Volpicelli 1992: the first trial of naltrexone for alcohol

What they tested

Seventy men who had just finished alcohol detox were randomly given naltrexone 50 mg a day or a placebo for 12 weeks. Neither they nor the staff knew who was taking which.

Everyone also kept going to outpatient treatment, so naltrexone was tested as an add-on to counselling, not on its own.

What they found

People on naltrexone reported less craving and fewer days with any alcohol than people on the placebo.

The biggest gap was after someone had a drink. Of the 20 placebo patients who drank, 19 (95%) went on to relapse. Of the 16 naltrexone patients who drank, 8 (50%) did.

Naltrexone wasn't linked to changes in mood or other psychiatric symptoms. Two people on it had significant nausea, and one said their arthritis pain got worse.

Limits

• It was small, and everyone in it was a man, so it can't say much about women or about how often rarer side effects happen.
• It lasted 12 weeks, which says nothing about what happens after people stop.
• Everyone was also in a treatment program, so the results apply to naltrexone plus counselling.

What it means if you're considering naltrexone

This trial and O'Malley's, both published in 1992, were the first placebo-controlled tests of naltrexone for alcohol. The FDA approved it for alcohol dependence two years later. The main finding, that naltrexone does most of its work by stopping one drink from turning into a relapse, is the idea later trials and the Sinclair Method build on.