| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 University of Leuven, Department of Public Health, Division of Nutritional Epidemiology, Kapucijnenvoer 33, B-3000, Leuven, Belgium
aauthor for correspondence: fax 32-16-336884, e-mail Pascale.Vanhoydonck{at}med.kuleuven.ac.be
Oxidative processes and endothelial cell dysfunction play an important role in the etiology of atherosclerosis. Oxidative modification of LDL in the subendothelial space of the vessel wall is thought to be important in the initiation of atherosclerosis. Oxidized LDL (OxLDL) may not only contribute to foam cell generation, but also stimulate the synthesis of adhesion molecules by endothelial cells. Expression of intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1) promotes the adherence and migration of new monocytes through the endothelial barrier to the subendothelial space (1)(2). Finally, some OxLDL will diffuse back from the atherosclerotic arterial wall to the blood, in which it can be measured. Coronary atherosclerosis is associated with increased amounts of circulating OxLDL (3). Circulating ICAM-1, VCAM-1, and von Willebrand factor (vWF) may be regarded as markers of endothelial function, and high concentrations of these markers are predictive of the risk, presence, and severity of vascular disease (4)(5)(6).
Interest is growing in the measurement of markers of atherosclerosis to predict disease risk and to investigate effects of different interventions. Data on the within-subject variation of these markers are necessary to adequately estimate the required sample sizes for intervention trials and/or the number of blood samples needed to obtain a stable estimate of typical concentrations. However, data on the reproducibility (i.e., within-subject and analytical variation) of oxidative and endothelial markers are still sparse. We therefore investigated the reproducibility of several diagnostic markers of oxidative processes (OxLDL, the endogenous antioxidants bilirubin and uric acid, and ferritin) and endothelial function (sICAM-1, sVCAM-1, and vWF) in healthy individuals. The markers were measured in samples taken on three different occasions within 1 week, and in men and women of different ages.
The study population consisted of 25 volunteers (12 men and 13 women; age range, 26–58 years; mean, 41 years). All were healthy as assessed by a medical questionnaire and were not taking any medications known to affect hemostatic values. The Ethics Committee approved the study protocol, and all participants gave informed consent before participation.
Three blood samples were collected within 1 week (day 0, day 3 or 4, and day 7) between 0800 and 1000 after an overnight fast (from 2200), and participants were asked to abstain from drinking alcohol during the day before blood sampling. All venipunctures were performed for each participant at the same hour of the day. Participants were asked not to change their dietary habits, smoking habits, or physical activities during the study. We collected 5 mL of venous blood into a tube containing a mixture of citrate, theophylline, adenosine, and dipyridamole (CTAD Diatubes; Becton Dickinson/Diagnostica Stago) and 10 mL of blood into a 10-mL clotting tube to obtain serum. The blood sample in the CTAD tube was centrifuged within 30 min after collection (1200g at room temperature for 10 min) to obtain platelet-poor plasma. Extra vitamin E (20 µmol/L) and butylated hydroxytoluene (10 µmol/L) were added to the aliquots as antioxidants to prevent the in vitro oxidation of LDL-cholesterol in plasma. Serum was obtained by centrifugation at 1200g (10 min at room temperature) 1 h after venipuncture. All samples were divided into aliquots, stored at -80 °C (deep freeze) before analysis, and thawed at room temperature for analysis.
Plasma concentrations of circulating OxLDL were quantified by an immunoassay based on competition with the binding of monoclonal antibody 4E6 to the wells of microtiter plates coated with OxLDL (7)(8). Plasma concentrations of vWF antigen were measured with a two-step enzyme immunoassay with fluorescent detection of the final product (ELFA Enzym Linked Fluorescent Assay; bioMérieux sa). The intensity of the fluorescence is proportional to the concentration of antigen present in the sample. The VIDAS vWF test is calibrated against WHO International Standard 91/666, and the concentrations are expressed as a percentage (1 IU/mL = 100%). Serum concentrations of soluble ICAM-1 (sICAM-1) and soluble VCAM-1 (sVCAM-1) were estimated with a sandwich enzyme immunoassay (ELISA; R&D Systems Europe, Ltd.) with monoclonal antibodies specific for sICAM-1 and sVCAM-1 precoated on the microplates.
Serum concentrations of total bilirubin, uric acid, and ferritin were also measured. Total bilirubin was measured by the 2,5-dichlorophenyldiazonium method on Hitachi 917 and Hitachi 747 analyzers (Roche Diagnostics). Uric acid was measured with an enzyme-linked colorimetric assay (Roche), and ferritin was measured by microparticle enzyme immunoassay technology using AxSYM Ferritin Reagents (Abbott Laboratories). All samples from each volunteer were analyzed in one run to avoid between-run analytical variation, and OxLDL samples were analyzed in duplicate. In addition, serum and plasma aliquots were combined in a pool sample. This pool sample was analyzed in triplicate in each run to calculate analytical variation (or intraassay variation).
We used SAS® software (Ver. 8.1; SAS Institute, Inc.) to perform statistical analyses. The analytical variance (Va) was subtracted from the combined within-subject and analytical variance (Via) to obtain the within-subject variance (Vi) and to subsequently estimate the within-subject CV (CVi). Via and the mean marker concentration were calculated from the three samples per volunteer. Va (and the corresponding mean concentration) was calculated from the pool samples measured in triplicate within runs. CVia, CVi, and CVa were then calculated from the SD (SD = 
variance) and the corresponding mean concentration. The between-subject variation (CVg) of the different markers was calculated based on the formula: Vg = Vtotal - Via - Va.
The mean (SD), within-subject SD, and variability components (CVa, CVi, CVia, and CVg) obtained from measurements of the oxidative and endothelial markers in the three samples obtained within 1 week from 25 healthy individuals are shown in Table 1
. The CVis for OxLDL and total bilirubin were relatively high (21% and 22%, respectively), whereas the CVis for endothelial markers such as sICAM-1, sVCAM-1, and vWF antigen were rather low (1.9%, 5.2%, and 5.0%, respectively). The CVa of endothelial markers contributed >70% of the CVia. The between-subject variation of all markers was much higher than the corresponding within-subject variation. The CVg for ferritin is given for males and females separately because of gender differences in ferritin concentrations.
|
The present study showed a relatively high within-subject variation in OxLDL. It has been reported that the main determinants of plasma concentrations of OxLDL are hypercholesterolemia, body mass index, dyslipidemia, and age (9). However, these determinants reflect differences among individuals and not within-subject variations. In this study, we obtained a CVi of 5.6% for LDL-cholesterol. Exclusion of the four cigarette smokers did not change the concentration of circulating OxLDL. A main part of the within-subject variation is therefore unexplained. We speculate that different exposures to oxidative stressors (such as passive smoking or dietary components such as antioxidants and fatty acids) or to other determinants (such as physical activity) may play a role. The main between-subject determinants of typical concentrations of sICAM-1 and sVCAM-1 are HDL-cholesterol, systolic blood pressure, age, and fibrinogen (10). Our data suggest that a main part of the combined within-subject analytical variation for sICAM-1 and s-VCAM-1 within 1 week is explained by analytical variation.
The most important reasons to assess reproducibility are to determine the statistical power of experiments, to calculate a study’s required sample size, and/or to decide on the number of blood samples needed from one individual to accurately assess typical concentrations. For OxLDL (SD, 2.1 mg/L), we calculated, according to the formula:
 |
is the significance level, (1 - ß) is the power, SD is the within-subject SD, and DIF is the difference to detect], that a sample size of 42 healthy individuals would be needed to detect a 10% difference between the means in a crossover study at 5% significance with 80% power:
 |
One may argue that the three measurements within 1 week may be too close together to estimate the variation in markers of oxidative processes and endothelial function. Our results show that multiple blood sampling may be best, especially for OxLDL and total bilirubin because of the high within-subject variation. Repeated measurements will reduce the error attributable to within-subject fluctuations and reflect more closely a person’s true mean value within a certain period.
CVias for vWF antigen (blood samples were taken every 6 days for 30 days) (11), total bilirubin (every 3 or 4 days for 2 weeks) (12), and ferritin (daily for 3 weeks) (13) have been reported previously for healthy individuals. The mean concentrations for these variables were consistent with our results. The combined within-subject and analytical CVs for vWF antigen and ferritin observed in our study were lower than those reported previously [9.1% vs 18.7% for vWF antigen (11) and 11% vs 14.5% for ferritin (13)] and were almost equal for total bilirubin [22% in the present study vs 22.0% in a previous study (12)]. The lower CVias in our study, especially for vWF antigen, may be attributable to the shorter time period between blood sampling in our study (three samples within 1 week).
In conclusion, we determined in healthy individuals the producibility of markers of endothelial function and of oxidative processes. This biological variation must be taken into account when determining the number of participants and/or blood samples in future trials and observational studies to ensure sufficient power.
Acknowledgments
We thank the volunteers for their participation in the study. We also thank the personnel of the division of Juvenile Health for taking the blood samples and the Centre of Experimental Surgery and Anaesthesiology for assisting with OxLDL, sICAM-1, and sVCAM-1 measurements. The study was supported by a grant from the Unilever Chair in Nutritional and Health.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  T. S. Perlstein, R. L. Pande, J. A. Beckman, and M. A. Creager Serum Total Bilirubin Level and Prevalent Lower-Extremity Peripheral Arterial Disease: National Health and Nutrition Examination Survey (NHANES) 1999 to 2004 Arterioscler. Thromb. Vasc. Biol., January 1, 2008; 28(1): 166 - 172. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
