Lower Serum Magnesium Is Associated with Mortality in Severe COVID-19: A Secondary Analysis of a Randomized Trial

Many abnormalities in laboratory tests have been described in severe coronavirus disease (COVID-19), but most of them probably just reflect the degree of organ dysfunction and are not true risk factors for death. The present study is a secondary analysis of a clinical trial carried out in patients hospitalized due to severe COVID-19 (ClinicalTrials.gov identifier No. NCT04443673). We explored the association of clinical laboratory tests and serum cytokines with death in COVID-19 patients, either considering only the initial measurement obtained shortly after the patient's arrival at the emergency room, or by means of the weighted average of all measurements during the entire hospitalization. The study included 56 patients with a mean age of 58.6 years (range from 31.8 to 86.2 years), with a fatality rate of 58.9% (33 patients). Among initial laboratory tests, only mean corpuscular volume (MCV), erythrocyte count, serum magnesium, and age showed a trend (p < 0.10, univariable logistic regression) for an association with a fatal outcome. However, in the multivariable logistic regression, only MCV and magnesium remained associated with death, with adjusted odds ratios (95% confidence intervals) of 1.253 (1.047-1.501, p = 0.014) and 0.091 (0.010-0.798, p = 0.03), respectively. Serum magnesium tended to decrease during the hospital stay in both groups, survivors and non-survivors. Compared with survivors, patients who died had a higher weighted average of urea, blood urea nitrogen (BUN), procalcitonin, MCV, neutrophils, neutrophil/lymphocyte ratio, fibrinogen/albumin ratio, C-reactive protein/albumin ratio, BUN/albumin ratio, IL-6, and IL-10, as well as decreased weighted average of albumin, lymphocytes, and monocytes, among others. In conclusion, patients with severe COVID-19 who had lower serum magnesium on their arrival at the emergency room were more prone to die. On the other hand, serum magnesium tended to decrease during the patients' hospital stay, independently of the outcome.Trial Registration: ClinicalTrials.gov NCT04443673.