Youth conduct problems are a serious public health concern, but little progress has been made in understanding their causes. This proposal addresses the genetic epidemiology of youth conduct problems to advance knowledge of their causes. We propose to analyze data from the National Longitudinal Study of Youth, which is a nationally-representative sample of 14-21 year olds first interviewed in 1979 (NLSY79). The NLSY79 includes 4,926 females who are now 37-45 years old. Since 1986, assessments have been conducted of all offspring of these women (Children of NLSY79 Study). Longitudinal data are now available on child conduct problems from ages 4 through 11 years for 5,122 offspring and on delinquent behavior through adolescence for 2,590 offspring. Because the participants are both vertically related (mother and offspring) and horizontally related (full siblings, half-siblings, etc.), their kinship links allow us to use behavior genetic methods to study genetic and environmental influences on youth conduct problems. We will test hypotheses derived from a causal model of the origins of conduct problems that integrates hypotheses from social learning theory, developmental theory, and behavior genetics. Our model shares T. E. Moffitt's assertion that the causes of adolescent delinquency are different for youth who follow different developmental pathways. We will provide the first strong test of this fundamental hypothesis by determining if genetic and environmental influences on adolescent delinquency vary as a function of the youth's level of conduct problems during childhood. We will examine sex differences in the development of conduct problems and test for possible sex differences in causal influences. Behavior genetic analyses typical treat environmental influences as anonymous, but we will estimate the extent to which a number of maternal, family, and community variables, which are important to both scientific theory and public policy, are causally related to youth conduct problems through genetic and environmental correlations. These include maternal age at first birth, maternal smoking during pregnancy, family socioeconomic status, neighborhood socioeconomic status, neighborhood social disorganization, and urbanicity.

R01MH070025

2004-01-01

2008-12-31