Hypermagnesemia is Associated with All-Cause Mortality in Patients with Chronic Kidney Disease

Abstract

Introduction: Hypomagnesemia has been associated with cardiovascular events and hypermagnesemia with overall mortality in the general population. However, in chronic kidney disease (CKD) the evidence is not as robust. The objective of our study was to investigate the relationship between magnesium levels and cardiovascular morbidity and mortality, all-cause mortality and the progression to end-stage kidney disease in a CKD population. Methods: Observational study of a cohort of 746 patients with CKD followed in a Nephrology outpatient unit. Baseline characteristics and analytical profile were collected at first visit, and patients were stratified according to their baseline magnesium levels into three categories: hypomagnesemia: 2.2 mg/dl. After a mean follow-up period of of 42.6 months, the following events were recorded: cardiovascular events, initiation of renal replacement therapy (RRT) and overall mortality of any cause or cardiovascular cause, and the relationship between magnesium levels and these outcomes was investigated. Results: 746 patients were analyzed, with a mean age of 70 ± 13 years, 62.9% were males, 45.2% had CKD stage 3 and 35.9% stage 4 at the start of follow-up. Mean baseline magnesium levels were 2.09 ± 0.33 mg/dl, and there was a close correlation between magnesium levels and serum creatinine, estimated glomerular filtration rate, phosphorus and PTH values. Calcitriol was associated with higher magnesium levels while calcium supplements and proton pump inhibitors were associated with lower magnesium levels. After a mean follow-up of 42.6 ± 19.6 months, 341 (45.7%) patients reached an event (Cardiovascular event, RRT initiation or death). Both patients with hypomagnesemia (Mg 2.2 mg/dl) had a higher risk of cardiovascular events (LogRank 4.45, p=0.035, and LogRank 7.45, p=0.006 respectively). In the multivariate analysis, they lost their predictive power. Patients with hypermagnesemia had a higher risk of all-cause mortality (Log Rank 13.11, p<0.001), while there was no association with hypomagnesemia. In the adjusted multivariate analysis, hypermagnesemia maintained its predictive power for all-cause mortality (HR=1.524, CI=1.002-2.319, p=0.049). After performing a propensity score matching for magnesium levels, we achieved two comparable groups of 94 patients, finding again a higher all-cause mortality in the hypermagnesemia group (Log Rank 17.48, p<0.001), that persisted in the Cox model adjusted for Calcium, Phosphorus and PTH. No association was found between magnesium levels and initiation of RRT. Conclusion: Magnesium values increase as CKD advances. Low magnesium levels predict cardiovascular events in patients with stage 3 and 4 CKD. On the other hand, patients with hypermagnesemia have a higher risk of all-cause mortality and cardiovascular events. Thus, with the available evidence to date, magnesium supplementation should be used with caution in these patients.