People who have done the Sinclair Method for a year or more often describe the same odd feeling. They can still remember drinking perfectly well. What has gone is the pull. “Friday night, wine” used to arrive on its own, and now the thought comes and nothing answers it. A 2026 review asks whether that change is only extinction, or whether naltrexone also changes the stored reward memory itself. This article is part of our Naltrexone hub.
The short version: the new idea is a hypothesis, the evidence behind it is thin but interesting, and it changes nothing about how you should take the medication. The longer version is below, with the sources, so you can judge it for yourself.
# The paper, in brief
In July 2026 Frontiers in Neuroscience published a mini-review by Iiro Jääskeläinen of Aalto University in Finland: Opioid antagonist treatment in alcohol use disorder: extinction, reward memory, and the possible importance of timing. Jääskeläinen worked with John David Sinclair, the researcher behind the method that carries his name, and the paper is dedicated to Sinclair’s memory.
It is a review, not a trial. It collects what is already known and makes one argument: the usual explanations for naltrexone may not cover everything we see, and the timing of the tablet relative to drinking may matter for a reason nobody has tested properly.
The review makes three points worth separating:
- Naltrexone and its relative nalmefene are among the best-supported medications for alcohol use disorder, but the average effect is modest and the response varies a great deal between people.
- Extinction, the mechanism the Sinclair Method is built on, explains a lot, and possibly not all of it.
- A second process, updating the alcohol reward memory while it is active, may add to it. This part is labeled as speculation by the author.
# What extinction explains, and where it runs out
Our Sinclair Method guide covers extinction in detail. In one paragraph: when you drink, your brain releases endorphins, its own opioids, and that reward teaches you to repeat the behavior. Naltrexone blocks the receptors the endorphins act on. Drink with the receptors blocked and the expected reward does not arrive. Repeat that enough times and the learned link between alcohol and reward weakens, the same way a dog stops salivating at a bell that no longer brings food.
Sinclair first set this out in 1990, and a Finnish trial led by Heinälä in 2001 supported it: naltrexone taken in a targeted way, with people still drinking, reduced relapse to heavy drinking. This is the core of why the Sinclair Method tells you to keep drinking at roughly your normal pattern while on the tablet. Extinction needs the drinking to happen under the blockade. Abstinence on naltrexone gives the brain nothing to unlearn from.
Classic extinction has a known weakness, though. In the lab, an extinguished response is not erased. A new “this no longer pays” memory is laid down on top of the old one, and the old one can come back: with time (spontaneous recovery), in a new setting (renewal), or after one strong reminder (reinstatement). That fragility is one reason researchers have asked whether something more lasting could also be involved, and it is where the review looks.
# Reward memory and reconsolidation, in plain terms
For a long time memory was thought to set like concrete: once stored, fixed. Work led by Karim Nader in 2000 (published in Nature) showed something different in rats. When a stored memory is brought back to mind, it becomes unstable for a short window and has to be stored again. That re-storing is called reconsolidation. If something interferes during the window, the memory can come back weaker or changed.
Researchers have since shown the same kind of window for drug-related memories. In 2008 Milton and colleagues found that blocking a specific brain receptor at the right moment could disrupt the reconsolidation of a drug-associated memory in rats, and that the same drug given too late did nothing. Timing was the whole effect.
The review’s proposal joins these two lines of work:
- Drinking itself is a strong reminder. It could bring the stored “alcohol means reward” memory back into that unstable window.
- If naltrexone is blocking opioid receptors during the window, the memory may be stored again with less reward attached.
- The result would be a memory that has been rewritten, rather than one that has only been covered over by a new one.
That would fit the description long-term users give: the facts of drinking are still there, and the value attached to them has faded. The fit makes it attractive. It does not make it true, and the review says as much.
# The rat study behind the timing question
The review’s most striking claim comes from Finnish rat work in the early 1990s using naloxone, a short-acting relative of naltrexone:
- Naloxone given well before alcohol access, so the blockade had worn off by the time the rats drank, did not produce lasting reductions in drinking.
- Naloxone given just before drinking, so the drinking happened under the blockade, reduced later drinking. That is extinction, as expected.
- Naloxone given immediately after drinking also seemed to reduce later drinking.
The third result is the one extinction cannot explain easily, because the drinking itself happened without any blockade. A process acting after the reward, on how the experience is stored, would explain it.
Before anyone builds on that, look at what the evidence is. The original work is a 1993 master’s thesis written in Finnish by Jääskeläinen and a 1995 conference abstract by Sinclair and Jääskeläinen. Neither is a full peer-reviewed paper. It was in rats, with a different drug, and it has not been repeated in people. The review itself says that whether this timing effect matters in human treatment “remains unknown”. It is a good question for researchers, and far too little to change anyone’s dosing.
# What this does not change
This is the part that matters if you take naltrexone.
Keep the one-hour rule. Take the tablet about an hour before your first drink, as your prescriber told you. Our dosing guide covers the window and the forgiving edges of it. Nothing in the review argues against it. Both extinction and the reconsolidation idea need the receptors blocked while alcohol is acting and in the period straight after, and a tablet taken an hour ahead covers both.
Do not switch to taking it after drinking. The post-drinking result used naloxone, which acts within minutes. Oral naltrexone takes roughly an hour to reach its peak. A tablet swallowed at the end of a session would arrive after the moment the rat studies were about, and there is no human evidence for doing this at all.
Do not drink more to speed it up. The method works on your normal drinking pattern. Adding extra sessions to “speed up the rewriting” has no support anywhere, and it adds risk.
Talk to your prescriber before any change. Naltrexone is a prescription medicine. Timing, dose and whether it suits you are clinical decisions.
The review makes one forward-looking suggestion: a fast-acting, short-lasting opioid blocker taken at the start of drinking, or just after, might target the reward-memory process more precisely than today’s long-acting tablets. No such product exists for alcohol use disorder. Treat it as a direction for future research, nothing more.
# Four processes that may run side by side
The review’s most useful framing is that naltrexone probably does not work through one mechanism. It suggests several may operate together, in different proportions for different people:
| Process | What changes | When it would act | How strong the evidence is |
|---|---|---|---|
| Less reward from alcohol | Drinking feels flatter while the drug is active | During the session | Strong: seen in lab and clinical studies |
| Less craving to cues | The pull from a pub, a time of day or a mood is weaker | Before and during drinking | Moderate: human cue-exposure studies, with mixed results (Monti, 2001) |
| Extinction | The learned “alcohol pays off” link weakens over many sessions | Across months | Supported by animal work and targeted-dosing trials |
| Reward-memory updating | The stored memory itself is saved again with less value | Just after the memory is recalled | Hypothesis: early animal hints only |
Seen this way, the Sinclair Method’s long timeline makes sense. The first process is immediate. The second shows up within weeks. Extinction takes months of repeated sessions. If the fourth is real, it would add to the third rather than replace it.
# Why naltrexone works so well for some people and not others
The review is also candid about the part advocates tend to skip. The largest recent analysis, McPheeters and colleagues in JAMA (2023), pooled 118 trials with nearly 21,000 people. Oral naltrexone at 50mg and acamprosate came out as the two best-supported medications. The average benefit was real and modest, and some people respond strongly while others notice little.
Several explanations have been put forward:
- Genes. A variant in the opioid receptor gene, OPRM1, was linked to better naltrexone response in a 2003 study by Oslin and colleagues (Neuropsychopharmacology). A later trial by the same group, which chose participants by genotype in advance, did not confirm it. The question is still open.
- How the drug is used. Daily dosing aimed at abstinence and targeted dosing aimed at extinction are different treatments that share a tablet. Trials that mix them, or that ask people not to drink, may blur the effect the Sinclair Method depends on.
- What else is happening. An early trial by O’Malley in 1992 found naltrexone worked better alongside coping-skills therapy when people did drink.
- Different mechanisms in different people. If four processes are in play, people will differ in which ones do most of the work.
None of this is a reason to avoid naltrexone. It is a reason to judge it on your own results over months rather than on anyone else’s story, including the enthusiastic ones.
# What “reaching extinction” tends to feel like
People describe the late stage of the Sinclair Method in similar terms, and the descriptions line up with the reward-memory idea closely enough to explain why this review spread so fast in TSM communities.
The memory stays, the pull goes. People can still recall what a good glass of wine tasted like. What is missing is the automatic “I want that now”.
Old triggers stop firing. The end of the working week, a bad day, a holiday, a celebration: situations that used to bring the thought of a drink without effort now pass without one. Some people test it on purpose and find nothing there.
It feels like something you never learned. A common comparison is cigarettes for a lifelong non-smoker: you know people enjoy them, and the idea does nothing for you.
It arrives slowly. Almost nobody describes a sudden switch. It shows up in hindsight, usually after month four or later, which is why the month-by-month timeline asks for patience.
These are reported experiences, not measurements, and they come mostly from people for whom the method worked. They are consistent with extinction alone as well as with memory updating. Telling the two apart would need controlled human studies, which is exactly what the review calls for.
# Tracking the change for yourself
Whichever mechanism does the work, the signs are the same things you can count: fewer drinks per session first, then fewer sessions, then days where drinking never came up. Our guide to tracking your Sinclair Method progress walks through what to look at at months one, three and six.
One addition this research suggests: keep a short list of your old automatic triggers, on paper or in your phone’s notes. Friday at six, a work win, an argument with someone, a particular pub. Every month or so, look at the list and ask which ones still pull. It turns the “nothing there anymore” feeling into something you can see changing, and it gives your prescriber something concrete at your next appointment.
# How AlcoLog helps you see the pattern
AlcoLog logs each drink and groups them into sessions, and its Medications card timestamps every naltrexone dose next to that drink log. That puts the dose-to-first-drink gap in view for every session, which is the timing both extinction and the reward-memory idea rely on. Over months, the Trends graph shows drinks per session falling and the calendar shows the days you did not drink.
Your data stays on your device. There is no account, and AlcoScore deliberately leaves medication out, so the medication log informs you rather than grading you.
