See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/285757563

Anatomical and Functional Findings in Female-to-Male Transsexuals: Testing

a New Hypothesis

Article  in  Cerebral Cortex · December 2015

DOI: 10.1093/cercor/bhv278

CITATIONS READS

59 542

3 authors:

Amirhossein Manzouri Kyriaki Kosidou

Karolinska Institutet Karolinska Institutet
41 PUBLICATIONS   378 CITATIONS    77 PUBLICATIONS   1,267 CITATIONS   

SEE PROFILE SEE PROFILE

Ivanka Savic
Karolinska Institutet
151 PUBLICATIONS   7,035 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Improving psychotherapeutic competences using socio-emotional training perceptual procedures View project

Gender identity and own body perception implications for the neurobiology of gender dysphoria View project

All content following this page was uploaded by Ivanka Savic on 02 April 2016.

The user has requested enhancement of the downloaded file.

 Cerebral Cortex Advance Access published December 4, 2015

Cerebral Cortex, 2015, 1–13

doi: 10.1093/cercor/bhv278
Original Article

ORIGINAL ARTICLE

Anatomical and Functional Findings in Female-to-Male

Transsexuals: Testing a New Hypothesis
A. Manzouri1, K. Kosidou2 and I. Savic1

Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016

1
Department of Women’s and Children’s Health and Neurology Clinic, Karolinska Institute and University
Hospital, Stockholm, Sweden and 2Gender team, Psychiatry Southwest, Karolinska University Hospital and
Center for Epidemiology and Community Medicine, Karolinska Institute, Stockholm, Sweden
Address correspondence to I. Savic, Karolinska Institutet, Department of Women’s and Children’s Health, Karolinska Hospital, Q2:07,
SE-171 76 Stockholm, Sweden. Email: ivanka.savic-berglund@ki.se

Abstract
Gender dysphoria (GD) is characterized by incongruence between onés gender assigned at birth and the gender that one
identifies with. The biological mechanisms of GD are unclear, especially in female-to-male transsexuals (FtM-TR). Here, we
investigate whether distinct structural and functional patterns along cerebral midline networks processing own-body
perception may constitute a biological correlate. Method: MRI of functional connectivity, cortical thickness, surface area, and
gray matter volume was carried out in 28 female-to-male transsexuals (FtM-TR) and 68 cis-sexual controls (34 male). FtM-TR
displayed thicker mid-frontal, precuneal-parietal, and lingual cortex than both male and female controls, whereas, in regions
with reported anatomical sex differences among the controls, FtM-TR followed patterns of the gender assigned at their birth.
FtM-TR also displayed weaker functional connections from the pregenual anterior cingulate to the insular cortex, and the
temporo parietal junction compared with both control groups. Distinct structural and functional pattern in the own-body image
network may represent biological markers for the dysphoric own-body perception in transgender individuals.

Key words: cortical thickness, gender dysphoria, MRI, own-body perception, resting-state fMRI

Introduction warranted. Increased information about GD is requested also by

The concept of self has theoretical, empirical, and clinical dimen-
sions. A condition where this concept is fundamental is Gender
identity disorder, referred to as gender dysphoria (GD) in DSM-5
(American Psychiatric Association 2013). GD is characterized by
a significant distress due to incongruence between onés gender
(sex) assigned at birth and the gender that one identifies with.
This distress often leads to a request for sex-reassignment sur-
gery (Blanchard 1993; Cohen-Kettenis and Gooren 1999). The
underlying biological mechanisms are largely unknown and
may be different for female-to-male (FtM-TR) and male-to-fe-
male (MtF-TR) transsexuals (Cohen-Kettenis et al. 1998; Schagen
et al. 2012). Despite a mounting public awareness about GD, there
is recognized stigmatization and hostility toward individuals
with GD, and better knowledge about this condition is highly
the medical community, as there are mounting requests for
cross-hormone treatment, at a steadily decreasing age, and also
among persons who are gender ambiguous, and not necessarily
gender dysphoric. This poses new diagnostic problems among
the professionals and calls for improved scientific support.
Possible Mechanisms Underlying GD
GD and Sexual Dimorphism of the Brain
The currently predominant hypothesis for GD presumes that the
perception of one’s sex is linked to sexual differentiation of the
brain and that in transsexuals there is a divergence between
sex differentiation of the brain and the reproductive organs,
due to genetic factors and/or to an early organizational effect of
testosterone during fetal development (Giedd et al. 1997; Green

© The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com

1
 2 | Cerebral Cortex
and Keverne 2000; van Goozen et al. 2002; Bentz et al. 2007; Swaab
2007). Such a scenario would be possible because sex differenti-
ation of the brain occurs later in development than sex differen-
tiation of the reproductive system (Swaab 2007). The hypothesis
is supported by some postmortem studies showing that MtF-TR,
like females, have smaller volume as well as a lower number
of neurons in the bed nucleus of the stria terminalis (BSTc) and
in the hypothalamic INH3 nucleus, compared with male controls
(Zhou et al. 1995; Kruijver et al. 2000; Garcia-Falgueras and Swaab
2008). More recent in vivo investigations are, however, inconclu-
sive. Depending on the type of metric and the study group (for a
summary of these findings, please see Supplementary Table 1),
some studies are showing cerebral features in transsexual per-
sons, which are in accordance with the gender assigned at their
birth, other results show features more congruent with the gen-
der transsexual persons identify with. For example, in MtF-TR,
the gray matter volumes (GMVs) and white matter volumes are
found to be similar to that of male cis-sexual controls but also dif-
ferent both male and female cis-sexual controls (Savic and Arver
2011). Examinations of cortical thickness (Cth), a more specific
metric than GMV, have found that MtF-TR exhibit a thicker cortex
than male controls, particularly in the parietal and occipital
lobes, which suggests a “female” pattern. However, the corre-
sponding measures in FtM-TR showed no cortical differences
from male or female controls (Luders et al. 2012; Zubiaurre-Elorza
et al. 2013). Of note is also that Zubiaurre-Elorza et al. recently
found increases in Cth in FtM-TR due to testosterone treatment
(Zubiaurre-Elorza et al. 2013, 2014), which could imply a contin-
ued and even exaggerated “female pattern”, which is contrary
to the expected “masculinizing” effect.
In sum, the notion that sexual differentiation of the brain
would be different among persons with GD is still unsettled.
One reason may be that surprisingly few studies focused on the
basic symptoms and fundamental subjective experiences asso-
ciated with GD—body dysphoria and own-body-related avoid-
ance (Cohen-Kettenis and Pfafflin 2010; Coleman et al. 2012;
American Psychiatric Association 2013), although some authors
addressed this crucial issue several years ago (Lawrence 2006;
Fisher et al. 2014).
GD and the Own-Body Image
We recently forwarded another, and not necessarily contradict-
ing hypothesis (Savic and Arver 2011) for GD. It posits that GD
could be associated with distinct functional and structural neural
signatures in cerebral networks involved in the own-body per-
ception in the context of self. One possibility is that in transgen-
dered individuals the typical physical traits of gender assigned at
birth are not incorporated into self-representation and that this is
associated with specific signatures in the brain. We forwarded
this hypothesis, based on observation that MtF-TR displayed
greater GMV in the right occipito-parietal junction and the infer-
ior frontal gyrus, and smaller thalamus and putamen volumes in
relation to both male and female controls (Savic and Arver 2011).
Thus, they displayed distinct patterns in structures which are, ac-
cording to a meta-analysis, parts of the own-body image network
(together with the anterior cingulate [ACC] and middle prefrontal
cortex [mPFC], the posterior cingulate, precuneus, and cuneus
[Northoff et al. 2006; Lopez et al. 2008; Northoff and Panksepp
2008]). Notably, several of these structures are also sexually di-
morphic (Savic and Arver 2011, 2014; Lentini et al. 2013). It is,
therefore, possible that the distinct signatures in the own-body
image network in GD may also infer that some structures in
this network appear less sexually differentiated, less “sexually
dimorphic.” This does not necessarily mean that an altered sex
dimorphism would be the mechanism underlying GD; this
could just be a parallel phenomenon.
In the present study, we tried to tie together the sparse and
sometimes contradictory information concerning cerebral anat-
omy and function in transgendered persons, and FtM-TR in par-
ticular, by utilizing several MRI methods (MRI measurements of
Cth, surface area [SA], and GMV, as well as rs-fMRI). Cth and SA
were investigated in addition to GMV, because they are described
to have differing evolutionary histories (Rakic 1988,1995), devel-
opmental trajectories (Sowell et al. 2007). and genetic determi-
nants and are differentially influenced by neuronal migration
(Panizzon et al. 2009). By examining the presumable own-body
image network in relation to cerebral sex dimorphism, and focus-
ing on FtM-TR in whom the available information about possible
cerebral correlates is particularly uncertain, we hope to add a
new dimension to the ongoing discourse on the mechanisms
underlying GD.
Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016
Our main hypothesis was that both brain anatomy and func-
tional connectivity would differ along the own-body perception
network as compared with the controls. Furthermore, if sex atyp-
ical features were to be detected, they were expected to be located
foremost within this network.
Methods
Material
Twenty-eight FtM-TR (age 23.5 ± 5.2 years, range 18–34 years;
education 13.0 ± 2.7 years, range 9–18 years), 34 heterosexual
cis-sexual men (HeM), (age 28.8 ± 4.8 years, range 20–39 years;
education 16.2 ± 2.4, range 12.5–21 years), and 34 heterosexual
cis-sexual women (HeW), (age 27.6 ± 4.1 years, range 20–32
years; education 16.5 ± 2.7 years, range 12–21 years) participated
in the study. None of the FtM-TR had received hormone treat-
ment or sex-reassignment surgery and none of them had any
neurological or psychiatric disorder. History of sex hormone
treatment was assessed by self-reports and controls of medical
journals, including previous prescriptions. The FtM-TR were re-
cruited by the Gender team at the Psychiatry clinic at Karolinska
University Hospital (Stockholm, Sweden), which specializes in
the evaluation and treatment of patients with GD. All consecu-
tively arriving adult patients aged 18–45 years who sought sex
correction therapy, and were diagnosed with a GD (DSM-5), spe-
cifically transsexualism (ICD-10) diagnostic criteria (http://www.
who.int/classifications/icd/en/), were approached to enter the
study, between January 2011 and July 2014. Exclusion criteria
consisted of previous or current hormonal treatment, any
known chromosomal or hormonal disorder, any severe psychi-
atric, neurological or systemic disorder, including autism spec-
trum disorder (ASD), and any medications with psychotropic
effects (antipsychotic or antiepileptic agents, lithium, benzodia-
zepines or opioid analgesics). The clinical evaluation at the
Gender team routinely includes the Mini International Neuro-
psychiatric Interview (M.I.N.I.) (Sheehan et al. 1998), which con-
firmed that none of the enrolled patients satisfied the criteria
for depression or other severe psychiatric disorder at the time
of recruitment.
The sexual orientation of all of the subjects was scored using
the Kinsey (Heterosexual–Homosexual Rating) scale (Kinsey et al.
1953), a 7-point scale ranging from 0 to 6. The approach is de-
scribed in detail in previous studies (Berglund et al. 2006b; Savic
and Lindstrom 2008a, please see also Table 1). All cis-sexual con-
trols scored 0–1 on the Kinsey scale (0.4 ± 0.8 for males, and
0.6 ± 0.9 for females), whereas the FtM-TR group were predominantly
 Gender Dysphoria, Brain Anatomy and Function Manzouri et al. | 3

Table 1 Demographic data

Unit HeM (N = 34) HeW (N = 34) FtM-TR (N = 28) F-value P-value

Mean SD Mean SD Mean SD

Age year 28.8 4.8 27.6 4.1 23.5 5.2 4.8 0.003
Education year 16.2 2.4 16.5 2.7 13.0 2.7 16.184 0.000
Estradiol ( plasma) pmol/L 506 475
Testosterone (bioactive) nmol/L 1.67 0.47
Kinsey scale 0.4 0.8 0.6 0.9 4.3 1.9 7.4 0.000
Social responsiveness score 42.9 5.9 46.9 5.9 49.9 10.4 1.8 0.13

Note: F-values from group comparisons (one-way ANOVA). MtF-TR scored significantly higher at the homosexual end of the Kinsey scale. The y were also younger and had
lower number years of education.

gynephyllic and scored between 4 and 6 (4.3 ± 1.9). In relation to (FoV = 24 cm, 60 interleaved axial slices, thickness = 2.9 mm, TE =
their sex assigned at birth, most of FtM-TR were, thus, predomin- 83 ms, TR 8000 ms, 60 diffusion gradient directions [b = 1000], flip

Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016

antly homosexual, but in relation to the sex they identified with
they were heterosexual. For the sake of simplicity and to accord
to current literature regarding sexual orientation in transsexual-
ity (Blanchard et al. 1987), we will relate Kinse’s scorings in
FtM-TR to the sex they were assigned at birth. More precisely,
13 FtM-TR scored Kinsey 6, whereas the remaining 15 scored
3.5–5 and none described self as truly bisexual.
Because there is considerable comorbidity between GD and
ASD (de Vries et al. 2010), we excluded subjects with ASD. Autistic
traits were assessed using the Social Responsiveness Scale (Con-
stantino and Gruber 2005), which consisted of 65 items scored on
a scale from 0 (never true) to 3 (almost always true) with higher
total scores suggesting the presence of greater autistic traits.
This scale was completed by a respondent who knew the subject
well (e.g., a parent, partner, or close friend). The cut-off score for
ASD was 65. One subject had to be excluded in the analyses be-
cause his ASD scores were above this level. The SRS was adminis-
tered to all transsexual participants, requiring that someone
close to the participant, such as a parent, partner or close friend,
complete it. The 65-item SRS questionnaire is scored on a scale
from 0 (never true) to 3 (almost always true). Total scores can
range from 0 to 195. Lower scores on the questionnaire signify
less autistic behavior over the 6 months prior to the testing,
whereas higher scores signify more autistic behavior over this
period.
The control group consisted of healthy subjects who did not
show any comorbidity with or heredity for neuropsychiatric dis-
orders, and who did not have a history of substance abuse or take
any ongoing medication. The study was approved by the ethical
committee at the Karolinska Institute and each subject provided
signed consent before participating in the study.
angle of 90°). Finally, clinical sagittal FLAIR images were taken
(TE/TR = 126.3/6000, TI = 1863, ETL = 140, ARC acceler. R = 2 × 2
[slice, phase], FoV = 24 cm, 512 × 512, slice thickness = 1.2 mm).
Results of the DTI measurement will be presented in a separate
publication. Images acquired with the 8-channel coil were used
only for the FreeSurfer segmentation analyses (see further), be-
cause these T1 images had better demarcation between white
and gray matter in the occipital cortex than T1 images acquired
with the 32-channel coil.
Cortical Thickness, GMV, and SA
The MR volumes were processed using FreeSurfer software ver-
sion 5.1 (Fischl and Dale 2000), (www.surfer.nmr.mgh.harvard.
edu), as described in detail in our previous studies (Savic and
Arver 2014; Savic 2015). For a detailed description, see Supple-
mental Material.
Possible group differences in Cth, GMV and SA were
evaluated for each vertex, using age as the nuisance variable,
and after employing Monte Carlo correction (5000 permuta-
tions). P was set to <0.01 corrected, to avoid large, confluent clus-
ters. Age and education differed significantly between the
control and the transsexual populations; see Table 1. Although
none of these factors, according to the qdec analysis using Free-
Sufer software, had a significantly different impact on regional
metrices (Cth, GMV, and SA) between the groups analyzed,
there were some subsignificant differences. Demeaned age
and education values were, therefore, used as nuisance vari-
ables, and the vertex-by-vertex group comparison (10-mm filter)
in the qdec analyses was carried out using a different-offset-
different-slope model for all the comparisons between patients
and controls.

Magnetic Resonance Imaging Conjunctional Clusters

Data Acquisition
Magnetic resonance imaging data were acquired on a 3-Tesla MRI
medical scanner (Discovery 3T GE-MR750, General Electric, Mil-
waukee, Wisconsin) equipped with a 32-channel and/or 8-chan-
nel phased array receiving coil. 3D T1-weighted SPGR images
were acquired with 1 mm3 isotropic voxel size (TE = 3.1 ms, TR =
7.9 ms, TI = 450 ms, FoV = 24 cm, 176 axial slices, flip angle of 12°).
Resting-state functional MRI was performed with a gradient echo
pulse sequence using a voxel size of 2.25 × 2.25 × 3 mm (TE = 30 ms,
TR = 2500 ms, FoV = 28.8 cm, 45 bottom up interleaved axial
slices, 3 mm thickness, flip angle of 90°). In addition, multislice
DTI (not used in the present analyses) was performed using an
echo planar imaging sequence with 1 × 1 mm in-plane resolution
To investigate whether there were common differences between
2 groups in relation to the third, conjunctional analyses were car-
ried out. For these analyses, the method called Minimum Statis-
tic compared with the Global Null (MS/GN) was used (Nichols
et al. 2005). The threshold level was P < 0.05 corrected (Monte
Carlo simulation).
Segmentation of the Subcortical Volumes
Subcortical segmentation generated with FreeSurfer was used to
calculate the volumes of 6 subcortical brain structures: the amyg-
dala, hippocampus, caudate, putamen, thalamus, and cerebel-
lum. When needed, the segmented brain structural masks were
modified manually by a rater who was not informed about the
 4 | Cerebral Cortex

identities of the subjects. Manual correction (especially with re-

Clusters calculated at P < 0.01, corrected for multiple comparisons (Monte Carlo permutations), filter 10 mm. Italics indicate clusters at P < 0.05 corrected. Demeaned age and education was used as nuisance covariates. The Talairach’s
gard to the putamen and amygdala) is often needed when

coordinates

−24 −76 −1
−52 −48 30
−33 −72 43
−25 51 −12

54 −41 −17
−29 −81 7
44 −80 −6
Talairach

23 −31 72
using FreeSurfer segmentation. In the present study, it had to

11 63 −2
FtM-TR-HeM (negative −log10(P) values)
HeM-FtM-TR ( positive −log10(P) values)

6 −84 2
be employed in 9 transgendered subjects and 16 controls. The
standard procedure employed has been described in detail pre-
viously (Fischl et al. 2002, 2004; Savic 2015). Ratios between the
respective VOI and the total intracranial volume (TIV) retrieved

size, cm2
Maximum vertex- Cluster
from the FreeSurfer program were entered into the statistical

9.2

11.2
4.3
8.9

11.6

6.8
4.7
10.1
9.9

3.2
analyses. After ensuring that the data were normally dis-
tributed, group comparisons of relative structural volumes
(VOI/TIV) were carried out with general linear model using

wise −log 10(P)

the individual relative values of the 2 homologous VOIs for
each type of structure as input values (P < 0.025, due to the

coordinates indicate location of maximum difference. HeW, female controls; HeM,male controls; FtM-TR, female-to-male transsexuals; R, right; L, left; * covers part of the L frontal lobe.
Bonferroni correction for the thalamus and cerebellum; the
P-level was 0.05 for the amygdala, hippocampus, putamen,

−5.0
−4.0
−3.9

−3.9

−4.5
−3.8
−3.8

−3.6
−4.3

−2.8
and caudate, because we had specific hypotheses for possible
group differences in these regions based on the previous results

Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016

coordinates
consistently showing gender differences in these regions). Age

−58 −28 35
−39 −65 42

54 −42 −16

15 −83 −8
Talairach
and education were used as nuisance covariates. The afore-

−38 32 8

10 63 −2
−9 39 18
FtM-TR-HeW (negative −log10(P) values)
HeW-FtM-TR ( positive −log10(P) values)
mentioned analyses were carried out with PASW Statistics 21
(SPSS, Inc.).

size, cm2
Resting-State Functional MRI

Maximum vertex- Cluster

To perform seed region analysis, the cross-correlation coeffi-

3.6
7.3
16.7
3.9

4.1
7.4
21.3
cients of the time series in a particular seed region-of-interest
are calculated for all other voxels in the brain. The approach
reveals the strength of functional connectivity with respect

wise −log10(P)
to this seed region (Biswal et al. 1995). The seed regions used
were the right and left amygdala, as it has been reported that
resting-state functional connectivity from the amygdala differs
−3.9
−3.2
−4.8
−4.9

−4.4
−3.8
−3.5
between men and women (Kilpatrick et al. 2006; Savic and
Lindstrom 2008b). In addition, we used seeds covering the right
and left TPJs, including the arcuate gyrus. The rationale for
coordinates

using the TPJ seed was our previous finding of significantly larger

−27 −90 13
−8 −69 48
Talairach

17 −47 57
GMV in this region in MtF-TR compared with both HeM and HeW
HeM-HeW (negative −log10(P) values)
HeW-HeM ( positive −log10(P) values)

controls (Savic and Arver 2011), and because this region is consid-
ered to be part of the own-body perception network. Finally, the
connections from a seed covering the pACC were evaluated
size, cm2

(BA 25, 32 and parts of 24), since this region is reportedly involved
Maximum vertex- Cluster

Note: The “Region” column describes the coverage of the respective cluster. R, right; L, left.
in distinguishing the self and own-body detection network
21.0
24.0

19.9
Table 2 Clusters showing significant group difference in cortical thickness

(Northoff et al. 2006; Northoff and Panksepp 2008). All seed re-
gions, except for the TPJ, were delineated with the guidance of
the WFU-pick atlas, and after adaptation to the gray matter tem-
wise −log10(P)

plate from the present population. The MNI coordinates for the
TPJ were 53–58, −30–−59, and 19–28 on the right side with a vol-
ume of 2160 mm2 and, on the left side, −55–−50, −30–−60, and
18–28 with a volume of 2304 mm2 (Savic and Arver 2011). The
3.1
2.6

4.4

amygdala seeds were spheres with 5-mm radii and coordinates

L precuneus (+L fusiform and sup parietal gyrus)

of −24, −7, −19 and 24, −7, −19, covering the entire amygdala
R superior parietal gyrus+part of the pre- and

with the exception of the most medial 4 mm, the basomedial

amygdala, which was excluded to avoid the susceptibility artifact
that was detected in some subjects. The MNI coordinates for the
R lingual+precalcarine cortex cuneus

pACC were X = 4–7 and −4 to −7, Y = 40–50, and Z = −10–11, with a

R fusiform+inferior occipital gyrus

volume of 3112 mm2.

L rostral middle frontal cortex

Spatial preprocessing and statistical analysis of functional

images were performed using SPM8 (Welcome Department of
R inferior temporal gyrus
L lateral occipital cortex

R superior frontal gyrus

Cognitive Neurology).
L supramarginal cortex

post-central gyrus

Group comparisons (male controls vs. female controls,

L pars triangularis

male controls vs. FtM-TR, and female controls vs. FtM-TR)

L parietal cortex

L lingual cortex

were carried out in SPM8 using one-way ANOVAs, with P < 0.001
voxel threshold, FDR corrected at cluster level, and P < 0.05,
while controlling for age, which was used as the nuisance
Region

covariate. Because our hypothesis was that functional con-

nections from the TPJ-AG and the pACC would be altered in
 Gender Dysphoria, Brain Anatomy and Function Manzouri et al. | 5

Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016

Figure 1. Group comparisons of cortical thickness (A), SA (B), and GMV (C). The contrasts were calculated at P < 0.01 corrected for multiple comparisons (Monte Carlo
permutation), using age and education as the covariates of no interest. The projection of cerebral hemispheres (MR images of the Freesurfer atlas) is standardized.
Scale is logarithmic and shows −log10(P). Warm colors indicate positive contrasts (higher values in FtM-TR), cool colors negative contrasts.

FtM-TR within the own-body perception network (the insular Results

cortex, the precuneus, the posterior cingulate, cuneus, and the
inferior parietal cortex), clusters were regarded as significant
at P < 0.001 voxel threshold, uncorrected, if appearing within
an explicit mask covering these regions, as defined by the
WFU-pick atlas.
Subjects
For demographics, see Table 1. Patients were significantly young-
er and less educated than controls. They were gynephillic and
scored in the homosexual range of the Kinsey scale in relation
 6 | Cerebral Cortex

to the gender they were assigned at birth. All the male and female
controls were heterosexual.
Cortical Thickness, SA, and GMV
Cth was greater in HeW than HeM in the left fusiform gyrus and
lateral occipital cortex, in the left precuneus and a part of the left
superior parietal cortex, as well as in the right superior parietal
cortex including a part of the postcentral gyrus (Table 2,
Fig. 1A). As with HeW, FtM-TR exhibited a thicker cortex than
HeM in the left lateral occipital lobe and in the left superior par-
ietal lobe. Furthermore, there were several areas, mainly around
the midline, in which FtM-TR showed significantly thicker cortex
compared with both control groups (Table 2, Fig. 1A). These areas
were located bilaterally in the superior, rostral, and middle front-
al gyri (including portions of the orbitofrontal gyrus), in the right
lingual+fusiform gyrus, the pericalcarine cortex and cuneus, and
in the right inferior temporal gyrus. The areas with thicker cortex
among FtM-TR in relation to both control groups were also dis-
minor extent in the right fusiform and inferior occipital gyrus.
The analysis of GMV showed, as in our previous studies (Savic
and Arver 2011; Lentini et al. 2012), that HeW had greater GMV
than HeM in the precentral gyrus (and also partly in the postcen-
tral gyrus) and in the right inferior parietal cortex (Table 4,
Fig. 1C), whereas in HeM, the GMV was greater in the left occipital
cortex (Table 5c). Similar differences were also detected when
comparing FtM-TR and HeM. Also the GMV in the right middle
and inferior temporal cortices were larger in HeW and FtM-TR
compared with HeM. In addition, around the fronto-occipital
midline (the right occipital cortex and the frontal pole), the
GMV was, just as Cth, in FtM-TR larger compared with both con-
trol groups (Table 4, Fig. 1C).
Thus, FtM-TR differed similarly from male controls, as did fe-
male controls with respect to all 3 metrics, Cth, SA, and GMV. In
addition, they differed from both control groups primarily in re-
gard to Cth, and to a minor extent also GMV, and this difference

Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016

closed through conjunctional analysis (Fig. 2). As opposed to
Cth, SA did not differ between FtM-TR and HeW. Just as among
HeW, SA was larger among FtM-TR than HeM in the right middle
temporal lobe, the superior frontal lobe (Table 3, Fig. 1B), and to a
was detected primarily along the fronto-occipital midline.
Subcortical Volumes
Next, we investigated subcortical volumes by specifically focus-
ing on possible sex differences and how they related to the corre-
sponding measures in transsexuals. The volumes analyzed were
those that were previously found to differ between male and fe-
male controls and HeW: the hippocampus, thalamus, caudate,
and putamen (Filipek et al. 1994; Goldstein et al. 2001; Raz et al.
2005; Savic and Arver 2011). The present data confirmed the pre-
vious observation of subcortical sex differences, with HeW hav-
ing significantly larger relative volumes for the right caudate
and right and left hippocampi than HeM, and HeM having larger
thalamus and putamen volumes than HeW (Table 5). FtM-TR fol-
lowed this “female” pattern. In addition, and congruous with sev-
eral previous publications of FtM-TR as well as MtF-TR (Luders
et al. 2009, 2012 Rametti et al. 2011; Zubiaurre-Elorza et al.
2013), putamen volume was larger in FtM-TR than in controls of
same natal gender (although significance was only reached for
the left side) (Table 5).

Figure 2. Conjunctional clusters illustrating regions in which cortical thickness of
FtM-TR differed from both control groups. Conjunctions are calculated for the
following contrasts: (FtM-TR–HeW) and (FtM-TR–HeM), (P < 0.05 corrected). Scale
is logarithmic and shows −log10(P). Warm colors indicate positive contrasts
(higher values in FtM-TR, cool colors negative contrasts).
Rs-fMRI
Given that our FtM-TR population differed from both control
groups with regard to cerebral midline structures, which are
known to be involved in internal body-self-representation (see
also discussion), we asked specifically whether the rs-functional
connections from the pregenual anterior cingulate ( pACC) seed
region, known to be a major node for self-perception (Northoff

Table 3 Group differences in SA

Region HeW–HeM FtM-TR–HeW FtM-TR–HeM

Max, Cluster size Coordinate Max, Cluster size Coordinate Max, Cluster size Coordinate
−log10(P) (cm2) −log10(P) (cm2) −log10(P) (cm2)

L superior frontal 3.4 14.0 −10 45 38

R middle temporal 3.0 6.7 46 6 −27 4.4 10.9 63 −44 0
R superior frontal 2.9 6.3 20 27 46 4.0 11.1* 11 53 26
R caudal and rostral 2.9 5.3 37 17 46 5.0 23.5 30 32 31
middle frontal
R occipital inferior 2.6 2.0 41 −80 5 4.0 9.0 46 −75 5

Note: The “Region” column describes the coverage of the respective cluster. R, right; L, left.
Clusters calculated at P < 0.01, corrected for multiple comparisons (Monte Carlo permutations), filter 10mm. Italics indicate clusters at P < 0.05 corrected. Demeaned age
and education was used as nuisance covariates. The Talairach’s coordinates indicate location of maximum difference. HeW, female controls; HeM, male controls; FtM-TR,
female-to-male transsexuals; R, right; L, left; * covers part of the L frontal lobe.
 Gender Dysphoria, Brain Anatomy and Function Manzouri et al. | 7

et al. 2006; Northoff and Panksepp 2008), could be distinct in FtM-

Clusters calculated at P < 0.01, corrected for multiple comparisons (Monte Carlo permutations), filter 10 mm. Italics indicate clusters at P < 0.05 corrected. Demeaned age and education was used as nuisance covariates. The Talairach’s coordinates
coordinates

59 −21 −16
49−31 −19
42 −79 −8

15 −64 −2
33 −29 39
35 50 −11
Talairach
TR and, thus, differ from HeM as well as HeW. Another seed re-

48 12 14

52 15 13
9 62 −4
gion used was the right and left temporoparietal junction includ-
ing the arcuate gyrus (TPJ) because the GMV for this region was
FtM-TR-HeM (negative −log10(P) values)
HeM-FtM-TR ( positive −log10(P) values)

found to be significantly different between MtF-TR and controls

in our previous study (Savic and Arver 2011). Also, the TPJ is con-
size, cm2

sidered to be part of the own-body perception network (see Dis-

Cluster

10.6 cussion). Finally, since the connectivity from the amygdala (both
9.0
11.1
6.2

4.2
5.9
8.8
4.1
4.3
right and left) has been reported to differ between males and fe-
males (Berglund et al. 2006a; Kilpatrick et al. 2006) examining
whether this connectivity in FtM-TR corresponded to the gender
Maximum vertex-

assigned at their birth was also of interest.

wise −log10(P)

The connections from left amygdala to the motor cortex

were stronger in HeM than in HeW, as they were from the right
amygdala to the hypothalamus. They were stronger also than in
FtM-TR, whereas no significant difference was detected be-
5.7
5.0
4.8
4.5

3.4
3.8
3.2
2.8
2.3
tween FtM-TR and HeW. FtM-TR displayed, however, singular

Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016

features in that their amygdala connections to the posterior cin-

indicate location of maximum difference. HeW, female controls; HeM, male controls; FtM-TR, female-to-male transsexuals; R, right; L, left; * covers part of the L frontal lobe.
gulate, the fusiform cortex, and the extrastriatal body area (EBA)
coordinates

42 −81 −5

were significantly weaker in comparison with both HeM and

Talairach

24 −57 3
8 62 −5

HeW, particularly on the right side; see Table 6 and Figure 3.

FtM-TR displayed weaker connections than both control groups
FtM-TR-HeW (negative −log10(P) values)
HeW-FtM-TR ( positive −log10(P) values)

also with regard to other seeds. Weaker functional connections

were detected primarily between the pACC and both the precu-
size, cm2

neus and thalamus. Furthermore, the left TPJ connections to the

Cluster

pACC, right cuneus, and precuneus were weaker in FtM-TR,

7.8
3.9

4.4

whereas their TPJ connections with the left inferior and middle
frontal gyri were stronger as compared with the control groups
(Fig. 3).
Maximum vertex-

In summary, the combined observations from analyses of cor-

tical thickness, subcortical volumes, and resting-state functional
wise log10(P)

connectivity indicated distinct patterns within the cortical mid-

line structures of FtM-TR compared with both male and female
cis-sexual controls. Furthermore, FtM-TR subjects displayed
4.0
3.5

3.2

weaker connections within the antero–posterior parts of the

own-body perception network as well as between the amygdala
and the temporo parietal parts of this network. Notably, no par-
coordinates

49 −41 −10

ticular sex atypical features were detected, and FtM-TR displayed

−7 −97 10
Talairach

27 −18 55

15 −47 58
27−83 8

similar difference in relation to HeM as did HeW.

Note: The “Region” column describes the coverage of the respective cluster. R, right; L, left.

Discussion
HeM-HeW (negative −log10(P) values)
HeW-HeM ( positive −log10(P) values)

The present study combines anatomical and functional MRI

size, cm2
Cluster

data in an effort to investigate the neurobiological underpin-

6.9
4.3

3.9
4.2

12.0

nings of GD. In this respect, this report differs from several pre-
vious brain imaging studies of this population, which focused
on single factors (e.g., GMV, fractional anisotropy). Furthermore,
whereas the main goal of several previous studies was to inves-
Maximum vertex-

tigate whether the brains of transsexuals are sexually differen-

wise −log10(P)

tiated in a manner that is opposite to the sex they were assigned

at birth, the present study addresses, in addition, whether
Table 4 Group differences in GM volume

transsexual subjects could have distinct anatomical and func-

−6.4
4.0

2.7
3.5

4.1

tional connectivity features in the own-body perception

networks.

Structural and Rs-Functional Connectivity Measures in

R pre+post-central gyrus

Relation to the Sex Assigned at Birth

R inferior temporal

R middle temporal
R pars opercularis
R inferior parietal
R superior frontal
R lateral occipital

R supra marginal
L lateral occipital

Control subjects exhibited the expected sex difference with

R pars orbitalis

regard to Cth, SA, and GMV; the structural volumes of the amy-
R post-central
R frontal pole

gdala, hippocampus, and caudate; and the pattern of rs-function-

R lingual

al connectivity from the amygdala (see Tables 2–6). In this

Region

respect, the FtM-TR group followed the pattern of HeW. These

observations are largely congruent with the available data in
 8 | Cerebral Cortex

Table 5 Structural volumes

Structural volumes (cm3) HeM (N = 34) HeW (N = 34) FtM-TR (N = 28) P and F values

L caudate volume 4.2 ± 0.5 4.0 ± 0.4 3.9 ± 0.4 P = 0.041, F = 3.3
R caudate volume 4.3 ± 0.6 4.0 ± 0.4 4.0 ± 0.4 P = 0.002, F = 5.3a,c
L putamen volume 5.4 ± 0.7 4.9 ± 0.5 5.2 ± 0.5 P = 0.020, F = 3.4a,b,c
R putamen volume 5.2 ± 0.6 4.8 ± 0.6 4.9 ± 0.5 P = 0.025, F = 3.3a,c
L hippocampus volume 4.3 ± 0.4 4.1 ± 0.4 4.0 ± 0.4 P = 0.006, F = 4.3a,c
R hippocampus volume 4.4 ± 0.4 4.1 ± 0.4 4.1 ± 0.4 P = 0.002, F = 5.1a,c
L amygdala 2.0 ± 0.2 1.7 ± 0.2 1.8 ± 0.2 P = 0.240, F = 1.43
R amygdala 2.0 ± 0.2 1.8 ± 0.2 1.8 ± 0.2 P = 0.530, F = 0.74
L thalamus volume 7.6 ± 0.6 6.9 ± 0.6 6.8 ± 0.8 P = 0.069, F = 2.44
R thalamus volume 7.4 ± 0.6 6.7 ± 0.6 6.8 ± 0.6 P = 0.049, F = 2.69a,c
L Cerebellum volume 58.5 ± 4.3 51.9 ± 4.8 53.7 ± 4.1 P = 0.071, F = 2.45
R Cerebellum volume 60.9 ± 4.3 53.6 ± 4.4 54.8 ± 5.5 P = 0.48, F = 0.84
L Pallidum volume 1.8 ± 0.3 1.6 ± 0.2 1.7 ± 0.1 P = 0.58, F = 0.67
R Pallidum volume 1.7 ± 0.2 1.5 ± 0.2 1.5 ± 0.2 P = 0.40, F = 1.01
TIV volume 1641.3 ± 135.0 1432.4 ± 123.3.4 1416.4 ± 112.1 P = 0.000, F = 27.3a,c

Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016

Note: TIV, total intracranial volume; P-values for structural volumes were based on calculations of ratios between the respective structural volume and the TIV, using the
general linear model and age as covariate. The P and F values indicate overall group analysis.
a = FtM-TR, significant difference from HeM; b = FtM-TR, significant difference from HeW; c = significant difference between HeW and HeM (P < 0.05).

the literature (Supplementary Table 1) and do not support the hy-
pothesis that the sexual differentiation of the brain in FtM-TR
would be away from the gender they were assigned at birth.
This brings attention to our second hypothesis, which queries
whether the own-body referential network could be distinct in
transsexuals. To discuss this hypothesis in relation to the present
data, it is necessary to recapitulate the current concepts about
this network.
The Own-Body Referential Network and the Observed
Differences between MtF-TR and Controls
The concept behind being cognizant that one’s own body is
part of one’s self involves an integrative process (Moseley et al.
2012) that requires a complicated interplay of several struc-
tures, which are mainly located along the antero–posterior axis
of the brain. Although these processes are tightly intercorrelated
and not entirely clarified, one may discern 3 different levels:
1. Configural body processing, body detection, mediated by the
visual and somatosensory perceptual areas, EBA, the right fu-
siform body area (Peelen and Downing 2005), and the right in-
fraparietal lobe (IPL) (Urgesi et al. 2004; Costantini et al. 2011).
2. Recognition and identification of a body as one’s own. This
process is suggested to be mediated by the right IPL, the TPJ,
the precuneus and posterior cingulate, the posterior orbital
gyrus, and the left lateral occipital gyrus (Blanke 2004,2012;
Northoff and Panksepp 2008), and according to some studies
also by the right insular cortex (Devinsky et al. 1989). This sup-
position is based on data from fMRI experiments comparing
brain activity when subjects are viewing their own body and
other familiar bodies showing activations of the posterior su-
perior temporal gyri, the right TPJ, the primary somatosen-
sory cortex, the medial premotor cortex, and the adjacent
medial prefrontal cortex (Hodzic, Kaas et al. 2009; Hodzic,
Muckli, et al. 2009). In addition, so-called body swap experi-
ments show that self-identification with a virtual body is as-
sociated with activation of parts of this network, including
the putamen (Brozzoli et al. 2014). Furthermore, several pa-
tient studies suggest that damage to the right angular gyrus
( part of the TPJ) as well as seizures originating from this
region can lead to out-of-body experiences (Devinsky et al.
1989; Heydrich et al. 2011; Solomon et al. 2015).
3. Self-referential processing, whereby the perception of one’s
own body is incorporated into the context of self, seems to re-
quire a major involvement of the pACC. A quantitative meta-
analysis of 87 studies representing 1433 participants showed
that the specificity of the self (e.g., hearing one’s own name,
seeing one’s own face compared with faces from known peo-
ple and others) involves an activation of the pACC. Other mid-
line regions, that is, the mPFC and the posterior cingulate
cortex, on the other hand, were recruited during the process-
ing of both self-specific and familiar stimuli. Notably, the area
of the pACC recruited during self-specific stimuli overlapped
with the default mode network (Northoff et al. 2006; Northoff
and Panksepp 2008).
In sum, it appears that the perception of one’s own body and
identifying it as one’s self heavily involves cerebral midline struc-
tures incorporating a parieto-occipital “body detection” network
and a fronto-parietal “body referential” network (identification of
own’s body as part of self ). In this context, it is of interest that we
found difference between FtM-TR subjects and both control
groups with regard to these same midline structures. FtM-TR dis-
played greater Cth in the pACC, the inferior and superior parietal
lobule including the TPJ, and also in the cuneus and pericalcarine
cortex (Fig. 1). Their rs-functional connections between pACC
and the precuneus, the right thalamus, and insular cortex were
significantly weaker than in controls. One tentative interpret-
ation of these findings is that in transsexuals (at least in FtM-
TR), there is a weak connection between the neuronal networks
mediating body perception and body ownership in the context
of self (the own-body referential network). The observed func-
tional disconnection from the amygdala to the right precuneus
and TPJ, right EBA, and the fusiform cortex is also of interest. It
accords with the reported emotional own-body estrangement
and disgust; it might reflect a coping mechanism that dissociates
bodily emotion from body image in transsexuals and represent a
neurobiological correlate to this phenomenon. The present data
thereby provide an explanatory mechanism for the GD among
FtM-TR, a group that has hitherto been less investigated than
MtF-TR, and in whom the majority of available scientific reports
Table 6 Group differences in resting-state functional connectivity from the selected seeds

Region HeM–FtM-TR HeW–FtM-TR HeW–HeM

Peak level Z Size (cm ) 3

Co-ordinate Peak level Z Size (cm ) 3
Co-ordinate Peak level Z Size (cm3) Co-ordinate

(a) The amygdala seed

L amygdala
Part of the R pre- and post-central gyrus 36 −14 18* −3.9 7.8 21 −24 63
R superior temporal gyrus+R eba 3.9 8.1 28 −27 −8*
L fusiform gyrus+L cuneus 3.9 10.4 −14 −45 0 4.4 12.8 −10 −42 −5
R amygdala
Hypothalamus 4.5 8.1 6 −3 18 −4.7 4.8 −4 10 4
R fusiform gyrus+R eba+R tpj 4.4 24.0 15 −42 −2 4.0 8.0 20 −66 −2
45 −46 −22
R inferior parietal cortex

Gender Dysphoria, Brain Anatomy and Function

R cuneus+precuneus 4.3 13.0 0 −91 12

HeM–FtM-TR HeW–FtM-TR HeW–HeM

(b) The pregenual anterior cingulate cortex ( pACC) seed
PACC
Left, right precuneus 2.4 3.6 −6 −56 48 3.9 4.1 −8 −56 42
Right thalamus, covering R caudate* 4.3 4.1 26 14 4 3.5 3.2 12 −2 14
Right insular cortex* 45 20 −6
Posterior cingulate 3.7 1.4 10 −58 26

HeM–FtM-TR HeW–FtM-TR HeW–HeM

(c) The temporo parietal junction (TPJ) seed
Left TPJ
Right ACC 4.0 2.3 2 58 6 ns
Right fusiform gyrus, cuneus, precuneus, EBA 3.5 7.6 14 −76 −12 3.7 5.1 6 −66 6
3.8 4.2 32 −88 6
Left inferior frontal gyrus −3.7 1.6 −28 16 26 −4.4 1.6 −22 26 22
Right middle frontal gyrus (and pACC) −4.2 2.0 24 52 14
Right TPJ ns ns ns

Manzouri et al.
Note: Clusters were detected at P < 0.05 FDR corrected at cluster level. Italics denote subsignificant clusters (P < 0.1 FDR corrected at cluster level). Negative Z-values indicate significant differences when running the respective contrast in
opposite direction; *Confluent cluster.
Ebj, extrastriatal body area; Tpj, temporo parietal junction; pACC, pregenual anterior cingulate cortex.

| 9
Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016
 10 | Cerebral Cortex
Figure 3. Resting-state functional connections in FtM-TR compared with controls.
Each raw indicates images showing group differences in functional connectivity
from the respective seed region ( pACC = pregenual anterior cingulate.
Ramy = right amygdala; R TPJ = right temporoparietal junction).
showed patterns of cerebral sex dimorphism congruent with fe-
male cis-controls. The body estrangement feeling is, however, re-
ported by both FtM-TR and MtF-TR, raising the question as to
whether, and to which extent the presently observed changes
in FtM-TR could be general to GD. On ongoing investigations of
MtF-TR and using the identical methodology suggest that this in-
deed might be the case.
In contrast to the reported studies of FtM-TR, which seem
to be predominantly based on gynephilic transsexuals (homo-
sexual in relation to the gender assigned at birth), the available
data from MtF-TR are mixed, and based on investigations
of gynephilic MtF-TR (Savic and Arver 2011) or androphilic trans-
sexuals (Zubiaurre-Elorza et al. 2013), or mixed andro and gyne-
phillic (Luders et al. 2012). In 2 independent studies, MtF-TR were
found to have areas with thicker cortex compared with male con-
trols (Luders et al. 2012; Zubiaurre-Elorza et al. 2013). While the
study by Zubiaurre-Elorza found major differences around mid-
line (in the orbitofrontal, insular, and occipital cortex), Luders
et al. reported differences also in the pre- and post-central
gyrus and in perisylvian regions. Our previous analyses of the
GMV in MtF-TR showed that, in several brain areas, they did
not differ significantly from controls of same gender assigned
at birth, but they did differ from both control groups in regard
to the GMV in the right inferior frontal cortex, the right TPJ, the
thalamus, and the putamen (Savic and Arver 2011). Given the
variability and sparsity of the previously published data on Cth,
more extensive studies using a well-characterized population of
MtF-TR are called for in order to further investigate whether the
presently observed changes in Cth and resting-state connectivity
are more general for GD, whether they apply to MtF-TR, and how
related they are to sexual orientation.
Methodological Issues and Possible Confounding Factors
The MR methodology and analysis methods employed were stan-
dardized and have been validated in several previous publica-
tions (Fischl 2012; Savic and Arver 2014; Savic 2015). With
respect to subject selection, it is important to emphasize that
none of the participants had undergone sex hormone treatment,
and they had estrogen and testosterone values, which were with-
in normal range for their gender assigned at birth, as shown in
Table 1. The results cannot be attributed to possible comorbidity
with psychiatric conditions, as this was a criterion for exclusion,
nor can they be explained by autistic traits, which were specific-
ally evaluated during the clinical assessment.
One might argue that living with the perception of being
“trapped in the wrong body” exposes transsexual persons to con-
stant psychosocial stress, which in itself is associated with cere-
bral changes. Our recent studies show, however, that chronic
stress is associated with a reduction of Cth in the mPFC (Savic
2015), which would not go along with the finding that FtM-TR dis-
played comparatively greater Cth in this region. Also, concomi-
tant depression could not explain the present observations, as
none of the transsexual subjects fulfilled the criteria for depres-
sion according to the clinical assessment or the M.I.N.I. scores.
A factor needing special comment is sexual orientation. This
issue has been discussed primarily in MtF-TR. Blanchard pro-
posed that in subjects who are transsexual homosexuals (homo-
Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016
sexual in relation to their sex assigned at birth), differences
would appear in sexually dimorphic structures, whereas in non-
homosexual transsexuals differences would emerge in brain
structures that do not show sex differences. Blanchard considers
nonhomosexual transsexuals as heterosexual and bisexual
(Blanchard et al. 1995; Smith et al. 2005). According to Kinsey
scores, 13 of our transsexual subjects were homosexual rating
Kinsey 6, 15 rated Kinsey <6. To further investigate possible im-
pact of sexual orientation on cerebral structure, we carried out
post hoc analysis of Cth and GMV and the left putamen volume
(metrices that showed significant difference between FtM-TR
and both control groups) in relation to the HeW and HeM. Inter-
estingly, the major results remained, and the significant increase
in the Cth, and GMV (P < 0.01 with Monte Carlo correction) was
detected around the frontal and occipital pole in FtM-TR rating
Kinsey 6 and those rating <Kinsey 6 in relation to both male
and female controls. Results from the analysis of Cth and GMV
are shown as supplemental information (Supplementary Figs
1A,B and 2A,B). With respect to the relative volume of the left pu-
tamen, both subgroups of FtM-TR showed higher values than
both control groups, but only at P < 0.1 Thus, the results seem
not related to sexual orientation. In this respect, it is also worth
noting that, to the best of our knowledge, the only report about
Cth in regard to sexual orientation showed “cortical thinning”
in the right precuneus and cuneus in homosexual men compared
with HeM (Abe et al. 2014), which is at variance to the correspond-
ing measures in our FtM-TR (“greater” Cth compared with both
HeM and HeW). Our FtM-TR population differed from both con-
trol groups also with respect to functional connectivity in the
own-body referential networks. No such difference was detected
between HeM and HeW. Due to movement artifact in 2 of the FtM-
TR, the subgroup size was too small to allow separate subgroup
comparisons with respect to Kinsey scale. It is worth noting,
however, that our previous data on amygdala connectivity
showed a pattern congruent with the opposite sex in homosexual
persons (Savic and Lindstrom 2008b). The amygdala connectivity
in our FtM-TR subjects followed the pattern of their sex assigned
at birth. Although it is unlikely that any of the present findings
were related to different sexual orientation, a categorical distinc-
tion between possible cerebral correlates of sexual orientation
and identity will require additional comparisons with cis-sexual
homosexual controls. Further studies are warranted to elucidate
this important distinction.
In regard to the possible etiology, it is important to note that
the cross-sectional design of the study precludes establishing
cause and effect. Therefore, we cannot determine whether a
 Gender Dysphoria, Brain Anatomy and Function Manzouri et al.
biological propensity for diffuse or ambiguous body identification
leads FtM-TR individuals to identify with the opposite gender or
if identification with the opposite gender contributes to diffuse or
ambiguous body identification. Given that an avoidance of view-
ing one’s own body is a common behavioral feature of GD (Cohen-
Kettenis and Pfafflin 2010; Bandini et al. 2013; Becker et al. 2015), it
is possible that it contributes to a less crystallized body identity,
at least for visual identification.
In summary, the present data suggest that transsexual per-
sons may differ from cis-sexual controls with respect to the
own-body image networks, which might be one neurobiological
substrate to their condition. Collectively, these convergent find-
ings posit a tentative neurobiological basis for the dysphoric ex-
perience of transsexuals, at least among FtM-TR. The data
strengthen the notion that there is an observable and measurable
biological basis for gender identity and puts GD into the realm of
human physiological variation as also recently suggested by
others (Hahn et al. 2014; Kranz et al. 2014). Whether and how
this neurobiological substrate varies among various populations
of transsexuals, if it is innate or acquired, and how it may be af-
fected by sex hormone treatment are important issues to target
in the near future.
Supplementary Material
Supplementary material can be found at: http://www.cercor.
oxfordjournals.org/.
Funding
This work was supported by grants from the Swedish Science
Council (I.S. Dnr 2007-3107); Stockholm Brain Institute (I.S.);
FORTE (I.S.); AFA (I.S.).
Notes
We are especially indebted to Dr. Cecilia Dhejne for her input dur-
ing discussion of patient recruitment, presented in other papers
with this population, where she is coauthor. We are also ex-
tremely grateful to Dr. Stefan Arver for patient recruitment and
selection. Conflict of interest: None.
Ethical Standards
The authors assert that all procedures contributing to this work
comply with the ethical standards of the relevant national and
institutional committees on human experimentation and with
the Helsinki Declaration of 1975, as revised in 2008.
References
Abe C, Johansson E, Allzen E, Savic I. 2014. Sexual orientation re-
lated differences in cortical thickness in male individuals.
PloS one. 9:e114721.
American Psychiatric Association. 2013. Diagnostic and
Statistical Manual of Mental Disorders: DSM-5. 5th ed. Wash-
ington (DC): American Psychiatric Association.
Bandini E, Fisher AD, Castellini G, Lo Sauro C, Lelli L,
Meriggiola MC, Casale H, Benni L, Ferruccio N, Faravelli C,
et al. 2013. Gender identity disorder and eating disorders:
similarities and differences in terms of body uneasiness. J
Sexual Med. 10:1012–1023.
Becker I, Nieder TO, Cerwenka S, Briken P, Kreukels BP, Cohen-
Kettenis PT, Cuypere G, Haraldsen IR, Richter-Appelt H.
| 11
2015. Body Image in Young Gender Dysphoric Adults: A Euro-
pean Multi-Center Study. Arch Sex Behav. 29:1–8.
Bentz EK, Schneeberger C, Hefler LA, van Trotsenburg M,
Kaufmann U, Huber JC, Tempfer CB. 2007. A common poly-
morphism of the SRD5A2 gene and transsexualism. Reprod
Sci. 14:705–709.
Berglund H, Lindstrom P, Savic I. 2006a. Brain response to puta-
tive pheromones in lesbian women. Proc Natl Acad Sci USA.
103:8269–8274.
Berglund H, Lindstrom P, Savic I. 2006b. Brain response to puta-
tive pheromones in lesbian women (vol 103, pg 8269, 2006).
Proc Natl Acad Sci USA. 103:11098–11098.
Biswal B, Yetkin FZ, Haughton VM, Hyde JS. 1995. Functional con-
nectivity in the motor cortex of resting human brain using
echo-planar MRI. Magn Reson Med. 34:537–541.
Blanchard R. 1993. Varieties of autogynephilia and their relation-
ship to gender dysphoria. Arch Sex Behav. 22:241–251.
Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016
Blanchard R, Clemmensen LH, Steiner BW. 1987. Heterosexual
and homosexual gender dysphoria. Arch Sex Behav.
16:139–152.
Blanchard R, Dickey R, Jones CL. 1995. Comparison of height and
weight in homosexual versus nonhomosexual male gender
dysphorics. Arch Sex Behav. 24:543–554.
Blanke O. 2012. Multisensory brain mechanisms of bodily self-
consciousness. Nature Rev. 13:556–571.
Blanke O. 2004. Out of body experiences and their neural basis.
BMJ. 329:1414–1415.
Brozzoli C, Ehrsson HH, Farne A. 2014. Multisensory repre-
sentation of the space near the hand: from perception to
action and interindividual interactions. Neuroscientist.
20:122–135.
Cohen-Kettenis PT, Gooren LJ. 1999. Transsexualism: a review of
etiology, diagnosis and treatment. J Psychosomatic Res.
46:315–333.
Cohen-Kettenis PT, Pfafflin F. 2010. The DSM diagnostic criteria
for gender identity disorder in adolescents and adults. Arch
Sexual Behav. 39:499–513.
Cohen-Kettenis PT, van Goozen SH, Doorn CD, Gooren LJ. 1998.
Cognitive ability and cerebral lateralisation in transsexuals.
Psychoneuroendocrinology. 23:631–641.
Coleman E, Bockting W, Botzer M, Cohen-Kettenis P,
DeCuypere G, Feldman J, Fraser L, Green J, Knudson G,
Meyer WJ, et al. 2012. Standards of care for the health of trans-
sexual, transgender, and gender-nonconforming people, Ver-
sion 7. Int J Transgenderism. 13:165–232.
Constantino JN, Gruber CP. 2005. Social Responsiveness Scale. Los
Angeles: Western Psychological Services.
Costantini M, Urgesi C, Galati G, Romani GL, Aglioti SM. 2011.
Haptic perception and body representation in lateral and
medial occipito-temporal cortices. Neuropsychologia.
49:821–829.
de Vries ALC, Noens ILJ, Cohen-Kettenis PT, van Berckelaer-
Onnes IA, Doreleijers TA. 2010. Autism spectrum disorders
in gender dysphoric children and adolescents. J. Autism
Develop. Disorders. 40:930–936.
Devinsky O, Feldmann E, Burrowes K, Bromfield E. 1989. Auto-
scopic phenomena with seizures. Arch Neurol. 46:1080–1088.
Filipek PA, Richelme C, Kennedy DN, Caviness VS Jr. 1994. The
young adult human brain: an MRI-based morphometric ana-
lysis. Cereb Cortex. 4:344–360.
Fischl B. 2012. FreeSurfer. NeuroImage. 62:774–781.
Fischl B, Dale AM. 2000. Measuring the thickness of the human
cerebral cortex from magnetic resonance images. Proc Natl
Acad Sci USA. 97:11050–11055.
 12 | Cerebral Cortex
Fischl B, Salat DH, Busa E, Albert M, Dieterich M, Haselgrove C,
van der Kouwe A, Killiany R, Kennedy D, Klaveness S, et al.
2002. Whole brain segmentation: automated labeling of
neuroanatomical structures in the human brain. Neuron.
33:341–355.
Fischl B, van der Kouwe A, Destrieux C, Halgren E, Segonne F,
Salat DH, Busa E, Seidman LJ, Goldstein J, Kennedy D, et al.
2004. Automatically parcellating the human cerebral cortex.
Cereb Cortex. 14:11–22.
Fisher AD, Castellini G, Bandini E, Casale H, Fanni E, Benni L,
Ferruccio N, Meriggiola MC, Manieri C, Gualerzi A, et al.
2014. Cross-sex hormonal treatment and body uneasiness
in individuals with gender dysphoria. J Sexual Med. 11:
709–719.
Garcia-Falgueras A, Swaab DF. 2008. A sex difference in the hypo-
thalamic uncinate nucleus: relationship to gender identity.
Brain. 131:3132–3146.
Giedd JN, Castellanos FX, Rajapakse JC, Vaituzis AC, Rapoport JL.
1997. Sexual dimorphism of the developing human brain.
Progress Neuro-Psychopharmacol Biol Psychiatry. 21:
1185–1201.
Goldstein JM, Seidman LJ, Horton NJ, Makris N, Kennedy DN,
Caviness VS Jr, Faraone SV, Tsuang MT. 2001. Normal sexual
dimorphism of the adult human brain assessed by in vivo
magnetic resonance imaging. Cereb Cortex. 11:490–497.
Green R, Keverne EB. 2000. The disparate maternal aunt-uncle
ratio in male transsexuals: an explanation invoking genomic
imprinting. J Theor Biol. 202:55–63.
Hahn A, Kranz GS, Kublbock M, Kaufmann U, Ganger S,
Hummer A, Seiger R, Spies M, Winkler D, Kasper S, et al.
2014. Structural connectivity networks of transgender people.
Cereb Cortex. 10:3527–3534.
Heydrich L, Lopez C, Seeck M, Blanke O. 2011. Partial and full own-
body illusions of epileptic origin in a child with right tempor-
oparietal epilepsy. Epilepsy Behavior. 20:583–586.
Hodzic A, Kaas A, Muckli L, Stirn A, Singer W. 2009a. Distinct cor-
tical networks for the detection and identification of human
body. NeuroImage. 45:1264–1271.
Hodzic A, Muckli L, Singer W, Stirn A. 2009b. Cortical responses to
self and others. Hum Brain Map. 30:951–962.
Kilpatrick LA, Zald DH, Pardo JV, Cahill LF. 2006. Sex-related dif-
ferences in amygdala functional connectivity during resting
conditions. NeuroImage. 30:452–461.
Kinsey AC, Pomeroy WB, Martin CE, Gebhard PH. 1953. Sexual
Behavior in the Human Female. Bloomington (IN): Indiana
University Press.
Kranz GS, Hahn A, Kaufmann U, Kublbock M, Hummer A,
Ganger S, Seiger R, Winkler D, Swaab DF, Windischberger C,
et al. 2014. White matter microstructure in transsexuals and
controls investigated by diffusion tensor imaging. J Neurosci.
34:15466–15475.
Kruijver FP, Zhou JN, Pool CW, Hofman MA, Gooren LJ, Swaab DF.
2000. Male-to-female transsexuals have female neuron
numbers in a limbic nucleus. J Clin Endocrinol Metab.
85:2034–2041.
Lawrence AA. 2006. Clinical and theoretical parallels between de-
sire for limb amputation and gender identity disorder. Arch
Sex Behavior. 35:263–278.
Lentini E, Kasahara M, Arver S, Savic I. 2013. Sex differences in the
human brain and the impact of sex chromosomes and sex
hormones. Cereb Cortex. 23:2322–2336.
Lopez C, Halje P, Blanke O. 2008. Body ownership and
embodiment: vestibular and multisensory mechanisms.
Neurophysiologie Clinique Clinical Neurophysiol. 38:149–161.
Luders E, Sanchez FJ, Gaser C, Toga AW, Narr KL, Hamilton LS,
Vilain E. 2009. Regional gray matter variation in male-to-fe-
male transsexualism. NeuroImage. 46:904–907.
Luders E, Sanchez FJ, Tosun D, Shattuck DW, Gaser C, Vilain E,
Toga AW. 2012. Increased cortical thickness in male-to-
female transsexualism. J Behav Brain Sci. 2:357–362.
Moseley GL, Gallace A, Spence C. 2012. Bodily illusions in health and
disease: physiological and clinical perspectives and the concept
of a cortical “body matrix.” Neurosci Biobehav Rev. 36:34–46.
Nichols T, Brett M, Andersson J, Wager T, Poline JB. 2005.
Valid conjunction inference with the minimum statistic.
NeuroImage. 25:653–660.
Northoff G, Heinzel A, de Greck M, Bermpohl F, Dobrowolny H,
Panksepp J. 2006. Self-referential processing in our brain—a
meta-analysis of imaging studies on the self. NeuroImage.
31:440–457.
Northoff G, Panksepp J. 2008. The trans-species concept of self
Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016
and the subcortical-cortical midline system. Trends Cogn
Sci. 12:259–264.
Panizzon MS, Fennema-Notestine C, Eyler LT, Jernigan TL, Prom-
Wormley E, Neale M, Jacobson K, Lyons MJ, Grant MD,
Franz CE, et al. 2009. Distinct genetic influences on cortical sur-
face area and cortical thickness. Cereb Cortex. 19:2728–2735.
Peelen MV, Downing PE. 2005. Within-subject reproducibility of
category-specific visual activation with functional MRI.
Human Brain Map. 25:402–408.
Rakic P. 1988. Defects of neuronal migration and the pathogen-
esis of cortical malformations. Prog Brain Res. 73:15–37.
Rakic P. 1995. The development of the frontal lobe. A view from
the rear of the brain. Adv Neurol. 66:1–6; discussion. 6–8.
Rametti G, Carrillo B, Gomez-Gil E, Junque C, Segovia S, Gomez A,
Guillamon A. 2011. White matter microstructure in female to
male transsexuals before cross-sex hormonal treatment. A
diffusion tensor imaging study. J Psychiatr Res. 45:199–204.
Raz N, Lindenberger U, Rodrigue KM, Kennedy KM, Head D,
Williamson A, Dahle C, Gerstorf D, Acker JD. 2005. Regional
brain changes in aging healthy adults: general trends, individ-
ual differences and modifiers. Cereb Cortex. 15:1676–1689.
Savic I. 2015. Structural changes of the brain in relation to occupa-
tional stress. Cereb Cortex. 6:1554–1564.
Savic I, Arver S. 2014. Sex Differences in cortical thickness and
their possible genetic and sex hormonal underpinnings.
Cereb Cortex. 12:3246–3257.
Savic I, Arver S. 2011. Sex dimorphism of the brain in male-to-fe-
male transsexuals. Cereb Cortex. 21:2525–2533.
Savic I, Lindstrom P. 2008a. PET and MRI show differences in cerebral
asymmetry and functional connectivity between homo- and
heterosexual subjects. Proc Natl Acad Sci USA. 105:9403–9408.
Savic I, Lindstrom P. 2008b. PET and MRI show differences in cerebral
asymmetry and functional connectivity between homo- and
heterosexual subjects. Proc Natl Acad Sci USA. 105:9403–9408.
Schagen SE, Delemarre-van de Waal HA, Blanchard R, Cohen-
Kettenis PT. 2012. Sibling sex ratio and birth order in early-
onset gender dysphoric adolescents. Arch Sexual Behavior.
41:541–549.
Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J,
Weiller E, Hergueta T, Baker R, Dunbar GC. 1998. The Mini-
International Neuropsychiatric Interview (M.I.N.I.): the devel-
opment and validation of a structured diagnostic psychiatric
interview for DSM-IV and ICD-10. J Clin Psychiat. 59(Suppl
20):22–33;quiz 34–57.
Smith YL, van Goozen SH, Kuiper AJ, Cohen-Kettenis PT. 2005.
Transsexual subtypes: clinical and theoretical significance.
Psychiatry Res. 137:151–160.
 Gender Dysphoria, Brain Anatomy and Function Manzouri et al.
Solomon D, McAbee J, Asberg K, McGee A. 2015. Coming out and
the potential for growth in sexual minorities: the role of social
reactions and internalized homonegativity. J Homosexuality.
62:1512–1538.
Sowell ER, Peterson BS, Kan E, Woods RP, Yoshii J, Bansal R, Xu D,
Zhu H, Thompson PM, Toga AW. 2007. Sex differences in cor-
tical thickness mapped in 176 healthy individuals between 7
and 87 years of age. Cereb Cortex. 17:1550–1560.
Swaab DF. 2007. Sexual differentiation of the brain and behavior.
Best Pract Res Clin Endocrinol Metab. 21:431–444.
Urgesi C, Berlucchi G, Aglioti SM. 2004. Magnetic stimulation of
extrastriate body area impairs visual processing of nonfacial
body parts. Current Biol. 14:2130–2134.
View publication stats
| 13
van Goozen SH, Slabbekoorn D, Gooren LJ, Sanders G, Cohen-
Kettenis PT. 2002. Organizing and activating effects of sex hor-
mones in homosexual transsexuals. Behav Neurosci.
116:982–988.
Zhou JN, Hofman MA, Gooren LJ, Swaab DF. 1995. A sex difference
in the human brain and its relation to transsexuality. Nature.
378:68–70.
Zubiaurre-Elorza L, Junque C, Gomez-Gil E, Guillamon A. 2014. Ef-
fects of cross-sex hormone treatment on cortical thickness in
transsexual individuals. J Sexual Med. 11:1248–1261.
Zubiaurre-Elorza L, Junque C, Gomez-Gil E, Segovia S, Carrillo B,
Rametti G, Guillamon A. 2013. Cortical thickness in untreated
transsexuals. Cereb Cortex. 23:2855–2862.
Downloaded from http://cercor.oxfordjournals.org/ at Karolinska Institutet on February 1, 2016