A Multi-Modal MRI Analysis of Cortical Structure in Relation to Gender Dysphoria, Sexual Orientation, and Age in Adolescents

Gender dysphoria (GD) is characterized by distress due to an incongruence between experienced gender and sex assigned at birth. Sex-differentiated brain regions are hypothesized to reflect the experienced gender in GD and may play a role in sexual orientation development. Magnetic resonance brain images were acquired from 16 GD adolescents assigned female at birth (AFAB) not receiving hormone therapy, 17 cisgender girls, and 14 cisgender boys (ages 12–17 years) to examine three morphological and microstructural gray matter features in 76 brain regions: surface area (SA), cortical thickness (CT), and T1 relaxation time. Sexual orientation was represented by degree of androphilia-gynephilia and sexual attraction strength. Multivariate analyses found that cisgender boys had larger SA than cisgender girls and GD AFAB. Shorter T1, reflecting denser, macromolecule-rich tissue, correlated with older age and stronger gynephilia in cisgender boys and GD AFAB, and with stronger attractions in cisgender boys. Thus, cortical morphometry (mainly SA) was related to sex assigned at birth, but not experienced gender. Effects of experienced gender were found as similarities in correlation patterns in GD AFAB and cisgender boys in age and sexual orientation (mainly T1), indicating the need to consider developmental trajectories and sexual orientation in brain studies of GD.

Participants self-reported the education of their mother and father separately by indicating the highest grade completed or the highest degree each parent received. These data were dichotomized into two categories: high school or less versus some post-secondary.

Parents' Marital Status
Participants self-reported their parents' marital status by selecting one of the following options: married to each other, cohabitating with each other, separated, divorced, single, or other (please specify). Participants also indicated the number of years and months that the father has not been living at home (if the father was not living at home). Responses were dichotomized into two categories: parents who were married/cohabitating and those who were not.

Wechsler Intelligence Subtests
Selected subtests from the Wechsler Intelligence Scale for Children (version V; [74]) were administered to participants 12 to 16 years old, and from the Wechsler Adult Intelligence Scale (version IV; [73]) to the 17-year-old participants. Four subtests covering verbal comprehension and visual spatial domains were completed and scaled scores were reported. For the verbal comprehension subtests, the vocabulary and either similarities or comprehension subtests were administered. For two participants, both similarities and comprehension were completed and the similarities scaled score was used. A verbal comprehension scaled score was computed by averaging the two verbal subtest scores (absolute range: 1-19; a score of 10 represents the 50th percentile). From the visual spatial domain, participants completed block design and either visual puzzles or object assembly (from the Wechsler Intelligence Scale for Children, version III; [72]). A visual spatial scaled score was computed by averaging the two visual spatial subtest scores (absolute range: 1-19).

Personal Development Scale and Menstrual Cycle
Pubertal stage in reference to sex assigned at birth was self-reported using the Personal Development Scale (PDS; [75]), which shows good reliability and validity. Participants reported the presence of five (cisgender boys) or four (cisgender girls and gender dysphoria assigned female at birth (GD AFAB)) secondary sex characteristics using a 4-point scale ranging from "no development" to "development has already completed." Participants reported on a growth spurt, skin changes (e.g., pimples), growing hair on their body in places other than their head (e.g., under arms, on/around genitals), growing hair on their face (boys), a change in their voice (boys), and breast growth (girls/GD AFAB). Cronbach's alpha was 0.95 for the boys and 0.70 for the girls/GD AFAB. A mean was calculated from these items where higher scores represented more advanced pubertal development.
Cisgender girls and GD AFAB were also asked whether they had started menstruating.
All AFAB participants (n = 33) answered affirmatively. Participants were further asked about the regularity of their menstrual periods and only two (out of four possible) answers were chosen by participants: "My menstrual periods are not regular (some months they occur and some months they do not occur, the amount of and number of days of bleeding is very different from month to month)" and "Very regularly (every month with approximately 2-6 days of bleeding)." A dichotomous menstrual cycle regularity variable was created to represent participants with regular cycles versus participants with irregular cycles.

Medication Use
Participants completed a medication log of all medications they were currently taking, the dosage, and any side effects. Eight participants were taking one (n = 6) or two (n = 2) medications. Types of medication included SSRIs (Prozac, Zoloft), stimulants (Concerta, Adderall, Ritalin), antipsychotic medication (Risperidone), and birth control pills (Seasonale, Micronor). A dichotomous variable was created representing those not taking any medication (including one participant who was taking Cold FX, a natural health product) versus those taking any medications.

Youth Self-Report
The Youth Self-Report (YSR) is a 118-item self-report questionnaire designed for youth between the ages of 11 and 18 years to assess general psychopathology [76][77][78]. Items were rated on a 3-point scale ranging from 0 to 2 for frequency of occurrence. Potential behavior problems were based on 102 of the items, which are used to index the degree of internalizing, externalizing, and total behavior problems, and the 16 other items reflect socially desirable behaviors ("filler" items that are not scored). Participants completed one of two versions of the YSR (1991 and 2001); the 1991 version contains two gender-specific items (Item 5: "I act like the opposite sex" and Item 110: "I wish I were of the opposite sex"), and the 2001 version contains one gender-specific item (Item 110). These items, along with any other responses indicating challenges related to gender questioning or dysphoria, were scored as 0s to avoid any artificial inflation of the general behavioral problem indices. T scores for the internalizing problems, externalizing problems, and total problems factors were used. Also, dichotomous variables for internalizing problems, externalizing problems, and total problems were calculated, representing whether the T scores were in the clinical range (>90th percentile; T score greater than 63) or not.

Statistical Analyses
Using SPSS version 27, group differences on continuous demographic variables were examined with one-way analyses of variance (ANOVA). In the presence of a significant omnibus effect, post hoc comparisons were conducted with least significant difference (LSD) tests. For categorical variables, Fisher's exact tests were conducted with follow-up Bonferroni-corrected ztests that compare column proportions in the presence of a significant omnibus effect. A twotailed critical p-value of 0.05 was used.
Correlations between each variable and the 228 surface area, cortical thickness, and T1 ROIs were conducted. Bonferroni corrections for multiple tests were utilized for each demographic variable (adjusted ps = 0.05/228 = 0.000219).

Results: Group Differences
For all continuous variables, there were no extreme deviations from normality based on skewness and kurtosis values, which were less than |2|.
One-way ANOVAs for verbal comprehension index, visual spatial index, PDS score, and externalizing problems T score indicated no significant between-group differences (see Table   S1). There was a significant difference for the internalizing problems T score such that the GD AFAB group scored significantly higher than both the cisgender girls (mean difference [MD] = 7.56, SE = 3.47, p = 0.035) and cisgender boys (MD = 16.05, SE = 3.64, p < 0.001), and the cisgender girls scored significantly higher than the cisgender boys (MD = 8.49, SE = 3.54, p = 0.021). There was also a significant difference for total problems T score such that the GD AFAB group (MD = 12.65, SE = 3.27, p < 0.001) and the cisgender girls (MD = 7.30, SE = 3.17, p = 0.026) scored significantly higher than the cisgender boys. The GD AFAB group did not score significantly higher than the cisgender girls (MD = 5. 35, SE = 3.12, p = 0.093). This is the full sample size for each group. If there were missing data, the sample size is indicated with the relevant variable. b Indicates absolute range. c A T score above 63 is considered to be in the clinical range. For Internalizing T scores, the lower bound T score is 27 for birth-assigned females and the lower bound T score is 30 for birthassigned males.
Fisher's exact tests for ethnicity, education of mother, education of father, externalizing problems in the clinical range, and regularity of menstrual cycle indicated these variables were not related to group (see Table S1). There was a significant association between marital status and group. The GD AFAB group was significantly more likely to have parents who were separated/widowed/divorced than married/cohabitating, the cisgender boys were significantly more likely to have parents who were married/cohabitating than separated/widowed/divorced, and there was no significant difference between the two categories related to marital status within cisgender girls. Medication use was also related to group. The GD AFAB participants were significantly more likely to be taking medication than not taking medication, whereas the cisgender boys and girls were significantly more likely to be not taking medication than taking medication, which is not surprising given mental health diagnoses were an exclusion criterion for cisgender participants. Variables representing whether internalizing problems and total problems were in the clinical range were both associated with group. For both, the GD AFAB group was significantly more likely to have a clinical flag than a non-clinical flag, the cisgender boys were significantly more likely to have a non-clinical flag than a clinical flag, and there was no significant difference between clinical or non-clinical flag within cisgender girls.

Results: Correlations with ROIs
The full correlation table can be found in the Excel file titled "SkorskaetalGDBrainPubCorrelationsDemoROIsN47.xlsx" [47]. Table S2 provides a summary of the number of significant correlations and the number of significant correlations after Bonferroni correction. Also, the largest correlation is provided, including the ROI for which this occurred. After correction for multiple comparisons, there were no significant correlations between any of the demographic variables and the ROI data, including demographic variables that showed significant group differences (i.e., internalizing problems T score and clinical flag, total problems T score and clinical flag, parent marital status, and medication use).  Figure S1. Examples of T1 Histograms. Panel A. A T1 histogram from the left and right hemispheres of the middle occipital gyrus is shown. In this case, the distribution of T1 values was positively skewed. A spline function was fit to the histogram to determine the maximum T1 value (mode). Panel B. A T1 histogram from the left and right hemispheres of the insula is shown. Here, the distribution of T1 values was more symmetrical around the mean. Figure S3. Scatterplots showing the brain score-behavior correlations for all groups and behavioral measures. These correlations are also shown in Fig. 3, Panel A as bar plots with 95% CIs. Left column is the strength of sexual attraction (MagSexAttr). Center column is the direction of sexual attraction (SexAttr) with a range from exclusively androphilic (11°) to exclusively gynephilic (79°). The dotted line at 45°indicates ambiphilic or asexual and the range is indicated by the dot-dashed line. Right column is age (months). The strength of the brain-behavior correlation is indicated at the top of each plot, and stable correlations are indicated by filled markers. CB = cisgender boys, GD = GD AFAB, CG = cisgender girls.