While we do not expect depression to be a paramount predictor, research from within both the autistic and the TGNC communities has suggested that both of these groups face many concerns related to their mental health (de Vries et al., 2011; Pitts et al., 2009), as well as barriers to accessing mental health care. Autistic adults have higher rates of all major psychiatric conditions when compared to non-autistic adults (Croen et al., 2015). Some research has shown that autistic adults who are referred for psychiatric care have an average of three current clinical diagnoses, with the most common diagnoses being attention-deficit hyperactivity disorder (ADHD), anxiety disorders, and major depressive disorder (MDD; Joshi et al., 2013). Autistic adults are also at higher risk for suicidal ideation and attempts to die by suicide (Cassidy et al., 2014; Paquette-Smith et al., 2014), and this is likely related to the high co-occurrence of mental health distress.

As with autistic adults, TGNC people are also more likely than their cisgender peers to have co-occurring mental health concerns such as anxiety, depression, eating disorders, psychosis, and substance use disorders (Barr et al., 2021; Reisner et al., 2016). TGNC people are more likely to have suicidal thoughts and to attempt to die by suicide (Grossman and D’Augelli, 2007; Testa et al., 2012). Though there has not been a great deal of research on the mental health of people identifying as both autistic and gender nonconforming, some evidence suggests autistic people who identify as TGNC or who are experiencing gender dysphoria have worse overall mental health than cisgender autistic people (George and Stokes, 2018; Murphy et al., 2020).

It is important to note that these disparities in mental health are likely related in part to poor treatment of autistic and TGNC people by society. Importantly, anticipation of discrimination appears to drive care avoidance (Kcomt et al., 2020). Both autistic (Gibbs et al., 2022; Weiss and Fardella, 2018) and TGNC (Griner et al., 2020; Marx et al., 2021) people are stigmatized and viewed negatively by peers and are often bullied and victimized because of their differences. Discrimination has a well-established connection with mental health disparities in people of color (Carter et al., 2017; Schmitt et al., 2014) as well as for sexual and gender minorities (Hendricks and Testa, 2012; Meyer, 2003). Evidence suggests that poor mental health in the autistic population is related to discrimination and minority stress in the same ways that it is related in other minoritized groups (Botha and Frost, 2020). As such, the complex trauma of repeated discrimination based on gender expression and autistic traits likely contributes to the development of mental health concerns for those reporting gender nonconformity and neurodivergence.

As mentioned previously, the anxiety that minority group members feel can lead to problems with EF. Executive functioning is a collection of cognitive processes that assist with the top-down control of thoughts and behaviors (Jones et al., 2012). It encompasses abilities related to guiding, directing, and managing cognitive, emotional, and behavioral functions. Executive functioning is related to adaptive functioning (Gilotty et al., 2002), which is the ability to perform daily tasks and to care for oneself. Thus, executive functioning skills are important for navigating the world and doing things like keeping track of tasks, completing tasks, weighing options, making decisions, and navigating the social world efficiently. Accordingly, stronger EF skills are associated with better quality of life, including improved ability to build and maintain relationships and to maintain and care for physical and mental health (de Vries and Geurts, 2015; Dijkhuis et al., 2017). Within this context, anxiety is one pathway through which EF may be affected.

Beyond anxiety alone, broader social mistreatment may also contribute to EF difficulties. When a person has intersecting marginalized identities they are trying to navigate, or they are attempting to manage tasks related to those identities that do not come naturally, their coping systems may be overwhelmed (Schmitt et al., 2014). This idea has been modeled by Hendricks and Testa (2012) and adapted by Botha and Frost (2020). When coping systems are taxed in this way, EF challenges may become more apparent, potentially negatively impacting quality of life.

At the same time, EF difficulties are not specific to minority stress; they are also common in a range of clinical conditions (e.g., ADHD, depression, anxiety) and may be comorbid for genetic and/or other environmental reasons. A meta-analysis by Snyder (2013) reports on differential effects on EF in adult clinical populations (primarily major depressive disorder). Notably, researchers have found these mental health conditions in both the autistic (Hollocks et al., 2019) and TGNC communities (Pinna et al., 2022).

This overlap matters for interpretation of EF findings in autism. It is difficult to study overall executive functioning in autistic people, as results from many studies have been inconsistent in demonstrating strengths and challenges in executive functioning for autistic people (St John et al., 2022). These findings may be unclear due to a lack of understanding of the impact of co-occurring mental health concerns and stressors, including gender dysphoria and navigating the gender-affirming medical care system or managing mental and physical health concerns. However, one clearer pattern is that autistic adults’ EF performance worsens under increased cognitive load (Bodner et al., 2019; Geurts and Vissers, 2012). This may have particular relevance for autistic TGNC individuals, as navigating an unwelcoming society as transgender or gender nonconforming can be a burdensome cognitive task that consistently co-occurs with other everyday cognitive demands (St John et al., 2022).

However, these autism-focused findings and interpretations point to a broader question: whether analogous “load” mechanisms operate in TGNC populations more generally. Addressing this, Strang et al. (2022) developed a model of executive functioning in TGNC people that extends the gender minority stress model (Hendricks and Testa, 2012) by specifying how gender-related stressors and identity-related demands may tax EF resources. This model follows the gender minority stress model (Hendricks and Testa, 2012), and states that TGNC people have additional demands on their executive functioning capacities as a result of their gender identity. It also suggests that factors such as having higher levels of autistic traits and internalizing symptoms may overwhelm already taxed executive functioning capacities in TGNC people, because of the high degree of overlap between identifying as TGNC and being autistic or having a co-occurring internalizing mental health condition, such as depression or anxiety. Strang et al.'s (2022) stated in their model that affirming environments that help meet TGNC people’s needs, and the use of gender-affirming care, may also impact executive functioning by reducing the demand for executive functioning capacities.

While Strang et al.'s (2022) model posits that internalizing symptoms may overwhelm EF capacities in TGNC youth, it does not delineate the potential differential unique contribution of anxiety versus depression. Building on the minority stress framework extended to autism (Botha and Frost, 2020), we hypothesize that anxiety may constitute a more direct and acute cognitive load on EF resources than depression in this population. Anxiety is characterized by hyper-vigilance and future-oriented threat appraisal-processes that continuously consume attentional and working memory resources critical for EF. In contrast, depressive symptoms, often marked by psychomotor retardation and anhedonia, may impact motivation and initiation more than the core planning/organizing aspects of EF. For TGNC autistic adults, navigating societal stigma, gender identity disclosure decisions, and complex healthcare systems are chronic, anxiety-provoking demands that likely place a sustained tax on EF systems. Therefore, we predict that in our sample of TGNC adults, self-reported anxiety will be the strongest independent predictor of EF difficulties, surpassing the contributions of depressive symptoms and autistic traits.

Prior work suggests that TGNC and autistic adults experience elevated internalizing symptoms and executive functioning challenges. Building on minority stress and attentional control theory, we tested whether anxiety would show the strongest independent association with executive functioning difficulties relative to depressive symptoms and autistic traits.

In TGNC adults, what is the relative contribution of anxiety symptoms, depressive symptoms, and autistic traits to self-reported executive functioning difficulties? Hypotheses. Based on minority stress and attentional control theory, we hypothesized that anxiety would account for the largest unique variance in executive functioning difficulties when modeled alongside depression and autistic traits. We further expected autistic traits to exacerbate the association between internalizing symptoms and executive functioning.

Because TGNC research often samples from gender dysphoria clinics, it may underrepresent individuals who do not experience sufficient distress to seek specialty care or who face barriers to access. To enhance representativeness and capture a broader range of experiences, we employed community-based recruitment strategies, including social media platforms, public forums, community flyers, word-of-mouth referrals, and outreach through queer resource organizations in Utah. This approach yielded a sample of 54 TGNC adults (aged 18 + years, English-speaking) who self-identified as TGNC. We retained self-diagnosed autistic participants given evidence for the validity for this approach comparable to clinical diagnosis in community samples and the number of significant barriers to receiving formal, clinical diagnosis as an adult (Ardeleanu et al., 2025; McDonald, 2020; Overton et al., 2024). This is especially important for transgender and gender diverse adults who experience additional barriers to diagnosis and acceptance, where exclusion risks systematic underrepresentation of people already marginalized by existing diagnostic systems (Ardeleanu et al., 2025; Adams et al., 2025). Consistent with minority stress theory and prior literature, we anticipated elevated anxiety among participants (Botha and Frost, 2020; Hendricks and Testa, 2012; Meyer, 2003).

Data analyzed for this paper were collected as part of a larger study [redacted]. We collected data online via a self-report Qualtrics (Provo, UT) survey including the measures listed below. We asked interested participants to reach out to the research team via email, where they were provided with instructions for participation and a link to the survey. Participants were then asked to email the team after completing the survey to receive the link to the secondary survey for entry into a drawing for a $50 Amazon gift card (chances of winning were 1 in 25). All participants completed the same measures, regardless of their recruitment origin or demographic factors.

The lead author of this paper is a queer autistic researcher and clinician who completed this project as part of a dissertation in pursuit of her doctoral degree. Additionally, prior to data collection, the research team sought feedback online from people who identify as both autistic and TGNC about the content and format of the survey used to collect data for this project. We integrated this feedback into the final version of the survey that was used for data collection.

Due to the small sample size, we retained all participants. However, missing data was small (three participants were missing “income,” one was missing “sex assigned at birth” [preferred not to specify], and one was missing “autism diagnosis”). One value was more than 1.5 interquartile ranges above the third quartile. We used SPSS (IBM Corp, 2022) for all statistical analyses.

Within our sample, 46.3% of participants had a T-score on the BRIEF-A that was above the cut score indicating clinical concern (i.e., 65). For the BDI-II, 53.7% of participants had a total score of 20 or higher, indicating moderate levels of depressive symptoms. For the BAI, 55.6% of participants had a total score of 16 or higher, indicating moderate levels of anxiety symptoms. On the AQ, 53.7% of participants’ total scores were above the cut score of 32 which indicates significantly elevated autistic traits (See Table 3 for more details).

We hypothesized that within our sample, executive functioning as measured by the BRIEF-A would be impacted by mental health functioning (i.e., anxiety and depression) as measured by the BDI-II and the BAI, and autistic traits as measured by the AQ. We entered these variables into a regression simultaneously, with executive functioning as the outcome variable. We did not enter autism diagnosis (formal or self-diagnosis) into the regression.

We report unstandardized coefficients (B) with standard errors and standardized coefficients (β). The overall model was significant R² = 0.48, (p < 0.001). Anxiety was the only predictor that contributed significantly to the model based on our alpha of 0.05 (see Table 4).

Depression (as expected according to Derakshan and Eysenck, 2009; Eysenck et al., 2007) did not reach significance in the model. However, depression and AQ showed imprecise estimates with p-values from 0.073–0.077. Larger samples are needed to estimate these effects more precisely.

In the full sample, anxiety showed a robust unique association with EF (β = 0.45, p = 0.001). Depression and autistic traits were not statistically significant at α = 0.05; however, estimates were imprecise and the evidence was inconclusive (ps = 0.073–0.077). When analyses were restricted to participants with a formal ASD diagnosis (n = 15), the model and predictors are not statistically significant. Estimates differed in magnitude in the diagnosed subgroup, but confidence intervals were wide and conclusions about subgroup differences are not warranted (Table 5).

This study is the first of its kind to examine these questions in an adult, TGNC, autistic sample recruited from the general population. We examined how anxiety, depression, and autistic traits relate to executive functioning (EF) in an adult TGNC sample with substantial autistic identification recruited from the general population. The primary finding was that anxiety showed the clearest unique association with EF, whereas depression and autistic traits were not statistically significant predictors in the main model, though both may have been difficult to detect with the small sample size.

When analyses were limited to participants with a formal ASD diagnosis (n = 15), effects were generally in the same direction but were not statistically detectable and had wide uncertainty, consistent with limited power; thus, the subgroup analysis does not provide strong evidence that associations differ by diagnostic status.

The lack of a statistically detectable association between autistic traits and EF is framed as plausible given mixed findings in the autism–EF literature: some studies report EF difficulties in autistic adults (St John et al., 2022), whereas others show relative strengths in certain EF domains (Geurts et al., 2009). The discussion highlights evidence that higher cognitive load can worsen EF in autistic people (Bodner et al., 2019; Geurts and Vissers, 2012), and it links this to the study’s conceptual rationale for expecting autistic traits to matter for EF, drawing on models in which anxiety disrupts cognitive performance (Derakshan and Eysenck, 2009; Eysenck et al., 2007). Specifically, the authors propose that the sustained cognitive demands of navigating TGNC-related stressors (e.g., monitoring safety, anticipating social reactions, managing complex gender-affirmation systems) may elevate anxiety, which then more directly interferes with EF (e.g., working memory, cognitive flexibility). Autistic traits (e.g., sensory sensitivities, social-communication differences) may intensify this cognitive burden, even if they do not emerge as an independent predictor of EF in this dataset.

Clinically, the findings suggest that when TGNC autistic people present with EF-related difficulties, interventions may be most effective if they prioritize anxiety reduction rather than presuming autism-linked EF impairment. This is particularly relevant given the substantial real-world planning and organization demands involved in gender affirmation tasks (Strang et al., 2022). If anxiety and vigilance are proximal pathways from minority stress to EF problems, then approaches that reduce anxiety (e.g., CBT targeting catastrophic thinking and hypervigilance) and supports that reduce stressor exposure (e.g., gender-affirming environments and care-navigation assistance) may indirectly improve everyday EF and access to gender-affirming care (Strang et al., 2022).

A primary limitation is sample size, which constrained statistical power to detect smaller-to-moderate effects and limited confidence in subgroup inferences. Relatedly, diagnostic certainty was limited because autism diagnosis status was self-reported and not clinician-verified; accordingly, conclusions about “autistic” TGNC adults should be interpreted as applying to a broad range of participants (e.g., experiencing diagnostic exploration, self-identifying as autistic, or self-reporting formal diagnosis) rather than those with a confirmed ASD. The study also relied on a dimensional trait measure (AQ) that may be inflated or confounded by current internalizing symptoms, as far as anxiety and depression can shape social functioning and self-perception; thus, AQ scores may partially reflect symptom burden rather than autistic traits alone (Ashwood et al., 2016). Finally, the exclusive use of self-report measures (including EF) and the inability to adjust for key covariates—especially ADHD, given its high comorbidity and strong links to EF—may have introduced measurement and omitted-variable bias that could attenuate or distort observed associations.

Contextual factors also limit generalizability. Because TGNC experiences vary substantially by sociopolitical context, the broad geographic recruitment and heterogeneous participant environments likely introduced meaningful, unmeasured variability tied to local attitudes and structural protections (Bränström and Pachankis, 2021; Hammond and Moretti, 2025; Worthen et al., 2017; Falck and Bränström, 2023; Williamson, 2023). In addition, data collection occurred amid a period of heightened anti-transgender legislation and rhetoric in the United States, which may have elevated distress and anxiety in ways that are time-specific and could influence both mental health and EF-relevant self-reporting (Houghtaling et al., 2025; Santhanam, 2023).