Aim.

1. To develop and apply "GCTA" to genome-wide SNP data on mothers and children to resolve and estimate the contributions of the offspring and maternal genotype to dyadic outcomes, including gestational age and birthweight. This will require extending GCTA methodology to condition the genetic correlations between mothers on those among their offsping. The result of this approach will be compared with our own published findings already obtained by biometrical-genetic modeling of data based on extended kinships of twins and other informative relationships from the Virginia 30,000 study, the Virginia Birth Record Study ("VABRS"), The Virgini Children of Twins Study ("VACOTS"), and the Virginia Twin Study of Adolescent Behavioral Development ("VTSABD"). 2. Explore effects of genome-wide mother-child genetic incompatiblity on outcomes of pregnancy.

Background.

Biometrical-genetic studies of twin kinships, half-sibships and cousinships (VABRS, PI TP York) have demonstrated that differences gestational age is the effect of differences in both fetal genotype and maternal genotype. VACOTS (PI JL Silberg) has shown how the maternal genotype for antisocial behavior influences offspring depression. VTSABD (PI LJ Eaves) demonstrated that genetic influences on maternal anti-social personality exacerbated childrens' anti-social behavior through their exposure to environmenal adversity.

We propose to extend the GCTA method for using genome-wide SNP data to resolve these effects with unrelated subjects from ALSPAC. We will also attempt a preliminary resolution of genes that contribute specifically and separately to maternal and fetal influences on these dyadic outcomes. The results will be evaluated in parellel with new biometrical genetic analyses of the VBRS data (greater than 1,000,000 births).

Hypotheses.

The principal goal of this project is methodological and demonstrative. However, it is guided by the need to exploit genome-wide polymorphism data to test hypotheses about the underlying causal relationships between maternal and offspring genotype and multiple outcomes in human development. The two systems selected for initial study (gestational age and birth-weight) are models that illustrate the methodological and substantive issues. We will also apply the approach to birth-length and height at age 7 as a model system that may reflect the developmental amelioration of the environmental impact of the maternal environment. We have shown (above) that both maternal and fetal genotypes affect gestational age. Further details of the proposed model and approach are given in the attached document.

The method can subsequently be used to resolve the effects of maternal and offspring genotype on other developmental outcomes where it is expected parental characteristics and behavior influence the phenotypes of their children.

Variables

Exposure variables: 1. Genome wide genetic relatedness of mothers and offspring derived from identity by state indexed by genome-wide SNP data in mothers and children (see Yang et al., 2011).

Outcome variables: a) Birthweight and gestational age; b) Length at birth and height at 7 years. If the method works, we hope to extend the approach to embrace the environmental effect of genetic differences in maternal psychopathology on offspring outcome (not included in current variable request).

Confounding variables: Maternal age at parturition; Maternal and paternal smoking in pregnancy; parity; maternal BMI.