0 145-6008/98/2202-0505$03.0~/0 zyxwvutsrqpo ALWHOLISM: zyxwvutsrqpo CLINICAI zyxwvutsrqponmlk AND EXP~..RIMENTA~ RESEARCH Vol. 22, No. 2 April zy 1998 A Family-Based Analysis of the Association of the Dopamine D2 Receptor zyxw (DRD2) with Alcoholism Howard J. Edenberg, Tatiana Foroud, Daniel L. Koller, Alison Goate, John Rice, Paul Van Eerdewegh, Theodore Reich, C. Robert Cloninger, John I. Nurnberger, Jr., Maria Kowalczuk, 60 Wu, T.-K. Li, P. M. Conneally, Jay A. Tischfield, William Wu, Shantia Shears, Raymond Crowe, Victor Hesselbrock, Marc Schuckit, Bernice Porjesz, and Henri Begleiter The possible association of the zyxwvutsrq OR02 locus, and in particular the Taql-A1 allele, with alcoholism remains controversial, in part be- cause of differences in allele frequencies among populations. To avoid problems associated with differences in allele frequencies in different populations, we tested whether the OR02 locus is associ- ated with alcohol dependence in a large family-based sample. Nei- ther the transmission/disequilibrium test nor the Affected Farnily- Based Controls test provide any evidence of linkage or association between the OR02 locus and alcohol dependence. Key Words: Alcoholism, Dopamine D2 Receptor, Association Study, Family Study. EVERAL LINES of evidence-including adoption, S half-sibling, twin, and family studies-point to a ge- netic component to the risk for alc~holism.’-~ Alcoholism is a complex disease, with no simple pattern of inheritance and with substantial environmental and social influences. The genes that affect the risk for alcoholism have been sought by several methods, including studies of candidate genes and searches of the whole genome. The zyxwvutsrqpo DRD2 gene on chromosome llq22-q23 encodes the dopamine D2 receptor.s36 The dopamine system has been considered a candidate for involvement in alcoholism, with postulated links to novelty seeking and central nervous system An association between the TuqI-A1 polymorphism in the DRD2 gene and alcoholism was first reported by Blum et aL9 Shortly thereafter, the first non- replication was reported.” This polymorphism has subsequently been examined by many groups, with differing results. There have been pos- itive reports of the association of the TuqI-A1 allele with From the Indiana University School of Medicine (H.J.E., T.F., D.L.K., J.I.N.Jr., M.K., B.W., T.-K.L., P.M.C., J.A.T.), Indianapolis, Indiana; Wash- ington University School of Medicine (A. G., J. R., P. I.: E., T. R., C. R. C., W W, S.S.), St. Louis, Missouri; University of Iowa School zyxwvuts of Medicine (R.C.), Iowa City, Iowa; University of Connecticut School of Medicine (KH.), Farming- ton, Connecticut; University of California at San Diego School of Medicine (M.S.),La Jolla, Calijomia; and SUNY Health Science Center at Brooklyn (B.P., H.B.), Brooklyn, New York. Received for publication August 5, 1997; accepted September 30, 1997 This national collaborative study zyxwvutsrqp is supported by the National Institute on Alcohol Abuse and Alcoholism (NLAAA) through U.S. Public Health Ser- vices Grants UlOAA08401, UlOAA08402, and UIOAA08403. Reprint requests: Howard J. Edenberg, Ph.D., Department of BiochemistT and Molecular Biology, Indiana University School of Medicine, 635 BamhiN Drive, Medical Science 418, Indianapolis, IN 46202-5122. Copyright zyxwvutsrqp 0 1998 by The Research Society on Alcoholism. Alcohol Clin Exp Rrs, Vol22, No 2, 1998: pp 505-512 and polysubstance use and Many of these studies analyze the so-called “prevalence” of the TuqI-A1 allele, defined as either the TuqI-AlIAl or TuqI-AlIA2 genotype, rather than the TuqI-A1 allele fre- quency. This approach, implying a dominant (or nearly dominant) mode of action, might not be appropriate for alcoholism, nor for association studies in general, because the likelihood of carrying a putative disease-susceptibility allele given the associated marker is twice as high in the homozygote as in the heter~zygote.“-~~ Some studies found an association with only a subset of alcoholics with “severe” disease, but different definitions of severity have been 1,12,15,2223 (and see Ref. 20). Neiswanger et al.“ found a significant population association between the TuqI-A1 allele and alcoholism in the absence of any evi- dence for family-based association. Alcoholics did not dif- fer from unscreened (population-based) control samples, but both of those groups differed from control groups screened to eliminate individuals with alcoholism and/or other psychopathologies.16 Many studies have reported no association of DRD2 polymorphisms with alcoholism 10,183 19,23-32* , among these are studies that compared alcoholics with screened, nonal- coholic c ~ n t r o l s . ’ ~ ~ ~ ” ~ ~ ~ ~ ~ Various reviews and meta-analy- ses have not reached consensus on the question of whether the TaqI-A1 allele is associated with an increased risk for alcoholism or for “severe” a l c ~ h o l i s m . ~ ~ - ~ ~ , ” - ~ ~ A poten- tial confounding factor in association studies is the differ- ence in TuqI-A1 allele frequencies among populations, This emphasizes the need for either careful ethnic match- ing of alcoholics and controls or the use of family-based analyses. Studies of possible linkage between polymorphisms in or near the DRD2 locus and alcoholism have produced nega- tive results. Bolos et a1.l’ reported no evidence of linkage in two families. Parsian et a1.12 found no evidence for either linkage or cosegregation of the TuqI-A1 allele with alco- holism or severe alcoholism in 17 nuclear families drawn from 12 multigenerational pedigrees. Neiswanger et al. l6 found no evidence of linkage in 20 families of male alco- holic probands under any model tested, using both para- metric and nonparametric methods. Affected family-based association (AFBAC) tests were also negative.16 Cook et which ranges from 0.09 to 0.80.18,’9,21,23,2S,2h,30,31,37,43,44 505