The first time, I looked forward to the night feeds. The room was dark, and the only sounds were her small, quick swallows and, now and then, a sigh. She was warm and heavy against me, and her head smelled sweet, like milk. That is the half I want to keep. In the other half, my hormones swung by the hour, I was hungrier than I had ever been, and I did not sleep. She wanted to be up at night like a little monster, and spent her days as a peaceful baby.

In the Madonna Litta, painted around 1490 and credited to Leonardo da Vinci or his workshop, a mother looks down at the baby she is feeding (Figure 1). Her red dress has a slit sewn over each breast, for nursing. The baby looks out at whoever is looking in. The painting draws one arrow: milk goes from the mother into the baby.

Figure 1. One arrow, mother to baby. Madonna Litta, about 1490, attributed to Leonardo da Vinci. State Hermitage Museum, Saint Petersburg. Public domain, via Wikimedia Commons.

At 2 a.m., in another house, a six-week-old with a cold lets go of the breast with a small wet snort, breathes, then latches again. A reel on her mother’s phone says the baby’s spit is talking back to the milk. The breast learns what the baby has caught, it says, and the milk answers with antibodies made for that germ.

Her mother sends the post to the twelve mothers from her prenatal class before the feed is over. Part of the claim was measured. Part of it was a guess.

In 2012, the newborn doctor Arieh Riskin and his team in Haifa, Israel, sampled the milk of mothers whose babies were in hospital with a fever. For 31 sick babies, all three months old or younger, the milk held a median of about 5,700 germ-fighting white blood cells per milliliter. After the babies got better, it held about 2,100. The milk of mothers of 20 healthy babies showed no such change.

The main antibody in milk barely changed. Riskin’s team picked mothers by the baby’s illness, so a mother could have been sick too. The study did not show how the breast would know.

A mother’s own infection is a separate case, and I have had one. Two months after my first daughter was born, the thermometer read 39.1 °C, and I was sure breastfeeding was over. My obstetrician said to keep going. A sick mother’s body makes antibodies against the germ, and some pass into her milk, antibodies a baby can borrow.

In 2013, a team in Perth, Australia, led by Foteini Hassiotou, followed 21 mothers. White blood cells made up 0 to 2 percent of the cells in healthy milk. During an infection in the mother or the baby, they rose as high as 94 percent, then fell back after recovery. In three of those infections, only the baby was sick. The team offered a route: at the end of each wave of milk in a feed, some flows back into the breast, “likely together with saliva from the baby’s mouth.” They did not measure that backflow (Figure 2).

▶

Figure 2. What milk carries and what science has yet to learn about it, from the evolutionary biologist Katie Hinde at TEDWomen 2016. Video and thumbnail: TED, CC BY-NC-ND 4.0.

Say you are one of the twelve. You wake at six to a baby who coughed all night, and the reel is the first thing on your screen. None of you can tell which half was measured. Without a baby, the post is a cold remedy a cousin sends to the family chat. Whichever half you trust is the half you act on at the next fever.

Part three left one question open: does anything travel back, from baby to mother? A trip back would close a loop: what systems thinking calls a chain of effects that comes back to change where it started. The reel claims a whole one.

A guess like Hassiotou’s can be checked: if spit flows back into the breast, bacteria from the baby’s mouth should turn up in the milk. In 2025, Pamela Ferretti, Ran Blekhman and colleagues studied 195 mothers and babies in Minneapolis. They asked which bacteria live in milk and which join the crowd in a baby’s gut. Mouth bacteria turned up in the milk anyway. Which way they had traveled was a harder question.

What a Night Feed Leaves in the Milk

In Minneapolis, a mother feeds her baby until it is full. Two hours later, she empties her right breast with a hospital pump. Within twenty minutes, the milk is in a freezer at minus 80 °C.

In all, 195 pairs did this. Mothers gave milk at one month and three months. Stool came from a diaper at one month and six months.

Milk is hard to search for bacteria. Those counted in part one are real, but next to stool they are few. Milk is mostly fat and the mother’s own cells, and a pea-sized scoop of stool is packed with the gut’s crowd.

Most earlier milk studies used 16S sequencing. It copies one gene that every bacterium carries, starting from a primer, a short single-stranded piece of DNA. A primer works like a key cut for one lock, the same trouble as a DNA kit that looks for only one key. Unlike a real lock, a poor fit still makes some copies, only fewer. The 16S studies mostly found Staphylococcus and Streptococcus, bacteria common on skin and in the mouth.

Ferretti’s team used shotgun sequencing, which reads all the DNA in a sample. In their milk, the most common bacterium was Bifidobacterium longum, a gut bacterium of breastfed babies, in 55 percent of the milk samples and 98 percent of the babies’ stool. For the gap, the team points first to the primer: the common ones were not built to fit Bifidobacterium.

A parent who reads that milk is mostly Staphylococcus may be reading the older method. In the next article about milk, check whether the team read one gene (“16S”) or all the DNA (“shotgun”).

One kind of B. longum, B. infantis, can live on the milk sugars a baby cannot digest: food sent ahead for bacteria. Here, the milk carried the bacteria too. Food and seed arrived together, and whether the seed takes is a question for the gut.

Who Stays in the Gut

A gardener tips a packet of mixed seed into one bed. In the packet, a dozen kinds sit together. In the soil, they compete, and one kind can take over.

A baby’s gut works like that bed. In the milk, three kinds of Bifidobacterium often lived side by side (Figure 3). In the babies’ guts, one of them, or something else entirely, usually took over. The team reads this as “a higher level of competition” in the gut than in the milk. Unlike a garden bed, a baby’s gut is planted by whatever arrives.

Figure 3. Seed in a dish: colonies of Bifidobacterium bifidum, one of the three kinds that shared the milk, after two to three days on agar. Agar is a jelly that bacteria grow on. Photo: Dr. Lucille K. Georg, CDC Public Health Image Library (ID 20527). Public domain, via Wikimedia Commons.

The team sorted each stool sample by its most common bacterium. At one month, about two in three were led by bacteria outside Bifidobacterium. By six months, under one in three were. The share led by B. longum rose from about one in six to almost half. Where B. infantis turned up, its share of the gut bacteria rose from about 3 percent to about 24.

Between one and six months, B. longum and B. breve grew in babies fed breast milk alone, and Clostridium perfringens, a cause of food poisoning, shrank. B. longum grew in babies who also had formula, a line for the parent who gives a bottle at night.

Across all 195 pairs, the amount of Bifidobacterium in a mother’s milk did not predict the amount in her baby’s gut, though in some pairs it did. Henri Tissier, in 1899, gave the milk the credit for what grew in breastfed babies. Here, how much seed the milk carried did not settle it.

The team names what else might decide: the crowd already in the gut, the baby’s own genes, and what there is to eat. Food is one of the checks part two described. The study tested none of the three. The seed fits the systems thinker Russell Ackoff’s test of a system: a part’s effect depends on the other parts. One seed lands differently in a different bed.

And the same species name in milk and in stool does not mean the same bacterium.

Which Way the Feed Ran

A family walks into a shelter calling for the beagle they lost last week. A dog barks from every pen. They walk past the shepherds and the terriers to the one beagle, then check for the torn left ear.

The species is the dog, and a strain is the beagle with the torn ear. Almost every baby had B. longum, so finding it in a mother’s milk and her baby’s stool is finding a dog in every pen. The same strain in both, torn ear and all, says much more. Unlike a dog, a strain is millions of nearly identical cells.

The team rebuilt the DNA of 77 strains, only 12 of them from milk, which held too little bacterial DNA to rebuild more. Twelve times, one strain turned up in a mother’s milk and in her own baby’s stool, across six pairs (Figure 4). Every strain rebuilt from milk turned up in that mother’s baby, so twelve is likely too low.

Figure 4. Six of the twelve matches, in four species. In each tree, an orange dot (a mother’s milk) lands on the same branch as purple dots (her own baby’s stool). Gray boxes mark strains that stayed in one baby from one month to six. From Ferretti et al. (2025), Fig. 3D, CC BY 4.0; cropped, nearby panels removed.

One match was Klebsiella pneumoniae, a gut bacterium that is often harmless but has been linked to sepsis, a dangerous blood infection, in newborns. Part three called this kind of microbe harmful only under the right conditions. None of the babies showed signs of illness, and no mother reported a breast infection.

Two other matches, Streptococcus salivarius and Rothia mucilaginosa, usually live in the mouth. They are what a baby’s spit would leave in the milk if Hassiotou’s backflow is real. So the arrow may run both ways: milk may seed the baby’s gut, and the baby’s mouth may seed the milk. If so, milk is half of a loop.

Which way came first is still unknown: the first milk samples came at one month, too late to see. A mother and a baby also share a home, and housemates share strains too. Mother and baby could each have picked up the strain elsewhere.

Strains also come and go: most in a baby’s stool at one month were gone by six. Babies born vaginally kept more of their strains than babies born by C-section.

For the mother who forwarded the reel, the study tests the route, not the message. Ask the reel who has seen spit bring news of the cold. A strain that travels can also carry more than itself.

Who Shares More Than Strangers

At one month, the babies’ stool carried more genes for resisting tetracycline than for any other antibiotic. So did the milk. Doctors avoid tetracycline in children under eight, because it stains growing teeth.

These genes let a bacterium survive an antibiotic, and they can pass between bacteria. Two thirds of the babies had taken no antibiotics by six months, and many with none on record still carried the genes. Babies whose gut was led by Bifidobacterium carried fewer.

Across all the pairs, milk and stool showed no link in resistance genes. But a pattern across a group can hide what happens inside each pair. So the team asked: does a baby share more resistance genes with its own mother than with a stranger?

Their test works like a daycare at pickup time with every baby handed to a random mother. Count the genes each baby shares with the mother holding it, then shuffle and count again, 1,000 times, to see what chance alone produces. Statisticians call this a permutation test (Figure 5). Unlike a real daycare, the team shuffled the milk samples’ labels on a computer, and no baby left its mother.

The real pairs shared more. A result as large as theirs turned up in fewer than 16 of the 1,000 shuffles.

Figure 5. The shuffle, done by hand. The statistician John Rauser asks whether mosquitoes prefer beer drinkers, shuffles the beer and water labels again and again, and finds the real difference out past almost every shuffle. Video: O’Reilly, Strata + Hadoop World 2014, Standard YouTube License.

The two pairs that shared the most genes also had matching strains. Their samples were among the richest in bacterial DNA, and more DNA means more chances to find a match.

The shuffle question works on more than genes. When spit and milk share a strain, or two people in one home share a cough or a claim, ask whether a stranger would share it too. If no one asks, the reel’s guess travels on beside its measured half.

Spit may reach the milk, then, but none of the studies here watched the spit tell the milk about a cold. At six o’clock the shuffle ends, and each baby goes home in the arms that fed it.

A Closing Invitation. The reel stood for a message from baby back to mother: a guess. The bacteria were real. Milk carries bacteria to a baby, a baby’s mouth may carry them back, and the crowd in the gut decides who stays.

1. Find the second path. This week, when a baby feeds near you or a toddler hands you a wet, half-eaten cracker, watch the trade both ways: milk in and spit back, a spoon from one mouth to the other. What just moved, and which way? Which cold last came home that way?
2. Split the reel. Before your next forward, or with the last one you sent, say aloud which half of the post was measured and which a guess. The milk reel: white cells measured, the route guessed. A teething tip splits too. Which half would you still send?
3. Bring one claim to the exam room. At any visit this month, the baby’s or your own, with the paper on the exam table crinkling, ask the doctor about one claim from the group chat: was it measured, and in whom? One claim checked now is one less guess at the next fever.

At the next 2 a.m. feed, a baby lets go of the breast to breathe, and a little of her spit may slip back with the milk. The milk does not read it. It carries it home.

Worth remembering Save as image

Sharing is not sending. A baby’s gut can hold the same bacteria as its mother’s milk, and both may have picked them up elsewhere. So when two people share a germ, a habit or a belief, ask what strangers share.

Where This Came From

Ferretti’s study is one of the first to match strains between milk and a baby’s gut. The team found only one earlier metagenomic study that did. Matching strains between a mother’s gut and her baby’s gut is older and better studied, and it finds far more sharing than milk does.

Intellectual Honesty Note. The births, the night feeds and the fever night, with its 39.1 °C reading, are my own, and the reading of the Madonna Litta is this piece’s. The mother at 2 a.m., the reel, its group chat and the twelve mothers are a composite; the reel paraphrases a common claim, not one post. Riskin’s abstract does not report the mothers’ own health. The mother at the pump is drawn from the study’s methods. The key and lock, the seed bed, the beagle and the daycare are illustrations. The formula line reports the study’s comparison, not advice on feeding. The study is observational: it cannot show that milk caused any gut change. Only 25 milk samples were sequenced at three months. The primer explanation is the team’s proposal, not a test. The 16 in 1,000 reads the paper’s bound (p < 0.016, from 1,000 shuffles) as a count.

References

Centers for Disease Control and Prevention. (2025). Influenza (flu) and breastfeeding. https://www.cdc.gov/breastfeeding-special-circumstances/hcp/illnesses-conditions/flu.html

Ferretti, P., Allert, M., Johnson, K. E., Rossi, M., Heisel, T., Gonia, S., Knights, D., Fields, D. A., Albert, F. W., Demerath, E. W., Gale, C. A., & Blekhman, R. (2025). Assembly of the infant gut microbiome and resistome are linked to bacterial strains in mother’s milk. Nature Communications, 16, 11536. https://doi.org/10.1038/s41467-025-66497-y

Georg, L. K. (n.d.). Bifidobacterium bifidum colonies on brain heart infusion agar [Photograph]. Centers for Disease Control and Prevention, Public Health Image Library, ID 20527. Public domain image via Wikimedia Commons. https://commons.wikimedia.org/wiki/File:Bifidobacterium_bifidum_CDC_20527.jpg

Harper, D. (n.d.). Antibody. In Online etymology dictionary. Retrieved October 6, 2026, from https://www.etymonline.com/word/antibody

Hassiotou, F., Hepworth, A. R., Metzger, P., Tat Lai, C., Trengove, N., Hartmann, P. E., & Filgueira, L. (2013). Maternal and infant infections stimulate a rapid leukocyte response in breastmilk. Clinical & Translational Immunology, 2(4), e3. https://doi.org/10.1038/cti.2013.1

Hinde, K. (2017, March 28). What we don’t know about mother’s milk [Video]. TED. YouTube. https://www.youtube.com/watch?v=Bo8YN3oB0Cw

Leonardo da Vinci (attributed). (ca. 1490). Madonna Litta [Painting]. State Hermitage Museum, Saint Petersburg. Public domain image via Wikimedia Commons.

National Cancer Institute. (n.d.). Organism strain. In NCI dictionary of cancer terms. Retrieved October 6, 2026, from https://www.cancer.gov/publications/dictionaries/cancer-terms/def/organism-strain

National Human Genome Research Institute. (n.d.). Antibody. In Talking glossary of genomic and genetic terms. Retrieved October 6, 2026, from https://www.genome.gov/genetics-glossary/Antibody

National Human Genome Research Institute. (n.d.). Primer. In Talking glossary of genomic and genetic terms. Retrieved October 6, 2026, from https://www.genome.gov/genetics-glossary/Primer

Rauser, J. (2014, October 17). Statistics without the agonizing pain [Video]. O’Reilly. YouTube. https://www.youtube.com/watch?v=5Dnw46eC-0o

Riskin, A., Almog, M., Peri, R., Halasz, K., Srugo, I., & Kessel, A. (2012). Changes in immunomodulatory constituents of human milk in response to active infection in the nursing infant. Pediatric Research, 71(2), 220–225. https://doi.org/10.1038/pr.2011.34

Bohnhoff’s team fed the mice Salmonella, the germ behind food poisoning. For healthy mice, it took about 100,000 germs to give 5 out of every 10 a gut infection. The mice that took fewer than three had been given one dose of an antibiotic, a drug that kills bacteria, the day before. The drug had killed many of the bacteria in their guts, and those bacteria had been keeping newcomers out, like guards at a door. So the ten-year-old, who asked for one number, gets two. Which one counts depends on who already lives there.

▶

Figure 1. The gut’s trillions of microbes and what they do for health, from BBC Global’s Health Decoded (2024), with science journalist Melissa Hogenboom. Video and thumbnail: BBC, Standard YouTube License.

The skin has guards of its own. In 2010, Tadayuki Iwase and colleagues in Tokyo studied Staphylococcus aureus, a bacterium behind skin and blood infections. Some strains of a common skin bacterium, they found, make a protein that breaks up the sticky layer S. aureus builds to cling on. Of 88 volunteers, those who carried those strains in their noses rarely carried S. aureus. When the team put such a strain into the noses of volunteers who did, the S. aureus was gone. Bohnhoff took the guards away, and the door opened. Iwase added one, and it shut.

▶

Figure 2. How the bacteria on skin make it a poor home for invaders, help hold the skin together and train the immune system, from Nature Video (2020). Video and thumbnail: Springer Nature, Standard YouTube License.

Say you stand at the pharmacy counter this month with a prescription for your child’s earache: an antibiotic, twice a day for ten days. The paper bag crinkles as the pharmacist folds it shut. Some ear infections get better without antibiotics. Is this one? The drug cannot tell the germ it was sent for from the crowd in your child’s gut. Ten days leave your child’s gut with fewer guards at its door until the crowd grows back.

Bohnhoff’s mice and Iwase’s volunteers show one idea from a 2024 review by Ziqi Ma and colleagues. The microbes defend the body, and they teach the immune system what to fight. The series’ first two parts followed five of the review’s ideas.

The review names two more. The microbes may help set the targets a body holds steady. And when the body and its microbes stop holding each other in check, the body tips toward illness. Both are loops, the kind systems thinking follows: each side acts on the other, and the other acts back.

A body holds some numbers steady, its temperature among them, the way an oven holds the heat it is set to. In a fever, the brain turns its own dial up by one or two degrees Celsius. The target moves first, and the body shivers until its temperature catches up.

Ma’s team argues that the microbes keep a second hand on the same dial. If ten days of pills can thin the guards in a child’s gut, how much of the turning were those guards doing?

What Keeps the Door Shut

In 1958, four years after Bohnhoff’s mice, surgeons in Denver ran the experiment in reverse. Ben Eiseman’s team gave four patients with a severe gut infection a healthy donor’s stool. They called the response “immediate and dramatic.”

The infection is now usually traced to Clostridioides difficile, which thrives once antibiotics have thinned its competition. In 2013, Els van Nood, Josbert Keller and colleagues in Amsterdam tested a donor’s stool against the standard antibiotic, vancomycin, in 42 patients whose infection kept returning. Thirteen of the sixteen patients in the donor group were cured by the first infusion. On vancomycin alone, four of thirteen were. The gap was so wide that the trial was stopped early. It was a small trial, but a stranger’s microbes had taken the space back.

Resident microbes keep newcomers out in several ways (Figure 3). They eat the food a newcomer needs. Some make proteins aimed at rivals, like Iwase’s nose strains. Some change the place itself: in the vagina, Lactobacillus bacteria make lactic acid, which keeps it too acidic for many germs.

Figure 3. The guards, up close: a crowd of Staphylococcus epidermidis, the skin bacterium whose strains cleared S. aureus from volunteers’ noses in Iwase’s study. Colorized image: National Institute of Allergy and Infectious Diseases (2011), via the CDC Public Health Image Library (PHIL 18135), public domain.

The microbes also teach. The Amish of Indiana and the Hutterites of South Dakota both came to America from Europe as German-speaking farmers, and their genes are close. They farm differently. Amish families keep dairy cows and work fields with horses, the barns a few steps from the house. Hutterites run large communal farms with modern machines. Michelle Stein, Carole Ober and colleagues found asthma in about 5 percent of Amish children, against 21 percent of Hutterite children. The dust in Amish homes held almost seven times as much endotoxin, a molecule from the outer coat of many bacteria.

In the airways of mice, Amish house dust protected against asthma attacks; Hutterite dust did not. In 60 children, 30 from each group, the Amish children’s immune cells leaned more toward fighting germs and less toward allergy. Two communities are not a trial. The mouse experiments are what point to the microbes in the dust.

By that definition, the immune system is a system in the systems scientist Donella Meadows’ sense, with parts, links and a purpose. The Amish dust suggests the microbes belong in it too, tuning the immune cells.

Medicine already counts several ways to gain protection after birth. A body can fight off an infection and remember it. A vaccine can teach it. A baby can take antibodies from milk; whether anything travels back the other way is the last part of this series.

Ma’s team adds the microbes to that list and calls their protection acquired microbial immunity (Figure 4). It is gained by living among the right microbes, as the Amish children did, or by medicine, as the Amsterdam patients did. It can be lost to one dose of the wrong drug.

Figure 4. Ways to gain protection after birth, with Ma’s addition in the bottom row. Adapted from Ma et al. (2024), Fig. 2, CC BY 4.0; simplified and redrawn, leaving out milk sugars and probiotic products.

That dose often comes home in a paper bag. Anyone handed an antibiotic for a cough, an earache or a sore throat can ask two questions before the first pill. Is this infection caused by bacteria? Antibiotics do nothing against a virus. Is there a narrower drug that leaves more of the gut alone? A friend with a toddler and a winter of earaches ahead may want the two questions too. Skipping them costs nothing that day; the cost comes later, in a gut with fewer guards. Ma’s team claims those guards also help set the body’s targets, starting with its temperature.

Who Set the Dial

In 1868, the German physician Carl Wunderlich published the temperatures of about 25,000 patients. He took them under the arm, with a foot-long thermometer that needed more than fifteen minutes a reading. He put normal at 37 °C.

In 2020, Myroslava Protsiv, Julie Parsonnet and colleagues at Stanford checked Wunderlich’s number against the records of about 189,000 people. The records ran from Civil War veterans to Stanford patients seen in 2017. Body temperature had fallen about 0.03 °C for every decade of birth. Like the fever, the target had moved. This move took two centuries, and the textbooks kept saying 37.

In 2023, Kale Bongers, Robert Dickson and colleagues at the University of Michigan compared about 30 mice raised without microbes with as many raised with them. The germ-free mice ran about 0.3 °C colder, and mice whose gut bacteria were wiped out with antibiotics ran colder too. In 116 patients in hospital, the mix of gut bacteria on arrival predicted how each patient’s temperature would change. A link like that does not prove a cause.

A body holds its temperature with a balancing loop, a loop that pushes back whenever a reading drifts from its target. Once the brain raises its set point, the temperature the body aims for, the body is for a while colder than its own new target. As on a cold night, it pulls blood from the skin, and the hands go cold. The thermometer climbs while the person under the blanket feels freezing.

Holding numbers steady has a name, homeostasis, from the physiologist Walter Cannon. The loop does the holding. An oven answers to one hand. A body holds many targets at once. In Ma’s view, the microbes are a second hand choosing where the loop holds, a role the team calls homeostatic reprogramming.

On a lab report, the range beside each number comes from many other people. At the next checkup, a useful question is not only “Is this in range?” but “Is this where I usually am?” Last year’s report, held next to this year’s, answers it in a minute, and a partner’s reports work the same way. Skip the comparison, and a number can drift for years inside the printed range before anyone looks.

My daughters have had no antibiotics so far. When my first daughter ran a fever at about 18 months, she would not touch her solid food and nursed mostly for comfort. The only medicine she got was acetaminophen. I took her temperature. It moved from 39.4 °C to 37.8 °C. The line the pediatrician’s office gave me was 38 °C, 100.4 °F: no higher. That line is the same for every child. It says nothing about where one child’s own normal sits.

If the microbes help choose where the loop holds, could thinner gut communities explain part of the drop since Wunderlich? Ma’s team offers the idea as a hypothesis. Protsiv’s team pointed to fewer chronic infections. Neither team tested the other’s.

Allan Konopka, who studies microbial communities, warns that biologists tend to expect a “balance of nature,” a steady state every system returns to. A normal that drifts over two centuries is easy to miss. So is one that two sides hold at once, like the low-oxygen layer by the gut wall in part two, which shifts with every meal.

Whose Hand Is on the Dial

Some gut bacteria turn part of a meal of eggs into a gas that smells like fish. The smell is the mild part. The part those bacteria use is choline, a nutrient in the yolk. The liver turns that gas into TMAO, a molecule the blood then carries.

In 2016, Weifei Zhu, Stanley Hazen and colleagues at the Cleveland Clinic followed more than 4,000 heart patients. Over three years, the top quarter for TMAO had 1.64 times the bottom quarter’s risk of a heart attack or stroke, after allowing for other risk factors. In the lab, TMAO made platelets, the blood cells that start a clot, quicker to react. The study measured TMAO in the blood and did not track diet, so it does not by itself make a case against eggs.

A bowl of oats or beans carries fiber for the gut’s bacteria. Some of that fiber ends as butyrate, the fatty acid the cells of the colon wall run on. One gut can make TMAO and butyrate at once. Which one wins depends on the meal and on which bacteria are there to meet it.

Ma’s last idea is one model of health and illness. The body needs four things from its microbes, and each can fail (Figure 5). It needs enough contact with microbes, kept in check, neither none at all nor an overgrowth. It needs a gut wall that holds, so bacteria stay out of the blood. It needs more helpful products than harmful ones, butyrate over TMAO. And it needs steady control as the microbes change, because bacteria mutate.

Figure 5. Four things the body needs from its microbes, and how each one fails. Adapted from Ma et al. (2024), Fig. 7, CC BY 4.0; simplified and redrawn.

In 2022, Yi Yang, Noah Palm and colleagues at Yale put one strain of Enterococcus gallinarum, a gut bacterium, into eight germ-free mice. Within three months, it had picked up 159 new mutations and split into two lines. One kept to the open gut. The other adapted to the gut lining, slipped past the immune system, and spread to the lymph nodes and the liver, where it caused more inflammation. A person’s gut is crowded with rivals; these mice had none. “In a rigorous sense, all human microbes exhibit conditional pathogenicity,” Ma’s team writes, ruling out any simple list of good and bad bacteria. Any of them can cause harm under the right conditions. What a microbe does depends on what it is connected to.

When all four needs hold, the body stays in charge, and the microbes’ hand on the dial helps it. When they fail, the microbes take over, and the body drifts toward illness. The model leaves open which side pushed first.

In an ordinary week, an antibiotic course thins the contact. A week of white bread and no beans leaves the butyrate makers hungry, and the wall’s cells with them. The snack within reach right now has a fiber line on its label. A pot of lentils simmering on a weeknight stove feeds everyone at the table, and the bacteria that feed the wall.

A Closing Invitation. Three germs could infect Bohnhoff’s mice because the door their microbes held had been opened. The microbes’ hand is on the body’s dial too: helping set a temperature, and deciding which way a meal runs.

1. Find your own normal. Tomorrow morning, before coffee, start a sticky note inside the medicine cabinet door with your temperature and a partner’s or child’s. The cool tip goes under the tongue, then the beep. Is it 37, or lower, like most people’s now? Was the last fever in your house read against 37?
2. Check the cart. At the next grocery run, before the checkout belt starts to move, look over what you picked: beans, oats, lentils, a whole-grain loaf? Which dinner this week has no fiber in it at all? Which can or bag could you add for that night?
3. Ask what number means call. Message the doctor’s office now, or ask at the next checkup: which temperature means call? If your normal runs below 37, does the line move with it? Then the next shivering night can be what it is: a body chasing a target it moved.

At the counter, ask the ten-year-old’s question the other way round: how many guards will ten days of pills leave? The label on the bag names one germ, not the crowd that helps hold the door and the dial.

Worth remembering Save as image

Good guards do more than guard. The microbes you carry keep germs out, and they may help set what your body treats as normal. So before you change your pills or your plate, ask what else the crowd was holding.

Where This Came From

Ma’s homeostatic reprogramming builds on Claude Bernard, who wrote in the 1800s of the milieu intérieur, the inner fluid a body keeps steady around its cells. Ma’s team adds the microbes as a fourth force beside the nervous, immune and metabolic systems, and asks whether “steady” was ever the right word.

Intellectual Honesty Note. The ten-year-old in the opening and the pharmacy counter scene are hypotheticals. My first daughter’s fever and the 38 °C line are my own. The fall in average body temperature is read here as a lower target; Protsiv’s team measured temperatures, not set points. Acquired microbial immunity, homeostatic reprogramming and the four needs are proposals in Ma’s review; the team calls the second a hypothesis. The hand on the dial and the oven are devices: a body’s set points are held by loops of nerves, hormones and cells, not turned by a hand. Bohnhoff’s team used a Salmonella strain resistant to streptomycin; Ma dates the study 1967, and the 1954 paper is the original.

References

Bohnhoff, M., Drake, B. L., & Miller, C. P. (1954). Effect of streptomycin on susceptibility of intestinal tract to experimental Salmonella infection. Proceedings of the Society for Experimental Biology and Medicine, 86(1), 132–137.

Bongers, K. S., Chanderraj, R., Woods, R. J., McDonald, R. A., Adame, M. D., Falkowski, N. R., Brown, C. A., Baker, J. M., Winner, K. M., Fergle, D. J., Hinkle, K. J., Standke, A. K., Vendrov, K. C., Young, V. B., Stringer, K. A., Sjoding, M. W., & Dickson, R. P. (2023). The gut microbiome modulates body temperature both in sepsis and health. American Journal of Respiratory and Critical Care Medicine, 207(8), 1030–1041.

Bryce, E. (2018, January 8). How does your immune system work? [Video]. TED-Ed. YouTube. https://www.youtube.com/watch?v=PSRJfaAYkW4

Cannon, W. B. (1929). Organization for physiological homeostasis. Physiological Reviews, 9(3), 399–431.

Centers for Disease Control and Prevention. (2025). Healthy habits: Antibiotic do’s and don’ts. https://www.cdc.gov/antibiotic-use/about/index.html

Eiseman, B., Silen, W., Bascom, G. S., & Kauvar, A. J. (1958). Fecal enema as an adjunct in the treatment of pseudomembranous enterocolitis. Surgery, 44(5), 854–859.

Hogenboom, M. (2024, October 8). How your gut microbes shape your health [Video]. BBC Global. YouTube. https://www.youtube.com/watch?v=aEhrcIZE1ZM

Iwase, T., Uehara, Y., Shinji, H., Tajima, A., Seo, H., Takada, K., Agata, T., & Mizunoe, Y. (2010). Staphylococcus epidermidis Esp inhibits Staphylococcus aureus biofilm formation and nasal colonization. Nature, 465(7296), 346–349. https://doi.org/10.1038/nature09074

Konopka, A. (2026, May 6). Systems thinking [Substack post]. https://thinkmicrobe.substack.com/p/systems-thinking

Ma, Z., Zuo, T., Frey, N., & Rangrez, A. Y. (2024). A systematic framework for understanding the microbiome in human health and disease: From basic principles to clinical translation. Signal Transduction and Targeted Therapy, 9, 237. https://doi.org/10.1038/s41392-024-01946-6

Meadows, D. H. (2008). Thinking in systems: A primer (D. Wright, Ed.). Chelsea Green.

National Cancer Institute. (n.d.-a). Homeostasis. In NCI dictionary of cancer terms. Retrieved October 6, 2026, from https://www.cancer.gov/publications/dictionaries/cancer-terms/def/homeostasis

National Cancer Institute. (n.d.-b). Immune system. In NCI dictionary of cancer terms. Retrieved October 6, 2026, from https://www.cancer.gov/publications/dictionaries/cancer-terms/def/immune-system

National Institute of Allergy and Infectious Diseases. (2011). Staphylococcus epidermidis [Digitally colorized scanning electron micrograph, PHIL 18135]. Centers for Disease Control and Prevention, Public Health Image Library. https://wwwn.cdc.gov/phil/Details.aspx?pid=18135

Nature Video. (2020, December 16). The skin microbiome: A healthy bacterial balance [Video]. YouTube. https://www.youtube.com/watch?v=MWE3U3FItlc

Protsiv, M., Ley, C., Lankester, J., Hastie, T., & Parsonnet, J. (2020). Decreasing human body temperature in the United States since the Industrial Revolution. eLife, 9, e49555.

Stein, M. M., Hrusch, C. L., Gozdz, J., Igartua, C., Pivniouk, V., Murray, S. E., Ledford, J. G., Marques dos Santos, M., Anderson, R. L., Metwali, N., Neilson, J. W., Maier, R. M., Gilbert, J. A., Holbreich, M., Thorne, P. S., Martinez, F. D., von Mutius, E., Vercelli, D., Ober, C., & Sperling, A. I. (2016). Innate immunity and asthma risk in Amish and Hutterite farm children. New England Journal of Medicine, 375(5), 411–421.

van Nood, E., Vrieze, A., Nieuwdorp, M., Fuentes, S., Zoetendal, E. G., de Vos, W. M., Visser, C. E., Kuijper, E. J., Bartelsman, J. F., Tijssen, J. G., Speelman, P., Dijkgraaf, M. G., & Keller, J. J. (2013). Duodenal infusion of donor feces for recurrent Clostridium difficile. New England Journal of Medicine, 368(5), 407–415.

Walter, E. J., Hanna-Jumma, S., Carraretto, M., & Forni, L. (2016). The pathophysiological basis and consequences of fever. Critical Care, 20, 200. https://doi.org/10.1186/s13054-016-1375-5

Wunderlich, C. A. (1868). Das Verhalten der Eigenwärme in Krankheiten. Otto Wigand.

Yang, Y., Nguyen, M., Khetrapal, V., Sonnert, N. D., Martin, A. L., Chen, H., Kriegel, M. A., & Palm, N. W. (2022). Within-host evolution of a gut pathobiont facilitates liver translocation. Nature, 607(7919), 563–570.

Zhu, W., Gregory, J. C., Org, E., Buffa, J. A., Gupta, N., Wang, Z., Li, L., Fu, X., Wu, Y., Mehrabian, M., Sartor, R. B., McIntyre, T. M., Silverstein, R. L., Tang, W. H. W., DiDonato, J. A., Brown, J. M., Lusis, A. J., & Hazen, S. L. (2016). Gut microbial metabolite TMAO enhances platelet hyperreactivity and thrombosis risk. Cell, 165(1), 111–124.