Serum calcium levels and long-term mortality in patients with acute stroke

Background: Calcium concentrations in serum are maintained within an exquisitely narrow range. Our aim was to examine the association between serum calcium and albumin-adjusted calcium (calcium_adj) levels and stroke outcome in a cohort of unselected patients with acute stroke. Methods: Consecutive patients hospitalized due to acute stroke (ischemic or intracerebral hemorrhage) throughout a large medical center were systematically assessed and followed for 1 year. Baseline total calcium and calcium_adj levels were collapsed into groups of low (<8.6 mg/dl), normal (8.7-9.9 mg/dl) and high (>10 mg/dl) levels and linear and quadratic relations with outcome were examined. Result: Among 784 patients (mean age 70.7 ± 12.5 years, 42.5% females), the mean ± SD total calcium level was 9.3 ± 0.6 mg/dl. For total calcium, the adjusted hazard ratio (HR) for all-cause death over 1 year was 1.83 [95% confidence interval (CI) 1.22-2.75] among patients with low versus normal levels. For calcium_adj, the adjusted HR for all-cause death among women was over 3-fold higher among patients with high calcium _adj levels versus those with normal levels (3.31; 95% CI 1.70-6.46), while no such associations were observed among men. In models developed to estimate the linear and quadratic relations, each unit increment in total calcium squared was associated with an increased adjusted HR of all-cause death over 1 year (p = 0.02) confirming nonlinear associations, and each unit increment in calcium_adj squared was associated with an increased adjusted HR of all-cause death over 1 year among women (p < 0.001) but not among men (p = 0.70). Conclusions: Serum calcium concentrations are a marker of mortality in acute stroke patients, but the associations are not linear, increasing at both extremes of calcium levels. Our findings suggest that long-term survival is optimal in a distinct range of serum calcium levels.