It has been shown that women's preference for masculinity in male faces changes across the menstrual cycle. Preference for masculinity is stronger when conception probability is high than when it is low. These findings have been linked to cyclic fluctuations of hormone levels. The purpose of the present study is to further investigate the link between gonadal steroids (i.e. testosterone, estradiol, and progesterone) and masculinity preference in women, while holding the cycle phase constant. Sixty-two female participants were tested in their early follicular cycle phase, when conception probability is low. Participants were shown face pairs and where asked to choose the more attractive face. Face pairs consisted of a masculinized and feminized version of the same face. For naturally cycling women we found a positive relationship between saliva testosterone levels and masculinity preference, but there was no link between any hormones and masculinity preference for women taking hormonal contraception. We conclude that in naturally cycling women early follicular testosterone levels are associated with masculinity preference. However, these hormonal links were not found for women with artificially modified hormonal levels, that is, for women taking hormonal contraception.
Facial attractiveness has been positively related to mating success (Jokela, 2009). Recent research has identified various facial cues that are responsible for attractiveness: symmetry, averageness, skin healthiness, facial adiposity and sexual dimorphism (e.g. Coetzee et al., 2009, Little et al., 2011 and Koscinski, 2012). In terms of sexual dimorphism, there is rather clear agreement that femininity is positively related to female attractiveness (e.g. Perrett et al., 1998, Rhodes, 2006 and Koscinski, 2013). However, there is no such agreement whether masculine facial traits make a male face more attractive (e.g. Rhodes, 2006, DeBruine et al., 2010, Koscinski, 2012 and Koscinski, 2013). A reason for inconsistent findings could be that women's preferences for male stimuli change across the menstrual cycle: in the late follicular phase women prefer more masculine men than in other phases. This shift in preferences has been shown for men's body odour (Grammer, 1993), body shape (Little et al., 2007) and voice (Puts, 2005, Feinberg et al., 2006 and Puts, 2006). Likewise women's preferences for male faces change. Penton-Voak et al. (1999) found that, in general, females preferred male faces with a slightly more feminine shape – however this preference was weaker if the woman was tested near ovulation. This implies that women's face preferences are modulated by hormonal changes which occur across the menstrual cycle in accordance with conception probability. Other studies found an overall preference for more masculine male faces (Johnston et al., 2001) or no overall preference (Penton-Voak and Perrett, 2000); however they all found the same cycle effect: women showed a stronger preference for more masculine face shapes if they were tested near ovulation than when tested in other cycle phases (Penton-Voak and Perrett, 2000 and Johnston et al., 2001). Females’ preference for faces of men with higher testosterone levels was also highest around ovulation (Roney and Simmons, 2008). Since facial masculinity has been linked to testosterone levels in men (Penton-Voak and Chen, 2004 and Roney et al., 2006), this result is in line with the studies reviewed above. However, the link between facial masculinity and testosterone level is debatable: some studies did not find an association between facial masculinity and current testosterone level (Peters et al., 2008) or only an association with testosterone increase after winning a competitive task but not with the baseline testosterone level (Pound et al., 2009). Furthermore, also the shift of women's preferences across the cycle is not undisputed (e.g. Harris, 2012).
