Do women with lower waist-to-hip ratios have higher intelligence?

Lassek and Gaulin analyzed data from the Third National Health and Nutrition Examination Survey (NHANES III), which was conducted by the US National Center for Health Statistics from 1988 to 1994.  They reported that women with lower waist-to-hip ratios (WHR) had higher intelligence and smarter children.^{(1, pdf)}  WHR decreases with increasing femininity and also lesser body fat as one goes from obese to normal.  So have the authors shown a relationship in women between intelligence and femininity (assessed by distribution of body fat) or between intelligence and degree of body fat or between intelligence and both amount plus distribution of body fat?  Let us see.

Some background on obesity and intelligence is relevant.

Obesity and intelligence

Some of the best data on this topic come from Scandinavia and the Netherlands.  In these regions, there is a strong socioeconomic status (SES) gradient in obesity, with obesity being more common in lower SES groups in both men and women.  This relationship between SES and obesity is observed in white women in all white populations.  What is responsible for this relationship?  Does lower SES increase the likelihood of obesity or does obesity increase the likelihood of lower SES or does a third factor simultaneously increase the likelihood of lower SES and excess body fat?  This question was resolved by a Danish adoption study where children had been adopted at an early age and grew up without contact with their biological parents.  As adults, the body weight of the adopted individuals matched the body weight of their biological parents, not that of the adopted parents.⁽²⁾  This shouldn’t be surprising since body mass is strongly influenced by genes.  The more interesting find was that the adult SES of the adopted individuals also matched the SES of their biological parents, not that of the adopted parents.⁽³⁾  SES is obviously not biologically transmitted, but something related to SES is biologically transmitted.  A path analysis revealed that this variable was intelligence and that lower intelligence increased the likelihood of lower SES as well as greater obesity.⁽⁴⁾  There is also plenty of evidence that the SES gradient observed in Scandinavia and the Netherlands is largely not accounted for in terms of education and lifestyle variables such as smoking and exercising (see data here; scroll down to the section on obesity).  In other words, the association between low SES and increased likelihood of obesity in whites is not merely accounted for by discrimination against the obese.

There are other data sets consistent with lower intelligence in obese individuals,^(5-10) but few compare to the Nordic data in clarifying the nature of the relationship between obesity and lower SES/lower intelligence.

Obese women undoubtedly offer a more hostile womb to a fetus, especially those with a tendency to pack on excess fat in the abdominal region.^(11-13) Physiological shifts in pregnant women make them lean toward a diabetic condition, and obese women are more likely to develop gestational diabetes.⁽¹⁴⁾  It has been shown that children born to women with diabetes have lower intelligence.^{(15, 16)}  Therefore, obese women are likely to give birth to children with lower intelligence both as a result of their own genetic make-up (seen from the Nordic data) and the more hostile womb they offer to the fetus.

The NHANES III data

In Lassek and Gaulin’s study sample, the ethnic composition was: 38% white, 29% African-American, 28% Hispanic and 5% other (16,325 females, aged 0–90 years; mean age, 29.9±25.8 years).  This is a problem because the dataset shows basically no relationship between SES and prevalence of obesity in African-American women and a weak relation in this regard among Hispanic women, and regardless of SES, these two groups have a high prevalence of overweight/obesity, especially in contrast to Scandinavian/Dutch white women.

The authors used the following measures of mental ability:

• The math and reading tests from the Wide Range Achievement Test—Revised, and the Digit Span and Block Design tests from the Wechsler Intelligence Scale for Children—Revised.; the Four Tests measure (sample ages 6-16 years)
• Years of education (sample ages 18-49).
• Serial Digit Learning Test and Serial Digit Substitution Test; the Two Tests measure  (sample aged 18-49)

Years of education is a problematic measure of cognitive ability.  Controlling for grade point average (GPA), a 4-year degree in chemistry requires greater intelligence than a 4-year degree in recreation studies.  Additionally, affirmative action weakens the relationship between cognitive ability and education in African-American and Hispanic women.

In the table below, it can be seen that ethnicity is the strongest predictor of intelligence when tests more closely assessing it are used, but a weaker predictor of intelligence when years of education is used.  Note that WHR is weakly related to cognitive ability.

Controlling for mother’s age, both parents’ education, family income and ethnicity, WHR in women but not body mass index (BMI; a proxy for percentage body fat) was negatively related to their intelligence as well as that of their offspring; offspring intelligence is higher even if one adjusts for mother’s IQ.  In a subset of the women that had no children, WHR explained 23% of the variance in total body fat and 28% of the variance in BMI, whereas BMI explained 89% of the variance in estimated body fat.  So the author’s results indicate that the distribution of fat but not the amount of fat is related to intelligence in women and their offspring.  This is a curious find, not consistent with the literature on obesity and intelligence that I have cited, and these citations are lacking in the paper.  Hence, people should wait for this study to be replicated using a better sample (Scandinavian/Dutch white; useful because of a low prevalence of obesity) before drawing any conclusions.  I am positive that better datasets will show that lower WHR is related to higher IQs to the extent that lower WHRs are corresponding to a normally feminine deposition of fat, i.e., minimal in the abdominal region and more so in the hips/upper thighs (gluetofemoral region).  If it can be shown that even in normal weight women without excess abdominal fat, a lower WHR (more feminine in this case) corresponds to higher intelligence, then this would be impressive, but I doubt that this will be shown for reasons that I will describe next.   

My anecdotal observations

My observations have been that the smartest women tend to be slim and somewhat masculinized.  To compare my observations with women known to be/have been very intelligent, I decided to look at pictures of women who have won the highest awards in math (the Field Medal) and the physical sciences (Nobel Prize) in recent years.  However, no woman has ever won the Field Medal.  Also, no woman has won a Nobel Prize in the physical sciences (physics, chemistry) in recent years (last one was in 1964).  Awards in recent years are important because earlier breakthroughs required lower intelligence than what is currently state of the art.  So I decided to include all female Nobel laureates in the physical sciences, but there were only 4 women with 5 Nobels awarded, and three of them were received by a mother-daughter pair.  So I decided to include all female Nobel Laureates in Physiology and Medicine, a field where given some luck and hardwork, someone with an IQ as low as the high 120s could come up with Nobel-worthy work, but what else I can do?  Ideally, all pictures should have featured the women as young adults, but I wasn’t successful in obtaining them in some cases.

Nobel laureates in science; from top to bottom: Marie Sklodowska Curie (1903, Physics; 1911, chemistry), Irene Joliot-Curie (1935, chemistry), Maria Goeppert Mayer (1963, physics), Dorothy Crowfoot Hodgkin (1964, chemistry), Gerty Radnitz Cori (1947), Rosalyn Sussman Yalow (1977), Barbara McClintock (1983), Rita Levi-Montalcini (1986), Gertrude Elion (1988), Christiane Nusslein-Volhard (1995), Linda B. Buck (2004).  Unless specified otherwise, the awards are for Physiology and Medicine.

On average, the Nobel laureates appear to be on the somewhat masculine side of the mean; none happened to be overweight/obese as young adults, and the great majority did not become obese as older women either.  Using their examples is merely anecdotal and does not allows us to draw any reliable conclusions, but it should be noted that slim women will usually have WHRs below the average for the female population if the prevalence of overweight and obesity is high, as is true for many populations today.  The average WHR of the top-50 high-fashion models as of March 30, 2007 was reported as 0.7, which is below average when compared to the general young adult female population, but one look at these models and it is clear that they have above average masculinization.  So, if one wanted to address the relation between femininity and overall intelligence in women, one had better use a more comprehensive measure of femininity/fat distribution than WHR alone.      

References

1. Lassek, W. D., and Gaulin, S. J. C., Waist-hip ratio and cognitive ability: is gluteofemoral fat a privileged store of neurodevelopmental resources?, Evol Hum Behav, Epub (ahead of print); dx.doi.org/10.1016/j.evolhumbehav.2007.07.005 (2007).
2. Sorensen, T. I., and Stunkard, A. J., Does obesity run in families because of genes? An adoption study using silhouettes as a measure of obesity, Acta Psychiatr Scand Suppl, 370, 67 (1993).
3. Teasdale, T. W., and Sorensen, T. I., Educational attainment and social class in adoptees: genetic and environmental contributions, J Biosoc Sci, 15, 509 (1983).
4. Stunkard, A. J., Socioeconomic status and obesity, Ciba Found Symp, 201, 174 (1996).
5. Zhang, H., and Li, Y., [Harmfulness of obesity in children to their health], Zhonghua Yu Fang Yi Xue Za Zhi, 30, 77 (1996).
6. Li, X., A study of intelligence and personality in children with simple obesity, Int J Obes Relat Metab Disord, 19, 355 (1995).
7. Campos, A. L., Sigulem, D. M., Moraes, D. E., Escrivao, A. M., and Fisberg, M., [Intelligent quotient of obese children and adolescents by the Weschler scale], Rev Saude Publica, 30, 85 (1996).
8. Cserjesi, R., Molnar, D., Luminet, O., and Lenard, L., Is there any relationship between obesity and mental flexibility in children?, Appetite, 49, 675 (2007).
9. Karnehed, N., Rasmussen, F., Hemmingsson, T., and Tynelius, P., Obesity and attained education: cohort study of more than 700,000 Swedish men, Obesity (Silver Spring), 14, 1421 (2006).
10. Chandola, T., Deary, I. J., Blane, D., and Batty, G. D., Childhood IQ in relation to obesity and weight gain in adult life: the National Child Development (1958) Study, Int J Obes (Lond), 30, 1422 (2006).
11. Prentice, A., and Goldberg, G., Maternal obesity increases congenital malformations, Nutr Rev, 54, 146 (1996).
12. Cnattingius, S., Bergstrom, R., Lipworth, L., and Kramer, M. S., Prepregnancy weight and the risk of adverse pregnancy outcomes, N Engl J Med, 338, 147 (1998).
13. Kieser, J. A., Groeneveld, H. T., and Da Silva, P. C., Dental asymmetry, maternal obesity, and smoking, Am J Phys Anthropol, 102, 133 (1997).
14. Johnson, S. R., Kolberg, B. H., Varner, M. W., and Railsback, L. D., Maternal obesity and pregnancy, Surg Gynecol Obstet, 164, 431 (1987).
15. Silverman, B. L., Rizzo, T., Green, O. C., Cho, N. H., Winter, R. J., Ogata, E. S., Richards, G. E., and Metzger, B. E., Long-term prospective evaluation of offspring of diabetic mothers, Diabetes, 40 Suppl 2, 121 (1991).
16. Yamashita, Y., Kawano, Y., Kuriya, N., Murakami, Y., Matsuishi, T., Yoshimatsu, K., and Kato, H., Intellectual development of offspring of diabetic mothers, Acta Paediatr, 85, 1192 (1996).