Piers Cornelissen has offered some pointed objections to concluding that WHR (waist-hip ratio) is more important than BMI (body mass index) in the evaluation of women’s body attractiveness. He posted these objections as a response to Caution: Curves Ahead. Since comments are easily overlooked on this blog due to the nature of the site’s formatting, I have decided to highlight his argument by reproducing it here.
A few points regarding Caution: Curves Ahead
1) BMI and WHR range effects.
Smith et al. (Smith, K.L., Cornelissen, P.L. & Tovée, M.J. (2007) Colour 3D Bodies and Judgements of Human Female Attractiveness. Evolution and Human Behaviour, 28, 48-54) used colour videos of women, who were rotated twice through 360 deg, as stimuli in an attractiveness rating paradigm. The relative range for WHR and BMI in these stimuli can be described by looking at the range of their z-scores: -2.55 to 2.72 and -1.70 to 1.89 respectively. In other words, there was more variability in WHR than BMI. However, the correlations between attractiveness and WHR / BMI in this study were -0.28 (p=0.06) and -0.73 (p<0.0001) respectively.
A similar result was reported in Tovee et al. (Tovée, M.J., Warren, T.T.L., Hancock, P. & Cornelissen, P.L. (2002). Visual cues to female attractiveness: Waveform analysis of body shape. Proceedings of The Royal Society, B Vol. 269, No. 1506., 2205-2213). In this study, using 2D gray level images in front view, stimuli were picked deliberately so that the WHR range outweighed the BMI range by a factor of ~3.
In conclusion, using videos / photos of whole bodies, we have repeatedly found that we can’t get WHR to work in explaining attractiveness ratings even when it has a bigger relative range than BMI.
2) False positives.
If Caucasian males were to rely primarily on WHR for mate choice when judging the bodies of potential partners, they would be prone to making false positive errors; sometimes they would pick women with amenorrhoea as partners who are infertile. As Fig. 1 shows in Tovee et al. (Tovée, M.J. and Cornelissen, P.L. (1999) Visual cues to female physical attractiveness. Proceedings of the Royal Society of London Series B-Biological Sciences, 266(1415), 211-218), it is perfectly possible to find individual females with optimally attractive WHRs but who have extremely low body fat, and who have amenorrhoea as a consequence. So, it is hard to see how WHR alone can be treated as a reliable signal.
3) Micrograft surgery BMI / WHR manipulations.
Clearly this is an elegant paradigm in principle. Indeed, based on their Fig. 2A, I would be hard pushed *not* to agree that the post-operative figures in Dixson et al. are more attractive (Dixson, B., Sagata, K., Linklater, W., & Dixson, A. (2009). Male preferences for female waist-to-hip ratio and body mass index in the highlands of Papua New Guinea American Journal of Physical Anthropology DOI: 10.1002/ajpa.21181). However, my reason for this decision would have to include the fact that the post-operative images are aesthetically more pleasing because they are rounder and smoother. This image property was neither quantified nor coded qualitatively in any way, and no such measures were included in the analyses either as outcome variables or covariates. Moreover, from an image analysis point of view, a smoothness metric could be obtained without having anything to do with WHR. Therefore, it would be useful to check that roundness / smoothness is not a confound in future research of this kind.
Secondly, even if the conclusion from these experiments is correct, we still have to explain why WHR fails as a cue when photos/videos of the *whole* body are available. Are we really suggesting that the answer lies in males *only* making their mate selection choice when they have a close up view – because that is the implication.
Most beauty research is done using static stimuli: static photographic images are used rather than, for example, dynamic video depictions. Since most real-world interactions with others do not involve static presentations, it is important to know how similar these two types of attractiveness ratings typically are.
Roberts, et.al. (2009) have recently reported finding a positive correlation between attractiveness ratings of static images (neutral photographs) and dynamic images (video recordings). Only a small number of studies have looked at this relationship, with mixed results:
A review of this table shows inconsistent findings in between subjects designs. The Roberts, et.al. study is the first to date to use a within subject design to examine this relationship. The table below shows the various correlations between ratings of static and dynamic stimuli utilizing a within subject design. As you can clearly see, there are strong correlations in this data set.
- The use of a novel design for exploring this research question.
- Stimuli and raters were from college student samples
- As the authors acknowledge, within subject designs – which involve repeated measures – are subject to carry-over effects (the alteration in the rating of the second stimulus (e.g., the video recording) that is a result of having seen the first stimulus (e.g., the neutral photograph).
- The nature of the task performed for the video recording can have an effect on the correlation. In this study, correlations were stronger between the static image and the dynamic images when the stimuli involved a hypothetical self-introduction in a bar setting than when the topic of the dynamic recording was a recent holiday.
In defense of the common use of static imagery in beauty research it should be emphasized that static depictions enable better control over extraneous variables and enable clearer comparisons of the influence of purely structural bodily/facial characteristics.
Roberts, S., Saxton, T., Murray, A., Burriss, R., Rowland, H., & Little, A. (2009). Static and Dynamic Facial Images Cue Similar Attractiveness Judgements Ethology, 115 (6), 588-595 DOI: 10.1111/j.1439-0310.2009.01640.x
In my last posting, I found myself musing about how WHR would influence ratings of body attractiveness if BMI was held constant. Recent research comparing the relative roles of BMI and WHR have tended to support a more prominent role for BMI over WHR. That is, the total amount of body fat seems to matter more than how that body fat is distributed. One recent study (Cornelissen, et.al., 2009) claims to have resolved the debate, concluding:
that although WHR appears to be an important predictor of attractiveness, this is largely explained by the direct effect of total body fat on WHR, thus reinforcing the conclusion that total body fat is the primary determinant of female body shape attractiveness.
I have found 3 recent or in press publications that have in many ways addressed my question [Singh, et.al. (in press); Dixson, et.al. (2010); and Dixson et.al. (in press)] and each reaches the opposite conclusion from Cornelissen; WHR is more important than BMI in determining female body attractiveness. Each uses before/after images of micrograft surgery in which fat is removed from the waist and implanted in the buttocks/hips (producing results similar to the liposuction on the right). This cosmetic surgery minimally impacts BMI but does reduce WHR. Using this methodology, each study concludes that WHR has a greater influence on attractiveness ratings than BMI.
- Novel methodology
- Results found in several cultures: China (Dixson et.al. (in press); Papua New Guinea (Dixson et.al. (2010); Samoa, Komodo Island, Cameroon, and New Zealand (Singh et.al. (in press)
- Not all before/after stimulus images show that a reduced WHR is more attractive to raters. WHR does not explain all of the variation in ratings.
Dixson et.al. (2010) suggest that studies which have found BMI to be more important than WHR have used stimuli with a wide range of BMI’s and a relatively restricted range of WHR’s – which likely would have the effect of inflating the influence of BMI. These three studies in effect do the reverse: use an expanded WHR range and a reduced BMI range: not surprisingly, they find the reverse outcome. It looks like this debate isn’t resolved after all….
Photo courtesy of Dr. Mordcai Blau and David A. Copeland 2009
CORNELISSEN, P., TOVEE, M., & BATESON, M. (2009). Patterns of subcutaneous fat deposition and the relationship between body mass index and waist-to-hip ratio: Implications for models of physical attractiveness Journal of Theoretical Biology, 256 (3), 343-350 DOI: 10.1016/j.jtbi.2008.09.
Dixson, B., Sagata, K., Linklater, W., & Dixson, A. (2009). Male preferences for female waist-to-hip ratio and body mass index in the highlands of Papua New Guinea American Journal of Physical AnthropologyDOI: 10.1002/ajpa.21181
Dixson, B., Baoguo, L., & Dixson, A. (in press). Female waist-to-hip ratio, body mass index and sexual attractiveness in China. Current Zoology.
Singh D, Dixson BJ, Jessop TS, Morgan B, Dixson AF. (in press). Cross-cultural consensus for waist- to-hip ratio and women’s attractiveness. Evol Hum Behav.
BMI – the ratio of body mass to height, typically correlates well with ratings of body attractiveness. WHR – a direct comparison of waist and hip measurements – also correlates with attractiveness. Recent research that compares the relative strengths of the two ratios generally finds that variation in BMI accounts for a greater proportion of variation in attractiveness ratings than does variation in WHR. The implication is that, at least in contemporary industrial/technological societies, levels of body fat matter more than how that body fat is distributed. I found myself reflecting on these ratios in relation to women’s body attractiveness today, and wondered how WHR would influence ratings of body attractiveness if BMI was held constant? My guess was that WHR would be more strongly correlated with attractiveness ratings when controlling for BMI in this way. (I couldn’t recall a study that explored this possibility and I also could not find one in the literature – if you know of one please post a link or citation.) My rationale was that if subjects are matched for BMI, then WHR variation would likely result from variation in estrogen efficacy. My hypothesis was that, other things being equal, curviness resulting from estrogen efficacy would more strongly influence attractiveness ratings.
So far my thinking has been pretty predictable. Then I reflected on estrogens’ role in developing the sexually dimorphic features that are found attractive in women’s faces (Smith, et.al, 2006). That’s when I realized that, to date, comparisons of WHR and BMI are done on ratings of body attractiveness alone. This practice is sensible, since cognitively, evaluations of faces and bodies are separate processes. But, since estrogens significantly influence both facial attractiveness and body attractiveness, these two ratings should be related. [There is some support for this relationship (Thornhill & Grammer, 1999).]
These musings leave me wondering: might WHR be a better predictor of overall attractiveness than BMI in women?
Image of the 3rd century Bikini Girls mosaic from the Villa Romana in Sicily courtesy of Roundtheworld. Wikipedia Commons.
Law Smith, M., Perrett, D., Jones, B., Cornwell, R., Moore, F., Feinberg, D., Boothroyd, L., Durrani, S., Stirrat, M., Whiten, S., Pitman, R., & Hillier, S. (2006). Facial appearance is a cue to oestrogen levels in women Proceedings of the Royal Society B: Biological Sciences, 273 (1583), 135-140 DOI: 10.1098/rspb.2005.3296
Thornhill, R. (1999). The Body and Face of Woman One Ornament that Signals Quality? Evolution and Human Behavior, 20 (2), 105-120 DOI: 10.1016/S1090-5138(98)00044-0
Using digitally manipulated levels of sexual dimorphism in human male and female faces (like the ones to the right), Glassenberg et.al. (2009) found that, compared to heterosexual women, homosexual women preferred greater masculinization in female faces [Brown-Forsythe t(303.38) = -2.92, p<.01] while heterosexual women preferred greater masculinization in male faces [t(375) = 6.77, p<.001]. Compared to heterosexual males, homosexual males preferred masculinization in both male and female faces [t(520) = -7.42, p<.001 and t(520) = -6.72, p<.001 respectively]. Calculations based on sociosexual orientation were mostly non-significant, though relatively small, significant, positive correlations were found in heterosexual males between unrestricted SO and a preference for feminization in female faces [R(125) = .20, p<.05] while homosexual males showed a positive correlation between unrestricted SO and a preference for masculinized male faces [R(259) = .17, p<.001]. These specifics aside, all raters preferred feminized female faces to masculinized female faces.
- large sample
- rated stimuli consisted of 3 face composites to ensure recognizable individuality. There was no effort to match stimuli for attractiveness prior to manipulating sexual dimorphism, so an attractiveness x dimorphism interaction would be missed in this design.
This study suggests that homosexuals’ preferences are neither identical to nor mirror-images of heterosexuals’ preferences. This data also suggests that researchers should control for sexual orientation when conducting attractiveness studies in which sex/gender are relevant variables.
Glassenberg, A., Feinberg, D., Jones, B., Little, A., & DeBruine, L. (2009). Sex-Dimorphic Face Shape Preference in Heterosexual and Homosexual Men and Women Archives of Sexual Behavior DOI: 10.1007/s10508-009-9559-6
French Elle’s April 2009 edition highlighted make-up free beauty. Given the social effects of the “perfectly” beautiful images in contemporary media, Elle’s edition is noteworthy. Three covers were used – the crop below right is of the one featuring Monica Bellucci. The photo above right is from Wikipedia Commons – showing Ms. Bellucci in make-up. So, what’s different? To my eye, the most obvious effects of makeup are: greater contrast between the eyes/lips and the surrounding skin; a more even skin tone, and more color in the cheeks. Increasing contrast between the eyes/lips and the surrounding skin enhances a human sex difference: having the effect of hyper-feminizing women (Russell, in press). Rather than lightening the skin, the usual technique is to darken the eyes and lips. Evening the skin tone makes faces appear more attractive, youthful, and healthy (Fink, et.al., 2006). Increasing cheek redness likely makes faces – especially women’s faces – appear healthier (Stephen, et.al., 2009).
I have discussed Russell’s work on sex differences in facial contrast in a recent blog entry – so I will not review that here. In a novel approach, Fink and colleagues converted skin tone information from digital photographs of actual faces into 3D illustrations that kept facial structure, hair color and style, and eye color constant (see below). Raters then evaluated the 3D creations for youthfulness, age, health, and attractiveness. The correlation between the estimated age of the 3D figures and the actual age of the photographed women was good (r=.708, p<.01). The estimated age range for the 3D figures appears to be 20-31; while the actual age range of the photographed women was 11-76 (mean=37.39, S.D.=17.35). This suggests that skin coloration alone contributed about 12 years to age estimates in this study and that other cues to aging may have a larger individual impact. Estimated age correlated with attractiveness (r=-.557, p<.01), healthy appearance (r=-.543, p<.01), and youthfulness (r=-.871, p<.01).
I am enthusiastic about the use of digital imaging in the study of beauty. But, to my eye, there is something just a little off with these images. My reaction leads me to one caution: as the use of computer generated stimuli become more common in beauty research, the risk of getting stuck in the uncanny valley becomes greater. While still mostly theoretical (but a quick search of Science Direct suggests that more empirical data is forthcoming), the notion of the uncanny valley is that as robots and 3D animations become almost human-like, they will produce an “uncanny” negative reaction – one that could interfere with beauty research.
Some carefully controlled research from Stephen and colleagues suggests a reason for the typical use of rouge in women’s cosmetic applications: to appear healthy. The research under discussion does not directly address reddened cheeks, but did find a tendency for increased levels of the color “oxygenated blood-red” in faces that appeared healthy to evaluators. All-in-all, recent research suggests that artfully applied makeup should increase ratings/evaluations of femininity, youthfulness, health, and attractiveness.
FINK, B., GRAMMER, K., & MATTS, P. (2006). Visible skin color distribution plays a role in the perception of age, attractiveness, and health in female faces☆ Evolution and Human Behavior, 27 (6), 433-442 DOI: 10.1016/j.evolhumbehav.2006.08.007
Stephen, I. D., Coetzee, V., Law Smith, M., & Perrett, D. I. (2009). Skin Blood Perfusion and Oxygenation Colour Affect Perceived Human Health. PLoS ONE, 4(4), e5083. doi: 10.1371/journal.pone.0005083
Russell, R. (in press). A sex difference in facial pigmentation and its exaggerationby cosmetics. Perception.
A common goal in most oculoplastic procedures is to increase symmetry. In an effort to establish baseline measures in attractive subjects, Ing et.al. (2006) measured ocular asymmetries in male and female models’ photos in fashion magazine advertisements (e.g., Cosmopolitan, Elle, Glamour, Vogue, Gentleman’s Quarterly, etc.). They found significant asymmetries in:
- horizontal fissure width (1)
- upper central lid fold (5)
- upper temporal lid fold (7)
- central eyebrow height (9)
- temporal eyebrow height (11)
- medial canthal to midline distance
- pupil to midline distance
- orbital distopia (asymmetrically displaced eyes)
While I applaud the effort to establish realistic expectations of beauty, I do not believe that the methods used in this study can reach valid conclusions regarding each of the numbered measures in the bulleted list above. Each of these measures can vary based on facial expression (if you like, you can demonstrate this point to yourself in front of a mirror). Even in cases where fashion models’ expressions appear neutral in a magazine ad, we cannot assume that subtle asymmetries are not the result of subtle expressions – as opposed to assuming they result from structural asymmetries.
That being said, attractive models are not always perfectly symmetrical. A cursory visual inspection of beauty shots (essentially, close-ups of faces intended to look beautiful) will reveal asymmetries in beautiful models that are visible to the naked eye.
Ing E, Safarpour A, Ing T, & Ing S (2006). Ocular adnexal asymmetry in models: a magazine photograph analysis. Canadian journal of ophthalmology. Journal canadien d’ophtalmologie, 41 (2), 175-82 PMID: 16767204