Monday, 3 December 2012

Autism prevalence increasing in New Jersey

A quote to begin with from the paper by Walter Zahorodny and colleagues* with special thanks to Natasa for the full-text of the paper on the prevalence of autism spectrum disorders (ASDs) in New Jersey: "Regardless of whether one acknowledges an increased ASD prevalence, all the epidemiological and administrative studies confirm that ASD is now among the most common, severe, developmental disorders".
Edison in his NJ workshop @ Wikipedia  

For me this quote encapsulates everything about the various autism prevalence (and incidence) debates which have been, and continue to be, rife particularly in cyberspace.

It kinda says, yep we don't know all the details behind the quite enormous rise in cases of autism - which lets face it, basically boils down to whether or not you believe environment in whatever form contributes to autism aetiology - but whatever the reason(s), autism is not an uncommon condition and has a huge range of implications stretching from the diagnosed individual, their families, their schools, their doctors through to society as a whole.

Indeed, the paper from Zahorodny has lots of other soundbites worthy of quoting, but just before I head out that way, a few snippets of information about the paper are in order:

  • This was a study all about comparing the estimated prevalence of ASD in the New Jersey Metropolitan Area (NJMA) in 2002 with some newer data from 2006.
  • The focus was on using "identical methods and procedures" between the two time points drawing on the ADDM method which included active-screening and subsequent independent case determination. 
  • Children born in 1998 and residing in the study area in 2006, classified as being educated under any special education classification (equivalent to receiving a SEN) were the starting point, filtered down to those with a documented or suspected ASD diagnosis, whose data were subsequently subjected to clinician review and analysis. "Agreement between previous ASD diagnosis and ASD case status by the surveillance method was 99%".
  • Results: In 2006, out of a population of 30,570 8-year olds, 3,332 child records were reviewed, 923 of which met criteria for further review, which identified 533 with study-determined ASD, equal to a prevalence of 17.4 per 1000 (1.7%). This compared with 2002 data (8-year olds born in 1994) suggestive of a prevalence rate of ASD of 10.6 per 1000 (1.1%). The data is a little confusing because two different estimates are given based on the use of curtailed and non-curtailed populations; one used a sort of comparator to assess any changes in population affecting the estimates. 
  • Allowing for the slight differences, the authors determined that 1 in 57 8-year olds in NJMA in 2006 had an ASD and a "startling" 1 in 35 boys.
  • Between the sexes, ASD prevalence in boys grew from 17 to 28.7 per 1000 between 2002 and 2006 and 4.1 to 5.9 per 1000 in girls. 
  • Regression, that most interesting of phenomenon with autism in mind, was relatively unchanged; the rate hovering around the 20% of cases mark.
  • Quite a few factors potentially influencing the numbers were also taken into account including migration. That on top of the fact that DSM-IV TR was the main assessment criteria across the time periods.

This is an interesting study for lots of reasons. Not least because using identical protocols and identical criteria, a change in the prevalence of ASD was noted over 4 years. One therefore, to quite a strong degree, also rules out changes in DSM diagnostic revisions as a cause of the changes and to some extent those arguments about widening criteria as was evidenced between DSM-III and DSM-IV. That being said, the issue of increased awareness and its impact cannot be ascertained from this data; indeed also diagnostic substitution to some extent also remains a possibility (i.e. from learning disability to autism) as does under-diagnosis of ASD in non-white groups. This last point however needs to be tempered by the authors demonstrating a 90% increase in cases of ASD among Black, non-Hispanic children but also indicating "a broadly based escalation of ASD, rather than an increase reflecting improved case-finding in specific populations".

I get the impression that the authors were hinting that the change they reported might be more administrative over real. So they say: "It is too soon to know at what point ASD prevalence will plateau", which implies to me that they are looking to the autism rate increase to somehow eventually 'bottom out' when the message finally gets through about autism awareness and what it is. I hope I don't seem too rude when I say that we've all been waiting for that plateau for quite a few years now so as to save those 'it must be better case ascertainment' arguments, but so far there is no end it sight at least according to the latest CDC estimates. I'm sure even the staunchest supporters of the better awareness et al camp would quite logically admit that at least part of the increase might not necessarily reflect just 'better awareness' and 'more diagnostic substitution' but rather some other factor or factors also being involved? Or maybe not....


* Zahorodny W. et al. Increasing autism prevalence in metropolitan New Jersey. Autism. November 2012.

---------- Zahorodny, W., Shenouda, J., Howell, S., Scotto Rosato, N., Peng, B., & Mehta, U. (2012). Increasing autism prevalence in metropolitan New Jersey Autism DOI: 10.1177/1362361312463977