Alcoholics Anonymous and other programmes: what the research says

“I’m not going to a group where you have to stand up and say you’re powerless. That’s exactly why I came to you — so that I wouldn’t be.”

I understand him. And at the same time I know this article will have to be more honest than is convenient for my own trade. Because when I look at what the research of the last few years says about mutual-help programmes, I get an answer that we therapists did not expect.

For a long time Alcoholics Anonymous was talked about as something between religion and folklore. Kind, possibly useful, scientifically not serious. That judgement no longer holds. Not because the programme changed, but because we finally learnt how to measure it at all.

What Alcoholics Anonymous actually is

Let us first be clear about what we are discussing, because people imagine all sorts of things.

Alcoholics Anonymous is not a health service, is not therapy and has no professional staff. It is groups of people who meet regularly, free of charge, without membership and without records. The programme consists of twelve steps — a sequence of personal actions running from admitting that things have got out of hand, through an inventory of one’s own conduct, to making amends and to helping someone else. Groups fund themselves from their own contributions and hold no opinion on anything outside their subject.

In Slovenia there are AA groups, and alongside them something that barely exists elsewhere: clubs of treated alcoholics, which grew out of Vladimir Hudolin’s model and have run here for decades. Their distinctive feature is that it is not only the person who drinks who attends, but at least one family member as well. From a family therapist’s point of view that is rather clever — the reasons come a few paragraphs on.

Why this was so hard to research

Imagine wanting to prove that exercise helps. You cannot force people to stop exercising. So you compare those who go with those who do not — and get a result that tells you nothing, because the people who go were different to begin with. More motivated. Healthier. They have time.

Mutual-help groups are the same. Whoever attends meetings regularly was, as a rule, more determined beforehand. So for years every study suffered from the same flaw: it was not measuring the programme, it was measuring the people who choose it. We call this self-selection bias, and in voluntary programmes it is the hardest obstacle there is.

There were two solutions. First: instead of the groups themselves, researchers randomised a professionally delivered encouragement — a short, manualised approach that points a person towards the twelve-step programme and helps them get in. That can be compared with cognitive behavioural therapy or motivational interviewing, because all three are professional acts. Second: statistical methods that try to subtract self-selection. I have written about how hard talking treatments are to research in general in the piece on why therapy is harder to study than a tablet.

What the results showed

In 2020 a Cochrane review appeared, pooling twenty-seven studies with several thousand participants. In medicine Cochrane is something like a supreme court for evidence — conservative, slow and inclined to say “we know too little”.

This time it did not say that. It said that professionally led encouragement into a twelve-step programme is, on one outcome, better than established psychological approaches, cognitive behavioural therapy included: continuous abstinence at one year. Roughly four in ten against roughly three in ten. It is not an enormous difference, but it is consistent, and it is rare in this field, where approaches usually turn out to be equivalent.

Now the second part, which enthusiasts like to leave out. On all the other outcomes measured — how many days a person drank, how severe the consequences were, how mental health fared — there was no difference between approaches. Just as good, not better.

And the third part: cost. Because the groups are free and run without a professional, the overall cost of care in this approach was considerably lower. For a public health system that matters more than it sounds.

What in a group actually works

For a therapist, more interesting than whether it works is the question of what in it works. Research into mechanisms points to three things above all, and none of them is spiritual.

The first, and by far the strongest, is a change in social network. A person who drinks usually has around them people around whom drinking happens, and a timetable organised around it. A group offers another network — people you meet at the riskiest hours, whom you do not have to ask for anything. When researchers measured how much of the effect ran through changed company, the answer was: a great deal.

The second is self-efficacy: the belief that in a concrete situation you can act differently. That does not come from advice; it comes from watching someone like you who managed it.

The third is less tangible and the one closest to me: in a group, shame goes down. Not through comfort, but because someone tells their story and nobody in the room looks away. I have written about why shame is so different from guilt, and why on its own it repairs nothing, in the piece on shame.

The other programmes that exist

The twelve steps are not the only option, and it helps to know what else there is.

SMART Recovery is a secular programme built on cognitive behavioural and motivational foundations. Instead of admitting powerlessness it works with skills: recognising triggers, working with thoughts, managing urges, weighing decisions. It is run by trained volunteers. There is less evidence than for AA, mainly because it is younger, but it is encouraging; in a study comparing four programmes, the outcomes differed less than anyone expected.

LifeRing and Women for Sobriety are also secular; the latter was created specifically for women and emphasises self-worth rather than powerlessness, which suits many women better.

Clubs of treated alcoholics here are their own thing. They work with the family, not only the individual, and they have a professional attached. For a family therapist the logic is obvious: if only one person changes, the system around them often pushes that new way back. When the whole family learns together, there is less of that pull backwards.

The twelfth step: why helping helps you

The last of the twelve steps is about offering your experience to someone who is where you once were. To an outside observer that looks like a nice gesture and nothing more. It is not.

Research has found that those who take on this role in a group — sponsoring, chairing a meeting, phoning a newcomer — hold on better than those who only attend. The phenomenon even has a name: the principle that helping benefits the helper. There are probably several reasons. When you explain to someone what you have learnt, you tell it to yourself as well. When someone rings you at eleven at night, you have a reason to be reachable and sober. And you acquire a role that is not “the problem person” — which, for someone who has held that role at home and at work for years, is a large change.

For a therapist that is a useful reminder. People do not recover only by receiving. They recover by becoming useful to someone again.

What actually happens at a first meeting

Because the imagined version is usually worse than the real one: at most meetings nobody has to say anything. You give a name, or you just listen. Nobody keeps a register, nobody rings anyone, contributions are voluntary. Some meetings are closed to those with the problem themselves, others are open to relatives too.

The hardest part is generally the car park outside. That is not a joke — more people give up there than over the content. Knowing it in advance makes it slightly easier.

The higher power part

Now to the sentence that stops most people.

The twelve steps do speak of a higher power and of surrender. For someone with no relationship to faith, that can be a big enough obstacle never to return. What is interesting is what the mechanism research found: the spiritual dimension is among the weaker ingredients in predicting outcome. What makes the difference is the company, not the belief.

In practice people interpret “higher power” very differently — for some it is the group itself, for others the idea that they are not the centre of the world. And those for whom even that reading does not sit have secular alternatives. Nobody is obliged to accept a language they do not understand in order to get help.

Where my enthusiasm stops

Now the section that argues against everything above.

First: the good results apply to alcohol dependence and should not simply be transplanted elsewhere. In opioid dependence, the treatment with the best evidence is medication prescribed by a doctor. Some groups reject that treatment as “still a drug” — and when they do, they do harm. If anyone in a group urges you to stop prescribed medication, that is bad advice and you should not act on it without your doctor.

Second: mandated attendance does not work. When a court or an employer sends someone to a group, the effect disappears. The programme rests on being voluntary, and that cannot be skipped.

Third: a group is not therapy. It does not treat trauma, does not resolve a marriage, does not work on depression. Many people need both, and where another mental health condition sits alongside, a group alone is usually not enough.

And fourth, most simply: a large share of people do not get on with groups. Drop-out is high and a first visit is unbearable for many. That is not a sign that you cannot do it — it is a sign that you tried one form out of several.

How to use this

If you are wondering whether to go: attend three different groups before you decide. Groups differ from each other more than programmes do — same organisation, entirely different atmosphere. A first visit tells you about that room, not about the method.

If you are in therapy and your therapist steers you away from groups without a reason, ask why. The evidence is such that they should have a good answer. What else is worth asking at the start I set out in the piece on choosing a therapist.

And if you are close to someone who drinks: the treated alcoholics club is one of the few doors here that you can walk through yourself, even while he is going nowhere.

What to remember

Alcoholics Anonymous is better researched than most people think, and on continuous abstinence it does better in reviews than established psychological approaches — on other outcomes just as well, not better, and at lower cost.

What works is mainly the change in the company around a person, not the spiritual side. So a secular alternative, when the programme’s language does not fit, is a perfectly legitimate choice, not a half-measure.

In opioid dependence, groups do not replace medication prescribed by a doctor. Coercion does not work. And a group is not therapy — but it is often what therapy lacks: people who are reachable on a Friday night.