Ethnic Androgen Differentials II

For some background, see Ethnic Androgen Differentials I

Testosterone Molecule

There are some important studies which provide a framework for understanding ethnic androgen differences. One such study is Zitzmann and Nieschlag (2001). They performed a review of the effects of testosterone (T). Some important points:

  • Stress, depressive illness and PTSD decreases T
  • Strength training exercise increases T post-workout but continuous strength training does not affect T significantly
  • Higher T does not necessarily increase aggression:

There is strong evidence for an interdependent feedback mechanism between testosterone and aggression, modified by experiences of victory and defeat, as well as by education, cultural and socioeconomic background.


The immense variety of individual response patterns to androgens is demonstrated by a controlled trial in which exceptionally high doses (600 mg/week) of testosterone cypionate were administered; maniac effects were reported in only 16% of the men. The psychological behaviour of the others remained unremarkable. Effects of external administration of testosterone on aggressive behaviour are controversial (55-58).

  • T may affect some cognitive abilities
  • T seems to affect verbal skills
  • Low T may enhance musical ability in males
  • Low status occupations seem to lower T levels due to lack of situational control


In violent sexual offenders, serum testosterone levels were significantly higher in the group of native Americans than in Caucasians (41, 108). Such analyses have led to conclusions that a predisposition to crime may be genetically based, have consistent racial variations and may be expressed as serum testosterone levels and brain size, which are supposed to cause different crime rates in Asian, Caucasian and African-Americans (109). Not biased by selection of subjects in conflict with the law, other studies demonstrate similar testosterone levels in African-American and Caucasian American men. (110, 111). In contradiction to the so-called predisposition theories on crime rates and race, another study involving 1127 men and adjusting data for age and body mass index (BMI), showed levels of total and bioavailable testosterone to be highest in Asian Americans, followed by African-Americans and lowest in Caucasian Americans (112). A greater risk for African-Americans of developing prostate cancer in comparison to Caucasians or Asians seems to exist; testosterone levels are not different, but hereditary factors concerning polymorphism in androgen receptors and 5a-reductase activity are held responsible (113). Supporting results were found comparing Americans of Chinese or Caucasian origin, explaining the lower expression of beard and body hair in Asians (114). Ethnic differences in sex hormone-dependent features are obviously restricted to genetic polymorphisms, leading to differential enzyme and receptor activity, but not to different testosterone levels (115, 116). The cultural background can also influence testosterone levels through nutritional factors (see Dietary influences; 117).

  • High T increases libido but does not translate into high sexual activity
  • The data are inconclusive if male homosexuals have lower T
  • High T does not necessarily correlate with aggression, both high and low levels of T can result in aggressive behaviour
  • Increasing age may decrease androgen levels
  • Obesity causes a decrease in T
  • Diet influences T
  • Living at higher altitudes seems to lower T
  • There seems to be a correlation between voice deepness and T
  • Balding may not necessarily be associated with T
  • Burns, infection, injury, disease, poor digestion and nutrient absorption can affect T
  • Marijuana does not seem to lower T
  • Excessive alcohol lowers T
  • A person’s T varies throughout the year (figure 1)




However, methodological problems of measuring testosterone in body fluids are often overlooked. Serum testosterone levels are determined routinely by radioimmunoassays or luminescence immunoassays. Because there is a strong diurnal variation with morning concentrations being approximately 20-30% higher than evening values (Fig. 1; (191-197)), samples should always be taken during the morning hours. Only then will single point samples be representative (198-200). Seasonal variations of testosterone levels should also be taken into consideration, at least in studies comparing several groups at different time points (Fig. 1; (201-203)). Because testosterone is a small molecule and present in relatively low concentrations, it is not surprising that there is marked inter- and intra-assay bias and variation (Fig. 2). Therefore, comparing testosterone levels determined with different assays and/or in different laboratories becomes questionable. Quality control uncovers this problem, but has not contributed to its solution. Direct measuring of free testosterone concentrations, which are found at much lower concentrations than the total amount of this steroid (about 2%), is subject to even greater difficulties.

In summary, although assays have been available for over 30 years, measurement of serum testosterone is still relatively unreliable and represents a significant bias when interpreting results and correlating them with biological phenomena. Unfortunately, peer-reviewed journals as the ultimate filter have not contributed much to improve the reliability of published values. Therefore, all reported variations of testosterone levels and their associations with physical and mental aspects must be viewed very critically.



Because of the great inter-individual variation, application of these results to individuals is questionable. A generalisation towards phenomenological structures would also require sublimation of various one-dimensional findings. Combining results concerning hormone interactions with physical or psychological phenomena to create new constructs explaining sociological observations can be dangerous and misleading. It is too easy a step from assigning high testosterone levels to aggressive people to the explanation of legal offences of violent nature and further, discrimination of racial or subcultural groups. This could mean, for example, that persons involved in regular marathon training are likely to show a submissive kind of conflict management and will not develop a balding hair pattern, or that watching blue movies would lead to premature balding which could then be prevented by a strict vegetarian diet or by moving to the mountains. It could mean that a man exposed to increased mental stress is likely to show extraordinary verbal fluency. Regarding all results mentioned, a person with high testosterone levels would then be living at sea level, would refrain from endurance training, enjoy high socioeconomic status as well as daily consumption of meat and sexual pleasures, but avoid gaining too much weight. He would be characterised by increased spatial cognition, but not be able to utter sentences fluently. He would have a deep voice, sing in the bass register, his beard growth would be prolific and his head would be bald. He would react rather aggressively to threatening situations. A person with low testosterone levels would be a mountaineer, exercising daily by long runs, living a frugal life with vegetarian food, without a partner and constantly harassed by his environment. He would have a full head of hair, a high-pitched voice, and would converse fluently about his submissive behavior and his lack of ability to solve mathematical problems (see Table 1). The described persons might exist, but they are likely to be rare, and just how high their actual testosterone levels would really be remains open to question. The rest of the male population cannot be judged by testosterone levels and we cannot guess their androgen levels by looking at their lives. Simplifying approaches are likely to produce results with simple structures, which, while logically consistent in themselves, do not conform to the external structures we experience. Many other factors of influence are omitted in these approaches such as experiences that are partially reinforced by hormonal feedback on behaviour that caused them and that social and physical environment and genetic background affect not only an individual’s personality but also a single parameter such as testosterone levels.

Source: Zitzmann, M., and E. Nieschlag. 2001. Testosterone levels in healthy men and the relation to behavioural and physical characteristics: facts and constructs. European Journal of Endocrinology 144(3): 183-197.


  • High T makes men aggressive: not necessarily, some low T high aggression males are calmed with T supplementation;
  • East Asians have the lowest T and Africans the highest, these are also genetically determined: not necessarily, some studies have found that Europeans have the lowest bio-available T levels and East Asians the highest. Also, nutrition plays a factor in T levels;
  • Peer-reviewed studies are authoritative and give credence to ethnic T variations: not necessarily, T is affected by (1.) the yearly timing of sample collection (2.) the method of analysis (3.) the daily timing of sample collection; and
  • Realistic portrayals of different ethnics can be found by analyzing testosterone: not necessarily, until better measurement methods are found, these studies if used improperly can create simplistic, racist and illogical views. Additionally, the data are inconclusive so no realistic portrayals can be derived.

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.