Welcome to Scientific American's

Mind Matters

where top scientists in neuroscience, psychology, and psychiatry explain and discuss the research driving their fields. Readers can join them. We hope you will.

This week we ponder

The Scent of a Man
- or -
Sniffing around for a human pheromone

Pheromones at work: A pair of monarch butterflies perpetuating the species. Photo courtesy the Tyler Arboretum, via Wikipedia Commons. Used by permission.



by David Dobbs, Editor, Mind Matters

Butterflies and many other animals use pheromones to attract mates, and research has shown that many nonhuman animals also use pheromones to communicate alarm, food availability, territory, and navigational information. But do humans deploy and respond to these chemical signals? Though much evidence (and logic) suggests we do, the question is far from settled. Now, as psychologist Michelle Wirth relates below, a new study about women's reactions to a chemical in men's sweat ("Smelling a Single Component of Male Sweat Alters Levels of Cortisol in Women," by Claire Wyart, Wallace W. Webster, Jonathan H. Chen, Sarah R. Wilson, Andrew McClary, Rehan M. Khan, and Noam Sobel, from the February 7, 2007 Journal of Neuroscience) raises some interesting questions while suggesting that we're not immune to pheromonal suasion -- the work of chemicals, it should be noted, that neither the conventional five senses nor the conscious mind can detect. Read it and wonder.

Men, your armpit excretions affect women more than you might think

by Michelle Wirth Affective Neuroendocrinology Lab University of Wisconsin, Madison, MI

Michelle Wirth

In the ongoing debate over whether human pheromones exist, this new study by Claire Wyart and colleagues adds a tidbit more evidence. Pheromones are hormone-like compounds emitted by one individual that, when detected by another individual's hormone receptors, affect physiology and behavior. In many animals, pheromones are detected by the vomeronasal organ (VNO) in the nose, which then sends signals to brain regions involved in homeostasis (roughly speaking, the regulation of physiology) and sexual behavior. Although odors from other individuals also affect behavior, the VNO is a separate system from the olfactory system, which detects smell. Until recently, the VNO in humans was thought to be vestigial and non-functional, and scientists still debate how significant a role the VMO plays. Thirty-five years ago, though, University of Chicago psychologist Martha McClintock (then a Wellesley undergrad) performed classic studies showing that nasal exposure to extracts of sweat from other women had the ability to synchronize women's menstrual cycles. Since then, a series of studies have demonstrated that women exposed nasally to androstadienone, an androgen-like compound found in male underarm sweat, show a host of psychological and physiological responses, even when the odor of this substance is disguised or when androstadienone is applied directly to the VNO so that women can't smell it. Androstadienone exposure made the women feel slightly less tense and nervous and reduced their heart rate, skin conductance and other measures of physiological excitability. Even more interesting, positron-emission tomography (PET) brain imaging studies showed that androstadienone exposure caused activation in the hypothalamus (a region involved in sexual behavior and reproductive hormone control, among other homeostatic functions) in straight women and gay men, whereas in straight men and gay women it activated only olfactory processing centers. Meanwhile, an estrogen-like putative pheromone produced by women did the reverse, activating the hypothalamus in straight men and gay women but only olfactory areas in the gay men and straight women. Some Answers, Many Questions In the Wyart study considered here, women sniffed androstadienone and a control substance on two different days. The amount of the control substance, baking yeast, had been predetermined for each subject to match the (un)pleasantness and intensity of the smell of androstadienone; still, one wonders how different these two substances smell. Could women tell them apart? More important, could women recognize the smell of androstadienone as smelling like men's sweat? These potential confounding factors could undermine the authors' arguments. Nonetheless, the researchers found that women's self-reported positive mood and sexual arousal were increased after androstadienone and not after yeast exposure, and (in contrast to previous studies) physiological arousal (as measured by heart rate, blood pressure, skin conductance and other measures all combined together) rose as well. Furthermore, sniffing androstadienone kept levels of cortisol, a stress hormone, steady throughout for the next hour, defying the normal drop in cortisol that generally occurs over the course of the day. Wyart and colleagues argue that this is the first evidence that androstadienone affects hormone levels in women, giving extra oomph to the proposal that this chemical is a human pheromone. However, a number of questions about this study arise in my mind. First, why didn't they measure the effect androstadienone had on other hormones? For example, if androstadienone affected estradiol and progesterone -- or even better -- luteinizing hormone and follicle stimulating hormone (pituitary hormones that control ovarian function), such effects would provide even stronger evidence that androstadienone serves a pheromone-like function. That no such hormone data are reported makes me wonder if they measured the reaction of other hormones but found no effect. In addition, time of day and stage of menstrual cycle, while matched for a given subject's two sessions, were not kept constant across subjects. Time of day is one of the biggest influences on cortisol level, which peaks in the morning and drops steeply through the day. Many studies don't even find that stress affects cortisol levels in the morning, because levels are already so high and in flux. The cortisol results reported here are thus hard to interpret without knowing what times of day that subjects were tested. Even more important, most other androstadienone studies have tested women in the late follicular phase of the menstrual cycle, shortly before ovulation -- a time in the cycle when women are most likely to conceive, when they have higher levels of sexual desire, and when they tend to be more drawn to masculine features in men (possibly including manly pheromones). In fact, some studies tested women in different cycle phases and found effects of androstadienone only in the late follicular phase. Wyart et alia's subjects presumably were tested at different points in their cycles. I wonder if this difference accounts for their findings of increased physiological arousal, in contrast to other studies that found lower heart rate and skin conductance in women exposed to androstadienone. (Wyart and colleagues report a rather sketchy composite measure of these physiological variables, possibly meaning that they didn't find significant results in any one variable.) Can You Be Relaxed and Excited? These conflicting findings about physiological arousal raise a final point: the authors do not address why androstadienone might increase physiological arousal and levels of a stress hormone, whereas other researchers have found calming effects of androstadienone. Wyart's results could be interpreted as evidence of "eustress" (a healthy stress, essentially) or positive excitement caused by androstadienone. Raised cortisol levels do not necessarily indicate distress or negative emotion: cortisol's main function is to mobilize energy in the body, so cortisol levels rise when we wake up in the morning, when we eat (which is a challenge to homeostasis), when we exercise, and when we prepare for important (but potentially fun) events like speeches or competitions. In this case, androstadienone may be triggering a hormonal response in women to prepare for courtship or sexual activity. However, the authors provide little interpretation for their results, leaving the reader to be a little confused as to why a hormone associated with stress would be increased by this supposedly sexy male pheromone. I'm convinced by the overall body of literature that androstadienone is a chemosignal capable of having subtle but perhaps important effects on women's physiology and mood. However, whether it is a true pheromone remains to be seen.
Michelle Wirth is a post-doctoral scholar at the University of Wisconsin's Affective Neuroendocrinology Lab, where she researches affective neuroscience.

OLD COMMENTS ON THIS POST (i.e. comments that were made on this post before we switched to the new commenting system - please leave subsequent comments below)