点击显示 收起
1 From the Centre de Recherche en Nutrition Humaine d'Auvergne, Unité Maladies Métaboliques et Micronutriments, INRA, Theix, Saint Genès Champanelle, France; the Laboratoire de Contrôle des Eaux, Institut Louise Blanquet, Faculté de Médecine et Pharmacie, Clermont-Ferrand, France; and the US Department of Agriculture/Agricultural Research Service Children's Nutrition Research Center, Baylor College of Medicine, Houston.
2 The contents of this article do not necessarily reflect the views or policies of the US Department of Agriculture, nor does mention of trade names, commercial products, or organizations imply endorsement by the US government. 3 Theraplix Laboratory supplied the magnesium pidolate. SAA of the Children's Nutrition Research Center is supported by the US Department of Agriculture/Agricultural Research Service under Cooperative Agreement 58-6250-6001. 4 Address reprint requests to C Feillet-Coudray, Unité Maladies Métaboliques et Micronutriments, INRA, 63122 Saint Genès Champanelle, France. E-mail: feillet{at}clermont.inra.fr.
ABSTRACT
Background: Studying magnesium pools in the body with use of stable isotopes may be helpful for evaluating magnesium status. Data on the evaluation of magnesium pools in humans are scarce.
Objective: We undertook this study to evaluate the effects of a magnesium supplementation program on the size of the exchangeable body pools of magnesium and on classic indexes of magnesium status in healthy women with normal magnesium status.
Design: Ten healthy women participated in a kinetic study with magnesium stable isotopes before and after 8 wk of magnesium supplementation. Each woman received 3 supplements containing 5.08 mmol (122 mg) elemental Mg/d (366 mg/d). Before and at the end of the supplementation period, each woman received an intravenous injection of 1.67 mmol (40 mg) 25Mg, and the plasma magnesium disappearance curve was followed for the next 7 d. Two methods were used to analyze the exchangeable pools of magnesium: 1) formal multicompartmental modeling and 2) a simplified estimation of the total mass of the rapidly exchangeable magnesium pool (EMgP).
Results: In these healthy women, exchangeable magnesium pools represented 11–12% of total body magnesium on the basis of multicompartmental analysis. The simplified estimation of EMgP overestimated the size of the exchangeable magnesium pools by 45–50%. Eight weeks of magnesium supplementation did not significantly modify the size of the exchangeable magnesium pools, whereas urinary magnesium excretion was significantly higher after 8 wk of supplementation.
Conclusion: Women with no clinical evidence of magnesium deficiency may not respond to short-term supplementation with increases in the mass of the exchangeable magnesium body pool or in magnesium turnover rates.
Key Words: Exchangeable magnesium pools • kinetic modeling • stable isotope • magnesium supplementation • healthy women • magnesium status
INTRODUCTION
In developed countries, marginal magnesium intake may increase the prevalence of magnesium deficiency (1,2). Many data suggest that magnesium deficiency may be implicated in various metabolic disorders, including cardiovascular disease, immune dysfunction, and free radical damage (3,4). It is thus necessary to better evaluate the relations among magnesium intake, indexes of magnesium status, and magnesium deficiency disorders (2). Whereas severe magnesium deficiency is easy to detect, however, mild or moderate deficiency is more difficult to diagnose.
At present, there is no consensus on the optimal method of evaluating magnesium status. Measuring plasma and erythrocyte magnesium concentrations allows one to detect only relatively large deficiencies, and these concentrations may not reflect whole-body magnesium stores. The more recently advocated method of measuring either plasma ionized magnesium with use of magnesium-specific electrodes or magnesium in leukocytes requires further validation (5). The parental loading test, which is based on the principle of magnesium retention according to the degree of magnesium deficiency, presupposes normal renal function and requires further validation as an indicator of magnesium status (6).
Exploring exchangeable magnesium pools with the use of stable isotopes may be a useful method for evaluating magnesium status. Such an approach was developed in humans to determine selenium and zinc status (7–9). Recently, we showed that the magnesium pool size and turnover rate vary in proportion to dietary magnesium intake in rats (10,11). Comparable data on the effects of magnesium intakes or magnesium deficiency on exchangeable magnesium pool masses in humans are not available. The objectives of the present study were to evaluate exchangeable magnesium pools by using stable isotopes in healthy women without evidence of magnesium deficiency and to evaluate the effect of 8 wk of nutritional magnesium supplementation on the size of these pools.
SUBJECTS AND METHODS
Subjects
Ten healthy women aged 20–34 y participated in the study (Table 1). Clinical examinations were performed by the medical staff of the Laboratory of Nutrition of Clermont-Ferrand. The subjects' kidney function was normal and no biochemical evidence of inflammatory conditions was found (as assessed by measurement of plasma creatinine, urea, and C-reactive protein). Fat-free mass was determined by bioelectrical impedance analysis at 5 and 100 kHz (Analycor; Eugedia, Cachan, France). Usual dietary magnesium intake was assessed before the study with use of 5-d estimated food-intake records and the GENI program (Micro 6, Villiers les Nancy, France). None of the women were taking oral contraceptives, other medications, or other vitamin or mineral supplements at the time of recruitment or for the duration of the study. Only one woman (subject no. 8) had previously given birth. The study was approved by the local Human Ethical Committee on 18 January 2000 under the number AU 282 (CCPPRB Auvergne). All participants were fully informed of the purposes of the study and gave their written, informed consent.
View this table:
TABLE 1 . Characteristics of the subjects1
Experimental design
Stable-isotope kinetic studies with 25Mg were performed before and after magnesium supplementation. One 10-mL ampule was given at breakfast, lunch, and dinner each day for 8 wk. Each ampule contained 5.08 mmol (122 mg) elemental Mg. Supplements consisted of magnesium pidolate (Mag 2; Theraplix, Rhone Poulenc, Melun, France). The magnesium content of each ampule was checked and the mean (±SD) concentration was 5.24 ± 0.23 mmol/ampule. The total daily supplementation was similar to the safe and adequate daily dietary magnesium intake recommended for French women (12) and the current recommended daily magnesium intake in the United States (13). The subjects maintained their usual diets during the period of magnesium supplementation. To ensure complete use of the supplements, any unused ampules of magnesium pidolate were collected periodically.
Preparation of stable-isotope solution
Magnesium has 3 stable isotopes with the following natural abundances: 24Mg, 0.789; 25Mg, 0.100; and 26Mg, 0.111 (14). The least naturally abundant stable isotope, 25Mg, was used in this study. This isotope was purchased in the oxide form, with an enrichment of 96.7% (Chemagas, Paris), and was dissolved to 0.032 mol/L in a 0.9%-NaCl solution. The pH of this solution was adjusted to 6.5. Bottles containing 50 mL were prepared by the hospital pharmacist. The sterility, pyrogenicity, and pharmaceutical quality of the solution were tested by the pharmacy of Clermont-Ferrand Hospital.
Tracer administration
Subjects were admitted to the Human Nutrition Laboratory of Clermont-Ferrand after they had fasted overnight. An intravenous catheter was inserted into each subject's right arm and 1.67 mmol (40 mg) 25Mg in 50 mL saline solution was perfused over 30 min (the end of the perfusion was considered time 0). An intravenous catheter was also inserted into each subject's left arm for blood sampling at -30, 15, 30, 60, 90, 120, 180, 240, 360, 480, and 600 min. On the following day (day 1) and on days 2–7, blood samples were collected after the subjects had fasted overnight. Total 24-h urinary magnesium excretion was measured for 2 d, the day before and the day of the isotopic loading.
Analysis
Blood samples (2 mL) were collected in heparin-containing tubes and the plasma was separated by centrifugation (2500 x g, 15 min, 20°C). Erythrocytes collected 30 min before the start of the isotope study were washed with saline solution and hemolyzed. Urine samples were acidified with pure HNO3 to a final concentration of 0.14 mol/L. Samples were stored at –20°C until analyzed.
The 25Mg content of plasma samples was determined by inductively coupled plasma mass spectrometry (ICP-MS; PlasmaQuad II systems; Fisons Instruments, Manchester, United Kingdom) (15). Before analysis, the plasma was diluted in 0.14 mol HNO3/L, and natural magnesium and beryllium were used as external and internal standards, respectively. Total magnesium concentrations in the analyzed samples were 50 µg/L. All isotope analyses were done at least twice. Within-day and between-day variations for 25Mg/26Mg measurement were 0.42% and 0.87%, respectively. The limit of detection for 25Mg/26Mg enrichment was 1%.
Plasma, urine, and hemolyzed erythrocytes were diluted in 0.1% LaCl3 for the measurement of total magnesium. Magnesium concentrations were then measured by flame atomic absorption spectrophotometry (model 560; Perkin-Elmer, St Quentin en Yvelines, France) at 285 nm. Plasma ionized magnesium was measured with an AVL 988/4 analyzer (AVL Medical Instruments, Eragny, France).
Kinetic analysis
Magnesium kinetics were determined by using a multicompartmental model as described by Avioli and Berman (16) and Sojka et al (17). A schematic of the model is shown in Figure 1. Compartmental modeling of the data was performed with the aid of the SAAM II (stimulation, analysis, and modeling) program (SAAM Institute, Inc, Seattle) (18). Plasma data were expressed as tracer/tracee, with tracer = 25Mg from the injection and tracee = (total Mg – 25Mg from the injection). The concentration of 25Mg from the intravenous injection and the concentration of total magnesium in plasma in mmol/L were measured by ICP-MS with use of an external calibration curve of natural magnesium. Given the type of quantitative calibration, ICP-MS gives the same concentration for the 3 stable isotopes of magnesium for natural unenriched magnesium solutions, whereas in a 25Mg-enriched solution, the obtained concentration of 25Mg will be greater than that of the other 2 isotopes. To calculate the concentration of 25Mg that came from the intravenous injection, the following formula was applied:
RESULTS
Magnesium intake and classic indexes of magnesium status
The subjects were 24 ± 4 y old, had a mean body mass index (in kg/m2) of 21.07 ± 2.11, and had a mean fat-free mass of 43.33 ± 4.13 kg (Table 1). The subjects' dietary magnesium intake was below the recently recommended dietary allowance for magnesium (0.25 mmol/kg body wt for adults; 12), but was similar to the mean daily magnesium intake of women observed in France (11.7 mmol/d) (2).
As shown in Table 2, classic indexes of magnesium status were in the lower end of the normal range at the beginning of the study (normal range: 0.69–0.92 mmol/L for plasma magnesium, 1.64–2.58 mmol/L for erythrocyte magnesium, and 1.33–8.19 mmol/d for urinary magnesium) (6,21). Plasma total magnesium was positively correlated with plasma ionized magnesium (r = 0.7837, P = 0.007). Plasma and erythrocyte magnesium concentrations were not significantly affected by 8 wk of magnesium supplementation, whereas plasma ionized magnesium and urinary magnesium excretion were significantly higher after 8 wk of supplementation.
View this table:
TABLE 2 . Magnesium status before and after 8 wk of magnesium supplementation in 10 women1
Compartmental model of magnesium metabolism in women
All subjects tolerated the intravenous magnesium infusion and kinetic study well. The plasma data before and after magnesium supplementation are shown in Figure 3. There was a rapid disappearance of tracer during the first 10 h, followed by a slow decline that extended though 170 h. This kinetic pattern was similar before and after magnesium supplementation.
FIGURE 3. . Mean (±SD) plasma data for healthy women before () and after () 8 wk of magnesium supplementation (n = 10). Compartmental modeling of the data was performed with use of the SAAM II program (SAAM Institute, Inc, Seattle) (18). Only the first phase is presented in the graph because it is the important phase in relation to magnesium kinetics.
The kinetic indexes, the masses of the magnesium pools, and the half-lives of the pools are given in Table 3. Pools 1 and 2 were primarily composed of extracellular fluid and pool 3 represented intracellular magnesium (16). In these healthy women, pool 3 accounted for 89% of the total exchangeable magnesium. Data for the one subject who had previously given birth (subject no. 8) were not significantly different from the data for the other women. This subject was most likely not physiologically different from the others because her pregnancy took place several years before the study. The range in pool sizes, rate constants, and half-lives among the subjects was large.
View this table:
TABLE 3 . Calculated 3-compartment model parameters of women before and after 8 wk of magnesium supplementation1
No significant differences were observed in M1, M2, and M3 before and after magnesium supplementation. No correlation was obtained for M1, M2, and M3 with plasma, erythrocyte, or urinary magnesium or with fat-free mass. Fractional transport rates of magnesium between pools, and irreversible losses from pools 1 and 3, were also not significantly affected by magnesium supplementation. In addition, the half-lives of the pools were not significantly changed by supplementation.
Rapidly exchangeable magnesium pool
At the beginning of the study, EMgP was 49% higher than total exchangeable magnesium as determined by the method of Avioli and Berman (16). EMgP was positively correlated with fat-free mass (r = 0.9153, P < 0.001). EMgP was not significantly modified by magnesium supplementation in these healthy women. When all individual values for EMgP were considered (ie, values determined both before and after supplementation), EMgP was positively correlated with M3 (r = 0.5334, P = 0.0334) and almost with M1 (r = 0.4915, P = 0.0532).
DISCUSSION
The first studies of the size of the exchangeable magnesium pool in humans were reported in the 1960s. In these studies, the radioisotope 28Mg was used as a tracer. Zumoff et al (22) and Silver et al (23) described in human volunteers a 3-pool model with relatively short half-lives. These half-lives differed from 15 min, 40 min, and 340 min (22) to 1 h, 3 h, and 14–35 h (23). Silver et al (23) hypothesized the existence of an additional magnesium pool with a very slow turnover (2% per day). Aikawa et al (24) and McIntyre et al (25) observed that, in humans, the exchangeable magnesium pool mass was <16% of the estimated body content of magnesium. In 1966, Avioli and Berman (16) proposed a multicompartmental model of 3 exchangeable magnesium pools. In that model, pools 1 and 2 have a relatively fast turnover and approximate extracellular fluid in distribution, whereas pool 3 is an intracellular pool containing >70% of the exchanging magnesium, with a turnover rate one-half that of the most rapid pool. There is also a fourth pool, which represents urinary excretion and endogenous fecal loss, and a fifth pool, which is a loss pathway representing deposition into tissues (17). The magnesium in pool 5 appears to be irreversibly lost because it is not possible to measure the release of magnesium from pool 5 into the circulation by use of stable isotopes. Using this model, Avioli and Berman (16) reported that exchangeable magnesium was 15% of the estimated body content in human volunteers, with pool sizes of 8.21 mmol (197 mg), 9.17 mmol (220 mg), and 116 mmol (2788 mg) for M1, M2, and M3, respectively. This result was later confirmed by Watson et al (26).
Because of the very short half-life of 28Mg (21.3 h), the turnover rates of the slowly exchangeable pool of magnesium could not be accurately determined with use of this tracer. Therefore, 28Mg kinetic studies were limited because of their apparent clinical inability to identify magnesium-deficient conditions (27). With improvements in analytic techniques, methods using the stable isotopes 25Mg and 26Mg were developed. Sojka et al (17) and Abrams and Ellis (19) validated the multicompartmental model described by Avioli and Berman (16) with use of stable isotopes. In children and adolescents, they observed that 26% of total body magnesium exchanged after 14 d and that pool masses were 11.5 mmol (275 mg), 48.6 mmol (1166 mg), and 142 mmol (3400 mg) for M1, M2, and M3, respectively.
In the present study, we showed that exchangeable magnesium pools represented 11–12% of total body magnesium. This value is lower than the results reported for adolescents (17,19), probably because of the rapid growth of bone and muscle tissue during early adolescence. Our results are similar, however, to those determined with use of 28Mg in adults (24,25). The half-lives of the exchangeable pools in our study were higher than those reported by Zumoff et al (22) and Silver et al (23), however, probably because of our use of the compartmental model of Avioli and Berman (16) as well as the more accurate measurements enabled by the use of stable isotopes. As previously observed, the short half-life of 28Mg does not permit accurate measurements over several days. However, we cannot exclude an imprecision of early tracer distribution, metabolism, and pool sizes in our study because the load of isotope administered was high (20% of the size of the magnesium pool). Unfortunately, high doses of the magnesium stable isotope must be administered to obtain enrichment distinguishable from natural abundance (10% for 25Mg and 11% for 26Mg).
We also examined EMgP, the combined pools of magnesium that exchange with plasma magnesium within 48 h of tracer administration. The method used, in which plasma enrichment data from days 3–7 are used, is simpler than the frequent early sampling required for compartmental analysis. However, this method is accurate only to the extent that 2 conditions are met: 1) the loss of tracer from EMgP occurs at a monoexponential rate from the moment of tracer administration to the end of the measurement period and 2) the tracer is homogeneously mixed throughout the EMgP during the measurement period (20). This method overestimates the size of the EMgP because the initial rapid loss of tracer from plasma is not accounted for by extrapolation of the monoexponential rate of loss during the measurement period, and it is extremely unlikely that the tracer exists at equal concentrations in all compartments of the EMgP at any time (20). In our study, the size of the exchangeable magnesium pool was overestimated by EMgP measurement by 45–50%. Because the regression analysis was based on only 4 measurement points postdosing, however, and because confidence limits about the regression line improve with the number of sampling points, the low number of sampling points may also have contributed to the overestimation of EMgP.
Abrams and Ellis (19) suggested that the body's fat-free mass correlates positively with exchangeable magnesium pool masses in children, illustrating the importance of accounting for growth in determining a subject's magnesium utilization. In our study, the total exchangeable pool mass determined by the model of Avioli and Berman (16) did not correlate with fat-free mass. Because the subjects in our study were adults who had achieved their ultimate height, however, the fact that we did not find such a correlation was not surprising.
We also examined the effect of 8 wk of magnesium supplementation on classic indexes of magnesium status. Magnesium supplementation resulted in significantly higher plasma ionized magnesium, but had no significant effect on the total plasma magnesium concentration. This suggests that plasma ionized magnesium is more sensitive to magnesium intake than is total plasma magnesium. Further investigations are necessary to achieve a standardized measurement of ionized magnesium and to validate the use of such a measure as an appropriate index of magnesium status. Erythrocyte magnesium concentrations were also not significantly modified by magnesium supplementation. In contrast, urinary magnesium excretion was significantly higher after magnesium supplementation. This is consistent with the normal renal regulation of magnesium body stores by the excretion of an excess in replete subjects and by the reabsorption of secreted magnesium in deficient subjects.
Magnesium supplementation did not modify the size of the exchangeable magnesium pools, as determined both by compartmental analysis and by the alternative technique developed by Miller et al (20). In agreement with our results, other studies reported no significant changes in the body's magnesium stores after magnesium supplementation in young healthy adults (28) and in athletes with low-normal serum magnesium concentrations (29). This finding differs from that which we previously reported for rats, in which the exchangeable magnesium pool size increased with magnesium intake (10,11). We observed in rats that a >2-fold increase in magnesium intake (from 200 mg Mg/kg diet, a marginally magnesium-deficient diet, to 500 mg Mg/kg diet) led to a 22% increase in the magnesium pool size. Thus, in humans, by contrast with rats, the size of the exchangeable magnesium pool may not be a sensitive index for identifying variations in magnesium status. It is also possible that magnesium stores were adequate in the healthy women in the present study before magnesium supplementation, even though their magnesium intake was below the recommended dietary allowance (12). This would explain why magnesium supplementation did not affect magnesium pool size but increased urinary magnesium excretion. Another possibility is that the 8-wk supplementation period was not long enough to identify an effect of magnesium supplementation on the size of the magnesium pool. More studies are necessary to better appreciate the relation between magnesium status and exchangeable magnesium pool size.
In conclusion, we determined the size of the exchangeable magnesium pool in healthy women with use of stable isotopes, both by multicompartmental analysis and by an alternative technique developed by Miller et al (20). The alternative technique was simpler, but overestimated the size of the exchangeable magnesium pool by 45–50%, making it unsuitable for this purpose. After 8 wk of magnesium supplementation, urinary magnesium excretion was significantly higher, whereas the magnesium pool size was not significantly affected. Full magnesium stores before magnesium supplementation may explain such results. Further studies are necessary to determine whether the magnesium pool size responds to magnesium supplementation in magnesium-deficient subjects.
ACKNOWLEDGMENTS
We are grateful to C Bouteloup for her help in this study, L Morin and M Brandolini of the Human Laboratory Nutrition of Clermont-Ferrand for their assistance in caring for the study subjects, the pharmacy of Clermont-Ferrand Hospital for preparing the stable isotopes for human use, and Children's Nutrition Research Center editor LA Loddeke for editorial advice.
REFERENCES