How to Stop Drinking Alcohol

Most people who want to stop drinking have already tried. Of the 290+ clients we have worked with across 12 countries over 7 years, 98.4% had already tried and failed to stop before they came to us, and alcohol is one of the habits they bring most often. The attempts were sincere: a dry month, a rule about weeknights, a switch to beer only, a promise made the morning after. The published record shows how common it is for those attempts to fail. Across studies of people stopping alcohol and tobacco, the curve that describes relapse has the same shape: losses are steepest soon after stopping and slow down as sober time builds up¹. In a 16-year study of 461 people who first sought help for a drinking problem, those who got no help were less likely to reach three years of remission and, if they did, more likely to relapse later².

The usual explanation for those failures is a lack of willpower. That explanation is wrong in a specific and useful way. Willpower, in the sense of deliberate control by the prefrontal cortex, is real, but it is the faculty most easily taken offline by the conditions in which most drinking happens: stress, fatigue, poor sleep and the presence of something that has been linked to drinking before^3-5. Alcohol adds a problem of its own. Even at doses that don't impair thinking in general, alcohol weakens inhibitory control and makes the next drink more appealing⁶. So the decision not to drink is made by a rested, sober person in the morning and has to be carried out by a tired, stressed person in the evening, sometimes after a first drink has already lowered the brake. People are poor at predicting how different they will feel in that second state⁷.

Modern addiction research explains what is happening underneath. Koob and Volkow describe addiction as a cycle of three stages, each run by different brain systems: drinking and its reward, in which dopamine and the brain's own opioids teach the habit and turn it into routine; withdrawal and low mood, in which the reward system responds less and the brain's stress systems take a bigger role; and preoccupation, in which craving rises and the prefrontal cortex can't hold it back⁸. Alcohol has its own version of each stage. It changes the balance between the brain's main calming and exciting chemicals, which produces tolerance and withdrawal. It disrupts sleep. It becomes the way a hard day ends. And it is legal, social and everywhere, so the triggers for it are hard to avoid.

The rest of this paper takes the questions people ask about stopping drinking one at a time. Section 2 answers whether it is safe to stop suddenly. Sections 3 to 7 explain why drinking is hard to stop: what alcohol does to the reward and habit systems, why the urge comes at the end of the day, what it does to sleep, how triggers are learned and how long cravings last. Sections 8 and 9 deal with cutting down and with stopping when everyone around you drinks. Section 10 describes what happens to the body and brain after stopping, and Section 11 what to do after drinking again. Section 12 sets out what the Behavior Control System targets for someone stopping drinking, day by day, and Section 13 reports our results and the limits of this paper. Where the program is described, we say what each day targets and why. We do not publish how.

It depends on how much and how often you drink, and only a doctor can judge that for you. Withdrawal happens when a brain that has adapted to regular heavy drinking suddenly has no alcohol. Heavy drinkers who stop or cut down sharply can develop symptoms that range from tremors, irritability, anxiety and agitation to hallucinations, seizures and delirium tremens⁹. Symptoms can start within 6 to 24 hours of the last drink¹⁰. That is why the standing rule in this paper, and in our program, is simple: anyone who drinks heavily every day talks to a doctor before stopping.

2.1The brake and the accelerator

The brain balances two main signalling chemicals. GABA is the main inhibitory transmitter: it calms nerve cells down and acts as the brain's brake. Glutamate is the main excitatory transmitter: it acts as the accelerator. Alcohol works on both. It acts largely through the GABA-A receptor, increasing the brake's effect, which is a large part of why the first drink takes the edge off so quickly¹¹. It also damps down glutamate signalling. The result is the familiar feeling of the first drink: shoulders drop, the noise in the head goes quiet and the day stops pressing.

When alcohol is present most days, the brain adapts to keep working. Davies's review describes both short-term and long-term changes in GABA-A receptors with continued alcohol exposure¹¹, and the glutamate system shifts the other way. This adaptation is what tolerance feels like: the same amount of alcohol does less. It also means that when the alcohol is suddenly gone, the brain is left with its brake turned down and its accelerator turned up. Saitz describes withdrawal as the result of alcohol-induced imbalances in brain chemistry that cause excessive nerve activity once alcohol is withheld⁹. The imbalance has been measured directly. In a magnetic resonance spectroscopy study of 47 alcohol-dependent patients in detoxification and 57 healthy controls, glutamate levels in prefrontal regions were significantly raised during acute withdrawal, and the changes returned to normal within a few weeks of abstinence; the same pattern was found in alcohol-dependent rats¹².

2.2What withdrawal looks like, and when

Mirijello and colleagues' clinical review gives the time course that doctors work from¹⁰. Symptoms can develop within 6 to 24 hours after drinking stops or drops sharply. Many cases are mild and pass within two to seven days of the last drink. Up to half of patients with alcohol use disorder have withdrawal symptoms when they stop. Of patients who show withdrawal symptoms, almost one in ten develops withdrawal seizures, usually 24 to 48 hours after the last drink. Delirium tremens occurs in about 5% of patients with withdrawal, usually 48 to 72 hours after the last drink but sometimes up to ten days later, and it can be fatal. The review lists previous withdrawal episodes, previous withdrawal seizures or delirium tremens, severe dependence, older age and other illness among the factors that raise the risk.

Withdrawal can also get worse with each repetition. Becker's review describes a kindling process in alcohol withdrawal: in many people with alcohol dependence, withdrawal symptoms become more severe after repeated withdrawal episodes, and both clinical and experimental evidence support the mechanism¹³. Becker concludes that kindling may also add to relapse risk and to alcohol-related brain damage. For someone who has stopped and started many times on their own, this is a second reason to involve a doctor. The previous attempts may have made the next withdrawal harder, not easier.

2.3What a doctor can do

Withdrawal has effective medical treatment. In the evidence-based guideline by Mayo-Smith for the American Society of Addiction Medicine, benzodiazepines reduced the severity of withdrawal, reduced delirium by 4.9 cases per 100 patients and reduced seizures by 7.7 per 100 patients, with dosing best guided by withdrawal scales¹⁴. Mirijello and colleagues describe benzodiazepines as the gold-standard treatment, with severe cases sometimes needing intensive care¹⁰. Heavy drinking can also leave people short of thiamine. Thiamine deficiency causes Wernicke's encephalopathy, a serious brain disorder that is still often missed, and Sechi and Serra emphasize thiamine for its prevention and for preventing Korsakoff's syndrome in people who misuse alcohol¹⁵. Decisions about either belong to a doctor.

A doctor can also prescribe medication that helps people stay stopped. In a systematic review and meta-analysis of 122 randomized trials by Jonas and colleagues, acamprosate and oral naltrexone were both associated with fewer people returning to drinking: about 12 people needed to be treated with acamprosate, and about 20 with naltrexone, to prevent one return to any drinking, with no significant difference between the two when compared directly¹⁶. Imaging work suggests one reason naltrexone helps: drinking alcohol releases the brain's own opioids in the nucleus accumbens and orbitofrontal cortex, and naltrexone blocks the receptors they act on¹⁷. Schuckit's review of alcohol use disorders lists detoxification, cognitive-behavioral approaches to prevent relapse and the judicious use of medication to reduce craving among the established treatments¹⁸. Our program does not prescribe or recommend medication. It works alongside whatever medical care a person has, and it never replaces it.

Being in control at work and losing to the evening drink is not a contradiction. The two happen in different brain systems and under different conditions. At work, decisions are made by goal-directed control, supported by the prefrontal cortex, in daylight, with a clear head and a reason to stay sharp. Drinking happens later, when the prefrontal cortex is tired and stressed, in settings the brain has linked to alcohol many times, and it is driven by reward and habit systems that respond to those settings before any decision is made^{8, 19}. The person hasn't lost their discipline. They are using it in a place where it has little say.

3.1What a drink does to the reward system

Alcohol acts on the reward system directly. In a PET imaging study by Boileau and colleagues, six healthy volunteers drank either alcohol in orange juice or orange juice alone; after alcohol, the scans showed increased dopamine release in the ventral striatum, including the nucleus accumbens, and the size of the change tracked the stimulant effect of the drink on heart rate²⁰. Alcohol also releases the brain's own opioids. Mitchell and colleagues scanned heavy drinkers and controls before and after a drink and found opioid release in the nucleus accumbens and the orbitofrontal cortex, both regions involved in valuing rewards; in the heavy drinkers, the change in the orbitofrontal cortex correlated with problem drinking and with how high they felt¹⁷. A drink therefore delivers two rewards at once: the calming effect on GABA described in Section 2 and a dopamine and opioid signal in the reward system.

Dopamine's main job in this system is teaching. In the work that defined the modern view, dopamine neurons responded to the difference between the reward received and the reward expected, and with experience the response moved from the reward itself to the earliest reliable sign that it was coming²¹. For a drinker, the earliest signs are often the end of the working day, the walk into the kitchen, the sound of a bottle opening and the people they usually drink with. Those become the moments the reward system responds to. Berridge and Robinson add a distinction that explains why drinking can continue long after it stops being much fun. The brain's systems for wanting a reward are separate from its systems for liking it, and repeated use can strengthen wanting, especially wanting set off by associated triggers, without strengthening liking²².

3.2What years of drinking do to it

Over time, the reward system responds less to everything else. Volkow and colleagues gave 20 recently detoxified people with alcohol dependence and 20 controls a dose of a stimulant and measured the dopamine response with PET. In the ventral striatum, where the nucleus accumbens sits, and in the putamen, the dopamine increase was 70% and 50% lower than in controls; in controls but not in the alcohol-dependent group, activity in the orbitofrontal cortex was linked to the size of the dopamine response, which the authors suggest may underlie a reduced sensitivity to rewards²³. Most of the people in that study also smoked, and a stimulant is not a natural reward, so the finding is a measure of a blunted system rather than a direct measure of how a sunset or a good meal feels. It fits what heavy drinkers describe: ordinary life feels flatter, and the drink becomes one of the few things that still lifts the evening.

Koob and Volkow's review places this change in the second stage of their cycle. In the withdrawal and negative-mood stage, the dopamine part of the reward system works less well, while brain stress chemicals such as corticotropin-releasing factor and dynorphin are recruited in the extended amygdala⁸. The person drinks less to feel good and more to stop feeling bad, a shift Section 4 examines in detail.

3.3From a choice to a habit

Repetition also changes who is in charge of the behavior. Everitt and Robbins describe addiction as a shift from goal-directed control, which weighs an action by its expected outcome, to habitual control, which runs a practiced response directly from the setting, with prefrontal control weakening along the way¹⁹. Habit research describes the same process in ordinary life: behavior repeated in a stable context comes to be triggered by that context with little reference to current goals²⁴. In alcohol dependence this shift has been measured. Sjoerds and colleagues compared 31 abstinent alcohol-dependent patients with 19 matched controls on a learning task built to separate the two kinds of control. The patients relied more on stimulus-response habits, with less engagement of brain areas linked to goal-directed action, the ventromedial prefrontal cortex and anterior putamen, and more of an area linked to habit, the posterior putamen²⁵.

Stress pushes the balance further toward habit. In a laboratory study, people who were stressed before learning a task kept working for a food they had just eaten their fill of, while unstressed people stopped; stress made their behavior insensitive to what the outcome was now worth⁴. This is the experience of deciding at nine in the morning that tonight will be different and finding at seven in the evening that the decision has no grip. The morning decision came from goal-directed control. The evening is run by a habit triggered by the setting, after a day of stress, and the system that made the decision isn't the one in charge.

3.4Why the first drink matters so much

Alcohol weakens the brake that would stop the next drink. Field and colleagues' review shows that alcohol impairs inhibitory control at doses that don't impair thinking more generally, that this effect is linked to alcohol's ability to prime further alcohol seeking, and that alcohol also changes how strongly alcohol-related things grab attention⁶. A plan to have one drink is therefore carried out by a person whose control has been lowered by that drink. This is the plainest reason why rules about quantity fail so often for people who already struggle to stop, a point Section 8 returns to.

The way the brain values time adds to it. Delay discounting is the tendency to value a reward less the further away it is. Petry compared people actively drinking heavily, people with alcohol dependence who were abstinent and controls, and found the steepest discounting in the active drinkers, intermediate rates in the abstinent group and the shallowest in controls; alcohol itself was discounted more steeply than money²⁶. A meta-analysis confirmed greater discounting across addictive behaviors²⁷. Bickel and colleagues describe this short time horizon, together with an excessive value placed on one reward, as the core of what they call reinforcer pathology, and argue that lengthening the horizon and building other rewards are both routes out²⁸. Tonight's drink is immediate and certain. A better morning, a clearer month and a healthier year are far away. The program works on both sides of that comparison.

Understanding comes first in the program because it changes how the rest of the work is experienced. A client who knows that the evening urge is a learned response, produced by systems that change with sustained abstinence, can watch it happen instead of treating it as proof of weakness. One of our clients, Bjorn, a senior operator in the Netherlands who came to us for alcohol, put it this way: “Now I know how my brain works and what it needs. That's the difference.” The demanding new pursuit on Day 17 rests on animal research in which an enriched environment, with novelty, exercise, social contact and challenge, both prevented the development of drug addiction and reduced relapse-like behavior²⁹, and on the case for alternative rewards set out by Bickel and colleagues²⁸.

The evening drink usually starts as relief and turns into need. At first it takes the edge off a hard day. With repetition, the brain's stress systems adjust to it, the hours without alcohol start to feel more tense than they used to, and the drink is increasingly used to get back to normal rather than to feel good. The end of the day is when stress has built up most, when self-control is most tired and when the setting, the time and the routine all point toward a drink. That's why the urge arrives at the same time most days.

4.1Drinking for relief

Alcohol's effect on stress has been studied for decades. Sher and Levenson's laboratory work on what they called the stress-response-dampening effect of alcohol linked individual differences in that effect to risk for alcoholism³⁰. Cooper and colleagues later showed that people drink for distinct reasons, and that two of them matter most: drinking to cope, to escape or reduce negative feelings, and drinking to enhance, to increase positive ones. In two random samples, one of adults and one of adolescents, these motives were the closest predictors of drinking and drinking problems, with different causes and different consequences³¹. Both are common among our clients. The evening drink after a hard day is coping. The drink after a good one, a deal closed or a week finished, is enhancement, and celebration can be as strong a trigger as stress.

Anxiety and drinking feed each other. Kushner, Abrams and Borchardt's review concluded that anxiety disorders and alcohol problems can each set off the other, and proposed a cycle in which alcohol's short-term relief of anxiety, combined with the longer-term anxiety produced by heavy drinking and withdrawal, drives both upward³². Many drinkers recognize the second half of that cycle as the anxious, on-edge morning after.

4.2How relief turns into need

Koob and Le Moal proposed that addiction is best understood through allostasis: the brain keeps apparent stability by shifting its set point, and the processes that normally counter a drug's effect fail to return to their usual range, leaving the person in a lasting deviation from normal reward function³³. Koob later described how brain stress systems, including corticotropin-releasing factor in the extended amygdala, are recruited as dependence develops and come to drive use through negative reinforcement: the drink is repeated to relieve a negative state that drinking itself has deepened³⁴. Becker's review of alcohol dependence makes the same point in clinical terms: withdrawal-related anxiety, which reflects adaptive changes in the body's stress response system, is one factor contributing to relapse, and the fear of withdrawal can itself keep drinking going³⁵.

The low mood and anxiety of early sobriety have a known course. Heilig, Egli, Crabbe and Becker's review notes that the depressed mood and raised anxiety of early withdrawal occur in the vast majority of alcohol-dependent patients and in most cases ease over three to six weeks of abstinence, but that the risk of relapse lasts much longer³⁶. Animal work reviewed in the same paper shows that a long history of dependence leaves subtler changes that alter emotional processing and increase sensitivity to stress well into protracted abstinence, and emerging human data support the clinical relevance of that negative emotionality for relapse. Explaining this in advance changes how a hard evening in the fifth week is read. It is an expected stage with a known cause and a known direction, not a sign that stopping isn't working.

4.3Stress predicts drinking again

Stress is one of the most consistent predictors of relapse to alcohol. Sinha and colleagues studied treatment-engaged, alcohol-dependent patients after a month of abstinence. Compared with controls, the patients showed a dysregulated hormonal stress system, more anxiety and more craving when exposed to stress and to alcohol triggers in the laboratory. They were then followed for 90 days. High craving provoked by stress and by alcohol triggers, and a measure of adrenal sensitivity taken in a relaxed state, each predicted a shorter time to relapse³⁷. Stress also acts on the faculty people use to resist. Even mild, uncontrollable acute stress can rapidly impair prefrontal function³, and stress shifts control toward habit⁴.

So the evening drink can't be treated as a question of willpower alone. The person needs ways to bring their stress down that work as quickly as a drink and that can be used at the moment the stress peaks. Naming the feeling is one of them: in an imaging study, putting a feeling into words reduced the amygdala's response to emotional images and increased activity in a region of the prefrontal cortex linked to inhibitory control³⁸. Slow breathing is another. In a study of 40 healthy adults, eight weeks of diaphragmatic breathing practice improved sustained attention, reduced negative mood and lowered cortisol³⁹. Those studies were of general populations, not of drinkers, and we cite them for the mechanism: short, learnable practices that lower arousal within minutes.

Alcohol puts you to sleep and then ruins the sleep it gave you. It shortens the time it takes to fall asleep, but as the body breaks it down, the second half of the night becomes lighter and more broken, and dreaming (REM) sleep is delayed and, at higher doses, reduced⁴⁰. After years of drinking, the brain has also learned to fall asleep with alcohol, so the first weeks without it often bring trouble falling asleep and waking in the night. Sleep improves slowly with time, and for some people it takes more than a year to settle fully⁴¹. Because poor sleep in early sobriety predicts drinking again⁴², we treat sleep as one of the main defences, not as a side effect to wait out.

5.1What a drink does to the night

Ebrahim and colleagues' review of alcohol and normal sleep found that at all doses studied, alcohol shortens the time taken to fall asleep and makes the first half of the night more consolidated, with more disrupted sleep in the second half. It delays the first period of REM sleep, and total REM sleep is reduced at moderate and high doses⁴⁰. Colrain, Nicholas and Baker reach the same conclusion and add what happens with heavy, long-term drinking: alcohol dependence is associated with chronic sleep disturbance, less slow-wave sleep and more REM sleep than normal, changes that last long into abstinence and may play a part in relapse⁴³. Their review also notes evidence of tolerance to alcohol's sedative effect with repeated use, which is part of why the nightcap tends to grow.

Sleep loss acts directly on the systems that drive drinking. One night without sleep reduced the availability of dopamine D2 and D3 receptors in the human ventral striatum⁴⁴. In another study, a night without sleep made the amygdala respond more strongly to negative images and weakened its connection with the medial prefrontal cortex, which normally keeps it in check⁴⁵. A poor night therefore produces, the next day, a reward system that responds less to ordinary things, an emotional system that reacts more to stress, and weaker control over both. A drink relieves all three for an hour, and makes the next night worse.

5.2Sleep after stopping, and why it matters

Sleep recovers after stopping, but slowly. Drummond and colleagues followed men with alcohol dependence at about 16 days, 19 weeks, 14 months and 27 months of abstinence. Sleep was short, fragmented and shallow early on, improved slowly over at least the first year, and some aspects were still abnormal after 27 months; insomnia and broken sleep at around five months may have been related to relapse by 14 months⁴¹. The later follow-ups were small, with nine men at 14 months and four at 27 months, so the long-term figures are tentative. The direction is clear enough to plan around: the first weeks without alcohol are likely to bring worse sleep, not better, and the person needs other ways to fall and stay asleep.

The link to relapse is the reason sleep comes first in the program. Brower and colleagues studied 172 men and women in treatment for alcohol dependence. Of them, 61% had insomnia in the six months before treatment, and those with insomnia were twice as likely to report often drinking to help them sleep (55% against 28%). Among the 74 patients followed for an average of five months after treatment, 60% of those with insomnia at the start drank again, against 30% of those without, and insomnia remained a strong predictor after other factors were taken into account. A history of drinking to sleep did not itself predict relapse⁴². The finding is an association, not proof that fixing sleep prevents relapse. But it shows that someone who stops drinking while their sleep is broken is starting with one of the strongest known risks still in place.

5.3The body clock

Sleep is timed by the body clock, and the clock is set mainly by light. Light acts on the master clock in the suprachiasmatic nucleus and on other regions to regulate circadian rhythms, sleep and mood⁴⁶. The circadian system also regulates reward processing, which is one reason disrupted rhythms and substance use are so often found together⁴⁷. Many of our clients who drink in the evening also keep late hours under artificial light and use screens in bed, so their clock is weakly anchored before alcohol is even considered. A fixed wake time anchors it from one end, daylight early in the day strengthens the anchor, and less light in the evening stops it being pushed later at the other end.

Triggers are the things your brain has learned to link with drinking. Every time you drink, the brain records what was around you: where you were, what time it was, who you were with, how you felt and what had just happened. Repeat that often enough and each of those things starts the urge on its own. That's why getting home, finishing a big piece of work, a particular friend, a bad phone call, Friday evening or a glass in the cupboard can bring on a craving out of nowhere. Good news can be as strong a trigger as bad news. And because stopping doesn't delete what the brain has learned, the triggers are still there after the drinking stops.

6.1How triggers are learned

The learning is the dopamine prediction signal described in Section 3: the reward response moves from the drink to the earliest reliable sign that a drink is coming²¹. Researchers measure the result by showing people the sights, smells or objects linked to their habit and recording their reaction, a method they call cue reactivity, which this paper calls trigger reactivity. A meta-analysis across addictions found that exposure to habit-related triggers reliably increased reported craving, with smaller effects on physical arousal⁴⁸. For alcohol specifically, Schacht, Anton and Myrick pooled 28 brain imaging studies covering 679 people with alcohol use disorders and 174 controls. Alcohol triggers reliably activated limbic and prefrontal regions, including the ventral striatum, the anterior cingulate and the ventromedial prefrontal cortex, and activation of the ventral striatum was the response most often linked to behavior and most often reduced by treatment⁴⁹.

Trigger reactivity may also help predict who drinks again. In a pilot study by Grüsser and colleagues, ten abstinent alcohol-dependent patients and controls viewed alcohol and neutral images during brain scans, and the patients were then followed for three months. Alcohol images activated the putamen, anterior cingulate and medial prefrontal cortex in the patients, and the activation was most pronounced in the five who went on to drink again. The amount they later drank was linked to the strength of that brain response, not to how much craving they reported or how long they had been abstinent⁵⁰. The study was small and the authors call it a pilot. Its point matters for anyone trying to stop: the strength of a trigger's hold can be greater than the person feels or reports.

6.2Why stopping doesn't erase them

Bouton's review of extinction explains why triggers outlast the drinking. Stopping doesn't erase the original learning. It lays new learning over it, and the new learning depends more on context than the old. When the person returns to the setting, mood or company in which they used to drink, or has a drink again, the old response can come back⁵¹. Two practical conclusions follow. Triggers that can be removed should be removed, and the ones that can't need a new, prepared response, used and repeated in the settings where the old response was learned.

Removal comes first because it is the cheapest and most reliable form of control. Duckworth, Gendler and Gross's review argues that strategies which change the situation before an impulse begins are more effective than strategies for resisting an impulse already under way⁵². Verplanken and Wood describe how a change in the context where a habit runs disrupts its automatic triggering and opens a window in which new behavior is easier to establish⁵³. If there is no alcohol in the house, the evening routine that ends at the fridge has nowhere to go.

6.3Noticing a trigger in time

A response only helps if it starts before the drinking routine has run. Quinn and colleagues compared the strategies people used to control strong and weak habits, and found that vigilant monitoring, watching for the signs of the habit and telling oneself not to do it, was the most effective strategy for strong habits, and that it worked by strengthening inhibitory control⁵⁴. In the program, this kind of awareness is trained weeks before the full response plan is written, so that noticing a trigger is routine by the time there is a prepared response to use.

6.4Planning a response for each trigger

The research on implementation intentions shows why the plan has to be specific. These are plans that state what a person will do when a particular situation arises. Gollwitzer and Sheeran's meta-analysis found that forming them had a medium-to-large effect on reaching goals, in part because they make the critical situation easier to notice and trigger the planned response with little deliberation when it arrives⁵⁵. The effect has been tested on drinking directly. In a randomized trial by Armitage with 248 adults from the general population, the two groups who formed implementation intentions reduced their alcohol consumption, while the two control groups did not; choosing a ready-made plan worked as well as writing one's own⁵⁶. Those participants were not trying to stop and most were not heavy drinkers, so the trial shows the mechanism rather than the size of the effect for someone stopping. A response that has to be chosen in the moment is chosen by the person least able to choose well. A response tied in advance to its trigger runs with far less effort.

There are two answers, because there are two time scales. A single craving builds, peaks and passes, and each time it ends in a drink, the link between that trigger and drinking is reinforced. We haven't found reliable research that gives a fixed number of minutes, and we don't quote one. Over weeks and months, the picture is more complicated than most people expect. The anxious, low feeling of early withdrawal usually eases over three to six weeks³⁶, but craving set off by alcohol triggers was higher at 60 days than at 7 days in one study⁵⁷, and the risk of relapse falls with time without disappearing^{1, 2}.

7.1What happens in a single craving

A craving is felt in the body: restlessness, tension, a sense that the evening needs something. The sense of the body's internal state is represented in the insula, and the anterior insula has been proposed as a basis for moment-to-moment awareness of how one feels⁵⁸. Kavanagh, Andrade and May's elaborated intrusion theory describes how such a feeling becomes a full craving episode⁵⁹. A thought about drinking arrives automatically, set off by a trigger or a bodily state. If it promises relief or pleasure, the mind elaborates it, often as vivid images of the drink, and that elaboration fills working memory and makes the person more aware of what they are missing. The theory gives two points to work on: the physical state that feeds the first thought, and the elaboration that turns a thought into an episode.

Observing an urge without acting on it changes what happens next. In an experiment with 123 people who smoked, a brief instruction in urge surfing, watching the urge rise and fall without acting on it, did not reduce the strength of the urges people reported, but it did lead to fewer cigarettes over the following week⁶⁰. That result is modest and useful. The practice changes what the person does with an urge, not whether the urge comes. In people with substance use disorders, a randomized trial found that mindfulness-based relapse prevention weakened the link between low mood and craving, and that this predicted less substance use four months later⁶¹. A larger trial is described in Section 11.

7.2Cravings over weeks

Craving set off by triggers does not simply fade with sober time. Li and colleagues assessed adult men in inpatient treatment for alcohol dependence at 7, 14, 30 and 60 days of abstinence, using four separate groups of 80 and a further group of 19 tested at every point. Craving in response to alcohol triggers was highest at 60 days, in both designs, though heart rate, blood pressure and skin responses did not differ⁵⁷. The authors call this incubation of craving, a pattern first described in animals and in people who smoke and use methamphetamine. The study was of men in an inpatient setting, and it stopped at 60 days, so it does not show when the curve turns down. What it shows is that the second month can bring stronger trigger-set craving than the first, at the same time as the withdrawal-related low mood is easing³⁶ and the glutamate imbalance of acute withdrawal has settled¹².

7.3The risk over months and years

Kirshenbaum, Olsen and Bickel's quantitative review of the relapse curve across alcohol and tobacco found the same shape across studies: the proportion still abstinent falls fastest soon after stopping and then flattens as abstinent time accumulates¹. Moos and Moos followed 461 people with drinking problems for 16 years from the point at which they first sought help. People who got no help were less likely to reach three years of remission and, if they did, more likely to relapse later. Among people who had remitted, those who drank more, saw their drinking as less of a problem, had less confidence in their ability to stay stopped and relied more on avoiding problems rather than dealing with them were more likely to relapse by 16 years². The authors' conclusion is direct: people who cut down on their own may need help to keep it that way.

Some people who once had a serious drinking problem do go on to drink at low levels. But for someone who has repeatedly tried to cut down and failed, a limit is hard to keep for a specific reason: each drink lowers the self-control needed to stop at the limit and makes the next drink more appealing⁶, and each drink also reactivates the learned link between drinking and its triggers⁵¹. A limit has to be applied by the person a drink has just made less able to apply it. That's why our program asks for complete abstinence, and why we'd suggest the same to anyone whose own rules haven't held.

8.1What the population data show

The honest starting point is that recovery without abstinence does happen. Dawson and colleagues analyzed a nationally representative US survey of 43,093 adults and looked at the 4,422 who had met the criteria for alcohol dependence before the past year. In the past year, 25.0% were still dependent, 27.3% were in partial remission, 11.8% drank without symptoms but in a pattern that put them at risk of relapse, 17.7% were drinking at low risk and 18.2% were abstaining. Only a quarter had ever received treatment. Greater severity of dependence made abstinent recovery more likely and recovery without abstinence less likely⁶².

So the question for an individual is not whether anyone ever learns to drink at low risk, but whether they are likely to be one of the people who do. The people who come to us have usually answered that question already, through their own history. Of our 290+ clients, 98.4% had tried and failed to stop before. Many of those attempts were attempts to cut down. For them, another set of rules is a repeat of what has already failed. The program's own position is plain: someone who has struggled to stay away from alcohol for more than a month without going back to it should stop completely, and moderation is unlikely to be the right path for them.

A set period off alcohol is a different thing again. De Visser, Robinson and Bond followed 857 British adults who took part in a month-long abstinence challenge. Taking part was linked to drinking less and to more confidence in refusing drinks six months later, even among those who didn't complete the month, and very few drank more afterwards⁶³. A month off is useful and rarely harmful. But it has an end date, and for someone whose drinking is driven by stress, sleep and well-learned triggers, the conditions that drove it are waiting on the first day after it ends.

8.2The thoughts that argue for one drink

Most drinking after a decision to stop is preceded by a thought that gives permission. One won't hurt. I've earned it. It's a special occasion. I'll start again on Monday. I'll get this under control and then drink like a normal person. These thoughts sound reasonable because they are produced in the hot state, when the value of the immediate reward is temporarily inflated and the person can't feel how differently they will see it tomorrow⁷. Arguing with them in the moment rarely works. Trying to push a thought away tends to make it return more often: in Wegner's experiments, people told not to think of a white bear later thought about it more than people who had been allowed to think about it all along⁶⁴.

Two approaches have better support. The first is to change one's relationship to the thought rather than its content: to notice it as a thought, not a fact or an instruction, a process acceptance and commitment therapy calls defusion. Hayes and colleagues' review found encouraging evidence for the approach while noting that the research base was still developing⁶⁵, and we use it with that caution. The second is to prepare a considered answer to the thought in advance, while calm, so the long-term view is ready when the thought arrives instead of having to be built under pressure. The program does both, for the specific permission thoughts each client has.

You stop by deciding in advance how each drinking situation will go, and by changing the balance of who you spend time with. Research on a large community followed for decades found that people were more likely to drink heavily when the people they were connected to drank heavily, and more likely to abstain when those people abstained⁶⁶. Treatment designed to shift a person's social circle toward support for not drinking has improved drinking outcomes for up to two years⁶⁷. The other half of the work is identity. A person who still sees themselves as a drinker who is temporarily not drinking is exposed every time someone offers a glass. A person who sees themselves as someone who doesn't drink has a much simpler decision.

9.1Drinking spreads through social ties

Rosenquist, Murabito, Fowler and Christakis analyzed 12,067 people in the Framingham Heart Study, assessed several times between 1971 and 2003, along with their friends, relatives, spouses, coworkers and neighbors. Clusters of drinkers and of abstainers were present at every point and extended to three degrees of separation. The clusters were not only explained by drinkers choosing each other: changes in the drinking of a person's network were followed by changes in their own. Relatives and friends had a significant effect; neighbors and coworkers did not⁶⁶. In the full paper, a person was 50% more likely to drink heavily if someone they were directly connected to drank heavily, and 29% more likely to abstain if that person abstained.

The influence can be turned to the person's advantage. In the Network Support Project, Litt and colleagues randomly assigned 210 alcohol-dependent men and women to a treatment designed to shift their social network toward support for abstinence, the same treatment with added incentives, or a case management control. The network support treatment changed participants' networks in that direction and improved drinking outcomes at the first follow-up⁶⁸. At two years, it produced up to 20% more days abstinent than the other conditions, and the network changes were accompanied by gains in confidence and coping that strongly predicted long-term outcomes⁶⁷.

Mutual-help groups are one of the most direct ways to change a social network, and the evidence for them is strong. A Cochrane review by Kelly, Humphreys and Ferri covering 27 studies and 10,565 participants found high-certainty evidence that manualized Alcoholics Anonymous and twelve-step facilitation increased rates of continuous abstinence at 12 months compared with other established treatments such as cognitive behavioral therapy, an effect that held at 24 and 36 months⁶⁹. Our program is not a twelve-step program, and many of our clients come to us after deciding that groups are not for them. Those who do attend a group can continue alongside the program.

9.2From someone trying not to drink to someone who doesn't

Recovery research describes stopping as a change of social identity. Best and colleagues' social identity model of recovery proposes that people move away from groups whose identity is bound up with drinking or drug use and toward groups whose identity supports recovery, and that the new memberships support the new view of oneself⁷⁰. Buckingham, Frings and Albery tested part of this with members of Alcoholics Anonymous and Narcotics Anonymous and with ex-smokers. Seeing the recovery identity as distinct from the old one, and preferring it, was related to lower relapse, less appetitive behavior and higher self-efficacy, though the authors describe the results as exploratory⁷¹. Oyserman's identity-based motivation model explains the underlying process: people act in ways that feel consistent with the identities active for them in the moment, and read difficulty according to whether the goal feels like one that people like them pursue⁷².

The program's self-image work makes the same point in one line: if you see yourself as someone trying to quit alcohol, you'll always remain trying. Identity work comes late in the program because a new self-description is only believable with evidence behind it, and by Day 61 the client has six or seven weeks of it.

9.3Social situations, prepared in advance

Drinking situations that matter to someone's work or family life, such as a client dinner, a wedding or a holiday with heavy-drinking relatives, can't always be avoided, and many are rare enough that they can't be practiced often. Mental rehearsal fills that gap. Driskell, Copper and Moran's meta-analysis found that mental practice reliably improves performance, though less than physical practice⁷³. Imagining a high-risk event in detail and rehearsing the response means meeting it the first time with a response that has already been used. Planning the situation itself, where to stand, what to hold, how to get home and when to leave, applies the situational strategies described in Section 6 to settings that can't be removed⁵².

Your body starts to recover within days, and the recovery continues for months. Heavy daily drinkers can have withdrawal in the first few days, which is why they need a doctor (Section 2). Over the first weeks, the low mood and anxiety of early sobriety usually ease³⁶, although sleep often gets worse before it gets better⁴¹. In a study of moderate to heavy drinkers who stopped for a month, insulin resistance, blood pressure, weight and two cancer-related growth factors all fell⁷⁴. In the brain, the glutamate imbalance of withdrawal settles within a few weeks¹², and grey matter begins to recover within the first two weeks⁷⁵, with further changes over the months that follow⁷⁶.

10.1The first month

Mehta and colleagues recruited healthy people who drank more than about 64 grams of alcohol a week for men, or 48 grams for women, with known liver disease and alcohol dependence excluded. Ninety-four chose to stop drinking for a month and 47 chose to carry on. The abstinence group had been drinking an average of about 258 grams a week, roughly 32 standard UK units. After one month, the abstinence group showed significant falls in insulin resistance, blood pressure, weight and two growth factors linked to cancer, VEGF and EGF, none explained by changes in diet, exercise or smoking. The group who kept drinking showed no significant changes⁷⁴. The participants chose their own group rather than being randomized, so the study shows an association. It is also one of the clearest measurements of what a month without alcohol does in people who were not dependent.

10.2The brain

Heavy drinking is associated with smaller volumes of grey and white matter, and some of that loss recovers with abstinence. Van Eijk and colleagues scanned 49 alcohol-dependent patients within the first 24 hours of detoxification and again after two weeks of supervised abstinence, and compared them with 55 healthy controls. On the first day, the patients had less grey matter, with the largest differences in the cingulate gyrus, frontal regions, cerebellum and insula. After two weeks, a significant though partial recovery of grey matter had already occurred in several regions, at different rates in different parts of the brain⁷⁵. Pfefferbaum and colleagues followed 58 men with alcoholism with scans at about 12 and 32 days of sobriety and again two to twelve months later. Improvements in cortical grey matter and in the fluid spaces of the brain appeared early in abstinence, further changes appeared later in those who stayed abstinent, and white matter volume fell in those who went back to drinking⁷⁶.

The reward system recovers more slowly, and the evidence on its timing is thinner. The blunted dopamine response described in Section 3 was measured in people who had recently stopped²³, and the persistence of altered emotional processing into protracted abstinence is described mainly from animal work³⁶. We tell clients what is known: recovery begins quickly, varies between people and between brain regions, and continues for months. We don't promise that any marker returns to normal by a given day, because the research doesn't support a date.

10.3Exercise, mood and fitness

Exercise is part of the first phase of the program, and the evidence for it needs stating carefully. Hallgren and colleagues' meta-analysis of 21 studies with 1,204 people with alcohol use disorders found that exercise did not reduce alcohol consumption on its own, but it did significantly reduce depressive symptoms and improve physical fitness⁷⁷. Lynch and colleagues' review proposes that exercise acts on the dopamine and glutamate signalling that drug use alters, with effects that depend on the stage of addiction and the type of exercise⁷⁸. We don't use exercise as a treatment for drinking. We use it because it lifts the low mood and restlessness of early sobriety, improves sleep and gives the evening a structure, and because visible progress in training is direct evidence to the client that they can change something about themselves⁷⁹.

Inge, a senior operator in Belgium who came to us for alcohol, described her first two months in one line: “I've been sober for 56 days today.” She has now been sober for over six months. We cite her not as evidence that the program works, which a single person can't provide, but because she describes the kind of craving the second phase aims for, a passing thought rather than a strong urge: “And even then, it wasn't a strong craving, just a thought about it. And I now know why it's important to not give in to the thoughts.”

A drink after stopping is not the end of the attempt unless you decide it is. What turns one drink into a full relapse is usually the story people tell themselves about it: that it proves they're weak, that they've always been like this, that the whole attempt has failed. People who explain a drink that way are more likely to keep drinking⁸⁰. The better response is practical. Go back to your plan the next day, work out what led to the drink, and change the part of the plan that didn't cover it. One extra point applies to alcohol: if a relapse becomes several days of heavy drinking, stopping again can bring withdrawal back, so talk to a doctor first¹³.

The research on this is known as the abstinence violation effect. Curry, Marlatt and Gordon studied people who had lapsed after stopping smoking and found that those who went on to full relapse put their lapse down more to internal, stable and global causes, such as a lasting personal weakness, than those who lapsed and recovered⁸⁰. Witkiewitz and Marlatt's reformulation of relapse prevention for alcohol and drug problems describes relapse as a dynamic process in which stress, craving, confidence, coping and mood interact, and in which small changes can produce sudden shifts in behavior⁸¹. In that model there is no single cause of relapse to remove and no point at which a person is permanently safe. What protects them is a set of coping responses that stay available as conditions change. Hendershot and colleagues' review of a decade of relapse prevention research describes the approach as part of most cognitive-behavioral treatments for addictive behaviors⁸².

Confidence in one's ability to stay stopped matters for the long run. In Adamson, Sellman and Frampton's systematic review of what predicts the outcome of alcohol treatment, alcohol-related self-efficacy was among the most consistent predictors, along with dependence severity, psychiatric symptoms, motivation and the treatment goal⁸³. Lower self-efficacy also predicted relapse over 16 years in Moos and Moos's study². A drink read as proof of weakness lowers that confidence. A drink that is analyzed and answered with a better plan protects it.

The combination of awareness training and prepared coping responses that runs through the program has direct support in a randomized trial. Bowen and colleagues assigned 286 people who had completed initial treatment for substance use disorders, including alcohol, to mindfulness-based relapse prevention, standard relapse prevention or treatment as usual. Both relapse prevention approaches reduced the risk of relapse to substance use and heavy drinking compared with treatment as usual at six months, and at twelve months the mindfulness-based approach was associated with fewer days of substance use and less heavy drinking than either⁸⁴. Our program is not either protocol, and the trial doesn't test it. But the elements the trial found useful are the ones a client carries out of the 90 days.

The Behavior Control System is the same 90-day program for every habit, applied to the loop beneath each person's drinking. Days 1 to 30 rebuild the ground drinking has been running on: sleep, the body clock, daylight, exercise, stress and a structured day. The client stops drinking on Day 15, after two weeks of that work, or on the date their doctor sets. Days 31 to 60 work on the moment a drink actually happens: every trigger is mapped, the ones that can be removed are removed, and each one left gets a prepared, rehearsed response, including the thoughts that argue for one drink and the stress that most often starts the evening. Days 61 to 90 work on identity and on the drinking culture, events and relationships that still pull the person back. Six months of aftercare follow. It asks for about thirty minutes of focused work a day: a short lesson each morning, a brief log each night, and a set of daily practices, which the program calls north stars, built into the routine the person already has.

12.1Before Day 1

Every program starts with a Root-Cause Diagnostic: an application describing the habits involved and what has already been tried, followed by a thirty-minute, one-on-one conversation with our team. For alcohol it has two jobs. The first is safety: it asks how much and how often the person drinks and about any previous withdrawal, and anyone who drinks heavily every day is told to see a doctor before they begin (Section 2). The second is to identify the loop beneath the drinking: what it does for the person, the earliest recurring trigger, the conditions that make it likely, why earlier attempts broke down and what is missing for change to hold. Two people who both drink every evening rarely have the same loop. One drinks to end a working day that never otherwise ends; another drinks only with a particular group; a third drinks alone after the household is asleep. The structure of the diagnostic parallels the three domains the Addictions Neuroclinical Assessment proposes for exactly this problem: the urge set off by triggers, negative emotion and executive function⁸⁵. It is not a clinical assessment and takes no medical measurements.

A preparatory phase, Phase 0, follows before Day 1 and does not count toward the 90 days. It separates the decision to stop from the start. The client sets out in concrete terms the life they are building and what drinking has cost it, makes a commitment to someone who matters to them, chooses a start date that marks a real boundary in their life, and proves to themselves that they can hold one full day of control. For a heavy daily drinker, that day, like the stop on Day 15, follows the doctor's advice. The choice of start date rests on the fresh start effect: temporal landmarks such as the start of a week, a month or a meaningful date were followed by increases in aspirational behavior⁸⁶. The day of control rests on Bandura's finding that direct experience of succeeding at a behavior is the most dependable source of belief that one can do it⁷⁹. A commitment that is recorded and made to another person holds better than a private one⁸⁷.

12.2Why the order matters for alcohol

The order follows from the research in Sections 2 to 11. Drinking stops after sleep and stress have had two weeks of work, because poor sleep and stress-provoked craving are among the strongest known predictors of drinking again^{37, 42}, and because the hours the evening drink used to fill need something already in them on the first night without it. The trigger work comes in the second month, when the low mood of early withdrawal is easing but craving set off by triggers may be at its strongest^{36, 57}. Identity and social context come last, because they change slowest and because a new self-description needs weeks of evidence behind it to be believed.

Replacement is the last step because a habit that served a function leaves a gap when it stops, and a gap left empty tends to be filled by whatever is most available. A replacement placed at the same time and in the same setting as the old drink can take over its context²⁴. Lally and colleagues found that the automaticity of a new daily behavior built gradually, reaching 95% of its plateau after a median of 66 days with wide variation between people, and that missing a single occasion did not materially affect the process⁸⁸. By Day 90 the replacement has been in place long enough to begin running on its own.

12.3In our clients' words

Two of our clients who came to us for alcohol have described their experience on the record. Inge, a senior operator in Belgium, had tried to stay sober before and held on by willpower each time. “I also understand now why my previous attempts at being sober didn't work,” she said, and of how it felt this time: “This time I feel different about sobriety. It doesn't feel like a punishment. It feels like a gift to myself and to my brain.” She has now been sober for over six months. Bjorn, a senior operator in the Netherlands, came to us having lost his motivation and joy in life. When he recorded his interview he said: “I've been sober for 71 days now and this time it was easier than previous attempts because of all the information and training exercises you get along with the program.” Their words are shared with consent. They are two people's experiences, not evidence of what will happen for anyone else.

Phalanx Labs has worked with 290+ clients across 12 countries over 7 years. Of those clients, 98.4% had already tried and failed to stop before they came to us. Our success rate is 97%, defined as the share of clients who ran the program and stayed off the habit through the program and the six months of aftercare that follow it. That rate covers all of our clients and all of the habits they came with. We don't publish a separate rate for alcohol. Because the definition matters as much as the number, we set out below what each figure counts, how it is known and what it does not establish.

13.1How outcomes are tracked

Throughout the program, the client keeps a nightly log and completes a weekly progress update. These are written records of what happened, what was difficult and what was handled; they are not scored instruments. At the end of each of the first two phases the client has a progress call with our team, the first around Day 28 and the second after Day 60. During aftercare we follow up with clients on whether they have stayed off the habit. A client counts as a success when they have run the program and stayed off the habit through the program and the full six months of aftercare.

13.2What our figures do not show

• Self-report. Whether a client has stayed off alcohol is known from what the client tells us, not from breath, blood or urine testing.
• Selection. Clients choose to apply, are accepted after a diagnostic in which we decline people we can't help, and are counted only if they ran the program. People who decide to work with us may differ in motivation or circumstance from people who don't.
• No comparison group. We can't say how many of the same people would have stopped through another route, such as medication, a mutual-help group or treatment, over the same period.
• Time window. The success rate covers the program and six months of aftercare. It does not describe what happens after that, and the relapse research in Section 7 shows that risk continues beyond it.
• All habits together. The rate is across every habit our clients bring. We don't report a rate for alcohol alone, and a single rate does not predict the outcome for any one person.

13.3What the research in this paper does not show

The studies cited here have their own limits, and readers should weigh them. Much of the alcohol research was done with people in treatment for alcohol dependence, often in inpatient settings and often mostly or only men, such as the craving study by Li and colleagues and the sleep and imaging studies by Drummond and by Pfefferbaum and their colleagues^{41, 57, 76}. Many of our clients drink heavily without ever having been in treatment, and the findings may apply to them with different strength. Several imaging studies were small, and some findings come from animals or from people who smoked rather than drank, such as the research on urge surfing and on how a lapse is explained^{60, 80}; we use those for the mechanism they describe and say so where we cite them. Some of what the program does has weaker direct evidence for drinking than its reputation suggests: exercise, for example, improved mood and fitness in people with alcohol use disorders but did not reduce their drinking on its own⁷⁷.

Other approaches have stronger trial evidence than ours. Acamprosate and naltrexone have been tested in many randomized trials¹⁶, and twelve-step facilitation has high-certainty evidence for continuous abstinence⁶⁹. Our method has not been tested in a randomized or controlled trial, its outcomes have not been assessed independently, and it has not been published in a peer-reviewed journal. The research supports the reasoning behind each element. It does not show that the combination works, or how much each element contributes. The figures in Table 5 are our evidence for the combination, with the limits stated above. A controlled evaluation would need independent outcome assessment, a comparison condition and follow-up beyond six months. We would welcome one.

See a doctor before you stop drinking if you drink heavily every day, if you have had withdrawal symptoms such as shaking, sweating or anxiety when you go without, if you have ever had a withdrawal seizure or delirium tremens, if you take benzodiazepines or other sleeping or anxiety medication, or if you have a heart, liver or other medical condition. A doctor can assess the risk, plan the stop, treat withdrawal if it comes, and discuss medication that helps people stay stopped^{10, 16}. If you or someone with you has a seizure, becomes confused or starts seeing things during withdrawal, call emergency services.

Phalanx Labs is a behavior and performance engineering firm. The Behavior Control System is a structured method for changing behavior. It is not therapy, clinical care or medical treatment, and it does not diagnose or treat any medical or psychiatric condition. Our team does not act as the client's doctor or therapist, and the program does not replace either. Where the diagnostic suggests withdrawal may be a risk, we tell the person to see a doctor before they begin, and the stop date follows medical advice rather than the other way round.

Some of the program's physical practices carry ordinary risks of their own. Cold and heat exposure and starting a training program are not suitable for everyone, and people with heart conditions, high blood pressure or other medical conditions should check with their doctor first. The program gives general information about supplements and tells clients to consult a doctor before taking any; it does not prescribe or recommend medication or supplements.

The program is also not designed for every situation. It is not a crisis service. It is not suitable as the only support for someone with a serious mental health condition, someone at risk of harming themselves or others, or someone whose drinking needs medically supervised treatment. The diagnostic exists partly to identify these situations, and when we can't help, we say so and suggest the person seek appropriate care.