Setting the AA Record Straight Part II: Myths and Facts About the Program
Part 1 of “Setting the AA Record Straight” attempted to sort out culpability for problems attributed to Alcoholics Anonymous, often unfairly, as well as to explain how AA and its 12 steps are used in various settings – all of which contribute to misconceptions about AA. Distinctions were made among AA support groups, which are not formal addiction treatment,12-step treatment as it goes on in traditional treatment programs, and research-based 12-step facilitation (TSF) – a professional treatment approach designed to facilitate (not pressure) clients to become involved in AA.
It’s important to understand these distinctions as Part 2 of this series addresses “Myths and Facts About the Program.” Any time issues are raised concerning AA, particularly when they involve research, it’s important to take into account which of the three different “presentations” of AA are under consideration, something often lacking in journalistic reporting.
Myth: We can’t prove AA’s efficacy and put it to the scientific test largely because it’s “anonymous.”
- Facts: While it’s true that some of AA’s operational rules make it difficult to study and recruit research participants, particularly as it exists in community groups, renowned addiction researcher and University of New Mexico researcher and professor emeritus William Miller, Ph.D. stated, “Anonymity is a normal condition in clinical research, and AA members can consent to participate in studies and have their anonymity protected. AA founder Bill W. welcomed research on AA and encouraged members to participate.” Dr. Miller referred to an official “Memo on Participation of AA Members in Research and Other Non-AA Surveys” from AA World Services that encourages such participation and states, “As long as there is frank communication and attitudes of open-mindedness and flexibility, it has proved possible to work out ways of participating in research which do not require A.A. members to compromise A.A.’s Traditions and which permit the researcher to arrive at valid findings.”
Anonymity is a normal condition in clinical research, and AA members can consent to participate in studies and have their anonymity protected.-William Miller, Ph.D.
Myth: Very little research has been conducted on AA – of the research that supports it, almost none meets acceptable scientific standards.
- Facts: Over the course of the last decade or so, a substantial amount of respectable research on AA has been published in what Harvard University’s John Kelly, Ph.D., has referred to as an “empirical awakening” for recovery support groups. William White, M.A., author of Slaying the Dragon: The History of Addiction Treatment and Recovery in America, has noted that, while many in the scientific community expected “AA and AA-oriented addiction treatment approaches to be blown out of the water, the growing body of more rigorous studies on AA has revealed that what started out as folk wisdom is now turning out to be pretty good science as well.” (Most of the research on 12-step groups has been on AA – little is known about Narcotics Anonymous or the 12 steps as applied to problems such as eating or sexual disorders.)
AA and Selection Bias
Authors of an important recent study published in the journal Alcoholism: Clinical & Experimental Research and described in detail in a NY Times article earlier this month, found a statistical way to study the impact of AA meeting attendance uninfluenced by what’s called “selection bias” – that is, not influenced by the fact that the people who attend AA may be more motivated from the outset, thereby teasing out the influence of AA itself. They pooled the results of multiple well-designed studies in which patients with alcohol use disorders were either randomly assigned to a form of treatment designed to increase participation in AA or assigned to a comparison condition that did not aim to facilitate AA involvement. The researchers then looked at percent days abstinent from alcohol as well as frequency of attending AA meetings, comparing abstinence rates with what they were upon starting each study. They found that an increase in attendance was significantly associated with an increase in days abstinent at three and 15 months, noting that attendance at an additional two AA meetings each week would be associated with about three and a half more days of abstinence per month. Because the study involved random assignment of people to an AA condition versus no AA condition, they concluded, “AA participation had a genuine benefit that was not attributable to self-selection bias.”
I asked Stanford University’s Keith Humphreys, Ph.D., lead author of the study, whether this was a test of AA in the “real world” since it involved people who were introduced to AA via research studies. He replied, “Millions of people have come to AA from referrals from health practitioners, so AA in the real world includes many such people. We cannot assume that the results are the same for those who go directly, but neither should we assume they are different.”
I’ve read a number of these studies, ones published in highly regarded peer-reviewed scientific journals and authored by respected researchers. Not claiming for one moment to be a researcher myself, I don’t always agree with their conclusions. For instance, noting high dropout rates over time in a number of these studies, I’ve questioned conclusions about encouraging more people to attend AA without also encouraging them to explore other forms of support. And I’ve been troubled by the dearth of research on non 12-step support groups such as SMART Recovery. However, it’s simply not true that practically none of these studies meet acceptable scientific standards.
While it is true that there’s next to no research in which people have been randomly assigned to take part in AA meetings versus other approaches, there have been at least eight randomized controlled trials in which 12-step facilitation treatment (TSF) was compared to and generally found to be at least as effective as other professional approaches that didn’t involve the 12 steps. Unfortunately, even though national surveys reveal that most treatment facilities report using TSF, experts in the field point out that when counselors attempt to involve their clients in 12-step self-help groups, they rarely use scientifically supported methods.
As for other types of studies, Dr. Miller points out, “There is a large correlational literature generally showing a modest positive relationship between attendance at AA meetings and abstinence.” Overall, this research suggests that people who go to meetings frequently are more likely to become abstinent and remain abstinent over the short and long term. And at least several studies indicate that clients in treatment programs have better outcomes when they also attend AA meetings. Some research also indicates that being involved in AA may considerably reduce the need for more costly professional care. It’s not just showing up at AA meetings that appears to be important – involvement in the program appears to make a difference. That includes things like having a sponsor, reaching out to others for help, and beginning to work the steps.
However, just because two things are associated (e.g., increased AA attendance and increased abstinence rates) does not mean that one causes the other. Thus, researchers and clinicians have heatedly debated whether this correlation reflects AA’s effectiveness or just says something about the people who choose to go to AA – for instance, that they’re more motivated to change than those who don’t go to 12-step meetings.
Myth: AA’s key components, such as peer support and spirituality, “don’t lend themselves to scientific metrics.”
- Facts: For some time now, respected researchers have been trying to tease out what it is about AA that helps those for whom it’s beneficial. Such studies have examined AA’s impact on social network support; effects on coping skills, negative emotions, and confidence in dealing with social high-risk drinking situations; impact on recovery motivation over time; and the relationship between increasing spirituality and stress management. According to Harvard’s John Kelly, Ph.D., who has conducted a number of these studies, “In all of these areas, AA involvement had a positive impact. And these studies were rigorously designed, with large samples, lagged (to ensure cause and effect), using psychometrically validated measures, and published in top addiction peer-reviewed journals.” According to Dr. Kelly, these studies included people from different types of treatment and programs across the U.S. Some came from rigorously controlled randomized controlled trials and some from residential and outpatient treatment. But the research design took measures to rule out “treatment effects” to get the unique effects of AA participation on outcome.
Does this mean AA is helpful for everyone? No, but such studies show scientifically that AA can be helpful, and how it helps those it does help.
Myth: AA has a success rate of between 5 and 10 percent.
- Facts: Dr. Miller believes that this figure is based on A.A.’s own anonymous survey data that “no one would accept as outcome research.” He said, “It’s a stretch even to interpret the data as a 90 to 95 percent drop-out rate, though that’s what is often claimed: of newcomers to AA, 90 to 95 percent won’t be there a year later.” Also noting that dropout from AA does not mean failure, he cited long-term follow-up research by his colleague, Scott Tonigan, Ph.D. at the Center on Alcoholism, Substance Abuse, and Addictions at the University of New Mexico, revealing that some people no longer attending AA still regard themselves as members, have internalized the 12-step program, and remain abstinent.
…some people no longer attending AA still regard themselves as members, have internalized the 12-step program, and remain abstinent.-William Miller, Ph.D.
When I was writing Inside Rehab, Tonigan – whom Miller refers to as “one of the most knowledgeable and prolific reviewers of the AA literature” – had just completed one such review and concluded, “Overall, studies suggest that between 55 and 80 percent of alcoholics encouraged to attend AA while in treatment will stop attending AA within nine months.” That stated, attendees do commonly go to meetings for a short time, drop out, and then come back later. And, in all fairness, early and high dropout rates are not unique to 12-step groups. Dr. Kelly noted, “It should be remembered that many people don’t benefit from cognitive-behavioral therapy either, and nearly half drop out within 90 days of starting treatment.”
But this still doesn’t tell us about AA’s “success rate” in helping people recover. Some insights can be gained from a study conducted by Tonigan and colleagues Under a National Institutes of Health/National Institute on Alcoholism and Alcohol Abuse grant that included more than 250 adults who had been to AA support groups in the community and in outpatient treatment. About 40 percent of them reported continuous AA attendance for 12 months. Of those who attended AA regularly during that time nearly half reported complete alcohol abstinence at the 12-month follow-up interview. Of the 60 percent of participants who discontinued AA attendance during the same 12-month period, about 30 percent reported complete alcohol abstinence at the 12 month interview. The difference in abstinence rates at 12-months between the regular AA attendees (48 percent) and AA dropouts (30.5 percent) is statistically significant. (These data come from personal communication with Dr. Tonigan; data from this study have been used in several published studies on different topics.)
A number of people expressed anger and indignation at Part 1 of this series, with some suggesting that I had “turned” into an AA supporter, but my views really have not changed. Part of my job is to follow the science, report on it, and try to clarify misconceptions relayed by the media. Much of what’s stated in this article was already stated (with somewhat different wording) in my book, Inside Rehab.
- Did AA work for me? Not particularly. Although it provided support when I first sought help many years ago, it wasn’t my primary route for resolving an alcohol use disorder.
- Is AA misused and misappropriated? Often.
- Is AA wrongly forced upon lots of people and part of one-size-fits all treatment in the US, thereby harming people? Yes.
- Does AA help a lot of people? Yes.
- Is there good research supporting its benefits for some people? Yes.
- Do we need more research on alternatives to AA? Indeed.
- Is it important to educate people to think critically about this? Absolutely.