Japanese Journal of Physiology, 52, 353–359, 2002

Human Olfactory Contrast Changes during the

Kyoko WATANABE*, Kana UMEZU*, and Takashi KURAHASHI*,†

* Department of Biophysical Engineering, Osaka University, Toyonaka, 560–8531 Japan; and † Precursory Research for Embryonic Science and Technology (PRESTO), Japan Science and Technology Corporation (JST), Toyonaka, 560–8531 Japan

Abstract: Several lines of studies have re- patch clamp experiment that uses the solitary ol- ported that olfactory perception is influenced by factory receptor cell. The results obtained from physical and hormonal conditions. In the present 18 trials (15 subjects) showed that olfactory con- study, we investigated changes of olfactory per- trast was significantly enhanced at the ovulatory ception during the menstrual cycle of the human. and/or menstrual phases. It is suggested that ol- Cyclopentadecanolide vapor was used and its factory contrast, which we defined as a new pa- perception intensity was measured by 6-point rameter, provides a useful tool in many kinds of category scale methods. We especially focused studies exploring the olfactory perceived ability. on a novel concept termed “olfactory contrast” [Japanese Journal of Physiology, 52, 353Ð359, that has been just very recently derived from the 2002]

Key words: sensory system, olfaction, receptor cell, sex , menstrual cycle.

Chemoreception is generally believed to be a very also confirmed the rhythmic change of the olfactory primary sensation, equipped in a wide variety of or- sensitivity during the menstrual cycle [5]. The detec- ganisms. In higher animals olfaction is tightly related tion thresholds of pentalide and varied to mating and feeding behaviors [1]. Therefore it has during the menstrual cycle [6]. In contrast, conflicting long been discussed that olfactory sensitivities are reports suggest that olfactory sensitivity is not rhyth- strongly dependent on the animal situation: Sensitivity mically changed throughout the menstrual cycle. becomes high when the animal seeks foods (see, e.g., Hummel et al. detected no changes in the olfactory Takagi [2]). Also, human olfaction has been thought sensitivity during the menstrual cycle when examin- to be influenced by body and external conditions ing it with phenylethyl alcohol, androstenone, and [3–8]. One amusing example is the relation between nicotine [7]. Furthermore, it was reported that thresh- the olfactory sensitivities and the menstrual cycles of old values against citral were uncorrelated with the women. menstrual cycle, but olfactory-related events obtained Several studies have investigated the olfactory sen- from the brain (CSERP: chemosensory event–related sitivity during the menstrual cycle, but the experimen- potentials) showed a remarkable change around the tal results express diverse variations among studies. It ovulatory phase [8]. was reported that women in ovulatory and premen- In the present study, we investigated the changes of strual phases could detect a lower concentration of olfactory perception during the menstrual cycle from Exaltolide, and during or just after the a different viewpoint. We carried out this study by fo- sensitivity became duller [3]. The olfactory sensitivity cusing on an index termed “olfactory contrast,” which is higher at than at menstruation when ex- has been just very recently derived from physiological amined with three involatile esters (pentadecalactone, experiments on the olfactory receptor cell [9]. Olfac- coumarin, and cinnamyl butylate) [4]. Doty et al. have tory contrast is defined as a slope of the dose–re-

Received on May 23, 2002; accepted on July 25, 2002 Correspondence should be addressed to: Takashi Kurahashi, Physiological Laboratory, Department of Frontier Biosciences, Osaka Univer- sity, Toyonaka, 560–8531 Japan. Tel and Fax: ϩ81–6–6850–6540, E-mail: [email protected]

Japanese Journal of Physiology Vol. 52, No. 4, 2002 353 K. WATANABE, K. UMEZU, and T. KURAHASHI sponse relation and therefore provides recognition in regard to temperature), we would sometimes read- ability for the changed intensity of odorant. In the pre- just the air-conditioning so make the subject more sent work, we obtained it from females with psy- comfortable. chophysical tests. Our purposes for this work are (1) Cyclopentadecanolide (CPD; oxacyclohexadecan- to determine if olfactory contrast is a useful index to 2-one, known as Exaltolide or Pentalide, donated from evaluate human sensation, (2) to determine if olfac- Soda Aromatic was used as an odorant). CPD solution tory contrast changes during the menstrual cycle, and was diluted to 0.1 g/ml by ethanol (99.5%, Katayama (3) to provide information regarding correlation be- Chemical Industries) 12 steps of concentration was tween the olfactory perception and change in the en- made by binary serial dilution. It was stocked in glass docrine factors. bottles and covered tightly at room temperature. The range of concentration prepared was 3 log. According MATERIALS AND METHODS to our preliminary experiments, we confirmed that the lowest concentration used in this study was lower than Subjects. By questionnaires and their basal the detection threshold. Before testing, both the exper- body temperature monitored for a month before the imenter and the subject washed their hands with a fra- investigation, we recruited 15 healthy oriental women grance-free soap. The odor solution (2 ␮l) was infil- (from among 17) as subjects who had stable menstrual trated into a filter paper 5 mm wide. After the com- cycles and who did not smoke, use perfume, or con- plete evaporation of ethanol vapor (40–60 s), the filter sume highly seasoned foods. The subjects included paper was presented to the subject. The stimulants people from different genetic origins: 11 Japanese, 2 were inhaled by their sniffing. Chinese, 1 Taiwanese, and 1 Singaporean. In the pre- Psychophysical procedure. The olfactory sent study, however, we obtained no systematic rela- perceived strength was rated by using a 6-point cate- tion between olfactory perception and their back- gory scale that was anchored with the phrases “no ground. The mean age was 22.8 years (rangeϭ19–32, smell” (leftmost word: zero) and “extremely strong” SDϭ4.1). All subjects were nonparous, and none were (rightmost word: five) at both ends (in Japanese). The taking oral contraceptives. The average length of men- odorants were presented with the ascending series. In strual cycle was 31.4Ϯ3.9 d. a lower concentration, a three-force choice method To avoid emotional artifact during the measure- was used in parallel. We odorized only one filter paper, ments, we did not inform the patients of the purpose then presented it with two nonodor (only ethanol) fil- of our experiment. Before the experiments, however, ter paper as the blank. The subject judged which filter protocols and species of smell (musk) were explained paper had an odorant, and she rated the intensity of so that the subjects completely agreed to paricipate in stimuli. However, if the subject made a wrong judg- the measurement. All subjects signed a letter of intent ment or couldn’t determine which paper had the odor, to establish voluntary consents. After the series of in- the intensity of concentration was judged to be zero. vestigations were completed, a detail of our study was When the subject could make a correct judgment for a explained individually to each subject. Throughout the sequential three presentations, the intensity values of experiments, we found no subject to be affected by a stimuli were recorded from the lowest concentration health problem caused by the experimental procedure. of those three presentations. In the present experi- Stimulus delivery. We asked subjects not to ment, we allowed subjects to use decimal numbers to use any fragrance on the day of the test. The tests increase the resolution for the results. This was were performed around noon (from 9 A.M. to 1 P.M.), needed especially in the present experiment because more than 2 h after breakfast and before lunch. During we had to obtain a slope of the dose–response rela- this 2-h interval, the subjects were not allowed to con- tionship. To avoid adaptation, filter papers were pre- sume any food or drink except for water and tea. The sented to the subjects at intervals of 60–180 s, and the time needed for one series of test was approximately subjects were allowed to smell the presented odor 30 min. Subjects who had more than 5 h of sleep par- with only one sniff. For a comparison, some experi- ticipated in the examination. Alcohol and spicy foods ments were performed on male subjects (nϭ3). Essen- were completely suppressed during the day before tially, the same protocols for the females were applied. tests. For all subjects, the tests were done in isolated The test was performed every other day during one rooms with the same environment. The condition of menstrual cycle. If a subject became unable to partici- each room was adjusted to a temperature of 23–27°C, pate (for example, a cold prevented smelling), we moisture 42–45%, and light 750–1,000 lx. But when a stopped the test until she recovered and started an- subject expressed uncomfortable feelings (especially other term of experiments over a single menstruation

354 Japanese Journal of Physiology Vol. 52, No. 4, 2002 Olfaction during the Menstrual Cycle

Fig. 1. (A) Relation between the concentration of the odorant and the perceived intensity. The horizon- tal axis is log concentration of odor (mM, at the filter paper). The vertical axis represents olfactory-perceived in- tensity rated by the category scale. A series of plots was fitted by a Sig- moidal function: yϭA1ϩ(A2ϪA1)/{1ϩ (log x0Ϫx p 10 ) }, where A1 is the bottom as- ymptote, A2 is the top asymptote, p is the Hill coefficient, and log x0 is the x value where y becomes (A1ϩA2)/2. (B) Derivative of the curve obtained in A. Note that the curve is monotonic, expressing a remarkable peak. The maximum value was used for the olfactory contrast value. These data were obtained from the subject 1 d after ovulation.

cycle. We also tried to use the LMS method [10, 11] RESULTS for some experiments (nϭ21). This method has been just recently becoming popular as a means of evaluat- Perceived intensities for CPD were obtained by the 6- ing perceptive intensity for human sensation. One big point category scale from 15 healthy women while advantage of this method is that we can exclude the their menstrual cycles were being monitored. Three upper limiting effect during the psychological tests. subjects were monitored for 2 cycles. As a conse- However, we failed to find a correct way to obtain a quence, the full length of data was obtained for 18 precise olfactory contrast because the maximum value menstrual cycles. varied day by day and between subjects. Figure 1 shows an example of the relationship be- During the test period, the subjects’ basal body tween odorant concentration and olfactory perception. temperatures were monitored every morning. The tim- The relation showed a remarkable subthreshold part at ing of ovulation was estimated both from the change low concentration and saturation at high concentra- in the and from the LH tion, indicating that the concentration range examined (luteinizing hormone) level within the urine measured here covered the putative perception range of the sub- by using a test kit (“dotest,” Rohto Pharmaceutical, ject. The relation could be fitted well by the single Osaka, Japan). Sigmoidal function. Analysis. We used a series of data sets without The olfactory contrast has been defined as a slope the first datum of the experiment for each subject, be- of the dose–response curve [9]. In other words, it ac- cause in the first test the subject were unable to prop- tually expresses how the small change of odorant erly judge the smell. strength can be recognized. In the present study, we The data were analyzed by using Origin 6.0J Pro- differentiated the concentration–perception curve fessional (OriginLab). The olfactory intensity values (Fig. 1B) and used a maximum value as an index for rated by subjects were plotted and approached by the the olfactory contrast. A derivative function of the Sigmoidal function as follows: concentration–perception relationship gave a monoto- Ϫ nic curve expressing one peak; thus we could obtain a yϭA ϩ(A ϪA )/{1ϩ10(log x0 x)p}, 1 2 1 specific point as a maximum value. where A1: bottom asymptote; A2: top asymptote; Figure 2 illustrates a series of raw data obtained log x0: the x value where y value becomes (A1ϩA2)/2; from a subject during her menstrual cycle. Through- p: Hill coefficient. The parameters x and y represent out the experiments, the data points were fitted well the log concentration of odor (mM) and perceived in- by the single Sigmoidal functions. It shows that the tensity, respectively. The olfactory contrast is obtained fundamental feature of the concentration–perception as a gradient of concentration–perception intensity relationships is constant throughout the days. How- curve. So we took the derivative of the function and ever, the slope of the curve was found to show a large got the maximum value as a definition for the olfac- deviation; therefore the dynamic range varied depend- tory contrast value of the day (see below). ing on the day. In Fig. 2 the curve was initially less steep and therefore expressed a wide dynamic range. As the days passed, the slope of the curves became

Japanese Journal of Physiology Vol. 52, No. 4, 2002 355 K. WATANABE, K. UMEZU, and T. KURAHASHI

Fig. 2. Change in concentrationÐper- ception intensity curve from a certain subject. The subject received tests every other day over one menstrual cycle, and we then obtained 16 data. Among them, 9 curves are represented here. The start- ing date of the menstruation was num- bered “day zero.” lut, ; mens, menstrual phase. The subject had a 29-d menstrual cycle in this experiment, and the ovulation was presumed by use of a kit to be 14 d before the start of menstru- ation (day Ϫ14).

Fig. 3. Four types of contrast changes during the course of sam- pling. Data points around the ovula- tory and menstruation periods were scanned to determine if they were sta- tistically significant in comparison with the other points. Initially, the mean and SD of olfactory contrast, except in the menstrual and ovulatory phases, were calculated. When the olfactory con- trast around menstruation and/or ovu- lation was larger than meanϩ2SD, the olfactory contrast value in this day was regarded as a significant point. Typi- cally, 4 data were represented. In A, olfactory contrast increased twice around the menstrual and ovulatory periods. B, once around only ovula- tion. C, once around only menstrua- tion. Type D showed no remarkable change. The vertical solid line repre- sents the day of ovulation. The period indicated by two dotted lines corre- sponds to the menstrual phase. A se- ries of plots was fitted with interpolat- ing polynomials performed by the sta- tistical software (see MATERIALS AND METHODS). steeper; the maximum was observed 4 d before men- three, contrast increased at both menstruation and struation started. After the maximum point, the slope ovulation (type A, nϭ5); around ovulation exclusively returned to being less steep. (type B, nϭ4); and menstrual phase alone (type C, It is very interesting that olfactory contrast obtained nϭ7). In the fourth, no change in contrast was recog- from the series of experiments showed a rhythmic nized in 2 out of 18 during the menstrual cycle (type change during the menstrual cycle (Fig. 3). Depend- D, nϭ2). We might think that these types could be ing on the period in which the olfactory contrast in- specified by individual persons, but this possibility is creased, we could divide the pattern into four types. In very unlikely. In fact, 2 out of 3 women who partici-

356 Japanese Journal of Physiology Vol. 52, No. 4, 2002 Olfaction during the Menstrual Cycle

Fig. 4. (A) Normalized contrasts around ovulation. The day of ovula- tion was set at zero (indicated by a dotted line marked “ovu”). Olfactory contrast values obtained from 9 sub- jects around ovulation were normal- ized between subjects. The peak val- ues of individual data were referred as 1. A series of them was then smoothed by interpolating polynomi- als. The filled squares show the mean of these daily smoothed values. The smooth line represents the least square fitting of data points with the

Gaussian function: yϭy0ϩA exp{Ϫ(xϪ 2 –— x0)/2␴ }/␴√2␲ ; where y0: base line; A: area; ␴: (SD); and x0: mean. MeanϭϪ0.82; SDϭ3.46. (B) Normal- ized contrasts around menstruation. The starting day of menstruation was set at zero (indicated by a dotted line marked “mens”). The filled squares show the mean of the daily smoothed values obtained from the data of 12 subjects around menstruation in the same way as in A. The smooth line represents the least square fitting of data points with the Gaussian function. MeanϭϪ1.08; SDϭ10.7. pated in the test for more than 2 menstrual cycles ex- recently, physiological experiments on the olfactory pressed different tendencies of olfactory contrast system have revealed that the olfactory recognition change in these terms (one showed type A and C and seems to be achieved by various factors other than a the other type A and B). So it is more natural to think detection threshold. One example is the olfactory con- that the data were grouped into multitypes because of trast that has been derived from the patch clamp ex- the experimental variations. Instead, the result ob- periment using olfactory receptor cells [9]. Differently tained here would suggest that olfactory contrast in- from previous concepts using sensitivities, olfactory creases twice generally both at the menstrual and the contrast is an index to detect the change in odorant ovulatory phases. In contrast, in men we observed no strength because it represents a slope of the dose–re- rhythmic change in olfactory contrast (nϭ3); the con- sponse relation. A large contrast means that a human trast changed irregularly, but we do not know why. can obviously recognize the difference between the To further analyze the contrast changes observed weak and strong stimuli; weak sensation becomes during the menstrual and ovulatory periods, we nor- weaker and strong becomes stronger. In other words, malized and averaged the curves obtained from sub- we saw temporal dynamics (how the change in the jects (Fig. 4). For both periods, the relation could be odor concentration can be received as sensation) while fitted by Gaussian curves expressing peak values past experiments were performed that focused on the around the day of ovulation (Fig. 4A) and the starting absolute sensitivities. It has long been pointed out that day of menstruation (Fig. 4B). One remarkable feature olfactory sensation expresses strong adaptation (or fa- of the distribution was that the standard deviation tigue; see Takagi [2]). In such a system, temporal in- (SD) was smaller for the fitting around ovulation (SDϭ formation would be especially important. Olfactory 3.46) than around menstruation (SDϭ10.7). This mat- contrast may detect these dynamic changes. ter is further discussed below. Since the menstrual cycle is precisely regulated by the endocrine systems, it is highly likely that the DISCUSSION change in the olfactory contrast is controlled by the change in the hormonal balance. Unfortunately, the In the present work we obtained the olfactory contrast present experiment is not suitable to make a conclu- from females with 6-point category scale methods. It sion in this regard. However, there are several physio- was found that the olfactory contrast expresses rhyth- logical experiments suggesting the relation between mic changes during the menstrual cycle. It is sug- the olfactory sensation and the endocrine factors. It is gested that the olfactory contrast we defined as a new reported that adrenaline causes a change in the steep- parameter provides a useful tool in many kinds of ness of the input–output relation of the isolated olfac- studies exploring the olfactory perceived ability. tory receptor cell via protein kinase A (PKA) [9, 12]. In the past, psychological studies have focused on Dopamine also affects high-voltage–activated inward how sensitive the subjects are. Olfactory sensitivities rectification (Ih) of the olfactory cell, which leads to were defined as an inverse of the threshold [3–8]. But the modulation of cell activity [13]. Furthermore,

Japanese Journal of Physiology Vol. 52, No. 4, 2002 357 K. WATANABE, K. UMEZU, and T. KURAHASHI GnRH modulates TTX-sensitive inward current and been raised as candidates for human : Ca2ϩ-dependent outward current in the olfactory cell ⌬4,16-androstadien-3-one (ER-670) and 1,3,5,(10), [14]. GnRH secretory neurons exist in lamina propria, 16-estratetraen-3-ol (ER-830) [22]. These compounds and olfactory receptor cells seem to have GnRH re- affected the emotional states of women and men, and ceptors [15]. Coupled with the result obtained in the androstadienone modulated the women’s mood states, present work, it is understandable to think that the en- even when subjects couldn’t recognize its odor [23]. docrine factors affect some points underlying the ol- In regard to one menstrual cycle, the olfactory con- factory signaling pathway. trast rises twice around ovulation and menstruation. A woman’s menstrual cycle is regulated by several This result may be puzzling because these periods sit- peptide and steroid hormones. Just before ovulation, uate opposite extremes in terms of the possibility for there is a transient surge for the multiple hormones the pregnancy. As described, many steroids with (e.g., LHRH or LH). In contrast, in the premenstrual musklike odors may have generally a possible func- phase sex steroids (progesterone and estrogen) fall rel- tion as a human pheromone. There is a possibility that atively gradually, which may correspond to long-last- the quality of perception to CPD is changed during ing effects. We found in the present study that the ol- these periods; e.g., at ovulation the musklike odor be- factory contrast increases in both the ovulatory and comes pleasant, but at menstruation it becomes un- the menstrual phases. Essentially, the same result was pleasant. The experiment done by Hummel et al. has obtained [16]. Moreover, it is interesting to see in the reported that androstenone perception in terms of he- present study that the time period in which the olfac- donic scale varies during the menstrual cycle [7]. Our tory contrast became high was longer in the menstrual results further add new information that this quality period than that observed in the ovulatory phase. If change can be enhanced by the increase in the olfac- the change of olfactory contrast is related to the tran- tory contrast. In other words, a human female who is sient increase of LHRH or LH concentration within around ovulation recognizes it as a pleasant smell and the body, the effects of these substances do not last for intends to make contact with the potential partner. In long. Similarly, the gradual changes of sex steroid an opposite period (which has less chance for preg- hormones during the premenstrual period may be re- nancy), the chemosignal alarming the existence of a sponsible for the long-lasting increase of olfactory human male may make the female unpleasant. It is contrast. These possibilities will remain under specu- worthwhile to investigate further the hedonic change lation, until the entire molecular systems controlling for other compounds having a musklike smell, includ- the olfactory contrasts are completely understood. ing CPD. The odorant used in the present study is cyclopen- We have defined the olfactory contrast as a new pa- tadecanolide, known by the trade names Exaltolide rameter in the study of an olfactory system. It was de- and Pentalide, which generally causes a musklike sen- rived from an electrophysiological experiment, but the sation in humans. The olfactory contrast was en- result in the psychological experiment was extremely hanced around the ovulation and menstruation peri- clear. It suggests that this new concept is sufficiently ods. These restricted enhancements may have some useful in many kinds of studies about an interaction relations with chemical communications between hu- between the olfactory and the endogenous systems. mans. In general, musk is well known as a sex pheromone emitted from the male musk deer [17]. We thank M. Fukuoka and Y. Koizumi (Shiseido) for their There is a possibility that a musklike sensation is gen- valuable advice, M. Moriyama for her establishing an early erally used for chemical communications in variable stage of the experimental protocol, and J. Fukuda for her animals, including the human. technical assistance. This work was supported by grants Several lines of evidence suggest that the human from the Human Frontier Science Program Organization (HFSPO), Ministry of Education, Culture, Sports, Science also utilizes chemical communications via pheromones. and Technology (MEXT), and Precursory Research for Em- For instance, it has been reported that the menstrual bryonic Science and Technology (PREST) (Japan Science cycles of women living in the same room are synchro- and Technology Corporation: JST). nized [18]. This synchrony is mediated by the com- pounds in the female axillary sweat [19]. For men- REFERENCES strual synchrony, it is necessary to have the strong ␣ ␣ 1. Doty RL: Odor-guided behavior in mammals. Experien- sensitivity for 5 -androst-16-en-3 -ol [20]. It is fur- tia 42: 257–271, 1986 ther reported that exposure to male axillary secretions 2. Takagi SF: Human Olfaction, University of Tokyo Press, repairs an aberrant length of female menstrual cycle Tokyo, 1989 to normal [21]. Furthermore, two compounds have 3. Vierling JS and Rock J: Variations in olfactory sensitiv-

358 Japanese Journal of Physiology Vol. 52, No. 4, 2002 Olfaction during the Menstrual Cycle

ity to Exaltolide during the menstrual cycle. J. Appl 13. Vargas G and Lucero MT: Dopamine modulates in- Physiol 22: 311–315, 1967 wardly rectifying hyperpolarization-activated current

4. Mair RG, Bouffard JA, Engen T, and Morton TH: Olfac- (Ih) in cultured rat olfactory receptor neurons. J Neuro- tory sensitivity during the menstrual cycle. Sens physiol 81: 149–158, 1999 Processes 2: 90–98, 1987 14. Eisthen HL, Delay RJ, Wirsig-Wiechmann CR, and 5. Doty RL, Snyder PJ, Huggins GR, and Lowry LD: En- Dionne VE: Neuromodulatory effects of gonadotropin docrine, cardiovascular, and psychological correlates releasing hormone on olfactory receptor neurons. J of olfactory sensitivity changes during the human men- Neurosci 20: 3947–3955, 2000 strual cycle. J Comp Physiol Psychol 95: 45–60, 1981 15. Wirsig-Wiechmann CR: Function of gonadotropin-re- 6. Ayabe-Kanamura S, Takashima Y, and Saito S: leasing hormone in olfaction. Keio J Med 50 (2): 81–85, Changes in detection thresholds of androstenol and 2001 pentalide during the menstrual cycle. In: Olfaction and 16. Wilson EO and Bossert WH: Chemical communication Taste XI, ed. Kurihara K, Suzuki N, and Ogawa H, among animals. Recent Progr Hormone Res 19: Springer-Verlag, Tokyo, pp 334–335, 1994 673–716, 1963 7. Hummel T, Gollisch R, Wildt G, and Kobal G: Changes 17. McClintock MK: Menstrual synchrony and suppression. in olfactory perception during the menstrual cycle. Ex- 229: 244–245, 1971 perientia 47: 712–715, 1991 18. Stern K and McClintock MK: Regulation of ovulation by 8. Pause BM, Sojka B, Krauel K, Fehm-Wolfsdorf G, and human pheromones. Nature 392: 177–179, 1998 Ferstl R: Olfactory information processing during the 19. Morofushi M, Shinohara K, Funabashi T, and Kimura F: course of the menstrual cycle. Biol Psychol 44: 31–54, Positive relationship between menstrual synchrony and 1996 ability to smell 5␣-androst-16-en-3␣-ol. Chem Senses 9. Kawai F, Kurahashi T, and Kaneko A: Adrenaline en- 25: 407–411, 2000 hances odorant contrast by modulating signal encod- 20. Cutler WB, Preti G, Krieger A, Huggins GR, Garcia CR, ing in olfactory receptor cells. Nat Neurosci 2: 133– and Lawley HJ: Human axillary secretions influence 138, 1999 women’s menstrual cycles: the role of donor extract 10. Green BG, Shaffer GS, and Gilmore MM: Derivation from men. Horm Behav 20: 463–473, 1986 and evaluation of a semantic scale of oral sensation 21. Monti-Bloch L and Grosser BI: Effect of putative magnitude with apparent ratio properties. Chem Senses pheromones on the electrical activity of the human 18: 683–702, 1993 vomeronasal organ and olfactory epithelium. J Steroid 11. Green BG, Dalton P, Cowart B, Shaffer G, Rankin K, Biochem Mol Biol 39: 573–582, 1991 and Higgins J: Evaluating the ‘Labeled Magnitude 22. Jacob S and McClintock MK: Psychological state and Scale’ for measuring sensations of taste and smell. mood effects of steroidal chemosignals in women and Chem Senses 21: 323–334, 1996 men. Horm Behav 37: 57–78, 2000 12. Kawai F, Kurahashi T, and Kaneko A: T-type Ca2ϩ chan- 23. Jacob S, Hayreh DJS, and McClintock MK: Context-de- nel lowers the threshold of spike generation in the newt pendent effects of steroid chemosignals on human olfactory receptor cell. J Gen Physiol 108: 525–535, physiology and mood. Physiol Behav 74: 15–27, 2001 1996

Japanese Journal of Physiology Vol. 52, No. 4, 2002 359