With the advent of new and powerful techniques in molecular genetics, the family study approach has become an increasingly important method of exploring the causes and mechanisms behind a wide variety of common, complex diseases. Foremost among these disorders are psychiatric illnesses, including alcoholism, and, to a lesser extent, other substance addiction.
Although many such disorders are known to aggregate within families, how they are transmitted and how environmental factors interact with inborn traits to cause the illnesses is not yet known. While proof of genetic causation of addiction is not necessary to validate the concept of addiction as a disease process, by learning more about the inheritance of vulnerability to addiction, clinicians should be able to define it more accurately, to subdivide addiction, and eventually to devise more effective means of treatment.
Family studies have played a significant role in expanding our understanding of the pathology and transmission of a number of medical illnesses, ranging from those based on changes in one allele, such as sickle cell disease, to those related to complex gene-environment interactions, such as cardiovascular disease, certain cancers, or diabetes.1 It is perhaps in the field of psychiatry, however, that family studies have most greatly contributed to an understanding of the validity and etiology of illness.
As Robins and Guze2 pointed out 20 years ago, family studies are a necessary step in the process of validating psychiatric illnesses and delineating them from other disorders. They specified that while the familiality of an illness could be accounted for by either environmental or hereditary factors (or both):
Independent of the question of etiology . . . the finding of an increased prevalence of the same disorder among the close relatives of the individual patient strongly indicates that one is dealing with a valid entity.
This criterion, certainly, is fulfilled by alcoholism. Despite varying criteria and definitions used, despite differing study methods, despite a variety of populations investigated, alcoholism has been shown repeatedly to be much more common among relatives of alcoholics than in the general population.'5 Furthermore, while alcoholism and other substance addiction is substantially more common among men than among women, the alcoholic's gender has not been found to influence the risk of alcoholism in relatives. This indicates dial while gender influences the liability of substance abuse, inheritance is not due to sex-linked genetic influences.4
By itself, the finding of familiality of an illness says nothing about mode of inheritance or transmission. To be sure, examination of the pedigrees of affected subjects may suggest patterns of inheritance consistent, for example, with a simple Mendelian trait; certainly many other medical illnesses such as color blindness, cystic fibrosis, and Huntington's chorea were recognized as being under genetic conü'ol in this manner. Sophisticated methods of segregation analysis are available to test hypotheses about forms of getietic inheritance of any disorder5 But the absence of such a pattern cannot be considered absence of genetic contribution. Incomplete penetrance of a characteristic or criologie heterogeneity, for example, might obscure the mode of inheritance.
Furthermore, it is increasingly well recognized that manifestation of most common illnesses depends on an interaction between biological factors in the host (differences in degree of resistance or susceptibility) and environmental exposure. At the simplest extreme, insulindependent diabetes has been theorized to result from viral infection of a genetically susceptible host6 - an example of a process requiring genetic predisposition plus exposure to an outside triggering factor.
There are more complex patterns of illness, as well. For example, emphysema is associated with longterm exposure to tobacco smoke, but the threshold required to provoke clinically significant lung changes is a biological variation among individuals. This is also an example of a medical disorder whose onset and course is directly influenced by behavior: the act of smoking brings about the exposure to smoke that causes the pathological changes in lung tissue.
The disorder (or group of disorders) that we call addiction may be seen as another example of this interaction between inborn, presumably genetic, variables and postnatal environmental influences. Since environmental factors relating to exposure to alcohol or other drugs are an obvious requirement for the development of substance abuse, the hypothesis of geneenvironment interaction therefore depends on our ability lo demonstrate the existence of genetic influences on substance abuse. Furthermore, the relative contribution of pre- and postnatal factors has long been debated.
While family studies alone usually cannot definitively seule the issue of mode of inheritance (genetic or learned), the degree to which traits are passed from generation to generation can be quantified and modeled.7 Reich el al8 have recently applied this model to the inheritance of alcoholism, showing that the transmissibility of the disorder appears to be increasing over time, an indication that purely genetic factors cannot fully explain the etiology of the illness.
While a substantial body of literature supports the role of genetic factors in alcohol and drug tolerance and drug-seeking behavior in laboratory animals,9 similar evidence for transmission of other drug addiction in humans has been harder to come by, for a number of reasons. Undoubtedly, as for alcoholism, part of the difficulty is due to secular trends in drug use. The rapid change in patterns of use over time makes adequate measure of the transmissibility between generations of specific drug addiction extremely difficult.
Drugs which a few years ago were difficult or expensive to obtain, such as cocaine, are now much more easily acquired, and in more potent forms, such as "crack" cocaine. Thus, individuals with a (hypothesized) genetic vulnerability' to addiction to a specific drug, such as cocaine or opioids, might in the past have been protected from expressing such a trail simply because they never had a chance to be exposed to the drug. Now, compared to even a few years ago, many more people have been exposed and are therefore at risk for development of addiction.10 If a powerful, highly heritable tendency to addiction existed, its existence might nonetheless have been obscured by the fact that nongenetic factors of exposure influenced or prevented its expression.11
Selected Family Studies of Drug Abuse
Secondly, drug addiction (compared to alcoholism) is relatively rare in the general population; the lifetime prevalence of opioid abuse and dependence, for example, was estimated to be 0.7% by the Epidemiologic Catchment Area study,12 versus 1 1-16% for alcohol abuse and dependence.1'1 Its relative rarity, coupled with legal sanctions against drug use and its consequent "underground" nature1" tend to make recruiting subjects and gathering accurate information more difficult.
Testing hypotheses about transmission of specific forms of drug addiction is complicated by the fact that drug users tend not to restrict themselves to one class of drug.1 ' thus making it difficult, if not impossible, to study the transmission of addiction to a single drug. As a result, most studies of familiality of drug addiction have either been forced to assume a single underlying vulnerability to alcohol and other drugs, or have lumped all illicit drugs together regardless of pharmacologic mode of action.
Because of differences in methodology, definitions of abuse, subject selection, restrictions in sample site, and inadequate or absent controls, it is difficult to reach firm conclusions about the inheritance of drug addiction: hypotheses ranging from parent-to-child transmission of vulnerability to addiction to a specific drug, to the possibility of inheritance of a generalized vulnerability to psychoactive substance addiction have been entertained,15-20 but so far have lacked rigorous testing (Table).
A final methodological difficulty common to all studies of the transmission of vulnerability to substance addiction, but apparently worse in the case of illicit substances, is the presence of substantial comorbidity, particularly depression and antisocial personality, among substance abusers.21-24 This adds yet another layer of complexity to analyses of the transmission of susceptibility to addiction, since the contribution of the comorbid conditions must also be considered."11
SEPARATION OF GENETIC AND ENVIRONMENTAL FACTORS
To settle the issue of inheritance, environmental factors must be separated out from inborn (presumably genetic) traits. In order to do this, two experimental designs have been employed: twin studies and adoption studies. In the first, concordance of a characteristic among monozygotic twins, who share an identical genetic makeup, is compared to concordance for that character in same-sex dizygotic twins, who on average share only half of their genes identical by descent. If environmental factors are considered roughly equal for twins. any excess concordance among monozvgotes would be evidence in support of genetic influences. Adoption studies also separate pre-and postnatal influences if the adoptees are raised by parents not biologically related to them. In this design, excess concordance with adoptive parents would argue for environmental influences, while concordance with birth parents would argue for genetic ones.
Both strategies, however, have limitations. For example, monozygotic twins, because of their closer resemblance, might be treated more similarly than dizygotic twins. Also, separation of adoptees from birth parents may occur at a variable time after birth, thus confounding genetic and environmental effects. Therefore, such studies should be interpreted with caution.25,26 Nonetheless, adoption and twin research results strongly favor the existence of inborn factors contributing to the development of alcoholism. For other drug addiction, evidence is scantier, but at least one adoption study27 suggests the existence of heritable factors.
However, as with other complex disorders having a strong genetic, component such as diabetes mellitus, obesity, or hypertension, it seems clear that such (actors are substantially modified by postnatal experience. Current theories regarding the etiology of insulin-dependent diabetes, for example, suggest that exposure to viral infection or other environmental factors is needed in genetically susceptible subjects for the disease to manifest.6 Even in the case of genetically simple disorders caused by a single allele, variation in severity, age of onset, or other aspects of the disease may occur, indicating the potential for substantial modification along the causal chain stretching from initial expression of the gene to the final, overt expression of the disease.28
For alcoholism, there is evidence that the final phenotype may be a result of heterogeneity at the genetic level as well as complex geneenvironment interactions. While most past adoption studies demonstrated elevated risk for alcoholism among adopted-awav children of alcoholics.29-31 the role of environmental factors has been difficult to evaluate. To more precisely study environmental and genetic influences on development of alcohol abuse. Cloninger and colleagues32-34 employed a crossfostering study design that followed all 862 men and 913 women of known paternity born to single women in Sweden and adopted by nonrelatives at an early age. They found that alcohol abuse by adoptive parents did not increase risk for alcohol abuse in their adoptive children, indicating no appreciable "'modeling" effect in children due to imitation of their parents.
Biologic background (having birth parents with alcohol abuse) did. however, exert an influence on risk for alcohol abuse in the offspring. In fact, the authors identified two subtypes with differing patterns of inheritance and different dependence on postnatal environment. The more common form, identified as Type I. is characterized by later ousel of drinking problems and appears to be related to loss of control over drinking. Ii also depends lo a .significant degree on gene-environment interaction to manifest: depending on the postnatal environment, the result may be either mild or severe alcohol abuse. Having both the Type 1 genetic background and exposure to provocative environmental factors more than doubled the risk of severe alcohol abuse, from 4.3% of men without either risk factor to 1 1.6% of those with both.
Bv contrast, Type 2 alcoholism is characterized by spontaneous alcohol-seeking, earlier onset ol alcohol abuse, and criminal behavior in the biologic father. Unlike Type I. Type 2 alcoholism appears to be transmitted primarily to men, and postnatal influences appear substantially less important. Men without genetic or environmental backgrounds had a 1.9% rale of Type 2 alcohol abuse, while for those with the genetic background, the risk was increased nine-fold, with little difference between those with and without the associated postnatal factors.
Based on these findings and other family studies. Cloninger35.36 concluded that risk for alcoholism is mediated in large part by inborn. heritable differences in temperament and learning styles. Evidence from twin and adoption studies is emerging to suggest thai final expression of these trails, too, depends on complex gene-environment interactions.37
FAMILY STUDIES AND GENETIC EPIDEMIOLOGY
With the recent advances in molectilar genetics has come a tremendously expanded ability to link phenotypic characters to specific chromosomal locations. While association between complex diseases and certain alleles has been demonstrated in a more general wav (for example, finding an association between risk for diabetes and presence of certain HIA antigens''), unequivocal evidence of genetic linkage in psychiatric disorders remains to be established. It is hoped that complex characteristics, or aspects of them, may ultimately be firmly linked to specific genes, and genetic studies with this goal are under wav currently for Alzheimer's disease, manic-depressive illness, schizophrenia, and alcoholism. For this tvpe of investigation, which nltimaielv requires measurement of both genetic and environmental factors, family studies are an unparalleled resource.38
In the case of alcoholism, certain aspects (at least) of the phenotype mav likely prove to be under relatively simple genetic control, tillimalelv allowing for the elucidation, for example, of the biochemical and molecular genetic basis for certain (onus of end-organ damage associated with alcoholism.39,40 Efforts have also been inaile to identify' biochemical, behavioral, and electrophysiological markers for those at risk for alcoholism, and over the next few years it is likely that many of these, as well, will be investigated for their mode of inheritance and potential linkage to specific chromosomal sites.11,12
For drugs other than alcohol, the genetic pattern may be similar. As with alcoholism, genetic factors may influence the development of addiction by controlling phenomena such as tolerance to drug effects, the rate of drug metabolism, or idiosyncratic reactions to drug use; they may also play a substantial role in the development of specific complications later in the course of addiction. More generally, genetic factors may influence personality style and thus more subtly influence the development of addiction by increasing the likelihood of exposure to novel stimuli such as drugs (drug-seeking behavior) or the degree to which the individual responds to social reinforcement (positive or negative) related to drug use. Finally, genetic factors may operate not only by promoting drug use. but by eliminating factors that protect against addiction, such as by ameliorating adverse effects of drug use.13 Thus, it is likely that, for any addiction, a multitude of protective and promoting factors will be found at the genetic level, with varying degrees of specificity of effect and sensitivity to environmental influence.
The clinical entity that we call "alcoholism" is a mixture of symptoms including:
* various types of end-organ damage (hepatic cirrhosis, cardiomyopathy),
* subjective feelings (guilt, loss of control),
* judgments by others (family objeclions, loss of friends), and
* behaviors that can be objectively defined (arrests for DWT binge drinking, morning drinking).
Addiction to other substances produces svmptoms either identical or which (as in the case of withdrawal phenomena) vary depending on the drug's specific pharmacologic properties. Should this heterogeneous collection of problems be lumped together as a disease? Does the research evidence that suggests the role of inborn, heritable tendencies toward substance dependence justify our defining it as a disease?
The purpose of disease classification is to allow clinicians to differentiate a given pathologic state horn others, so that its course mav be predicted with reasonable accuracy and proper treatments initiated. Diagnosis further allows the clinician to group pathologic states into homogeneous categories so that etiology may be explored, the effectiveness of treatment tor specific conditions evaluated, and superior interventions found.
In the final analysis, defining a clinical condition as a disease is an exercise in pure empiricism: Does use of a given set of criteria, choosing some svmptoms and not others, help the clinician in prognosis and selection of treatment? The five steps outlined by Robins and Guzefor establishing diagnostic validity (clinical description, use of diagnostic iesis. delimitation from other disorders, follow-up, and family studies) are merely practical steps to achieve these ends.
Thus, the justification for considering alcoholism or other substance addiction a disease is practical. It gives the clinician information about prognosis, potential complications, and what sort of treatment is (or is not) likely to help. Applying the "disease concept" to alcoholism helps clinicians in this manner.4"' For this purpose, the finding of a heritable, biologically based tendency toward addiction is helpful in validating the concept of altoholism, but is not necessary in order to consider alcoholism or other drug addiction a disease. But the existence of such characteristics is of great significance, nonetheless.
We may conceptualize the clinical entity of alcoholism and other drug addiction as being at the end of a lengthy causal chain beginning with the expression of genes and gene products, which subsequently interact in a highly complex way with a variety of environmental influences. Tracing the chain of events leading to alcoholism should enable us to better understand the process by which it manifests, and should ultimately allow us to improve our definition of the illness by dividing it into more homogeneous subtbrms. As we gain a deeper understanding of the interacting biological, psychological, and social factors contributing to substance abuse, we should be able to better identify persons at high risk and design superior interventions at both the biological and behavioral levels for prevention and treatment.15
1. Riseli N. Genetic linkage and complex discuses, with special reference to psychiatric disorders. Genet Epidemiol. 1990: 7:316.
2. Robins E, Gaze SB. Establishment of diagnostic validity in psychiatric illness: ils application to schizophrenia. Am J Psychiatry 1970: 126 (7) : 107-111.
3. Cotton NS. The familial incidence of alcoholism. J Stud Alcohol. 1979; 10(1);89116.
4. Dinwiddie SH, Cloninger CR. Family and adoption studies ol alcoholism. In: Goedde HW. Agarwal DP. eds. Alcoholism Biomedical And Genetic Aspects. New York. NV: Pergamon Press Inc: 1989:259-276.
5. Rice J. A quantitative genetic approach to psychiatric research. In: Gibbons RD. Dvskan MW, eds. Statistical and Methodological Advances in Psychiatric Research. New York, NY: SP Medical & Scientific Books: 1983:91-129.
6. Kuller LH. Becker DJ. Cruickshanks KJ. el al. Evolution of the Pittsburgh studies of the epidemiology of insulin-dependent diabetes mellitus. Genet Epidemiol. 1990:7:105-119.
7. Rice JP. Genetic epidemiology: models ol multifactorial inheritance and path analysis applied to qualitative traits. In: Moolgavkar SII, Prentice RI.. eels. Modern Statistical Methods in Chronic Disease Epidemiology New York, NY: John Wilev & Sous: 1986:225-243.
8. Reich T. Cloninger CR. van Ecrdewegh P. Rice JP. Mullanev J. Secular trends in the familial transmission of alcoholism. Alcohol Clin Exp Res. 1988: 12(4):458-164.
9. Crabbe JC, McSwigan JD. Belknap JK: The role of genetics in substance abuse:. In: Galizio M, Maisto SA. eds. Determinants of Substance Abuse, New York, NY: Plenum Press: 1985:13-61.
10. Kozel NJ. Adams EH. Epidemiology of drug abuse: an overview. Science. 1989: 234:970-974.
11. Rotmsaville BJ. The role of psychopathology in the familial transmission of drug abuse. In: Pickens RW. Svikis DS. eds. Biological Vulnerability to Drug Abuse. Washington, DC: National Institute on Drug Abuse: 1988: 108-119. NIDA Research Monograph 89.
12. Anthony JC, Helzer JE. Syndromes of drug abuse-dependence, in: Robins LN. Regier DA eds. Psychiatric Disorders in America. New York. NY: The Free Press: 1990:1 16-154.
13. Robins L.N. Helzer JE, Weissman MM. el al. Lifetime prevalence of specific psychiatric disorders in three sites. Arch Cen Psychiatry 1984: 41(10):9 19-956.
14. Dinwiddie SH, Reich T, Cloninger CR. The relationship of solvent use to other substance use. Am J Drug Alcohol Abuse. 1991: 17(2): 173-186.
15. Scherer SE. Mukherjee BN. "Moderale" and hard drug users among college students: a study of their drug use patterns and characteristics. Br J Addict. 1971: 60:250-269.
16. Smart RG. Fejer D. Drug use among adolescents and their parents. J Abnorm Psychol. 1972; 79: 153-160.
17. Scherer SE. Self-reported parent and child drug abuse. Br J Addict. 1973: 08:303364.
18. Annis HM. Patterna of infrafamilial drug use. Br J Addict. 1971; 09:301-309.
19. Fawzi IF, Coombs RH. Gerber B. Generational continuity in the use of substances: the impact of parental substance use on adolescent substance use. Addict Behim 1983: 8:109-114.
20. Meiler WH. Rinehart R. Cadoret RJ. Troughton F. Specific familial transmission in substance abuse. InI J Addict. 1988; 23(10): 1 029- 1 039.
21. Khaiuzian FJ. Trecce C. DSM-III psychiatric diagnosis of narcotic acidic Is: recent findings. Arch Gen Psychiatry. 1985: 42(11): 1 007-1 1)71.
22. Mirin SM, Weiss RD. Mie had J. Psvchopathology in substance abusers: diagnosis and treatment. Am. J Drug Alcohol Abuse. 1988; 14 (2): 139-157.
23. Rounsaville BJ. Weissinau MM. CriisChristophi K, Wilber C. Weber H. Diagnosis and symptoms ol depression in opiate addicts. Arch Gen Psychiatry 1982: 39(2): 151-156.
21. Lewis CE. Halikas JA. Morse C. Rinunci JD. Alcoholism in narcotic addicts with antisocial personality. Br J Addict. 1987: 82:305-311.
25. Gotlesman II: Exploring drug abuse with genetic strategies: cautionary tales. In: Pickens RW Svikis DS, eds. Biological Vulnerability Io Drug Ahnst: Washington, DC.: National Institute on Drug Abuse; 1988:150-104. NIDA Research Monograph 89.
26. Svikis DS, Pickens RW. Methodological issues in family, adoption, and twin research. In: Svikis DS. Pickens RW (eds). Biological Vulnera hi lily Io Drug Abuse. Washington, DC: National Institute on Drug Abuse: 1988:120-133. NIDA Research Monograph 89.
27. Cadore I RJ, Troughton K. O'Gorman TVA'. Hevwood K: An adoption study of genetic and environmental factors in drug abuse. Arch Gen Psychiatry 1980: 43(12):1 131-1 136.
28. Propping P. Studying biological vulnerability and heterogeneity - a long way ahead. J Psychiatrie ReS. 1987: 21 (4):473478.
29. Goodwin DW, Schulsinger F, Hermansen I.. Guze SB. Winokur G. Alcohol problems in adoptees raised apart from alcoholic biological parents. Arch Gen Psychiatry. 1 973: 28:238-243.
30. Goodwin DW. Adoption studies of alcoholism. In: Goedde HW, Agarwal DP. eds. Cenetics and Alcoholism. New York. NY: Alan RLiss Inc. 1987: 61-70.
31. Cadoret EJ, Cain CA. Grove WM. Development of alcoholism in adoptees raised apart from alcoholic biologic parents. Arch Gen Psychiatry 1980; 37(5) :56 1-563.
32. Cloninger CR, Bohumil .M, Sigvardsson S. Inheritance of alcohol abuse: crossfostering analysis of adopted men. Arch Got Psychiatry. 1981; 38(8):86l-868.
33. Bobinan M. Sigvardsson S. Cloninger CR. Maternal inheritance of alcohol abuse: cross-fostering analysis of adopted women. Arch Gen Psychiatry. 1981: 38(9):965-969.
34. Cloninger CR. Bobinan M, Sigvardsson S, von Knorring A-I.. Psychopathology in adopted-out children of alcoholics: the Stockholm adoption study. Recent Dec Alcohol. 1985:3:37-51.
35. Cloninger CR. Neurogenetic adaptive mechanisms in alcoholism. Science. 1987: 236:410-416.
36. Cloninger CR. Enologie factors in substance abuse: an adoption study perspective. In: Pickens RW, Svikis DS. eds. Biological Vulnerability to Drug Abuse. Washington, OC: National Institute on Drug Abuse; 1988:52-72. NEDA Research Monograph 89.
37. Kaprio J, Koskenvuo M. Rose Rf. Change in cohabitation and in ua pair similarity of monozygotic (MZ) convins for alcohol use. extraversión, and neuroiicism. Behav Genet. 1990: 20(2):265-276.
38. Dorman JS, Trueco M. LaPorte RE. Kuller LH. Family studies: the key to understanding the genetic and environmental etiology of chronic disease? Genet Epidemiol. 1988: 5:305-310.
39. Lange LC, Bergmann SR. Sobe] BE. Identification of fatty acid ethvl esters as products of rabbit myocardial ethanol metabolism. J Biol Chum. 1981; 256:968973.
40. Blass [D. Gibson GE. Abnormality of a thiamine-requiring enzyme in patients with Wernicke-Korsakoff syndrome. N Engl J Med. 1 977: 297: 1 367-1370.
41. Begleiter II, Potjesz B. Neuroelei trie processes in individuals at risk for alcoholism. Alcohol Alcohol. 1990; 25:251-256.
42. Schuckit VLA. Biological markers in alcoholism. Prog Neuropsychopharmacol Biol Psychiatry. 1986; 10:191-199.
43. Pickens RW. Svikis DS. Genetic vulnerability to drug abuse. In: Pickens RW. Svikis DS. eds. Biological Vulnerability Io Drug Abuse. Washington. DC: National Institute on Drug Abuse; 1988:1-7. NIDA Research Monograph 89.
44. Guze SB. Cloninger CR, Martin R, Clayton PJ. .Alcoholism as a medical disorder. Compi Psychiatry 1986: 27(0):501-510.
45. Cloninger CR. Dinwiddie SH, Reich T. Epidemiology and genetics of alcoholism. In: Tasman A, Hales RE, Frances AJ, eds. Review uj Psychiatry, VHI. Washington, DC: ARA Press Ine,' 1989; 293-308.
Selected Family Studies of Drug Abuse