Reference values for magnesium and potassium contents of mononuclear cells and erythrocytes were estimated in cord blood and in children from infancy ...
1323-1 327 (1990)
Reference Values of Magnesium and Potassium in Mononuclear Cells and Erythrocytes of Children W. B. Geven,’ G. M. Vogele-Mentink,’ J. L. WIIIems,2 Th. de Boo,3 W. Lemmens,3 and L. A. H. Monnens1 Reference values for magnesium and potassium contents of mononuclear cells and erythrocytes were estimated in cord blood and in children from infancy through adolescence. No differences were detected between results for boys and girls. The mononuclear magnesium content was independent of age and was within the adult range of values. No significant correlation was shown between magnesium in serum and in mononuclear cells. Mononuclear potassium also showed no age-related differences. The correlation between magnesium and potassium contents in mononuclear cells was significant; however, the correlation was lower when the magnesium and potassium contents were expressed in terms of protein content: micromoles or millimoles per gram of protein, respectively. The concentration of magnesium in erythrocytes was significantly lower in cord blood and dunng the first month of life, compared with that at older ages, and showed no significant correlation with serum magnesium. The concentration of erythrocyte potassium was independent of age and showed a low but significant correlation with erythrocyte magnesium content. Magnesium and potassium are the main intracellular cations, with only small percentages of their total amounts in the body being located in extracellular fluid. After Baron and Ahmed (1) advocated the use of leukocytes for measurements of intracellular ions, an increasing number of papers have proposed the use of magnesium content of mononuclear cells as an index of intracellular magnesium (2). Paunier et al. (3), investigating the fetalmaternal relationship of intra- and extracellular magnesium and potassium concentrations, found a significantly lower magnesium concentration in erythrocytes of cord blood in comparison with the maternal erythrocytes. The potassium concentration in cord blood erythrocytes, however, was higher than in maternal erythrocytes. The good correlation of magnesium and potassium in lymphocytes nd erythrocytes in both cord and maternal blood showed emarkable similarity. But Abraham et al. (4) reported the bsence of a significant correlation of magnesium and otassium in lymphocytes and erythrocytes. The aim of this study
was to establish
riagnesium and potassium in erythrocytes and mononulear cells in comparison with the concentrations in serum rom cord blood in childhood through adolescence. Investiation into the possible age-dependency of intracellular sagnesium and potassium values shows considerable gaps. Vithout reference values, individual data obtained in chilren with suspected magnesium deficiency or magnesiumosing nephropathy cannot be put into context. This study
of 1 Pediatrics,2
Clinical Chemistry, and3 Medical
Iijmegen, The Netherlands. Received February 2, 1990; accepted May 9, 1990.
also supplies potassium
new data on the relation of magnesium in mononuclear cells and erythrocytes.
Materials and Methods Samples: Approximately 5-10 mL of heparinized blood was taken during venipuncture from subjects whose conditions had no known relation to hypomagnesemia, lowered concentrations of potassium in serum, or magnesium or potassium deficiency. Most of the patients in the younger age groups (Groups I and II) were hospitalized; the older ones (Groups Ill and IV) were mostly outpatients. Previously measured concentrations of serum potassium and magnesium in these subjects in Groups III and IV were within the normal ranges. All patients had normal oral feeding. Ten milliliters of cord blood was taken from neonates born vaginally after uneventful pregnancies and births.
We grouped the samples as follows: Group I, eight samples of cord blood; Group II, eight samples from the first month postpartum; Group III, 15 samples from ages one to six months; and Group IV, 57 samples from ages six to 200 months (6-12 months, eight samples; 12-36 months, 12 samples; 36-72 months, 13 samples; 72-120 months, 10 samples; 120-200 months, 14 samples). Every child in this study was analyzed once. All groups contained as many boys as girls, or differed by no more than one in groups with an unequal number of samples. Thus we could evaluate the
of age and sex on the reference values. methods: Magnesium was measured by atomic absorption spectrophotometry (Model 5000; PerkinElmer, Norwalk, CT). Other measurements were done by routine laboratory methods. In 17 samples, taken at random, the total number of leukocytes in the venous blood Laboratory
was counted (Coulter Counter; Luton, England) and identified
Coulter Electronics Ltd., microscopically in a smear.
Mononuclear cells were isolated according to Elm and Johnson (5). The heparinized blood was diluted with equal volumes of buffered saline and glucose solution (BSG; containing 8.1 g of NaC1, 1.53 g of NaHPO4 2H20, 0.194 g of NaH2PO4 . H20, and 2 g of glucose per liter with an osmolality of 290 mosmol/kg and pH 7.4, adjusted with NaOH). The cord blood was diluted with 20 mL of BSG because of the high hemoglobin concentrations and viscosity. Then, the mixture was carefully layered onto FicollPaque (density 1.077 kgfL) and centrifuged at 400 x g for 35 mm. The interphase was collected, washed in BSG, and centrifuged at 400 x g for 10 mm. To the pellet we added 4.5 mL of BSG and counted the leukocytes. A paraformaldehyde fixation in a cytospin preparation was made and the viability of cells was checked by staining with trypan blue (6). Exactly 4 mL of the cell suspension was centrifuged at 600 x g for 10 mm. To the pellet we added 700 L of de-ionized water and sonicated the sample three times, 15 s each at 45-s intervals. The tubes were cooled in ice. The magnesium content of the lysate was determined by atomic CLINICAL CHEMISTRY, Vol. 36, No. 7, 1990
absorption spectrophotometry with lanthanum oxide (5). Protein was measured according to Lowry et al. (7). DNA was estimated with use of 4,6-diamidino-2phenylindole 2HC1 (8). From the counted total number of mononuclear cells in the known volume of venous blood and in the cell suspension, we calculated the percentage of harvested
fmollcell, potassium tometry,
mononuclear cells. Magnesium was expressed as mol/g of protein, and mmollg of DNA. The content of the lysate, measured by flame phowas expressed as fmollcell, mmollg of protein, and
mmol/g of DNA. After centrifugation, erythrocytes were collected and washed four times in CsC1 (155 mmoIIL, osmolality 290 mosmol/kg) (9). We then lysed 100 L of packed cells in 400 pL of de-ionized water. We suspended 100 L from this lysate in an Eppendorf cup, evaporated the liquid, and weighed the residue. The lysate was frozen until time of assay. Magnesium and potassium were expressed as micromoles or millimoles, respectively, per gram of erythrocytes dry weight. Statistical methods: We used Spearman’s rank correlation coefficients to detect possible relationships between different variables. The Kruskal-Wallis test (10) was used to detect possible differences between the age groups for all analytes. Control for normality was performed according to Stephens (11), and we used Wilcoxon’s two-sample test to detect possible differences between boys and girls. Results were considered to be significant when P 0.05) in all age groups above the age of six months, all children older than six months were combined into one group (IV). Table
1 lists the reference
of the different
(6.10) 16.3-52.1 (35.6)
(4.04) 42.5-49.6 (46.3)
(35.4) (51.4) (34.3) (32.4) 0.29-0.51 0.25-0.56 0.21-0.68 0.21-0.53 (0.40) (0.32) (0.36) (0.33) 2.92-12.13 2.26-11.23 2.82-10.31 2.67-8.56
mmol/g prot mmol/g DNA mo-Protein pg/cell
Erc-Mg prool/g erc
(5.28) Erc-K mmol/g erc
0.23-0.27 (0.24) mo, mononuclear cells.
Because of the limited number of samples in Groups I-rn, we list the highest, lowest, and median values. Serum magnesium (Figure 1) in cord blood (Group I) was significantly lower than in samples from age groups older than one month (Groups ifi and IV), with P