Original Article
Telling facial metrics: facial width is associated with testosterone levels in men

https://doi.org/10.1016/j.evolhumbehav.2013.03.005Get rights and content

Abstract

High facial width-to-height ratio (fWHR) has been associated with a cluster of behavioural traits in men, including aggression and status-striving. This association between face structure and behaviour may be caused by testosterone. Here we investigated the relationship of both baseline and reactive testosterone levels to fWHR. In addition, we investigated the link between testosterone and three well-characterised sexually dimorphic facial metrics. Testosterone was measured in one sample of males (n = 185) before and after a speed-dating event. An additional sample provided only baseline testosterone measures (n = 92). fWHR was positively associated with testosterone reactions to potential mate exposure and marginally associated with baseline testosterone in Sample 1. We found a positive association with baseline testosterone and fWHR in Sample 2. In addition, face-width-to-lower-height ratio was positively associated with testosterone in both samples, suggesting that, in particular, facial width (scaled by two measures of facial height) is associated with testosterone. Importantly, our results also indicate that there is no association between adult testosterone and the sexual dimorphism of face shape. Thus, while our findings question the status of sexual dimorphism as a proxy measure of testosterone, they do provide evidence that testosterone is linked to fWHR and might underlie the relationship between fWHR and behaviour.

Introduction

Recently, the facial width-to-height ratio (fWHR: bizygomatic width divided by upper-face height; see Fig. 1) has been identified as a facial metric with links to a range of behavioural traits in men. fWHR has been shown to predict aggression (Carré & McCormick, 2008; but see Deaner et al., 2012, Özener, 2012), deception (Haselhuhn & Wong, 2012), and untrustworthiness (Stirrat & Perrett, 2010), but also more positive behaviours such as achievement striving (Lewis, Lefevre, & Bates, 2012) and self-sacrifice towards the in-group (Stirrat & Perrett, 2012). Moreover, several studies have demonstrated that individuals are able to discern other individuals’ tendency to aggression (Carré, McCormick, & Mondloch, 2009) and trustworthiness (Stirrat & Perrett, 2010) from fWHR. While this research indicates that the face and, in particular, fWHR act as a cue to behaviour, it is currently unclear what constitutes the physiological mechanisms underlying this anatomy–behaviour association. Here, in two samples, we examine whether levels of baseline and reactive testosterone are associated with fWHR. We also investigate the association of testosterone with three well-characterised sexually dimorphic facial metrics; namely, 1) lower-face to whole-face-height, 2) cheekbone prominence, and 3) face-width to lower-face-height (see Fig. 1 and Lefevre et al., 2012, Penton-Voak et al., 2001), alongside a global morphometric measure of facial masculinity (Pound, Penton-Voak, & Surridge, 2009).

Recent work reported sexual dimorphism in fWHR, with males showing higher fWHR than females (Carré and McCormick, 2008, Weston et al., 2007). While this claim of sexual dimorphism in fWHR has since been challenged (Lefevre et al., 2012, Özener, 2012, Stirrat et al., 2012), several of the behaviours linked to fWHR (e.g. aggression, dominance) are also known to be sexually dimorphic. For example, males show consistently higher levels of direct physical aggression than females (Archer, 2006). Testosterone levels, which themselves show large sex differences (e.g. Mazur, Susman, & Edelbrock, 1997), have been proposed as a common underlying factor linking fWHR to behaviour (e.g. Carré & McCormick, 2008).

Consistent with the suggestion that testosterone mediates the association between fWHR and behaviour, a number of behavioural characteristics similar to those associated with fWHR have been linked to both direct and indirect measures of testosterone. For example, basal circulating testosterone levels are associated with levels of dominance in men (e.g. Josephs et al., 2003, Mazur and Booth, 1998). Additionally, second-to-forth digit ratio (2D:4D) – a putative proxy measure of in-utero testosterone exposure (Manning et al., 1998, Manning et al., 2000, Williams et al., 2003) – has been linked to levels of both self-reported (Bailey & Hurd, 2005) and lab-induced aggression (Millet & Dewitte, 2007).

Work seeking to associate testosterone with facial shape has also provided some support for a potential link with fWHR. In an early study, Verdonck, Gaethofs, Carels, and de Zegher (1999) observed that testosterone administration enhanced craniofacial growth in delayed puberty male adolescents. In particular, these authors report enhanced growth of upper and total face height, and the mandible (jawbone) and ramus (upper part of the jawbone) length. However, these results should be treated with some caution as the sample size was small (N = 7) and the boys assessed were all showing delayed puberty, making their development likely not fully comparable to healthy controls.

In addition, several studies have linked testosterone levels to perceived facial masculinity in men. Roney, Hanson, Durante, and Maestripieri (2006) report a moderate correlation (r = .34) between ratings of masculinity in natural, unmodified faces and baseline testosterone levels. Similarly, using a forced-choice paradigm, Penton-Voak and Chen (2004) report a weak, but significant, association between testosterone levels and masculine appearance. In pairs of either natural or composite faces the face higher in testosterone was chosen as more masculine 53% and 57% of the time respectively. The authors argue that only men with very high or very low levels of testosterone may be visually distinguishable in terms of their masculinity.

However, other studies find no links between testosterone and masculinity. A study using almost identical methods to Roney et al. (2006), but with a much larger set of men, found no association between perceived facial masculinity and testosterone levels (Peters, Simmons, & Rhodes, 2008). Similarly, Neave, Laing, Fink, and Manning (2003) reported links of perceived facial masculinity with second-to-forth digit ratio (2D:4D), but not with measured baseline testosterone levels; and Ferdenzi, Lemaître, Leongómez, and Roberts (2011) found no association between perceived facial masculinity and 2D:4D ratio.

Fink et al. (2005) incorporated a formal measure of global facial shape, calculating the difference between the average face-shape of men with high and low 2D:4D ratio. This study found that men with a low digit ratio (associated with higher prenatal testosterone) exhibited a wider jaw and zygomatic arch (i.e. face width) as compared to those with a higher digit ratio. However, direct reports of an association between circulating testosterone levels and facial metrics are (to our knowledge) limited to a single study (Pound et al., 2009). In this study, the authors computed a global facial masculinity index using five empirically derived, sexually dimorphic facial metrics (see Penton-Voak et al., 2001), and measured both baseline and reactive levels of testosterone. Reactive testosterone measures were taken after participants won in a manipulated competitive betting task, a scenario previously shown to significantly elevate testosterone levels (e.g. Archer, 2006). While failing to replicate associations between testosterone measures and perceived facial masculinity, Pound et al. (2009) observed that reactive, but not baseline, testosterone was positively associated with a global facial masculinity measure. It is noteworthy that the association between baseline testosterone and facial masculinity also showed a positive trend towards statistical significance in this study, suggesting that the relationship between face shape and baseline testosterone might not be completely absent, but instead may reflect a weaker link than the association with reactive testosterone.

The above findings reported by Pound et al. (2009) are in line with the “challenge hypothesis” (Archer, 2006, Wingfield et al., 1990), which states that testosterone rises in challenging situations within the mating context, serving as a sexually selected physiological mechanism that calibrates the optimal effort put into intra-sexual competition. Moreover, only the increase in testosterone, rather than baseline testosterone per se, is suggested to associate with aggression and other status-related traits. Accordingly, reactive testosterone may serve as a better predictor of behaviour as compared to baseline levels of testosterone.

Determining whether an association exists between facial structure and testosterone is important in order to advance understanding of the underlying basis for links between behaviour and facial characteristics. Particularly, associations between testosterone and fWHR should be examined in light of recent work indicating that fWHR is not sexually dimorphic (Lefevre et al., 2012, Özener, 2012, Stirrat et al., 2012). Moreover, in line with the challenge hypothesis and the work described above, facial metrics associated with aggression and status (e.g. fWHR) may be more closely linked to testosterone reactivity in response to competitive mating opportunities compared to baseline testosterone measures (Ronay and von Hippel, 2010, Roney et al., 2007, Roney et al., 2003). Accordingly, here we sought to establish whether fWHR, alongside three sexually dimorphic facial metrics (Lefevre et al., 2012, Penton-Voak et al., 2001) and a global measure of facial masculinity (Pound et al., 2009), predicts baseline testosterone (samples 1 and 2) and testosterone reactivity in response to competitive exposure to potential mates in a speed-dating context in (sample 1).

Section snippets

Sample 1

Facial metrics and testosterone measures from 188 Caucasian men (mean age = 33.6 years, SD = 7.5, range: 20 to 54 years) who participated in the Berlin Speed Dating Study (Asendorpf, Penke, & Back, 2011) were analysed. All participants were singles drawn from the general population whose motivation to participate was the chance to find a real-life romantic or sexual partner.

Sample 2

This sample consisted of 79 Caucasian male undergraduate students (mean age: 20.50 years, age range: 18–25 years).

Speed-dating procedure in Sample 1

Participants

Facial metrics and testosterone

Testosterone levels more than three standard deviations from the mean were removed prior to analyses, since outliers of such magnitude are likely caused by imperfect testosterone sampling (e.g. contamination). This reduced the dataset by 3, and 4 participants for baseline and post-exposure testosterone levels, respectively. Furthermore, all testosterone variables were positively skewed and were therefore transformed to normal distribution using square-root transforms.1

Discussion

Evidence from two samples supports a link between circulating testosterone levels and behaviourally-relevant facial structures in adult men; namely, facial width. Of particular interest, facial width-to-height ratio (fWHR) showed positive marginal associations with baseline testosterone in sample 1 and positive significant associations baseline testosterone in sample 2. Moreover, in sample 1 fWHR was positively associated with testosterone reactivity following exposure to potential mates. These

Supplementary Materials

The following are the Supplementary data to this article.

Sample1_forEHB.sav

Sample2_forEHB.sav

References (47)

  • B. Özener

    Facial width-to-height ratio in a Turkish population is not sexually dimorphic and is unrelated to aggressive behaviour

    Evolution and Human Behavior

    (2012)
  • I.S. Penton-Voak et al.

    High salivary testosterone is linked to masculine male facial appearance in humans

    Evolution and Human Behavior

    (2004)
  • M. Peters et al.

    Testosterone is associated with mating success but not attractiveness or masculinity in human males

    Animal Behaviour

    (2008)
  • J.R. Roney et al.

    Rapid endocrine responses of young men to social interactions with young women

    Hormones and Behavior

    (2007)
  • J.R. Roney et al.

    Behavioral and hormonal responses of men to brief interactions with women

    Evolution and Human Behavior

    (2003)
  • M. Stirrat et al.

    Male facial width is associated with death by contact violence: Narrow-faced males are more likely to die from contact violence

    Evolution and Human Behavior

    (2012)
  • J.H.J. Williams et al.

    Second to fourth finger ratio and possible precursors of developmental psychopathology in preschool children

    Early Human Development

    (2003)
  • J.B. Asendorpf et al.

    From dating to mating and relating: Predictors of initial and long-term outcomes of speed-dating in a community sample

    European Journal of Personality

    (2011)
  • M.D. Back et al.

    Knowing your own mate value: Sex-specific personality effects on the accuracy of expected mate choices

    Psychological Science

    (2011)
  • M.D. Back et al.

    Why mate choices are not as reciprocal as we assume: The role of personality, flirting, and physical attractiveness

    European Journal of Personality

    (2011)
  • J.M. Carré et al.

    In your face: Facial metrics predict aggressive behaviour in the laboratory and in varsity and professional hockey players

    Proceedings of the Royal Society B

    (2008)
  • J.M. Carré et al.

    Facial structure is a reliable cue of aggressive behavior

    Psychological Science

    (2009)
  • V. Coetzee et al.

    Deciphering faces: Quantifiable visual cues to weight

    Perception

    (2010)
  • Cited by (196)

    • Gender, personality, and performance

      2024, Journal of Behavioral and Experimental Economics
    • Audit partner facial traits, gender, and career outcomes

      2023, Accounting, Organizations and Society
    • Does the primate face cue personality?

      2023, Personality Neuroscience
    View all citing articles on Scopus
    View full text