| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Alcohol Treatment Center, CHUV, Lausanne, Switzerland;2 Institute of Forensic Medicine, CHUV, Lausanne, Switzerland;
aaddress correspondence to this author at: Alcohol Treatment Center, Mont-Paisible 16, CHUV, 1011 Lausanne, Switzerland; e-mail jean-bernard.daeppen{at}inst.hospvd.ch
Carbohydrate-deficient transferrin (CDT) measured by capillary zone electrophoresis (CZE), particularly asialo-transferrin (Tf), is purported to better differentiate between excessive and moderate drinkers than does CDT measured by turbidimetric immunoassay (TIA) (1)(2). The use of biological markers such as CDT is of particular interest for identifying young heavy drinkers because other clinical signs of heavy drinking are generally absent and heavy drinking is a leading cause of morbidity and mortality in this age group (3)(4). Several authors have shown interest in the ability of CDT to identify nondependent heavy drinkers (5)(6); we therefore describe here the performance of CZE measurements of asialo- and disialo-Tf and TIA analysis of CDT in a large community sample of 19-year-old men, of whom 21% were heavy drinkers.
From a sample of 1018 men attending a mandatory 1-day army recruitment process for all Swiss males at age 19 years, 1004 (98.6%) agreed to complete a research questionnaire. Of these, 581 young men (57.9%) consented to give blood for the measurement of asialo-Tf (CZE), disialo-Tf (CZE), and CDT (TIA). The Ethics Committee of the Lausanne University Medical School approved the study protocol. Volunteers were compensated for participation in the study.
Volunters gave written informed consent and then completed an instrument entitled "Health and Lifestyle Questionnaire", which included questions assessing the typical quantity and frequency of alcohol consumption during the 12 months preceding the survey and the frequency of drunkenness over the last 30 days. One drink was defined as a 250-mL can or bottle of beer, a 120-mL glass of wine, or a 40-mL shot of liquor straight or in a mixed drink, and corresponded to 
12 g of pure ethanol.
A study investigator was present during administration of the questionnaire to verify that participants answered all items. Serum samples were obtained by centrifugation of peripheral blood collected in 10-mL tubes. Samples were stored at –20 °C before analysis.
Total CDT was measured by anion-exchange chromatography and TIA with the Axis-Shield CDT (TIA) reagent set (7). To separate and measure Tf isoforms, we used a previously described and validated CZE method (8)(9) with the Ceofix CDT reagent set (Analis) on a Hewlett Packard (HP) 3D-CE instrument. The CZE conditions are described in Table 1 of the Data Supplement that accompanies the online version of this Technical Brief at http://www.clinchem.org/content/vol51/issue6/. CZE electropherograms showing the serum Tf profiles for a heavy drinker before and after addition of anti-Tf polyclonal antibody to the serum are shown in Fig. 1 of the online Data Supplement, and CZE electropherograms showing the Tf profiles of a teetotaler and of 2 heavy drinkers are shown in Fig. 2 of the online Data Supplement.
Peaks representing the different Tf isoforms were quantified as the amounts of the asialo-, disialo-, trisialo-, tetrasialo-, pentasialo-, and hexasialo-Tf (CZE) as a percentage of the total Tf content, in terms of valley-to-valley areas under the curve. The intraday CV values (n = 6) for "low" (0.6% by CZE) and "high" disialo-Tf (4.8% by CZE) were 9.8% and 1.2%, respectively, and the interday CVs (n = 5) for low (0.6% by CZE) and high disialo-Tf (4.8% by CZE) were 11% and 2.3%, respectively. The intra- and interday CVs for asialo-Tf (0.5% by CZE; n = 6) were 6.8% and 11%, respectively. The limit of quantification of each Tf (CZE) isoform was 0.1%, expressed a percentage of total Tf isoforms.
Continuous data are reported as the mean (SD) and the median (interquartile range). We used a 
2 test to compare categorical variables and Mann–Whitney U-tests to compare continuous variables because this nonparametric statistic makes no assumption about the distributional properties of variables. We also determined the areas under the ROC curves (AUROC), the sensitivity, and the specificity for disialo-Tf (measured by CZE) and CDT (measured by TIA) in identifying heavy drinkers.
There were 121 (20.8%) heavy drinkers in the sample: 31 (5.3%) who reported typical alcohol consumption of >21 drinks/week over the last 12 months, 52 (8.9%) who said they had been drunk at least 3 times over the last month; and 38 (6.5%) who reported both. Mean (SD) alcohol consumption in heavy drinkers was 26.4 (8.4) drinks (300 g of ethanol) per week. Among the remaining participants, 435 (74.9%) were categorized as moderate drinkers, reporting, on average, 6.0 (4.7) drinks (65 g of ethanol) per week, and 25 (4.3%) were considered abstinent (mean reported quantity and frequency = 0). The abstaining participants were retained as part of the moderate-drinker group.
Our results indicate that asialo-Tf (CZE) could not differentiate between moderate and heavy drinkers because 574 (98.8%) of the participants had a asialo-Tf (CZE) value of 0% and only 3 moderate drinkers and 4 heavy drinkers had positive values. We did, however, find significant differences between heavy and moderate drinkers for disialo-Tf as measured by CZE {mean (SD), 0.8 (0.6)% [median (interquartile range), 0.7 (0.5–0.9)%] vs 0.6 (0.2)% [0.6 (0.5–0.8)%]; P <0.01} and for CDT as measured by TIA {2.5 (0.8)% [2.3 (2.0–2.8)%] vs 2.1 (0.5)% [2.0 (1.8–2.3)%]; P <0.001}. The areas under the ROC curves for disialo-Tf (CZE) and CDT (TIA) are shown in Fig. 1
. ROC curve analysis indicated low sensitivities and specificities for disialo-Tf measured by CZE (AUROC = 0.58) and for CDT measured by TIA (AUROC = 0.66) in identifying heavy drinkers, with optimal cutoffs (inflection point on ROC curve) of 0.62% (sensitivity, 60.3%; specificity, 52.1%) and 2.2% (sensitivity, 58.7%; specificity, 63.2%), respectively. As also shown in Fig. 1
, the sensitivities and specificities at the usual cutoffs of 0.7% for disialo-Tf measured by CZE (sensitivity, 47.1%; specificity, 64.4%) (1) and 2.6% for CDT measured by TIA (sensitivity, 34.7%; specificity, 87.2%) (7) were not optimal.
|
We also explored the sensitivities and specificities of disialo-Tf (CZE) and CDT (TIA) for differentiating the 121 heavy drinkers from the 25 abstainers after excluding the 435 moderate drinkers (not reported in Fig. 1
). At the optimal cutoff of 0.59%, the sensitivity and specificity of disialo-Tf (CZE) in identifying heavy drinkers were 64.5% and 68.0%, respectively (AUROC = 0.66); at the optimal cutoff of 2.0% for CDT (TIA), the sensitivity and specificity of CDT (TIA) were 76.9% and 64.0% (AUROC = 0.75), respectively.
Generalizing the relationship between alcohol use and asialo-Tf (CZE), disialo-Tf (CZE), and CDT (TIA) concentrations to the total sample of 1004 men would be possible only if the participants who refused to give blood had alcohol use and alcohol-related problems that were similar to those who consented to give blood. Although not reported in Fig. 1
or the supplemental data, the results also indicated that the 581 participants who agreed to give blood were consuming significantly more alcohol [mean (SD), 10.0 (12.61) vs 8.1 (10.39) drinks per week; P <0.05] and were more likely to report having been drunk at least 3 times over the last 30 days (15.5% vs 11.4%; P <0.01) than were the 423 participants who refused to give blood.
Our results indicate that asialo-Tf (as measured by CZE) is of no diagnostic value for young men. These results contrast with previous work suggesting that asialo-Tf (CZE) was highly efficient in identifying patients reporting a mean alcohol consumption >50 g/day (AUROC = 0.91) (1). Two factors might help to explain the differences observed between that earlier study and the present one: In the earlier study, participants were older, consumed more alcohol, and were more likely to have altered hepatic function related to alcohol (1)(8). These conditions apparently increase the sensitivity of CDT (10). The other factor, as suggested by Legros et al. (8), was that in the earlier study (1), the performance of the biological markers was amplified because the participants were compared after exclusion of those who were abstinent or alcohol dependent and those who reported drinking 30–50 g of ethanol/day. These findings suggest that the superiority of asialo-Tf (CZE) over CDT (TIA) is evident only in individuals who consume excessive amounts of alcohol, such as very heavy drinkers, who are predominantly alcohol dependent.
Our results confirm that either heavy alcohol consumption or regular drunkenness significantly increases disialo-Tf (CZE) and CDT (TIA), but does not do so sufficiently to differentiate heavy from moderate drinkers. Disialo-Tf (CZE) and CDT (TIA) had relatively similar sensitivities and specificities for identifying young heavy drinkers. According to previous findings (1)(8), the exclusion of moderate drinkers amplified the sensitivities and specificities of disialo-Tf (CZE) and CDT (TIA) for correctly classifying heavy drinkers and teetotalers.
Our data also confirm the predictive limitations of asialo-Tf (CZE) and disialo-Tf (CZE), as described earlier for CDT (TIA), i.e., that the sensitivity of CDT is decreased at younger ages (10)(11) and that CDT performs poorly for the identification of heavy alcohol consumption in college students (12)(13).
Although these results may broadly apply to young men, it is important to recognize several limitations when generalizing these findings to other populations. These results may not hold true for samples of other individuals, such as women, older persons, or those recruited within medical settings. Our study sample consisted mostly of Caucasians; thus, the findings may not apply to other ethnic groups. The differences we observed in drinking patterns between those who agreed to give blood and those who refused preclude generalizing the findings to the overall sample. Finally, although great effort was made to optimize the accuracy of these data, the information obtained regarding alcohol use and alcohol-related problems was based solely on the estimates and recollections of the participants.
Acknowledgments
We are grateful to Magali Dovat for skillful technical assistance with asialo-Tf (CZE) and disialo-Tf (CZE) measurements and to George Danko, PhD, for careful help in the editing of the manuscript.
References
-glutamyl transferase (GGT), and mean corpuscular volume (MCV). Alcohol Clin Exp Res 2000;24:1414-1419.[CrossRef][Web of Science][Medline]
[Order article via Infotrieve]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
