Flawed reasoning: The claim of a lower COVID-19 mortality among the Amish relies on a comparison of infection fatality ratio between the Amish and the general U.S. population by Steve Kirsch. However, risks of underreporting of COVID-19 cases and deaths, as well as other risks of bias, invalidate the comparison.
FULL CLAIM: “New Study Finds Zero Amish Children Diagnosed with Cancer, Diabetes or Autism”; Amish children are largely unvaccinated; “a study [...] found that Covid death rates among Amish communities are 90 times lower than for the rest of America”
REVIEW
As of 15 July 2023, COVID-19 caused more than one million deaths across the U.S. 69.5% of the total population have completed their primary vaccine series since the vaccination campaign began in December 2020, and statistical models estimated that two years of vaccination prevented millions of deaths in the U.S[1].
However, the COVID-19 vaccination campaign was also accompanied by the spread of countless pieces of COVID-19 vaccine misinformation discouraging vaccination. The persistent spread of such claims during the pandemic may have also helped propel a growing distrust of vaccines in general, laying the ground for the narrative that all vaccines, not just COVID-19 vaccines, are useless or even dangerous. Previous reviews by Health Feedback explained that such claims are contradicted by scientific evidence.
One of the claims woven into this anti-vaccine narrative revolves around a discussion about the health of the Amish, which circulated on social media in June 2023. The Amish are a Protestant group living in tight-knit communities that maintain limited contact with the rest of society. They usually follow a preindustrial way of life, avoiding the use of modern technology.
One example of this claim can be found in this Leading Report article, published on 9 July 2023, alleging that a “new study [found] zero Amish children diagnosed with cancer, diabetes or autism” because they are “strictly 100 percent unvaccinated”. The article also claimed that the “Covid death rates among Amish communities are 90 times lower than for the rest of America” because they didn’t get vaccinated. Leading Report is recognized as a Questionable Source by Media Bias/Fact Check, regularly promoting conspiracy theories and misinformation.
It’s worth noting that the claim that the health of the Amish is much better because they don’t vaccinate isn’t new; Vaxopedia and Snopes published articles in 2016 refuting the claim.
This recent claim revolving around this allegedly “new study” is also inaccurate and misleading, as we will show below.
Contrary to the claim, there is no new scientific study
This claim lends users the impression that a newly peer-reviewed published scientific study provides supporting evidence for the claim. A closer look at the source of the claim in the article showed that there isn’t actually a “new study”. Instead, the article repeats the main arguments of entrepreneur Steve Kirsch’s testimony in front of the “Medical Freedom Panel 2023” at the Pennsylvania State Senate on 9 June 2023. In his testimony, Kirsch, who has repeatedly propagated inaccurate or unsubstantiated information on COVID-19 and vaccines in the past, didn’t provide any scientific publications to support his claims.
Many Amish parents vaccinate their children
First, published scientific studies show that the Amish aren’t “100 percent unvaccinated”. But it is challenging to obtain consistent estimates of vaccine coverage in the Amish because they live in close communities without telecommunication, therefore different studies report different levels of coverage, as shown below.
- In 2011, a study of 359 Amish parents found that only 14% “reported that none of their children were vaccinated”[2].
- A 2017 study of 84 Amish parents found that 97% of them accepted at least some vaccinations for their children[3].
- A 2021 study of 391 parents showed that 41% of respondents accepted at least some vaccinations for their children[4].
The studies mentioned above relied on collecting information by a questionnaire, which in turn depends on the families’ willingness to complete the questionnaire and return it by mail. This method of collecting information could introduce a sampling bias, because those who responded may not have been representative of the Amish population as a whole.
Nonetheless, what these results show is that the Amish aren’t entirely unvaccinated as the article asserted. Therefore, any attempt to link vaccines to the presence or absence of a given health condition among the Amish should at least take into consideration their vaccination status. This is something that neither Kirsch nor those who propagated his claim did.
The Amish still develop health conditions such as autism or cancer
It is inaccurate to say that there are no health conditions such as autism, cancer, or diabetes among the Amish. A conference paper from the International Society for Autism Research dating to 2010 reported an autism prevalence of 1 in 271 children in Amish communities. This was lower compared to the overall U.S. prevalence of 1 in 91 children at that time, but is by no means zero, contrary to the claim.
Braxton Mitchell, an epidemiologist at the University of Maryland who specializes in studying diseases in the Amish, confirmed to PolitiFact that cases of autism, cancer, and diabetes do occur among the Amish.
A study of a cohort of Amish from 1996 to 2003 found that the cancer incidence was 389 cases per 100,000 person-years, which was 60% of the cancer incidence in the general population. This shows that Amish people do develop cancer, albeit at a lower rate, contradicting the claim of zero cancer among them.
A presentation at the American Public Health Association annual meeting in 2005 reported that Amish women experienced breast cancer, albeit to a lower extent than women in the general population. However, preliminary data suggested that their mortality rate from breast cancer was higher than for women of the general population. Because Amish women die at a higher rate from preventable cancers, tumor detection strategies directed at Amish women like mobile mammograms have been developed.
It is also inaccurate to claim that there is no diabetes among the Amish. A study reported cases of diabetes at a lower prevalence compared to the general population[5].
To summarize, all the premises upon which the claim is based—the absence of vaccination and of several health conditions among the Amish—are directly contradicted by available scientific evidence, thus invalidating the claim that Amish are healthier because they don’t vaccinate.
The lower incidence of various health conditions such as cancer and diabetes could be due to certain lifestyle factors associated with the Amish, such as a diet low in processed food, reduced smoking, and increased physical activity[6].
Another possible explanation is that Amish communities are relatively isolated with limited interaction with health services. Therefore, it is plausible that many cases of conditions like autism or cancer are simply not detected. It’s important to keep in mind that both explanations for the phenomenon aren’t mutually exclusive.
The Amish had a greater risk of dying during the COVID-19 pandemic compared to the U.S. population
The claim that Amish people were much less likely to die from COVID-19 than the rest of the population is based on a deeply flawed methodology.
According to Kirsch’s testimony at the Pennsylvania Senate, he drove to Lancaster County, home to many Amish communities, and asked around for people who were known to have died from COVID-19. He repeated this question on Twitter and, according to him, got no answer.
There are significant problems with this approach. Given that COVID-19 symptoms resemble symptoms of other respiratory diseases, it’s possible that some people who died from COVID-19 weren’t diagnosed due to communities’ relative isolation or limited access to healthcare, but Kirsch didn’t take that possibility into consideration.
Kirsch also didn’t account for the risk of sampling bias. There are over 30,000 Amish in Lancaster County. We weren’t able to find evidence that Kirsch interviewed all of them or at least ensured that those he interviewed were representative of the Amish population. It’s also unclear how he verified that the persons he allegedly interviewed knew of all the COVID-19 deaths occurring in the other communities scattered across Lancaster County, in order to ensure a satisfactory level of reporting of his survey. Science Feedback reached out to Kirsch to inquire about these methodological questions; the review will be updated if additional information becomes available.
According to his own statement, Kirsch calculated a COVID-19 infection fatality ratio (IFR) for the Amish and compared it to the IFR of the entire U.S. population. In order to calculate an IFR, Kirsch must have obtained an estimate of the number of infections among the Amish. This doesn’t mean the number of detected cases, but the number of all SARS-CoV-2 infections, both diagnosed and undiagnosed. Again, Kirsch didn’t explain how he achieved such a feat.
The Amish way of life has several major differences to the wider U.S. population. In particular, the Amish mostly live in smaller, scattered communities, whereas 80% of the U.S. population lives in urban areas. Obviously, population density directly affects the risk of infection and access to healthcare. Whether these differences were considered by Kirsch before making his comparison of infection fatality ratio between the Amish and the U.S. population is unclear.
In addition to these methodological shortcomings, scientific studies attest that mortality was high among the Amish during the pandemic. A 2021 study gathered obituaries submitted to The Budget, one of the longest-established Amish newspapers. According to the authors, death certificates typically don’t include the religion of the decedent, this is why resorting to obituaries proved necessary to identify deaths occurring specifically among the Amish.
The authors analyzed the number of deaths reported to the Budget and found that Amish communities experienced a peak of excess mortality in November 2020[7].
Excess mortality represents the number of all-cause deaths beyond what would be expected based on previous years’ records. Because excess mortality doesn’t depend on the accurate diagnosis of COVID-19, it provides a fair representation of the direct and indirect mortality associated with the disease.
In a follow-up study covering 2020 and 2021, the authors compared the excess mortality among the Amish and the general U.S. population. They found that the Amish communities experienced a higher excess mortality most of the time[8].
Therefore, data actually indicates that the Amish suffered more losses during the COVID-19 pandemic than the rest of the population, in contradiction with Kirsch’s conclusions.
Conclusion
There is no basis in fact for the claim that the Amish don’t experience medical conditions such as autism, cancer, and a lower COVID-19 mortality rate because they don’t vaccinate.
This claim relies on false assumptions and flawed calculations. The scientific evidence and data actually show that a large proportion of the Amish population vaccinate their children, at least to some degree. The Amish follow a healthier lifestyle that includes daily physical activity, non-industrially processed food, and less exposure to urban air pollution. This may reduce the incidence of certain chronic diseases in this population.
Nevertheless, cases of autism, cancer, and diabetes are still recorded among the Amish. The available evidence also suggests that the Amish experienced higher excess mortality during COVID-19 compared to the rest of the U.S. population, contrary to the claims made by Kirsch and Leading Report.
REFERENCES
- 1 – Sah et al. (2022) Estimating the impact of vaccination on reducing COVID-19 burden in the United States: December 2020 to March 2022. Journal of Global Health.
- 2 – Wenger et al. (2011) Underimmunization in Ohio’s Amish: parental fears are a greater obstacle than access to care. Pediatrics.
- 3 – Kettunen et al. (2017) Evaluation of low immunization coverage among the Amish population in rural Ohio. American Journal of Infection Control.
- 4 – Scott et al. (2021) Vaccination patterns of the northeast Ohio Amish revisited. Vaccine.
- 5 – Hsueh et al. (2000) Diabetes in the Old Order Amish: characterization and heritability analysis of the Amish Family Diabetes Study. Diabetes Care.
- 6 – Hairston et al. (2013) Comparison of BMI and Physical Activity Between Old Order Amish Children and Non-Amish Children. Diabetes Care.
- 7 – Stein et al. (2021) Closed but Not Protected: Excess Deaths Among the Amish and Mennonites During the COVID-19 Pandemic. Journal of Religion and Health.
- 8 – Stein et al. (2023) Pathways to Immunity: Patterns of Excess Death Across the United States and Within Closed Religious Communities. Journal of Religion and Health.
