The Atlantic Monthly, August 1990, pp. 52-58
Second Thoughts About a Gene for Alcoholism
Claims of a genetic basis for alcoholism, a leading theorist argues, are not scientifically supportable and ignore the crucial link between personal values and self-destructive or antisocial behavior
Morristown, New Jersey
Major news stories about discoveries of the genetic sources of emotional and behavioral problems surface every year or so. Since 1987 such reports have appeared on the front page of The New York Times in connection with manic-depressive disorder, schizophrenia, and alcoholism. Last year, however, the Times published a story titled "Scientists Now Doubt They Found Faulty Gene Linked to Mental Illness." This story received less attention than the announcements of positive findings, for it appeared not on the front page of the newspaper but buried deep inside. The article revealed that both new data and further analysis of the original data "cast serious doubt" on the earlier finding that a defective gene was, in fact, associated with manic-depressive disorder. The article furthermore noted that "the new findings underscore the difficulty of assigning specific causes to such a complex and variable illness," and that problems also plague efforts to identify precisely any sort of genetic role in schizophrenia. One of the authors of the original study said, "We are sort of back to square one."
The Blum-Noble "Alcoholism Gene"
The study that the Times reported linking alcoholism to a specific gene was published in the Journal of at American Medical Association on April 18 of this year, as the journal's lead article. It was accompanied by press releases, a highly publicized news conference in Los Angeles, and video interviews with the study's authors, which the AMA transmitted by satellite in its weekly television news release. The study's chief authors were Kenneth Blum, a pharmacologist at the University of Texas Health Science Center, in San Antonio, and Ernest Noble, a psychiatrist and biochemist at the UCLA Alcohol Research Center and a former director of the National Institute of Alcohol Abuse and Alcoholism.
Unlike the manic-depressive-disorder and schizophrenia studies, which were conducted within individual families or communities, the alcoholism study involved seventy unrelated cadavers, thirty-five of which had been alcoholics and thirty-five of which were controls. According to the researchers, the alcoholic cases were of an extremely "virulent" typemany of the thirty-five people had died of cirrhosis. Genetic material from all seventy brains was analyzed. A genetic marker was found in 69 percent of the alcoholics in the study but in only 20 percent of the nonalcoholics.
Blum and Noble concluded that the Al "allele," or variant, of the dopamine D2 receptor gene was associated with alcoholism. Dopamine is one of a number of neurotransmitters, or chemicals produced in the body that convey information throughout the nervous system. Neurotransmitters communicate by attaching to receptors on nerve cells which are tailored specifically to them. One important question not answered by the gene discovery is exactly what dopamine receptors, the physiological mechanisms affected by the gene, have to do with alcoholism. Another question is how this discovery fits in with what has previously been established about the heritability of alcoholism.
Theories of the Heritability of Alcoholism
The dopamine receptor gene is not clearly implicated in the major existing genetic theories of human alcoholism. According to these theories, a certain personality type predisposes a minority of male problem drinkers to both crime and alcoholism; an inherited insensitivity to alcohol allows alcoholics to drink more while being less aware of the effects of the alcohol they are consuming; alcoholics are not able to metabolize alcohol normally; and alcoholics have inherited neurological and intellectual dysfunctions. Not only are the originators of some of these theories the sole researchers to have found evidence to support them, but also several of the theories and findings directly contradict the premises or reasoning of others.
The inherited-personality theory has been promoted by a research group at the School of Medicine at Washington University, in St. Louis, under the direction of Robert Cloninger, a psychiatrist. This may be the most popular current notion about how alcoholism is inherited. However, it must face the formidable difficulties involved in associating entire personality syndromes, such as criminality, with particular genes.
The idea that alcoholics inherit an insensitivity to alcohol has been presented by Mark Schuckit, a psychiatrist at the University of California at San Diego Medical School. Another research team (led by Barbara Lex, of Harvard Medical School) has supported Schuckit's hypothesis in a study comparing small numbers of women from alcoholic and nonalcoholic families. But James Wilson and Craig Nagoshi, of the University of Colorado, in a much larger study, found the lessened sensitivity to alcohol to hold for some age-and-sex groups among offspring of alcoholics but not for others.
Schuckit also at one time presented evidence that alcoholism results from an inherited flaw in the way alcoholics metabolize alcohol. According to this theory, alcoholics' bodies do not break down alcohol properly and, as a result, build up abnormal levels of acetaldehyde (one of the products of alcohol's oxidization) when they drink.
The acetaldehyde hypothesis got strong play in the widely read book Under the Influence, by James Milam and Katherine Ketcham, which was published in 1981. Milam and Ketcham forcefully argued that acetaldehyde is a primary biological and genetic basis for alcoholism, which they claimed is completely biologically determined. However in recent years researchers have lost enthusiasm for this model, which has been contradicted by subsequent research. For example, several research teams investigating the offspring of alcoholics were unable reliably to identify abnormal acetaldehyde buildups after their subjects consumed alcohol. Schuckit himself no longer focuses on acetaldehyde as the most likely mechanism for the heritability of alcoholism.
According to another alcohol-metabolism model, the oxidized alcohol product, or aldehyde, interacts with neurotransmitters to create tetrahydroisoquinolines, or TIQs, which are thought to create the intense craving for alcohol that alcoholics demonstrate. However, the idea that alcoholics produce abnormal amounts of TIQs has never been demonstrated satisfactorily and several prominent investigators have abandoned this path of research. Nonetheless, the TIQ model is often taught to alcoholics at treatment centers in the United States. When several TV news shows interviewed alcoholics (attractive middle-class women who would not have been diagnosed as advanced alcoholics, as the cases that Blum and Noble examined had been) in connection with the JAMA gene study, the women claimed they had inherited an exaggerated reaction to alcohol that made them drink more. They were likely referring to TIQs, about which they were probably lectured in treatment.
Another set of theories about the heritability of alcoholism concerns abnormal brain waves and other cognitive or neurological impairments that children of alcoholics are said to inherit. The most prominent brain-wave researcher, Henri Begleiter, a psychiatrist at the Downstate Medical Center, in Brooklyn, has found such abnormalities in alcoholics' sons who have never drunk alcohol. However, other factorssuch as diet or physical abuse may explain why many children of alcoholics show abnormal brain activity. Moreover, different research teams have found different abnormalities in the brain-wave patterns of different groups of alcoholics and children of alcoholics. Finally, no link has been shown between the abnormal brain waves and the development of alcoholism, because the children with the suspect brain waves have not been followed up to an age where their rate of alcoholism can be compared with that of children of alcoholics without these specific abnormalities.
Children of alcoholics have also been studied with tests that measure perception, coordination, and intellectual abilities, and are often found deficient in these areas. But Schuckit finds the results of this research inconsistent. Moreover, one of his own studies, using subjects who on average had completed more than three years of college, did not find such impairment, which may be owing to social differences between these subjects and, for example, children of alcoholics identified through the juvenile justice system.
Plainly, the various heritability theories about alcoholism tend in quite different directions. Schuckit has strongly criticized the idea of an inherited "antisocial' personality disposition to alcoholism, which he has not found in a group of college students and staff who are sons of alcoholics. The insensitivity model seems to be the opposite of the TIQ model: according to the former, those likely to become alcoholics drink more and respond less, whereas according to the latter, those predisposed to alcoholism respond more. According to the insensitivity model, those predisposed to alcoholism may drink heavily for extended periods because they cannot detect the physical effects of the alcohol they consume, and consequently become dependent on alcohol. Most of the modern theories about the heritability of alcoholism recognize that no matter what their genetic makeup, people must drink a lot for a long time to become alcoholics. In this sense genetic research in alcoholism actually contradicts stories commonly told at Alcoholics Anonymous meetings and popular treatment myths about people who become alcoholic with their first drink.
How Does the Blum-Noble Gene Fit In?
The Blum-Noble team probed nine different genes, all hypothetically linked to alcoholism, in the DNA of the brains they examined. They found that only the Al allele of the dopamine D2 receptor gene correlated significantly with the diagnosis of alcoholism in their subjects. Where does this discovery leave other theories of the heritability of alcoholism? The absence of a statistically significant relationship between any of the eight other genes and the occurrence of alcoholism actually undercuts most of the other theories about the way alcoholism is inherited. Thus this discovery yields more disconfirmation than support for researchers in the field.
On the other hand, if a range of inherited mechanisms leads to alcoholism, how can the dopamine receptor gene alone account for more than two thirds of a group of cases of alcoholism? Commenting in the same issue of JAMA in which the alcoholism-gene study appeared, Enoch Gordis, who is the director of the National Institute on Alcohol Abuse and Alcoholism, and three colleagues wrote that the association between the gene and alcoholism was "surprisingly strong" in view of the "presumed heterogeneity of alcoholism and its likely polygenic causes."
Gordis, who has long championed genetic views of alcoholism, noted problems with the Blum-Noble research that make it "possible that the differences between the groups were caused by characteristics of the groups that were unrelated to alcoholism. This type of comparison, although provocative, cannot substitute for. . . complex family studies." The standard practice in gene-mapping research is to identify genes in a range of living related persons. This "family-tree" approach can then trace whether or not various relatives with a specific genetic marker develop a given disease. To disprove a specific genetic link is easier than to prove one: a finding that relatives who do not share the genetic marker in question become alcoholic as frequently as relatives who do share it would be strong evidence against that genetic link.
All sophisticated observers who have commented on the JAMA study have cautioned that additional research must be designed to examine the nature of the relationship between the dopamine D2 receptor gene and alcoholism. Blum and Noble are now mapping this gene in a study of living relatives, and their results are awaited eagerly.
Another form of replication would be to compare the incidence of the Blum-Noble marker in a random sample of patients in alcoholism clinics with its incidence in a general population. Would such investigations affirm the results of the original research? Many observers interviewed about the Blum-Noble finding seemed skeptical. Donald Goodwin, the psychiatrist whose research first pointed to the heritability of alcoholism, says that "the history of this kind of work so far has been a failure to replicate."
How Might a Gene Cause Alcoholism?
For a theory to be accepted scientifically, not only must the finding or findings that support it be replicated in rigorously designed research, but also the hypothesized connection between a mechanism and an effect must be plausible and consistent with everything else we know. How, exactly, does the dopamine receptor gene influence the alcoholic's biochemistry and behavior? This connection is not at all obvious, since dopamine receptors are not directly affected by alcohol. Indeed, it is not even known how the Blum-Noble marker affects the dopamine receptors. The alcoholism-associated allele might increase or reduce the number of dopamine receptor sites in brain cells, or affect how well dopamine binds at these sites. One theory is that if the allele reduces the number of receptor sites or minimizes binding at them, then the person's brain might be deficient in dopamine activity. Drinking might compensate for this deficiency by increasing the drinker's levels of dopamine.
But does this complex series of links account for the alcoholic's motivation to drink? Some researchers propose that dopamine is related to pleasure-seeking activity. This is a very long way, however, from explaining how people with the Al allele become alcoholic. Alcoholics often receive negative feedback about their drinking from other people, from their own internal standards, and from physical problems they encounter. Why don't these experiences overcome the presumed pleasurable nature of the dopamine activity and cause the person to curtail drinking? Put in another way, the question is whether a complex behavior can be said to be motivated solely by the pleasure (or comfort or relief) it produces. For example, orgasms are intensely pleasurable and most people can produce them at will, yet relatively few people become compulsive fornicators or masturbators, at least in the long run.
Moreover, dopamine stimulation is simply not so direct or exclusive an effect of drinking that it seems likely to be the root cause of compulsive drinking. Many human activities other than drinking stimulate dopamine release. Some may argue that alcoholics are part of a larger group of people who seek such stimulation, and who therefore experience a higher level of dopamine activity in their brain cells. But why do only some people seek this effect from alcohol?
In an interview about the JAMA study Gordis speculated that the gene marker "may not be specific for alcoholism but it might have a more general influence on appetite, personality, and behavior." Noble has concurred: "The good Lord did not make an alcoholic gene, but one that seems to be involved in pleasure-seeking behaviors." These statements seem skeptical about the existence of a link between the gene and alcoholism. The idea that people with the A1 allele may be predisposed to pleasurable stimulation, of which alcoholism may be one example, also falls short of supporting claims that alcoholics have an innate response to alcohol that dooms them to alcoholism once they begin drinking.
Even if Blum and Noble's finding holds up, most people with the A1 allele will not become alcoholics. According to the study, about 25 percent of the population have this form of the gene. Estimates of the proportion of alcoholics in the U.S. population range up to 10 percent. However, the "virulent" type of alcoholism that Blum and Noble talk about probably occurs in less than five percent of the population. Thus if every full-blown alcoholic had this allele, then only a fifth of the 25 percent that have it would be severely alcoholic. Since only 69 percent of the alcoholics in the Blum-Noble study actually had the allele, even fewer than a fifth of those with the marker will conform to the diagnosis of alcoholism used in the study.
How Can This Discovery Be Used?
Blum and Noble have announced that the identification of a gene that places people at risk for alcoholism should open new treatment and prevention options. The most likely step, given the current cultural climate, would be to tell those with the genetic marker that since they have an elevated risk of alcoholism, they should not drink. But this idea has its problems. In the first place, since less than a fifth of the people with the gene will actually become alcoholics, the warning is unnecessary for most of those with the marker and could make their lives needlessly difficult.
Moreover, unforeseen, and lamentable, consequences might occur. We frequently tell young people not to drink, and they frequently don't listen. But if a person has been told that drinking will lead to alcoholism, the prophecy may ultimately prove to be self-fulfilling. According to Peter Nathan, who was until last year the director of the Rutgers Center of Alcohol Studies, "It has become increasingly clear that, in many instances, what alcoholics think the effects of alcohol are on their behavior influences that behavior as much [as] or more than the pharmacologic effects of the drug." For example, alcoholics have been shown to drink excessively and even behave drunkenly when they are told they are drinking an alcoholic beverage even though the drink does not actually contain alcohol.
In other words, indoctrinating young people with the view that they are likely to become alcoholics may take them there more quickly than any inherited reaction to alcohol would have. In fact a majority of children of alcoholics do not become alcoholic themselves, for whatever reason. No epidemiologic study has ever found that as many as half of such children develop a drinking problem of their own, and most research places the figure at 25 percent or less. That many children don't inherit their parents' alcoholism is indicated by the rapid growth of the "adult children of alcoholics" movement, whose typical member is a woman who has never had a drinking problem. Moreover, adolescents and young adults who have drinking problems often overcome them without abstaining, even if they have an alcoholic parent.
Blum and Noble have also suggested that their genetic finding will lead to medical therapies for alcoholism. That is, if the source of alcoholism is biochemical, then a drug therapy might be designed to eliminate the alcoholic's craving for alcohol. If a vitamin stimulated dopamine activity, for example, or if a drug blocked the binding of dopamine to receptors, then alcoholics could be freed from their biological motivation for excessive drinking and might drink moderately.
Yet it is unlikely that treatment centers and AA would recant their commitment to total abstinence for alcoholics. Nonetheless, it is in these quarters that we can expect the most enthusiastic reception for the Blum-Noble research, because genetic causation of alcoholism is seen to support their fundamental assumption that alcoholism is an involuntary disease.
The New York Times and other media claimed that the JAMA study appeared to offer "strong new evidence" for the heritability of alcoholism (this was, of course, the point of the AMA press release), an issue that has "been widely debated for decades." Actually, enthusiasm for the idea that alcoholism is inherited has prevailed in the United States for some time. AA and most proponents of the disease theory (including most of the people who manage America's 4,600 private alcohol-treatment programs) have for a long time insisted that alcoholism is an inherited disease. For about the past decade the number of researchers investigating the topic has been growing, along with the assumption that alcoholism is at least in part genetically caused. For example, the National Institute on Alcohol Abuse and Alcoholism published a pamphlet in 1985 titled "Alcoholism: An Inherited Disease."
Even many psychologists now assert that the problem drinking of at least some alcoholics has a significant genetic component. Moreover, the public has accepted the idea that alcoholism is inherited. A 1987 Gallup poll found that nearly 90 percent of Americans believe alcoholism is a disease, and more than 60 percent think it may be inherited. Both figures represent a steep rise from earlier ones; even five years previously a solid majority of respondents to the second question had said they didn't believe that alcoholism is inherited. Thus the JAMA report did little more than affirm what most Americansincluding alcoholics, treatment personnel, and researchersalready believe.
Although the heritability of alcoholism is often presented as a modern discovery, alcoholism (along with sexual promiscuity, intelligence, criminality and insanity) was widely believed to be an inherited trait in the nineteenth century. That view receded in this century; in 1938 Karl Menninger could state, "The older psychiatrists . . . considered alcoholism to be an hereditary trait. Of course, scarcely any scientist believes so today, although it's still a popular theory. Alcoholism cannot possibly be an hereditary trait...." Today, obviously, the tide has shifted again. As Robin Murray, a British psychiatrist, notes, "Students of alcoholism must continually beware lest they fall victim to the extravagant swings of intellectual fashion that bedevil the field, and nowhere is such vigilance more necessary than in considering the possible etiological role of heredity." Murray's views are particularly interesting because he is Britain's leading investigator of the heritability of alcoholism. Murray and his research team have found that the rates of coincidence of alcoholism are similar for identical and fraternal twins, a result that substantially undermines genetic hypothesesand one that is almost never cited by American researchers.
Will views on this question swing back? Oddly, the attention attracted by the JAMA study may encourage them to, as bold genetic claims draw sophisticated clinical geneticists into the field. (As far as I am aware, the leading current investigators in human studies of the heritability of alcoholism are not trained as genetic researchers. ) Paul Billings, the director of the Clinic for Inherited Disease at New England Deaconess Hospital, was interviewed about the JAMA study for The New Republic. Billings commented, "If this type of genetic analysis was carried out for a disease or a behavior less attractive than alcoholism, it would never get published. It tells you nothing of significance."
What Does Cause Alcoholism?
Doubts about genetic research aside, a large body of findings about alcoholism and its correlates cannot possibly be translated directly into biological and genetic terms. One such area of research concerns drinkers' expectations. Alcoholics have stronger expectations about how alcohol will affect them than other drinkers do. Alcoholics believe that alcohol transforms their personalities, making them more attractive to others, more relaxed, more alert, and more sexually responsive, even though in alcoholics it usually has the opposite effect in all these areas.
Mark Goldman, a psychologist at the University of South Florida, and his colleagues have extended research on expectations to college problem drinkers and even adolescents. Goldman and his colleagues found they could predict the likelihood that adolescents would develop drinking problems on the basis of their expectations about alcoholbefore they had begun drinking. Furthermore, even elementary school children have distinct beliefs about how alcohol will affect them. Thus when Cathleen Brooks, the president of the National Association for Children of Alcoholics, describes her first drink, at age eleven, in euphoric terms''I remember the warmth, I remember the well-being"she is more likely reflecting what she had seen or been taught than what her body inherited.
Expectations about the effects of alcohol relate to parental, peer-group, and cultural influences. The psychiatrist George Vaillant's widely cited 1983 book, The Natural History of Alcoholism, examined the drinking histories of more than 600 men over forty years. Vaillant found that Harvard students were a third to a quarter as likely to develop alcoholism as inner-city Boston ethnics, and that Irish-Americans in Boston were seven times as likely to become alcoholic as Italian-Americans.
If a researcher were to find a biological marker for alcoholism as strong and reliable as the ethnic and social markers Vaillant identified, he or she would likely win the Nobel Prize. Yet the discrepancy between Irish and Italian alcoholism rates is far from the largest ethnic difference. Two sociologists, Barry Glassner and Bruce Berg, expected to find the legendary moderation of Jewish drinkers much attenuated in an assimilated upstate New York Jewish community. They found instead that none of their eighty-eight Jewish subjects had ever had a drinking problem. The most dire count of the city's Jewish alcoholics implies an alcoholism rate of about 0.1 percent for that Jewish community, or one hundredth the rate reported for Americans at large.
Another sociologist examined police blotters in New York's Chinatown for the period 1933-1949. Among the 17,515 arrest records, not one specifically reported violent or disorderly drunkenness. What is most remarkable about Chinese sobriety is that the Chinese, like other Asian groups, show a high incidence of flushing after imbibing alcohol. This visible reddening has been connected to acetaldehyde buildup in the drinker and is inherited genetically. Some researchers have proposed that such flushing makes the Chinese less likely to drink excessively. Working against this hypothesis is the fact that Native Americans and Eskimos, likewise prone to flushing, have the highest incidence of alcoholism among American ethnic and racial groups. The difference in the incidence of alcoholism among Native Americans and Chinese-Americans reflects how poorly or how well the group's indigenous values have fit in with those of the broader American culture.
One of the other cultural (and individual) factors that most clearly elevates alcoholism rates is, paradoxically, the acceptance of conceptions of alcoholism as disease-like. That is, alcoholism-prone groups such as Native Americans and Irish-Americans invest alcohol with tremendous power and readily accept that they cannot control its effects. Groups with a low incidence of alcoholism, such as Chinese-Americans and Jews, do not tolerate "loss of control" as an excuse for problem drinking or antisocial drunkenness. Glassner and Berg discovered that Reform and nonpracticing Jews among their subjects viewed alcoholism as a psychological dependence or weaknessexactly the attitude the modern genetic movement disavows. Yet people from social backgrounds in which antisocial drunkenness is strongly disapproved of and people who believe they can and should control their drinking are less likely to lose themselves in the experience of intoxication.
Rats can be bred to drink large quantities of alcohol. But rats do not have values and cultures that contravene the urge to drink excessively. While human beings clearly differ in how their bodies process and respond to alcohol, these differences do not translate into alcoholism independent of individual needs, options, and values. Someone who has a strong reaction to alcohol, or who cannot sense when he or she has had too much to drink, may just as easily choose habitually to stop after one or two drinks as to become intoxicated.
To deny these commonsense insights in the light of putative genetic findings is to deny the connection between human values and self-destructive or antisocial behavior as well as the crucial role that both individuals and social groups have in regulating such behavior. Even a researcher like Robert Cloninger, who espouses genetic theories, recognizes this role. According to Cloninger, "The demonstration of the critical importance of sociocultural influences in most alcoholics suggests that major changes in social attitudes about drinking styles can change dramatically the prevalence of alcohol abuse, regardless of genetic predisposition."
Addictive drinking is one of a range of dependencies that people may acquire in attempting artificially to regulate their sense of themselves and their world. Some people become compulsively enmeshed in destructive drinking as they pursue sensations that they are progressively less able to attain through any other means. And yet we cannot take the power and the seeming inevitability of this self-destructiveness for proof that it is written in the genes.