What Are the Alternatives to AA?

The short answer

Most people who enter the rooms of Alcoholics Anonymous find themselves full of resistance. Most don’t make it 30 days, which is a real and substantial hindrance to recovery, because AA is one of the most available and affordable recovery routes.

Fortunately, there are real alternatives, and while it is true that Alcoholics Anonymous reduces drinking and encourages abstinence, for those who do stay in and do the work long-term, the vast majority of those who try AA do not. A recent Cochrane review found high-certainty evidence that manual-guided twelve-step facilitation, a clinician-delivered protocol (importantly, NOT what is delivered in the free rooms of community AA), beat other established treatments such as CBT when it comes to producing continuous abstinence. So "alternatives to AA" is a fit and goal problem (since not everyone is seeking abstinence), not an efficacy problem.

The AA alternatives fall into four groups:

  1. Other mutual-help communities that drop the higher-power and powerlessness language (SMART Recovery, LifeRing, Women for Sobriety, Recovery Dharma), where cohort studies have not detected any difference in outcomes from twelve-step groups, in samples too small and too self-selected to rule one out;

  2. Professional therapies such as cognitive behavioral therapy, acceptance and commitment therapy, mindfulness-based recovery therapies (MBSR, and MBRP), and motivational enhancement

  3. Medications, the most under-used option in the whole field and the one most likely to be missing from your plan; and approaches that aim at reduced drinking rather than total abstinence. Most people who change their drinking combine several.

Before anything else, one safety point. If you drink heavily every day, especially if you have done so for a number of years, do not simply stop on your own. Alcohol withdrawal can cause seizures and can kill you. Get a medical plan before you stop. And if you have already cut down or stopped and you develop shaking, heavy sweating, vomiting, a racing heart, confusion, hallucinations, or a seizure, call 911 or go to an emergency department now rather than waiting it out. Risk is highest in roughly the first six to seventy-two hours, and the most dangerous form, delirium tremens, usually appears later than people expect.

If you have tried AA and it did not take, that tells you something about fit. It tells you nothing about you and your actual capacity to achieve full recovery and remission.

What the evidence actually says

About AA and Other Mutual Help Groups

Let’s start with the thing many people looking for an alternative do not expect: AA has good evidence behind it. John Kelly, Keith Humphreys, and Marica Ferri published the Cochrane review in 2020, covering 27 studies and 10,565 participants (DOI: 10.1002/14651858.CD012880.pub2). For manual-guided AA and twelve-step facilitation compared with other clinical treatments such as cognitive behavioral therapy, continuous abstinence at 12 months came out higher (risk ratio 1.21, 95% CI 1.03 to 1.42; 2 studies, 1,936 participants), rated high-certainty evidence, and the advantage held at 24 and 36 months.

The rest of that review is where it gets complicated, and the complications run in both directions. On percentage of days abstinent, AA and twelve-step facilitation performed about the same as the comparison treatments at 12 months (mean difference 3.03, 95% CI −4.36 to 10.43; 4 studies, 1,999 participants, very-low certainty), but better at 24 months (12.91, 95% CI 7.55 to 18.29; 2 studies, 302 participants, low certainty) and at 36 months (6.64, 95% CI 1.54 to 11.75; 1 study, 806 participants, low certainty). One study also found alcohol addiction severity favoring AA/TSF at 12 months. On drinking intensity, measured as drinks per drinking day, there was essentially no difference (−0.17, 95% CI −1.11 to 0.77; 1 study, 1,516 participants, moderate certainty), and the same for alcohol-related consequences (−2.88, 95% CI −6.81 to 1.04; 3 studies, 1,762 participants, moderate certainty). The review concluded AA/TSF probably produces substantial healthcare cost savings, at moderate certainty, with the clearest advantage among people with worse prognostic characteristics.

Two things are worth pulling out of that. First, the outcomes AA is explicitly designed around, the abstinence-shaped ones, are where its advantage shows up, and the longer-term estimates point the same way, though they rest on one or two studies at low or very low certainty and should not be leaned on hard. Second, on drinking intensity and alcohol-related harm, the moderate-certainty estimates show other treatments holding their own. So if total abstinence is your goal, the evidence says give AA a serious try. If it is not your goal, or the room is not survivable for you, what you are giving up is an edge on continuous abstinence, not an edge on everything.

There is also a distinction the review makes that almost never survives being quoted. The high-certainty result is for manualized twelve-step facilitation, meaning a clinician working from a standardized protocol. Where the facilitation was delivered without such a protocol, the abstinence estimate at three to nine months was too imprecise to interpret in either direction (risk ratio 1.71, 95% CI 0.70 to 4.18; 1 study, 93 participants, low certainty), though the same comparison did favour AA/TSF on percentage of days abstinent (3.00, 95% CI 0.31 to 5.69) and on drinks per drinking day (−1.76, 95% CI −2.23 to −1.29). One thing the review does not do is estimate an effect for peer-led AA attendance on its own, separate from professional facilitation. That is a genuine gap in the evidence rather than a negative finding about church basements, and anyone who tells you Cochrane proved AA meetings work, or proved they don't, is going past what the review measured.

One thing the review does not do is estimate an effect for peer-led AA attendance on its own, separate from professional facilitation.

The secular mutual-help groups now have outcome evidence, and it is more encouraging than most people realise. The relevant work is a research program rather than a single study. Sarah Zemore and colleagues surveyed adults with a lifetime alcohol use disorder in 2015, comparing members of Women for Sobriety, LifeRing, and SMART Recovery with current twelve-step attendees (DOI: 10.1016/j.jsat.2016.10.004). Members of the alternatives were less religious and generally higher in education and income, and LifeRing and SMART members were less likely to hold the strictest abstinence goal. Despite attending fewer in-person meetings, members of all three reported equivalent involvement and higher satisfaction and group cohesion than twelve-step members. One caveat does a lot of work here: the alternative-group members were recruited through their own group directors, while the twelve-step comparison came from a general online meeting hub. Recruiting through group leadership selects for the satisfied and highly involved, which is close to being the finding.

The same cohort was then followed forward, which is the part that bears on the question people actually ask. Zemore's 2018 longitudinal analysis tracked 647 adults at 6 and 12 months and detected no significant differences in efficacy between Women for Sobriety, LifeRing, SMART, and twelve-step groups (DOI: 10.1016/j.jsat.2018.02.004). Read that as a failure to detect a difference rather than as proof of equivalence: the sample is modest, self-selected, recruited online, and no equivalence margin was set in advance, so a real difference could hide inside it. The authors themselves say the results "tentatively suggest" comparable effectiveness. It also found something that cuts against easy conclusions: people whose primary group was SMART fared worse across outcomes, and LifeRing members had lower odds of total abstinence. Both effects disappeared once baseline recovery goal was controlled for, which suggests the difference was selection, people with weaker abstinence motivation choosing those groups, rather than the groups themselves. The authors' own summary is worth quoting in spirit: the alternatives look about as effective, and this population has the best odds when committing to total abstinence.

A 2025 follow-up pooled the 2015 and 2021 cohorts, 1,152 people in all, and reached the same conclusion with more power (DOI: 10.1016/j.drugpo.2025.104921). Which group someone chose was unrelated to their outcomes. What predicted outcomes was how involved they were: greater involvement strongly predicted alcohol abstinence (OR = 2.62), lower odds of alcohol problems (OR = 0.39), and fewer drinking days. The authors note that SMART's program has varied across time and geography, so they counsel some caution reading the SMART-specific result.

For SMART Recovery on its own, Alison Beck and colleagues ran a systematic review in 2017 and found 12 studies, only 3 of which evaluated effectiveness (DOI: 10.1037/adb0000237). Positive effects appeared; small samples and inconsistent methods meant the authors could not draw conclusions about efficacy. Nine years on, that review is the standard citation for "SMART is unproven," and the Zemore cohort work has largely overtaken it.

Put the three together and the 2026 answer to "is SMART as good as AA?" is: nobody has run a randomised comparison, two prospective cohorts have looked for a difference and not found one, and what tracks with outcomes in that data is not which room you picked but how involved you got. Two honest caveats on that last point. These cohorts were recruited through group leadership and skew toward higher education and income, so they may not describe you. And involvement and outcome move together in a way the study design cannot untangle, since people whose recovery is going well are also the people who keep showing up and take on a service role. It is a good bet, not a proven lever.About Other Therapies

No professional therapy has been shown to beat the others. Molly Magill and colleagues published a meta-analysis of cognitive behavioral therapy for substance use disorders in 2026, covering 52 randomized trials across 9,442 participants (DOI: 10.1016/j.brat.2026.105101). CBT worked, but the effect depended on what it was compared against. Against usual care or minimal treatment, it produced a small benefit on consumption outcomes (g = 0.14, 95% CI 0.01 to 0.26). Against other evidence-based treatments, it did not outperform them. Be careful with that second result: a non-significant difference in a heterogeneous subgroup is consistent with rough parity, but it does not establish equivalence, and it is about CBT specifically rather than professional therapy as a category. Added on top of usual care rather than substituted for it, CBT produced larger effects on consumption (g = 0.44, 95% CI 0.01 to 0.88, p=.048) and psychosocial functioning (g = 0.56, 95% CI 0.08 to 1.03), though those estimates are barely significant and carry near-total heterogeneity (I² = 87.6%).

Kelly and Magill land in the same place, and it is not the place treatment marketing wants you to land. No modality has emerged as the one that works. My own read, and I want to flag that this is a clinical view rather than something either study tested, is that what separates people who get better is much less about which brand of help they picked and much more about whether they got enough of something and whether they stayed.

About Medications

Medication is the alternative most people have never been offered. Melissa McPheeters and colleagues published a systematic review and meta-analysis in JAMA in 2023 covering 118 clinical trials and 20,976 participants (DOI: 10.1001/jama.2023.19761). The review assessed a range of individual medications; two reached first-line status. For acamprosate, you would need to treat about 11 people to prevent one additional person from returning to any drinking, though the confidence interval on that estimate runs from 1 to 32, which is very wide. For oral naltrexone at 50 mg a day, about 18 people for the same outcome (interval 4 to 32), and about 11 people to prevent one returning to heavy drinking (interval 5 to 41). Injectable naltrexone reduced drinking days by about five over a 30-day treatment period, with an interval whose lower bound is half a day. Several other medications were assessed with varying degrees of support, including some used off-label specifically to reduce heavy drinking rather than to produce abstinence, which is worth asking a prescriber about if reduction is your goal. Side effects of the two first-line drugs are real and mostly gastrointestinal: diarrhea with acamprosate, nausea and vomiting with naltrexone. The authors conclude that alongside psychosocial treatment, those two belong in the conversation far more often than they appear in it.

Those intervals matter, and I would rather show them than quote a single tidy number. What they mean is that the average benefit is real and the size of it is genuinely uncertain. Naltrexone also does something worth knowing about: it does not require you to be abstinent before you start, which makes it one of the few options that fits a person who has not decided yet what they want.

A newer, still-emerging option: GLP-1 receptor agonists. These are the same drug class built for diabetes and weight loss — semaglutide (Ozempic, Wegovy), tirzepatide (Mounjaro, Zepbound), liraglutide (Victoza, Saxenda), and the older exenatide (Byetta, Bydureon) — and there is now real, if early, evidence they affect drinking too. Two small randomized trials have tested semaglutide directly. Christian Hendershot and colleagues ran a 9-week trial of 48 adults with AUD who were not seeking treatment and found weekly semaglutide reduced drinks per drinking day (β −0.41, 95% CI −0.73 to −0.09) and weekly craving (β −0.39, 95% CI −0.73 to −0.06), though it did not move the average number of drinks per calendar day or the number of drinking days (JAMA Psychiatry, 2025; DOI: 10.1001/jamapsychiatry.2024.4789). A larger Danish trial went further: Mette Kruse Klausen and colleagues randomized 108 treatment-seeking adults with AUD and comorbid obesity to semaglutide 2.4 mg weekly or placebo, both on top of standard therapy, for 26 weeks, and found a 13.7-percentage-point greater reduction in heavy drinking days in the semaglutide group (95% CI −22 to −5.4) (Lancet, 2026; DOI: 10.1016/S0140-6736(26)00305-3). Two trials, run independently in different populations, pointing the same direction is worth taking seriously.

It is not a clean story, and the messy part is informative. An earlier trial of a related drug, exenatide, randomized 127 treatment-seeking adults over 26 weeks and found no significant effect on its primary outcome, heavy drinking days, across the full sample. A benefit turned up only in an unplanned subgroup of participants with obesity (Klausen et al., JCI Insight, 2022; DOI: 10.1172/jci.insight.159863). That null result is exactly what led the follow-up Lancet trial to enroll only people with comorbid obesity — the field is still working out who this helps, not just whether it helps.

There is also a much larger, much less certain signal sitting behind the trials. A Swedish national registry study tracked 227,866 people with an AUD diagnosis for a median of nearly nine years and found that periods of semaglutide use were associated with 36% lower odds of alcohol-related hospitalization (adjusted hazard ratio 0.64, 95% CI 0.50–0.83) and liraglutide use with 28% lower odds (aHR 0.72, 95% CI 0.57–0.92) — numerically as strong as, or stronger than, naltrexone and acamprosate combined in the same data (HR 0.74, 95% CI 0.61–0.89) (Lähteenvuo et al., JAMA Psychiatry, 2024; DOI: 10.1001/jamapsychiatry.2024.3599). Read that as association, not proof: it is a within-person design, which reduces some of the usual confounding, but it cannot rule out that people who end up prescribed a GLP-1 drug differ in other ways from people who don't.

Put together: this is genuinely emerging evidence, not settled evidence. No GLP-1 drug carries FDA approval for alcohol use disorder; every use for drinking today is off-label. If you are already taking one of these medications for diabetes or weight and you also drink more than you want to, it is worth raising both trials with your prescriber directly. If you have no metabolic reason to be on one, the evidence is not yet strong enough that I would tell you to seek it out ahead of the two first-line, FDA-approved options above.

One safety note that belongs with any mention of these drugs. Naltrexone is an opioid blocker. If you are using any opioid, including prescribed painkillers, starting it can throw you into acute withdrawal, so it generally requires a period of being opioid-free first, and it will blunt opioid pain relief in an emergency. It should not be used in acute hepatitis or liver failure, which are not rare in people who have been drinking heavily for years, and baseline liver testing before starting is standard. If you stop naltrexone after being on it, your opioid tolerance is lower than it was, which raises overdose risk if you use opioids again. Acamprosate is cleared by the kidneys and is not appropriate, or needs a reduced dose, in significant renal impairment. GLP-1 drugs carry their own profile: nausea, vomiting, and diarrhea are common, especially early in treatment or right after a dose increase; they are not approved for anyone with a personal or family history of medullary thyroid carcinoma or multiple endocrine neoplasia syndrome type 2; and they carry a real, if uncommon, risk of pancreatitis and gallbladder disease. None of this is a reason to avoid asking. All of it is a reason the asking happens with a prescriber and not on the internet.

The alternative that isn't a group at all: change the goal. I want to be careful here, because the cleanest illustration I have is off-target. A 2026 individual-participant meta-analysis in JAMA Psychiatry pooled 12 randomized trials of medications for stimulant use disorder, 2,000 participants (DOI: 10.1001/jamapsychiatry.2026.1092). The medications mostly did not beat placebo, which is the headline. The part that carries over is about measurement: 31.2% of participants met the trials' reduced-use criterion while 13.3% achieved abstinence. Three caveats travel with that. The criterion was "five or more using days a month down to one through four or abstinence," so the larger figure contains the smaller rather than describing a separate group. Reduced use was self-reported while abstinence was urine-confirmed, so part of the gap is measurement rather than change. And it is a stimulant study, from trials run between 2001 and 2011, not an alcohol study. What survives is the authors' conclusion: scoring recovery as abstinence-or-nothing makes a large group of people who genuinely changed invisible.

For alcohol specifically, the case for reduction as a legitimate target rests on a different and more directly relevant literature, the work linking drops in WHO risk drinking levels to improved functioning and health outcomes. The practical argument stands on its own: fewer heavy drinking days is less liver damage, fewer injuries, and fewer of the nights that end badly, and none of that requires a study to notice.

Where the experts disagree

The twelve-step tradition: the alternatives are a way of avoiding surrender. The strongest version of this argument is about mechanism. AA's central demand, admitting you cannot manage this alone, is what does the therapeutic work, and programs that soften it let people keep the illusion of control that got them here. That deserves better than a reflexive dismissal, and it should be tested against mechanism research rather than against outcome data, which is consistent with almost any mechanism and evidences none. The mechanism literature does identify things AA seems to work through, including changes in a person's social network, confidence in staying sober around drinkers, and better coping with negative feelings, with spirituality showing up as an additional pathway among the most severely affected people. What nobody has done is operationalise and test "surrender" or "powerlessness" directly. So the traditional account has not been refuted. It has not been tested either, and it is worth being clear about which of those is true.

The critics: AA is a religious program the healthcare system keeps prescribing. The objection is about the mismatch between AA's spiritual framework and a secular medical system that routinely mandates attendance, sometimes through the courts. Critics also note that AA's structure makes it nearly impossible to study people who leave, so the evidence base skews toward people who stayed. On the mandate question specifically, the Cochrane review offers no help in either direction, because it excluded coerced participants by design. That strengthens the critics' point rather than answering it: the evidence base for AA comes from people who chose it.

Kelly and Humphreys, whose own work is the strongest evidence for AA, did not draw the maximal conclusion from it. Their review reports certainty grades honestly, distinguishes manualized facilitation from the fellowship, and hedges the cost-savings finding. Notice that the researchers with the most reason to overclaim didn't.

And the disagreement I care most about: whether abstinence is the only legitimate goal. A large part of the treatment field still treats any non-abstinence goal as denial. The evidence does not support that as an absolute, and treating it as one has a measurable cost, because people who will not accept an abstinence goal frequently accept no help at all rather than the wrong help. The counterargument is real too: for some people, particularly those with severe physical dependence, moderation goals genuinely fail and delay care that would have worked. Both of these are true, which is why goal choice belongs in a conversation with a clinician rather than in a doctrine.

My take: you are not shopping for a group, you are shopping on four axes

Here is what I have seen go wrong for something like twenty years. Someone tries AA, it doesn't fit, and they conclude that help doesn't work for them. They do not go looking for a different kind of help. They go back to drinking with a new piece of evidence about how broken they are. The failure was a matching failure, and it got filed as a character failure.

So stop asking "what should I join?" and start asking four separate questions.

  1. What is your goal? Complete abstinence, or less drinking? These point to different places. AA, LifeRing, and Women for Sobriety are abstinence-based. SMART accommodates both. Moderation-oriented programs and naltrexone-based approaches are built for the second. Pick the goal honestly rather than picking the one you think you are supposed to want. A goal you are lying about is a plan that will not survive month two.

  2. What is the mechanism you actually need? Some people need belonging, and a room full of people who get it is the entire medicine. Some need skills, in which case CBT and SMART's tools beat any amount of fellowship. And some need naltrexone, because no quantity of insight changes a receptor. Most people need more than one of these.

  3. What format can you sustain? Daily meetings, weekly therapy, an app, one prescription and a check-in. The best program you will not attend is worse than a mediocre one you will.

  4. What is in your way? Cost, childcare, work hours, whether you can say any of this out loud where you live. This is the axis people skip, and it decides more outcomes than the first three.

In the Unhooked Method’s SAPRO lens I use - Stimulus, Perception, Activation, Response, Outcome - each option works at different links. Environmental change works at the Stimulus, cognitive work at Perception, and medication at Activation. Skills and fellowship change the Response (and the environment to some extent). When someone tells me nothing has worked, what I usually find is that they tried three things that all worked at the same link and left the rest of the chain untouched.

And then my EAT process, Explore, Accept, Transform, has an answer for the thing most people are actually stuck on. If you bounced off AA, the Explore step is not "find a better group." It's an autopsy. Was it the God language, the powerlessness framing, the abstinence requirement, the meeting times, or the fact that you were not ready? Each points to a different next move. "AA didn't work for me" is a description, not a diagnosis.

One last thing, said plainly. The single most common gap I see in people's plans is that nobody ever offered them medication.

What to actually do

✅ If you're physically dependent on alcohol, talk to a doctor before you stop. This one comes first because it is the only item on the list that can kill you. Alcohol withdrawal can cause seizures and can be fatal. "Just quit and go to a meeting" is genuinely dangerous advice for a daily heavy drinker, and the medical plan comes before the recovery plan.

✅ Ask your doctor about naltrexone or acamprosate this month. These are first-line treatments with real evidence, and most people who drink problematically have never been offered either. Tell your prescriber about any opioid use, including prescribed painkillers, and about any liver or kidney problems, because those determine which of the two is appropriate and when you can start.

✅ Name what specifically didn't work about AA before you pick the next thing. The higher-power language, the powerlessness framing, the abstinence requirement, the meeting format, or the timing. Each points somewhere different. SMART Recovery if you want tools and no spirituality. LifeRing if it's the God language and not the fellowship. If the room itself is the barrier because it is all men, Women for Sobriety. Recovery Dharma leans contemplative rather than theistic, and is the larger of the two Buddhist-influenced networks since Refuge Recovery split in 2019. Secular Organizations for Sobriety is another secular option. Celebrate Recovery is the Christ-centred one, though note that it keeps the twelve steps and the powerlessness framing, so it is a fit if AA was not spiritual enough for you and a poor fit if the steps themselves were the problem. One honesty note on this list: only SMART, LifeRing and Women for Sobriety have been studied against twelve-step groups. The rest are here on fit, not on evidence.

✅ Weigh the practical asymmetry honestly. AA has vastly more meetings than any alternative, in vastly more places, at more hours, for free. In a small town or on a night shift, "the group with the best fit" may not exist near you and the one that does may be AA. Online meetings narrow this gap but do not close it.

✅ Try at least three meetings of anything before you judge it. That number is my rule of thumb rather than a study result, but groups vary enormously by room and not just by organization. And in the cohort data, how involved people got tracked with their outcomes more closely than which organization they picked, so lean toward something you can actually throw yourself into.

✅ Add a professional layer, not just a peer one. A therapist trained in CBT or motivational interviewing, plus a group, plus a medication where appropriate. The trials do not directly test stacking, but no single-modality option has outperformed the others, which makes combining them the reasonable bet.

✅ Write down your goal and say plainly whether it is abstinence. Then choose programs that support that goal rather than ones that will spend six months trying to change it. Note the honest counterweight: in the one longitudinal study of the alternatives, people committing to total abstinence had the best odds.

✅ If a reduction goal is what fits, look at the programs built for it. Moderation Management, HAMS, and clinician-delivered moderation protocols are options, none of them carrying comparative outcome evidence of the kind cited above. The Sinclair Method also comes up a lot; note that it uses naltrexone dosed before drinking rather than daily, which is a different protocol from the daily 50 mg regimen the trial evidence above supports, with a much thinner evidence base of its own, so set it up with a prescriber rather than from a forum. So is the coaching program I founded at IGNTD, which I have a financial interest in and am listing here as one item on a menu rather than the menu. Ask a clinician whether a reduction goal is safe for you specifically, because with significant physical dependence it often is not.

FAQ

Does AA actually work, or is that just tradition? It works. The 2020 Cochrane review found high-certainty evidence that manual-guided twelve-step facilitation produced higher rates of continuous abstinence at 12 months than other established treatments (risk ratio 1.21, 95% CI 1.03 to 1.42), with consistent findings at 24 and 36 months (DOI: 10.1002/14651858.CD012880.pub2). Read the qualifier: that result is for a clinician-delivered protocol. Peer-led AA on its own performed comparably to other treatments rather than better. The review also concluded AA/TSF probably produces substantial healthcare cost savings, at moderate certainty.

Is SMART Recovery as effective as AA? Better answered now than it used to be. The 2017 systematic review of SMART found only 3 studies evaluating effectiveness and could not draw conclusions (DOI: 10.1037/adb0000237), and that review is still the one people quote. The relevant evidence since is the PAL cohort: a longitudinal study found no efficacy differences between SMART, LifeRing, Women for Sobriety and twelve-step groups (DOI: 10.1016/j.jsat.2018.02.004), and a 2025 pooled analysis of 1,152 people found group choice unrelated to outcomes while involvement strongly predicted them (DOI: 10.1016/j.drugpo.2025.104921). Nobody has run a randomised comparison. Prospective cohort evidence is what exists, and it points to parity.

Can I get sober without any group at all? Yes, and plenty of people do. Medication and individual psychotherapy each have their own evidence base, though no trial has tested them as a deliberately group-free package, so what you have is well-supported components rather than a tested combination. The Cochrane review found professional treatments such as CBT performing comparably to twelve-step approaches on drinking intensity and alcohol-related harm while trailing on continuous abstinence (DOI: 10.1002/14651858.CD012880.pub2). Groups help a lot of people. They are not a requirement.

What if I don't want to quit entirely? Then say so out loud and find a program built for that, because a mismatch between your goal and your program is one of the most reliable ways to end up with no help at all. Moderation-oriented programs exist, and naltrexone in particular is used in approaches that do not require abstinence up front. Talk to a clinician about whether a reduction goal is safe for you specifically, because for people with significant physical dependence it often is not.

Are the secular groups actually as supportive as AA? Their members report higher satisfaction and group cohesion than twelve-step members, despite attending fewer in-person meetings (DOI: 10.1016/j.jsat.2016.10.004). Treat that as suggestive rather than as a ranking, because the alternative-group members were recruited through their own group directors while the comparison group came from a general online hub, and recruiting through leadership selects for exactly the satisfied, involved members the study then measures. It measures how members feel about their groups. Drinking outcomes were never in it.

Related in the Behavior Change Atlas

If you need help right now

If you or someone you love is struggling with alcohol, help is available: call or text the SAMHSA National Helpline at 1-800-662-4357 (free, confidential, 24/7), or find treatment at findtreatment.gov. If you are in emotional crisis or having thoughts of harming yourself, reach the 988 Suicide & Crisis Lifeline (call or text 988). And if you drink heavily every day, talk to a doctor before stopping: alcohol withdrawal can cause seizures and can be life-threatening.

References

All citations verified against PubMed via the PubMed MCP tools at draft time (2026-08-03).

  1. Kelly JF, Humphreys K, Ferri M. Alcoholics Anonymous and other 12-step programs for alcohol use disorder. Cochrane Database Syst Rev. 2020;3(3):CD012880. DOI: 10.1002/14651858.CD012880.pub2 (PMID: 32159228)

  2. Magill M, Nichols LM, Kiluk BD, Alton E, Tartak O, Ray LA. A meta-analysis of cognitive behavioral therapy for substance use disorder: treatment effects by comparator type and consumption and psychosocial outcomes. Behav Res Ther. 2026;203:105101. DOI: 10.1016/j.brat.2026.105101 (PMID: 42361733)

  3. McPheeters M, O'Connor EA, Riley S, Kennedy SM, Voisin C, Kuznacic K, Coffey CP, Edlund MD, Bobashev G, Jonas DE. Pharmacotherapy for alcohol use disorder: a systematic review and meta-analysis. JAMA. 2023;330(17):1653–1665. DOI: 10.1001/jama.2023.19761 (PMID: 37934220)

  4. Zemore SE, Kaskutas LA, Mericle A, Hemberg J. Comparison of 12-step groups to mutual help alternatives for AUD in a large, national study: differences in membership characteristics and group participation, cohesion, and satisfaction. J Subst Abuse Treat. 2017;73:16–26. DOI: 10.1016/j.jsat.2016.10.004 (PMID: 28017180)

  5. Zemore SE, Lui C, Mericle A, Hemberg J, Kaskutas LA. A longitudinal study of the comparative efficacy of Women for Sobriety, LifeRing, SMART Recovery, and 12-step groups for those with AUD. J Subst Abuse Treat. 2018;88:18–26. DOI: 10.1016/j.jsat.2018.02.004 (PMID: 29606223)

  6. Zemore SE, Lui CK, Mericle AA, Li L, Martinez P, Timko C. Second-wave mutual-help groups: examining effectiveness for individuals with alcohol use disorders in the longitudinal, U.S. national PAL Study cohorts. Int J Drug Policy. 2025;147:104921. DOI: 10.1016/j.drugpo.2025.104921 (PMID: 40750480)

  7. Beck AK, Forbes E, Baker AL, Kelly PJ, Deane FP, Shakeshaft A, Hunt D, Kelly JF. Systematic review of SMART Recovery: outcomes, process variables, and implications for research. Psychol Addict Behav. 2017;31(1):1–20. DOI: 10.1037/adb0000237 (PMID: 28165272)

  8. Amin-Esmaeili M, Farokhnia M, Mojtabai R, Leggio L, Johnson RM, Susukida R. Evaluating reduced use and abstinence as outcomes in pharmacotherapy trials for stimulant use disorder: a meta-analysis of 12 randomized controlled trials. JAMA Psychiatry. 2026. DOI: 10.1001/jamapsychiatry.2026.1092 (PMID: 42234418)

Limitations stated honestly: The Cochrane AA review's high-certainty finding rests on 2 studies for the 12-month continuous-abstinence outcome and applies to manualized twelve-step facilitation rather than to peer-led AA, which are not the same intervention; its other estimates range from very-low to moderate certainty and the article now states each grade where it quotes one. The review also excluded coerced participants, so it says nothing about court-mandated attendance. The Magill meta-analysis carries near-total heterogeneity, and its add-on effect sizes have confidence intervals that nearly touch zero. The medication numbers describe average trial populations rather than any individual reader, their confidence intervals are wide, and both drugs have real contraindications. The Zemore 2017 comparison is cross-sectional and recruited the alternative-group members through their own group directors, which selects for satisfaction. The PAL longitudinal studies are prospective cohorts, not randomized trials, recruited the same way, and the 2025 authors themselves counsel caution about the SMART-specific result because SMART's program has varied across time and geography. No randomized comparison of AA against the secular alternatives on drinking outcomes exists. The SMART-specific review is nine years old. The reduced-use meta-analysis concerns stimulants rather than alcohol, used trials conducted 2001–2011, measured its two outcomes by different methods, and its reduced-use figure includes the abstinent group. The alcohol-specific evidence on drinking-level reduction is pointed to in the body but was not reviewed to citation standard for this page, and is flagged as such. None of this substitutes for an individual clinical assessment.

Disclosure: I am the founder of IGNTD, which offers non-abstinence-based recovery programs. It is listed above as one option among several, and I have no financial relationship with AA, SMART Recovery, LifeRing, Women for Sobriety, or any medication manufacturer.

Written by Dr. Adi Jaffe, PhD (UCLA), author of The Abstinence Myth and Unhooked. Last medically reviewed: 8-8-26.