AA and the compliance effect

Establishing causal relationships is harder than it at first seems. Often tantalising correlations are observed that people then jump on to posit cause erroneously.

For example, studies showed that people who consumed a fish oil pill daily had a greater life expectancy. This seems plausible right? Maybe the Omega-3 is helping in some way. However, the only way to know for sure is to do a randomised trial — that is to say, follow one group of people who take fish oil daily, and another “control” group who do not but are otherwise more of less identical to the first control group (same age ranges, same geographic location, wealth bracket, weight ranges, gender distribution etc).

If fish oil really is the causative agent of the extra life expectancy, this kind of trial will show it. The Cochrane group did this trial and found the fish oil supplements did……nothing. That’s right, they made absolutely no difference to the life expectancy of either group.

So why did other studies seem to show that people who took fish oil lived longer? The answer is the “compliance effect”. These were just observational studies, not randomised trials, and the people who took the fish oil daily opted to do so (they “self-selected” in the parlance of these trials), they “complied”.

Let’s pause for a moment…what kind of people would choose to take fish oil supplements daily if taking fish oil was the generally perceived healthy thing to do of the time? …People who cared a lot more than the average Joe about their health of course!

This was the key. These people opting to do this supposed healthy behaviour cared a lot about their health, which also means they are more likely to do other healthy behaviours including exercising more and eating healthier foods, going for medical checkups more often and so on. It is this that actually causes the increased life expectancy.

Even more shockingly, in a widely studied case, numerous epidemiological studies showed that women taking combined hormone replacement therapy (HRT) also had a lower-than-average incidence of coronary heart disease (CHD), leading to the conclusion that HRT was protective against CHD. But randomized controlled trials showed that HRT actually caused a small but statistically significant increase in risk of CHD. Why was there a correlation in in the epidemiological studies then? It turned out that women taking HRT were more likely to be from higher socioeconomic groups, with better-than-average diet and exercise regimens. The use of HRT and decreased incidence of coronary heart disease were coincident effects of a common cause (i.e. the benefits associated with a higher socioeconomic status), rather than a direct cause and effect, as had been supposed.


How does this relate to AA you may wonder? Well, most of the studies that seem to support AA are observational studies. There seems to bean undeniable correlation that increased involvement in AA goes with more sobriety (here sobriety is defined by the percent days abstinent and the drinks per drinking day, and involvement with AA is measured by things like meeting attendance, having a sponsor, spirituality adopted, reading associated literature). However without a control group is impossible to directly posit that AAI (AA Involvement) is the causative agent of that sobriety rather than just correlated with it.

The addicts who surrendered (noticed the similarity of this word to “complied”) and eventually did everything on the AA checklist, warts and all, most likely got so desperate that they’d do anything to be sober. They become utterly willing without exception or excuse to change their lives. If their sponsor told them to pray while standing on their heads each morning whilst wearing a tutu they would have done at that point. However it may be that it is the willingness itself that was the true causative agent. The addicts who really got to the brink and swallowed the whole pill, most likely finally cared enough about their addiction to also make life changes universally. In other words both sobriety and AAI could be coincident effects of the true causative agent that was driving sobriety, utter desperation and willingness to do whatever it takes to get clean

In my opinion this invalidates most of these observational studies of AA’s effectiveness. It is important to note that it doesn’t mean AA isn’t effective either however. It may well be the causative agent and the correlation is a hint at that. It simply means that the studies are fairly useless to determine if it is or isn’t without carefully controlled randomised trials.

It’s also important to note that pretty much the same objections could be thrown in the direction of psycho-therapeutic approaches too.


Is AA involvement just fish oil for the soul then? The old-timers that survive the process and reach sobriety were just the most willing who finally become willing enough to do anything and who would have succeeded no matter what they had done to get sober at that stage? They could have white-knuckled it, gone to MET, CBT, counselling, whatever and as long as they had that willingness they would have made it to sobriety regardless. Their treatment and their sobriety are just coincident effects that correlate, but their underlying willingness and desperate drive to change was the true cause of both.

I’m of the opinion that AA is of value. Most of the things it offers seem fairly sensible tools for an addict:

  • A social network for addicts who are often very isolated and in need of a bunch of non-drinkers to fill the void.
  • A sense of purpose and that they can be worthwhile members of community via helping other addicts and doing acts of service such as being the treasurer or even just making the tea and coffee
  • A network of people who will offer objective, impartial advice that could perhaps make the addict see through the fog.
  • Some relief of the shame by knowing you aren’t the only one who did or thought X
  • Tools very similar to CBT to get more aware of triggering thoughts, places, people and things
  • Regular reminders of their goal when the tide of life often makes one forget this

and many other things. They are pretty common sense things an addict needs and very plausible.


Having said that, is AA perfect? By no means. I think the compliance effect mentioned above means that along with the good elements, addicts can often end up taking on a lot of “fish oil” along with the actual “cure”.

A downward spiral of shaming can happen. Initially an addict may not be fully willing to change and do anything, so they refuse to do action X (say praying on their knees) because they don’t believe it will help and they still aren’t quite desperate enough to “just do whatever”. However because they’re not yet desperate enough they have a relapse. Again a correlation is presented to them and it’s just too tempting (especially in the pain of relapse) to re-write this as “not doing X caused the relapse”, when in fact it may have been just the lack of willingness to change or any of many other factors. Their sponsor will tell them “they are in denial” or “not fully surrendering”, and guilt them into doing X that they wouldn’t previously do saying things like “all sober members do X, and you can’t see what’s best for you”. After a few iterations of this, the addict has either given up or he’s staying sober doing the full AA checklist and attributing doing X to being a cause of his sobriety along with everything else.

It’s probably most likely that this checklist consists of both things that do genuinely aid the addicts sobriety (say changing his social network from his old addict friends to a new group of people who don’t engage in drinking so much), and also some junk that has only thinnest of plausibility connections to staying sober (such as doing 10 Hail Mary’s each day). He will now enter the ranks of the old-timers and people will point newcomers to look at him as further evidence that doing X equals sobriety and the belief is reinforced ever further down the generations.


Summary: it’s very hard as someone in active addiction to work out the parts of AA that are fish oil and the parts that are genuinely helpful. Most addicts by the time they hit their bottom will be happy just to swallow the lot, after all if X is just a placebo then it’s not gonna do them any harm. They’ll get sober that way and that’s all that matters in the end. There may be things that they really really do need to do that they are unwilling to do and being on that side of the fence seems far more harmful in the long run.

I think as long as the actions expected are not harmful and don’t compromise your principles then this is probably a minor worry. But if you know in your heart that X is something that truly won’t help your recovery or may even harm your recovery then I think there is no reason to be shamed into doing it. If a sponsor said “going to any lengths means spending $30k on a month in a TSF rehab” and that would leave you destitute and unhappy. If you are being guilted into 14 meetings per week and have no time to actually build a happy life or social network outside of AA. Or discouraged to stop seeing your therapist or taking meds that you need. If your family and relationships are suffering and life seems to be getting worse. Don’t be afraid to toe the middle line….It’s a balancing act I think. Going to one extreme and quitting or severely slacking off from your recovery is not a good idea and you have to able to be honest enough with yourself about that to not rationalise it, but equally becoming an AA extremist to the point were your life actually seems less fulfilling would probably not help anyone.