| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1
Lab. de Referència de Catalunya;
2
Servei de Bioquim., Hosp. General Vall d'Hebron and
3
Dept. d'Epidemiol, IMIM, Barcelona, Spain;
a address for correspondence: Unitat de Lípids i Epidemiol. Cardiovascular, IMIM, Carrer Dr. Aiguader, 80, 08003 Barcelona, Spain, fax +34–3-2213237, e-mail MCOVAS{at}IMIM.ES
Oxidative stress resulting from increased free-radical production and (or) defects in antioxidant defenses is implicated in the pathogenesis of several diseases (1). Biological effects of these highly reactive compounds are controlled in vivo by a wide spectrum of antioxidant mechanisms. Among these, the enzymes superoxide dismutase (SOD) and glutathione peroxidase (GPX) act as endogenous antioxidants (2). Introducing SOD and GPX measurements in laboratory medicine application requires identification of the biological sources of variation for: assessing the desired performance characteristics of a test, deriving the utility of the conventional population-based reference interval, and defining significant differences that imply changes in serial results from an individual (3).
Our study was designed to estimate biological variation components of SOD in erythrocytes and GPX in whole blood to examine the role of these enzymes in the diagnosis, screening, and monitoring of patients.
Eighteen healthy individuals (8 men and 10 women) between 24 and 48 years of age were included in the study. They agreed to maintain uniform daily habits during the study. Venous blood samples were collected into lithium heparinized tubes once a week, between 0800 and 0900 h, for 5 consecutive weeks. Specimens were drawn from the volunteers after 20 min in a sitting position. Owing to the short time stability of the samples, their analyses were performed on the day of collection.
Analytical procedures: SOD activity in erythrocytes was measured by the rate of inhibition of 2-(4-iodophenyl)-3-(4-nitrophenol)-5-phenyltetrazolium chloride (INT) reduction (Ransod SD 125, Randox Lab.) and expressed in U/g of hemoglobin (Hb). GPX activity in whole blood was measured by a modification of the method of Paglia and Valentine (4) (Ransel RS 505, Randox Lab.) and expressed in U/L. Enzyme activities were measured in a Cobas Mira Plus analyzer (Hoffmann-La Roche) at 37 °C. Hb concentration was determined with a Sysmex K1000 hematological analyzer (Toa Electronics). An internal control system based on commercial lyophilized blood (Ransod control SD 126 and Ransel control SC 692, Randox Lab.) was used for validation of analytical runs.
Analytical between-run imprecision was determined from 20 day-to-day measurements of control samples and expressed as CVab. Before analyzing the data to establish biological variation, the Cochran test (5) and the Reed test (6) were used to exclude outlying values. From the samples, 2 of 180 values for SOD and 2 of 180 values for GPX were ruled out. One individual was ruled out when the Reed test was applied to the GPX set of data. The index of heterogeneity (3) and the Levene test (7) were applied to check the homogeneity of variances. Two individuals were ruled out according to the Levene test.
Biological within-subject (intraindividual) variation
(CVi) was estimated from the total within-subject
variance (intraindividual plus analytical) minus between-run analytical
variance (3) with the formula:
 | (1) |
The term (s2i+a) was computed
by averaging data from all volunteers
(
s2i+a/n). The between-subject
(global) biological variation (CVg) was obtained
(3) with the formula:
 | (2) |
Index of individuality (II) as the ratio:
 | (3) |
Critical differences for serial results (CD) (P <0.05)
(3), applying the formula:
 | (4) |
Desirable quality specifications for analytical imprecision
(CVD) according to the following expression
(8):
 |
 |
After logarithmic transformation to normalize the data, Student's t-test was applied to examine differences between sexes, and regression analysis and correlation performed between the two enzyme activities in the volunteers' samples.
A slight but statistically significant positive correlation was found
between SOD in erythrocytes and GPX in total blood (r =
0.301, P = 0.0001). Between-run imprecision
(CVab) was 6.12% for SOD and 6.57% for GPX. Desired
quality specifications for imprecision were 6.19% for SOD analysis in
erythrocytes and 3.86% for GPX analysis in whole blood. Bias was
3.36% and 5.75% for SOD and GPX, respectively. Table 1
shows biological variation components and data derived from
them in both sexes and in the whole population.
|
The positive correlation found between SOD in erythrocytes and GPX in whole blood appears logical since GPX is, together with catalase, one of the enzymes that destroys the H2O2 generated by SOD action (2). The analytical between-run CVab accounts for 25.73% of the total variation (analytical plus biological) for SOD and 18.52% for GPX. These percentages are similar to those obtained for other enzymes with well-established biological variation estimates in studies where volunteer data were obtained in the same run (9). These results support that estimation of biological variation components with different lots does not substantially affect reliability of the information obtained.
The lower total biological variation (CVi +
CVg) suggests than SOD is less sensitive to environmental
or physiological changes than GPX. CVi of SOD and GPX was
similar and comparable with those obtained for other enzymes such as
serum alkaline phosphatase, amylase, or aspartate aminotransferase and
lower than those obtained for serum 
-glutamyltransferase or creatine
kinase (10). CVi for SOD in erythrocytes
obtained in the women' s group (12.8%) lay between the two values
reported by Gallagher et al. (17.5% in free diet, and 10.91% in
controlled diet) (11), who studied SOD activity in serum
of five healthy females in the same time span evaluated here. The lower
CVi for SOD in erythrocytes obtained in this study in
non-diet-controlled women could be explained by the homogeneity of data
in a higher number of participants.
We applied the index of heterogeneity (IH) in the whole set of data, and values obtained were clearly lower than the critical value 1.45 (12), indicating homogeneity between within-subject variances. However, observation of the individual CVi revealed that some volunteers showed discrepant values compared with the global set of data. The Levene test was therefore applied and two individuals subseqently had to be ruled out. Thus, IH is not applicable in our set of data to establish homogeneity of variances.
II is considered to be the key for determining the practical utility of population-based reference ranges. If <0.6, the use of a reference interval is of little value for diagnostic purposes. If it is >1.4, then the reference interval is valuable (3). II for GPX was 0.45, suggesting that it has little value as a diagnostic or screening tool. The differences in SOD II between sexes are principally related to the differences in SOD between-subject variation. At present, possible causes (influence of life-style or hormonal status) for these sex differences remain unknown.
To our knowledge, this is the first report on the biological variation of GPX in whole blood and SOD in erythrocytes. Data obtained from the biological variation of these enzymes support SOD in erythrocytes as the scavenger enzyme of choice for diagnosis of an alteration in antioxidant status in a pathological situation, as well as for screening in population studies. In view of its strong individuality and smaller critical differences compared with SOD, GPX determination in total blood would be useful for monitoring antioxidant status in pathological situations and changes in life-style such as diet or exercise. Improvement in measurement of GPX in blood is required to achieve the analytical goal established in this work.
Acknowledgments
We thank Joan Vila and Jordi Sunyer for their assessment in statistical analyses and Christine O'Hara for the English revision of the manuscript. This work was supported in part by grant from the CIRIT (1995/SGR/00167).
Footnotes
Unitat de Lípids i Epidemiol. Cardiovascular. Inst., Municipal d'Investigació Mèdica (IMIM)
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
