Autism spectrum disorder (ASD) is a neurodevelopmental condition affecting approximately 1-2% of the general population. While research has explored ASD prevalence in cisgender individuals, data on transgender and gender-diverse individuals is limited, particularly from large, non-clinic-based cohorts. This study aims to address this gap by examining the rates of autism diagnosis and associated traits in a large sample of transgender and gender-diverse individuals compared to cisgender individuals. The researchers sought to determine if transgender and gender-diverse individuals exhibit higher rates of autism diagnosis, and elevated or diminished levels of traits associated with autism such as systemizing, empathy, and sensory sensitivity, compared to cisgender individuals. The investigation also included assessing the rates of co-occurring neurodevelopmental and psychiatric conditions (ADHD, bipolar disorder, depression, OCD, learning disorder, and schizophrenia) in both groups. The importance of this research lies in its potential to inform clinical practice, improve mental health care access, and tailor support for transgender and gender-diverse individuals, particularly given the overlap of marginalization faced by both autistic and gender-diverse populations.

Previous research, primarily based on clinic-based studies with small sample sizes, has suggested a link between autism and gender diversity. Some studies showed increased rates of gender diversity in autistic individuals compared to the general population, while others indicated elevated autism diagnosis rates among individuals attending gender dysphoria (GD) clinics. However, these studies often lacked matched control groups and focused on individuals with GD, not the broader transgender and gender-diverse population. Moreover, studies varied significantly in methodology, age ranges, and nationalities, hindering generalization of findings. The existing literature highlights a need for large-scale studies employing diverse methodologies to ascertain the prevalence of autism in transgender and gender-diverse individuals within the general population, rather than only clinical populations. The researchers note a lack of large-scale national or regional registries with information on both gender identity (not restricted to those with GD) and autism diagnosis.

This study utilized data from five independently recruited cross-sectional datasets and one longitudinal dataset, all employing convenience sampling. The datasets included diverse recruitment strategies and participant demographics. The Channel 4 (C4) dataset (N=514,100) was the largest, recruited via an online questionnaire related to a television program on autism. The Musical Universe (MU) dataset (N=85,670) was collected through a website focused on music, personality, and cognition. The Autism Physical Health Survey (APHS) dataset (N=2312) was drawn from an internet-based survey. The IMAGE dataset (N=1803) involved participants from a genetic study on autism and mathematical ability. Finally, the LifeLines dataset (N=37,975) comprised a subset of participants from a large longitudinal cohort study. Participants self-reported their gender identity using various classifications across the datasets; this included options for male, female, transgender, other, and non-binary depending on the specific questionnaire used. Across the datasets, information on autism diagnosis and co-occurring mental health conditions were collected through various self-report questionnaires. The questionnaires also included measures of autistic traits (AQ-10, AQ-50), systemizing (SQ-10), empathy (EQ-10), and sensory sensitivity (SPQ-10). Statistical analyses included chi-square tests, logistic regressions (adjusted for age and educational attainment), and ANOVAs, with bootstrap resampling used to compare odds ratios (ORs) across datasets. "Brain Types," categorized by the standardized difference between empathy and systemizing scores, were also analyzed.

The study revealed a significantly higher prevalence of autism diagnosis among transgender and gender-diverse individuals compared to cisgender individuals across all five datasets. Odds ratios (ORs) ranged from 3.03 to 6.36 after adjusting for age and educational attainment. These findings held consistent across diverse recruitment strategies, suggesting robustness. Transgender and gender-diverse individuals exhibited significantly higher scores on self-report measures of autistic traits (AQ-10, AQ-50), systemizing (SQ-10), and sensory sensitivity (SPQ-10), and significantly lower scores on empathy (EQ-10), compared to cisgender individuals. These differences were observed in both autistic and non-autistic individuals. The effect sizes for gender differences were often larger than those observed between autistic and non-autistic individuals, as well as between cisgender males and cisgender females. Analysis of "Brain Types" further supported this, showing a higher propensity for Systemizing brain types among transgender and gender-diverse individuals compared to cisgender males and females. Moreover, transgender and gender-diverse individuals showed elevated rates of multiple other neurodevelopmental and psychiatric conditions (ADHD, bipolar disorder, depression, OCD, learning disorders, and schizophrenia) compared to cisgender individuals in at least two datasets. Multiple regression analyses indicated that autism and depression had the strongest associations with transgender and gender-diverse identity in the two largest datasets, although this result shows some discrepancy between datasets.

The consistent findings across multiple large datasets with varying recruitment methods strongly suggest a genuine association between gender identity and autism, rather than a false positive. The study's large sample size minimized bias commonly seen in smaller studies. The similar ORs observed between internet-based datasets and a previous study using GD-clinic data suggest that this is true for multiple sample types. The elevated autistic traits, sensory sensitivity, and systemizing alongside reduced empathy observed in transgender and gender-diverse individuals further corroborate this association. The higher rates of other co-occurring conditions highlight the need for comprehensive mental health care considering the complex intersectionality of these factors. However, the discrepancy in findings related to co-occurring conditions between the datasets highlights the need for further research to clarify their relative contributions. This could be due to sample size differences, ascertainment bias, or other cohort characteristics. For example, the C4 dataset is likely to contain individuals with high co-morbidity while the MU study could suffer from under-representation of individuals with specific vulnerabilities. The study does not establish causality, and further research is needed to understand the underlying mechanisms of this association; such mechanisms could include sociological factors related to the experience of gender identity, or biological factors related to prenatal hormone exposure and brain development.

This study provides strong evidence for an increased prevalence of autism diagnosis, autistic traits, and co-occurring neurodevelopmental and psychiatric conditions in transgender and gender-diverse individuals compared to cisgender individuals. The findings underscore the importance of gender-affirming and neurodiversity-informed clinical care for this population. Future research should explore the causal mechanisms underlying this association, investigate differences in autistic characteristics between transgender and cisgender individuals, and develop tailored interventions to support the unique needs of autistic transgender and gender-diverse individuals. This also includes exploring better methods of diagnostic screening for this population given the increased rates of undiagnosed autism.

Several limitations should be considered. The study relied on self-reported data, which could be subject to biases. The exclusion of intersex individuals in some datasets may limit the generalizability of findings. Variations in the questions used to ascertain gender identity and autism diagnosis across datasets could introduce some inconsistency. The convenience sampling method employed may not fully reflect the broader population. The study's design does not permit causal conclusions. The statistical power in the LifeLines dataset was somewhat limited given the lower representation of transgender and gender-diverse individuals.