Functional Connectivity in Certain Brain Regions May Be Associated with Hormone Therapy Outcomes in Transgender Patients

Symbol of a transgender and female and male gender symbols drawn with chalk on a black background
The researchers used clinical and MRI data before cross-sex hormone therapy as features to train a machine learning model to predict individuals’ post-therapy body congruence (the degree to which photos of their bodies match their self-identities).

Study data published in NeuroImage: Clinical suggest that pre-therapy neuroimaging profiles may be able to predict outcomes of cross-sex hormone therapy in transgender individuals. In a cohort study of transgender women and men, post-therapy reductions in gender dysphoria were predicted by greater pre-therapy connectivity within the cingulo-opercular and fronto-parietal networks. Such data may provide insight into the “body-brain” effects of hormone therapy and support the use of certain pre-therapy characteristics to predict whether individuals may need additional support during hormone replacement therapy.

Participants were recruited from a clinic specializing in gender-affirming medical care in Stockholm, Sweden. Participants underwent magnetic resonance imaging (MRI) at 2 time points: (1) prior to hormone intervention and (2) an average of 14 months post-therapy initiation. During the MRI scans, participants were asked to complete a body morph task, in which they were presented with a set of images taken of their own bodies.

The body photographs were “morphed” towards 5 different iterations of female-presenting bodies and 5 different iterations of male-presenting bodies. While being presented with this continuum of images, participants were asked to rate each one on how closely it represented their own self-image. Results from these body image tasks were used to produce a “body index” (BI) score, or the degree to which each patient felt their body’s physical characteristics aligned with their gender identity.

More negative scores on the BI were indicative of greater congruence between body image and gender identity. The mean MRI-determined activity within 7 regions of interest (ROIs) was compared between the pre- and post-hormone therapy time points. A least absolute shrinkage and selection operator (LASSO) regression model was used to predict post-therapy BI scores using pre-therapy imaging and clinical data.

Data from 16 trans women and 9 trans men were used in analyses. Mean age at enrollment was 25.2 ± 7.8 years. Mean pre-therapy BI score was -10.4 ± 21.8; mean post-therapy score was -23.1 ± 25.7 (P =.002). The majority of patients (n=18; 72%) experienced a significant decrease in BI score over time, indicating better congruence between body image and gender identity. In regression models which incorporated both clinical features and imaging data, functional connectivity in the fronto-parietal network (P <.005) and the cingulo-opercular network (P <.006) were significantly associated with post-therapy BI ratings.

Models which incorporated all 7 ROIs with clinical features were less accurate in predicting post-therapy BI scores. The best-fitting algorithm combined pre-therapy clinical features with functional connectivity within the fronto-parietal and cingulo-opercular networks (P =.001). Predictive clinical features included age, body mass index, years of education, sexual orientation, and duration of hormone therapy. However, clinical features alone did not predict post-therapy congruence.

Per these data, network connectivity in certain brain regions was associated with hormone therapy outcomes in transgender individuals. Specifically, the front-parietal and cingulo-opercular networks were predictive of post-therapy BI scores. It remains unclear why these specific regions were implicated in therapy outcomes.

Regarding study limitations, investigators cited the small cohort size, relatively short follow-up duration, and use of BI score as a measure of body congruence. While the BI is a validated measure of body congruence, it likely fails to capture the full spectrum of patient experiences.

Even so, “[this study] illustrates the potential for predicting hormone therapy responsiveness in transgender individuals with [gender incongruence],” the investigators wrote. “Results could help identify the need for personalized therapies in individuals predicted to have low body-self congruence after standard therapy.”


Moody TD, Feusner JD, Reggente N, et al. Predicting outcomes of cross-sex hormone therapy in transgender individuals with gender incongruence based on pre-therapy resting-state brain connectivity. Neuroimage Clin. Published online December 2, 2020. doi:10.1016/j.nicl.2020.102517