Association of serum alkaline phosphatase with mortality in non-selected European patients with CKD5D: an observational, three-centre survival analysis

OBJECTIVES: For characterisation of mineral bone disease in chronic kidney disease (CKD), laboratory surrogates have been suggested. This observational should investigate the association of total and skeletal alkaline phosphatase (AP) with mortality of patients undergoing maintenance renal replacement therapy. SETTING: Three renal outpatient centers in eastern-central Germany (secondary and tertiary care). PARTICIPANTS: Complete survival and laboratory datasets were available in 407 of 493 patients. Age and gender distribution was equivalent to a general population with end-stage CKD (CKD5D). Patients were included between 2008 and 2010 if at least 2 weeks of maintenance treatment were documented. PRIMARY OUTCOME MEASURES: Mortality was estimated by setting the end of dialysis date as event date. Events other than death (change to another centre, life-sustaining renal function or transplantation) were censored. RESULTS: The OR to die within follow-up for patients in the higher of two total AP strata was 2.70 (95% CI 1.76 to 4.15). In univariate Kaplan-Meier analysis, total AP had a strong association with all-cause mortality (LogRank=24.1, p<0.001). Mean total AP and individual lowest skeletal AP, but not mean skeletal AP entered step-wise Cox models for survival from dialysis start (χ(2)=22.4; p<0.001) after adjusting for age, Kt/V, diabetes and vintage. Mean values of skeletal, total AP and parathyroid hormone were 14.8±8.9 µg/L, 91.9±55.3 U/L and 188±164 ng/L, respectively. Skeletal and total AP were highly correlated (R=0.86; p<0.001). CONCLUSIONS: This unselected CKD5D population exhibited a clinical significant association of total AP with crude mortality and a stronger death risk association of total AP and individual lowest skeletal AP with crude mortality.