The Neuroscience of Alcoholics Anonymous

My last two posts looked at what bottom-up research on the brain’s response to simple threats can tell us about clinical work with anxious patients. This week I’ll shift viewpoints to consider an ambitious attempt by the psychiatrist Marc Galanter to use social and cognitive neuroscience to explain the healing that takes place in Alcoholics Anonymous. While necessarily incomplete, his model provides a picture of the recovery process that should be useful to clinicians.

I spent many formative years of my career at a private psychiatric hospital working with a talented group of psychiatrists and other professionals well above the average in training, skills, and commitment to helping patients. At one of my first case conferences there, the staff pressed an alcoholic school principal to explain why she drank. The emotional tone was one of shock and disapproval that someone of our social class could degenerate and harm her mind, body, family, and career. My colleagues recommended in-depth psychotherapy to resolve the issues underlying her behavior. When I suggested AA, they poo-pooed it as a “self-help program.” That angered me enough to remark that if they didn’t want a patient like this to go to AA, they needed to come up with a treatment that would provide all the things AA provides.

Just what does AA provide to facilitate recovery? Since 1935, when Bill Wilson helped Dr. Robert Smith get sober, and the two of them founded AA, many theories have purported to explain the fellowship’s healing. Proposed factors include self-efficacy, social networks, personal narratives, reducing negative affect, and the always-contentious issue of spirituality. AA remains controversial— The Atlantic last year published an article entitled, “The Irrationality of Alcoholics Anonymous.”

Marc Galanter, a leader in the study of alcoholism and recovery for forty-five years, has marshalled recent neuroscience findings to suggest what goes on in the brain of a person recovering in AA. Such a project seems fraught with peril—extrapolating laboratory and neuroimaging findings to explain a complex social group phenomenon runs the risk of specious conclusions. As I have shown in previous posts, even the simplest responses to threats involve at least several circuits in different parts of the brain. Addiction and recovery are vastly more complex—they occur at levels ranging from the molecular and genetic to the psychology of large groups, and any explanation has to leave out a great deal. Galanter has created a useful picture of recovery and in the process illustrated the possibilities of applying today’s evolving neuroscience to the complexities of patient care.

Galanter summarizes the pathophysiology of the mesolimbic dopaminergic tract, the reward system central to all addiction, and goes on to note that neuroscience is also clarifying how addicts get better: cognitive therapy activates areas of the prefrontal cortex associated with cognitive control and regulation of emotion and reduces activity in limbic regions associated with craving.

Mutuality is central to AA—one popular slogan is, “Identify, don’t compare.” Galanter explores this by referring to research on mirror neurons, which were discovered in monkeys and are thought to function similarly in humans. They fire when an animal observes another animal’s behavior and are thought to underlie the finding that observing another person’s experience of pain or reward activates neural areas in the observer associated with those same experiences. While it makes sense to think mirror neurons and such “emotional empathy” underlie an AA member’s identification with her sponsor’s success in recovery, it would be a great oversimplification to see mirror neurons alone as an explanation for the healing power of identifying with a sober mentor.

Galanter notes that other brain sites are involved in the cognitive process of recognizing mental states in one’s self and others. Such “mentalizing” requires an ability to appreciate the emotions of other people and to recognize the difference between self and other. These cognitive aspects of empathy have been associated with activity at several cortical sites. I would add that the ability to identify with others, either in a close, long-term relationship with an AA sponsor or therapist, or with speakers relating their experiences of addiction and recovery from the podium, requires some permeability of the self’s boundaries and an openness to sharing. This requires a sense of safety. In simple terms, the amygdala, which handles threats and is involved in generating the emotions of fear and anxiety, has to be calmed down. The rituals of AA as well as the Twelve Traditions help maintain such an environment of safety, and it would be interesting to look at the brain mechanisms involved.

Recovery in any social program involves learning, which requires the creation of new memories and the modification of old ones. These involve the hippocampus and parts of the prefrontal cortex. Galanter notes that some learning in AA is likely to take place outside of conscious awareness, and that procedural memory for skills or habits that are executed unconsciously is mediated by the ventral striatum. I would add that this region is also a component of the brain’s dopamine reward system, which, as I noted earlier, is involved in addiction. Galanter also suggests that the consistent ideology of the A.A. fellowship is important for transmitting transmit specific beliefs to the recovering person.

He then turns to cognitive psychology to discuss schemas, which are mental formations that organize categories of information and their relationships. Research suggests roles for the hippocampus and the left prefrontal cortex in integrating meaning between related events. In what strikes me as something of a leap, he suggests that these processes may contributed to the A.A. member’s new conceptualization of a sober self, derived from A.A. experience. (He notes that this could apply to psychotherapy as well.) He also discusses the role of storytelling, including the self-disclosure involved in recounting one’s own history of addiction and recovery. Self-disclosure is associated with activation of parts of the mesolimbic reward system, which we again recognize as central to addiction.

Galanter notes that the twelve-step programs, unlike most professional psychotherapies, espouse moral values beyond abstinence, including taking responsibility for the harm one has caused and making amends. Both conscious and nonconscious processes are involved in the attribution of moral significance to one’s experience; the former involve various cortical areas, and the latter also the amygdala. AA also constitutes an in-group, and activation of the anterior insula is associated with helping in-group members.

Puzzlingly for me, Galanter discusses AA’s expectation of “turning our lives and our will over ” to a higher power in terms of cognitive dissonance theory. The new attendee at AA typically believes she can drink safely at times but is also starting to understand that she cannot reliably control her drinking. Galanter posits that “the acceptance of a Higher Power that governs and guides an AA member toward a reconstructed abstinent recovery. . . obviates living with two dissonant beliefs, control and lack of control.” The anterior cingulate cortex is involved with reconciling conflicting perceptions. I would suggest that, while one contribution of acceptance of a higher power may be to reconcile cognitive dissonance, other aspects of AA’s spiritual orientation may also be important. In his book on the neuroscience of positive emotions, George Vaillant discusses many aspects of spirituality in healing—see my earlier post on this.

Has Galanter created a useful model of AA’s healing, or just a potpourri of experimental findings pasted onto to a complex biopsychosocial phenomenon? While the model is incomplete and relies on simplifications at many levels, this is because our understanding of human neurobiology, despite exciting recent advances, is still at a fairly primitive level. Most useful for me is his dissection of the clinical/behavioral phenomena involved in addiction and recovery, which he summarizes in a table, associating each with particular neural sites. They are: craving/relapse; suppression of craving; mirroring; mentalizing; cognitive vs. emotional empathy; memory acquisition, integration, and retrieval; unconscious learning; procedural memory; schema integration; incentivizing and storytelling; fellow member identification; and resolving cognitive dissonance.

While one can quibble over the details—he leaves out the structure and sense of safety AA provides, and the altruism involved in helping one’s self by helping others—I believe it is useful for clinicians to keep a model like this in mind as a reminder that recovery takes place at many levels both in and out of treatment sessions. Some of the phenomena on Galanter’s list are specific to addiction or to twelve-step recovery, but similar models could be constructed for healing in various psychotherapeutic situations and in other interpersonal spheres. For this he is to be applauded.

Two of the phenomena involved in AA recovery—procedural memory and self-disclosure—appear to overlap biologically with the pathophysiology of addiction, since procedural memory involves the ventral striatum, which includes the nucleus accumbens, a key component of the dopamine reward system, and self-disclosure activates the reward system itself. An interesting question is whether recovery in AA or by other means may modify the brain’s response to addictive cues or substances.

Galanter, then has given us an interesting picture of what may be going on in the brains of people with alcoholism who respond to the AA program. While most people who recover from addictions do so without the help of twelve-step programs, Galanter’s model helps us think about what takes place during the complex process of change that is recovery.

Further reading: Marc Galanter, Alcoholics Anonymous and Twelve-Step Recovery: A Model Based on Social and Cognitive Neuroscience, Am J Addict 2014;23:300-307.

Recent Posts