“Reading, after a certain age, diverts the mind too much from its creative pursuits. Any man who reads too much and uses his own brain too little falls into lazy habits of thinking.” – Albert Einstein

I recently came across an episode of the podcast TED Radio Hour titled The Biology of Sex that purports to explore how very much non-dimorphic and non-binary sex and gender are. If that statement makes you skeptical of the contents of the podcast, perhaps it should. If it doesn't, I hope you'll take time to read the following. I take the time below to analyze a single statement in that podcast in order to shed some light on a common problem in research and discussion of LGBT topics.

The podcast opens with the story of Emily Quinn, a person who has an androgen insensitivity disorder, who has XY chromosomes, and has a vagina and testes but no ovaries or uterus. It is not clear from the podcast if it is an actual vagina or something that looks like one but is not. At the end of Emily's TED talk (partially presented in the episode) Emily states that the number of people in the "intersex" category is 1-2 people per 100.

I became really curious and skeptical about the 1-2 in 100 births figure, especially because I had spent some time a few years ago reading about disorders of sexual development and remembered much lower figures, so I decided to track the figure down.

I found the source of the number in an article referenced in the Intersex article on Wikipedia. It was a claim made by Blackless, et al. in an article in the American Journal of Human Biology and then in a book, both published in the year 2000. The article is a meta-analysis of various studies on disorders of sexual development that attempt to estimate their individual prevalence.

I glanced through that article and then looked for commentary on it. The most simplistic reply to the article was a letter to the editor of The Journal of Sex Research. The author states:

"Many reviewers are not aware that this figure includes conditions which most clinicians do not recognize as intersex, such as Klinefelter syndrome, Turner syndrome, and late‐onset adrenal hyperplasia. If the term intersex is to retain any meaning, the term should be restricted to those conditions in which chromosomal sex is inconsistent with phenotypic sex, or in which the phenotype is not classifiable as either male or female. Applying this more precise definition, the true prevalence of intersex is seen to be about 0.018%, almost 100 times lower than Fausto‐Sterling's estimate of 1.7%.

But that appears to be only the tip of the iceberg with the problems with the 1-2 per 100 figure. A more extensive letter to the editor of the same journal that published the article containing the original number systematically reviews the claims made in Blackless, et. al.'s article and finds a number of problems in the approach that cause an overestimate of the prevalence by an order of magnitude, even if all the conditions included in the review are included. Some of their statements (sorry, this is extensive):

"Walzer and Park (1977) is a study of male infants only; thus, their results need to be roughly halved"
"Hamerton et al. (1975), Lin et al. (1976), and Lubs and Ruddle (1970) find zero instances of XX male. However, these lines are merely left blank on the table and the incidence of XX is calculated without these studies."
"... all of the AIS cases were "females'' with inguinal hernias. Thus, that study's estimated incidence of 0.049 per 1,000 male births (or approximately 0.025 per 1,000 total births) should not be supplemented. Blackless et al.'s calculation increases the then available numbers bymore than threefold, to 0.076 per 1,000 live births."
"There is much more literature on the relatively common congenital adrenal hyperplasia (CAH). The data for classic CAH is well summarized by Blackless et al. (p. 155). However, it is their presentation of nonclassic CAH (p. 156) that led me to examine this review article so closely in the first place. Blackless et al. indicate that nonclassic CAH occurs with a frequency of 1.5%. This figure makes up nearly 87% of their total 1.728 non-dimorphic births per 100, and is thus crucial to the research. While Blackless et al. caution that revisions may be required because only one study is employed, the authors owe it to their readers to provide as much information as possible. It should first be noted that non-classic CAH–typically a mild androgen excess–is very often entirely asymptomatic in males (White and Speiser, 2000). If we are to count female hirsutism (a common symptom of the syndrome) as a form of non-dimorphism blurring sexual difference, it hardly seems fair to count male hirsutism in the same fashion. Second, the original study cited by Blackless et al. (Speiser et al., 1985) does not state that the frequency of late-onset CAH is 1.5%. Rather, it indicates a high frequency for four American ethnic groups (Ashkenazi Jews 3.7%, Hispanics 1.9%, Yugoslavs 1.6%, and Italians 0.3%), a lower 0.1% incidence for a Caucasian sample excluding the above groups, zero incidence in a small group of Blacks and American Indians, and a 0.9% overall frequency across these groups. I gather that Blackless et al. take the average of the groups (misquoting the original research for the Caucasian population as 0.01/1,000, and omitting the zero incidence figure) to arrive at their total of 1.50%"

In summary, the certain studies have been ignored in favor of other studies that showed higher results, and numbers for syndromes that don't even cause symptoms (such as excess androgens in males) are exaggerated and included in the overall figure. In fact, that final category seems to make up a majority of the estimated frequency.

After covering various other issues, the author of the letter attempts their own summary, being generous in many areas, and concludes:

"The total I arrive at is 0.373 per 100 live births, a difference of an order of magnitude from Blackless et al.'s estimate of 1.728 ... While this data will undoubtedly transform over the next years, the numerous errors and omissions made by Blackless et al. suggest that they were too keen to find a relatively high incidence of sexual nondimorphism."

So the 1-2 per 100 number does not appear to be accurate. At best it is the result of a clumsy analysis of various studies. At worst, an intentional exaggeration.

I find it irresponsible that the podcast presents the number as if it represents reality, when in fact it is a highly-controversial and almost certainly inaccurate and exaggerated number.

Perhaps the worst part of how this number is used is that most who cite it use it to claim that sex is not actually binary - yet neglect to mention that these conditions are disorders, many of which cause infertility. In other words, they do not represent non-binary sexes but rather disordered variations that prevent fully-functioning sexual development.

Listening to the podcast, there were a number of other points where I felt the producers were similarly avoiding important counter-points that might provide a more complete and accurate understanding of the question they purport to be answering.

Unfortunately, the problems in this podcast, and in the study I've focused on, are, in my experience, much more common on this and similar topics of research and discussion than is usually recognized. In any other field of research, this level of sloppiness would not be tolerated. Yet it is here, and it's time we asked why - because such sloppiness has long-lasting effects on the lives of real people.