J. I. Nurnberger^{1, 4}; R. Wiegand¹; L. Bierut²; K. Bucholz²; T. Foroud^{4, 1}; H. Edenberg^{3, 4}; E. T. Meyer¹; J. Kramer⁵; D. Dick²; W. Reich²; S. Kuperman⁵; V. Hesselbrock⁷; A. Goate²; B. Porjesz⁶; H. Begleiter⁶
1. Psychiatry, Indiana University School of Medicine, Indianapolis, IN, USA.
2. Psychiatry, Washington University School of Medicine, St. Louis, MO, USA.
3. Biochemistry and Molecular Biology, Indiana University School of Medicine, Indianapolis, IN, USA.
4. Medical and Molecular Genetics, Indiana University School of Medicine, Indianapolis, IN, USA.
5. Psychiatry, University of Iowa School of Medicine, Iowa City, IA, USA.
6. Psychiatry, State University of New York Health Sciences Center, Brooklyn, NY, USA.
7. Psychiatry, University of Connecticut School of Medicine, Farmington, CT, USA.

1022 subjects from 448 families in the Collaborative Study of the Genetics of Alcoholism (COGA) were assessed at two time points separated by an interval of 4-5 years. The age of subjects at Time 2 was 17.4 + 3.4 years. At Time 2, 230 subjects were diagnosed with DSM IV alcohol dependence or abuse by personal interview and/or interview of a parent about the subject.

Characteristics at Time 1 were used to predict affected status at Time 2. Multilevel regression modeling was employed.

Controlling for age, sex, and Time 1 diagnosis, parental affected status predicted Dependence/Abuse at Time 2 (Odds Ratio 1.4, Confidence Interval 1.0-1.8, F = 3.9, p<.05); the diagnosis of Conduct Disorder at Time 1 also predicted Dependence/Abuse at Time 2 (OR 1.7, CI 1.0-2.9, F = 4.4, p = .04). The diagnoses of Major Depression and Drug Dependence were also predictive. Affected Status at Time 2 was associated with Age at First Drink as reported at Time 1 (chi-square = 30.8, p < .0001). In younger adolescents, GABRA2 (the alpha 2 subunit of the GABA-A receptor) genotype may be expressed as conduct problems (F = 4.0, p = .02) and sometimes alcohol abuse. In older adolescents, GABRA2 genotype is associated with alcohol dependence (F = 11.2, p < .0003), especially when combined with Alcohol Dehydrogenase 4 genotype. CHRM2 haplotype (muscarinic cholinergic receptor 2) also predicts DSM IV alcohol dependence (F=6.12, p<.0056) along with separation anxiety and ADHD; this may be a separate vulnerability pathway.

Additional subjects and additional vulnerability genes are now being studied prospectively. The intention is to develop models for risk and protective factors for the development of alcohol problems, incorporating genotypic information.