Our ancestors might not have run around with BMI charts (and I doubt many people today do either). However, the BMI system is a reasonable shorthand method of classifying different body types, rather than (fun as it might be) spending the next six months debating the relative merits of individual porn stars, and whether or not we might each classify them as skinny or otherwise.

Now, elementary Evolutionary theory says that the primary reason men are attracted to women is their ability to give birth to healthy children. That makes sense to me, and I won’t disagree with you there.

But, the analysis found that amongst the people downloading the pictures (presumably men), the lower BMI women were the more popular, especially when compared to both the average women in the population and even Beauty queens – thus, the Sociobiological theory breaks down.

As you correctly note, large breasts and fat reserves make for a healthy Mother, and this seems to account for larger (Rubenesque) women being considered very beautiful in the past, as can be evidenced from paintings of the Wives and Mistresses of Princes, Kings and Nobility from ages gone by.

So, this together with the analysis indicates (to me at least) that the paradigm has shifted, and women with large waist/hip ratios are no longer considered as attractive as they once were, especially when added to anecdotal evidence of young girls spending more time dieting, and a reported increase in cases of Anorexia amongst Western girls.

Medical evidence tells us that low BMI women are less likely to conceive, and are also more likely to suffer complications in pregnancy – Thus, a girl could have stunning facial beauty, but if she doesn’t have the body mass to support the development and nourishment of a baby to its full term, then this makes no evolutionary sense at all because then her genes will simply not be passed on to the next generation. So, again, the fact that people are actively selecting low BMI women in preference to higher BMI women seems to break the ‘appearance’ theory.

More controversially, when you talk about killing Cavemen, you’ve completely lost me… How or why does that change people’s choice of low BMI women over women with a BMI closer to the population average? I’m guessing you’re confusing some naïve pop-science interpretations of Polygamy and Evolutionary theory. But then, think about it, if large numbers of Pleistocene people were killed because some rival felt that he had a better claim over a man’s wife than her existing partner, then their would most likely be almost nobody left alive today. Peacetime Rape and Murder have always been serious crimes throughout our history, and are even universally condemned amongst all primitive tribes, therefore it’s highly doubtful that our own ancient ancestors would have condoned such activities.

Apart from the facts that it’s only in the past few thousand years that people have realised that semen causes pregnancy; and that Monogamy is a relatively new idea, spread very rapidly in recent years by church Missionaries – Why take the risk that you might lose in combat (and thus end up dead yourself), against an irate husband who has tradition, his family, his honour (and presumably the support of his female partner) on his side, versus you as a trouble-making interloper, determined to disturb the peace for everyone else in the village. Not to mention that you might be victorious in combat, only to have the woman spurn your advances towards her.

Therefore, isn’t it far more sensible to seduce the woman, and then disappear off to a secluded spot with her, whilst her original partner is away doing whatever Pleistocene men did? Thus, you get fun with the girl; you avoid putting yourself in danger from her husband or his family. And, much more importantly in evolutionary terms, should you impregnate the woman, her husband becomes cuckold to your offspring (and ergo your genes), which now has a much greater chance of surviving to adulthood since the infant will have two parents to support it (plus whatever contributions you make to its upbringing), rather than only having a raped single mother, whose original partner you just slaughtered a few months back.

Back to your hypothetical example of twins, this would seem to be somewhat irrelevant, since if we give one underweight twin sister large breast implants, this will increase her total weight (around 1 or 2Kg if the implants are really big), and therefore increase her BMI score. Additionally, the size of the girl relative to her breasts could also create an optical illusion effect, whereby the pendulous breasts the girl has just acquired, make the rest of her body appear to be even smaller than it was before. This enhancement may make her more desirable to lovers of large breasts, and thus give her a larger pool of sexual partners to choose from, in comparison to her less well endowed twin – but the point you seem to be missing is that the analysis predicts that the lower BMI girl would be the more desirable of the two.

As a very quick experiment, I increased all the weights of the porn girls from the original study by 2Kg; but that only shifts the modal BMI upwards by about 0.8 points, still leaving a large number of the women underweight, according to WHO guidelines.

Finally, since these are all porn actresses that were comparing, it seems highly unlikely that skinny arms and legs would be concealed under clothing or even ignored by Pornography fans – especially when added to the fact that statistically the large sample size involved would almost completely eliminate the chances of so many underweight girls turning up in the sample merely by chance.

So, we still don’t have an answer as to why the low BMI girls are so popular.

My assumptions:

Fitness to bear healthy children is at least partially underlying biological attractiveness – before any social overlays.

Cavemen diddn’t run around with BMI charts.

Therefore, there must be some means of rapidly computing a females fitness to bear children, based on appearance.

(Is it worth trying to kill the caveman you’ve just seen to steal his wife)

Symmetry, muscle tone, body-fat and other factors will go into this.

If however you take two underweight twin sisters, and give one large breast implants, this may well confuse the computation of fitness.

Both directly – breasts are fat reserves, which can supply nutrition during critical child development, and indirectly by causing other indications that the female is unsuited to bear children (skinny legs and arms indicating low body fat) to be ignored or not properly weighted.

Again, assuming this mechanism got locked in sometime in the hunter-gatherer stage, where lack of nutrition causing failed pregnancies was a big issue.

The Pornography industry has pretty much every type of woman you can imagine doing nearly anything you could imagine on camera; and a quick Google for pictures and information about the low BMI adult entertainers doesn’t appear to indicate anything either way, because it seems at first glance that some low BMI Porn stars have had breast augmentation, and some haven’t.

Furthermore, since surgically enhanced women have larger breasts – fat is less dense than saline solution and is usually less dense than silicone gel – I’d expect the Porn stars who’ve had boob jobs to have higher BMI’s.

Now, any Internet based survey can be criticized for being self-reported data, but in this case, it’s actually the skewing of the information that we’re interested in, since the data used isn’t just a random selection of adult entertainers examined to see if there’s a correlation between skinny models and girls having sex on camera.

The data is based on the popularity of Adult entertainers amongst people downloading Pornographic actress wallpapers, with the results of the analysis showing that girls with borderline poor health BMI’s are the much more popular choices, compared with more average sized females.

Thus, the conundrum at the end of the article still doesn’t have an answer.

Self-reported weights.