Evidence of heterogeneity of genetic effect in Iowa adoption studies

This paper describes male adoption studies at the University of Iowa using private and public adoption agencies within the state of Iowa from 1974 to the present time. This research involves four large studies, the first two of which demonstrated significant genetic as well as environmental effects in the etiology of alcoholism as well as significant correlations between adoptee conditions of antisocial personality and alcohol abuse. Findings in the first two studies were similar. However, the third study, using a sample from Catholic-sponsored adoption agencies across the state, failed to show a genetic effect. The final and fourth study was designed in part to look for heterogeneity in the manifestation of a genetic factor from one sample to another in the Iowa studies. This was done by sampling from two agencies, one of which had shown a genetic effect and the other that had not. In this fourth study, the agency that had not shown a genetic effect in the previous study also failed to show an effect, but the agency that had shown a significant genetic effect in the past did demonstrate a significant genetic effect. There were two remaining agencies in the fourth study for which no comparison with past samples could be made. One of these agencies showed a marginally significant genetic effect and the other showed no effect. Statistical analysis of this last study suggested that observed variability in the odds ratio from sample to sample was due to differences in manifestation of the genetic effect.