Background: Several studies found associations between higher plasma calcium and phosphorus and mortality in dialysis patients. However, different predefined categories and reference values were applied and the precise shape of these relationships remains unclear.
Methods: We evaluated 1,621 patients from NECOSAD, a prospective multicenter cohort study of incident dialysis patients (60 +/- 15 years, 61% male, 64% hemodialysis). We used multivariate Cox regression and restricted cubic spline regression to study the effects of time-updated plasma concentrations on mortality in a flexible manner.
Results: 486 patients (30%) died during follow-up. Elevated phosphorus concentration was associated with higher mortality (p = 0.0009). The association of high calcium with mortality was borderline significant (p = 0.07). Within the studied ranges, we could not identify a threshold where an appreciable change in mortality risk occurred.
Conclusions: Mortality risk started to increase at a relatively low phosphorus concentration (4.5 mg/dl). Low-normal calcium combined with low-normal phosphorus concentration was associated with the lowest mortality.
Copyright 2008 S. Karger AG, Basel.