In the December, 2015, edition of this Science Update we responded to a recent article appearing in The Atlantic.1 Its author, Gabrielle Glaser, claimed that AA and its 12-step programs lack scientific foundation, asserting that “nothing about the 12-step approach draws on modern science …..” We presented the data supporting the opposite case, citing several published scientific reports that she did not mention. In the present installment, we review the basis on which she asserts her claim that the success rate of AA is only 5-8 percent. Relying on a single secondary source2 for this claim, Ms. Glaser writes, “That is just a rough estimate [of AA’s effectiveness], but it’s the most precise one I’ve been able to find.” Because flawed science can cause harm, we offer a critique of the scientific basis she cites for her claim.
At the outset, Ms. Glaser’s source presents neither new data nor any of the studies we have cited that report first-hand observations targeting AA’s effectiveness. Rather, her source itself refers only to data gathered by others, mostly for purposes other than judging AA’s effectiveness. This forms the basis for three separate, questionable, calculations that arrive at the 5-8% figure. In each calculation, all dropouts—counted after as few as one AA meeting—are treated as AA failures. By analogy, this seems to us like counting insulin for diabetes as a failed treatment after only one insulin injection. In our view, looking at outcome rates for active AA members offers a more accurate estimate of AA’s effectiveness. But let us examine the 5-8% figure.
In the first calculation, The Atlantic article’s source multiplies a 25% AA attendance figure by a 22% abstinence figure to arrive at a 5.5% estimate of AA’s effectiveness. Where do these figures come from? Another second-hand source3 that also cites the work of others: two publications from the Rand Corporation that examined, among other things, attempts at controlled drinking and offered little focus on AA’s effectiveness. At 4-year follow-up the Rand group identified patients with at least one year abstinence who had been regular members of AA 18 months after the start of treatment: 42% of the regular AA members were abstinent, not the “calculated” 5.5% figure. The Rand Reports are public and both Ms. Glaser and The Atlantic editors could have read them rather than rely on a third-hand source.
The second calculation repeats the 25% AA attendance rate multiplying it by another “abstinence rate” of 21%. This rate is taken from an article by Harris and colleagues4 who surveyed 150 alcoholics entering a residential treatment program because they were not abstinent. Based on the reports of those entering, the study concluded that the sample did “not represent ‘typical’ AA recruits.” Despite this, the third-hand calculation method uses two percentages lifted out of context from the Harris study—16% who had reported ever taking at least one step of the 12-step program divided by 75% who had ever attended an AA meeting—and gives a figure of 21%. This calculation has no bearing on abstinence from alcohol, nor does it apply to AA participation over time. Ms. Glaser and her editors at The Atlantic might have looked into these available data in greater detail in the interest of accuracy.
The third calculation applies the 21% “abstinence” rate claimed above to an alleged 40% sustained abstinence rate noted in yet another report, a paper by Fiorentine (1999).5 Ms. Glaser’s source quotes Fiorentine as writing “’approximately 40 percent of individuals categorized as having continued active participation in AA maintained high rates of abstinence.’” Our reading of Fiorentine’s paper fails to find any such statement. Curiously, Fiorentine reports on a study of drug addicted individuals, only a portion of whom were identified as having an alcohol problem, to offer an estimate of the success of AA. That being said, the data Fiorentine presents is as follows: 77.7% of individuals who attended AA 12-step meetings at least weekly reported being free of drug use for 6 months prior to a 24-month follow-up, a finding corroborated by urinalysis at the time of the interview, and 74.8 % reported being free of alcohol use during the same time period. These figures suggest that a high observed abstinence rate is associated with regular participation in AA. Neither The Atlantic editors nor Ms. Glaser indicate an awareness of these factual discrepancies.
Let us summarize the science that The Atlantic article relied upon to make the assertion of a very low success rate of AA. The Harris data do not address either current abstinence or AA involvement over time going forward. Both the Rand Reports and the Fiorentine study, respectively, note that individuals in 12-step groups had high rates of self-reported sobriety: 42% at four years and 75% between 18 – 24 months respectively. These rates are far higher than 5-8%. These data all reside in publicly available scientific reports. In the interest of scientific accuracy, The Atlantic had access to them just as we have had.
Instead, The Atlantic relied on an author who used a single source that relied on questionable calculations that appear to us to have clouded, rather than clarified, the true data. We look forward to better science from The Atlantic and are always pleased to further the cause of better science.
Is a research psychologist with the Denver Veterans Affairs Medical Center, Laboratory for Clinical and Translational Research in Psychiatry and is an Assistant Clinical Professor of Clinical Psychology in Psychiatry at the University of Colorado Anschutz Medical Center. Additionally, Dr. Emrick cares for private patients, with his specializing in the treatment of addictions, mood disorders, and posttraumatic stress disorder (PTSD). He has authored more than 25 articles and book chapters on the treatment of alcoholism and other addictive disorders, with a significant portion of his publications dealing with research on Alcoholics Anonymous. He is a charter member of the Research Society on Alcoholism and is a member of Sigma Xi. He has lectured on the treatment of alcoholism throughout the US as well as in Canada, England and the former Soviet Union. He has appeared on the Oprah Winfrey show and Nightline as an expert on alcoholism treatment. For nearly 35 years, he served on the medical staff of the Denver Veterans Affairs Medical Center, working mostly in the substance abuse program. He was director of that program for nearly nine years, after which he became the medical center’s first specialist in coordinating the treatment of veterans who suffer from both PTSD and a substance use disorder.
Is Professor of Psychiatry, with tenure, at the University of Colorado School of Medicine and Physician at the Department of Veterans Affairs (DVA) Medical Center, Denver. Trained in psychiatry at The Cambridge Hospital/Harvard Medical School, Dr. Beresford studied directly with George Vaillant, M.D., author of the landmark study The Natural History of Alcoholism. For over thirty years Dr. Beresford has focused his professional efforts on the medical/ neuropsychiatric problems due to alcohol use disorders, especially as seen in medical and surgical wards and clinics. His systematic, empirical studies opened liver transplantation to a stigmatized patient group who might not otherwise have had access to this life saving procedure. Dr. Beresford’s newest book, Psychological Adaptive Mechanisms, (Oxford University Press, 2012) breaks new ground in taking observed psychological adaptation into the clinic as a useful clinical tool. This has wide ranging application in alcoholism research owing to property of uncontrolled ethanol use to move humans from high level adaptive Maturity to much more rigid and ineffective Immature level adaptations.