ABO genes are differentially distributed in socio-economic groups in England.

No direct evidence is available concerning what average genetic differences, if any, characterize the segments of socially stratified human populations, although theoretical considerations suggest that genetic differentiation within such populations is to be expected. We have now analysed two large samples of data from blood donors in England to test whether the distributions of the ABO and Rhesus blood group phenotypes are random with respect to socio-economic groups as determined by occupational classification. We have found that in both native and migrant sections of the populations of two widely separated regions (south-west England and part of Yorkshire) and in both sexes, the A phenotype is highly significantly more, and the O phenotype significantly less, frequent than expected in social classes I and II, while the converse is seen in social classes III-V. An individual of the A phenotype has thus about 15% greater probability than chance would dictate of being placed in classes I and II. The distribution of the Rh+ and Rh- phenotypes does not differ significantly between classes. It seems unlikely that this nonrandom distribution of the ABO phenotypes among socio-economic groups results from sampling, historical or migrational effects and we conclude that the observed association is likely to result from pleiotropic effects of the ABO alleles (or genes closely linked to them) on attributes influencing occupational type, social mobility and social class.