White matter microstructure in transsexuals and controls investigated by diffusion tensor imaging

Abstract

Biological causes underpinning the well known gender dimorphisms in human behavior, cognition, and emotion have received increased attention in recent years. The advent of diffusion-weighted magnetic resonance imaging has permitted the investigation of the white matter microstructure in unprecedented detail. Here, we aimed to study the potential influences of biological sex, gender identity, sex hormones, and sexual orientation on white matter microstructure by investigating transsexuals and healthy controls using diffusion tensor imaging (DTI). Twenty-three female-to-male (FtM) and 21 male-to-female (MtF) transsexuals, as well as 23 female (FC) and 22 male (MC) controls underwent DTI at 3 tesla. Fractional anisotropy, axial, radial, and mean diffusivity were calculated using tract-based spatial statistics (TBSS) and fiber tractography. Results showed widespread significant differences in mean diffusivity between groups in almost all white matter tracts. FCs had highest mean diffusivities, followed by FtM transsexuals with lower values, MtF transsexuals with further reduced values, and MCs with lowest values. Investigating axial and radial diffusivities showed that a transition in axial diffusivity accounted for mean diffusivity results. No significant differences in fractional anisotropy maps were found between groups. Plasma testosterone levels were strongly correlated with mean, axial, and radial diffusivities. However, controlling for individual estradiol, testosterone, or progesterone plasma levels or for subjects' sexual orientation did not change group differences. Our data harmonize with the hypothesis that fiber tract development is influenced by the hormonal environment during late prenatal and early postnatal brain development.

Keywords: diffusion tensor imaging; gender identity disorder; testosterone; transsexual; white matter microstructure.

Copyright © 2014 the authors 0270-6474/14/3415466-10$15.00/0.

Figures

Figure 1.

Effects of gender identity and sex on regional MD. Depicted are pairwise comparisons between the four study groups (FCs, MCs, FtM transsexuals, and MtF transsexuals), overlaid onto eight axial planes (from z = −30 to 40 mm, MNI space); see schema on the top right. Six comparisons, from top to bottom: FC > FtM, FC > MtF, FC > MC (first column); FtM > MtF, FtM > MC (second column); and MtF > MC (third column), e. g. white matter tracts in red–yellow show significantly higher MD in FCs compared with FtM transsexuals in the top left comparison (p < 0.05, FWE corrected). Left is right.

Figure 2.

Box plots representing estimated MDs (10−3mm2/s, centered around global mean) of the four study groups (FCs, MCs, FtM transsexuals, and MtF transsexuals). Values refer to regional MD corresponding to the crosshair depicted on the left axial template. Because results are corrected for TIV, values are mean centered. Locations are given in millimeters, MNI standard space. Values are derived from the ANCOVA model using TIV as the covariate of no interest. The transition MC–MtF–FtM–FC in significantly increasing MD values becomes apparent in the body of corpus callosum (top row) and the right superior frontal blade (middle row; p < 0.05, FWE corrected). In the right CST (bottom row), significant differences are observed between the two biological sexes (e.g., FC and FtM > MtF and MC), whereas no significant differences are seen between FCs and FtM transsexuals or between MtF transsexuals and MCs. Left is right.

Figure 3.

Effects of gender identity and sex on regional AD; see schema in Figure 1. Depicted are six comparisons, from top to bottom: FC > FtM, FC > MtF, FC > MC (first column); FtM > MtF, FtM > MC (second column); and MtF > MC (third column), e. g. white matter tracts in red–yellow show significantly higher AD, and white matter tracts in blue–green show significantly lower AD in MtF transsexuals compared with MCs in the bottom right comparison (p < 0.05, FWE corrected). Left is right.

Figure 4.

Effects of gender identity and sex on regional RD; see schema in Figure 1. Depicted are six comparisons, from top to bottom: FC > FtM, FC > MtF, FC > MC (first column); FtM > MtF, FtM > MC (second column); and MtF > MC (third column), e.g., white matter tracts in red–yellow show significantly higher RD, and white matter tracts in blue–green show significantly lower RD in FtM transsexuals compared with MtF transsexuals in the first comparison of the second column (p < 0.05, FWE corrected). Left is right.

Figure 5.

Positive association between T plasma levels and regional MD, AD, and RD, i.e., white matter tracts in red–yellow show significant positive associations (p < 0.05, FWE corrected). Data are based on a multiple regression analysis of the entire sample using group as variable of no interest. Left is right.

Figure 6.

Scatter plots depicting the association between diffusivities and age in four major tracts, in right and left CST and Fmajor and Fminor. FCs, Red crosses; FtM transsexuals, blue circles; MtF transsexuals, green triangles; MCs, black squares. For correlations, see Table 3.