Testosterone and human variation

My last few posts have presented the following hypothesis:

1. During the hunter-gatherer stage of prehistory, food gathering was less important further away from the equator. This was because the longer winters restricted gathering of fruits, vegetables, roots, etc.

2. At these higher latitudes, women had to shift from food gathering to food processing or to activities unrelated to food procurement. For a man, this raised the cost of providing for a second wife, making it prohibitive for all but the ablest hunters. Thus, with increasing distance from the equator, polygyny became less common and men provided more for their wives and children.

3. This north-south gradient accentuated with the advent of agriculture. In the tropical zone, year-round farming made women much more self-reliant in feeding themselves and their children. Conversely, it made polygyny much less costly for men, thereby increasing male-male competition for mates.

4. In these tropical agricultural societies, men and women initially adapted by pushing the envelope of behavioral plasticity. Over time, however, natural selection favored those genotypes that lay closer to the new behavioral mean, i.e., increased predisposition to polygyny, increased aptitude for male-male competition for mates, and decreased paternal investment.

5. This cultural and biological evolution seems to have gone the furthest among the agricultural societies of sub-Saharan Africa, where 20-50% of all sexual unions are polygynous. These societies typically dealt with the resulting wife shortage by giving priority to men who were at least ten years past puberty. Younger men (15-25 years old) could have sex lives only through surreptitious means or by seizing women during domestic or intertribal warfare.

6. This behavioral shift probably involved a number of genetic changes, including higher testosterone levels in young males. Raising these levels seems to lower the threshold for expression of polygynous behavior while raising the one for expression of paternal investment.

High testosterone levels are widely attested among populations that descend from sub-Saharan agriculturalists. When Ross et al. (1986) studied white and black American students (mean age = 20 yrs), they found mean testosterone levels to be 19% higher in the blacks than in the whites and free testosterone levels 21% higher. In a later study, Ross et al. (1992) found that young East Asian men had intermediate testosterone levels but less 5α-reductase—an enzyme that converts testosterone into the physiologically more active DHT.

Testosterone levels are higher in black boys than in white boys as early as the prepubertal age class (5-9 yrs) (Abdelrahaman et al., 2005). Black males seem to reach maximum t levels during adolescence and early adulthood (Ross et al., 1986; Winters et al., 2001). The black-white difference then shrinks after 24 years of age and is gone by the early 30s (Gapstur et al., 2002). It actually seems to reverse in later years (Nyborg, 1994, p. 111-113).

Broadly speaking, lifetime exposure to testosterone correlates with the incidence of prostate cancer and the highest incidences in the world are among African American men (Brawley and Kramer, 1996). Other populations of black African descent (i.e., West Indians and sub-Saharan Africans) exhibit lower incidences, but these have been shown to reflect underreporting and are probably just as high (Glover et al., 1998; Ogunbiyi and Shittu, 1999; Osegbe, 1997). It should be noted that the human body’s exposure to testosterone depends not only on blood serum levels but also on how well this hormone interacts with androgen receptors, which seem to be more numerous and more receptive in black subjects (Kittles et al., 2001).

Despite these findings, several studies have raised doubts about the existence of a black-white difference in t levels:

Ross et al. study

Ross et al. (1986) found that the black-white difference shrank to 15% for testosterone levels and to 13% for free testosterone levels after adjusting for time of sampling, age, weight, alcohol use, cigarette smoking, and use of prescription drugs. Thus, these factors may have artificially boosted testosterone levels in black subjects.

Of these factors, adjusting for weight and age actually widened the black-white difference. Adjusting for time of sampling narrowed it but only to a small degree. It was thus the lifestyle factors—smoking and drinking—that seem to be associated with high t levels in black subjects.

These lifestyle factors have been investigated by Martin et al. (2002), who found that pubertal development in males and females was associated with sensation seeking, smoking, and drinking, even after controlling for age. This resonates with my own experience. At school, boys who didn’t drink or smoke were derided as immature ‘sucks’. And strangely enough they looked less mature, being smaller and less hirsute. The smokers and drinkers were generally bigger and already sporting sideburns in Grade 8.

It looks as though higher t levels cause boys to seek social environments where drinking and smoking is more common, probably because such environments provide more opportunities for male dominance and access to women. If we adjust for these lifestyle factors, we end up biasing the sample toward subjects with low t levels.

Rohrmann et al. study

Rohrmann et al. (2007) tested black Americans, white Americans, and Mexican Americans in three age classes: 20-44; 45-69; and 70+. In the 20-44 age class, white Americans had the lowest level of testosterone (5.17 ng/mL), black Americans an intermediate level (5.35), and Mexican Americans the highest level (5.55)

As noted above, the black-white difference in t levels shrinks after 24 years of age, is gone by the early 30s, and seems to reverse at older ages. So it should not show up in the average t levels of 20 to 44 year-olds. In addition, median age differed among the three groups of 20 to 44 year-olds, being 31.8 for the whites, 29.5 for the blacks, and 28.6 for the Hispanics. These age differences probably account for the higher t levels of the Mexican American subjects. Moreover, the youthfulness of the Mexican-American group is probably understated by the use of median ages rather than mean ages. If means had been used, the age differences would have probably been greater, given the more youthful age pyramid of the Hispanic-American population.

Finally, it is odd that the 45-69 year old black men had higher t levels (5.62) than the 20-44 year old black men (5.35). This is contrary to the literature on hormonal variation with age and suggests the authors may have massaged their data, perhaps by eliminating high testosterone outliers.

Ellison et al. study

Ellison et al. (2002) measured salivary testosterone in young men (15-30 years) from the United States, Congo, Nepal, and Paraguay. The American subjects had the highest t levels (335 pmol/l), followed by the Congolese (286), the Nepalese (251), and the Paraguayans (197).

But who were these young American subjects? They are simply identified as … young Americans—a demographic that is now less than 60% of European descent. In Boston, where the study was conducted, the youth population is now largely of African or part-African descent. In 2005, Boston public schools were 46% black and 31% Latino (mainly Puerto Ricans and Dominicans).

Perhaps a disproportionate number of subjects were white Harvard students. On the other hand, the authors state, "the USA participants were recruited by public advertisement, thus the potential for self-selection bias should be noted.” The demographic profile should therefore have resembled that of others who give blood in exchange for payment, i.e., disproportionately poor and non-white. In any event, the results are unusable without any indication of racial composition.

African studies

Several studies have found lower testosterone levels in African populations than in North Americans. This difference might be partly due to the effects of malnutrition or infectious diseases, notably among the Zimbabwean subjects studied by Lukas et al. (2004). The main reason, however, is that these studies used largely middle-aged or even elderly subjects. Lukas et al. (2004) report a mean age of 42.18 (S.D. = 13.75). Their scatter plot (Fig. 2) suggests a logarithmic decline in t values with age, but they had too few subjects below 25 for any meaningful analysis of that age group. The same criticism applies to a study by Campbell et al. (2003) of t levels in Ariaal pastoralists from northern Kenya. Those subjects had a mean age of 46.8 (S.D. = 14.3 years).

In addition, some of these studies were done on hunter-gatherers, like the !Kung of Namibia and the Ituri Forest pygmies of the Congo, who have low levels of polygyny and weak male-male competition for mates (e.g., Winkler and Christiansen, 1993). Their low t levels are thus to be expected. This factor may also partly explain why Ellison et al. (2002) found lower t levels in young Congolese than in young Americans. The Congolese subjects were Lese, an Ituri Forest Bantu population with high admixture from Pygmy hunter-gatherers.


References

Abdelrahaman, E., Raghavan, S., Baker, L., Weinrich, M., and Winters, S.J. (2005). Racial difference in circulating sex hormone-binding globulin levels in prepubertal boys. Metabolism, 54, 91-96.

Brawley, O.W. and Kramer B.S. (1996). Epidemiology of prostate cancer. In Volgelsang, N.J., Scardino, P.T., Shipley, W.U., and Coffey, D.S. (eds). Comprehensive textbook of genitourinary oncology. Baltimore: Williams and Wilkins,

Campbell, B., O’Rourke, M.T., and Lipson, S.F. (2003). Salivary testosterone and body composition among Ariaal males. American Journal of Human Biology, 15, 697-708.

Ellison, P.T., Bribiescas, R.G., Bentley, G.R., Campbell, B.C., Lipson, S.F., Panter-Brick, C., and Hill, K. (2002). Population variation in age-related decline in male salivary testosterone. Human Reproduction, 17, 3251-3253.

Gapstur, S.M., Gann, P.H., Kopp, P., Colangelo, L., Longcope, C., and Liu, K. (2002). Serum androgen concentrations in young men: A longitudinal analysis of associations with age, obesity, and race. The CARDIA male hormone study. Cancer Epidemiology, Biomarkers & Prevention, 11, 1041-1047.

Glover, F., Coffey, D., et al. (1998). The epidemiology of prostate cancer in Jamaica. Journal of Urology, 159, 1984-1987.

Kittles, R.A., Young, D., Weinrich, S., Hudson, J., Argyropoulos, G., Ukoli, F., Adams-Campbell, L., and Dunston, G.M. (2001). Extent of linkage disequilibrium between the androgen receptor gene CAG and GGC repeats in human populations: implications for prostate cancer risk. Human Genetics, 109, 253-261.

Lukas, W.D., Campbell, B.C., and Ellison, P.T. (2004). Testosterone, aging, and body composition in men from Harare, Zimbabwe. American Journal of Human Biology, 16, 704-712.

Martin, C.A., Kelly, T.H., Rayens, M.K., Brogli, B.R., Brenzel, A., Smith, W.J., and Omar, H.A. (2002). Sensation seeking, puberty, and nicotine, alcohol, and marijuana use in adolescence. Journal of the American Academy of Child & Adolescent Psychiatry, 41, 1495-1502.

Nyborg, H. (1994). Hormones, Sex, and Society. The Science of Physiology. Westport (Conn.): Praeger.

Ogunbiyi, J. and Shittu, O. (1999). Increased incidence of prostate cancer in Nigerians. Journal of the National Medical Association, 3, 159-164.

Osegbe, D. (1997). Prostate cancer in Nigerians: facts and non-facts. Journal of Urology, 157, 1340.

Pettaway, C.A. (1999). Racial differences in the androgen/androgen receptor pathway in prostate cancer. Journal of the National Medical Association, 91, 653-660

Rohrmann, S., Nelson, W.G., Rifai, N., Brown, T.R., Dobs, A., Kanarek, N., Yager, J.D., Platz, E.A. (2007). Serum estrogen, but not testosterone levels differ between Black and White men in a nationally representative sample of Americans. The Journal of Clinical Endocrinology & Metabolism, 92, 2519-2525

Ross, R.K., Bernstein, L., Lobo, R.A., Shimizu, H., Stanczyk, F.Z., Pike, M.C., and Henderson, B.E. (1992). 5-apha-reductase activity and risk of prostate cancer among Japanese and US white and black males. Lancet, 339, 887-889.

Ross, R., Bernstein, L., Judd, H., Hanisch, R., Pike, M., and Henderson, B. (1986). Serum testosterone levels in healthy young black and white men. Journal of the National Cancer Institute, 76, 45-48.

Winkler, E-M., and Christiansen, K. (1993). Sex hormone levels and body hair growth in !Kung San and Kavango men from Namibia. American Journal of Physical Anthropology, 92, 155-164.

Winters, S.J., Brufsky, A., Weissfeld, J., Trump, D.L., Dyky, M.A., and Hadeed, V. (2001). Testosterone, sex hormone-binding globulin, and body composition in young adult African American and Caucasian men. Metabolism, 50, 1242-1247.