| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
1 Dept. of Biol., Mathias Belius Univ., Tajovského 40, 974 01 Banská Bystrica, Slovak Republic
Dept. of Clin. Biochem., F.D. Roosevelt Hosp., Nám. L. Svobodu 1, 975 17 Banská Bystrica, Slovak Republic
a Author for correspondence.
To the Editor:
Determination of carnitine
(L-ß-hydroxy-
-trimethylaminobutyric acid) in
biological samples plays an important role in the diagnosis of diseases
with carnitine deficiency (1). In the past few years, and
also in this Journal, several spectrophotometric methods for
assaying carnitine in serum have been described (2)(3)(4)(5)(6). To
our knowledge, however, no data have been published on the biological
variation of this serum analyte.
For any new test, data on the biological variation generated from the healthy population may be used (a) in setting desirable performance standards or analytical goals, (b) in assessing the utility of conventional population-based reference intervals, and (c) in critically evaluating the significance of changes in serial results from an individual (7).
To investigate the analytical and biological variation of free and total carnitine, as well as of the acyl/free carnitine ratio, we collected blood samples once a week for 4 weeks from each of 14 healthy subjects (7 men and 7 nonpregnant women, ages 21–23 years), students of Mathias Belius University. The subjects agreed to maintain current dietary habits, body weight, and exercise program (if any) for the duration of the study. The criteria of the hospital Ethics Committee were respected in this experiment.
To minimize sources of preanalytical variation, venous blood specimens were drawn between 0800 and 0900 h after ~12 h of fasting. The subjects remained seated for at least 20–30 min before the blood was drawn. Usually, specimens were obtained by a single phlebotomist and with minimal stasis into Monovette Serum Gel blood-collection tubes (Sarstedt). Serum specimens for both the free and total carnitine assays were prepared and stored the same way. After clotting, each specimen was centrifuged at 1500g for 15 min and the serum obtained was eluted through CentrifreeR columns (Amicon) in an angle-head rotor at 2000g for 30 min. Aliquots (800 µL) of protein-free serum filtrates were stored frozen at -20 °C until assayed.
Free and total carnitine concentrations were determined by a spectrophotometric enzymatic assay adapted for use on the Cobas Mira analyzer (Hoffmann-La Roche). The assay uses carnitine acetyl-transferase (CAT; EC 2.3.1.7) and 5,5-dithiobis(2-nitrobenzoic acid) (DTNB) as a thiolgroup color reagent. The Cobas analyzer was programed as two-reagent chemistry with a primary reagent and an enzymatic start reagent. Diluted sample (80 µL) was mixed with 200 µL of the primary reagent (10 mL of 0.2 mmol/L Na2HPO4, pH 7.8, 0.2 mL of 10 mmol/L DTNB, and 0.4 mL of 15 mmol/L acetyl-CoA), and the reaction was started with 24 µL of sixfold-diluted start reagent [CAT from pigeon muscle (1 mL, protein 5.3 g/L, CAT 120 kU/g protein) was diluted with 0.9 mL of phosphate buffer (0.5 mol/L, pH 7.5) to give a final concentration in start reagent of 63.6 kU/L]. All specific reagents used were from Sigma Chemical Co. The reaction mixture was incubated for 5 min at 37 °C and absorbance was read at 405 nm. Calibrators containing 10–100 µmol/L of L-carnitine were prepared by programed dilutions of the stock solution. The Cobas assay was linear for carnitine concentrations up to 250 µmol/L and the limit of detection was ~2 µmol/L.
Free carnitine values were measured directly. Total carnitine was quantified after nonautomated deesterification by alkaline hydrolysis (6). In our hands, the analytical recovery of L-octanoylcarnitine added to serum filtrate, determined as free carnitine after hydrolysis, was 94–105% (mean 98%, n = 6). All specimens were assayed in replicate in the same analytical run.
Dixon's test was used to exclude outlying values from a single
subject. The analytical (CVA), within-subject
(CVI), and between-subject (CVG) components of
variation were calculated by nested analysis of variance
(8). Useful indices, i.e., analytical goals for
imprecision (CVA < 1/2CVI), indices of
individuality (CVI/CVG), and critical
differences required for significant (P 
0.05) changes in
serial results [2.77 (CVA2 +
CVI2)1/2], were also obtained
(7). The significance of the differences between means and
between variances was evaluated by using the unpaired t-test
and the Fisher F-test, respectively.
The results for free carnitine, total carnitine, and the acyl/free
carnitine ratio are summarized in Table 1
. As expected, men showed higher means for these analytes than
women. However, this difference was significant only for the total
carnitine concentration (P <0.02). All values obtained for
each subject fell within our laboratory's reference intervals for free
and total carnitine.
|
In several clinical conditions, an acyl/free carnitine ratio should be calculated from the two carnitine assays (acylcarnitine = total carnitine - free carnitine) (9)(10). Our values for this ratio are somewhat higher than those found for healthy subjects in several other laboratories (11)(12)(13), probably because of different methodologies used, different physiological characteristics of the subjects, or both.
Replicate analyses for each sample were used to investigate within-run analytical variation. The CVA values of the analytical methods for free and total carnitine are lower than the respective goals, but those of the acyl/free carnitine ratio are greater. The high CVA component of carnitine assays could limit the clinical utility of the acyl/free carnitine ratio (7).
The mean CVI value for each carnitine quantity is <12.3%; for men and women, the acyl/free carnitine ratio is the most variable quantity. The relative distribution of free carnitine and carnitine esters varies according to fasting status, adiposity, renal function, and muscular exercise (11). Nevertheless, these data indicate the existence of a reliable homeostatic mechanism in steady-state conditions. The CVG values indicate moderate variation in carnitine concentrations between healthy subjects. No significant differences between men and women were observed in within- or between-subject variances for any of the quantities studied.
Indices of individuality were <1.4 for each of the carnitine groups,
meaning that individual results are more useful than population-based
data (7). For diagnosis and screening, however, free
carnitine values <20 µmol/L and total carnitine 30 µmol/L,
determined by nonradioenzymatic methods, are considered to indicate
carnitine deficiency (4). In carnitine-deficient patients,
a low free carnitine concentration in serum is often associated with an
increased acyl/free carnitine ratio
(4)(9)(10)(11).
Finally, the critical differences obtained in this study, calculated
from the mean CVI values, are also shown (Table 1
).
For free and total carnitine the critical difference was less than that
for the acyl/free carnitine ratio. The former assays may therefore be
more suitable for monitoring purposes, especially in carnitine
supplementation therapy. However, the critical differences presented
here are only a guide to clinical practice; other laboratories should
take into consideration their own between-day imprecision of carnitine
assays.
Acknowledgments
We thank Mária Dibalová for help in sample analyses and Samuel Koróny for statistical assistance.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  Z. Dagher, N. Ruderman, K. Tornheim, and Y. Ido Acute Regulation of Fatty Acid Oxidation and AMP-Activated Protein Kinase in Human Umbilical Vein Endothelial Cells Circ. Res., June 22, 2001; 88(12): 1276 - 1282. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
