Wednesday, 22 April 2015

MMR vaccine was not associated with increased risk of autism

"In this large sample of privately insured children with older siblings, receipt of the MMR [measles-mumps-rubella] vaccine was not associated with increased risk of ASD [autism spectrum disorder], regardless of whether older siblings had ASD. These findings indicate no harmful association between MMR vaccine receipt and ASD even among children already at higher risk for ASD."

That was the conclusion reached in the study by Anjali Jain and colleagues [1] (open-access) based on an analysis of data for over 95,000 children with an older sibling held on "an administrative claims database associated with a large US health plan (the Optum Research Database)." With the aim of reporting "ASD occurrence by MMR vaccine status in a large sample of US children who have older siblings with and without ASD" researchers trawled through the collected data looking "for autistic disorder or other specified pervasive developmental disorder (PDD) including Asperger syndrome, or unspecified PDD (International Classification of Diseases, Ninth Revision, Clinical Modification 299.0x, 299.8x, 299.9x)" as a function of MMR vaccine receipt between birth and the 5 years of age.

They found that nearly 2000 of the 95,000+ children included for study had a sibling diagnosed with autism (2%). As per the results of quite a bit of research in this area, having an older sibling with autism was associated with an increased risk of autism among study participants compared with having an older sibling without autism (6.9% vs. 0.9%).

Then to the MMR...

Vaccination rates were lower for those with an older sibling with autism compared with those with an older sibling without autism (73% at 2 years & 86% at 5 years vs. 84% at 2 years & 92% at 5 years respectively). This difference was present despite similar rates of 'vaccination-associated allergies' being reported. The presence of seizures and the variable of preterm birth were somewhat different between the sibling with ASD and sibling without ASD groupings but the authors don't appear to be reporting such differences to be statistically significant. The issue of vaccine uptake as a function of having an older sibling with autism has been covered before on this blog (see here).

And then to the headlines: 'No association found between MMR vaccine and autism, even among children at higher risk'. This was based on the calculation of unadjusted relative risks (RRs) - "cumulative incidence rate ratios by taking the ratio of the proportion of children who had an ASD diagnosis in an exposed group (either 1 MMR dose or 2 MMR doses) to the proportion of children who had an ASD diagnosis in the unvaccinated group at a given age" and adjusted RRs - "hazard rate ratios estimated from a single Cox proportional hazard regression model that used age since birth as the time scale and included MMR receipt as a time-varying covariate ascribing follow-up time to either the unvaccinated group, the 1-dose group, or the 2-dose group, depending on immunization status at any given age." Further: "An interaction term between MMR receipt and older sibling ASD status was included to allow adjusted RRs to vary by older sibling ASD status." What this boiled down to based on either one dose of MMR received at age 2 or two doses of MMR by age 5 was pretty much nothing in terms of any elevated statistical risk of autism among cases as a function of either having an older sibling with autism or not.

The editorial from Bryan King [2] accompanying the Jain results is also quite illuminating specifically on the possibility that some of the results might be "arguing that MMR vaccine actually reduces the risk of ASD in those who were immunized by age 2 years." I tread very carefully in discussing this sentiment given the confidence intervals (CIs) reported and the idea that delayed MMR vaccination past some pervceived 'critical period' might artificially decrease the risk of autism being diagnosed. I'll come back to this shortly.

The Jain results add further weight to the idea that childhood vaccination is probably not going to be a significant risk factor for the subsequent development of autism. The impressive participant numbers included for study cannot be readily ignored. The data is however not perfect as per the authors critique and issues such as "children in our study who are considered unvaccinated may have received vaccines in settings such as schools or public health clinics in which claims were not submitted." MMR might also not mean MMR in terms of a single combined vaccine as per the comment: "The date of administration of the trivalent MMR (or the last-administered component of monovalent vaccines) was used to determine age at administration for each dose (first or second)."

Vaccination and autism has appeared before on this blog as per my mega-post back in 2014 (see here) and previous musings on the 'too many too soon' suggestion (see here). In both those posts, the data pointed towards no population wide link between vaccination and autism as well as reiterating the important public health message on the value of vaccination. Other subsequent studies have found similar things [3]. This does not rule out isolated events potentially temporally linked to vaccine administration (as per the sentiments of 'Four Kingdoms' post from Dr Tom Insel and past other high-profile cases [4]) but certainly does not mirror the narcolepsy - H1N1 influenza vaccine story for example (see here for some background).

Just before I go, I do want to go back to the issue of whether childhood MMR vaccination (or other vaccines) might actually reduce the risk of autism. Taking into account the research literature looking at something like congenital rubella with autism in mind (see here) one might make a case for suggesting that the rubella vaccine part of MMR vaccine has potentially impacted on the numbers of cases of autism down the years [5]. Whether vaccination with MMR or other immunisations might more directly confer some protection against the development of autism is something that perhaps needs further exploration alongside the possible mechanisms of effect. I'd also put forward the idea that variables such as post-vaccination paracetamol (acetaminophen) use might also require a little more study too [6].


[1] Jain A. et al. Autism Occurrence by MMR Vaccine Status Among US Children With Older Siblings With and Without Autism. JAMA. 2015; 313: 1534-1540.

[2] King BH. Promising Forecast for Autism Spectrum Disorders. JAMA. 2015; 313: 1518-1519.

[3] Uno Y. et al. Early exposure to the combined measles-mumps-rubella vaccine and thimerosal-containing vaccines and risk of autism spectrum disorder. Vaccine. 2015 Jan 3. pii: S0264-410X(14)01689-2.

[4] Poling JS. et al. Developmental regression and mitochondrial dysfunction in a child with autism. J Child Neurol. 2006 Feb;21(2):170-2.

[5] Berger BE. et al. Congenital rubella syndrome and autism spectrum disorder prevented by rubella vaccination--United States, 2001-2010. BMC Public Health. 2011 May 19;11:340.

[6] Schultz ST. et al. Acetaminophen (paracetamol) use, measles-mumps-rubella vaccination, and autistic disorder: the results of a parent survey. Autism. 2008 May;12(3):293-307.

---------- Jain, A., Marshall, J., Buikema, A., Bancroft, T., Kelly, J., & Newschaffer, C. (2015). Autism Occurrence by MMR Vaccine Status Among US Children With Older Siblings With and Without Autism JAMA, 313 (15) DOI: 10.1001/jama.2015.3077