Social inference brain networks in autistic adults during movie-viewing: functional specialization and heterogeneity.

The autistic group was small (34 participants), and recruitment ended early because of the COVID-19 pandemic, limiting power to detect group differences and subgroup effects. Many group differences were small, and some were significant only in the exploratory closely matched sample, so the neural findings were weak and not consistently found. The study used passive movie viewing without concurrent behavioural measurement, so it cannot show what participants were actually inferring about the characters during the scan. The social behaviour measures were administered outside the scanner and included self-report questionnaires, which may not capture subtle social cognition or brain-behavior relationships. Participants were excluded if full-scale IQ was below 80, and the study required enough scanner tolerance to avoid sedation, so it does not represent autistic people with more severe cognitive or motor difficulties. The movie was aimed at a child audience and had relatively simple social inferences, which may not reveal subtle mental-state difficulties in adults. The study compared two broad groups and did not identify subgroups, so it cannot explain which patterns of neural variability correspond to which patients.

The study was funded by the National Institute of General Medical Sciences (R01MH110630). The supplied text does not state author conflicts of interest.

Do not read the brain differences as proof that autistic adults lack empathy or have impaired pain processing, and do not use this study to choose or change therapy. The pain network showed similar connectivity between groups, and mental-state network effects were weak and inconsistent across samples.