What You Can Change _. And What You Can't - Martin E. Seligman [110]
With regard to medications, the most widely used drug is Antabuse (disulflram). Antabuse and alcohol don’t mix: When an alcoholic takes a dose of Antabuse and then drinks alcohol, he becomes horribly nauseated and short of breath. This discourages alcohol drinking; but the alcoholic can always decide to eliminate Antabuse rather than alcohol. To avoid this, Antabuse is surgically implanted under the skin. In controlled studies, however, there is no difference in later drinking between alcoholics who have implanted Antabuse or a placebo. Both groups continue heavy drinking when the implantation ends.22
The use of Antabuse is but one in the larger portfolio of treatments that aim to produce an aversion to alcohol. In electrical aversion, shock is paired with taking a drink in the hope of making the taste of alcohol aversive; this treatment does not seem to work. In chemical aversion, drugs that make alcoholics sick to their stomach are paired with the taste of alcohol. There is better theoretical rationale for this move (recall the potency of the sauce béarnaise effect discussed in chapter 6, on phobias). Over thirty thousand alcoholics have received such treatments, but without clear effect—the treated groups do about as well as matched controls would be expected to do. Astonishingly, there has been only one study with random assignment to aversion or control. Overall, then, I cannot recommend aversion therapy for alcoholism. There is simply no scientifically worthy evidence that aversion treatment improves on the natural course of recovery.23
Naltrexone provides new hope for the successful medication of alcoholics. Joseph Volpicelli, an addiction researcher at the University of Pennsylvania, proposed that alcohol drinking stimulates the body’s opiate system and so causes a high. By blocking the brain’s opiate system chemically, the high should be blocked. In a twelve-week study of seventy male alcoholics, half got naltrexone (an opioid blocker) and half got a placebo. Fifty-four percent of the placebo-treated men relapsed, but only 23 percent of the naltrexone-treated men did. Most of the men had at least one drink during the study; the placebo-treated men went on to binge, but most of the naltrexone-treated men stopped after one drink—just what you’d expect if naltrexone had indeed blocked the high. A second study, at Yale, has replicated these effects. Caution is in order because of the lack of long-term follow-up, but this is the most promising development to date in the otherwise fruitless history of medication for alcoholics.24
THE SINGLE most frustrating question is whether Alcoholics Anonymous and the self-help groups that have spun off from it work. Alcoholics Anonymous cannot be evaluated with any certainty for several reasons. First, AA cannot be scientifically compared to the “natural course of recovery,” my criterion for all the comparisons above, because the group itself is such a ubiquitous part of the natural course of recovery in America. Try to create a control group of alcoholics whose members do not already go to AA. When you do, you will find that they are usually less severely alcoholic than those who enter AA, and thus a useless control. Second, AA does not welcome scientific scrutiny. It promises anonymity to its members and inculcates loyalty to the group—two conditions that make long-term follow-up difficult and disinterested self-reporting unusual. AA bears more resemblance to a religious sect than to a treatment, a fact that is not irrelevant to its success. Third, the people who stay in AA and attend thousands of meetings tend to be the people who stay sober. This does not necessarily mean that AA is the cause. The causal arrow may go the opposite way: Drunkenness