusted for 15 confounders (e.g., age, sex, APACHE scores, comorbidities), and restricted cubic spline (RCS) models assessed linear and non-linear associations, with subgroup analyses by disease severity. RESULTS: In fully adjusted models, ionized calcium showed a significant non-linear association with ICU mortality (OR: 0.90 per 1 mmol/L increase, 95% CI: 0.83 - 0.97, P = 0.009). Piecewise regression identified a threshold at 1.2 mmol/L: below this, each 1 mmol/L increase reduced mortality risk by 14% (OR: 0.86, 95% CI: 0.79 -0.94, P = 0.001); above it, risk increased by 97% (OR: 1.97, 95% CI: 1.12 -3.49, P = 0.019). This protective effect was stronger in patients with lower APACHE II scores (P-interaction = 0.021). Magnesium (OR: 0.95, 95% CI: 0.79 -1.14, P = 0.594) and total calcium (OR: 1.02, 95% CI: 0.93 -1.11, P = 0.689) showed no significant associations. CONCLUSIONS: Ionized calcium exhibits a U-shaped relationship with ICU mortality, with an optimal range near 1.2 mmol/L, particularly in less severe cases. These findings suggest prioritizing ionized calcium monitoring in ICU settings and warrant prospective validation.