Prevalence Rates of Autism Spectrum Disorders Among the Old Order Amish

17 May

One of the topics that comes up over and over online is “The Amish don’t vaccinate” and “the Amish don’t have autism”. Both statements are incorrect. The Amish have no religious prohibition against vaccination and they do have autism.

The question of autism amongst the Amish has been studied and is being presented at the IMFAR autism conference this week. The paper,
Prevalence Rates of Autism Spectrum Disorders Among the Old Order Amish, demonstrates a preliminary prevalence of 1 in 271 as the prevalence of autism amongst Amish children in two Amish communities: Holmes County, Ohio and Elkhart-Lagrange County, Indiana.

J. L. Robinson , Hussman Institute for Human Genomics, University of Miami Miller School of Medicine, Miami, FL
L. Nations , Hussman Institute for Human Genomics, University of Miami Miller School of Medicine, Miami, FL
N. Suslowitz , Center for Human Genetics Research, Vanderbilt University, Nashville, TN
M. L. Cuccaro , Human Genetics, University of Miami School of Medicine, Miami, FL
J. Haines , Center for Human Genetics Research, Vanderbilt University, Nashville, TN
M. Pericak-Vance , Hussman Institute for Human Genomics, University of Miami Miller School of Medicine, Miami, FL

The prevalence rate of Autism Spectrum Disorders (ASD) appears to be steadily increasing. The latest report from the Center for Disease Control estimates the rate of ASD is 1 in 91 children (Kogan, 2009), up from 1 in 150 in 2007. Understanding the seeming changes in ASD prevalence require careful exploration of genetic and environmental factors. A method that has proven useful in dissecting the etiology of complex diseases is the study of isolated populations. One population isolate that has been studied extensively is the Amish, with well over 250 genetic studies. Expanding studies of autism to the Amish may provide important information about etiology. A crucial first step in this process is a feasibility study to determine ASD prevalence rates in this population.


This study presents preliminary data on the estimated prevalence of ASD among the Amish in two Amish dominant counties as part of a larger epidemiological study. All children between ages 3 to 21 in those counties will be screened for the presence of an ASD.


Screening occurred in, two of the largest Amish communities in the United States. Trained clinicians ascertained door to door using a published Amish Directory as a guide. Families were approached and asked to participate in a brief interview regarding their children. Two primary screening instruments were used: the Social Communication Questionnaire (SCQ) and the DSM-IV-TR Checklist (a tool created by the authors). A Vaccination History and a brief family history including questions specific to the ASD phenotype were also taken. Children screening positive on either the SCQ or DSM-IV-TR Checklist were seen for a more comprehensive clinical evaluation by two licensed psychologists. This evaluation included the Autism Diagnostic Observational Schedule (ADOS) and Autism Diagnostic Interview (ADI).


From September 2008 to October 2009, 1899 Amish children were screened in the two Amish communities. A total of 25 children screened positive for ASD on either the SCQ or the DSM-IV-TR checklist. A total of 14 screened positive for ASD on both screeners. Of those 25 children, 14 were evaluated and seven children were confirmed as having a diagnosis of ASD using the ADI and/or ADOS, and clinical judgment. Interestingly, four of the seven only met ASD criteria on the ADOS but not the ADI. Three of the four who were not diagnosed by the ADI only missed criteria on the Behavioral Domain, which may be attributable to the reporting style of Amish caregivers.


Preliminary data have identified the presence of ASD in the Amish community at a rate of approximately 1 in 271 children using standard ASD screening and diagnostic tools although some modifications may be in order. Further studies are underway to address the cultural norms and customs that may be playing a role in the reporting style of caregivers, as observed by the ADI. Accurate determination of the ASD phenotype in the Amish is a first step in the design of genetic studies of ASD in this population.

18 Responses to “Prevalence Rates of Autism Spectrum Disorders Among the Old Order Amish”

  1. David N. Brown May 18, 2010 at 00:16 #

    I have a couple pieces up on this at “Evil Possum”. I observed two things. First, most of Olmsted’s sources (if not all) were “alternative medicine” practitioners or sympathizers. Second, he was very fuzzy on distinctions between Old Order Amish and related but distinct groups like the Mennonites and Beasley Amish. My conclusions have been that Olmsted never did enough research to be conversant with Lancaster County culture, and mostly had the “Amish anomaly” idea handed to him by others.

    • Sullivan May 18, 2010 at 00:52 #

      David N. Brown,

      one thing I should have noted–the idea of “The” Amish is a misnomer. Genetically, the Amish are basically island populations that split generations ago. The various communities could very well have differences in genetics which could result in differences in the populations.

      For example, the Ohio Amish have a high incidence of cystic fibrosis. I’ve been told (by someone who should know) that the Lancaster Amish have no cystic fibrosis.

  2. Brooke May 18, 2010 at 00:51 #

    It will be interesting to see where this “reporting style” issue winds up. After all, if Old Order Amish have cultural norms that preclude Autism being regarded as a ‘disorder’ than the wobbly sociological underpinnings of the diagnosis will be on display.

  3. Joseph May 18, 2010 at 02:26 #

    Methodologically, some things are not clear. Just from the numbers (25 & 14) it would appear that they only evaluated the 14 children who managed to clear both screening tools. One screening tool is well known (the SCQ), and it has a so-so false-negative rate. The other screening tool was apparently created by the authors for the study, and we can only take an educated guess at its false-negative rate.

    If you combine two screening tools the way they apparently did, your overall false-negative rate will obviously go up. For example, if the combined false-negative rate is 50%, then the prevalence the authors found is only half the true prevalence.

    (They could take some clues from the NHS adult study, where false-negatives in screening were — probabilistically — a non-issue.)

  4. Sullivan May 18, 2010 at 02:32 #


    the authors make it clear that this is a “preliminary” study. It is also just an abstract (to my knowledge at least), not a peer reviewed paper.

    I wouldn’t take “1 in 271” as a hard number at all.

    The main interest to me in this study is the fact that they (1) Actually tried to interview a large number of Amish families and (2) they appear to have found some autistics amongst these families.

    It is a start.

    I am very interested to see how many of these children had ASD (or any) diagnoses prior to this study. Assuming, of course, that the diagnoses from this study are valid.

  5. Corina Becker May 18, 2010 at 02:59 #

    Mennonites and Amish are related, so I’m not surprised that a casual observer would mix up the groups, but for researchers to? Olmsted rather fails at his research.

    And yes, there is no religious prohibition of vaccines or modern medicine. We all disagree about things like cars, computers, and other technologies, since we have different opinions on what constitutes as “simple” life, but medicine? Medicine is good, saves the children, saves the livestock and heals people.

    The impression that I get from my limited experience with the Old Orders is that yes, they will have to take in the cultural differences to properly assess the communities. In isolated communities where social standards and gender roles are firmly established and clearly defined, with routines for chores and long-standing traditions, I think that it’s harder for “milder” cases to stand out, and to properly assess the group as a whole.
    There’s also the instilled wariness of outsiders. Tend to stick to one’s own.

    Corina Becker
    ~autistic adult of Mennonite descent.

  6. David N. Brown May 18, 2010 at 07:39 #

    Two more very interesting things I unearthed comparing Olmsted’s “reporting” to independent sources on the Amish. First, a very prominent source was one Frank C. Noonan, who claimed to have treated “thousands and thousands” of Amish at a “wellness center” without ever seeing a case of autism. Suspiciously, no name or location was given for the wellness center, hence, none of what Noonan claimed, up to and including the EXISTENCE of the center, can be verified from what Olmsted provided.

    Second, apart from Noonan, it appears that Olmsted’s other sources claiming experience treating the Amish confirmed specific cases of unvaccinated autistic Amish. One of these cases was in TEXAS. I double-checked this, and yes, there is a documented population of fewer than 200 Old Order Amish in Texas. Therefore, Olmsted had proof that at least one Amish subpopulation had the same ca. 1/150 autism rate reported elsewhere.

  7. Prometheus May 19, 2010 at 00:27 #

    The Old Order Mennonites and Old Order Amish are – as mentioned above – “island populations” that have been genetically separated from the general population in the US for over 150 years.

    In addition, as a number of studies have shown, the Old Order groups have also had at least one and often more than one “bottenecking” event, where their populations went below two hundred individuals (below one hundred in some groups). Consequently, there is a lot of cosanguinity (inbreeding) in the Old Order Mennonites and Amish.

    As a result, not only are the Old Order Mennonites and Amish not comparable to the general population in the US (or elsewhere), the individual groups are not comparable to each other.

    One expected result of such small populations with little (almost zero, in fact) immigration into the group is that some gene alleles will become more rare (or even lost) and others will become more prevalent than in the general population – they may even become “fixed” (everyone having two identical copies of the same allele). That’s why some groups have higher incidence of Maple Syrup Urine disease, others have higher incidence of Cystic Fibrosis and others… well, you get the idea.

    It may be true that some of the Old Order Mennonite and Amish groups have a lower prevalence of autism than the general population, but the most likely explanation is that they have a lower prevalence of autism because they have lost some of the alleles that cause or predispose to autism.

    Assuming that any differences between the Old Order Amish (or even “regular” Amish) and the general population are cause by vaccination is an amazing leap of illogic.


  8. tony bateson May 26, 2010 at 17:52 #

    Well here we go again. The unconscious contradictions amongst your correspondents are illuminating. If autism rates are lower in the Amish as your writers suggest “is it down to alleles?”, well I get the idea, it’s anything but vaccines. How about that “one unvaccinated Amish in the Texas families is autistic”. Every one tells me about one here or one there! No, its the hundreds or thousands I am interested in if vaccines are neutral to autism as the medical establishment say.

    The Chief Advisor to the UK Government said, and its on the video record that 10/12% of kids may not have been vaccinated in the UK since the late 1980s when MMR was introduced. Where are the one in a hundred autistic kids out of around a million plus unvaccinated kids please?

    People are not vaccinating despite your attempts to frighten them to death and as the population of unvaccinated and unautistics grow so the truth will emerge. It will be called ‘Wakefield’s Revenge’.

    Tony Bateson, Oxford, UK.

    • Sullivan May 26, 2010 at 18:23 #

      tony bateson,

      will you ever address the facts presented to you? For example, vaccine uptake is not as low as you claim. As noted before, the vaccine uptake for MMR in the UK is about 92%. Here is the first hit on my google search for: vaccine uptake united kingdom.

      Uptake of one dose of MMR (MMR1) by 24 months for children born in the calendar year of 2005 is 92.1%. This is the same as the previous reporting calendar year (children born in 2004).

      If the “Chief Advisor” said differently, he was mistaken.

      Vaccination rates in Lancaster county have increased substantially over the past 20 years. People who would know (i.e. those who are actually doing the vaccination) put the vaccine rate rising from 20% to 70%. By your logic, there should be an increase in the autism rate in Lancaster County. Please demonstrate this.

      Multiple people have shown up on discussion forums you participate in–people claiming to have unvaccinated autistic children. I have seen zero effort on your part to contact those people. Likely this is because you realize that the chances of those people responding to you are slim. Nontheless, you perseverence in seeking unvaccinated autistics is underwhelming.

      “Wakefield’s Revenge”. Thanks for that one! I’ve already forwarded it to some friends.

    • Sullivan May 26, 2010 at 18:27 #

      “No, its the hundreds or thousands I am interested in if vaccines are neutral to autism as the medical establishment say.”

      Just how big do you think the Amish communties are in the United States? Thousands of autistic Amish? Do me a favor and calculatet what 1% of 113 is. (113 is the estimated size of the Amish population in Texas). If anyone claimed “hundreds” or “thousands” of autistic Amish in Texas, we would know they were…how doe Generation Rescue put it?….”expanding the truth”.

  9. Joseph May 26, 2010 at 21:37 #

    The Chief Advisor to the UK Government said, and its on the video record that 10/12% of kids may not have been vaccinated in the UK since the late 1980s when MMR was introduced.

    @Tony: Where can we find this video? Preferably, do you have the exact statement?

    • Sullivan May 26, 2010 at 22:52 #


      tony bateson’s previous statement was “An estimation based upon these figures suggests 10%/12% of children born since 1966 have not had childhood vaccinations”

      So, now it has changed from since 1966 to the 1980’s?

      Following the same link I gave above, the uptake rate by 12 months in Scotland is about 97%.

  10. David N. Brown May 27, 2010 at 00:11 #

    The link actually shows 135 “Old Order” Amish in Texas, as of 2008. There could be more, or less, now. Two major ironies:
    1. The Texas Amish subpopulation is so small that it would not have been surprising if no members were autistic.
    2. Olmsted reported one of his sources had only encountered 1 autistic in the Ohio Amish population, which is ca. 15,000.

    Finally, a thought on terms: I think, for discussion purposes, “Amish” and “Old Order Amish” can be treated as synonymous. On a “colloquial” basis, it is the latter people will think of.

  11. Michelle Allison January 28, 2019 at 19:46 #

    Here’s what I would like to know; of those 25 cases that we’re screened further, how many of them were vaccinated. That was left out of the study.

    • Sullivan (Matt Carey) February 20, 2019 at 23:23 #

      Because it has to be vaccines.

      Here’s the thing–Dan Olmsted’s work promoting the Amish as an example of “they don’t vaccinate, they don’t get autism” was junk journalism. And you and others bought it.

      But that doesn’t bother you. The question of whether these people were vaccinated bothers you.

      And that’s why you keep believing.


  1. Tweets that mention Autism Blog - Prevalence Rates of Autism Spectrum Disorders Among the Old Order Amish « Left Brain/Right Brain -- - May 17, 2010

    […] This post was mentioned on Twitter by Kev, Bioinsilico. Bioinsilico said: Prevalence Rates of Autism Spectrum Disorders Among the Old Order Amish – Left Brain/Right Brain (blog): P… #biotech […]

  2. ?Low Functioning Autism Treatment » Helping A Child With Autsim | Helping A Child With Autsim - May 25, 2012

    […] my expectations. My theory is based around my assumption that most leaders have an indisposition as to that penchant. That wasn’t highly appreciated. I don’t judge that I should like to […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: