Trends in documented co-occurring conditions in children with autism spectrum disorder, 2002-2010.

The data were cross-sectional records, so the study could not follow individual children to see when each diagnosis was made. Co-occurring conditions may have been recorded better in later years, so the increase could partly reflect documentation rather than true change. Only conditions written in health or education records were counted, so undocumented conditions were missed. Conditions noted in only one child in a surveillance year were excluded, which may slightly underestimate the average number of conditions. The authors assumed documented conditions truly co-occurred with ASD, but some may have been used to explain ASD traits before diagnosis. The study covered only eight sites and is not nationally representative. It did not assess whether co-occurring conditions appeared before or after ASD diagnosis.

Do not read the rise in documented co-occurring conditions as proof that autism is becoming more severe or that therapy is needed. The study used records and tested no intervention; the authors say the trend may reflect more conditions being identified, documented, or diagnosed.