Gene-Environment Interaction Effects in Neurodevelopment

Cognitive development is known to be highly heritable, but various genome wide association studies (GWAS) have failed to find common single nucleotide polymorphisms (SNPs) which can explain a large proportion of the variance in cognitive ability. We hypothesized that an interaction between genes and exposure to potentially toxic metals may account for some proportion of this missing heritability. We genotyped 749 infants in Mexico and 636 infants in Bangladesh. Criteria for inclusion into the study are given below. We then performed a Genome Wide Environmental Interaction Study (GWEIS) with neurodevelopment outcomes. There were 552,487 SNPs in common between the two cohorts. ]]>
In Bangladesh, pregnant women were recruited between 2008 and 2011 from rural health clinics in the Sirajdikhan and Pabna Sadar Upazilas of Bangladesh. These women were 18 years or older, planned to receive prenatal care through the Dhaka Community Hospital, and intended to live in the same house through the pregnancy. Each of their infants was tested on the Bayley Scales of Infant Development at approximately 2 years of age. A Mental Composite Score was calculated as the sum of cognition, expressive language, and receptive language scores. The objective of this analysis was to associate individual single nucleotide polymorphisms (SNPs) with the Mental Composite Score, which is a continuous variable. Genotyping was performed at the Broad Institute using the OmniExpressExome-8 BeadChip from Illumina (San Diego, CA). The covariates we adjusted for were gestational age, infant gender, maternal education level, child's age at time of assessment, household smoking, and the first two principal components.In Mexico, participants recruited between 2007 and 2011 were healthy pregnant women, at least 18 years of age, received healthcare through the Mexican social security system, and lived within the Mexico City metropolitan area. They also had access to a telephone and planned to continue living in the city for the next three years. Each of their infants was tested on the Bayley Scales of Infant Development at approximately 2 years of age. A Mental Composite Score was calculated as the sum of cognition, expressive language, and receptive language scores. The objective of this analysis was to associate individual single nucleotide polymorphisms with the Mental Composite Score, which is a continuous variable. Genotyping was performed at the Broad Institute using the HumanOmni1-Quad BeadChip from Illumina (San Diego, CA). The covariates we adjusted for were gestational age, infant gender, maternal education level, child's age at time of assessment, household smoking, and the first two principal components.In the Mexico cohort, a total of 1054 pregnant women initially consented to be enrolled in the study. Of this sample, we excluded subjects for high levels of relatedness, high levels of missingness, and outlying principal components. In all, 749 infants were genotyped and passed quality control. In the Bangladesh cohort, a total of 1458 pregnant women initially consented to be enrolled in the study. These women were at least 18 years old and planned to deliver under Dhaka Community Hospital Trust supervision. We again excluded subjects for high levels of relatedness, high levels of missingness, and outlying principal components. In all, 636 infants were genotyped and passed quality control.]]>