Current issues of ACP Journal Club are published in Annals of Internal Medicine


Brief advice to stop drinking for adults with high GGT levels reduced alcohol consumption at 1 year

ACP J Club. 1992 Jan-Feb;116:6. doi:10.7326/ACPJC-1992-116-1-006

Source Citation

Nilssen O. The Tromsø study: identification of and a controlled intervention on a population of early-stage risk drinkers. Prev Med. 1991 May-Jun;20:518-28.



To examine the effect of 2 types of intervention on a population of early-stage risk drinkers.


Randomized controlled trial of 1-year duration.


Community health survey, Tromsø, Norway.


Male and female subjects, ages 20 to 62 years, were invited to participate. Participants completed questionnaires and a brief physical examination. Inclusion criteria were a γ-glutamyltransferase (GGT) level between 50 (45 for women) and 200 U/L, and consumption of beverage alcohol 2 to 3 times a week, or consumption corresponding to 1 bottle of wine on 1 occasion at least twice a month. Subjects were excluded if hospital records showed alcoholism, hepatobiliary disease, or major psychiatric illness, or if subjects were taking antiepileptic medication. 338 of 27 198 invitees were eligible, and 320 attended the 1-year follow-up.


Participants were randomized to a minor- or major-intervention or a control group. After receiving a mailed invitation for a second GGT, 99% of intervention subjects returned. Minor-intervention subjects were told the most common reasons for an elevated GGT value (including alcohol), asked to consider reasons for their elevated GGT, and given written advice about GGT and alcohol. Results of the second GGT were mailed with a suggestion to repeat the test at 1 year. Major-intervention subjects were told that alcohol elevated their GGT. Details of alcohol amounts consumed were recorded, blood was taken, and monthly consultations (with GGT) were offered. They were told how to reduce consumption and were given written advice.

Main outcome measures

GGT levels and alcohol consumption.

Main results

About 18 months after screening, mean GGT levels decreased by 22% and 16% to 61.1 and 58.8 U/L, for the minor- and major-intervention groups, respectively, compared with an increase in GGT of 9% to 86.5 U/L for the controls (P for difference between groups < 0.001). Self-reported drinking on a detailed questionnaire for the control group was more than double that of the 2 intervention groups (P < 0.001). In the major-intervention group, consumption decreased dramatically (33 to 13 g/d).


Heavy drinkers with elevated GGT levels reduced consumption when offered brief advice and a 1-year follow-up, a single session being as effective as the offer of monthly sessions.

Source of funding: Blue Cross Center for Treatment of Alcoholics, Håkøya, Tromsø, Norway.

Address for article reprint: Dr. O. Nilssen, Institute of Community Medicine, University of Tromsø, Postuttak, 9000 Tromsø, Norway.


The Tromsø study confirms growing evidence indicating that people reduce alcohol intake after brief interventions. Other variables such as hospital days and sick days have also improved in other randomized, controlled studies on different populations (in communities, and with hospitalized or office patients [1, 2]).

Physicians should briefly intervene with patients who have abnormal results on screening instruments such as the GGT or CAGE test or who drink too heavily (> 28 drinks weekly for men, 16 for women; 6 or more at one time, 2 to 3 times a month). The Tromsø study provides an intervention model that includes components of motivation, education, and support, and links health outcomes to diminished drinking. Interventions should include self-help materials and the opportunity for feedback. The GGT test is appropriate for office practice in the way that it was used in this study. The stigma attached to identification of drinking as a problem must be addressed. Unlike treatment for alcoholism, abstinence need not be an initial recommendation.

This study, taken with others, demonstrates that use of a simple short routine inquiry (e.g., Graham's "Life-style Risk Assessment" [3]) would find at-risk drinkers and that brief interventions would diminish drinking and prevent serious difficulties. This work is reimbursable, and physicians can easily learn the requisite skills. I expect doctors will accept the challenge presented by recent publications.

William D. Clark, MD
Addiction Resource CenterBath, Maine, USA

William D. Clark, MD
Addiction Resource Center
Bath, Maine, USA


1. Barbor T, Grant M. From clinical research to secondary prevention. Alcohol Health Res World. 1989;13:371-4.

2. Wallace P, Cutler S, Haines A. Randomized controlled trial of general practitioner intervention in patients with excessive alcohol consumption. BMJ. 1988;297:663-9.

3. Graham A. Screening for alcoholism by life-style risk assessment in a community hospital. Arch Intern Med. 1991;151:958-64.