Scientists have found long-sought proof that people release potent chemical signals that can have profound effects on other people.
The research settles a 40-year debate about whether humans produce and can respond to pheromones, molecules that are usually airborne and odorless and which, in other species, influence such physiological processes and behaviors as mate choice, the recognition of one's own family members, and the ability to "smell" the difference between friend and foe.
Specifically, the research shows that women's underarm odors can alter the timing of other women's reproductive cycles. It explains why women who live together often develop synchronous menstrual periods, and could spur development of "natural" fertility drugs or contraceptives.
The finding may also lead to the discovery of compounds in sweat that could be incorporated into fragrances to alter body chemistry or mood.
"This is definitely going to make people sit up and take notice," said Charles Wysocki, a neuroscientist at the Monell Chemical Senses Center in Philadelphia. Previous studies by scientists at Monell and elsewhere showed similar results but later were recognized as flawed. The new work, Wysocki said, seems to answer the question for good.
"The evidence has now become quite strong that humans produce and detect pheromones," agreed Edward Johnson, a chemosensory biologist at Idaho State University in Pocatello.
The discovery was especially gratifying to Martha McClintock, the University of Chicago researcher who, with colleague Kathleen Stern, describes the work in today's issue of the journal Nature. As an undergraduate almost 30 years ago, McClintock observed that many women in her dormitory menstruated in synchrony - a phenomenon that became the focus of her senior thesis.
For decades McClintock immersed herself in the task of identifying the timing mechanism. She and others suspected pheromones, but proof was hard to come by.
Pheromones have been documented in many species, ranging from insects to elephants, as sex attractants, kinship identifiers or alarm signals. In many species they are detected by a specialized organ inside the nose or mouth called the vomeronasal organ, or VNO.
There was ample evidence that human pheromones exist; babies show a clear preference for pieces of clothing that have been worn by their own mothers, for example, and research suggests that men and women choose their mates in part by sniffing out partners with compatible immune systems. Several years ago, researchers in Utah even said they had identified the first human pheromones - and turned their discovery into a line of perfumes that today boasts revenue of $40 million a year.
But the Utah work has been criticized by many experts. And the Monell work, on menstrual cycles, did not take into consideration the fact that many out-of-phase cycles will naturally converge over time.
Moreover, scientists have remained uncertain whether the human VNO, a pair of tiny pits in the nose, is a functional organ or an evolutionary vestige.
To find out if human pheromones exist and can affect menstrual timing, McClintock and Stern asked nine women to wear gauze pads in their armpits all day. (Sweat is a common source of pheromones in mammals.) The pads, changed daily, were cut into pieces and frozen, and a daily tally was kept of each woman's menstrual phase.
Then, every day for four months, the researchers rubbed thawed gauze pads above the upper lips of 20 volunteer women who had agreed to have any of 30 different "natural essences" rubbed under their noses. "Sweat" was among the 30. "We buried it in the list," McClintock said.
For two months, 10 women sniffed sweat from women in the early phase of their menstrual cycle, while the other 10 sniffed sweat from women in a later phase of their cycle. Then the groups switched and spent two months getting the opposite scent. Throughout, the sniffers' own menstrual phases were tracked.
The women smelled nothing, but the results were striking: Those exposed to "early phase" sweat saw their own cycles shortened by an average of 1.7 days per month, and as much as 14 days a month. Those who sniffed "later phase" sweat saw their cycles lengthened by an average of 1.4 days a month, and up to 12 days a month.
Computer models indicated there must be two substances in the sweat - one that lengthens cycles and one that shortens them - and that together they can quickly lead to groups of women having synchronous periods.