Aetiology of Alcohol Problems

The aetiology of alcohol problems.

Topics covered:

  • Introduction
    • Background to the current notions of alcoholism
  • Sociocultural factors Individual vulnerability
    • Genetic influences
    • Twin studies
    • Adoption studies
    • Biological predisposition
    • Alcohol sensitivity
    • Biochemical markers
    • Neurophysiology
    • Molecular genetics
    • Psychological factors
    • Personality
    • Psychodynamic processes
    • Learning
    • Psychiatric comorbidity
    • Brain dysfunction
  • Conclusions 
  • References


Alcohol is one of the most widely used psychoactive substances; approximately eight out of every 10 persons living in Europe and the Americas would report drinking in their lifetime. (1) The norm is for drinking to start in adolescence: close to 50 per cent of American students have had drinking experiences by the age of 13 years and the rate increases to 81.7 per cent at age 17. (2) In Canada, 78 per cent of persons aged 15 years or more are current alcohol users, with 12 per cent of the men and 3 per cent of the women falling within the drinker-at-risk category (i.e. 14 units per week or more). This is the level of use associated with the highest probability of occurrence of untoward consequences. Higher frequencies of drinking (i.e. four or more times per week) are observed in the older age groups, while a heavy intake per single drinking occasion (i.e. five units or more) is more often reported by the population younger than 35 years of age. (3)

The figures from a random selection of general population surveys of alcohol abuse and alcohol dependence, conducted in four separate countries indicate that the largest proportion of alcohol users do not develop those clinical disorders. In fact, it has been shown that drinkers who limit their average intake to no more than two standard units per day may even derive some health benefits from alcohol, such as a lower risk of coronary disease. (4) Screening questionnaires such as CAGE, a four-item instrument which rates the probability of alcoholism, identify around 20 per cent of the adult population as probable alcohol misusers. More stringent procedures, which apply full diagnostic criteria, yield point prevalence rates ranging from 5.4 to 7.4 per cent, depending on whether the clinical disorder surveyed is strictly alcohol dependence or any form of harmful drinking (i.e. dependence plus alcohol abuse). These figures are total population percentages; the rates for men are about double those for women.

Alcohol can be used in a relatively harmless manner and there exist public health guidelines on ‘safe' drinking practices. The recommendations vary considerably from country to country, but they all assume a greater vulnerability to alcohol effects in the female gender. In the United Kingdom, for instance, hazardous drinking is thought to start at 21 units/week for men and 16 units/week for women; (5) in Canada the upper limit for moderate drinking is estimated at 14 and 12 units/week respectively (6) and in the United States the equivalent guidelines are 14 and 7 drinks per week. (7)

The expression ‘alcohol problems' encompasses a wide range of untoward occurrences, from maladaptive, impaired, or harmful social behaviours, to health complications and the condition of alcohol dependence. Alcohol problems are not incurred just by chronic excessive drinkers, but also by persons who drink heavily on isolated occasions (e.g. accidents, violence, poisoning, etc.). Given their high frequency and social costs, these consequences of acute inebriation represent the most significant public health burden of drinking. (8)This article focuses rather on the causes of problems of a clinical nature, the ones presented by individuals who engage in patterns of repeated excessive drinking, i.e. ‘alcohol dependence' and ‘alcohol abuse' (DSM-IV nomenclature) or ‘harmful drinking' (ICD-10 nomenclature).

Alcohol misuse is a bio-psychosocial phenomenon par excellence; it results from the contribution of multiple individual and environmental risk factors. Causal mechanisms have been proposed by researchers in fields as diverse as social sciences, behavioural psychology, psychopathology, genetics, pharmacology, toxicology, and neurobiology. But none of them, in isolation, provides a complete explanation for the occurrence of this problem. is a depiction of the interactions assumed to play a role in the genesis of alcoholism. It identifies excessive drinking as the sine qua non component in the equation. Heavy drinking is sometimes fostered by sociocultural conditions, but it might result also from individual predispositions of a genetic, neurobiological, or psychopathological nature. It is usually the combined effect of several of these factors that determines the development of an alcohol dependence, or the occurrence of harmful drunken behaviour in a given individual.

Background to the current notions of alcoholism

Two very different schools of thought attract the largest following at the moment—the social learning and the disease models of alcoholism. The former is a relatively recent development, mainly since the 1970s. (9) It contends that alcohol misuse is an acquired behaviour which the individual is capable of correcting with adequate cognitive–behavioural training; even to the point of relearning to drink in a ‘controlled' manner. (10) Thus, the learning model tends to play down the importance of predisposing factors, and to ignore the variance in individual biological responses to alcohol, or the brain changes caused by chronic intoxication. Indeed, it assumes that all alcoholics are basically the same and that they are not different from any other drinker in their capacity to exert control over alcohol use.

In marked conceptual contrast, the inability to restrict the quantity or frequency of drinking is a key notion in the disease or ‘medical' model of alcoholism. (11) This anomaly is thought to result from heavy alcohol exposure, or else to pre-exist in constitutionally vulnerable individuals. This model further suggests that, once in place, such loss of control over drinking is irreversible.

Learning theorists currently object to the disease concept on scientific grounds, but this modern argument is very much the continuation of old ideological debates on the nature of alcoholism. Pioneer medical authors such as T. Trotter in the early 1800s, Magnus Huss, who coined the term alcoholism in 1849, and the founders of the American and British Societies for the Study of Inebriety towards the end of the nineteenth century had already maintained that the disorder involves the loss of voluntary control over drinking, (12) and this is still the opinion of an important sector of the scientific community today. (13) Moralists of the turn of the century opposed that notion because it exempts the drinker from personal blame. Of course, they viewed any form of drinking as intrinsically wrong, and alcohol itself as an evil which should be eliminated through prohibition.

The leaders of the 1960s ‘antipsychiatry' movement thought this disease entity to be nothing but a self-serving concoction of the medical profession. (11) According to these ideologists, alcoholics are always capable of choosing whether or not to drink, and must assume full responsibility for their behaviour. Thus, by denying the existence of a pathological process, today's supporters of the social learning model find themselves defending the same ideas as the old temperance preachers and the detractors of psychiatry.

This article is an eclectic summary of the aetiological theories on alcoholism, most of which view alcohol dependence as ill health.

Sociocultural factors

It is an undisputed fact that the prevalence of alcohol problems varies markedly across different cultural and social settings; that certain environmental conditions appear to facilitate their occurrence while others seem to prevent them. (14) Macrocultural influences such as values, beliefs, and mores; social role functions; local economy; customs and dietary habits; rapid social change; and cultural stress do shape and dictate the way alcohol is used in human societies. But even within a single society, there is variance in the alcohol problems profile of different subgroups. For instance, drinking, heavy drinking, alcohol use disorders, and treatment for alcoholism are more frequently recorded in men than women, (1) the risk of hospital admission for alcoholic psychosis, acute intoxication, and liver cirrhosis is elevated in unskilled and blue collar workers when compared with higher occupational categories, (15) alcoholics are over-represented in occupations with flexible work schedules, in those less supervised, and in the ones which facilitate access to alcohol, (16) and although there are a larger proportion of regular alcohol users among the older, the wealthier, and the better educated, frequency of heavy drinking (i.e. episodes of intoxication, 5+ drinks at a time) is inversely correlated with age, income, and level of education. (3,17)

Cultural beliefs about drinking and related social norms largely determine the manner in which alcohol is used. Disorderly conduct and drunken violence are more likely to occur in societies which, while allowing drinking, do view alcohol as an evil substance. (18) Similar consequences can be expected if drunkenness is culturally considered as a ‘time out,' when socially unacceptable behaviours are tolerated or excused. (19) In fact, the social condoning of drunkenness is considered as an epidemiological risk factor. (14)

The very availability of alcohol and the social promotion of frequent or heavy drinking are examples of social risk. But environmental facilitation per se does not explain the genesis of an alcohol dependence in specific individuals. This disorder is best understood as the result of social prompting and individual vulnerabilities. However, an individual predisposition is likely to play a lesser role in high availability societies with strong drinking traditions. Conversely, in ‘dry' cultures, where drunkenness or even drinking itself are viewed as deviant behaviours, alcohol abusers would tend to be more ‘abnormal' individuals capable of disregarding the stronger social barriers against drinking. (12)

Bullet list 1: Sociocultural Risk Factors:

  • Male role
  • Lower education
  • Lower income
  • Marital breakdown
  • Certain occupations
  • Idleness
  • Cultural ambivalence towards drinking
  • Self-fulfilling prophecy
  • Socially condoned drunkenness
  • Anomie/marginalization
  • Social stress

Individual vulnerability

Genetic influences

The most powerful predictor of alcohol misuse in any individual is the occurrence of alcoholism in first-degree relatives. Men and women belonging to families with alcoholic parents and/or siblings are twice as likely to develop the disorder than those without such family history. The risk is threefold when the disorder is present also in second- or third-degree relatives. (20) More comprehensive analyses demonstrate that family aggregation is true as well for addictive substances other than alcohol. (21) The excess probability observed within single family groups suggests that alcohol misuse could be a genetically transmitted behaviour, and several research approaches have been used to explore this aetiological pathway.

Twin studies

The comparison of concordance rates for alcoholism in monozygotic and dizygotic twins permits to test the genetic basis of the disorder. Since monozygotic pairs have a common genetic stock, while dizygotic ones share on average only 50 per cent of their genes, any condition of genetic origin would be found to co-occur more in them than in fraternal twins. Most studies to date have elicited significantly higher concordance in identical twins, although the rates never reach the perfect 100 per cent level.

Using general population prevalence rates as reference, to have an alcoholic monozygotic sibling markedly increases the chances of developing alcoholism; risk ratio values ranging from 11.8 to 3.9 have been found. (22) Dizygotic co-twins of alcoholics also present elevated risk ratios, but to a lesser degree, in keeping with their lower concordance rates. (22) The earlier studies included male twins only and the applicability of their findings to female siblings was uncertain. However, more recent data from a large sample of female twins indicates an equally strong genetic influence in women's alcoholism. (23)

The significance of those observations notwithstanding, it must be noted that between 40 and 70 per cent of the identical co-twins of alcoholics do not present the disorder, and that such variance can only be due to non-genetic causes.

Adoption studies

The study of adopted-away children of alcoholic parents is a powerful approach to the elucidation of the nature of the family transmission of alcoholism. A genetic vulnerability, of course, should express itself even when such children are brought up by non-alcoholic adoptive parents. All adoption studies consistently report an increased risk in these offspring, regardless of the family environment where they grow up. The probability of occurrence in the probands is much higher than in the control adoptees. This was demonstrated with Danish (risk ratio 3.6), Swedish (risk ratio 1.3), and American (risk ratios 3.5 and 3.6) samples. (22) As in the case of the twin studies, the original research with adoptees was inconclusive with respect to female alcoholism, either because women were excluded from the samples or because their small numbers limited the power of statistics. However, additional work has demonstrated the significant influence of parental alcoholism on women as well, particularly when the affected biological parent is the mother. (23,24)

The offspring of alcoholic parents are unquestionably exposed to an excess risk, but the understanding of the exact pathway of familial transmission is still limited. A more detailed analysis of the adoption data shows that the biological parents of alcoholic adoptees, when compared to the controls, are not only more likely to present alcoholism but also antisocial behaviour. (25) Antisocial personality is also significantly more prevalent in alcohol-abusing adoptees than in controls. The evidence indicates that the alcoholism–antisocial personality tandem is much more ‘inheritable' than alcoholism alone. In decreasing order of significance, alcohol abuse in adoptees is predicted by their own antisocial personality, biological parents's alcoholism, biological parents's antisocial history, and, to a much lesser degree, history of alcoholism or psychopathology in the adoptive parents.

Biological predisposition

Alcohol sensitivity

Compared to controls, a significantly larger percentage of men with family history of alcoholism present a lesser physiological response to alcohol; in terms of both subjective sensations and objective measurements (postconsumption plasma cortisol levels and body sway). A lower sensitivity is assumed to lead to heavier drinking and predicts the eventual development of alcoholism. The authors of this prospective follow-up study concluded that the ‘innate' low response to alcohol is an independent risk factor, regardless of family history. (26)

Related research findings indicate that men with multigenerational family history of alcoholism tend to derive a greater anxiolytic effect from alcohol. Under experimental conditions, and after drinking an intoxicating dose, their cardiovascular response to acute stress is more attenuated than that of controls. (27) Such increased effects can be expected to reinforce heavy drinking.

Biochemical markers

On average, alcoholics have lower than normal levels of platelet serotonin and of its metabolite 5-hydroxyindole acetic acid in cerebrospinal fluid; especially the problem drinkers who exhibit impulsive and violent behaviour. This serotonin deficiency profile was also found to be abnormally prevalent in young adults with a family history of alcoholism. (28)

The monoamine oxidase enzyme is of interest to alcohol researchers because it participates in the breakdown of neurotransmitters which are involved in the brain action of alcohol. Abnormally low levels of monoamine oxidase platelet activity were found to be linked to a family history of alcoholism, as well as to an earlier age of onset and a higher severity of the disorder. (29)

Both these biochemical deviations seem to be more significant among male than female subjects, and to be strongly associated with antisocial behaviour.

Event-related potentials

Alcoholics have been found to present a lower than normal amplitude of the P300 wave, when such brain potential is evoked through complex visuomotor tasks. The P300 potential is thought to measure attentional and memory processes, and its amplitude tends to increase with age and neurological maturity. Subsequent work demonstrated that low P300 amplitude could be observed also in young offspring of alcoholics, both male and female, who had not yet started drinking. It is consequently suggested that such neurophysiological finding might be a biological marker of biological vulnerability to alcoholism. (30) However, P300 anomalies are not specific to alcoholism and are observed also in other psychiatric disorders.

Molecular genetics

As is the case with most health disorders, a very active search is being conducted to identify the genes responsible for the alcohol misuse phenotype. It was originally reported that a variant (Taql-A1 allele) of the human dopamine receptor gene DRD2 was more prevalent in alcoholics than in controls (linkage disequilibrium), and that this allele could be a marker for heightened responsiveness to pharmacological stimulation (i.e. increased baseline positive reinforcement). However, subsequent research has shown that such genetic variance exists across ethnic groups, and that it is not even a consistent finding in all samples of alcoholics. (31)

There is little doubt that alcohol abuse will be found to be a polygenic disorder. The genetically influenced traits assumed to underlie responses to alcohol are quantitative traits. A section of DNA in a chromosome thought to influence a particular quantitative trait is known as a quantitative trait locus. The mapping of quantitative trait loci permits to locate and measure the effects of a single quantitative trait locus on a trait (phenotype). The quantitative trait locus findings available to date pertain mostly to laboratory animals, and markers of genetically transmitted responses such as alcohol-induced hypothermia and alcohol drinking preference have been reliably located in specific chromosomes. (28)

The first genome-wide screens for alcoholism in humans have produced evidence of several chromosomal regions linked to alcohol dependence. A conclusive finding so far is a ‘protective' locus near the alcohol dehydrogenase gene cluster in chromosome 4. (32) Ethanol and acetaldehyde metabolizing enzymes, of course, are known to vary across different ethnic populations, and such heterogeneity is assumed to play a role in the cross-ethnic variance of alcoholism prevalence rates. (33) The Collaborative Study on the Genetics of Alcoholism researchers report (34) that the linkage analysis of severe alcohol dependence has identified a locus on chromosome 16, near the marker D16S675.

Bullet list 2: Evidence of a constitutional predisposition to alcohol misuse

  • Significantly increased concordance rates in monozygotic twins
  • Elevated risk for children of alcoholics, even when raised by adoptive parents
  • Greater innate tolerance for intoxicating effects of alcohol
  • Associated biochemical and neurophysiological anomalies (markers) Linkage of both resilient and vulnerable phenotypes with chromosomal loci.

Psychological factors


It is now widely accepted that alcoholics do not present a homogeneous premorbid personality profile. However, some distinctive trait clusters have been identified which seem to characterize different types of alcoholics. (35) One such group (type 1) tend to score low in novelty seeking and high in harm avoidance and reward dependence. Another group (type 2) is formed by the natural thrill seekers, who appear to ignore harmful consequences and punitive responses. This latter cluster, which prevails mostly in males with early-onset alcoholism, is also typical of antisocial personalities. Of all personality features, conduct disorder and antisocial behaviour are the strongest predictors of alcohol misuse. (36) However, more than half the alcoholic population do not have such a personality background, presenting rather with a non-specific mixture of the different personality types described in clusters A, B, and C of the DSM-IV classification. (37)

Psychodynamic processes

In keeping with a ‘topographic' notion of psychic structure, early psychodynamic writings viewed alcoholism and other addictions as regressive behaviours caused by unconscious conflicts about libidinal pleasures, homosexuality, and aggression. More recent formulations emphasize ego and self-developmental problems, and consider psychoactive substance abuse as a response to psychological suffering; an attempt at re-establishing homeostasis. This is known as the self-medication hypothesis of addictions, (38) according to which, persons with self-regulatory deficiencies in the areas of self-care, self-esteem, self-object relations, and affect tolerance, would drink to palliate their distress.


An important sector of the scientific community considers alcohol abuse as a behavioural pattern which has been learned through mechanisms of classical (i.e. Pavlovian) and operant conditioning. According to this interpretation, the perpetuation of heavy drinking results from its association with conditioned stimuli (cues), and from the action of positive (pleasant effects) or negative (stress reduction) behavioural reinforcement. (39) Additional components of this equation are the so-called alcohol ‘expectancies'. Alcohol abusers tend to overemphasize the pleasant aspects of drinking and to exclude the negative ones; the learning theory of alcoholism assumes that such a cognitive set is also acquired through social exposure. (28)

Psychiatric comorbidity

Community and clinical epidemiology findings point to the presence of other psychiatric disorders as one of the most significant psychological risk factors in alcoholism. The risk is particularly high in persons with schizophrenia, bipolar disorder, major depression, social phobia, panic disorder, post-traumatic stress, attention-deficit hyperactivity disorder, and antisocial and borderline personality disorders. (40)

A major confounder in the interpretation of these findings is the poor specificity of psychiatric symptoms in the alcoholic population. A large proportion of the disorders diagnosed are alcohol induced and tend to dissipate in conditions of abstinence. This evidence led some authors to conclude that most of the excess psychopathology observed in alcoholics is secondary to alcoholism rather than a pre-existing risk factor. (41) It has also been suggested that the coexistence of alcoholism with other psychiatric illnesses (e.g. affective disorders) does not necessarily mean that one is causing the other, but rather that they both result from a common genetic influence. (42)

Brain dysfunction

Altered neuropsychological function can be seen as an additional risk factor in alcoholism: minimal brain damage, attention deficit, learning disabilities, head injuries, fetal alcohol effects, or the actions of other drugs of abuse are examples of brain conditions likely to increase individual vulnerability. Moreover, a transketolase deficiency (possibly genetic), which affects carbohydrate metabolism in the brain, is believed to predispose towards the occurrence of alcoholic organic brain complications. (43)


Several risk factors have been identified which appear to predict the occurrence of the following alcohol-related phenomena: excessive drinking, untoward drunken behaviour, the development of alcohol dependence and its severity, alcohol-induced organ damage, and some neuropsychiatric complications. However, the causal link between such factors and their putative consequences is only partially established, and a sizeable portion of the variance remains largely unexplained. 


1. Edwards, G., Anderson, P., Babor, T.F., et al. (1994). Alcohol policy and the public good. Oxford University Press.

2. National Institute on Drug Abuse (1998). Monitoring the future study: trends in drug use among 8th, 10th and 12th graders. NIDA Notes, 13, 15.

3. Health and Welfare Canada (1992). Alcohol and other drug use by Canadians: a national alcohol and other drug survey (1989) technical report. Health and Welfare Canada,. Ottawa.

4. Hanna, E.Z., Chou, S.P., and Grant, B.F. (1997). The relationship between drinking and heart disease morbidity in the United States: results from the National Health Interview Survey. Alcohol: Clinical and Experimental Research, 21, 111–18.

5. Royal College of General Practitioners (1986). Alcohol—a balanced view. Royal College of General Practitioners, London.

6. Royal College of Physicians and Surgeons of Canada (1990). Health and public policy committee statement on alcohol and alcohol-related problems. Royal College of Physicians and Surgeons of Canada, Ottawa.

7. O'sConnor, P.G. and Schottenfeld, R.S.(1998). Patients with alcohol problems. The New England Journal of Medicine, 338, 592–602.

8. Single, E., Robson, L., Xie, X., and Rhem, J. (1996). The costs of substance abuse in Canada. Canadian Centre on Substance Abuse, Ottawa.

9. Marlatt, G.A. and Donovan, D.M. (1982). Behavioural psychology approaches to alcoholism. In Encyclopedic handbook of alcoholism (ed. E.M. Pattison and E. Kaufman), pp. 560–76. Gardner Press, New York.

10. Annis, H.M. and Davis, C.S. (1988). Assessment of expectancies. In Assessment of addictive behaviors (ed. D.M. Donovan and G.A. Marlatt), pp. 84–111. Guilford Press, New York.

11. Keller, M. (1976). The disease concept of alcoholism revisited. Journal of Studies on Alcohol, 37, 1694–717.

12. Jellinek, E.M. (1960). The disease concept of alcoholism. Hillhouse Press, New Brunswick, NJ.

13. Lyvers, M. (1998). Drug addiction as a physical disease: the role of physical dependence and other chronic drug-induced neurophysiological changes in compulsive drug self-administration. Experimental and Clinical Psychopharmacology, 6, 107–25.

14. Blacker, E. (1966). Sociocultural factors in alcoholism. International Psychiatry Clinics, 3 (2), 51–80.

15. Hemmingsson, T., Lundberg, I., Romelsjo, A., et al. (1997). Alcoholism in social classes and occupations in Sweden. International Journal of Epidemiology, 26, 584–91.

16. Mandell, W., Eaton, W.W., Anthony, J.C., et al. (1992). Alcoholism and occupations: a review of 104 occupations. Alcohol: Clinical and Experimental Research, 16, 734–46.

17. Dawson, D.A., Grant, B.F., Chou, S.P., et al. (1995). Subgroup variation in U.S. drinking patterns: results of the 1992 national longitudinal alcohol epidemiological study. Journal of Substance Abuse, 7, 331–44.

18. Merton, R.K. (1972). Social theory and social structure. Free Press, New York.

19. Heath, D.B. (1993). Recent developments in alcoholism: anthropology. Recent Developments in Alcoholism, 11, 29–43.

20. Dawson, D.A., Harford, T.C., and Grant B.F. (1992). Family history as predictor of alcohol dependence. Alcoholism: Clinical and Experimental Research, 16, 572–5.

21. Bierut, L.J., Dinwiddie, S.H., Begleiter, H., et al. (1998). Familial transmission of substance dependence: a report from the Collaborative Study on the Genetics of Alcoholism. Archives of General Psychiatry, 55, 982–8.

22. Heath, A.C. (1995). Genetic influences in alcoholism; a review of adoption and twin studies. Alcohol Health and Research World, 19, 166–71.

23. Kendler, K.S., Heath, A.C., Neale, M.C., et al. (1994). A twin-family study of alcoholism in women. American Journal of Psychiatry, 151, 707–15.

24. Allgulander, C., Nowak, J., and Rice, J.P. (1992). Psychopathology and treatment of 30 344 twins in Sweden. Acta Psychiatrica Scandinavica, 86, 421–2.

25. Cadoret, R.J., Yates, W.R., Troughton, E., et al. (1995). Adoption study demonstrating two genetic pathways to drug abuse. Archives of General Psychiatry, 52, 42–52.

26. Schuckit, M.A., Tsuang, J.W., Anthenelli, R.M., et al. (1996). Alcohol challenges in young men from alcoholic pedigrees and control families: a report from the COGA project. Journal of Studies on Alcohol, 57, 368–77.

27. Finn, P.R., Earleywine, M., and Pihl, R.O. (1992). Sensation-seeking, stress reactivity and alcohol dampening discriminate the density of a family history of alcoholism. Alcoholism: Clinical and Experimental Research, 16, 585–90.

28. National Institute on Alcohol Abuse and Alcoholism (1997). Ninth Special Report to the US Congress on Alcohol and Health From the Secretary of Health and Human Services. US Department of Health and Human Services, Washington, DC.

29. Anthenelli, R.M., Tipp J., Li, T.K., et al. (1998). Platelet monoamino oxidase (MAO) activity in subgroups of alcoholics and controls: results from the COGA study. Alcoholism: Clinical and Experimental Research, 22, 598–604.

30. Hill, S.Y., Steinhouser, S.R., Lowers, L., et al. (1995). Eight year longitudinal follow-up of P300 and clinical outcome in children from high risk for alcoholism families. Biological Psychiatry, 37 (11), 823–7.

31. Gelernter, J., Kranzler, H., and Cubells, J.F. (1998). DRD2 allele frequencies and linkage disequilibria, including the -141C Ins/Del promoter polymorphism, in European-American, African-American and Japanese subjects. Genomics, 51, 21–6.

32. Goate, A.M. and Edenberg, H.J. (1998). The genetics of alcoholism. Current Opinion in Genetics and Development, 8, 282–6.

33. Maezawa, Y., Yamauchi, M., Toda, G., et al.(1995). Alcohol metabolizing enzyme polymorphisms and alcoholism in Japan. Alcoholism: Clinical and Experimental Research, 19, 951–4.

34. Foroud, T., Bucholz, K.K., Edenberg, H.J., et al. (1998). Linkage of an alcoholism-related severity phenotype to chromosome 16. Alcoholism: Clinical and Experimental Research, 22, 2035–42.

35. Cloninger, C.R. (1987). Neurogenetic adaptive mechanisms in alcoholism. Science, 236, 410–16.

36. Harford, T.C. and Parker, D.A. (1994). Antisocial behaviour, family history and alcohol dependence syndrome. Alcoholism: Clinical and Experimental Research, 18, 265–8.

37. Driessen, M., Veltrup, C., Wetterling, T., et al. (1998). Axis I and Axis II comorbidity in alcohol dependence and the two types of alcoholism. Alcoholism: Clinical and Experimental Research, 22, 77–86.

38. Khantzian, E.J. (1985). The self-medication hypothesis of addictive disorders. American Journal of Psychiatry, 142, 1259–64.

39. Wilson, G.T. (1988). Alcohol use and abuse: a social learning analysis. In Theories on alcoholism (ed. C.D. Chaudron and D.A. Wilkinson), pp. 239–287. Addiction Research Foundation, Toronto.

40. Kessler, R.C., McConagle, K.A., Zhao, S., et al. (1994). Lifetime and 12 month prevalence of DSM-III-R psychiatric disorders in the United States. Archives of General Psychiatry, 51, 8–19.

41. Schuckit, M.A. and Hesselbrock, V. (1994). Alcohol dependence and anxiety disorders: what is the relationship? American Journal of Psychiatry, 151, 1723–34.

42. Winokur, G., Coryell, W., Endicott, J., et al. (1996). Familial alcoholism in manic-depressive (bipolar) disease. American Journal of Medical Genetics, 67, 197–201.

43. Martin, P.R., Adinoff, B., Weingartner, H., et al. (1986). Alcoholic organic brain disease: nosology and pathophysiologic mechanisms. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 10, 147–64.