Magnesium in Health and Disease: Y. Itokawa & J. Durlach, eds. © 1989 John Libbey & Co Ltd. pp 191-197.
The reason we are interested in dietary magnesium is that reduced intake, primarily in the drinking water, is associated with an increase in the myocardial infarction death rate. There may also be an association with cardiac arrhythmias, hypertension, toxemia of pregnancy, and possibly some neoplastic diseases. Animal models of magnesium deficiency appear to be suitable for study of this condition. Much useful information has been obtained from such models, and we can expect much more to come. It is worth noting that a wide variety of animals, including cattle, dogs, rats, mice, and hamsters have been found to be vulnerable to magnesium deficiency in ways that are apparently similar to humans. The ease of producing deficiency, however, varies considerably among the species and also varies as a function of age. It is presumed, but not established, that the effects of species and age reflect differences in magnesium reserves and the kinetics of exchange between compartments. Most of the work that is discussed here has been carried out with hamsters. This shows its effects with clarity. We have used the dog for studies on myocardial infarction because the surgical procedures and other aspects of the methodology for the study of myocardial infarction are well established for this species. While an adult hamster shows severe signs of magnesium deficiency after 2 weeks on a diet devoid of added magnesium, adult dogs show only minimal signs after 3 months on such a diet.
It has been known for some time that severe magnesium deficiency in animals is associated with focal myocardial necrosis and decalcification (Heggtveit, 1969). Perhaps the intuitively most likely mechanism for this effect is depletion of tissue magnesium, which would lead to serious deleterious consequences because of the role of magnesium in many enzymatic reactions, especially those involving nucleotides. While depletion of tissue magnesium may be an important mechanism of injury in some tissues, it does not appear to be the case for heart muscle. We have shown (Chang and Bloom, 1985) that myocardial levels of magnesium are not affected by reduced dietary levels, even if the deficiency is severe enough to greatly lower the serum magnesium level, and even lead to death of the animal. Because of the methodology used (determination of magnesium levels in bulk myocardium), it is possible that changes in a small intracellular magnesium pool were missed. However, such an event remains to be demonstrated.
While consumption of a magnesium deficient diet may not lead to reduced myocardial levels of magnesium, it does lead to an increase in myocardial levels of both sodium and calcium. These changes area observed even in moderate magnesium deficiency, in the absence of myocardial necrosis (hamster on Mg10, dog on Mg0). Furthermore, it appears that the rise in sodium may precede the rise in calcium (Chang and Bloom, 1985). It has been inferred from the above observations that magnesium deficiency leads to reduced activity of the myocyte (Na,K)-ATPase, leading to an accumulation of sodium in cardiac myocytes (Bloom, 1988}, and that the high level of intracellular sodium potentiates sodium-calcium exchange, resulting in an increased level of intracellular calcium as well. This conclusion is also supported by studies on the kinetics of potassium flux across heart muscle cells (Madden et al, 1982). It must be emphasized that the degree of calcium accumulation can be great enough to cause morphologically obvious cardiac myocyte calcification, usually associated with cell death. The influx of calcium is presumably a critical event, since high levels of cytosolic calcium are associated with deleterious metabolic effects, including impaired ATP production (Bloom, 1981). Increased calcium influx therefore increases vulnerability of cells to injury. Sodium may be important in the same way, but the metabolic basis by which accumulation of sodium ions could be deleterious is not clear.
There are several possible tests of our basic hypothesis that consumption of a moderately magnesium deficiency diet leads to increased vulnerability to injury, and that such an increase is secondary to changes in sodium or calcium metabolism. We can ask, for example, if controlled amounts of ischemia produce more myocardial necrosis in magnesium deficient animals than in control animals. We have, in fact, found this to be the case (Chang et al, 1985). Furthermore, the amount of myocardial necrosis that was observed, per unit of muscle made ischemic, varied directly with the level of myocardial sodium as well as with the level of myocardial calcium, strongly implicating these in the pathogenesis of the increased vulnerability.
The basic hypothesis could also be tested by determining if magnesium-deficient animals are more vulnerable to the necrogenic effects of isoproterenol. As with myocardial infarction, this was found to be the case (Bloom, 1988). Since isoproterenol-induced myocardial necrosis is thought to occur through a calcium overload mechanism (Bloom and Davis, 1972; Bloom, 1981), the synergism of magnesium deficiency and isoproterenol in producing myocardial necrosis attains additional significance.
The efficacy of administered magnesium to patients with toxemia of pregnancy has implicated magnesium in the control of hypertension. Recent studies in which reduced levels of magnesium in serum or red blood calls correlated with the occurrence of hypertension (Dyckner and Wester, 1983; Resnick at al, 1984) further suggest that this is the case. We can hypothesize that consumption of a low dietary magnesium level leads to increased levels of calcium in vascular smooth muscle, just as it does in myocardium. The vasodilator actions of magnesium (Overbeck at al, 1969) also support this conclusion. Measurement of the vascular smooth muscle electrolyte levels has not been made. Such measurements are technically difficult, but not impossible. Our general theory regarding the relationship between dietary magnesium, tissue electrolytes, and vulnerability to injury leads us to hypothesize that magnesium deficiency might increase the tone of vascular smooth muscle, its vulnerability to injury, and if sufficiently severe, cause necrosis of vascular smooth muscle. The hamster is particularly sensitive to magnesium deficiency and might therefore ha particularly valuable as an indicator of vascular smooth muscle injury in this condition. We have examined severely magnesium deficient hamsters for arterial vascular lesions and found that these are indeed present. The most striking lesion found was fibrinoid necrosis of arteries and arterioles (Bloom, 1985). This lesion in hamsters was similar to that found in human toxemia of pregnancy.
If it is true that the mechanism by which MD leads to increased vulnerability to injury is through inhibition of (Na,K)-ATPase, leading to sodium accumulation, then the pathologic changes caused by magnesium deficiency should be aggravated by the administration of an (Na,K)-ATPase inhibitor during. Furthermore, if the importance of sodium accumulation is the rise in intracellular calcium it engenders, then the pathologic changes caused by magnesium deficiency should be reduced by the administration of a calcium channel blocking agent. We have found both of these to be true. Digoxin, an inhibitor of (Na,K)-ATPase, increased both the abundance and size of myocardial lesions due to magnesium deficiency. Nifedipine, a calcium channel blocker, had the opposite effect (Ahmad and Bloom, unpublished observations).
Experimental animals provide considerable evidence implicating increased levels of intracellular sodium and calcium in the pathogenesis of the effects of magnesium deficiency on the cardiovascular system. By extension, we predict that calcium overload is also important in humans suffering from magnesium deficiency. What has not bean demonstrated, however, is the occurrence of increased levels of sodium or calcium in the vascular smooth muscle or myocardium of humans consuming a suboptimal level of dietary Mg. For obvious reasons, it may not be possible to make such measurements. However, it may be possible to obtain some information on this point indirectly.
At the time of myocardial infarction, there is a systemic surge of sympathetic activity. One of the effects of adrenergic stimulation of the heart is an increase in cardiac myocyte calcium levels, leading to a positive inotropic effect. This may contribute to the metabolic demands on the heart and may, for this reason, actually be harmful. The protective effects of calcium channel blockers and beta adrenergic blockers in patients with myocardial infarction are presumably due, at least in part, to their effects in counteracting this sympathetic surge. The important point for the present discussion, however, is that this increase in the cardiac myocyte calcium level could lead to morphologically demonstrable deposits of calcium in myocytes, especially if the heart muscle was previously loaded with an abnormally large burden of calcium, as might be the case among people drinking water containing a low level of magnesium. The is, if humans are similar to animals in the effects of dietary magnesium on cardiac myocyte calcium metabolism, then we might expect to find myocyte calcification in the hearts of patients who die of acute myocardial infarction in regions where the drinking water contains low levels of dietary magnesium. The frequency of such myocyte calcification should be lower, according to this line of reasoning, in regions where the drinking water contains large amounts of magnesium.
To determine if this is the case, we have examined the hearts obtained at autopsy of patients dying with acute myocardial infarction, and scored them as positive or negative for myocardial calcification (Bloom and Peric-Golia, unpublished). We have done this for patients dying at the Veterans Administration Medical Center in Salt Lake City, Utah, and for patients dying at the George Washington University Hospital in Washington, DC. We selected these areas because published values for the water composition in these areas showed that the Salt Lake City water supply contains a much higher level of Mg than does the District of Columbia water supply, furthermore, in agreement with other data on the association between magnesium content of drinking water and the myocardial infarction death rate, the myocardial death rate has been found to be much lower in Utah than in the District of Columbia (Moriyama et al, 1971).
We examined autopsy material of patients who died with microscopically demonstrable acute myocardial infarcts. We took all such cases at the George Washington University Hospital for the year 1986. There were 23 such cases. We then took 23 consecutive cases at the Salt Lake City Veterans Administration Medical Center, beginning January 1, 1986, in which there was microscopically demonstrable acute myocardial infarction. Cases were examined in a double blind fashion.
The results were very clear. There were no cases of cardiac myocyte calcification among the 23 cases from the Salt Lake City area, while there were 15 such cases among the 23 from the Washington, DC area. This difference was highly significant. Of course there may be confounding factors. The Salt Lake City population was from a Veterans Administration hospital, and all were white males. The Washington, DC population, on the other hand, was composed of both men and women, white and black. Among the 8 white males that were in the Washington, DC group there were 5 cases in which cardiac myocyte calcification was found, and was significantly different from the Salt Lake City cases.
Several specific morphologic features of the calcium deposits in cardiac myocytes in the Washington, DC population are worthy of note. In all situations, the calcification was limited to cells. That is, there was no interstitial deposition of calcium which might suggest an artifact of processing. Some examples of calcification of old scar tissue were noted, but these were not scored as positive, since our criteria specified calcification of heart muscle cells. The most striking of our observations was the occurrence of calcification of myocytes both outside the area of infarction, as well as within the area of infarction. The is, apparently normal ('viable') cardiac myocytes, sometimes some distance from the area of infarction, showed dense accumulations of calcium in a granular form scattered throughout their cytoplasm. The granules of calcium were often distributed in such a way as to accentuate the cross striations of the myocytes. Within the area of infarction, calcification was noted as accumulations of granular calcium in necrotic myocytes. These, of course, showed no cross striations, and no nuclei. The granules of calcium were randomly distributed within these necrotic myocytes.
In addition to these two patterns of calcification, there were two cases in which granular calcium was found within the walls of small blood vessels. Some of these blood vessels were capillaries, and the calcium was present in the endothelium. Other blood vessels were larger but could not be identified with certainty as to type. In these, calcium granules were present both in the intima and elsewhere in the wall, possibly in vascular smooth muscle cells.
The finding of acute myocardial infarction associated calcification in myocytes of patients in the Washington, DC area, but not in the Salt Lake City, Utah area, implies that dietary magnesium affects the deposition of calcium in heart muscle. Of course other confounding factors may explain this finding, but no such confounding factor is apparent. For example, the difference in myocyte calcification could be related to the elevation above sea level, which is much greater in the Salt Lake City area than in the Washington, DC area, but there is no basis for such a belief. Nevertheless, such a possibility must be considered.
The significance of calcification of myocytes that are not necrotic must be addressed. It would be expected that if enough calcium were deposited in a myocyte to make it morphologically obvious, then it should be sufficient to produce a lethal metabolic perturbation. This is obviously not the case. It is concluded that these calcium deposits are sequestered in a compartment that prevents them from exerting their deleterious consequences. It is likely that such sequestration could only occur if the calcium influx occurred gradually, since rapid of a large amount of calcium appears to rapidly produce necrosis (Bloom, 1981). Thus, we conclude the patients in the Washington, DC area have baseline calcium myocyte levels that are elevated. This support the basic hypothesis that the increased myocardial infarction death rate in the Washington, DC population is due to reduced dietary intake of magnesium, that this relative magnesium deficiency exerts its deleterious effect through action on calcium metabolism.
One other thing that this finding suggests, is that animal models of magnesium deficiency are, indeed, appropriate, and that we say learn more about the human condition by studying them.
Bloom, S. and Davis, D (1972): Calcium as mediator of isoproterenol-induced myocardial necrosis. Am. J. Path. 69, 459-470
Bloom, S. (1951): Reversible and irreversible injury: Calcium as a major determinant. In: Cardiac Toxicology, ed T. Balazs, pp. 179-199 Boca Baton: CRC Press
Bloom, S. (1985): Coronary arterial lesions in Mg-deficient hamsters. Magnesium 4, 82-95
Bloom, S (1988): Mg deficiency cardiomyopathy. Am. J. Cardiovasc. Path. 2, 7-17
Chang, C. and Bloom, S. (1985): Severe Mg deficiency, electrolyte homeostasis, and myocardial necrosis in hamsters. J. Am. Coll. Nutrition 4, 173-185
Chang, C., Varghese, J., Downey, J. and Bloom, S. (1985): Magnesium deficiency and vulnerability to myocardial infarction. J. Am. Coll. Cardiol. 5, 280-289
Dyckner, T. and Wester, P. (1953): Effect of magnesium on blood pressure. Brit. Med. J. 286, 1847-1849
Heggtveit, HA. (1969): Myopathy in experimental Mg deficiency. Ann. NY Acad. Sci. 162, 758-765
Madden, J.D., Smith, G.A. and Llaurado, J.G. (1982): Myocardial K kinetics in rats on Mg deficient diet. J. Am. Coll. Nutr. 1, 323-329
Moriyama, I., Krueger, D. and Stamler, J. (1971): Cardiovascular Diseases in the United States. pp 49-118. Cambridge: Harvard University Press
Overbeck, H.W., Daugherty, R.M. and Haddy, P.J. (1989): Continuous infusion indicator dilution measurement of limb blood flow and vascular response to magnesium sulfate in normotensive and hypertensive men. J. Clin. Invest. 48, 1944-1956
Resnick, L.M., Gupta, R.K. and Laragh, J.H. (1984): Intracellular free magnesium in erythrocytes of essential hypertension: Relation to blood pressure and serum divalent cations. Proc. Natl. Acad. Sci. USA 81, 6511-6515
This page was first uploaded to The Magnesium Web Site on September 9, 2002