Comparison of two methods for the assessment of intra-erythrocyte magnesium and its determinants 65 3 and requires precision. However, despite the accuracy of the personnel, some PBS could have remained in the samples, which explains the systematic bias and subsequently, the lower concentrations measured by the direct method. The fact that the systematic bias is much lower after adjusting for erythrocyte count supports this hypothesis. The advantages of the indirect method, including lower costs and lower labor intensity, outweigh the relative small bias that was observed. In a randomized controlled trial, the effect of magnesium supplementation on IEM concentrations was studied. They showed that after a magnesium supplementation period of 3 weeks in women with low IEM concentrations (<1.97 mmol/L), plasma magnesium and 24-h urinary magnesium excretion were significantly increased, but no significant increase was found in IEM concentrations 26. The authors concluded that low IEM concentrations may not reflect systemic magnesium deficiency, but rather the activity of metabolic determinants, such as the Na+/Mg2+ exchanger 27. However, the supplementation period of 3 weeks might have been too short to detect differences in erythrocytes concentrations of magnesium. In the present study, dIEM, iIEM and plasma magnesium were not correlated with 24-h urinary magnesium excretion, as estimate of dietary absorption, suggesting that dietary magnesium intake is not directly reflected in the circulation and in cells, at least not in erythrocytes. Evidence regarding involvement of magnesium in glucose and insulin metabolism is accumulating 7. Previous studies showed that lower circulating magnesium concentrations were associated with increased risk of T2D in the general population 4,28. Furthermore, long-term magnesium supplementation improved insulin sensitivity in T2D patients 29, suggesting that the previously found associations are causal in nature. Magnesium concentrations in plasma, as well as concentrations in erythrocytes and platelets, were lower in diabetic patients and were even more reduced in diabetic patients with hypertension 30,31. We therefore hypothesized that glucose and HbA1c are inversely associated with IEM and plasma magnesium concentrations because of its potential to stimulate cellular glucose utilization. In the present study, higher glucose and HbA1c were independently associated with lower plasma magnesium concentrations. This is in line with previous studies in diabetic patients 32,33 and also with our recent findings in the general population 4,34. Although glucose and HbA1c were inversely associated with plasma magnesium, we have not observed such associations with IEM. Our findings may indicate that plasma magnesium, rather than IEM, is involved in glucose metabolism. However, future studies, preferable in patients with T2D who are at risk of hypomagnesaemia, should further elucidate whether
RkJQdWJsaXNoZXIy MTk4NDMw