Table 2.

Potential Future Research Directions

Research Topics	Methodological Considerations
Level 1: Nosological and Diagnostic Challenges
Nosological reflection on sex/gender-differential criteria: Qualitative	•Qualitative research on female presentations •Developing new instruments that reflect narrow constructs and that collect a sufficiently wide range of behavioral exemplars (beyond classical autistic symptoms but also associated and co-occurring features) •Applying multi-group confirmatory factor analysis (CFA) or item response theory (IRT) models to test for sex/gender differences at the narrow construct level; if existing, testing whether this is due to the lack of female-specific or sex/gender-independent behavioral exemplars that measure these narrow constructs •Delineating core vs. non-core/associated behavioral exemplars, or narrow constructs, for males and females respectively, by examining endorsement rates, or building measurement models with broad and narrow constructs and then examining the loading of each narrow construct onto the broad constructs
Nosological reflection on sex/gender-differential criteria: Quantitative	•Developing objective measures of autistic traits, free from rater bias, to assist the decision about sex/gender-norming •Adopting both sex/gender-independent and sex/gender-dependent statistical thresholds for research related to autistic traits
Nosological reflection on sex/gender-differential criteria: Developmental	•Investigating lifespan development in males and females, especially in relation to sex-linked biological effects and gendered socio-cultural influences
Factors associated with under- and/or misidentification of females with autism	•General population epidemiological studies on autism prevalence/incidence using tools better capturing subtle (higher-functioning) presentations •Epidemiological studies on co-occurrence and shifts of diagnoses over time to elucidate how co-occurring conditions contribute to diagnostic overshadowing or substitution •Exploring how co-occurring conditions or cognitive/temperamental features influence the presentation and identification of autism •Qualitative work to identify mechanisms and consequences of “camouflage” (i.e., masking and/or compensation) •Developing quantitative measures for camouflage •Developing instruments sensitive to females with autism to assist identification

Level 2: Sex/Gender-Independent and Sex/Gender-Dependent Characteristics
Similarities and differences between males and females with autism	•Using well-powered, comparable male–female group-size study designs •Adopting (at least) 4-group designs to compare how males and females with autism differ respectively from neurotypical (or other groups of) males and females •Longitudinal design to capture differences in developmental trajectories •Multi-level and multi-domain investigation to identify convergence and divergence
Understanding how the findings are influenced by intellectual level and co-occurring conditions	•Comparing above findings in intelligence and co-occurring condition-stratified samples, and including these factors as predictors

Level 3: General Models of Etiology: Liability and Threshold
Clarifying sex/gender-differential etiological load, threshold, and genetic heterogeneity	•Replicating findings of sex/gender-differential familial (sibling) autism recurrence rate and autistic traits, especially in general-population (rather than clinical) samples •Identifying endophenotypic, genetic, and epigenetic factors associated with the increased diagnoses/traits in families of female compared to male probands (if confirmed) •Examining the above issues in large infant-sibling samples to reveal associated characteristics early in life •Using data from multiplex families and families with broader autism phenotype to compare male-only, sex/gender-mixed (female-containing), and female-only families •Examining how co-occurring medical, neurodevelopmental, psychiatric conditions, and cognitive-behavioral features are related to sex/gender-differential etiological structures
Testing sex/gender-differential shifts of liability distribution	•Investigating what normative sex/gender differences in the general population constitute risk and protection for developing autism (e.g., by identifying sex/gender-differential correlations between autistic traits and biological/neurocognitive characteristics) •Testing whether females have more fractionable underlying cognitive and biological structures that provide better protection

Level 4: Specific Etiological-Developmental Mechanisms
Whether normative sex differences in genetic, epigenetic, and pre-/perinatal environmental factors contribute to autism etiologies	•Testing and identifying overlap between normative sex differences and known autism characteristics, including downstream effects from both •Investigating multi-level characteristics related to autism in individuals with clinical conditions of altered sexual differentiation •Epidemiological studies to test whether factors associated with sexual differentiation contribute to risk/protection for autism separately for males and females
Whether gendered socio-cultural factors contribute to the emergence, lifespan development, and identification of autism	•Testing whether early gendered environment (e.g., interaction style) affects the onset of autistic features in high-risk infants •Investigating whether gendered environment modulates lifespan development of individuals with autism •Investigating how gendered socio-cultural contexts (e.g., gender stereotypes) influence autism diagnosis in the real world and cross-culturally
Moderating effects of sex/gender in etiologies	•Testing for moderating effects of sex/gender and comparing findings stratified by sex/gender •Identifying sex/gender-specific etiological factors •Investigating whether gene–environment interplay produces multiple “hits” in the emergence of autism, and how sex/gender moderates this