Human ovulatory phase-increasing odors cause positive emotions and stress-suppressive effects in males
Subject areas
Introduction
In many species, female body odors reflect their reproductive status, and odors around the time of ovulation attract males, resulting in behavioral and/or physiological changes. For instance, male hamsters are attracted to estrous vaginal discharge, facilitating their copulatory behavior.1 In some non-human primates, ovulatory odors from females induce inspection or exploratory behavior in males.2,3 Common marmosets exhibit an increase in male testosterone levels in response to the ovulatory scent emitted from females.4 Olfactory cues play a crucial role in successful mating in many animals, prompting the intriguing question of whether humans utilize olfaction for male-female interaction.
1.
O’Connell, R.J. ∙ Singer, A.G. ∙ Macrides, F. ...
Responses of the male golden hamster to mixtures of odorants identified from vaginal discharge
Behav. Biol. 1978; 24:244-255
2.
Cerda-Molina, A.L. ∙ Hernández-López, L. ∙ Rojas-Maya, S. ...
Male-induced sociosexual behavior by vaginal secretions in Macaca arctoides
Int. J. Primatol. 2006; 27:791-807
3.
Clarke, P.M.R. ∙ Barrett, L. ∙ Henzi, S.P.
What role do olfactory cues play in chacma baboon mating?
Am. J. Primatol. 2009; 71:493-502
4.
Ziegler, T.E. ∙ Schultz-Darken, N.J. ∙ Scott, J.J. ...
Neuroendocrine response to female ovulatory odors depends upon social condition in male common marmosets, Callithrix jacchus
Horm. Behav. 2005; 47:56-64
In humans, several studies have suggested that female body odor signals periods of high fertility. For instance, vaginal secretions collected during the preovulatory and ovulatory phases are rated as more pleasant than those from the menstrual, early luteal, or late luteal phases.5 Additionally, axillary odors during the fertile phase are perceived by men as more pleasant, sexy, or attractive compared to those during the menstrual and luteal phases.6,7 Another study indicated that high-fertility scents collected from the armpit were chosen as more attractive than low-fertility scents.8 The fact that hormonal contraception results in the loss of these effects suggests that attractiveness is dependent on female hormones.9 It has been suggested that women’s behavior and physical attractiveness may serve as identifiable cues to fertility,10 and that men’s preferences are likely to increase for physical cues associated with high estrogen levels.11,12 However, in a recent study by Zetzsche et al., no difference in pleasantness was observed in axillary samples collected overnight for 12 h at donor’s home during the menstrual cycle.13 In the same study, chemical analysis using thermal desorption (TD) tubes showed no correlation between axillary odor composition and fertility states. Therefore, while female axillary odors may convey fertility-related information, some inconsistencies still exist.
5.
Doty, R.L. ∙ Ford, M. ∙ Preti, G. ...
Changes in the intensity and pleasantness of human vaginal odors during the menstrual cycle
Science. 1975; 190:1316-1318
6.
Singh, D. ∙ Bronstad, P.M.
Female body odour is a potential cue to ovulation
Proc. Biol. Sci. 2001; 268:797-801
7.
Havlíček, J. ∙ Dvořáková, R. ∙ Bartoš, L. ...
Non-advertized does not mean concealed: Body odour changes across the human menstrual cycle
Ethology. 2006; 112:81-90
8.
Gildersleeve, K.A. ∙ Haselton, M.G. ∙ Larson, C.M. ...
Body odor attractiveness as a cue of impending ovulation in women: Evidence from a study using hormone-confirmed ovulation
Horm. Behav. 2012; 61:157-166
9.
Haselton, M.G. ∙ Gildersleeve, K.
Can men detect ovulation?
Curr. Dir. Psychol. Sci. 2011; 20:87-92
10.
Gangestad, S.W. ∙ Thornhill, R.
Human oestrus
Proc. Biol. Sci. 2008; 275:991-1000
11.
Havliček, J. ∙ Cobey, K.D. ∙ Barrett, L. ...
The spandrels of Santa Barbara? A new perspective on the peri-ovulation paradigm
Behav. Ecol. 2015; 26:1249-1260
12.
Kuukasjärvi, S. ∙ Eriksson, C.P. ∙ Koskela, E. ...
Attractiveness of women’s body odors over the menstrual cycle: The role of oral contraceptives and receiver sex
Behav. Ecol. 2004; 15:579-584
13.
Zetzsche, M. ∙ Weiß, B.M. ∙ Kücklich, M. ...
Combined perceptual and chemical analyses show no compelling evidence for ovulatory cycle shifts in women’s axillary odour
Proc. Biol. Sci. 2024; 291:20232712
In this study, we explore various methods of axillary odor collection and, using a more careful and rigorous approach, aim to investigate whether body odor communicates information about the menstrual cycle. First, we extended previous sensory evaluation studies to investigate not only pleasantness of odors but also perceived odor quality across the menstrual cycle. Second, we aimed to identify chemical profile of axillary odor at each phase of the menstrual cycle using gas chromatography-mass spectrometry (GC/MS). We also aim to examine whether ovulatory phase odors induce positive emotions and psychological effects in males. The overarching goal is to address the question of whether olfactory communication exists to facilitate relationships between males and females in humans.
Results
Axillary odor during the ovulatory phase is more pleasant than that during the menstrual phase
To assess the intensity and pleasantness of axillary odor across different phases of the menstrual cycle, axillary odors were collected on gauze from 21 female donors at each of the four stages of their menstrual cycle (i.e., M[enstrual], F[ollicular], O[vulatory], L[uteal]-phases; a total of 84 odor samples) (Figure S1A). Each phase of the menstrual cycle was determined by the luteinizing hormone (LH) surge and body temperature, and the amounts of estradiol and progesterone in saliva were evaluated for some donors (Figure S1B). Given that previous reports suggest differences in body odor under stress from normal body odor,14,15 we used the state-trait anxiety inventory (STAI) score to assess the anxiety state of the donors at the time of sample collection. No significant difference in anxiety state, as indicated by the STAI score, was observed among the four phases (see STAR Methods). Out of the 84 odor samples, each male rater (n = 21) provided odor ratings (intensity and pleasantness) on samples from 10 to 11 donors (40–44 samples in total). Linear mixed model (LMM) analyses revealed that axillary odor during the O-phase was rated significantly more pleasant than those during M and L-phases (M vs. O: t(846) = 3.1, p = 0.006, L vs. O: t(846) = 2.4, p = 0.048, Figure 1 and Table S1). There was also a tendency for pleasantness to be higher compared to the F-phase (F vs. O: t(846) = 2.2, p = 0.077, Figure 1 and Table S1). When excluding raters with lower olfactory ability (Methods), the O-phase showed significantly higher pleasantness compared to the M, F, and L-phases (M vs. O: t(636) = 2.8, p = 0.017, F vs. O: t(636) = 2.5, p = 0.043, L vs. O: t(636) = 2.4, p = 0.045, Table S1). These findings suggest that the O-phase tends to have a pleasant odor, with this tendency being more pronounced among raters with higher olfactory abilities. No statistically significant differences in intensity were observed among the phases.
14.
de Groot, J.H.B. ∙ Kirk, P.A. ∙ Gottfried, J.A.
Titrating the Smell of Fear: Initial Evidence for Dose-Invariant Behavioral, Physiological, and Neural Responses
Psychol. Sci. 2021; 32:558-572
15.
Mujica-Parodi, L.R. ∙ Strey, H.H. ∙ Frederick, B. ...
Chemosensory Cues to Conspecific Emotional Stress Activate Amygdala in Humans
PLoS One. 2009; 4, e6415
Next, the axillary odor quality at each of the four phases was investigated using the check-all-that-apply (CATA) sensory profiling method. Raters with low odor identification quality were excluded because they may not accurately assess odor quality (STAR Methods). The selection rate of 15 descriptors, based on quantitative descriptive analysis (QDA, see STAR Methods), chosen by male raters (n = 16) to describe the axillary samples is presented in Table 1. Among the 15 descriptors, with a focus on those with higher frequency of selection, “vinegary odor” was checked 24–28 times in M, F, and L-phases, and 17 times in the O-phase, with a lower number of selections in the O-phase. In contrast, “fragrant odor” was examined 37–40 times in the M-, F-, and L-phases, and 53 times in the O-phase, with a higher number of selections in the O-phase. All these results suggest that axillary odors during the O-phase are relatively pleasant.
| Descriptor | Male raters, n = 16 | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Descriptor selection rate (%) | Number of selections | Selection rate (%) | Cochran’s Q test (p value) | Effect size (η2Q) | ||||||||
| M | F | O | L | M | F | O | L | |||||
| Pleasant | Fragrant | 25% | 37 | 37 | 53 | 40 | 22% | 22% | 32% | 24% | 0.044 | 0.016 |
| Woody/Pencil-like | 11% | 16 | 23 | 15 | 17 | 23% | 32% | 21% | 24% | 0.41 | 0.0056 | |
| Sweet | 6% | 11 | 8 | 12 | 9 | 28% | 20% | 30% | 23% | 0.73 | 0.0026 | |
| Milky | 5% | 11 | 7 | 11 | 8 | 30% | 19% | 30% | 22% | 0.65 | 0.0032 | |
| Citrus | 4% | 5 | 6 | 13 | 6 | 17% | 20% | 43% | 20% | 0.14 | 0.011 | |
| Grassy/Green | 4% | 6 | 4 | 10 | 6 | 23% | 15% | 38% | 23% | 0.31 | 0.0070 | |
| Earthy | 3% | 4 | 7 | 4 | 2 | 24% | 41% | 24% | 12% | 0.22 | 0.0086 | |
| Yogurt | 2% | 4 | 0 | 8 | 3 | 27% | 0% | 53% | 20% | 0.027 | 0.018 | |
| Unpleasant | Sweaty | 29% | 51 | 55 | 45 | 46 | 26% | 28% | 23% | 23% | 0.40 | 0.0058 |
| Vinegary | 14% | 26 | 24 | 17 | 28 | 27% | 25% | 18% | 29% | 0.17 | 0.010 | |
| Sebaceous | 13% | 21 | 23 | 21 | 25 | 23% | 26% | 23% | 28% | 0.82 | 0.0019 | |
| Damp clothing | 8% | 15 | 13 | 12 | 14 | 28% | 24% | 22% | 26% | 0.90 | 0.0019 | |
| Dusty | 6% | 7 | 10 | 11 | 11 | 18% | 26% | 28% | 28% | 0.72 | 0.0027 | |
| Musty | 5% | 13 | 10 | 3 | 7 | 39% | 30% | 9% | 21% | 0.048 | 0.016 | |
| Stinky | 5% | 9 | 8 | 7 | 8 | 28% | 25% | 22% | 25% | 0.96 | 0.00061 | |
Table 1
Sensory evaluation of the odor quality of female axillary odor during the menstrual cycle
Evaluation of the sensory profile of female axillary odors at each phase of the menstrual cycle using the check-all-that-apply (CATA) method (male raters, n = 16). Descriptor selection rate indicates “number of descriptor selections/number of raters (16 males) × number of axillary odor samples evaluated by each male rater (40 or 44 samples).” Number of selections indicates the “number of descriptor selections for each phase samples.” Selection rate indicates “number of selections/number of total selection from M to L phase.” Cochran’s Q-test was used to evaluate the differences in axillary odors between the four different phases.
Chemical composition analysis of axillary odor identified increasing or decreasing compounds for each phase
To explore the chemical profiles of female axillary odors during the M-, F-, O-, and L-phases, we analyzed volatiles emitted from the axilla during the menstrual cycle in 21 female donors, the same donors as in the previous section, using GC/MS. Initially, we compared four different adsorbents, including sterile gauze, polydimethylsiloxane (PDMS) membranes, Monotrap, and Twister, for a female (Table S2 and Method for detailed sample preparation and analytical methods for each adsorbent). We found that the PDMS membrane adsorbed the largest number of compounds (Figure S2) and that its close fit to the armpit allowed for the efficient adsorption of low-volatile compounds such as long-chain and unsaturated fatty acids secreted from the axillary surface (see Tables S3 and S4). Therefore, we used the PDMS membrane to sample axillary odors.
Subsequently, 84 axillary odor samples (four-phase samples from 21 female donors) collected using the PDMS membrane were subjected to analysis using GC/MS combined with a thermal desorption system (Figure S3). Among numerous peaks identified following GC/MS analyses, we focused on compounds that were structurally identified using an authentic sample or that existed in more than five donors, at least in one phase. The resulting 98 peaks were further analyzed (average occurrence rate in all phases of all compounds = 71%, see STAR Methods). To mitigate the effect of inter-individual data variability, the peak area values were standardized using a robust standardization method. To discern changes in the secretion of the 98 compounds during the menstrual cycle, Ward’s method cluster analysis was performed using the mean of the robust Z scores for each time period. The 98 compounds were arranged so that compounds with similar changes between the four phases were located in close proximity (Figure 2). This analysis reveals several groups of compounds that tend to decrease during the M-phase (top cluster) or increase during the O- and F-phases (top or middle cluster, respectively).
Figure 2 The secretion levels of the 98 compounds in each phase of the menstrual cycle
To identify increasing or decreasing compounds at each phase, we searched for compounds that significantly increased above the median of all phases for each phase. A one-sample Wilcoxon signed-rank test (one-tailed) was performed to determine whether the amount of the compound in each phase was greater than 0 (median of the robust Z score). Out of the 98 compounds, seven significantly increased during the O-phase (Figures 2 and 3; Tables S3 and S4). Three compounds were identified as (E)-geranylacetone (ID#49) (Z = 2.1, p = 0.016, r = 0.47), tetradecanoic acid (ID#86) (Z = 1.7, p = 0.043, r = 0.38), and (Z)-9-hexadecenoic acid (ID#94) (Z = 2.1, p = 0.019, r = 0.45) (Figure S4). In the M-phase, four compounds significantly increased, and the identified compounds were octanal (ID#15) (Z = 1.9, p = 0.031, r = 0.41) and 1-tetradecanol (ID#65) (Z = 1.7, p = 0.045, r = 0.37) (Tables S3 and S4). The remaining two compounds with small p value (like ID#2) were unknown (not identified) but representative m/z was obtained (Table S10). In the F-phase, 26 compounds significantly increased, and of the 15 compounds that could be identified, the three with the lowest p values were 1,3-butanediol (ID41) (Z = 2.6, p = 0.005, r = 0.56), octadecane (ID#45) (Z = 2.7, p = 0.0032, r = 0.60), and 1-pentadecanol (ID#72) (Z = 2.6, p = 0.0043, r = 0.58) (Figures 2 and 3; Tables S3 and S4). In the L-phase, four compounds significantly increased, and the identified compounds were hydrogen sulfide (ID#1) (Z = 2.1, p = 0.02, r = 0.45), 2-pyrrolidone (ID#58) (Z = 2.3, p = 0.01, r = 0.51), and hexahydro-2H-azepin-2-one (ID#68) (Z = 1.9, p = 0.029, r = 0.41) (Tables S3 and S4).
Figure 3 Variability in volatile compounds that significantly increased during ovulatory or follicular phases
Men perceive the odor of ovulatory phase-increasing compounds as more pleasant and comforting
In conjunction with the sensory evaluation (refer to Figure 1) and GC/MS analysis (refer to Figures 2 and 3), (E)-geranylacetone, tetradecanoic acid, and (Z)-9-hexadecenoic acid emerged as potential compounds contributing to the reduction of the unpleasantness of axillary odor during the O-phase. To probe how the increase in these three compounds during the O-phase influences the quality of axillary odor for men, we examined changes in odor quality when the odor-O (odor-O: three O-phase-increasing compounds) was introduced to a basal axillary odor, closely mimicking axillary odor16 (odor-A: a basal axillary odor), consistently present in the female axilla throughout the menstrual cycle with little or no difference (refer to Star Methods and Table S6). Sensory evaluations for four odor stimuli (Ctrl, no odor; odor-O, three O-phase-increasing compounds; odor-A, basal axillary odor; odor-A+O, a mixture of odor-A and odor-O; Table S6) were conducted on 72 male raters. The evaluation terms included the 15 descriptors in Table 1, along with intensity, pleasantness, preference, feminine/masculine, relax, and arousal/excitement. Figure 4 illustrates the results for 13 terms where significant differences were observed in the sensory evaluation (all results are available in Table S7). Odor-O was weakly perceptible in intensity (t(64) = 5.4, q < 0.0005, r = 0.56), and the overall odor quality was deemed pleasant with fragrant (t(64) = 3.9, q < 0.005, r = 0.44), citrus (t(64) = 3.1, q < 0.05, r = 0.36), and relaxing (t(60) = 3.6, q < 0.005, r = 0.42). The addition of odor-O to odor-A led to significantly higher ratings for pleasantness (t(71) = 10.1, q < 0.0005, r = 0.77) and preference (t(71) = 9.5, q < 0.0005, r = 0.75), a more feminine rating for feminine/masculine (t(69) = 4.9, q < 0.0005, r = 0.51), and lower ratings for intensity (t(71) = 4.0, q < 0.0005, r = 0.43) (odor-A+O in Figure 4A). Fragrant (t(71) = 2.9, q < 0.05, r = 0.32), citrus (t(71) = 2.6, q < 0.05, r = 0.29), woody (t(71) = 3.7, q < 0.005, r = 0.40), sweet (t(71) = 2.3, q < 0.05, r = 0.27), and relaxing (t(65) = 3.6, q < 0.005, r = 0.41) received significantly higher ratings, whereas sweaty (t(71) = 5.0, q < 0.0005, r = 0.51), damp clothing (t(71) = 2.9, q < 0.05, r = 0.33), sebaceous (t(71) = 2.3, q < 0.05, r = 0.26), and stinky (t(71) = 4.8, q < 0.0005, r = 0.50) received significantly lower ratings in odor-A+O (Table S12). This observation aligns with the finding that male raters associated fragrant and citrus scents when rating the ovulatory female axillary odor (Table 1).
16.
Hara, T. ∙ Kyuka, A. ∙ Shimizu, H.
Butane-2,3-dione: the Key Contributor to Axillary and Foot Odor Associated with an Acidic Note
Chem. Biodivers. 2015; 12:248-258
Figure 4 Three ovulatory phase-increasing compounds positively affected the unpleasant quality of basal axillary odor
We posited a hypothesis that the three O-phase-increasing compounds (odor-O) are pivotal female axillary odors evoking pleasant and positive emotional changes in males. Thus, we tested the effects of adding each of the odor-O compounds to odor-A for 37 male raters in terms of pleasantness. A single odor-O compound significantly reduced the unpleasantness of odor-A, although the combined effect of the odor-O mix was more pronounced than that of a single compound (Figure 4B, A + O vs. A + G: t(36) = 2.1, p = 0.016, d = 0.48; A + O vs. A + T: t(36) = 1.8, p = 0.023, d = 0.45; and A + O vs. A + H: t(36) = 2.1, p = 0.012, d = 0.48, odor-G, T, and H are (E)-geranylacetone, tetradecanoic acid, and (Z)-9-hexadecenoic acid, respectively; Tables S6, S12, and Figure S5). To evaluate whether the same unpleasantness-suppression effect could be observed in compounds that increase during phases other than the O-phase, a similar experiment was conducted using the three compounds that increase during the F-phase (odor-F). Odor-F also demonstrated the effect of reducing the unpleasantness of the basal axillary odor; however, the effect was not as significant as that of odor-O (Figure 4C, A + O vs. A + F: q = 6.2, p < 0.005, d = 0.70).
Psychological and physiological changes induced by the O-phase-increasing compounds on men
To investigate the psychological and physiological effects of the O-phase-increasing odor, we conducted an experiment with the timeline depicted in Figure 5A. In essence, male raters (n = 19) entered a private experimental room to avoid encountering other participants, sat calmly in a chair, and were initially asked to perform a baseline multiple mood scale (MMS, Table S8) (a type of psychological test) assessment and saliva collection (base condition). In session 1–3, the raters then wore a headset with a microphone, in which the experimenter applied an odor to the microphone portion without informing the raters, and rated female face images (impression rating) and completed MMS, followed by saliva collection immediately after removing the headset. The 19 male rater had counter-balanced exposure to Ctrl or odor-O in session 1 and 2. Of the 19 male raters, 9 were exposed to odor-A, and 10 were exposed to odor-A+O in session 3.
Figure 5 Physiological and psychological effects of the ovulatory compounds on males
The analysis of MMS in sessions 1 and 2 (Figure 5A) revealed that “hostility” was significantly lower in the odor-O condition compared to the Ctrl condition (no odor) (q = 3.7, p = 0.05, d = 0.66) (Figure 5B, top panel, second from the left; Table S14). The impression rating tasks during sessions showed a significant increase in “boredom” (q = 4.2, p = 0.025, d = 0.90) and a decrease in “concentration” (q = 4.5, p = 0.016, d = 0.84) in the Ctrl condition compared to the base condition (before session 1), with no significant difference observed between the odor-O and base conditions (Figure 5B, top panel; Table S14). These findings suggest that odor-O has a positive psychological effect that might decrease the feeling of “hostility.”
In session 3, a comparison of the scales for each odor condition and the corresponding control condition demonstrated a significant decrease in “relaxation” in the odor-A condition (t(6) = 2.5, p = 0.047, d = 0.69). However, this decrease was mitigated by adding odor-O into odor-A (odor-A+O condition) (Figure 5B, middle and lower panel, forth column from the right; Table S14). These results suggest that odor-O has a positive psychological effect that increases the feeling of relaxation when added to the basal axillary odor.
To delve into the relaxation effect of odor-O, we focused on α-amylase in saliva, known to be elevated in malodor stress.17 We observed a significant negative correlation between α-amylase and the “relaxation” score in MMS at all time points (p = 0.0005) (base, session 1 to 3, n = 19 × 4, Figure 5C), indicating that the amount of α-amylase in saliva increases as the “relaxation” score decreases. Comparing α-amylase levels in the base, Ctrl (no odor), and odor-O conditions revealed no significant differences (Figure 5D, left). Comparing the Ctrl(A) (Ctrl data from raters who smelled odor-A in session 3, n = 7) and odor-A conditions showed a significant increase in α-amylase under odor-A (t(6) = 2.9, p = 0.027, d = 0.94, Figure 5D, middle, Table S14), whereas this increase was suppressed by adding odor-O (odor-A+O condition) (Ctrl data from raters who smelled A + O in session 3, n = 10). These results demonstrate that odor-O has the effect of suppressing the increase in salivary amylase caused by a baseline axillary odor, correlating with the level of relaxation.
17.
Hirasawa, Y. ∙ Shirasu, M. ∙ Okamoto, M. ...
Subjective unpleasantness of malodors induces a stress response
Psychoneuroendocrinology. 2019; 106:206-215
Effects of the O-phase-increasing compounds on male evaluation of female faces
In the same experimental timeline (Figure 5A), we examined the impact of the O-phase-increasing odor on the impressions male raters had for female faces. Images of eight female faces were prepared (50% smiling, Figure 5E), and raters were instructed to rate their impression of each face in four categories (categories: beautiful, elegant and intellectual, want to spend time with, want to keep gazing at). Because all four categories of 8 images exhibited similar rating trends under Ctrl conditions (Figure 5E, purple, blue, green, and orange lines), we calculated the average ratings for these categories (Figure 5E, red line) and employed it as the “attractiveness” in the subsequent analysis. Face images were divided into two groups based on attractiveness under Ctrl conditions: a low-rated group (4 images) and a high-rated group (4 images). It was observed that the attractiveness of faces in the low-rated group decreased with odor-A and significantly increased with odor-A+O compared to the no-odor condition (F (1, 8) = 9.2, p = 0.016, η2 = 0.004, Figure 5F, left, Table S14). When the differences between odor-A and no-odor (odor-A - Ctrl(A) in Figure 5F) and odor-A+O and no-odor (A + O - Ctrl(A + O) in Figure 5F) were calculated and compared, it was evident that the attractiveness of low-rated images was significantly higher under odor-A+O than under odor-A (F (1, 16) = 15, p = 0.0013, η2 = 0.1, Figure 5F, left). In the faces of the high-rated group, there were no significant differences between the odor stimuli condition and the no-odor state (Figure 5F, right). These results suggest that when odor-O was added to the basal axillary odor, a positive behavioral effect was observed in terms of impression ratings.
Discussion
This study investigated the role of olfactory cues in human intersexual communication, with a focus on the ovulatory phase of the menstrual cycle, and identified ovulatory phase-increasing odorants that elicited positive emotional, physiological, and behavioral effects, facilitating male-female interaction.
Key points to consider when analyzing body odors are comprehensiveness and reproducibility in a well-controlled condition. Conventional research on axillary odor often relies on a single collection or analytical method, resulting in the detection of a limited number of compounds.18,19,20,21 In this study, we compared four collection adsorbents and GC/MS pretreatment methods. Based on the diversity of collected compounds and consistency with previous findings,22 we adopted a method combining PDMS and a thermal desorption system with GC/MS analysis. This method proved to be the most efficient in detecting the largest number of compounds among those tested and could reproducibly detect compounds from an average of more than 70% of the donors. However, since competitive effects are known to occur with adsorbents, the composition of the compounds adsorbed on PDMS cannot be said to be identical to that of the original axillary odor. Through careful optimization of the analytical method allowed for the identification of compounds during the menstrual cycle that affect male-female interaction.
18.
Zeng, X.N. ∙ Leyden, J.J. ∙ Spielman, A.I. ...
Analysis of characteristic human female axillary odors: Qualitative comparison to males
J. Chem. Ecol. 1996; 22:237-257
19.
Natsch, A. ∙ Derrer, S. ∙ Flachsmann, F. ...
A broad diversity of volatile carboxylic acids, released by a bacterial aminoacylase from axilla secretions, as candicate molecules for the determination of human-body odor type
Chem. Biodivers. 2006; 3:1-20
20.
Gallagher, M. ∙ Wysocki, C.J. ∙ Leyden, J.J. ...
Analyses of volatile organic compounds from human skin
Br. J. Dermatol. 2008; 159:780-791
21.
Curran, A.M. ∙ Rabin, S.I. ∙ Prada, P.A. ...
Comparison of the volatile organic compounds present in human odor using SPME-GC/MS
J. Chem. Ecol. 2005; 31:1607-1619
22.
Dormont, L. ∙ Bessière, J.M. ∙ Cohuet, A.
Human Skin Volatiles: A Review
J. Chem. Ecol. 2013; 39:569-578
The current knowledge on the identified three O-phase-increasing compounds is as follows: (1) (E)-geranylacetone: it is produced by oxidation of squalene and emits a floral or green odor quality and is among the prominently recognized substances in humans.23,24,25,26,27 (2) Tetradecanoic acid: it has an oxidized sebum odor quality and is detected in skin, saliva,28 breast milk, and amniotic fluid. Its primary production pathway is believed to be triglyceride hydrolysis by skin bacteria lipase,29 or it may be derived from blood owing to its ubiquitous presence in cells and plasma.30,31 Human amniotic fluid and breast milk contain this odorant to which newborns exhibit a suckling response.32,33 (3) (Z)-9-hexadecenoic acid: commonly termed palmitoleic acid, it is nearly odorless. It is derived from palmitic acid and is presumed to be a precursor to (E)-2-nonenal, a representative odor of “aging odor,” which is generated through its breakdown by skin bacteria.34,35 Although the three compounds aforementioned increased during the ovulatory phase, it is important to note that this study was designed as an exploratory investigation. To minimize type II error, multiple testing correction was not applied, which unfortunately increased susceptibility to type I error.
23.
Leal, H.M. ∙ Hwang, J.K. ∙ Tan, K. ...
Attraction of Culex mosquitoes to aldehydes from human emanations
Sci. Rep. 2017; 7:17965
24.
Zhao, Z. ∙ Zung, J.L. ∙ Hinze, A. ...
Mosquito brains encode unique features of human odour to drive host seeking
Nature. 2022; 605:706-712
25.
Drabińska, N. ∙ Flynn, C. ∙ Ratcliffe, N. ...
A literature survey of all volatiles from healthy human breath and bodily fluids: the human volatilome
J. Breath Res. 2021; 15, 034001
26.
Syed, Z. ∙ Leal, W.S.
Acute olfactory response of Culex mosquitoes to a human- and bird-derived attractant
Proc. Natl. Acad. Sci. USA. 2009; 106:18803-18808
27.
Owsienko, D. ∙ Goppelt, L. ∙ Hierl, K. ...
Body odor samples from infants and post-pubertal children differ in their volatile profiles
Commun. Chem. 2024; 7:53
28.
de Lacy Costello, B. ∙ Amann, A. ∙ Al-Kateb, H. ...
A review of the volatiles from the healthy human body
J. Breath Res. 2014; 8, 014001
29.
Nicolaides, N.
Skin Lipids: Their Biochemical Uniqueness
Science. 1974; 186:19-26
30.
Aparicio, E. ∙ Martín-Grau, C. ∙ Hernández-Martinez, C. ...
Changes in fatty acid levels (saturated, monounsaturated and polyunsaturated) during pregnancy
BMC Pregnancy Childbirth. 2021; 21, 778
31.
Aktas, M. ∙ Elmastas, M. ∙ Ozcicek, F. ...
Erythrocyte Membrane Fatty Acid Composition in Premenopausal Patients with Iron Deficiency Anemia
J. Oleo Sci. 2016; 65:225-231
32.
Contreras, C.M. ∙ Gutiérrez-García, A.G. ∙ Mendoza-López, R. ...
Amniotic fluid elicits appetitive responses in human newborns: Fatty acids and appetitive responses
Dev. Psychobiol. 2013; 55:221-231
33.
Gutiérrez-García, A.G. ∙ Contreras, C.M. ∙ Díaz-Marte, C.
Myristic acid in amniotic fluid produces appetitive responses in human newborns
Early Hum. Dev. 2017; 115:32-37
34.
Haze, S. ∙ Gozu, Y. ∙ Nakamura, S. ...
2-Nonenal newly found in human body odor tends to increase with aging
J. Invest. Dermatol. 2001; 116:520-524
35.
Porter, N.A.
Mechanisms for the Autoxidation of Polyunsaturated Lipids
Acc. Chem. Res. 1986; 19:262-268
Changes in axillary volatile compounds within the menstrual cycle may be attributed to fluctuations in hormone concentrations, including estradiol and progesterone, within the cycle. In particular, estradiol receptors are expressed in many tissues and are believed to be involved in systemic metabolism and sweat gland secretory activity. Studies have revealed that amino acids and their derivatives in blood tend to increase during the menstrual phase and decrease during the luteal phase, whereas phospholipids and other substances in blood increase during the follicular phase and decrease during the luteal phase.36 In this study, we demonstrate that long-chain fatty acids, including tetradecanoic acid and (Z)-9-hexadecenoic acid, increase from the follicular phase to around ovulation and decrease during the luteal phase. This possibly reflects the degradation of blood-derived lipids on the skin surface, which increase during the follicular phase. Moreover, several compounds also increased during the follicular phase, and some of these tended to further increase during ovulation, suggesting the strong effects of estradiol on the variability of amounts of each compound during the menstrual cycle.
36.
Draper, C.F. ∙ Duisters, K. ∙ Weger, B. ...
Menstrual cycle rhythmicity: metabolic patterns in healthy women
Sci. Rep. 2018; 8:14568
This study revealed that the odor of the three O-phase-increasing compounds induces a sense of comfort and relaxation in men. Furthermore, when the odor of the basal axillary odor was combined with the three O-phase increasing compounds, it not only enhanced the attractiveness of low-attractive-level female faces but also suppressed the increase in salivary amylase, an indicator of stress-related bad odor, in comparison to the odor of the basal axillary scent alone. It has been reported that pleasant essential oil (like rose) enhance facial attractiveness,37 but it is noteworthy that such effects may also be present in components of body odor that fluctuate with the menstrual cycle. If other compounds that increase during ovulation that could not be identified in this study can be clarified, the function of ovulatory axillary odor will be elucidated in more detail. Previous reports have suggested that when others smell body odor collected in high-stress situations, they experience anxiety from ambiguous facial expressions,38 and anxiety and nervousness increase,39 whereas when others smell body odor collected in a happy state, a facial expression and perceptual processing style indicative of happiness were shown to be induced.40 Additionally, women who smelled their partner’s body odor experienced less stress.41,42 Body odors from newborns has been reported to activate reward-related cerebral areas in their mothers43 and mother’s odor did not cause typical brain responses to fearful stimuli in their child.44 These findings are implying that while body odor conveys negative emotions, it also acts in a positive direction and inhibit negative emotions.
37.
Seubert, J. ∙ Gregory, K.M. ∙ Chamberland, J. ...
Odor valence linearly modulates attractiveness, but not age assessment, of invariant facial features in a memory-based rating task
PLoS One. 2014; 9, e98347
38.
Zhou, W. ∙ Chen, D.
Fear-Related Chemosignals Modulate Recognition of Fear in Ambiguous Facial Expressions
Psychol. Sci. 2009; 20:177-183
39.
Albrecht, J. ∙ Demmel, M. ∙ Schöpf, V. ...
Smelling chemosensory signals of males in anxious versus nonanxious condition increases state anxiety of female subjects
Chem. Senses. 2011; 36:19-27
40.
de Groot, J.H.B. ∙ Smeets, M.A.M. ∙ Rowson, M.J. ...
A sniff of happiness
Psychol. Sci. 2015; 26:684-700
41.
Granqvist, P. ∙ Vestbrant, K. ∙ Döllinger, L. ...
The scent of security: Odor of romantic partner alters subjective discomfort and autonomic stress responses in an adult attachment-dependent manner
Physiol. Behav. 2019; 198:144-150
42.
Hofer, M.K. ∙ Collins, H.K. ∙ Whillans, A.V. ...
Olfactory cues from romantic partners and strangers influence women’s responses to stress
J. Pers. Soc. Psychol. 2018; 114:1-9
43.
Lundström, J.N. ∙ Mathe, A. ∙ Schaal, B. ...
Maternal status regulates cortical responses to the body odor of newborns
Front. Psychol. 2013; 4:597
44.
Schaal, B. ∙ Saxton, T.K. ∙ Loos, H. ...
Olfaction scaffolds the developing human from neonate to adolescent and beyond
Philos. Trans. R. Soc. Lond. B Biol. Sci. 2020; 375, 20190261
In addition to body odor, several reports indicate various physical changes in females during the menstrual cycle. The voice during the ovulation period is perceived as the most attractive to men,45 and it has been demonstrated that the frequency of the voice also increases during the fertile phase.46,47 Photographs of a female face during the ovulation phase are perceived as more desirable than other times.48 The redness of the face has been reported to decrease after the first day of menstruation and increases toward ovulation, although it is not detectable by the naked eye.49 It appears that physical features that enhance appeal to the opposite sex are heightened during the ovulation phase. Together with our findings, although females with highly attractive faces appear to be less affected by olfactory information, other females may use the ovulatory phase-increasing odor and other physical features synergistically to attract the interest of the opposite sex during the reproductive period.
45.
Pipitone, R.N. ∙ Gallup, G.G.
Women’s voice attractiveness varies across the menstrual cycle
Evol. Hum. Behav. 2008; 29:268-274
46.
Bryant, G.A. ∙ Haselton, M.G.
Vocal cues of ovulation in human females
Biol. Lett. 2009; 5:12-15
47.
Pavela Banai, I.
Voice in different phases of menstrual cycle among naturally cycling women and users of hormonal contraceptives
PLoS One. 2017; 12, e0183462
48.
Roberts, S.C. ∙ Havlicek, J. ∙ Flegr, J. ...
Female facial attractiveness increases during the fertile phase of the menstrual cycle
Proc. Biol. Sci. 2004; 271:S270-S272
49.
Burriss, R.P. ∙ Troscianko, J. ∙ Lovell, P.G. ...
Changes in women’s facial skin color over the ovulatory cycle are not detectable by the human visual system
PLoS One. 2015; 10, e0130093
In other mammals, ovulatory body odors function as a releaser pheromone, causing apparent behavioral changes such as attraction and providing information regarding the reproductive stage of female individuals. However, a uniform intrinsic attractive behavior is not observed between males and females in highly rationalized human society. Our study highlights an intriguing hypothesis that the role of ovulatory odors might have shifted in the course of evolution in humans from the classical behavior-releasing pheromone to a communicative tool that positively affects emotions in males and induces relaxation and a positive impression toward females, thereby facilitating male-female interaction. Such a role could function as a discreet but effective cross-sex olfactory signal in a highly developed human society governed by intellect rather than instinct.
Limitations of the study
It is known that the composition of axillary odor is partially influenced by genetic factors,50 and that the perception of body odor is also affected by genetic background51,52 and cultural context.53,54 Therefore, by conducting sensory evaluations and chemical analysis of axillary odor across diverse ethnic groups beyond the Japanese population, while accounting for the effects of genetic polymorphisms associated with odor perception, it should be possible to perform a detailed investigation into the factors that influence the evaluation of pleasantness or unpleasantness of menstrual cycle-related body odor.
50.
Toyoda, Y. ∙ Sakurai, A. ∙ Mitani, Y. ...
Earwax, osmidrosis, and breast cancer: why does one SNP (538G> A) in the human ABC transporter ABCC11 gene determine earwax type?
FASEB J. 2009; 23:2001-2013
51.
Keller, A. ∙ Zhuang, H. ∙ Chi, Q. ...
Genetic variation in a human odorant receptor alters odour perception
Nature. 2007; 449:468-472
52.
Menashe, I. ∙ Abaffy, T. ∙ Hasin, Y. ...
Genetic elucidation of human hyperosmia to isovaleric acid
PLoS Biol. 2007; 5, e284
53.
Ferdenzi, C. ∙ Razafindrazaka, H. ∙ Baldovini, N. ...
Influence of gender and culture on the perception of acidic compounds of human body odor
Physiol. Behav. 2019; 210, 112561
54.
Schleidt, M. ∙ Hold, B. ∙ Attili, G.
A cross-cultural study on the attitude towards personal odors
J. Chem. Ecol. 1981; 7:19-31
Resource availability
Lead contact
Further information and requests for resources should be directed to and will be fulfilled by the lead contact, Kazushige Touhara (ktouhara@g.ecc.u-tokyo.ac.jp).
Material availability
This study did not generate new unique reagents.
Data and code availability
•
GC/MS raw data supporting the current study will not be deposited in a public repository because we are still analyzing the data for projects not reported in this study; however, these data will be available from the corresponding author upon reasonable request.
•
We used publicly available code for the analyses.
Acknowledgments
We thank all Touhara members for help in conducting experiment. This work was supported by the ERATO Touhara Chemosensory Signal Project to K.T. (JPMJER1202), the JST Mirai program to K.T. (JPMJMI17DC and JPMJMI19D1), and JSPS KAKENHI grants (18K14651 and 22K06418 to M.S.; 21K13546 to Y.H.; 18K02477, 18H04998, and 21H05808 to M.O.; and JP18H05267 and JP23H05410 to K.T.), and Grant for Women Scientists in Challenging Research by Japan Society for Bioscience, Biotechnology, and Agrochemistry (JSBBA) to M.S.
Author contributions
N.O., M.S., and K.T. designed research; N.O., M.S., Y.O., Y.H., M.O., and R.K. performed research; M.S., N.O., H.T., and K.T. analyzed data; M.S., N.O., and K.T. wrote the paper.
Declaration of interests
The authors declare that they have filed a patent application related to the content of this manuscript (PCT/JP2024/024888).
STAR★Methods
Key resources table
| REAGENT or RESOURCE | SOURCE | IDENTIFIER |
|---|---|---|
| Chemicals, peptides, and recombinant proteins | ||
| Geranylacetone | Tokyo Chemical Industry Co., Ltd. | Cat#G0236 |
| Tetradecanoic acid | Tokyo Chemical Industry Co., Ltd. | Cat#M0476 |
| (Z)-9-hexadecenoic acid | Tokyo Chemical Industry Co., Ltd. | Cat#H0072 |
| Acetic acid | FUJIFILM Wako Pure Chemical Co. | Cat#017-00251 |
| Propanoic acid | Tokyo Chemical Industry Co., Ltd. | Cat#P0500 |
| Butanoic acid | Tokyo Chemical Industry Co., Ltd. | Cat#B0754 |
| 2-methylpropanoic acid | Tokyo Chemical Industry Co., Ltd. | Cat#I0103 |
| 3-methylbutanoic acid | Tokyo Chemical Industry Co., Ltd. | Cat#M0182 |
| Octanoic acid | Kanto Chemical Co., Inc. | Cat#07132-33 |
| Decanoic acid | Tokyo Chemical Industry Co., Ltd. | Cat#D0017 |
| Software and algorithms | ||
| IBM SPSS version 27 | IBM Co. | https://www.ibm.com/support/pages/downloading-ibm-spss-statistics-27 |
| GraphPad Prism Ver9.5.1 | GraphPad Software | https://www.graphpad.com/updates |
| GCMS solution software | Shimadzu | https://www.shimadzu.com/an/products/gas-chromatograph-mass-spectrometry/gc-ms-software/gcmssolution/index.html |
| NIST 17 | National Institute of Standards and Technology | https://chemdata.nist.gov/dokuwiki/doku.php?id=chemdata:nist17 |
| Wiley Registry of Mass Spectral Data, 9th edition | Wiley | https://sciencesolutions.wiley.com/solutions/technique/gc-ms/wiley-registry-of-mass-spectral-data/ |
| Other | ||
| Open Essence Kit | FUJIFILM Wako Pure Chemical Corporation | https://labchem-wako.fujifilm.com/us/category/00368.html |
| PDMS membranes | Goodfellow Cambridge Limited | Cat#751-624-16 |
| Gas Chromatography Mass Spectrometer -TQ8030 | Shimadzu | GCMS-TQ8030 |
| A thermal desorption system OPTIC4 | GL Science | OPTIC-4 |
| Salivary Estradiol ELISA Kit | Salimetrics | https://salimetrics.com/assay-kit/salivary-estradiol-elisa-kit/ |
| Salivary Progesterone ELISA Kit | Salimetrics | https://salimetrics.com/assay-kit/salivary-progesterone-elisa-kit/ |
| Salivary α-Amylase Kinetic Enzyme Assay Kit | Salimetrics | https://salimetrics.com/assay-kit/salivary-alpha-amylase-enzymatic-assay/ |
| One Step Ovulation Urine Test Strip | Doctor’s Choice | https://blerdology.co/hairannkennsayaku/he-20-ovltnc.html |
Experimental model and study participant details
Ethics
The study received approval from the Ethics Committee of the University of Tokyo and adhered to the principles of the Declaration of Helsinki. Informed written consent was obtained from all participants before conducting the experiments.
Female axillary odor donors
A total of 29 females participated as odor donors, providing axillary odors during each phase of their menstrual cycles. Inclusion criteria required regular menstrual cycles (averaging between 25 and 30 days), overall good health without medication use (including oral hormonal contraceptives), absence of pregnancy, no gynecological diseases, chronic conditions, inflammations of the axillary surface, or smoking habits. Axillary odors were collected during the menstrual cycle (menstrual, follicular, ovulatory, and luteal phases). Eight females were excluded based on the following criteria: (i) an unstable menstrual cycle during the study, (ii) poor physical condition, (iii) scheduling conflicts, and (iv) a negative ovulation test result. Therefore, 21 females (age: mean ± SD = 23.7 ± 2.9 years) who provided samples from all four phases were included for further analysis. The donor profiles are shown in Table S9. Donors were instructed to abstain from sleeping with or having sexual intercourse with another person, consuming garlic, spicy or herb-infused foods, and alcohol from the day before each collection day until the end of the collection period. Additionally, they were asked to refrain from using antiperspirants, perfumes, scented deodorants, and scented soap powder on each collection day until the end of the collection period. All female donors were required to shave their axillary hair before the odor collection day. In this way, we try our best to prevent sampling of exogenous compounds, but it is fair to say that it is difficult to exclude 100%. Therefore, we carefully considered a possibility of exogenous origin after we identified compounds.
Odor raters
Sensory evaluation of female axillary odor
Twenty-one raters (males: n = 21, age: mean ± SD = 24 ± 2.9 years) were recruited for evaluating axillary odor samples. Raters were heterosexual, single, free of respiratory, brain and nervous system, mental, and chronic diseases, not taking anabolic steroids, and nonsmokers. Raters were informed that they would be evaluating various odors, including female-related odors. In this study, the influence of sex and gender was not evaluated.
Odor evaluation using O-phase-increasing odor
Seventy-two single men (age: mean ± SD = 29 ± 5.4 years, the score of Open Essence Kit (OPE score): 10 ± 1.2) were recruited as odor raters. All raters were heterosexual, did not suffer from respiratory, brain and nervous system, mental, and chronic diseases, did not take anabolic steroids, and were nonsmokers. The ability to identify odors was measured using the Open Essence Kit, and all raters scored 8 or higher, indicating no problem in identifying odor quality.
In this study, the influence of sex and gender was not evaluated.
Method details
Menstrual cycle assessment
The identification of menstrual phases relied on multiple methods, including LH surge assessment using LH test kits (One Step Ovulation Urine Test Strip, Doctor’s Choice, USA), basal body temperature (BBT) measurements (n = 21), donor’s self-reported onset and end dates of menstrual bleeding, and salivary estradiol and progesterone measurements (n = 12, for donors participating in the later period).
Donors provided information on the onset of their last menstrual period, the length of the menstrual phase, and the usual cycle length before axillary odor collection. This was used to estimate the length of their cycle and the onset of their next menstrual period. During the collection period, donors were also asked to report the first and last dates of their menstrual bleeding during the collection period via email and axillary odor was collected during their menstrual phase. After menstruation, LH test kits were used twice daily (morning and evening) to detect the LH surge. When positive results was prompted, donors were instructed to notify the laboratory and axillary odor was collected within one day. BBT was measured using a thermometer (A basal thermometer M-172L, OMRON, Japan) immediately upon waking and recorded in a BBT management application (WOMANCOM, https://kisotaion.web-db.ws/system/servlet/womancom.user, Japan) during the collection period. Additionally, saliva samples (1 mL) were collected from 12 donors at the time of axillary odor collection, and the levels of female hormones (estradiol and progesterone) were measured using enzyme-linked immunosorbent assay kits (SALIVARY 17β-ESTRADIOL ENZYME IMMUNOASSAY KIT, SALIVARY PROGESTERONE ENZYME IMMUNOASSAY KIT, Salimetrics, USA). As saliva sampling for hormone analysis began with the latter half of the 21 participants, hormone data are presented for 12 donors. In addition to the previously mentioned periods, axillary odor was collected during the period after menstruation ended and before LH tests turned positive, as well as during the period after a positive LH test turned negative and before the onset of menstruation. Samples collected during these periods were classified as follicular, ovulation or luteal phase samples, based on the results of BBT.
Stress related emotional state assessment in axillary odor sampling
We used STAI scores to determine the anxiety of the donors at the time of sample collection. Stress-related emotional state was assessed using the anxiety state scale of STAI-State (STAI-S), which consists of 20 questions. Anxiety is a negative emotional state, characterized by the reaction of an organism to stressors.55 STAI-S has been reported to be a sensitive tool to evaluate the increase of anxiety in response to various stressors.56 STAI-S scores ranged from 20 (low anxiety) to 80 (high anxiety) and a score of 42 or higher indicates the presence of anxiety symptoms. The mean STAI-S scores for each phase were as follows: M-phase = 32.6 ± 5.48 points, F-phase = 33.7 ± 9.16 points, O-phase = 34.9 ± 10.3 points, and L-phase = 34.9 ± 7.72 points. No significant difference in anxiety state was observed among the four phases (p = 0.423, Friedman’s test).
55.
Spielberger, C. ∙ Gorsuch, R. ∙ Lushene, R. ...
Consulting Psychologists Press, 1983
56.
Rossi, V. ∙ Pourtois, G.
Transient state-dependent fluctuations in anxiety measured using STAI, POMS, PANAS or VAS: a comparative review
Anxiety Stress Coping. 2012; 25:603-645
Collection of axillary odor for sensory evaluation
Before odor collection, the surface of each axilla underwent cleaning with a moist hot towel and 70% ethanol using sterile gauze to minimize potential confounding factors, especially from cosmetic products. Four sterile gauzes (5 cm × 5 cm) (Oosaki Medical Corp., Japan) were placed on both axillae, secured with baby skin surgical tape (3M Japan, Japan) for 4 h, with the order of gauze changing every hour. After removal, eight gauzes were immediately frozen at −80°C in a 50-mL glass vial (NICHIDENRIKA-GLASS CO., LTD., Japan) until sensory evaluation. Donors changed into clothes (UNIQLO, Japan) washed with fragrance-free detergent and remained at rest in a room with a temperature of 26°C and constant humidity during odor collection.
Procedure for sensory evaluation of female axillary odor
Sensory evaluations occurred twice daily at 10:00 a.m. and 2:00 p.m. Gauze samples were thawed 2 h before each session. Two of eight pieces from each donor were made into a set, secured with a double clip, and placed in a glass vial. Each rater evaluated axillary odor samples from four menstrual phases from 10 or 11 different female donors (40 or 44 axillary odor samples) and one control sample (a new sterile gauze). Each rater was seated in a chair in front of a desk in a well-ventilated, temperature-controlled experimental room. The raters were asked to fill out the odor questionnaire immediately after sniffing each sample. The odor questionnaire consisted of 22 questions (described in the next section) related to the sensory characteristics of the odor samples. The order of gauze samples was counterbalanced, and the evaluations took 60–90 min, with short breaks allowed every 12 samples.
All raters were required to refrain from eating odor-producing foods and drinking alcoholic beverages for 24 h before the testing because these activities may affect olfactory performance. They also refrained from using perfume, scented body cream, hand lotion, lipstick, and scented deodorant on the day of the evaluation.
Odor questionnaire for axillary odor
The questionnaire consists of two-halves. The first half assesses odor intensity (0 = not perceived to +6 = extremely intense) and pleasantness (−4 = extremely unpleasant to +4 = extremely pleasant). In the second half, raters use a CATA format57 to rate the odor quality of axillary samples using 20 descriptors (Sweaty, Vinegary, Fragrant, Sebaceous, Woody/Pencil-like, Damp clothing, Musty, Sweet, Earthy, Stinky, Grassy/Green, Citrus, Yogurt, Milky, Spicy, Sulfuric, Grapefruit peel-like, Dusty, Fishy, Meaty). Descriptor selection was based on the QDA method (details in the “generation of descriptive terms” section). Descriptors with a selection rate below 2% were excluded from subsequent analyses (Table 1).
57.
Adams, J. ∙ Williams, A. ∙ Lancaster, B. ...
Advantages and uses of check-all-that-apply response compared to traditional scaling of attributes for salty snacks
7th Pangborn Sensory Science Symposium 12–16 August Minneapolis USA
2007
For the CATA analysis, data from raters scoring 8 or higher on the Open Essence Kit (OPE) (FUJIFILM Wako Pure Chemical Corporation, Japan) were used to rigorously evaluate odor quality (single men: n = 16, age: mean ± SD = 23 ± 2.2). The Open Essence Kit measured odor identification ability, an enhanced version of OSIT-J,58 widely used in Japan, involves 12 odors familiar to the Japanese population.59 Raters chose 1 of 6 alternatives for each odor: 4 odor names, including a correct name, “not detected,” and “unknown.” Scores were calculated based on the total number of correct answers for the 12 odors. Scores range from 0 to 12, with scores of 8 or higher indicating satisfactory odor identification ability for all age groups.60
58.
Nishida, K. ∙ Kobayashi, M. ∙ Ogihara, H. ...
Clinical Usefulness of Smell Identification Test Card : Open Essence
Nihon Jibiinkoka Gakkai Kaiho. 2010; 113:751-757
59.
Suzuki, M. ∙ Hashimoto, M. ∙ Yoshioka, M. ...
The odor stick identification test for Japanese differentiates Parkinson’s disease from multiple system atrophy and progressive supra nuclear palsy
BMC Neurol. 2011; 11:157
60.
Fujio, H. ∙ Inokuchi, G. ∙ Fukuda, Y. ...
A New Screening Tool, An Odor Identification Test Card to Detect Age―related Changes in Olfaction Hisami
Nippon Jibiinkoka Gakkai Kaiho. 2018; 121:38-43
Generation of descriptive terms
We selected 20 descriptors using the following procedure.
(1)
First, the sensory terms were generated via a pilot test comprising an open-ended questionnaire that was administered to 15 raters (male: n = 8, female: n = 7). They evaluated 17 samples of axillary odor during the menstrual cycle (menstrual phase: n = 4, follicular phase: n = 4, ovulatory phase: n = 6, luteal phase: n = 3). The raters sniffed the axillary odor samples in Teflon squeeze bottles (Big Boy, Japan) and immediately described the quality of the odor after sniffing each sample. The list of descriptors consisted of 103 terms. Of the 103 terms, those used only once were omitted.
(2)
The next procedure was conducted by five experts (male: n = 2, female: n = 3), involved in sensory evaluation and analytical chemistry, to summarize the terms of the descriptor list. When several terms suggested a synonym, the most common word was selected. After eliminating redundancies and controversial terms, the descriptor list had a total of 45 terms.
(3)
To ensure that the 45 terms were sufficient to evaluate axillary odors, we conducted a pilot evaluation. Five raters assessed 21 odor samples (menstrual phase: n = 4, follicular phase: n = 4, ovulatory phase: n = 5, luteal phase: n = 8) and rated the 45 terms on a seven-point scale (0 = not perceived to +6 = extremely perceived). We retained characteristic descriptors that were likely to be important to discriminate between samples, as well as descriptors that were commonly perceived in most samples and were considered important in assessing axillary odors. Furthermore, we re-confirmed the definitions of terms that were believed to have multiple definitions and could cause individual perception-related differences. Terms with high-frequency correlations were consolidated. Thus, the final selection resulted in 20 descriptors.
GCMS analysis
Four different sampling techniques and associated GC/MS analysis
To determine the appropriate sampling method for the analysis of volatile organic compounds (VOCs) in axillary odors, we compared four different sampling methods. These included (1) sterile 100% cotton gauze (Oosaki Medical Corp., Japan), which is widely used for body odor collection18,19; (2) PDMS membranes (Cat. No.: 751-624-16, Goodfellow Cambridge Limited) that are suitable for capturing body odor VOCs due to their high adsorption capacity and flexibility61,62; (3) Monotrap (GL Sciences, Japan); and (4) Twister (GERSTEL Inc., USA) adsorbents, which are widely used for odor analysis and are easy to handle.63
18.
Zeng, X.N. ∙ Leyden, J.J. ∙ Spielman, A.I. ...
Analysis of characteristic human female axillary odors: Qualitative comparison to males
J. Chem. Ecol. 1996; 22:237-257
19.
Natsch, A. ∙ Derrer, S. ∙ Flachsmann, F. ...
A broad diversity of volatile carboxylic acids, released by a bacterial aminoacylase from axilla secretions, as candicate molecules for the determination of human-body odor type
Chem. Biodivers. 2006; 3:1-20
61.
Stevens, D. ∙ Cornmell, R. ∙ Taylor, D. ...
Spatial variations in the microbial community structure and diversity of the human foot is associated with the production of odorous volatiles
FEMS Microbiol. Ecol. 2015; 91:1-11
62.
Riazanskaia, S. ∙ Blackburn, G. ∙ Harker, M. ...
The analytical utility of thermally desorbed polydimethylsilicone membranes for in-vivo sampling of volatile organic compounds in and on human skin
Analyst. 2008; 133:1020-1027
63.
Soini, H.A. ∙ Bruce, K.E. ∙ Wiesler, D. ...
Stir Bar Sorptive Extraction: A New Quantitative and Comprehensive Sampling Technique for Determination of Chemical Signal Profiles from Biological Media
J. Chem. Ecol. 2005; 31:377-392
To compare each method, axillary odors were obtained from a female (laboratory member who met the inclusion/exclusion criteria in the method of “female axillary odor donors,” age = 27) axilla using all four methods on the same day. The PDMS membranes were first placed on the axillae surface, covered with PET film cut from Flek-Sampler (OMI ODOR-AIR SERVICE Corporation, Japan), and secured to the skin using waterproof films (New Tac Kasei, Japan) for 2 h (refer to “collection of axillary odors to compare the four phases of menstruation”). Subsequently, sterile gauze (5 cm × 5 cm) was placed on the surface of the axillae and held in place for 2 h using baby skin surgical tape. Monotrap or Twister was placed on the surface of the axillae and covered with a plate (TissueTec Cryomold No. 2, Sakura Finetech Inc Japan.) and secured with waterproof film for 2 h. Following removal, PDMS membranes, Monotrap, and Twister samples were transferred to a 2-mL glass vial (Thermo Fisher Scientific Inc., USA) and immediately frozen at −30°C. The gauze samples were immediately frozen at −80°C in a 50-mL glass vial (NICHIDENRIKA-GLASS CO., LTD., Japan). While the samples were attached to the axilla, the donor was instructed to change into clothes (UNIQLO, Japan) that were washed with fragrance-free detergent.
We analyzed the axillary odor samples obtained from the four sampling methods using the appropriate equipment and conditions for each sample. The sterile gauze sample was analyzed by GCMS-QP2010 Ultra (Shimadzu, Japan) (SH-Stabilwax column of 30 m × 0.53 mm i.d. with a film thickness of 1 μm) combined with the large volume static headspace system (7650HS-CTS, Entech, USA). The gauze samples were placed in 500-mL inert glass vials (EN39-75500W, Entech, USA) and equilibrated in the oven for 20 min (80°C) before extracting the headspace into the capillary traps. The headspace (200 cc) was directed to the capillary traps, followed by backflushing directly to the GC analytical column. The column temperature was programmed to rise from 40°C (5 min hold) to 230°C (66 min hold) at 10°C/min. The interface temperature was maintained at 230°C, and the ion source temperature at 230°C. Mass spectra were obtained in full-scan mode (range: m/z 29–400) at 70ev electron impact.
PDMS membrane, Twister, and Monotrap samples were analyzed using GCMS-TQ8030 (Shimadzu, Japan) (Stabilwax column of 60 m × 0.32 mm i.d. with 0.5-μm film thickness) combined with a thermal desorption system OPTIC4 (GL Science, Japan). The conditions of thermal desorption controlled by OPTIC4 are as follows. The vaporization chamber increased from 35°C to 230°C at 5°C/s, septum purge flow 5mL/min, sample sweep time 30 s, column flow 2mL/min, transfer time 2 min, split flow 10mL/min, cryotrap low temperature −150°C (130 s hold), cryotrap high temperature 250°C, cryotrap heat ramp rate 60°C/s, and the GCMS column temperature was programmed to rise from 50°C (2.5 min hold) followed by 10°C/min to 150°C, then 5°C/min to 230°C (61.5 min hold). The interface temperature was maintained at 230°C, and the temperature of the ion source was 230°C. Mass spectra were obtained in full-scan mode (range: m/z 29–400) at 70ev electron impact. Volatile compounds were detected using the automatic peak detector (GCMS solution software; Shimadzu, Japan), and individual peaks were identified using the NIST 17 (National Institute of Standards and Technology, USA) and the Wiley Registry of Mass Spectral Data, 9th edition (Wiley, USA). Visual inspection and fragment matching were used to confirm peaks with >70% similarity in the library search. Moreover, about half of the compounds were confirmed by the MS and retention time of purchased authenticated compounds with >95% similarity. A peak retention index (RI) was subsequently calculated to compare the analysis results for each condition. If a peak detected in the axillary odor sample was also detected in the control sample (experimental room), only the peak detected with a peak area value more than twice that of the control was adopted as the body odor compounds and the total number of such peaks were calculated (Figure S2). The threshold of more than twice was set to avoid picking up noise while ensuring that no potential candidates were excluded.
Collection of axillary odors to compare the four phases of menstruation
PDMS membranes were utilized to collect axillary odor for GC/MS. Small pieces of PDMS membrane (15 mm × 5 mm × 0.45 mm) were cut from a sheet and conditioned in an oven at 260°C for 1 h with nitrogen flowing. These conditioned pieces were immediately placed in Low Adsorption (LA) Qsert vials (Cat no: 29663-U, MERCK, USA) and stored at −30°C before axillary odor collection. Before applying the PDMS membrane to the axilla, the surface of both axillae was wiped with a moist hot towel and cleaned with 70% ethanol using sterile gauze. Conditioned PDMS membranes were then placed on the surface of both axillae, covered with polyethyleneterephthalate (PET) film cut from Flek-Sampler (OMI ODOR-AIR SERVICE Corporation, Japan), and secured to the skin with waterproof films (New Tac Kasei, Japan). The PDMS membranes were left on the skin for 2 h, following which they were removed using clean tweezers, inserted into an OPTIC DMI sample insert (60-μL micro vial) (GL Science, Japan), and placed in 2-mL glass vials (Thermo Fisher Scientific Inc., USA). Blank PDMS membranes, placed on a PET film on a desk in the experimental room, were taken and analyzed as controls. Each glass vial was injected with 1 μL of 100 ppm toluene-D8 solution diluted with methanol (Cat no: 233382, Sigma-Aldrich, US) as an internal standard. After sampling, the PDMS membranes were immediately stored at −80°C. Odor collection with the PDMS membrane was performed before axillary odor collection with gauze in the “collection of axillary odor for sensory evaluation” section. During odor collection, donors were instructed to change into clothes (UNIQLO, Japan) washed with fragrance-free detergent and were kept at rest in a room with a temperature of 26°C and constant humidity (48 ± 5%).
We collected 84 axillary odor samples using PDMS membranes from November 2017 to March 2018, and all GC/MS data were obtained from January 2018 to April 2018 using the aforementioned methods. Volatile compounds were detected using the automatic peak detector (GCMS solution software; Shimadzu, Kyoto, Japan), and individual peaks were verified using NIST 17 (National Institute of Standards and Technology, USA) and the Wiley Registry of Mass Spectral Data, 9th edition (Wiley, USA). Compounds with fatty acid characteristics, even if they did not match the database exactly, were labeled as “long-chain fatty acid” (Table S4) “Unknown compound” refers to a compound with an RI or mass spectrum that matched between samples but could not be identified. Subsequently, the peaks were aligned, and the area of each peak was calculated using the GCMS solution software (Shimadzu, Japan) by automatic integration. The compounds information is shown in Tables S10 and S11. For statistical analysis, 98 compounds that could be identified by authentic compounds or had a minimum occurrence rate of 24% (equivalent to 5 donors) or more in the phase in which they were detected in the largest number of people (average occurrence rate in all phases of all compounds = 71%) were used. Of the 98 compounds, of which 52 compounds were identified by authentic compounds, 28 compounds were detected in the control sample.
Because the samples were carefully stored and analyzed in a short period of time, there was little difference in the peak areas of the internal standard toluene-D8 between axillary samples, so absolute peak areas were used in subsequent statistical analyses. When compounds were not detected by GCMS solution (LOD&LOQ, a signal-to-noise ratio of 3 was considered), we set 0 for peak area.
Separation of E- and Z-isomers of geranylacetone
A mixture of geometrical isomers of geranylacetone (Tokyo Chemical Industry Co., Japan) was purified using column chromatography. Silica gel was impregnated with 8% AgNO3,64, (hexane/ethyl acetate = 9:1) to obtain its E- and Z-isomers. The ratio of E- and Z-isomers of the fractionated samples was determined at each timepoint using GC/MS, and finally, the (E)-geranylacetone was purified to 99.6%. The geometry of each isomer was confirmed by observing diagnostic nuclear Overhauser effect (NOE) correlations. The purified (E)-geranylacetone (purity >99.6%) was used in this study.
64.
Williams, C.M. ∙ Mander, L.N.
Chromatography with silver nitrate
Tetrahedron. 2001; 57:425-447
Odor evaluation using O-phase-increasing odor
Odor solution
For experiments involving the O-phase-increasing three compounds, we prepared an odor mixture solution containing the same compounds with their respective ratios as those detected from odor-O. All samples were soluble in Ethanol (EtOH), and it was thus used as the solvent. The final concentration of the solution was adjusted so that the 60 μL of odor-O solution used in each experiment contained three times the median of the O-phase-increasing compounds detected by the 2-h collection of axillary odors using the PDMS membrane. The concentration of the model axillary odor solution was set at approximately 1.8 times the intensity of the axillary odor when gauze was collected for 4 h in Figure 1 (equivalent to a total of 7.2 h). Therefore, the concentration of the ovulation-inducing component solution was set at 3 times the concentration of the 2-h PDMS collection (6 times the total concentration), which was estimated to be slightly lower due to individual differences. For quantification, diluted authentic compounds were applied to the same size of PDMS membranes used for the axillary, analyzed in the same manner, and quantified from a calibration curve.
A model mixture of axillary odors,12 the mixture of seven fatty acids (Table S6), modified to match our results, was used as the basal axillary odor (odor-A). Due to practical constraints during reagent preparation, the concentrations used in this study are slightly different from those reported in the original publication. GC/MS analysis did not reveal phase specificity in the compounds used as odor-A, except propanoic acid, which is increased in the F-phase (Table S3). The odor intensity (median = 2.89, SE = 0.147, Figure 4A) of the basal axillary odor was set at approximately 1.8 times the odor intensity (mean = 1.61, SE = 0.107, Figure 1B) of the female axillary gauze collected for 4 h.
12.
Kuukasjärvi, S. ∙ Eriksson, C.P. ∙ Koskela, E. ...
Attractiveness of women’s body odors over the menstrual cycle: The role of oral contraceptives and receiver sex
Behav. Ecol. 2004; 15:579-584
Furthermore, we prepared a mixture of odor-A and odor-O as odor-A+O, and a solution containing one compound of each of the three O-phase-increasing compounds in odor-A as odor-A+G, odor-A+T, and odor-A+H. As the control in the odor phase, we prepared an odor mixture solution containing the compounds detected in the axillary odor of a follicular female (odor-F), and the final concentration was adjusted three times the median of the F-phase-increasing compounds detected by the 2-h collection of axillary odor using the PDMS membrane. Additionally, a mixture of odor-A and odor-F was prepared as odor-A+F. The required amount of each odorant solution was pipetted onto the top of a mouillette (5.5 × 8.5 × 2.5 cm, Hiromichi, Japan) attached in a loop on a 32-mm double clip (ASKUL Corporation, Japan) and allowed to wait for 10 min until the solvent EtOH had completely evaporated.
Sensory evaluation
Odor ratings of (Ctrl, no odor; odor-O, three O-phase-increasing compounds; odor-A, basal axillary odor; odor-A+O, mixture of odor-A & odor-O, Table S6) were obtained for intensity (0 = not perceived to +6 = extremely intense), pleasantness (−3 = very unpleasant to +3 = very pleasant), preference (−3 = extremely dislike to +3 = extremely like), feminine/masculine (−3 = very masculine to +3 = very feminine), relax (0 = not perceived to +6 = extremely perceived), arousal/excitement (0 = not perceived to +6 = extremely perceived), and the top 15 descriptors in Table 1 (0 = not perceived to +6 = extremely perceived) were performed.
Thirty-seven single men of 72 single men participated in the evaluation (age: mean ± SD = 30 ± 5.5 years, OPE score: 9.8 ± 1.2). Odor ratings of (odor-A, basal axillary odor; odor-A+O, mixture of odor-A & odor-O; odor-A+G, mixture of odor-A & (E)-geranylacetone; odor-A+H, mixture of odor-A & (Z)-9-hexandecenoic acid; odor-A+T, mixture of odor-A & tetradecanoic acid; odor-A+F, mixture of odor-A & odor-F, Table S6) were obtained for intensity (0 = not perceived to +6 = extremely intense) and pleasantness (−3 = very unpleasant to +3 = very pleasant). All other descriptors are presented in Figure S5.
Physiological and psychological experiments
To explore the psychological and physiological effects of the O-phase-increasing odor, we conducted an experiment with the timeline shown in Figure 5A for single men (n = 19, age: mean ± SD = 27 ± 4.5 years, OPE score: 9.2 ± 1.7). The experiments started at 10:30 a.m. or 1:00 p.m. The order of odor stimulation was counterbalanced, and we employed a double-blind experimental design. In Base session, each rater entered the experimental room (private room), sat quietly in a chair, and received experimental instructions displayed on monitors in front of the raters and explained verbally by a female experimenter. First, raters completed the baseline MMS, a type of questionnaire (details provided in the “MMS” section). They were then instructed regarding the saliva collection procedure, and baseline saliva was collected (details are provided in the “Saliva assay” section). Subsequently, the rater was instructed to wear a headset with an odor (Session1: Ctrl or odor-O, Session2: odor-O or Ctrl, Session3: odor-A or odor-A+O, The odors presented in each session are noted in Table S13) on the microphone65 for 22–25 min and was subjected to three sessions of impression ratings of female facial images using a computer (details are provided in the “impression rating of female facial images” section) and MMS. Saliva was collected immediately after removing the headset. Of the 19 single men, 9 received odor-A and 10 received odor-A+O in Session 3. The raters were unaware that the microphone of the headset was scented. We explained to raters upon informed consent that a debriefing will be performed upon request, but nobody requested.
65.
Leleu, A. ∙ Demily, C. ∙ Franck, N. ...
The Odor Context Facilitates the Perception of Low-Intensity Facial Expressions of Emotion
PLoS One. 2015; 10, e0138656
MMS
The MMS, a self-reporting questionnaire developed for Japanese individuals to measure multiple mood-emotional states at a given time, was used to assess the emotions of the participants.66 The MMS has proven to be a sensitive tool for assessing emotions in response to various stimuli. MMS can measure general emotional states, unlike STAI-S, which focuses on anxiety states, and the positive and negative affect schedule (PANAS), which measures more abstract emotions and moods such as positive or negative effects.67 There are five subcategories within the eight major emotional categories (Depression/Anxiety, Hostility, Boredom, Liveliness, Relaxation/Tranquility, Friendliness/Amicability, Concentration/Focus, Startle/Surprise) in MMS, each of which is scored from 0 (not perceived) to 3 (clearly perceived). Thus, the score for each major category ranges from 0 to 15 points (details are shown in Table S8).
66.
Terasaki, M. ∙ Klshimoto, Y. ∙ Koga, A.
Construction of a multiple mood scale
Japanese J. Psychol. 1992; 62:350-356
67.
Ogawa, T. ∙ Monchi, R. ∙ Kikuya, M. ...
Development of the General Affect Scaled
Shinrigaku Kenkyu. 2000; 71:241-246
Saliva assay
Saliva was collected without orofacial movement using polypropylene tubes (Cryovial 2 mL, Salimetrics LLC, USA).68 The collected samples were frozen at −80°C until analysis. Salivary α-amylase activity (U/min) was measured using the Salivary α-Amylase Kinetic Enzyme Assay Kit (Salimetrics LLC, USA) at the Yanaihara Institute Inc. in Japan. It is worth noting that one participant could not collect saliva, and two participants faced challenges collecting saliva during the experiment on several occasions.
68.
Nagy, T. ∙ van Lien, R. ∙ Willemsen, G. ...
A fluid response: Alpha-amylase reactions to acute laboratory stress are related to sample timing and saliva flow rate
Biol. Psychol. 2015; 109:111-119
Impression rating of female facial images
We obtained one smiling and one neutral image each of eight young female faces from Getty Images (https://www.gettyimages.co.jp/), AIST facial expression database 2017 (https://unit.aist.go.jp/harc/kao_CECRT/kaohyoujoDB_HP.html), iStock (https://www.istockphoto.com/jp), Shutterstock (https://www.shutterstock.com/ja), and PIXTA (https://pixta.jp/). We then used Future Baby Maker | FaceFilm (Jiangguo Yan, App Store) to create a morphing image in 9 steps with 0% for neutral images and 100% for smiling images, and 50% smiling face images of eight females were used in this experiment (Figure 5D). During the odor stimulation sessions 1–3, raters rated their impression of each face using four categories (beautiful, elegant & intellectual, want to spend time with, want to keep gazing at; ratings: 0 = not perceived, 5 = extremely strong). Each image was presented on the monitor for 2 s, followed by a 10-s presentation of the page on which the ratings for the four categories were administered. If a response was not received within 10 s, the next image was forcefully displayed. Presentation (Neurobehavioral Systems, USA) was used for data collection.
Quantification and statistical analysis
Statistical analysis of sensory evaluation of female axillary odor
The effect of menstrual cycle status on the perceived odor intensity and pleasantness was analyzed using linear mixed models (LMMs). In each model, the response variable was the rating scores, and the fixed factor was the menstrual cycle phase at the time of odor collection (four categories: M, F, O, and L-phase), which was used as dummy variables. The models specified random intercepts for odor raters and odor donors. All analyses were conducted in R, version 3.5.2, using the lmerTest package (lmerTest package; https://cran.r-project.org/web/packages/lmerTest/index.html) (Figure 1 and Table S1). All statistical tests were two-tailed, and the significance level was set at p < 0.05. For the CATA questions, the number of selections and descriptor selection rate (%) were determined for each descriptor. Cochran’s Q-test was performed for 15 descriptors (descriptor selection rate >2%) to evaluate significant differences in axillary odors from four different phases. IBM SPSS version 27 (IBM Co., USA) and GraphPad Prism Ver9.5.1 (GraphPad Software, USA) were used. A priori power analysis using G∗Power 3.1 indicated that 23 raters were enough in the condition of effect size f = 0.25, α level = 0.05, Power (1-β err prob) = 0.8 in ANOVA (Repeated measures, within factors).
Statistical analyses of 98 axillary odor compounds in each phase of the menstrual cycle
A priori power analysis using G∗Power 3.1 indicated that 22 donors were enough in the condition of effect size d = 0.57, α level = 0.05, Power (1-β err prob) = 0.8 in Wilcoxon signed-rank test (one sample case). We converted the peak area of 98 compounds into a robust Z score for standardization. Robust z-scores were calculated as follows: robust z = (x – median)/normalized interquartile range (NIQR), where “x” is the respective peak area value, and median and NIQR were those of the respective female donors. The cluster analysis used in Figure 2 is ward method with squared euclidean distance. IBM SPSS version 27 (IBM Co., USA) was used for these analyses. The p-value was calculated using a one-sample Wilcoxon signed-rank test (one-tailed) to determine whether the amount of the compound in each phase was greater than 0 (median of the robust Z score), and the significance level was set at p < 0.05 (Tables S3 and S4). To test whether it is possible to discriminate between the ovulatory phase and other phases using the 98 compounds that were used in this analysis, conditional logistic regression on the corresponding data for 21 donors was performed using the COXREG command of SPSS ver. 27 (Conditional logistic regression using COXREG (ibm.com)). First, the problem of multicollinearity was encountered, so the Spearman’s coefficient for the 98 compounds was calculated. Based on > 0.7, 13 variables (#5, 6, 14, 19, 24, 56, 83, 89, 90, 91, 92, 93, 94) were removed. A logistic regression analysis was conducted using forward selection (likelihood ratio) for the remaining 85 variables (Table S5).
Statistical analyses of sensory evaluations and physiological and psychological experiments
For the analyses shown in Figure 4, a priori power analysis using G∗Power 3.1 indicated that 67 raters were sufficient for t test (Difference between two dependent means (matched pairs)) in Figure 4A with an effect size dz = 0.35, α level = 0.05, Power (1-β err prob) = 0.8, while 34 raters were sufficient for Figures 4B and 4C under the same conditions with an effect size dz = 0.5. Paired t test with false discovery rate (FDR) was used for the 21 descriptors in Figure 4A (control vs. odor-O, odor-A vs. odor-A+O). Dunnett’s multiple comparison test among the four types of odors is shown in Figure 4B (odor-A+O vs. A + G, A + H, A + T) and Tukey’s multiple comparison test is shown in Figure 4C (odor-A vs. A + O, A + F, and odorA + O vs. odorA + F). All statistical tests were two-tailed, and the significance level was set at q < 0.05 or p < 0.05. IBM SPSS version 27 (IBM Co., USA) and GraphPad Prism Ver9.5.1 (GraphPad Software, Boston, USA) were used for statistical analyses. The Statistical data is shown in Table S12.
For the analyses shown in Figure 5, a priori power analysis using G∗Power 3.1 indicated that 24 raters were enough in the condition of effect size f = 0.25, α level = 0.05, Power (1-β err prob) = 0.8 in ANOVA (Repeated measures, within-between interaction). One-way ANOVA (within-participants) was performed using the Greenhouse-Geisser method with Tukey’s multiple comparison test as a post hoc test (base vs. Ctrl vs. odor-O) or paired t test (Ctrl(A) vs. odor-A, Ctrl(A + O) vs. odor-A+O) (Figures 5B and 5D). We also performed two-way ANOVA for within-subject design, or if some values are missing, those data were analyzed by fitting a mixed effect model (REML) (Ctrl(A) vs. odor-A, Ctrl(A + O) vs. odor-A+O), and two-way ANOVA or REML (Ctrl(A) vs. Ctrl(A + O), odor-A – Ctrl (A) vs. odor-A+O – Ctrl(A + O)) adjusted by the Greenhouse-Geisser method (Figure 5F). Simple linear regression analysis was performed for salivary α-amylase and relaxation score of MMS (Figure 5C). All statistical tests were two-tailed, and the significance level was set at p < 0.05. IBM SPSS version 27 (IBM Co., USA). GraphPad Prism Ver9.5.1 (GraphPad Software, Boston, USA) were used for statistical analysis. The Statistical data is shown in Table S14.
Supplemental information (7)
Document S1. Figures S1–S5 and Tables S1–S8
Table S9. Odor donor profiles, related to Figure 1
Table S10. Compounds information, related to Table S3 and S4
Table S11. Robust Z value of 98 compounds for each phase, related to Figure 2
Table S12. Statistical data, related to Figure 4
Table S13. Odor rater profiles, related to Figure 5
Table S14. Statistical data, related to Figure 5
References
O’Connell, R.J. ∙ Singer, A.G. ∙ Macrides, F. ...
Responses of the male golden hamster to mixtures of odorants identified from vaginal discharge
Behav. Biol. 1978; 24:244-255
Cerda-Molina, A.L. ∙ Hernández-López, L. ∙ Rojas-Maya, S. ...
Male-induced sociosexual behavior by vaginal secretions in Macaca arctoides
Int. J. Primatol. 2006; 27:791-807
Clarke, P.M.R. ∙ Barrett, L. ∙ Henzi, S.P.
What role do olfactory cues play in chacma baboon mating?
Am. J. Primatol. 2009; 71:493-502
Ziegler, T.E. ∙ Schultz-Darken, N.J. ∙ Scott, J.J. ...
Neuroendocrine response to female ovulatory odors depends upon social condition in male common marmosets, Callithrix jacchus
Horm. Behav. 2005; 47:56-64
Doty, R.L. ∙ Ford, M. ∙ Preti, G. ...
Changes in the intensity and pleasantness of human vaginal odors during the menstrual cycle
Science. 1975; 190:1316-1318
Singh, D. ∙ Bronstad, P.M.
Female body odour is a potential cue to ovulation
Proc. Biol. Sci. 2001; 268:797-801
Havlíček, J. ∙ Dvořáková, R. ∙ Bartoš, L. ...
Non-advertized does not mean concealed: Body odour changes across the human menstrual cycle
Ethology. 2006; 112:81-90
Gildersleeve, K.A. ∙ Haselton, M.G. ∙ Larson, C.M. ...
Body odor attractiveness as a cue of impending ovulation in women: Evidence from a study using hormone-confirmed ovulation
Horm. Behav. 2012; 61:157-166
Havliček, J. ∙ Cobey, K.D. ∙ Barrett, L. ...
The spandrels of Santa Barbara? A new perspective on the peri-ovulation paradigm
Behav. Ecol. 2015; 26:1249-1260
Kuukasjärvi, S. ∙ Eriksson, C.P. ∙ Koskela, E. ...
Attractiveness of women’s body odors over the menstrual cycle: The role of oral contraceptives and receiver sex
Behav. Ecol. 2004; 15:579-584
Zetzsche, M. ∙ Weiß, B.M. ∙ Kücklich, M. ...
Combined perceptual and chemical analyses show no compelling evidence for ovulatory cycle shifts in women’s axillary odour
Proc. Biol. Sci. 2024; 291:20232712
de Groot, J.H.B. ∙ Kirk, P.A. ∙ Gottfried, J.A.
Titrating the Smell of Fear: Initial Evidence for Dose-Invariant Behavioral, Physiological, and Neural Responses
Psychol. Sci. 2021; 32:558-572
Mujica-Parodi, L.R. ∙ Strey, H.H. ∙ Frederick, B. ...
Chemosensory Cues to Conspecific Emotional Stress Activate Amygdala in Humans
PLoS One. 2009; 4, e6415
Hara, T. ∙ Kyuka, A. ∙ Shimizu, H.
Butane-2,3-dione: the Key Contributor to Axillary and Foot Odor Associated with an Acidic Note
Chem. Biodivers. 2015; 12:248-258
Hirasawa, Y. ∙ Shirasu, M. ∙ Okamoto, M. ...
Subjective unpleasantness of malodors induces a stress response
Psychoneuroendocrinology. 2019; 106:206-215
Zeng, X.N. ∙ Leyden, J.J. ∙ Spielman, A.I. ...
Analysis of characteristic human female axillary odors: Qualitative comparison to males
J. Chem. Ecol. 1996; 22:237-257
Natsch, A. ∙ Derrer, S. ∙ Flachsmann, F. ...
A broad diversity of volatile carboxylic acids, released by a bacterial aminoacylase from axilla secretions, as candicate molecules for the determination of human-body odor type
Chem. Biodivers. 2006; 3:1-20
Gallagher, M. ∙ Wysocki, C.J. ∙ Leyden, J.J. ...
Analyses of volatile organic compounds from human skin
Br. J. Dermatol. 2008; 159:780-791
Curran, A.M. ∙ Rabin, S.I. ∙ Prada, P.A. ...
Comparison of the volatile organic compounds present in human odor using SPME-GC/MS
J. Chem. Ecol. 2005; 31:1607-1619
Dormont, L. ∙ Bessière, J.M. ∙ Cohuet, A.
Human Skin Volatiles: A Review
J. Chem. Ecol. 2013; 39:569-578
Leal, H.M. ∙ Hwang, J.K. ∙ Tan, K. ...
Attraction of Culex mosquitoes to aldehydes from human emanations
Sci. Rep. 2017; 7:17965
Zhao, Z. ∙ Zung, J.L. ∙ Hinze, A. ...
Mosquito brains encode unique features of human odour to drive host seeking
Nature. 2022; 605:706-712
Drabińska, N. ∙ Flynn, C. ∙ Ratcliffe, N. ...
A literature survey of all volatiles from healthy human breath and bodily fluids: the human volatilome
J. Breath Res. 2021; 15, 034001
Syed, Z. ∙ Leal, W.S.
Acute olfactory response of Culex mosquitoes to a human- and bird-derived attractant
Proc. Natl. Acad. Sci. USA. 2009; 106:18803-18808
Owsienko, D. ∙ Goppelt, L. ∙ Hierl, K. ...
Body odor samples from infants and post-pubertal children differ in their volatile profiles
Commun. Chem. 2024; 7:53
de Lacy Costello, B. ∙ Amann, A. ∙ Al-Kateb, H. ...
A review of the volatiles from the healthy human body
J. Breath Res. 2014; 8, 014001
Aparicio, E. ∙ Martín-Grau, C. ∙ Hernández-Martinez, C. ...
Changes in fatty acid levels (saturated, monounsaturated and polyunsaturated) during pregnancy
BMC Pregnancy Childbirth. 2021; 21, 778
Aktas, M. ∙ Elmastas, M. ∙ Ozcicek, F. ...
Erythrocyte Membrane Fatty Acid Composition in Premenopausal Patients with Iron Deficiency Anemia
J. Oleo Sci. 2016; 65:225-231
Contreras, C.M. ∙ Gutiérrez-García, A.G. ∙ Mendoza-López, R. ...
Amniotic fluid elicits appetitive responses in human newborns: Fatty acids and appetitive responses
Dev. Psychobiol. 2013; 55:221-231
Gutiérrez-García, A.G. ∙ Contreras, C.M. ∙ Díaz-Marte, C.
Myristic acid in amniotic fluid produces appetitive responses in human newborns
Early Hum. Dev. 2017; 115:32-37
Haze, S. ∙ Gozu, Y. ∙ Nakamura, S. ...
2-Nonenal newly found in human body odor tends to increase with aging
J. Invest. Dermatol. 2001; 116:520-524
Porter, N.A.
Mechanisms for the Autoxidation of Polyunsaturated Lipids
Acc. Chem. Res. 1986; 19:262-268
Draper, C.F. ∙ Duisters, K. ∙ Weger, B. ...
Menstrual cycle rhythmicity: metabolic patterns in healthy women
Sci. Rep. 2018; 8:14568
Seubert, J. ∙ Gregory, K.M. ∙ Chamberland, J. ...
Odor valence linearly modulates attractiveness, but not age assessment, of invariant facial features in a memory-based rating task
PLoS One. 2014; 9, e98347
Zhou, W. ∙ Chen, D.
Fear-Related Chemosignals Modulate Recognition of Fear in Ambiguous Facial Expressions
Psychol. Sci. 2009; 20:177-183
Albrecht, J. ∙ Demmel, M. ∙ Schöpf, V. ...
Smelling chemosensory signals of males in anxious versus nonanxious condition increases state anxiety of female subjects
Chem. Senses. 2011; 36:19-27
de Groot, J.H.B. ∙ Smeets, M.A.M. ∙ Rowson, M.J. ...
A sniff of happiness
Psychol. Sci. 2015; 26:684-700
Granqvist, P. ∙ Vestbrant, K. ∙ Döllinger, L. ...
The scent of security: Odor of romantic partner alters subjective discomfort and autonomic stress responses in an adult attachment-dependent manner
Physiol. Behav. 2019; 198:144-150
Hofer, M.K. ∙ Collins, H.K. ∙ Whillans, A.V. ...
Olfactory cues from romantic partners and strangers influence women’s responses to stress
J. Pers. Soc. Psychol. 2018; 114:1-9
Lundström, J.N. ∙ Mathe, A. ∙ Schaal, B. ...
Maternal status regulates cortical responses to the body odor of newborns
Front. Psychol. 2013; 4:597
Schaal, B. ∙ Saxton, T.K. ∙ Loos, H. ...
Olfaction scaffolds the developing human from neonate to adolescent and beyond
Philos. Trans. R. Soc. Lond. B Biol. Sci. 2020; 375, 20190261
Pipitone, R.N. ∙ Gallup, G.G.
Women’s voice attractiveness varies across the menstrual cycle
Evol. Hum. Behav. 2008; 29:268-274
Pavela Banai, I.
Voice in different phases of menstrual cycle among naturally cycling women and users of hormonal contraceptives
PLoS One. 2017; 12, e0183462
Roberts, S.C. ∙ Havlicek, J. ∙ Flegr, J. ...
Female facial attractiveness increases during the fertile phase of the menstrual cycle
Proc. Biol. Sci. 2004; 271:S270-S272
Burriss, R.P. ∙ Troscianko, J. ∙ Lovell, P.G. ...
Changes in women’s facial skin color over the ovulatory cycle are not detectable by the human visual system
PLoS One. 2015; 10, e0130093
Toyoda, Y. ∙ Sakurai, A. ∙ Mitani, Y. ...
Earwax, osmidrosis, and breast cancer: why does one SNP (538G> A) in the human ABC transporter ABCC11 gene determine earwax type?
FASEB J. 2009; 23:2001-2013
Keller, A. ∙ Zhuang, H. ∙ Chi, Q. ...
Genetic variation in a human odorant receptor alters odour perception
Nature. 2007; 449:468-472
Menashe, I. ∙ Abaffy, T. ∙ Hasin, Y. ...
Genetic elucidation of human hyperosmia to isovaleric acid
PLoS Biol. 2007; 5, e284
Ferdenzi, C. ∙ Razafindrazaka, H. ∙ Baldovini, N. ...
Influence of gender and culture on the perception of acidic compounds of human body odor
Physiol. Behav. 2019; 210, 112561
Schleidt, M. ∙ Hold, B. ∙ Attili, G.
A cross-cultural study on the attitude towards personal odors
J. Chem. Ecol. 1981; 7:19-31
Rossi, V. ∙ Pourtois, G.
Transient state-dependent fluctuations in anxiety measured using STAI, POMS, PANAS or VAS: a comparative review
Anxiety Stress Coping. 2012; 25:603-645
Adams, J. ∙ Williams, A. ∙ Lancaster, B. ...
Advantages and uses of check-all-that-apply response compared to traditional scaling of attributes for salty snacks
7th Pangborn Sensory Science Symposium 12–16 August Minneapolis USA
2007
Nishida, K. ∙ Kobayashi, M. ∙ Ogihara, H. ...
Clinical Usefulness of Smell Identification Test Card : Open Essence
Nihon Jibiinkoka Gakkai Kaiho. 2010; 113:751-757
Suzuki, M. ∙ Hashimoto, M. ∙ Yoshioka, M. ...
The odor stick identification test for Japanese differentiates Parkinson’s disease from multiple system atrophy and progressive supra nuclear palsy
BMC Neurol. 2011; 11:157
Fujio, H. ∙ Inokuchi, G. ∙ Fukuda, Y. ...
A New Screening Tool, An Odor Identification Test Card to Detect Age―related Changes in Olfaction Hisami
Nippon Jibiinkoka Gakkai Kaiho. 2018; 121:38-43
Stevens, D. ∙ Cornmell, R. ∙ Taylor, D. ...
Spatial variations in the microbial community structure and diversity of the human foot is associated with the production of odorous volatiles
FEMS Microbiol. Ecol. 2015; 91:1-11
Riazanskaia, S. ∙ Blackburn, G. ∙ Harker, M. ...
The analytical utility of thermally desorbed polydimethylsilicone membranes for in-vivo sampling of volatile organic compounds in and on human skin
Analyst. 2008; 133:1020-1027
Soini, H.A. ∙ Bruce, K.E. ∙ Wiesler, D. ...
Stir Bar Sorptive Extraction: A New Quantitative and Comprehensive Sampling Technique for Determination of Chemical Signal Profiles from Biological Media
J. Chem. Ecol. 2005; 31:377-392
Leleu, A. ∙ Demily, C. ∙ Franck, N. ...
The Odor Context Facilitates the Perception of Low-Intensity Facial Expressions of Emotion
PLoS One. 2015; 10, e0138656
Terasaki, M. ∙ Klshimoto, Y. ∙ Koga, A.
Construction of a multiple mood scale
Japanese J. Psychol. 1992; 62:350-356
Ogawa, T. ∙ Monchi, R. ∙ Kikuya, M. ...
Development of the General Affect Scaled
Shinrigaku Kenkyu. 2000; 71:241-246
Nagy, T. ∙ van Lien, R. ∙ Willemsen, G. ...
A fluid response: Alpha-amylase reactions to acute laboratory stress are related to sample timing and saliva flow rate
Biol. Psychol. 2015; 109:111-119
Figures (5)
Article metrics
Supplemental information (7)
PDF (1.07 MB)
Document S1. Figures S1–S5 and Tables S1–S8
Spreadsheet (13.77 KB)
Table S9. Odor donor profiles, related to Figure 1
Spreadsheet (18.98 KB)
Table S10. Compounds information, related to Table S3 and S4
Spreadsheet (80.19 KB)
Table S11. Robust Z value of 98 compounds for each phase, related to Figure 2
Spreadsheet (17.02 KB)
Table S12. Statistical data, related to Figure 4
