Sunday, October 22, 2006

Leprosy and Scandinavian babes   posted by agnostic @ 10/22/2006 10:10:00 PM

Speaking of the ubiquity of recent selection, I'm in the middle of looking at the recent selection of various traits that have to do with sexiness in different parts of the world. The primary selection pressure concerning physical attactiveness is, of course, pathogen prevalence -- what stronger force could there be for making us care about how hot a person is? Since the sexiest birds evolved in the tropics -- areas of high pathogen prevalence -- there is no reason to think that humans would turn out any differently, to a first approximation. Hamilton & Zuk (1982) proposed that ornateness was a signal of health in more germ-ridden species, and that mates would choose the more ornate individuals to give their offspring a leg up in chasing the moving target of evading pathogens. (They looked at variation in pathogen prevalence in North American birds, not those from the tropics, but it's clear which species are more decorated.)

Continuing in that vein, Gangestad & Buss (1993) showed that the degree to which individuals (both male and female) emphasized "good looks" in a potential mate correlated highly with local pathogen prevalence. The correlation is not caused by latitude or climate per se but truly by the pathogen prevalence -- the raw correlation is +0.38 (p less than .05), but when controlling for the possible confounds of latitude, geographical region (e.g., Africa, Europe, etc.), and average income, the correlation becomes +0.72 (p less than .001). If the effect were merely due to differences in latitude, continent, etc., then the correlation should decrease in magnitude, whereas it nearly doubles. The obvious corrollary, which Gangestad & Buss don't mention, is that populations adapted to historically more pathogen-stricken areas will on average be more physically attractive, since both sexes emphasize (and thus, probably choose based on) good looks in a partner. To my knowledge, no one has showed that yet, but I plan to shortly by analyzing data on how different countries have faired in the Miss Universe contest.

However, Scandinavia would appear to be a glaring exception -- though pathogen pressures have been less severe than in the Mediterranean, they fair well in beauty contests and are thought of by non-Scandinavian whites (and maybe other groups) as more beautiful than other Northern European groups. (See the Appendix for the Miss Universe-derived scores of Northern European countries.) And yet, there is an exception to this exception -- Scandinavia was the one holdout in Europe for leprosy, which continued pretty much up to the start of the 20th C. From William McNeill's Plagues and Peoples (p.185):

Hansen's disease appears to have established itself in Europe and the Mediterranean coastlands in the sixth century A.D. Thereafter, together with other infections classified as leprous, it remained of major importance until the fourteenth century. Leprosaria were established outside thousands of medieval towns. By the thirteenth century one estimate puts their number in all of Christendom at 19,000.

And though leprosy faded in most of Europe in the 14th C (p.186),

Hansen's disease did continue to exist, on a significant scale, in Scandinavia and more sparsely in other parts of Europe as well.

The Encyclopaedia Britannica article on leprosy agrees that leprosy continued in Scandinavia after it had begun to subside elsewhere in Europe during the 14th C. McNeill guesses that it could have continued due to the practice of huddling close together every night to keep each other warm, since leprosy requires prolonged contact with an infected person to be transmitted to the next host. It is no accident, then, that the physician who discovered the microbial cause of leprosy, Armauer Hansen, was a Norwegian who worked in the historically largest city of Norway -- Bergen -- since it is difficult to conduct painstaking research where there are few subjects to investigate. So, though low overall in pathogen prevalence, Scandinavia had a much higher prevalence of leprosy from Medieval times almost right up to the present.

Of course, leprosy is a disfiguring skin disease, one that would have made a sufferer a virtual Darwinian dead-end, as no one would want to mate with them. Scandinavian males today, as shown by Buss' (1989) [pdf] cross-cultural survey, do not care much about whether or not a potential partner is good-looking (and of course, neither do the females), in line with the historically low pathogen prevalence of their environment. However, no matter how willing Scandinavians may be to overlook sub-perfect breast development, graying hair, and so on, damn few would be so openminded as to willingly mate with a leper. Our modern sensibilities are offended by the sight of mere genital warts -- imagine viewing a nude person with festering sores encrusting a fair fraction of their body. Talk about a selection pressure! Especially when this pressure lasted for at least 700 to 1000 years, or from 28 to 50 generations depending on exactly how long and whether a generation equals 20 or 25 years; and particularly in a population that experienced little inward gene-flow.

The only study to attempt to quantify the narrow-sense heritability (or h^2) of good looks, using a twin design, is McGovern et al (1996), which showed a DZ concordance rate of 0.33 and a MZ concordance rate of 0.65. The broad-sense heritability = 2(MZ - DZ) or 0.64, and since the MZ rate is almost exactly twice the DZ rate, we can assume that most of this genetic component is additive, so let h^2 = 0.64. The Breeder's Equation

says that the response to selection R, equals the selection differential S (or how high the parents' mean is above the population mean) multiplied by h^2. If leprosy was a strong enough pressure so that the parents' mean would have been at just the 52.2 percentile, then S = 0.0558 SD; this multiplied by h^2 of 0.64 = 0.035712 SD increase in attractiveness per generation. Multiply that by 28 generations of selection, and the overall increase is 1 SD, comparable to a gain of 15 IQ points or 3 to 4 inches of height. If the selection lasted for the rough upper-bound of 50 generations, then the parents in each generation would only had to have placed in the 51.2 percentile -- barely above-average in either case.

Some caveats: this assumes minimal inward gene-flow from populations not also subject to leprosy pressures, which would have blunted the response to selection. However, inward gene-flow from other Northern European countries wouldn't have mattered much up to about 1400, since other regions were also plagued by leprosy. From 1400 to 1900, I can't find estimates of gene-flow into Scandinavia (though I'm sure they're in the literature), but it must have been very low, as this region is one of the most homogenous. The pressure probably relaxed around 1900, after the infectious nature had been discovered and medical intervention began wiping out the disease. Still, that's only 4 or 5 generations ago, hardly enough time to undo 28 to 50 generations of directional selection, and surely good looks were not selected against during the 20th C. Last, it would be desirable to uncover independent confirmation that Scandinavians were more affected by leprosy pressures than other Northern Europeans. There are genes in the HLA that are implicated in susceptibility to leprosy, but I couldn't find allele frequency distribution data on ALFRED. By hypothesis, Scandinavians would show greater frequency of such alleles compared to other Northern Europeans (at least the populations included in the chart below).

Finally, we don't have a convenient metric of attractiveness, let alone data on the difference in means across various populations, but this simple calculation shows that it's entirely feasible for Scandinavians to have risen 1 SD above their previous Northern European mean. Of course, populations outside Northern Europe may have already undergone substantial selection for attractiveness (also due to pathogen pressure) before leprosy hit the scene in Northern Europe, and continued to experience such selection up to 1900 and maybe beyond. So, the claim is not that selection has made Scandinavians the most attractive population on Earth, but at the very least the most attractive of Northern Europe.

Appendix: While admitting that attractiveness will always remain somewhat subjective and more prone to errors of measurement than gauging height, there are universals in what features are considered attractive. Citations would be irrelevant (though you can look up Buss, Thornhill, Singh, and others if you're skeptical). Hair-splitting aside, it's not as if maculated skin, beady eyes, or a waist that's pronouncedly wider than the hips are considered attractive by the world's cultures. The Miss Universe pageant is one of the few arenas where individuals representing the world's populations are subject to the same criteria, and where the scores reflect the preferences of the same target audience. True, there could be a non-trivial chance component since only one individual serves as a delegate for their entire population, not to mention the whims of the judges. Before the Civil Rights and national liberation movements of the 1960s, there may also have been a bias toward those of European descent (Miss Universe began in 1955). Nevertheless, the latter does not factor in here, as all countries are Northern European. And to remove as much of the chance component as possible, only countries who had entered 20 Miss Universe pageants were considered -- 1 win out of 4 contests might represent a fluke, but Sweden's 3 wins and 28 semi-finalist placings out of 54 contests cannot be a fluke.

The hotness score H was determined as follows:

Where W = number of wins, S = number of semi-finalist placings (S is multiplied by 0.1 to make sure semi-finalist status did not trump winner status), C = number of contests entered, and P = size of country's population in millions as of the past three years. W, S, and C were culled from a website of Miss Universe statistics, and the population estimates were gotten from the country's Wikipedia entry. The numbers inside the parentheses represent the country's raw placing score per contest entered per unit of population, and since it varies by orders of magnitude, the (natural) log was taken, and this multiplied by -1 to remove the negative sign (for ease of inspection). This simplifies to the second line of the graphic. Lower H scores represent greater hotness -- akin to 1st place being the greatest and 10th place the worst in a 10-contestant race. Below is a chart of Northern European countries ranked in ascending order of H scores (right column) -- that is, in descending order of hotness. Again, don't take the ranking too seriously, but it's clear nonetheless that the Scandinavian countries really are, in some sense, hotter than the other Northern European countries.

Country W _ S _ .1S _ W+.1S _ C _ P _ H

Iceland 0 _ 4 _ .4 _ .4 _ 40 _ 0.3 _ 3.401
Finland 2 _ 18 _ 1.8 _ 3.8 _ 51 _ 5.3 _ 4.265
Sweden 3 _ 28 _ 2.8 _ 5.8 _ 54 _ 9.1 _ 4.439
Norway 1 _ 17 _ 1.7 _ 2.7 _ 49 _ 4.7 _ 4.446
Wales 0 _ 5 _ .5 _ .5 _ 30 _ 3 _ 5.193
Scotland 0 _ 5 _ .5 _ .5 _ 29 _ 5.1 _ 5.690
Denmark 0 _ 7 _ .7 _ .7 _ 41 _ 5.4 _ 5.757
Switzerland 0 _ 10 _ 1.0 _ 1.0 _ 46 _ 7.3 _ 5.817
Netherlands 1 _ 11 _ 1.1 _ 2.1 _ 48 _ 16.3 _ 5.920
Ireland 0 _ 7 _ .7 _ .7 _ 46 _ 6 _ 5.977
Austria 0 _ 7 _ .7 _ .7 _ 38 _ 8.3 _ 6.111
Belgium 0 _ 7 _ .7 _ .7 _ 52 _ 10.4 _ 6.650
England 0 _ 19 _ 1.9 _ 1.9 _ 36 _ 50.4 _ 6.862
Germany 1 _ 21 _ 2.1 _ 3.1 _ 55 _ 82.4 _ 7.288
France 1 _ 12 _ 1.2 _ 2.2 _ 55 _ 63.6 _ 7.371
Poland 0 _ 4 _ .4 _ .4 _ 23 _ 38.1 _ 7.692
Luxembourg 0 _ 0 _ 0 _ 0 _ 27 _ 0.5 _ UND