Evidence-Based Reviews

Pharmacotherapy of alcohol dependence: How and when to use disulfiram and naltrexone

Author and Disclosure Information



Fuller and Roth10 addressed this concern by designing a double-blind study that consisted of 3 groups. Group 1 received disulfiram 250 mg/d, group 2 received 1 mg/d (a pharmacologically ineffective dose), and group 3 received a placebo. Importantly, patients in both groups 1 and 2 were informed that they were receiving disulfiram and were warned of a possible ethanol-disulfiram reaction. At a 6-month follow-up, patients receiving disulfiram showed a small but significant increase in abstinence compared with those receiving placebo. Interestingly, there was no difference between patients receiving 250 mg of disulfiram and those receiving the ineffective 1 mg dose, suggesting that the primary action of disulfiram is the fear of a reaction with alcohol, not the agent’s pharmacologic effects.

A number of studies have shown that patients who agree to take disulfiram and continue taking it are generally highly motivated; such patients unsurprisingly experience better treatment outcomes.

In the largest controlled, blinded study of disulfiram performed to date, Fuller et al11 evaluated disulfiram treatment in 605 male veterans randomly assigned to disulfiram 250 mg a day, disulfiram 1 mg a day, or placebo. No significant differences were found between groups in total abstinence or time to first drink. Among patients who participated in all assessments and who drank at least once during the study, however, those receiving 250 mg had fewer drinking days than those in either control group.

The authors concluded that disulfiram may help reduce drinking frequency after a return to drinking, but does not contribute to continuous abstinence or to delay in time to first drink. Notably, despite a medication compliance rate of only 20%, a significant relationship existed between medication compliance and complete abstinence, regardless of treatment group. O’Farrell and colleagues14 have studied ways to increase compliance with disulfiram, developing the concept of the “Antabuse contract,” in which the patient takes disulfiram in front of a significant other as part of a couple’s therapy program.

How long should a patient continue taking disulfiram? Unfortunately, the ideal length of disulfiram treatment has not been established. While most randomized trials only administer disulfiram for a few months, research by Ojehagen et al15 has shown that long-term treatment (greater than 12 months) with disulfiram is significantly related to positive drinking outcomes during the 2 years following treatment.

Naltrexone The two initial landmark studies of naltrexone for alcohol dependence were published nearly a decade ago. Volpicelli et al7 conducted a double-blind, placebo-controlled trial of naltrexone 50 mg/d for 12 weeks with 70 male veterans who also received intensive psychosocial alcohol rehabilitation. Study participants who received naltrexone had fewer drinking days, less craving for alcohol, and a lower rate of full-blown relapse than did patients who received placebo. The major effect of naltrexone occurred among patients who sampled alcohol; only half of those on naltrexone progressed from first drink to a full-blown relapse, compared to a 95% rate of relapse among those who initiated drinking while receiving placebo.

O’Malley et al8 also found naltrexone to be more effective than placebo and found an interesting interaction with psychosocial treatment. Individuals who received weekly abstinence-oriented supportive therapy were more likely to be continuously abstinent from alcohol at the end of the 12-week study period. Those who received naltrexone along with cognitive behavioral coping skills therapy were least likely to progress to a full-blown relapse if they did drink.

More recently, Anton et al12 conducted a double-blind, placebo-controlled trial of naltrexone in 131 patients, all of whom received cognitive behavioral therapy (CBT). Those receiving naltrexone had a significantly longer period prior to relapse, fewer drinking days, and fewer drinks per drinking day than the placebo group. No significant differences were reported between the groups in time prior to the first drink.

These studies suggest that naltrexone may diminish the likelihood of progression from first drink to full-blown relapse. This may occur through the agent’s attenuation of the reinforcing effects of alcohol.13 While naltrexone does not fully block alcohol the way it blocks opioid drugs, the reduction in alcohol’s positive effects may help naltrexone responders to contain their “slips” and prevent progression to a full-blown relapse.

The difference in adverse effects

Disulfiram Adverse effects (see Table 2) arise from 3 main causes:

  1. Medical complications during an ethanol-disulfiram reaction;
  2. Toxicity due to disulfiram or its metabolites;
  3. Interactions between disulfiram and other medications.

Table 2


Mechanism of actionInterrupts metabolism of alcohol, leads to a buildup of acetaldehydeOpiate antagonist, may attenuate reinforcing property of alcohol
Recommended dose*25-500 mg50 mg
Side effects and adverse eventsDrowsiness, impotence, headache, acne, metallic aftertaste, hepatitis, neuritis, ethanoldisulfiram reactionNausea, vomiting, headache, anxiety, dizziness, fatigue, insomnia, elevated liver enzymes
*Physicians Desk Reference 55th ed. Montvale, NJ: Medical Economics, 2001