The relationship of facial width-to-height ratio and aggressiveness is strongest for males at 27–33 and females at 34–61

Tracking sexual dimorphism of facial width-to-height ratio across the lifespan: implications for perceived aggressiveness. Stephanie Summersby, Bonnie Harris, Thomas F. Denson and David White. Royal Society Open Science, Vol 9 Iss 5, May 4 2022. https://doi.org/10.1098/rsos.211500

Abstract: The facial width-to-height ratio (FWHR) influences social judgements like perceived aggression. This may be because FWHR is a sexually dimorphic feature, with males having higher FWHR than females. However, evidence for sexual dimorphism is mixed, little is known about how it varies with age, and the relationship between sexual dimorphism and perceived aggressiveness is unclear. We addressed these gaps by measuring FWHR of 17 607 passport images of male and female faces across the lifespan. We found larger FWHR in males only in young adulthood, aligning with the stage most commonly associated with mate selection and intrasexual competition. However, the direction of dimorphism was reversed after 48 years of age, with females recording larger FWHRs than males. We then examined how natural variation in FWHR affected perceived aggressiveness. The relationship between FWHR and perceived aggressiveness was strongest for males at 27–33 and females at 34–61. Raters were most sensitive to differences in FWHR for young adult male faces, pointing to enhanced sensitivity to FWHR as a cue to aggressiveness. This may reflect a common mechanism for evaluating male aggressiveness from variability in structural (FWHR) and malleable (emotional expression) aspects of the face.

4. General discussion

The present study was a large-scale investigation into sexual dimorphism in the FWHR across the lifespan. We found that sexual dimorphism was present in our sample of over 17 000 Australian citizens. In younger to middle adulthood, the FWHR in men was greater than the FWHR in women. From the age of 48 onwards, this pattern was reversed such than women had larger FWHRs than men. One qualification is that the differences between FWHRs in men and women were very small across the age groups. These small effect sizes are consistent with that reported for sexual dimorphism in a meta-analysis (d = 0.11; [5]).

The predicted sexual dimorphism was therefore only observed in a surprisingly narrow age band in early adulthood. Yet this pattern in young adults is consistent with the view that FWHR is an evolutionarily important cue to physical formidability, as sexual dimorphism in this age band aligns with the period of life most commonly associated with mate selection and intrasexual competition. The reversal of this dimorphism in middle and late-adulthood, with progressively larger female compared to male FWHR, is more difficult to explain. It is possible that there are broader physical changes in ageing that explain the pattern. For example, because body mass index (BMI) is moderately correlated with the FWHR (r = 0.31; [5]), one possibility is that age-related BMI changes are different for males and females. Other possibilities are that the reversal in dimorphism is connected to age-related structural changes to the faces, such as differences in the rate of face lengthening with age [42]. Other possibilities are that increasingly fewer males with higher FWHR apply for passports later in life—perhaps because many men with the largest FWHR may be removed from society via incarceration [43] or early mortality relative to women—or that the difference is affected by changes in head pose behaviour in males and females of different ages [44].

We also tested the relationships between the FWHR and perceived aggressiveness in men and women across the lifespan. In their meta-analysis, Geniole et al. [5] found that the relationship between the FWHR and perceived aggressiveness was stronger for younger faces than older faces, but there was no evidence of moderation by sex. By contrast, we found that the relationships between the FWHR and perceived aggressiveness for males was strongest for the youngest age group of faces (27–33 years old), but from 34–61 years old, this relationship was strongest for female faces. These results suggest that the effect of FWHR on perceived aggressiveness ratings varies as a function of age and sex.

Moreover, the effect of FWHR on perceived aggressiveness was somewhat independent from physical variation in FWHR in these age groups. Aggressiveness ratings to faces in the 34 to 40 age range show greater modulation for female faces, despite there being more physical FWHR variation in male faces (see the electronic supplementary material, figure S4). This shows increased sensitivity to FWHR in the youngest male faces when people evaluate perceived aggressiveness, albeit restricted to a relatively narrow age band. This is consistent with results showing that people are more sensitive to threatening emotional expressions in male compared with female faces [45,46] and may point to a common mechanism responsible for processing FWHR-related and expression-related cues to threat. In face perception research more broadly, it has been proposed that our social impressions of structural aspects of faces are shaped by social learning of facial expressions [47,48], for example, that trustworthiness judgements from structural properties of faces are linked to transient changes such as smiling or warm expressions (e.g. [49]). Future work examining whether other threat cues are also modulated by face age can potentially help to resolve whether similar social learning mechanisms are involved in perceived aggressiveness.3

Another potential explanation of these findings is that the apparent increase in sensitivity to FWHR cues in young males was owing to participants being mostly undergraduate students. For face identity processing at least, there is consistent evidence that people develop perceptual expertise specifically for faces fitting the viewer's demographic profile, including faces of the same age as the viewer [51,52]. This raises the possibility that the apparent perceptual sensitivity to FWHR in young faces that we observe may be specific to the younger participants in our study. However, we note that this ‘own age effect’ is reported mostly in identity memory-based recognition tasks and is not consistently found for other types of identity processing task formats [53], or for other types of face judgements [54].4 Moreover, participant's age is not known to affect perceptions of aggression [56]. Nevertheless, this is an intriguing question that could be addressed in future work.

The enhanced effect of FWHR on aggressiveness for young men is consistent with evolutionary perspectives on the FWHR as a cue to physical formidability. However, the relationship between the FWHR and aggressiveness for women in middle and late-adulthood is more difficult to explain. Physical variation in FWHR in these age groups was greater for male faces (electronic supplementary material, figure S4), but the effect this variation had on aggressiveness ratings was higher for female faces (figure 4). The differences in ratings of aggression for younger and older men and women may be related to age-specific facial adiposity (the perception of weight in the face). Higher facial adiposity had been associated with higher perceptions of male facial masculinity [57], and lower perceptions of female facial femininity [58]. Thus, one possibility is that age-related changes in facial adiposity are different for males and females, and could be contributing to the sex differences in FWHR and perceived aggressiveness.

Another plausible explanation is that differences in sensitivity to FWHR cues were mediated by widely held stereotypes of masculinity and femininity in younger and older men and women. Men with relatively larger FWHRs are considered masculine, unattractive and physically formidable [5]. As in previous research, we found that these men were also considered likely to become aggressive if provoked [5]. This perception was largest among young men and faded with time. This finding is consistent with the stereotype of older men becoming weaker and less formidable with age, while younger men with relatively large FWHRs were probably viewed as ‘fighting fit’. By contrast, younger women are stereotyped as more feminine, attractive and passive than older women. Thus, participants' judgements of aggressiveness may have been relatively unaffected by the FWHR of the young women. However, stereotypes of older women can be particularly harmful, as they lead to appearance-based discrimination [59]. In this case, the larger FWHRs may have elicited a global negative evaluation bias because the women were not stereotypically attractive, feminine women.

These proposals should be treated as speculative for now. Indeed, there should be some caution when interpreting associations between FWHR and social impressions more generally. The use of discrete anthropometric ratios can sometimes be misleading owing to their associations with other surrounding measures in the body and face (e.g. [60]). This can lead to effects emerging as a consequence of interactions with other interrelated traits (e.g. attractiveness, facial masculinity, facial maturity and anger resemblance, [56]) rather than a consequence of the ratio itself. Therefore, it is important for future research to explore potential traits that may be interacting with the FWHR to impact judgements such as perceived aggressiveness in younger and older low- and high-FWHR faces.

Notwithstanding, to our knowledge our results provide the most comprehensive analysis of FWHR across the lifespan to date. We show that the sexual dimorphism of this trait is consistent with a secondary sexual characteristic that signals formidability in young males. We also show that perceivers were particularly sensitive to FWHR variation in young faces when evaluating perceived aggressiveness. Understanding the causes of face age dependency on perceptions of aggressiveness would be a worthwhile focus of future work.