Ross et al (1986) is one of the most oft-cited papers that HBDers use to attempt to show that blacks have higher levels of testosterone than whites, which then—supposedly—goes on to explain higher rates of crime, aggression, and prostate cancer. However, people 1) say this only from reading the abstract (and not reading the full paper) and 2) even if they could read the paper they would not know where the flaws were to point them out and discredit the study based on flawed methodology. I see this study getting cited every now-and-then and I’m sick of seeing it.
Ross et al (1986: 45) state that they “recently reviewed the evidence that endogenous levels of certain steroid and polypeptide hormones are causally related to a group of human cancers, including cancer of the prostate gland.” I’ve shown how even injecting a man with exogenous testosterone does not worsen his prostate cancer (Eisenberg et al, 2015; Boyle et al, 2016) and testosterone doesn’t cause prostate cancer (Stattin et al, 2003; Michaud, Billups, and Partin, 2015). So this has been falsified. Even if blacks had the testosterone levels that they claimed it still would not cause higher rates of PCa (prostate cancer) incidence.
They solicited study participants from two colleges around the Los Angeles metro area. The two universities they got their sample from were the University of Southern California and California State University of Los Angeles. They recruited individuals through postings on the school bulletin board in in the school newspaper. They got 50 blacks and 50 whites. They then write something that’s troubling to me: “A convenient time for blood collection was arranged, and students were met by a nurse epidemiologist (R. H.) at either the Student Health Center or another mutually convenient location” (Ross et al, 1986: 45). This is dumb. The students were assayed at all times between 10 am and 3 pm; testosterone levels are highest at 8 am though one study on older men shows that assaying between 8 am and 2 pm doesn’t matter (Crawford et al, 2015). However, for the purposes of discussing this paper this is irrelevant.
Table 1 from Ross et al (1986) tells us a lot about the flaws in the study—most importantly, the assay times. A majority were assayed between 10 am and 2 pm—which would depress testosterone though people assayed nearer to 10 am would have higher levels than people assayed nearer to 2 pm. Alcohol consumption only decreases testosterone while drunk, and a majority of the cohort did not consume alcohol within 12 hours of being assayed.
They come to the conclusion that the mean total testosterone level was 19 percent higher than whites whereas free testosterone was 21 percent higher. In regard to the assay collection times, Ross et al (1986: 47) write:
There was a negative correlation between time of sampling (No. of min elapsed since 0600 hr) and testosterone levels for whites (r=-O.4I) but not for blacks (r = -0.08). Adjustment for this variable caused a small reduction in geometric mean differences in levels of testosterone and free testosterone between blacks and whites. After simultaneous adjustment by analysis of covariance for time of sampling and age, weight, alcohol use, smoking, and use of prescription drugs, there
remained a 15% difference in total testosterone levels and a 13% difference in free testosterone levels between blacks and whites.
Even though they ‘adjusted for this variable’, it’s still a huge confound. Testosterone assays must be taken nearer to 8 am; the fact that people were assayed all over the place in the span of a 5 hour time period while testosterone levels decrease throughout the day is a huge red flag.
They then say that they are “uncertain why young black men have higher levels of circulating testosterone than white men“, though small sample sizes, a large range of variation in assay time, and a nonrepresentative sample is why. Other, more robust, analyses show a smaller ‘gap’, about 2.5 to 4.9 percent, favoring blacks (Richard et al, 2014). All in all, this study has huge flaws and should not be pointed to—especially today in 2017—because much larger analyses with much better methodologies have been carried out and some studies show no difference while others show a small difference favoring blacks but that still would not explain higher rates of testosterone, aggression and crime.
Ross et al (1986) is used by hereditarians such as Lynn (1990), Rushton (1997) and Hart (2007). Lynn (1990) states that these large testosterone differences discovered by Ross et al (1986) lend credence to Rushton’s r/K selection theory in which blacks have more children than whites who have more children than East Asians. Evidence for this assertion, states Lynn (1990) is the fact that blacks have higher rates of prostate cancer than whites who have higher levels of testosterone than East Asians, however this has been disproven by ethnicmuse.
Rushton (1997: 170) states that blacks had 19 percent higher levels of testosterone citing Ross et al (1986), however, Rushton didn’t cite the adjusted level which ended up being 15 percent, and, again, doesn’t mean anything to their hypothesis.
And Hart (2007) yet still repeats the same old stuff that “these differences in sexual behavior may be a consequence of the fact that blacks, on average, have higher levels of testosterone than whites“. These three researchers, clearly, are citing this study uncritically because it fits with their ‘racial hierarchy’. In fact, Rushton (1999) asked if testosterone was a ‘master switch’. In this paper, he cites Ellis and Nyborg (1992) who find that blacks had 3 percent higher levels of testosterone than whites. They gave the following values:
For the purposes of Rushton’s interpretation, writes Fish (2013), “These uncorrected figures are, of course, not consistent with their racial r- and K-continuum.” This, of course, is a big deal. Rushton cites this study as if it lends credence to his claims but it actually found the same result as Richard et al (2014). Thirty years after Ross et al (1986) we have numerous other studies showing a small gap between the races or no gap at all. We have numerous other studies showing that testosterone does not cause aggression, crime or violent behavior. However, people will still point to the abstract of Ross et al (1986) and think that they have proven that blacks have higher levels of testosterone than whites which proves how and why they have higher levels of testosterone, commit more crime and are all around more violent.
Though, as I have written about before, what Ross et al (1986) set out find the answer for (testosterone influencing higher levels of prostate cancer) can most definitely be explained by diet and lack of certain vitamins such as vitamin D, since low levels of this steroid hormone (it’s not a vitamin) cause prostate cancer (Schwartz and Hulka, 1990; Zhao and Feldman, 2001; Khan and Partin, 2004; Garland et al, 2006). Diet can explain a lot of the variation, as I have argued in the past.
In sum, Ross et al (1986) is the only study that I’ve found on racial testosterone differences that shows that extreme directionality favoring blacks over whites. This should set off some alarms in some people’s heads. People—psychologists included—need to learn about these hormones, how they’re produced in the body, and what they’re used for. Hormones don’t cause behavior, hormones influence behavior.
This fear of testosterone is largely overblown. We need testosterone for proper normal functioning. We need testosterone to be socially dominant; if you have lower levels you’ll be less socially dominant. This fear of testosterone—especially when it comes to race and it’s so-called causes—is largely pushed by Lynn, Rushton, Hart, and Ellis. I have spent a lot of time and thousands of words showing that they are wrong and testosterone is not a hormone to fear. It does not cause crime. It does not cause aggression. It does not cause prostate cancer. I’ve rebutted quite a few hereditarians on testosterone as well as testosterone and race, and if I come across more I will rebut them as well.