Children with autism are less likely to be fully vaccinated. Thanks, antivaxers!

In the world of clinical trials (and medicine in general) children are considered a vulnerable class of patients. That’s why the Common Rule and the federal government’s rules for human subjects research classify children as a vulnerable class for whom protections in clinical trials and other human subjects research must be more stringent.

Of course, if children are considered a vulnerable class of patient, then children with disabilities or neurodevelopmental disorders are an even more vulnerable class of patients. For example, autistic children are definitely vulnerable. That’s one reason why “autism biomed” quackery irritates me so much. Thanks to such quackery, vulnerable autistic children are routinely subjected to dangerous quackery like chelation therapy, chemical castration, and even bleach enemas. All of these quack interventions come from the antivaccine movement, who view vaccines as the cause of autism. Their fervent, undying belief that vaccines cause autism is the basis of many of these quack “autism biomed” treatments. It’s also a basis of another harm inflicted on autistic children by their antivaccine parents.

From a CNN story:

Children with autism spectrum disorder are significantly less likely to be fully vaccinated than children unaffected by autism, new research finds. And the same is true of their younger sisters and brothers.

“This study is showing that children with autism and their younger siblings might be at greater risk of vaccine-preventable diseases,” said Ousseny Zerbo, lead author of the study and a postdoctoral fellow with the Kaiser Permanente Northern California Division of Research.

The study was published in JAMA Pediatrics.

This study seeks to answer a very simple question that you and I know the most likely answer to: After receiving an autism spectrum disorder diagnosis, do children obtain all of their remaining scheduled vaccines, and are the younger siblings of these children vaccinated according to vaccine recommendations?

Yes, you and I know the anser to both of these questions: No and no. It’s fairly obvious to anyone who pays a lot of attention to the antivaccine movement. Parents of autistic children have a disturbing propensity to become antivaccine, and, as a result, they don’t vaccinate after that diagnosis. That applies to both their autistic children and to any subsequent children they have. But, to you and me, most likely this conclusion is due to just our personal observations, which of course can be influenced by confirmation bias.

The authors of this study note that there is evidence for this impression:

Despite numerous scientific studies reporting no association between childhood vaccination and ASD, there remain concerns about such a connection for some of the public.16 In recent years, rates of undervaccination and vaccine refusal have been on the rise in the United States and have been associated with vaccine-preventable disease outbreaks. Rates of undervaccination among the subpopulation of children with ASD have not been fully investigated. A survey conducted among 98 parents of children with ASD and 65 parents of children without ASD in Canada found that a lower proportion of children with ASD received their measles, mumps, and rubella (MMR) or diphtheria and tetanus toxoids and acellular pertussis and inactivated poliovirus (DTaP-IPV) vaccines compared with children without ASD.25 Because the first dose of MMR and the first 3 doses of DTaP-IPV are recommended before the age when ASD can be reliably diagnosed (which is at least 2 years), it was not clear from that study if the lower observed vaccination rates among the children with ASD were a consequence of the child’s ASD diagnosis. In a recent letter to the editor, Glickman and colleagues reported no significant difference between rates of vaccination of 71 children with ASD and those of 135 children without ASD. However, they found that families with children with ASD were less likely to vaccinate subsequent children. Other studies also found that parents of children with ASD were more likely to either delay or refuse vaccination for their younger children. In a survey of 197 parents, Bazzano and colleagues found that half of the parents of children with ASD changed vaccination practices for their younger children because of beliefs that vaccines contributed to their child’s ASD. After surveying 486 parents of children with ASD, Rosenberg and colleagues30 found that almost 20% of parents declined or delayed MMR immunization in the younger siblings of children with ASD. Previous studies were limited by small samples, lack of comparable control groups, or restriction to specific vaccines.

Hence this study.

The design of the study was a retrospective matched cohort study using data from six integrated health care delivery systems across the United States within the Vaccine Safety Datalink (VSD). Participants were children born between January 1, 1995, and September 30, 2010, and their younger siblings born between January 1, 1997, and September 30, 2014. The end of follow-up was September 30, 2015. The VSD is a database that I’ve written about before. Basically, it’s an active surveillance system run by the CDC in which prospectively collected data are collected and examined for evidence of complications or adverse reactions due to vaccines. The study itself compared vaccination rates between a matched cohort of 3,729 children with autism spectrum disorder (ASD) and 592,907 children without ASD.

Vaccine status was determined thusly:

To assess the vaccination patterns of children after receiving an ASD diagnosis, we only included children who were at least 7 years old as of September 30, 2015, and in whom ASD was diagnosed at 5 years or younger. We limited this assessment to vaccines routinely recommended between ages 4 and 6 years and ages 11 and 12 years. For the comparison of vaccination patterns of younger siblings of children with ASD and younger siblings of children without ASD, we included children (siblings) who were at least 1 year old as of September 30, 2015, and assessed vaccines routinely recommended by the ACIP at ages 1 to 11 months (vaccine doses given at birth were not assessed), 1 to 2 years, 4 to 6 years, and 11 to 12 years. Children who received all of their vaccine doses within the ACIP-recommended age limits were considered fully vaccinated. For example, we considered a child who received at least 1 dose of DTaP, at least 1 dose of MMR, at least 1 dose of IPV, and at least 1 dose of varicella vaccine between ages 4 and 7 years as fully vaccinated for vaccines recommended at ages 4 to 6 years (Table 1) regardless of vaccination history before age 4 years. We required that children be health plan members during the periods the vaccines were recommended (eg, to assess vaccination status at ages 1-2 years, we required that the children be health plan members between ages 1 and 2 years). For the younger siblings, we assessed vaccination status regardless of ASD diagnosis. Vaccine refusal was identified using ICD-9-CM codes V64.05 and V64.06.

The results:

The proportion of children who received all recommended vaccine doses (Table 1) between ages 4 and 6 years was lower in children with ASD compared with children without ASD (81.6% [2331 of 2855] vs 94.1% [455,435 of 483,961], respectively) (Table 2). The proportion receiving each individual vaccine was also lower among children with ASD compared with children without ASD. For MMR vaccine, 84.0% (2397 of 2855) of those aged 4 to 6 years with ASD were vaccinated compared with 95.9% (464,245 of 483,961) of those without ASD. After adjusting for maternal age at the time of the child’s birth and race/ethnicity (which were both associated with ASD in our bivariate analyses) and the matching variables (month and year of birth, sex, and site), children with ASD were significantly less likely to be fully vaccinated (adjusted RR, 0.87; 95% CI, 0.85-0.88) compared with children without ASD. Adjusted RRs were also significant for each individual vaccine.

Or, to put it more simply, children with ASD were 13% less likely to be fully vaccinated. That might not sound like a lot on the surface, but it’s definitely potentially significant. Even worse, parents of children diagnosed with ASD are less likely to vaccinate and more likely to delay vaccines or fail to vaccinate—or refuse vaccines altogether. Here’s a graph telling the tale. It describes “Parental Vaccine Refusal of Any Vaccine Dose for Younger Siblings by Age Category and by Child Autism Spectrum Disorder (ASD) Status of Older Siblings”:

Rates of vaccine refusal

Rates of vaccine refusal

As you can see, younger siblings of children with ASD are roughly three times more likely to have had their parents refuse one or more vaccines in the 4-6 year old category, but, as the authors point out, each age category of vaccination before age ten years. They also noted that the highest rates of undervaccination:

Undervaccination in siblings of children with ASD

Undervaccination in siblings of children with ASD

In a news report of the study, its first author, Ousseny Zerbo, noted:

“I was not expecting to see this big of a difference,” says Zerbo of the results. “From this study, having lower rates of vaccination among children with autism and their siblings suggests that they might have higher risk of vaccine-preventable diseases.”

To be honest, I was rather surprised that the difference wasn’t even bigger, but maybe that’s because I spend a lot more time dealing with the antivaccine movement. In that, I’m with Dr. Andrew Adesman, chief of developmental and behavioral pediatrics at Cohen Children’s Medical Center of New York:

“It is concerning that children with autism spectrum disorder and their younger siblings were under-vaccinated compared to the general population, since the key to preventing various childhood infectious diseases is to make sure as many children as possible are fully vaccinated,” Adesman said.

In fact, Adesman said he was “a bit surprised the differences were not even greater, especially given that this study was conducted in California, where under-vaccination had been a significant problem until the state recently changed its laws requiring vaccine exemptions.”

However, “since all of the data for this study was collected in California prior to when the new laws went into effect — July 1, 2016 — it is likely that the under-vaccination rates that these researchers found have dramatically improved, and may even have disappeared,” Adesman added.

It will indeed be interesting to update the study with data after 2016 to see whether this difference in vaccine uptake among children with ASD and their younger siblings. We already know that the law referred to above, SB 277, which eliminates non-medical so-called “personal belief exemptions” to school vaccine mandates, has already started to work to increase vaccine uptake rates.

I also can’t help but conclude with a little snark. Antivaxers frequently argue for a “vaccinated vs. unvaccinated” study, comparing health outcomes between these two groups of children. Some go so far as to demand a highly unethical randomized, double-blind, placebo-controlled trial of the current vaccine schedule. Others grudgingly accept that such a trial would be unethical because it would leave the control group vulnerable to vaccine-preventable disease and propose an epidemiological study, blithely ignoring just how large and expensive such a study would have to be to be adequately powered to detect differences in ASD prevalence in the two groups. Not that this has stopped antivaxers from trying to do “studies” of their own, ranging from the risibly bad to the just dubious.

In reality, real studies of vaccinated versus unvaccinated children inevitably show either no difference in health outcomes of interest to antivaxers associated with vaccination or even that vaccinated children tend to be healthier. Some studies show that the benefits of vaccines go beyond the disease vaccinated against. MMR, for instance, protects against more than just measles. More amusingly, however, and more specifically addressing the study I’m discussing today, if a “vaxed/unvaxed” study were done looking at whether vaccinated children were more likely to be diagnosed with autism it is conceivable that the lower rate of vaccination in autistic children could produce a result suggesting that vaccination protects against autism! Sure, good controls and careful design should prevent such an outcome, but antivaxers should be careful what they ask for.