Uselessness of Blood Groups

August 5, 2005

Certain people continue to draw conclusions about ancestry from frequencies of blood groups and other blood-related genes, a common practice decades ago when the field of population genetics was still in its infancy. However, David Goldstein, Professor of Genetics at Duke University, stated in a recent interview that "blood groups are not now considered a good marker for population relationships, and they provide very little information about individual ancestry."

Here are four blood markers once used to measure African admixture in Southern Europeans, and the specific reasons why they aren't reliable for this purpose:

HbS (sickle cell)

African admixture in Sicily has been long suspected because of the presence of the sickle gene. Nevertheless, the degree of African admixture cannot be derived from the study of HbS frequency, since this gene was most likely expanded by the selective pressure of malaria, for a long time endemic to the region. We have examined 142 individuals from the Sicilian town of Butera (12% sickle trait) to search for other markers of the globin gene cluster less likely to be selected for by malaria. The TaqI polymorphism in the intervening sequences between the two gamma genes is informative. We have found only two instances of this African marker (TaqI(-)) among 267 normal chromosomes, demonstrating that the admixture occurred at a much lower level than previously thought. [Ragusa et al. 1992]

Fy(a-b-) (Duffy-null)

The Duffy system is also a single locus with two antigens, Fy a and Fy b. The only rare phenotype is Fy(a-b-), which has a higher frequency in countries where there is a high incidence of Plasmodium falciparium malaria. This phenotype gives a degree of immunity to the disease because the malarial parasite requires Duffy antigens to enter the red cells. Duffy antibodies are almost exclusively IgG. This system is named after the family of the antibody producer, Duffy. []

GM and KM allotypes

Fulani and Masaleit, sympatric tribes in eastern Sudan, are characterized by marked differences in susceptibility to Plasmodium falciparum malaria. To determine whether the two tribes differ in the frequency of immunoglobulin GM/KM allotypes, which are associated with immunity to several pathogens, serum samples from 50 Fulani and 50 age- and sex-matched Masaleit subjects were allotyped for several GM/KM determinants. The distribution of GM phenotypes as a whole, as well as a particular combination of KM and GM phenotypes, differed significantly between the two tribes (P = 0.03). These data suggest that GM allotypes may contribute to the genetic aetiology of malaria. [Pandey et al. 2007]

cDe (Rhesus)

But note that the cDe gene is nevertheless present in a fraction of the gene pool almost everywhere in the world. [...] Does it follow that once upon a time everybody in sub-Saharan Africa was homozygous for cDe, and in the rest of the world nobody had this gene? Should Europeans, Asiatics, and Americans who carry the cDe gene be presumed to have some Negro ancestry? There is no basis whatsoever to think so. The gene cDe is almost cosmopolitan in distribution, though for some unknown reason it reaches its highest frequency in Africa. [Dobzhansky, 1962]

Frequencies of the cDe gene in various populations (after Mourant 1954):


Spaniards ..... 3.7%
English ....... 2.8%
Germans ....... 2.6%
Danes ......... 1.8%
Italians ...... 1.6%
Basques ....... 0.5%


Bushmen ...... 89.0%
Hutu ......... 62.9%
Shona ........ 62.7%
S.A. Bantu ... 59.6%
Kikuyu ....... 59.5%
Egyptians .... 17.3%


S. Chinese .... 4.1%
E. Pakistan ... 3.9%


Aborigines .... 8.5%
Javanese ...... 6.5%
Papuans ....... 2.0%
Marshallese ... 0.5%


Navajo ....... 28.0%
Eskimos ....... 2.1%

Updated 11/03/2009