Increased serum alkaline phosphatase as a predictor of long-term mortality after stroke

Background: Although the critical role of alkaline phosphatase in bone mineralization is clearly understood, the potentially adverse effect of high alkaline phosphatase levels on the cardiovascular system was only recently suggested. In this study, we hypothesized that increased levels of serum alkaline phosphatase may be associated with poor outcome after stroke in terms of mortality. Methods: We prospectively included patients with acute stroke admitted consecutively to our hospital, from October 2002 to September 2008. A total of 2,029 patients were selected for the analyses. In the analyses of mortality, the patients were divided by baseline measurements into quintiles of alkaline phosphatase levels (<57, 57-69, 70-81, 82-97, >97 IU/L). Results: In the Cox proportional hazard models, compared with the first alkaline phosphatase quintile, adjusted hazard ratios of the third, fourth, and fifth quintiles for all-cause death were 1.67 (95% confidence interval 1.12-2.49), 1.79 (1.20-2.67), and 2.83 (1.95-4.10). When we divided the patients into ischemic and hemorrhagic stroke, the association was also significant for both subtypes of stroke. In terms of vascular death, compared to the first alkaline phosphatase quintile, the adjusted hazard ratios of the fourth and fifth quintiles of alkaline phosphatase were 1.81 (95% confidence interval 1.14-2.86) and 2.78 (1.87-4.15). Conclusion: Our study demonstrated that increased serum levels of alkaline phosphatase are an independent predictor of all-cause and vascular death after either ischemic or hemorrhagic stroke.