As director of the National Institute on Drug Abuse (NIDA), Alan Leshner toured the country with a PowerPoint presentation featuring brain scans. The show was a slightly more sophisticated version of the Partnership for a Drug-Free America's famous ad showing an egg frying in a pan. As he flashed magnetic resonance images (MRIs) on a screen, Leshner would say, in effect, "This is your brain on drugs."
Leshner's message was threefold. First, certain drugs are inherently addictive. Second, scientists have discovered the neurochemical processes through which these drugs cause addiction. Third, that understanding will make it possible to develop drugs that cure or prevent addiction. Leshner's traveling PowerPoint show epitomized NIDA's reductionist approach to drug abuse: Take a brain, add a chemical, and voilà, you've got substance dependence.
Leshner left NIDA at the end of November. Coincidentally, Enoch Gordis, head of the National Institute on Alcohol Abuse and Alcoholism (NIAAA) since 1986, retired around the same time. Like Leshner, Gordis sees addiction as a biological problem with a pharmaceutical solution. He believes scientists have "the ability based on new knowledge from neuroscience research to develop pharmacologic treatments that act on brain mechanisms involved in alcohol dependence."
The view of addiction espoused by Leshner and Gordis is at odds with what we know about the actual behavior of drug users and drinkers -- including evidence from government-sponsored research. These studies indicate that treatment is neither necessary nor sufficient for overcoming addiction. The main factor in successful resolution of a drug or alcohol problem is the ability to find rewards in ordinary existence and to form caring relationships with people who are not addicts. By looking instead for a magical elixir just over the horizon, NIDA and the NIAAA give short shrift to the individual circumstances that are crucial to understanding why some people abuse drugs.
'A Medical Illness'
NIDA's official mission is, in its own words, "to lead the Nation in bringing the power of science to bear on drug abuse and addiction." Leshner, who has a Ph.D. in physiological psychology, took the agency's helm in 1994. During his tenure NIDA's budget doubled to $781 million, money devoted mainly to biological research that approaches addiction as a disease.
Although drug use "begins with a voluntary behavior," Leshner said in a 2001 interview with The Journal of the American Medical Association, it ceases to be voluntary after it repeatedly affects the "pathway deep within the brain" common to all drug addiction. "There's no question it's a medical illness," he said, "and once you have it, it mandates treatment. It's a myth that millions of people get better by themselves."
Leshner's model of addiction emphasizes the special power of drugs. After all, he did not travel around the country with MRI images showing how shopping, gambling, or eating potato chips affects the brain. Thus it was startling to see him concede that drug abuse may be fundamentally similar to excessive involvements with other activities that give pleasure or relieve stress. "Over the past 6 months," he said in the November 2 issue of Science, "more and more people have been thinking that, contrary to earlier views, there is a commonality between substance addictions and other compulsions." Some of us have been making this point for years, and it does not fit very well with the idea that drugs create addicts by transforming their brains.
As evidence for this view, Leshner would point to MRI scans of experienced drug users, which he claimed differed in characteristic ways from images of ordinary brains. He also cited studies of drug-induced brain changes in animals. He liked to display a map -- reminiscent of a phrenology chart -- showing which areas of the brain are involved in drug use and addiction.
But Leshner's seemingly scientific claims have never jibed with reality. Consider what the sociologist Lee Robins and the psychiatrist John Helzer found when they headed a team that interviewed veterans who had been addicted to heroin in Vietnam. Only one in eight became readdicted at any time during the three years after they came home. This was not because the rest were abstinent: Six in 10 used a narcotic after returning to the U.S., and a quarter of the previously addicted men used heroin regularly.
Yet only one in five of those who used a narcotic after they got home, including only half of those who used heroin regularly, became readdicted.
The Vietnam situation, of course, was unique. Young men were torn from their homes, sent to a strange and dangerous environment, and offered easy access to heroin. Then they returned to normal life. Still, the results surprised Robins and her associates, who commented: "It is uncomfortable presenting results that differ so much from clinical experience with addicts in treatment. But one should not too readily assume that differences are due to our special sample. After all, when veterans used heroin in the United States…only one in six came to treatment." In other words, looking only at addicts who are treated provides a skewed view of addiction. Indeed, the vets who were treated after they got home actually were more likely to pick up the habit again.
Rats vs. People
Any doubts about the relevance of the Vietnam veterans study are allayed by findings from long-term studies of drug users in the U.S. Long-term cocaine users, for example, generally do not become addicts. And when they do go through periods of abuse, they typically cut back or quit on their own. They may not do so as rapidly as others (and they themselves) wish they would. But addicts act very much like other human beings:
They pursue pleasure or relief, and most will change their behavior when it causes them serious harm, so long as they have reasonable alternatives.
According to the National Household Survey on Drug Abuse (overseen by the Substance Abuse and Mental Health Services Administration), about 3 million Americans have used heroin. Of these, one in 10 report using the drug in the last year, and one in 20 say they've used it in the past month. The percentages for cocaine are similar. In both cases, daily use is so rare that the government does not provide figures for it. These findings indicate that the vast majority of heroin and cocaine users either never become addicted or, if they do, soon manage to moderate their use or abstain.
This pattern has been confirmed again and again by government-sponsored research. At NIDA, however, studies of human behavior have taken a back seat to research involving brain scans, special breeds of rats, and monkeys tethered to drug-dispensing catheters.
Given NIDA's biological orientation, it may seem odd that the main form of treatment the agency advocates (pending development of a wonder drug for addiction) involves adopting a new set of quasi-religious beliefs and meeting regularly with like-minded individuals. But NIDA's take on addiction has much in common with the view promoted by Alcoholics Anonymous (A.A.) and its imitators. Both see addiction as a disease involving loss of control that can be overcome only through abstinence.
NIDA's support for drug treatment based on A.A.-like principles, the dominant approach in the United States, flies in the face of its avowed commitment to rigorous science -- a conflict illustrated in the last issue of NIDA's newsletter published under Leshner. A front-page article announced the disastrous long-term consequences of heroin use, based on a study that followed a group of addicts for more than 30 years. "The death rate among the members of the group is 50 to 100 times the rate among the general population of men in the same age range," the article said.
"Even among surviving members of the group," the lead researcher added, "severe consequences such as high levels of health problems, criminal behavior and incarceration, and public assistance were
associated with long-term heroin use."
Yet the subjects of this study were criminal offenders in California who were forced to attend abstinence-oriented, A.A.-style group sessions between 1962 and 1964. In other words, they benefited from just the sort of treatment NIDA advocates. Undaunted, Leshner began his column in the same issue of the newsletter with the cheery news that "NIDA's quarter century of research has produced a basic unequivocal message -- drug addiction is a treatable brain disease." Yet today's preferred treatment is indistinguishable from the programs those California convicts attended in the 1960s.
Sugar: The Miracle Cure
If Leshner and Gordis are right, A.A.-style therapy will ultimately be replaced, or at least supplemented, by drugs that block addiction. The leading candidate so far is naltrexone, which is reputed to curb the urge for both heroin and alcohol. Naltrexone has been approved for treatment of alcohol dependence, and Gordis, an M.D., promoted the drug as the first in the pharmacopoeia he envisioned for alcoholism.
A study published in December made that prospect seem unlikely. The researchers divided 600 alcoholics into three groups: One received naltrexone for a year, another was given naltrexone for three months followed by nine months of sugar pills, and the third group took just the placebo. The subjects began the study drinking, on average, on two out of every three days, 13 drinks on each occasion. One year after their treatment began, these men were drinking one-quarter as frequently and consuming somewhat less when they did drink. But the reduction was about the same for the men who took the fake pills as it was for those who were given naltrexone.
Announced in The New England Journal of Medicine, these findings were incomprehensible to anyone who accepts the view of alcoholism promoted by the NIAAA. Aside from the evidence against naltrexone's effectiveness, it was stunning that sugar pills enabled severe alcoholics to reduce their drinking without abstaining completely, which alcoholism experts in the United States teach is impossible. Yet every major study of alcoholism carried out during Gordis' tenure at the NIAAA yielded the same sort of results. It's just that Gordis spent much of his energy denying what his own agency had found.
In 1992 the NIAAA surveyed more than 42,000 randomly selected Americans in the National Longitudinal Alcohol Epidemiologic Survey. Census Bureau interviewers questioned each respondent about his or her lifetime drug and alcohol use. Of special interest were 4,585 respondents who at some time in their lives were "alcohol dependent" (what most people call alcoholic). Of this group, only about a quarter were ever treated for alcoholism (including A.A. as treatment). But the treated group was no more likely to have improved, as measured by either abstinence or drinking without abuse. In fact, more treated (33 percent) than untreated alcoholics (28 percent) were continuing to abuse alcohol.
One reason untreated alcoholics did better was that many more of them reduced their drinking without abstaining. Among people who at some point in their lives had qualified as alcohol dependent but were never treated, nearly "6 in 10" or "more than half" (58 percent) in the untreated group were drinking without a diagnosable problem. Including all the treated and untreated alcoholics in this random sample of Americans, half were drinking without abusing alcohol.
Driven Not to Drink
The NIAAA sponsored another ambitious study -- the largest trial of psychotherapy ever conducted. Completed in 1996, the study was known as Project MATCH because it was aimed at determining whether different treatments could be "matched" to specific types of alcoholics to produce optimum results. One of the therapies, based on A.A.'s 12 steps, was called "12-step facilitation." A second was dubbed "coping skills therapy." The third was "motivational enhancement therapy." Nearly half of the 1,700 or so subjects underwent hospital treatment first; the rest entered the MATCH treatments directly.
All the therapies performed equally well, but one was considerably simpler than the others: Motivational enhancement involved four sessions with each alcoholic, compared to 12 for the two other types of therapy (although, on average, subjects attended only two-thirds of the sessions scheduled for any of the therapies). Motivational enhancement brings into focus and strengthens the individual's own drive for sobriety, but it leaves the mechanics of sobriety to the alcoholics themselves.
Although the Project MATCH subjects had few counseling sessions (especially in motivational enhancement therapy), their drinking was periodically assessed following treatment. These interactions with the project, intended solely for research purposes, seem to have had the effect of keeping alcoholics focused on controlling their drinking.
Whatever treatment alcoholics received in Project MATCH, few abstained for even a year. Gordis and his colleagues instead emphasized dramatic reductions in drinking by the subjects. Whereas they averaged 25 days of drinking a month prior to treatment, after a year they were drinking only six days out of the month. Moreover, the average number of drinks they consumed each time they drank dropped from 15 to three.
In all three of these prominent studies -- the naltrexone trial, the NIAAA's national survey, and Project MATCH -- the results were essentially the same. Even with clinical alcoholics, minimal treatments were as successful as more elaborate ones, and the best indicator of success was the alcoholics' ability to cut back their drinking rather than quit altogether. But how can sugar pills or a few sessions of motivational enhancement help alcoholics control their drinking? The basic ingredients for successful treatment are 1) identifying a problem with the agreement of the addict, 2) believing change is possible, 3) placing primary responsibility on the addict for carrying out the change, 4) accepting reductions in use as well as abstinence, and 5) following up to let addicts know someone cares and wants to make sure they stay on course.
In the face of studies that cast doubt on traditional notions about alcoholism, Gordis seemed to consider it his duty to explain why they actually confirmed the conventional wisdom. Project MATCH in particular presented a serious P.R. problem for the NIAAA: It spent more than $30 million without fulfilling its purpose of identifying principles for matching alcoholics to treatments. This is how Gordis spun the results: "The good news is that treatment works. All three treatments…produced excellent overall outcomes."
Although Gordis relied on reduced drinking as a measure of success to put the best gloss on Project MATCH, he has always quashed any revision of the abstinence-oriented goals that characterize virtually all American alcoholism treatment. Responding to a 1997 U.S. News and World Report story on Moderation Management, a program for reducing alcohol consumption among problem drinkers, Gordis sternly warned that "current evidence supports abstinence as the appropriate goal for persons with the medical disorder 'alcohol dependence' (alcoholism)."
While abstinence may be a desirable goal for these individuals, not many accomplish it. Project MATCH engaged the top clinical practitioners and researchers in the United States in designing and supervising treatment for alcoholics. As a result of this attentive, sophisticated care, which is unlikely to be matched by any program an alcoholic could find in the real world, about a quarter of the subjects abstained for as long as a year.
Gordis' attitude seems to be: "Most alcoholics won't abstain after treatment, but they should! And we are not going to accept anything less than this worthy, if unreachable, goal." His attitude is especially disturbing since Project MATCH found that reductions in drinking were beneficial. The subjects' liver functioning typically improved, and they displayed fewer problems associated with drinking. Surely, better health and less destructive behavior are worthy goals.
Since Gordis spoke for the U.S. alcohol treatment establishment, his rigidity condemned American alcoholics to limp along, most continuing to drink, with little chance of finding assistance in limiting their drinking or reducing its negative consequences. We will never eliminate drinking and drug use. But we might be able to reduce the harm they sometimes cause if we could eliminate the pseudoscientific moralism dispensed by the likes of Leshner and Gordis.