donkey o.d. too

My main site, donkey o.d. is moving here. Pardon the dust...

Tuesday, June 13, 2006

Olivia Judson - Is He More Attractive If She’s on the Pill?

June 11
10:30 pm

I hereby declare this week to be … Mad Scientist Week! In that spirit, I’m going experimental. Over the next five days, I will describe a few experiments I’d like to see done, and why. The first is to answer this question: Can taking the oral contraceptive pill change a woman’s taste in men?

Before I get into that question, however, I’d like to say a little about my motivation for the series.

Part of what we know about the world comes from experiments. But many important experiments never get done. Experiments that are very expensive, or that will take years to get results, are often not funded by the public purse. Ideology gets in the way, too. We know far less than we should about human sexuality, for example, because many people shy away from doing the experiments we’d need to find the information. Then, some interesting experiments aren’t practical. It’s much easier for us to conduct experimental evolution on bacteria, some of which have a generation time as short as 20 minutes, than on a fish like the orange roughy, which takes around 25 years to reach maturity.

Finally, ethics prohibit many experiments. Gone, I hope, are the days of shame: the days when surgeons performed head transplants on monkeys. (Yes, such experiments were done, and not that long ago. Unsurprisingly, the transplanted heads died soon after arriving on their new bodies, though at least one of the heads had a chance to bite the finger of a surgeon who was foolish enough to poke it in the mouth.) Gone, too, are the days when the U.S. government deliberately irradiated part of the population (yes, this happened, too), or failed to treat people with syphilis just to see how the disease unfolds (ditto).

But there are still plenty of fascinating experiments that we should do and could do, but haven’t yet done. For this week’s series of columns, I’ve picked questions I think are important or just plain interesting and — as far as I’ve been able to discover — still unanswered. My aim is to play with ideas and suggest ways to explore the world.

And so, back to the pill. I’m prompted to ask the question — can taking the oral contraceptive pill change a woman’s taste in men? — because of a curious result. In 1995, a team of scientists published what is now universally known as “the smelly T-shirt experiment.” In brief: The team collected DNA from 49 women and 44 men, all students at Bern University, in Switzerland. The DNA was then analyzed for a particular set of genes known as the major histocompatibility complex, or M.H.C. These genes are highly variable and are thought to affect several traits, including what you smell like and the workings of your immune system.

The M.H.C. genes are also suspected to influence whom you find attractive. There are two good reasons why this might be the case. The first is that members of the same family have similar M.H.C. genes, and it’s not a good idea (at least if you’re a mammal — some insects are less constrained) to mate with members of your own family, because it can lead to genetic disorders in the children. Avoiding people whose M.H.C. genes are like yours may thus help you avoid sex in the family (not everyone knows their relatives). Second, people with lots of different M.H.C. genes are thought to have immune systems that are better at fighting infectious diseases. If your mate had an M.H.C. different from yours, your children would inherit a broader arsenal of weapons for their immune systems, making them more likely to survive diseases. Because the M.H.C. also affects what you smell like, it is thought that smell might mediate sexiness, with women preferring the smells of men whose M.H.C. genes are different from their own.

To test this, the men in the Swiss study were issued clean white cotton T-shirts and told to wear them to bed two nights in a row. During this time, the T-shirt wearers were asked to abstain from (among other things) smoking, sex, garlic and deodorant. The day after the second night, the T-shirts were put into boxes, and the women were invited to sniff the shirts (each woman was given six of them) and rate the smells.

The results were provocative. Thirty-one women preferred the smells of men whose genes were different from their own. Eighteen women did not — they liked the smells of people whose genes were more similar to theirs. These 18, it turned out, were all on the pill; none of the 31 was.

Provocative — but hardly compelling. For starters, the number of people involved in the experiment was tiny: perhaps those 18 happened to be a bit weird — perhaps, if you looked at 10,000 women, the results would be different. Another problem is that the effect of the pill was not tested directly; it was invoked to explain an anomaly. The effect might have had nothing to do with the pill. Perhaps the women’s desire to take the pill in the first place reflects a wiring of the brain that also somehow determines which smells they would like.

Therefore, I propose an experiment to test the pill directly. The basic idea would be to select a large number of women — say 1000 — of the same age and ethnic group who are not on the pill, and give them the smelly T-shirt test. Then put the same women on the pill (they should all use the same brand), and test them again. We’d probably also want to control for a possible change in preference over time — so we’d want half the women to start off on the pill, and then come off, while the others did the reverse. This is not a complete design — I haven’t detailed everything we’d want to control for — but you get the idea.

Of course, even if the pill were to influence which smells a woman likes, it wouldn’t necessarily be that important in deciding whom she mates with. Woman does not love by smell alone; Romeo does not always smell of roses. Indeed, though the smelly T-shirt experiment suggested that women like the smells of men with M.H.C. genes different from their own, other research has found that when it comes to choosing someone to marry, there is no evidence that M.H.C. makes a difference.

A final caveat: there is some additional circumstantial evidence to suppose the pill may have an effect. A few studies have shown that a woman’s desire for sexual activity changes over the course of her menstrual cycle; such changes are presumably under hormonal control, and the pill affects hormone levels. (Indeed, the pill has been accused of altering the desire for sex, too.) But almost all the studies I know of have involved fewer than 40 people — which makes it hard to draw conclusions. Also, the scientific literature tends to overestimate such effects, because small studies that fail to find an effect often don’t get published. (This is the problem of so-called “negative” results.)

But suppose the pill does have a big effect. Suppose a young woman is on the pill, falls in love, gets married and wants to have children. Then she comes off the pill — and finds she no longer desires her husband. If that is true, it is important. Perhaps we should find out.