INFLUENCE OF AGE, ALCOHOL CONSUMPTION AND ABSTINENCE ON THE SENSITIVITY OF CARBOHYDRATE-DEFICIENT TRANSFERRIN, {gamma}-GLUTAMYLTRANSFERASE AND MEAN CORPUSCULAR VOLUME

Götz Mundle*, Klaus Ackermann, Jörg Munkes, Daniela Steinle and Karl Mann

University of Tübingen, Department of Psychiatry, Addiction Research Centre, Osianderstrasse 24, 72076 Tübingen, Germany

Received 18 September 1998; in revised form 25 January 1999; accepted 16 March 1999


    ABSTRACT
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Duration of abstinence before blood test, alcohol consumption and age was examined in 177 male alcohol-dependent patients as factors influencing serum carbohydrate-deficient transferrin (CDT), serum {gamma}-glutamyltransferase (GGT) and mean corpuscular volume (MCV). The strongest influence on all markers was the factor ‘duration of abstinence before blood test’. In patients who had been abstinent for >4 days before the blood test, the markers had low sensitivities (GGT, 33%; CDT, 14%; MCV, 42%), whereas in patients with <= 4 days of abstinence the markers had reasonably good sensitivities (GGT, 72%; CDT, 56%; MCV, 48%). GGT was more sensitive than CDT (P < 0.05) and MCV (P < 0.001). The combined use of CDT and GGT had sensitivity of over 90%. Mean alcohol consumption in the 30 days prior to the blood test had a significant effect on CDT and GGT, but not on MCV. Age did not have a clear effect on CDT and GGT. For MCV, a significant and linear increase with age was shown. We conclude that GGT is the most sensitive of these three markers. Using GGT and CDT combined, sensitivity can be enhanced to over 90%. The period of abstinence before the blood test has a strong influence on CDT and GGT. If a longer period of abstinence is suspected, MCV should also be measured, in order to detect evidence of earlier heavy drinking.


    INTRODUCTION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Excessive alcohol consumption and alcoholism are widely observed risk factors for health damage and social problems. Early identification of excessive alcohol consumption and alcoholism could improve the possibility of early treatment and reduce health damage and health care costs.

Although self-report is in general a valid and important diagnostic instrument (Maisto and O'Farrell, 1985Go; Babor et al., 1989Go; Rosman et al., 1992Go; Mundle et al., 1996Go), more objective and effective screening instruments, such as biological markers, are needed. Traditional markers are the membrane-bound liver enzyme {gamma}-glutamyltransferase (GGT) and the mean corpuscular volume (MCV), an index of red blood cell size (Conigrave et al., 1995Go). The difficulty of diagnostic accuracy has been a limiting factor, although these markers are widely available. The most effective marker has been GGT with a sensitivity of about 40–60% and a specificity of about 80% (Mihas and Tavassoli, 1992Go). A new marker for high alcohol consumption and alcoholism is carbohydrate-deficient transferrin (CDT). Early studies on CDT were promising and suggested high sensitivity and high specificity (Stibler, 1991Go). Studies comparing the diagnostic value of CDT with other biological markers usually suggested the superiority of CDT, particularly because of its specificity (Litten et al., 1995Go; Sillanaukee, 1996Go). Recent studies comparing the sensitivity of GGT and CDT, however, have shown a similar or even higher sensitivity for GGT in detecting heavy alcohol consumption and alcoholism (Helander and Tabakoff, 1997Go; Huseby et al., 1997Go; Yeastedt et al., 1998Go). Most studies have suggested that GGT and CDT appear to be independent and may therefore compliment each other. Their combined use is reported to increase diagnostic sensitivity with little loss of specificity (Litten et al., 1995Go; Sillanaukee, 1996Go; Yeastedt et al., 1998Go).

Factors influencing biological markers are duration of abstinence, age and alcohol consumption. The duration of abstinence before blood testing can cause differences in study results, especially for CDT, which has a relatively short half-life of 15 days (Behrens et al., 1988Go; Stibler et al., 1988Go). Nevertheless, the time since the last drink is included in only a few studies. General conclusions from study results are therefore limited in scope (Litten et al., 1995Go). The effects of age have also been reported in only a few studies. These indicate that GGT and MCV increase with age (Whitfield et al., 1978Go; Sillanaukee et al., 1998Go). Yersin et al. (1995) observed an increase in sensitivity in older patients from 56 to 74% for GGT and an increase from 17 to 39% for MCV. Daeppen et al. (1998) found a 64% average increase in GGT over a 15-year period in alcohol-drinking, but non-alcoholic, males, independent of alcohol consumption. The effect of age on CDT is inconsistent. Nystrom et al. (1992) recommended serum CDT for detecting early heavy drinking in young students, although the sensitivity was relatively low. Yersin et al. (1995) and Huseby et al. (1997) found that CDT values decreased with age; test results were highest in younger patients. Agelink et al. (1998) found that CDT was more sensitive in patients over 40 years of age than in younger patients. Conigrave et al. (1995) concluded that the sensitivity of all three markers is lower for hazardous consumption than for alcohol dependence: 20–60% vs 60–90% for CDT and GGT, and 20–30% vs 40–50% for MCV.

The aims of this study were to compare the sensitivities of CDT, GGT and MCV in a large clinical sample of male alcohol-dependent patients entering a 6-week in-patient treatment programme (Mann and Batra, 1993Go) and to analyse the factors influencing the test results of all three markers.


    PATIENTS AND METHODS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Patients
The study consisted of 177 male alcohol-dependent patients participating in the alcoholism treatment programme of the Addiction Research Centre, University of Tübingen, Germany. Corresponding with other samples (Mann et al., 1998Go), which were studied in this treatment and research unit, patients showed good social functioning despite alcohol dependence, and most lived in stable social circumstances; 75% were able to maintain jobs and 59% were married. The mean age (± SD) was 41.6 ± 9.2 years, the mean onset of dependency was 32.4 ± 9.4 years. The days of abstinence before the blood test ranged from 0 to 98 days with a mean of 16.8 ± 25.1 days. The mean alcohol consumption of those who drank in the 30 days before the blood test was 181 ± 122 g/day.

A baseline evaluation, carried out at the beginning of the in-patient programme, included a structured interview on alcohol history and social status (Mann, 1992Go), whose validity and reliability has been shown to be good (Mann et al., 1997Go). In addition, a physical examination and laboratory and psychological tests were performed. The diagnosis of alcoholism was established according to DSM III-R (American Psychiatric Association, 1987Go) and ICD-10 (World Health Organization, 1992Go). Exclusion criteria were drug dependence, severe psychiatric co-morbidity (i.e. schizophrenic or depressive episode) or significant medical illness (i.e. liver cirrhosis, tuberculosis or tumour). Patients using medication that might affect liver enzymes were also excluded from the study.

Biochemical analysis
Blood tests were performed at the beginning of the in-patient programme within 24 h of admission. They were performed at the clinical laboratory at the Medical Department of the University of Tübingen. For measurement of serum GGT and MCV the commercial kit reagents from Boehringer Mannheim (Mannheim, Germany) and Bayer Technikon H1 (Germany) were used according to the International Federation of Clinical Chemistry's recommended methods. GGT levels above normal were defined as GGT > 28 U/l at a measurement temperature of 25°C. The upper limit for MCV was 96 fl. Serum CDT was analysed using a commercial kit from Axis Biochemicals AS, which gives the result as a percentage of the amount of total transferrin. This kit uses micro-column ion-exchange separation of the transferrin isoforms and then quantification of the CDT isoforms by radioimmunoassay (Lyngbye et al., 1997Go). Normal CDT levels were defined as <=2.0%.

Statistical analysis
Differences in sensitivities between CDT, GGT and MCV were compared using the McNemar {chi}2-test. To analyse the effect of the duration of abstinence before the blood test, patients were divided in two groups (<=4 days and >4 days). Differences between these two groups were calculated by Fisher's exact test.

To assess potential group differences of mean levels of CDT, GGT and MCV due to age and alcohol consumption, patients were divided into three subgroups (alcohol consumption: <60 g/day; 60–200 g/day; >200 g/day; age: <35; 35–50 and >50 years).

Factorial analyses of variance were performed to investigate the influence of alcohol consumption and age for each marker. To control for days of abstinence, these were entered as a second factor in the analyses of variance. This was necessary, because days of abstinence did not show a normal distribution and therefore could not be entered as a covariate. Age was entered as a covariate in the analyses of variance of alcohol consumption, and alcohol consumption was entered as a covariate in the analyses of age. t-Tests for the adjusted means were calculated to analyse differences between the subgroups according to age and mean alcohol consumption.

To assess the correlations between markers, Spearman rank correlations were calculated with days of abstinence, age and alcohol consumption partialled out.

Prior to these analyses, CDT and GGT underwent logarithmic transformation to normalize distributions and to make variance more homogeneous. The Statistical Analysis System (SAS Institute, 1989) was used for calculations.


    RESULTS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Duration of abstinence
Sensitivities of the single markers and of different combinations of markers according to duration of abstinence are shown in Table 1Go. Differences in sensitivities of CDT and GGT between the two abstinence groups were highly significant, whereas this was not so with MCV. Comparing the three markers, GGT showed a significantly higher sensitivity than CDT (GGT/CDT {chi}2 = 4.74, P < 0.05) and MCV (GGT/MCV {chi}2 = 13.94, P < 0.001) in the subgroup of patients who were abstinent <=4 days before blood tests. In the subgroup of patients with more than 4 days of abstinence, CDT showed a significantly lower sensitivity than GGT (GGT/CDT {chi}2 = 8.54, P < 0.01) and MCV (MCV/CDT {chi}2 = 16.50, P < 0.001). The results in Table 1Go also show that sensitivity was significantly greater in the subgroup of subjects abstinent for <=4 days before blood testing for the different combinations of markers.


View this table:
[in this window]
[in a new window]
 
Table 1. Sensitivities of {gamma}-glutamyltransferase (GGT), mean corpuscular volume (MCV) and carbohydrate-deficient transferrin (CDT) for alcohol dependence according to daysof abstinence before blood testing
 
Alcohol consumption
The marker data as a function of level of alcohol consumption are shown in Table 2Go for both subgroups of patients in whom the blood tests were performed <= or >4 days after abstinence. In the two factorial analyses of variance, mean alcohol consumption per day in the 30 days before the blood tests showed a highly significant effect for GGT [F(2,173) = 8.61, P < 0.001] and a significant effect for CDT [F(2,173) = 4.61, P < 0.05]. Post-hoc analysis of adjusted means of GGT showed that the group with low alcohol consumption under 60 g/day significantly differed from the middle consumption group [t(176) = 3.814, P < 0.001] and from the high consumption group [t(176) = 3.539, P < 0.001]. The results of post-hoc analysis of adjusted means for CDT showed significant differences between the low consumption group and the high consumption group [t(176) = 2.711, P < 0.01]. The differences between the middle and upper groups were also significant [t(176) = 2.403, P < 0.05]. No significant main effects were found for MCV for mean alcohol consumption [F(2,172) = 1.54, P = 0.218].


View this table:
[in this window]
[in a new window]
 
Table 2. Means values of {gamma}-glutamyltransferase (GGT), mean corpuscular volume (MCV) and carbohydrate-deficient transferrin (CDT) according to alcohol consumption and days of abstinence before blood testing
 
To control for the influence of days of abstinence they were entered as a second factor. This factor showed highly significant effects for GGT [F(1,173) = 7.01, P < 0.001] and CDT [F(1,173) = 19.22, P < 0.001], but not for MCV [F(1,172) = 0.31, P = 0.581]. To control for age, this factor was entered as a covariate and showed a significant effect for MCV [F(1,172) = 9.38, P < 0.01].

Age
The laboratory marker data as a function of age of patients are shown in Table 3Go for both the <= and >4 days of abstinence subgroups. In the two factorial analyses of variance, age showed a significant effect for MCV [F(2,173) = 4.04, P < 0.05], but not for GGT [F(2,172) = 0.10, P = 0.901] or CDT [F(2,172) = 0.66, P = 0.519]. Post-hoc analysis of adjusted means for MCV revealed that the younger group significantly differed from the middle group [t(176) = 2.295, P < 0.05] and from the older group [t(176) = 2.605, P < 0.05].


View this table:
[in this window]
[in a new window]
 
Table 3. Mean values of {gamma}-glutamyltransferase (GGT), mean corpuscular volume (MCV) and carbohydrate-deficient transferrin (CDT) according to age and days of abstinence before blood testing
 
The second factor, days of abstinence, showed a highly significant effect for GGT [F(1,172) = 20.27, P < 0.001] and CDT [F(1,172) = 19.64, P < 0.001], but not for MCV [F(1,172) = 2.19, P = 0.141]. To control for mean alcohol consumption, this factor was entered as covariate and showed a significant effect for CDT [F(1,172) = 11,07, P < 0.01], but not for GGT [F(1,172) = 3.5, P < 0.10].

Correlations
Significant Spearman rank correlations among the three biological markers were found for MCV and CDT (r = 0.15, P < 0.05%) and for MCV and GGT (r = 0.29, P < 0.001) (Table 4Go). There was no significant correlation between CDT and GGT (r = –0.11).


View this table:
[in this window]
[in a new window]
 
Table 4. Spearman rank correlations of biological markers
 

    DISCUSSION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
The present results support the hypothesis that GGT might be the most sensitive marker for alcoholism (Helander and Tabakoff, 1997Go; Yeastedt et al., 1998Go). Comparing the sensitivity of all markers in our study, GGT showed a significantly higher sensitivity than CDT. Since no control group was included in this study and because differences of specificity could not be investigated, these conclusions are limited. It is also clear from the literature that there seems to be no doubt that the advantage of CDT compared to GGT is its high specificity (Conigrave et al., 1995Go; Litten et al., 1995Go).

The combination of markers enhanced sensitivity for alcoholism. A sensitivity of over 90% could be achieved by the combination of CDT and GGT. The combination of all three markers provided little further information. The increase in sensitivity using CDT and GGT conjointly confirms earlier studies, and is mainly explained by the fact that these markers increase independently and are not correlated (Litten et al., 1995Go; Huseby et al., 1997Go; Yeastedt et al., 1998Go). Also, in our study, there was no significant correlation between CDT and GGT. The combination of GGT and MCV was not as effective, because these markers correlated significantly.

MCV seems to be a marker of second choice for alcohol-dependent men. As CDT is not yet widely available and remains expensive, MCV is still, despite its low sensitivity, a useful tool for screening, particularly when used conjointly with GGT (Monteiro and Masur, 1985Go; Levine, 1990Go; Seppä et al., 1996Go; Sillanaukee et al., 1998Go).

The strongest influence on the test results of CDT and GGT was duration of abstinence before blood testing. In patients with only a few days of abstinence CDT and GGT showed significantly higher sensitivities. This effect is probably caused by the relatively short half-life of CDT of 15 days (Behrens et al., 1988Go; Stibler et al., 1988Go) and of GGT of about 4 weeks (Orrego et al., 1985Go). Therefore, duration of abstinence before blood testing should be considered in research studies and in clinical practice, otherwise results of CDT and GGT could be difficult to interpret (Litten et al., 1995Go; Agelink et al., 1998Go). Since MCV has a relatively longer half-life of 2 to 3 months (Renwick and Asker, 1982Go), it was not as susceptible to the influence of this factor. In clinical practice, MCV should also be measured if there is a longer period of abstinence suspected (Haffner et al., 1989Go).

In accordance with the literature (Conigrave et al., 1995Go; Litten et al., 1995Go; Sillanaukee, 1996Go) the degree of alcohol consumption had a significant effect on GGT and CDT. Even in alcohol-dependent patients, low alcohol consumption of up to 60 g/day was difficult to detect and mean levels of CDT or GGT were normal if these patients had been abstinent more than 4 days before blood sampling. This result indicates again that GGT and CDT are not ideal markers in cases of low alcohol consumption. The absence of a significant correlation between alcohol consumption and MCV is probably due to the time frame of 1 month which we used for the alcohol consumption figure. In general, other studies indicate a positive correlation between the amount of alcohol consumed and MCV (Chick et al., 1981Go; Tonnesen et al., 1986Go).

A significant influence of age on test results was only found for MCV, but not for CDT or GGT. MCV showed significantly higher levels for older patients and lower results for younger ones. This finding is consistent with those of previous studies (Whitfield et al., 1978Go; Yersin et al., 1995Go; Sillanaukee et al., 1998Go) and indicates that MCV increases linearly with age. The finding that GGT is not affected by age contradicts other studies (Schiele et al., 1977Go; Van Der Meulen et al., 1993Go; Yersin et al., 1995Go) showing an increasing effect of age, but supports studies showing no clear age effect for GGT. Whitfield et al. (1978) found increasing values of GGT up to the age of 50 years, with a decrease thereafter. In the study by Sillanaukee et al. (1998), the correlation of age and GGT was only just significant (P = 0.049) and weak (r = 0.06). Persson et al. (1990) failed to demonstrate an association between age and GGT in 188 teetotallers. Daeppen et al. (1998) investigated patients only up to age 35 years. Changes in GGT over a 15-year period were minor. A linear association between age and GGT is therefore still unproven. It is possible that middle-aged patients have higher GGT values than younger and older patients. In our study, the highest GGT values were obtained in patients aged 35–50 years.

Our result suggesting that CDT is not significantly correlated with age is in accordance with those of Sillanaukee et al. (1998). In this latter study, a significant correlation was found between age and MCV, but not between age and CDT. The findings that both younger and older patients have higher CDT values (Yersin et al., 1995Go; Huseby et al., 1997Go; Agelink et al., 1998Go) are indirectly supported by our study. Younger and older patients showed higher CDT levels, compared to middle-aged patients, though these effects did not reach significance and therefore a clear age effect could not be proven.

In conclusion, the results of our study confirm that CDT and GGT are effective markers for alcoholism. The two important factors influencing GGT and CDT are the duration of abstinence before blood testing and quantity of alcohol consumed. Clear age effects are shown only for MCV. If the limitations of each marker are understood and taken into consideration, biological markers, especially used in combinations, are useful aids in diagnosis of alcoholism.


    ACKNOWLEDGEMENTS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
The authors wish to thank Dr I. Besenthal, Assistant Director of the Clinical Laboratory of the Medical Department for his support in this work and Dr U. Ren for his help with the manuscript. The study was supported by the Federal Goverment of Germany (BMFT 07 FDA 02).


    FOOTNOTES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
* Author to whom correspondence should be addressed. Back


    REFERENCES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 REFERENCES
 
Agelink, M. W., Kirkes-Kerstin, A., Zeit, T., Bertling, R., Malessa, R. and Klieser, E. (1998) Sensitivity of carbohydrate-deficient transferrin (CDT) in relation to age and duration of abstinence. Alcohol and Alcoholism 33, 164–167.[Abstract]

American Psychiatric Association (1987) Diagnostic and Statistical Manual of Mental Disorders, 3rd edn, revised. American Psychiatric Association, Washington, DC.

Babor, T. F., Kranzler, H. R. and Lauerman, R. J. (1989) Early detection of harmful alcohol consumption: comparison of clinical, laboratory, and self-report screening procedures. Addictive Behaviour 14, 139–157.[ISI][Medline]

Behrens, U. J., Worner, T. M. and Lieber, C. S. (1988) Changes in carbohydrate-deficient transferrin levels after alcohol withdrawal. Alcoholism: Clinical and Experimental Research 12, 539–544.[ISI][Medline]

Chick, J., Kreitmann, N. and Plant, M. (1981) Mean cell volume and gamma-glutamyltranspeptidase as markers of drinking in working men. Lancet i, 1249–1251.

Conigrave, K. M., Saunders, J. B. and Whitfield, J. B. (1995) Diagnostic tests for alcohol consumption. Alcohol and Alcoholism 30, 13–26.[Abstract]

Daeppen, J. B., Smith, T. L. and Schuckit, M. A. (1998) Influence of age and body mass index on {gamma}-glutamyltransferase activity: a 15-year follow up evaluation in a community sample. Alcoholism: Clinical and Experimental Research 22, 941–944.[ISI][Medline]

Haffner, H. T., Becker, I. S. and Mann, K. (1989) The sensitivity of clinico-chemical markers of alcoholism following short-term motivated alcohol abstinence. Blutalkohol 26, 114–122.[Medline]

Helander, A. and Tabakoff, B. (1997) Biochemical markers of alcohol use and abuse: experiences from the Pilot Study of the WHO/ISBRA Collaborative project on state and trait markers of alcohol. Alcohol and Alcoholism 32, 133–144.[Abstract]

Huseby, N. E., Nilssen, O., Erfurth, A., Wetterling, T. and Kanitz, R. D. (1997) Carbohydrate-deficient transferrin and alcohol dependency: variation in response to alcohol intake among different groups of patients. Alcoholism: Clinical and Experimental Research 21, 201–205.[ISI][Medline]

Levine, J. (1990) The relative value of consultation, questionnaires and laboratory investigation in the identification of excessive alcohol consumption. Alcohol and Alcoholism 25, 539–553.[ISI][Medline]

Litten, R. Z., Allen, J. P. and Fertig, J. B. (1995) Gamma-glutamyltranspeptidase and carbohydrate-deficient transferrin: alternative measures of excessive alcohol consumption. Alcoholism: Clinical and Experimental Research 19, 1541–1546.[ISI][Medline]

Lyngbye, J., Eide, A., Walter, H., Semler, F., Lesch, O. M. (1997) Clinical value of two methods for the determination of carbohydrate-deficient transferrin. Clinical Laboratory 43, 53–55.

Maisto, S. A. and O'Farrell, T. J. (1985) Comment on the validity of Watson et al.'s ‘Do alcoholics give valid self-reports?’. Journal of Studies on Alcohol 46, 447–453.[ISI][Medline]

Mann, K. (1992) Alcohol and the Brain [Alkohol und Gehirn]. Springer, Berlin,

Mann, K. and Batra, A. (1993) The community based treatment of alcoholics; evaluation of a combined inpatient and outpatient program. Psychiatrische Praxis 20, 102–105.[ISI][Medline]

Mann, K., Kapp, B. and Hunkeler, P. (1997) Validity and reliability of alcoholics' self-reports. Alcoholism: Clinical and Experimental Research 21, 94A.

Mann, K., Ackermann, K., Jung, M., Morlock, P. and Mundle, G. (1998) Aggressiveness, onset of dependence, and treatment outcome in socially well-adapted alcoholics. Alcohol and Alcoholism 33, 16–19.[Abstract]

Mihas, A. A. and Tavassoli, M. (1992) Laboratory markers of ethanol intake and abuse: a critical appraisal. American Journal of Medical Sciences 303, 415–428.[ISI]

Monteiro, M. G. and Masur, J. (1985) Diagnostic of alcoholism: how useful is the combination of gamma glutamyl transferase with different biochemical markers? Drug and Alcohol Dependence 16, 31–37.[ISI][Medline]

Mundle, G., Ackermann, K., Günthner, A. and Mann, K. (1996) Treatment outcome in alcoholism: a comparison of self-reports and biological markers. Alcoholism: Clinical and Experimental Research 20, 813.

Nystrom, M., Perasalo, J. and Salaspuro, M. (1992) Carbohydrate-deficient transferrin (CDT) in serum as a possible indicator of heavy drinking in young university students. Alcoholism: Clinical and Experimental Research 16, 93–97.[ISI][Medline]

Orrego, H., Blake, J. E. and Israel, Y. (1985) Relationship between gamma-glutamyltranspeptidase and mean urinary alcohol levels in alcoholics while drinking and after alcohol withdrawal. Alcoholism: Clinical and Experimental Research 9, 10–13.[ISI][Medline]

Persson, J., Magnusson P. H. and Borg, S. (1990) Serum gamma-glutamyl transferase (GGT) in a group of organized teetotallers. Alcohol 7, 87–89.[ISI][Medline]

Renwick, J. H. and Asker, R. L. (1982) The time course of response of erythrocyte volume to ethanol and to its withdrawal. Clinical Laboratory of Haematology 4, 325–326.

Rosman, A. S., Lieber, C. S., Litten, R. Z. and Allen, J. P. (1992) An Overview of Current and Emerging Markers of Alcoholism Measuring Alcohol Consumption: Psychosocial and Biochemical Methods. Humana Press, Totowa, NJ.

Institute, Inc. (1989) SAS/STAT User's Guide, Release 6.03, 4th edn. SAS Institute, Cary, NC.

Schiele, F., Guilmin, A. M., Detienne, H. and Siest, G. (1977) Gamma-glutamyl-transferase activity in plasma: statistical distributions, individual variations, and reference intervals. Clinical Chemistry 23, 1023–1028.[Abstract/Free Full Text]

Seppä, K., Heinila, K., Sillanaukee, P. and Saarni, M. (1996) Evaluation of macrocytosis by general practitioners. Journal of Studies on Alcohol 57, 97–100.[ISI][Medline]

Sillanaukee, P. (1996) Laboratory markers of alcohol abuse. Alcohol and Alcoholism 31, 613–616.[Abstract]

Sillanaukee, P., Aalto, M. and Seppä, K. (1998) Carbohydrate-deficient transferrin and conventional alcohol markers as indicators for brief intervention among heavy drinkers in primary health care. Alcoholism: Clinical and Experimental Research 22, 892–896.[ISI][Medline]

Stibler, H. (1991) Carbohydrate-deficient transferrin in serum: a new marker of potentially harmful alcohol consumption reviewed. Clinical Chemistry 37, 2029– 2037.

Stibler, H., Dahlgreen, L. and Borg, S. (1988) Carbohydrate-deficient transferrin (CDT) in serum in women with early alcohol addiction. Alcohol 5, 393–398.[ISI][Medline]

Tonnesen, H., Hejberg, L., Frobenius, S. and Andersen, J. R. (1986) Erythrocyte mean cell volume — correlation to drinking pattern in heavy alcoholics. Acta Medica Scandinavica 219, 515–518.[ISI][Medline]

Van Der Meulen, E. A., Van Sittert, N. J., De Koningh, A. G. J., Lugtenburg, D. and Van Strik, R. (1993) General approach to correction for bias in analytical performance in longitudinal studies illustrated by estimating the effect of age on {gamma}-glutamyltransferase activity. Clinical Chemistry 39, 1375–1378.[Abstract/Free Full Text]

Whitfield, J. B., Hensley, W. J., Bryden, D. and Gallagher, H. (1978) Effects of age and sex on biochemical responses to drinking habits. Medical Journal of Australia 2, 629–632.[Medline]

World Health Organization (1992) International Classification of Diseases and Related Health Problems, 10th edn. World Health Organization, Geneva.

Yeastedt, J., LaGrange, L. and Anton, R. F. (1998) Female alcoholic outpatients and female college students: a correlation study of self-reported alcohol consumption and carbohydrate-deficient transferrin levels. Journal of Studies on Alcohol 59, 555–559.[ISI][Medline]

Yersin, B., Nicolet, J. F., Dercrey, H., Burnier, M., van Melle, G. and Pecoud, A. (1995) Screening for excessive alcohol drinking. Comparative value of carbohydrate-deficient transferrin, gamma-glutamyltransferase and mean corpuscular volume. Archives of Internal Medicine 155, 1907–1911.[Abstract]