Sexual dimorphism measures were also taken from points marked on facial features (Figure 1). While pictures were initially screened for head tilt there was still the potential for outliers in facial asymmetry. For the Hadza images, male (67 individuals) and female (69 individuals) participants had their photograph taken with a digital camera under variable outside lighting conditions. For the macaque images, a digital video camera was used to capture images of adult males (105 individuals) and females (111 individuals) from the free-ranging population of rhesus macaques on Cayo Santiago, Puerto Rico. Direct effects of sex hormones on bone growth might be responsible, either by regulation of Hox genes in digit development or independently of such genes. The mechanisms by which prenatal hormones influence digit ratio are not clearly understood. In other words, women with a greater anogenital distance, indicating greater prenatal androgen exposure, had a smaller digit ratio. In a non-clinical sample of women, digit ratio correlated with anogenital distance in the expected direction. Men with less sensitive androgen receptors may compensate for this by secreting more testosterone via reduced inhibitory feedback on gonadotropins. Researchers have raised concerns that, although the general trend points towards a correlation between digit ratio and early androgen exposure, many results have not been statistically significant. Despite some exceptions, male tallness is preferred across human societies while there is no such simple preference for female height54,55,56,57. Himba nomads from northern Namibia showed preference for equal height instead of "taller male–shorter female" stereotype43. Although there is some research showing environment-related variation in facial femininity preferences37, more feminine female faces are perceived as more attractive unanimously across populations12,30,38,39. All the experiment protocol for involving humans was in accordance to guidelines of national/international/institutional or Declaration of Helsinki. The difference between male and female intercepts and male and female slopes were included in the regression model, as fixed effects, to test the difference between males and females. Females were used as the reference category, and we report the slope of the regression of attractiveness on SShD. These SNPs were shown to impact specific regions of the face that show strong evidence of sex differences – the mandible, lips, and upper face. We now understand that the level of free testosterone in individuals is at least partly under genetic control (Ohlsson et al., 2011; Zhai et al., 2011; Coviello et al., 2012; Jin et al., 2012; Prescott et al., 2012; Chen et al., 2013; Ruth et al., 2016). Furthermore, several animal studies suggest an influence of NR2F2 on craniofacial morphology. The nearby NR2F2 gene has been implicated in testicular Leydig function, essential for male testosterone production (Hu et al., 2013). (D) Surface warp showing the effect (exaggerated) of the minor allele (A) on lower facial shape.
