ABSTRACT
BACKGROUND: Abnormal chronic kidney disease-mineral and bone disorder (CKD-MBD) markers have been associated with adverse outcomes in hemodialysis (HD) patients. Dialysate calcium concentration (D-Ca) likely influences serum calcium and phosphorus levels. Optimal D-Ca level remains unclear. We hypothesized that higher D-Ca is associated with cardiovascular events and mortality among Japanese HD patients. METHODS: Enrollment data of chronic HD patients in the prospective observational study JDOPPS, phases 1-5 (1999-2015), provided exposures and covariates. All-cause mortality, non-arrhythmic cardiovascular events (NonAR-CVE), or their composites were analyzed by D-Ca, and divided into 2.5, 2.75, and 3.0 mEq/L. To minimize confounding by indication, analyses were restricted to facilities in which at least 90% of patients received the same D-Ca prescription. Association of D-Ca level with outcomes was evaluated in Cox models stratified by phase and accounting for facility clustering. Covariates describing patient demographics, comorbidities, laboratory values, CKD-MBD therapy, and facility attributes provided adjustment. RESULTS: Of 9,201 patients included, 25.0% had D-Ca of 2.5 mEq/L; 6.8% D-Ca 2.75; and 68.2% D-Ca 3.0. Median follow-up time was 2.03 years. D-Ca was not associated with all-cause mortality, with hazards ratios for 2.5 vs. 3.0 mEq/L of 0.90 and 95% CI (0.73-1.11), nor with other outcomes. One effect modification occurred, protective for lower D-Ca on NonAR-CVE in the absence of cardiovascular comorbidities (p = 0.032), although corresponding D-Ca effects were not significant after multiple comparisons adjustment (p = 0.261 [D-Ca 2.5] and 0.125 [D-Ca 2.75]). CONCLUSION: Lowering D-Ca level below 3.0 mEq/L seems not to have a meaningful effect on patient outcomes.