Ionised and total serum magnesium in renal transplant patients

Background: Hypomagnesemia in renal tran plant patients is almost always documented through total serum values (MgT), but it has recently become user-friendly to assay the biologically active, ionised fraction (Mg-). We verified the prevalence of true ionised magnesemia and the correspondence between total and ionised Mg assays in our transplanted patients, taking into account renal Mg excretion and the possible role of other reputed factors of hypomagnesemia (cyclosporine, secondary hyperparathyroidism and acid-base balance). Methods: Thirty-eight transplanted patients (25M/13F, aged 41+/-11 years) and 38 age and sex matched controls were enrolled. Blood chemistries included: ionised Mg and Ca, total Mg and Ca, phosphate, creatinine, albumin, bicarbonate, alkaline phosphatase, parathyroid hormone and, in patients, cyclosporine (CyA). A 24-h urine collection (for Ca and Mg) and a fasting spot sample (for pH, Mg, Ca, phosphate, creatinine) were also obtained. Results: Patients with mild renal failure (creatinine: Cr=1.75+/-0.83 mg/dL), mild persistent secondary hyperparathyroidism and almost normal tubular acidification capacity had MgT lower than controls (0.76+/-0.08 vs 0.82+/-0.08 mmol/L; p<0.002), with 10 cases (26%) of total hypomagnesemia. Mg- was also significantly low (0.51+/-0.08 vs 0.53+/-0.05 mmol/L; p<0.03), but there were only four cases (10%) of true ionised hypomagnesemia. MgT and Mg-, although correlated (with a low r value: =0.49; p<0.001), showed poor correspondence in individual patients and MgT was not useful to identify cases of true ionised hypomagnesemia. Neither assay correlated with renal function. Daily urinary excretion of Mg was normal (3.5+/-1.3 vs 3.0+/-0.24 mmol/day; n.s.), with no case of definite hypomagnesuria. Fasting excretion fraction (EF) of Mg, calculated with both assays, was increased in approximately 60% of patients (EFMgT 4.9+/-2.6 vs; 2.32+/-0.7%; p<0.0001; EFMg++ 7.74+/-4.9 vs 3.63+/-1.18%; p<0.0001) and positively correlated with serum Cr (r=0.62; p<0.0001 with EFMgT; and r=0.467; p<0.005 with EFMg++) but not with CyA. Neither Mg assay correlated with serum CyA, calcium, phosphate, PTH or bicarbonate. Conclusions: In long term renal transplant patients not taking diuretics, the prevalence of true ionised hypomagnesemia is low. Renal insufficiency, typically associated with Mg retention, is the major cause of increased EF,g and, as such, plays an antagonistic role to CyA and other factors of renal Mg wasting. Because MgT and Mg- are not closely related, assay of the ionised fraction seems advisable in case of total hypomagnesemia. However, because diagnosis of depletion can hardly rely on serum assay alone, a fuller evaluation (urinary excretion and other clinical and biochemical signs of hypomagnesemia) is suggested before diagnosis is made.