Wednesday, 4 March 2015

Asthma and autism: a spanner in the works?

As happens so many times in autism research, spanners are thrown in works. Take the paper from Ousseny Zerbo and colleagues [1] who concluded that: "children with autism have elevated prevalence of specific immune-related comorbidities". Nothing surprising about that finding based on the volumes of other research which seemed to have reached similar conclusions (see here).

Then the spanner: "asthma was diagnosed significantly less often" in autism cases compared with asymptomatic controls. Asthma diagnosed significantly LESS often? I was taken aback by this finding bearing in mind the small but significant difference in cases reported (13.7% vs. 15.9% respectively). Quite a few times on this blog I've talked about research which indicated something quite the opposite when it came to asthma and autism as per post such as this one and this one. One might quibble about what was correlating with what (autism rates elevated in cases of asthma or asthma rates elevated in cases of autism) but there did seem to be something of a relationship between the conditions noted.

OK, let's step back a little. The Zerbo paper (from a group who've appeared before on this blog) and a few details...

  • A case-control study based on data derived from "among members of Kaiser Permanente Northern California (KPNC) born between 1980 and 2003". Further: "Cases were children and young adults with at least two autism diagnoses recorded in outpatient records (n=5,565)" who were matched 5:1 based on "birth year, sex, and length of KPNC membership (n=27,825)." Sample size was therefore pretty good.
  • Databases were combed for outcomes - "asthma, allergies, and autoimmune diseases" - and differences were detailed complete with odds ratios (ORs).
  • Results: as just described, there was something of a difference between the autism and not-autism groups in that: "Allergies and autoimmune diseases were diagnosed significantly more often among children with autism than among controls." Asthma diagnosis aside, psoriasis got a particular mention occurring "more than twice as often in cases than in controls." This is not the first time that the skin condition psoriasis has been mentioned with autism in mind (see here).
  • Conclusion: "results support previous observations that children with autism have elevated prevalence of specific immune-related comorbidities."

I'm having a hard time with the 'asthma diagnosed less often in autism' sentiments of the Zerbo paper. I talked about the Kotey paper [2] last year (2014) and their headline: "Asthma is approximately 35 % more common in autistic children." That finding complemented the various data coming out of the fabulous Taiwanese database (see here) including asthma and other autism comorbidity being potentially related. Yes, one could highlight the differences between these studies in terms of geography and also the type of data collected (reports vs. medical records). One might also consider the impact of other variables such as mold exposure too [3] potentially linked to asthma. But could science have been that wrong?

Well, there have been other examples in the research literature where asthma does not seem to be over-represented when it comes to a diagnosis of autism. Take the paper from Mrozek-Budzyn and colleagues [4] as an example. The findings reported by Bakkaloglu and colleagues [5] are another case where there was very little reported (accepting the much lower participant number included for study). And then there is the recent report from Kristen Lyall and colleagues [6] to consider: "Prevalence of asthma and overall allergies did not differ between cases and controls" based on their cohort, albeit flanked by a finding that: "food allergies and sensitivities may be more common in children with ASD [autism spectrum disorder]."

That all being said, I'm not going to fall into the scientific trap that normally follows null findings over positive findings specifically with autism in mind. My recent musings on 'what have we learned about autism' (see here) kinda said it all - plurality, different trajectories and outcome and comorbidity clusters - in terms of the potential way forward for autism research. The idea that certain 'types' of autism might be preferentially more strongly related to certain types of comorbidity is perhaps a better way of approaching this issue. As per the growing evidence base on air pollution potentially 'correlated' with some autism, asthma might be something of an important variable to study (see here). Oh, and then there are certain pharmaceutics to potentially consider [7] too in light of other associations (see here).

Music to close: Pulp and er, don't do drugs kids!


[1] Zerbo O. et al. Immune Mediated Conditions in Autism Spectrum Disorders. Brain Behav Immun. 2015 Feb 11. pii: S0889-1591(15)00019-7.

[2] Kotey S. et al. Co-occurrence of autism and asthma in a nationally-representative sample of children in the United States. J Autism Dev Disord. 2014 Dec;44(12):3083-8.

[3] Karvonen AM. et al. Moisture Damage and Asthma: A Birth Cohort Study. Pediatrics. 2015. Feb 16.

[4] Mrozek-Budzyn D. et al. The frequency and risk factors of allergy and asthma in children with autism--case-control study. Przegl Epidemiol. 2013;67(4):675-9, 761-4.

[5] Bakkaloglu B. et al. Atopic features in early childhood autism. Eur J Paediatr Neurol. 2008 Nov;12(6):476-9.

[6] Lyall K. et al. Asthma and Allergies in Children With Autism Spectrum Disorders: Results From the CHARGE Study. Autism Res. 2015. Feb 26.

[7] Becker KG. & Schultz ST. Similarities in features of autism and asthma and a possible link to acetaminophen use. Med Hypotheses. 2010 Jan;74(1):7-11.

---------- Zerbo O, Leong A, Barcellos L, Bernal P, Fireman B, & Croen LA (2015). Immune Mediated Conditions in Autism Spectrum Disorders. Brain, behavior, and immunity PMID: 25681541