genetic epidemiology of psychiatric disorders · tical manual of mental disorders, 3rd edition, ......

Eptdemiologic ReviewsCopyright O 1997 by The Johns Hopkins University School of Hygiene and Public HealthAll rights reserved

Vol. 19, No. 1Printed In U.SA.

Genetic Epidemiology of Psychiatric Disorders

Kathleen Ries Merikangas and Joel David Swendsen

INTRODUCTION

Perhaps more than any other domain, the search foretiologic factors in psychiatry has been characterizedby an enduring "nature versus nurture" debate, withresearchers in medicine, psychology, and public healthtraditionally emphasizing separate causal factors in thegeneration of mental disorders. More recently, re-searchers have increasingly embraced a biopsychoso-cial model based on a growing body evidence thatpsychiatric disorders are determined by an interactionof multiple factors. Genetic epidemiology reflects thiscomprehensive approach by clarifying how both ge-netic and environmental components produce the com-plex phenotypes of various forms of disorders. How-ever, its application to psychiatry is still relatively newand its methods and contributions within this domainare not well understood. The purpose of this presen-tation is to review issues and methodological conceptsbasic to the genetic epidemiology of psychiatric dis-orders, and to present examples of its application tothis field.

DEFINITIONS OF PSYCHIATRIC DISORDERSAND PREVALENCE IN THE GENERALPOPULATION

Since the time of Sydenham's admonition to clas-sify diseases with the same care that botanists exhibitin the development of their phytologies (1), nosologybased on observable characteristics has been central toclinical medicine. Although progress in the character-ization and assessment of psychiatric syndromes has

Received for publication November 11, 1996, and accepted forpublication September 3, 1997.

Abbreviations: ABO, blood type locus,; ALDH, aldehyde dehy-drogenase; BRCA1, breast and ovarian cancer susceptibility gene;DRD2, dopamlne receptor gene; DSM-III-R, Diagnostic and Statis-tical Manual of Mental Disorders, 3rd edition, revised; DSM IV,Diagnostic and Statistical Manual of Mental Disorders, 4th edition;GABAA, gamma-aminobutyric acid beta 1 receptor genes; 5HT,serotonin receptor gene; ICD-10, The ICD-10 Classification of Men-tal and Behavioral Disorders.

From the Genetic Epidemiology Research Unit, Yale UniversitySchool of Medicine, New Haven, CT.

Reprint requests to Dr. Kathleen R. Merikangas, Director, GeneticEpidemiology and Psychiatry, Yale University School of Medicine,40 Temple Street, Suite 7B, New Haven, CT 06510-3223.

strengthened the reliability and validity of current no-sology, phenotypic imprecision has often been consid-ered as the major culprit in inconsistencies observed inpsychiatric genetic research. Much of the criticismleveled against early psychiatric genetic investiga-tions, and indeed biologic psychiatric studies in gen-eral, has concentrated on the subjectivity of psychiat-ric diagnosis. The accepted solution to this problemhas gradually developed over the past quarter century,reflecting a trend toward explicit "operational" defini-tions. That is, clinicians increasingly diagnose mentaldisorders from a clearly defined constellation of symp-toms that are experienced for a specified duration oftime. The first set of criteria in psychiatry that recog-nizably conformed to this pattern was developed byRobins and colleagues at Washington University in1972, and became widely known as the "Feighnercriteria" (2). Other sets of operational criteria rapidlyfollowed from other groups, culminating in publica-tion of the recent World Health Organization's TheICD-10 Classification of Mental and Behavioral Dis-orders (3) and the American Psychiatric Association'sDiagnostic and Statistical Manual of Mental Disor-ders: DSM-IV (4).

In summary, operational definitions such as thosecontained in "official" systems of classifications suchas DSM-TV (and more recently ICD-10) provide ac-ceptably high levels of reliability. Patients are, there-fore, more likely to be diagnosed in a similar mannerby different clinicians due to the use of uniform cri-teria and standards for diagnosis. However, the intro-duction of operational definitions for psychiatric re-search has appeared to do little to overcome a moredifficult obstacle, that of biologic validity. For exam-ple, when the threshold of diagnostic criteria for majordepression is systematically lowered, affected individ-uals still differ from those without depressive symp-toms according to numerous validators, including fam-ily history, longitudinal stability, suicide attempts, andpsychosocial impairment (5). In this way, operationaldefinitions of psychiatric disorders, although reliable,may not always reflect discrete and homogeneousclinical entities. A related problem is the lack of dis-tinct boundaries between psychiatric disorder catego-

144

Genetic Epidemiology of Psychiatric Disorders 145

ries. For example, many patients suffer from a condi-tion that is characterized simultaneously by bothanxiety and depressive symptoms, yet most nosologicsystems would classify these individuals as having twoseparate (albeit co-occurring) diagnoses. Kendell (6)noted that it is unlikely that the etiologic secrets ofmajor psychiatric disorders will be unlocked withoutaccurate and valid identification of the syndromesthemselves. While progress in classification is advanc-ing, the findings and conclusions of all psychiatricinvestigations must still be qualified against imperfectbiologic validity.

With the inherent problems of psychiatric nosologyin mind, information about the frequency of psychiat-ric disorders in the general population may be used toprovide an initial frame of reference for investigatingpatterns of familial aggregation in clinically ascer-tained samples. Table 1 presents the lifetime preva-lence rates from the most recent large-scale epidemi-ologic study of psychiatric disorders in the UnitedStates using contemporary diagnostic criteria (7).These data reflect retrospective estimates of 14 diag-nostic categories defined by the Diagnostic and Sta-tistical Manual of Mental Disorders: DSM-III-R (8),and were obtained from community residents aged15-54 years. The results indicate that almost half ofthe entire sample met diagnostic criteria for a psychi-atric disorder at some point over their lifetime. Fur-thermore, these prevalence rates of psychiatric disor-ders also vary by demographic characteristics such asage, ethnicity, and gender. For example, prominentgender differences for depression and alcoholism arepresented in table 1, and stratification by these vari-

ables is often required in epidemiologic research. Thehigh rates of disorder must also be interpreted withinthe context of psychiatric disorder severity; even themost common forms of disorder examined by theNational Comorbidity Survey (7) are often severeenough to prevent basic functioning in social, occupa-tional, and family domains.

The frequency and severity of psychiatric disordersin the general population offer compelling reasons topursue etiologic studies that may identify key vari-ables for disorder prevention and intervention. It isimportant to underscore, however, that data about therespective roles of genes and the environment may bedifficult to interpret. The expression of genetic factorsis rarely independent of the environment, and thediscussion of genetic or environmental forces, in theabsence of the other, may be limited in meaning fromboth the theoretical and clinical standpoints. It is forthis reason that research paradigms in genetic epide-miology are particularly useful for not only elucidatingthe separate roles of nature and nurture, but also theimportance of their interaction in determining pheno-typic expression.

STUDY PARADIGMS IN GENETICEPIDEMIOLOGY

Widi their roots in the methods of population andclinical genetics as well as chronic disease epidemiol-ogy, investigations in genetic epidemiology are typi-cally based on one of four research paradigms (a moredetailed description of these paradigms can be foundin Khoury et al. (9)). The basic strategy of these

TABLE 1. Roaulta from the National Comorbidity Survey*

Disorder

Anxiety disordersPanic disorderAgoraphobia without panicSocial phobiaGeneralized anxiety disorderAny anxiety disorder

Affective disordersMajor depressive episodeManic episodeAny effective disorder

Substance use disordersAlcohol dependenceDrug dependenceAny substance U3e disorder

Psychosis (nonaftective)

Any psychiatric disorder

Male

2.0 (±0.3)3.5 (±0.4)

11.1 (±0.8)3.6 (±0.5)

19.2 (±0.9)

12.7 (±0.9)1.6 (±0.3)

14.7 (±0.8)

20.1 (±1.0)9.2 (±0.7)

35.4 (±1.2)

0.6 (±0.1)

48.7 (±0.2)

Lletlme prevalence (standard error)

Female

5.0 (±1.4)7.0 (±0.6)

15.5 (±1.0)6.6 (±0.5)

30.5 (±1.2)

21.3 (±0.9)1.7 (±0.3)

23.9 (±0.9)

8.2 (±0.7)5.9 (±0.5)

17.9 (±1.1)

0.8 (±0.2)

47.3 (±1.5)

Total

3.5 (±0.3)5.3 (±0.4)

13.3 (±0.7)5.1 (±0.3)

24.9 (±0.8)

17.1 (±0.7)1.6 (±0.3)

19.3 (±0.7)

14.1 (±0.7)7.5 (±0.4)

26.6 (±1.0)

0.7 (±0.1)

48.0 (±1.1)

1 From Kessler et al. (7).

Epidemiol Rev Vol. 19, No. 1, 1997

146 Merikangas and Swendsen

designs is either to hold the environment constantwhile allowing genetic factors to vary, or the reverse.Each approach is characterized by inherent strengthsand limitations, and progress in all areas continues toclarify the etiology of psychiatric disorders.

Family studies

The observation that some disorders aggregate infamilies serves as prerequisite evidence suggesting apossible genetic component. The basic family studyapproach involves identifying individuals with a par-ticular psychiatric disorder (the proband) and thendetermining the rates of disorder in the proband'srelatives. These morbidity statistics can then be com-pared with the rates of disorder in families of unaf-fected individuals (controls). The common indicator offamilial aggregation is the prevalence ratio, which isdefined as the ratio of the prevalence rate of a disorderamong the relatives of cases to the prevalence rate ofa disorder among the controls (9).

While family studies are an important starting pointof genetic epidemiology, data from family studies canbe difficult to interpret for several reasons. Like allresearch in psychiatry it is dependent on the diagnosticclassification system, and some disorders may showmarkedly different patterns of familial transmissiondepending on minor changes in the diagnostic thresh-old. However, a problem even more specific to familystudies is that important environmental factors are also"familial." Key environmental variables, such as so-cial support, chronic and acute life stress, economicstatus, community environment, and many others, tendto vary along family lines and are known to haveindependent effects on mental health. For this reason,family studies may look beyond basic familial aggre-gation to examine specific patterns of transmissionthat, while still confounded with environmental fac-tors, more clearly suggest genetic influences. Thesespecific patterns of transmission within families mayvary according to whether the genes are dominant orrecessive, autosomal or X-linked, or multifactorial (in-cluding nongenetic factors (10)).

Although the family-study design has typically beenemployed to elucidate the degree and mode of trans-mission of most disorders, there are numerous otherpurposes for the application of such data. The majoradvantage of studying diseases within families is thatthe assumption of homotypy of the underlying factoreliminates the effects of heterogeneity which arepresent in comparisons between families. Family stud-ies can therefore be employed to examine the validityof diagnostic categories by assessing the specificity oftransmission of symptom patterns and disorders, com-pared with between-family designs (11). Data from

family studies may also provide evidence regardingetiologic or phenotypic heterogeneity. Phenotypic het-erogeneity is suggested by variable expressivity ofsymptoms, whereas etiologic heterogeneity is demon-strated by homotypic expression of different etiologicfactors between families. Moreover, the family-studymethod permits assessment of associations betweendisorders by evaluating specific patterns of co-aggregation of two or more disorders within families.Controlled family studies have been employed to datein investigating the comorbidity of panic disorder anddepression (12), alcoholism and depression (13), af-fective disorders and schizophrenia (14), and numer-ous other applications.

Twin studies

The concordance rate is the measure of associationthat has been used in twin studies to compare thepresence or absence of a trait or disorder withinmonozygotic twins (who share the same genotype)with that of dizygotic twins (who share an average of50 percent of their genes in common). The concor-dance rate is calculated by dividing the number oftwins who have the trait or disorder by the number oftwins in which at least one has the trait or disorderwithin each zygosity group (i.e., monozygotic or dizy-gotic). Concordance rates can be calculated either us-ing the pairwise method, whereby a twin-pair that isconcordant for a disorder is counted as one pair in thenumerator and in the denominator, or by using theprobandwise method, whereby concordant twins arecounted as two pairs both in the numerator and in thedenominator (but only when each affected twin wasidentified from the official register of cases indepen-dently (see Gottesman (15)). The ratio of concordancerates of monozygotic to dizygotic twins yields anestimate of the extent to which the trait is attributableto genetic factors.

To support a genetic etiology, the concordance ratesfor monozygotic twins should be significantly greaterthan those for dizygotic twins, and consistent with theconcept of familial aggregation. The degree of con-cordance between cotwins of either type can also beused to provide information about the magnitude ofgenetic or environmental effects. However, the prob-lem of same-environment confounds has also beenraised against twin-study paradigms. Althoughmonozygotic twins, whether they are raised togetheror apart, have been shown to have similar concordancerates for some traits (16), a more enduring criticism isthat the intrauterine environment is more similar formonozygotic twins than for dizygotic twins. The pos-sibility of environmental factors that may co-vary withzygosity is, therefore, an important consideration.



Although the traditional application of the twin de-sign focuses on the estimation of the heritability of atrait, there are several other research questions forwhich the twin study may be of value. Differences inconcordance rates between monozygotic and dizygotictwins may be investigated at the level of symptoms orsymptom clusters in order to study the validity ofsymptom complexes. Varying forms or degrees ofexpression of a particular disease in monozygotictwins may be an important source of evidence of thevalidity of the construct or disease entity. For example,McGuffin (17) and Kendler et al. (18) have, respec-tively, employed the twin-study design to investigatethe validity of the diagnostic categories of schizophre-nia and depression. In addition, Kendler et al. (19)showed that monozygotic twins were not only moreoften concordant for depression than dizygotic twins,but that they were concordant for specific depressionsubtypes, underscoring the heterogeneity of these dis-orders and need for nosology that reflect these entities.

Adoption studies

Family and twin studies are genetically informativebecause they hold the environment "constant" whileexamining the rates of disorder across different levelsof genetic relationship. An alternative approach is tovary the environment while comparing individualswith the same degree of genetic similarity. Adoptionstudies are part of this latter approach in that thepsychiatric similarity between an adoptee and his orher biologic versus adoptive relatives is directly com-pared. Another similar paradigm compares the bio-logic relatives of affected adoptees with the biologicrelatives of unaffected (or control) adoptees. Perhapsthe most powerful approach to studying the joint con-tribution of genetic and environmental factors is thecross-fostering adoption-study design which comparesrates of disorder in adoptees without biologic risk,who are raised by affected adoptive parents, withadoptees at biologic risk, who are raised by nonaf-fected adoptive parents. However, adoption studies arealso characterized by certain characteristics that maybias results. Biologic parents of adopted children areknown to have higher rates of psychopathology, alco-holism, or criminality than other parents, and adoptedchildren may themselves be at greater risk for psychi-atric disorders (e.g., Bohman (20) and Lipman et al.(21)). Although such criticisms may be valid reasonsto carefully interpret the rates of disorder found inthese studies, they do not negate the value of adoptionstudies to clarify genetic and environmental effects (inparticular for disorders showing specificity of trans-mission).

Estimates of heritability derived from adoption stud-ies may also be used to examine the validity of dif-ferent phenotypic definitions. For example, the origi-nal Danish-American studies by Kety et al. (22) onschizophrenia were particularly influential but werecriticized by some because of the breadth of the phe-notypic definition (which included vague categoriessuch as "latent" and "uncertain" schizophrenia). Asubsequent reassessment of the Danish material byKendler et al. (23) using DSM-IH-R criteria was,therefore, valuable. Although the more stringent crite-ria yielded far lower rates of schizophrenia or schizo-typal personality disorder, this definition provided bet-ter separation between the relatives of schizophrenicsversus controls. Thus, a narrower and more reliabledefinition of the disorder led to an increase in thegenetic effect, thereby validating its definition.

Genetic marker studies

The two basic study designs for identifying thegenes involved in disease etiology are associationstudies and linkage studies. Both study paradigmsexamine links between genetic markers and a specificdisease or trait. Genetic markers are a measurablehuman trait controlled by a single gene with a knownchromosomal location. It must also be polymorphicwith at least two alleles having a gene frequency of atleast 1 percent. The association study tests for a non-random relation between a genetic marker and a dis-ease using a case-control design and analytic method.

Association studies generally employ unrelated in-dividuals selected from the general population. Thechief impediment to association studies is selection ofa control group which is similar to the cases on allrelevant factors except disease status. Ethnicity hasbeen a particularly troublesome confounding factor inassociation studies. The transmission disequilibriumtest, a modification of the traditional association studyin which the nontransmitted alleles of parents of anaffected individual are used as controls, offsets someof the limitations in the selection of control samplesfor association studies (24).

Linkage studies examine the association betweengenetic markers and disease genes within families, andare based on the principle that two genes that lie closein proximity on a chromosome are transmitted to off-spring together. The specific alleles segregating in oneparticular family may differ from those in anotherfamily; for example, an association between bipolarillness and the ABO (blood type) locus may manifestin some bipolar families as association with the Aallele, while others manifest the B allele. Whereasassociation is a property of alleles, linkage is a prop-



erty of loci thereby involving all alleles at that locus(9, 25).

Linkage is quantified through use of the lod score;the lod score is the ratio of the likelihood of observingco-segregation of a disease and marker under linkageversus no linkage within families (26). The sib-pairmethod is an alternative strategy for testing linkage.Whereas the lod-score method assumes knowledge ofthe mode of transmission, the sib-pair method ismodel-free.

The choice of a method for identifying genes de-pends upon disease frequency, degree of genetic com-plexity, and strength of the contribution of genes to thedisease. The linkage method is still the most powerfulmethod for identifying genes for rare disorders withclear Mendelian patterns of inheritance. In contrast,association studies may require far smaller samplesizes compared with linkage studies when the popula-tion attributable risk is moderate (27).

Advances in neuroscience should enhance our un-derstanding of the pathophysiology of the major psy-chiatric disorders, and with continued family- andtwin-study research should lead to a reduction in het-erogeneity and other sources of genetic complexity ofthe psychiatric disorders. The application of linkagestudies and association studies to more homogeneoussubtypes of mental disorders with strong underlyinggenetic basis should prove to be more fruitful than thepresent state of knowledge would indicate. Recentsuccesses in identifying genes for complex disorders,such as apolipoprotein E (28) and several other loci forAlzheimer's disease, should serve to increase opti-mism regarding our ultimate ability to identify thegenetic factors which contribute to psychiatric disor-ders.

CHARACTERISTICS OF PSYCHIATRICDISORDERS THAT IMPEDE GENETIC STUDIES

Family, twin, adoption, and genetic marker studiescompose the basic research paradigms of genetic epi-demiology. However, the application of these para-digms is complicated by several factors germane topsychiatry itself. As mentioned in the beginning of thispresentation, one of the most far-reaching impedi-ments to genetic research is the reliability and validityof diagnostic categories. For example, twin studies ofmale and female alcoholics have revealed a signifi-cantly higher heritability for alcohol dependence thanfor alcohol abuse (29, 30). While narrow definitions ofalcoholism may provide a more valid phenotype forfuture genetic analyses, other disorders may requirebroader definitions. For example, the apparently lowfamiliality for some illnesses, such as obsessive-compulsive disorder, may be due to the use of narrow

diagnostic criteria that may not detect transmission ofobsessive-compulsive disorder "spectrum" withinfamilies (31). While the use of stricter criteria mayreduce the absolute degree of familial aggregation of adisorder, the relative difference in familial aggregationin relatives of cases compared with controls is likely toremain constant. As the appropriateness of using var-ious thresholds is rarely clear in advance of conductingepidemiologic investigations, research in this domainis necessarily dependent on diagnostic definitions thatare in the constant process of refinement.

In addition to the paramount issues of nosology, thetransmission of mental illness within families may notfollow patterns seen for other diseases. One importantreason is that assortative mating with respect to psy-chiatric disorders has been well-established (32); thatis, individuals with a specific form of disorder may bemore likely to have children with persons having thesame illness, or having the illness in their families. Inaddition to assortative mating, cross-mating amongindividuals with different types of psychiatric prob-lems is also frequently observed. For example, schizo-phrenic females are more likely than normal women tomarry men with substance abuse and behavior prob-lems, alcoholic men more often marry women withdepression or anxiety, and these women often have afamily history of alcoholism (32). Nonrandom matingleads to an increase in variability of a given trait in thepopulation (33). With respect to psychiatric disorders,nonrandom mating leads to a clustering of familieswith high density of disorder at one extreme andclustering of unaffected families at the other. Thebimodal distribution induced by nonrandom matingwould be expected to impede our ability to discrimi-nate the role of genetic factors in familial aggregation.

A third barrier to research in genetic epidemiologyis the strong co-occurrence among certain disorderswithin individuals. Comorbidity between psychiatricdisorders appears to be the rule rather than the excep-tion: numerous studies of clinical samples have dem-onstrated the large proportion of patients who simul-taneously meet diagnostic criteria for more than asingle disorder (e.g., Babor et al. (34), Chambless et al.(35), and Hasegawa et al. (36)) and multiple diagnoseswithin individuals appear to be quite frequent in epi-demiologic surveys of the general population (7, 37).Comorbidity, therefore, confounds the study of "pure"disorder etiology, but also poses important questionsas to the specificity of risk factors and the appropri-ateness of diagnostic boundaries. Cohort effects com-prise another limiting factor as it is often unclear ifobserved effects are artifacts or true differences. Acohort effect is defined as differences in disease prev-alence in a particular group of individuals, generally a



birth cohort, who progress simultaneously through therisk period for a particular disease (38). For any dis-ease which requires a particular environmental expo-sure for its development, the disease frequency maydiffer dramatically according to the variation in thedegree of exposure to the particular environmentalagent. The dramatic increase in availability of certaindrugs, for example, may complicate family studies ofalcoholism because of the tendency for individuals touse and/or abuse multiple substances over time. Dif-ferent generations may manifest the agent as a func-tion of drug availability at the time that substanceproblems are developing, and, therefore, it is not clearwhether individuals having drug abuse problemsshould be classified as affected or not affected in afamily study of alcoholism. Evidence from family andtwin studies may ultimately help distinguish whetherthere is a generalized liability to abuse substances ofall classes versus specificity in the use and abuse of aparticular class of drugs. However, cohort effects pres-ently pose difficult questions in the application orinterpretation of research in this domain. A relatedissue concerns recall biases associated with age, andfor this reason, epidemiologic studies increasingly in-clude late adolescent and young adult cohorts in orderto minimize reporting biases of childhood disorders.

A final major impediment concerns the genetic com-plexity of psychiatric disorders. Despite the dramaticsuccess of molecular genetics in the identification ofthe genetic basis of Huntington's disease (39),Duchenne's muscular dystrophy (40), cystic fibrosis(41), and breast cancer (42), the application of thesemethods to psychiatric disorders has been quite disap-pointing. Furthermore, although recent success inidentifying vulnerability genes for complex diseasessuch as diabetes (43, 44) has generated enthusiasm,replication attempts have thus far been unsuccessful.The reasons for these difficulties reside in the criticaldifferences that exist between the psychiatric disordersand the disorders to which the tools of moleculargenetics have been successfully applied. Linkage hasbeen reported for diseases which are rare (i.e., <0.01percent population prevalence), exhibit Mendelianpatterns of inheritance, and can be clearly diagnosedwith extremely high specificity and sensitivity (45). Incontrast, psychiatric disorders are complex disorders,which are conditions characterized by high populationprevalence, a lack of clear distinction between affectedand unaffected individuals (often with arbitrary thresh-olds for case definition), and failure to adhere toMendelian patterns of transmission. For these reasons,Risch and Botstein (46) concluded that the chief im-pediment to identifying genes for psychiatric disordersis their underlying genetic complexity. Psychiatric dis-

orders are likely to be attributable to a large number ofgenes, and the interactions among genes and withenvironmental factors contribute to phenotypic heter-ogeneity.

EVIDENCE FOR THE ROLE OF GENETICFACTORS IN THE ETIOLOGY OF PSYCHIATRICDISORDERS

Although there are formidable challenges to apply-ing genetic epidemiology to the field of psychiatry,progress continues to be made in each of the fourparadigms described previously. We will now presenta review of research in the genetic epidemiology of themajor classes of psychiatric diagnoses: schizophrenia,affective disorders, anxiety disorders, and substancedependence.

Schizophrenia

Schizophrenia is a form of psychotic disorder char-acterized by delusions, hallucinations, or disturbancesin affect and thought processes. More is known aboutthe genetic basis of schizophrenia than perhapsany other psychiatric condition, with genetically-informative studies stemming from early this century.Concerning familial aggregation of this disorder,Gottesman (15) pooled data from approximately 40family studies reported between 1920 and 1987 andconcluded that there is considerable support for theclaim that schizophrenia is familial. The risk ofschizophrenia and related disorders to first-degree rel-atives (siblings, parents, children) was on average 9.3times greater than the risk of schizophrenia to thegeneral population. However, because early familystudies used outdated diagnostic criteria and widelydiffering methodologies, the conclusions that can bedrawn from pooling earlier data are limited. In a morerecent review, Kendler and Diehl (47) examined onlyrecent family studies that have included a controlgroup, direct in-person interviews, and blind diagnosesof relatives (see table 2). Based on the average risks tofirst-degree relatives of proband and control groupsacross studies, a remarkably similar prevalence riskratio of 9.4 is observed. The conclusion that schizo-phrenia is highly familial is augmented by twin andadoption studies support for a genetic etiology. Asshown in table 2, McGuffin et al. (48) reviewed sixtwin studies demonstrating an average probandwiseconcordance rate of 46 percent for monozygotic twinsand 13 percent for dizygotic twins. Similarly, adoptionstudies demonstrated that the average risk to first-degree relatives for schizophrenia spectrum disorders(broadly defined) was 15.5 percent compared with 3.6percent for controls (giving a grand mean prevalence



TABLE 2.Typo

otIrvesti gallon

Family

Twin

Adoption

Evidence for genetic factors in schizophrenia*

Standardcomparison

Relatives of probands versusrelatives of controls

Monozygotic probandwise concordanceversus dizygotic probandwiseconcordance

Adoptee-biologic relatives versus adoptee-adoptive relatives

No.o(studies

reviewed

9

12

4

Averageprevalence

ratio

8.9

4.4

4.3

Ratiorange

2.7-18.5

2.2-12.0

1.9-10.6

Data compiled by Render and Diehl (47), and McGuffin et al. (48).

risk ratio of 4.3). Very recent adoption studies con-tinue to support the genetic transmission of schizo-phrenia and related disorders (e.g., Kendler et al. (23)),and have produced findings of similar magnitude.

With such clear evidence of a large genetic role inschizophrenia, it is not surprising that much researchhas focused on identifying specific genetic markers forthis disorder. Although many association studies ofschizophrenia have been performed using diversepolymorphisms, the majority of work in this area hasfocused on linkage studies. Fueled by initial findingsof associations of specific cytogenic abnormalitieswith schizophrenia, many recent DNA marker studieshave focused on regions of chromosomes 3, 5, 6, 8,11,22, and sex chromosomes (for review, see Kendler andDiehl (47), Gurling (49), Mowry et al. (50), andNurnberger and Byerley (51)). However, very fewresults have been replicated for either association stud-ies or linkage studies despite many exciting initialfindings, and the newer approach of scanning theentire human genome with DNA markers has, to date,produced only equivocal findings (see Tsuang andFaraone (52)).

In summary, while considerable evidence exists thatschizophrenia is a genetically transmitted disorder, weare presently far from identifying with confidencewhich genes are implicated. It is also important tounderscore that many of the studies in genetic epide-miology indicate that schizophrenia is strongly depen-dent on nongenetic factors. For example, as the aver-

age concordance rate for monozygotic twins isapproximately 50 percent, an individual may have astrong genetic vulnerability to the disease but notmanifest the illness. Tienari et al. (53) recently re-ported that the genetic propensity for schizophreniawas only manifested if the individual lived in a dis-turbed family environment. Similarly, other investiga-tions have revealed that negative life events increaseboth the chance of relapse and the severity of symp-toms (for a review, see Norman and Malla (54)), thatmarital status is a powerful predictor of schizophreniacourse (Jablensky et al. (55)), and that additional di-verse environmental factors may be implicated(Kendler et al. (56)). Taken together, these investiga-tions support a diathesis-stress model of the disorderwhereby a genetic vulnerability serves as a key etio-logic factor that is dependent, at least in part, onenvironmental factors for its ultimate expression.

Mood disorders

Mood disorders are comprised of a heterogeneousgroup of syndromes, of which major depression andbipolar disorder (manic depression) are major sub-types. Depressive episodes involve not only depressedmood or anhedonia, but also a variety of somatic orcognitive symptoms that signify a marked changefrom previous functioning. Manic episodes involve aperiod of abnormally elevated, expansive, or irritablemood that is severe enough to cause marked impair-

TABLE 3. Evidence for genetic factors in mood disorders*

Typeof

investigation

Standardcomparison

No. ofstudies

reviewed

Averageprevalence

ratio

Ratiorange

Family

Twin

Adoption


Monozygotic probandwise concordanceversus dzygotic probandwiseconcordance


4 (bipolar)6 (unipolar)5 (bipolar)2 (unipolar)

1 (bipolar)2 (unipolar)

10.85.05.31.5

2.63.3

3.7-17.71.5-18.93.6-13.41.2-1.9

1.7-4.8

* Data compiled by Maier (14), Merikangas and Kupfer (10), and Taylor et al. (58).



ment in essential life domains. Table 3 summarizesevidence for genetic factors in bipolar disorder andunipolar depression as reviewed by previous investi-gations (10, 57, 58). The overall conclusion from thisresearch is that both major depression and bipolardisorder have important genetic components. Con-trolled family studies show a fivefold risk to relativesof depressed patients, and greater than a tenfold risk torelatives of bipolar patients. The concordance rate forbipolar monozygotic twins is over five times that ofdizygotic twins, and depressed twin concordanceshows less dramatic but still notable differences. Theprevalence ratios based on the few existing adoptionstudies also confirm that the magnitude of twin con-cordance or familialism is not due purely to environ-mental factors and that genes play a prominent role.

Concerning marker studies, progress has been elu-sive in identifying the specific genetic mechanismsimplicated in mood disorders. Association studieshave focused on blood groups, human leukocyte anti-gens, rapid eye movement sleep, and diverse othermarkers. More recently, these investigations have alsomoved to using molecular biology to examine specificDNA markers including tyrosine hydroxylase, dopa-mine, and monoamine oxidase. Linkage studies havealso focused on human leukocyte antigen haplotypesand markers associated notably with chromosomes 4,11, 15, 18, 21, or the X chromosome. Although workis continuing to focus on both previous and new loci(e.g., Blackwood et al. (59) and Ginns et al. (60)),reviews of association studies and linkage studies con-clude a present lack of consistent or replicated findings(Merikangas and Kupfer (10) and Craddock andMcGuffin (61)).

Anxiety disorders

The anxiety disorders compose a diverse group ofsyndromes that share core features of anxiety, fear, orbehavioral avoidance, and are of sufficient severity toimpair the individual's daily functioning. At present,

relatively few studies have examined anxiety disordersfrom the perspective of genetic epidemiology, andthere are virtually no data from certain paradigms(such as adoption studies). However, existing researchindicates that most anxiety disorders aggregate in fam-ilies, and several investigations have offered specificsupport for genetic etiology (examples of three anxietydisorders are shown in table 4). Perhaps the mostconsistent support for the role of genetic factors hasbeen found for panic disorder. A review of six con-trolled family studies using direct interviews providesan average prevalence risk ratio of 9.4 (62), and newinvestigations continue to report high levels of aggre-gation (e.g., Battaglia et al. (63)). Although there hasbeen some inconsistency reported by twin studies ofpanic disorder (see McGuffin et al. (48)), two recentstudies applying modern diagnostic criteria demon-strated considerably higher rates for monozygotic,compared with dizygotic, twins (64, 65). Furthermore,current estimates derived from the Virginia Twin Reg-istry show panic disorder to have the highest herita-bility of all anxiety disorders at 44 percent (66).

Genetic factors are implicated in other anxiety dis-orders, although comparatively few investigationshave been completed. For example, social phobia ag-gregates in families, and twin studies show a higherconcordance for monozygotic twins (see table 4).Other phobias (i.e., specific phobia, agoraphobia)have also been shown to be familial, with the threephobia subtypes having similar prevalence risk ra-tios and specificity of transmission (for review, seeMerikangas and Angst (67) and Woodman and Crowe(68)). More recent data from the Virginia Twin Studyreport the estimated total heritability for phobias to be35 percent (66). The application of genetic epidemi-ology to understanding other anxiety disorders hasbeen limited not only due to a dearth of controlledstudies, but also because of uncertainty about the ap-propriateness of phenotypic descriptions. For example,the few family studies of obsessive-compulsive disor-

TABLE 4. Evidence for genetic factors In anxiety disorder*

Typeoi

Investigation

Standardcomparison

No. ofstudies

reviewed

Averageprevalence

ratio

Ratiorange

Family Relatives of probands versusrelatives of controls

Twin Monozygotic probandwise concordanceversus dizygotic probandwiseconcordance

6 (panic)2 (social phobia)2 (obsessive-

compulsive disorder)2 (panic)1 (social phobia)1 (obsessive-

compulsive disorder)

9.43.11.0

2.41.61.9

4.2-17.83.0-3.21.0-1.1

2.2-2.5

1 Data compfled by Chapman et al. (87), Weissman (62), and Woodman and Crowe (68).



der that have used standardized assessments with nar-row operationalized criteria have demonstrated no in-creased risk to relatives (see table 4). By contrast, anexamination of obsessional symptoms in cotwins ofobsessive-compulsive disorder probands revealed anincreased risk to monozygotic twins over dizygotictwins. Discrepancies such as these will be clarified asmore controlled family, twin, and adoption studies arecarried out, and as the validity of narrow versus broaddefinitions is established.

Concerning marker studies of anxiety disorders, thehigh heritability rates seen for panic disorder has madeit the natural focus of research in this area, and manyclinical and neurobiologic challenge studies haveserved as a guide by implicating the adrenergic system(for a review, see Goddard et al. (69)). However,recent linkage studies have excluded the possibilitythat panic disorder was due to mutations in adrenergicreceptor loci on chromosomes 4, 5, or 10 (70), andother work has similarly excluded linkage with gamma-aminobutyric acid beta 1 (GABAA) receptor genes (71).Recent reports from a genomic survey of panic disorderusing 600 markers have not yielded evidence of linkage(51). While the status of current biologic marker studiesof panic are still in their infancy, there is reason to beoptimistic as the Human Genome Project (and the iden-tification of numerous highly polymorphic markers) willsoon lead to major increases in the precision of thehuman genome map (72).

Substance use disorders

Substance use disorders involve the maladaptiveand typically regular use of psychoactive substancesthat affect the central nervous system. Substance abusecauses recurrent problems in social, occupational, psy-chologic, or physical functioning, or is characterizedby recurrent substance use in hazardous situations(such as drinking and driving). Substance dependencemay include any of these problems, as well as in-creased tolerance, inability to control use, or with-drawal symptoms. The majority of studies concerning

the genetic epidemiology of addictive behaviors hasfocused on alcoholism rather than drug-related prob-lems, and a shared etiology is suspected by someresearchers (e.g., Blum et al. (73)). Table 5 summa-rizes the family, twin, and adoption studies of alco-holism as recently reviewed by McGue (74) andMerikangas (75). Not only does alcohol abuse anddependency aggregate in families (comprising a sev-enfold risk to first-degree relatives), but twin andadoption studies indicate that this aggregation is partlydue to genetic factors. At present, the evidence for agenetic predisposition to alcoholism is stronger formen than for women (but generally significant forboth). The more recent investigations continue to offerconsistent support not only for familial aggregation(e.g., Araujo and Monteiro (76) and Maier et al. (57)),but also clarify previously weak areas of evidence bydemonstrating a greater concordance for monozygoticover dizygotic female twins (Kendler et al. (77)). Theheritability of alcoholism (narrowly defined) has beenestimated at 59 percent by some researchers (66),while the heritability of problem drinking (using broaddefinitions) has been estimated at 8-44 percent infemales and 10-50 percent in males (78). However,the genetic information derived from these twin stud-ies is complex, and recent twin evidence also suggeststhat the heritability of alcoholism (at least in males) isgreater when the individual has a comorbid psychiatricdiagnosis (79).

Similar to other domains, the search for specificmarkers through association and linkage studies inalcoholism has produced equivocal findings. The ma-jority of association studies to date have focused onthe dopamine (DRD2) and serotonin (5-HT) receptorgenes as well as the aldehyde dehydrogenase (ALDH)locus (for reviews, see Blum et al. (73), McGue (74),Gelernter et al. (80), Goldman (81), and Pato et al.(82)). Although several investigations have replicatedsignificant associations between alcoholism and thesemarkers, the majority of investigations are either pre-liminary, nonconfirmatory, or have revealed potential

TABLE 5. Evidence for genetic factore in alcohol abtne and/or dependence*

Typeo!

Investigation

Standardcomparison

No. ofstudies

reviewed

Averageprevalence

ratio

Ratiorange

Family

Twin

Adoption


Monozygotic probandwise concordanceversus dizygotic probandwiseconcordance


5

6 (male)5 (female)

5 (male)4 (female)

7.0

1.61.1

2.42.4

2.5-20.1

1.1-2.20.6-1.4

1.0-3.60.5-6.3

1 Data compiled by McGue (74) and Merikangas (75).



sampling biases that may independently explain ob-served associations.

While genetic marker studies for alcoholism (andpsychiatric disorders in general) are still in their in-fancy, the results from other paradigms support theconclusion that genetic factors play a moderate role formale, and at least a modest role for female, drinkingproblems (74). In addition, the role of environmentalfactors in the etiology of alcoholism has been sup-ported by numerous studies from a genetic epidemio-logic perspective. For example, twin studies have in-dicated that concordance rates for alcohol-relateddisorders are greater for certain geographic areas, pos-sibly due to socioeconomic factors (83). Adoption-study paradigms have shown not only that a disturbedadoptive family environment interacts with a geneticpredisposition for alcoholism to affect the risk for thedisorder (84), but that the adoptive family environ-ment can predict alcohol abuse or dependency inde-pendent from genetic vulnerability (85). Finally, al-though the majority of work in this area has focused onalcoholism, research on other substance use disordersis growing at a fast pace. Current work indicates thatthese disorders are familial (e.g., Skre et al. (86)) andthat they have complex etiologies involving both ge-netic and environmental components (see Cadoret(85)).

SUMMARY AND FUTURE STEPS

Advances in standardized definitions of major psy-chiatric disorders has dramatically enhanced our abil-ity to reliably characterize behavioral phenotypes forgenetic studies. The application of family, twin, adop-tion, and genetic marker paradigms to this domain hasoffered new opportunities for understanding the re-spective roles of genetic and environmental factors forclasses of disorders that affect large percentages of thegeneral population. However, previous research hasbeen impeded by several factors, notably a lack ofbiologic validity of diagnostic categories as well as thegenetic complexity of psychiatric disorders. In fact,the discovery of the breast and ovarian cancer suscep-tibility gene (BRCA1) witnesses the importance ofmaximizing disease homogeneity among study sub-jects. The application of genetic study paradigms toguide definitions of the thresholds and boundaries ofpsychiatric syndromes (and to refine the precision ofphenotypic definitions) will hopefully enhance theidentification of homogeneous subtypes, and therebyincrease the power of linkage studies in elucidatingtheir genetic basis.

ACKNOWLEDGMENTS

This work was supported in part by National Institutesof Mental Health Research Fellowship 5T32MH14235(Dr. Swendsen), Research Scientist Development AwardK02 MH00499, K02DA00293, and 5R01AA09978 (Dr.Merikangas), and National Institutes of Drug Abuse grantsDA05348 and AA07080.

REFERENCES

1. Sydenham T. The works of Thomas Sydenham: translatedfrom the Latin edition of WA Greenhill, with a life of theauthor, by RG Latham. London, England: Sydenham Society,1848:13.

2. Feighner JP, Robins E, Guze SB, et al. Diagnostic criteria foruse in psychiatric research. Arch Gen Psychiatry 1972; 16:57-63.

3. World Health Organization. The ICD-10 classification ofmental and behavioral disorders. Geneva, Switzerland: WorldHealth Organization, 1993.

4. American Psychiatric Association. Diagnostic and statisticalmanual of mental disorders: DSM-IV. 4th ed. Washington,DC: American Psychiatric Association, 1994.

5. Angst J, Merikangas KR, Preisig M. Subthreshold syndromesof depression and anxiety in the community. J Clin Psychiatry1997;57(Suppl 8):6-10.

6. Kendell RE. Clinical validity. Psychol Med 1989; 19:45-55.7. Kessler RC, McGonagle KA, Zhao S, et al. Lifetime and

12-month prevalence of DSM-HI-R psychiatric disorders inthe United States: results from the National Comorbidity Sur-vey. Arch Gen Psychiatry 1995;51:8-19.

8. American Psychiatric Association. Diagnostic and statisticalmanual of mental disorders: DSM-IH-R. 3rd ed, rev. Wash-ington, DC: American Psychiatric Association, 1987.

9. Khoury MJ, Beaty TH, Cohen BH. Fundamentals of geneticepidemiology. New York, NY: Oxford University Press,1993:75.

10. Merikangas KR, Kupfer DJ. Mood disorders: genetic aspects.In: Kaplan HI, Sadock BJ, eds. Comprehensive textbook ofpsychiatry. 6th ed. Baltimore, MD: Williams and Wilkins,1995:1102-16.

11. Tsuang MT, Faraone SV, Lyons MJ. Identification of thephenotype in psychiatric genetics. Eur. Arch Psychiatry ClinNeurosci 1993;243:131-42.

12. Leckman JF, Merikangas KR, Pauls DL, et al. Anxiety disor-ders and depression: contradictions between family study dataand DSM-m conventions. Am J Psychiatry 1983;140:880-2.

13. Merikangas KR, Leckman JF, Prusoff BA, et al. Familialtransmission of depression and alcoholism. Arch Gen Psychi-atry 1985;42:367-72.

14. Maier W, Lichtermann D, Minges J, et al. Continuity anddiscontinuity of affective disorders and schizophrenia: resultsof a controlled family study. Arch Gen Psychiatry 1993;50:871-83.

15. Gottesman n. Schizophrenia genesis: the origins of madness.New York, NY: Freeman, 1991.

16. Gottesman n, Sheilds J. Schizophrenia, the epigenetic puzzle.New York, NY: Cambridge University Press, 1982.

17. McGuffin P, Farmer AE, Gottesman II. Twin concordance foroperationally defined schizophrenia: confirmation of familial-ity and heritability. Arch Gen Psychiatry 1984;41:541-5.

18. Kendler KS, Neale MC, Kessler RC, et al. A population-basedtwin study of major depression in women: the impact ofvarying definitions of illness. Arch Gen Psychiatry 1992;49:257-66.

19. Kendler KS, Eaves LJ, Walters EE, et al. The identificationand validation of distinct depressive syndromes in a



population-based sample of female twins. Arch Gen Psychi-atry 1996;53:391-9.

20. Bohman M. Some genetic aspects of alcoholism andcriminality: a population of adoptees. Arch Gen Psychiatry1978;35:269-76.

21. Lipman EL, Offord DR, Boyle MH, et al. Follow-up ofpsychiatric and educational morbidity among adopted chil-dren. J Am Acad Child Adolesc Psychiatry 1993;32:1007-12.

22. Kety SS, Rosenthal D, Wender PH. The types and prevalenceof mental illness in the biological and adoptive families ofadopted schizophrenics. In: Rosenthal D, Kety SS, eds. Thetransmission of schizophrenia: proceedings of the SecondResearch Conference of the Foundation's Fund for Researchin Psychiatry, Dorado, Puerto Rico, 26 June to 1 July 1967.New York, NY: Pergamon Press, 1968:345-62.

23. Kendler KS, Gruenberg AM, Kinney DK Independent diag-noses of adoptees and relatives as defined by DSM-UJ in theprovincial and national samples of the Danish Adoption Studyof Schizophrenia. Arch Gen Psychiatry 1994;51:456-68.

24. Spielman RS, McGinnis RE, Ewens WJ. Transmission test forlinkage disequilibrium: the insulin gene region and insulin-dependent diabetes mellitus (IDDM). Am J Hum Genet 1993;52:506-16.

25. Vo'gel F, Motulsky AG. Human genetics: problems and ap-proaches. New York, NY: Springer-Verlag, 1986.

26. Risch N. Mapping genes for psychiatric disorders. In: GershonES, Cloninger CR, eds. Genetic approaches to mental disor-ders. Washington, DC: American Psychiatric Press, 1994:47-61.

27. Risch N, Merikangas K. The future of genetic studies ofcomplex human diseases. Science 1996;273:1516-17.

28. Saunders AM, Strittmatter WJ, Schmechel D, et al. Associa-tion of apolipoprotein E allele e4 with late-onset familial andsporadic Alzheimer's disease. Neurology 1993;43:1467-72.

29. Caldwell CB, Gottesman U. Sex differences in the risk ofrelapse for alcoholism: a twin study. Presented at the 21stAnnual Meeting of the Behavior Genetics Association, StLouis, Missouri, June 1991.

30. Pickens RW, Svikis DS, McGue M, et al. Heterogeneity in theinheritance of alcoholism: a study of male and female twins.Arch Gen Psychiatry 1991;48:19-28.

31. Black DW, Noyes R, Goldstein RB, et al. A family study ofobsessive-compulsive disorder. Arch Gen Psychiatry 1992;49:362-8.

32. Merikangas KR, Brunetto W: Assortative mating and psychi-atric disorders. In: Papadimitriou GN, Mendlewicz J, eds.Genetics of mental disorders. Part II: Clinical issues. London,England: Baillidre Tindall, 1996:175-85.

33. Crow JF, Felsenstein J. The effect of assortative mating on thegenetic composition of a population. Eugen Q 1968; 15:85-97.

34. Babor TF, Wolfson A, Boivin D, et al. Alcoholism, culture,and psychopathology: a comparative study of French, FrenchCanadian, and American alcoholics. In: Helzer JE, Canino GJ,eds. Alcoholism in North America, Europe, and Asia. NewYork, NY: Oxford University Press, 1992:182-95.

35. Chambless DL, Chemey J, Caputo GC, et al. Anxiety disor-ders and alcoholism: a study with inpatient alcoholics. J Anx-iety Disord 1987;l:29-40.

36. Hasegawa K, Mukasa H, Nakazawa Y, et al. Primary andsecondary depression in alcoholism—clinical features andfamily history. Drug Alcohol Depend 1991;27:275-81.

37. Regier DA, Farmer ME, Rae DS, et al. Comorbidity of mentaldisorders with alcohol and other drug abuse: results from theEpidemiologic Catchment Area (ECA) Study. JAMA 1990;264:2511-18.

38. Mausner JS, Bahn AK. Epidemiology: an introductory text.Philadelphia, PA: WB Saunders,1974:81-2.

39. Gusella JF, Wexler NS, Conneally PM, et al. A polymorphicDNA marker genetically linked to Huntington's disease. Na-ture 1983;306:234-8.

40. Kunkel LM. Analysis of deletions in DNA from patients with

Becker and Duchenne muscular dystrophy. Nature 1986;322:73-7.

41. Beaudet A, Bowcock A, Buchwald M, et al. Linkage of cysticfibrosis to two tightly linked DNA markers: joint report froma collaborative study. Am J Hum Genet 1986;39:681-93.

42. Hall JM, Lee MK, Morrow J, et al. Linkage of early onsetfamilial breast cancer to chromosome 17q21. Science 1990;259:1684-9.

43. Davies JL, Kawaguchi Y, Bennett ST, et al. A genome-widesearch for human type 1 diabetes susceptibility genes. Nature1994;371:130-6.

44. Hashimoto L, Habita C, Beressi JP, et al. Genetic mapping ofa susceptibility locus for insulin-dependent diabetes mellituson chromosome llq. Nature 1994;371:161—4.

45. Risch N. Genetic linkage and complex diseases, with specialreference to psychiatric disorders. Genet Epidemiol 1990;7:3-16.

46. Risch N, Botstein D. A manic depressive history. (News). NatGenet 1996;12:351-3.

47. Kendler KS, Diehl SR. The genetics of schizophrenia: a cur-rent, genetic-epidemiologic perspective. Schizophr Bull 1993;19:261-85.

48. McGuffin P, Owen MJ, O'Donovan MC, et al. Seminars inpsychiatric genetics. Washington, DC: American PsychiatricPress, 1994.

49. Gurling H. The genetics of schizophrenia. Baillieres ClinPsychiatry 1996;2:15-47.

50. Mowry BJ, Levinson DF. Genetic linkage and schizophrenia:methods, recent findings and future directions. Aust N ZJ Psychiatry 1993;27:200-18.

51. Nurnberger JI, Byerley WF. 1995 Park City Molecular Psy-chiatry Conference, 5-8 February 1995, Park City, Utah.Psychiatr Genet 1995;5:5-7.

52. Tsuang MT, Faraone SV. Genetic heterogeneity of schizo-phrenia. Seishin Shinkeigaku Zasshi 1995;97:485-501.

53. Tienari Pi Wynne LC, Moring J, et al. The Finnish adoptivefamily study of schizophrenia: implications for family re-search. Br J Psychiatry 1994;23:20-6.

54. Norman RM, Malla AK. Stressful life events and schizophre-nia. I. A review of the research. Br J Psychiatry 1993; 162:161-6.

55. Jablensky A, Sartorius N, Ernberg G, et al. Schizophrenia:manifestations, incidence and course in different countries. AWorld Health Organization ten-country study. Psychol MedMonogrSuppl 1992;20:l-97.

56. Kendler KS, Karkowski-Shuman L, Walsh D. Age at onset inschizophrenia and risk of illness in relatives: results from theRoscommon Family Study. Br J Psychiatry 1996;169:213-18.

57. Maier W, Minges J, Lichtermann D. Alcoholism and panicdisorder co-occurrence and co-transmission in families. EurArch Psychiatry Clin Neurosci 1993;243:205-ll.

58. Taylor MA, Berenbaum SA, Jampala VC, et al. Are schizo-phrenia and affective disorder related? Preliminary data froma family study. Am J Psychiatry 1993;150:278-85.

59. Blackwood DH, He L, Morris SW, et al. A locus for bipolaraffective disorder on chromosome 4p. Nat Genet 1996;12:427-30.

60. Ginns El, Ott J, Egeland JA, et al. A genome-wide search forchromosomal loci linked to bipolar affective disorder in theOld Order Amish. Nat Genet 1996;12:431-5.

61. Craddock N, McGuffin P. Approaches to the genetics ofaffective disorders. Ann Med 1993;25:317-22.

62. Weissman MM. Family genetic studies of panic disorder.J Psychiatr Res 1993;27:69-78.

63. Battaglia M, Bertella S, Politi E, et al. Age at onset of panicdisorder influence of familial liability to the disease and ofchildhood separation anxiety disorder. Am J Psychiatry 1995;152:1362-4.

64. Kendler KS, Neale MC, Kessler RC, et al. Panic disorder inwomen: a population-based twin study. Psychol Med 1993;23:397-406.



65. Skre I, Onstad S, Torgersen S, et al. A twin study of DSM-m-R anxiety disorders. Acta Psychiatr Scand 1993;88:85-92.

66. Kendler KS, Walters EE, Neale MC, et al. The structure of thegenetic and environmental risk factors for six major psychi-atric disorders in women: phobia, generalized anxiety disor-der, panic disorder, bulimia, major depression, and alcohol-ism. Arch Gen Psychiatry 1995;52:374-83.

67. Merikangas KR, Angst J. Comorbidity and social phobia;evidence from clinical, epidemiologic, and genetic studies.Eur Arch Psychiatry Clin Neurosci 1995;244:297-303.

68. Woodman CL, Crowe RR. The genetics of anxiety disorder.Baillieres Clin Psychiatry 1996;2:47-57.

69. Goddard AW, Woods SC, Charney DS. A critical review ofthe role .of norepinephrine in panic disorder focus on itsinteraction with serotonin. In: Westenberg HGM, Den BoerJA, Murphy DL, eds. Advances in the neurobiology of anxietydisorders. New York, NY: John Wiley and Sons, 1996:107-37.

70. Wang ZW, Crowe RR, Noyes R Jr. Adrenergic receptor genesas candidate genes for panic disorder: a linkage study. Am JPsychiatry 1992;149:470-4.

71. Schmidt SM, Zoega T, Crowe RR. Excluding linkage betweenpanic disorder and the gamma-aminobutyric acid beta 1 re-ceptor locus in five Icelandic pedigrees. Acta Psychiatr Scand1993;88:225-8.

72. Crowe RR. The Iowa Linkage Study of panic disorder. In:Gershon ES, Cloninger CR, eds. Genetic approaches to mentaldisorders. Washington, DC: American Psychiatric Press,1994:291-309.

73. Blum K, Sheridan PJ, Wood RC, et al. Dopamine D2 receptorgene variants: association and linkage studies in impulsive-addictive-compulsive behavior. Pharmacogenetics 1995;5:121-41.

74. McGue M. Genes, environment, and the etiology of alcohol-ism. In: Zucker R, Boyd G, Howard J, eds. The Developmentof alcohol problems: exploring the biopsychosocial matrix.Research monograph no. 26. Rockville, MD: US Departmentof Health and Human Services, Public Health Service, Na-tional Institutes of Health, National Institute on AlcoholAbuse and Alcoholism, 1994:1-39.

75. Merikangas KR. The genetic epidemiology of alcoholism.Psychol Med 1990;20:11-22.

76. Araujo NP, Monteiro MG. Family history of alcoholism andpsychiatric co-morbidity in Brazilian male alcoholics and con-trols. Addiction 1995;90:1205-11.

77. Kendler KS, Neale MC, Heath AC, et al. A twin-family studyof alcoholism in women. Am J Psychiatry 1994;151:707-15.

78. Heath AC, Martin NG. Genetic influences on alcohol con-sumption patterns and problem drinking: results from theAustralian NH&MRC twin panel follow-up survey. Ann N YAcad Sci 1994;708:72-85.

79. Pickens RW, Svikis DS, McGue M, et al. Common geneticmechanisms in alcohol, drug, and mental disorder comorbid-ity. Drug Alcohol Depend 1995;39:129-38.

80. Gelernter J, Goldman D, Risch N. The Al allele at the D2dopamine receptor gene and alcoholism: a reappraisal. JAMA1993;269:1673-7.

81. Goldman D. Candidate genes in alcoholism. Clin Neurosci1995;3:174-81.

82. Pato CN, Macciardi F, Pato MT, et al. Review of the putativeassociation of dopamine D2 receptor and alcoholism: a meta-analysis. Am J Med Genet 1993;48:78-82.

83. Romanov K, Kaprio J, Rose RJ, et al. Genetics of alcoholism:effects of migration on concordance rates among male twins.Alcohol Alcohol Suppl 1991;l:137-40.

84. Cutrona CE, Cadoret RJ, Suhr JA, et al. Interpersonal vari-ables in the prediction of alcoholism among adoptees: evi-dence for gene-environment interactions. Compr Psychiatry1994;35:171-9.

85. Cadoret RJ, Yates WR, Troughton E, et al. Adoption studydemonstrating two genetic pathways to drug abuse. Arch GenPsychiatry 1995;52:42-52.

86. Skre I, Onstad S, Edvardsen J, et al. A family study of anxietydisorders: familial transmission and relationship to mood dis-order and psychoactive substance use disorder. Acta PsychiatrScand 1994;90:366-74.

87. Chapman TF, Mannuzza S, Fyer AJ. Epidemiology and familystudies of social phobia. In: Heimberg RG, Liebowitz MR,Hope DA, et al., eds. Social phobia: degrees, assessment, andtreatment. New York, NY: Guilford Press, 1995:21-40.


genetic epidemiology of psychiatric disorders · tical manual of mental disorders, 3rd edition, ......

Documents