Chronic kidney disease (CKD) is a major health problem worldwide.1 It is estimated that more than 10% of the world population suffers from CKD2 and it is expected to become one of the leading causes of death in the next decades.3 The progressive loss of renal function leads to a variety of complications affecting the cardiovascular system, bone health, inflammatory and metabolic regulation, and others. Renal dysfunction causes abnormalities in phosphate (P) and calcium metabolism.4 The abnormalities in calcium-phosphate regulation are largely responsible for the development of vascular calcifications and the increase in vascular stiffness in CKD,5 which in turn causes an increase in pulse wave velocity (PWV), an independent predictor of cardiovascular mortality in CKD.6

Experimental work7,8 has shown that magnesium (Mg) supplementation reduces vascular calcifications (VC). In CKD, hypomagnesemia is associated with cardiovascular and other comorbidities that have been summarized by Rodelo-Haad et al.9 In CKD patients, lower levels of Mg are associated with a two-fold increased risk of progression to end-stage renal disease.10 Moreover, in patients on renal replacement therapy, a better survival rate is observed in those patients with slightly high Mg levels.10 However, there are no clinical studies to substantiate the potential beneficial effect of Mg supplementation on vascular stiffness as assessed by PWV.

The effect of Mg on bone has been a matter of discussion. Clinical studies have suggested an association between oral Mg and fractures,11 yet Mg may promote osteogenesis by the activation of Notch signaling.12 Altogether, evidence suggests that Mg administration may have benefits in patients with CKD.

Hence, our objective was to analyze in CKD patients stages 3–4 the impact of the administration of Mg supplements in clinical aspects of renal disease such as vascular compliance as measured by PWV, VC, and parameters of mineral and bone health.

This study has evaluated the effect of the administration of a supplement of Mg for 15 months to patients with CKD stages 3–4 and VC. In particular, CKD progression, bone status, and parameters related to mineral metabolism and VC have been analyzed. The adherence to treatment was suggested by a significant increase in the urinary excretion of Mg in patients receiving Mg supplements; however, serum levels of Mg remained stable. Estimated glomerular filtration rate was not modified by the administration of Mg, as well as bone mineral density and bone markers of bone resorption or bone formation. Vascular calcification, assessed by X-rays, did not improve after Mg supplementation. Nevertheless, in patients receiving Mg, the serum level of Mg and also the FE Mg were inversely correlated with PWV, suggesting an improvement in vascular compliance. Interestingly, strong associations between FE Mg and PINP levels and PWV were also observed. Neither BMD nor VC were modified in the group of patients receiving oral Mg for 15 months.

Patients were treated daily with 360mg Mg carbonate (28.7% of elemental Mg). The administration of Mg carbonate, alone or in combination with calcium carbonate, has been already associated with no increase or, at least, not symptomatic increase in serum Mg levels.14–17 In our study, serum Mg levels remained unchanged in the treated group, with no reported undesirable side effects.

Despite no changes in serum Mg levels, we found that the Mg-treated arm exhibited a marked increase in the urinary excretion of Mg, expressed as FE Mg. Altogether, the results observed in both serum and urine Mg in the intervention group might be interpreted as the result of the replenishment of the intracellular Mg stores following supplementation, with the excess of Mg being eliminated by urine excretion. This supports the notion of the usefulness of urine as a source to estimate Mg status, as it has been recently reported.18 In this regard, it is important to consider whether Mg serum levels might accurately represent the body Mg content. As it has been previously published,19 total serum Mg might not precisely reflect the actual Mg availability. This is due to the fact that ionized Mg is the biologically active form and it may be affected by factors such as pH, the presence of other ligands, or even alterations in the stability of the sample.20 In addition, it has been reported that less than 1% of total Mg is found in serum.21 This fact might be particularly relevant in CKD where, despite showing hypermagnesemia due to low filtration induced Mg retention, intracellular Mg stores may not be repleted.9

One of the main goals of this work was to assess bone status following the administration of a Mg supplement. Bone mineral density was evaluated at three levels: column, femoral neck, and hip. The effect of Mg in bone has not been totally clarified so far and it has been suggested a relationship between Mg and osteoporosis.22 However, Spiegel et al. did not find significant changes in vertebral BMD in dialysis patients following the administration of Mg carbonate/Ca carbonate for 18 months.23 More recently, two systematic reviews and meta-analyses concluded that a higher Mg intake is associated with increased hip and femoral neck BMD.24,25 In our study, we did not find changes in BMD at the three locations evaluated. However, we determined the levels of PINP, a marker of bone formation,26 and although circulating PINP did not change after the follow-up, it showed a strong and positive correlation with urine Mg excretion. In this regard, it should be noted that the measurement of total PINP may be affected since monomeric PINP accumulates in CKD.27 However, this limitation was overcome in our study given that no changes in the GFR occurred throughout the study.

Furthermore, plasma levels of DKK1 were significantly reduced in those patients receiving oral Mg at the end of the study. The role of DKK1 in bone metabolism has been already studied. In fact, deletion of DKK1 is associated with an increase in bone formation,28 whereas DKK1 overexpression induces osteopenia.29 Thus, it could be hypothesized that Mg supplementation might induce an activation of Wnt/β-catenin pathway by reducing the Wnt activators DKK1 and SFRP1, thus contributing to better bone health. This notion is supported by the findings reported by us, showing that Mg acts as an osteoinductor in vitro, promoting the differentiation of mesenchymal stem cells into osteoblasts in a process that is mediated by Notch signaling.12 However, this effect seems to be dual depending on Mg concentration.30 Although it could not be ruled out that the bone response to Mg is also altered under uremic conditions, we have observed that the administration of high dietary Mg improved bone parameters in CKD rats, suggesting an osteoblastogenic effect similar to that found in vitro.8 Nonetheless, more studies are needed to clarify the mechanisms involved in this effect.

The development and progression of VC is one of the most frequent, complex, and serious complications in CKD. Calcification was assessed in our study by calculating the Adragao index in X-ray images of hands and pelvis. The impact of Mg administration on the development and progression of VC has been studied in both clinical and experimental approaches. In uremic animals, several works have repeatedly reported the effect that Mg exerts preventing and even reversing VC.8,31 Different physicochemical and cell-mediated mechanisms have been proposed to mediate this effect,32 but additional effects as an inhibitor of inflammation and oxidative stress at vascular level have been recently suggested.33 Clinical studies have shown non-uniform results. In two quite recent works, Sakaguchi and cols. found that oral Mg administration slowed down coronary arterial calcification, yet Bressendorf et al. failed to find any change in this parameter. Such differences may be attributable to methodological differences in the MAGiCAL-CKD trial.34,35 In our work, we did not detect any differences in VC between the control and the Mg-treated group, although the change over time in the intervention group showed a trend to be lower. Magnesium oxide was administered to patients in the work carried out by Sakaguchi, containing a substantially higher amount of elemental Mg (198mg) when compared with the Mg compound used in our study. In line with this effect at vascular level, we found an interesting association between serum and urine Mg and PWV, indicator of arterial stiffness. In this regard, several studies performed following different experimental approaches have reported variable results.36–38 Experimental studies have demonstrated that the administration of dietary Mg to uremic rats is associated with lower oxidative stress and reduced expression of inflammatory markers at vascular level, along with a better hemodynamic profile.33 Altogether, these results allow to hypothesize a possible direct effect of Mg on the vascular wall.

We must admit some limitations in our study. First, the study population comprised a low number of patients because of the strict inclusion criteria and the fact that some patients did not complete the study. In particular, some patients were excluded from the experimental group due to lack of adherence to treatment, which might have been due to the side effects attributed to the administration of low doses of Mg, such as abdominal pain and diarrheal stools. Second, although unintentionally, the number of women was limited in our study, which may lead to a gender bias. However, the dose of Mg administered covers the requirements for both male and female and therefore, similar effects derived from Mg supplementation might be expected in both genders. Third, the sum of the low sensitivity for longitudinal changes with the limited sample could be related to the extreme values that some patients had in the control group. Fourth, the follow-up was 15 months, period that may be considered relatively short compared with other studies. Finally, the Mg compound administered may contain lower elemental or bioavailable Mg in comparison with other molecules. Undoubtedly, these factors alone or in combination might have contributed to underestimate the effect of Mg on the parameters studied due to methodological aspects or lack of statistical power.

In sum, the results obtained here add information on the effects of Mg in parameters such as vascular stiffness, a key factor in the progression of CKD, and to the occurrence of undesirable cardiovascular events. Thus, it might be considered monitoring serum Mg levels and urinary excretion of Mg. Nevertheless, more studies are needed to fully understand the role of this element in the pathophysiology of CKD and its complications.

This work has been funded by the Spanish Society of Nephrology (Ayuda a la Investigación en Nefrología, Año 2018), the Programa Nacional de I+D+I 2013-2016 co-financed with European Funds (FEDER) from the Spanish Government (grants PI17/01010, PI18/00138, PI20/0660, and PI20/01645), Consejeria de Salud de la Junta de Andalucia (PI-0154-2017 and PI-0071-2021), Consejería de Transformación Económica, Industria y Conocimiento (PY20_00773), RICORS2040, and European Uremic Toxin Work Group (EUTox).

Authors have no conflicts of interest to declare.