Characteristics of toddlers with early versus later diagnosis of autism spectrum disorder.

This was not a test of treatment. It cannot show that diagnosing earlier caused milder delays, or that early diagnosis detects children who would do better with therapy. The early diagnosis group was small, with 20 children, so differences involving that group are less stable. Not all children were screened before 18 months, so some later diagnosed toddlers might have been found earlier if they had been screened at 12 or 15 months. The reasons for later diagnosis were not measured reliably. Some children may have had regressive onset, later emerging symptoms, caregiver reporting differences, or delayed evaluations. Developmental scores may be biased by age. The MSEL used developmental quotient scores, which can be affected by chronological age, so older children may look more delayed. The sample came from a universal screening study with free evaluations, which may not match routine care.

Diana Robins and Deborah Fein are owners of M-CHAT, LLC, which receives royalties from companies that incorporate M-CHAT-R/F into commercial products or distribute products containing it. The authors state that the data reported did not incur royalties. Other authors declared no conflicts. The article is listed as NIH extramural funded.

Do not conclude that later diagnosis causes more severe delays, or that early diagnosis identifies milder ASD. The study compared children who were diagnosed at different ages, not children randomly assigned to diagnosis timing. Later diagnosis may reflect symptoms that appeared after screening, caregiver delays, scheduling delays, or measurement bias.