Decline in autism among high-income white households not caused by a decrease in childhood vaccination, vaccines don’t cause autism

Despite overwhelming evidence that childhood vaccination doesn’t cause autism spectrum disorder (ASD), this belief still persists among some groups. A 2020 study, for instance, found that more than one in three parents of children with ASD believed vaccines were responsible for the disorder^[1].

In August 2024, political economist Toby Rogers—who has previously spread vaccine misinformation—promoted this belief during an interview on CHD.TV, the video platform of Children’s Health Defense (CHD). CHD is an organization known for promoting vaccine misinformation, as reported multiple times by Science Feedback.

During the interview, Rogers cited the findings of a 2020 study by Nevison and Parker that observed a decline in ASD prevalence among high-income white populations in California between 2000 and 2013^[2]. He further claimed that childhood vaccination declined in the same group within the same period.

By juxtaposing these two observations, Rogers implied a potential causal relationship between reduced vaccine coverage and fewer cases of ASD. An excerpt from this interview was posted to Facebook in the form of a reel in early August 2024, which was viewed more than 700,000 times.

In this review, we’ll show how Rogers’ implication isn’t supported by scientific evidence.

Available data show that vaccines don’t cause autism

On multiple occasions, Science Feedback summarized the medical and epidemiological results showing that there is no causal association between childhood vaccination and autism. Several studies assessed whether receiving childhood vaccines was associated with an increased risk of ASD, and they found no causal relationship.

One common claim is that young children receive too many vaccines and that this would lead to neurological disorders including ASD. However, studies comparing the prevalence of ASD with the number of vaccine doses or vaccine antigens received during childhood found no associations between vaccination and ASD.

The study Rogers referenced didn’t show a correlation between vaccination and cases of ASD

Rogers’ claim solely relies on the citation of the 2020 study by Nevison and Parker mentioned earlier. Although Rogers presented Nevison as an expert on autism, Nevison is actually a geologist at the University of Colorado and most of her academic work focuses on the study of greenhouse gasses.

Even though her expertise is unrelated to autism, Nevison authored several publications on autism, several of which listed Mark Blaxill, an anti-vaccine activist, as a co-author. Nevison also wrote for The Defender, a publication of the anti-vaccine organization Children’s Health Defense. Thus, Rogers’ choice to cite the 2020 study by Nevison and Parker to promote the false claim that vaccines cause autism can be viewed in light of her association with anti-vaccine groups.

The Nevison and Parker study referenced by Rogers monitored the number of ASD cases reported to the California Department of Developmental Services (DDS) in each California county by a child’s birth year. The authors explained that the DDS provides services including day care programs, and behavioral management consultation.

According to the authors, children with ASD must “demonstrate significant functional disability in three out of seven life challenges, which include self-care, language, learning, mobility, self-direction, capacity for independent living and economic self-sufficiency” in order to qualify for DDS services. As a result, one caveat about the DDS dataset is that it tends to primarily focus on people with severe forms of ASD.

The authors observed that the prevalence of ASD increased over time. For children born in 1994, the prevalence was 0.4% among all race groups; for children born in 2013, the prevalence reached 1.5% among all race groups.

The study also reported a decrease in ASD prevalence among children born to white families in wealthier California counties from 2000 to 2013. This negative correlation—in other words, the higher the median household income of a county, the lower the ASD prevalence among children born after 2000—wasn’t observed for other races. A weak correlation also existed in the case of Asian families, but the effect was much more conspicuous for white households.

However, the authors didn’t consider vaccination as a possible explanation for this observed change in ASD prevalence. In fact, the study doesn’t mention the word “vaccine” at all, and no data about vaccine coverage was included in the study’s analysis.

Rather, the authors suggested other possible explanations for the decrease in ASD prevalence. For instance, they considered that early ASD diagnosis and intensive behavioral intervention provided to children from richer families with better access to healthcare “may decrease the severity of the ASD diagnosis”, thereby disqualifying those children for DDS. Another possible explanation provided by the authors is that wealthier families simply began to drop out of the DDS program and opted for privately funded services.

If these explanations are correct, then the observed decrease in ASD prevalence would actually reflect a change in the composition of the DDS database, rather than a true reduction in ASD prevalence. However, Rogers made no mention of these explanations in the video.

No data provided to support the claim that vaccine coverage in high income white people in California decreased

To begin with, it’s important to stress that even if there were a correlation between a decline in ASD prevalence and a decrease in vaccine coverage, it wouldn’t be sufficient evidence to demonstrate that one is indeed the cause of the other. Establishing that a correlation exists is an important step to establish a causal association between two phenomena, but on its own is insufficient, as Science Feedback explained in this article.

Rogers’ claim hinges on his statement that vaccine coverage in high-income white people from California decreased over the years since 2000, and that at the same time, ASD prevalence decreased in the same population. But as we explained above, the study by Nevison and Parker, didn’t provide data on vaccine coverage. Rogers didn’t provide such data either. We reached out to Rogers to ask for it. In his reply to our email, Rogers claimed Science Feedback was among “Stasi narrative check organizations”, but provided no evidence to support his claim that vaccine coverage had declined in high-income white households in California.

In a follow-up email, Rogers sent us data showing an increase in vaccine exemptions due to personal beliefs since 2000—something that we discuss below. But he didn’t provide supporting data on an actual reduction of vaccine coverage in children born 2000 or later in white households from wealthier counties in California.

We decided to look into the official vaccination database to determine if U.S. statistics on vaccine coverage supported his claim. The National Immunization Survey Child (NIS-Child) is an annual survey that began in 1994 with the objective of providing official statistics on vaccine coverage in U.S. children. It’s operated by the University of Chicago, under the direction of the U.S. Centers for Disease Control and Prevention (CDC).

Although the raw data of the survey is available on the CDC website, we couldn’t find an analysis that presented vaccine coverage from 2000 to 2013 by state and household income and race at the same time.

However, analyses at the national level are available.

Using data from NIS-Child, a study from Walsh and colleagues showed that the vaccine coverage of several childhood vaccines increased or remained steady among high-income white people^[3]. Between 1998 and 2000, the coverage for the DTaP vaccine was around 90%, 94% for the MMR, and 93% for the polio vaccine. Between 2011 to 2013, it was at 91% for the DTaP, 94% for the MMR, 96% for the polio vaccine.

Thus, vaccine coverage in high-income white families in the U.S. on the whole didn’t decrease after 2000. Nevertheless, we cannot rule out the possibility that California is an outlier in this regard.

For instance, a study showed that the number of vaccine exemptions for personal belief (PBE) increased in California from 2007 to 2013, especially in regions with greater proportions high-income white households^[4].

However, this study didn’t provide data at the individual level. So we don’t know with certainty whether the increase in PBEs occurs in high-income white households, causing a decline in vaccination in that population.

Furthermore, PBEs are just one of the possible reasons why a child may be unvaccinated. Parents unwilling to vaccinate their child may use other strategies, such as medical waivers and homeschooling. Therefore, an increase in PBEs could be compensated by a decrease in homeschooling or medical waivers, such as the overall vaccination coverage doesn’t change much.

In summary, Rogers’ claim that the level of childhood vaccination among high-income white families in California led to lower rates of ASD is unsubstantiated. Country-level data also doesn’t support the hypothesis of a nationwide decline in vaccination among high-income white families.

Conclusion

Roger’s implication about a causal link between a decrease in vaccine coverage and a decline in ASD prevalence is unsubstantiated. The Nevison and Parker study used as supporting evidence did report that the prevalence of ASD declined among white families from richer California counties. However, the study didn’t include data on vaccine coverage and didn’t establish an association with vaccination. Rogers didn’t provide any other data on vaccine coverage either.

More importantly, even if a correlation between vaccine coverage and ASD prevalence was observed, this still wouldn’t be sufficient to demonstrate a causal relationship. A large body of published scientific evidence has shown that childhood vaccination doesn’t cause ASD nor increase the risk of its development.

 

UPDATE (12 October 2024):

Information on vaccine exemption data that Rogers sent us in a follow-up email has been added to the nineteenth paragraph.

 

REFERENCES

• 1 – Sahni et al. (2020) Vaccine Hesitancy and Illness Perceptions: Comparing Parents of Children with Autism Spectrum Disorder to other Parent Groups. Child Health Care.
• 2 – Nevison & Parker (2020) California Autism Prevalence by County and Race/Ethnicity: Declining Trends Among Wealthy Whites. Journal of Autism and Developmental Disorders.
• 3 – Wlsh et al. (2016) Since The Start Of The Vaccines For Children Program, Uptake Has Increased, And Most Disparities Have Decreased. Health Affairs.
• 4 – Yang et al. (2015) Sociodemographic Predictors of Vaccination Exemptions on the Basis of Personal Belief in California. American Journal of Public Health.