Direction of Causation

Abstract

The past 15 years have seen a broadening of the focus of studies in behavioral genetics and genetic epidemiology. Rather than concentrating exclusively on a single behavioral domain (e.g., IQ or personality or psychiatric disorder or a behavioral risk-factor for major chronic disease), researchers have attempted to assess numerous behavioral domains and environmental risk-factors in a single study. There has been a corresponding growth of interest in identifying ‘intervening’ or ‘mediating’ variables in the causal pathway pathway from genotype and environment to outcome. In alcoholism research, for example, the hypothesis has been advanced (e.g., Tartar et al., 1985; Cloninger, 1987) that the genetic influence on risk of alcoholism (Heath et al., 1990) is in part explained by the inheritance of temperamental or personality factors which in turn influence an individual’s risk of becoming alcoholic. The growing interest in genotype-environment correlation, particularly the extent to which an individual’s genotype shapes the environment to which she or he is exposed (Eaves et al., 1977; Plomin et al., 1977; Scarr and McCartney, 1983; Plomin and Bergeman, 1991), has led to the formulation of competing hypotheses about how inherited differences in personality, temperament, or other behavioral variables might influence the environment to which an individual exposes herself (e.g., adverse life-events). These environmental agents in turn may increase the risk of psychiatric disorder (e.g., depression: Kendler et al., 1992d). In such cases we need to go beyond merely demonstrating that an outcome variable is influenced by genetic factors, in tests of univariate genetic models (Chapter 8), or merely demonstrating that this variable loads on the same genetic factor(s) as measures of temperament and personality or measures of environmental exposure, in a multivariate genetic analysis (Chapter 12).