CARBOHYDRATE-DEFICIENT TRANSFERRIN: MARKER OF ACTUAL ALCOHOL CONSUMPTION OR CHRONIC ALCOHOL MISUSE?

Alcohol & Alcoholism Vol. 33, No. 6, pp. 646-650. 1998 CARBOHYDRATE-DEFICIENT TRANSFERRIN: MARKER OF ACTUAL ALCOHOL CONSUMPTION OR CHRONIC ALCOHOL MISUSE? IDUN-MERETE MIKKELSEN*, ROLF-DIETER KANITZ1, ODD NILSSEN 2 and NILS-ERIK HUSEBY Institute of Medical Biology, University of Troms0, N-9037 Troms0. Norway, 'Kliniken des Kreises Pinneberg. Kreiskrankenhaus Elmshorn, Germany and institute of Community Medicine, University of Troms0, N-9037 Troms0, Norway Abstract — Carbohydrate-deficient transferrin (CDT) is a useful indicator of excessive alcohol consumption with higher sensitivity and specificity than other markers that are used. In the present study, CDT was analysed in 161 patients hospitalized in a surgical ward to evaluate whether history of drinking and chronic alcohol misuse are important determinants of CDT elevations. Fifty-one of the patients were diagnosed as alcohol-dependent and they all reported a long history of alcohol abuse Several of these, as well as many of the non-dependent patients, reported a high, recent alcohol consumption (>60 g/day for the previous 2 weeks). CDT performed better in detecting patients with alcohol dependency than in detecting patients with high alcohol consumption irrespective of dependency, showing higher sensitivity (47 vs 37%), likelihood ratio (4.7 vs 3.4), and a statistically significant difference in the receiveroperating characteristic curve areas (P = 0.04 in a two-tailed comparison test). In two subgroups, one with alcohol-dependent and one with non-dependent patients, consuming similar amounts of alcohol (range: 60—170 g/day), the sensitivity of CDT was 52 and 5%, respectively. We conclude that CDT is a better marker for patients with chronic alcohol misuse than as a marker for high actual alcohol consumption alone. INTRODUCTION Carbohydrate-deficient transferrin (CDT) is an 'abnormal' transferrin frequently used as a marker of excessive and chronic alcohol consumption. Stibler (1991), when introducing the test, suggested that serum CDT levels will increase after excessive drinking (more than 50-80 g per day) for a period of 2-3 weeks. However, frequency and intensity of alcohol consumption have also been suggested to influence CDT (Allen et al., 1994), and it is also known that the clinical utility of the CDT measurements is altered according to the population examined. Thus, the sensitivity of CDT in detecting alcohol abuse is lower among patients in general clinical settings than among patients admitted to detoxification treatments (Allen et al., 1994). Low sensitivities have been reported in general population studies (Nilssen et al., 1992; Gronbaek et al., 1995). In young, healthy subjects drinking voluntarily for a short time excessive amounts of alcohol, no or low- *Author to whom correspondence should be addressed. frequent increases in CDT were found (Salmela et al., 1994; Lesch et al, 1996). However, CDT is frequently and rapidly elevated in chronic alcoholics relapsing after a period of not drinking (Schmidt et al., 1997), indicating that alcoholdependent subjects might be especially susceptible to effects from re-exposure to ethanol resulting in elevated CDT (Borg et al., 1994; Rosman et al., 1995). Lesch et al. (1996) stated that CDT is a marker for chronic alcohol abuse with high sensitivity and specificity among alcohol-dependent subjects. They also reported that there was no correlation between blood-alcohol concentration and elevations of CDT. Niemela et al. (1995) described higher CDT levels in early phases of alcoholic liver diseases and Yamauchi et al. (1993) concluded that CDT is a useful marker of noncirrhotic alcoholic liver diseases. These studies suggest that other features beyond alcohol consumption alone, such as long-term alcohol abuse and chronicity of the alcohol problem have effects on CDT levels, as recently reported by Saini et al. (1997). In two previous reports, we have evaluated CDT 646 © 1998 Medical Council on Alcoholism (Received 5 March 1998; in revised form 24 April 1998; accepted 1 June 1998) CDT AS A MARKER FOR ALCOHOL INTAKE OR CHRONIC MISUSE? PATIENTS AND METHODS Patients The patients were hospitalized in a surgical department at the University Hospital of Liibeck and have been described in detail in our previous studies (Huseby et al., \991a,b). In the present study, we excluded those patients who had not reported (or did not want to report) their recent alcohol consumption (meaning daily consumption over the previous 2 weeks), and those patients from whose serum samples we now lacked sufficient amounts for renewed %CDT analysis. The patients consisted of 161 otherwise unselected subjects admitted to an acute surgical ward. Among these, 51 were diagnosed as alcoholdependent according to SCAN interviews (WHO, 1992a) (section 11: Abuse) complying with both ICD-10 (WHO 19926) and DSM-III-R (American Psychiatric Association, 1987) criteria. The remaining 110 patients did not meet these criteria and were classified as non-dependent. The patients were also interviewed at admittance about their alcohol consumption for the previous 2 weeks (DOSE: mean consumed g of pure alcohol/day). The alcohol-dependent patients had > 10-year history of drinking and had had at least one detoxification treatment period. All non-dependent patients, including the high consumers, had < 10-year history of drinking and had not been subject to detoxification. The alcohol consumption for the whole patient group ranged between 0-800 g/day, median value: 36 g/day. The alcohol consumption for the alcohol-dependent patients ranged between 0-800 g/day with a median DOSE value of 63 g/day. Among the non-dependent patients, the alcohol consumption ranged between 0-162 g/day, median DOSE was 26 g/day. Most patients in this group (91 of the 110) consumed quantities of ethanol of < 60 g/day. The remaining 19 non-dependent patients reported a higher consumption (60-162 g alcohol/day), with mean and median values of 92 and 78 g/day, respectively. A comparable group (n = 21) was selected from among the alcohol-dependent patients (consuming >60 g/day but less than 170 g/day); their alcohol intake mean and median values were 92 and 84 g/day, respectively. The alcohol-dependent patients from the surgical ward had a median age of 45.0 years, and the non-dependent patients had a median age of 43.5 years. All patients voluntarily joined the programme, which complied with the Helsinki declaration of 1975. Blood sampling and analysis A venous blood sample was drawn from the patients at inclusion and serum was kept frozen at —70°C until analysed. CDT measurement was performed by the AXIS %CDT Turbidimetric Immunoassay (TIA) kit from AXIS Biochemicals ASA (Oslo, Norway). This kit separates the transferrin isoforms by micro columns and quantifies CDT by turbidimetry. The analytical procedure was performed as described by the manufacturer. The cut-off value was selected as 6% according to the manufacturer's ins (...truncated)