Tuesday, June 17, 2008

More genetic variation within than between?

Research in human genetics has highlighted that there is more genetic variation within than between human groups, where those groups are defined in terms of linguistic, geographic, and cultural boundaries.

Statement 2. Guiding principles on using racial categories in human genetics (Soo-Jin Lee et al., 2008)

The above statement is part of an open letter to human geneticists, recently published in Genome Biology. The authors are a number of Stanford University scholars from the humanities, social sciences, life sciences, law and medicine.

Yes, the statement is true. It has been confirmed by a wide range of studies, the most well known being by Richard Lewontin (1972). At most human genes, there is far more variation within human populations than between them. Lewontin concluded that 85% of human genetic variation exists only between individuals and cannot be apportioned on a population basis. In this, he was repeating a conclusion that others had made before him, notably Frank Livingstone (1962), and that others still have since confirmed. Clearly, racial categories seem to be very hazy entities—so hazy, in fact, as to seem useless.

Or maybe not. It turns out that the same haziness exists between many species. When we compare pairs of species that are closely related but anatomically distinct, we often see more genetic variation within each species than between them, so much so, that we cannot reliably assign individuals to either by genetic criteria alone (see older post).

How come? When two populations diverge and eventually become separate species, they usually do so because they have adapted to different environments with different selection pressures. These selection pressures, however, do not act on the entire genome. In fact, they act only on a small set of genes—the ones that need to function differently in either environment.

We see the same thing with artificial selection. Kennel clubs maintain distinct dog breeds by insisting that dogs in any one breed meet various criteria. This distinctness disappears, however, when we look at their genomes.

… genetic and biochemical methods … have shown domestic dogs to be virtually identical in many respects to other members of the genus. … Greater mtDNA differences appeared within the single breeds of Doberman pinscher or poodle than between dogs and wolves. Eighteen breeds, which included dachshunds, dingoes, and Great Danes, shared a common haplotype and were no closer to wolves than poodles and bulldogs. These data make wolves resemble another breed of dog.

… there is less mtDNA difference between dogs, wolves, and coyotes than there is between the various ethnic groups of human beings, which are recognized as a single species.
(Coppinger & Schneider, 1995)

Artificial selection uses a smaller set of criteria than does natural selection. Nonetheless, both forms of selection act on only a small fraction of the genome, partly because most genes are of marginal selective value and partly because many genes can function identically in a wide variety of species.

Of course, even at these other genes two populations may start to show diverging patterns of variation once they have become reproductive isolated species. But this is a long process that occurs through each of them slowly and separately accumulating its own mutations at genes of low selective value. This is not the case with our species. We began to spread out of Africa only 50,000 or so years ago.


Coppinger, R. and R. Schneider (1995). Evolution of working dogs. In J. Serpell (ed.), The Domestic Dog: Its Evolution, Behaviour and Interactions with People. Cambridge: Cambridge University Press, pp. 21-47.

Soo-Jin Lee, S., Mountain, J., Koenig, B., Altman, R., Brown, M., Camarillo, A., Cavalli-Sforza, L., Cho, M., Eberhardt, J., Feldman, M., Ford, R.,Greely, H., King, R., Markus, H., Satz, D., Snipp, M., Steele, C., and Underhill, P. (2008). The ethics of characterizing difference: guiding principles on using racial categories in human genetics. Genome Biology, 9, 404 doi:10.1186/gb-2008-9-7-404

Lewontin, R.C. (1972). The apportionment of human diversity. Evolutionary Biology, 6,

Livingstone, F.B. (1962). On the non-existence of human races. Current Anthropology, 3, 279-281.

Women's hair color. It doesn't pay to be average

Gene expression has run a second post on blonde preference. This time the analysis is based on the front covers of Maxim magazine, whose readership is 27.5 years old on average and thus provides a better guide of what young men prefer (Playboy’s readership is 32.5 on average).

The post makes two points:

1) After increasing from the mid-1960s to about the year 2000, preference for blonde women seems to have levelled off over the past 11 years.

2) In comparison to the white American population, there is an overrepresentation of women with either blonde or dark brunette hair on the front covers of Maxim. Women with intermediate hair shades are underrepresented.

The second finding is interesting because it bears out the frequency-dependent nature of hair-color preference. Men seem to have a stronger preference for less common shades.

The Rise of the Blonde Playmate

Gene Expression has a post on changing preferences for hair color over time. The hair color of Playboy playmates was charted from 1954 to 2007, on the assumption that this monthly series reflects male sexual tastes. It turns out that the percentage of blonde playmates has risen over the years: from a low of about 35% in the mid-1960s to a high of 60% around the year 2000. A similar study was done fifteen years ago, with similar results (Rich & Cash, 1993).

These proportions are well above the actual proportion of blonds among white Americans. When Rich and Cash (1993) studied a sample of undergraduates, the proportions were 68.1% brunette, 26.8% blond, and 5.1% red. This breakdown parallels those put forward in two British studies: 68% brunette, 25% blond, 1% red, and 6% black (Takeda et al., 2006); and 74% brunette; 18% blond, and 8% red (Mather et al., unpublished). Jason Malloy says a similar breakdown appears in Beddoe (1885).

Thus, blond hair preference does not reflect the actual distribution of hair color. In fact, in light of current demographic trends, the scarcer blondes become, the more they seem to be preferred.

This frequency dependence has been shown in humans. Thelen (1983) presented male participants with slides showing attractive brunettes and blondes and asked them to choose, for each series, the woman they would most like to marry. One series had equal numbers of brunettes and blondes, a second 1 brunette for every 5 blondes, and a third 1 brunette for every 11 blondes. Result: the rarer the same brunette was in a series, the likelier the men would choose her.

Just as blondes are more strongly preferred as they become scarcer, they seem to be less popular when more abundant. Havelock Ellis (1928, pp. 182-183) noted a weaker preference for blonde women in England than in France, which he ascribed to the higher prevalence of blondness among the English.

Indeed, if this preference were not frequency-dependent, sexual selection should have steadily increased the incidence of blond hair, to the point of eliminating all other colors. Yet blonds do not form the majority of any human population. Even Swedes are no more than 40% blond, and this is using a definition that already includes a variety of shades (platinum blond, ash blond, sunny blond, sandy blond, golden blond, strawberry blond, zebra blond, dirty blond, brownish blond, see Wikipedia article).

In addition to America’s changing demographics, the rise of the blonde playmate may also reflect the spread of Playboy magazine into American Catholic communities where it was initially excluded and where blond hair is less common (many of these communities being of southern European origin). In the wake of Vatican II (1962-1965), Catholic Americans have adopted increasingly secular attitudes towards sex, and consumption of soft porn has become less stigmatized, if not tolerated.


Beddoe, J. (1885). The Races of Britain: A Contribution to the Anthropology of Western Europe, Arrowsmith, Bristol & Trübnermm, London.

Ellis, H. (1928). Studies in the Psychology of Sex. Vol. IV, "Sexual Selection in Man." Philadelphia: F.A. Davis Company.

Mather, F., Manning, J.T., & Bundred, P.E. (unpublished). 2nd to 4th digit ratio, hair and eye colour in Caucasians: Evidence for blond hair as a correlate of high prenatal oestrogen.

Rich, M.K., & Cash, T.F. (1993). The American image of beauty: Media representations of hair color for four decades. Sex Roles, 29, 113-124.

Takeda, M.B., Helms, M.M., & Romanova, N. (2006). Hair color stereotyping and CEO selection in the United Kingdom. Journal of human behavior in the social environment, 13, 85-99

Thelen, T.H. (1983). Minority type human mate preference. Social Biology, 30, 162-180.

The Big Bang of modern human evolution

The Smithsonian magazine has an interesting article on the origin of modern humans and the events leading to their rapid spread ‘out of Africa’:

Then, about 80,000 years ago, says Blombos archaeologist Henshilwood, modern humans entered a "dynamic period" of innovation. The evidence comes from such South African cave sites as Blombos, Klasies River, Diepkloof and Sibudu. In addition to the ocher carving, the Blombos Cave yielded perforated ornamental shell beads—among the world's first known jewelry. Pieces of inscribed ostrich eggshell turned up at Diepkloof. Hafted points at Sibudu and elsewhere hint that the moderns of southern Africa used throwing spears and arrows. Fine-grained stone needed for careful workmanship had been transported from up to 18 miles away, which suggests they had some sort of trade. Bones at several South African sites showed that humans were killing eland, springbok and even seals. At Klasies River, traces of burned vegetation suggest that the ancient hunter-gatherers may have figured out that by clearing land, they could encourage quicker growth of edible roots and tubers. The sophisticated bone tool and stoneworking technologies at these sites were all from roughly the same time period—between 75,000 and 55,000 years ago.

The above archaeological findings match African mtDNA data presented by Watson et al. (1997):

We find that most African mitochondrial sequences appear to be the result of demographic expansions that started ~60,000-80,000 years ago, the earliest of which led to the colonization of Eurasia. Only a minority (13%) of sequences fall outside these expansion clusters, echoing a time, before the expansions, when the human mitochondrial gene pool was possibly more diverse (in terms of mean sequence divergence) than it is today.

These population expansions appear to have been driven by cultural innovations that gave some African populations an edge over others and, eventually, over archaic humans in Europe and Asia. Unfortunately, instead of speaking of an expansion out of Africa, the Smithsonian article prefers the term ‘exodus’ — as if life had become so intolerable in Africa that humans were forced to pick up and leave. In fact, almost the opposite happened: modern humans did so well in Africa that they elbowed out their archaic rivals not only on their own continent but on others as well.

The Smithsonian article also speculates about the reasons behind their demographic success. These humans had become better at constructing mental models of the world around themselves. They could imagine the things they had to make and tinker with them in their minds. And they could share this mental tinkering with others through language.

This improvement in mental functioning may have been made possible initially by increased consumption of fatty acids in seafood. Then, as humans developed a new cultural environment and its attendant demands on brainpower, there would have been selection for genes to ‘hardwire’ this neurological change, i.e., the Baldwin effect.

Virtually all of these sites had piles of seashells. Together with the much older evidence from the cave at Pinnacle Point, the shells suggest that seafood may have served as a nutritional trigger at a crucial point in human history, providing the fatty acids that modern humans needed to fuel their outsize brains: "This is the evolutionary driving force," says University of Cape Town archaeologist John Parkington. "It is sucking people into being more cognitively aware, faster-wired, faster-brained, smarter." Stanford University paleoanthropologist Richard Klein has long argued that a genetic mutation at roughly this point in human history provoked a sudden increase in brainpower, perhaps linked to the onset of speech.


Gugliotta, G. (2008). The Great Human Migration. Why humans left their African homeland 80,000 years ago to colonize the world. Smithsonian magazine, July.

Watson, E., Forster, P., Richards, M., and Bandelt, H-J. (1997). Mitochondrial footprints of human expansions in Africa. American Journal of Human Genetics, 61, 691-704.

Maps of European hair and eye color

A reader sent me the following e-mail:

I want to ask you some questions regarding your maps.

To be honest both maps look very questionable. Also I didn't see any source for these studies, especially the one for eye color looks completely invented. I'm sure that the percentage of light eyes in Italy (for example) is higher than 1-19%, surely in Italy there are more people with light eyes than Spain, Portugal and especially than Turkey or Lebanon , as there are more people with light eyes in Southern England than Bosnia.

My suppositions are confirmed by the studies of Coon or Lundman for example:

Map by Carleton Coon

Map by Bertil Lundman

The maps on my website come from my article on European hair and eye color diversity (Frost 2006) and are reproduced from an anthropology textbook (Beals & Hoijer 1965, pp. 213-214). Beals and Hoijer, in turn, cite a textbook by another anthropologist, Frederick Hulse (1963: p. 328). Unfortunately, Hulse does not indicate the provenance of his data. I suspect he was using data from military recruits, with a lot of interpolation. Or perhaps he was using even earlier maps.

Since my 2006 article came out, I have found one such map in a work by Biasutti (1959, cap. I Nazioni europee e la loro composizione etnica p. 43), which in turn is taken from Günther (1929). At Gene Expression, Razib has turned up another map by Carleton Coon (1982), which is ascribed to ‘Elmer Rising, 1939.’

All of these maps resemble each other in their broad outlines, but the minor differences are significant. I still have not found the original data on which any of them are based.


Beals, R.L. and H. Hoijer. (1965). An Introduction to Anthropology, 3rd edition, New York: MacMillan Co.

Biasutti, R. (1959). Razze e Popoli della Terra. Torino: Unione Tipografico-Editrice

Coon, C.S. (1982). Racial Adaptations. Burnham.

Coon, C.S. (1939). The Races of Europe. New York: The Macmillan Co.

Frost, P. (2006). European hair and eye color - A case of frequency-dependent sexual selection? Evolution and Human Behavior, 27, 85-103.

Günther, H.F.K. (1929). Rassenkunde des deutschen Volkes. Munich: J.F. Lehmann

Hulse F.S. (1963). The Human Species. An Introduction to Physical Anthropology. New York: Random House.

Lundman, B.J. (1977). The Races and Peoples of Europe. New York: IAAEE

Testosterone and Greenland Inuit

Human populations vary in their incidence of polygyny. In general, the more it costs a man to provide for a second wife, the less the population will be polygynous. This cost is highest among arctic hunter-gatherers, where women depend on meat from men to feed themselves and their children. It is lower among tropical hunter-gatherers, where women need less male provisioning because they can gather fruits, vegetables, and tubers. Finally, the cost becomes negative among tropical agricultural peoples, where a woman is able to raise enough food through year-round farming to feed herself and her children, with little assistance from a male provider.

In the last group of societies, men best serve their reproductive fitness by having as many wives as possible. Result: a relative shortage of wives and intense male-male competition for the few women available. Highly polygynous societies usually deal with this problem by raising the age of marriage for men, typically by ten years past the age of puberty. Such a constraint may be imposed formally or informally. Often, a man may not marry until he has proven himself as a warrior or has attained a certain standing within the community.

This social rule has the effect of concentrating male sexual competition among young adults. Over time, the resulting selection pressure may have led to the higher levels of blood testosterone that we see in tropical agricultural peoples of sub-Saharan Africa and Papua-New Guinea, where 20 to 50% of all marriages are polygynous (see previous post). In particular, it may explain why this hormonal advantage seems to be limited to young men. Testosterone levels are higher in black boys than in white boys as early as 5 to 9 years of age (Abdelrahaman et al., 2005). In black males, these levels peak during adolescence and early adulthood (Ross et al., 1986; Winters et al., 2001). The black-white difference then shrinks after 24 years of age and is gone by the early 30s (Gapstur et al., 2002). It actually seems to reverse in later years (Nyborg, 1994, p. 111-113).

Broadly speaking, lifetime exposure to testosterone correlates with the incidence of prostate cancer. The highest incidences in the world are among African American men (Brawley and Kramer, 1996). It was once thought that lower incidences prevail in other populations of black African descent (i.e., West Indians and sub-Saharan Africans), but this difference has since been shown to reflect underreporting (Glover et al., 1998; Ogunbiyi and Shittu, 1999; Osegbe, 1997). Conversely, the lowest incidences are among East Asians (Ross et al., 1992).

Testosterone has a wide range of physiological, morphological, and behavioral effects that in one way or another improve male reproductive success, particularly under conditions of intense sexual competition. These effects may be achieved not only by raising the concentration of testosterone in the bloodstream, but also by making androgen receptors more effective or by converting testosterone into the physiologically more active DHT (5α-dihydrotestosterone). Because such changes multiply the actual impact of testosterone on the human organism, we may underestimate this impact by looking only at blood testosterone levels. For instance, East Asians have the lowest incidences of prostrate cancer, yet their blood testosterone levels are intermediate between those of white and black Americans. They do, however, have less 5α-reductase than either white or black Americans, this being the enzyme that converts testosterone into DHT (Ross et al., 1992).

There is other evidence that human populations vary not only in blood testosterone levels, but also in testosterone/DHT conversion and in androgen receptor activity. It has been shown that androgen receptors seem to be more numerous and more receptive in African-American subjects (Kittles et al., 2001). Now, another study has looked at the other extreme of human variation. Giwercman et al. (2007) have found that Inuit Greenlanders have very low levels of prostate cancer because they have fewer alleles of the sort that increase androgen receptor activity or facilitate testosterone to DHT conversion.

The difference was striking, even in comparison with Swedish subjects—whose incidence of prostrate cancer falls within the world average. The authors concluded:

Our results suggest that Greenlanders are genetically predisposed to a lower activity in testosterone to 5α-dihydrotestosterone turnover and to lower AR activity, which, at least partly, could explain their low incidence of prostate cancer.


Abdelrahaman, E., Raghavan, S., Baker, L., Weinrich, M., and Winters, S.J. (2005). Racial difference in circulating sex hormone-binding globulin levels in prepubertal boys. Metabolism, 54, 91-96.

Brawley, O.W. and Kramer B.S. (1996). Epidemiology of prostate cancer. In Volgelsang, N.J., Scardino, P.T., Shipley, W.U., and Coffey, D.S. (eds). Comprehensive textbook of genitourinary oncology. Baltimore: Williams and Wilkins,

Gapstur, S.M., Gann, P.H., Kopp, P., Colangelo, L., Longcope, C., and Liu, K. (2002). Serum androgen concentrations in young men: A longitudinal analysis of associations with age, obesity, and race. The CARDIA male hormone study. Cancer Epidemiology, Biomarkers & Prevention, 11, 1041-1047.

Giwercman, C., A. Giwercman, H.S. Pedersen, G. Toft, K. Lundin, J-P. Bonde, and Y.L. Giwercman. (2007). Polymorphisms in genes regulating androgen activity among prostate cancer low-risk Inuit men and high-risk Scandinavians. International Journal of Andrology, 31, 25-30.

Glover, F., Coffey, D., et al. (1998). The epidemiology of prostate cancer in Jamaica. Journal of Urology, 159, 1984-1987.

Kittles, R.A., Young, D., Weinrich, S., Hudson, J., Argyropoulos, G., Ukoli, F., Adams-Campbell, L., and Dunston, G.M. (2001). Extent of linkage disequilibrium between the androgen receptor gene CAG and GGC repeats in human populations: implications for prostate cancer risk. Human Genetics, 109, 253-261.

Nyborg, H. (1994). Hormones, Sex, and Society. The Science of Physiology. Westport (Conn.): Praeger.

Ogunbiyi, J. and Shittu, O. (1999). Increased incidence of prostate cancer in Nigerians. Journal of the National Medical Association, 3, 159-164.Osegbe, D. (1997). Prostate cancer in Nigerians: facts and non-facts. Journal of Urology, 157, 1340.

Ross, R.K., Bernstein, L., Lobo, R.A., Shimizu, H., Stanczyk, F.Z., Pike, M.C., and Henderson, B.E. (1992). 5-apha-reductase activity and risk of prostate cancer among Japanese and US white and black males. Lancet, 339, 887-889.

Ross, R., Bernstein, L., Judd, H., Hanisch, R., Pike, M., and Henderson, B. (1986). Serum testosterone levels in healthy young black and white men. Journal of the National Cancer Institute, 76, 45-48.

Winters, S.J., Brufsky, A., Weissfeld, J., Trump, D.L., Dyky, M.A., and Hadeed, V. (2001). Testosterone, sex hormone-binding globulin, and body composition in young adult African American and Caucasian men. Metabolism, 50, 1242-1247.