ArticlePDF AvailableLiterature Review

Abstract and Figures

It is believed that during the intrauterine period the fetal brain develops in the male direction through a direct action of testosterone on the developing nerve cells, or in the female direction through the absence of this hormone surge. According to this concept, our gender identity (the conviction of belonging to the male or female gender) and sexual orientation should be programmed into our brain structures when we are still in the womb. However, since sexual differentiation of the genitals takes place in the first two months of pregnancy and sexual differentiation of the brain starts in the second half of pregnancy, these two processes can be influenced independently, which may result in transsexuality. This also means that in the event of ambiguous sex at birth, the degree of masculinization of the genitals may not reflect the degree of masculinization of the brain. There is no proof that social environment after birth has an effect on gender identity or sexual orientation. Data on genetic and hormone independent influence on gender identity are presently divergent and do not provide convincing information about the underlying etiology. To what extent fetal programming may determine sexual orientation is also a matter of discussion. A number of studies show patterns of sex atypical cerebral dimorphism in homosexual subjects. Although the crucial question, namely how such complex functions as sexual orientation and identity are processed in the brain remains unanswered, emerging data point at a key role of specific neuronal circuits involving the hypothalamus.
Content may be subject to copyright.
Provided for non-commercial research and educational use only.
Not for reproduction, distribution or commercial use.
This chapter was originally published in the book Sex Differences in the Human Brain,
Their Underpinnings and Implications. The copy attached is provided by Elsevier for the
author’s benefit and for the benefit of the author’s institution, for non-commercial
research, and educational use. This includes without limitation use in instruction at
your institution, distribution to specific colleagues, and providing a copy to your
institution’s administrator.
All other uses, reproduction and distribution, including without limitation commercial
reprints, selling or licensing copies or access, or posting on open internet sites, your
personal or institution’s website or repository, are prohibited. For exceptions,
permission may be sought for such use through Elsevier’s permissions site at:
From: Ivanka Savic, Alicia Garcia-Falgueras and Dick F. Swaab,
Sexual differentiation of the human brain in relation to gender identity and sexual
In: Ivanka Savic, editor,
Sex Differences in the Human Brain, Their Underpinnings and Implications.
Academic Press 2010, p. 41.
ISBN: 978-0-444-53630-3
ÓCopyright 2010 Elsevier Inc.
Academic Press.
Author's personal copy
I. Savic (Ed.)
Progress in Brain Research, Vol. 186
ISSN: 0079-6123
Copyright © 2010 Elsevier B.V. All rights reserved.
Sexual differentiation of the human brain in relation
to gender identity and sexual orientation
Ivanka Savic
, Alicia Garcia-Falgueras
and Dick F. Swaab
Department of Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden
Netherlands Institute for Neuroscience, an Institute of the Royal Netherlands Academy of Arts and Sciences,
Amsterdam, The Netherlands
Medical Psychology Unit, Institute of Neuroscience, Autonomous University of Barcelona, Bellaterra, Barcelona, Spain
Abstract: It is believed that during the intrauterine period the fetal brain develops in the male
direction through a direct action of testosterone on the developing nerve cells, or in the female
direction through the absence of this hormone surge. According to this concept, our gender identity
(the conviction of belonging to the male or female gender) and sexual orientation should be
programmed into our brain structures when we are still in the womb. However, since sexual
differentiation of the genitals takes place in the first two months of pregnancy and sexual
differentiation of the brain starts in the second half of pregnancy, these two processes can be
influenced independently, which may result in transsexuality. This also means that in the event of
ambiguous sex at birth, the degree of masculinization of the genitals may not reflect the degree of
masculinization of the brain.
There is no proof that social environment after birth has an effect on gender identity or sexual
orientation. Data on genetic and hormone independent influence on gender identity are presently
divergent and do not provide convincing information about the underlying etiology. To what extent
fetal programming may determine sexual orientation is also a matter of discussion. A number of studies
show patterns of sex atypical cerebral dimorphism in homosexual subjects. Although the crucial question,
namely how such complex functions as sexual orientation and identity are processed in the brain remains
unanswered, emerging data point at a key role of specific neuronal circuits involving the hypothalamus.
Keywords: Gender identity; Homosexuality; Human brain; Sexual orientation; Sexual differentiation;
Corresponding author.
Tel.: þ31 20 5665500; Fax: þ31 20 5666121;
DOI: 10.1016/B978-0-444-53630-3.00004-X 41
Author's personal copy
General concepts
Gender identity and sexual orientation represent
two fundamental functions in human neurobiol-
ogy. These functions have hitherto mainly been
discussed in relation to the specific signs of sexual
dimorphism in the brain and the potential
mechanisms thereof. By mapping differences
between men and women in cerebral anatomy,
function, and neurochemistry, neuroscientists are
trying to identify sex typical and sex atypical
actors in transsexual and homosexual individuals.
This has been done in postmortem analyses
of the brain, and investigations of neuronal
anatomy, connectivity, and function by means of
positron emission tomography (PET) and mag-
netic resonance imaging (MRI). The extracted
networks are then mapped onto those known to
be related to sexual behavior in animals to
formulate biological underpinnings of homo-
and transsexuality in humans. This widely used
approach has several difficulties with this
approach: (1) gender identity cannot be investi-
gated in animals; (2) sexual behavior in animals is
reflex-like and cannot simply be translated to
sexual orientation and attraction in humans;
(3) reliable sex differences in the human brain
require investigations of large populations and
have only recently been demonstrated reliably;
(4) the majority of studies on sex differences do
not account for sexual orientation of the investi-
gated participants; (5) studies of homo- and
transsexual persons are very limited, and only
few comparisons have hitherto been presented
between homo- and transsexual subjects.
An alternative and parallel approach is pin-
pointing the specific neuronal networks related
to gender identity and sexual orientation, ana-
lyzing the factors programming these networks
and possible differences between control, homo-,
and transsexual subjects. Emerging fMRI and
PET studies suggest that sexual arousal is
mediated by specific core neuronal networks,
which may be also involved in sexual
Sexual organization and activation of the human
The process of sexual differentiation of the
brain brings about permanent changes in brain
structures and functions via interactions of the
developing neurons with the environment,
understood in its widest sense. The environment
of a developing neuron is formed by the sur-
rounding nerve cells and the childs circulating
hormones, as well as the hormones, nutrients,
medication, and other chemical substances
from the mother and the environment that
enter the fetal circulation via the placenta.
Along with the genetic code, all these factors
may have a lasting effect on the sexual differen-
tiation of the brain.
The testicles and ovaries develop in the sixth
week of pregnancy. This occurs under the influ-
ence of a cascade of genes, starting with the sex-
determining gene on the Y chromosome (SRY).
The production of testosterone by a boys testes is
necessary for sexual differentiation of the sexual
organs between weeks 6 and 12 of pregnancy. The
peripheral conversion of testosterone into dihy-
drotestosterone is essential for the formation of a
boys penis, prostate, and scrotum. Instead, the
development of the female sexual organs in the
womb is based primarily on the absence of andro-
gens (Swaab et al., 2003).
Once the differentiation of the sexual organs
into male or female is settled, the next thing that
is differentiated is the brain, under the influence,
mainly, of sex hormones on the developing brain
cells. The changes (permanent) brought about in
this stage have organizing effects; later, during
puberty, the brain circuits that developed in the
womb are activated by sex hormones. This para-
digm of sexual differentiation of the brain was
coined by Phoenix et al. (1959) and has dominated
the view on cerebral sex dimorphism during the
last decades.
The fetal brain is protected against the effect of
circulating estrogens from the mother by the pro-
tein α-fetoprotein, which is produced by the fetus
Author's personal copy
and binds strongly to estrogens but not to testos-
terone (Bakker et al., 2006, 2008). However, not
only estrogens reach the brain via circulation, but
the brain itself is capable of producing estrogens.
In human beings testosterone may thus not only
have a direct effect on a masculine brain, but, once
converted into estrogens by aromatase, may also
act on developing neurons. In addition, there are
sex differences in brain steroid receptor distribu-
tion not only in adulthood (Ishunina and Swaab,
2008; Kruijver and Swaab, 2002; Kruijver et al.,
2001; Swaab et al., 2001) but also during develop-
ment (Chung, 2003), which may be genetically
determined. In addition, in rat hormone receptor
genes a sex difference in methylation pattern
occurs during development (Schwarz et al.,
2010). In rats, the formation of estradiol in the
brain by aromatization of circulating testosterone
is the most important mechanism for virilization of
the brain (Gorski, 1984), but, as seen below, it
does not determine human gender identity or sex-
ual orientation.
There may also be direct genetic effects that
affect the sexual differentiation of the brain with-
out involving the sex hormone receptors.
Sex hormones and human brain development
During fetal development, the brain is influenced
by sex hormones such as testosterone, estrogens,
and progesterone (Swaab, 2004). From the earliest
stages of fetal brain development, many neurons
throughout the entire nervous system already
have receptors for these hormones (Chung,
2003). The early development of boys shows two
periods during which testosterone levels are
known to be high. The first surge occurs during
mid-pregnancy: testosterone levels peak in the
fetal serum between weeks 12 and 18 of pregnancy
(Finegan et al., 1989) and in weeks 3441 of preg-
nancy the testosterone levels of boys are ten times
higher than those of girls (De Zegher et al., 1992;
Van de Beek et al., 2009). The second surge takes
place in the first three months after birth. At the
end of pregnancy, when the α-fetoprotein level
declines, the fetus is more exposed to estrogens
from the placenta, this exposure inhibiting
the hypothalamushypophysealgonadal axis of
the developing child. Loss of this inhibition once
the child is born causes a peak in testosterone in
boys and a peak in estrogens in girls (Quigley,
2002). The testosterone level in boys at this time
is as high as it will be in adulthood, although a
large part of the hormone circulates bound. Also
at this time the testosterone level is higher in boys
than in girls. During these two periods, therefore,
girls do not show high levels of testosterone.
These fetal and neonatal peaks of testosterone,
together with the functional steroid receptor activ-
ity, are, according to the current dogma, thought
to fix the development of structures and circuits in
the brain for the rest of a boys life (producing
programming or organizing effects). Later,
the rising hormone levels that occur during pub-
erty activate circuits and behavioral patterns
that were built during development, in a masculi-
nized and de-feminized direction for male brains
or in a feminized and de-masculinized direction
for female brains.
The brain structure differences that result
from the interaction between hormones and
developing brain cells are thought to be the
major basis of sex differences in a wide spectrum
of behaviors, such as gender role (behaving as a
man or a woman in society), gender identity (the
conviction of belonging to the male or female
gender), sexual orientation (heterosexuality,
homosexuality, or bisexuality), and sex differ-
ences regarding cognition, aggressive behavior,
and language organization. Factors that interfere
with the interactions between hormones and the
developing brain systems during development in
the womb may permanently influence later
As sexual differentiation of the genitals takes
places much earlier in development (i.e., in the
first two months of pregnancy) than sexual differ-
entiation of the brain, which starts in the second
half of pregnancy and becomes overt upon
Author's personal copy
reaching adulthood, these two processes may be
influenced independently of each other. In rare
cases, these two processes may be incongruent,
providing one possible mechanism for transsexu-
ality, that is, people with male sexual organs who
feel female or vice versa. It also means that in the
event of an ambiguous sex at birth, the degree of
masculinization of the genitals may not always
reflect the degree of masculinization of the brain
(Hughes et al., 2006; Swaab, 2004, 2008). In addi-
tion, gender identity may be determined by pre-
natal hormonal influences, even though the
prenatal hormonal milieu might be inadequate
for full genital differentiation (Reiner, 1999).
Programmed gender identity is irreversible
The irreversibility of programmed gender identity
is clearly illustrated by the sad story of the John
JoanJohn case (i.e., the case of David Reimer).
In the 1960s and 1970s, in the context of the con-
cept of behaviorism, it was postulated that a child
is born as a tabula rasa and is subsequently forced
in the male or female direction by societys con-
ventions. Although it is true that, in humans, self-
face recognition appears to emerge at around 18
months of age (Keenan et al., 2000) and that by
the age of 23 years children are able to correctly
label themselves and others according to gender
(Ahmed et al., 2004), there is no evidence that
external or social events might modify these pro-
cesses. However, J. Money argued that Gender
identity is sufficiently incompletely differentiated
at birth as to permit successful assignment of a
genetic male as a girl. Gender identity then differ-
entiates in keeping with the experiences of rear-
ing (Money, 1975). This view had devastating
results in the JohnJoanJohn case (Colapinto,
2001). Money maintained that gender imprinting
does not start until the age of 1 year, and that its
development is well advanced by the age of 34
years (Money and Erhardt, 1972). This was,
indeed, the basis for the decision to make a girl
out of an 8-month-old boy who lost his penis due
to a mistake during minor surgery (i.e., an opera-
tion to correct phimosis). The testicles of this child
were removed before he reached the age of 17
months in order to facilitate feminization. The
child was dressed in girls clothes, received psy-
chological counseling, and was given estrogens in
puberty. According to Money, this child devel-
oped as a normal female. However, Milton Dia-
mond later made it clear that this had not been the
case at all. In adulthood, this child changed back
to male, married, and adopted several children
(Diamond and Sigmundson, 1997). Unfortunately,
he had a troubled life and committed suicide in
2004. This story illustrates the enormous program-
ming influence of the intrauterine period on gen-
der. Other cases have been described in the
literature (Bradley et al., 1998), due to enzymatic
disorders (al-Attia, 1996; Cohen-Kettenis, 2005;
Praveen et al., 2008) or to cloacal exstrophy
(Reiner, 2005), that support the existence of
early permanent programming of brain sex by
biological factors and androgen exposure, rather
than by social environment and learning (Jürgensen
et al., 2007; Swaab, 2004).
The mechanism of sexual differentiation of the
brain: neurobiological factors
In male rats, testosterone is turned into estrogens
by local aromatization in the brain, and these estro-
gens then masculinize certain brain areas. This
finding agrees with the observation that, in partially
androgen insensitive (testosterone feminized
Tfm) male rats, no reversion of the sex difference
was present in the preoptic area (Gorski, 1984)
and the bed nucleus of the stria terminalis
(Garcia-Falgueras et al., 2005). These animals
retained a male neuroanatomy. Other brain
nuclei, such as the posteromedial amygdala, the
ventromedial hypothalamus, and the locus coeru-
leus were, however, feminized in Tfm male rats
(Morris et al., 2005; Zuloaga et al., 2008).
In humans, however, the main mechanism
appears to involve a direct effect of testosterone
Author's personal copy
on the developing brain. Complete androgen
insensitivity syndrome is caused by mutations in
the receptor gene for androgens. Despite their
genetic (XY) masculinity, affected individuals
develop as phenotypical women and experience
heterosexual sexual orientation, fantasies, and
experiences, without gender problems (Wisniewski
et al., 2000). On the other hand, when a male
fetus has a 5α-reductase-2 or 17β-hydroxy-steroid-
dehydrogenase-3 deficiency preventing peripheral
testosterone from being transformed into dihy-
drotestosterone, a girl with a large clitoris is
born. These children are generally raised as girls.
However, when testosterone production
increases in these XY children during puberty,
this clitoris grows to penis size, the testicles
descend, and the childs build begins to masculi-
nize and become muscular. Despite the fact that
these children are initially raised as girls, the
majority (60%) change into heterosexual males
(Cohen-Kettenis, 2005; Hughes et al., 2006;
Imperato-McGinley et al., 1979; Praveen et al.,
2008; Wilson et al., 1993), apparently due to the
organizing effect of testosterone on early brain
development. Boys who are born with a cloacal
exstrophythat is, with bladder exstrophy and a
partly or wholly absent penisare usually chan-
ged into girls immediately after birth. A survey
showed that in adulthood only 65% of these chil-
dren who were changed into girls continued to
live as girls, and when individuals with gender
dysphoria were excluded, the figure dropped to
47% (Meyer-Bahlburg, 2005; Reiner and
Gearhart, 2004). From these examples, it appears
that the direct action of testosterone on the
developing brain in boys and the lack of it in
the developing brain in girls are crucial factors
in the development of male and female gender
identity and sexual orientation, although other
sexually dimorphic functions stillneedtobe
investigated in these people. Conversely, studies
on cloacal exstrophy suggest that the postnatal
testosterone peak is not crucial for gender
identity development, given that these children
generally undergo operation shortly after birth.
Recent data show that environmental com-
pounds during early development may interfere
with sexual differentiation of the human brain.
Plastic softeners, that is, phthalate esters, are
pervasive environmental chemicals with anti-
androgenic effects. Exposure to these compounds
is accompanied by reduced masculine play in boys
(Swan et al., 2010). Higher prenatal polychlori-
nated biphenyls (PCB) levels were related with
less masculine play in boys, while higher prenatal
dioxin levels were associated with more feminized
play in boys as well as in girls (Vreugdenhil et al.,
2002). The effect of such environmental endocrine
disruptors on sexual differentiation of brain sys-
tems should be further studied in future.
Sex differences in the human brain
A sex difference in brain weight is already present
in children from the age of 2 years (Swaab and
Hofman, 1984) and sex differences can thus be
expected throughout the brain from early in
development. In the adult human brain structural
sex differences can be found from the macroscopic
level (Goldstein et al., 2001) down to the ultrami-
croscopic level (Alonso-Nanclares et al., 2008).
Functionally, too, a large number of sex differ-
ences in different brain regions have recently
been described (Allen et al., 2003; Amunts et al.,
1999, 2007; Savic, 2005; Savic and Lindstroom,
2008). Sexual differentiation of the human brain
is also expressed in behavioral differences, includ-
ing sexual orientation (homo-, bi-, and hetero-
sexuality) and gender identity (Allen and Gorski,
1992; Hines, 2003; LeVay, 1991; Swaab, 2003), and
in differences at the level of brain physiology and
in the prevalence of neurological and psychiatric
disorders (Bao and Swaab, 2007; Savic and Engel,
1998; Swaab, 2003). In the current review we focus
on the sex differences in the human hypothalamus
and adjacent areas.
When observed by Swaabs group, the struc-
tural difference in the intermediate nucleus
of the human hypothalamus (InM) (Braak and
Author's personal copy
(a) (b) (e)
Somatostatin neuron number in BSTc ( 103)
50 FMT
IC IC 40 A S2
(c) (d)
20 M2
T6 S7
S6 T1
10 P
Braak, 1987; Brockhaus, 1942; Koutcherov et al.,
2007)was foundtobe2.5 timeslargerinmen than
in women and to contain 2.2 times as many cells
(Swaab and Fliers, 1985). This InM nucleus was at
first termed the sexually dimorphic nucleus of
the preoptic area (SDN-POA) (Swaab and
Fliers, 1985). In the preoptic area, Allen et al.
(1989) described four interstitial nuclei of the
anterior hypothalamus (INAH-1 to 4, while
INAH-1 is identical to the InM/SDN-POA) and
found a larger volume of the INAH-3 and INAH-
2 subdivisions in men compared to women
(respectively 2.8 and 2 times greater). The fact
that they could not find a sex difference in
INAH-1 (InM), as found by Swaabsgroup
(Swaab and Fliers, 1985), could be fully explained
by the strong age effect on the sex differences of
this nucleus (Swaab, 2003; Swaab and Hofman,
1988). In fact, the sex difference develops only
after the age of 5 years and disappears tempora-
rily after the age of 50 years (Swaab and Fliers,
1985; Swaab et al., 1992). Further analysis of
INAH-1 galanin cell population in the transsexual
people and controls is ongoing and confirms the
presence of a clear sex difference in adult controls
up to 45 years of age.
The uncinate nucleus (Un) was localized and
delineated using three different stainings, that is,
thionin, neuropeptide-Y, and synaptophysin.
We found sex differences in volume and neuron
number in the INAH-3 subdivision while no
differences were found for INAH-4 (Fig. 1;
Garcia-Falgueras and Swaab, 2008). The presence
of a sex difference in INAH-3 volume fully agreed
with previously reported data (Allen et al., 1989;
Byne et al., 2000, 2001; LeVay, 1991), as did the
Fig.1. Representative immunocytochemical staining of the somatostatin neurons and fibers in the bed nucleus of the stria terminalis,
central subdivision (BSTc) of a reference man (a), a reference woman (b), a homosexual man (c), and a male-to-female transsexual
(d). *, Blood vessels; LV, lateral ventricle; IC, internal capsule. Bar represent 0.35 mm. (e) Graph of BSTc number of neurons in
different groups according to sexual orientation and gender identity (M, heterosexual male reference group; HM, homosexual male
group; F, female reference group; TM, male-to-female transsexual people; T1-T6, transsexual subjects; A, AIDS patient;
P, postmenopausal woman; S7, Gender Identity Disorder subject). The sex hormone disorder patients S1, S2, S3, S5, S6, and M2
indicate that changes in sex hormone levels in adulthood do not change the neuron numbers of the BSTc. There is a statistical
difference between the M and the TM group (p < 0.04) while no difference was between the heterosexual male reference group and the
homosexual group. The female to male transsexual (FMT) subject is in the male range. From Kruijver et al. (2000) with permission.
Author's personal copy
sex difference for the number of neurons in
INAH-3. A number of different names have
been used to refer to the two Un subnuclei
(Garcia-Falgueras and Swaab, 2008): (1) periven-
tricular and uncinate nucleus (the former closer to
the third ventricle than the latter) (Braak and
Braak, 1987); (2) INAH-4 (closer to the third ven-
tricle than the INAH-3) (Allen et al., 1989); and,
most recently, (3) lateral and medial subdivisions of
the Un (Koutcherov et al., 2007). In view of the
evidence provided by neurochemical markers such
as neuropeptide-Y and synaptophysin and the fact
that they appear as one structure in some subjects,
there are indeed arguments in favor of considering
these two subdivisions a single structure called the
Un. It has been suggested the INAH-3 was the
homologue of the rat central nucleus of the medial
preoptic area (Koutcherov et al., 2007) that, in this
animal, is clearly related to the brain network
for input and output of male sexual behavior
(Schober and Pfaff, 2001; Swaab, 2004). On the
other hand, the INAH-1 (InM) may be a candidate
for that homology. Further research with specific
markers is required to solve this issue.
Moreover, similar to the BSTc, the INAH-3 was
found in male-to-female (MtF) transsexual people to
be small (of female size and cell number), while the
INAH-4 subdivision did not show gender-related
differences, or any morphological sex difference
between men and women (Fig. 1; Garcia-Falgueras
and Swaab, 2008). Other sex differences have been
found in the human anterior commissure, the inter-
thalamic adhesion and in the corpora mammillaria
(Allen and Gorski, 1991; Swaab, 2003).
Sex hormone receptors and neurosteroids
Sex hormone receptors, too, are expressed in a
sexually dimorphic way in the human hypothala-
mus and adjacent areas.
In most hypothalamic areas that show androgen
receptor staining, nuclear staining, in particular, is
less intense in women than in men. The strongest sex
difference was found in the lateral and the medial
mammillary nucleus (MMN; Fernandez-Guasti
et al., 2000). The mammillary body complex is
known to be involved in several aspects of sexual
behavior, such as arousal of sexual interest and
penile erection (Fernandez-Guasti et al., 2000;
MacLean and Ploog, 1962; Swaab, 2003). In
addition, a sex difference in androgen receptor
staining was present in the horizontal diagonal
band of Broca, SDN-POA, medial preoptic area
(mPOA), dorsal and ventral zone of the periven-
tricular nucleus (PVN), supraoptic nucleus (SON),
ventromedial hypothalamic nucleus, and infundib-
ular nucleus (INF). However, no sex differences
were observed in androgen receptor staining in the
adult bed nucleus of the stria terminalis (BSTc),
the nucleus basalis of Meynert, and the islands
of Calleja (Fernandez-Guasti et al. 2000).
No differences related to male sexual orientation
were found in nuclear androgen receptor activity in
the mammillary complex, this activity not being
found to differ in heterosexual men compared with
homosexual men, but it was significantly stronger in
men than in women. A female-like pattern was
found in 26- and 53-year-old castrated men and in
intact old men. These data indicate that the amount
of nuclear receptor staining in the adult mammillary
complex is dependent on the circulating levels of
androgens rather than on gender identity or sexual
orientation. This idea is supported by the findings
that a male-like pattern of androgen receptor stain-
ing was found in a 36-years-old bisexual non-
castrated MtF transsexual (T6) and a heterosexual
virilized woman aged 46 (Kruijver et al., 2001), while
a female-like pattern for INAH-3 volume and num-
ber of cells was found in the former patient (T6)
(Garcia-Falgueras and Swaab, 2008).
Various sex differences have been observed for
estrogen receptor α (ERα) staining in the hypotha-
lamus and adjacent areas of young adult human
subjects. More intense nuclear ERα immunoreac-
tivity was found in young men compared with
young women, for example, in the SDN-POA,
the SON, and the PVN. Women showed a stron-
ger nuclear ERα immunoreactivity in the supra-
chiasmatic nucleus (SCN) and MMN. No sex
Author's personal copy
differences in nuclear ERα staining were found in,
for example, the bed nucleus of the stria terminalis
(BSTc), the islands of Calleja (Cal), or the INF.
More intense nuclear ERβ staining was found in
men in, for example, the neurons of the BSTc,
the islands of Calleja, and the InM/SDN-POA.
Women showed more nuclear ERβ staining in the
SCN, the SON, the PVN, the INF, and the MMN
(Ishunina et al., 2007). Observations in subjects with
abnormal hormone levels showed, in most areas,
ERβ immunoreactivity distribution patterns that
were consistent with the level of circulating estro-
gens, suggesting that the majority of the reported
sex differences in ERβ immunoreactivity are
activational rather than organizational in
nature (Kruijver et al., 2002, 2003).
In the BSTc, differences in sex hormone recep-
tors such as ERα,ERβ, androgen receptor (AR),
and progesterone receptor (PR) are present from
fetal age onward. More nuclear ERβ was observed
in females than in males during the fetal/neonatal
ages, whereas there were no overt sex differences
in the other three sex hormone receptors detected.
In adult men, ERα and PR immunoreactivity was
more pronounced in the BSTc of men than in
women (Chung, 2003). Hence, the sensitivity of
the BSTc for the different sex hormones depends
strongly on sex and age.
There is a vast array of factors that may lead to
gender problems (Table 1). Twin and family
research has shown that genetic factors play a
part (Coolidge et al., 2002; Gómez-Gil et al.,
2010a; Hare et al., 2009; van Beijsterveldt et al.,
2006). Rare chromosomal abnormalities may lead
to transsexuality (Hengstschläger et al., 2003) and
it was found that polymorphisms of the genes for
ERα and ERβ, AR repeat length polymorphism
and polymorphisms in the aromatase or CYP17
gene also produced an increased risk (Bentz et al.,
2008; Hare et al., 2009; Henningsson et al., 2005).
Abnormal hormone levels during early devel-
opment may play a role, as suggested by the high
frequency of polycystic ovaries, oligomenorrhea
and amenorrhea in female-to-male (FtM) trans-
sexuals. This observation suggests early intrauter-
ine exposure of the female fetus to abnormally
Table 1. Prenatal factors that influence gender identity (the conviction of being a man or a woman) and that may result in transsexuality
Genetic factors Rare chromosomal disorders (Hengstschläger et al., 2003)
Twin studies (van Beijsterveldt et al., 2006; Coolidge et al., 2002; Gómez-Gil et al., 2010a; Hare et al., 2009)
Polymorphisms in ERβ, androgen receptor, and aromatase genes (Bentz et al., 2008; Hare et al., 2009;
Henningsson et al., 2005)
Hormones Phenobarbital/diphantoin taken by pregnant mother (Dessens et al., 1999)
Hormones, cloacal exstrophy (Meyer-Bahlburg, 2005; Reiner and Gearhart, 2004)
5α-reductase-2 or 17β-hydroxy-steroid-dehydrogenase-3 deficiency (Cohen-Kettenis, 2005; Hughes et al., 2006;
Imperato-McGinley et al., 1979; Praveen et al., 2008; Wilson et al., 1993)
Girls with CAH (Dessens et al., 2005; Meyer-Bahlburg et al., 1995, 1996; Zucker et al., 1996)
Complete androgen insensitivity syndrome results in XY heterosexual females with feminine identity
(Wisniewski et al., 2000)
DES sons: 25% gender problems (
Immune Fraternal birth order (Gómez-Gil et al., 2010b)
Social factors Postnatally no evidence (Cohen-Kettenis and Gooren, 1999; Colapinto, 2001; Diamond and Sigmundson, 1997;
Swaab, 2004)
Abbreviations: CAH, congenital adrenal hyperplasia; DES, diethylstilbestrol.
Author's personal copy
high levels of testosterone (Padmanabhan et al.,
2005). A recent study did not confirm a signifi-
cantly increased prevalence of polycystic ovary
syndrome. However, there was a significantly
higher prevalence of hyperandrogynism in FtM
transsexuals, also indicating the possible involve-
ment of high testosterone levels in transsexuality
(Mueller et al., 2008). A girl with congenital adre-
nal hyperplasia (CAH), who has been exposed to
extreme levels of testosterone in utero, will also
have an increased chance of becoming transsex-
ual. Although the likelihood of transsexuality
developing in such cases is 3001000 higher than
normal, the risk for transsexuality in CAH is still
only 13% (Zucker et al., 1996), whereas the
probability of serious gender problems is 5.2%
(Dessens et al., 2005). The consensus is, therefore,
that girls with CAH should be raised as girls, even
when they are masculinized (Hughes et al., 2006).
Epileptic women who were given phenobarbital
or diphantoin during pregnancy also have an
increased risk of giving birth to a transsexual
child. Both these substances change the metabo-
lism of the sex hormones and can act on the sexual
differentiation of the childs brain. In a group
of 243 women who had been exposed to such
substances during pregnancy, Dessens et al.
(1999) found three transsexual children and a few
others with less radical gender problems; these are
relatively high rates for such a rare condition. On
the DES (diethylstilbestrol, an estrogen-like
substancesee later) childrens website they
claimed that transsexuality occurs in 35.5% and a
gender problem in 14% of the DES cases (links
GIRES and DES SONS webpages). This is alarm-
ing, but needs, of course, to be confirmed in a formal
study. There are no indications that postnatal social
factors could be responsible for the occurrence of
transsexuality (Cohen-Kettenis et al., 1998).
In addition, homosexual MtF transsexual peo-
ple were found to have a later birth order and
more brothers than sisters (Gómez-Gil et al.,
2010b), suggesting the presence of immunological
processes during pregnancy directed toward pro-
ducts of the Y chromosome.
It should be noted that only in 23% of cases does a
childhood gender problem lead to transsexuality in
adulthood. With regard to sexual orientation,
the most likely outcome of childhood gender
identity disorder is homosexuality or bisexuality
(Cohen-Kettenis and Gooren, 1999; Coolidge et al.,
2002; Wallien and Cohen-Kettenis, 2008). Moreover
for the diagnosis of transsexuality other disorders
inducing temporal transsexual desiressuch as
bipolar psychosis, schizophrenia, and personality
disordersshould be excluded (à Campo et al.
2003; Habermeyer et al., 2003; Mouaffak et al., 2007).
Transsexuality and the brain
The theory on the origins of transsexuality is
based on the fact that the differentiation of sexual
organs takes place during the first couple of
months of pregnancy, before the sexual differen-
tiation of the brain. As these two processes have
different timetables, it is possible, in principle, that
they take different routes under the influence of
different factors. If this is the case, one might
expect to find, in transsexuals, female structures
in a male brain and vice versa, and indeed, we
did find such reversals in the central nucleus of
the BSTc and in the INAH-3 (Figs. 1 and 2), two
brain structures that, in rats, are involved in many
aspects of sexual behavior. However, a gender
identity test for rats does not exist, and this hypo-
thesis can therefore be studied only in humans.
We found a clear sex difference in the human
BSTc and INAH-3. In men, the BSTc area was
twice that found in women and contained twice as
many somatostatin neurons (Garcia-Falgueras
and Swaab, 2008; Kruijver et al., 2000; Zhou
et al., 1995). The same was true of the INAH-3,
which was found to be 1.9 times larger in men than
in women and to contain 2.3 as many neurons
(Fig. 2; Garcia-Falgueras and Swaab, 2008). In
relation to sexual orientation, no difference was
found in the size or number of neurons in the
BSTc area, while for the INAH-3 the volume has
previously been found to be related to sexual
Author's personal copy
(a) (b)
3V 3V
(c) (d)
Number of neurons in INAH-3
4000 T4
S7 S3
T6 S9
3V T5T8T9 T10S10
M F MtF CAS PreM PostM
Fig. 2. Representative immunocytochemical staining of the NPY innervation of the uncinate nucleus (INAH-3 and INAH-4) of a
reference man (a), a reference woman (b), and a male-to-female transsexual (c). Note that the size is larger in the male group (a) than
in the other two groups (b and c). Bar represent 500 mm. (d) Distribution of the INAH-3 number of neurons among different groups
according to their gender identity and hormonal changes in adulthood. M, control male group; F, control female group; MtF, male-to-
female transsexual group; CAS, castrated male group; PreM, premenopausal women; PostM, postmenopausal women T1-T10,
transsexual subjects; S3, S5, S8, S9, S10, castrated subjects because of prostate cancer. Bars represent means and standard errors
of the mean. Statistically significant differences were found between men (M) and women (F) (p < 0.029) and between men (M) and
male-to-female transsexual MtF groups (p < 0.002). The female to male transsexual subject (FTM), in the male group, had a
masculine INAH-3 number of neurons and the untreated S7 subject, in the MtF group, had a similar number of neurons to the
other transsexuals examined. (a, b, c and d) Adapted from Garcia-Falgueras and Swaab (2008) with permission.
orientation, being larger in heterosexual than in and his BSTc and INAH-3 indeed turned out to
homosexual men (Byne et al., 2001; LeVay, 1991). have all the male characteristics. We were able
In MtF transsexuals, we found a completely female to exclude the possibility that the reversal of sex
BSTc and INAH-3. Until now, we have only been differences in the BSTc and INAH-3 were
able to obtain material from one FtM transsexual, caused by changing hormone levels in adulthood
Author's personal copy
(Garcia-Falgueras and Swaab, 2008; Kruijver et al.,
2000; Zhou et al., 1995), and it therefore seems that
we are dealing with a developmental effect. Our
observations thus support the above-mentioned
neurobiological theory about the origin of trans-
sexuality. The size of the BSTc and the INAH-3
and their number of neurons match the gender
that transsexual people feel they belong to, and
not the sex of their sexual organs, birth certificate
or passport. Unfortunately, the sex difference in the
BSTc volume does not become apparent until early
adulthood (Chung et al., 2002), meaning that this
nucleus cannot be used for early diagnosis of
One person we studied had untreated male gen-
der dysphoria (S7), took no hormones and kept
his transsexual feelings under wraps. He appeared
to have a large INAH-3 volumein the male
rangebut a female INAH-3 number of neurons
(Garcia-Falgueras and Swaab, 2008; Fig. 2d) and
a female BSTc somatostatin neuron number
(Kruijver et al., 2000). Hence, this individuals
hypothalamic characteristics were mid-way
between male and female values.
In transsexual MtF patients who receive hormo-
nal treatment, some intermediate values, between
those typical for men and women, have been
found for lateralization and cognitive performance
(Cohen-Kettenis et al., 1998). Recently, functional
reversals have been reported in the brains of
transsexual people. A PET study innon-homosexual
MtF transsexual people (i.e., erotically attracted to
women), who were not treated hormonally,
showed that a number of brain areas in the trans-
sexual hypothalamus were activated by phero-
mones in a sex-atypical way. Although the
functional reactions in the hypothalamus to an
estrogen-derived pheromone were predominantly
female, MtF transsexual people also showed some
characteristics of a male activation pattern
(Berglund et al., 2008). Also studies of mental
rotation task, in which men typically outperform
women, showed an in-between pattern in MtF
transsexuals. Compared to control males, the acti-
vation in MtF transsexuals during the task was,
like in female controls, lower in the superior par-
ietal lobe. MtF transsexuals differed, however, also
from the females, and showed higher activation in
orbital and right dorsolateral prefrontal regions and
lower activation in the left prefrontal gyrus. Inter-
estingly, the reduced parietal activation in MtF
transsexuals was correlated with years of estrogen
treatment (Carrillo et al., 2010), suggesting that a
major reason for the observed female feature
could have been the hormone supplement treat-
ment. When viewing erotic stimuli, MtF
transsexuals before treatment tended to display
female-like cerebral processing on functional
magnetic resonance imaging (fMRI). The core
network consisting of the occipitotemporal cortex,
anterior cingulate cortex, medial prefrontal cor-
tex, pre- and postcentral cortex, thalamus,
hypothalamus, and bilateral amygdala was acti-
vated in males, females, and MtF transsexuals.
The three latter regions, however, were more acti-
vated in male controls than in female controls and
MtF transsexuals (Gizewski et al., 2009). One pos-
sible explanation could be that both females and
MtF transsexuals reported a lower degree of sex-
ual arousal, and particularly the hypothalamus
activation is reported to arousal-dependent.
Transsexual persons have recently been investi-
gated with diffusion tensor imaging (DTI), which
measures fractional anisotropy (FA) and provides
information about neuronal fiber tracts. The study
showed significantly higher FA values in the medial
and posterior parts of the right superior long-
itudinal fasciculus (SLF), the forceps minor, and
the corticospinal tract in male controls and FtM
transsexuals compared to control females
(Rametti et al., 2010). In contrast to these two
studies, which suggested sex atypical parietal acti-
vations and fronto-parietal neuronal connections,
no difference from sex matched controls were
detected in a comparative study of regional gray
and white matter volumes, with exception for an
increase in gray matter volume in the left putamen
in MtF transsexuals compared to both male and
female controls (Luders et al., 2009). Recently,
Savic and coworkers combined voxel-based
Author's personal copy
morphometry and structural volumetry to find that
MtF transsexuals have reduced structural volumes
of the putamen and thalamus compared to both
male and female controls. In addition, their gray
matter fraction in the right insular cortex, and the
right temporo-parietal junction was larger than in
both control groups. Together, these anatomical
findings question the dogma that transsexual per-
sons simply have an inverted sex dimorphism of
the brain in relation to their biological sex. The
findings also raise question as to whether trans-
sexuality may be associated with changes in the
cerebral networks involved in self-perception
the temporo-parietal junction, the thalamus, and
the insular-inferior frontal cortex (Northoff et al.,
Sexual orientation: heterosexuality,
homosexuality, and bisexuality
Sexual orientation in humans is also determined
during early development, under the influence of
our genetic background and factors that influence
the interactions between the sex hormones and
the developing brain (Table 2).
The apparent impossibility of getting someone
to change their sexual orientation is a major argu-
ment against the importance of the social environ-
ment in the emergence of homosexuality, as well
as against the idea that homosexuality is a lifestyle
choice. The mind boggles at the methods used in
the attempt to bring about changes in sexual
orientation: hormonal treatments such as castra-
tion, administration of testosterone or estrogens
(treatments that appeared to affect libido but not
sexual orientation); psychoanalysis; apomorphine
administered as an emetic in combination with
homoerotic pictures; psychosurgery (lesions in
the hypothalamus); electroshock treatment; che-
mical induction of epileptic insults and imprison-
ment. As none of these interventions has led to a
well-documented change in sexual orientation
(LeVay, 1996), there can be little doubt that our
sexual orientation is fixed by the time we reach
adulthood and is beyond further influence.
Changes in sexual orientation in adulthood have
been describedfor example, from heterosexual
to pedophilebut only in cases of brain tumors in
the hypothalamus and prefrontal cortex (Burns
and Swerdlow, 2003; Miller et al., 1986). However,
these devastating changes in the hypothalamus are
too large to interpret them in terms of functional
changes in particular neuronal circuits. There are
also claims that pedophiles and homosexual men
have switched to heterosexual behavior as a result
of stereotactical psychosurgery (lesions in the
nucleus ventromedialis) (Dieckmann and Hassler,
1977), but these interventions are not only ethi-
cally questionable, they also do not meet any
Table 2. Prenatal factors that may influence sexual orientation (homosexuality, heterosexuality, bisexuality)
Genetic factors Twin studies (Bailey and Bell, 1993; Bockalandt and Vilain, 2007; LeVay and Hamer, 1994)
Molecular genetics (Swaab, 2004)
Hormones Girls with CAH (Meyer-Bahlburg et al., 1995, 1996; Swaab, 2004; Zucker et al., 1996)
DES (Cohen-Kettenis et al., 1998; Ehrhardt et al., 1985; Swaab, 2004)
Chemical Prenatal exposure to nicotine, amphetamines, or thyroid medication (Ellis and Cole-Hardin, 2001; Ellis and
factors Hellberg, 2005)
Immune Homosexual orientation in men is most likely to occur in men with a large number of older brothers (Blanchard,
response? 2001; Bogaert, 2003)
Social factors? Stress in the mother during pregnancy (Bailey et al., 1991; Bogaert, 2003; Ellis et al., 1988) Being raised by
transsexual or homosexual parents does not affect sexual orientation (Green, 1978)
Abbreviations: CAH, congenital adrenal hyperplasia; DES, diethylstilbestrol.
Author's personal copy
scientific standards. There are also some recent
reports postulating that the sexual orientation of
homosexual women, more than that of homosex-
ual men, may sometimes change, either sponta-
neously or under the influence of psychotherapy
(Spitzer, 2003). The effectiveness of therapy and
the absence of bisexuality has, however, never
been convincingly demonstrated in these cases.
The presence of a substantial genetic compo-
nent in the development of sexual orientation is
apparent from family and twin studies (Bailey and
Bell, 1993; Bocklandt and Vilain, 2007). However,
exactly which genes play a role is not yet clear.
According to LeVay and Hamer (1994), the size
of the genetic component in homosexuality for
both sexes is over 50%. A number of genetic
studies have suggested maternal transmission,
indicating X-linked inheritance. The X chromo-
some has accumulated genes involved in sex,
reproduction, and cognition. A meta-analysis of
four linkage studies suggested that Xq28 plays an
important role in male homosexuality (Hamer
et al., 1993). However, 16 years after the initial
findings the exact genes involved have not yet
been identified (Bocklandt and Vilain, 2007).
A different technique also indicated a role for
the X chromosome in male sexual orientation.
Women with gay sons appeared to have an
extreme skewing of X-inactivation when they are
compared to mothers without gay sons (Bocklandt
et al., 2006). Although this unusual methylation
pattern supports a possible role of the X chromo-
some in male homosexuality, its mechanism of
action is far from clear. Given the complexity of
the development of sexual orientation, it is likely
to involve many genes. A genome-wide linkage
screening indeed identified several chromosomal
regions and candidate genes for further explora-
tion (Mustanski et al., 2005).
Whatever the exact nature of the genetic factor,
it is interesting that such a factor has stayed pre-
sent in the population throughout human history,
given that homosexuals do not tend to procreate
as much as the rest of the population. A good
explanation could be that the genetic factors that
are responsible for homosexuality also have a
beneficial effect on the procreation of the popula-
tion. Indeed, Camperio Ciani et al. (2004) have
found that women on a homosexual males
mothers side tend to be more fertile. This antag-
onistic inheritance that promotes fecundity in
females and a homosexual orientation in males is
partly linked to the X chromosome (Iemmola and
Camperio Ciani, 2009).
Abnormal hormone levels originating from the
child itself during intrauterine development may
influence sexual orientation, as is apparent from the
large percentage of bisexual and homosexual girls
with CAH (Meyer-Bahlburg et al., 1995, 1996;
Zucker et al., 1996). Between 1939 and 1960, some
two million pregnant women in the United States
and Europe were prescribed diethylstilbestrol
(DES) in order to prevent miscarriage. DES is an
estrogen-like substance that actually turned out not
to prevent miscarriage; furthermore, it also found, in
small dosages, not only to give a slightly elevated
risk of cervical cancer but also to increase the chance
of bisexuality or lesbianism in adult woman
(Ehrhardt et al., 1985; Meyer-Bahlburg et al.,
1996; Titus-Ernstoff et al., 2003) although this was
not confirmed in an other study (Ellis et al., 1988).
The chance that a boy will be homosexual
increases with the number of older brothers he
has. This phenomenon is known as the fraternal
birth order effect and is putatively explained by an
immunological response by the mother to a pro-
duct of the Y chromosome of her sons. The chance
of such an immune response to male factors would
increase with every pregnancy resulting in the
birth of a son (Blanchard, 2001; Bogaert, 2003).
Prenatal exposure to nicotine, amphetamine, or
thyroid-gland hormones increases the chances of
giving birth to lesbian daughters (Ellis and
Cole-Harding, 2001; Ellis and Hellberg, 2005).
A stressed pregnant woman has a greater chance
of giving birth to a homosexual son (Ellis and
Cole-Harding, 2001; Ellis et al., 1988) or a lesbian
daughter (Bailey et al., 1991)(Table 2).
Although it has often been postulated that post-
natal development is also important for the
Author's personal copy
direction of sexual orientation, there is no solid
proof for this. On the contrary, children who were
born after artificial insemination with donor sperm
and who were raised by a lesbian couple are het-
erosexually oriented (Green, 1978). There is also
no proof for the idea that homosexuality is the
result of a deficient upbringing, or that it is a life-
style choice or an effect of social learning (LeVay,
1996). It is curious, therefore, that some children
are still forbidden to play with homosexual friends,
an unthinkable attitude left over from the idea that
homosexuality is contagious or can be learned.
Sexual orientation and the brain
Clinical observations have shown the involvement
of a number of brain structures in sexual orienta-
tion. It has been reported that in some patients
with Klüver-Bucy syndrome, which involves
lesions of the temporal lobe, orientation changed
from heterosexual to homosexual. Shifts in sexual
orientation (to homosexual and pedophile) have
also been reported in connection with tumors in
the temporal lobe and hypothalamus. Lesions in
the preoptic area of the hypothalamus in male
rodents, such as ferrets and rats, produce shifts in
sexual orientation (Swaab, 2003). Lesions of the
same structure in their female conspecifics do not
change sexual behavior. Instead, female rats
become aggressive toward male intruders and
start approaching their female conspecifics upon
lesion of the ventromedial hypothalamic nuclei
(Kindon et al., 1996; Leedy, 1984; Paredes and
Baum, 1995).
Of interest is also that male rat knockouts lack-
ing Ca-TRP channels (TRPC2), which are neces-
sary for pheromone signal transduction, do not
approach to fertile females, but do mount male
rats (Zufall, 2005). These data have two implica-
tions: first, intact pheromone signal detection, as
well as an intact hypothalamic transduction seems
necessary for heterosexual behavior. Second, the
hypothalamic nuclei mediating sexual behavior
seem, at least in some rodents, to differ between
males and females. The exact function of the these
nuclei is not well known, but it seems to be crucial
for the approach to a sexual partner, since it is
implicated in the recognition and integration of
sensory stimuli such as sexual clues, in arousal
mechanisms and in copulatory behavior and its
motor expression (Schober and Pfaff, 2007;
Swaab, 2003).
Several structural and functional differences in
the brain have been described in relation to sexual
orientation (for a review see Swaab, 2008).
Swaabs group found the first difference in the
SCN, or brain clock, which turned out to be twice
as large in homosexual compared with heterosex-
ual men (Swaab and Hofman, 1990). In an experi-
ment with rats a similar difference could be
induced, by pharmacologically disturbing the inter-
action between testosterone and the developing
brain around the time of birth, using the aromatase
inhibitor 1,4,6-androstatrien-3,17-dione (ATD) in
the neonatal period. This experiment yielded
bisexual adult rats, which had larger numbers of
cells in their SCN (Swaab et al., 1995). The differ-
ence in the SCN was therefore not caused by a
change in sexual behavior, as was suggested at
the time, but by a disturbed interaction between
sex hormones and the developing brain. In 1991,
LeVay reported that homosexual men, just like
heterosexual women, have a smaller volume of
hypothalamic nucleus (INAH-3) (LeVay, 1991).
No differences were found in the BSTc volume or
number of somatostatine neurons in homosexual
men compared to heterosexual men (Kruijver
et al., 2000; Zhou et al., 1995). In 1992, Allen and
Gorski reported that the anterior commissure of
homosexual men is larger than that of heterosexual
men (Allen and Gorski, 1992). This structure, which
is larger in women than in men, takes care of left
right connections within the temporal cortex and is
thus involved in sex differences in cognitive abilities
and language. The difference in its size may possibly
be related to the sex-atypical hemispheric asymme-
tries observed in homosexual men and homosexual
women by Savic and Lindström (2008). Witelson
et al. (2008) recently reported that the isthmal
Author's personal copy
One-group random effect analysis
HeW Lesbian HeM Homosexual
area of corpus callosum was larger in the homosex-
ual compared to heterosexual men, which also
could contribute to the observed differences in
hemispheric asymmetry.
Emerging studies with functional imaging show
differences in the hypothalamus activation in rela-
tion to sexual orientation. The first brain imaging
paper to point out differences in the hypothalamus
in relation to sexual orientation by means of fluor-
odeoxy glucose (FDG)PET, by Kinnunen et al.
(2004), did not receive much scientific or public
attention, although it may have clinical conse-
quences. The hypothalamus of homosexual men
turned out not to be as responsive to a classic
antidepressant (fluoxetine) as that of heterosexual
men, which suggests a different kind of activity of
the serotonergic system. Savic et al. (2001) used
androstadienone, a pheromone-like compound
derived from progesterone and excreted in
perspiration in concentrations. Smelling of this
compound activated the hypothalamus of hetero-
sexual women and homosexual men in the same
way, but did not elicit any hypothalamus response
in heterosexual men. Apparently in heterosexual
men the hypothalamic pathway is not stimulated
by a male body-scent, which suggests that
pheromone-like compounds in humans may con-
tribute to determining our behavior in relation to
our sexual orientation (Savic et al., 2005). In a
follow-up study (Berglund et al., 2006), lesbian
women, as compared to heterosexual women,
reacted in a sex-atypical, almost reciprocal way
(Fig. 3). These observations, too, show that there
are hypothalamic circuits that function in a way
that depends on our sexual orientation. The
hypothalamic circuits are incorporated in the
core network system for sexual arousal (Karama
et al., 2002). Interestingly, when balancing for the
degree of sexual arousal, this network seems simi-
lar in homo- and heterosexual subjects. Just like
the pheromone responding core network, the trig-
gering stimulus is reciprocal in homosexual com-
pared to heterosexual subjects. Indeed, viewing
erotic videos of heterosexual or homosexual con-
tent produced activation in the hypothalamus,
detectable by fMRI, but only when subjects were
Fig. 3. Illustration of group-specific activations with the two putative pheromones (AND and EST). AND, androstadienone. EST,
estratetraenol, is derivative of estrogene. The Sokoloff color scale illustrates Z-values reflecting the degree of activation (0.05.0).
Because the same brain section is chosen, the figures do not always illustrate maximal activation for each condition (Upper). Cerebral
activation during smelling of AND and EST. Clusters of activated regions are superimposed on the standard MRI brain (midsagittal
plane). HeW, heterosexual women; HeM, heterosexual men. Note that there are hypothalamic circuits that function in a way that
depends on our sexual orientation. From Berglund et al. (2006) with permission. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this book.)
Author's personal copy
viewing videos of their respective sexual orienta-
tion (Paul et al., 2008). Accordingly Ponseti et al.
(2006, 2009) found that neuronal response of the
ventral striatum and the centromedian thalamus
was stronger to prefer relative to non-preferred
stimuli. Using fMRI, Kranz and Ishai found that
face perception is modulated by sexual prefer-
ence. Looking at a female face made the thalamus
and medial prefrontal cortex of heterosexual men
and homosexual women react more strongly,
whereas in homosexual men and heterosexual
women these structures reacted more strongly
to the face of a man (Kranz and Ishai, 2006).
A sexual-orientation-related difference in proces-
sing neuronal networks was suggested only by
Hu et al. (2008). However, their subjects viewed
erotic film involving mixed and same sex couples,
evoking different levels of sexual arousal and dis-
gust in homo- and heterosexual subjects, which
may account for the detected differences. While
being compelling in pinpointing the neuronal cir-
cuits for sexual attraction and arousal, these data
cannot explain why the object of arousal differs.
Savics previous studies raised the question of
whether certain sexually dimorphic features in the
brain, which are unlikely to be directly involved in
reproduction, may differ between homosexual and
heterosexual individuals. This issue was explored
by studying hemispheric asymmetry, using volu-