Role of Body Fat and Body Shape on Judgment of Female Health and Attractiveness: An Evolutionary Perspective

The main aim of this paper is to present an evolutionary perspective for why women’s attractiveness is assigned a great importance in practically all human societies. We present the data that the woman’s body shape, or hourglass figure as defined by the size of waist-to-hip-ratio (WHR), reliably conveys information about a woman’s age, fertility, and health and that systematic variation in women’s WHR invokes systematic changes in attractiveness judgment by participants both in Western and non-Western societies. We also present evidence that attractiveness judgments based on the size of WHR are not artifact of body weight reduction. Then we present cross-cultural and historical data which attest to the universal appeal of WHR. We conclude that the current trend of describing attractiveness solely on the basis of body weight presents an incomplete, and perhaps inaccurate, picture of women’s attractiveness. “... the buttocks are full but her waist is narrow ... the one for who[m] the sun shines ...” (From the tomb of Nefertari, the favorite wife of Ramses II, second millennium B.C.E.) “... By her magic powers she assumed the form of a beautiful woman ... her hips and breast were full, the waist slender.” (Bhagavata Purana, second or third century C.E.) The iconic representation of a beautiful and sensuous woman as an hourglass figure defies prevalent belief among laypersons and scholars that beauty is arbitrary, PSYCHOLOGICAL TOPICS 15 (2006), 2, 331-350 332 ever changing in the eye of the beholder. If so, why does the label of “hourglass”, hand gestures to depict the shape of women, and even the numbers (36-24-36) often used in vernacular speech to describe a woman effortlessly invoke the image of a youthful, attractive and enticing woman? Furthermore, such impact is not evident of what is in vogue that thin is beautiful; simply describing a woman’s height and weight does not necessarily conjure up the image that she is also young and beautiful. While slenderness is often prized, it is practically impossible to imagine that most people would judge a woman with 32-32-32 measurements as attractive regardless of how skinny she is. The hourglass figure remains critical for judgments of youthfulness and beauty even in skinny models. So what explains the universal and enduring appeal of the hourglass figure? One explanation based on evolutionary psychological theory is that female beauty as represented by the hourglass figure taps into important biological information about various factors regulating women’s reproductive potential and fertility. In this paper we will describe basic tenets of evolutionary psychology pertaining to the nature and significance of female attractiveness. Then we will summarize experimental and clinical evidence demonstrating a link between the hourglass figure and hormonal and endocrinological mechanisms regulating reproductive potential, fertility and risk for major diseases. Such evidence is crucial to support the claim that attractiveness is a reliable cue to a female’s reproductive capability and good health. Next, we will present evidence that changes in the hourglass figure alone systematically affect female attractiveness judgments of lay and professional men and women not only in our society but in various and diverse societies. Finally, we will present evidence showing that equating beauty with the hourglass figure is not a novel or recent phenomenon shaped by the mass media; the allure of an hourglass figure is evident across generations and in diverse ancient cultures. Importance of Female Attractiveness in Mate Selection The fundamental assumption of Evolutionary Psychology is that natural selection not only shaped specialized bodily organs to solve the problems of efficient survival, it also shaped specialized mechanisms to solve problems encountered by humans in ancestral environments. Thus, Evolutionary Psychology argues that the human mind is a collection of special-purpose mechanisms designed by natural selection to solve the problems of survival and reproduction that were recurrently faced by our ancestors. One of the adaptive problems that our male ancestors regularly faced was to assess a female’s mate value, or the degree to which she would enhance his reproductive success. Potential mates necessarily varied in mate value, just like potential foods vary in their nutritional value and shelter and housing vary in their potential utility value. Female mate value was determined by numerous variables such as hormonal profile, reproductive age, fecundity, parity and resistance to PSYCHOLOGICAL TOPICS 15 (2006), 2, 331-350 333 diseases, none of which could be directly observed. It has been proposed that information about some of these variables is reliably conveyed by specific characteristics of female bodies and that natural selection therefore produced psychological mechanisms in men to attend to bodily features in assessing a female’s mate value (Symons 1979). It has been further proposed that as females vary in their mate value, the intensity of male sexual attraction was designed to vary directly with perceived cues of female mate value (Symons 1995). Although people are not consciously aware to such a link, therein lies the power of physical attractiveness. To demonstrate that the female hourglass figure is a reliable cue to her mate value, it is essential to establish that this figure has a plausible link to physiological mechanisms regulating reproductive capability and good health. Furthermore, variation in the hourglass figure should not only be correlated with variation in reproductive potential, but such variations should systematically affect the judged degree of female attractiveness. The nature of body fat distribution, which largely determines the hourglass shape, meets most of the above stated criteria. Waist-to-Hip Ratio (WHR): An Indicator of Women’s Age, Fecundity and Health Status Fat distribution in humans depends both on age and their sex; the sexes are similar in infancy, early childhood, and old age, but differences in fat distribution are greater from the early teens until late middle age (Vague 1956). The fat distribution in humans is regulated by sex hormones, and fat can be used from a region of the body at the same time as it is being accumulated at another (Pond 1981). Extensive evidence by Bjorntorp (1987, 1991) and by Rubuffe-Scrive (1987, 1988) have demonstrated the ways sex hormones affect specific regional adiposity and regulate utilization and accumulation of fat. Simply stated, estrogen inhibits fat deposition in the abdominal region and stimulates fat deposition in the gluteofemoral region more than in other body regions. Testosterone, in contrast, stimulates fat deposition in the abdominal region and inhibits deposition in the gluteofemoral region. It is this sexually dimorphic body fat distribution that primarily sculpts the typical hourglass figure in women after pubertal onset; women have greater amounts of body fat in the lower part of the body (gynoid aka “pear-shaped” body fat), whereas men have greater amounts of fat in the upper body (android aka “apple-shaped”body fat). Figure 1 shows the sex difference in body shape in men and women as a function of age but independent of body weight. It should be obvious that in spite of the lack of typical sex differences associated with the two sexes (i.e., breasts, long hair, etc.) one can readily identify the sex of the schematic body figure after puberty (Figure 1). PSYCHOLOGICAL TOPICS 15 (2006), 2, 331-350 334 Figure 1. Schematic Drawings of Sexually Dimorphic Body Shape From Age 1-9 Years (Panel A) and from Age 10-18 (Panel B) A widely-used anthropometric technique to ascertain the degree of gynoid and android fat distribution is to measure circumference of the waist (narrowest portion between the ribs and iliac crest) and hips (at the level of the greatest protrusion of the buttocks), and using these measurements to compute a waist-to-hip ratio (WHR). WHR is a stable and highly reliable measure and is significantly correlated with fat distribution measures using computed tomography scanning (Despres, Prudhomme, Pouliot, Temblay & Buchard 1991). Before puberty, both sexes have similar WHRs. After puberty, females deposit more fat in the hips and buttocks; WHR therefore becomes significantly lower in females than in males. WHR has a bimodal distribution with relatively little overlap between the sexes. The typical range of PSYCHOLOGICAL TOPICS 15 (2006), 2, 331-350 335 WHR for healthy pre-menopausal women has been shown to be .67-.80, whereas healthy men have WHRs in the range of .85-.95 (National Academy of Sciences, 1991). Women typically maintain a lower WHR than men throughout adulthood, although after menopause their WHR approaches the masculine range (Kirschner & Samojlik 1991). Thus, the size of WHR, unlike overall body weight, can be used as a reliable proxy of women’s general reproductive status (preor post-pubertal and menopause) and youthfulness. There is no other observable body feature that tracks women’s age. WHR is also a reliable indicator of reproductive capability of pre-menopausal women. Compared to women with high WHR, women with a low WHR have fewer irregular menstrual cycles (Van Hooff, et al., 2000), optimal sex hormone profiles (Jasienska, et al., 2004), ovulate more frequently (Moran, et al., 1999), and have lower endocervical pH, which favors sperm penetration (Jenkins, Brooke, Sargeant & Cooke 1995). Low WHR is also an independent predictor of pregnancy in women attending an artificial insemination clinic (Zaadstra, et al., 1993) and in women attempting in-vitro embryo fertilization transfer (Waas, Waldenstrom, Rossner & Hellberg, 1997). WHR size can be used to roughly estimate whether a woman is in the early stages of pregnancy induced by another male; investing resources in a child fathered by another male does not enhance a male’s reproductive success. The size of WHR increases even in the early stages of pregnancy. Finally, women with low WHR have lower risk of heart diseases, stroke, type II diabetes, gallbladder disease, kidney diseases, various cancers (breast, endometrial, ovarian), and premature death. (for a review see Singh, 1993a, 2006; Kissebah & Karkower, 1995). Women with low WHR also cope with stress (as measured by cortisol release) more efficiently (Epel, et al., 2000) and report fewer episodes of depression than women with higher WHR (Nelson, Palmer, Pedersen & Miles, 1999). Women less susceptible to health problems would likely have more energy to attend to their family and children, and because many health problems are heritable, their offspring will receive the genetic gift of good health. In summary, WHR size provides reliable information about the reproductive age, fertility and health status of a woman at a glance. Given that WHR is a signal of health and reproductive capability, do changes in the size of WHR cause changes in the judgment of female attractiveness? WHR and Attractiveness Judgments To investigate this issue Singh (1993a) developed 12 line drawings of female figures representing three body weight categories (under-, normal and overweight) and two levels of feminine (0.7 and 0.8) and two levels of masculine (0.9 and 1.0) WHR within each weight category (Figure 2). In the initial series of research (Singh 1993a, 1993b), judgments of attractiveness, healthiness and youthfulness were obtained for these figures from men and women of diverse ages (18-85 years old), PSYCHOLOGICAL TOPICS 15 (2006), 2, 331-350 336 professions (white collar workers, lawyers, physicians), educations (undergraduates and post-graduate degree) and ethnicities (African-American, Mexican American, Euro-American) backgrounds. The findings indicated that variations in the WHR in female figures produces systematic changes in the judgment of attractiveness. Participants rated figures with a lower, more feminine, WHR as more attractive than figures with a higher, more masculine, WHR in each of the weight categories. These findings, using line drawings developed by Singh (1993a), now have been replicated (for exception see Yu & Shepard, 1998; Marlow & Wetsman, 2001) with AfricanAmericans (Markey, et al., 2002), in Great Britain (Furnham, Tan & McManas, 1997), Germany (Henss, 1995), Australia (Connolly, Slaughter & Mealey, 2004), Indonesia (Singh & Luis, 1995), the Azore Island, the African country of Guinea Bissau (Singh, 2004), and in the Shiwiar Tribes of East Ecuador (Sugiyama, 2004). Figure 2. Female Body Figures Differing in Body Weight (U = Underweight, N = Normal Weight and O = Overweight) and Four Levels of WHR (0.7, 0.8, 0.9, and 1.0) In the original studies, the relationship between attractiveness and WHR depended on body weight category; the normal weight figure with 0.7 WHR was judged most attractive, followed by the underweight figure, whereas the overweight figure was judged not to be attractive (Singh, 1993a, b). The impact of WHR on PSYCHOLOGICAL TOPICS 15 (2006), 2, 331-350 337 attractiveness judgments is obscured by body weight deviation from the average weight, regardless of whether the weight is extremely low (underweight) or high (overweight). Therefore, Singh (1993b) concluded that, “...neither body weight, nor WHR alone can explain attractiveness. To be attractive, women must have a low WHR and deviate little from normal weight” (pp. 310-311). Bodyweight, WHR and Attractiveness Some researchers have ignored this complex relationship between body mass index (weight/height squares = BMI) and have opted to treat as if WHR and BMI are independent variables and have reported that body weight accounts for more variance than WHR (Tassinary & Hansen, 1998; Tovee, Reinhardt, Emery & Cornelissen, 1998). Tovee and Cornelissen (1999) have even suggested that an attractiveness judgment based on WHR is an artifact of BMI. They argue that reduction of the size of WHR, for example, form 0.8 to 0.7, reduces BMI regardless of body weight category and that this reduction in BMI is responsible for attractiveness ratings for figures with low WHR. These arguments are, however, based on misunderstandings about the nature of WHR. WHR describes the nature of body fat distribution and therefore the effect of WHR, on attractiveness judgments, independent of body weight, is meaningless. The statement that women suffering from anorexia have low WHR but are not fertile is based on failure to take into account the effect of body weight (Tovee, et al., 1997). An extreme example of this kind of reasoning would be to ask why human skeletons are not fertile even though they have low WHR. The degree of overall obesity, as measured by BMI, does affect the size of WHR especially in instances of very low and high BMI. Frequently, the effect of WHR size, for example, on risk of diseases is found to be significant for normal weight range; BMI representing overweight and obesity overshadows the impact of WHR (Han, Morrison & Lean 1999). It is, therefore, critical to take into account total body fat, or BMI, a feature which people in ancestral environments would have encountered when examining the effects of WHR (the nature of body fat distribution) on attractiveness. Given the cycles of famine, feast, and workload, obesity would have been rare. There would have been some emaciated people suffering from disease or malnutrition and they would not likely have been desirable mates. The selection process would have stabilized the preference for population-typical average body weight. Finally, normal body weight alone, without taking into account the nature of the distribution of body fat (WHR), can neither distinguish male from female nor convey any reliable information about the age of a woman. Additionally, due to seasonal availability of food supply, body weight would have frequently fluctuated in ancestral environments. WHR, unlike total body weight, does not significantly change if weight loss or gain is not greater than eight kg (Kissebah, Evans, Peiris & PSYCHOLOGICAL TOPICS 15 (2006), 2, 331-350 338 Wilson, 1985). WHR is, therefore, reliable signal of women’s health and fertility, in spite of short-term body weight fluctuation. However, as WHR and BMI are correlated, one needs to assess their relative contributions to attractiveness ratings. Some researchers have used post-hoc “variance accounted” for methods to assess the relative contribution of WHR and BMI (Tovee, Reinhardt, Emery & Cornelissen, 1998). However, these methods are overly sensitive to the range of the independent variable in the stimulus set. The optimal solution is to use stimulus for which BMI and WHR are independent. A recent surgical technique provides a means to examine the role of different sizes of WHR on attractiveness, independent of BMI. The surgical procedure includes liposuction of the circumference of the waist and then using purified fat cells to graft to the buttocks (Roberts, Weinfeld & Nguyen, 2005). This procedure both narrows the waist and enhances the buttocks, having a synergetic effect on WHR without altering BMI. In his practice, aesthetic plastic surgeon Roberts measures preand post-operative waist and buttocks circumferences and records preand postoperative BMI. Post-operative measurements and photographs are taken about two to three months after surgery to allow for healing and scar disappearance. Although all patients have a lower post-operative WHR, some patients gain weight and the other patients lose weight post-operatively. These patients provide a unique opportunity to examine the independent effects of WHR and body weight, or BMI on attractiveness. Roberts, working with more than 200 patients, notes that, “It has been my experience...that body weight has a negligible effect on how attractive the body looks post-surgically or the judgment of attractiveness of their own by the patient. Rather, the critical variable in attractiveness is the proportionality of the shape and the size of the waist and buttocks... A review of our data shows the best results have a waist-to-hip ratio of approximately 0.7 (their pre-operative WHR averaged 0.85)” (personal communication, April, 28, 2005). The lack of impact on BMI in normal weight women is strikingly obvious when preand post-operative photographs of a patient are observed (Figure 3). In spite of an increase in postoperative BMI, the post-operative photograph is judged to be more attractive than low BMI pre-operative photograph (Figure 3). PSYCHOLOGICAL TOPICS 15 (2006), 2, 331-350 339 Figure 3. Sample of a Representative Preand Post-operative Photographs The pre-operative photograph (left) has lower BMI than post-operative photograph (right) and has a higher BMI and low WHR. Post-operative photograph is rated more attractive than pre-operative photograph, in spite of higher BMI. Temporal Stability of WHR Appeal It is a commonly held belief that what makes a woman beautiful changes over time in a given society. The most commonly cited example is body weight; people point out that 50 years or so ago, plump women were judged to be beautiful whereas as at present being thin is a prerequisite for female attractiveness. Indeed, current popular fashion models and Miss America winners are significantly more slender than famous sex symbols of yesteryear, such as Marilyn Monroe and Jayne Mansfield. While the body has shrunk, what about body shape? In an oft-quoted study of Playboy centerfolds and Miss America winners, Garner, Garfinkel, Schwartz, and Thompson (1980) concluded there is a significant trend for thinness and that the ideal female body form is evolving away from the hourglass shape and becoming tubular. Morris, Cooper and Cooper (1989) also found that British fashion models over the period of 1967-1987 exhibited “a trend toward a more tubular or androgynous body shape.” If true, fashion models should also show a gradual increase towards a WHR size of 0.80 or in the masculine range. In a previous study, Singh (1993a) examined the body weights and WHRs of Miss America winners PSYCHOLOGICAL TOPICS 15 (2006), 2, 331-350 340 from 1923 to 1987 and of Playboy centerfolds from 1955 to 1990. Indeed, there was a significant trend in reduction of body weight, but WHRs of both Miss America winners and Playboy centerfolds remained in the 0.68-0.72 range. A similar trend is evident in BMI and WHR of Playboy centerfolds and Miss Hong Kong winners from 1987 to 2004 (Singh, 2006). As evident in Figure 4, the BMIs of Playboy centerfolds have fluctuated between 18 and 19.25, and Miss Hong Kong winners have lower BMIs (17.25-17.75) than Playboy centerfolds while maintaining their WHRs in the range of 0.65-0.71. Figure 4. Body Mass Index (BMI) and WHR of Playboy Centerfold and Miss Hong Kong from 1987-2004