Related Experiment Videos
Suggestive linkage on chromosome 1 for a quantitative alcohol-related phenotype
Danielle M Dick1, John Nurnberger, Howard J Edenberg
1Indiana University School of medicine, Indianapolis, Indiana 46202-525, USA.
Background:
Alcohol dependence is a clinically and etiologically heterogeneous disorder. Accordingly, a variety of subtypes of alcohol-dependent individuals have been proposed, and multiple operational definitions of alcohol use, abuse, and dependence have been used in linkage analyses directed toward detecting genes involved in alcohol use and problems. Here, we develop quantitative phenotypes that characterize drinking patterns among both alcoholic and nonalcoholic subjects, and use these phenotypes in subsequent linkage analyses.
Methods:
More than 9000 individuals from alcoholic and control families were administered a semistructured interview and personality questionnaire as part of the initial stage of the Collaborative Study on the Genetics of Alcoholism (COGA). A principal component analysis was conducted on items that captured many of the dimensions of drinking and related behaviors, including aspects of alcohol use, antisocial behavior and affective disturbance when drinking, and personality. Factor scores were computed for all individuals. Nonparametric linkage analyses were conducted on these factor scores, in the initial COGA sample consisting of 987 individuals from 105 extended families, and in a replication sample consisting of 1295 individuals from 157 extended families.
Results:
Three factors were identified, accounting for 68% of the total variance. The most promising regions of linkage appeared for factor 2, on which higher scores indicate a later age of onset of regular drinking and higher harm avoidance. Chromosome 1 yielded consistent evidence of linkage in both samples, with a maximum lod score of 3.3 when the samples were combined for analysis. Consistent suggestion of linkage also was found to chromosome 15.
Conclusions:
Developing novel phenotypes that more accurately model the effect of influential genes may help efforts to detect genes involved in complex disorders. Applying principal component analysis in the COGA sample provided support for some regions of linkage previously reported in COGA, and identified other new, promising regions of linkage.