Krystal 2001: the VA trial where naltrexone didn't beat placebo

What they tested

Veterans with long-standing, severe alcohol dependence were randomly assigned to 12 months of naltrexone 50 mg once a day, 3 months of naltrexone followed by 9 months of placebo, or 12 months of placebo. There were 209 people in each group, and neither they nor the staff knew who was taking what.

Everyone was offered individual counselling and programs to help them take their pills, and was encouraged to attend Alcoholics Anonymous.

What they found

At 13 weeks, the two naltrexone groups went an average of 72.3 days before relapse and the placebo group 62.4 days. That difference was not significant (95 percent confidence interval for the difference, -3.0 to 22.8).

At 52 weeks, the three groups did not differ significantly in the percentage of days with any drinking or in the number of drinks per drinking day.

The authors concluded their findings do not support naltrexone for men with chronic, severe alcohol dependence.

Limits

• Almost everyone was a man and a veteran with long-standing, severe dependence, so the result may not apply to women or to people with milder drinking problems.
• Everyone had stopped drinking before starting, so it doesn't speak to taking naltrexone while still drinking.

What it means if you're considering naltrexone

This large, year-long negative trial shows naltrexone doesn't help everyone. Even with support for pill-taking, it found no clear benefit in this group.

Pooled analyses of many trials still find naltrexone helps on average, so this trial is best read as a warning that the benefit may be small or absent for some people, particularly men with very long-standing, severe dependence.