31P nuclear magnetic resonance and zero-point titration compared for measuring free magnesium concentration in erythrocytes.

Intracellular ionized magnesium concentrations ([Mg2+]i) were measured in erythrocytes by 31P nuclear magnetic resonance (NMR) and zero-point titration in 14 controls and seven patients with renal magnesium loss. The mean intracellular ionized magnesium concentration in controls measured by 31P NMR was 0.20 (SD 0.03) mmol/L cell water, compared with 0.55 (SD 0.12) mmol/L cell water by zero-point titration. Total erythrocyte magnesium content measured with the lysate method was 0.63 mmol/L cell water higher than estimated by 31P NMR, probably because not all magnesium complexes are fully visible to the NMR technique. We found a positive correlation between plasma ultrafiltrable magnesium and [Mg2+]i irrespective of the [Mg2+]i assay used. [Mg2+]i measured with 31P NMR correlated modestly but significantly with [Mg2+]i determined by zero-point titration (r = 0.58, P less than 0.02). Washing erythrocytes before the zero-point titration decreased the ATP content and the cell water fraction, which led to overestimation of [Mg2+]i by zero-point titration. Although absolute values for [Mg2+]i differ with the assay used, both methods determined significantly lower values for [Mg2+]i in patients with isolated renal magnesium loss.