Existing twin registries or combinations of them are inadequate to conduct large, in-depth studies with sufficient statistical power, especially if potential confounding or the disease prevalence dictates stratification based on gender, ethnicity, age or other factors.