AJH
1997;10:1420 –1424
Serum Ionized Magnesium
Relation to Blood Pressure and Racial Factors
Lawrence M. Resnick, Orit Bardicef, Bella T. Altura,
Michael H. Alderman, and Burton M. Altura
To study potential ionic factors predisposing to
vascular disease in hypertension, particularly
among black subjects, we used a recently
developed combined magnesium and calcium
specific, ion selective electrode apparatus to
measure extracellular ionized calcium (Ca-ion),
ionized magnesium (Mg-ion), and Ca-ion/Mg-ion
ratios in the serum of fasting, nonmedicated white
and black normotensive (n 5 61) and hypertensive
(n 5 23) subjects, studied consecutively in a
tertiary referral center.
Both race and blood pressure status had
independent effects on the distribution of Mg-ion
values. Although Mg-ion levels for the group as a
whole were lower in hypertensive versus in
normotensive subjects (0.571 6 0.012 v 0.601 6
0.005 mmol/L; P < .01), this was only true of white
subjects (0.579 6 0.021 v 0.620 6 0.006 mmol/L;
P 5 .0095). The lack of a significant difference in
Mg-ion levels between black hypertensive versus
normotensive subjects (0.553 6 0.012 v 0.577 6
0.007 mmol/L, P 5 NS) was attributable to the
significantly lower Mg-ion levels present in
normotensive blacks compared to those in
normotensive white subjects (0.577 6 0.007 v
0.620 6 0.006 mmol/L, P 5 .0001). Resultant Ca-ion/
Mg-ion ratios were elevated in all black subjects
and in white hypertensive subjects.
These data support the presence among
hypertensives and among black subjects
(independently of blood pressure) of a consistent
depletion of circulating magnesium and of an
imbalance of calcium and magnesium that may
potentiate vascular disease among these
subjects. Am J Hypertens 1997;10:1420 –1424
© 1997 American Journal of Hypertension, Ltd.
here is a well known excess of hypertension
and its sequelae in the black population. This
predisposition extends not only to the prevalence and severity of hypertension,1 but to
the organ system complications associated with it,
including left ventricular hypertrophy,2 hypertensive
nephropathy,3 stroke,4 and overall cardiovascular
T
mortality.5 Despite these consistent epidemiologic
data, the biologic basis underlying this excess morbidity and mortality in black populations remains undefined.
Our group has attempted to understand hypertension and its cardiovascular complications by studying
mineral ion metabolism and its hormonal and intra-
Received January 16, 1997. Accepted July 15, 1997.
From the Cardiovascular Center, New York Hospital-Cornell
Medical Center, New York, New York (LMR); Department of Physiology, State University of New York, Brooklyn, New York (BTA,
BMA); Department of Epidemiology, Albert Einstein College of
Medicine, Bronx, New York (MHA), and Division of Endocrinology
and Hypertension, Wayne State University Medical School, Detroit,
Michigan (LMR, OB).
Address correspondence and reprint requests to Lawrence M.
Resnick, MD, Division of Endocrinology and Hypertension, Wayne
State University Medical School, University Health Center– 4H, 4201
St. Antoine, Detroit, MI 48201.
© 1997 by the American Journal of Hypertension, Ltd.
Published by Elsevier Science, Inc.
KEY WORDS:
Magnesium, blood pressure, race,
calcium, blacks.
0895-7061/97/$17.00
PII S0895-7061(97)00364-6
AJH–DECEMBER 1997–VOL. 10, NO. 12, PART 1
cellular consequences.6 Specifically, we have demonstrated the presence of excess intracellular free calcium levels and suppressed free magnesium levels in
hypertension that were closely linked to the height of
the blood pressure, to the degree of cardiac hypertrophy, and to the hyperinsulinemia and insulin resistance characteristic of hypertension.7 Indeed, we have
suggested an ‘‘ionic hypothesis’’ in which hypertension, as well as its metabolic and cardiovascular consequences, are all different clinical manifestations of a
common defect in cell ion handling.8
We wondered to what extent these concepts might
also provide insights into the greater prevalence and
pathologic consequences of hypertensive vasculopathies among black subjects. Indeed, magnesium deficiency, as well as calcium excess have been implicated
in the pathogenesis of hypertension and atherosclerosis.9,10 Therefore, we studied the distribution of extracellular ionized calcium and magnesium in fasting
black and white subjects, using a newly available ionspecific electrode apparatus.11
Our preliminary results demonstrate that both elevated blood pressure and racial factors are associated
with an altered distribution of serum ionized magnesium, but not ionized calcium values. Thus, hypertension is associated with lower Mg-ion values, and black
individuals, independently of their blood pressure status, exhibit lower serum ionized magnesium levels
compared with white control subjects. We suggest that
the resulting increased calcium:magnesium ratio
among these groups may contribute to the incidence
and severity of vascular disease in black or hypertensive populations.
METHODS
Nonmedicated normotensive and essential hypertensive black and white subjects free of other known
diseases were studied after an overnight fast at the
Cardiovascular Center of the New York Hospital—
Cornell Medical Center. Normotensive subjects were
recruited from among a working population designed
to reflect the demographic characteristics of the general population in New York City, as part of a prospective study on the incidence of cardiovascular disease in a general population. Essential hypertensive
subjects were diagnosed on the basis of elevated blood
pressure readings found repeatedly to be .150/95
mm Hg, in the absence of physical diagnostic or laboratory evidence of secondary hypertension. Blood
samples for analysis (see below) were obtained as part
of ongoing protocol studies approved by the Scientific
Advisory and Human Rights committees.
Blood was drawn in the seated position in the Cardiovascular Center of the New York Hospital—Cornell Medical Center. Blood samples were analyzed for
serum total and ionized magnesium (Mg-ion) and for
SERUM IONIZED MAGNESIUM IN HYPERTENSION
1421
serum ionized calcium (Ca-ion). Total magnesium levels were measured by standard automated techniques.
Blood for serum ionized magnesium (Mg-ion) was
drawn into air evacuated glass tubes containing an
inert cell separating matrix. After clotting and centrifugation the tubes were inverted and serum was
drawn off into a syringe anaerobically, the latter being
capped and sent for analysis. A magnesium ion selective electrode (ISE) with a neutral carrier based membrane was used to measure serum ionized magnesium. As previously described,11 this new ISE for free
magnesium gives equivalent values of Mg-ion in samples of blood, serum, or plasma. The same selective
ion apparatus, equipped with a calcium-specific electrode, was also used to concurrently measure serum
ionized calcium values.
All data were analyzed by two-way analysis of variance (ANOVA), with statistical significance assessed
by post hoc t tests for all means (Bonferroni-Dunn) to
test the effects of both blood pressure status (hypertensive versus normotensive subjects) and race (black
versus white subjects) (Super Anova, Abacus Concepts, Berkeley, CA). All data are reported as the
means 6 SEM.
RESULTS
The age, sex distribution, weight, and renal function of
the hypertensive subjects did not differ significantly
between black versus white subjects, or between normotensive versus hypertensive individuals. Both systolic and diastolic pressure were elevated in hypertensive compared with normotensive controls, but did
not differ between black versus white subjects in either blood pressure group.
For the group as a whole, independent of race,
Mg-ion values were significantly lower in hypertensive (n 5 23, HiBP), compared with normotensive (n 5
61, NlBP) subjects (0.601 6 0.005 v 0.571 6 0.012
mmol/L, P , .01). When analyzed further, however, it
was observed that this difference was only significant
among white individuals (NlBP: 0.620 6 0.006 v HiBP:
0.579 6 0.021 mmol/L, P 5 .0095), but not among
black subjects, in whom Mg-ion levels among normotensive and hypertensive individuals did not significantly differ (NlBP: 0.577 6 0.007 v 0.553 6 0.012
mmol/L, P 5 .1609) (Table 1).
Indeed, the most significantly different results were
observed in normotensives, among whom black subjects exhibited significantly lower Mg-ion values than
did whites (0.577 6 0.007 v 0.620 6 0.006 mmol/L, P 5
.0001 (Figure 1A). There were no significant differences in male versus female subjects, between either
different blood pressure or racial groups. These results
were also observed when serum total magnesium values were analyzed; lower values were found among
black versus white normotensives (0.818 6 0.012 v
1422
AJH–DECEMBER 1997–VOL. 10, NO. 12, PART 1
RESNICK ET AL
TABLE 1. SERUM LEVELS OF TOTAL MG (MG-TOT), IONIZED MG (MG-ION), AND IONIZED
CALCIUM/IONIZED MG RATIOS IN NORMOTENSIVE (NlBP) AND HYPERTENSIVE (HiBP)
SUBJECTS GROUPED ACCORDING TO RACE
Group
White
NlBP
HiBP
Black
NlBP
HiBP
Mg-tot (mmol/L)
Mg-ion (mmol/L)
Ca-ion/Mg-ion
0.878 6 0.007
0.809 6 0.027
0.620 6 0.006
0.579 6 0.021†
1.955 6 0.019
2.104 6 0.071†
0.818 6 0.012*
0.781 6 0.016
0.577 6 0.007*
0.553 6 0.012
2.097 6 0.028*
2.191 6 0.037
* P , .0001 v white subjects; † P , .01 v NlBP.
0.878 6 0.007 mmol/L, P 5 .0001), but not hypertensives (0.78 6 0.02 v 0.82 6 0.02, NS).
Fasting serum ionized calcium levels, obtained concurrently, did not differ among any of the groups,
being the same for normotensives as a whole versus
hypertensives (1.21 6 0.004 v 1.21 6 0.012 mmol/L,
P 5 .713), for white normotensives versus hypertensives (1.21 6 0.004 v 1.22 6 0.015 mmol/L, P 5 .5217),
and for black versus white normotensives (1.21 6
0.005 v 1.20 6 008 mmol/L, P 5 .7632).
Ratios of Ca-ion to Mg-ion were also calculated as
an index of the net divalent ion contribution to vascular tone.12 A greater Ca-ion/Mg-ion ratio was found in
hypertension compared with normotensives for all
subjects (HiBP: 2.13 6 0.04 v NlBP: 2.02 6 0.02, P 5
.004), and among hypertensive v normotensive whites
(HiBP: 2.104 6 0.07 v NlBP: 1.955 6 0.02, P 5 .0025),
but not blacks (HiBP: 2.191 6 0.04 v 2.097 6 0.03, P 5
.1184). This was again because normotensive black
subjects already had an elevated Ca-ion/Mg-ion ratio
compared with similarly normotensive whites (blacks:
2.097 6 0.03 v 1.955 6 0.02, P 5 .0001) (Figure 1B).
DISCUSSION
Hypertension, as well as its clinical-pathological consequences, are more prevalent in black populations,
FIGURE 1. Serum ionized magnesium values (Mg-ion, left, A) and
the ratio of serum ionized calcium to
serum ionized magnesium (Ca-ion/
Mg-ion, right, B) in normotensive
white versus black subjects.
and are associated with an earlier onset, increased
severity, and greater target organ damage.1 Mild to
moderate hypertension is 1.5 to 2 times more prevalent in blacks than in whites, with severe hypertension
being 5 to 7 times more prevalent.1 The annual rate of
referral for end-stage renal disease in one study was
4.2 times greater for blacks than for whites, and in
hypertensives, was 17.7 times higher in blacks than in
whites.3 Left ventricular hypertrophy, by electrocardiographic criteria, exhibits a 2- to 4-fold greater prevalence in hypertensive blacks compared with their
white counterparts.2 Stroke is 2 to 3 times more likely
in black versus white populations, even without concurrent hypertension.4 Despite these consistent epidemiologic data, the biologic basis underlying this excess
morbidity and mortality in black populations has remained undefined.
Our approach to this problem has emphasized the
role of mineral ions in general, and of magnesium in
particular, in the pathophysiology of hypertension,
insulin resistance, and diabetes mellitus.6,7,9 We have
focused on magnesium as well as calcium because
magnesium is the major intracellular divalent cation,
regulating enzyme function, modulating cellular responsiveness to physiological stimuli such as calcium,
and thus steady state organ activities including car-
AJH–DECEMBER 1997–VOL. 10, NO. 12, PART 1
diac output, vasoconstrictor tone, pancreatic insulin
secretion, and peripheral insulin sensitivity.9,13 In normal populations, lower serum magnesium levels with
age are associated with higher blood pressures.14
Pathologically, both essential hypertension and non–
insulin-dependent diabetes mellitus (NIDDM) are associated with significantly suppressed intracellular
free magnesium levels,7 which, in turn, are inversely
related to the height of the pressure and the degree of
insulin resistance—the lower the magnesium, the
higher the pressure and the greater the hyperinsulinemic response to oral glucose loading. Consistent with
this, the ability of oral and parenteral magnesium
administration to lower blood pressure and to ameliorate diabetes have been long known and have recently
been rediscovered.15 Furthermore, experimentally induced magnesium deficiency can cause vasoconstriction, increase platelet aggregation, produce insulin resistance, and accelerate atherosclerosis.9,10,16 Altogether these observations led us to reason that
chronically lower steady state levels of magnesium, or
an imbalance between circulating calcium and magnesium,9,11 might predispose to and potentiate hypertensive, diabetic, or atherosclerotic vascular disease.
As such, we hypothesized that these alterations would
most likely be found in patient groups predisposed to
or suffering from these diseases. Our results suggest
that this may indeed be the case.
Technically, the lack of a routinely available, accurate, and physiological measurement of magnesium
has long hampered progress in this area. The measurement of free cytosolic magnesium has only recently
been possible with the advent of nuclear magnetic
resonance (NMR) spectroscopic techniques,17 and fluorescent probes. In the extracellular space, although
serum ionized calcium measurements have long been
available, we have only recently described a new, ion
specific electrode apparatus for measuring serum ionized magnesium in healthy subjects.11 Using this technique, we have demonstrated significant suppression
of circulating ionized magnesium in mild, fasting diabetic subjects, even when no such differences in serum total magnesium were apparent. These extracellular magnesium levels were, in turn, closely related
to concurrently measured intracellular free magnesium values.18
In this study, we used ion specific electrode techniques to measure both serum ionized calcium and
magnesium in normotensive and in hypertensive subjects. We observed: 1) lower circulating levels of serum
ionized magnesium in white individuals with essential hypertension, compared with normotensive white
control subjects; 2) racial differences in serum ionized
magnesium levels among normotensive subjects in
otherwise healthy black adults having lower serum
ionized magnesium levels compared with those of
SERUM IONIZED MAGNESIUM IN HYPERTENSION
1423
normotensive white control subjects; and 3) a higher
ratio of serum ionized calcium to serum ionized magnesium in hypertensive versus normotensive white
subjects and in black versus white normotensives.
These data further support the association of magnesium depletion with hypertension and suggest that
racial differences in magnesium status may also help
to explain the predisposition of black populations to
vascular diseases and their complications. This latter
notion is further supported by the finding of higher
Ca-ion/Mg-ion ratios in the same patient groups having lower absolute Mg-ion levels, as this ratio has been
directly related to the potential for arterial calcification
in experimental models of atherosclerosis.10,11
A few caveats need to be considered in interpreting
these preliminary data. First, because lower magnesium levels were found in black subjects in the absence
of overt hypertensive disease, magnesium depletion
of this modest degree cannot per se account for hypertension, with other as yet undefined concomitant
factors presumably needed for the emergence of elevated blood pressure. This is also consistent with suppressed levels of intracellular free magnesium found
in other nonhypertensive conditions such as non–insulin-dependent diabetes mellitus.6,8,18 We have not
assessed intracellular free magnesium values in a sufficient number of normotensive and hypertensive
black subjects to be able to determine whether the
equally low serum ionized magnesium values in normotensive and hypertensive black subjects are also
true for the intracellular ionic species. Thus, lower
magnesium levels, whether intracellular or extracellular, may be a necessary but not sufficient condition for
the onset of hypertension. Second, the use of serum
ionized magnesium measurements as an accurate reflection of tissue magnesium levels needs to be better
documented. The close correspondence between intracellular free magnesium levels, as assessed by NMR
spectroscopy in peripheral red blood cells, and the
concurrently measured serum ionized magnesium
level18 needs to be extended to the evaluation of intracellular free magnesium levels in organ systems
that are critical targets for hypertensive and diabetic
vascular disease, such as brain, heart, and kidney.
Preliminary reports in essential hypertensive subjects
suggest an intracellular free magnesium deficit in situ
in brain and skeletal muscle, similar to previous reports in peripheral blood cells.19 Third, these measurements, both serum ionized magnesium and intracellular free magnesium, as well as serum ionized calcium, all need to be made in the fasting state, as was
done here. This is because intracellular free magnesium and calcium levels change following oral glucose
ingestion and in vitro glucose and insulin incubation,20 which appears to be equally and inversely true
for serum magnesium and calcium in the extracellular
1424
AJH–DECEMBER 1997–VOL. 10, NO. 12, PART 1
RESNICK ET AL
compartment.21 Failure to evaluate subjects in the fasting state may thus obscure differences between hypertensive and normotensive subjects. Fourth, much
larger numbers of subjects will need to be evaluated to
confirm and extend these findings. Lastly, the pathophysiologic mechanism(s) underlying these data is
unknown, and measurements of dietary mineral intake, circulating mineral regulating hormones, other
markers of overall mineral balance, and other factors
not measured in this preliminary study, whether dietary– environmental or genetic–physiologic, need to
be considered in future studies.
intake modulates blood lipid levels and atherogenesis.
Proc Natl Acad Sci USA 1990;87:1840 –1844.
11.
Altura BT, Shirey TL, Young CC, et al: A new method
for the rapid determination of ionized Mg21 in whole
blood, serum and plasma. Meth Find Exper Clin Pharmacol 1992;14:297–301.
12.
Nakayama K, Fleckenstein A, Byon YK, FleckensteinGrun G: Fundamental physiology of coronary smooth
musculature from extramural stem arteries of pigs and
rabbits (electric excitability, tension development, influence of Ca, Mg, H and K ions). Eur J Cardiol 1978;
8:319 –335.
13.
Gueux E, Rayssiguier Y: The effect of Mg deficiency on
glucose stimulated insulin secretion in rats. Horm
Metab Res 1983;15:594 –597.
14.
Peterson B, Schroll M, Christiansen C, Tranavol I: Serum and erythrocyte magnesium in normal elderly
Danish people. Acta Med Scand 1977;201:31–34.
15.
Paolisso G, Sgambato S, Pizza G, et al: Improved insulin response and action by chronic magnesium administration in aged NIDDM subjects. Diabetes Care 1989;
12:265–269.
16.
Nadler JL, Malayan S, Luong H, et al: Intracellular free
magnesium deficiency plays a key role in increased
platelet reactivity in type 2 diabetes mellitus. Diabetes
Care 1992;15:835– 840.
REFERENCES
1.
Saunders E: Hypertension in blacks. Primary Care
1991;18:607– 624.
2.
Beaglehole R, Tyroler HA, Cassel JC, et al: An epidemiological study of left ventricular hypertrophy in the
biracial population of Evans County, Georgia. J Chron
Dis 1975;28:549 –559.
3.
Rostand SJ, Kirk KA, Rutsky EA, Pate BA: Racial differences in the incidence of treatment for end-stage
renal disease. N Engl J Med 1982;306:1276 –1279.
4.
Modan B, Wagener DK: Some epidemiological aspects
of stroke: mortality/morbidity trends, age, race, socioeconomic status. Stroke 1992;23:1230 –1236.
17.
Francis CK: Making a difference. Managing hypertension in minority patients. Am J Med 1990;88(suppl
3B):1S–2S.
Gupta RK, Benovic JL, Rose JB: The determination of
the free magnesium level in the human red blood cell
by 31P-NMR. J Biol Chem 1978;253:6172– 6176.
18.
Resnick LM, Altura BT, Gupta RK, et al: Intracellular
and extracellular magnesium depletion in non insulin
dependent diabetes mellitus. Diabetologia 1993;36:767–
770.
19.
Resnick LM, Bardicef O, Barbagallo M, et al: 31P-NMR
spectroscopic studies of oral glucose loading and in situ
skeletal muscle ion content in essential hypertension
(abst). Hypertension 1995;26:552.
20.
Resnick LM, Gupta RK, Laragh JH: Possible effects of
diet and other factors on 31P-nuclear magnetic resonance measurements of intracellular magnesium in hypertension. Clin Sci 1989;73:565–566.
21.
Rosenboom AL: Serum calcium and magnesium decline during oral glucose tolerance testing in children
with preclinical diabetes mellitus, less than in normals.
Metabolism 1977;26:1033–1039.
5.
6.
Resnick LM: Cellular ions in hypertension, insulin resistance, obesity and diabetes: a unifying theme. J Am
Soc Nephrol 1992;3:S78 –S85.
7.
Resnick LM, Gupta RK, Laragh JH: Intracellular free
magnesium in erythrocytes of essential hypertension:
relation to blood pressure and serum divalent cations.
Proc Natl Acad Sci USA 1984;81:6511– 6515.
8.
Resnick LM: Ionic basis of hypertension, insulin resistance, vascular disease, and related disorders: mechanism of syndrome X. Am J Hypertens 1993;6:123s–134s.
9.
Altura BM, Altura BT: Magnesium ions and contraction
of vascular smooth muscle: relationship to some vascular diseases. Fed Proc 1981;40:2672–2679.
10.
Altura BT, Brust M, Bloom S, et al: Magnesium dietary