Mind & Body · Saturday, 29 August 2026

01 Briefing what happened

Nothing in your gut is defending you. The protection is that there is no room.

Mind & Body 1 min 49 sources

Most infections fail before your immune system hears about them, because the residents already hold the food and the landing sites. A study of 14,979 adults now shows how long one course of antibiotics can keep those seats empty.

350

gut species in people who took no antibiotics

measured across eight years in 14,979 Swedish adults [1][2]

fewer species after one clindamycin course

counted a year later, against people who took none [2]

10x

higher C. diff risk around antibiotics

while taking them and for the month afterwards [6]

1 in 5

get diarrhoea from a course of antibiotics

mostly mild; the severe form is usually C. diff [7]

The lead story — what happened

Your gut, skin, nose and vagina are already fully occupied. Residents hold the food and the landing sites, so most arrivals never get a foothold. [3][4][10][13][17]
Scientists call it colonization resistance. The idea dates to the 1950s, when Marjorie Bohnhoff proposed that an antibiotic could leave a host more open to infection. [4]
The mechanism is competition, not defence. Early colonisers use up the growth-limiting resources, which holds back later arrivals that need the same things. [3][5]
They also make the place hostile. Fibre fermented into short-chain fats acidifies an invader from the inside, and bile acids suppress others. [3][4]
The gut is not one habitat but many. In mice, oxygen falls from 58 mmHg in the stomach to about 3 mmHg in the far colon, and each zone suits different residents. [3]
The lining helps by being restless. It renews itself every four to seven days, and a mucus layer several hundred microns thick keeps most bacteria off the cells. [3]
A course of antibiotics empties seats. Among 14,979 Swedish adults, those who had taken none in eight years carried about 350 gut species; users carried fewer. [1][2]
Clindamycin did the most damage. Each course in the previous year was linked to 47 fewer species detected, and shifts in nearly 300 of the 1,340 species measured. [2]
The mark lasts. Species abundances were still altered in people whose single course came four to eight years before the sample. [1]
Diversity rebounds fastest in the first two years and then flattens. The study's senior author, Tove Fall, put it plainly: it seems like you do not recover completely. [2]
Empty seats are what Clostridioides difficile needs. People are up to 10 times more likely to get a C. diff infection while on an antibiotic and in the month after. [6]
About 1 in 5 people who take antibiotics get diarrhoea from them. The severe form is usually C. diff inflaming the large intestine. [7][8]
The exposure is ordinary; the opening is not. C. diff waits outside as spores that survive months or years on surfaces, and finding them in a home is not unusual. [6]
Carrying C. diff without being ill is more common than the infection itself. Those people test positive and often have no symptoms at all. [46]
Refilling the room is now the treatment. Stool transplants work by restoring secondary bile acids and taking the free nutrients back off the pathogen. [9]

What is pushing on this

How full the room is Steady↓

antibiotic users carried fewer species than non-users across eight years [1]

Damage from clearing Building↑

clindamycin shifted nearly 300 of 1,340 species measured [2]

How invisible it is High→

commensals block arrivals by holding the niche and eating the nutrients [4]

Ability to refill Easing↑

fibre feeds the natives back, while added strains can slow it [26]

Who is involved

Resident bacteria — hold the food, the surfaces and the chemistry Clostridioides difficile — waits outside as a spore for seats to open The Swedish cohorts — 14,979 adults whose stool was matched to their prescriptions [1] Stool-derived products — two are now approved in the US for recurrent C. diff [23]

How it unfolded

1950s Marjorie Bohnhoff proposes that an antibiotic opens a host to infection [4]
2022-23 the FDA approves the first two stool-derived therapies [23]
Dec 2024 OpenBiome halts shipped frozen transplant preparations [23]
Mar 2026 Nature Medicine reports the eight-year mark in 14,979 adults [1]

Where this points

The next test is whether the damage a drug does to the resident community starts being weighed when it is prescribed, rather than only counted after an infection follows it. [34]

The rest of the day

35 more stories on this beat.

Each with its own sources. None of these is a link to the story above.

02
Who holds the nose decides the staph

A study of about 1,100 people found seven types of nasal community. Persistent carriage went with more staph, and with less of three Corynebacterium species, Dolosigranulum pigrum and Staphylococcus epidermidis. [10]

Why it matters — Same rule as the gut, on a different surface: staph settles where nobody else is sitting.
03
A third of adults carry staph

One London orthopaedic hospital clears it on purpose before surgery. Patients use a nasal antibiotic ointment for five days and chlorhexidine cloths for two. [11]

Why it matters — When a knife is about to open a route inward, the harmless resident becomes the risk.
04
Staph spread between babies in intensive care

Three years of genome-level surveillance in a neonatal unit traced strains passing between infants through shared spaces and caregivers. The clusters that reached the most infants over time were also the ones associated with invasive infection. [12]

Why it matters — Being colonised comes first; the infection is what happens when the lodger moves.
05
Skin's quiet tenant gets re-read

Staphylococcus epidermidis was long filed as an opportunist that infects implants. A new review describes it as an active resident that calms inflammation and holds the skin barrier together. [13]

Why it matters — The same organism defends in one place and infects in another; the address decides which.
06
Skin bacteria rebuild the skin's own fats

Residents eat the oils your skin makes and convert them into new compounds, including short-chain fats. Those products then shape which other microbes can grow there. [14]

Why it matters — Occupancy is chemical as well as physical: they change the room, not just fill it.
07
A wound reopens the seat plan

When skin breaks, the resident community reorganises through every phase of repair. [15] S. epidermidis and Lactobacillus plantarum prompt the skin to make its own antibacterial proteins, while S. aureus and Pseudomonas slow closure with biofilm. [16]

Why it matters — A cut is a vacancy, and what fills it partly decides how fast it shuts.
08
The vagina defends itself with acid

About 95% of the resident bacteria there are lactobacilli. They make lactic acid and hydrogen peroxide, holding the pH near 3.8 to 4.2, which most newcomers cannot grow in. [17]

Why it matters — The protection is a by-product of what lactobacilli do to feed themselves.
09
Losing the acid is the illness

Bacterial vaginosis is not an invasion by one germ. It is the loss of hydrogen-peroxide-producing lactobacilli and the overgrowth of anaerobes already present. [18] The WHO names douching as a risk factor. [19]

Why it matters — Named for what arrives, caused by what left.
10
Thin defences raise infection risk

A community dominated by Lactobacillus crispatus protects against picking up chlamydia. Where those species are depleted, acquisition risk rises and the body clears the infection less well. [20]

Why it matters — The same defence and the same failure, measured against a specific infection.
11
Products aim to refill the vagina

Live biotherapeutic products are being designed to restore Lactobacillus dominance rather than only kill the overgrowth. [21] A review of fermented foods found 87% of food starter uses drew on only two of the 54 strains with clinical evidence. [47]

Why it matters — The strains that work in trials and the ones in the shops are mostly not the same strains.
12
Stool transplants held up over years

129 patients with recurrent C. diff took freeze-dried capsules at two acute-care hospitals. 103 of them, or 80%, had no recurrence at three months, and follow-up ran a median of 182 weeks. [22]

Why it matters — Refilling the room is a standard treatment now, not an experiment.
13
Two products approved, one supply closed

The FDA approved the first two stool-derived therapies in 2022 and 2023. OpenBiome stopped shipping frozen preparations on 31 December 2024, which made conventional transplants harder to obtain. [23]

Why it matters — The licensed version arrived as the improvised version got scarce.
14
Transplants tested in Parkinson's

A phase 2 trial randomised 72 people newly diagnosed and not yet on medication. At 35 weeks, motor scores improved by 3.8 points in the donor group against 0.1 in those given back their own stool. [24]

Why it matters — The claim is spreading well past C. diff on much thinner evidence.
15
Transplant tried alongside cancer drugs

In a phase 1 trial, 20 people with advanced kidney cancer took donor capsules with immunotherapy. Half had serious immune side effects, and half of those assessable responded. [25]

Why it matters — Twenty people is a safety signal, not a result.
16
Fibre, not capsules, feeds recovery

UCLA Health's physicians point to research showing probiotics after antibiotics can delay recovery. A few outside strains disturb the natives that are trying to regrow. [26]

Why it matters — Adding tenants is not the same as the room being full again.
17
Two trials tried to soften the damage

A 91-day randomised trial gave a mixed synbiotic through a week of two antibiotics, and stool butyrate rose 119% against placebo. [27] A 32-patient postbiotic trial reported 40% higher diversity, and its authors include the product company's chief executive and co-founders. [28]

Why it matters — Both are real results; only one of them is at arm's length from the seller.
18
Probiotic evidence stays thin

A meta-analysis of 34 trials in 2,390 adults found possible gains in cognitive scores, rated very low certainty. [29] Another pooled 29 trials in 1,633 older adults and measured bacteria, inflammation markers and fatty acids, not how people fared. [30]

Why it matters — Changing the roster is not the same as making someone better.
19
The regulator names who it can harm

The NIH's Office of Dietary Supplements says there are no official recommendations covering probiotic use by healthy people. It also notes the FDA has reported that probiotics might cause infections or life-threatening illness in preterm infants. [31]

Why it matters — The safety case is thinnest exactly where the marketing is warmest.
20
The gut is where resistance lives

Drug-resistant bacteria sit quietly in the gut and later cause bloodstream, surgical and urinary infections. Resistant infections were directly responsible for 1.27 million deaths in 2019. [32]

Why it matters — The reservoir is inside the patient long before the infection has a name.
21
Which residents keep resistance out

Where bacteria of the Lachnospiraceae family were abundant, multidrug-resistant Enterobacterales were less likely to settle. Among liver transplant patients, lower diversity tracked with carrying them. [33]

Why it matters — Named residents, not a vague notion of good bacteria: occupancy is specific.
22
It is not only antibiotics

A review names the problem microbiotoxicity: unintended microbiome damage caused by non-antibiotic drugs, and by consumer, dietary and agricultural products. [34]

Why it matters — The bill is being run up in places nobody is counting.
23
The gut hit shows up in the lungs

In mice, amoxicillin caused long-lasting gut disruption and a long-term drop in one arm of lung immune defence, well after the drug had stopped. [35]

Why it matters — A local clearance with a distant, delayed cost.
24
Children take the longer hit

A review of 25 studies in low- and middle-income countries found children's disruption ran deeper and lasted longer than adults'. Resistance genes persisted after the community itself had recovered. [36]

Why it matters — The room refills before the resistance leaves it.
25
Babies swap strains at nursery

Researchers tracked 134 people and 1,013 stool samples across three nurseries. Babies were passing strains to each other within a month, and antibiotic treatment was the condition that most increased the influx of new strains. [37]

Why it matters — Filling a room is a social process, not a private one.
26
Milk carries strains, not just food

507 samples from 195 mother-infant pairs showed milk and infant gut sharing strains, driven by Bifidobacterium longum. Antibiotic-resistance genes overlapped within pairs too. [38]

Why it matters — The seats are handed over, and so is what is written on them.
27
A latecomer can still get a seat

Bifidobacterium infantis given to breastfed US infants at two to four months colonised rapidly and was still there a month after the supplement stopped. [39]

Why it matters — Being first is powerful but not absolute; the right food still buys entry.
28
Feeding mode shapes the assembly

Breastfed infants tend to hold a stable community dominated by bifidobacteria. Formula-fed infants show more diversity earlier and move sooner toward an adult-like pattern. [40]

Why it matters — More diversity is not automatically better here, because stability was doing a job.
29
Weaning rewrites it again

Solid food at around six months drives shifts that are known to persist into adulthood. The weaning window has had far less study than birth. [48]

Why it matters — A second handover, and much less watched.
30
C. diff after big surgery

Among 76,083 people having liver or pancreatic surgery, 1.1% got C. diff within 30 days, and nearly half of those cases came in the first week. After liver resection it more than tripled the odds of dying. [41]

Why it matters — Antibiotics, hospital and a cut stack the openings into one week.
31
Hospitals vary on the basics

A survey of 350 French infection-control staff found 70% used contact precautions for patients with symptoms. 71% cleaned with spore-killing products, and 5% used products that do not kill spores. [42]

Why it matters — Spores survive what ordinary cleaning leaves behind.
32
Diet enters the C. diff story

A Cell Host and Microbe review argues C. diff thrives on nutrients that only become available once the community is disrupted, including some enriched by modern diets. [9]

Why it matters — Not just an empty room, but an emptied room with food left out on the table.
33
Klebsiella lives in the gut first

Reviewers describe gut carriage of Klebsiella pneumoniae as the step before pneumonia, urinary and bloodstream infection. Resident bacteria are what normally hold it down. [43]

Why it matters — Another familiar hospital pathogen that begins as a lodger.
34
Critically ill patients look different

A study compared 88 patients infected with resistant organisms, 100 merely carrying them and 86 healthy controls. Both patient groups had lost Bacteroides, Faecalibacterium and Roseburia and gained Enterococcus. [49]

Why it matters — The thinned community and the resistant organism turn up in the same people.
35
Nobody agrees which chairs matter

Methods for identifying the keystone gut bacteria that hold a community together are inconsistent between studies, which is why diet advice built on them is shaky. [45]

Why it matters — We can watch a room empty without knowing which seats were load-bearing.
36
IBS and the fatty acids

Researchers linked patterns of gut bacteria and short-chain fatty acids to different presentations of irritable bowel syndrome, in a field where earlier findings had failed to replicate. [44]

Why it matters — An honest example of evidence that is genuinely unsettled.

02 Lesson why it matters

The defence that nobody is performing

Some things are protected not by a guard but by there being no room, and whoever fills that room is defending their own dinner.

The twist

The thing keeping you safe is not on your side. It is holding its own dinner, and your protection is what falls off the edge of the table.

How it works

Every surface of you is already occupied
Residents hold the food and the landing sites
Their waste makes the place hostile to arrivals
A newcomer finds no meal, no seat, wrong chemistry
Most invasions fail with nothing firing
Clear the residents and the defence leaves unnoticed

Where you've seen this

A full car park

nobody guards it; there is simply nowhere left to stop

A lawn

grass keeps weeds out by taking the light first, not by fighting

A busy street

shopkeepers minding their own tills make it safer, and none was hired to

An old language

it survives while people still need it, not because anyone defends it

The catch

Occupancy is not loyalty. The skin bacterium that holds your barrier together also causes implant infections, and a full room is only safe while everyone stays in their own seat.

And the whole of it

Almost everything holding you up today is held by someone not thinking about you: a supplier, a neighbour, a bacterium. None of them owe you the protection, and that is exactly why it is so easy to remove by accident while aiming at something else.

03 Lab your turn

Room at the Table

Rehearse clearing an infection while watching the invisible defence you empty to do it.

04 Truth what's really going on

What is really going on

For most people the barrier against infection is not the immune system but a full room, and the drug that clears an infection clears the tenants too.

Why it works on us — Every visible thing here is an action - a prescription, a symptom, a cure - while the protection is only an absence of space.

Who gains

Clostridioides difficile — It feeds on nutrients that only become available once the resident community has been disrupted, so the treatment for one infection is its opening. [9]
Makers of probiotic, synbiotic and postbiotic products — Every finding that antibiotics injure the microbiome creates a market for repair, and one trial reporting a benefit was co-authored by the seller's chief executive. [27][28]
Stool-derived therapy companies — Two US approvals in 2022 and 2023 turned a donated, improvised procedure into a licensed product line, months before the free frozen supply stopped shipping. [23]
Hospitals that clear staph before surgery — Clearing a patient's own staph beforehand lowers the surgical infection the hospital would otherwise have to treat. [11]

Who pays

People given clindamycin — Each course in the year before sampling was linked to 47 fewer species detected, the largest effect of any drug in the study. [2]
Children in lower-income countries — Their disruption ran deeper and lasted longer than adults', and resistance genes stayed behind after the community itself recovered. [36]
Preterm infants — The FDA has reported that probiotics might cause infections or life-threatening illness in them, and no official guidance covers healthy use. [31]
Patients after liver surgery — C. diff followed 1.1% of these operations, nearly half within a week, and after liver resection it more than tripled the odds of dying. [41]
People who need a conventional stool transplant now — OpenBiome stopped shipping frozen preparations on 31 December 2024, which made the procedure harder to access in 2025. [23]

What nobody knows yet

Open questions from across today’s stories — ours included.

01

Whether the species that never came back were doing anything.

The Swedish work measured which species were present eight years on, not what any of them did for the people carrying them. [1][2]
02

How much of the skin or the gut is actually the bigger surface.

One review puts the gut ecosystem at about 400 square metres; another says skin covers at least 30 and surpasses the gut. They cannot both be right, and neither shows its working. [3][13]
03

How long recovery really takes, and for whom.

Diversity rebounded fastest in the first two years and then flattened, and the study's senior author says it seems you do not recover completely. Nobody can yet say who does. [2]
04

Which residents are load-bearing.

Methods for identifying keystone gut bacteria are inconsistent between studies, so restore your microbiome has no agreed target to aim at. [45]
05

Whether probiotics do anything measurable for a healthy person.

A meta-analysis of 34 trials rated the outcome evidence very low certainty, and one of 29 trials in older adults measured bacteria and markers rather than how people fared. The NIH says no official recommendations cover healthy people. [29][30][31]
06

How much weight a postbiotic result should carry when its authors sell it.

The 32-patient trial reporting 40% higher diversity lists the product company's chief executive and co-founders among its authors. [28]
07

Whether stool transplants outside C. diff will hold up.

The Parkinson's trial randomised 72 people and the kidney-cancer trial 20. Both are small, and neither has long follow-up yet. [24][25]
08

Whether the lung finding is true in people.

The long drop in one arm of lung immune defence after gut disruption was measured in mice, not humans. [35]

05 Hope carry this

The strongest defence most people have was never bought, never noticed and never fired a shot. The room was simply full, and a room can be filled again.