Innumeracy in the NY Times

2 minute read

From an otherwise-horrifying story about the plight of albinos in East Africa:

Ideally, they would like guarded camps, like one Burundi has started, where albinos can take refuge. But because Tanzania has an estimated 170,000 albinos, it would be a huge undertaking. Albinism is common among East Africans; 1 birth in 3,000 is albino, versus 1 in 20,000 in the United States.

So…that makes 510 million people in Tanzania, right? Wrong. That’s more than 10 times bigger than it ought to be.

The frequency of albinism really is high in sub-Saharan Africa. Most instances are OCA2-type albinism. Variation around OCA2 also generates blue eyes, but that’s a different allele. Albinism is a consequence of knockout mutations that stop the normal action of the gene.

Why is this type of albinism so common in Africa? That’s not entirely clear. The incidence is as high as 1/3900 births in southern Africa, but substantially less in West-Central Africa. This type of albinism is only around 1/36000 in Americans of European descent, and only 1/15000 in African-Americans. That 1/3900 frequency doesn’t sound like much, but since only homozygotes express albinism, it equates to an allele frequency of around 1.6 percent. That means that over 3 percent of people carry the allele in populations where it is most common.

Alleles that are bad and relatively common deserve some explanation. People born with this type of albinism have certain disadvantages, more severe in the cultural and physical environments of the past. These are not limited to traditional fear of albinos, but also include risks of cancer and blindness. The African variety of OCA2 albinism (partially redundant since “OCA2” stands for oculocutaneous albinism, type 2) is almost all attributable to a single deletion mutation, which occurred sometime before 3000 years ago.

There are two possible explanations. Genetic drift might have elevated the frequency of this allele as early Bantu agriculturalists dispersed from West-Central Africa into East and South Africa. In that scenario, there’s nothing special about albinism; it’s just a marker that happened to surf up to a high frequency along with population movement.

Or, heterozygotes who carry the allele might have some fitness advantage. The nature of this advantage isn’t obvious, but OCA2 knockouts do commonly occur in other species, hinting that the gene might have pleiotropic effects on other phenotypes. Even the blue-eyed allele of OCA2-HERC2 seems unlikely to have been selected for its effect on eye color, considering that effect is usually recessive.

It’s a case where the common error – that each phenotype is “caused” by a single gene, and vice versa – may easily lead people astray.