Aims: The aim of this project is to comprehensively assess the properties of genomic variation in light of their use for Mendelian Randomisation. It will follow on from research conducted by Davey-Smith et al (2007, PLoS Med) using Genome-wide data that was outside the scope of the paper when it was published. For more information see 'Justification' section.

Hypotheses: We hypothesise that clinic variables taken at any one age will show high levels of correlation (greater than would be expected by chance). In contrast, one would not expect the same inter-correlations to exist either within a collection of genotypic data or across genetic variation and randomly selected phenotype data.

Exposure, outcome and confounding variables: This is not an association study for risk, per se, rather an assessment of the properties of genetic and phenotypic data. As such there will be no outcomes and exposures.

Data: The data that is required for this study will be all that is contained within a "standard" ALSPAC clinic assessment at around age 7 year along with genome-wide common variation for individuals found within that clinic.