Ethnic Androgen Differentials I

Testosterone Molecule

Masculinizing: Total Testosterone (T), Circulating Free Testosterone (cFT)

Feminizing: Estrone (E1) & Estradiol (E2)

The race realist assertion that (East) Asians have less testosterone (T) than Europeans who have less than Africans is mistaken on multiple levels. Here is Rushton (2000b):

The races differ in their level of sex hormones. Hormone levels are highest in Blacks and the lowest in Orientals. This may tell us why Black women have premenstrual syndrome (PMS) the most and Orientals the least. The races also differ in testosterone level which helps to explain men’s behavior. In one study of college students, testosterone levels were 10 to 20% higher in Blacks than in Whites. For an older sample of U.S. military veterans, Blacks had levels 3% higher than Whites (see the 1992 issue of Steroids). In a study of university students, Black Americans had 10 to 15% higher levels than White Americans. The Japanese (in Japan) had even lower levels.

Rushton (2000a) states it thus:

One study of matched groups of 50 black and 50 white male college students in California found that testosterone levels were 19 percent higher in blacks than in whites (Ross, Bernstein, Judd, Hanisch, Pike, & Henderson, 1986). A 3 percent difference favoring blacks has been found among an older group of 3, 654 white and 525 black male Vietnam era military veterans (Ellis & Nyborg, 1992)…… Measurements of two metabolites if dihydrostesterone shows marked lower levels in the serum of Japanese natives and 10 to 15 percent higher concentrations in American blacks (Hixson, 1992).

Here Rushton (2000a, 2000b) uses three studies; the first is Ross et al. (1986) which found that aged matched African American (AA) college students had 19% more testosterone than European American (EA) college students. His first study thus has the following problems:

  • African Americans are a mixed sub-clade (with European, American, East Asian) of the African clade and most likely does not represent unmixed Africans;
  • European Americans are a sub-clade of the European clade and might not necessarily represent unmixed Europeans as some 20-30% have African ancestry [Stuckert (1958)] with additional others having American and East Asian ancestry;
  • Small sample size (n = 100, 50 AA); and
  • Sample bias: university students do not represent the average population. This is a bias typical of peer-reviewed studies.

Granted the study’s authors make the same mistakes and even more. Blood samples were collected at times convenient to the students between 10 A.M. and 3 P.M. Timing affects testosterone levels. While the students were age matched, they were not weight matched.

ead1From their table 1 data above, the African Americans (AA) were over 11 lbs heavier than the European Americans (EA). And yes, weight can affect T-levels if the additional weight indicates adiposity. The study authors adjusted for weight:

Adjustment by analysis of covariance for time of sampling, age, weight, alcohol use, cigarette smoking, and use of prescription drugs somewhat reduced the differences. After these adjustments were made, blacks had a 15% higher testosterone level and a 13% higher free testosterone level.

So Rushton (a double doctorate psychologist who knows about variable adjustments) did not read properly.

ead2Note as well that table 2 indicates more feminizing E1 and E2 for AA males. Notice also that AAs have 18.6% more total-T and 21.1% more free-T (cFT) than EAs. Thus Rushton is referring to the total-T values when he uses 19% which is yet another problem. He should be more concerned with the cFT values.

The second study Rushton (2000a, 2000b) uses is Ellis & Nyborg (1992) which examines 4462 army veterans of African American and European American descent. Army veterans may be more predisposed to aggressiveness and higher testosterone levels thus presenting a possible bias.

From the abstract:

Racial and ethnic variations in serum testosterone levels were investigated among a large sample of male Vietnam era veterans. Based on geometric means, significant average differences were found between 3,654 non-Hispanic white and 525 black individuals. The geometric mean for testosterone levels among 200 Hispanic individuals was similar to that of non-Hispanic white individuals. Regarding two other racial/ethnic groups (Asian/Pacific Islanders and Native Americans), no significant differences were found, due perhaps to small sample sizes. Results were interpreted as having considerable potential for explaining some of the race differences in the incidences of cardiovascular diseases, hypertension, and prostate cancer.

Given Rushton’s thesis, this study presents nothing in his defense. A 3% difference is insignificant and can be simple fluctuations due to any number or combination of conditions. Why the study would call this “significant average differences” is unknown. Given no significant differences found in East Asian sub-clades (Hispanics, Asian/Pacific Islanders, Native Americans), this study hurts his cause. Unfortunately I do not have access to this study to probe further.

The third study is Hixson (1992) which is actually not a study but a medical report. In Rushton (2000a), the reference is stated as:

Hixson, J. R. (1992, October 20). Benign prostatic hypertrophy drug to be tested in prostate CA prevention. The Medical Post.

Even if Hixson (1992) was a study, the Japanese can never be indicative of all East Asians. Thus we can conclude that Rushton’s thesis in unstable and unscientific.

Problems with ethnic androgen differential studies is summed up by Kehinde et al. (2006) which states:

However, several studies have sought racial–ethnic differences in circulating androgen levels with conflicting results [2, 13, 14]. Thus, while Ross et al. [2] found a statistically significant higher TT (19%) and free testosterone (21%) levels in young African-American men compared to Caucasians, Wright et al. [13], Ellis and Nyberg [14] found higher but statistically insignificant testosterone, free testosterone and SHBG levels in comparable groups of African-American men and Caucasians living in various parts of the USA. These studies may however been flawed by small sample size, for example, Ross et al. [2] studied only 50 young men (20–30 years old) from each racial group and Wright et al. [13] studied a total of 33 men in both racial groups who were 20–40 years old. Other flaws of some of the earlier studies include inappropriate sample collection and storage, for example while Ellis and Nyborg [14] studied 525 African-American and 3654 Caucasian men, unfortunately the blood samples were collected at various times during the day in this study. It is now well established that diurnal variation exists in circulating testosterone concentrations in men with the highest levels found at 8.00 am and lowest levels at 8.00 pm [15, 16]. All of these flaws may be responsible for the conflicts in the association between steroid and steroid metabolites levels and aetiology of prostate cancer.

It had been suggested that, prolonged exposure (i.e. from <30 years of age) to high androgen levels in African-American and Caucasians compared to the Chinese is the basis for the observed difference in the effect of androgens in various racial/ethnic groups although other environmental, metabolic, dietary and genetic factors may also have important contributory factors [17].

The sources used in Kehinde et al. (2006) are:

2. Ross RK, Bernstein L, Judd H et al. Serum testosterone levels in healthy young black and white men. JCNI 1986; 76: 45–48.
13. Wright NM, Renault J, Willi S et al. Greater secretion of growth hormone in black than in white men: possible factor in greater bone mineral density. A clinical research study. J Clin Endocrinol Metab 1995; 80: 2291–2292.
14. Ellis L, Nyberg H. Racial/ethnic variation in male testosterone levels; a probable contributor to group differences in health. Steroids 1992; 57: 72–75.
15. Winters SJ, Brufsky A, Weissfeld J et al. Testosterone, sex hormone binding globulin and body composition in young adult African American and Caucasian men. Metabolism 2001; 50: 1242–1247.
16. Veldhius JD, King J, Urban RJ et al. Operating characteristics of the male hypothalamo-pituitary-gonadal axis: pulsatile release of testosterone and FSH and their temporal coupling with lutenising hormone. J Clin Encocrinol Metab 1987; 65: 929–941.
17. Santner SJ, Albertson B, Zhang GY et al. Comparative rates of androgen production and metabolism in Caucasian and Chinese subjects. J Clin Endocrinol Metab 1998; 83: 2104–2109.

NOTE: Helmuth Nyborg is a Danish race realist psychologist.


L. Ellis., and H. Nyborg. 1992. Racial/ethnic variations in male testosterone levels: a probable contributor to group differences in health. Steroids 57(2): 72-75.

Ross, R., Bernstein, L., Judd, H., et al. 1986. Serum Testosterone Levels in Healthy Young Black and White Men. Journal of the National Cancer Institute 76(1): 45-48.

Rushton, J. 2000a. Race, Evolution, and Behavior: A Life History Perspective. Port Huron, USA: Charles Darwin Research Institute.

Rushton, J. 2000b. Race, Evolution, and Behavior: A Life History Perspective. 2nd special abridged edition. Port Huron, USA: Charles Darwin Research Institute.

Stuckert, Robert. 1958. African Ancestry of the White American Population. The Ohio Journal of Science 58(3): 155-160.


Intelligence and metabolism

From Dr. Ray Peat, a post that touches on may issues of race realism’s fascination with IQ:

Even sophisticated people can fall into stereotyped thinking when they write about issues of intelligence. For example, no one considers it a sign of genius when a slum kid is fluent in both Spanish and English, but when some of history’s brightest people are discussed, the fact that they learned classical Greek at an early age is always mentioned. No one mentions whether they were competent in idiomatic Spanish.

In 1912 Henry Goddard, a pioneer in intelligence testing (and author of The Kallikak Family: A Study in the Heredity of Feeble- Mindedness), administered intelligence tests to immigrants and determined that 83 percent of Jews and 87 percent of Russians were “feeble-minded.” By the standards of the time, it was highly inappropriate for the child of extremely poor Jewish immigrants from eastern Europe to be so bright.

During this time Lewis Terman was studying bright children, and wanted to disprove some of the popular stereotypes about intelligent people, and to support his ideology of white racial superiority. In 1922 he got a large grant, and sorted out about 1500 of the brightest children from a group of 250,000 in California. He and his associates then monitored them for the rest of their lives (described in Genetic Studies of Genius). His work contradicted the stereotype of bright people as being sickly or frail, but, contrary to his expectation, there was an association between maladjustment and higher I.Q.; the incidence of neurotic fatigue, anxiety, and depression increased along with the I.Q. The least bright of his group were more successful in many ways than the most bright. He didn’t really confront the implications of this, though it seriously challenged his belief in a simple genetic racial superiority of physique, intellect, and character.

I.Q. testing originated in a historical setting in which its purpose was often to establish a claim of racial superiority, or to justify sterilization or “euthanasia,” or to exclude immigrants. More recently, the tests have been used to assign students to certain career paths. Because of their use by people in power to control others, the I.Q. tests have helped to create misunderstanding of the nature of intelligence. A person’s “I.Q.” now has very strong associations with the ideology of schooling as a road to financial success, rather than to enrichment of a shared mental life.

Terman’s book (volume 1 only) is freely available here. Volume 5 here.

For more on IQ tests and intelligence from a race realist POV, see a series of posts by Satoshi Kanazawa on his blog E pur si muove.

Genetic Variance

Ancestry by DNA Ancestry Map

AncestrybyDNA has a chart of their genetic tests and variance which illustrates (not conclusively though) that certain populations are more mixed than others.

Those which they have found to be more diverse as per their own definitions of ‘well-defined populations’ are:

  • South Asians;
  • North Africans;
  • Pacific Islanders;
  • American Indians and;
  • Mexicans.

Some data seems problematic as in the low East Asian contribution to indigenous (true) Americans, their inclusion of South Asians and Middle Easterners in the European clade and the separation of the ‘Native’ American and East Asian clades. They don’t state any rigorous reason for their cladistic definitions (see BioGeographical Ancestry). Assumptions also present problems but this is inherent to all extrapolations.

African contribution (over 5%) was found in:

  • African Americans;
  • North Africans;
  • Puerto Ricans;
  • Middle Easterners;
  • Iberians;
  • South Asian Indians; and
  • Mexicans.

East Asian contribution (over 5%) was found in:

  • North Africans;
  • Turks;
  • South Asian Indians;
  • Chinese;
  • Japanese;
  • Atayals;
  • South East Asians;
  • Pacific Islanders; and
  • American Indians.

Indigenous American contribution (over 5%) was found in:

  • Puerto Ricans;
  • Iberians;
  • Italians;
  • South Asian Indians;
  • South East Asians;
  • Pacific Islanders; and
  • Mexicans.

If one were to join the American and East Asian clades, the following populations have the most Asian ancestry:

  • Puerto Ricans (12.4%);
  • Italians (10%);
  • Turks (12.4%);
  • South Asian Indians (36%);
  • South East Asians (88.5%);
  • Pacific Islanders (71.6%); and
  • Mexicans (51.2%).