| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters |
Faculteit Farmaceutische Wetenschappen, Laboratorium voor Analytische Chemie, Harelbekestraat 72, B-9000 Gent, Belgium
a Author for correspondence. Fax 32-9-264 81 98; e-mail Linda.Thienpont{at}rug.ac.be.
To the Editor:
In the past, serum complexed magnesium (S-cMgcom) did not receive much attention, and the few investigations that were performed gave discrepant results. Walser (1) found for S-cMgcom a value of 13% of serum total magnesium (S-cMg), whereas Speich et al. (2) reported a value of 5.5% (Note: The % notation refers to S-cMg throughout the text; these and the other values relate to serum samples from apparently healthy donors). The development of ion-selective electrodes (ISEs) for measurement of serum ionized magnesium (S-cMg2+) (3) led to increased interest in research on S-cMgcom (4)(5)(6)(7)(8). This is because in theory S-cMgcom can be calculated easily from the difference: ultrafilterable total magnesium (UF-cMg) - S-cMg2+. However, recent investigations in this area failed to resolve the original discrepancy. Altura and Altura (4) found a value of 14% for S-cMgcom with the Nova 8 analyzer. However, because their ultrafiltration was done aerobically and without pH control, they obtained these values only after sparkling native samples (at pH 7.4) with CO2. If they had not sparkled the samples with CO2, they would have calculated a negative value for S-cMgcom, namely -1% (5). Filos and Okorodudu (6), using the AVL 988-4 analyzer, and Külpmann and Gerlach (7), using the Kone Microlyte 6 analyzer, found a value of 5%, whereas Huijgen et al. (8), using the Kone Microlyte 6 analyzer, reported a value of 8%. These differences may stem from two limitations in the calculation of S-cMgcom from UF-cMg and S-cMg2+. First, the value for UF-cMg depends on the ultrafiltration procedure applied (4)(5). Second, the measurement of S-cMg2+ by ISE has not been standardized, hence the true value of S-cMg2+ is unknown. Indeed, different values for S-cMg2+ are obtained with different ISE systems (9). Moreover, in the absence of adequate quality-control materials, different instruments from the same manufacturer might give different values for S-cMg2+ (10). In addition, because of the combined effect of differences in ultrafiltration procedures and variable calibration of individual instruments, values reported for S-cMgcom cannot be related to the calibration of certain ISE systems. Therefore, we conclude that calculating S-cMgcom from UF-cMg and S-cMg2+ is not valid at this time.
Instead of calculating S-cMgcom, we measured S-cMgcom in the serum system by adding known complexing anions, such as citrate, phosphate, bicarbonate, and lactate, to serum samples. In this way, we avoided ultrafiltration and made S-cMgcom determination independent of the calibration of the ISE system used. We added 20 µL of a concentrated solution of sodium salts of bicarbonate, citrate, lactate, and monohydrogen phosphate as complexing agents to one aliquot of a serum sample to obtain twice the original concentration of the respective complexants. To the reference aliquot, we added 20 µL of an equivalent NaCl solution. We then measured S-cMg2+ in both aliquots with the AVL 988-4 instrument (AVL List GmbH). Using this procedure, we found a value of 10.4% for S-cMgcom (± 1%, 95% confidence interval), which is between the 5.5% found by Speich et al. (2) and the 13% found by Walser (1). The calculation we used was as follows: measured mean of S-cMg2+ in samples with NaCl solution, 0.609 mmol/L; measured mean of S-cMg2+ in samples with complexant solution, 0.522 mmol/L.
Because of sample dilution (20 µL/500 µL serum), both values must be multiplied by a factor of 1.04. The difference between the values gives the concentration of S-cMgcom: 1.04(0.609 - 0.522) = 0.0905 mmol/L.Because S-cMgcom is measured in serum water, it must be multiplied by a factor of 0.94 before it can be expressed as % S-cMg, the mean of which was 0.819 mmol/L. From this follows: % S-cMgcom = [(0.0905 x 0.94)/0.819] x 100 = 10.4%.
Note that S-cMgcom, measured by ISE, needs to be corrected for the water displacement effect by the serum proteins when it is related to S-cMg. Furthermore, our approach assumes, for example, that the added phosphate (1.09 mmol/L) complexes the same amount of magnesium as the phosphate already present (mean, 1.09 mmol/L). This is justified because the functions calculated from the theoretical complexation constants are quasi-linear in the range we investigated (8). In addition, complexation decreased proportionally when the concentrations of added complexants were only one-half of the concentrations of the original complexants.
On the other hand, when reliable values for S-cMgcom and UF-cMg become available, we should be able to calculate S-cMg2+ by subtracting S-cMgcom and serum protein-bound magnesium (S-cMgpb) from S-cMg (Note: S-cMgpb = S-cMg - UF-cMg). If S-cMg2+ correlates reasonably well with S-cMg, a practical standardization approach for S-cMg2+, based on S-cMg, should be possible. To investigate this, we measured the S-cMg and UF-cMg in 12 serum samples with an ion chromatography reference method (11); for S-cMgpb, we found a value of 31.5% (± 1.6%, 95% confidence interval), which is in good agreement with the 33.7% reported by Speich et al. (2) (Note: Like S-cMgcom, UF-cMg must also be corrected). According to the above proposal, our data would yield a calculated value of 58.1% for S-cMg2+. We measured S-cMg2+ with the AVL 988-4 and found an excellent correlation between S-cMg and pH-normalized S-cMg2+ in a panel of 57 serum samples (r = 0.9223; P <0.001; S-cMg range, 0.74–0.92 mmol/L).
From these preliminary results, we concluded that standardization of S-cMg2+ on the basis of S-cMg is a realistic option. We will undertake additional experiments to substantiate our observation that, for serum samples from apparently healthy donors, S-cMg2+ accounts for ~58% of S-cMg (the fraction is usually assumed to be 65%).
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
