Tuesday, 18 December 2012

Mortality and autism

I know the topic of this post is not exactly great dinner table conversation, particularly at this time of year during the season of goodwill to all.  Nevertheless I'm posting today on the paper by Deborah Bilder and colleagues* looking at mortality and causes of death in relation to autism spectrum disorder (ASD) simply because there are important lessons which can be learned and applied with relative ease as a result of some of their collected data.

Regular readers might know that one of my interests on this blog is the issue of health inequality when it comes to conditions like autism and indeed beyond. It's a sad fact that whilst autism generates quite a lot of discussion, research, politics and emotion, the message that people with autism are people first, and hence are at least facing the same risk (if not more) of developing health-related complications, can get a little lost in the noise. I'm as guilty as anyone in losing this message with my constant ramblings on the wonders of autism research.

Papers like the one from Bilder et al however reinforce the view that autism, in some cases, might very well place a person at increased risk of early mortality and outside of just seeing a person with autism, we could do a lot by putting the person and their physical health and wellbeing first and foremost.

A few bullet-points from the Bilder paper:

  • The study was based in quite an important geographical area for autism research, Utah. Important because as I've blogged about before, Utah has been the focus of some milestone studies on autism, in particular the UCLA-University of Utah Epidemiologic Study of Autism. Indeed some of the original data from that initiative has quite recently become the source of some discussion as part of the autism numbers game and the transition from DSM-III to DSM-IV.
  • In particular, the authors sought to follow-up the cases of autism identified during the UCLA-University of Utah study and estimate the mortality hazard rate ratio (HRR) whilst look at the patterns of mortality and what factors were linked to death. The authors define HRR as: "an estimate of the excess risk associated with the identified exposure (i.e. ASD case status) as a constant effect over a specific time period (i.e. interval since case ascertainment) while adjusting for selected covariates (i.e.birth risk factors)".
  • Based on a cohort of 305 adults with autism (DSM-III and DSM-IV-TR diagnosed/reclassified participants), 29 people (9.5%) had died by the end of 2011. 
  • Most of the deceased were male, and the age at time of death varied between 7 - 46 years old.
  • The HRR overall was 9.91 compared to population controls (N=2466). When taking into account the different genders, the HRR for males was 7.92 and for women with an ASD a staggering 20.71 (without covariates).
  • As one would expect, there were various causes of death indicated on death certificates, ranging from seizures to cardiac-related disorders. Epilepsy was a prominent feature for quite a few of the cohort as per the range of physical disease comorbidity observed in cases. That being said, and without making any judgements, death listed as a result of an "open wound of hip" in one case invites further investigation. 
  • Death by unnatural causes were relatively few although "two individuals died from an adverse event related to medication".

The Bilder paper does make somber reading when you realise that behind every statistic there was a person and a family. Indeed looking at the list of deceased, their intellectual and autism classification, their cause of death and importantly, their accompanying physical disease, one is struck by the heterogeneity present.

With my cold, dispassionate research spectacles on a few points caught my eye.

It was interesting to note how many cases cite cerebral palsy (CP) for example, as a comorbidity to their autism diagnosis (I counted 6 cases of the 29 deceased). I'm not inferring that CP was the cause of death, even though it was listed as a factor in at least one case, but rather that those cases with comorbid CP might indicate some elevated risk of early mortality as per the literature on excess mortality in CP** alone.

Reports of comorbid Sanfilippo syndrome were also present in more than one case of autism, two actually. Given the rarity with which this syndrome appears - anywhere between 1 in 100,000 to 1 in 1,000,000 in the general population - to see even one case in the current cohort should automatically be a cause of some interest bearing in mind presentation can include autistic features and the syndrome previously being mentioned in the autism research literature*** (2 / 222 cases of autism). Given the poor prognosis of Sanfilippo syndrome, one might reasonably assume some impact on mortality where autism also appears alongside.

I also observed that cancers and neoplasms did not seem to prominently figure among the deceased as a cause of death. One could argue that given the relatively young mean age of the deceased, this might play a role in mortality and indeed the lack of accompanying information on things like cancer treatment and survival among this cohort. But I'm interested in this phenomenon, especially when one considers what the population at large - as in everyone whether with autism or not - generally tend to die of (see this post for more information).

The low prevalence of death by unnatural causes also caught my attention. I've talked fairly recently about the wandering and autism research which was published and how wandering can, in a few cases, lead to the very worst outcome when autism is involved. As far as I can surmise and allowing for the difference in age groups looked at, wandering was not implicated in any of the cases looked at by Bilder. I'm not able to provide an in-depth view of wandering leading to the circumstances of early mortality in autism, but one speculates whether an overall increase in cases of autism might be a factor in the very public recording of death implicating wandering. Same goes for issues like suicide, recently discussed with autism in mind****.

Finally, I can't leave this post without commenting on those two people where death was linked to medication use. In both cases, individuals were in their late 30s-early 40s, and one case also carried a diagnosis of schizophrenia. Accepting that medicines - particularly those with a psychotropic effect - are still a voyage of discovery in terms of their precise effects and mode of action, pharmacotherapy is as much about good medicines management as it is about the effects of the drugs themselves. I note for example one case where poisoning due to exposure to "unspecified" drugs is cited alongside the word 'obesity'. The relationship between these concepts has been discussed before including the important issue of side-effects.

The take-home message from this post is best left to another comment from the paper:

"The elevated mortality risk associated with ASD in the study cohort appeared related to the presence of comorbid medical conditions and intellectual disability rather than ASD itself suggesting the importance of coordinated medical care for this high risk sub-population of individuals with ASD."

I would struggle to disagree.

----------

* Bilder D. et al. Excess mortality and causes of death in autism spectrum disorders: a follow up of the 1980s Utah/UCLA Autism Epidemiologic Study. J Autism Dev Disord. September 2012.

** Strauss D. et al. Causes of excess mortality in cerebral palsy. Dev Med Child Neuol. 1999; 41: 580-585.

*** Ververi A. et al. Clinical and laboratory data in a sample of Greek children with autism spectrum disorders. J Autism Dev Disord. 2012; 42: 1470-1476.

----------

ResearchBlogging.org Bilder D, Botts EL, Smith KR, Pimentel R, Farley M, Viskochil J, McMahon WM, Block H, Ritvo E, Ritvo RA, & Coon H (2012). Excess Mortality and Causes of Death in Autism Spectrum Disorders: A Follow up of the 1980s Utah/UCLA Autism Epidemiologic Study. Journal of autism and developmental disorders PMID: 23008058