We thank Romano for his comments, which mainly focused on the well-established discrepancy between serum and tissue Mg levels. We concur with him that in many illnesses, including those he refers to, the serum Mg levels do not represent the status of body stores of this cation as reflected in measurements in muscle, erythrocytes, or lymphocytes. Also, we agree that the Mg loading test is a better predictor of Mg tissue levels than the measurement of the serum Mg. Reference to all the above-mentioned concerns was made in the original version of the manuscript, but space constraints prevented us from addressing each issue in the published article.
Nevertheless, our main goal was to investigate whether the serum Mg level could predict the in-hospital outcome of patients with an acute myocardial infarction, and the answer to this question is conveyed by the title of our article.1 Serum levels are often