Heritability estimation is possible from concordant twin pairs alone, based on the proportion of like-sexed pairs among all concordant affected pairs. The method is limited to conditions found in both sexes in the prevalence range 0.1% to 10%, and a relatively large population size is required to give an adequate sample of twin pairs. However, the method has the considerable advantage that zygosity determination is not required and that any bias due to incomplete diagnosis/ascertainment is likely to be small. The method is particularly suited to diseases where registration is obligatory and computerised so that the register can be scanned for pairs of individuals with the same date of birth, place of birth and birth surname.
