The Role of Biological Sex and Age in the Intrinsic Functional Properties of Males and Females with Autism

Background: The male preponderance in the prevalence of autism spectrum disorder (ASD) has sparked research into its sex differences. While sex differences have been reported to vary with age in cognitive-behavioral studies of typical and atypical development, very little is known about neural sex differences across neurodevelopment in ASD. Initial resting-state functional magnetic resonance imaging (R-fMRI) findings suggest that females and males with ASD present with different neurophenotypes (Alaerts et al., 2016, Ypma et al., 2016). However, no study has considered developmental sex differences in the intrinsic functional brain of those with ASD.

Objectives: We aimed to explore the role of biological sex and age in the intrinsic functional properties of the brain in males and females with ASD relative to neurotypical males and females.

Methods: We leveraged the combined large-scale Autism Brain Imaging Data Exchange (ABIDE I and II) repositories to select de-identified, low-motion R-fMRI from the publically available ABIDE repositories to include females with ASD aged 5 to 30 years, IQ- and age-matched males with ASD as well as age-matched neurotypical control males and females. We examined five voxel-wise R-fMRI metrics that have been found to be affected in ASD and to have distinct patterns by sex. R-fMRI data was preprocessed using the Configurable Pipeline for The Analysis of Connectomes. We set a general linear model at each voxel assessing the main effect of sex, age and diagnosis and their three-way interaction while accounting for head motion, acquisition sites and mean measure signal. We employed Gaussian random field theory correction with stringent statistical thresholds (Z≥3.1; p<0.05).

Results: Data selection yielded the largest-to-date neuroimaging sample of females with ASD (N=91), along with a sample of males with ASD (N=422) and controls (N males=472, N females=218). Across most R-fMRI metrics, there were significant main effects of diagnosis, sex, and age, along with their significant three-way interaction. A region in the right inferior parietal lobule (IPL) showed consistently across a range of R-fMRI metrics decreases as a function of increasing age in females with ASD compared to males with ASD. The opposite pattern was evident in neurotypicals. This region significantly mapped onto cognitive components related to sensory visceral, inhibitory control and theory of mind functions. Across all results sex differences were most pronounced in adults compared to younger ages.

Conclusions: Results of this cross-sectional design indicate that relative to neurotypical males and females, age-related sex differences in ASD occur in IPL consistently across R-fMRI metrics. The pattern in males with ASD resembles that seen in neurotypical females, whereas the pattern in females with ASD resembles that in neurotypical males, suggesting that age-related factors come into play in models of ‘gender-incoherence’ of ASD. Future work is warranted to confirm current results in longitudinal samples.