Edited by: Hubert Vaudry, University of Rouen, France
Reviewed by: Alexander S. Kauffman, University of California San Diego, USA; Hubert Vaudry, University of Rouen, France
This article was submitted to Neuroendocrine Science, a section of the journal Frontiers in Endocrinology.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
The odorous steroid androstadienone, a putative male chemo-signal, was previously reported to evoke sex differences in hypothalamic activation in adult heterosexual men and women. In order to investigate whether puberty modulated this sex difference in response to androstadienone, we measured the hypothalamic responsiveness to this chemo-signal in 39 pre-pubertal and 41 adolescent boys and girls by means of functional magnetic resonance imaging. We then investigated whether 36 pre-pubertal children and 38 adolescents diagnosed with gender dysphoria (GD; DSM-5) exhibited sex-atypical (in accordance with their experienced gender), rather than sex-typical (in accordance with their natal sex) hypothalamic activations during olfactory stimulation with androstadienone. We found that the sex difference in responsiveness to androstadienone was already present in pre-pubertal control children and thus likely developed during early perinatal development instead of during sexual maturation. Adolescent girls and boys with GD both responded remarkably like their experienced gender, thus sex-atypical. In contrast, pre-pubertal girls with GD showed neither a typically male nor female hypothalamic activation pattern and pre-pubertal boys with GD had hypothalamic activations in response to androstadienone that were similar to control boys, thus sex-typical. We present here a unique data set of boys and girls diagnosed with GD at two different developmental stages, showing that these children possess certain sex-atypical functional brain characteristics and may have undergone atypical sexual differentiation of the brain.
In humans, the odorous steroid 4,16-androstadien-3-one (androstadienone) has been studied intensively as a putative male modulator chemo-signal. Androstadienone, probably synthesized in the gonads (
In line with the effects androstadienone had on the autonomic nervous system and on behavior, studies conducted by Savic et al. (
Secretion of androstadienone increases substantially during puberty. As a result, sex-related changes in olfactory sensitivity to androstenes have been reported during adolescence (
Sisk and colleagues (
Of note, further studies by Savic et al. (
At the Center of Expertise on Gender Dysphoria in Amsterdam, the current treatment protocol allows adolescents diagnosed with GD that persisted from childhood into adolescence to start treatment with gonadotropin-releasing hormone analogs (GnRHa) from the age of 12 years, to suppress endogenous gonadal stimulation and thus the development of irreversible sex characteristics of the natal sex (
A second aim of the current study was therefore to explore whether children and adolescents, diagnosed with GD, would show brain responses that reflect their expressed/experienced gender rather than their natal sex, and whether these would vary as a function of their developmental phase. Four groups of subjects, all diagnosed with GD, participated in the current study: girls and boys that were pre-pubertal and treatment-naïve, and girls and boys that were adolescent in age (though in hypogonadal state due to GnRHa treatment). None of our participants received cross-sex hormones at the time of data acquisition.
The initial study sample consisted of a total of 158 participants. Four subjects were excluded from further analysis, because of anatomical anomalies (one adolescent), technical errors during data collection (one child), or because the diagnosis GD had been revised since their participation in the study (two boys with GD in remission). All participants diagnosed with GD, were recruited via the Center of Expertise on Gender Dysphoria at the VU University Medical Center in Amsterdam. The control participants were recruited via several primary and secondary schools in the Netherlands and by inviting friends and relatives of the participants with GD.
The children sample consisted of 19 control girls [mean years of age (
The adolescent groups consisted of 21 control girls (
Sexual orientation was difficult to assess, especially in the pre-pubertal sample, because most children were simply too young to be able to report their sexual orientation. Therefore, current or presumed future sexual attraction was assessed by asking whether the participant had ever been in love with somebody, and if yes, whether that person was a boy or a girl. Normal olfactory function was ascertained by means of an extended version of the “Sniffin’ Sticks” test battery (32-item odor identification test and olfactory threshold measurement) (
Androstadienone (Steraloids Inc., Newport, RI 02840, USA) was diluted in propylene glycol (Sigma) to a concentration of 10 mM, according to the “high” concentration used in our previous study (
Scans were performed on a 3.0-T GE Signa HDxt scanner (General Electric, Milwaukee, WI, USA). A gradient echo, echo planar imaging sequence was used for functional imaging (19.2 cm2 field of view, TR of 1950 ms, TE of 25 ms, an 80° flip angle, isotropic voxels of 3 mm, and 36 slices). Before each imaging session, a local high-order shimming technique was used to reduce susceptibility artifacts. A scanning session consisted of six alternating ON-OFF cycles over 108 volumes in a classical block design (one block consisted of nine volumes), lasting 3.6 min. For co-registration with the functional images, a T1-weighted scan was obtained (3D FSPGR sequence, 25 cm2 field of view, TR of 7.8 ms, TE of 3.0 ms; slice thickness of 1 mm, and 176 slices).
Data analysis was performed with SPM8 software (Statistical Parametric Mapping; Wellcome Department of Imaging Neuroscience, Institute of Neurology at the University College London, UK) implemented in Matlab R2009b (Math Works Inc., Natick, MA, USA). Functional images were slice-timed and realigned to the mean image, followed by unwarp. Applying the “New Segment” and “Create Template” options of the DARTEL (Diffeomorphic Anatomical Registration Through Exponentiated Lie Algebra) toolbox, structural images were segmented. Then, gray matter and white matter images were used for creating age-group specific templates (one for the children and the adolescents sample each), registered in Montreal neurological institute (MNI) space. Functional images were spatially normalized to their respective group-template, applying each individual’s DARTEL flow field, and finally, images were smoothed by means of a 5-mm full width half maximum (FWHM) isotropic Gaussian kernel.
Individual image data were analyzed using boxcar regressors convolved with a synthetic hemodynamic response function and a first-order time-modulation (TM) regressor to test for possible effects of adaptation/sensitization to androstadienone. In order to account for assumed late “wash-out” effects during OFF blocks and an early peak response to the odor stimulation during ON blocks, first-level contrast images were built by subtracting the second half (b) of the OFF blocks (four volumes) from the first part (a) of the ON blocks (four volumes). Accordingly, this was done with the associated TM regressor blocks. Further, based on the image realignment process, individual head jerks were identified (>1 mm displacement) (
First, in order to test whether the sex difference in response to androstadienone was present in both developmental control groups, and whether that sex difference in responsiveness varied as a function of adaptation/sensitization to the odor, we conducted a
Second, by means of four
Analyses were restricted to the hypothalamus area as region of interest (ROI), defined [with Marsbar (
Demographic, self-report, and subject characteristics are presented in Table
Pre-pubertal children |
Adolescents |
|||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Ctrl girls | Ctrl boys | Girls with GD | Boys with GD | Ctrl girls | Ctrl boys | Girls with GD | Boys with GD | |||||||
Group size | 19 | 20 | 17 | 19 | 21 | 20 | 21 | 17 | ||||||
Age in years | Mean (SD) | 9.7 (0.9) | 9.5 (1.1) | 9.6 (1.1) | 10.4 (0.9) | 3.1 (3.71) | 16.3 (0.9) | 15.9 (0.6) | 16.1 (0.8) | 15.3 (1.2) | 4.6 (3.75) | |||
Pubertal stage | Mean (SD) | P | 1 | 1 | 1 | 1 | – | – | 4.2 (0.7) | 4.7 (0.7) | 4.7 (0.6) | 3.1 (1.1) | 17.4 (3.74) | |
G/M |
1 | 1 | 1 | 1 | – | – | 4.1 (0.8) | 4.1 (0.8) | 4.1 (1.1) | 3.1 (0.8) | 5.1 (3.74) | |||
Sexual orientation | % ( |
Gynephilic | – | 66.0 (13) | 17.6 (5) | 21.1 (4) | – | 100 (20) | 100 (21) | – | ||||
Androphilic | 89.5 (17) | – | 29.4 (3) | 42.1 (8) | 100 (21) | – | – | 70.6 (12) | ||||||
Ambiphilic | – | – | 5.9 (1) | – | 1.1 (3.71) | 0.364 | – | – | – | 5.9 (1) | 33.3 (3.74) | |||
Don’t know | 10.5 (2) | 35.0 (7) | 47.1 (8) | 36.8 (7) | – | – | – | 17.6 (3) | ||||||
Missing | – | – | – | – | – | – | – | 5.9 (1) | ||||||
Androstadienone |
Mean (SD) | 6.1 (2.4) | 6.9 (1.5) | 6.2 (2.7) | 5.7 (2.6) | 0.7 (3.55) | 0.548 | 5.7 (2.3) | 4.8 (1.9) | 5.5 (1.9) | 4.6 (2.0) | 1.1 (3.63) | 0.351 | |
Sniffin’ sticks | Mean (SD) | Threshold | 7.3 (3.9) | 6.2 (4.1) | 8.5 (3.4) | 8.6 (3.4) | 1.6 (3.58) | 0.212 | 9.9 (3.3) | 8.5 (3.4) | 8.1 (2.9) | 9.1 (3.4) | 1.1 (3.68) | 0.356 |
Identification | 21.5 (3.6) | 18.5 (3.8) | 19.1 (4.0) | 19.9 (5.3) | 1.8 (3.66) | 0.162 | 25.2 (3.8) | 23.9 (2.8) | 23.5 (3.9) | 23.7 (2.8) | 1.0 (3.71) | 0.419 |
In order to determine whether hypothalamic activation upon smelling androstadienone is dependent on puberty, we conducted separate ANOVAs in the pre-pubertal and the adolescent groups, and compared hypothalamic activation in control girls to that of control boys during exposure to androstadienone. Results are displayed in Table
Effect | |||||||
---|---|---|---|---|---|---|---|
Children | Gender | 2.5 | -6 | -12 | -5 | 30 | 0.202 |
Odor | 2.4 | -2 | -15 | -2 | 39 | 0.244 | |
ON-OFF | 1.8 | 0 | -10 | -14 | 1 | 0.456 | |
ONTM-OFFTM | 3.2 | -6 | -12 | -5 | 92 | ||
2.9 | 6 | -7 | -8 | ||||
Adolescents | Gender | 1.7 | -3 | -7 | -2 | 17 | 0.091 |
Odor | 1.7 | 6 | -7 | -6 | 8 | 0.655 | |
ON-OFF | 3.4 | 6 | -7 | -8 | 1 | 0.522 | |
ONTM-OFFTM | 2.5 | -6 | -12 | -5 | 115 | ||
Children | No effects | ||||||
Adolescents | Gender | 3.5 | 6 | -7 | -9 | 76 | |
Odor | 1.7 | -6 | -7 | -6 | 1 | 0.627 | |
ON-OFF | – | – | – | – | – | – | |
ONTM-OFFTM | 4.1 | 6 | -7 | -9 | 214 | ||
Children | Gender | 1.8 | 2 | -14 | -2 | 4 | 0.447 |
Odor | 1.9 | 0 | -10 | 0 | 3 | 0.436 | |
ON-OFF | 1.7 | -4 | -6 | -4 | 12 | 0.437 | |
ONTM-OFFTM | 1.9 | 0 | -10 | -12 | 31 | 0.329 | |
2.4 | 0 | -14 | -12 | ||||
Adolescents | Gender | 2.2 | 6 | -8 | -6 | 11 | 0.237 |
Odor | 3.0 | 0 | -14 | -12 | 1 | 0.341 | |
ON-OFF | 1.8 | 2 | -14 | -2 | 36 | ||
ONTM-OFFTM | – | – | – | – | – | – | |
Children | No effects | ||||||
Adolescents | Gender | – | – | – | – | – | – |
Odor | 1.9 | 3 | -12 | -9 | 12 | 0.535 | |
ON-OFF | – | – | – | – | – | – | |
ONTM-OFFTM | 2.1 | 3 | -10 | -2 | 26 | 0.323 |
Pre-pubertal girls and boys did not differ in terms of general hypothalamic responsiveness (condition effect ON > OFF) to androstadienone, but directional
Accordingly, similar comparisons were done in the adolescent groups. Again, the sex difference in hypothalamic activation (girls > boys) was significantly dependent on the factor time (
In order to test whether individuals diagnosed with GD would show a hypothalamic activation that reflected their experienced gender, and whether this brain response would be different for pre-pubertal and adolescent subjects, four ANOVAs were conducted, in which control girls were compared to girls with GD, and control boys were compared to boys with GD (see Table
Effect | |||||||
---|---|---|---|---|---|---|---|
Children | No effects | ||||||
Adolescents | Gender | – | – | – | – | – | – |
Odor | 2.5 | 6 | -9 | -5 | 25 | 0.203 | |
ON-OFF | 1.7 | -6 | -7 | -5 | 2 | 0.512 | |
ONTM-OFFTM | 2.3 | -6 | -9 | -8 | 25 | 0.258 | |
Children | Gender | 3.0 | -6 | -10 | -6 | 35 | 0.065 |
Odor | 2.6 | 3 | -16 | -9 | 27 | 0.168 | |
ON-OFF | 1.6 | -2 | -16 | -6 | 4 | 0.550 | |
ONTM-OFFTM | 3.5 | -6 | -10 | -6 | 126 | ||
Adolescents | Gender | 2.2 | -2 | -4 | -10 | 5 | 0.336 |
Odor | 1.5 | 2 | -14 | -12 | 4 | 0.699 | |
ON-OFF | – | – | – | – | – | – | |
ONTM-OFFTM | 1.7 | 6 | -8 | -4 | 7 | 0.504 | |
1.6 | -6 | -12 | -8 | 16 | 0.553 | ||
Children | No effects | ||||||
Adolescents | Gender | – | – | – | – | – | – |
Odor | 1.9 | 3 | -12 | -9 | 12 | 0.535 | |
ON-OFF | – | – | – | – | – | – | |
ONTM-OFFTM | 1.9 | -6 | -7 | -11 | 1 | 0.459 |
None of the comparisons between pre-pubertal control girls and pre-pubertal girls with GD revealed any differences in hypothalamic activation upon smelling androstadienone. However, no sex-typical effects (i.e., a female-typical hypothalamus response), when compared to control boys, could be confirmed. Thus, pre-pubertal girls with GD neither differed significantly from their experienced (control boys) nor from their natal sex (control girls) in terms of hypothalamic activation when smelling androstadienone.
In contrast, the comparison of adolescent control girls to girls diagnosed with GD revealed a significant effect of
No significant effects were revealed when comparing pre-pubertal boys with GD with control boys. In contrast, a significant effect of
When adolescent boys with GD were compared to adolescent control boys we observed a significant effect of condition (ON > OFF) (
The present study is, to our knowledge, the first to demonstrate that sex differences in hypothalamic activation upon smelling androstadienone are already present before puberty, and thus may be considered as a sex difference established during early brain development. We found that pre-pubertal as well as adolescent control girls showed a stronger hypothalamic activation compared to boys (at both developmental stages), and that this sex difference was crucially modulated by effects of sensitization to androstadienone.
Previous psychophysiological studies in children and adolescents have shown that olfactory sensitivity to androgenic odors differed between boys and girls during development, due to the increased production of endogenous androgens by boys during puberty and the assumed resulting adaptation to their own body odors (
Gender dysphoria has been hypothesized to develop due to an altered sexual differentiation of the body and the brain during early development (
Pre-pubertal girls with GD did not differ in hypothalamic activation from control girls indicating that they did not show a male-typical response. However, they also did not show a female-typical hypothalamic activation, when compared to control boys, indicating that they did not differ from either of the control groups. It is possible that pre-pubertal girls with GD constitute a rather heterogeneous group with respect to future persisting GD. It has been shown that only about 15.8% of the childhood GD cases will eventually lead to adult GD (
The comparison of adolescent female controls versus adolescent girls with GD revealed very similar results as the control group comparisons, i.e., stronger hypothalamic activations in those subjects experiencing a female gender identity, and this effect was mainly driven by effects over time, reflecting sensitization. Thus, adolescent girls with GD responded remarkably like their experienced gender (control boys). While speculative, these findings fit with the idea that this group of girls with GD (who are more homogenous in terms of future persisting GD compared to the pre-pubertal groups) may have had a more male-typical perinatal hormonal environment, resulting in the development of certain typically male functional sex characteristics of the brain. The hypogonadal state of adolescent girls with GD using GnRHa at the moment of data acquisition is not likely to account for these findings, since our results in the pre-pubertal control groups suggest that hypothalamic responsiveness to androstadienone is probably independent of puberty, and thus not affected by circulating sex hormones. Furthermore, animal studies showed that sexually dimorphic responses to volatile urinary odors were not dependent on circulating sex steroids, but rather developed under the influence of organizational effects of sex hormones (
It should be noted, though, that sexual orientation of the participants with GD might present a confounding factor. Berglund et al. (
The hypothalamic response to androstadienone in the pre-pubertal boys with GD was not significantly stronger than that shown by the pre-pubertal control boys. Moreover, the reverse contrast (control girls > boys with GD) revealed that pre-pubertal control girls showed significantly stronger hypothalamic activation, implying that boys with GD responded according to their natal sex. Again, it is likely that the younger, pre-pubertal sample of boys with GD constitutes a rather heterogeneous group in terms of future persistence of their gender dysphoric feelings (
Adolescent boys with GD showed significantly stronger, thus sex-atypical hypothalamic activations, compared to control boys, although this effect was not modulated by any effects of time, thus sensitization to androstadienone, as we did observe in the control girls. Accordingly, the reverse comparison (control girls > boys with GD) revealed no significant effects. The female-typical activation in adolescent boys with GD is in line with a previous study by Berglund et al. (
Some limitations of the present study should be mentioned and may be addressed in future research. In most comparisons, the gender differences in hypothalamic response to androstadienone showed relatively small effect sizes. These effect sizes, in combination with relatively small group sizes (especially that of the pre-pubertal girls with GD with
In summary, the present study is the first to demonstrate that sex differences in hypothalamic activation upon smelling the chemo-signal androstadienone are not acquired during sexual maturation, under the influence of gonadal hormones during puberty, but may be considered hard-wired responses, which already can be observed in pre-pubertal children. Moreover, the current study is the first to explore sex-atypical hypothalamic responses to androstadienone in male and female individuals with GD at two different developmental stages. Our results indeed suggest that individuals with GD possess certain functional brain characteristics of their experienced gender and may have undergone atypical neuronal sexual differentiation.
Julie Bakker and Sarah M. Burke designed the study set-up. Sarah M. Burke performed the neuroimaging experiments, conducted the data analyses with advice of Dick J. Veltman, and wrote the manuscript. Daniel T. Klink performed the clinical assessments. Peggy T. Cohen-Kettenis, Dick J. Veltman, and Julie Bakker supervised the project. All authors contributed to interpretation of the data and revisions of the manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We thank Prof. Thomas Hummel and his Smell and Taste lab at the University Medical School in Dresden (Germany) for his hospitality and the kind introduction into the field of olfactory fMRI. We thank Paul Groot from the Academic Medical Centre Amsterdam for his help with the data analyses, and we thank Prof Michael Baum from Boston University for his helpful comments on earlier versions of this manuscript. Last but not least, we are especially thankful to Dipl.-Ing. Johannes Burke, who developed and built the olfactometer equipment and programed the application software. This study was funded by a VICI grant (453-08-003) from the Dutch Science Foundation (Nederlandse Organisatie voor Wetenschappelijk Onderzoek) to Dr. Julie Bakker. Dr. Julie Bakker is a senior research associate of the Belgian Fonds National de la Recherche Scientifique.