- Email: [email protected]
Lack of support for relation between woman's masculinity preference, estradiol level and mating context
Hormones and Behavior 78 (2016) 1–7
Contents lists available at ScienceDirect
Hormones and Behavior journal homepage: www.elsevier.com/locate/yhbeh
Lack of support for relation between woman's masculinity preference, estradiol level and mating context Urszula M. Marcinkowska a,⁎, Peter T. Ellison b, Andrzej Galbarczyk a, Karolina Milkowska a, Boguslaw Pawlowski c, Inger Thune d, Grazyna Jasienska a,e a
Department of Environmental Health, Faculty of Health Sciences, Jagiellonian University Medical College, 20 Grzegorzecka St., 31-531 Krakow, Poland Department of Human Evolutionary Biology, Harvard University, 11 Divinity Av., Cambridge, MA 02138, USA Department of Human Biology, University of Wrocław, Kuźnicza 35, Wrocław, Poland d Department of Oncology, Oslo University Hospital, Ullevål, Oslo, Norway e Department of Anthropology, Yale University, 10 Sachem Street, New Haven, CT 06511-3707, USA b c
a r t i c l e
i n f o
Article history: Received 29 December 2014 Revised 5 October 2015 Accepted 20 October 2015 Available online 21 October 2015 Keywords: Human sexual preferences Facial preferences Sex steroid hormones Estradiol Sexual dimorphism Masculinity
a b s t r a c t It has been proposed that women's preferences for male facial sexual dimorphism are positively correlated with conception probability and differ between short- and long-term mating contexts. In this study, we tested this assumption by analyzing relationships between estradiol levels to the women's preferences of male faces that were manipulated to vary in masculinity. Estradiol was measured in daily saliva samples throughout the entire menstrual cycle collected by Polish women with regular menstrual cycles. In our analyses, we included the three most commonly used definitions of the fertile window in the literature. After computing the overall masculinity preference of each participant and measuring hormone levels, we found that i) the timing of ovulation varied greatly among women (between −11 and −17 days from the onset of the next menses, counting backwards), ii) there was no relationship between daily, measured during the day of the test (N = 83) or average for the cycle (N = 115) estradiol levels and masculinity preferences, iii) there were no differences in masculinity preferences between women in low- and high-conception probability phases of the cycle, and iv) there were no differences in masculinity preferences between short- and long-term mating contexts. Our results do not support the idea that women's preferences for a potential sexual partner's facial masculinity fluctuate throughout the cycle. © 2015 Elsevier Inc. All rights reserved.
Introduction Many studies suggest that women's preferences for male traits such as masculinity, symmetry and dominance differ between fertile and non-fertile phases of the menstrual cycle. When in fertile phase, women were found to exhibit stronger preferences for more masculine voices (Puts, 2005), body odors (Thornhill et al., 2013) and body shapes (Little et al., 2007) or increased height (Pawlowski and Jasienska, 2005). Many studies have also documented that women exhibit stronger preferences for facial masculinity around the time of ovulation (e.g., PentonVoak and Perrett, 2000). This fluctuation in preference for masculinity may have evolved as a trade-off between choosing cooperative, warm, and caregiving (i.e., less masculine) partners that are more suited to long-term relationships and partners with purported higher genetic
⁎ Corresponding author at: Department of Environmental Health, Faculty of Health Sciences, Jagiellonian University Medical College, 20 Grzegorzecka St., 31-531 Krakow, Poland. E-mail address: [email protected] (U.M. Marcinkowska).
http://dx.doi.org/10.1016/j.yhbeh.2015.10.012 0018-506X/© 2015 Elsevier Inc. All rights reserved.
quality (i.e., more masculine, Perrett et al., 1998). Despite numerous studies, common agreement regarding the evolutionary bases (or the occurrence) of this fluctuation is lacking (for a debate on this topic, see DeBruine et al. (2010); Gildersleeve et al. (2014a, 2014b); Harris (2011); Wood (2014)).
Hormonal fluctuations As the menstrual cycle is characterized by changing levels of steroid hormones, it is likely that changes in women's preferences throughout the cycle may be primarily attributable to fluctuating hormone levels (Bobst et al., 2014). The menstrual cycle has an average length of 28 days and comprises two ovarian phases: the follicular phase that begins on the first day of menstrual bleeding and continues until ovulation, and the luteal phase that begins after ovulation and continues until the day before next menstrual bleeding (Hawkinsa and Matzuk, 2008). In fully functioning cycles, during the follicular phase, a mature follicle is produced, which then releases an egg in a process called ovulation. The ovary secretes estradiol, progesterone and inhibins. Mid-
2
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7
cycle drop in estradiol is a reasonable estimate of the day of ovulation (Lipson and Ellison, 1996).
When does the fertile phase occur? A body of evidence has found that pre-menopausal women in the most fertile phase of their cycle (and who do not take hormonal contraceptives) have stronger preferences for more masculine faces than they do in their less fertile phases (for a review, see DeBruine et al., 2010). Some studies, however, did not replicate these results (for a metaanalysis, see Gildersleeve et al., 2014aand Wood et al., 2014). One limitation of studies on masculinity preferences and their fluctuations throughout the menstrual cycle is a lack of consistency in the methods used to determine the fertile window. Authors usually do not measure levels of steroid hormones and do not use physiological methods to detect ovulation, but assign women to being in “fertile” or “non-fertile” phase based on a cycle length. This methodology is problematic due to a high variability in menstrual cycle length (Chiazze et al., 1968) and intra- and inter-individual variation in levels of reproductive steroid hormones (Jasienska and Jasienski, 2008). Penton-Voak et al. (1999) defined the fertile phase as ranging from the end of previous menses to 14 days prior to next menses. In a later study, Penton-Voak and Perrett (2000) defined the high-conception probability phase as between 6 and 14 days from the onset of the previous menses. Macrae et al. (2002) and Garver-Apgar et al. (2008) proposed using a backward-counting method to estimate the fertile window; their proposed high-conception probability phases were, respectively, from 2 or 4 days prior to the ovulation. Some authors combined different approaches to determine the fertile window (i.e., Harris, 2011), which resulted in a 5-day window around ovulation (for more examples see meta-analysis by Gildersleeve et al., 2014b). To reliably use these methods, it must be assumed that cycle length and, particularly, the length of the luteal phase is consistent among participants. From studies of hormonal levels in naturally cycling women, we know that this assumption is problematic; multiple factors (Wilcox et al., 2000, Jasienska, 2013) cause cycle lengths to vary within and among participants (Jasienska and Jasienski, 2008). Although it is well established that levels of steroid hormones fluctuate during a cycle, there are surprisingly few studies on mate preferences that incorporate actual measurements of hormonal levels in naturally cycling women (Pisanski et al., 2014, Roney and Simmons, 2008).
Fluctuation of preferences Ovarian steroid hormones play an important role in female sexual motivation and behavior (Durante and Li, 2009) and overall preference of male facial attractiveness is the highest during the fertile phase of the cycle (Danel and Pawlowski, 2006). Other studies showed that general preference shifts occur only when attractiveness is judged in a shortterm, and not in a long-term context (Little and Jones, 2012) or only for women who were in a committed relationship (Little et al., 2008). However, the agreement on the robustness of women's cyclical preferences shifts is lacking. In the most recent meta-analysis of studies investigating menstrual cycle shifts in women's preferences, Gildersleeve et al. (2014a) reported “… analyses revealed robust cycle shifts that were specific to women's preferences for hypothesized cues of (ancestral) genetic quality…”. In contrast, a meta-analysis by Wood et al. (2014) was “largely non-supportive [of this hypothesis]. Specifically, fertile women did not desire sex in short-term relationships with men purported to be of high genetic quality (i.e., high masculinity …)”. It has been suggested though, that lack of support for cyclical shifts in Wood et al. (2014) could have been caused by differences in methodological and analytic decisions made by the authors (Gildersleeve et al., 2014b, Wood and Carden, 2014).
Hormones and masculinity preferences Although sex steroid hormones were suggested to be associated with visual processing (Little, 2013) there are surprisingly few studies of preferences that directly measured hormone levels, and those that did measure hormonal levels had conflicting results. Welling et al. (2007), based on within-participants measurements (N = 70), identified a positive correlation of preferences for masculinity and testosterone but failed to find a relationship with estradiol or progesterone. Similar relationships were documented by Bobst et al. (2014) for 27 cycling women. Rosen and Lopes (2009) found in a betweenparticipants study that courtship language preferences fluctuated depending on estradiol levels of participants (N = 17) but not on progesterone or testosterone levels. Salivary estradiol fluctuations have also been found to best predict changes in women's (N = 62) preferences for vocal masculinity (Pisanski et al., 2014). Finally, Roney and Simmons (2008) reported a positive correlation of woman's (N = 75) estradiol levels and preference for the faces of males with high circulating testosterone levels. It should be noted that the studies on preferences that measured hormone levels were not free from methodological problems, including having as participants young women, who are known to have irregular cycles and cycles with low hormone levels, estimating cycle length, taking a single or only a few samples for hormonal assessments from each participant, collecting samples at different times of a day, comparing hormonal values obtained from saliva samples to previously published estimates that were obtained from serum samples, or having small number of participating women. In our sample we aimed to repeat preference measurements correcting for all above mentioned flaws.
Study aim In this study, we sought to identify the putative fluctuation of facial masculinity preferences exhibited by women throughout the menstrual cycle. Because we measured estradiol levels in saliva samples collected daily for the entire cycle, we were able to detect the day of ovulation based on physiology, rather than estimating ovulation by counting the days of the cycle (i.e., the method used in most previous studies). We compared the masculinity preferences of women from high- and lowconception probability groups (between participant measurement) and also tested how preferences differed within those two groups depending on the mating context (within-participant measurement). Furthermore, we tested for relation between masculinity preferences and estradiol levels of participants. To add to the ongoing debate about the existence of fluctuating masculinity preferences, we conducted 4 separate analyses that compared results based on previously used definitions of the fertile window. Definitions came from studies that did not include hormonal measurements.
Methods Participants A total of 115 women from Poland (average age = 29.9, SD = 3.46) participated in this study. Women were recruited between June 2001 and June 2003 by posted advertisements in public places and by newspaper, radio and television. On the basis of an initial interview, the women that were selected to participate in this study were limited to those who met following criteria: between 24 and 36 years of age, regular menstrual cycles and no fertility problems, no gynecological or chronic disorders (e.g., diabetes, hypo/hyperthyroidism), not pregnant or lactating in the 6 months prior to recruitment, and not using hormonal contraception or taking any hormonal medication in the 6 months prior to recruitment.
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7
Questionnaire and visual stimuli After agreeing to participate in the study and completing a written consent form, participants answered a series of socio-demographic questions including relationship status (“single”, “in a relationship” or “hard to say”) and were instructed to collect saliva samples daily for the entire menstrual cycle. During the subsequent visit women judged 18 photographs of men (9 slides, each depicting 2 manipulated male faces). Visual stimuli were manipulated according to previous studies (Perrett et al., 1998). Women were presented with 2 morphed faces that were either 40% feminized or 40% masculinized version of same composite face (Penton-Voak et al., 1999, for the description of the morphing technique, see e.g., Marcinkowska et al., 2014). Composite male faces differed only in the particular shape that was manipulated to be more or less masculine, and all other characteristics (i.e., skin texture or symmetry) were held constant. Photographs were presented to women in a random order and were evaluated twice, once in a shortterm and once in a long-term relationship context. Masculinized and feminized versions of same face were presented on each page. Surveying was conducted in participants' homes or in the lab depending on a preference of a participant. Time and date of testing was marked by the member of the research group and included in the personal file of the participant. Survey completion timing varied from − 29 to − 1 day, counting backwards from the onset of the following menses (Mean = − 13.16 days). During the first evaluation, participants were asked to judge the attractiveness of male faces in the context of a potential short-term relationship described as follows: “You are looking for the type of person who would be attractive in a short-term relationship. This implies that the relationship may not last a long time. Examples of this type of relationship would include a single date accepted on the spur of the moment, an affair within a long-term relationship, and possibility of a one-night stand”. During the second evaluation, attractiveness was judged based on a potential long-term relationship, introduced as follows: “You are looking for the type of person who would be attractive in a long-term relationship”. Examples of this type of relationship would include someone you may want to move in with, someone for whom you may consider leaving a current partner, and someone you may, at some point, wish to marry (or have a relationship similar to marriage). In both evaluations participants were asked to choose the more attractive of the two faces and then indicate the extent to which they preferred a particular face by indicating if they had ‘almost no preference’, ‘slight preference’, ‘average preference’ or ‘strong preference’. Responses were recoded using a scale from 1 to 8, in which 1 indicated a strong preference for a feminized face and 8 indicated a strong preference for a masculinized face. Next, the mean preference for masculinity was calculated for each participant. Scores were converted to percentages, such that 0% represented a maximal preference for feminized faces and 100% represented a maximal preference for masculinized faces. Preferences variation was based on a between participants analysis. Hormonal measures Steroid hormones were measured in saliva samples collected daily by the participants themselves in their homes during their menstrual cycles. Prior to sample collection, each woman was instructed how to collect and store samples of saliva. Each participant received a set of 38 plastic tubes sealed with a stopper, 38 sugar-free sticks of laboratory-tested chewing gum, and calendars in which they recorded the collection date and time of each saliva sample. Participants collected daily morning saliva samples from the first day of menstruation to the day prior to the next menstrual bleeding. A total of 5.3% of daily samples were missing due to inappropriate collection or loss during laboratory procedures (Jasienska et al., 2006). Saliva samples were analyzed for 17-β estradiol (E2) by the Reproductive Ecology Laboratory at Harvard University, following a previously described procedure (Jasienska et al.,
3
2006). E2 was analyzed in saliva samples from 24 days of each cycle using an I-125-based radioimmunoassay kit (#39100, Diagnostic Systems Laboratories, Webster, Texas, USA) with published (Jasienska et al., 2004) modifications to the manufacturer's protocol. The quality of laboratory analysis was monitored by including samples of known concentrations (“pools”) with low (follicular), medium (luteal) and high (gestational) concentrations. The sensitivity of the E2 assay was 4 pmol/l. Average intra-assay variability was 9%, and inter-assay variability ranged from 23% for lower values (15 pmol/l) to 13% for higher values (50 pmol/l). Prior to statistical analyses, menstrual cycles were aligned based on the day of the mid-cycle E2 drop (i.e., day 0), which provides a reasonable estimate of the day of ovulation, according to published methods (Lipson and Ellison, 1996). Ovulation day defined in such way was used to divide participants into low and highconception probability groups. E2 values from 18 consecutive days of each cycle that were aligned on day 0 were used in analyses. Participants, whose mid-cycle estradiol drop could not be reliably identified, or the cycle had more than one estradiol drop; were not included in the analysis. The method used in this study is very detailed and narrows the possibility of erroneous fertility phase classification. Statistical analyses Participants were divided into two groups (high- and lowconception probability) based on their hormonal levels and day of the cycle during which they took a test of masculinity preference (Detecting ovulation). The placement of participants into these two groups was based on 3 separate definitions of a fertile window, so that our results could be compared to previous studies (hence the subdivision of the Results section into 3 analyses), all based on the counting days around the ovulation day. Whether the average masculinity preference within a group differed from random (50%) was tested using single sample t-tests. In the first analysis (t-test for independent samples), the average masculinity preference of women from the high-conception probability group was compared to the average preference of women from the low-conception probability group. This comparison was performed separately for short- and the long-term mating contexts (Difference in the masculinity preferences of women in high- and low-conception probability groups). In the second analysis (t-test for dependent samples), we compared the preferences of women within the two conception probability groups, depending on their mating context. Specifically, we examined whether a woman's preferences in a short-term mating context differed from her preferences in a longterm mating context (this comparison was performed separately for the high- and low-conception probability groups, see Difference in masculinity preference between short- and long-term mating contexts). Lastly, we tested with simple linear regression whether average estradiol levels throughout the cycle (Average estradiol and masculinity preference) and morning levels of estradiol on the day of completing the survey (Estradiol on the day of the survey and masculinity preference) were related to masculinity preferences in short- and long-term mating contexts. Confidence intervals and effect sizes can be found in Tables 2 and 3. Results On average, women in low- and high-conception probability phase preferred more masculine faces, in both short- and long-term mating contexts (58.33%, SD = 9.66, p b 0.001 and 57.87%, SD = 9.57, p b 0.001, respectively). Participant age did not influence masculinity preferences (N = 115, β = 0.013, p = 0.89 in a short-term context, and β = − 0.021, p = 0.83 in a long-term context). Preference of women did not differ depending on relationship status (paired vs. single) in either a short- (t = 0.488, p = 0.63) or long-term (t = 1.58, p = 0.12) mating context.
4
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7
probability groups in either a short- or long-term context (t = 0.165, p = 0.87 and t = 0.548, p = 0.58, respectively, Table 2). Analysis 3 Following Penton-Voak and Perrett (2000), the high-conception phase was from 8 days before to the day of ovulation (N = 42), and the low-conception phase included all other days of the cycle (N = 73). As in the previous analysis, calculation is a translation of forward calculations (Penton-Voak and Perrett, 2000). Masculinity preference did not significantly differ between the two groups, regardless of the mating context (short-term: t = − 0.165, p = 0.87; long-term: t = 0.044, p = 0.96, Table 2). Difference in masculinity preference between short- and long-term mating contexts
Fig. 1. Timing of ovulation day: number of women that ovulated on each of the observed cycle days counted backwards from the onset of the next menses.
Analysis 1 When the fertile phase was considered to be between 2 days prior to and 2 days after ovulation, there was no significant difference in masculinity preference between short- and long-term mating contexts in either the low- or high-conception probability groups (t = 1.09, p = 0.28 and t = 1.92, p = 0.07, respectively, Table 3).
Detecting ovulation Based on estradiol levels, we determined the day of ovulation for each participant. Ovulation was considered to occur on the day of the greatest drop in E2 following the estradiol peak (Lipson and Ellison, 1996). Ovulation day varied among participants, falling between the 11th and 17th days of the cycle (counting backwards from the first day of next menstrual bleeding, Fig. 1). Cycle length varied between 23 and 38 days (average = 28.5 days, SD = 3.52, Table 1). Depending on the day of the cycle on which a participant evaluated the photographs, they were divided into high- and low-conception probability groups according to 3 different grouping methods used in the literature. Difference in the masculinity preferences of women in high- and lowconception probability groups Analysis 1 Following Macrae et al. (2002) we divided women into highconception (2 days before and until the day of ovulation, N = 15) and low-conception probability groups (all other days, N = 100). Groups did not differ significantly in their average preference for masculinity either in the short-term (t = −0.43, p = 0.66) or long-term context (t = 0.24, p = 0.81, Table 2). Analysis 2 Following Harris (2011), the high-conception probability phase was considered to be 6 days prior and 1 day after the ovulation (N = 43), and the low-conception probability phase included all other days (N = 72). Calculations in the current work are a translation of the previously used forward counting method (Harris, 2011), so that range of menstrual cycle days is centered around hormonal measures of when ovulation occurred. There were no statistically significant differences in masculinity preferences between low- and high-conception
Analysis 2 When the fertile phase was considered to be between 6 days prior to and 1 day after ovulation, there was no significant difference in preference between short- and long-term mating contexts in either the lowor high-conception probability groups (t = 1.23, p = 0.22 and t = 1.35, p = 0.18, respectively, Table 3). Analysis 3 When the high-conception probability phase was considered to be between 8 days prior to and the day of ovulation, there was no significant difference in preference between short- and long-term mating contexts for low- or high-conception probability groups (t = 1.75 p = 0.08 and t = 0.68, p = 0.50, respectively, Table 3). Relation between estradiol and masculinity preference Morning estradiol levels on the day of the survey ranged between 4 and 78 pmol/L (SD = 14.35, Table 1). Based on daily measurements, average levels of estradiol per cycle were calculated for each woman. Average estradiol levels were between 5.4 and 59.5 pmol/L (SD = 9.33). Estradiol on the day of the survey and masculinity preference Only women who provided saliva sample for the day of completing the face rating task were included in this analysis (N = 83). There was no statistically significant relationship between the level of estradiol on the day of the survey and masculinity preference (long-term context: β = 0.106, SE = 0.110, p = 0.34 and short-term context: β = 0.031, SE = 0.11, p = 0.78, Table 4). Average estradiol and masculinity preference Mean estradiol level throughout the menstrual cycle was not related to masculinity preferences in either mating context (long-term: β = 0.021, SE = 0.094, p = 0.82 and short-term: β = 0.002, SE = 0.092, p = 0.98, Table 4).
Table 1 Descriptive statistics for cycle length and hormonal profile.
Length of the cycle (days) Timing of the ovulation day Estradiol levels (pmol/L) in the day of completing the questionnaire Average estradiol levels (pmol/L) for the cycle
N
Average
Minimum
Maximum
SD
115 115 83 115
28.5 −13.44 20.5 18.51
23 −17 4 5.4
38 −11 78 59.5
3.52 1.53 14.35 9.36
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7
5
Table 2 Average masculinity preference in two different mating contexts and differences between high and low conception probability group (between participant design). Preference in short-term mating context High conception defined as:
From −2 to ovulation day (Analysis 1) From −6 to +1 days around ovulation (Analysis 2) From −8 to ovulation day (Analysis 3)
High conception probability phase
Low conception probability
Mean SD
Mean SD
N
57.3
6.59 15 58.5
58.1
10.03 43 58.4
57.7
9.59 42 58.7
Preference in long-term mating context t
p
Effect size
CI 95%
N
High conception probability phase
Low conception probability
Mean SD
Mean SD
t
p
Effect size
CI 95%
10.07 100 0.436 0.664 −0.12 [−0.67, 0.42] 56.3
10.44 56.9
9.51
72 0.165 0.869 −0.03 [−0.41, 0.35] 56.2
9.65 57.3
9.56 0.548 0.585 −0.11 [−0.49, 0.26]
9.75
73 0.165 0.869 −0.10 [−0.48, 0.28] 56.8
9.81 56.9
9.49 0.044 0.965 −0.01 [−0.39, 0.37]
Discussion Regardless of mating context and conception probability, participants judged masculinized faces to be slightly more attractive. Male facial masculinity preferences, however, did not depend on woman's potential fertility throughout the menstrual cycle. When conception probability was assessed based on daily measurements of estradiol levels, women with high- and low-conception probabilities did not differ in their preferences, regardless of the assumed length of the relationship (i.e., within-subjects short- or long-term mating context trials). A number of previous studies also did not find cyclical shifts in women's preferences (Harris, 2013, Peters et al., 2009, Scott et al., 2013, Welling et al., 2008). Furthermore, masculinity preferences were not related to average levels of estradiol in the cycle or estradiol levels on the morning of the survey. Neither age nor relationship status of the participant influenced her preference for masculinity. In this study, we measured daily estradiol levels in women's saliva throughout the entire menstrual cycle. Day of ovulation was estimated based on hormonal measurements and varied among participants (between the 11th and 17th days, counting backwards from the end of the cycle). Our results provide an important critique of the studies that do not take into consideration inter-individual variation in cycle length and ovulation timing. We have documented that interindividual variation in timing of ovulation is high, even among women with regular menstrual cycles. Additionally, self-reported lengths of cycle are prone to measurement error (e.g., 43% of women misreported their cycle length by 2 or more days, Small et al., 2007), and women often report incorrect days of their previous menses (Wegienka and Baird, 2005). In our study, we measured estradiol levels; therefore, it is not possible that the null effect of conception probability on preferences is due to a miscalculation of the ovulation day. Moreover, we have analyzed the preferences of women in the same manner as some previous authors (Analyses 1 to 3 in the Results), and failed to find significant differences between groups with high and low probabilities of conception.
10.45 0.24
0.812 −0.06 [−0.6, 0.49]
There are several possible reasons why preference shifts throughout the menstrual cycle were not observed. As subtle facial trait differences in men may not have been relevant to the reproductive success of women in the past, it is unlikely that selection would promote preference shifts in women every few weeks or days. It is also possible that a woman's preference for a stable, investing partner overrides any cyclical shift of preference. Men with highly exaggerated secondary sexual features may be perceived as unlikely to invest in potential offspring; masculinity and testosterone are positively correlated with enhanced sexual activity (Halpern et al., 1993) and so interest in extra-pair liaisons may reduce interest in monogamous relationships (McIntyre et al., 2009). Moreover, even examining only the short-term preferences of women in relationships, it may still be possible that extra-pair copulations with men with pronounced masculinity are too risky to outweigh the potential cost of losing the investment of stable partner. In such scenario, masculinity preference shifts throughout the menstrual cycle would not be adaptive. Although, it is probable that women's sexual desires might still be triggered during fertile phase, they would not necessarily act on them. Participants in our study showed slight preferences for masculinized men; however, the preference was not especially strong in either short- or long-term mating contexts (varying from 54.4 to 58.7 on a scale in which 50% is no preference and 100% is a very strong preference for masculinized faces). Our results regarding inter-individual differences support the findings of Roney et al. (2011), who also grouped women into high- and low-conception probability groups based on estradiol levels; these authors also failed to find a difference in masculinity preferences between groups. They did find, however, a difference in preferences for androgen-dependent facial cues. It is possible that facial masculinity does not perfectly correspond to high circulating levels of testosterone (Peters et al., 2009) and hence may trigger different preference responses. Furthermore, it is possible that previously found correlation between men's genetic quality and masculinity (Penton-Voak and Perrett, 2000) is not as straightforward as assumed hitherto (immunocompetence handicap hypothesis, Folstad and Karter, 1992). There has been only one study reporting small but significant correlation between
Table 3 Average masculinity preferences in different mating contexts for two fertility groups (within participant analysis). High conception phase defined as:
Between −2 and ovulation day (Analysis 1) Between −6 and +1 days around ovulation (Analysis 2) Between −8 and ovulation day (Analysis 3)
Menstrual phase of the participant
Low-conception probability phase (N = 100) High-conception probability phase (N = 15) Low-conception probability phase (N = 72) High-conception probability phase (N = 43) Low-conception probability phase (N = 73) High-conception probability phase (N = 42)
Preference in long-term mating context
Preference in short-term mating context
Mean
SD
Mean
SD
56.96 56.32 57.3 56.2 56.9 56.8
9.48 10.45 9.56 9.65 9.49 9.81
58.49 57.31 58.4 58.1 58.7 57.7
10.07 6.59 9.51 10.03 9.76 9.59
t
p
Effect size
CI 95%
1.73 0.54 1.23 1.35 1.75 0.68
0.086 0.594 0.221 0.18 0.08 0.50
−0.16 −0.11 −0.12 −0.19 −0.19 −0.09
[−0.43, 0.12] [−0.83, 0.61] [−0.44, 0.21] [−0.61, 0.23] [−0.51, 0.14] [−0.52, 0.34]
6
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7
Table 4 Correlation between masculinity preference and mean estradiol levels throughout the cycle and level of estradiol from the day of taking the survey, in two mating contexts. N
Mean estradiol levels during the cycle Morning estradiol levels on the day of completing the survey
115 83
masculinity and good health (Rhodes et al., 2003). It is possible that the lack of shift in masculinity preferences found in our study, is not univocal with lack of fluctuation preference towards purportedly good genes (following Harris, 2011). It could be that within woman measures of shifts across the cycle are likely to be much more sensitive, and the result of our study can be treat as a between subjects effect. The null results reported in our study might also have been caused by the fact that the manipulated photos did not simulate variation in facial masculinity in a manner that was sufficiently pronounced to be perceived by women raters. The usage of 2D, immobile composite faces rather than real faces may have contributed to the discrepancy between our and previously published results (Ellison, 2008). Furthermore, given that within woman measures of shifts across the cycle are likely to be much more sensitive to small changes than between women measurement, it might be possible that our null results are a result of inter-individual, cross-sectional design (unlike the studies that found significant positive relation between hormones and masculinity preference, e.g. Roney and Simmons, 2008). We would want however to underline, that due to detailed hormonal measurements we were able to, very precisely, segregate our participants into high and low conception probability phases, what should increase the probability of finding the putative effect. In our study, ovulation was assessed based on estradiol levels, and ovulation was not confirmed by an LH surge. We followed published methods of identifying the ovulation by the largest mid-cycle drop of estradiol levels, which occurs in response to hormonal changes related to ovulation (Lipson and Ellison, 1996). Use of this methodology may result in misclassifying an anovulatory cycle as an ovulatory cycle. However, we excluded women without a reliably identified mid-cycle estradiol drop from our analyses. In the majority of studies on fluctuating preferences, hormonal measurements are not included; therefore obtaining daily estradiol levels is one of the most detailed methodologies to date. It is also possible that collecting more in-depth sociodemographic data from participants would allow us to detect more subtle preference changes (and covariance of such changes with, for example, sexual openness or number of children). Results of this study strongly suggest the necessity of incorporating physiological measures of steroid sex hormones in future studies of preference fluctuations. Caution should be exercised when evaluating existing studies that report robust cyclical shifts in preference for facial masculinity in men based solely on counting days methods (especially forward counting one).
Funding statement This study was supported by grants from the National Science Centre (grant no. 2014/12/S/NZ8/00722 to UMM and grant no. N N404 273440 to GJ) and the Ministry of Science and Higher Education (grant no. IdP2011 000161 to GJ).
Acknowledgments We would like to thank all participants, Dr. Susan Lipson, Dr. Anna Ziomkiewicz and our study assistants. We also thank Prof. David Perrett and Dr. Ben Jones for providing the visual stimuli.
Long-term mating context
Short-term mating context
β
SE
p
β
SE
p
0.021 0.106
0.094 0.110
0.82 0.34
−0.002 0.031
0.092 0.111
0.98 0.78
References Bobst, C., Sauter, S., Foppa, A., Lobmaier, J.S., 2014. Early follicular testosterone level predicts preference for masculinity in male faces — but not for women taking hormonal contraception. Psychoneuroendocrino. 41, 142–150. Chiazze, L., Brayer, F.T., Macisco, J.J., Parker, M.P., Duffy, B.J., 1968. Length and variability of human menstrual cycle. J. Amer. Assoc. 203 (6), 377–380. Danel, D., Pawlowski, B., 2006. Attractiveness of men's faces in relation to women's phase of menstrual cycle. Collegium Antropol. 30 (2), 285–289. DeBruine, L., Jones, B.C., Frederick, D.A., Haselton, M.G., PentonVoak, I.S., Perrett, D.I., 2010. Evidence for menstrual cycle shifts in women's preferences for masculinity: A response to Harris (in press) “Menstrual cycle and facial preferences reconsidered”. Evol. Psych. 8, 768–775. Durante, K.M., Li, N.P., 2009. Oestradiol level and opportunistic mating in women. Biol. Lett. 5, 179–182. Ellison, P.T., 2008. Beauty: in the gonads of the beholder — and the beheld — commentary. Horm. Behav. 53, 11–13. Folstad, I., Karter, A.J., 1992. Parasites, bright males, and the immunocompetence handicap. Am. Nat. 139, 603–622. Garver-Apgar, C.E., Gangestad, S.W., Thornhill, R., 2008. Hormonal correlates of women's mid-cycle preferences for the scent of symmetry. Evol. Hum. Behav. 29, 223–232. Gildersleeve, K., Haselton, M.G., Fales, M., 2014a. Do women's mate preferences change across the ovulatory cycle? A meta-analytic review. Psychol. Bull. http://dx.doi.org/ 10.1037/a0035438 Gildersleeve, K., Haselton, M.G., Fales, M., 2014b. Meta-analyses and P-curves support robust cycle shifts in women's mate preferences: reply to Wood and Carden (2014) and Harris, Pashler, and Mickes (2014). Psychol. Bull. 140, 1272–1280. Harris, C.R., 2011. Menstrual cycle and facial preferences reconsidered. Sex Roles 64, 669–681. Harris, C.R., 2013. Shifts in masculinity preferences across the menstrual cycle: still not there. Sex Roles 69, 507–515. Halpern, C.T., Udry, J.R., Suchindran, C., 1993. Testosterone and pubertal development as predictor of sexual-activity: a panel analysis of adolescent males. Psychosom. Med. 27, 445–465. Hawkinsa, S.M., Matzuk, M.M., 2008. The menstrual cycle — basic biology. Ann. N. Y. Acad. Sci. 1135, 10–18. Jasienska, G., Ziomkiewicz, A., Ellison, P.T., Lipson, S.F., Thune, I., 2004. Large breasts and narrow waists indicate high reproductive potential in women. Proc. R. Soc. London B. 271, 1213–1217. Jasienska, G., Ziomkiewicz, A., Thune, I., Lipson, S., Ellison, P., 2006. Habitual physical activity and estradiol levels in women of reproductive age. Eur. J. Cancer Prev. 15, 439–445. Jasienska, G., Jasienski, M., 2008. Interpopulation, interindividual, intercycle, and intracycle natural variation in progesterone levels: a quantitative assessment and implications for population studies. Am. J. Hum. Biol. 20, 35–42. Jasienska, G., 2013. The Fragile Wisdom. An Evolutionary View on Women's Biology and Health. Harvard University Press, Cambridge, MA. Lipson, S.F., Ellison, P.T., 1996. Comparison of salivary steroid profiles in naturally occurring conception and non-conception cycles. Hum. Reprod. 11 (10), 2090–2096. Little, A.C., Jones, B.C., 2012. Variation in facial masculinity and symmetry preferences across the menstrual cycle is moderated by relationship context. Psychoneuroendocrino. 37, 999–1008. Little, A.C., Jones, B.C., Burriss, R.P., 2007. Preferences for masculinity in male bodies change across the menstrual cycle. Horm. Behav. 31, 633–639. Little, A.C., Jones, B.C., DeBruine, L.M., 2008. Preferences for variation in masculinity in real male faces change across the menstrual cycle: women prefer more masculine faces when they are more fertile. Pers. Indiv. Differ. 45, 478–482. Little, A.C., 2013. The influence of steroid sex hormones on the cognitive and emotional processing of visual stimuli in humans. Front. Neuroindocrin. 34, 315–328. Macrae, C.N., Alnwick, K.A., Milne, A.B., Schloerscheidt, A.M., 2002. Person perception across the menstrual cycle: hormonal influences on social-cognitive functioning. Psychol. Sci. 13, 532–536. Marcinkowska, U.M., Kozlov, M., Cai, H., Contreras-Garduño, J., Dixson, B.J., Oana, G., Kaminski, G., Li, N.P., Lyons, M.T., Onyishi, I.E., Prasai, K., Pazhoohi, F., Prokop, P., Rosales Cardozo, S.L., Sydney, N., Yong, J.C., Rantala, M.J., 2014. Cross-cultural variation in men's preference for sexual dimorphism in women's faces. Biol. Lett. 10, 20130850. McIntyre, M., Gangestad, S.W., Gray, P.B., Chapman, J.F., Burnham, T.C., O'ROurke, M.T., Morrison, E.R., Clark, A.P., Gralewski, L., Campbell, N., Penton-Voak, I.S., 2009. Women's probability of conception is associated with their preferences for flirtatious but not masculine facial movement. Arch. Sex. Behav. 39, 1297–1304. Pawlowski, B., Jasienska, G., 2005. Women's preferences for sexual dimorphism in height depend on menstrual cycle phase and expected duration of relationship. Biol. Psychol. 70, 38–43.
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7 Penton-Voak, I.S., Perrett, D.I., 2000. Female preferences for male faces changes cyclically: Further evidence. Evol. Hum. Behav. 21, 39–48. Penton-Voak, I.S., Perrett, D.I., Castles, D.I., Kobayashi, T., Burt, D.M., Murray, L.K., Minamisawa, R., 1999. Menstrual cycle alters face preference. Nature 399, 741–742. Peters, M., Simmons, L.W., Rhodes, G., 2009. Preferences across the menstrual cycle for masculinity and symmetry in photographs of male faces and bodies. PLoS One 4, e4138. Perrett, D.I., Lee, K.J., Penton-Voak, I.S., Rowland, D.R., Yoshikawa, S., Burt, D.M., Akamatsu, S., 1998. Effects of sexual dimorphism on facial attractiveness. Nature 394, 884–887. Pisanski, K., Hahn, A.C., Fisher, C.I., DeBruine, L.M., Feinberg, D.R., Jones, B.C., 2014. Changes in salivary estradiol predict changes in women's preferences for vocal masculinity. Horm. Behav. 66 (3), 493–497. Puts, D.A., 2005. Mating context and menstrual phase affect women's preferences for male voice pitch. Evol. Hum. Behav. 26, 388–397. Rhodes, G., Chan, J., Zebrowitz, L.A., Simmons, L.W., 2003. Does sexual dimorphism in human faces signal health? Proc. R. Soc. London B. 270 (Suppl. 1), S93–S95. Roney, J.R., Simmons, Z.L., 2008. Women's estradiol predicts preference for facial cues of men's testosterone. Horm. Behav. 53, 14–19. Roney, J.R., Simmons, Z.L., Gray, P.B., 2011. Changes in estradiol predict within-women shifts in attraction to facial cues of men's testosterone. Psychoneuroendocrino. 36 (5), 742–749. Rosen, M.K., Lopes, H.H., 2009. Menstrual cycle shifts in attentional bias for courtship language. Evol. Hum. Behanv. 30, 131–140. Scott, I.M.L., Clarke, A.O., Boothroyd, L.G., Penton-Voak, I.S., 2013. Do men's faces really signal heritable immunocompetence? Behav. Ecol. 24, 579–589.
7
Small, C.M., Manatunga, A.K., Marcus, M., 2007. Validity of self-reported menstrual cycle length. Ann. Epidemiol. 17, 163–170. Thornhill, R., Chapman, J.F., Gangestad, S.W., 2013. Women's preferences for men's scents associated with testosterone and cortisol levels: Patterns across the ovulatory cycle. Evol. Hum. Behav. 34, 216–221. Wegienka, G., Baird, D.D., 2005. A comparison of recalled date of last menstrual period with prospectively recorded dates. J. Women's Health 14, 248–252. Welling, L.L.M., Jones, B.C., DeBruine, L.M., Conway, C.A., Law Smith, M.J., Little, A.C., Feinberg, D.R., Sharp, M., Al Dujaili, E.A.S., 2007. Raised salivary testosterone in women is associated with increased attraction to masculine faces. Horm. Behav. 52, 156–161. Welling, L.L., Jones, B.C., DeBruine, L.M., Smith, F.G., Feinberg, D.R., Little, A.C., Al Dujaili, E.A.S., 2008. Men report stronger attraction to femininity in women's faces when their testosterone levels are high. Horm. Behav. 54, 703–708. Wilcox, A., Dunson, D., Baird, D.D., 2000. The timing of the fertile window in the menstrual cycle: day specific estimates from a prospective study. Brit. Med. J. 321, 1259–1262. Wood, W., 2014. Author reply: once again, menstrual cycles and mate preferences. Emotion 6, 258–260. Wood, E., Carden, L., 2014. Elusiveness of menstrual cycle effects on mate preferences: Comment on Gildersleeve, Halton, and Fales (2014). Psych. Bull. 140, 1265–1271. Wood, W., Kressel, L., Joshi, P.D., Louie, B., 2014. Meta-analysis of menstrual cycle effects on women's mate preferences. Emotion 0, 1–21.
Contents lists available at ScienceDirect
Hormones and Behavior journal homepage: www.elsevier.com/locate/yhbeh
Lack of support for relation between woman's masculinity preference, estradiol level and mating context Urszula M. Marcinkowska a,⁎, Peter T. Ellison b, Andrzej Galbarczyk a, Karolina Milkowska a, Boguslaw Pawlowski c, Inger Thune d, Grazyna Jasienska a,e a
Department of Environmental Health, Faculty of Health Sciences, Jagiellonian University Medical College, 20 Grzegorzecka St., 31-531 Krakow, Poland Department of Human Evolutionary Biology, Harvard University, 11 Divinity Av., Cambridge, MA 02138, USA Department of Human Biology, University of Wrocław, Kuźnicza 35, Wrocław, Poland d Department of Oncology, Oslo University Hospital, Ullevål, Oslo, Norway e Department of Anthropology, Yale University, 10 Sachem Street, New Haven, CT 06511-3707, USA b c
a r t i c l e
i n f o
Article history: Received 29 December 2014 Revised 5 October 2015 Accepted 20 October 2015 Available online 21 October 2015 Keywords: Human sexual preferences Facial preferences Sex steroid hormones Estradiol Sexual dimorphism Masculinity
a b s t r a c t It has been proposed that women's preferences for male facial sexual dimorphism are positively correlated with conception probability and differ between short- and long-term mating contexts. In this study, we tested this assumption by analyzing relationships between estradiol levels to the women's preferences of male faces that were manipulated to vary in masculinity. Estradiol was measured in daily saliva samples throughout the entire menstrual cycle collected by Polish women with regular menstrual cycles. In our analyses, we included the three most commonly used definitions of the fertile window in the literature. After computing the overall masculinity preference of each participant and measuring hormone levels, we found that i) the timing of ovulation varied greatly among women (between −11 and −17 days from the onset of the next menses, counting backwards), ii) there was no relationship between daily, measured during the day of the test (N = 83) or average for the cycle (N = 115) estradiol levels and masculinity preferences, iii) there were no differences in masculinity preferences between women in low- and high-conception probability phases of the cycle, and iv) there were no differences in masculinity preferences between short- and long-term mating contexts. Our results do not support the idea that women's preferences for a potential sexual partner's facial masculinity fluctuate throughout the cycle. © 2015 Elsevier Inc. All rights reserved.
Introduction Many studies suggest that women's preferences for male traits such as masculinity, symmetry and dominance differ between fertile and non-fertile phases of the menstrual cycle. When in fertile phase, women were found to exhibit stronger preferences for more masculine voices (Puts, 2005), body odors (Thornhill et al., 2013) and body shapes (Little et al., 2007) or increased height (Pawlowski and Jasienska, 2005). Many studies have also documented that women exhibit stronger preferences for facial masculinity around the time of ovulation (e.g., PentonVoak and Perrett, 2000). This fluctuation in preference for masculinity may have evolved as a trade-off between choosing cooperative, warm, and caregiving (i.e., less masculine) partners that are more suited to long-term relationships and partners with purported higher genetic
⁎ Corresponding author at: Department of Environmental Health, Faculty of Health Sciences, Jagiellonian University Medical College, 20 Grzegorzecka St., 31-531 Krakow, Poland. E-mail address: [email protected] (U.M. Marcinkowska).
http://dx.doi.org/10.1016/j.yhbeh.2015.10.012 0018-506X/© 2015 Elsevier Inc. All rights reserved.
quality (i.e., more masculine, Perrett et al., 1998). Despite numerous studies, common agreement regarding the evolutionary bases (or the occurrence) of this fluctuation is lacking (for a debate on this topic, see DeBruine et al. (2010); Gildersleeve et al. (2014a, 2014b); Harris (2011); Wood (2014)).
Hormonal fluctuations As the menstrual cycle is characterized by changing levels of steroid hormones, it is likely that changes in women's preferences throughout the cycle may be primarily attributable to fluctuating hormone levels (Bobst et al., 2014). The menstrual cycle has an average length of 28 days and comprises two ovarian phases: the follicular phase that begins on the first day of menstrual bleeding and continues until ovulation, and the luteal phase that begins after ovulation and continues until the day before next menstrual bleeding (Hawkinsa and Matzuk, 2008). In fully functioning cycles, during the follicular phase, a mature follicle is produced, which then releases an egg in a process called ovulation. The ovary secretes estradiol, progesterone and inhibins. Mid-
2
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7
cycle drop in estradiol is a reasonable estimate of the day of ovulation (Lipson and Ellison, 1996).
When does the fertile phase occur? A body of evidence has found that pre-menopausal women in the most fertile phase of their cycle (and who do not take hormonal contraceptives) have stronger preferences for more masculine faces than they do in their less fertile phases (for a review, see DeBruine et al., 2010). Some studies, however, did not replicate these results (for a metaanalysis, see Gildersleeve et al., 2014aand Wood et al., 2014). One limitation of studies on masculinity preferences and their fluctuations throughout the menstrual cycle is a lack of consistency in the methods used to determine the fertile window. Authors usually do not measure levels of steroid hormones and do not use physiological methods to detect ovulation, but assign women to being in “fertile” or “non-fertile” phase based on a cycle length. This methodology is problematic due to a high variability in menstrual cycle length (Chiazze et al., 1968) and intra- and inter-individual variation in levels of reproductive steroid hormones (Jasienska and Jasienski, 2008). Penton-Voak et al. (1999) defined the fertile phase as ranging from the end of previous menses to 14 days prior to next menses. In a later study, Penton-Voak and Perrett (2000) defined the high-conception probability phase as between 6 and 14 days from the onset of the previous menses. Macrae et al. (2002) and Garver-Apgar et al. (2008) proposed using a backward-counting method to estimate the fertile window; their proposed high-conception probability phases were, respectively, from 2 or 4 days prior to the ovulation. Some authors combined different approaches to determine the fertile window (i.e., Harris, 2011), which resulted in a 5-day window around ovulation (for more examples see meta-analysis by Gildersleeve et al., 2014b). To reliably use these methods, it must be assumed that cycle length and, particularly, the length of the luteal phase is consistent among participants. From studies of hormonal levels in naturally cycling women, we know that this assumption is problematic; multiple factors (Wilcox et al., 2000, Jasienska, 2013) cause cycle lengths to vary within and among participants (Jasienska and Jasienski, 2008). Although it is well established that levels of steroid hormones fluctuate during a cycle, there are surprisingly few studies on mate preferences that incorporate actual measurements of hormonal levels in naturally cycling women (Pisanski et al., 2014, Roney and Simmons, 2008).
Fluctuation of preferences Ovarian steroid hormones play an important role in female sexual motivation and behavior (Durante and Li, 2009) and overall preference of male facial attractiveness is the highest during the fertile phase of the cycle (Danel and Pawlowski, 2006). Other studies showed that general preference shifts occur only when attractiveness is judged in a shortterm, and not in a long-term context (Little and Jones, 2012) or only for women who were in a committed relationship (Little et al., 2008). However, the agreement on the robustness of women's cyclical preferences shifts is lacking. In the most recent meta-analysis of studies investigating menstrual cycle shifts in women's preferences, Gildersleeve et al. (2014a) reported “… analyses revealed robust cycle shifts that were specific to women's preferences for hypothesized cues of (ancestral) genetic quality…”. In contrast, a meta-analysis by Wood et al. (2014) was “largely non-supportive [of this hypothesis]. Specifically, fertile women did not desire sex in short-term relationships with men purported to be of high genetic quality (i.e., high masculinity …)”. It has been suggested though, that lack of support for cyclical shifts in Wood et al. (2014) could have been caused by differences in methodological and analytic decisions made by the authors (Gildersleeve et al., 2014b, Wood and Carden, 2014).
Hormones and masculinity preferences Although sex steroid hormones were suggested to be associated with visual processing (Little, 2013) there are surprisingly few studies of preferences that directly measured hormone levels, and those that did measure hormonal levels had conflicting results. Welling et al. (2007), based on within-participants measurements (N = 70), identified a positive correlation of preferences for masculinity and testosterone but failed to find a relationship with estradiol or progesterone. Similar relationships were documented by Bobst et al. (2014) for 27 cycling women. Rosen and Lopes (2009) found in a betweenparticipants study that courtship language preferences fluctuated depending on estradiol levels of participants (N = 17) but not on progesterone or testosterone levels. Salivary estradiol fluctuations have also been found to best predict changes in women's (N = 62) preferences for vocal masculinity (Pisanski et al., 2014). Finally, Roney and Simmons (2008) reported a positive correlation of woman's (N = 75) estradiol levels and preference for the faces of males with high circulating testosterone levels. It should be noted that the studies on preferences that measured hormone levels were not free from methodological problems, including having as participants young women, who are known to have irregular cycles and cycles with low hormone levels, estimating cycle length, taking a single or only a few samples for hormonal assessments from each participant, collecting samples at different times of a day, comparing hormonal values obtained from saliva samples to previously published estimates that were obtained from serum samples, or having small number of participating women. In our sample we aimed to repeat preference measurements correcting for all above mentioned flaws.
Study aim In this study, we sought to identify the putative fluctuation of facial masculinity preferences exhibited by women throughout the menstrual cycle. Because we measured estradiol levels in saliva samples collected daily for the entire cycle, we were able to detect the day of ovulation based on physiology, rather than estimating ovulation by counting the days of the cycle (i.e., the method used in most previous studies). We compared the masculinity preferences of women from high- and lowconception probability groups (between participant measurement) and also tested how preferences differed within those two groups depending on the mating context (within-participant measurement). Furthermore, we tested for relation between masculinity preferences and estradiol levels of participants. To add to the ongoing debate about the existence of fluctuating masculinity preferences, we conducted 4 separate analyses that compared results based on previously used definitions of the fertile window. Definitions came from studies that did not include hormonal measurements.
Methods Participants A total of 115 women from Poland (average age = 29.9, SD = 3.46) participated in this study. Women were recruited between June 2001 and June 2003 by posted advertisements in public places and by newspaper, radio and television. On the basis of an initial interview, the women that were selected to participate in this study were limited to those who met following criteria: between 24 and 36 years of age, regular menstrual cycles and no fertility problems, no gynecological or chronic disorders (e.g., diabetes, hypo/hyperthyroidism), not pregnant or lactating in the 6 months prior to recruitment, and not using hormonal contraception or taking any hormonal medication in the 6 months prior to recruitment.
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7
Questionnaire and visual stimuli After agreeing to participate in the study and completing a written consent form, participants answered a series of socio-demographic questions including relationship status (“single”, “in a relationship” or “hard to say”) and were instructed to collect saliva samples daily for the entire menstrual cycle. During the subsequent visit women judged 18 photographs of men (9 slides, each depicting 2 manipulated male faces). Visual stimuli were manipulated according to previous studies (Perrett et al., 1998). Women were presented with 2 morphed faces that were either 40% feminized or 40% masculinized version of same composite face (Penton-Voak et al., 1999, for the description of the morphing technique, see e.g., Marcinkowska et al., 2014). Composite male faces differed only in the particular shape that was manipulated to be more or less masculine, and all other characteristics (i.e., skin texture or symmetry) were held constant. Photographs were presented to women in a random order and were evaluated twice, once in a shortterm and once in a long-term relationship context. Masculinized and feminized versions of same face were presented on each page. Surveying was conducted in participants' homes or in the lab depending on a preference of a participant. Time and date of testing was marked by the member of the research group and included in the personal file of the participant. Survey completion timing varied from − 29 to − 1 day, counting backwards from the onset of the following menses (Mean = − 13.16 days). During the first evaluation, participants were asked to judge the attractiveness of male faces in the context of a potential short-term relationship described as follows: “You are looking for the type of person who would be attractive in a short-term relationship. This implies that the relationship may not last a long time. Examples of this type of relationship would include a single date accepted on the spur of the moment, an affair within a long-term relationship, and possibility of a one-night stand”. During the second evaluation, attractiveness was judged based on a potential long-term relationship, introduced as follows: “You are looking for the type of person who would be attractive in a long-term relationship”. Examples of this type of relationship would include someone you may want to move in with, someone for whom you may consider leaving a current partner, and someone you may, at some point, wish to marry (or have a relationship similar to marriage). In both evaluations participants were asked to choose the more attractive of the two faces and then indicate the extent to which they preferred a particular face by indicating if they had ‘almost no preference’, ‘slight preference’, ‘average preference’ or ‘strong preference’. Responses were recoded using a scale from 1 to 8, in which 1 indicated a strong preference for a feminized face and 8 indicated a strong preference for a masculinized face. Next, the mean preference for masculinity was calculated for each participant. Scores were converted to percentages, such that 0% represented a maximal preference for feminized faces and 100% represented a maximal preference for masculinized faces. Preferences variation was based on a between participants analysis. Hormonal measures Steroid hormones were measured in saliva samples collected daily by the participants themselves in their homes during their menstrual cycles. Prior to sample collection, each woman was instructed how to collect and store samples of saliva. Each participant received a set of 38 plastic tubes sealed with a stopper, 38 sugar-free sticks of laboratory-tested chewing gum, and calendars in which they recorded the collection date and time of each saliva sample. Participants collected daily morning saliva samples from the first day of menstruation to the day prior to the next menstrual bleeding. A total of 5.3% of daily samples were missing due to inappropriate collection or loss during laboratory procedures (Jasienska et al., 2006). Saliva samples were analyzed for 17-β estradiol (E2) by the Reproductive Ecology Laboratory at Harvard University, following a previously described procedure (Jasienska et al.,
3
2006). E2 was analyzed in saliva samples from 24 days of each cycle using an I-125-based radioimmunoassay kit (#39100, Diagnostic Systems Laboratories, Webster, Texas, USA) with published (Jasienska et al., 2004) modifications to the manufacturer's protocol. The quality of laboratory analysis was monitored by including samples of known concentrations (“pools”) with low (follicular), medium (luteal) and high (gestational) concentrations. The sensitivity of the E2 assay was 4 pmol/l. Average intra-assay variability was 9%, and inter-assay variability ranged from 23% for lower values (15 pmol/l) to 13% for higher values (50 pmol/l). Prior to statistical analyses, menstrual cycles were aligned based on the day of the mid-cycle E2 drop (i.e., day 0), which provides a reasonable estimate of the day of ovulation, according to published methods (Lipson and Ellison, 1996). Ovulation day defined in such way was used to divide participants into low and highconception probability groups. E2 values from 18 consecutive days of each cycle that were aligned on day 0 were used in analyses. Participants, whose mid-cycle estradiol drop could not be reliably identified, or the cycle had more than one estradiol drop; were not included in the analysis. The method used in this study is very detailed and narrows the possibility of erroneous fertility phase classification. Statistical analyses Participants were divided into two groups (high- and lowconception probability) based on their hormonal levels and day of the cycle during which they took a test of masculinity preference (Detecting ovulation). The placement of participants into these two groups was based on 3 separate definitions of a fertile window, so that our results could be compared to previous studies (hence the subdivision of the Results section into 3 analyses), all based on the counting days around the ovulation day. Whether the average masculinity preference within a group differed from random (50%) was tested using single sample t-tests. In the first analysis (t-test for independent samples), the average masculinity preference of women from the high-conception probability group was compared to the average preference of women from the low-conception probability group. This comparison was performed separately for short- and the long-term mating contexts (Difference in the masculinity preferences of women in high- and low-conception probability groups). In the second analysis (t-test for dependent samples), we compared the preferences of women within the two conception probability groups, depending on their mating context. Specifically, we examined whether a woman's preferences in a short-term mating context differed from her preferences in a longterm mating context (this comparison was performed separately for the high- and low-conception probability groups, see Difference in masculinity preference between short- and long-term mating contexts). Lastly, we tested with simple linear regression whether average estradiol levels throughout the cycle (Average estradiol and masculinity preference) and morning levels of estradiol on the day of completing the survey (Estradiol on the day of the survey and masculinity preference) were related to masculinity preferences in short- and long-term mating contexts. Confidence intervals and effect sizes can be found in Tables 2 and 3. Results On average, women in low- and high-conception probability phase preferred more masculine faces, in both short- and long-term mating contexts (58.33%, SD = 9.66, p b 0.001 and 57.87%, SD = 9.57, p b 0.001, respectively). Participant age did not influence masculinity preferences (N = 115, β = 0.013, p = 0.89 in a short-term context, and β = − 0.021, p = 0.83 in a long-term context). Preference of women did not differ depending on relationship status (paired vs. single) in either a short- (t = 0.488, p = 0.63) or long-term (t = 1.58, p = 0.12) mating context.
4
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7
probability groups in either a short- or long-term context (t = 0.165, p = 0.87 and t = 0.548, p = 0.58, respectively, Table 2). Analysis 3 Following Penton-Voak and Perrett (2000), the high-conception phase was from 8 days before to the day of ovulation (N = 42), and the low-conception phase included all other days of the cycle (N = 73). As in the previous analysis, calculation is a translation of forward calculations (Penton-Voak and Perrett, 2000). Masculinity preference did not significantly differ between the two groups, regardless of the mating context (short-term: t = − 0.165, p = 0.87; long-term: t = 0.044, p = 0.96, Table 2). Difference in masculinity preference between short- and long-term mating contexts
Fig. 1. Timing of ovulation day: number of women that ovulated on each of the observed cycle days counted backwards from the onset of the next menses.
Analysis 1 When the fertile phase was considered to be between 2 days prior to and 2 days after ovulation, there was no significant difference in masculinity preference between short- and long-term mating contexts in either the low- or high-conception probability groups (t = 1.09, p = 0.28 and t = 1.92, p = 0.07, respectively, Table 3).
Detecting ovulation Based on estradiol levels, we determined the day of ovulation for each participant. Ovulation was considered to occur on the day of the greatest drop in E2 following the estradiol peak (Lipson and Ellison, 1996). Ovulation day varied among participants, falling between the 11th and 17th days of the cycle (counting backwards from the first day of next menstrual bleeding, Fig. 1). Cycle length varied between 23 and 38 days (average = 28.5 days, SD = 3.52, Table 1). Depending on the day of the cycle on which a participant evaluated the photographs, they were divided into high- and low-conception probability groups according to 3 different grouping methods used in the literature. Difference in the masculinity preferences of women in high- and lowconception probability groups Analysis 1 Following Macrae et al. (2002) we divided women into highconception (2 days before and until the day of ovulation, N = 15) and low-conception probability groups (all other days, N = 100). Groups did not differ significantly in their average preference for masculinity either in the short-term (t = −0.43, p = 0.66) or long-term context (t = 0.24, p = 0.81, Table 2). Analysis 2 Following Harris (2011), the high-conception probability phase was considered to be 6 days prior and 1 day after the ovulation (N = 43), and the low-conception probability phase included all other days (N = 72). Calculations in the current work are a translation of the previously used forward counting method (Harris, 2011), so that range of menstrual cycle days is centered around hormonal measures of when ovulation occurred. There were no statistically significant differences in masculinity preferences between low- and high-conception
Analysis 2 When the fertile phase was considered to be between 6 days prior to and 1 day after ovulation, there was no significant difference in preference between short- and long-term mating contexts in either the lowor high-conception probability groups (t = 1.23, p = 0.22 and t = 1.35, p = 0.18, respectively, Table 3). Analysis 3 When the high-conception probability phase was considered to be between 8 days prior to and the day of ovulation, there was no significant difference in preference between short- and long-term mating contexts for low- or high-conception probability groups (t = 1.75 p = 0.08 and t = 0.68, p = 0.50, respectively, Table 3). Relation between estradiol and masculinity preference Morning estradiol levels on the day of the survey ranged between 4 and 78 pmol/L (SD = 14.35, Table 1). Based on daily measurements, average levels of estradiol per cycle were calculated for each woman. Average estradiol levels were between 5.4 and 59.5 pmol/L (SD = 9.33). Estradiol on the day of the survey and masculinity preference Only women who provided saliva sample for the day of completing the face rating task were included in this analysis (N = 83). There was no statistically significant relationship between the level of estradiol on the day of the survey and masculinity preference (long-term context: β = 0.106, SE = 0.110, p = 0.34 and short-term context: β = 0.031, SE = 0.11, p = 0.78, Table 4). Average estradiol and masculinity preference Mean estradiol level throughout the menstrual cycle was not related to masculinity preferences in either mating context (long-term: β = 0.021, SE = 0.094, p = 0.82 and short-term: β = 0.002, SE = 0.092, p = 0.98, Table 4).
Table 1 Descriptive statistics for cycle length and hormonal profile.
Length of the cycle (days) Timing of the ovulation day Estradiol levels (pmol/L) in the day of completing the questionnaire Average estradiol levels (pmol/L) for the cycle
N
Average
Minimum
Maximum
SD
115 115 83 115
28.5 −13.44 20.5 18.51
23 −17 4 5.4
38 −11 78 59.5
3.52 1.53 14.35 9.36
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7
5
Table 2 Average masculinity preference in two different mating contexts and differences between high and low conception probability group (between participant design). Preference in short-term mating context High conception defined as:
From −2 to ovulation day (Analysis 1) From −6 to +1 days around ovulation (Analysis 2) From −8 to ovulation day (Analysis 3)
High conception probability phase
Low conception probability
Mean SD
Mean SD
N
57.3
6.59 15 58.5
58.1
10.03 43 58.4
57.7
9.59 42 58.7
Preference in long-term mating context t
p
Effect size
CI 95%
N
High conception probability phase
Low conception probability
Mean SD
Mean SD
t
p
Effect size
CI 95%
10.07 100 0.436 0.664 −0.12 [−0.67, 0.42] 56.3
10.44 56.9
9.51
72 0.165 0.869 −0.03 [−0.41, 0.35] 56.2
9.65 57.3
9.56 0.548 0.585 −0.11 [−0.49, 0.26]
9.75
73 0.165 0.869 −0.10 [−0.48, 0.28] 56.8
9.81 56.9
9.49 0.044 0.965 −0.01 [−0.39, 0.37]
Discussion Regardless of mating context and conception probability, participants judged masculinized faces to be slightly more attractive. Male facial masculinity preferences, however, did not depend on woman's potential fertility throughout the menstrual cycle. When conception probability was assessed based on daily measurements of estradiol levels, women with high- and low-conception probabilities did not differ in their preferences, regardless of the assumed length of the relationship (i.e., within-subjects short- or long-term mating context trials). A number of previous studies also did not find cyclical shifts in women's preferences (Harris, 2013, Peters et al., 2009, Scott et al., 2013, Welling et al., 2008). Furthermore, masculinity preferences were not related to average levels of estradiol in the cycle or estradiol levels on the morning of the survey. Neither age nor relationship status of the participant influenced her preference for masculinity. In this study, we measured daily estradiol levels in women's saliva throughout the entire menstrual cycle. Day of ovulation was estimated based on hormonal measurements and varied among participants (between the 11th and 17th days, counting backwards from the end of the cycle). Our results provide an important critique of the studies that do not take into consideration inter-individual variation in cycle length and ovulation timing. We have documented that interindividual variation in timing of ovulation is high, even among women with regular menstrual cycles. Additionally, self-reported lengths of cycle are prone to measurement error (e.g., 43% of women misreported their cycle length by 2 or more days, Small et al., 2007), and women often report incorrect days of their previous menses (Wegienka and Baird, 2005). In our study, we measured estradiol levels; therefore, it is not possible that the null effect of conception probability on preferences is due to a miscalculation of the ovulation day. Moreover, we have analyzed the preferences of women in the same manner as some previous authors (Analyses 1 to 3 in the Results), and failed to find significant differences between groups with high and low probabilities of conception.
10.45 0.24
0.812 −0.06 [−0.6, 0.49]
There are several possible reasons why preference shifts throughout the menstrual cycle were not observed. As subtle facial trait differences in men may not have been relevant to the reproductive success of women in the past, it is unlikely that selection would promote preference shifts in women every few weeks or days. It is also possible that a woman's preference for a stable, investing partner overrides any cyclical shift of preference. Men with highly exaggerated secondary sexual features may be perceived as unlikely to invest in potential offspring; masculinity and testosterone are positively correlated with enhanced sexual activity (Halpern et al., 1993) and so interest in extra-pair liaisons may reduce interest in monogamous relationships (McIntyre et al., 2009). Moreover, even examining only the short-term preferences of women in relationships, it may still be possible that extra-pair copulations with men with pronounced masculinity are too risky to outweigh the potential cost of losing the investment of stable partner. In such scenario, masculinity preference shifts throughout the menstrual cycle would not be adaptive. Although, it is probable that women's sexual desires might still be triggered during fertile phase, they would not necessarily act on them. Participants in our study showed slight preferences for masculinized men; however, the preference was not especially strong in either short- or long-term mating contexts (varying from 54.4 to 58.7 on a scale in which 50% is no preference and 100% is a very strong preference for masculinized faces). Our results regarding inter-individual differences support the findings of Roney et al. (2011), who also grouped women into high- and low-conception probability groups based on estradiol levels; these authors also failed to find a difference in masculinity preferences between groups. They did find, however, a difference in preferences for androgen-dependent facial cues. It is possible that facial masculinity does not perfectly correspond to high circulating levels of testosterone (Peters et al., 2009) and hence may trigger different preference responses. Furthermore, it is possible that previously found correlation between men's genetic quality and masculinity (Penton-Voak and Perrett, 2000) is not as straightforward as assumed hitherto (immunocompetence handicap hypothesis, Folstad and Karter, 1992). There has been only one study reporting small but significant correlation between
Table 3 Average masculinity preferences in different mating contexts for two fertility groups (within participant analysis). High conception phase defined as:
Between −2 and ovulation day (Analysis 1) Between −6 and +1 days around ovulation (Analysis 2) Between −8 and ovulation day (Analysis 3)
Menstrual phase of the participant
Low-conception probability phase (N = 100) High-conception probability phase (N = 15) Low-conception probability phase (N = 72) High-conception probability phase (N = 43) Low-conception probability phase (N = 73) High-conception probability phase (N = 42)
Preference in long-term mating context
Preference in short-term mating context
Mean
SD
Mean
SD
56.96 56.32 57.3 56.2 56.9 56.8
9.48 10.45 9.56 9.65 9.49 9.81
58.49 57.31 58.4 58.1 58.7 57.7
10.07 6.59 9.51 10.03 9.76 9.59
t
p
Effect size
CI 95%
1.73 0.54 1.23 1.35 1.75 0.68
0.086 0.594 0.221 0.18 0.08 0.50
−0.16 −0.11 −0.12 −0.19 −0.19 −0.09
[−0.43, 0.12] [−0.83, 0.61] [−0.44, 0.21] [−0.61, 0.23] [−0.51, 0.14] [−0.52, 0.34]
6
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7
Table 4 Correlation between masculinity preference and mean estradiol levels throughout the cycle and level of estradiol from the day of taking the survey, in two mating contexts. N
Mean estradiol levels during the cycle Morning estradiol levels on the day of completing the survey
115 83
masculinity and good health (Rhodes et al., 2003). It is possible that the lack of shift in masculinity preferences found in our study, is not univocal with lack of fluctuation preference towards purportedly good genes (following Harris, 2011). It could be that within woman measures of shifts across the cycle are likely to be much more sensitive, and the result of our study can be treat as a between subjects effect. The null results reported in our study might also have been caused by the fact that the manipulated photos did not simulate variation in facial masculinity in a manner that was sufficiently pronounced to be perceived by women raters. The usage of 2D, immobile composite faces rather than real faces may have contributed to the discrepancy between our and previously published results (Ellison, 2008). Furthermore, given that within woman measures of shifts across the cycle are likely to be much more sensitive to small changes than between women measurement, it might be possible that our null results are a result of inter-individual, cross-sectional design (unlike the studies that found significant positive relation between hormones and masculinity preference, e.g. Roney and Simmons, 2008). We would want however to underline, that due to detailed hormonal measurements we were able to, very precisely, segregate our participants into high and low conception probability phases, what should increase the probability of finding the putative effect. In our study, ovulation was assessed based on estradiol levels, and ovulation was not confirmed by an LH surge. We followed published methods of identifying the ovulation by the largest mid-cycle drop of estradiol levels, which occurs in response to hormonal changes related to ovulation (Lipson and Ellison, 1996). Use of this methodology may result in misclassifying an anovulatory cycle as an ovulatory cycle. However, we excluded women without a reliably identified mid-cycle estradiol drop from our analyses. In the majority of studies on fluctuating preferences, hormonal measurements are not included; therefore obtaining daily estradiol levels is one of the most detailed methodologies to date. It is also possible that collecting more in-depth sociodemographic data from participants would allow us to detect more subtle preference changes (and covariance of such changes with, for example, sexual openness or number of children). Results of this study strongly suggest the necessity of incorporating physiological measures of steroid sex hormones in future studies of preference fluctuations. Caution should be exercised when evaluating existing studies that report robust cyclical shifts in preference for facial masculinity in men based solely on counting days methods (especially forward counting one).
Funding statement This study was supported by grants from the National Science Centre (grant no. 2014/12/S/NZ8/00722 to UMM and grant no. N N404 273440 to GJ) and the Ministry of Science and Higher Education (grant no. IdP2011 000161 to GJ).
Acknowledgments We would like to thank all participants, Dr. Susan Lipson, Dr. Anna Ziomkiewicz and our study assistants. We also thank Prof. David Perrett and Dr. Ben Jones for providing the visual stimuli.
Long-term mating context
Short-term mating context
β
SE
p
β
SE
p
0.021 0.106
0.094 0.110
0.82 0.34
−0.002 0.031
0.092 0.111
0.98 0.78
References Bobst, C., Sauter, S., Foppa, A., Lobmaier, J.S., 2014. Early follicular testosterone level predicts preference for masculinity in male faces — but not for women taking hormonal contraception. Psychoneuroendocrino. 41, 142–150. Chiazze, L., Brayer, F.T., Macisco, J.J., Parker, M.P., Duffy, B.J., 1968. Length and variability of human menstrual cycle. J. Amer. Assoc. 203 (6), 377–380. Danel, D., Pawlowski, B., 2006. Attractiveness of men's faces in relation to women's phase of menstrual cycle. Collegium Antropol. 30 (2), 285–289. DeBruine, L., Jones, B.C., Frederick, D.A., Haselton, M.G., PentonVoak, I.S., Perrett, D.I., 2010. Evidence for menstrual cycle shifts in women's preferences for masculinity: A response to Harris (in press) “Menstrual cycle and facial preferences reconsidered”. Evol. Psych. 8, 768–775. Durante, K.M., Li, N.P., 2009. Oestradiol level and opportunistic mating in women. Biol. Lett. 5, 179–182. Ellison, P.T., 2008. Beauty: in the gonads of the beholder — and the beheld — commentary. Horm. Behav. 53, 11–13. Folstad, I., Karter, A.J., 1992. Parasites, bright males, and the immunocompetence handicap. Am. Nat. 139, 603–622. Garver-Apgar, C.E., Gangestad, S.W., Thornhill, R., 2008. Hormonal correlates of women's mid-cycle preferences for the scent of symmetry. Evol. Hum. Behav. 29, 223–232. Gildersleeve, K., Haselton, M.G., Fales, M., 2014a. Do women's mate preferences change across the ovulatory cycle? A meta-analytic review. Psychol. Bull. http://dx.doi.org/ 10.1037/a0035438 Gildersleeve, K., Haselton, M.G., Fales, M., 2014b. Meta-analyses and P-curves support robust cycle shifts in women's mate preferences: reply to Wood and Carden (2014) and Harris, Pashler, and Mickes (2014). Psychol. Bull. 140, 1272–1280. Harris, C.R., 2011. Menstrual cycle and facial preferences reconsidered. Sex Roles 64, 669–681. Harris, C.R., 2013. Shifts in masculinity preferences across the menstrual cycle: still not there. Sex Roles 69, 507–515. Halpern, C.T., Udry, J.R., Suchindran, C., 1993. Testosterone and pubertal development as predictor of sexual-activity: a panel analysis of adolescent males. Psychosom. Med. 27, 445–465. Hawkinsa, S.M., Matzuk, M.M., 2008. The menstrual cycle — basic biology. Ann. N. Y. Acad. Sci. 1135, 10–18. Jasienska, G., Ziomkiewicz, A., Ellison, P.T., Lipson, S.F., Thune, I., 2004. Large breasts and narrow waists indicate high reproductive potential in women. Proc. R. Soc. London B. 271, 1213–1217. Jasienska, G., Ziomkiewicz, A., Thune, I., Lipson, S., Ellison, P., 2006. Habitual physical activity and estradiol levels in women of reproductive age. Eur. J. Cancer Prev. 15, 439–445. Jasienska, G., Jasienski, M., 2008. Interpopulation, interindividual, intercycle, and intracycle natural variation in progesterone levels: a quantitative assessment and implications for population studies. Am. J. Hum. Biol. 20, 35–42. Jasienska, G., 2013. The Fragile Wisdom. An Evolutionary View on Women's Biology and Health. Harvard University Press, Cambridge, MA. Lipson, S.F., Ellison, P.T., 1996. Comparison of salivary steroid profiles in naturally occurring conception and non-conception cycles. Hum. Reprod. 11 (10), 2090–2096. Little, A.C., Jones, B.C., 2012. Variation in facial masculinity and symmetry preferences across the menstrual cycle is moderated by relationship context. Psychoneuroendocrino. 37, 999–1008. Little, A.C., Jones, B.C., Burriss, R.P., 2007. Preferences for masculinity in male bodies change across the menstrual cycle. Horm. Behav. 31, 633–639. Little, A.C., Jones, B.C., DeBruine, L.M., 2008. Preferences for variation in masculinity in real male faces change across the menstrual cycle: women prefer more masculine faces when they are more fertile. Pers. Indiv. Differ. 45, 478–482. Little, A.C., 2013. The influence of steroid sex hormones on the cognitive and emotional processing of visual stimuli in humans. Front. Neuroindocrin. 34, 315–328. Macrae, C.N., Alnwick, K.A., Milne, A.B., Schloerscheidt, A.M., 2002. Person perception across the menstrual cycle: hormonal influences on social-cognitive functioning. Psychol. Sci. 13, 532–536. Marcinkowska, U.M., Kozlov, M., Cai, H., Contreras-Garduño, J., Dixson, B.J., Oana, G., Kaminski, G., Li, N.P., Lyons, M.T., Onyishi, I.E., Prasai, K., Pazhoohi, F., Prokop, P., Rosales Cardozo, S.L., Sydney, N., Yong, J.C., Rantala, M.J., 2014. Cross-cultural variation in men's preference for sexual dimorphism in women's faces. Biol. Lett. 10, 20130850. McIntyre, M., Gangestad, S.W., Gray, P.B., Chapman, J.F., Burnham, T.C., O'ROurke, M.T., Morrison, E.R., Clark, A.P., Gralewski, L., Campbell, N., Penton-Voak, I.S., 2009. Women's probability of conception is associated with their preferences for flirtatious but not masculine facial movement. Arch. Sex. Behav. 39, 1297–1304. Pawlowski, B., Jasienska, G., 2005. Women's preferences for sexual dimorphism in height depend on menstrual cycle phase and expected duration of relationship. Biol. Psychol. 70, 38–43.
U.M. Marcinkowska et al. / Hormones and Behavior 78 (2016) 1–7 Penton-Voak, I.S., Perrett, D.I., 2000. Female preferences for male faces changes cyclically: Further evidence. Evol. Hum. Behav. 21, 39–48. Penton-Voak, I.S., Perrett, D.I., Castles, D.I., Kobayashi, T., Burt, D.M., Murray, L.K., Minamisawa, R., 1999. Menstrual cycle alters face preference. Nature 399, 741–742. Peters, M., Simmons, L.W., Rhodes, G., 2009. Preferences across the menstrual cycle for masculinity and symmetry in photographs of male faces and bodies. PLoS One 4, e4138. Perrett, D.I., Lee, K.J., Penton-Voak, I.S., Rowland, D.R., Yoshikawa, S., Burt, D.M., Akamatsu, S., 1998. Effects of sexual dimorphism on facial attractiveness. Nature 394, 884–887. Pisanski, K., Hahn, A.C., Fisher, C.I., DeBruine, L.M., Feinberg, D.R., Jones, B.C., 2014. Changes in salivary estradiol predict changes in women's preferences for vocal masculinity. Horm. Behav. 66 (3), 493–497. Puts, D.A., 2005. Mating context and menstrual phase affect women's preferences for male voice pitch. Evol. Hum. Behav. 26, 388–397. Rhodes, G., Chan, J., Zebrowitz, L.A., Simmons, L.W., 2003. Does sexual dimorphism in human faces signal health? Proc. R. Soc. London B. 270 (Suppl. 1), S93–S95. Roney, J.R., Simmons, Z.L., 2008. Women's estradiol predicts preference for facial cues of men's testosterone. Horm. Behav. 53, 14–19. Roney, J.R., Simmons, Z.L., Gray, P.B., 2011. Changes in estradiol predict within-women shifts in attraction to facial cues of men's testosterone. Psychoneuroendocrino. 36 (5), 742–749. Rosen, M.K., Lopes, H.H., 2009. Menstrual cycle shifts in attentional bias for courtship language. Evol. Hum. Behanv. 30, 131–140. Scott, I.M.L., Clarke, A.O., Boothroyd, L.G., Penton-Voak, I.S., 2013. Do men's faces really signal heritable immunocompetence? Behav. Ecol. 24, 579–589.
7
Small, C.M., Manatunga, A.K., Marcus, M., 2007. Validity of self-reported menstrual cycle length. Ann. Epidemiol. 17, 163–170. Thornhill, R., Chapman, J.F., Gangestad, S.W., 2013. Women's preferences for men's scents associated with testosterone and cortisol levels: Patterns across the ovulatory cycle. Evol. Hum. Behav. 34, 216–221. Wegienka, G., Baird, D.D., 2005. A comparison of recalled date of last menstrual period with prospectively recorded dates. J. Women's Health 14, 248–252. Welling, L.L.M., Jones, B.C., DeBruine, L.M., Conway, C.A., Law Smith, M.J., Little, A.C., Feinberg, D.R., Sharp, M., Al Dujaili, E.A.S., 2007. Raised salivary testosterone in women is associated with increased attraction to masculine faces. Horm. Behav. 52, 156–161. Welling, L.L., Jones, B.C., DeBruine, L.M., Smith, F.G., Feinberg, D.R., Little, A.C., Al Dujaili, E.A.S., 2008. Men report stronger attraction to femininity in women's faces when their testosterone levels are high. Horm. Behav. 54, 703–708. Wilcox, A., Dunson, D., Baird, D.D., 2000. The timing of the fertile window in the menstrual cycle: day specific estimates from a prospective study. Brit. Med. J. 321, 1259–1262. Wood, W., 2014. Author reply: once again, menstrual cycles and mate preferences. Emotion 6, 258–260. Wood, E., Carden, L., 2014. Elusiveness of menstrual cycle effects on mate preferences: Comment on Gildersleeve, Halton, and Fales (2014). Psych. Bull. 140, 1265–1271. Wood, W., Kressel, L., Joshi, P.D., Louie, B., 2014. Meta-analysis of menstrual cycle effects on women's mate preferences. Emotion 0, 1–21.