Amish house dust protected mice from asthma, and the lesson isn’t to keep a dirtier home

Key takeaways

• In a 2016 study of 60 farm children, none of the 30 Amish children had asthma, compared with 20% of the 30 Hutterite children, though the two groups share ancestry¹².
• In mice, Amish house dust blocked allergic airway disease, but only when the animals’ innate immune sensors worked¹.
• Most human evidence shows links, not proof: farm life, a dog in the home and older siblings track with less asthma or hay fever, and C-section birth tracks with slightly more asthma⁴⁹¹¹.
• In a small Stanford-led trial, antibiotics taken around a flu shot blunted key antibody responses in adults with little prior flu immunity⁷⁸.
• Raw milk isn’t a safe way to borrow the farm effect: from 1987 to 2010, FDA counted at least 133 U.S. outbreaks tied to it, with 269 people hospitalized¹².

Amish families in Indiana still farm the old way. Each family works its own dairy farm, horses pull the plows, and children spend their days helping with the animals and playing in the barns²³.

Hutterite families in South Dakota share big communal farms instead, worked with modern industrial machinery²³.

Otherwise, the two groups’ lives look alike. They share ancestry, have large families, breastfeed, vaccinate their children and keep no pets indoors¹³. Yet the researchers who studied them report that asthma is about four times less common among Amish children, and allergic sensitization about six times less common¹.

In 2016, a team that included University of Chicago scientists took a closer look in the New England Journal of Medicine¹³. They studied the children’s immune cells, then put the house dust into mice.

That experiment sits at the center of an appealing idea: microbes teach the immune system how to behave. This is what the evidence shows, and where it stops.

Two ways of farming, two very different asthma rates

The study matched 30 Amish children from Indiana with 30 Hutterite children from South Dakota, ages 7 to 14, by age and sex².

None of the Amish children had asthma, compared with 20% of the Hutterite children, or six children by our math². At the stricter of two cutoffs, 7% of Amish children had allergy antibodies in their blood, compared with 30% of Hutterite children².

The homes differed, too. Median levels of endotoxin, a piece of the outer wall of many bacteria, were about 6.8 times higher in Amish house dust¹².

So did the children’s blood. Amish children had more neutrophils, front-line infection fighters, and fewer eosinophils, cells tied to allergic inflammation². When researchers stimulated the children’s innate immune cells in the lab, all 23 signaling molecules they measured came out lower in the Amish group². The innate layer is your fast, inborn defense.

Then came the mouse test. For four to five weeks, researchers put house-dust extracts into the noses of mice used to model allergic asthma¹². Amish dust, but not Hutterite dust, reduced the airway twitchiness and allergic cell buildup that mark the disease¹.

The protection vanished in mice bred without MyD88 and Trif, two molecules the innate immune system uses to sense microbes¹. That points to innate immune signaling as the route the dust worked through, at least in mice¹.

Keep the limits in view. The study had 60 children, dust from only 10 homes in each community, and no children under 6³. Mouse results show a possible mechanism, not what happens in a child.

The National Institutes of Health funded the work, along with the St. Vincent Foundation and the foundation of the American Academy of Allergy, Asthma & Immunology². Several authors reported industry ties².

The hygiene hypothesis began with older siblings, not soap

In 1989, British researcher David Strachan reported a pattern in a national group of children born in 1958⁴. The more older brothers and sisters a child had, the less likely the child was to have hay fever at age 11 and again at 23⁴. He proposed that infections caught from older siblings early in life might help prevent allergy⁴.

The idea became known as the hygiene hypothesis, though Strachan’s paper never used that phrase⁴. A 2022 review he coauthored traces how it changed:

• The first version was too simple. Scientists framed it as a seesaw between two kinds of helper T cells. That model faltered partly because allergies and autoimmune diseases often occur together⁴.
• “Old friends.” In 2003, the researcher G.A.W. Rook argued that the key exposures are microbes humans lived with through their evolution, not the common infections children pass around⁴.

Farm studies give the idea its strongest support. The GABRIELA study surveyed 79,888 children ages 6 to 12⁴. Hay fever ranged from 2% in farm children with more than two siblings to 12% in non-farm children with no siblings⁴. A pooled analysis of 39 studies found about 25% lower asthma rates in children exposed to farming⁴.

A messier house is another matter. The review found few studies of cleanliness itself, with only weak links⁴. In a deworming trial of 1,566 children, clearing worm infections raised allergy skin-test results but didn’t change actual allergic disease⁴.

The authors concluded that the sibling and farm effects hold up, but no specific microbe or lifestyle factor has been shown to explain them⁴. A later article in this series covers hygiene and household germs.

How gut bacteria teach immune cells to stand down

Your immune system must attack real threats and leave harmless things alone, such as food, pollen and gut bacteria. Regulatory T cells, or Tregs, help with the second job by calming immune reactions⁵.

A team led by Kenya Honda at Japan’s RIKEN research institute found that a mix of Clostridium strains from mice boosted Treg activity in germ-free mice, which are raised with no microbes at all⁶. Those mice seemed more resistant to allergies and gut inflammation⁶.

In 2013, the team reported in the journal Nature that it had taken stool from a healthy volunteer and narrowed its bacteria down to 17 Clostridia strains⁵. Given by mouth to mice, the mix raised Tregs in the colon and eased inflammation in mouse models of colitis and allergic diarrhea⁵. No single strain worked nearly as well as the group⁵.

The human piece is still an association. People with ulcerative colitis carried lower levels of these strains, though accounts differ on how many. RIKEN’s summary says all 17 were significantly lower, while The Scientist reported a statistically clear drop in only 5⁵⁶.

Two cautions apply. Clostridia include the germs behind tetanus and botulism, so the team chose strains that don’t cause disease⁶. And one coauthor was chief operating officer of Vedanta Biosciences, a company some of the authors helped launch in 2010 to develop treatments from this line of research⁶. That doesn’t make the science wrong, but independent human trials should come first.

The clearest human test: antibiotics and a flu shot

Most microbiome studies in people only show links. A 2019 Stanford-led study in the journal Cell ran an experiment instead⁷⁸.

In the 2014–15 flu season, 22 healthy adults ages 18 to 45 got a flu shot⁷. Eleven of them also took a five-day course of strong, broad-spectrum antibiotics and were vaccinated partway through⁷. Their gut bacteria counts plunged roughly 10,000-fold, and the mix of species stayed less varied for as long as a year⁷.

Even so, antibody gains at 30 days were similar in both groups⁷. These volunteers already had fairly high flu antibodies from past seasons, which may have carried them through⁷.

The next season, the team enrolled 11 people with low flu antibodies and no flu shot in at least three years, and five of them took antibiotics⁷. In those five, antibodies against the vaccine‘s H1N1 strain rose significantly less⁷⁸.

In both trials, the antibiotic takers showed more signs of inflammation⁸. Their blood levels of secondary bile acids, compounds gut bacteria make from the bile your liver produces, fell about 1,000-fold⁸.

Bali Pulendran, the senior researcher, drew a reassuring lesson: immunity stayed sturdy against flu strains people had met before, and he urged a yearly flu shot⁷. With only five people in the key group, the finding needs larger studies.

What shapes a child’s immune training

Most of this evidence comes from observational studies, which follow people without assigning them to anything. They can show links but can’t prove cause.

Birth by c-section. A 2018 analysis in PLOS Medicine pooled studies of nearly 30 million people¹¹. Children born by cesarean section had slightly higher odds of asthma. By our math from the review’s raw counts, about 3.8% of them had asthma, compared with 3.1% of children born vaginally, roughly 8 more per 1,000¹¹. The review couldn’t separate planned from emergency C-sections¹¹.

Antibiotics in the first year. In the PARSIFAL study, antibiotic use in a baby’s first year went with a higher risk of most allergic diseases⁴. A baby with a bacterial infection still needs treatment; the concern is courses that aren’t needed.

Dogs and farm animals. A Swedish registry study covered more than 1 million children born from 2001 to 2010⁹. A dog at home in the first year went with 13% lower odds of asthma at age 6⁹. About 4.2% of the children had asthma at 6, so by our math that’s roughly 5 fewer cases per 1,000 children¹⁰. Farm animals, mostly cattle and sheep, went with about half the odds⁹¹⁰. Dog exposure also went with slightly more pneumonia¹⁰.

This article covers how the main factors stack up on our four-level evidence scale.

Factor	What studies show	Evidence rating	A practical step
Farm life with animals	Tied to about 25% less asthma across 39 studies; cause unproven4	Strong evidence	No one has found a way to copy it at home
Older siblings	Tied to less hay fever, a pattern seen for decades4	Strong evidence	Nothing to change; early germ-sharing is normal
A dog in year one	13% lower odds of asthma, but more pneumonia, in one large registry study910	Promising but early	A small plus if your family wants a dog
Birth by C-section	Slightly more asthma (about 3.8% vs. 3.1%), but no proof the surgery itself is the cause11	Not enough evidence	Decide on safety for mother and baby
Antibiotics in year one	Tied to more allergic disease in one large study4	Promising but early	Ask whether an infection is likely bacterial
A less-clean home	Few studies, weak links4	Not enough evidence	Keep ordinary hygiene
Probiotics in pregnancy and infancy	Some strains may lower eczema risk, per our probiotics article	Promising but early	Plan it with a clinician

Raw milk: the farm benefit people hope for, and the risk they get

In an early 2025 survey of 1,716 U.S. adults by the Annenberg Public Policy Center, 28% thought raw milk was more nutritious, and 4% had drunk it in the past year¹⁴.

Some hope it carries the farm effect. In farm studies, children who drank unpasteurized farm milk did have less eczema and allergic sensitization⁴.

But those were observational links. FDA points out that PARSIFAL tied farm milk, not raw milk specifically, to less allergy, and that the study’s own authors advised against raw milk as prevention¹². And according to the Asthma and Allergy Foundation of America, Amish and Hutterite families both drank raw milk, yet their children’s asthma rates differed sharply³.

The risks are on record. From 1987 to September 2010, FDA counted at least 133 U.S. outbreaks tied to raw milk, with 2,659 illnesses, 269 hospitalizations, three deaths, six stillbirths and two miscarriages¹². By our math, about 1 in 10 people sickened in those outbreaks went to the hospital.

FDA says those numbers likely undercount the true toll, and they’re striking for a product that made up less than 1% of the milk Americans drank¹². In a 2005 E. coli outbreak tied to raw milk sold in Washington state, five children ages 1 to 13 were hospitalized, and four developed hemolytic uremic syndrome, a dangerous kidney complication¹². FDA names children, pregnant women, older adults and people with weakened immune systems as most at risk¹².

Bird flu added a new worry. In 2024, CDC reported that H5N1 virus had turned up in raw milk from both sick and healthy-looking dairy cows, and said the risk to people who drink it is unknown¹³. In September 2025, CDC added that no human H5N1 infection had been traced to drinking raw cow’s milk and that the risk of catching it this way is likely low, but it still advised avoiding raw milk because so much remains unknown¹⁵. Our farmers market guide covers more recent raw milk outbreaks.

Pasteurization kills these germs, including H5N1, and CDC says pasteurized milk offers most of the same nutritional benefits¹³. FDA adds that raw milk holds no probiotics; bifidobacteria in it signal fecal contamination¹². Federal law bars selling raw milk for drinking across state lines, and CDC notes that legal in your state doesn’t mean safe¹³.

What adults can do for their microbes and immune system

A few everyday steps fit the evidence, though their immune payoff in adults hasn’t been well tested.

• Feed your gut bacteria fiber and variety. Our articles on fiber and microbiome diversity give food lists and explain what diversity can and can’t tell you.
• Add fermented foods if you enjoy them. Our fermented-foods article covers a small Stanford trial that linked them with lower inflammation markers.
• Use antibiotics when they’re needed, not by reflex. The flu-shot study shows how hard broad antibiotics can hit the gut⁷. If you need one for a bacterial infection, this research is no reason to skip it. A companion article on inflammation and infection explains when antibiotics help.

How strong is the evidence?

Claim	Evidence level	Why
Farm microbes train children’s innate immunity	Promising but early	One 60-child study plus mouse tests12
Certain gut bacteria raise calming Tregs and curb allergy	Promising but early	Shown in mice; the human link is an association56
Gut bacteria shape vaccine responses in people	Promising but early	One small trial; effect seen only with low prior immunity78
Raw milk prevents asthma or allergy	Not enough evidence	The links involve farm milk; study authors advise against raw milk412
Raw milk is safe for children	Evidence says no, or unsafe	Outbreaks, hospitalizations and kidney complications in children12

Questions readers ask

The bottom line

Microbes help teach the immune system when to fight and when to stand down. The Amish-Hutterite contrast, farm surveys and mouse work point the same way, and one small trial suggests gut bacteria can shape a vaccine response. Most of the human evidence, though, shows links rather than cause, and no one has turned the farm effect into a safe recipe.

So borrow what fits. Feed your gut plenty of plants, use antibiotics when they’re truly needed, and make birth, pet and vaccine choices with your family’s safety first. Skip the shortcuts that carry real risk, raw milk above all.

Related reading

• What your gut bacteria do, given that they don’t outnumber you 10 to 1
• A diverse gut is usually a good sign, though it isn’t a report card
• Beyond the Gut, Most Probiotic Claims Outrun the Evidence. A Few Hold Up
• Sick guts hold different microbes, and proving the microbes did it is harder
• Who is most at risk of weak immunity from poor nutrition

Sources

1. Stein MM, Hrusch CL, Gozdz J, et al. “Innate Immunity and Asthma Risk in Amish and Hutterite Farm Children.” New England Journal of Medicine, 2016;375(5):411 (University of Arizona repository record). pubmed.ncbi.nlm.nih.gov
2. Frontline Medical News. “Better innate immunity in Amish vs. Hutterite children due to different allergen exposure.” MDedge, August 3, 2016. mdedge.com
3. Asthma and Allergy Foundation of America. “Does Growing Up on Amish Farms Protect Children Against Asthma?” AAFA Community blog, August 18, 2016. community.aafa.org
4. Perkin MR, Strachan DP. “The Hygiene Hypothesis for Allergy – Conception and Evolution.” Frontiers in Allergy, 2022;3:1051368 (St George’s, University of London open access copy). openaccess.sgul.ac.uk
5. RIKEN. Research highlight on Atarashi K, et al., “Treg Induction by a Rationally Selected Mixture of Clostridia Strains From the Human Microbiota,” Nature, 2013;500:232–236. RIKEN, September 27, 2013. riken.jp
6. The Scientist. “Gut Microbes Treat Illness” (on Atarashi K, et al., Nature, 2013). The Scientist, 2013.
7. Stanford Medicine. “People’s Response to Flu Vaccine Influenced by Gut Microbes.” Stanford Medicine News, September 5, 2019. med.stanford.edu
8. Hagan T, Cortese M, Rouphael N, et al. “Antibiotics-Driven Gut Microbiome Perturbation Alters Immunity to Vaccines in Humans.” Cell, 2019;178(6):1313–1328 (Universidade Federal de Goiás repository record). repositorio.bc.ufg.br
9. Fall T, Lundholm C, Örtqvist AK, et al. “Early Exposure to Dogs and Farm Animals and the Risk of Childhood Asthma.” JAMA Pediatrics, 2015;169(11):e153219 (SciLifeLab publication record). publications-affiliated.scilifelab.se
10. Van Niel CW. Summary of Fall T, et al., JAMA Pediatrics, 2015. NEJM Journal Watch, November 2, 2015. clinician.nejm.org
11. Medical Xpress. “C-Sections Linked to Long-Term Risks and Benefits” (on Keag OE, Norman JE, Stock SJ, PLOS Medicine, 2018, doi:10.1371/journal.pmed.1002494). January 2018. m.medicalxpress.com
12. U.S. Food and Drug Administration. “Raw Milk Misconceptions and the Danger of Raw Milk Consumption.” FDA, content current as of March 5, 2024. fda.gov
13. Centers for Disease Control and Prevention. “Talking to Patients about Unpasteurized (Raw) Milk and Highly Pathogenic Avian Influenza.” CDC, content dated June 3, 2024, page republished April 2026. cdc.gov
14. Annenberg Public Policy Center of the University of Pennsylvania. “With Bird Flu in Raw Milk, Many in U.S. Still Do Not Know Risks of Consuming It.” March 14, 2025. annenbergpublicpolicycenter.org
15. Centers for Disease Control and Prevention. “Influenza A Viruses May Infect GI Tract and Cause Digestive Symptoms.” CDC, September 8, 2025. cdc.gov