Sniffing out human pheromones

Many of us love the idea of the quick dating fix offered by body sprays that purportedly contain human pheromones. After all, spraying on a sexy scent might give you the aromatic edge you need to attract that elusive mate.  

But while Internet companies might claim their sprays have scientific backing, the truth may not smell so sweet. In a review of the scientific literature on pheromones published in Proceedings of the Royal Society B, zoologist Tristram Wyatt of the University of Oxford in England argues that the majority of human pheromone research has little to recommend it. He suggests starting from scratch, with a series of strict guidelines for good pheromone research.

To understand why human pheromones are a controversial topic, you have to begin with what the word “pheromone” means. Most scientists agree that a pheromone is a chemical signal that helps individuals within the same species to communicate. Perhaps the chemical produces a specific behavior, or initiates puberty or heat. But after that, things get murky. Some say that the signal needs to produce an innate behavioral response in the recipient. Others say the response could be innate or learned. Some say that the chemical must have evolved to be a signal. Others say it could be by accident. “The problem with pheromones is that depending on who you ask, you get a different definition,” says Darren Logan, an olfactory geneticist at the Wellcome Trust Sanger Institute in Hinxton, England.

Some researchers even go so far as to say that pheromones in general don’t exist. “It’s an attractive concept, the idea that there are mysterious things floating around controlling behavior,” says Richard Doty, who studies smell and taste at the University of Pennsylvania. But, he notes, “I really question the scientific validity of the concept of pheromones to begin with. Everyone defines it as what they want it to be. The term has lost any real meaning.”

This public view of pheromones has been driven largely by a few popular studies. People stare at stick figures while sniffing scents to see if they can determine the gender of the drawing. Women inhale the smell of sweaty shirts to influence their mate choice. But many of these studies are subject to problems. “There’s a huge publication bias. We know that many interesting results will never be replicated,” says Mats Olsson, an experimental psychologist at the Karolinska Institute in Stockholm. “It’s a general problem in science. But in the pheromone field this has been a larger problem because the results have been so sensational.” In fact, Olsson says the field has become so controversial that he avoids using the term pheromone altogether.

Replication aside, Wyatt points out that none of the chemicals identified as human pheromones actually appear to be biologically relevant in humans. No putative pheromones have been isolated from a human secretion, synthesized and then put under human noses to produce reliable effects on behavior. “I saw good scientists being diverted on a wild goose chase,” Wyatt explains, “devoting time and effort to studying molecules for which there was no basis.”

Consider the pig. One of the most oft-studied  human pheromones is androstenone, a pheromone produced by male pigs that sends females into heat. It is also found in human sweat. It seemed logical that they would also be human pheromones.

But there’s more to sweat than pheromones. “There are thousands of chemicals in human armpits,” Logan says. “We share some of them [with pigs] because we are mammals, but they may not play a role in our behavior.” In the case of androstenone, genetic differences in odor receptors result in variable smell experiences. Some report a “sweaty, urinous” smell, while others report that it smells sweet. Still others can’t smell it at all.

But androstenone is used commercially in pig farming, so “you could buy [the chemical] off the shelf without having to do the difficult chemistry and assays,” Wyatt says. So the molecule was easy to obtain for study. But Wyatt notes there’s not really much evidence that adrenostenone is likely to be a human pheromone. He says that many of the papers suffer from small sample sizes and poor experimental design. And after a few years of studies with the chemical, there was a pheromone echo chamber. “When the papers went out to review they went to other people who had already used the molecules and had no reason to question that they were pheromones.”

Androstenone, then, has never really been assessed for what human behavior it might cause. It may be associated with women sitting closer to men. But is this a chemical signal? Or a learned response to body odor that people are used to or taught to prefer?

We often respond to scents not because of something innate, but because of something in our culture, notes Olsson. “Say I observed persistent avoidance behavior of a sweaty person on a bus,” he explains. “I can identify the molecule, and that this is responsible for avoidance, but would it qualify as a pheromone? From a psychological point of view, I would say this person smells a lot, and this is not acceptable in our society.” Humans may display a behavior, but it doesn’t mean there’s a pheromone in the mix.

To establish human chemical signals with consistent behavioral effects, Wyatt argues, scientists need to go back to first principles. He proposes a series of experiments to establish the presence of a pheromone in humans. First, confirm that a chemical odorant produces a specific behavior or bodily response. Then, identify exactly which chemical is producing the response. Synthesize that chemical in a laboratory. Finally, show that this synthesized chemical is required to produce the response, and that no other chemicals are also needed. Check all those boxes and you might have found a human pheromone.

This sounds relatively simple and straightforward, and maybe it is if you do the tests in lab animals. But trying to run such an experiment to identify a pheromone in human adults might be more challenging. First, which secretion do you want to test? Sweat may be the least embarrassing of the options, and perhaps the easiest to obtain, but will it contain the chemical you are looking for? Second, what behavioral response should result? If you want to induce a sexual response, should you look for hormone levels? Physical signs of arousal? How do you make sure that the response is the result of the chemical, and not a matter of our culture teaching us to prefer one scent over another?

There’s also the matter of finding the money to do these studies in the first place, Logan notes. “If you’re curing cancer, there’s money for that,” he says. “But if you’re trying to figure out if a chemical makes you feel slightly sexy, who’s funding that?”

Wyatt believes the best chance for finding the first established human pheromone might be in newborn babies. A team in France led by Benoist Schaal has been examining chemical secretions from the breasts of new mothers. The researchers have shown that chemicals produced in glands around the nipple produce increases in mouth motion in newborns, potentially leading to better latching and more effective early breastfeeding attempts between mother and child. If there is a single chemical that is responsible, it could fit the definition of a pheromone: A chemical signal between two members of the same species that produces a reliable behavioral effect.

Establishing that there is a pheromone that acts between mother and child may not be as sexy as one that stimulates sexual behavior, but it may also be more attainable. A newborn child has no cultural indoctrination for what she should or should not prefer. It is possible that there is a chemical signal involved that is present in all mothers, though exactly what that chemical is hasn’t yet been identified.

Wyatt believes studies such as Schaal’s are the best hope of finding pheromones in humans. “Molecules have been shown to be effective pheromones in mice, sheep, elephants, goats and rabbits,” he notes. “We’re just another kind of smelly mammal.” The trick will be reliably linking the right chemical with the right behavior. “The chemistry is getting ever more sophisticated,” Wyatt says. “But if we’re going to get anywhere with humans, I think we need to start again.”