Chronic alcoholism is associated with a marked deficit in total magnesium, but little is known about the status of the physiologically active form, ionized magnesium. We assessed serum ionized magnesium (measured with AVL and Nova ion-selective electrodes) and total magnesium concentrations in chronic alcoholics at admission and after abstinence, and compared these results with those for a control group. At admission, the total and Nova- ionized magnesium concentrations in alcoholics were significantly lower than in controls. After 3 weeks of abstinence there was a significant increase in total magnesium concentration and both the AVL- and Nova- ionized magnesium concentrations. Total magnesium concentrations positively correlated with the AVL results in both alcoholics and controls, but only in controls with the Nova results. The altered status of ionized magnesium in alcoholics was clearly instrument-dependent. In another study we compared abnormal low and high AVL- and Nova-ionized magnesium results in serum samples from randomly selected patients. Major intermethod differences were found for samples with normal total magnesium concentration: most had abnormally low Nova results and normal AVL results. The agreement for the clinical interpretation of ionized magnesium results based on the reference interval for each method was only 32%. The extent of the inter-method difference was associated with a particular patient rather than with a specific diagnosis. Changes in serum total and ionized magnesium (measured with Nova ion- selective electrode) were monitored in 3 patients who transiently developed severe to profound hypomagnesemia due to cisplatin or interleukin-2 therapies. Independent of the etiology, the ionized magnesium fraction increased as the concentration of total magnesium decreased. When total magnesium was 0.35 mmol/L or less, the ionized magnesium concentration exceeded the total magnesium concentration. Similar overestimation of serum ionized magnesium was found for other randomly selected patients who had abnormally low concentrations of total magnesium, suggesting an error in the measurement of ionized magnesium with the Nova method. We assessed the effect of smoking on serum ionized magnesium concentration as measured with AVL and Nova ion-selective electrodes. A significant intra-method difference between smokers and non-smokers was found only for Nova results. A dose-dependent decrease in Nova results was observed with an increase in cigarettes/day. Nova results inversely correlated with white blood cell (WBC) counts. Nicotine, cotinine; arachidonic acid, and related metabolites did not affect the performance of either ion-selective electrode. Smoking may induce a serum factor, possibly related to WBCs, that negatively interferes with the response of the NOVA electrode.
