BACKGROUND/AIMS: Although high serum alkaline phosphatase (ALP) levels were reported as predictive factors for death risk in dialysis patients on the basis of large databank analyses, the real scenario in a single hemodialysis (HD) center is unknown. METHODS: In this study, a 5-year cohort of 1126 prevalent HD patients in the largest HD center in Taiwan was studied. The associations of ALP levels expressed as baseline, time-average, and time-dependent with all-cause mortality and cardiovascular mortality were evaluated by using adjusted Cox regression models. RESULTS: At baseline, levels of serum parathyroid hormone, calcium, and liver enzymes are increased in parallel with ALP quartiles. The hazard ratio (HR) for all-cause mortality was significantly increased in time-average and time-dependent ALP quartile in the unadjusted Cox analysis. The significance disappeared when multivariate adjusted Cox analysis was used. Similarly, HR was not significantly increased for cardiovascular mortality with ALP quartile expressed as baseline, time-average, and time-dependent in three models of Cox analyses. CONCLUSION: Our study demonstrated that serum ALP levels were not associated with increased death risk in prevalent HD patients over a 5-year interval.
BACKGROUND/AIMS: Although high serum alkaline phosphatase (ALP) levels were reported as predictive factors for death risk in dialysis patients on the basis of large databank analyses, the real scenario in a single hemodialysis (HD) center is unknown. METHODS: In this study, a 5-year cohort of 1126 prevalent HDpatients in the largest HD center in Taiwan was studied. The associations of ALP levels expressed as baseline, time-average, and time-dependent with all-cause mortality and cardiovascular mortality were evaluated by using adjusted Cox regression models. RESULTS: At baseline, levels of serum parathyroid hormone, calcium, and liver enzymes are increased in parallel with ALP quartiles. The hazard ratio (HR) for all-cause mortality was significantly increased in time-average and time-dependent ALP quartile in the unadjusted Cox analysis. The significance disappeared when multivariate adjusted Cox analysis was used. Similarly, HR was not significantly increased for cardiovascular mortality with ALP quartile expressed as baseline, time-average, and time-dependent in three models of Cox analyses. CONCLUSION: Our study demonstrated that serum ALP levels were not associated with increased death risk in prevalent HDpatients over a 5-year interval.