Developmental changes of autistic symptoms.

Shirley’s team looked back at medical charts of people with autism. They compared ADI-R scores from two points in time. The gap between tests ranged from childhood to adult years.

They asked: do the three core autism areas—social, communication, and repetitive behaviors—get better, worse, or stay the same? Age and IQ were checked to see if those mattered.

Every area improved. Social symptoms showed the biggest gain. The change happened no matter the person’s age or IQ. Most kept the autism label, but their scores were milder.

Edgin et al. (2017) saw the same drift toward milder scores, but in a whole country registry. Their drop happened before diagnosis, while Shirley shows within-person change after.

Sutton et al. (2022) looks like a contradiction: Italian adults lost adaptive skills during COVID. The difference is context—natural gain versus sudden loss of services.

Konke et al. (2026) adds a why: toddlers who can wait for treats keep better adaptive skills even when autism traits stay. Self-control may be one engine behind Shirley’s upward curve.

Expect autistic social traits to ease over years, even without therapy. Use this to calm families worried about lifelong severity. Still, teach waiting and self-management early—those skills may speed the natural upswing. Re-test with ADI-R before adding new services; yesterday’s “severe” score may already be outdated.