Problems in the assessment of magnesium depletion in the rat by in vivo 31P NMR.

Prior in vitro studies, utilizing 31P nuclear magnetic resonance (31P NMR) to measure the chemical shift (sigma) of beta-ATP and lengthening of the phosphocreatine spin-spin (T2) relaxation time, suggested an assessment of their efficacy in measuring magnesium depletion in vivo. Dietary magnesium depletion (Mg2+ decreases) produced markedly lower magnesium in plasma (0.44 vs 1.13 mmol/liter) and bone (130 vs 190 mumol/g) but much smaller changes in muscle (41 vs 45 mumol/g, P less than 0.01), heart (42.5 vs 44.6 mumol/g), and brain (30 vs 32 mumol/g). NMR experiments in anesthetized rats in a Bruker 7-T vertical bore magnet showed that in Mg2+ decreases rats there was a significant change in brain beta-ATP shift (16.15 vs 16.03 ppm, P less than 0.05). These chemical shifts gave a calculated free [Mg2+] of 0.71 mM (control) and 0.48 mM (Mg2+ decreases). In muscle the change in beta-ATP shift was not significant (Mg2+ decreases 15.99 ppm, controls 15.96 ppm), corresponding to a calculated free Mg2+ of 0.83 and 0.95 mM, respectively. Phosphocreatine T2 (Carr-Purcell, spin-echo pulse sequence) was no different with Mg2+ decreases in muscle in vivo (surface coil) (Mg2+ decreases 136, control 142 ms) or in isolated perfused hearts (Helmholtz coil) (control 83, Mg2+ decreases 92 ms). 31P NMR is severely limited in its ability to detect dietary magnesium depletion in vivo. Measurement of beta-ATP shift in brain may allow studies of the effects of interaction in group studies but does not allow prediction of an individual magnesium status.