Wednesday, 5 November 2014

(Partly) explaining the increase in the prevalence of autism

"Changes in reporting practices can account for most (60%) of the increase in the observed prevalence of ASDs [autism spectrum disorders] in children born from 1980 through 1991 in Denmark."

Prince of Denmark @ Wikipedia 
That was the headline conclusion reported by Stefan Nygaard Hansen and colleagues [1] based on an analysis of births in Denmark between 1st January 1980 and 31st December 1991 (N=677,915) followed up until 31st December 2011 (or "until ASD diagnosis, death, emigration"). The suggestion being that an update to the ICD (International Classification of Diseases) criteria in 1994 - ICD-8 replaced by ICD-10 - played a significant role alongside "the inclusion of outpatient contacts" and voilà: "the apparent increase in ASDs in recent years is in large part attributable to changes in reporting practices". Some associated media coverage of this study can be found here.

This is certainly interesting data which again taps into those oh-so valuable Scandinavian health registries which are producing all-manner of interesting correlations and results (see here for another example). The question of what factor(s) might be driving the quite astounding increase in cases of autism worldwide (see here) has been a constant source of discussion and argument in certain circles. I have covered some of those debates before on this blog (see here).

Changes to the various diagnostic schedules as being a factor behind the alterations to the autism prevalence rate is by no means a new idea as per discussions about the the rival DSM (Diagnostic and Statistical Manual of Mental Disorders) criteria down the years. Indeed, looking at how the DSM-IV revision (which also came into play in 1994) might have influenced the autism numbers compared against the previous DSM-III criteria tells a similar tale (see here). The most recent DSM-5 revision also looks like it will be playing its own part on the epidemiology of autism at some point (see here) but we wait to see how that eventually pans out.

As per the press release for the Hansen study, there is still a question about the remaining percentage (~40%) of cases of autism not readily attributed to the changes in reporting practices.  A quick trawl through some of the other peer-reviewed work produced by Erik Parner, one of the co-authors on the Hansen paper, provides a few clues as to where researchers have been looking. I've covered some of these variables previously including: shifts in age at diagnosis (see here), the sibling recurrence rate (see here), maternal infection requiring hospitalisation (see here), etc. I might add that this list is not exhaustive as the words 'better autism awareness' are also likely to be banded around (see here for some recent interesting data on this).

I note also one sentence in the press release: "Environmental factors like using certain medications during pregnancy might also contribute, but more research is required to quantify the contribution of hypothesized risk factors" which I suspect ties into some work looking, for example, at prenatal valproate exposure and offspring neurodevelopmental outcomes [2]. Again, this is a topic which has been covered on this blog on more than one occasion (see here and see here) with no medical advice given or intended. I might add that valporate is not the only pharmaceutic which has come under the research spotlight (see here).

The Hansen results make a valuable contribution to the discussions about what might have been driving the increase in autism cases at least in Denmark [3]. Given their focus on participants with birth dates between 1980 and 1991 and the reporting changes included between 1994 and 1995, I'd be interested to see how future work from this group following up subsequent birth cohorts goes, assuming no significant changes in diagnostic schedules or other important adjustments to reporting procedures. I'd also rather hope to see a little more data about possible tie-ins with comorbidity for example, given the input of another Scandinavian team on the ESSENCE of autism...

To close: given that today is 'Remember, remember the 5th of November', I'm inclined to provide you yet again with a link to the rousing speech by 'V' from a favourite film of mine bearing in mind I'm not advocating such revolution. Oh, and please do be careful this Bonfire night...


[1] Hansen SM. et al. Explaining the Increase in the Prevalence of Autism Spectrum Disorders. JAMA Pediatr. 2014. November 3.

[2] Christensen J. et al. Prenatal valproate exposure and risk of autism spectrum disorders and childhood autism. JAMA. 2013 Apr 24;309(16):1696-703.

[3] Parner ET. et al. A comparison of autism prevalence trends in Denmark and Western Australia. J Autism Dev Disord. 2011 Dec;41(12):1601-8.

---------- Hansen, S., Schendel, D., & Parner, E. (2014). Explaining the Increase in the Prevalence of Autism Spectrum Disorders JAMA Pediatrics DOI: 10.1001/jamapediatrics.2014.1893