Authors/Institutions: M Hesselbrock, V Hesselbrock, G Chan, L Wetherill, T Foroud, L. Almasy and COGA colleagues University of Connecticut Health Center, Department of Psychiatry, Farmington, CT 06030-2103
Purpose: This presentation will focus on four novel phenotypes of alcoholism based upon the clinical signs and symptoms of ‘alcoholism’. Three phenotypes were conceptually derived, a fourth was empirically derived. Content validity, reliability and heritability for the four phenotypes will be presented.
Methods: Data were based upon SSAGA interviews of adults participating in the COGA study. Items related to drinking behavior and the consequences of heavy alcohol use, including many non-DSM diagnostic items, were selected. Four phenotypes were developed by summing the endorsement of items selected. Three phenotypes conceptual phenotypes included ‘Alcoholism’ –non-DSM clinical signs of chronic alcohol use (12 items); ‘Biological Consequences’ (15 items) – with a focus on items related to hangover, withdrawal, tolerance, and medical consequences; ‘Alcoholism Severity’ – includes 14 sample items from other non-DSM criterion sets, including blackout, psychosocial problems, etc. A fourth phenotype (20 items) was based upon a latent class analysis of alcohol related items in the SSAGA (Bucholz et al, 1996)
Data: Subjects included 3130 adult (65.8%males) moderate and heavy drinkers from COGA. The sample was ethnically diverse: 73.7% Caucasian; 16.8% African American; 6.3% Hispanic.
Results: For each scale, item- total correlations were computed. Items not contributing to the total score correlation were removed. Chronbach Alpha’s ranged from .78-.82. Each scale score was compared to the presence of a DSM-IV alcohol dependence diagnosis, the number of lifetime DSM alcohol symptoms endorsed, maximum number of drinks consumed (lifetime), frequency of externalizing symptoms endorsed, the Subjective Response to Alcohol scale (Schuckit), lifetime depressive symptoms, and persistence of alcohol dependence. Moderate correlations (.20-.40) were found for each phenotype with each of these other features commonly associated with alcohol dependence. Heritabilities were for each phenoytype and were in the expected range, ~.30.
Conclusion: Four phenotypes were developed for use in genetic and other analyses of alcohol dependence. These four phenotypes capture aspects of the disorder not included in the current DSM-IV and DSM% diagnostic criteria. Likely these phenotypes are useful for identifying subsets of genes related to severity of the disorder and may also be useful for examining the course of alcohol use disorders.