American Journal of Anal yt ical Chemistry, 2011, 2, 996-1002
doi:10.4236/ajac.2011.28117 Published Online December 2011 (
Copyright © 2011 SciRes. AJAC
Determination of Magnesium in Whole Blood and Serum
of Ischemic Heart Disease (IHD) Patients by Flame Atomic
Absorption Spectrometry
Hassan T. Abdulsahib
Chemistry Department, College of Science, University of Basra, Basra, Iraq
Received October 10, 2011; revised November 15, 2011; accepted November 23, 2011
Flame atomic absorption spectrometric determination of magnesium in whole blood and serum of ischemic
heart disease patients and control with different ages and sex was proposed. The limiting interfering phos-
phate/magnesium ratio have been estimates. 2% w/v AlCl3·6H2O was found to be very influential in remov-
ing phosphate interference effects. The detection limit was 0.065 μg/ml. Magnesium added to blood and se-
rum sample, and carried through this method may be recovered completely (96% - 100%) recovery percent-
age. The suggested method is simple, fast and selective. The statistical analysis of magnesium levels in blood
and serum showed that blood and serum magnesium levels in patients were lower than magnesium contents
of control group. Magnesium levels in blood and serum of males were significantly higher than females and
the magnesium levels were age independent. These findings indicate that there was an association between
blood and serum magnesium deficiency which can induce an entire array of path physiological phenomena
known to be important in ischemic heart disease.
Keywords: Magnesium, Whole Blood, Serum, Ischemic Disease, Atomic Absorption
1. Introduction
Magnesium is an essential micronutrient for humans and
plays many important roles, of principal importance are
its roles in the function of over 300 enzymes and in neu-
romuscular transmission [1]. Abnormal dietary deficiency
of Mg as well as abnormalities in Mg metabolism play
important roles in different types of heart diseases such as
ischemic heart disease [2]. Magnesium deficiency has
been associated with most of the major risk factors for
atherosclerotic coronary artery disease, coronary artery
spasm, and coronary artery thrombosis and many studies
in large population have shown that higher intake of this
mineral is associated with lower risk of high blood pres-
sure, stroke, and ischemic heart disease [3].
There is growing evidence that magnesium deficiency
is important in the pathogenesis of ischemic heart disease
[4]. In addition to the indirect epideomologic evidence,
there is experimental evidence that supports the role of
magnesium deficiency in ischemic heart disease, autop-
sies of patients who died from ischemic heart disease
have shown lower cardiac magnesium levels than autop-
sies of those who died from other causes [5].
Bersohn [6] also commented on the much greater sus-
ceptibility to ischemic heart disease of the African whites
than the blacks and considered the higher serum Mg of
the blacks a possibly significant factor.
Magnesium deficiency may usually be reflected in
low-magnesium diet, blood loss, excessive sweating,
drug and/or alcohol abuse or due to certain medication
use (such as loop diuretics and thiazides, cytotoxic drugs,
aminoglycosides, digoxin, steroids), or some physio-
logical conditions of over utilization of magnesium such
as pregnancy or infancy growth. Mental stress can also
lead to magnesiuresis due to high serum adrenalin [7].
Several analytical methods are normally needed for
the determination of magnesium in whole blood and se-
rum [8]. Even with determination of this element by
atomic absorption spectroscopy, various dilutions and
some anti interference reagents are required [9]. In this
work, a sensitive method for the determination of mag-
nesium in serum and whole blood is introduced, In which,
the interference effects of phosphate was completely
removed by the addition of (2% w/v AlCl3·6H2O). The
procedure is relatively simple, rapid and the addition of
anti interference reagents could easily lead to real values.
Therefore, the present study was elucidated some
physiological and biochemical actions of magnesium in
the mechanism of cardiac disease by the determination of
magnesium in whole blood and serum samples of ische-
mic heart disease patients using a flame atomic absorp-
tion spectrometric method.
2. Materials and Methods
2.1. Apparatus
Ashimadzu atomic absorption spectrophotometer model
(AA-630-12) was used with an air-acetylene burner (slot
dimensions 100 × 0.62 mm). Instrument settings were:
lamp current, 10 mA; wave length, 285.2 nm; slit width,
0.2 nm. These conditions were maintained constant
throughout the measurements.
2.2. Reagents
Magnesium stock solution Add 200 mg. of distilled
magnesium metal to a 100 ml volumetric flask. Dissolve
using the minimal quantity of hydrochloric acid. Dilute
to volume with deionized water. Magnesium working
standards: Transfer (0.5 - 5) ml of the stock solution to
ten a 100 ml. volumetric flask and dilute to volume with
deionized water. This gives a concentrations of (10 - 100)
μg Mg/ml.
Aluminum chloride solution Dissolve 8 gm. of re-
agent-grade aluminum chloride hexahydrate in 100 ml.
of deionized water.
2.3. Preparation of Samples
Thirty one patients with ischemic heart disease fulfilling
the Clinical care unit in Al-sader technical hospital. Twenty
three aged and sexes matched control group were taken
from apparently healthy people attending the hospital,
for pre-emplyment examination .For the groups, patients
and control, serum and whole blood magnesium were
measured upon presentation before starting treatment.
Pipet 1 ml of serum into a digestion tube. Add 2 ml of
concentrated nitric acid and 2 drops of concentrated per-
chloric acid. Place the tube in a hotplate at 130˚C to
150˚C. The initial temperature of hotplate is 120˚C.
When the volume is reduced to about 1 ml, turn the hot
plate to high 180˚C. Continue the digestion until the
perchloric acid fumes. Add 3 drops of nitric acid and
leave on the solution until dry salts start to form. Remove
the tube from the hotplate and wash down the walls of
the tube with approximately 2 - 4 ml. of deionized water.
Add 1 drop of 8% aluminum chloride. Place the tubes in
a boiling water bath for 5 min. Remove, cool, and dilute
to 12.5 ml. with deionized water. Centrifuge and transfer
two 2 ml aliquots of the supernatant liquid.
2.4. Statistical Analysis
The statistical analysis were carried out using two-way
analysis of variance with unbalanced repeated measure-
ments .Statistical significance between individual time
points was made by using Revised Least Significant
Difference (RLSD) test. The probability level for sig-
nificance was 5% less.
3. Results and Discussion
3.1. Interferences
The interference of ions present in serum in concentra-
tions greater than 0.5 μg/ml. was considered. These in-
clude Fe (0.1 - 1.8 μg/ml), Cu (1.96 - 2.63 μg/ml), Ca (90
- 110 μg/ml), NH4 (1.43 - 3.0 μg/ml), P (85 - 155 μg/ml),
S (22 - 56 μg/ml), Cl (3500 - 3710 μg/ml) [10], Zn (1.25
μg/ml) [11]. The sensitivity of the method makes it pos-
sible to use an aliquot equivalent to 1 ml. of serum and
the quantities of ions checked were based on these rela-
tionships. The interference, which appears to be incon-
sistent, is compensated by adding it to the magnesium
standard at the time of analysis. Phosphate was found to
cause the major interference, as shown in Figure 1. In an
effort to understand the nature of this interference, this
effect cannot be due exclusively to a magnesium phos-
phate complex since aluminium is affected by phosphate
even in the absence of magnesium. The interference is
not dependent on ionic strength.
Investigations to mask or remove the phosphate led to
Figure 1. Interferences effects of phosphate on (1 μg/ml)
magnesium absorbance signal.
Copyright © 2011 SciRes. AJAC
the discovery that aluminum effectively precipitates all
the phosphate. This was confirmed by spectrographic
analysis. The excess precipitant comes down as the
hydroxide. At least 1 mole of aluminum is required per
mole of phosphorus but large excesses must be avoided.
Boiling is necessary for complete precipitation and re-
duces the inclusion of magnesium in the precipitate.
This controlled precipitation is complete enough to
avoid such problems. At the levels used in the above
experiment, the effect of combinations of calcium
phosphate on the absorbance of magnesium was com-
pletely eliminated by addition 2% (w/v) AlCl3·6H2O as
shown in Table 1.
3.2. Digestion
Protein must be removed to avoid precipitation in the
acid medium during measurements. Wet-ashing seemed
to be the most feasible approach. It was found that 1 ml.
of serum could be completely digested with minimum
amounts of nitric and perchloric acids provided combus-
tion tubes were used. In order to remove ammonium salts
which form during the breakdown of protein, a few addi-
tional drops of nitric acid are added to the sample toward
the end of the digestion. All chlorides are volatilized
during this digestion. If the hot plate is preheated to di-
gestion takes about 3 hr.
3.3. Precision and Recoveries
The average deviation of 10 determinations on a single
serum sample was ±3.47% with 5 of these determina-
tions having a variability under 2%. Recovery studies
were made by adding known quantities of magnesium to
a serum, the results of these experiments, which are
summarized in Table 2, indicate that the over all recov-
ery in the analytical process is 96.0% - 100%.
3.4. Whole Blood and Serum Magnesium
Hypomagnesemia (defined as concentration of magne
Table 1. Effect of (2% w/v) Aluminium chloride on magne-
sium atomic absorption signal in whole blood samples.
Mean of absorbance
Whole blood
samples No. With AlCl3·6H2O Without AlCl3·6H2O
sium in serum less than the normal range; (0.7 - 1.1 mM,
1.7 - 2.5 mg/dl) [12] is often not present in patients with
chronic depletion of magnesium because of very slow
equilibration of magnesium among serum and blood
samples, which contains a small fraction of magne-
sium .These findings may explain at least part of the
beneficial effect of administration of magnesium in some
In our laboratory assessment of magnesium status
usually begins with measurement of concentration of
magnesium in the ischemic heart disease patients and
control blood and serum samples Table 3 show their
characters. All the measurements of the serum and blood
samples subjected to statistical analysis using the inde-
pendent samples RLSD-test to compare between meas-
urements of the study groups( patients and control ), sex
(males and females) and age groups (19 - 69 years).
The statistic results showed that there is a significant
difference (P < 0.001) in magnesium levels in all the
Table 2. Recovery percentage of the direct method of whole
blood samples.
Mg content
added, μg
Mg found
I 55.7
II 42.2
III 33.9
Table 3. Ischemic heart patients and charactors.
Patients groups Character
Age range (years)
Median age (years)
19 - 69
Associated diseases
Diabetes Millets
Thazide diuretics
Mg containing antacid
Copyright © 2011 SciRes. AJAC
compared males and females serum and blood measure-
ments as shown in Figure 2 and Figure 3 the concentra-
tion of magnesium whole blood and serum about 15%
excess in males higher than in females. From these find-
ings it would appear that while variation does occur in
concentrations of magnesium in whole blood and serum
with difference in sex, no definite pattern of distribution
is present.
Figure 2. Mg concentrations (μg/ml) in whole blood and
serum of control group.
Figure 3. Mg concentrations (μg/ml) in whole blood and
serum of IHD patients.
No statistically significant difference between age
groups were seen in the mean value of magnesium con-
centrations in blood and serum for both patients and con-
trol. (P < 0.01) as shown in Figure 4 and Figure 5.
Figure 6 illustrates that there is a statistical relation-
ship between the magnesium levels in whole blood and
serum of ischemic heart disease patient and control
group.The concentration of magnesium in patients serum
and blood was about 35% lower than in the control
Although this statistically significant difference, they
imply that less than one percent of the variation in serum
magnesium levels can be explained by differences in
dietary intakes.. This relationship provides a quantitative
Figure 4. Mg concentrations (μg/ml) in whole blood and
serum of control group.
Figure 5. Mg concentrations (μg/ml) in whole blood and
serum of IHD patients group.
Figure 6. Mg concentrations (μg/ml) in whole blood and
serum of IHD patients.
indication of the affinity of magnesium for a particular
clinical fraction.
There may be a further explanation for the post-infarct
decrease in serum magnesium concentration. Flink et al.
[13] correlated the decrease of serum magnesium with a
concomitant increase of free fatty acids. The rise in free
Copyright © 2011 SciRes. AJAC
fatty acid concentration is explained by an increased
lipolysis which is induced by the enlarged catecholamine
secretion involved in infarction. The same observation, a
low magnesium concentration at increased free fatty acid
concentration in serum was also found at ethanol with
drawal in dogs [14] and after infusion of adrenalin in
ewes [15]. Since the free fatty acids are able to bind
magnesium it has been assumed that serum magnesium
is thereby lowered [13]. Perhaps a more reasonable ex-
planation for lowering of serum magnesium during in-
creased lipolysis is an increased binding of magnesium
by free fatty acids within the adipocytes. In vitro ex-
periments with adipocytes support this view. Intact adi-
pocytes showed an increased magnesium uptake after
treatment with adrenalin [16].
Some aspects of the pathogenesis of cardiac disease
are reviewed in the light of current knowledge of the
physiological and biochemical actions of magnesium on
heart function.. In human during myocardial infarction,
there is a reduction of magnesium content even in non
infracted areas of the myocardium, followed by a tran-
sient reduction of serum magnesium concentration.
These effects are explained by the action of catechola-
mine on the myocardial cells, resulting in a loss of mag-
nesium accompanied by a catecholamine-induced of
many magnesium loss and/or increased lipolysis which
binds magnesium as magnesium soaps in the adiposities.
As a consequence serum magnesium may be decreased.
A reduced serum magnesium concentration may enhance
the action of catecholamines on the heart muscle as well
as the action of vasopressive hormones, thus provoking
contraction of coronary artery smooth muscle cells and
favoring the development of arrhythmia.
Although many patients with chronic heart disease are
magnesium depleted [17], a variety of superimposed
stressors cause additional decreases in concentrations of
magnesium in serum by redistribution [18]. Concentra-
tions of magnesium in serum decrease significantly dur-
ing Increasing concentrations of catecholamines also
may have contributed to the decrease. Further decreases
in concentrations of magnesium in serum are caused by
additional hemodilution, binding to albumin in the pump
prime, and redistribution secondary to catecholamine in-
duced increases in concentrations of free fatty acid [19].
A considerable number of experimental, epidemiol-
ogical and clinical studies are now available which point
to an important role of Mg2+ in the etiology of cardio-
vascular pathology. In human subjects, hypomagnesemia
is often associated with an imbalance of electrolytes such
as Na+, K+ and Ca2+. Abnormal dietary deficiency of
magnesium as well as abnormalities in magnesium me-
tabolism play important roles in ischemic heart disease.
Nonetheless, studies reporting Mg2+ assays on a vari-
ety of tissues suggest that many patients with cardiovas-
cular disease exhibit depletion of Mg2+ compared with
healthy individuals [17].
A commonly cited explanation for the clinical efficacy
of magnesium relates to its effect on the blood. Clini-
cians from the British Commonwealth [20] and Russia
[21] have reported that magnesium therapy of patients
with ischemic heart disease is associated with decreased
β-lipoproteins, increased α-lipoproteins, and increased
lecithin/cholesterol ratio, or a drop in serum cholesterol.
The efficacy of magnesium in ischemic heart disease
has been attributed to its antithrombotic activity, both
enhancement of fibrinolysis and inhibition of coagulation
being considered as mediating mechanisms. Substan-
tially increased fibrinolytic activity of the blood of car-
diac ischemia patients on Mg therapy has been reported
by Parsons et al. [22]. This observation recalls the ex-
planation of the efficacy of the earlier use of Mg in pre-
vention of postoperative thrombosis [23] and in treat-
ment of peripheral thrombotic disease [24]. The anti-
thrombotic effect of magnesium has also been attributed
to stabilization of platelet membranes [25] and to inhibi-
tion of platelet aggregation [26]. Supportive of the plate-
let membrane theory is recent work that has shown that
Mg is necessary to maintain the disc shape of platelets
[27]. Addition of MgCl2 to fresh human blood, under
conditions that maintained electrolytes and enzyme sys-
tems as nearly normal as possible, resulted in reduction
in the size and number of platelet clumps and an increase
in the number of discrete platelets [28].
Unfortunately, it is difficult to evaluate the true human
requirement of magnesium: depending age and protein
intake the values given by various authors range from 0.1
to 0.4 mmole/kg per day [29]. This may be partially due
to considerable individtial variation in intestinal absorp-
tion and renal excretion of magnesium. For instance, the
normal magnesium intake via food and tap water by re-
ceiving an additional 200 - 600 mg magnesium, which
increased serum magnesium by approximately 10% [30].
Although this result is not consistent with the tendency
of magnesium to be associated with the differences in
dietary intakes. This relationship provides a quantitative
indication of the affinity of magnesium for a particular
clinical fraction, magnesium deficiency may contribute
to pathological processes [31]. Clinicians should con-
sider using magnesium supplementation to prevent defi-
ciency in patients at risk and to treat deficiency when it
occurs.When the individual magnesium intake from food
is low, e.g. 200 - 250 mg, then the additional daily mag-
nesium intake in magnesium rich regions, e.g. 50 mg,
may have a “cardio protective” effect while in magne-
sium poor regions a chronic magnesium deficiency is
likely to develop [32].
Copyright © 2011 SciRes. AJAC
The inability of the senescent myocardium to respond
to ischemic stress could be due to several reasons. mag-
nesium supplemented K+ cardioplegia modulates Ca2+
accumulation and is directly involved in the mechanisms
leading to enhanced post ischemic functional recovery in
the aged myocardium following ischemia. While many
of these mechanisms remain controversial and in some
cases speculative, the beneficial effects related to conse-
quences of magnesium supplementation are apparent.
Further research is needed for the incorporation of these
findings toward the development of novel myocardial
protective role of magnesium to reduce morbidity and
mortality of patients suffering from a variety of cardiac
3.5. Conclusions
The development of procedures for the direct analysis of
magnesium in blood and serum remains a great challenge.
The insufficient sensitivity and vulnerability to matrix
interferences of these technique can be overcome by the
addition of 2% (w/v) AlCl3·6H2O which was responsible
for the disappearance of phosphate interference. Since
the method described in this study gave correct results
for magnesium, it can be used for the direct determina-
tion of magnesium in whole blood and serum .Results
obtained for magnesium levels in whole blood and serum
show that a direct inversely correlation observed between
magnesium levels and ischemic heart disease patients,
and a significance difference observed between magne-
sium levels and the sex of individuals, but no correlation
observed between magnesium levels and age. The data
indicate that Mg deficiency can lead to ischemic heart
disease because its essential role in a wide range of fun-
damental cellular reactions in patients with ischemic
heart disease. Furthermore, the therapeutic value of Mg
in the management of coronary risk factors and ischemic
heart disease has been clarified. Dietary Mg supplemen-
tation should be considered as a preventive element in
ischemic heart disease.
4. Acknowledgements
The researcher would like to thank all the staff, dectors,
nurses, laboratory technician, in the Al-Sader technical
hospital for their kind assistance and providing facilities
during data and samples collection procedure. The au-
thors also would like to express thanks to N. A. Awad for
his helpful comments on the manuscript.
5. References
[1] National Research Counsil, “Diet and Health,” National
Academy Press, Washington DC, 1989.
[2] S. Chakraborti, T. Chakraborti, M. Mandal, A. Mandal, S.
Das and S. Ghosh, “Protective Role of Magnesium in
Cardiovascular Disease: A Reiew,” Molecular and Cel-
lular Biochemistry, Vol. 238, No. 1-2, 2002, pp. 163-179.
[3] N. E. Saris, E. Mervaala, H. Karppanen, J. A. Khawaja
and A. Lewenstam, “Magnesium. An Update on Physio-
logical, Clinical and Analytical Aspects,” Analytica
Chimica Acta, Vol. 294, No. 1-2, 2000, pp. 1-26.
[4] G. E. Burch and T. D. Giles, “The Importance of Magne-
sium Deficiency in Cardiovascular Disease,” American
Heart Journal, Vol. 94, No. 5, 1977, pp. 649-657.
[5] P. C. Elwood and W. H. Beasley, “Myocardial Magne-
sium and Ischaemic Heart Disease,” Artery, Vol. 9, 1981,
pp. 200-204.
[6] I. Bersohn, “Atherosclerosis and Coronary Heart Disease:
Possible Incrimination of Magnesium Deficiency in Their
Promotion,” Medical Procedures, Vol. 4, 1958, p. 62.
[7] M. S. Seelig, “Possible Roles of Magnesium in Disorders
of the Aged,” In: R. W. Gelson and F. M. Sinex, Eds.,
“Intervention in the Aging Process. Part A: Quantitation,
Epidemiology,” Clinical Research, AR Liss, Inc, New
York, 1983, pp. 279-305.
[8] J. G. Wang, “Analysis of Clinical Biochemistry,” Hunan
Scientific & Technological Publishing Company, Chang-
sha, 1980.
[9] “Analytical Methods for Flame Spectroscopy,” Varian
Techtron Pty. Limited, Mulgrave, 1979.
[10] P. W. Sunderman and F. Boerner, “Normal Values in
Clinical Medicine,” Saunders, Philadelphia, 1950.
[11] W. S. Spector, “Handbook of Biological Data,” Wright
Air Development Center, New York, 1956.
[12] R. Elin, “Laboratory Tests for Assessment of Magnesium
Status in Humans,” Magnesium and Trace Elements, Vol.
10, 1991-1992, pp. 172-181.
[13] E. B. Flink, J. E. Brick and S. R .Shane, “Alterations of
Long-Chain Free Fatty Acid and Magnesium Concentra-
tions in Acute Myocardial Infarction,” Archives of Inter-
nal Medicine, Vol. 141, No. 4, 1981, pp. 441-443.
[14] E. B. Flink, S. R. Shane, R. R. Scobbo, N. G. Blechschmidt
and P. McDowell, “Relationship of Free Fatty Acids and
Magnesium in Ethanol Withdrawal in Dogs,” Metabolism,
Vol. 28, No. 8, 1979, pp. 858-865.
[15] D. A. Elliott and M. A. Rizack, The Journal of Biological
Chemistry, Vol. 249, 1973, pp. 3985-3990.
[16] H. Ebel and T. Günther, “Role of Magnesium in Cardiac
Disease,” Journal of Clinical Chemistry & Clinical Bio-
chemistry, Vol. 21, 1983, pp. 249-265
[17] M. Haigney, B. Silver, E. Tanglao, H. Silverman, D. Hill,
E. Shapiro, G. Gerstenblith and S. Schulman, “Noniva-
sive Measurement of Tissue Magnesium and Correlation
with Cardiac Levels,” Circulation, Vol. 92, 1995, pp.
Copyright © 2011 SciRes. AJAC
Copyright © 2011 SciRes. AJAC
[18] H. Rasmussen, P. Aurup, S. Hojberg, K. Jensen and P.
McNair, “Magnesium and Acute Myocardial Infarction,”
Archives of Internal Medicine, Vol. 146, No. 5, 1986, pp.
872-874. doi:10.1001/archinte.1986.00360170068010
[19] M. R. England, G. Gordon, M. Salem and B. Chernow,
“Magnesium Administration and Dysrhythmias after
Cardiac Surgery. A Placebo-Controlled, Double-Blind,
Randomized Trial,” JAMA, Vol. 268, No. 17, 1992, pp.
2395-2402. doi:10.1001/jama.1992.03490170067027
[20] R. S. Parsons, T. C. Butler and E. P. Sellars, “Hardness of
Local Water Upplies and Mortality from Cardiovascular
Disease,” Lancet, Vol. 278, No. 7195, 1961, p. 213.
[21] P. M. Savenkov, A. K. Martynov, G. I. Kertsman, O. B.
Bobrooskaya and A. D. Pogorelsky, “The Use of Magne-
sium Composition in Patients with Atherosclerosis of
Vessels of the Heart, Brain, and Lower Extremities,”
Kardiologia, Vol. 11, 1971, p. 85.
[22] R. S. Parsons, T. Butler and E. P. Sellars, “The Treatment
of Coronary Artery Disease,” Med. Proc., Vol. 5, 1959, p.
[23] K. H. Hackethal and Z. Den, “Aussprache Vortragen I
and II,” Chirurgie, Vol. 35, 1951, p. 270.
[24] H. G. Heinrich, “Prophylaxe und Therapie Thrombotischer
Zustande Mit Magnesium,” Zschs Gesell Inn Medicine,
Vol. 12, 1957, p. 777.
[25] B. Schitzler, “Thromboseprophylaxe Mit Magnesium,”
Munchen Medical Wochschr, Vol. 99, 1957, p. 81.
[26] J. Durlach, “Pilule et Thrombose (des Plaquettes,des Es-
trogens et du Magnesium),” Reviews Franclin Endocni-
nology Clinical, Vol. 11, No. 45, 1970.
[27] N. G. Ardlie, E. E. Nisizawa and M. Guccione, “Effect of
Ca and Mg on Platelet Function,” Federation Procedures,
Vol. 29, 1970, p. 423.
[28] R. S. Tonks, “Haematology-Magnesium, Adenosine Di-
phosphate and Blood Platelets,” Nature, Vol. 210, 1966,
p. 106.
[29] M. S. Seelig, Magnesium Bull, Vol. 3, 1981, pp. 26-47.
[30] R. Fehlinger, L. Franke, E. Glatzel, E. Meyer, M.
Michalik, S. M.Rapoport, M. Rüstow, C. Schulz and G.
Schumann, Magnesium Bull, Vol. 3, 1981, pp. 298-306.
[31] F. Liao, A. R. Folsom, F. L. Brancati, “Is Low Magne-
sium Concentration a Risk Factor for Coronary Heart
Disease? The Atherosclerosis Risk in Communities
(ARIC) Study,” American Heart Journal, Vol. 136, No. 3,
1998, pp. 480-490. doi:10.1016/S0002-8703(98)70224-8
[32] P. S. Gartside and C. J. Glueck, “The Important Role of
Modifiable Diatary and Behaivioral Characteristic in the
Causation and Prevention of Coronary Heart Disease
Hospitalization and Mortality: The Prospective NHANES
Followup Study,” Journal of the American College of
Nutrition, Vol. 14, 1995, pp. 71-79.