Naltrexone is an opioid antagonist that’s available most commonly as a 50 mg oral tablet, and also as Vivitrol, a long-acting injectable formulation. It is approved by the Food and Drug Administration as a treatment for opioid use disorder. That use, which we won’t address here, is fraught with controversy. But naltrexone was also approved by the FDA as a treatment for alcohol use disorder in 1994.
Since then, for people who want to quit drinking, doctors in the United States have generally prescribed it as an anti-craving medicine and told patients to abstain while taking it.
However, it is unclear whether naltrexone is effective if you follow those instructions. Trials by John David Sinclair in Finland, Henry Kranzler in the US and others indicate that naltrexone is most effective in reducing alcohol consumption—if you drink alcohol about an hour after taking it. This is the Sinclair Method, also known as “targeted naltrexone.”
So how well does it work?
Its model views addiction as a conditioned response to dopamine being released into the brain’s reward pathway, in this case caused by an increase in beta-endorphin which results from drinking alcohol. The Sinclair Method seeks to block the effect of the beta-endorphin when alcohol is consumed, which, in turn, blocks the release of dopamine, which leads to the “extinction” of the conditioned response over a period of a few months.
So how well does it work?
The largest trials of the Sinclair Method were conducted between 2008 and 2011, using a close relative of naltrexone called nalmefene. Named ESENSE 1, ESENSE 2 and SENSE, they resulted in nalmefene being approved in Europe as a medication for the reduction of alcohol consumption.
ESENSE 1 (Mann et al., 2013; NCT00811720) was conducted in Austria, Finland, Germany, and Sweden at 39 sites with 604 subjects: 306 in the nalmefene (18 mg) group and 298 in the placebo group. It lasted six months. All subjects met DSM-IV-TR criteria for “alcohol dependence,” as it was then termed, ranging from mild to moderate to severe. All also received the BRENDA psychosocial treatment to encourage medication compliance and behavioral change.
Patients were instructed to take their medication (nalmefene or placebo) one-to-two hours prior to a high-risk drinking situation.
Between the month before initial screening and the last month of the study, the number of heavy-drinking days (60 grams of ethanol or more for men, 40 or more for women) in the nalmefene group fell from an average of 19.4 days to 8.2 per month; in the placebo group it went from 19.6 to 10.7 per month. The average amount drunk per day in the nalmefene group went from 84 grams of ethanol to 33; in the placebo group it went from 85 grams to 45. Overall, this showed significantly greater reduction in drinking in the nalmefene group.
Dropout rates were high, however, at 31 percent of subjects in the placebo group and 53 percent in the nalmefene group—a significant difference.
Notably, there were large reductions in both groups between initial screening and the beginning of treatment, one-to-two weeks later; 18 percent reduced their drinking to less than six heavy-drinking days per month in this brief period. It’s actually a common phenomenon: Once people decide to initiate treatment their habits tend to change even before it starts. Similar phenomena were noted in Project MATCH.
ESENSE 2 (Gual et al., 2013; NCT00812461) was conducted in Belgium, the Czech Republic, France, Italy, Poland, Portugal and Spain at 57 sites with 718 subjects: 358 in the nalmefene (18 mg) group and 360 in the placebo group. All other conditions were the same as ESENSE 1.
The number of heavy-drinking days in the nalmefene group went from 19.8 to 6.6 per month, and in the placebo group from 18.3 to 7.5—still a statistically significant difference between groups.
With average amount drunk per day—which went from 93 grams of ethanol to 30 in the nalmefene group, and from 89 to 33 in the placebo group—the difference was not statistically significant. However, even more participants in ESENSE 2 reduced their drinking between screening and treatment initiation; 33 percent had reduced heavy-drinking days to less than six per month. When these subjects were removed from calculations, nalmefene did have a significant effect on reducing average daily consumption.
Dropout rates were 41 percent in the nalmefene group and 38 percent in the placebo group—a much smaller difference than in ESENSE 1.
Overall, it was a significant—but far from dramatic—difference in favor of nalmefene.
The SENSE study (van den Brink et al., 2014; NCT00811941) lasted 12 months, twice as long as the two ESENSE studies. Conducted at 60 sites in the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Russia, Slovakia, Ukraine and the United Kingdom, it had 675 subjects with alcohol dependence: 509 in the nalmefene group and 166 in the placebo group. All other conditions were the same as the ESENSE studies.
At six-month followup, there were no significant differences between the nalmefene and placebo groups in terms of heavy-drinking days or total consumption.
However, at 13-month followup, the nalmefene group showed better outcomes. It had gone from 15.2 heavy-drinking days per month to 4.7, a reduction of 10.5. The placebo group had gone from 14.7 heavy drinking days per month to 5.7, a reduction of 9.0. Average ethanol consumption in the nalmefene group had gone from 75 grams per day to 22.2, a reduction of 52.8 grams. The placebo group had gone from 75 grams to 28.7, a reduction of 46.3.
Overall, it was a significant—but far from dramatic—difference in favor of nalmefene.
This time, 39 percent of participants had reduced their drinking to less than six heavy-drinking days per month between screening and treatment initiation. When these patients were removed from calculations, the effect of nalmefene over placebo was also significant at the six-month followup on both measures.
Dropout rates were 38 percent in the nalmefene group and 32 percent in the placebo group—a significant difference once again. Some patients dropped out due to nalmefene’s side effects—similar in profile to those of generic naltrexone, and which some subjects in the placebo group also perceived. But the most common reason was withdrawal of consent.
In any case, those three major trials were enough to convince the European Medicines Agency, in 2012, to approve nalmefene for the reduction of alcohol consumption in adults with alcohol dependence.
However, six members of the agency objected to the approval. And a number of scholars, including Fitzgerald et al. (2016), Naudet et al. (2016), and Palpacuer et al. (2015), published papers criticizing the decision.
For one thing, generic naltrexone, which was much cheaper than nalmefene, had long been approved in Europe for supporting abstinence—but was also already being used off-label to reduce drinking. A lot of people viewed nalmefene as a money-making scheme for the pharmaceutical company which manufactured it.
Another charge was that subjects who improved greatly prior to treatment had been removed after the fact in the data reanalysis, to inflate the efficacy numbers.
Finally, the effect size when compared to placebo was very small.
The small effect size in some studies could be due to hit-and-miss medication compliance.
It is worth noting here that medication compliance was not reported in any of the three trials. The role of medication compliance has generally been badly overlooked in studies of targeted naltrexone. And it is possible that the Sinclair Method’s efficacy has been underrepresented as a result.
Advocates of the method say that total compliance—always taking the pill one hour before drinking—is essential to obtain and maintain the extinction effect. So the small effect size in some studies of targeted opioid antagonists for alcohol use disorder could be due to hit-and-miss medication compliance.
Research priorities in the US have sometimes reflected how abstinence is culturally prized here much more than in Europe.
A 2016 study by Amy W. Helstrom et al. attempted to test whether naltrexone plus abstinence reduced craving for alcohol. Subjects were 100 patients and 100 matched controls undergoing residential treatment at the Hazelden rehab—where subjects had no access to alcohol.
This study was not blinded, not placebo-controlled, and not randomized. The patients themselves chose whether they wanted to be in the naltrexone group or the control group. Those given naltrexone all knew they were taking it, and that it was intended to reduce their cravings.
A randomized placebo-controlled trial is needed to determine whether there is a medication effect when naltrexone is taken while abstaining.
Craving was rated using the Penn Alcohol Craving Scale, from 0 to 6. At the beginning of the trial, the average craving rating in the naltrexone group was 2, compared to 1.1 in the control group—a significant difference. By the end, the figure for both groups had fallen below 1, and was not significantly different between groups.
However, it is currently impossible to know whether this drop in craving was due to the effect of naltrexone, to a placebo effect, or simply to the effect of time. A randomized placebo-controlled trial is needed to determine whether there is actually a medication effect when naltrexone is taken while abstaining. Still, this study showed that naltrexone plus abstinence does not lead to increased craving, as author Roy Eskapa had claimed it would.
Some people say they have experienced large reductions in craving by taking naltrexone while abstaining. But it is impossible to tell if this is simply a strong placebo effect, or if naltrexone causes a large craving reduction in some individuals. More and better research is needed on this topic.
We know that the craving reduction in people who drink on naltrexone is due, at least mainly, to the extinction effect. If naltrexone actually does reduce craving in abstainers, what could be the mechanism?
One hypothesis is that alcohol-related cues cause a release of beta-endorphin, which in turn releases dopamine. This theory proposes that naltrexone blocks the reaction of the opioid system to alcohol-related cues. A 2014 brain-imaging study by Mann et al. seems to confirm that naltrexone blocks cravings induced by alcohol-related cues.
There is evidence of a great deal of individual variation in response to naltrexone as a treatment for alcohol use disorder.
A 2019 brain-imaging study by Amanda Elton et al. found that naltrexone enhances the connectivity between the ventromedial prefrontal cortex (an area of the brain involved in craving) and the left frontoparietal network (an area of the brain involved in executive control). This increased connectivity appears to contribute to a reduction of craving for alcohol caused by naltrexone, according to the authors, a process which involves the kappa-opioid receptor system.
There is evidence of a great deal of individual variation in response to naltrexone as a treatment for alcohol use disorder. A 2014 paper by James C. Garbutt et al. reviewed possible factors.
One factor is genetic. A variant in the gene which codes the development of mu-opioid receptors appears significant. People who have the Asn40Asp variant of the gene respond better to naltrexone treatment for alcohol use disorder than those who don’t. This variant both increases mu-opioid receptor binding affinity and decreases the number of these receptors.
People with a family history of alcohol use disorder also appear to respond better to naltrexone treatment for alcohol than those with no such history. People with high levels of alcohol craving prior to treatment also appear to respond better. And out of left field, people who like sweets also appear to do better with naltrexone than those who dislike them.
These potential influences on efficacy are not always confirmed by other studies, however.
A 2019 study by Katie Witkiewitz et al. found that the Sinclair Method has different effects on people who drink for pleasure (high-reward drinkers) and those who drink to avoid pain (high-relief drinkers). These researchers divided drinkers into four categories: high-reward/high-relief, high-reward/low-relief, low-reward/high-relief, and low-reward/low-relief. They found that high-reward/low-relief drinkers showed the greatest response to naltrexone for number of heavy-drinking days and drinks per drinking day.
Only you can choose if naltrexone is right for you.
Nothing here is medical advice, and seeking medical advice is advisable. But evidence suggests that if you are a daily drinker, it does not really matter if you take naltrexone daily, an hour before drinking, or get the Vivitrol shot—the opioid receptors are still blocked either way, and the extinction effect still takes place. Fifty mg of naltrexone gives a near-total blockade of the opioid receptors for the first 24 hours, and it takes about four days for the blockade to be reduced by half.
If, on the other hand, you tend to spread out your drinking days, abstaining in-between, then taking naltrexone one hour before drinking, i.e., targeted naltrexone—the Sinclair Method—may be the best option.
And if you tend to abstain for multiple days, then go on multiple-day benders drinking morning to night, you might find that Vivitrol removes medication compliance concerns.
Potential side-effects may also influence such decisions.
There is some solid evidence for naltrexone to treat alcohol use disorder. But the evidence is much stronger for people whose goal is to reduce, rather than quit, drinking. And even then, its efficacy may be only somewhat better than placebo, and individual responses vary greatly.
It could be worth a try. But only you can choose if naltrexone is right for you.
If you take it and don’t like the effect on your alcohol buzz, there are many other ways to reduce alcohol-related harm.
Then again, if you take it while abstaining and your cravings go way down as you want them to, it doesn’t really matter if it is a medication effect or a placebo effect. If the Sinclair Method helps you reduce your consumption to where you want it to be, that is a win.
The biggest obstacle may be your medical provider. Unfortunately, as Jonathan Avery, MD noted in the American Journal of Psychiatry in 2022, many physicians are unaware of the research behind naltrexone and its targeted use.
Photograph by eflon via Flickr/Creative Commons 2.0