There is increasing concern that the non-random sampling of participants into major cohort studies may lead to spurious GWAS and Mendelian randomisation results.
The UK Biobank, ALSPAC and the 1958 cohort sampled individuals in different ways. The UK Biobank invited around 9 million individuals to attend a clinic, and around 0.5 million responded. These individuals are not a random sample of those invited to take part, for example they are more educated. Whereas ALSPAC sampled all mothers who gave birth within the Avon area. The 1958 cohort sampled all people born within a specific week in 1958.

This project will investigate whether there are any systematic genetic differences between the participants of the three studies.