Developmental regression and autism

researchOverall, 36.9% of children had some type of regression”. That was one of the primary findings reported by Robin Goin-Kochel and colleagues* based at the Baylor College of Medicine in the United States looking at the rate of developmental regression in a large cohort of children with autism derived from the Simons Simplex Collection (SSC).

Regression, defined as a loss of previously acquired skills, continues to court research interest in relation to the diagnosis of autism spectrum disorder (ASD). Acceptance that regression can and does occur in a sizeable proportion of children with ASD has been a long and often difficult road, although such a concept is now widely considered real. Details remain scant however about why regression occurs and who it is most likely to impact on the most .

Goin-Kochel and colleagues retrospectively examined parent reports for over 2000 children included as part of the SSC, a genetic repository devised to aid autism research. Looking at categorisations of ‘full’ or ‘subthreshold’ losses of skills such as language, researchers were able to determine that over a quarter of their cohort experienced some degree of language loss or loss of other important skills based on reports utilising one of the gold standard assessment instruments for autism, the Autism Diagnostic Interview-Revised (ADI-R) administered with a supplemental interview schedule.

Focusing on skill losses before 3 years of age, researchers also reported that those with language loss scored lower than those without language loss on various measures of cognitive / adaptive functioning and autism severity. The slightly better news was that most children were able to regain skills previously lost – over 90% of those with communication loss “recovered their language skills”.

The Goin-Kochel paper has some significant strengths based on their large participant group and their reliance on a well-defined cohort of children with ASD derived from the SSC. Their use of the ADI-R and other standardised assessment instruments adds to the impact of the findings. The authors do note however that a reliance on parent or caregiver reports can be a potential source of bias and in particular, how the recollection of the timing of events can be influenced by various factors. This is not to say that parent report is not valuable to such studies but rather that more prospective research charting the appearance of regression as it happens, would provide additional, confirmatory data to the findings.


* Goin-Kochel RP. et al. Developmental regression among children with autism spectrum disorder: Onset, duration, and effects on functional outcomes. Res Autism Spectr Disord. 2014; 8: 890-898.

For more discussion about this study see: