Studies of Human Physique and Sexual Attractiveness: Sexual Preferences of Men and Women in China

AMERICAN JOURNAL OF HUMAN BIOLOGY 19:88–95 (2007)

Original Research Article Studies of Human Physique and Sexual Attractiveness: Sexual Preferences of Men and Women in China

1 2 3 1 BARNABY J. DIXSON, ALAN F. DIXSON, * BAOGUO LI, AND M.J. ANDERSON 1Department of Conservation and Research for Endangered Species, Zoological Society of San Diego, San Diego, California 2School of Biological Sciences, Victoria University of Wellington, Wellington, New Zealand 3College of Life Sciences, and Key Laboratory Resource Biology and Biotechnology in Western China, Ministry of Education, Northwest University, Xi’an, China

ABSTRACT Men and women at Northwest University (n ¼ 631), Xi’an, China, were asked to rate the attractiveness of male or female figures manipulated to vary somatotype, waist-to- hip ratio (WHR), secondary sexual traits, and other features. In study 1, women rated the average masculine somatotype as most attractive, followed by the mesomorphic (muscular), ectomorphic (slim), and endomorphic (heavily built) somatotypes, in descending order of preference. In study 2, the amount and distribution of masculine trunk (chest and abdominal) hair were altered progressively in a series of front-posed figures. Women rated figures with no or little trunk hair as most attractive. Study 3 assessed the attractiveness of front-posed male figures which varied only in length of their nonerect penis. Numerical ratings for this trait were low, but moderate lengthening of the penis (22% or 33% above average) resulted in a significant increase in scores for attractiveness. In study 4, Chinese men rated the attractiveness of back- posed female images varying in waist-to-hip ratio (WHR from 0.5–1.0). The 0.6 WHR figure was most preferred, followed by 0.7, while figures with higher ratios (0.9 or 1.0) were significantly less attractive. Study 5 rated the attractiveness of female skin color: men expressed a marked preference for images which were lighter in color, as compared to images of average or darker skin colors. These results, the first of their kind reported for a Chinese population, support the view that sexual selection has influenced the evolution of human physique and sexual attractiveness in men and women. Am. J. Hum. Biol. 19:88–95, 2007. ' 2006 Wiley-Liss, Inc.

There is increasing evidence that sexual with higher levels of 17-b-estradiol and selection has influenced the evolution of greater reproductive potential in women human morphology, especially via sexual pref- (Jasien´ ska et al., 2004). Women with higher erences for features which signal female re- follicular-phase levels of estrogen are also productive potential or masculine strength judged to be facially more attractive by men and fitness. In women, for example, the waist- (Law-Smith et al., 2006). Measurements of the to-hip ratio (WHR) may provide an honest sig- body mass index (BMI) of women are also cor- nal of fat distribution in relation to repro- related with female attractiveness; the rela- ductive health (Singh, 1993, 2002; Singh and tive importance of BMI and WHR in this Young, 1995). A low female WHR (0.7) is respect was debated by various authors (e.g., judged as most attractive by men in some cul- Toveé et al., 1999; Streeter and McBurney, tures (e.g., North America: Singh and Young, 2003; Schu¨ tzwohl, 2006). 1995), but not in all human populations (e.g., Skin color is another trait which may be the Matsigenka of Peru: Yu and Shepard, related to female attractiveness. Darwin (1871) 1998; the Hadza of Tanzania: Wetsman and noted that women in various parts of the world Marlowe, 1998). Recent work on the Hadza confirmed the importance of a low WHR in *Correspondence to: Alan F. Dixson, School of Biological female attractiveness, provided that side- Sciences, Victoria University of Wellington, PO Box 600, posed images are used which allow men to Wellington, New Zealand. E-mail: [email protected] Received 17 January 2006; Revision received 5 August include judgments of the shape and size of the 2006; Accepted 8 August 2006 female buttocks (Marlowe et al., 2005). Nar- Published online in Wiley InterScience (www.interscience. row waists and large breasts are correlated wiley.com). DOI 10.1002/ajhb.20584

VC 2006 Wiley-Liss, Inc. SEXUAL PREFERENCES IN CHINA 89 tend to have lighter skin than men. He quoted significant role in visual display either within, Schweinfurth, for example, concerning the or between, the sexes (Short, 1980; Diamond, Monbuttoos of Africa: ‘‘Like all her race she 1997; Potts and Short, 1999). It may play some had a skin several shades lighter than her hus- role in female sexual preference, although band’s, being something of the colour of half- with some exceptions (Stulhofer, 2006), the roasted coffee.’’ Some evidence to support evidence for this remains largely anecdotal. effects of sexual selection on female skin color The current study was undertaken in order has been obtained (van den Berghe and Frost, to obtain information on the sexual preferences 1986; Frost, 1988; Symons, 1995). Female skin of young men and women in a Chinese popula- color changes with age and reproductive condi- tion. Questionnaires were used to measure tion (e.g., in pregnant and multiparous female preferences for images which varied in women), so that it might provide men with vis- masculine physique (somatotype), distribution ual cues of female age and reproductive condi- of chest and abdominal hair, and penile size, as tion (Symons, 1995). However, few cross-cul- well as men’s preferences for female waist-to- tural data have been collected to assess men’s hip ratio and skin color. The results, which are preferences for female skin color. the first of their kind obtained for a Chinese Where masculine physique is concerned, population, are discussed in relation to studies there is evidence that taller men achieve conducted in other parts of the world. greater reproductive success (Pawlowski et al., 2000), and that a broad chest, narrow waist, METHODS and muscular torso are rated as attractive by women (Mealey, 1997; Lynch and Zellner, 1999; Subjects Maisey et al., 1999). These traits may have This study was conducted at Northwest played an important role in human evolution, University, in the city of Xi’an, Shaanxi Prov- given the requirement for men to compete with ince, China. Students were asked to partici- one another, to provide protection for female pate by filling out a questionnaire and rating partners, and to assure competence in hunting various images (see below) for sexual attrac- and other physically demanding survival skills tiveness. A total of 631 subjects participated (Buss, 2003; Bramble and Lieberman, 2004). in the study (311 men and 320 women). Men are more hirsute than women, and in a va- riety of nonhuman primate species, reproduc- Questionnaires tively mature males exhibit androgen-dependent capes of hair (Dixson, 1998). Sexual dimor- Each questionnaire had a cover sheet phism in the distribution of trunk (chest and (written in Mandarin Chinese) for demographic abdominal) hair in human beings might play data, including the respondent’s sex, age, and some role in masculine sexual attractiveness. marital status. Participation was voluntary, Human beings are unique among primates in and at no time were respondents asked to give being bipedal, so that features on the chest and their names. Women were asked to rate images abdomen are displayed more prominently to of males, varying in somatotype, distribution of conspecifics. Masculine trunk hair was rated as body hair, and size of genitalia. To produce highly sexually attractive by women in studies these images, photographs of back-posed and conducted in the UK (Dixson et al., 2003). front-posed men were scanned from Sheldon Women also rated more hirsute male images as et al. (1954). Men were asked to rate the attrac- being older than those lacking chest hair. Cross- tiveness of images of women, varying in WHR cultural studies consistently show that women and skin color. Each image was manipulated prefer to marry men who are older than them- using Photoshop version 7.0 in order to stand- selves (Buss, 2003). However, there are very few ardize images for height, posture, and skin cross-cultural data on the potential attractive- color. Skin color was matched to that of a Chi- ness of masculine trunk hair, and this trait may nese sample by scanning photographs into the vary considerably between human populations. computer and using these as the basis for color- The assumption of an upright gait also ing images. Where front-posed images were means that human genitalia are displayed in used, the faces were blocked out, as our studies a different way, and perhaps more promi- did not concern facial stimuli. nently, than in typically quadrupedal mammals. A number of authors commented on the relatively large size of the human penis, for Study 1. Study 1 assessed the preferences of example, and the possibility that it plays some women for four back-posed male figures of

American Journal of Human Biology DOI 10.1002/ajhb 90 B.J. DIXSON ET AL. varying somatotype (ectomorphic, endomor- and 15 units of contrast) to produce two phic, mesomorphic, and average). images which were darker than the standard image, and two that were lighter in color. Data analysis and statistics Study 2. Study 2 assessed women’s preferences for front-posed images of males differing In studies 1–3, a repeated-measures analy- in levels of hirsuteness of the trunk (chest and sis of variance (ANOVA) was applied to deter- abdomen). Five images were presented, which mine whether attractiveness ratings for the varied only in degree of trunk hair (from none various stimuli exhibited significant differen- to pronounced hirsuteness). ces. Post hoc Scheffe´ tests were then used to test for significant differences in attractiveness ratings in pairs of stimuli, but only if the Study 3. Study 3 examined women’s prefer- initial ANOVA had yielded a significant result ences for five male images varying only in (P ¼ < 0.05). In studies 4 and 5, w2 tests were length of the (nonerect) penis. In one figure, used to compare observed vs. expected scores the penis was its original size (i.e., the same for preferences expressed by men for the vari- size as the original photograph used to model ous female images. the image in the computer). In the remaining four images, penile length was modeled to be RESULTS either smaller or larger than the original size. A total of 320 women answered the ques- We had intended to alter penile lengths to be tionnaire: 68% of them were 20 years or less 80%, 120%, 130%, and 140% of original size. in age, and the remainder were aged 21–30 In the event, measurement of the actual years. Only one of these women was married. images confirmed that they were 78%, 122%, The ages of male respondents (n¼311) were 133%, and 143% of original size, respectively. distributed as follows: <20 years old ¼ 44%; In all three studies, women rated each of 21–30 years ¼ 54.5%; 31–40 years ¼ 1%; >40 the stimulus images on a six-point scale where ¼ ¼ ¼ years 0.5%. Six of these men (i.e., 2%) were 0 unattractive, 1 only slightly attractive, married. Hence our sample consisted predomi- 2 ¼ mildly attractive, 3 ¼ moderately attrac- ¼ ¼ nantly of young and unmarried university stu- tive, 4 very attractive, and 5 extremely dents in their late teens or twenties. attractive. The scale appeared next to each image, as in previous studies (Dixson et al., Study 1: Women’s ratings of masculine 2003). somatotypes ANOVA showed a significant effect for wom- Study 4. Study 4 examined men’s preferen- en’s ratings of the attractiveness of the four ¼ < ces for back-posed images of the same women, male figures (F[3,957] 293, P 0.0001). The differing only in waist-to-hip ratios. Using average somatotype was judged as most Photoshop, six images were produced (WHR ¼ attractive, followed by the mesomorphic soma- 0.5, 0.6, 0.7, 0.8, 0.9, and 1.0). These images totype and the ectomorphic image. The endo- were presented in random order on the same morphic somatotype was clearly rated as the page of the questionnaire; respondents were least attractive image (Fig. 1). Post hoc asked to select only the image that they found Scheffe´ tests revealed the endomorphic image most attractive sexually. On a subsequent to be significantly less attractive than any of < page, the same stimuli were presented again, the other three somatotypes (P 0.001 for but respondents were asked to select the one each paired comparison). The male image they found most attractive for a long-term depicting an average somatotype was rated as relationship. significantly more attractive than the mesomorphic figure (P < 0.001), and the meso- morph achieved a higher rating than the ecto- < Study 5. Study 5 assessed male preferences morphic image, but only slightly so (P 0.05; for variations in female skin color. Five images Fig. 1). of an identical back-posed female figure Study 2: Women’s ratings of male images (WHR, 0.8) were used. One standard image varying in hirsuteness had skin typical for a Chinese woman. Skin tone was altered (using Adobe Photoshop) in a ANOVA showed that the five male (meso- stepwise fashion (by 10 units of brightness morphic) figures differing only in amounts of

American Journal of Human Biology DOI 10.1002/ajhb SEXUAL PREFERENCES IN CHINA 91

Fig. 2. Women’s mean ratings (þ SEM) for attractiveness of front-posed male figures which vary only in hirsuteness of trunk (chest and abdomen). None, no trunk hair; Max, pronounced hirsuteness. *P < 0.05, ***P < 0.001. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley. com.] Fig. 1. Women’s mean ratings (þ SEM) for sexual attractiveness of back-posed male figures of four different somatotypes: ENDO, endomorphic; ECTO, ectomor- ings were mean ¼ 1.87 and mean ¼ 1.84 for phic; MESO, mesomorphic; AVER, average body build. images in which the penis was 22% or 33% < < *P 0.05, ***P 0.001. [Color figure can be viewed in longer in the original, respectively (it should the online issue, which is available at www.interscience. ¼ wiley.com.] be recalled that a rating of 1 slightly attractive, and 2 ¼ mildly attractive). All images used in study 3 depicted the same mesomor- trunk (chest and abdominal) hair varied sig- phic front-posed figure, lacking trunk hair, nificantly in their attractiveness to women and were identical except for their genitalia. (F[4,1276] ¼ 76.69, P < 0.0001). Most attrac- The image depicting an unaltered penile tive was the image lacking any trunk hair, length and lacking trunk hair also formed with a progressive decline in attractiveness as part of study 2, on masculine hirsuteness. In more hair was added to the image (Fig. 2). The that study, this figure scored a mean score of image with the most chest hair was rated as 2.48, whereas in the study on penile size the significantly less attractive than any of the identical image was rated as significantly less other four images (P < 0.001 for each paired attractive, with a mean score of 1.38. An comparison). unpaired t-test revealed that this difference was significant (P < 0.0001). Study 3: Penile size and attractiveness Study 4: Men’s ratings of female WHR ANOVA revealed an overall effect of penile length on attractiveness ratings (F[4,1276] ¼ The majority of men rated the female image 29.59, P < 0.0001). Women rated the figure depicting a WHR of 0.6 as most attractive sex- with the shortest penis (reduced to 78% of nor- ually, and most attractive for a potentially mal length) as least attractive, followed by the long-term relationship. Thus 163 and 148 original and largest images, with the interme- men, respectively, rated the 0.6 WHR image diate (þ22% and þ33%) lengths yielding the as most attractive (Fig. 4), compared to the highest ratings (Fig. 3). However, none of the total of 52 expected by chance (w2 test ¼ P < images was rated as being moderately attrac- 0.001). The second most favored choice was tive (score ¼ 3) on average. The highest rat- the 0.7 WHR image, significantly so in the

American Journal of Human Biology DOI 10.1002/ajhb 92 B.J. DIXSON ET AL.

Fig. 3. Women’s preferences for images of male figures varying only in length of (nonerect) penis (1–5 in order of increasing size). Data are means (þ SEM). ***P < 0.001. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley. com.] case of a prospective long-term relationship (P ¼ 0.001). The high WHR images (0.9 and 1.0) were consistently rated as less attractive than expected (P < 0.001), as was the extremely low WHR (0.5) image (P < 0.001 in Fig. 4B). Ratings for the 0.8 WHR image did not differ from those expected by chance. Study 5: Men’s ratings of female skin color There was a striking effect of skin color on men’s ratings of attractiveness of the five Fig. 4. Men’s preferences for back-posed female fig- back-posed female images (Fig. 5). The two ures varying in WHR from 0.5–1.0. A: Subjects were images with lighter-than-average skin colors asked to choose most sexually attractive image. B: Sub- jects were asked to choose most attractive image for were chosen as most attractive by 272 (87%) of potentially long-term relationship. Dashed line, ex- the 311 men who filled out the questionnaire. pected number of males preferring a given image; histo- By contrast, the two darker-than-average grams, observed scores. **P < 0.001, ***P < 0.001 (w2 female images were preferred by 15 (5%) of test). [Color figure can be viewed in the online issue, the men, and the average image was chosen which is available at www.interscience.wiley.com.] by the remaining 24 (8%). The observed preferences for the five images differed signifi- versity students); thus, only 4 of our subjects < cantly (P 0.001) from those expected by were over 30 years of age, and only 7 of them chance in every case (Fig. 5). were married. However, having acknowledged DISCUSSION these limitations, our results provide some use- ful comparisons and contrasts with the find- The studies reported here are, to our knowl- ings of similar studies conducted in the US, edge, the first of their kind conducted in Europe, and elsewhere. China. It should be apparent that they refer to Chinese women rated images of average or a limited sample of the population (631 uni- mesomorphic (i.e., muscular) masculine body

American Journal of Human Biology DOI 10.1002/ajhb SEXUAL PREFERENCES IN CHINA 93 In this study, Chinese women showed a clear preference for male images lacking hair on the chest and abdomen. This result contrasts with that obtained in the UK, where women rated hirsute male images as highly attractive and also as being older than images of men lacking trunk hair. Clearly it would be incorrect to attribute a generalized function to masculine trunk hair, as a signal of hormonal (androgenic) status or reproductive maturity. Very few human populations have been examined, and the expression of this trait probably differs significantly in different parts of the world. Chinese men may be less hirsute than men of some European populations, for example, although it is exceedingly difficult to obtain quantitative information on this question. Body depilation is also practiced by some men, as reported by Boroughs et al. (2005) for a sample 118 male university students in the Fig. 5. Men’s preferences for back-posed female fig- US (Florida). Over half (63.61%) of these men ures varying in skin color. Average, a figure whose color regularly removed body hair, especially from is typical of Chinese women. Two figures were lighter in the groin, chest, and abdomen (Boroughs color, and two figures were darker than average. Dashed line, expected number of males preferring a et al., 2005). given image; histograms, observed scores. ***P < 0.001 To our knowledge, no previous question- (w2 test). [Color figure can be viewed in the online issue, naire study measured women’s preferences for which is available at www.interscience.wiley.com.] male images which differ in genital morphology. Many authors drew attention to the rela- types as most attractive. Next preferred was tively large size of the human penis, and to the slimly built (ectomorphic) male physique, the fact that it is more prominently displayed, with the endomorphic (obese) male scoring due to the upright posture and bipedal gait of lowest for attractiveness. These results are human beings (Short, 1980; Diamond, 1991, similar to those obtained in the UK and Sri 1997; Jolly, 1999; Potts and Short, 1999; Miller, Lanka (Dixson et al., 2003) which showed that 2000; Gra¨slund, 2005). It was suggested that women rate images of endomorphic males as penile size may influence attractiveness of the unattractive. The preferred (mesomorphic and human male, via visual as well as tactile cues average) male physiques incorporate features, (e.g., Diamond, 1997; Jolly, 1999; Potts and such as a broad chest and narrow waist, which Short, 1999; Miller, 2000). Our current find- were identified as attractive features for ings, from China, indicate that women rate women in a number of studies (e.g., Mealey, the attractiveness of male images differently 1997; Schultze et al., 1991; Lynch and Zellner, if the length of the (nonerect) penis is altered. 1999; Maisey et al., 1999). Signals of mascu- The lowest ratings were for the image in which line strength and health may have been penile length was reduced, but the greatest selected for early in human evolution, because length increase (þ43%) was rated as less the possessors of such traits offered tangible attractive than smaller increases (þ22% and benefits to women. Men who were better able þ33%; Fig. 5). The comparatively low scores to protect their female partners (and off- given by Chinese women may reflect a cultural spring), to succeed in physical contests with or experiential effect (e.g., ambivalence to view- other males, and to undertake rigorous activ- ing such images or lack of sexual experience). ities (such as endurance hunting) may have In studies conducted in Cameroon and the US experienced significantly greater reproductive (California), we noted that women gave higher success (Buss, 2003). By contrast, obese or ratings, although the overall pattern was the very slim men may have less stamina and same; the lowest ratings were for extremes may be less healthy than those having a more (small or large) of penile length (Dixson et al., athletic physique (Katzmarzyk et al., 1998; unpublished data). The mechanisms controlling Bolunchuk et al., 2000). these choices are currently unknown. It might

American Journal of Human Biology DOI 10.1002/ajhb 94 B.J. DIXSON ET AL. be argued, for example, that images showing a tiveness. The results of the current study sup- longer penis are more attractive because they port the notion that lower WHRs are more signal some degree of sexual arousal. However, attractive to Chinese males than images with it would be difficult to explain the lessened a high (0.9 or 1.0) WHR. Whether WHR alone attractiveness of the largest image on that ba- or BMI is the critical variable in making such sis. The women viewed all images simultane- judgments remains to be determined. ously (arranged in random order). It is possible The evolution of lighter skin color in women, that viewing and rating each image separately as compared to men, may have been influenced might have influenced their results. Setting by sexual selection (Darwin, 1871). Changes in aside cultural and methodological speculations, pigmentation (e.g., affecting the skin of the face it remains the case that very little is known and abdomen) may occur in women as they about how important women judge penis size to mature and pass through successive pregnan- be. In a recent study conducted in Croatia, cies (Wong and Ellis, 1984; Sodhi and Sausker, Stulhofer (2006) reported that, in a sample of 1988). Male preferences for lighter skin tones 556 sexually experienced women, 75% rated might be selected for because skin color pro- penile length or girth as being either vides cues concerning female age and reproduc- ‘‘somewhat important’’ or ‘‘very important’’ for tive potential (Symons, 1995). In the current their judgment of partners. We believe that it is study, Chinese males expressed marked prefer- time to set prudery aside and conduct more ences for images of women with the lightest skin research on this topic. tones. Ethnographic comparisons confirm that Men in our study population showed pro- such preferences exist in various human popula- nounced preferences for back-posed female tions (van den Berghe and Frost, 1986; Frost, images, dependent on their WHR or skin tone. 1988). Our findings support this conclusion. A WHR of 0.6 was judged as most attractive, However, it is important to extend such studies with 0.7 as the next most favored image, espe- to a wide range of human populations, because cially when judging attractiveness for a poten- human skin pigmentation varies so greatly in tially long-term relationship. By contrast, different parts of the world, and because cul- images with a high WHR (0.9 or 1.0) were tural biases may also influence judgments of rated as significantly less attractive by Chi- attractiveness where this trait is concerned. nese men. Singh (1993, 2002) argued that a low WHR (0.7) typically reflects a healthy and LITERATURE CITED fertile reproductive status in women, based on Bolunchuk WW, Siders WA, Lykken GI, Lukaski HC. 2000. the distribution of subcutaneous and intra-ab- Association of dominant somatotype of men with body dominal fat reserves. Support for this view structure, function during exercise, and nutritional comes from studies which reported correla- assessment. Am J Hum Biol 12:167–180. Boroughs M, Cafri G, Thompson JK. 2005. Male body depi- tions between female waist size, breast size, lation: prevalence and associated features of body hair and levels of 17-b-estradiol in relation to removal. Sex Roles 52:637–644. reproductive health (Jasien´ ska et al., 2004). Bramble DM, Lieberman DE. 2004. Endurance running However, cross-cultural studies revealed male and the evolution of Homo. Nature 432:345–352. Buss DM. 2003. The evolution of desire, 2nd ed. New York: preferences for a higher WHR in some cases Basic Books. (e.g., among the Matsigenka of Peru: Wu and Darwin C. 1871. The descent of man and selection in rela- Shepard, 1998; and the Hadza of Tanzania: tion to sex. London: John Murray. Wetsman and Marlowe, 1998). A more recent Diamond J. 1991. The rise and fall of the third chimpanzee. London: Vintage. study of the Hadza by Marlowe et al. (2005) Diamond J. 1997. Why is sex fun? The evolution of human showed that preferences for higher WHR sexuality. London: Weidenfeld and Nicholson. diminished when men were asked to rate side- Dixson AF. 1998. Primate sexuality: comparative studies of posed images which varied in female buttock the prosimians, monkeys, apes and human beings. Oxford: Oxford University Press. size. Whether WHR or body mass index (BMI: Dixson AF, Halliwell G, East R, Wignarajah P, Anderson weight scaled for height) is the most impor- MJ. 2003. Masculine somatotype and hirsuteness as tant factor influencing men’s ratings of female determinants of sexual attractiveness to women. Arch images remains a controversial issue (Tassi- Sex Behav 32:29–39. Frost P. 1988. Human skin color: a possible relationship nary and Hansen, 1998; Manning et al., 1999; between its sexual dimorphism and its social perception. Toveé et al., 1999, 2001; Henns, 2000; Singh, Perspect Biol Med 32:38–58. 2002; Streeter and McBurney, 2003). Because Gra¨slund B. 2005. Early humans and their world. New WHR and BMI tend to be positively corre- York: Routledge. Henns R. 2000. WHR and female attractiveness. Evidence lated, it has proven difficult to tease apart the from photographic stimuli and methodological considera- two variables in judgments of female attrac- tions. Perspect Indiv Diff 28:501–513.

American Journal of Human Biology DOI 10.1002/ajhb SEXUAL PREFERENCES IN CHINA 95

Jasien´ ska G, Ziomkiewicz A, Ellison PT, Lipson SF, Thune Short RV. 1980. The origins of human sexuality. In: Repro- I. 2004. Large breasts and narrow waists indicate high duction in mammals, volume 8. Austin CR, Short RV, reproductive potential in women. Proc R Soc Lond [Biol] editors. Cambridge: Cambridge University Press. p 1–33. 271:1213–1217. Singh D. 1993. Body shape and women’s attractiveness. Jolly A. 1999. Lucy’s legacy: sex and intelligence in human Hum Nat 4:297–321. evolution. Cambridge, MA: Harvard University Press. Singh D. 2002. Female mate value at a glance: relationship Katzmarzyk PT, Malina RM, Song TMK, Bouchard C. 1998. of waist-to-hip ratio to health fecundity and attractive- Somatotype and indicators of metabolic fitness in youth. ness. Neuroendocrinol Lett [Suppl] 23:81–91. Am J Hum Biol 10:341–350. Singh D, Young RK. 1995. Body weight, waist-to-hip ratio, Law-Smith MJ, Perret DI, Jones BC, Cornwell RE, Moore breasts, and hips: role in judgements of female attrac- FR, Feinberg DR, Boothroyd LG, Durrani SJ, Stirrat MJ, tiveness and desirability for relationships. Ethol Socio- Whiten S, Pitman RM, Hillier SG. 2006. Facial appear- biol 16:483–507. ance is a cue to oestrogen levels in women. Proc R Soc Sodhi VK, Sausker WF. 1988. Dermatoses of pregnancy. Lond [Biol] 273:135–140. Am Fam Physician 37:131–138. Lynch SM, Zellner DA. 1999. Figure preference in two Streeter SA, McBurney DH. 2003. Waist-to-hip ratio and generations of men: the use of figure drawings illus- attractiveness: new evidence and a critique of ‘‘a critical trating differences in muscle mass. Sex Roles 40:833– test.’’ Evol Hum Behav 24:88–98. 843. Stulhofer A. 2006. How (un)important is penis size for Maisey DS, Vale ELE, Cornellisen PL, Toveé MJ. 1999. women with heterosexual experience? Arch Sex Behav Characteristics of male attractiveness for women. Lancet 35:5–6. 353:1500. Symons D. 1995. Beauty is in the adaptions of the beholder: Manning JT, Trivers RL, Singh D, Thornhill R. 1999. The the evolutionary psychology of human female sexual mystery of female beauty. Nature 399:214–215. attractiveness. In: Sexual nature, sexual conflict. Abram- Marlowe FW, Apicella CL, Reed D. 2005. Men’s preferences son PR, Pinkerton SD, editors. Chicago: Chicago Univer- for women’s profile waist to-hip ratio in two societies. sity Press. p 80–118. Evol Hum Behav 25:371–378. Tassinary LG, Hansen KA. 1998. A critical test of the WHR Mealey L. 1997. Bulking up: the roles of sex and sexual ori- hypothesis of female sexual attractiveness. Psychol Sci entation as attempts to manipulate sexual attractive- 9:150–155. ness. J Sex Res 34:223–228. Toveé MJ, Maisey DS, Emery JL, Cornellisen PL. 1999. Vis- Miller GF. 2000. The mating mind: how sexual choice ual cues to female sexual attractiveness. Proc R Soc Lond shaped the evolution of human nature. New York: Dou- [Biol] 266:211–218. bleday. Toveé MJ, Brown JE, Jacobs D. 2001. Maternal waist-to- Pawlowski B, Dunbar RIM, Lipowicz A. 2000. Tall men hip ratio does not predict child gender. Proc R Soc Lond have more reproductive success. Nature 403:156. [Biol] 268:1007–1010. Potts P, Short RV. 1999. Ever since Adam and Eve. The evo- van der Berghe PL, Frost P. 1986. Skin color preference, lution of human sexuality. Cambridge: Cambridge Uni- sexual dimorphism and sexual selection: a case of gene versity Press. culture co-evolution? Ethnic Rac Stud 9:87–113. Schultze S, Knussman R, Christiansen K. 1991. Male sex Wetsman A, Marlowe FW. 1998. How universal are prefer- role identification and body build. Homo 42:203–215. ences for female waist-to-hip ratios? Evidence from the Schu¨ tzwohl A. 2006. Judging female figures: a new meth- Hadza from Tanzania. Evol Hum Behav 20:228–229. odological approach to male attractiveness of female Wong R, Ellis CN. 1984. Physiologic skin changes in preg- waist-to-hip ratio. Biol Psychol 71:223–229. nancy. J Am Acad Dermatol 10:929–943. Sheldon WH, Dupertuis CW, McDermott E. 1954. Atlas of Yu W, Shepard GH. 1998. Is beauty in the eye of the men. New York: Harpers. beholder? Nature 396:321–322.

American Journal of Human Biology DOI 10.1002/ajhb