Mental Health of Youth With Autism Spectrum Disorder and Gender Dysphoria

This secondary analysis used data from a large administrative dataset formed by the PEDSnet learning health system network of 8 pediatric hospital institutions.²⁴  We included patients in the analysis if they (1) were aged 9 to 18 years and (2) had ≥2 inpatient or outpatient encounters at a PEDSnet member institution between 2009 and the data extraction date (March 2022), with at least 1 encounter in the previous 18 months. Main variables of interest included the presence of ASD, GD, anxiety, depression, eating disorder, suicidality (including ideation and attempt), and self-harm diagnoses as recorded in the patient’s electronic medical record (EMR; see Supplemental Table 4 for included diagnosis codes). Additional variables included sex, age, ethnicity, race, and health insurance type category (private, public, other) as documented in the patient’s EMR, and the PEDSnet institution where the patient was seen. EMR-reported sex represented the sex assigned at birth for most patients and was used to be consistent with previous PEDSnet research.⁷  The “other” health insurance category included any insurance type that was not listed as private or public (eg, self-pay, charity).

Analyses were conducted in Stata 17.0²⁵  and included descriptive statistics and a set of adjusted mixed logistic regression models, modeling each individual mental health diagnosis as the outcome and combinations of ASD and GD diagnoses as the predictor (ASD alone, GD alone, cooccurring ASD/GD, neither). All models adjusted for age, EMR sex, ethnicity and race, health insurance type, and for clustering by PEDSnet institution. Ethnicity and race were included in our models given our recently published work with this cohort⁸  and other published studies that suggest the existence of ethnic and racial disparities in access to evaluation and treatment of ASD,^26,27  gender dysphoria,^28–30  and other mental health concerns, which are attributable to cultural norms, bias, and structural racism in health care settings.³¹  We also calculated average predicted probabilities of each mental health diagnosis by ASD and GD diagnosis category. Finally, we conducted pairwise comparisons to test for statistically significant differences in the odds of each mental health diagnosis between each of the different groups (eg, comparing ASD only with GD only, GD only with ASD + GD). Results were considered statistically significant if the P value was less than .05 after the Bonferroni correction, which accounts for the family-wise error rate when making multiple comparisons.

All study procedures were reviewed and approved by the Seattle Children’s institutional review board.

Among the 919 898 patients meeting inclusion criteria, the mean age was 13.6 years (SD = 2.6). Just over half (50.8%) were reported as male in the EMR, 15.9% identified as Hispanic/Latino/a/x/e ethnicity, 55.2% identified as white, 40.3% used private insurance, and 32.5% used public insurance (Table 1).

With respect to ASD and GD diagnoses, 4.4% (n = 40 249) had an ASD diagnosis alone, 0.5% (n = 4925) had a GD diagnosis alone, 0.05% (n = 464) had both ASD and GD diagnoses, and 95.0% had neither diagnosis (n = 874 230; Table 2). In the overall sample, 15.8% had an anxiety diagnosis, 4.6% had a depression diagnosis, 1.3% had an eating disorder diagnosis, 3.0% had a suicidality diagnosis, and 1.1% had a self-harm diagnosis; however, the prevalence of each mental health diagnosis was higher among those with ASD and/or GD, and greatest among those with cooccurring ASD and GD (Table 2).

The average predicted probabilities of each mental health diagnosis by ASD and GD category from the adjusted mixed logistic regression are illustrated in Fig 1. Broadly, the predicted probability of each mental health diagnosis was highest among those with cooccurring ASD and GD diagnoses and lowest among those with neither diagnosis.

Results of the pairwise comparisons between ASD and GD diagnosis groups using the Bonferroni correction for multiple tests are presented in Table 3. In general, those with ASD alone, GD alone, and cooccurring ASD and GD showed significantly greater odds of each mental health diagnosis compared with those with neither ASD nor GD.

Youth with only an ASD diagnosis had 4.46 times greater odds (95% CI, 4.33–4.59), youth with only a GD diagnosis had 6.07 times greater odds (95% CI, 5.61–6.57), and youth with cooccurring ASD and GD diagnoses had 19.31 times greater odds (95% CI, 14.11–26.42) of also having an anxiety diagnosis compared with those with neither ASD nor GD diagnoses. Youth with only a GD diagnosis also had significantly greater odds of an anxiety diagnosis compared with those with only an ASD diagnosis (adjusted OR [aOR], 1.36; 95% CI, 1.25–1.48), and youth with cooccurring ASD and GD diagnoses had significantly greater odds of an anxiety diagnosis compared with both youth with only an ASD diagnosis (aOR, 4.33; 95% CI, 3.16–5.93) and only a GD diagnosis (aOR, 3.18; 95% CI, 2.30–4.39).

Youth with only an ASD diagnosis had 1.89 times greater odds (95% CI, 1.78–2.01), youth with only a GD diagnosis had 11.47 times greater odds (95% CI, 10.53–12.50), and youth with cooccurring ASD and GD diagnoses had 18.81 times greater odds (95% CI, 14.34–24.68) of also having a depression diagnosis compared with those with neither ASD nor GD diagnoses. Similar to the results for anxiety, youth with only a GD diagnosis also had significantly greater odds of a depression diagnosis compared with those with only an ASD diagnosis (aOR, 6.07; 95% CI, 5.47–6.73), and youth with cooccurring ASD and GD diagnoses had significantly greater odds of a depression diagnosis compared with both youth with only an ASD diagnosis (aOR, 9.95; 95% CI, 7.54–13.13) and only a GD diagnosis (aOR, 1.64; 95% CI, 1.23–2.18).

Youth with only an ASD diagnosis had 5.22 times greater odds (95% CI, 4.84–5.62), youth with only a GD diagnosis had 4.70 times greater odds (95% CI, 4.08–5.41), and youth with cooccurring ASD and GD diagnoses had 5.97 times greater odds (95% CI, 3.94–9.05) of also having an eating disorder diagnosis compared with those with neither ASD nor GD diagnoses. No other differences emerged between youth with only ASD, only GD, and cooccurring ASD and GD diagnoses.

Youth with only an ASD diagnosis had 2.15 times greater odds (95% CI, 2.01–2.31), youth with only a GD diagnosis had 8.48 times greater odds (95% CI, 7.71–9.32), and youth with cooccurring ASD and GD diagnoses had 11.22 times greater odds (95% CI, 8.46–14.89) of also having a suicidality diagnosis compared with those with neither ASD nor GD diagnoses. Youth with only a GD diagnosis (aOR, 3.94; 95% CI, 3.51–4.41) and with cooccurring ASD and GD diagnoses (aOR, 5.21; 95% CI, 3.90–6.96) each had significantly greater odds of a suicidality diagnoses compared with youth with only an ASD diagnosis. There was no significant difference in the odds of a suicidality diagnosis when comparing youth with only a GD diagnosis and those with cooccurring ASD and GD diagnoses.

Youth with only an ASD diagnosis had 4.32 times greater odds (95% CI, 3.95–4.74), youth with only a GD diagnosis had 9.64 times greater odds (95% CI, 8.56–10.86), and youth with cooccurring ASD and GD diagnoses had 11.27 times greater odds (95% CI, 7.95–15.98) of also having a self-harm diagnosis compared with those with neither ASD nor GD diagnoses. Similar to the results for suicidality, youth with only a GD diagnosis (aOR, 2.23; 95% CI, 1.93–2.58) and with cooccurring ASD and GD diagnoses (aOR, 2.61; 95% CI, 1.82–3.73) each had significantly greater odds of a self-harm diagnoses compared with youth with only an ASD diagnosis, and no statistically significant difference emerged when comparing the odds of self-harm diagnosis among youth with only a GD diagnosis and those with cooccurring ASD and GD diagnoses.

The results of this study build on our previous research focused on demographic differences in cooccurring ASD and GD, showing that youth with cooccurring ASD and GD diagnoses are more likely to be diagnosed with anxiety and depression compared with their peers with 1 or neither of these diagnoses. Importantly, such research helps to better understand the cumulative impact of these diagnoses and can be used to inform collaborations between health care providers who specialize in ASD and gender-affirming care to create developmentally appropriate, gender-affirming services and support for youth with cooccurring ASD and GD.^32,33 

In line with previous research,²²  we found that although youth with either ASD or GD alone were significantly more likely than youth with neither diagnosis to have anxiety or depression diagnoses, youth with cooccurring ASD and GD were significantly more likely to have these diagnoses compared with youth with only 1 of these diagnoses. Research suggests that the association between ASD and mental health concerns may be attributable to difficulties with sensory processing, executive functioning, and communication, all of which can contribute to social isolation and emotional dysregulation.³³  Similarly, gender-diverse youth face markedly elevated rates of marginalization, victimization, and social isolation related to their gender identity, which play a major role in the development of mental health concerns such as anxiety and depression.³⁴  Given these findings, providers should remain aware of these heightened risks and be prepared to implement mental health screening, interventions, and supports that meet the needs of this population.

With regards to eating disorders, we found that youth with 1 or both ASD and GD diagnoses were significantly more likely than youth with neither diagnosis to have an eating disorder diagnosis. These findings support current research showing the increased prevalence of disordered eating behaviors among youth with either ASD or GD.^7,14,21  However, in our cohort, youth with cooccurring ASD and GD were not more likely to have an eating disorder compared with youth with ASD or GD alone. This contradicts the limited available research showing that transgender youth and young adults (ages 14–25 years) with ASD are more likely to report having previously received an eating disorder diagnosis compared with transgender youth without ASD.²²  One reason for this discrepancy could be that our study focused on younger youth (ages 9–18 years), who may go on to receive an eating disorder diagnosis before age 25 years. As a result, further research is needed to understand when youth with cooccurring ASD and GD may be at greatest risk for eating disorders to identify appropriate preventive interventions.

Our results for suicidality and self-harm diagnoses followed similar patterns. Youth with cooccurring ASD and GD, as well as those with only one of these diagnoses, were also more likely to have suicidality or self-harm diagnoses compared with those with neither ASD nor GD. This, too, is in line with previous research on suicidality and self-harm among populations with either ASD or GD.^7,13,16,17  Furthermore, although youth with a GD diagnosis or with cooccurring ASD and GD were more likely to have these diagnoses compared with youth with only an ASD diagnosis, we did not observe such differences between youth with GD only and youth with ASD and GD, suggesting that gender dysphoria may be driving these associations with suicidality and self-harm. These findings suggest the critical need for gender-affirming care services that include suicide screening tools and other preventive interventions for youth experiencing GD.

Our study has several limitations. First, although our use of the PEDSnet database allowed us to understand mental health concerns among youth with 2 relatively uncommon diagnoses, we were still limited by small numbers in some subgroups, particularly when considering less prevalent mental health diagnoses. Similarly, we were unable to test for variation by demographic characteristics, particularly ethnicity and race and insurance status, which we have found in our previous work to be associated with a lower likelihood of cooccurring ASD and GD diagnoses.⁸ 

We must also acknowledge potential quality and accuracy concerns when using EMR data. For example, significant barriers to receiving timely and accurate diagnoses of ASD, GD, and other mental health concerns exist. Thus, it is possible that patients were misclassified in our analysis because of undiagnosed ASD, GD, and/or mental health conditions, leading to an underestimation of their true prevalence. In addition, because PEDSnet diagnosis data are extracted from billing codes and problem lists, we did not have access to medical histories or individual clinician notes that may have documented these concerns. Furthermore, a diagnosis may not be recorded at a PEDSnet institution because of receiving care in another location, lack of provider diagnosis code placement, or the fear of stigma or discrimination should these diagnoses be documented in their records.

Furthermore, our study focused on the presence of these diagnoses in the EMR and thus should not be interpreted as ongoing access to ASD services, gender-affirming care, or mental health care. Importantly, many youth currently face significant delays in accessing timely and effective mental health care, and this can be especially true for youth with GD and ASD.^22,35  The combination of social and communication differences and stigma experienced by youth with ASD and GD may also influence their ability to seek and engage in mental health services.³⁶  Because of these differences, providers and families may have difficulty recognizing the severity or significance of mental health symptoms.³⁷  Youth with GD may also be hesitant to disclose their identity to providers because of previous negative experiences in health care settings.^38–40  Additionally, mental health evaluations have been, and continue to be, used to determine eligibility for access to gender-affirming medical care, which may make youth less likely to disclose their mental health concerns.^34,41  It is therefore important to continue developing research to understand the intersection of ASD, GD, and mental health concerns, as well as variations in access to care, to design and provide services that meet the mental health needs of youth with cooccurring ASD and GD.

Youth with cooccurring ASD and GD diagnoses are more likely to be diagnosed with mental health concerns compared with their peers, particularly anxiety and depression. In addition to guiding future research, the results of this project highlight the clear need for collaborations between ASD and gender-affirming care providers to provide more robust mental health services that are both developmentally appropriate and gender-affirming for youth with cooccurring ASD and GD.

The research reported in this publication was conducted using PEDSnet, A National Pediatric Learning Health System, and includes data from the following PEDSnet institutions: Children’s Hospital Colorado, Children’s Hospital of Philadelphia, Cincinnati Children’s Hospital Medical Center, Nationwide Children’s Hospital, Nemours Children’s Health, Seattle Children’s Hospital, Stanford Children’s Health, and Ann and Robert H. Lurie Children’s Hospital of Chicago. Thank you to the PEDSnet Data Coordinating Center for providing us with these data. Thank you to Daksha Ranade and Victoria Soucek for their assistance in accessing and preparing the data for analysis.:

aOR

adjusted OR

CI

confidence interval

ASD

autism spectrum disorder

EMR

electronic medical record

GD

gender dysphoria