Rösner 2010 (Cochrane): naltrexone and nalmefene for alcohol

What they tested

A Cochrane review, which pools the results of many trials using strict rules, gathered every double-blind randomized trial comparing naltrexone or nalmefene (a related drug) with a placebo (dummy pill) or another treatment on drinking outcomes. Searches ran through January 2010, and the authors asked drug makers and researchers about unpublished trials.

Two authors pulled out the data independently, and trial quality was checked by two authors.

What they found

Naltrexone lowered the risk of heavy drinking to 83% of the placebo risk (RR 0.83, 95% CI 0.76 to 0.90) and reduced drinking days by about 4% (MD -3.89).

It also improved heavy drinking days (MD -3.25), amount of alcohol consumed (MD -10.83) and gamma-glutamyltransferase, a liver blood test linked to drinking (MD -10.37). Its effect on return to any drinking (RR 0.96, 95% CI 0.92 to 1.00) missed statistical significance.

The main side effects were stomach problems such as nausea (RD 0.10) and sedation such as daytime sleepiness (RD 0.09).

Based on few studies, the effects of injectable naltrexone and of nalmefene missed statistical significance.

Industry-funded trials (RR 0.90, 95% CI 0.78 to 1.05) did not differ significantly from non-profit trials (RR 0.84), and a statistical test didn't find signs of publication bias (P = 0.765).

Limits

• The searches ended in January 2010, so later trials aren't included.
• Results for injectable naltrexone and nalmefene rest on a limited number of studies.
• The authors described the treatment effects as moderate in size.

What it means if you're considering naltrexone

This is one of the largest pooled analyses of naltrexone for alcohol. Its clearest effect is on heavy drinking rather than on keeping people from drinking at all, which matches the site's guides: naltrexone leans toward reducing heavy drinking, while acamprosate leans toward abstinence.

The authors called naltrexone effective and safe, and argued that moderate effects should be weighed against how often alcohol problems relapse and how few treatments exist.