Variations in men's facial (e.g., symmetry, masculinity) and body (e.g., shoulder-to-hip ratio) morphology are related to women's ratings of attractiveness. Specifically, women tend to rate more symmetrical and masculine faces, and higher SHR[1] as attractive during fertile phases of their menstrual cycle and for short-term mating partners[2], [3], [4], [5], [6], [7], [8]. Several recent findings[9], [10], [11], [12] demonstrate that faces that are rated as attractive activate neural reward substrates in females, and that this neural activation is partly modulated by hormonal status of women[11]. These findings support the hypothesis that men's facial and body morphology serves as an honest biosignal to genetic fitness.

Women have a larger proportion of fat stores in comparison to non-human female primates and men[13]. Prior to puberty, a sexual dimorphism in fat distribution is not apparent, however, in post-menarche women estrogen inhibits storage of fat around abdominal areas and fat is stored in the gluteofemoral (thighs and buttock) region[14]. Women have a lower waist-to-hip ratio (WHR) than men and men's preference for lower WHR, or hour-glass figures, appears to be cross-cultural[15], [16]c.f.[17], [18] and adaptive because WHR is positively correlated with fecundity, levels of the reproductive hormones estradiol and progesterone[19], higher likelihood of conception[20], and availability of neurodevelopmental resources for offspring[21], but negatively correlated with lower likelihood of illnesses such as heart disease and various cancers[15], and lower incidence of depression[22]. Differences in WHR can also be used to index pregnancy and the capacity for unencumbered childbirth[1]. Consequently, it's been hypothesized that distribution of body fat represents a secondary sexual characteristic that, in the same way variation in men's facial and body morphology does, can influence ratings of attraction and fecundity (see also[23] for arguments regarding female BMI).

In order to determine whether the WHR surgical procedure had an effect on brain activation we first computed a contrast between pre-surgical and post-surgical bodies. Relative to viewing pre-surgical bodies, viewing post-surgical bodies revealed activation in right orbital frontal cortex (OFC) (Z = 4.11, p<.01 cluster corrected), lateral occipital cortex (Z = 3.82, p<.01 cluster corrected), and the anterior cingulate gyrus (Z = 3.71, p<.01) (Fig. 1a–b). Because the OFC has been associated with reward evaluation we created a bilateral OFC region of interest (ROI) mask to investigate the extent of activation in this area specifically. This analysis revealed that post-surgical images activated the left (Z = 3.59, p<.01) and the right (Z = 3.77, p<.01) OFC, but pre-surgical images did not (Fig. 2). An ROI of activation in bilateral amygdala, another area often associated with reward processing[26], [27], [28], [29] and recently associated with implicit social judgment making[30], [31] revealed no voxels survived statistical threshold for pre- or post-surgical bodies. This finding demonstrates that changes associated with cosmetic surgery (aimed at optimizing female WHR) have specific effects on men's OFC, an area of the brain that is associated with evaluations of rewards.

We then regressed each woman's change in BMI and WHR between pre-surgery and post-surgery onto the participants' brain activation to determine how changes in BMI and WHR differentially affected the brains of men. While changes in BMI only revealed activation in visual brain substrates (right lingual gyrus Z = 2.72, p<.001 uncorrected; left fusiform gyrus Z = 2.25, p<.001 uncorrected; and right lateral occipital cortex Z = 2.21, p<.001 uncorrected) (Fig. 3a), changes in WHR revealed activation in the right para-anterior cingulate gyrus (Z = 2.58, p<.001 uncorrected) (Fig. 3b). This finding suggests that changes in BMI activate low-level visual areas that are tuned to noticing variations in body configuration, but not involved in the aesthetic evaluation of the body. On the other hand, changes in WHR activated the anterior paracingulate gyrus, which has been associated with reward processing[26], [29], social evaluation and decision-making[32], [33], and responses under uncertainty and adaptive personal significance[34].

Lastly, we regressed each participant's idiosyncratic ratings of attractiveness on their brain activation to post- minus pre-surgical bodies to investigate how explicit ratings of body attractiveness predict brain activation. This revealed activation in various forebrain structures including the left (Z = 3.40, p<.001 uncorrected) and right (Z = 4.30, p<.001 uncorrected) OFC, left putamen (Z = 3.70, p<.001 uncorrected), left nucleus accumbens (Z = 4.22, p<.001 uncorrected), left (Z = 4.93, p<.001 uncorrected), right (Z = 3.90, p<.001 uncorrected) caudate, and right thalamus (Z = 3.47, p<.001 uncorrected) (Fig. 4). This activation pattern, particularly activation of the nucleus accumbens, suggests that attractiveness ratings were associated with activation in neural reward centers[9], [10], [26], [27], [28], [35], [36], [37], [38], [39], [40] that have also been associated with drug/alcohol-induced reward^c.f. [41], [42], [43].

Our observed activation patterns in the brain suggest that female body configuration represents a salient stimulus to men and that optimal female body configurations activate areas of men's brains that are associated with reward processing and appetitive behaviors. This activation may represent the proximate neural mechanism of attraction to females that express curvaceous body types and also further account for cross-cultural findings showing that optimal WHR (∼.7) as being consistently rated as attractive. Lastly, this finding may extend our understanding of some men's proclivity to develop a preoccupation with stimuli depicting optimally designed women (e.g., pornography)[44], [45]. Interestingly, our findings did not demonstrate that BMI had a large effect on brain activation except in areas associated with simple visual evaluations of shape and size. This does not downplay the importance of BMI in evaluations of female attractiveness, but may suggest that BMI's role in these evaluations is less the product of evolved psychological mechanisms and more the part of culturally driven, or societal based norms and perceptions.

The authors would like to thank Magdalena Babiszewska for her assistance with data collection. The authors would also like to thank Rebecca Burch, J. Mark Davis, Jennifer Davis, Andrew Gallup, Gordon Gallup, Aaron Goetz, Austen Krill, Katherine Reding, Todd Shackelford, Kyrre Wahtne and the staff at MARIARC for comments on an earlier draft and assistance with fMRI data acquisition.