****Please note this post was originally an assignment for my undergraduate class in my senior year. We were asked to explore a contrarian viewpoint on a common health based claim. ****
Is a 0.7 Waist-to-Hip Ratio Actually Attractive?
There is not one single trait that perfectly predicts attractiveness. A number of studies have competing opinions on what trait best predicts attractiveness. A commonly held belief in attractiveness literature is that a 0.7 waist to hip ratio is the most attractive ratio on a woman. This number is cited in mainstream press, fitness content, and beauty content over and over again. How credible are the studies that came up with this number?
The 0.7 waist-to-hip ratio is the most cited number used to explain perfect proportions on a woman. It is argued that the 0.7 W.H.R has been observed throughout human history as the most attractive body proportions on a woman (Singh 2006). It is cited in the greater literature as a morphological feature that acts as a proxy to signal to good health and fertility. What is special about the W.H.R? (a) First, waist and buttocks are uniquely human features, as such may be an important adaptive trait that merits scientific inquiry; (b) both men and women need stored fat to deal with food shortage. Sex hormone encourages regional adiposity, where women store more fat around the hips and need access to this “storage” for survival and preferential access during pregnancy and lactation (Bjorntorp 1987); and (c) women who reach menopause, and are no longer fertile, begin to have W.H.R similar to men. It is a baseline assumption in evolution-based theories that what we find attractive and beautiful in women are cues to her reproductive capabilities (Buss 1987, Kenrick 1989, Symons 1979). A comparative analysis of the literature on W.H.R and B.M.I studies reveal systematic flaws within research design and raise important questions for future studies. This is a post of a research paper I wrote, it errs on the side of LONG. Jump down to conclusion if you just want the juicy bits.
LOW WAIST-TO-HIP RATIO (W.H.R) AND ATTRACTIVENESS
Low W.H.R as an Adapted preference
The literature on attraction and W.H.R argues that a low waist-to-hip ratio [specifically a 0.7 W.H.R] is the most attractive ratio for a woman. This argument is rooted in the logic that men and women select mates who enable them to enhance reproductive success and a low waist-to-hip ratio signals for such success. There is evidence that shows body fat distribution, when measured by W.H.R, correlates with perceived youthfulness, reproductive capabilities, endocrinological status, and long-term health risk in women (Singh 1993). Evolutionary psychologists assert that women select mates based on status, ability to protect, and good genetics [assessed from physical attributes] (Singh 1993). The reproductive value of a woman is concealed within her body [her ovaries and uterus] (Singh 1993). Without any direct cues of her ovulating, her most fertile period, attractiveness is used as a proxy (Singh 1993). Since women’s fertile cues are concealed, men now and historically have attached greater importance to a woman’s looks; this is, in part, adaptive behaviour (Buss 1987, Feingold 1990, and Singh 1993).
In present day however, it would be naive to say that the general fascination with attraction is entirely an evolutionary adaptation. What is considered beautiful is also influenced by culture and media. Singh however, argues that the culturally constructed notions of beauty are secondary to evolutionary preferences (1993). Culturally constructed attributes are important only after her reproductive potential has been assessed based on her morphological features (Singh 1993). W.H.R is one such morphological feature and the most important, according to Singh, that allows men to quickly assess a woman’s fecundity. A low W.H.R is therefore an adapted preference amongst men to filter and choose the most fertile and capable mother of their offspring. It is debatable if crediting this preference entirely to evolutionary behaviour is correct.
A low waist-to-hip ratio has been a preferred attribute historically. For a low W.H.R to hold water as an evolved preference it should have been documented as a celebrated trait historically with marginal change. A low W.H.R, specifically a 0.7 W.H.R, has been the ratio of models in Playboy Centerfolds (Singh 1993) , Miss America (Singh 1993), Miss Hong Kong (Singh 2006), and Miss India (Singh 2006) winners across 30-60 years and seen in historical figurines and statues across many, many decades (Singh 1993, Singh 2006). Although ideal weight for such models and pageants has decreased the 0.7 W.H.R, with a small waist being emphasized has remained the same (Singh 1993, Singh 1994, Singh and Young 1995, Singh 2006). Contrary to popular belief, the preferred silhouette of the “Miss ___” contests and Playboy Centerfolds, have not become androgynous; the W.H.R of the contestants and models is an average of 0.7 (Singh 1993, Singh 2006).
Singh fails to ask certain key questions: when looking at Playboy Centerfolds, an overwhelming amount of heterosexual men find these models attractive, however, would they consider them to be long-term or short-term mates? If low W.H.R is truly the best evolutionary trait to signal fecundity it should also serve to be a proxy to indicate mothering capabilities. Second, Singh throughout his 10-year+ careers in looking at W.H.R ignore runway models, which have increasingly become androgynous, with W.H.R akin to men, and are considered to be some of the most beautiful people. There are socialized beauty-forces at play that Singh consistently fails to account.
College Age Men Prefer Low W.H.R
A majority of the W.H.R studies have been carried out on college-aged men. As such these findings are often only generaliazable to them. Male ratings of female attractiveness are significantly correlated with W.H.R (Singh 1993). A survey was conducted amongst 106 college-aged men asking them to rank 12 line drawings of women from most attractive to least (Singh 1993). They were also asked to rank them on the following attributes: good health, youthful looking, attractive, sexy, desire for children, and capability of having kids (Singh 1993). Findings reveal that W.H.R is related to health and attractiveness; however, body weight is correlated to reproductive capability. Singh concludes that fatness and thinness is not attractive and that normal weight women with 0.7 W.H.R were ranked the most attractive. Youthfulness as an attribute is an anomaly since attractive women are often considered youthful, however youthful women are not always considered attractive (Singh 1993).
To make up for the drawback of line-drawings in 1994 Singh and his research team supplemented the line drawing survey with 4 photographs of a real woman without her face. Her body varied in W.H.R [0.6, 0.7, 0.8, and 0.9] and did not vary in weight class. A sample of 64 college aged men ranked the photo on physical attractiveness, youthfulness, good health, would be a good companion, capability of having children, faithfulness, kind and understanding, intelligence, aggressiveness, and need to lose weight (Singh 1994). In this sample Singh also included a lower W.H.R [0.6] (Singh 1994). In his results the 0.6 photo was ranked as most attractive consistently, the differences in attractiveness ranking were significant (Singh 1994). Interestingly, this sample of men did not assign this figure with a high rating of intelligence, kindness, faithfulness, and understanding (Singh 1994). These results did not line up with Singh’s previous 1993 results showing that a 0.7 W.H.R predicts attractiveness.
In 1995 Singh and Young, conducted another survey, using line drawings again. 230 undergraduate college-aged men participated in a survey and were asked to rank line drawings that varied by W.H.R, weight class, and also breast and hip size. Narrow waist and small hips were rated higher than figures with large waist and large hips (Singh and Young 1995). Figures with large breasts and small hips were also rated higher than other figures (Singh and Young1995). Large breasted figures with low W.H.R are judged as highly attractive, feminine looking, and healthy, and are preferred for both short-and long-term relationships (Singh and Young 1995). These results also mimic the typical “ideal beauty” seen on T.V., in Hollywood movies, and in magazines of the late 20th and 21st century. However, a large breasted, tiny waist-ed, and narrow hipped woman do not represent the type of woman found attractive and fertile in the 15th and 16th century in Germany and Italy, respectively. The most attractive women have been represented in the artistic work of Rubens and Titian. Their artwork display: small breasted, large hipped, and heavy-set women. Although, such an example is anecdotal, Singh’s findings arguably make less of a statement about adapted evolutionary traits and more of statement about learned and socialized attraction.
There are a few methodological concerns in Singh’s studies. I address most of these in the conclusion. Of note however, is the cover story given to participants. The cover story given to participant in the above studies was a bit deceptive. The cover story indicated that physical attributes and personality traits were being studied (Singh 1993, Singh 1994, Singh and Young 1995).
Older Men Prefer Low W.H.R
Singh and other studies did look at samples of older men. In order to make up for the pitfalls of surveying only a small sample of college age men, Singh tested older men to see if these results were common across generations (1993). For W.H.R to be a proxy to reproductive health, the results should stand Trans-generationally. He found that the results were strikingly similar for all age groups, the normal weight 0.7 W.H.R figure was ranked as most attractive amongst a sample of 89 men, of varying S.E.S backgrounds, aged 25-89 (Singh 1993). Again, the sample here is small, and likely not generaliazable to the greater North American population.
Karremans et al also looked at a small sample of older men (2010). They used a sample of 19 blind men, 19 sighted and 19 blindfolded men in the Netherlands. The blind men and blindfolded men were asked to feel the bodies of two mannequins, one with a W.H.R of 0.7 and the other with a W.H.R of 0.84. The blind men [who are blind from birth], blindfolded, and sighted men all rated the woman with the 0.7 W.H.R as more attractive (Karremans et al 2010). However, the strength of the preference is stronger amongst sighted than blind men (Karremans et al 2010). The sample in this study is small, however it suggests that to a certain extent sighted men have and do develop preferences of attraction outside of their evolutionary desires. On the flip side [considering the preference of blind men] it also suggests that a low W.H.R is not entirely conditioned by socialization [visual media, culture etc] evolutionary forces may be at play. Essentially, attraction is explained by multiple interacting factors.
Singh in 2005 surveyed a group of older men; asking them to evaluate pre and post surgical pictures of women who had undergone buttocks augmentation. In such a surgery fat cells from the waist are injected in to the butt (Singh 2005). The W.H.R of these women went from 0.85 down to 0.7 after the surgery. The men who were evaluating the attractiveness of these pictures were not told these were pre and post images. The post surgical photos were rated to be more attractive, and this difference was significant (Singh 2005).
Singh in 2006 sampled a group of 19 male plastic surgeons [they would be categorized as “older men”]. Singh argues that since plastic surgeons help people create ideal bodies, they should be familiar with the idea of an attractive female, which, to repeat, Singh argues is shaped by the forces of natural selection. These surgeons were given a sheet of paper with six line drawings of female figures: underweight, normal, and overweight with 0.7 and 0.9 W.H.R in each weight class. They were asked to alter the body and make it more attractive. If the body was already attractive they were to leave it as is. 3 of 19 surgeons drew lines on the normal weight 0.7 figures. Breast size was increased on all models, fullness was removed from overweight models and fullness was added to underweight models (Singh 2006).
Cross-Cultural 0.7 W.H.R is Preferred
For a trait such as a low waist-to-hip ratio to be considered an evolutionary adaptive trait it should stay consistent across cultures, because, as noted by Sugiyama “universal standards of beauty reflect adaptations of reproductive value only if they take in to account local context” (2004).
Singh took his line drawing test and asked men in Guinea-Bissau [poor and minimum exposure to Western media], Azores Island [government controlled television], Indonesia, and U.S blacks and whites to ranke the figures from most to least attractive (Singh 2006). All groups have very similar ratings, despite cultural differences, a low 0.7 W.H.R is seen as more attractive (Singh 2006). Singh argues that this convergence of perceived attractiveness across these groups “…cannot be attributed to media exposure” (Singh, 364, 2006).
Amongst the people of Shiwiar [an Amazonian community living on the border of Ecuador and Peru], W.H.R is higher relative to women in industrialized societies, no one had a W.H.R lower than 0.8 (Sugiyama 2003). Among this group of people, heavier women were ranked as more attractive, better mothers, sexier, and more fertile (Sugiyama 2004). However, when looking only at W.H.R, low W.H.R women were ranked as most attractive, more healthy, more sexy, better mothers, and better wife’s. The issue with this study is that the sample is small, n=30, as a result the differences between perceived attraction are not significant (Sugiyama 2004).
Hadza, are subsistence based tribal community outside of Tanzania. Hadza men were asked to rank women based on different weight classes W.H.R. In 1997 Hadza men showed preference for only heavier women (Westman and Marlowe 1999). In 1998 they tested the Hadza men again (Marlowe and Westman 2001). Images of women varied by weight class and W.H.R [from 0.4 to 1.0 waist-to-hip ratio]. The 1.0 waist-to-hip ratio was overwhelmingly preferred by Hadza men on characteristics of attractiveness, health, and desirability as a wife. Again the significance of this study comes in to question since the sample size is really small, n=31 [a control of 16 college aged men in the U.S. were used]. (Marlowe and Westman 2001). In another cross-cultural study with 308 male and female subjects from Greece, Uganda, there was an over all preference for 0.7 W.H.R [when ranking 8 line drawings that varied in weight class and W.H.R from 0.4-1.0] (Furnham et al 2002).. Notably, Ugandan’s significantly preferred the 0.5 W.H.R in the heavy weight category (Furnham et al 2002). Although this study has a larger sample size relative to other cross cultural study, their selection of countries and the definition of “Cross Cultural” seems arbitrary and was not explained well in the study.
BODY MASS INDEX (B.M.I) AND ATTRACTIVENESS
“If the utilitarian interpretation of the W.H.R hypothesis is correct, a W.H.R of 0.7 should be ranked as most attractive and fecund relative to larger and smaller W.H.Rs, regardless of waist or hip size. If the aesthetic-preference interpretation of the W.H.R hypothesis is correct, then a W.H.R of 0.5 should be ranked as the most attractive and more fecund relative to larger W.H.Rs, regardless of weight” (Tassinary and Hansen, 1998). As we will see the following review of disprove the importance of W.H.R as an adapted evolutionary preference, and instead suggest B.M.I is a better predictor for attractiveness.
B.M.I is a Significant Predictor of Attractiveness
In 1998 Tassinary and Hansen attempted to replicate Singh’s findings using the 12 line drawings from Singh’s 1993 study in addition to 27 other line drawings of the female figure that varied by weight, waist size, and hip size. 83 college-aged women and 53 college-aged men were asked to rank the drawings with respect to attractiveness. The data show little evidence in support for Singh’s conclusions. Tassinary and Hansen’ study found that relative hip size and weight are better predictors of attractiveness regardless of W.H.R. Essentially they find that attractiveness and fertility can be related or unrelated, either positively or negatively to W.H.R depending on the waist and hip size, and the weight of a woman (Tassinary and Hansen 1998). As such they are not in agreement with the evolution-based argument that a woman’s attractiveness is necessarily a sign of mate value. There are some issues with Tassinary and Hansen study, namely the use of poor quality line drawings, and a small sample size of undergraduate students is barely representative.
Cornelissena et al, supports Tassinary and Hansen’ findings (2007). Cornelissena et al looked at the eye movement of men and women when examining photographed female bodies. When respondents were asked to rate attractiveness and judge relative body fat the eyes focused on similar parts of the body. Their eyes focused on the central and upper abdomen and chest – but not the pelvic or hip areas. When asked to judge W.H.R the eyes focused on the pelvic and hip regions. This suggests that the patterns to judge attractiveness and body fat are similar and that the pattern to view W.H.R were significantly different than the above two (Cornelissena et al 2007). This finding is consistent with the finding that W.H.R had little influence over attractiveness judgments. Again, this study should be met with some skepticism, since n=60 is a small sample the and mean age of participants is 21.
In studies where heavier bodies were made curvier– representing W.H.R as low as 0.5, B.M.I came out as a better predictor of attractiveness (Tovee et al 2002). Using 360 degree colour video clips of 43 white woman also revealed that attractiveness to both male and female respondents depend on body fat percentage and tanned skin, and not other attributes such as W.H.R and cardiovascular fitness [which was also measured in this study] (Smith et al 2007). Again the sample here is small and samples a younger group of people.
For a trait such as waist-to-hip ratio to be considered an evolutionary adaptive trait it should stay consistent across cultures. A few studies conducted cross-cultural analysis to find that W.H.R was not an important predictor in attractiveness, instead B.M.I was. Swami and Tovee, point out the there is an inherent weakness in using line drawings since they are unable to properly convey to the respondent a woman’s attractiveness (2007). Their study used picture profiles of 50 real women who ranged in 5 B.M.I categories and W.H.R of 0.68 -0.98. Malaysian men in low S.E.S, high S.E.S, and men from Great Britain evaluated the images in a similar way (Swami and Tovee 2007). Overwhelming B.M.I came out as a better predictor of attractiveness than W.H.R Swami and Tovee concludes the W.H.R is a weak predictor of female attractiveness (Swami and Tovee 2007). The sample size in this study is small; as such the findings should be read with some skepticism.
CONCERNS AND CONCLUSION
Concerns for the W.H.R Studies
(1) Causal direction: most of the studies did not take in to account causal direction. That is, a particular body could be attractive given societal norms, and the person may rank a body as more “capable of having children” because they find this body attractive. (2) Socialized factors: Many of the Singh studies make a note of factors that could be attributed to socialization and not evolutionary adaptation. For example, (a) models have gotten skinnier over the past three decades, and lower weight categories of the line drawings are seen as attractive and healthier. It is valid to infer that socialization plays some part in this assessment; (b) large breasted women are seen as attractive, this could very likely be due to socialization.
(3) Absolute conclusions: the W.H.R studies, and in particular the Singh studies, made absolute statements attributing attraction to an inherent evolutionary driving force. At times discrediting the role of visual media. This is risky given the lack of statistical rigor in their studies. A more cautious and less brash approach to their conclusions is merited. (4) Sample: Although this has been pointed out ad nausea, the sample size in most of the W.H.R studies is very small and as such the significance of the results come in to question. Furthermore, college-aged students participated, on a compulsory or voluntary basis through an undergraduate course. There is obvious sample selection bias here that should be addressed. (5) Line drawings: line drawings were used consistently in the studies that found significant relationships between low W.H.R and attractiveness. The line drawings are poor quality. The changes in W.H.R and weight class are often too subtle and errors in ranking are possible. (6) Sexuality: none of the W.H.R studies inquired about the sexuality of their participants. This seems to be a major error as perceived attraction could change dramatically with a person’s sexuality. (7) Tinkering with W.H.R say for a moment a 0.7 W.H.R was the most important trait that predicted attraction. As such, it would be possible to change men’s evaluation of a woman’s attractiveness by only manipulating her W.H.R. This was done by Tassinary and Hansen (1998) and the figure was not perceived to be attractive. [It is worth mentioning that some women would look absurd with a 0.7 W.H.R ]
Concerns for the B.M.I Studies
There are similar concerns for the B.M.I study. In general however, the researchers with the B.M.I. studies were less likely to make inflammatory conclusions and advised their readers to take caution with their findings. Line drawings were also used less frequently, and female respondents were included in most studies. (1) Socialized factors: the B.M.I studies rarely offered “socialized attraction” as an alternative explanation to their findings. (2) Sample: similar to the W.H.R studies the sample size of the B.M.I studies was consistently low and often used a sample of college age students. (3) Sexuality: none of the W.H.R studies inquired about the sexuality of their participants. This seems to be a major error as perceived attraction could change dramatically with a person’s sexuality.
Concerns for Both W.H.R and B.M.I Studies
Apart from the semantic issues within each study, the biggest flaw across the W.H.R, B.M.I and attractiveness studies is the isolated manner in which they assess attractiveness. There are many attributes and characteristics that contribute to a person’s attractiveness. The physical-body is a major one, in addition to: facial features [symmetry], poise, confidence, articulateness, intelligence, charisma, the ability to make someone feel good, the person’s relative status, and their clothes. This is not an exhaustive list; however this list does point out additional attributes that contribute to someone’s attractiveness and their potential as a mate.
The low W.H.R and attractiveness studies are interesting and should be reexamined using larger more representative samples. The review of the studies show that there is little agreement and considerable noise as to what trait predict attractiveness. Perhaps, one explanation is that there is no “one trait,” rather a confluence of traits and behaviours [some which we have been socialized to find attractive] that predict beauty. The absolute claims made in some of the low W.H.R studies and the general riskiness of such studies are the implications on beauty standards. When science and media begin to set rigid beauty standards [and when on body ignore the impact of media] the youngest women in our societies are influenced. Of course such studies should continue to be published, however the language and tone of their findings should reflect the statistical power of the study.