(c) 1992 David Wolfgang-Kimball
Pheromones are volatile substances which are released by one animal and detected by another, causing some sort of physiological reaction. These reactions can manifest themselves in a variety of different ways: some pheromones modulate sexual activity, some affect aggression, some play roles in territory marking, and other pheromones have similarly diverse effects on the target animal. Pheromones have been demonstrated in a very large number of organisms ranging from amoebas to fish to mammals, including primates. However, the question of whether human olfactory signals exist has been a question of much debate and few definite conclusions. In this paper I will look at some possible examples of odor signaling in humans.
Mammals of all sorts use olfactory signals to indicate willingness to copulate, define territory, mark their young, and signal aggressive intent. These processes can be seen in many animals used as models for human systems, including rats, monkeys (both Old World and New World), hamsters, and mice. The fact that pheromones are important biological signals in a plethora of other species indicates that the possibility of human pheromones should not be discarded lightly.
Although humans generally rate olfaction as their least important sensory modality, we still spend billions of dollars, years of our life, and a considerable amount of effort to modify the way we smell (at least in industrialized countries). These efforts typically include scrubbing with deodorant soaps and scented shampoos, applying deodorants to those parts of our bodies we feel need deodorizing, and finally applying perfumes and sprays to replace those natural odors we just discarded down the shower drain. This points out an obvious contradiction: if olfaction is considered unimportant and possibly even obsolete, why do we work so hard to change the way we smell? The first question to address is where do these odors we produce come from? Whereas animals release pheromones from their skin, urine, feces, and to some extent breath, most research on pheromones in humans indicates that the main odor-producing organ is the skin. For the purposes of this paper, the skin is what I will focus on. These odors are largely produced by the skin's apocrine sebaceous glands, which develop during puberty and are usually associated with sweat glands and tufts of hair. These glands are located everywhere on the body surface, but tend to concentrate in six areas1:
1) The axillae (underarms)
2) The nipples of both sexes2
3) The pubic, genital, and circumanal regions
4) The circumoral region and lips
5) The eyelids
6) The outer ear
The first four of these regions are generally associated with varying amounts of hair growth, which makes perfect sense, as the extremely large surface area of a tuft of hair is a very effective means of spreading an odor by evaporation. The fact that body hair and apocrine glands appear simultaneously at puberty is significant and suggests that body odor and its dispersal may be linked to sexual development. These supposedly non-functional structures, coupled with the olfactory system, would be called part of a pheromonal system in any other mammal.
The substances produced by these glands are relatively imperceptible by the human nose; what we smell when we detect skin odor is not the fresh glandular secretions but rather the bacterial breakdown products of these glandular secretions. The sebaceous secretions themselves consist mostly of lipids such as squalene and other esters. When degraded by enzymes of bacteria naturally present on human skin, free fatty acids result, including those that smell hircine and are generally regarded as unpleasant. The most prominent examples of these hircine fatty acids have the general formula (CH3(CH2)nCOOH) and are called butyric acid (n=2), caproic acid (n=4), and caprylic acid (n=6).
The first studies I will discuss relate to evidence for the existence of pheromone signaling in human babies and children. The first interesting studies regarding children come from Michael Kalogerakis and Irving Bieber. They proposed a theory that olfaction is related to sexual identification in young children. Kalogerakis performed a study on young boys, two to four years of age, which strongly indicated that at some point in early childhood, a boy will begin to show an aversion to the odors of their father, and will simultaneously feel attraction to the odors of their mother. According to Bieber, this indicates a shift in sexual interest and acts as a biological trigger for the Oedipus response. Kalogerakis supports this theory with a case study of a boy named Jackie, who originally was closer to his father, but at the age of three years, three months, began to show a distinct preference for his mother's smells, especially at times right after she and Jackie's father had been having intercourse. At four years of age, Jackie would become nauseous at the smell of his father. This behaviour continued, tapering off slowly until Jackie was six, and his sexual identity had presumably been established.
Another intriguing study was carried out by Michael J. Russell of UCSF in 1976. He enlisted the help of ten recent mothers, whom he asked to wear a cotton pad in their bra for three hours before testing. Russell then tested the sleeping babies' ability to differentiate between pads worn by their own mothers and those worn by strange mothers. At the age of two days, only one of the ten babies responded to either type of pad, and he responded to both with a sucking response. At the age of two weeks, eight babies responded by sucking to a stranger's pad, and seven responded to their mother's pad. Also, one child responded only to its mother's pad. At the age of six weeks, however, things had changed. Eight babies responded to their mother's pad, one responded to a stranger's pad, and one did not react to it's mother's pad but did react with a jerk and a cry to the stranger's pad. These results may indicate either that a baby inprints on its mother's odor, or as Russell suggests, that the mother unconsciously marks her baby with a distinctive scent, a phenomenon observed in many other primates. This latter possibility is supported by the common parental observation that a child will reject their favorite blanket or stuffed animal after it has been washed, presumably because it has lost specific odors acquired in previous contacts.
A final childhood phenomenon worth mentioning is one observed by Dr. Alex Comfort. Comfort noticed that in the past three centuries, the age of onset of menstruation for girls has had a direct correlation with the amount of time that young girls spend with boys. In pre-Victorian times, menstruation began at an early age, only slightly above the average age of onset now. However, in Victorian times, when mingling between the sexes was minimized as much as possible, the average age of onset climbed a few years. In post-Victorian times, as boys and girls were allowed to mingle more freely and coeducation appeared, the average age fell once again. Admittedly, this could be due to a number of other factors, but it is Comfort's opinion that it is due to the exposure to odors of the opposite sex. In fact, this phenomenon has been documented in mice and is called the Vandenbergh effect: female mice raised alone in sterile cages have a much higher age of maturation than that of female mice raised alone in cages filled with a male mouse's bedding material. When the bedding belonged to a castrated male mouse, this effect was not observed.
There are variations in odor perception between human adult males and females. Le Magnen and Doty found that this is most evident in the case of women's acute ability to smell musk3, which are steroids, large cycloketone or lactones, often with side chains which are most likely involved with their biological specificty of action. All of these compounds are very similar to the male sex hormone testosterone (see appendix for structures). Whereas women are very sensitive (1 part in 1,000,000,000) to the musky odors of civetone (from the anal glands of the civet cat and used in many perfumes), exaltolide (a synthetic musk), and boar taint substance (a sexual attractant produced in the preputial glands of the boar), men are relatively insensitive (1 part in 1,000,000) to these substances. Moreover, women's sensitivity to these substances varies as a function of where they are in their menstrual cycle: during menstruation, women are no more sensitive to musks than men, but about ten days after menstruation (ovulation -- a woman's peak fertility period), women reach their maximum sensitivity. In addition, women on the pill, women who have had ovarectomies, pregnant women, and post-menopausal women are relatively insensitive to these substances. Le Magnen deduced from these results that sensitivity to musk in women is critically dependant on the levels of estrogen in the blood: during ovulation, serum estrogen is at a peak, whereas serum levels of estrogen are low during menstruation, pregnancy, in post-menopausal women, women who have had ovarectomies, and birth-control pill users. Further, it is the action of progesterone which causes nasal congestion during menstruation and pregnancy4, and might be responsible for the reduced sensitivity at these times.
Why is this relevant? Men secrete musky odorants in abundance. The -3-ol precursor of boar taint substance is found in male urine, and substances similar to testosterone, such as androstenone, are secreted in the smegma and from the apocrine glands of the underarms5 and pubic area of males. As is usually the case, bacterial action may be necessary for the release of the odorants. The fact that men's bodies secrete these substances and that women are maximally sensitive to them when they are most fertile indicates that there may be a olfactory-sexual role for these substances in human sexuality.
Indeed, a study performed by J. Richard Udry at the University of North Carolina attempted to delineate the relationship between coitus, orgasm and position in the menstrual cycle. He found that women do indeed engage in sexual intercourse about six times more frequently at about the time of ovulation, when women's sensitivity to the male musk odor is highest. In addition, the women are much more likely to have an orgasm at these times. Further, the women Udry studied women were several times less likely to have sexual intercourse or have an orgasm during and two to three days after menstruation, which is when women's sensitivity to the musky smell of men is lowest. Coupled with women's odor sensitivity, these results could indicate a possible pheromonal trigger for sexual behaviour.
There are several other effects in adult humans which might hinge on pheromones. Some of the most interesting results come from work done by Martha McClintock at Harvard. She performed a study on menstrual cycles in women who lived together in dormitories and found that when women are housed together, their menstrual cycles tend to synchronize and lengthen. She also found that the lengthening effect was attenuated in dirext relation to the amount of time these women spent with men. In one woman's case, her regular cycle was six months long, but when she started seeing a man, it dropped to four and a half weeks. After she stopped seeing this man, her cycle once again lengthened. Of course, in an experiment like this, it is difficult to eliminate diet, work and sleep habits as factors, but the fact that this is such a widespread phenomenon indicates that something more basic is probably at work here. It is to be stressed that airborne odors or pheromones were not directly demonstrated in this study, but there is an identical phenomenon in mice that has been shown to be pheromonal in nature. This effect is called the Lee-Boot phenomenon, in which groups of female mice housed together experience increases and synchrony in their estrus cycles. When a female mouse is housed alone, this effect does not occur, but when a solitary female mouse is kept in a cage supplied with bedding from a cage full of female mice, the Lee-Boot effect is once again observed, indicating that the cues are chemosensory in nature. The attenuation of cycle elongation in women in response to male contact is also echoed in mice, and is called the Whitten effect. Once again this effect has been shown to be due to olfactory signals.
Michael Russell provided some more insight on the phenomenon of menstrual synchrony. A colleague of his, on reading McClintock's paper, mentioned that she too had noticed the same phenomenon among her friends, except that in every case, it was her own menstrual cycle which determined the synchronization of her friends'. Upon hearing this, Russell asked his colleague if he could use her underarm scent to help confirm and extend McClintock's findings. She consented, and proceeded to wear sterile cotton pads under her arms regularly. Russell the recruited sixteen female volunteers, each of whom came in three times a week for four months to have a liquid applied to her upper lip. One group of women had pure alcohol applied to their lips, and the other group had a mixture of alcohol and Russell's colleague's underarm scent from the previous day applied. The group which received pure alcohol did not experience changes in their menstrual cycle, but those that had the mix of alcohol and underarm scent applied showed a radical change in their cycles: The average time lag between cycles had been 9.3 days, but after four months, this had decreased to 3.4 days, and fully half the women were in exact synchrony with Russell's colleague, discounting the the aforementioned one day time lag. None of these women had ever even met Russell's colleague. McClintock's study showed that women who lived together reported menstrual synchronization, and Russell's study provided a likely mechanism: underarm scent. Another possible interpretation of this study leads to the conclusion that there may be dominant women with regard to menstrual synchrony, a phenomenon observed in many animals.
Dr. Russell provided yet another interesting result. At the same time he was performing his experiments on babies' ability to discriminate between their own mothers and strange mothers, he performed another experiment on whether young adults could discriminate between their odors and others' and between male and female odors. Twenty-nine college age students, 16 male and 13 female, were asked to wear a clean undershirt for twenty-four hours without using soap, deodorants, or perfumes. After twenty-four hours passed, the shirts were collected and put in buckets with the armpit right above a strategically placed sniffing hole. Two tests were then performed: the subjects were presented with three shirts, one theirs, one from a strange female and one from a strange male. The subjects were then asked to ientify which shirt was theirs, taking as much time as needed. The subjects were then asked to identify which shirt belonged to the strange male and which shirt belonged to the strange female. The subjects generally sniffed each bucket once in succession, and then repeated the process. The results were impressive: 81% of the males and 69% of the females identified their own shirts correctly, for an average success rate of 75%, which is highly significant when compared to the chance percentage of 33%. In the second sex-identifying test, the subjects performed just as well: 81% of the males and 69% of the females were correct, for an average of 75%. Once again, this result was very significant, as chance would dictate a 50% success rate. When asked to characterize the odors of the shirts, the subjects generally said the males' shirts smelled musky and the females' shirts smelled sweet. This observation jibes well with the previous discussion of variations in odor perception.
One final effect needs to be mentioned due to large amount of research on it. There have been many studies on whether or not human vaginal secretions might contain some kind of sex pheromone (or "copulin", as one researcher calls them). Several researchers have found that human vaginal secretions contain various small (C2 to C6) fatty acids, with acetic acid predominating. Richard P. Michael found that about 30% of the women (he called them 'producers') produced a significant amount of those small fatty acids (not including acetic acid) that induce copulatory behaviour in infra-human monkeys. In addition, these "copulins" increased up until ovulation, and then decreased as menstruation approached. Michael also noted that women on birth-control pills did not show this mid-cycle increase, and had a lower overall fatty acid content. Michael theorized that these fatty acids or "copulins" were a sexual trigger in humans, but this has never been demonstrated, although the producers' secretions did increase copulatory behaviour in rhesus monkeys. When David Goldfoot's group in Wisconsin tried to confirm these results, however, they were unsuccessful.
Are pheromones in humans a myth or are they real? At this point, it is difficult to say either a definite yes or a definite no. The field is obviously very confused, and for every paper one finds that seems to demonstrate the existence of human pheromones, one can find another equally compelling study refuting their existence. In this paper I have tried to consider a few compelling bits of evidence, but it should be noted that none of these results are yet widely accepted, and no pheromone has yet been isolated and conclusively linked to a physiological effect in humans. Further, much of the work in this field is of a qualitative nature, without adequate controls or firm statistical basis.
However, some of the results mentioned above are quite compelling. McClintock's study and Russell's extension seem to strongly indicate there is some odorant that affects women's menstrual cycles. The fact that men secrete musk-like substances that women are maximally sensitive to during ovulation coupled with the finding that there is a demonstrated increase in coitus during this period is also very intriguing. "Copulins" may or may not be human sexual releasers, and they seem to stimulate copulatory behavior in monkeys, although this result has not been confirmed.
To close, I would like to propose a new way of looking at pheromones, specifically in humans. With our highly developed intellect and rich compliment of emotions, ambitions, motivations and desires, it may not be profitable to look at human pheromones the same way we look at animal pheromones. Instead of looking for odorants that cause a definite physiological response, it may behoove us to look at how possible pheromones affect our attitudes. We are not machines that blindly fall into some stereotyped behaviour in response to an odor, but we may be machines that are nudged towards a type of behaviour by pheromones in concert with our higher intellect.
1This is an overgeneralization; there are substantial differences in apocrine gland distribution and quantity between the various races. The six areas outlined here are generally found in caucasians, but blacks and Aborigines tend to have more and larger glands, with a higher number on the chest and abdomen than is found in an average caucasian. In addition, Aborigines have a much more powerful scent gland in the circumanal region. Asians, on the other hand, tend to have smaller and far fewer apocrine glands than either caucasians or blacks, and many have none at all. In fact, only about 10% of Japanese people have any underarm odor at all, and at one point having scent glands in the underarms qualified a Japanese male for a military exemption and a free ticket to a medical center where they could receive treatment.
2Interesingly, the mammary glands themselves are highly modified apocrine glands
3 Musk is a basic ingredient of all perfumes and colognes.
4This phenomenon might be responsible for womens' reputed proclivity for unusual foods during pregnancy and menstruation.
5A note about underarms: many of the authors of the refernces for this paper have pointed out that underarms are the ideal location for the dispersion of odors and /or pheromones. This is because 1) They are among the warmest parts of the body, and are among the first parts to perspire. 2) They are amply endowed with apocrine and sweat glands. 3)There is usually a strong growth of hair, which is a very effective means of dispersing an odor (as noted above). 4) Underarms are high on the torso and thus well-situated to disperse odors in the region of other people's noses. 5) Finally, being under the arms, armpits are protected from excessive evaporation. To release odors, the arms must be raised or in motion. Comfort speculates that underarms may even be specialized for this purpose.
1) Hopson, Janet. Scent signals: The Silent Language of Sex. New York: William Morrow and Company, 1979
2) Stoddart, D. Michael. Mammalian Odours and Pheromones. London: Edward Arnold Ltd., 1976
3) Shorey, H.H. Animal Communication by Pheromones. New York: Academic Press, 1976
4) Vandenbergh, John G. (ed). Pheromones and Reproduction in Animals. New York: Academic Press, 1983
5) Doty, Richard L. (ed). Mammalian Olfaction, Reproductive Processes, and Behavior. New York: Academic Press, 1976
6) Theimer, Ernst T. (ed). Fragrance Chemistry: The Science of the Sense of Smell. New York: Academic Press, 1982
7) Wells, F. V. and Marcel Billot. Perfumery Technology Art: Science: Industry. Chichester: Ellis Horwood Ltd, 1981
8) Comfort, Alex. "Likelihood of Human Pheromones." Nature, vol. 220, pp. 432-479
9) McClintock, Martha K. "Menstrual Synchrony and Suppression." Nature, vol. 229, pp. 244-245
10) Russell, Michael J. "Human Olfactory Communication." Nature, vol 260, pp.520-522
11) Udry, J. Richard and Naomi M. Morris. "Distribution of Coitus in the Menstrual Cycle." Nature, vol. 220, pp. 593-596
12) Michael, Richard P. et al. "Volatile Fatty Acids, 'Copulins', in Human Vaginal Secretions." Psychoneuroendocrinology, vol. 1, pp. 153-163
13) Huggins, George P and George Preti. "Volatile Constituents of Human Vaginal Secretions." American Journal of Obsterics and Gynacology, vol. 126, pp. 129-136
14) Kalogerakis, Michael G. "The Role of Olfaction in Sexual Development." Psychosomatic Medicine, vol. 25, pp. 420-432
15) Bieber, Irving. "Olfaction in Sexual Development and Adult Sexual Organization." American Journal of Psychotherapy, vol. 13, pp. 851-859
16) Michael, Richard P. et al. "Human Vaginal Secretions: Volatile Fatty Acid Content." Science, vol. 186, pp. 1217-1219