three or more alcoholic family members the investigators conducted ge - PDF document

1 Vol. 26, No. 3, 2002215 three or more al
Vol. 26, No. 3, 2002215 three or more alcoholic family members, the investi-gators conducted genetic analyses using microsatellitemarkersÑDNA regions located across all chromo-somes, in which short repeated sequences exist inmanyvgenotyping. More than 1.2 million genotypes havefrom families of alco-holics and1,238 people from control families. Bymonitoring the inheritance patterns of such markerinvestigators could identify chromosomal regions thatinfluence (i.e., show genetic linkage with) certainalcohol-related traits. other aspects of the COGA study are described in moredetail in the article by Bierut and colleagues, pp. sults of Genetic AnalysesDNA Regions with Susceptibility Genes. dependence as the phenotype have revealed regions onseveral chromosomes that appear to contain genesaffecting the risk for alcoholism. The primary analysesalcoholism. The primary COGA definition of beingaffected with alcoholism requires a person to meetthe Feighner criteria (Feighner et al. 1972) for definitealcoholism. If siblings who are alcoholic share morealleles at a marker than would be expected based onchance, this suggests that genes within the chromoso-mal region containing the marker contribute to theAnalyses of 987 people from 105 families in theinitial sample provided evidence that regions on 3chromosomes contained genes that increase the riskfor alcoholism (Reich et al. 1998). The strongestevidence was for regions on chromosomes 1 and 7,with more modest evidence for a region on chromo-some 2. The DNA regions identified through theseanalyses were broad, as is typical for studies of com-plex genetic diseases, and therefore are likely to con-tain numerous genes. Much additional work isequired to narrow the regions and attempt to deter-mine which specific gene or genes play a role inaffecting the risk for alcoholism. Therefore, addi-tional markers within these regions of interest wereanalyzed in the same people. Subsequent analyses thatincluded the additional markers supported the initialfindings (Foroud et al. 2000) but did not narrow thechromosomal regions in which genes influencingalcoholism susceptibility are likely to lie. The data from the second part of the split sam-pleÑthe replication sample, which comprised 1,295people from 157 familiesÑgenerally supported theinitial findings (Foroud et al. 2000). Thus, the repli-cation sample again provided evidence that genesincreasing the risk of alcoholism were located in thesame regions of chromosomes 1 and 7, albeit withless statistical support. When the initial and replica-tion samples were combined, these chromosomalegions remained the strongest candidates for con-vidence for the region on chromosome 2 increasedwith the additional markers in the initial sample, butthe replication sample provided no additional evi-dence for alcoholism susceptibility genes in this chro-mosomal region. Conversely, the strongest evidencein the replication sample for a region containinggenes affecting the risk for alcoholism was on chro-mosome 3, which had shown no evidence of beinglinked with alcoholism in the initial sample. Becauseholism, however, it is not surprising that an indepen-dent sample, even one collected by the same group ofesearchers, might replicate some previously identi-fied genes and also identify some novel alcoholismoud and colleagues (2000) also analyzed thecombined data set from the initial and replicationsamples using a more restricted definition of alco-holism as specified in the ICDÐ10. This restrictiongreatly reduced the number of sibling pairs in thecomparison. The region on chromosome 1 providedthe strongest evidence for a susceptibility gene in thecombined sample. In addition, this new evaluationdetected a region on chromosome 8 that was linked DNA Regions with Protective Genes. nterestingly,sample produced evidence for a protective region onchromosome 4, in the general vicinity of the alcoholdehydrogenase ( ADH A related analysis, usinga technique that treats alcoholism as the extreme of adistribution of an underlying quantitative trait,showedevidence for linkage to this same region he Collaborative Study on the Genetics of Alcoholism 2Several ADH genes exist, each of which has several alleles.across a population, such as height. 216 illiams et al. 1999). This finding suggests that vari-ants of a gene or genes within this region reduced the ADH alleles are known toaffect the risk for alcoholism; however, the knownprotective alleles occur at high frequency in Asianpopulations but are rare in the Caucasian populationthat makes up most of the COGA sample (Edenberg2000). Therefore, these analyses may have identified anew protective ADH allele or another protective genelocated nearby. The number of unaffected siblingpairs genotyped in the replication sample was toothat may reflect aprotective influence against alcoholism is the maximumThis phenotype is quantitativeand heritable, and a low number of drinks consumedin a 24-hour period may reflect a reduced tolerancefor high levels of alcohol. An advantage of a quantita-tive phenotype is that everyone in a study can con-phenotype in both the initial and replication data sets(and in the combined sample) showed the strongest evi-dence for linkage in the same region of chromosome 4where the ADH genes reside (Saccone et al. 2000). Thisphenotype may be related to thepro-tective region identified in the unaffected sibling pairsand to protective effects of certain ADH (Edenberg 2000). DNA Regions Related to Symptoms of Alcoholism. are highly diverse, ranging from more biologicalsymptoms (e.g., tolerance and withdrawal) to moresocial symptoms (e.g., social and legal problems).unique mix of those symptoms; therefore, a diagnosisof alcoholism does not reflect a uniform phenotype.Consequently, researchers have constructed other,more defined phenotypes from the data obtained inthe COGA interviews. These include an analysis ofsymptoms related to alcoholism that produced pheno-types which appeared to reflect the severity of alcoholproblems. Analysis of these phenotypes providedevidence for a DNA region on chromosome 16 thatwas associated with an increased risk for more severealcohol problems (Foroud et al. 1998). DNA Regions Associated with Co-Occurring Disorders. depressive disorder or depressive syndrome (Nurn-berger et al. 2001). Depression alone showed modestevidence of linkage to a region on chromosome 7.The phenotype characterized by co-occurring alco-holism and depression showed evidence of linkage toa region on chromosome 2, primarily in the replica-tion sample. The most interesting finding was for thebroad Òalcoholism or depressionÓ phenotype, with strong evidence for linkage to the same region ofchromosome 1 that was linked to alcoholism alone(Nurnberger et al. 2001). This suggests that a gene orgenes within this chromosomal region increase therisk for both alcoholism and depression. (For moreinformation on these analyses, see the article byurnberger and colleagues, pp. 233Ð240, in this DNA Regions Linked with Electrophysiologicaleasures. The COGA investigators also evaluatedelectrophysiological variables, such as EEGs andERPs, from study participants. EEGs measure overallbrain activity, whereas ERPs are brain waves elicitedin response to specific stimuli (e.g., a light or sound).Analysis of such electrophysiological data may reveala subset of genes that affect these quantitative, biolog-ical phenotypes related to alcoholism (Porjesz et al.1998, 2002). One component of an ERP is a brainwave called P300, which typically occurs 300 mil-liseconds after a stimulus. Previous studies had foundthat a reduced amplitude of the P300 wave is a heri-table phenotype that correlates with alcohol depen-dence and other psychiatric disorders (Porjesz et al.1998). The genetic analyses of the COGA partici-pants identified four regions, on chromosomes 2, 5,amplitude of the P300 (Begleiter et al. 1998). addition to these findings, recent analyses demon-strate strong evidence for a locus that affects brain waveoscillations as measured by electroencephalography(Porjesz et al. 2002). Thus, a gene or genes that affectbrain rhythms lies in a region of chromosome 4 thatcontains a cluster of genes encoding proteins (i.e., andidate Genes. COGA researchers have also ana-lyzed candidate genesÑgenes suspected to play a rolein the development of alcoholism based on other stud-ies. Some of these candidate genes encode componentsof various brain chemical systems that allow communi-cation among nerve cells. Two of these genes are the DRD ) and a serotonintransporter gene ( HTT) . However, the analyses found Vol. 26, No. 3, 2002217 no evidence that DRD (Edenberg et al. 1998 a ) or that HTT either alcoholism in general or to a more severe form ofalcoholism (Edenberg et al. 1998 b ). erspective Where does the COGA study go from here? Theincreasing availability of the DNA sequence of theentire human genome and knowledge of variations inthat sequence among people are greatly aiding the cur-ent phase of the research. Particularly important to thecurrent work is the use of the sequence data to identifywhich genes are located within the regions that haveshown linkage with alcoholism and the other pheno-Where the available data are incomplete or insufficient,COGA researchers are seeking these polymorphismsthemselves. Of particular value are single-nucleotidepolymorphisms (SNPs)Ñsites at which people differ ina single base pairÑin or near genes within the regionsof interest. COGA investigators are doing additionalgenotyping of SNPs in and near candidate genes in theegions of linkage for further analysis of linkage andof alleles). This should allow the investigators to greatlynarrow the regions and to identify individual genes inwhich variations affect the risk for alcoholism and theother phenotypes they are studying. The COGA data set is a rich resource for further re-earch. For example, it has already provided a test of newmethods for genetic analysis, as presented at the GeneticAnalysisWorkshop 11 (Begleiter et al. 1999). In addi-tion, COGA researchers are currently re-interviewingparticipants as part of a 5-year followup. This strategywill allow the investigators to increase the reliability ofthe data and to refine the phenotypes, which in turn willenhance the power of the genetic analyses. inally, the large number of children and adolescentsin the original sample will prove invaluable as theseoung people pass through the age of greatest risk fordeveloping alcoholism. The value of the COGA data asa national resource for studies of alcoholism shouldincrease with the re-interviews and with the develop-ment of new methods for both the determination andanalysis of various genotypes. These efforts ultimatelyare expected to lead to the identification of genes thataffect the risk for alcoholism and related phenotypes.
cknowledgments The Collaborative Study on the Genetics of Alcoholisminvolves nine centers across the United States. Theprincipal investigator of COGA is H. Begleiter, Stateniversity of New York, Health Science Center atooklyn, and co-principal investigator is T. Reich,ashington University. The study sites and their prin-cipal investigators and co-investigators are: Indiananiversity (T. K. Li; J. Nurnberger, Jr.; P.M. Conneally;H.J. Edenberg); University of Iowa (R. Crowe, S.uperman); University of California at San Diego (M.Schuckit); University of Connecticut (V. Hesselbrock);tate University of New York, Health Science Center atooklyn (B. Porjesz, H. Begleiter); Washingtonniversity in St. Louis (T. Reich, C.R. Cloninger, J.Rice, A. Goate); Howard University (R. Taylor);utgers University (J. Tischfield); and SouthwestThe author thanks Dr. Tatiana Foroud for helpful References American Psychiatric Association (APA). Diagnostic and StatisticalManual of Mental Disorders, Third Edition, Revised. Washington, DC:APA, 1987.American Psychiatric Association (APA). Diagnostic and StatisticalManual of Mental Disorders, Fourth Edition. ETAL Alcohol Health & Research ETAL Electro-encephalogr

2 aphy and Clinical Neurophysiology ETAL G
aphy and Clinical Neurophysiology ETAL GeneticEpidemiology , H.J.Regulation of the mammalian alcohol dehydrogenase Progress in Nucleic Acid Research and Molecular Biology ETAL DRD2 ) withalcoholism. Alcoholism: Clinical and Experimental Research 22:505Ð512, 1998 a. ETAL HTT affect the risk for alcohol dependence. Alcoholism: Clinical and Experimental Research 22:1080Ð1085, 1998 b. he Collaborative Study on the Genetics of Alcoholism 218 ETAL Archives of General Psychiatry ETAL Alcoholism: ETAL Alcoholism: Clinical and Experimental Research ETAL Genetics andAlcoholism: The COGA Project. ETAL American Journal of Psychiatry ETAL Alcoholism: ETAL Proceedings of the National Academy of Sciencesof ETAL American Journalof Medical Genetics (Neuropsychiatric Genetics) ETAL AmericanJournal of Medical Genetics (Neuropsychiatric Genetics) ETAL American Journal of HumanGenetics 65:1148Ð1160, 1999. International Classification ofDiseases and Related Health Problems, Tenth Revision, Vols. 1Ð3. For a free subscription toAttn.: Fax: (703) 312–5230. What Can Be Done to Stop Underage Drinking? findings relevant to this and other questions can be found in cohol Alert , the quarterlybulletin publishedbythe National Institute on Alcohol Abuse and Alcoholism. lcohol Alert provides timely information on alcohol research and trissue addresses a specifictopic in alcohol research and summarizes critical findings in a brief, four-page, easy-to-read format. nderage Drinking: A Major Public Health Challenge (No. 59)Ñdescribes some of the most harm-ful consequences of underage drinking as well as prevention and treatment approaches that can beapplied successfully to meet the unique needs of this age group. 214 he Collaborative Study on the Genetics ofalcoholism (i.e., alcohol dependence) and alcohol-related ing to alcoholism risk, COGA required a large samplesize to allow detection of the genetic ÒsignalÓ throughthe Ònoise.Ó Of particular concern was the likely vari-of genetic and environmental factors contributing toalcoholism risk; therefore, the contribution of any oneariation in risk. The investigators chose a family studydesign to allow the use of multiple methods of geneticanalysis. Systematic recruitment from outpatient andinpatient alcoholism treatment facilities and assessmentof families initially was carried out at six sites across the and Statistical Manual of Mental Disorders, Thirddition, Revised Psychiatric Association [APA] [1987]) may have dif-ferent sets of symptoms, greatly complicating geneticanalyses. Therefore, COGA researchers gathered adetailed psychiatric history of each participant, alongwith electrophysiological data (electroencephalograms[EEGs] and event-related potentials [ERPs]). Theseegleiter et al. 1995, 1998; Hesselbrock et al. 2001) nternational Classification ofDiseases and Related Problems (ICDÐ10) of the Worldealth Organization (WHO) (1992Ð1994).phenotypes that would be collected. Therefore, he Collaborative Study on the Genetics of Alcoholism: An Update Howard J. Edenberg, Ph.D. HJ. EDENBERG, PH 1individual. T he Collaborative Study on the Genetics of Alcoholism (COGA) is a large-scale family study designed to identify genes that affect the risk for alcoholism (i.e., alcohol dependence) and alcohol-related characteristics and behaviors (i.e., phenotypes1). This collaborative project is funded by the National Institute on Alcohol Abuse and Alcoholism. Data collection, analysis, and/or storage for this study take place at nine sites across the United States. Because alcoholism is a complex genetic disorder, the COGA researchers expected that multiple genes would contribute to the risk. In other words, there will be no single Ògene for alcoholismÓ but rather variations in many different genes that together, interacting with the environment, place some people at significantly higher risk for the disease. This genetic and onmental variability (i.e., heterogeneity) makes the task of identifying individual genes difficult. However, the COGA project was designed with these difficulties in mind and incorporated strategies to meet the challenges. ticle briefly reviews these strategies and summarizes some of the results already obtained in the ongoing COGA study. S tudy Design B ecause of the expected complexity of factors contributing to alcoholism risk, COGA required a large sample size to allow detection of the genetic ÒsignalÓ through the Ònoise.Ó Of particular concern was the likely variability within the sample of both the number and type onmental factors contributing to alcoholism risk; therefore, the contribution of any one factor would only account for a small fraction of the ariation in risk. The investigators chose a family study design to allow the use of multiple methods of genetic analysis. Systematic recruitment from outpatient and inpatient alcoholism treatment facilities and assessment of families initially was carried out at six sites across the United States, with a seventh site more recently. The study also included a large sample of control families that were randomly selected from the community. For the analyses, the researchers chose a split-sample designÑtwo groups of subjects (i.e., an initial sample and a replication sample) were analyzed independently; this approach allows investigators to examine the reproducibility of the initial study findings. Because of the complexity of the risk factors for alcoholism and of the disorder itself, the COGA project was designed to gather extensive data from the participants. Although standard diagnostic systems for alcoholism can reliably determine who needs treatment, the diagnostic criteria used in these systems comprise problems in many domains of functioning. This means that two people with the same diagnosis (e.g., alcohol dependence as defined in the Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised [DSMÐIIIÐR] of the American Psychiatric Association [APA] [1987]) may have different sets of symptoms, greatly complicating genetic analyses. Therefore, COGA researchers gathered a detailed psychiatric history of each participant, along with electrophysiological data (electroencephalograms [EEGs] and event-related potentials [ERPs]). These multiple domains of data (described in detail in egleiter et al. 1995, 1998; Hesselbrock et al. 2001) provide a rich resource for exploring phenotypes related to alcoholism. In addition, they allow analyses under standard diagnostic systems, such as the 4th edition of the DSM (DSMÐIV) (APA 1994) and the 10th edition of the I nternational Classification of Diseases and Related Problems (ICDÐ10) of the W orld Health Organization (WHO) (1992Ð1994). The strategies for genetic analyses in the COGA study also had to accommodate the anticipated efore, COGA investigators chose an unbiased survey of the entireor participants from families with T he Collaborative Study on the Genetics of Alcoholism: An Updat Ho ward J. Edenberg, Ph.D. H OWARD J. EDENBERG, PH.D., Õs Professor in the Department of Biochemistry and Molecular Biology and the Department of Medical and Molecular Genetics, Indiana University School of Medicine, Indianapolis, Indiana. The Collaborative Study on the Genetics of Alcoholism is supported by National Institute on Alcohol Abuse and Alcoholism grant U10ÐAAÐ08403. 1 The term ÒphenotypeÓ refers to any observable characteristic or behavior of an individual. 214 Alcohol Research & Health thr ee or more alcoholic family members, the investigators conducted genetic analyses using microsatellite markersÑDNA regions located across all chromosomes, in which short repeated sequences exist in many variants (i.e., alleles). This process is called genotyping. Moee been generated on 2,310 people om families of alcoholics and from control families. By monitoring the inheritance patterns of such marker alleles within families with alcoholic members, the investigators could identify chromosomal regions that influence (i.e., show genetic linkage with) certain alcohol-related traits. The methods used in these genetic analyses and other aspects of the COGA study are described in more detail in the article by Bierut and colleagues, pp. 208Ð213, in this issue. Re sults of Genetic Analyses DNA Regions with Susceptibility Genes. G enetic analyses using the diagnostic criteria for alcohol dependence as the phenotype have revealed regions on several chromosomes that appear to contain genes affecting the risk for alcoholism. The primary analyses were based upon determining the extent of allele sharing among siblings who meet diagnostic criteria for alcoholism. The primary COGA definition of being affected with alcoholism requires a person to meet both DSMÐIIIÐR criteria for alcohol dependence and eighner criteria (Feighner et al. 1972) for definite alcoholism. If siblings who are alcoholic share more alleles at a marker than would be expected based on chance, this suggests that genes within the chromosomal region containing the marker contribute to the risk of alcoholism. Analyses of 987 people from 105 families in the initial sample provided evidence that regions on 3 chromosomes contained genes that increase the risk for alcoholism (Reich et al. 1998). The strongest evidence was for regions on chromosomes 1 and 7, with more modest evidence for a region on chromosome 2. The DNA regions identified through these analyses weeoad, as is typical for studies of complex genetic diseases, and therefore are likely to contain numerous genes. Much additional work is required to narrowegions and attempt to deter-mine which specific gene or genes play a role in affecting the risk for alcoholism. Therefore, additional markers within these regions of interest wee analyzed in the same people. Subsequent analyses that included the additional markers supported the initial findings (Foroud et al. 2000) but did not narrowchromosomal regions in which genes influencing alcoholism susceptibility are likely to lie. The data from the second part of the split sampleÑthe replication sample, which comprised 1,295 people from 157 familiesÑgenerally supported the initial findings (Foroud et al. 2000). Thus, the replication sample again provided evidence that genes increasing the risk of alcoholism weesame regions of chromosomes 1 and 7, albeit with less statistical support. When the initial and replication samples weeomosomal regions remained the strongest candidates for containing genes influencing the risk of alcoholism. Evidence for the region on chromosome 2 increased with the additional markers in the initial sample, but the replication sample provided no additional evidence for alcoholism susceptibility genes in this chromosomal region. Conversely, the strongest evidence in the replication sample for a region containing genes affecting the risk for alcoholism was on chromosome 3, which had shown no evidence of being linked with alcoholism in the initial sample. Because of the large number of genes that may contribute to the risk for a complex genetic disease such as alcowee,dent sample, even one collected byoup of researchers, might replicate some previously identified genes and also identify some novel alcoholism susceptibility loci. Foroud and colleagues (2000) also analyzed the combined data set from the initial and replication samples using a more restricted definition of alcoholism as specified in the ICDÐ10. This restriction greatly reduced the number of sibling pairs in the comparison. The region on chromosome 1 provided the strongest evidence for a susceptibility gene in the combined sample. In addition, this new evaluation detected a region on chromosome 8 that was linked with the risk for alcoholism. DNA R egions with Protective Genes. I nterestingly, analyses of nonalcoholic sibling pairs in the initial sample produced evidence for a protective region on chromosome 4, in the general vicinity of the alcohol dehydrogenase ( ADH ) genes. 2 A related analysis, using a technique that treat

3 s alcoholism as the extreme of a distrib
s alcoholism as the extreme of a distribution of an underlying quantitative trait,3 showed evidence for linkage to this same region T he Collaborative Study on the Genetics of Alcoholism 2 Alcohol dehydrogenase is an enzyme that helps break down alcohol in the body. Several ADH genes exist, each of which has several alleles. 3Quantitative traits are characteristics that are distributed along a continuum across a population, such as height. Vol. 26, No. 3, 2002 215 (W illiams et al. 1999). This finding suggests that variants of a gene or genes within this region reduced the risk of becoming alcoholic. ADH alleles ar e known to affect the risk for alcoholism; however, the known protective alleles occur at high frequency in Asian populations but are rare in the Caucasian population that makes up most of the COGA sample (Edenberg 2000). Therefore, these analyses may have identified a new protective ADH allele or another pr otective gene located nearby. The number of unaffected sibling pairs genotyped in the replication sample was too small to analyze. Another phenotype that may reflect a protective influence against alcoholism is the maximum number of drinks a person has consumed in a 24-hour period (MAXDRINKS). This phenotype is quantitative and heritable, and a low number of drinks consumed in a 24-hour period may reflect a reduced tolerance for high levels of alcohol. An advantage of a quantitative phenotype is that everyone in a study can con-tribute to the MAXDRINK phenotype in both the initial and replication data sets (and in the combined sample) showed the strongest evidence for linkage in the same region of chromosome 4 where the ADH genes r eside (Saccone et al. 2000). This finding suggests that the gene or genes influencing the MAXDRINKS phenotype may be related to the protective region identified in the unaffected sibling pairs and to protective effects of certain ADH alleles (Edenberg 2000). DNA R egions Related to Symptoms of Alcoholism. The symptoms used to establish a diagnosis of alcoholism are highly diverse, ranging from more biological symptoms (e.g., tolerance and withdrawal) to more social symptoms (e.g., social and legal problems). Each person diagnosed with alcoholism exhibits a efore, a diagnosis of alcoholism does not reflect a uniform phenotype. This lack of uniformity complicates genetic analyses. Consequently, researchers have constructed other, more defined phenotypes from the data obtained in the COGA interviews. These include an analysis of symptoms related to alcoholism that produced phenotypes which appeared to reflect the severity of alcohol problems. Analysis of these phenotypes provided evidence for a DNA region on chromosome 16 that was associated with an increased risk for more severe alcohol problems (Foroud et al. 1998). DNA R egions Associated with Co-Occurring Disorders. M any people in the COGA families of alcoholics also met the DSMÐIIIÐR diagnostic criteria for major depressive disorder or depressive syndrome (Nurnberger et al. 2001). Depression alone showed modest evidence of linkage to a region on chromosome 7. The phenotype characterized byholism and depression showed evidence of linkage to a region on chromosome 2, primarily in the replication sample. The most interesting finding was for the broad Òalcoholism or depr essionÓ phenotype, with very strong evidence for linkage to the same region of chromosome 1 that was linked to alcoholism alone (Nurnberger et al. 2001). This suggests that a gene or genes within this chromosomal region increase the risk for both alcoholism and depression. (For more information on these analyses, see the article by Nurnberger and colleagues, pp. 233Ð240, in this issue.) DNA R egions Linked with Electrophysiological Measures. The COGA inv estigators also evaluated electrophysiological variables, such as EEGs and ERPs, from study participants. EEGs measure overall brain activity, whereas ERPs are brain waves elicited in response to specific stimuli (e.g., a light or sound). Analysis of such electrophysiological data may reeal a subset of genes that affect these quantitative, biological phenotypes related to alcoholism (Porjesz et al. 1998, 2002). One component of an ERP is a brain wave called P300, which typically occurs 300 milliseconds after a stimulus. Previous studies had found that a reduced amplitude of the P300 wave is a heritable phenotype that correlates with alcohol dependence and other psychiatric disorders (Porjesz et al. 1998). The genetic analyses of the COGA participants identified four regions, on chromosomes 2, 5, 6, and 13, that appear to contain genes affecting the amplitude of the P300 (Begleiter et al. 1998). In addition to these findings, recent analyses demonstrate strong evidence for a locus that affects brain wave oscillations as measured by electroencephalography (Porjesz et al. 2002). Thus, a gene or genes that affect brain rhythms lies in a region of chromosome 4 that contains a cluster of genes encoding proteins (i.e., receptors) which interact with the brain chemical gamma-aminobutyric acid (GABA). C andidate Genes. COGA r esearchers have also analyzed candidate genesÑgenes suspected to play a role in the development of alcoholism based on other studies. Some of these candidate genes encode components of various brain chemical systems that allow communication among nerve cells. Two of these genes are the dopamine D2 receptor gene ( DRD 2 ) and a ser otonin transporter gene ( HT T) . H owever, the analyses found 216 Alcohol Research & Health no evidence that DRD 2 affected the risk for alcoholism (Edenberg et al. 1998 a ) or that HT T was linked to either alcoholism in general or to a more severe form of alcoholism (Edenberg et al. 1998 b ). P erspective Wher e does the COGA study go from here? The increasing availability of the DNA sequence of the entire human genome and knowledge of variations in that sequence among people are greatly aiding the cur-rent phase of the research. Particularly important to the current work is the use of the sequence data to identify which genes are located within the regions that have shown linkage with alcoholism and the other phenotypes examined in the COGA analyses and to identify variations (i.e., polymorphisms) within those genes. Where the available data are incomplete or insufficient, COGA researchers are seeking these polymorphisms themselves. Of particular value are single-nucleotide polymorphisms (SNPs)Ñsites at which people differ in a single base pairÑin or near genes within the regions of interest. COGA investigators are doing additional genotyping of SNPs in and near candidate genes in the regions of linkage for further analysis of linkage and linkage disequilibrium (i.e., the nonrandom association w the investigators to greatly narrow the regions and to identify individual genes in which variations affect the risk for alcoholism and the other phenotypes they are studying. The COGA data set is a rich resource for further re-search. For example, it has already provided a test of new methods for genetic analysis, as presented at the Genetic Analysis Workshop 11 (Begleiter et al. 1999). In addition, COGA researchers are currently re-interviewing participants as part of a 5-year followup. This strategy will allow the investigators to increase the reliability of the data and to refine the phenotypes, which in turn will enhance the power of the genetic analyses. Finally, the large number of children and adolescents in the original sample will prove invaluable as these young people pass through the age of greatest risk for developing alcoholism. The value of the COGA data as a national resource for studies of alcoholism should increase with the re-interviews and with the development of new methods for both the determination and analysis of various genotypes. These efforts ultimately are expected to lead to the identification of genes that affect the risk for alcoholism and related phenotypes.
A cknowledgments The Collaborativ e Study on the Genetics of Alcoholism involves nine centers across the United States. The principal investigator of COGA is H. Begleiter, State University of New York, Health Science Center at Brooklyn, and co-principal investigator is T. Reich, Washington University. The study sites and their principal investigators and co-investigators are: Indiana University (T. K. Li; J. Nurnberger, Jr.; P.M. Conneally; H.J. Edenberg); University of Iowa (R. Crowe, S. Kuperman); University of California at San Diego (M. Schuckit); University of Connecticut (V. Hesselbrock); State University of New York, Health Science Center at Brooklyn (B. Porjesz, H. Begleiter); Washington University in St. Louis (T. Reich, C.R. Cloninger, J. Rice, A. Goate); Howard University (R. Taylor); Rutgers University (J. Tischfield); and Southwest Foundation (L. Almasy). The author thanks Dr. Tatiana Foroud for helpful comments on the manuscript. References American Psychiatric Association (APA). Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised. Washington, DC: APA, 1987. American Psychiatric Association (APA). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington, DC: APA, 1994. EGLEITER, H.; REICH, T.; HESSELBROCK, V.; ET AL. The Collaborative Study on the Genetics of Alcoholism. Alcohol Health & Research World 19:228Ð236, 1995. BEGLEITER, H.; PORJESZ, B.; REICH, T.; ET AL. Quantitative trait loci analysis of human event-related brain potentials: P3 voltage. Electro encephalography and Clinical Neurophysiology 108:244Ð250, 1998. BEGLEITER, H.; REICH, T.; NURNBERGER, J., JR.; ET AL. Description of the Genetic Analysis Workshop 11 Collaborative Study on the Genetics of Alcoholism. Genetic Epidemiology 17(Suppl.1):S25ÐS30, 1999. EDENBERG, H.J. genes. Progress in Nucleic Acid Research and Molecular Biology 64:295Ð341, 2000. EDENBERG, H.J.; FOROUD, T.; KOLLER, D.L.; ET AL.A family-based analysis of the association of the dopamine D2 receptor ( DRD2 ) with alcoholism. Alcoholism: Clinical and Experimental Research 22:505Ð512, 1998 a. E DENBERG, H.J.; REYNOLDS, J.; KOLLER, D.L.; ET AL. A family-based analysis of whether the functional promoter alleles of the serotonin HTT affect the risk for alcohol dependence. Alcoholism: Clinical and Experimental Research 22:1080Ð1085, 1998 b. T he Collaborative Study on the Genetics of Alcoholism Vol. 26, No. 3, 2002 217 F EIGHNER, J.P.; ROBINS, E.; GUZE, S.B.; ET AL.Diagnostic criteria for use in psychiatric research. Archives of General Psychiatry 26:57Ð63, 1972. FOROUD, T.; BUCHOLZ, K.K.; EDENBERG, H.J.; ET AL. alcoholism-related severity phenotype to chromosome 16. Alcoholism: Clinical and Experimental Research 22:2035Ð2042, 1998. FOROUD, T.; EDENBERG, H.J.; GOATE, A.; ET AL. Alcoholism susceptibility loci: Confirmation studies in a replicate sample and further map-ping. Alcoholism: Clinical and Experimental Research 24:933Ð945, 2000. HESSELBROCK, V.; FOROUD, T.; EDENBERG, H.; ET AL. Genetics and Alcoholism: The COGA Project. New York: Marcel Decker, 2001. NURNBERGER, J.I., JR.; FOROUD, T.; FLURY, L.; ET AL. Evidence for a locus on chromosome 1 that influences vulnerability to alcoholism and American Journal of Psychiatry 158:718Ð724, 2001. PORJESZ, B.; BEGLEITER, H.; REICH, T.; ET AL. event-related potential as a phenotypic marker for a predisposition to Alcoholism: Clinical and Experimental Research 22:1317Ð1323, 1998. PORJESZ, B.; ALMASY, L.; EDENBERG, H.J.; ET AL. rium between the beta frequency of the human EEG and a GABAA receptor gene locus. Proceedings of the National Academy of Sciences of the USA 99:3729Ð3733, 2002. REICH, T.; EDENBERG, H.J.; GOATE, A.; ET AL.A genome-wide search for genes affecting the risk for alcohol dependence. American Journal of Medical Genetics (Neuropsychiatric Genetics) 81:207Ð215, 1998. SACCONE, N.L.; KWON, J.M.; CORBETT, J.; ET AL.A genome screen of

4 maximum number of drinks as an alcoholi
maximum number of drinks as an alcoholism phenotype. American Journal of Medical Genetics (Neuropsychiatric Genetics) 96:632Ð637, 2000. WILLIAMS, J.T.; BEGLEITER, H.; PORJESZ, B.; ET AL. linkage analysis of multivariate qualitative and quantitative traits. II. Alcoholism and event-related potentials. American Journal of Human Genetics 65:1148Ð1160, 1999. World Health Organization (WHO). International Classification of Diseases and Related Health Problems, Tenth Revision, Vols. 1Ð3. Geneva: WHO, 1992Ð1994. For a free subscription to Alcohol Alert, write to: National Institute on Alcohol Abuse and Alcoholism, Attn.: Alcohol Alert, Publications Distribution Center, P.O. Box 10686, Rockville, MD 20849–0686. Fax: (703) 312–5230. Alcohol Alert is also available on NIAAA’s Web site (http://www.niaaa.nih.gov). What C an Beone to Stop Underage Drinking? N ew findings relevant to this and other questions can be found in Al cohol Alert , the quar terly bulletin published by the National Institute on Alcohol Abuse and Alcoholism. A lcohol Alert pro vides timely information on alcohol research and treatment. Each issue addresses a specific topic in alcohol research and summarizes critical findings in a brief, four-page, easy-to-read format. U nderage Drinking: A Major Public Health Challenge applied successfully to meet the unique needs of this age gr oup(No.ful consequences of underage drinking as well as prevention and treatment approaches that can be . 218 Alcohol Research & Health T he Collaborative Study on the Genetics of Alcoholism (COGA) is a large-scale family study designed to identify genes that affect the risk for alcoholism (i.e., alcohol dependence) and alcohol-related characteristics and behaviors (i.e., phenotypes1). This collaborative project is funded by the National Institute on Alcohol Abuse and Alcoholism. Data collection, analysis, and/or storage for this study take place at nine sites across the United States. Because alcoholism is a complex genetic disorder, the COGA researchers expected that multiple genes would contribute to the risk. In other words, there will be no single Ògene for alcoholismÓ but rather variations in many different genes that together, interacting with the environment, place some people at significantly higher risk for the disease. This genetic and onmental variability (i.e., heterogeneity) makes the task of identifying individual genes difficult. However, the COGA project was designed with these difficulties in mind and incorporated strategies to meet the challenges. ticle briefly reviews these strategies and summarizes some of the results already obtained in the ongoing COGA study. S tudy Design B ecause of the expected complexity of factors contributing to alcoholism risk, COGA required a large sample size to allow detection of the genetic ÒsignalÓ through the Ònoise.Ó Of particular concern was the likely variability within the sample of both the number and type onmental factors contributing to alcoholism risk; therefore, the contribution of any one factor would only account for a small fraction of the ariation in risk. The investigators chose a family study design to allow the use of multiple methods of genetic analysis. Systematic recruitment from outpatient and inpatient alcoholism treatment facilities and assessment of families initially was carried out at six sites across the United States, with a seventh site more recently. The study also included a large sample of control families that were randomly selected from the community. For the analyses, the researchers chose a split-sample designÑtwo groups of subjects (i.e., an initial sample and a replication sample) were analyzed independently; this approach allows investigators to examine the reproducibility of the initial study findings. Because of the complexity of the risk factors for alcoholism and of the disorder itself, the COGA project was designed to gather extensive data from the participants. Although standard diagnostic systems for alcoholism can reliably determine who needs treatment, the diagnostic criteria used in these systems comprise problems in many domains of functioning. This means that two people with the same diagnosis (e.g., alcohol dependence as defined in the Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised [DSMÐIIIÐR] of the American Psychiatric Association [APA] [1987]) may have different sets of symptoms, greatly complicating genetic analyses. Therefore, COGA researchers gathered a detailed psychiatric history of each participant, along with electrophysiological data (electroencephalograms [EEGs] and event-related potentials [ERPs]). These multiple domains of data (described in detail in egleiter et al. 1995, 1998; Hesselbrock et al. 2001) provide a rich resource for exploring phenotypes related to alcoholism. In addition, they allow analyses under standard diagnostic systems, such as the 4th edition of the DSM (DSMÐIV) (APA 1994) and the 10th edition of the I nternational Classification of Diseases and Related Problems (ICDÐ10) of the W orld Health Organization (WHO) (1992Ð1994). The strategies for genetic analyses in the COGA study also had to accommodate the anticipated efore, COGA investigators chose an unbiased survey of the entireor participants from families with T he Collaborative Study on the Genetics of Alcoholism: An Updat Ho ward J. Edenberg, Ph.D. H OWARD J. EDENBERG, PH.D., Õs Professor in the Department of Biochemistry and Molecular Biology and the Department of Medical and Molecular Genetics, Indiana University School of Medicine, Indianapolis, Indiana. The Collaborative Study on the Genetics of Alcoholism is supported by National Institute on Alcohol Abuse and Alcoholism grant U10ÐAAÐ08403. 1 The term ÒphenotypeÓ refers to any observable characteristic or behavior of an individual. 214 Alcohol Research & Health thr ee or more alcoholic family members, the investigators conducted genetic analyses using microsatellite markersÑDNA regions located across all chromosomes, in which short repeated sequences exist in many variants (i.e., alleles). This process is called genotyping. Moee been generated on 2,310 people om families of alcoholics and from control families. By monitoring the inheritance patterns of such marker alleles within families with alcoholic members, the investigators could identify chromosomal regions that influence (i.e., show genetic linkage with) certain alcohol-related traits. The methods used in these genetic analyses and other aspects of the COGA study are described in more detail in the article by Bierut and colleagues, pp. 208Ð213, in this issue. Re sults of Genetic Analyses DNA Regions with Susceptibility Genes. G enetic analyses using the diagnostic criteria for alcohol dependence as the phenotype have revealed regions on several chromosomes that appear to contain genes affecting the risk for alcoholism. The primary analyses were based upon determining the extent of allele sharing among siblings who meet diagnostic criteria for alcoholism. The primary COGA definition of being affected with alcoholism requires a person to meet both DSMÐIIIÐR criteria for alcohol dependence and eighner criteria (Feighner et al. 1972) for definite alcoholism. If siblings who are alcoholic share more alleles at a marker than would be expected based on chance, this suggests that genes within the chromosomal region containing the marker contribute to the risk of alcoholism. Analyses of 987 people from 105 families in the initial sample provided evidence that regions on 3 chromosomes contained genes that increase the risk for alcoholism (Reich et al. 1998). The strongest evidence was for regions on chromosomes 1 and 7, with more modest evidence for a region on chromosome 2. The DNA regions identified through these analyses weeoad, as is typical for studies of complex genetic diseases, and therefore are likely to contain numerous genes. Much additional work is required to narrowegions and attempt to deter-mine which specific gene or genes play a role in affecting the risk for alcoholism. Therefore, additional markers within these regions of interest wee analyzed in the same people. Subsequent analyses that included the additional markers supported the initial findings (Foroud et al. 2000) but did not narrowchromosomal regions in which genes influencing alcoholism susceptibility are likely to lie. The data from the second part of the split sampleÑthe replication sample, which comprised 1,295 people from 157 familiesÑgenerally supported the initial findings (Foroud et al. 2000). Thus, the replication sample again provided evidence that genes increasing the risk of alcoholism weesame regions of chromosomes 1 and 7, albeit with less statistical support. When the initial and replication samples weeomosomal regions remained the strongest candidates for containing genes influencing the risk of alcoholism. Evidence for the region on chromosome 2 increased with the additional markers in the initial sample, but the replication sample provided no additional evidence for alcoholism susceptibility genes in this chromosomal region. Conversely, the strongest evidence in the replication sample for a region containing genes affecting the risk for alcoholism was on chromosome 3, which had shown no evidence of being linked with alcoholism in the initial sample. Because of the large number of genes that may contribute to the risk for a complex genetic disease such as alcowee,dent sample, even one collected byoup of researchers, might replicate some previously identified genes and also identify some novel alcoholism susceptibility loci. Foroud and colleagues (2000) also analyzed the combined data set from the initial and replication samples using a more restricted definition of alcoholism as specified in the ICDÐ10. This restriction greatly reduced the number of sibling pairs in the comparison. The region on chromosome 1 provided the strongest evidence for a susceptibility gene in the combined sample. In addition, this new evaluation detected a region on chromosome 8 that was linked with the risk for alcoholism. DNA R egions with Protective Genes. I nterestingly, analyses of nonalcoholic sibling pairs in the initial sample produced evidence for a protective region on chromosome 4, in the general vicinity of the alcohol dehydrogenase ( ADH ) genes. 2 A related analysis, using a technique that treats alcoholism as the extreme of a distribution of an underlying quantitative trait,3 showed evidence for linkage to this same region T he Collaborative Study on the Genetics of Alcoholism 2 Alcohol dehydrogenase is an enzyme that helps break down alcohol in the body. Several ADH genes exist, each of which has several alleles. 3Quantitative traits are characteristics that are distributed along a continuum across a population, such as height. Vol. 26, No. 3, 2002 215 (W illiams et al. 1999). This finding suggests that variants of a gene or genes within this region reduced the risk of becoming alcoholic. ADH alleles ar e known to affect the risk for alcoholism; however, the known protective alleles occur at high frequency in Asian populations but are rare in the Caucasian population that makes up most of the COGA sample (Edenberg 2000). Therefore, these analyses may have identified a new protective ADH allele or another pr otective gene located nearby. The number of unaffected sibling pairs genotyped in the replication sample was too small to analyze. Another phenotype that may reflect a protective influence against alcoholism is the maximum number of drinks a person has consumed in a 24-hour period (MAXDRINKS). This phenotype is quantitative and heritable, and a low number of drinks consumed in a 24-hour period may reflect a reduced tolerance for high levels of alcohol. An advantage of a quantitative phenotype is that everyone in a study can con-tribute to the MAXDRINK phenotype in both the initial an

5 d replication data sets (and in the comb
d replication data sets (and in the combined sample) showed the strongest evidence for linkage in the same region of chromosome 4 where the ADH genes r eside (Saccone et al. 2000). This finding suggests that the gene or genes influencing the MAXDRINKS phenotype may be related to the protective region identified in the unaffected sibling pairs and to protective effects of certain ADH alleles (Edenberg 2000). DNA R egions Related to Symptoms of Alcoholism. The symptoms used to establish a diagnosis of alcoholism are highly diverse, ranging from more biological symptoms (e.g., tolerance and withdrawal) to more social symptoms (e.g., social and legal problems). Each person diagnosed with alcoholism exhibits a efore, a diagnosis of alcoholism does not reflect a uniform phenotype. This lack of uniformity complicates genetic analyses. Consequently, researchers have constructed other, more defined phenotypes from the data obtained in the COGA interviews. These include an analysis of symptoms related to alcoholism that produced phenotypes which appeared to reflect the severity of alcohol problems. Analysis of these phenotypes provided evidence for a DNA region on chromosome 16 that was associated with an increased risk for more severe alcohol problems (Foroud et al. 1998). DNA R egions Associated with Co-Occurring Disorders. M any people in the COGA families of alcoholics also met the DSMÐIIIÐR diagnostic criteria for major depressive disorder or depressive syndrome (Nurnberger et al. 2001). Depression alone showed modest evidence of linkage to a region on chromosome 7. The phenotype characterized byholism and depression showed evidence of linkage to a region on chromosome 2, primarily in the replication sample. The most interesting finding was for the broad Òalcoholism or depr essionÓ phenotype, with very strong evidence for linkage to the same region of chromosome 1 that was linked to alcoholism alone (Nurnberger et al. 2001). This suggests that a gene or genes within this chromosomal region increase the risk for both alcoholism and depression. (For more information on these analyses, see the article by Nurnberger and colleagues, pp. 233Ð240, in this issue.) DNA R egions Linked with Electrophysiological Measures. The COGA inv estigators also evaluated electrophysiological variables, such as EEGs and ERPs, from study participants. EEGs measure overall brain activity, whereas ERPs are brain waves elicited in response to specific stimuli (e.g., a light or sound). Analysis of such electrophysiological data may reeal a subset of genes that affect these quantitative, biological phenotypes related to alcoholism (Porjesz et al. 1998, 2002). One component of an ERP is a brain wave called P300, which typically occurs 300 milliseconds after a stimulus. Previous studies had found that a reduced amplitude of the P300 wave is a heritable phenotype that correlates with alcohol dependence and other psychiatric disorders (Porjesz et al. 1998). The genetic analyses of the COGA participants identified four regions, on chromosomes 2, 5, 6, and 13, that appear to contain genes affecting the amplitude of the P300 (Begleiter et al. 1998). In addition to these findings, recent analyses demonstrate strong evidence for a locus that affects brain wave oscillations as measured by electroencephalography (Porjesz et al. 2002). Thus, a gene or genes that affect brain rhythms lies in a region of chromosome 4 that contains a cluster of genes encoding proteins (i.e., receptors) which interact with the brain chemical gamma-aminobutyric acid (GABA). C andidate Genes. COGA r esearchers have also analyzed candidate genesÑgenes suspected to play a role in the development of alcoholism based on other studies. Some of these candidate genes encode components of various brain chemical systems that allow communication among nerve cells. Two of these genes are the dopamine D2 receptor gene ( DRD 2 ) and a ser otonin transporter gene ( HT T) . H owever, the analyses found 216 Alcohol Research & Health no evidence that DRD 2 affected the risk for alcoholism (Edenberg et al. 1998 a ) or that HT T was linked to either alcoholism in general or to a more severe form of alcoholism (Edenberg et al. 1998 b ). P erspective Wher e does the COGA study go from here? The increasing availability of the DNA sequence of the entire human genome and knowledge of variations in that sequence among people are greatly aiding the cur-rent phase of the research. Particularly important to the current work is the use of the sequence data to identify which genes are located within the regions that have shown linkage with alcoholism and the other phenotypes examined in the COGA analyses and to identify variations (i.e., polymorphisms) within those genes. Where the available data are incomplete or insufficient, COGA researchers are seeking these polymorphisms themselves. Of particular value are single-nucleotide polymorphisms (SNPs)Ñsites at which people differ in a single base pairÑin or near genes within the regions of interest. COGA investigators are doing additional genotyping of SNPs in and near candidate genes in the regions of linkage for further analysis of linkage and linkage disequilibrium (i.e., the nonrandom association w the investigators to greatly narrow the regions and to identify individual genes in which variations affect the risk for alcoholism and the other phenotypes they are studying. The COGA data set is a rich resource for further re-search. For example, it has already provided a test of new methods for genetic analysis, as presented at the Genetic Analysis Workshop 11 (Begleiter et al. 1999). In addition, COGA researchers are currently re-interviewing participants as part of a 5-year followup. This strategy will allow the investigators to increase the reliability of the data and to refine the phenotypes, which in turn will enhance the power of the genetic analyses. Finally, the large number of children and adolescents in the original sample will prove invaluable as these young people pass through the age of greatest risk for developing alcoholism. The value of the COGA data as a national resource for studies of alcoholism should increase with the re-interviews and with the development of new methods for both the determination and analysis of various genotypes. These efforts ultimately are expected to lead to the identification of genes that affect the risk for alcoholism and related phenotypes.
A cknowledgments The Collaborativ e Study on the Genetics of Alcoholism involves nine centers across the United States. The principal investigator of COGA is H. Begleiter, State University of New York, Health Science Center at Brooklyn, and co-principal investigator is T. Reich, Washington University. The study sites and their principal investigators and co-investigators are: Indiana University (T. K. Li; J. Nurnberger, Jr.; P.M. Conneally; H.J. Edenberg); University of Iowa (R. Crowe, S. Kuperman); University of California at San Diego (M. Schuckit); University of Connecticut (V. Hesselbrock); State University of New York, Health Science Center at Brooklyn (B. Porjesz, H. Begleiter); Washington University in St. Louis (T. Reich, C.R. Cloninger, J. Rice, A. Goate); Howard University (R. Taylor); Rutgers University (J. Tischfield); and Southwest Foundation (L. Almasy). The author thanks Dr. Tatiana Foroud for helpful comments on the manuscript. References American Psychiatric Association (APA). Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised. Washington, DC: APA, 1987. American Psychiatric Association (APA). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington, DC: APA, 1994. EGLEITER, H.; REICH, T.; HESSELBROCK, V.; ET AL. The Collaborative Study on the Genetics of Alcoholism. Alcohol Health & Research World 19:228Ð236, 1995. BEGLEITER, H.; PORJESZ, B.; REICH, T.; ET AL. Quantitative trait loci analysis of human event-related brain potentials: P3 voltage. Electro encephalography and Clinical Neurophysiology 108:244Ð250, 1998. BEGLEITER, H.; REICH, T.; NURNBERGER, J., JR.; ET AL. Description of the Genetic Analysis Workshop 11 Collaborative Study on the Genetics of Alcoholism. Genetic Epidemiology 17(Suppl.1):S25ÐS30, 1999. EDENBERG, H.J. genes. Progress in Nucleic Acid Research and Molecular Biology 64:295Ð341, 2000. EDENBERG, H.J.; FOROUD, T.; KOLLER, D.L.; ET AL.A family-based analysis of the association of the dopamine D2 receptor ( DRD2 ) with alcoholism. Alcoholism: Clinical and Experimental Research 22:505Ð512, 1998 a. E DENBERG, H.J.; REYNOLDS, J.; KOLLER, D.L.; ET AL. A family-based analysis of whether the functional promoter alleles of the serotonin HTT affect the risk for alcohol dependence. Alcoholism: Clinical and Experimental Research 22:1080Ð1085, 1998 b. T he Collaborative Study on the Genetics of Alcoholism Vol. 26, No. 3, 2002 217 F EIGHNER, J.P.; ROBINS, E.; GUZE, S.B.; ET AL.Diagnostic criteria for use in psychiatric research. Archives of General Psychiatry 26:57Ð63, 1972. FOROUD, T.; BUCHOLZ, K.K.; EDENBERG, H.J.; ET AL. alcoholism-related severity phenotype to chromosome 16. Alcoholism: Clinical and Experimental Research 22:2035Ð2042, 1998. FOROUD, T.; EDENBERG, H.J.; GOATE, A.; ET AL. Alcoholism susceptibility loci: Confirmation studies in a replicate sample and further map-ping. Alcoholism: Clinical and Experimental Research 24:933Ð945, 2000. HESSELBROCK, V.; FOROUD, T.; EDENBERG, H.; ET AL. Genetics and Alcoholism: The COGA Project. New York: Marcel Decker, 2001. NURNBERGER, J.I., JR.; FOROUD, T.; FLURY, L.; ET AL. Evidence for a locus on chromosome 1 that influences vulnerability to alcoholism and American Journal of Psychiatry 158:718Ð724, 2001. PORJESZ, B.; BEGLEITER, H.; REICH, T.; ET AL. event-related potential as a phenotypic marker for a predisposition to Alcoholism: Clinical and Experimental Research 22:1317Ð1323, 1998. PORJESZ, B.; ALMASY, L.; EDENBERG, H.J.; ET AL. rium between the beta frequency of the human EEG and a GABAA receptor gene locus. Proceedings of the National Academy of Sciences of the USA 99:3729Ð3733, 2002. REICH, T.; EDENBERG, H.J.; GOATE, A.; ET AL.A genome-wide search for genes affecting the risk for alcohol dependence. American Journal of Medical Genetics (Neuropsychiatric Genetics) 81:207Ð215, 1998. SACCONE, N.L.; KWON, J.M.; CORBETT, J.; ET AL.A genome screen of maximum number of drinks as an alcoholism phenotype. American Journal of Medical Genetics (Neuropsychiatric Genetics) 96:632Ð637, 2000. WILLIAMS, J.T.; BEGLEITER, H.; PORJESZ, B.; ET AL. linkage analysis of multivariate qualitative and quantitative traits. II. Alcoholism and event-related potentials. American Journal of Human Genetics 65:1148Ð1160, 1999. World Health Organization (WHO). International Classification of Diseases and Related Health Problems, Tenth Revision, Vols. 1Ð3. Geneva: WHO, 1992Ð1994. For a free subscription to Alcohol Alert, write to: National Institute on Alcohol Abuse and Alcoholism, Attn.: Alcohol Alert, Publications Distribution Center, P.O. Box 10686, Rockville, MD 20849–0686. Fax: (703) 312–5230. Alcohol Alert is also available on NIAAA’s Web site (http://www.niaaa.nih.gov). What C an Beone to Stop Underage Drinking? N ew findings relevant to this and other questions can be found in Al cohol Alert , the quar terly bulletin published by the National Institute on Alcohol Abuse and Alcoholism. A lcohol Alert pro vides timely information on alcohol research and treatment. Each issue addresses a specific topic in alcohol research and summarizes critical findings in a brief, four-page, easy-to-read format. U nderage Drinking: A Major Public Health Challenge applied successfully to meet the unique needs of this age gr oup(No.ful consequences of underage drinking as well as prevention and treatment approaches that can be . 218 Alcohol Research & Heal