Links Between Autism, Childhood Gender Nonconformity, and Adult Sexual Orientation

Background: Autism, gender diversity, and sexual orientation diversity can each be indexed categorically (autistic vs. non-autistic, cisgender/gender incongruent, heterosexual/nonheterosexual) and dimensionally (continuous trait scores). Research has linked autism with both gender and sexual orientation diversity, but the basis of these associations remains unclear. Because childhood gender nonconformity is a robust dimensional correlate of adult sexual orientation in the general population, we examined whether recalled childhood gender nonconformity accounts for associations between autism and adult sexual orientation, and whether autism influences the childhood gender nonconformity × adult sexual orientation link. Methods: Across two preregistered studies and two direct replications, we examined associations among autism, recalled childhood gender nonconformity, sexual attraction, and sexual identity in cisgender adults. Study 1 examined dimensional associations with autistic traits in 400 non-autistic adults and was replicated in an independent sample of 396 non-autistic adults. Study 2 compared 200 diagnosed autistic adults with 200 non-autistic adults and was replicated in an independent sample of 244 autistic and 229 non-autistic adults. Results: Across all studies and samples, childhood gender nonconformity predicted adult sexual orientation. Study 1 showed no evidence that autism traits were associated with sexual attraction, sexual identity, or gender nonconformity. Autism traits did not moderate the association between childhood gender nonconformity and adult sexual orientation, and childhood gender nonconformity did not mediate these null associations. In contrast, Study 2 showed that autistic adults reported less heterosexual attraction, were more likely to identify as nonheterosexual, and recalled greater childhood gender nonconformity than non-autistic adults. Childhood gender nonconformity mediated associations between autism diagnostic status and sexual orientation, whereas moderation effects were largely absent. Conclusions: These findings suggest that autism-sexual orientation links emerge at the categorical level of autism diagnosis only and are accounted for statistically by childhood gender nonconformity. Community Brief Why is this an important issue? Childhood gender nonconformity (preferring activities and interests not stereotypically associated with one’s biological sex) is strongly associated with sexual orientation diversity in adulthood. Also, autism is associated with sexual orientation and gender diversity in adults, but little is known about how autism relates to childhood gender nonconformity, or whether autism influences the link between childhood gender nonconformity and adult sexual orientation. This matters because some have speculated that gender diversity has an autism-specific basis in autistic people. If so, then the usual link between childhood gender nonconformity and adult sexual orientation may not hold in autistic people. What was the purpose of this study? To assess whether recalled childhood gender nonconformity accounts statistically for the association between autism and adult sexual orientation. We tested whether this association is found for autistic traits in non-autistic adults, autism diagnostic status, or both. What did the researchers do? We conducted two studies and repeated each study with a new group of participants (“replication studies”). Study 1 examined autistic traits, recalled childhood gender nonconformity, sexual attraction, and sexual identity in 400 cisgender non-autistic adults (and another 396 participants in the replication study). Study 2 compared 200 autistic and 200 non-autistic cisgender adults (and another 244 autistic and 229 non-autistic participants in the replication study) using the same measures. What were the results? Across studies, people who recalled more childhood gender nonconformity reported less heterosexual patterns of adult sexual attraction and were more likely to identify as nonheterosexual. In Study 1, autistic traits in non-autistic adults were not significantly related to adult sexual attraction or identity, or recalled childhood gender nonconformity. In Study 2, autistic adults reported greater sexual orientation diversity and recalled higher childhood gender nonconformity than non-autistic adults. Childhood gender nonconformity partly explained the statistical association between autism diagnosis and adult sexual orientation. Findings from each study replicated closely. Importantly, autism (traits or diagnoses) did not influence the robust association between childhood gender nonconformity and adult sexual identity or attraction except in one instance. What do these findings add to what was already known? The research shows for the first time that (a) autistic traits among non-autistic people do not predict sexual orientation diversity, whereas autism diagnosis does, and (b) childhood gender nonconformity and adult sexual orientation are related in similar ways in autistic and non-autistic adults. The results can’t be used to predict any individual child’s future adult sexual orientation or categorical gender identity, and gender nonconformity and nonheterosexuality shouldn’t be conflated (they are not synonymous, even though they are related). As such, the observed association between childhood gender nonconformity and adult sexual orientation in our study should not be used to dismiss any individual child’s gender nonconformity as evidence they are “just gay,” or used as a reason to discourage or attempt to change their gender presentation. What are potential weaknesses in the study? The studies were cross-sectional, so they cannot prove developmental or causal pathways. Childhood gender nonconformity was measured retrospectively (adults reported what they remembered about childhood). How will these findings help autistic adults now or in the future? These findings may help clinicians, educators, families, and researchers avoid interpreting autistic people’s gender expression and sexuality as autism-specific simply because the person is autistic. Instead, they suggest autistic people’s gender expression and sexual orientation should be understood as in the wider population, while ensuring that communication and support are accessible to autistic people. These results might provide useful context for some autistic adults when they reflect on their own childhood gender nonconformity or adult sexuality, although experiences are individual and group-level results cannot explain any individual’s identity.