The interaction reflected that faces that were misclassified according to facial measures demonstrated greater asymmetry than faces that were classified as sex typical (see Figure 2). A univariate ANOVA was conducted with asymmetry as the dependent variable, and face-type (European/Hadza/Macaque), sex (male/female), and classification (male/female) as factors. The features of the individual faces were then morphed to the relevant average shape before superimposing the images to produce a photographic quality result. Colour differences between the images are irrelevant for measurements as they involve only shape information. Any images with asymmetry scores higher than 120 were then excluded from the analysis for all sets. However, if only high quality individuals are capable of bearing the handicap of growing large traits or symmetric traits we would expect size and symmetry of traits to correlate. Higher negative scores indicate more female-like morphology, whereas higher positive scores indicate a more male-like facial shape. The magnitude of sexual dimorphism may be affected by overall morphological variation of populations. There are two contrasting views on cross-cultural variation in preferences of human sexual dimorphism. Since Darwin, numerous proposals have been suggested to explain sexual dimorphism, the most popular of which is sexual selection. Again, this was true only for differences between sex averages but not for the overall range of sexual dimorphism. Although we have reported an association between sex differences in body height and distance between sex-specific facial averages (see Fig. 6A), this does not explain the variation either of total SShD or its allometric component. Furthermore, the extent to which sex differences in facial shape can be related to body size, and the contribution of shape variation due to size, seem to vary across populations (Fig. 1). Face shape may not be the only characteristic that provides information about the level of sexual dimorphism in African populations. The differences between male and female facial shapes based on faces from all eight populations are shown in Fig. PC1 captures mostly the variation between national samples with European samples on the right and African samples on the left, while PC2 allows faces to align along sexual dimorphism axis. Populations differed in the degree of sexual shape dimorphism measurable from the face (SShD). The non-allometric component of SShD was acquired by regressing the original shape coordinates on height and then projecting the residualized facial coordinates on the sex difference vector calculated on these residuals. The allometric variation in SShD was calculated by regressing the original facial coordinates on height and projecting the estimated values from this regression on the vector of sex differences. The relationship between overall, allometric, and non-allometric sexual shape dimorphism (SShD) and rated facial attractiveness. The relationship between attractiveness ratings and facial sexual dimorphism was examined with linear mixed effect models. The sex difference in body height predicts the distance between male and female facial averages. The upper panel shows overall shape sexual dimorphism, while the two lower panels depict allometric and non-allometric components. Maintaining body mass is costlier in regions with uncertain food availability. Taller and heavier men are not only perceived as more masculine49 but taller men also possess more masculinized facial structure20. A conjoint phenomenon might be the existence of lower variability of facial morphologies in less populated rural societies compared to higher facial variation in large-scale urban societies. The null preferences for masculine male faces in small-scale traditional societies were explained by encountering fewer different faces and by having fewer social interactions with potential mates over a lifespan19. For instance, Jamaican women exhibited a greater preference for masculine over feminine male faces compared to British women41. Amniocentesis samples in males with CAH show that prenatal levels of testosterone are in the high-normal range and levels of the weaker androgen androstenedione are several fold higher than in controls, indicating that males with CAH are exposed to greater prenatal concentrations of total androgens. However, when examining the extent of sexual dimorphism evident after birth, adults were found to demonstrate more dimorphism than children, who show the same extent of dimorphism as fetuses. Additionally, our data does not necessarily support the idea that sexual dimorphism represents a single continuum in faces. Taken together these findings suggest that feminine traits may be less powerful signals of good-genes than masculine traits, although we note there that here femininity in female faces is correlated with symmetry, another proposed aspect of quality. The link between sexual dimorphism and good-genes advertisement has produced many more theories. Symmetry and sexual dimorphism may be linked by an underlying biological factor. Rating sessions took place in each of the investigated populations, and raters judged only opposite-sex faces from their own population. The photographs were subsequently post-produced to adjust the eyes horizontally at the same height. Likewise, women of the Tanzanian Hadza tribe showed no preference for large body size in potential mates52 and they were more likely to marry men shorter than themselves compared to British women53. Greater mass may also be disadvantageous for hunting as it may make one more visible to prey; as indicated by the negative correlation between body size and food returns in African hunter-gatherers51. High levels of testosterone are primarily responsible for development of masculine facial morphologies, while development of female secondary sexual traits is attributed to a high estrogen-to-testosterone ratio10,11,12,13,14. If this causal model is correct, then any factor affecting testosterone levels should also impact facial morphology along with a host of other sexually dimorphic traits. This accelerated dimorphism post-puberty is the result of changes in circulating hormone levels, which regulate the development and differentiation of male and female primary and secondary sex characteristics, such as voice, body shape and facial morphology (Hines, 2011).