1 in 100, Part 2

by Benison O'Reilly on Saturday, October 24th, 2009

A couple of weeks ago I blogged about a recent British study which found an estimated 1 in 100 adults were on the autism spectrum. Today I will follow up with a report on a 2007 American survey of children that reached similar conclusions.  In fact, the US survey, which questioned a random sample of over 78,000 parents of children aged 3-17 years, actually found a prevalence of 110 per 10,000 or 1 in 91 in this age group. However, I’m sticking to my title, and after reading the survey methods and results I’d probably caution against hanging a huge amount of weight on that 1 in 91 figure.

The study, published in a recent issue of the medical journal Pediatrics, was part of 2007 National Survey of Children’s Health. ¹ Parents or guardians were asked if they had ever been told by a doctor or other health care provider that their child had ‘autism, Asperger disorder, pervasive developmental disorder, or other autism spectrum disorder’. If parents answered yes, they were asked if their child currently had autism or ASD and, if so, to provide a qualitative ranking of severity. Children were only classified as having ASD if their parents answered yes to both the first two questions (i.e. the parents claimed their child currently had ASD).

What seems surprising about this survey is that an unexpectedly large number of parents (representing 38.2% of those children who met the ever-reported criterion) answered no to that second question; that is, they claimed their child had previously been diagnosed with an ASD, but did not currently have the condition.  Thirty-eight percent sounds very high. It would be wonderful if we could declare this a success story for early intervention, but the authors propose a few other reasons:  possible misdiagnosis  in very young children (subsequently revised);  that ASD was once suspected at developmental screening but later ruled out and  thus the child was never really ‘diagnosed’;  diagnostic substitution  (where some children with another developmental condition might have at one stage been claimed to have an ASD to access ASD-specific funding/services);  and finally that some parents may have answered ‘no’ to the second question because their child no longer received special education or autism-specific services for the condition.

Researchers also found that the odds for boys having an ASD were 4 times as large as the odds for girls (consistent with previous research) and that non-Hispanic black and non-Hispanic multiracial children had 57% and 42% lower odds, respectively, of having an ASD than non-Hispanic white children (possibly due to poorer access to ASD diagnostic and intervention services).

Because ASD status in this survey based only on parental report and not confirmed by a health care worker, I’d suggest we treat that 1 in 91 figure with a bit of caution (although a quick search on Google suggests few others have thought to). However, whatever the case, the prevalence of ASD in this survey was found to be higher than previously reported and it appears that 1 in 100 (or 1%) figure is looking more and more likely (there is apparently a Center for Disease Control study yet to come).  The authors of the Pediatrics paper suggest that more inclusive survey questions, increased public awareness, and improved screening and identification by health care workers may go some way to explain this finding.

1. Kogan MD, et al. Prevalence of parent-reported diagnosis of autism spectrum disorder among children in the US, 2007. Pediatrics 2009; 124: 0000 [accessed October 6, 2009]

Tags: autism, prevalence, research