OUP user menu

Who do we think we are? The brain and gender identity

Joe Herbert
DOI: http://dx.doi.org/10.1093/brain/awn257 3115-3117 First published online: 24 October 2008

No-one, wrote Frank Beach, a notable contributor to the experimental study of hormones and sexual behaviour, ever died from lack of sex. But the personal, social and legal aspects of sexual behaviour are a pervasive pre-occupation in all humans. The variety and vagaries of sex can have severe implications, and the existence of homosexuality and disorders of gender identity demand some sort of explanation (Bancroft, 2008). Neuroscience can ask itself, therefore, why it has contributed so little to understanding human sexuality. One reason is our overall ignorance about the brain, which hinders attempts to relate particular patterns of brain activity to an observable behaviour in a way that contributes to understanding. Another is the effect of sexual mores on the study of sexuality itself: studying sex is still considered a slightly risqué career, and made difficult by the politics, constraints and prejudices of human societies. It took the AIDS epidemic to convince many governments and funding bodies that studying sex was important and respectable. Most of our information on the neurobiology of sex comes from animal studies (Becker et al., 2005), but nearly all of what we know about variations in human sexuality, including hetero- and homo-sexuality, and disorders of gender identity (transsexualism), comes from clinical material, anecdotes or even fiction (the three overlap).

It is now well-known that sex-determination is heavily reliant upon the sry gene encoded on the Y chromosome. But genes, of course, do nothing themselves: they activate molecules that are the mechanisms, and prominent amongst these is testosterone. During development, the male fetus secretes testosterone—and production may continue into early post-natal life. This has dramatic effects on the internal reproductive organs, promoting the masculine arrangement. But, here we are more interested in what it does to the brain. Experimental studies show that both patterns of sexual behaviour and preference for females are directed towards the ‘male’ type by exposure to testosterone during this critical period (reviewed in Pardridge et al. 1982; Gorski, 2002). Whether this applies in all its detail to humans is still debated (Swaab, 2007). Even amongst closely related species, reproductive strategies (particularly female) are curiously varied, which makes extrapolations to humans a risky business (Baum, 2006). Nevertheless, there are experimental grounds for suggesting that variations in prenatal exposure to testosterone may influence later sexual preference (orientation) in man. But on an equally important issue, that of gender identity (whether we think of ourselves as male or female), the experimental literature is silent. For good reason: animals cannot report, as humans can, on their gender identity, and no-one has found an indirect way to reveal the gender identity of a monkey, let alone a rat. There are no experimental models of transsexuality. So, studies on the neuroscience of human transsexuality are limited to man alone.

The logic behind them is simple, though necessarily limited. Define a sex difference in some feature of the brain, preferably one known to be associated with sexual behaviour; then show that this difference is in the expected direction in those reporting gender dysphoria (the belief that one's ‘true’ or ‘core’ gender identity is the opposite of both chromosomal sex and bodily habitus). Since many gender dysphorics are either given or self-administer steroids more typical of the opposite sex, it is important to exclude these as a determinant of the atypical neural feature. A clear association between the neural and behavioural phenotypes suggests that this is causal. This is, essentially, what Garcia-Falgueras and Swaab report in the current issue (page 3132). The medial preoptic area of the hypothalamus (MPOA) is known to be involved in masculine patterns of sexual behaviour in rodents and monkeys, and to be sexually dimorphic, though, it should be pointed out, there are suggestions that it promotes sexual performance (the ability to copulate) more than motivation for masculine-type sex (reviewed in Balthazart and Ball, 2007). Garcia-Falgueras and Swaab find that part of this area, INAH3 in humans, is larger and has a greater number and density of neurons in males than females, and that male-to-female transsexuals (MtF) resemble females.

But these overall statistics tell only part of the story. Their data show considerable overlap between males, females and MtF. So, either the precision of their methods is not very high, or there is an indistinct relation between INAH3 and individual gender identity. This is odd: people usually have no doubt about their gender, though there are degrees of self-regarded ‘masculinity’ and ‘femininity’. There are other questions: why should more neurons in INAH3 predispose to male sexual identity? It seems unlikely that such simple methods will ever tell us much about the neural basis of human sexuality. The difficulty of interpreting measures such as volume is brought into focus by their first result: they report that brain weight is greater in males than females (not surprising), but that MtF subjects have intermediate values. What are we to make of this? And why are similar differences in INAH3 associated, in previous studies, with male homosexuality (Le Vay, 1991)? Gender identity does not predict sexual orientation: all permutations are possible (Hellman et al., 1981). Indeed, in the cases studied by Garcia-Falgueras and Swaab, only 7 out of 11 were orientated towards females (i.e. non-homosexual for their chromosomal sex, but homosexual for their gender identity). How could the same area of the brain regulate these two independent aspects of sexuality—though, it has to be said, there are indistinct boundaries between homosexuality, gender identity and other aspects of sexuality? For example, young boys that prefer to cross-dress are likely to be homosexual, not transsexual, in adult life (Green, 1985).

There is some disagreement about whether testosterone levels in the adult can alter the size or chemical structure of the MPOA in rodents (Davis et al., 1996; Ulibarri and Yahr, 1993). Garcia-Falgueras and Swaab try to eliminate this as a factor in their study by looking at INAH3 in five orchidectomized (non-transsexual) men. Their results are not very helpful: there are no differences in INAH3 volume or neuronal number between these subjects and the control or MtF males, or females. Neuropeptide Y (NPY)-stained sections show a reduction in volume compared to control males. The interval between castration and death was very short in most of their cases. It is curious that structures in the hypothalamus, a part of the brain closely associated with internal events in the body rather than those outside it, could determine sexual preference or gender identity. Previous studies by the same lab (Zhou et al., 1995), as well as others, show structural differences associated with varieties of sexual behaviour and attitude in other areas of the brain, for example the bed nucleus of the stria terminalis (BST), which is linked both with the hypothalamus and the amygdala. The latter has more direct access to external stimuli, such as those determining sexual attractiveness. It seems that there may be a neural ‘system’, rather than a single nodal area, that determines or influences the different parameters of human sexuality. Add to this the evident contribution of the cerebral cortex, involved in social awareness, attitudes, decision-making and the use of sex as a social instrument and we can see that the boundaries of the ‘sexual’ brain are as indistinct as the definition of sex itself.

But the significance of the Garcia-Falgueras and Swaab paper is really as much political and even legal as it is neuroscientific. If there are demonstrable and functionally relevant features in the brain that underlie beliefs or proclivities that determine a person's behaviour from an early age, and may be immutable, then the case for a redefinition of gender and for reassignment surgery in transsexuals is strengthened. There is still conflict between those who consider human sexuality to be either biologically or socially determined (the two, of course, are not mutually exclusive—as some seem to think—and must work together). Future studies may uncover more exact neurobiological explanations for the strange human phenomenon of gender dysphoria, and this might help our understanding of more mainstream sexuality (Brunetti et al., 2008). Our view of homosexuality was altered by Le Vay's findings on INAH3 (1991). One day, someone may uncover a neural feature (which we might call an abnormality) underlying other, less acceptable, traits such as paedophilia: what then? There are already suggestions that head injury before the age of 13 years is more common than usual in paedophiles (Blanchard et al., 2003).


Wellcome Trust.


I thank J. Bancroft, R. Green and M. Hines for their help with this commentary.