So where were we? Oh yes, the origins of gender.
Gee, I sure do like to choose the easy topics, don’t I? (What was I thinking??)
In our last episode, Part 2 of this post series, we had discussed the hypothesis that human brains are intrinsically female and are masculinized during fetal development by the actions of testosterone, resulting in male “typical” behavior and male gender.
This hypothesis was supported over the years, in part, by the 156 published case studies (Mazur, 2005) of chromosomal males (46,XY) with CAIS, Complete Androgen Insensitivity Syndrome. These individuals had no functional androgen receptor (AR) because of a mutation in their AR gene. Without working ARs, the testosterone produced by their internalized testes could not be recognized by the body, so these individuals were female-bodied, assigned female at birth, raised female and reportedly had female genders.
The gender hypothesis was questioned, however, when we considered the case report (T’Sjoen et al., 2010) of an individual with CAIS who, unlike all the other CAIS cases, reported a male gender identity.
So then we had to ask, if testosterone is indeed the factor that induces male gender in human brains, then how could this individual, whose brain cannot recognize testosterone, have a masculinized brain and a male gender?
Damn good question, don’t you think?
(If you click “Continue reading –>” I’ll take that as a ‘yes’)
Okay, so at this point in my last post, I made my little confession that I had simplified things. Let’s fill in some details now.
As we know, Mother Nature is ripe with biological variation. Trans and intersex folks are a testament to that. (And so are clowns. I mean, really, where did Nature ever come up with the idea for clowns??)
In this story of brain masculinization, gender, behavior, genetics and sex, there are quite a few places where biological variation can occur and confuse the hell out of the situation. In order to understand how the biological systems can vary, especially with respect to each other, we need a little background information.
Fetal Development of the External Genitalia & Reproductive System
Timing is everything.
Oh wait, I thought time was money and location was everything.
Anyway, for the purposes of our discussion, timing is important. With that in mind, let’s consider this from a trans perspective.
As trans people, we sometimes try to understand how our gender could end up in a mis-match with our biological sex. If testosterone is needed to masculinize the brain and is also needed to masculinize the body, then why wouldn’t both brain and body be masculinized (or not masculinized in the absence of testosterone) at the same time within a given individual? (This is where the bit about timing comes in.)
Well, what if determination of biological sex occurs at a different time during fetal development than determination of gender? To be able to answer that question, we need to know the timing of each.
With regard to the former, the human fetus, at approximately eight weeks of gestation, is ambiguous with regard to its biological sex. It is bi-potential at this stage because it can develop down the male path or the female path, depending on the genetic and hormonal signals it receives.
At this time, the gonads are undifferentiated; they can become either testes or ovaries. The duct-work does not have bi-potential development like the gonads, but rather, two systems are present. It’s genetic signals, at this point, that start the sexual differentiation process down one path or the other (Tanagho & McAninch, 2000).
In 46,XY “male” fetuses, the undifferentiated gonads take the path toward becoming testes. They secrete a protein called Anti-Müllerian Hormone (AMH; also known as Müllerian Inhibiting Substance – MIS) which does exactly what its name implies — it causes regression of the Müllerian ducts. The fetal testes also begin synthesizing and secreting testosterone at this time (until Week 24) which induces development of the Wolffian ducts into the epididymides (plural for epididymis), the vas deferens (the tubes that get cut during a vasectomy), the ejaculatory ducts and a pair of secondary sex glands called seminal vesicles (that produce part of the semen during ejaculation).
In 46,XX “female” fetuses, the undifferentiated gonads develop into ovaries, the Wolffian ducts regress and the Müllerian ducts differentiate to become the Fallopian tubes (i.e. oviducts), the uterus, cervix and the upper third of the vagina.
Differentiation of the bi-potential external genitalia occurs at this time as well. At roughly eight weeks of gestation, the genitalia of all fetuses have the same appearance, regardless of chromosomal sex. In the 46,XY male fetuses, differentiation of the genitalia requires the production of testosterone by the newly differentiated testes. Testosterone is then converted by the enzyme 5α-reductase to the more potent androgen, dihydrotestosterone (DHT), which drives differentiation of the external genitalia to the male form.
This bi-potentiality of the external genitalia is what allows female-bodied individuals, like trans men, when they administer testosterone, to develop virilized genitalia of analogous structures. The clitoris is equivalent to the penis and the labia to the scrotum and they can be enhanced with testosterone hormone replacement therapy, although the degree of enhancement is dependent upon the genetics of each individual.
As you can see, a lot of different things are happening fairly quickly with regard to development of biological sex, and the window of opportunity is relatively small for processes to go according to plan.
Of course, we know that things don’t always go according to plan in nature, and in the pathways that determine internal and external biological sex, there are myriad of steps where biological variability can occur, resulting in intersex conditions.
Below is a somewhat outdated table of intersex conditions and one can see that disorders of sexual development can arise from chromosomal abnormalities, genetic mutations of genes involved in sex determination, issues with gonadal differentiation, mutations of enzymes at different steps in steroidogenic pathways (i.e. production of testosterone or DHT), mutations in the androgen receptor gene, incomplete regression of Müllerian duct structures, and others. (If you click on the table, you’ll see it in a new window and it will be more legible.)
Okay, now we know about development of biological sex, but how does that relate to development of gender? One way to consider that question is to look at gender identity in individuals with variations in sexual development. For that, we can consider one particular type of intersex condition that has been studied relatively more than some of the others with regard to gender-specific behavior and gender identity.
Congenital Adrenal Hyperplasia
The most common intersex condition in chromosomal (46,XX) females is congenital adrenal hyperplasia (CAH). The different types of CAH are caused by genetic mutations in one of six different enzymes in the steroidogenic pathway, but at least 90% of cases of females with CAH that are born with masculinized genitalia have a mutation of the CYP21 gene that encodes the 21-hydroxylase enzyme (reviewed in White & Speiser, 2000).
As seen below, the 21-hydroxylase enzyme converts progestins (progesterone and 17-OH-progesterone) to adrenal steroids, including cortisol. In the cases of CAH where 21-hydroxylase is mutated, the adrenal steroids cannot be made from the progestins, so the entire pathway shifts and the progestins are converted into androgens instead (Forest, 2004).
In addition, without feedback on the pituitary gland from adrenal steroids (because they aren’t being produced), the pituitary reacts by trying to induce the adrenal glands to make the steroids that are missing by increasing release of ACTH (adrenocorticotrophic hormone). The increased ACTH induces even more progestins to be synthesized by the adrenals which are subsequently converted into even more androgens (testosterone and dihydrotestosterone) that virilize the genitalia.
A range of masculinization of 46,XX individuals with CAH can occur, from “mild” clitoral enlargement to ambiguous genitalia to development of a penis and fusion of the labioscrotal folds to form a scrotum. Internally, the female reproductive organs are all present but Wolffian duct structures are rarely seen, theoretically because a higher level of testosterone is required to maintain Wolffian ducts than what occurs in these individuals. On the other hand, the 21-hydroxylase mutation in 46,XY fetuses usually does not affect genital development. (White & Speiser, 2000)
Going back to our topic of gender, one could hypothesize that the elevated androgens from CAH that virilize external genitalia of 46,XX fetuses could potentially masculinize the developing brain and affect gender. It appears that plenty of people have considered this theory, and I was surprised by the number of publications that were devoted to this topic.
Studies of Masculine Behavior and Gender in 46,XX Individuals with CAH
First, let’s consider the reasoning behind these studies. With the range of masculinization of genitalia in 46,XX individuals with CAH (from “mild” clitoral enlargement to ambiguous genitalia to fully masculinized genitalia), researchers studied these patients to try to correlate androgen exposure with masculine behavior and gender identity. The majority of the individuals in these documented cases were assigned female and raised as girls, but some of these individuals reported gender dysphoria later in life. The goal of these types of studies, therefore, was to give health professionals guidance on how to advise parents to raise these children (i.e. as girls or boys). At least, that’s what it says in these publications.
This method is better than what occurred in many of these cases in the not-too-distant past (and still happens today in some places), where surgery would be performed to “correct” the external genitalia, the chosen sex being based on the best results the surgeon could achieve with the given starting material. Fortunately, the standards of care for these individuals are changing and surgery is less-often imposed on these babies who cannot speak for themselves.
Going back to these evaluations of gender-specific behavior and gender identity, I really can’t evaluate the science. When it comes to measurements of behavior and gender, especially in children (which were studied in most of these publications), I don’t have enough insight to know what constitutes good versus bad science. For example, I wouldn’t know a good survey from a bad one when these kids or their parents are interviewed about their behavior.
What I do know, however, is that more masculine or less feminine behavior in girls does not necessarily translate to a male gender identity. In addition, I am wary of some of these studies because they are authored by psychologists who are notorious in eyes of the trans community, such as Ken Zucker, Michael Bailey and John Money.
Having said that, and looking collectively at the published reports, there appears to be a greater level of masculine (“boy-typical”) behavior expressed by 46,XX girls with CAH than by their sisters or by girls without CAH, such as playing more with “boys’ toys” like cars rather than with dolls, showing less interest in feminine dress, jewelry and make-up (i.e. more tomboyish), exhibiting more aggressive behavior and, like boys, showing greater spatial abilities and a higher incidence of left-handedness (Meyer-Bahlberg et al, 1996; Zucker et al., 1996; Berenbaum et al., 2000; Hines et al., 2002; Berenbaum & Bailey, 2003; Hall et al., 2004; Dessens et al., 2005; references therein).
In some of these studies, the degree of masculine behavior was correlated with some phenotypic (i.e. physical) aspect of CAH, such as genotype (i.e. mutations of the affected genes) or degree of genital masculinization, whereas in some, there was no correlation.
There are some caveats to keep in mind from these studies, however.
First, the researchers in these papers assumed that the degree of virilization of the genitalia was proportional to the amount of testosterone to which these individuals were exposed during fetal development. The big problem with that assumption is that it’s exactly that — an assumption! It’s never been proven, and with the current technology, it can’t be proven. Although the elevated progestins produced by fetuses with CAH can be measured in the amniotic fluid during gestation, the androgens metabolized from those progestins in the fetus cannot (Hines et al., 2002).
That leads to the second caveat, which is that the levels of progestins in the amniotic fluid of women bearing fetuses with CAH also cannot be used to indicate how much androgen the fetal external genitalia are exposed to because the efficiency of the conversion of progestins to androgens (including DHT) in the CAH fetuses can’t be measured, nor can the timing of the prenatal exposure to androgens as it relates to the timing of fetal sexual differentiation and development (White & Speiser, 2000).
Finally, the third caveat, which is the biggest caveat of them all, is: WE DO NOT KNOW WHEN MASCULINIZATION OF THE HUMAN BRAIN OCCURS (Cohen-Bendahan et al., 2005). Ooh, sorry, I didn’t mean to yell, but this point is important. It’s kinda tough to consider timing when we don’t know what that timing is.
Nevertheless, with the accumulation of reports that behavior is masculinized in females with CAH, presumably due to prenatal exposure to androgens, one would expect that gender identity would be masculinized as well.
To address that point, Dessens and coworkers (2005) performed a review of the literature of documented 46,XX females with genitalia masculinized from CAH. They found that of 250 individuals with CAH who were 46,XX and had been raised as females, 13 reported gender dysphoria (5.2%), four of which who wished to change their gender. Of those 13, however, the degree of masculinization of their genitalia was not correlated with the level of gender dysphoria.
Of the individuals who were 46,XX with CAH who had severely masculinized genitalia and were raised as males (33), four of them (12.1%) either identified as female or reported gender dysphoria. To look at those numbers flipped around, there were 288 total cases of 46,XX individuals, 42 of which appeared to have male gender identities (22.3%).
There were two main conclusions from this review of the literature :
–(1) Elevated androgen levels that induced varying degrees of masculinization of the genitalia did not similarly masculinize the brains of a majority of the individuals.
–(2) The percentage of the 46,XX individuals who did report a male gender identity was relatively higher than what we would see in a similar population of 46,XX females who don’t have CAH.
In other words, prenatal exposure to testosterone in individuals with CAH does appear to have some masculinizing effects on behavior and gender identity in 46,XX females.
Other Evidence for Testosterone as the “Male Gender Maker”
While we are considering CAH, what about other biological situations where testosterone is elevated in females during fetal development?
There is the study by Hines and coworkers (2002) where they measured maternal concentrations of testosterone during pregnancy and then correlated those with the male-typical behaviors in the daughters when they reached 3.5 years of age. They found that women with higher circulating testosterone during pregnancy had daughters with greater masculine gender role behavior than daughters from women with lower concentrations of testosterone in their blood.
I’m not sure how much weight to give to this study, however, because, according to the authors,
“… it should be noted that the relationship observed between T and behavior was small, accounting for only about 2% of the variance in the gender role behavior of preschool girls, leaving ample scope for influences from other factors.”
So what can we give weight to? Well, I always say, if you can’t learn from the humans, get yer sock monkeys out!
Would you believe that gender behavior has been studied in sock monkeys??! Well, what other choices do we have? It isn’t ethical to conduct these kinds of experiments on humans, so we’re left with sock monkeys, Macaca socketta.
Well, actually, it’s rhesus monkeys that I’m referring to, and they do appear to be a good species for studies of gender-typical behavior. According to Wallen and Hassett in their review (2009),
“Rhesus monkeys are an ideal model for their similarities with humans in prenatal development, sex differences in juvenile behavior, and their complex social structure.”
In fact, rhesus monkeys engage in similar sex-specific play, interests and behaviors as boys and girls. For example, when studying the toy preferences of juvenile rhesus monkeys, the young male monkeys showed significant preferences for wheeled toys (like toy cars and construction vehicles), similar to boys, whereas the young female monkeys had no strong preference between plush or wheeled toys, similar to girls (Hassett et al., 2008).
(The Wallen & Hassett review is an interesting look at behavior as affected by nature vs. nurture, which can more easily be studied in monkeys than in humans; it’s a free article for anyone who would like to read it – see below.)
So with the rhesus monkey, researchers have a way to study the timing of prenatal testosterone exposure on masculinization of genitalia and post-natal behavior.
In one study, female rhesus monkeys were treated with testosterone to expose their fetuses to androgens at two different time points during gestation (early versus late) and with two different durations of treatment (long – 25 days versus short – 15 days).
In the female offspring, those that had been exposed early in gestation had masculinized genitalia (similar to that seen in cases of CAH in humans), with no differences between long and short treatment durations. On the other hand, female offspring that had been exposed late in gestation, no matter which duration, had no masculinization of their external genitalia (references in Wallen & Hassett, 2009).
The effects on behavior, however, were reversed, so that female offspring exposed early in gestation had masculinized genitalia and female-typical behavior whereas those exposed later in gestation had typical female genitalia but male-typical behavior (although not as masculinized as unexposed male offspring). In addition, sexual mounting behavior was affected in the female offspring by the duration of the fetal exposure to testosterone but not by the timing (i.e. not by exposure early versus late in gestation).
The conclusion from this study was,
“The finding that male-like behaviors occurred even without male-like genitalia suggested that these behaviors emerged because of direct effects of prenatal testosterone on the brain and not because of differential social treatment by group members based on the genitals of the individual. These studies clearly demonstrated that not only were physical and behavioral differentiation separable, but the critical periods for sexual differentiation of behavior varied depending on the specific behavior.”
Summary and Next Post
Whew. That’s a lot of information to digest. At least, it seems that way. What is the bottom line regarding our question about brain masculinization and gender?
There are certainly some data that prenatal exposure of 46,XX females with androgens, as in the cases of CAH, results in some brain masculinization, as evidenced by a relatively greater incidences of male-typical behavior and male gender identity.
The studies with the monkeys allow us to assume that masculinization of the brain occurs at a time during fetal development that is later than development of the genitalia, which could explain a few things for the origins of transsexualism. The big gap we have from the monkey studies, however, is that we don’t know a whit about monkey gender identity. After all, behavior ≠ gender identity.
In this particular post of this seemingly never-ending series, we have looked at cases where androgens were abnormally high in genetic females. In the next (and hopefully final) post of the series, we’ll go back to where we started and look at a reduction in androgens or androgen signalling in genetic males and the effects on behavior and gender identity.
Acknowledgment: I would like to thank Jamison Green for kindly procuring and providing pdf files of some of the references cited in this post series.
Berenbaum, Duck, Bryk, 2000. Behavioral effects of prenatal versus postnatal androgen excess in children with 21-hydroxylase-deficient congenital adrenal hyperplasia. J Clin Endocrinol Metab 85:727-733 (free article)
Berenbaum & Bailey, 2003. Effects on gender identity of prenatal androgens with genital appearance: evidence from girls with congenital adrenal hyperplasia. J Clin Endocrinol Metab 88:1102-1106 (free article)
Hall, Jones, Meyer-Bahlburg et al., 2004. Behavioral and physical masculinization are related to genotype in girls with congenital adrenal hyperplasia. J Clin Endocrinol Metab 89:419-424 (free article)
Tanagho & McAninch (eds), 2000. Smith’s General Urology, 15th Edition, Lange Medical Books/McGraw Hill, New York.