tenone, methyl anthranilate, skatole, and a no-odor control) and were asked to rate photographs of a male stimulus and to rate their own mood in the presence of ...
Copyright 1984 by the American Psychological Association, Inc.
Journal of Comparative Psycholojfy 1984. Vol. 98, No. 2, 219-222
Brief Communication Human (Homo sapiens) Responses to the Pig [Sus scrofa) Sex Pheromone 5 Alpha-androst-16-en-3-one Erik E. Filsinger
J. Jay Braun
Center for Family Studies Arizona State University
Arizona State University
Woodrow C. Monte
Darwyn E. Linder
Food Science and Nutrition Laboratories Arizona State University
Arizona State University
Previous studies have suggested that the pig sex pheromone 5 alpha-androst16-en-3-one may function as a human sex pheromone. Two hundred male and female college students were assigned to one of four odor conditions (androstenone, methyl anthranilate, skatole, and a no-odor control) and were asked to rate photographs of a male stimulus and to rate their own mood in the presence of each odorant. There was a significant overall sex by treatment condition interaction. Men in the androstenone condition rated the stimulus male as more passive, and women in the androstenone condition rated themselves as less sexy; these effects were specific to the androstenone condition. Alpha androstenone and alpha androstenol are the ketone and alcohol forms, respectively, of delta-16 steroids which have been identified as pig pheromones (Melrose, Reed, & Patterson, 1974). These pig pheromones are of interest from a comparative perspective because in men, as in boars, they are synthesized by the testes, and they have been identified in human male axillary sweat secreted by apocrine glands (Brooksbank, Brown, & Gustafsson, 1974; Gower, 1972), which in lower animals appear to be specialized for pheromone secretion (Comfort, 1971; Hurley & Shelley, 1960). These chemicals have been identified in the urine of both human sexes after puberty, with much greater concentrations in male urine (Brooksbank & Haslewood, 1961). Furthermore, the ratio of alpha androstenone to testosterone in the blood of men has been found to be similar to the ratio found in boars (Glaus & Alsing, 1976). These observations provide circumstantial support for the suggestion that the odors of these steroids, having a clear pheromonal function in pigs, might influence human behavior as well. The discriminability of male and female human sweat as demonstrated by Russell (1976) lends further circumstantial support. To date, four published studies have investigated the influence of alpha androstenol on the We thank Nancy Bowman, Debbie Brill, Mary Cams, Maggie McDowell, and Renee Weir for their assistance in collection of the data. Requests for reprints should be sent to Erik E. Filsinger, 201 Center for Family Studies, Arizona State University, Tempe, Arizona 85287.
responses of human subjects to various materials designed to assess the subjective impressions of the subjects toward others. Three of the studies claimed to have detected an influence of alpha androstenol on the subjects (Benton, 1982; Cowley, Johnson, & Brooksbank, 1977; KirkSmith, Booth, Carroll, & Davies, 1978). The fourth study found no indication of an influence of the odorous steroid on human judgments (Black & Biron, 1982). The present study was intended to evaluate the effects of the androstenone form of the pig pheromone on human responses in its presence. One of the investigators noted a previously unreported instability of the alcohol form whereas the ketone form was found to be stable. This finding contributed to the decision to use androstenone. It should be noted that Melrose's (1972) patent for artificial insemination in swine references androstenone. In addition, the present study used a broader range of control odorants than the studies in which positive effects of androstenol on human responses were reported (Benton, 1982; Cowley et al., 1977; KirkSmith et al., 1978). Also, in contrast to the one study (Black & Biron, 1982) that found negative effects, we used a variety of rating scales rather than a single item and obtained the results in the presence of the odorants rather than at a later time. Method Subjects Two-hundred undergraduates (98 men, 102 women) at Arizona State University served as
219
220
BRIEF COMMUNICATION
subjects for the experiment. The mean age of the subjects was 20.12 (SD = 3.87). All subjects received credit toward research participation requirements for an introductory psychology course and had signed up for a "Personality Judgment Experiment." Sixteen sign-up slots alternating men and women were available for each day of the experiment. Materials For the first phase, four sets of packets of stimulus materials were prepared. Each packet consisted of four sheets of paper containing instructions, photographs, a written description of a hypothetical male college student, and rating scales that the subjects used for evaluating the stimulus person and for evaluating their own mood. The independent variable (odor) was presented by impregnating the packets with a moderate level of one of three odors, or no odor, representing the four major conditions of the experiment. The odors used were 5 alpha-androst-16-en3-one (alpha androstenone, the experimental odor)—synthesized from androstenone acetate (Sigma A-7633) by a modification of the method of Brooksbank and Haslewood (1961), purified by column chromatography, and verified by high-pressure liquid chromatography; methyl anthranilate (a fruity odor; Sigma M-3888); 3methylindole (skatole, a fecal odor; Sigma M2127). The control odors were selected to represent the positive and negative dimensions of a "pleasantness" continuum, and they provided odor controls that had not been employed in the previous studies. Boilable cooking pouches 8 X 12 in. of 2.5-mil construction (Dazey Products Co., Industrial Airport, Kansas) were used to package the photographs, instructions, and rating scales; after being deodorized in a vacuum after printing, they were either impregnated with one of the odorants or left clean as a noodor control. The bags were found to be impervious to the odorant chemicals even at temperatures about 85 °C. Prior to the experiment, the odorants had been applied to the contents of the bags in the following manner: An aliquot of an ethanol solution of an odorant stimulus was introduced by a micropipet. All compounds were diluted to 1% solution in 100% ethanol. To the appropriate packets were entered 1 mg of alpha androstenone, 20 mg of methyl anthranilate, or 1 mg of skatole. Following introduction of an odorant, the bag was immediately heat sealed. The bags were placed according to group in a drying oven at 65 °C for 24 hr to allow for the diffusion of the odorant throughout the packet and to test for leakage. Preliminary odor tests of varying
concentrations were used to determine the appropriate concentration of the odorant stimuli. Levels were chosen that were above threshold, but were not obtrusive. The sealed packet contained four sheets of paper. The first sheet was a set of instructions which the subjects were to follow for handling the additional sheets, and it contained an invented neutral description of "Paul" as a typical male upperclassman. The second page of the packet was a collage of four black and white photographs of "Paul." The third sheet in the packet was a set of rating scales which the subjects were instructed to use to evaluate Paul on the basis of the information provided. The rating scales consisted of eight items from the affective dimension of the Semantic Differential (Osgood, Suci, & Tannenbaum, 1957). The key items of the activity and potency dimensions of the Semantic Differential were also included, that is, activepassive and weak-strong. In addition, the bipolar stems ugly-handsome and sexually attractive-sexually unattractive were also presented. All items were presented on a 7-point bipolar scale. The fourth sheet in the packet was a Self Rating Scale which contained two variables of interest. One was a set of bipolar items from the Bond and Lader (1974) Mood Rating Scale. The items reflected mood states, for example, alertdrowsy, calm-excited, and strong-feeble. In addition, an item "sexy-not sexy" was embedded in the general self-rating scales. The subjects were instructed to rate the way "you feel about yourself right now" on each item by placing a vertical line on a 10-cm horizontal line. These were subsequently scored by placing a 10-point grid over the scale and assigning numbers of 09 to the position where the subject had indicated he or she felt with regard to that item. Design Four conditions were used: androstenone treatment, no-odor control, positive-odor control (methyl anthranilate), and negative-odor control (skatole). A balanced Latin square design was used to determine the order of the four conditions on 4 days over 4 weeks. Subjects were completely blind to the purpose of the experiment. Receptionists were given a standard protocol to follow in all interactions with the subjects and were blind to odor conditions on a given day. Procedure Subjects received the standardized instructions from the receptionists and then completed
221
BRIEF COMMUNICATION a personal information form. Upon completion of the form, they were handed the plastic envelope containing the experimental materials, escorted into a room in which there was a chair, a desk, scissors, and a sign on the wall telling them to cut open the packet, remove the materials, and follow the written instructions contained within. After completion of the scales, the subjects placed the materials in a refrigerator-like storage unit in the room and proceeded to a third room. A second receptionist conducted a test for anosmia to the odor conditions. The four odors were presented in a balanced random order to all subjects: no odor, androstenone, methyl anthranilate, and skatole. The test consisted of having the subjects smell a filter paper strip that had been exposed in a 500-ml glass-stoppered bottle to an atmosphere saturated with each odorant at room temperature. The subjects rated each odorant on three 7-point scales: no-odorstrong, familiar-unfamiliar, and unpleasantpleasant. Subjects who indicated "no odor" to androstenone were dropped from subsequent analyses. Results The seven dependent variables (ratings of Paul and of self) were analyzed in a multivariate analysis of variance (MANOVA; Winer, 1971) with experimental groups and sex as two factors in a 4 (Group) x 2 (Sex) design. Twenty-two individuals who reported "no odor" to androstenone (13 men and 9 women) were dropped from this analysis. The MANOVA indicated a significant overall Group X Sex interaction, F(21, 465) = 1.56, p < .05. This indicated that the treatments had different effects for each sex. The main effect for group was significant at the .06 level, F(21, 465) = 1.53. No significant main effect for sex was found, F(7, 162) = 1.31, p = .25. Means for each variable are presented in Table 1.
Individual analyses of variance for the variables were performed within the MANOVA framework. Significant univariate effects were explored by pairwise comparisons according to the Student-Newman-Kuels a posteriori procedure (Winer, 1971). The analyses of variance for affect ratings of Paul, judgments of his handsomeness, sexual attractiveness, and strength showed no significant effects for group, sex, or Group X Sex interactions. There was a significant Group x Sex interaction for the activity-passivity rating of Paul, F(3,168) = 2.85, p < .04. A separate analysis of the men and women indicated a significant group difference for the men, F(3, 79) = 3.68, p < .01, but not for the women, F(3, 89) = 0.18, p = .96. For men, Paul was seen as more passive by the androstenone group (M = 4.07) and more active by the methyl anthranilate group (M = 5.65). These means differed from each other but not from the means for the other groups. A significant main effect on self-rated sexiness was found for group, F(3, 168) = 2.78, p < .04. Pairwise comparisons showed the androstenone group to be significantly lower on selfrated sexiness than the other groups, which were not different from one another. The androstenone group mean was 4.40 versus means around 5.50 for the other groups. Separate analyses for each sex indicated that most of the effect was contributed by the female subjects. The means for the male subjects across the groups were not significantly different, F(3, 79) = 0.59, p = .62, whereas for the female subjects the androstenone group had a mean of 3.85 which was significantly lower than the means for the other groups, F(3, 89) = 3.00, p < .04. Women in the androstenone group reported themselves less sexy. Discussion Although there appears to be an impact of androstenone on humans, that impact was not
Table 1 Means on the Ratings of Paul and of Self in Each Odor Condition No odor
Androstenone
Women Women Men (n = 28) (n = 24) (n = 17) (n = 20)
Methyl anthranilate
Men Rating (n = 20) Of Paul Affect 40.54 42.62 43.13 42.40 41.30 Handsome 4.46 4.70 4.40 4.50 4.65 Active 4.86 5.04 4.07 5.00 5.65 Sexually attractive 3.79 3.96 3.20 3.50 3.55 Strong 4.61 4.75 4.93 4.85 4.70 Of self Mood 91.96 89.67 91.67 97.15 93.55 5.25 5.13 3.85 5.90 Sexy 5.38 Note Subjects who reported "no odor" to androstenone are not included. Men
Women {n = 24)
Skatole Men Women (n = 20) (n = 25)
40.00 4.58 4.96 4.03 4.63
41.30 4.55 4.30 3.80 4 15
40.44 3.88 5.16 3.76 4.92
93.29 5.46
99.30 5 75
92.84 5.44
222
BRIEF COMMUNICATION
readily predictable from the hypothesis that it is a sex pheromone. In swine, androstenone induces estrous sows to immobilize in a matingstance display even when presented in aerosol form without a boar being present (Melrose et al., 1974). The effects in the present study indicated that androstenone had a differential impact on men and women, but the effects did not form a simple and coherent pattern related to sexual activity or excitement. In fact, the effects were somewhat contradictory to the sexual arousal hypothesis, with the female subjects reporting themselves to feel less sexy and the male subjects attributing passivity to the target male. It is conceivable that the subjects did not have the appropriate response available to them in the relatively sterile laboratory situation; that is, perhaps and androstenone has arousal properties, but the situation did not permit the typical responses to emerge. It is also possible that the stable form, androstenone, used in the present experiment is less effective than the more labile alcohol form, androstenol. References Benton, D. (1982). The influence of androstenol—a putative human pheromone—on mood throughout the menstrual cycle. Biological Psychology, 15, 249256. Black, S. L., & Biron, C. (1982). Androstenol as a human pheromone: No effect on perceived physical attractiveness. Behavioral and Neural Biology, 34, 326-330. Bond, A., & Lader, M. (1974). The use of analogue scales in rating subjective feelings British Journal of Medical Bulletin, 47, 211-218. Brooksbank, B. W. L., Brown, R., & Gustafsson, J. A. (1974). The detection of 5 alpha-androst-16-en-3alpha-ol in human male axillary sweat Expenentia, 30, 864-865.
Brooksbank, B. W. L., & Haslewood, G. A. D. (1961). The estimation of androst-16-en-3-alpha-ol in human urine: Partial synthesis of androstenol and of its beta-glucosiduronic acid. Biochemistry Journal, 80, 488-496. Claus, R., & Alsing, W. (1976). Occurrence of 5-alphaandrost-16-en-3-one, a boar pheromone, in man and its relationship to testosterone. Journal of Endocrinology, 68, 483-484. Comfort, A. (1971). Likelihood of human pheromones. Nature, 230, 432-433. Cowley, J. J., Johnson, A. J., & Brooksbank, B. W. L. (1977). The effect of two odorous compounds on performance in an assessment-of-people test. Psychoneuroendocrinology, 2, 159-172. Gower, D. B. (1972). 16-Unsaturated C19 steroids: A review of their chemistry, biochemistry, and possible physiological role. Journal of Steroid Biochemistry, 3, 45-103. Hurley, H. J., & Shelley, W. B. (1960). The human apocrine sweat gland in health and disease. Springfield, IL: Charles C Thomas. Kirk-Smith, M., Booth, D. A., Carroll, D., & Davies, P. (1978). Human social attitudes affected by androstenol. Research Communications in Psychobgy, Psychiatry, and Behavior, 3, 379-384. Melrose, D. R. (1972). Artificial insemination of pigs (U.S. Patent No. 3,681,490). Washington, DC: U.S. Government Printing Office. Melrose, D. R., Reed, H. C. B., & Patterson, R. L. S. (1974). Androgen steroids as an aid to the detection of oestrus in pig artificial insemination. British Veterinary Journal, 130, 61-67. Osgood, C. E., Suci, G J., & Tannenbaum, P H. (1957). The measurement of meaning. Urbana: University of Illinois Press. Russell, M. J. (1976). Human olfactory communication. Nature, 260, 520-522. Winer, B. J. (1971). Statistical principles in experimental design (2nd ed.). New York: McGraw-Hill.
Received July 19,1983 Revision received December 16, 1983 •
