What Are Postbiotics and Do They Work?

The word postbiotics now appears on supplement labels, in gut health articles and across search results. Yet it does not describe a single thing. Different writers use it for different substances. For most of its life, the term had no agreed meaning at all.

That vagueness is not a small problem. It changes what a product must prove before the word can appear on its front.

The confusion has deep roots. Over the past decade, substances made from non-living microorganisms have travelled under many competing names. Paraprobiotics, ghost probiotics, metabiotics, heat-killed probiotics and bacterial lysates all described broadly similar ideas. Each name arrived from a different research group with a different emphasis.

That history has a practical cost. Anyone searching for evidence has to know every one of those older labels. Otherwise, whole sets of relevant trials never surface.

The idea itself is far older than any of the vocabulary. Preparations made from broken-down bacteria have been given to children for decades. They were used to prevent repeated chest infections, through immune effects still not fully explained. In Japan, the same principle goes back further still.

Similar products were formally named biogenics in 1998, having been sold there for more than a century. What changed recently was the pace, not the principle. Most research using the word has appeared since 2018.

More than half of it arrived after a panel of scientists finally agreed on a definition. The science and the marketplace expanded at the same moment, which is rarely comfortable.

Interest grew for practical reasons rather than fashionable ones. Live bacterial products are difficult to make and harder to keep. Many are sensitive to oxygen and heat, and numbers fall steadily across shelf life. A second problem sits alongside the technical one.

Reports have questioned how well live products work, and how safe they are, in seriously ill patients. Then researchers noticed something genuinely strange. Preparations of killed microorganisms sometimes matched the living versions.

Occasionally they outperformed them. That pattern acquired a name of its own, the probiotic paradox. A deeper shift was happening beneath it all.

Gut bacteria share metabolic pathways widely, so the same compounds come from many different species. Healthy people also carry strikingly different mixtures of microbes. Defining a normal, healthy gut has therefore proved close to impossible. Attention moved accordingly, towards preserving useful functions rather than hunting for particular microbes.

This leaves an ordinary reader in an awkward position. A tub on a shelf carries a word that scientists are still arguing over.

The honest answer to whether postbiotics work does not exist at the level of the word. It belongs to one preparation, at one dose, in one group of people, for one outcome. Learning to see that difference turns a confusing label into something anybody can judge.

What Are Postbiotics?

A panel of scientists settled the wording in 2021, and it repays close reading. A postbiotic is a preparation of inanimate microorganisms and/or their components that confers a health benefit. Every part of that sentence is doing work.

Start with the word. It joins ‘post’, meaning after, to ‘biotic’, meaning relating to living things. What it describes, quite literally, is after-life. ‘Inanimate’ was chosen over ‘inactive’ on purpose.

Inactive suggests inert material with nothing left to give. Inanimate records only that living microbes were present and have since been killed. Nothing in the word implies that function was lost along the way.

‘Preparation’ is the clause that catches most products out. Formulation, growth medium and killing method all shape whatever benefit appears.

So the term describes one specific product rather than an organism in general. Change the process, and you have changed the thing. That position openly contradicts an earlier definition. Under the 2011 wording for paraprobiotics, the killing method was not supposed to matter.

Once a benefit was shown, classification did not depend on it. ‘Components’ entered the definition because whole microbes may not be necessary. An effect can come from parts alone, such as pili or cell wall structures.

Components means cellular material here. That covers cell wall compounds, membrane lipids, peptidoglycans and teichoic acids, many of which provoke immune responses. Metabolites are something else again.

Those are substances a microbe produces, such as short-chain fatty acids, vitamins or bacteriocins. Such breadth could easily have become a loophole. The panel closed it by setting six requirements a preparation must satisfy. Fail one and the product is something else, whatever the label says.

1. Molecular Characterisation of the Original Microbes: A fully annotated genome sequence is expected. This allows precise identification and screening for genes that raise safety concerns.
2. A Full Description of How It Was Killed: The killing procedure must be described, along with the growth medium used.
3. Confirmation That Inactivation Actually Happened: Run a negative cell viability test on the finished product.
4. Evidence of a Health Benefit in the Host: This must come from a controlled, high-quality trial. Reasoning about mechanism does not count.
5. A Detailed Description of the Composition: What the preparation actually contains has to be described.
6. Assessment of Safety in the Target Host: Safety is judged for the intended use, not in general.

The fourth requirement separates this term from its neighbours. A demonstrated benefit is not an optional marketing extra.

It is written into what the word means. A product using the word without a trial behind it is misusing it, not stretching it. What postbiotics exclude is as revealing as what they admit. Purified metabolites, with no cellular material present, do not qualify.

They already have chemical names, such as butyric acid or lactic acid. Vaccines, purified proteins, peptides, exopolysaccharides and filtrates without cell components fall outside too. Chemically synthesised compounds are excluded as well.

So are viruses, including the bacteriophages that infect bacteria. The reasoning behind that exclusion is unusually neat. A metabolite was never alive, so it cannot have an after-life. Cellular material, by contrast, can only come from something that lived.

A second argument closes the door properly. Suppose purified metabolites from a microbe counted as postbiotics. A chemically identical preparation from a non-microbial source then could not.

The label would depend on a molecule’s origin rather than on what the molecule is. Undefined cultures are excluded for a different reason entirely.

Many traditional fermented foods rely on wild, mixed populations nobody has identified. Those cannot be used to make a postbiotic at all. Fermented products made with defined, characterised organisms are a different matter. Several assumptions people carry about the category turn out to be wrong.

A strain does not have to qualify as a probiotic while alive. Every older term carried the word probiotic, which implied proving benefit twice. That burden makes little sense, since a strain might do nothing alive and something useful once killed.

The host need not be human either. Companion animals and livestock both sit within the definition. The gut is not the only target.

These preparations are applied at body surfaces, including the mouth, skin, urogenital tract and nasopharynx. Injections fall outside the term completely. Sterility is not required, which surprises people. What matters is deliberate, controlled processing intended to kill the original strain.

Most such preparations contain no living cells, although some survivors may persist. No precise limit has been set on those survivors, which the panel treated as a question for regulators.

The range of possible source organisms is far wider than most people expect. Any microbe could serve, provided it is identified down to strain level. Even organisms considered dangerous while alive have been used. Mixtures of respiratory pathogen lysates are given deliberately as immune stimulants.

A preparation may also combine several taxonomically distinct strains. Each one still has to be properly identified and characterised. What reaches the consumer is often more than killed cells.

Growth medium components remain behind after harvesting and inactivation. In three trials of inactivated bacteria in children, the sachets contained the killed organisms together with neutralised spent culture medium. The liquid the bacteria grew in was therefore part of the product being tested. Manufacture follows a recognisable sequence.

Cells are cultured, broken open, separated out and then preserved. The strain, growth medium, and breaking method determine what ends up inside. One pair of studies makes that point vividly.

A particular Lactiplantibacillus strain produced material that supported gut lining renewal and shifted the microbes present. Grown on bitter melon instead, the same strain yielded material that reduced obesity in rats.

It altered both their gut bacteria and their blood chemistry. Feed a microbe differently, and you get a different product from the same organism.

How Postbiotics Differ from Prebiotics and Probiotics

Four terms sit in this family, and each carries its own agreed definition. A probiotic is a live microorganism that, given in adequate amounts, confers a health benefit. The word live is not decorative there.

It is the whole definition. A prebiotic is a substrate selectively used by host microorganisms that confers a health benefit. In everyday terms, that means food for your bacteria rather than food for you.

A synbiotic combines live microorganisms with substrates those microorganisms use. A postbiotic, as established already, is a preparation of inanimate microorganisms or their components. Reading them side by side reveals a shared design. All four describe something consumed or applied, not something produced inside the body.

That single principle does more work than any other in the family. It also settles an argument that keeps resurfacing.

Some authors describe postbiotics as the compounds gut bacteria generate from prebiotics or probiotics. Under the agreed definition, that reading does not hold. A postbiotic is the material administered, not the by-products that follow it. A preparation may well cause substances to be produced inside the body. Those substances are the consequence, not the postbiotic itself.

The distinction is easy to lose, because both stories involve the same compounds. Butyrate made by your own bacteria is chemically the same butyrate found anywhere else. What differs is everything around it.

Observational work shows how far apart the two situations sit. In over 200 children, a rich population of butyrate-producing bacteria was linked to earlier resolution of cow’s milk allergy. That is an association between resident organisms and an outcome. It is not evidence that supplying those compounds in a product does the same.

Where a compound comes from can even change where it is absorbed. Dietary B vitamins are taken up in the small intestine. Those made by microbes are absorbed in the large intestine.

Those are different environments and possibly different fates. Origin turns out to matter more than chemistry alone suggests.

The boundary with probiotics is blurrier in practice than on paper. Probiotics must be alive and present in effective numbers when taken. Yet most probiotic preparations also contain large numbers of dead and damaged microorganisms. By the end of shelf life, dead cells can outnumber living ones ten to a hundred times over.

What those non-viable cells contribute has received remarkably little attention. A shopper buying a live product is often buying a substantial quantity of dead ones. Even so, a probiotic that dies on the shelf does not become a postbiotic.

The definition requires a deliberate, controlled and reproducible killing step. Cell death across months of storage is not that. Nor can it be assumed to leave behind the same biologically active material.

Dose exposes the gap between the categories rather neatly. One head-to-head trial gave the live product as one billion colony-forming units daily. A colony-forming unit is a countable clump of bacteria capable of growing. The killed version was given as 2.5 billion cells daily, which worked out to 50 milligrams.

Counts of living units and counts of dead cells are not comparable quantities. Two numbers on two labels can look similar and mean entirely different things.

Fermented foods occupy their own uncomfortable position here. They may contain substantial numbers of non-viable microbial cells, particularly after storage or processing. Pasteurised soy sauce and baked sourdough both illustrate that. The microbes that did the work are no longer alive by the time anybody eats them.

Synbiotics are kept separate too, at least for now. Some researchers argue postbiotics should be folded into that definition, since they may also strengthen the microbial community. The agreed position has not moved on that point.

Practical differences follow from being dead rather than alive. Inanimate preparations are extremely stable, often for years at room temperature. They suit places without reliable refrigeration. They also survive transport from factory to shelf without losing potency.

They can be taken alongside antibiotics without losing effect, because there is nothing left to kill. That single property removes a familiar frustration for anyone on a course of them. They cannot colonise the gut either.

Any effect therefore stops when the supply stops. Removing the requirement to be alive also widens the pool of usable organisms.

Probiotics are restricted to species with a recognised history of safe consumption. Most microbes living in a healthy human gut do not hold that status. They cannot be given as live products even where research suggests they would help. Faecalibacterium prausnitzii shows what that costs.

It is widely regarded as beneficial, yet it is not currently considered safe to consume alive. Its usefulness is thought to come from the butyrate it produces and an anti-inflammatory protein it makes. Neither of those requires the organism to be breathing.

How Postbiotics May Work

Being dead imposes a hard constraint that shapes everything else. Whatever a preparation does, it cannot make anything new inside the body. Every active molecule had to be produced by the microbes before they were killed.

It must also survive the killing process. Then it has to be present in enough quantity to matter. That rules out one whole class of explanation immediately. Any mechanism needing ongoing metabolic activity is simply unavailable.

A living microbe can fight off a competitor by producing acid as it grows. A dead one cannot do that at any dose.

Within those limits, five routes of action are generally considered for postbiotics. They may act alone or together, and different preparations may use entirely different ones.

Changing the Microbes Already There

Some molecules carried in these preparations act directly on resident organisms. Lactic acid and bacteriocins, which are antimicrobial peptides made by bacteria, both show activity in living systems. Resident microbes may also feed on the lactic acid and convert it into short-chain fatty acids.

The preparation then acts as raw material rather than as a drug. Yet the evidence for genuine microbiome change is thin. An inanimate preparation cannot colonise, so any influence is likely to be temporary. Animal studies show only small shifts in which species are present.

None report what those shifts do to gut function. Strikingly, changing the microbiome may not be required at all. One killed preparation improved constipation measures in rats with no detectable change in their gut bacteria.

Strengthening the Gut Barrier

The gut lining is held together by protein complexes called tight junctions. Two of the best studied are zonula occludens-1 and occludin.

When those proteins fall, the barrier loosens. Chemicals, antigens and microbes can then cross where they should not. Secreted bacterial proteins appear to support that barrier. So do exopolysaccharides, which are sugar polymers released outside the bacterial cell.

Exopolysaccharides from one Lactiplantibacillus strain raised both proteins in mice and in human cell lines. They also suppressed several inflammatory signalling molecules at the same time. Short-chain fatty acids show similar barrier effects at low concentrations.

Those demonstrations sit in cell culture rather than in people. That distinction matters more than it might appear. It recurs throughout this whole subject.

Signalling to the Immune System

Immune activity is generally driven by microbial surface structures meeting receptors on immune cells. Those receptors include Toll-like receptors, which recognise patterns common to microbes. Crucially, those structures survive only if the killing method leaves them intact. Heat, pressure and chemical treatment do not treat them equally.

The signal therefore depends on which organism was used and how it was processed. Two products can share an ingredient list and behave differently.

Direction varies as well. Some preparations push immune responses one way, and some push them the other. Lipoteichoic acid, a cell wall component, illustrates the ambiguity well. Some reports show it dampening inflammation, while others show it damaging intestinal tissue.

Acting on the Body’s Metabolism

Effects here can come from enzymes and metabolites carried on or inside the dead cells. Bile salt hydrolase offers a clear example. This microbial enzyme modifies bile acids, which widens the range circulating in the body.

Those altered bile acids then interact with host receptors. Glucose handling, fat handling and energy use all sit downstream of that. Short-chain fatty acids contribute too, with propionate affecting insulin sensitivity and fat production. Butyrate has been shown to raise antioxidant activity in the colon of healthy people.

Perhaps most surprisingly, killing a bacterium can strengthen a metabolic effect. Pasteurisation improved the effect of Akkermansia muciniphila on fat mass and insulin resistance in mice.

Talking to the Nervous System

Microbes produce compounds that can act on nerves in the gut and the brain. These include serotonin, dopamine, acetylcholine and gamma-aminobutyric acid.

How much of that survives in an inanimate preparation is poorly documented. Sometimes it does not survive at all. Work in mice and human gut tissue compared live and heat-killed Bifidobacterium dentium. Viability turned out to be essential for that organism’s effect on serotonin.

Killing it removed the very thing being measured. Not every effect is portable across the boundary between alive and dead. Two production details decide how much of any of this actually applies.

The first is whether cells are left whole or broken open. An intact wall shields molecules from digestive enzymes and may keep them in the body longer. Breaking the cell may instead expose useful compounds and make them more available.

Intact cells may also interact with immune receptors differently than loose fragments. Evidence has not settled either possibility. The second detail is the process itself, which alters function rather than just ending life. Spray drying strips the pili from one widely studied Lacticaseibacillus strain.

Pili are hair-like surface projections used to grip the gut lining. Losing them was linked to more inflammatory signalling and less stimulation of cell growth.

So two preparations of the same organism can behave differently in the same body. The manufacturing line is part of the biology, not a detail beneath it. None of this amounts to a settled account of how postbiotics work. The available data are insufficient, and the substances involved vary widely.

For most preparations, several mechanisms probably act together in ways nobody has separated. That is not unusual in medicine, where many effective treatments work through disputed routes. Knowing the mechanism has never been required to show that something works.

What the Human Evidence Shows

Human data on these preparations are limited, uneven in quality and mixed in outcome. The category rests on a thinner base than its shelf presence suggests. That is not a reason to dismiss it, but it is a reason to look at individual results. Irritable bowel syndrome has produced the most informative work.

It is also where the limits of that work are easiest to see. One trial gave a killed strain of Bifidobacterium bifidum to 443 adults for eight weeks. Success meant at least 30% improvement in abdominal pain, plus adequate relief of symptoms.

Among those treated, 34% reached it. On placebo, 19% did.

No serious side effects appeared in either group. That strain had already been shown to work while alive. Producing it as a killed preparation and testing it again reversed a widespread assumption. Killing the organism did not cost it its effect.

A later trial went further by testing both forms at once. Two hundred adults with diarrhoea-predominant irritable bowel syndrome took part. They received live Bifidobacterium longum, the heat-treated version of that same strain, or a placebo.

All three were identical in size, colour, and weight, and participants took them daily for 84 days. The two active groups were close to indistinguishable. Symptom severity fell by 59.6% on the live strain and 60.8% on the killed one.

Placebo produced a 21.2% reduction. Judged by a 30% fall in abdominal pain, responders reached 90.63% and 88.06%. The placebo group reached 28.79%. Days with normal stools rose from around one a week to over four.

Use of anti-diarrhoeal medication fell by more than half compared with placebo. Read at that level, the result looks overwhelming.

A different threshold in the same trial tells a more careful story. Researchers also measured abdominal pain against a 2.2-point change, the amount considered clinically meaningful. By that standard, 53.13% responded on the live strain and 47.76% on the killed one. Placebo reached 16.67%.

So the treatment clearly beat placebo. Yet under half of one treated group reached the stricter mark. Both numbers come from the same trial, the same people and the same 84 days.

Only the yardstick changed. The trial’s own authors added further caution. Anxiety scores improved significantly, but participants had no meaningful anxiety at the start. That leaves the practical relevance of the finding open.

All study sites also sat in a single country, with diet left unstandardised. Local patterns in fibre intake, parasitic infection and rural living could not be measured out. Results from one population do not automatically travel to another.

Not every preparation performs the same way under the same conditions. A non-living lysate of two gut bacteria was tested in 389 patients over 26 weeks.

Overall symptom response reached 17.4%, against 14.4% on placebo. Abdominal pain showed no significant difference either. Only a subgroup with diarrhoea-type disease improved. One condition, two preparations, two entirely different outcomes.

Long-running diarrhoea produced a result that is easy to miss and hard to ignore. A controlled study compared a heat-treated preparation directly against live lactobacilli. The killed version significantly outperformed the living one on symptoms.

That is not the same as saying dead is better. It reminds us that the two are different interventions, not two grades of the same product. Evidence in young children is narrower than the marketplace implies.

One review found only seven controlled trials in under-fives, covering 1,740 children. Heat was the killing method in every single one. Nobody has tested the alternatives properly in that age group. A killed strain of Lacticaseibacillus paracasei roughly halved the risk of diarrhoea.

Sore throat and laryngitis also fell, consistently across the pooled trials. Other outcomes in the same analysis did not change.

Neither runny nose nor middle ear infection showed a significant difference. A benefit in one respiratory outcome did not extend to the next. Treatment results proved steadier than prevention results. Across four trials in 224 children, a killed strain of Lactobacillus acidophilus shortened diarrhoea by around 20 hours.

Fermented infant formula shows how quickly apparent promise flattens out. One trial followed 913 French infants across 94 centres. Diarrhoea episodes did not differ from the control formula.

The authors reported fewer cases of dehydration. Several paediatric studies report laboratory changes rather than changes anybody could notice. Fermented formula lowers stool acidity and raises stool antibody levels. Whether either change is itself a benefit has never been established.

Across the reviewed formulas, safety and tolerance were consistent, and clinical benefit was not. Older adults have produced one of the clearest dose patterns available. In 280 people over 65, a killed preparation was taken across 20 weeks.

Colds affected 47.3% on placebo, 34.8% on the lower amount and 29.0% on the higher one. Quality-of-life scores climbed with the amount taken.

A stepwise pattern like that is harder to explain away than a single positive result. Larger studies in other groups have repeatedly failed to reproduce it. A four-month trial in 172 preschool children found no prevention of respiratory infection. Salivary antibody levels rose in the treated children.

Their illness rate did not fall. That gap between a laboratory marker and a real outcome recurs constantly in this literature. In almost 2,200 healthy adults, a killed preparation did not reduce influenza.

Only participants under 40 showed a reduction, and that came from a subgroup analysis. Adults with chronic obstructive pulmonary disease produced a similar disappointment. Researchers tested a bacterial lysate in 288 patients.

It failed to reduce flare-ups, which was the outcome the study existed to measure. Metabolic health has produced the single most counterintuitive human result available. Overweight, insulin-resistant volunteers took live or pasteurised Akkermansia muciniphila daily for three months. The pasteurised organism improved insulin sensitivity and lowered insulin levels and total cholesterol.

White blood cell counts fell, and body weight declined slightly. The living organism produced no significant overall benefit.

Killing the bacterium did not weaken it. On these measures, it appears to have done the opposite. Set against each other, these results resist any verdict on postbiotics as a group. Positive and negative findings both exist in healthy populations.

No recommendation can currently be made either way. Data rarely extend beyond a single study for any given preparation. Confirmation trials attract little funding, because repeating a finding offers no scientific novelty.

Commercial sponsors have limited interest in reconfirming a product already on the market. Confirmation is exactly what tells you an initial result was robust. Trials also measure different things, which makes comparison difficult within a single condition. No agreed standard set of outcomes covers benefits and harms.

For irritable bowel syndrome alone, three separate thresholds exist for judging improvement. The one a researcher picks changes how impressive the result looks, as the figures above showed. For postbiotics, the number quoted matters less than the yardstick behind it.

How to Judge Food and Supplement Claims

Reading a label in this category is harder than it should be. No regulator anywhere has produced a definition or framework specific to these products.

In the European Union, no rules cover probiotics, prebiotics, synbiotics or postbiotics as such. The United States Food and Drug Administration has not addressed them directly either. Products are handled under whichever category they were developed as. That vacuum has consequences on the shelf.

Long-established products containing killed microorganisms may list something like lactobacilli in the ingredients. They do not always state that those microorganisms are no longer alive. A shopper reading that list would reasonably assume the bacteria are living.

Nothing on the pack corrects the assumption. The word also travels ahead of the rules. One infant formula was launched labelled ‘with postbiotics’, in a country whose food code had not addressed them at all.

Verifying what is in a product is genuinely difficult, even for laboratories. Counting killed cells relies on techniques whose relationship to traditional bacterial counts is not established. Full chemical analysis needs specialist equipment and several purification steps. No consumer can check any of this independently.

Since regulation offers little help, judgement has to come from the product itself. The agreed definition supplies a workable set of questions.

Each one is answerable from a label, a website or a published trial. An answer that cannot be found is itself an answer.

• Which Organism, Named to Strain Level: A genus and species are not enough on their own. Different strains of one species behave differently in the body.
• How the Microbes Were Killed: Heat is the most common method, though pressure, radiation, and chemical treatments exist. Each produces a different composition and, therefore, different effects.
• Whether the Process Matches the Tested Product: A preparation should be made the same way as the tested one. If the process is altered, the benefit has to be demonstrated again.
• What Dose, Expressed in Which Units: Living products are counted in colony-forming units. Killed ones are counted as cells or weighed in milligrams. Those figures cannot be compared directly.
• Who It Was Tested In: A benefit has to be shown in the target group. Results in hospitalised infants do not transfer to healthy adults.
• Which Outcome Was Actually Measured: Some trials report symptoms people notice. Others report stool acidity or antibody levels, whose value to the person is often unproven.
• Whether a Human Study Exists for This Exact Preparation: Responsible use of the term requires controlled studies in people. Evidence for a similar product is not evidence for this one.

Fermented foods are where this reasoning most often goes wrong. They are the main way microbial products already reach people for gut health. Most consumers have no idea those components are present.

Traditional fermentation uses wild microbes to leaven, preserve and build flavour. The identity and effects of most compounds they produce remain unknown. Sourdough bread makes the difficulty concrete. Baking at high temperature kills the microbes that fermented the dough.

Any benefit would come from cell fragments and heat-stable compounds left behind. Those have often been called postbiotics, and they share many of the characteristics. Yet the agreed definition excludes products made by wild microbes nobody has identified.

Either those microbes get characterised, or defined organisms are used instead. Heat treatment of a fermented food is a trade-off, not an upgrade.

It extends shelf life while killing the living fermentation microbes. It can also destroy heat-sensitive components, including certain vitamins and lactase. Some shoppers additionally perceive the treated food as less fresh. Whole foods remain the most reliable route to microbial activity in the gut.

World Health Organization guidance recommends at least 25 grams of fibre a day. Estimated intake worldwide is closer to 20 grams. In trials that raised fibre intake, seven out of ten did it through food rather than supplements.

Supporting the bacteria you already have is not the same as buying their products. Some products carry the language while falling outside the definition entirely. A cell-free filtrate or an isolated compound contains no cellular material.

It should carry the chemical name of its contents instead. A supplement containing isolated lactic acid contains lactic acid, not postbiotics. The presence of such an ingredient need not imply any health claim at all. Preparations of this kind are added to food for texture and yield.

One was used in cheddar cheese to raise moisture content and improve output. Others act as preservatives, extending shelf life and inhibiting spoilage organisms.

Regulators intervened where evidence was formally weighed. European authorities assessed eight bacterial lysates developed for respiratory conditions. They restricted them to preventing repeated infections, rather than treating them or pneumonia. Manufacturers were required to supply further safety and effectiveness data from new studies.

Combination products present the last interpretive problem. Many tested formulas differ from their comparator in several ways at once. Added prebiotics, milk sugars and modified fats can all appear in the same product. No conclusion can then be drawn about which ingredient produced any observed effect.

Who Should Be Careful

The safety argument for killed preparations starts from something real. Microorganisms that cannot multiply cannot cause bloodstream infection. That risk does exist with live probiotics, although it remains extremely rare. Case reports in newborns include blood infection, gut inflammation, pneumonia and meningitis.

Those reports are why part of the scientific community resists live products in young children. Removing the ability to multiply removes a genuine hazard. It does not remove every hazard.

These preparations are not inert substances, by definition. Safety also cannot be inferred from the safety of the original organism.

Components released from dead bacteria carry risks of their own. Lipopolysaccharides from some bacteria can provoke sepsis and toxic shock. That material normally sits embedded in the outer membrane of a living cell. Death is what releases it.

Two trials in children show this is not a theoretical concern. In 35 infants with eczema and cow’s milk allergy, researchers compared three formulas. Plain formula caused no problems, and neither did formula containing the live strain.

The formula containing the heat-killed strain produced significantly more diarrhoea. The killed version caused harm. A second trial studied infants aged 6 to 12 months at high risk of dying from diarrhoea.

Diarrhoea affected 26% of those given micronutrients plus a heat-killed organism. Micronutrients alone produced 15%. Placebo also produced 26%. The authors concluded that adding the killed organism had a negative effect.

Vomiting and abdominal swelling were more common in that group as well. Elsewhere in this literature, one trial reported severe dehydration linked to a killed preparation.

The safety record is thinner than the effectiveness record, which is unusual. Of seven controlled trials in children under five, only three examined side effects at all. The remaining four did not report on adverse events. Those three found no significant differences, which is reassuring rather than conclusive.

Risk is not uniform across products describing themselves in the same way. Different killing methods leave different numbers of surviving cells behind. Heat, high pressure and oxygen exposure would each be expected to differ.

No published comparison exists, so those differences remain unmeasured. Individual components carry their own questions. Lipoteichoic acid and other cell wall material may provoke excessive inflammatory responses. Regulation compounds the uncertainty around postbiotics rather than resolving it.

Frameworks built for live microbes do not apply to killed preparations. European safety assessment focuses on live organisms with a history of safe consumption. Only three killed preparations have been assessed as novel foods so far.

Two were authorised, and one was not. Safety assessments are still required to confirm the components themselves are not toxic.

One genuine regulatory advantage is worth stating plainly. Killed preparations cannot acquire or pass on antibiotic resistance genes, which live bacteria can do inside the body. Certain groups have clearer reasons to take advice first. People with weakened immune systems have been advised to avoid these preparations altogether.

Suitable trials in that population have not been carried out. Neither have the animal studies that would normally precede them. Pregnancy raises its own questions, since pregnant women are the intended recipients of some products.

Researchers have repeatedly called for work on the associated risks. Concerns have also been raised about older people. That is part of the reason killed versions are offered as a substitute at all.

Children remain the group where caution is most justified. Every harm recorded above occurred in infants or young children. A review of acidified and fermented infant formulas found a theoretical risk of acid build-up. Actual cases appeared only in unwell infants, such as those with short bowel syndrome.

One question underlies all of this and remains unanswered. No safe dose range has been established for postbiotics as a group.

Existing safety data concentrate on a handful of killed probiotic strains. Research on the other components remains scarce.

Safety and effectiveness point in the same direction, but it’s quite specific. Neither question can be answered by the words on the front of a pack. Both belong to a named preparation, made in a described way, tested in people who resemble the person taking it.

That is an unglamorous position for a category sold on promise. It is also the most useful thing the research currently offers. Products will change, and the definition may change with them, since the field has not finished arguing. The question that survives all of it is small and stubborn.

Which organism, killed how, at what dose, in whom, for which outcome, and where is the study? A label that cannot answer that is selling a word.

Sources

• Adams CA. The probiotic paradox: live and dead cells are biological response modifiers. Nutr Res Rev. 2010;23(1):37-46.
• Andresen V, Gschossmann J, Layer P. Heat-inactivated Bifidobacterium bifidum MIMBb75 (SYN-HI-001) in the treatment of irritable bowel syndrome: a multicentre, randomised, double-blind, placebo-controlled clinical trial. Lancet Gastroenterol. Hepatol. 2020;5:658-666.
• Braido F, et al. Sub-lingual administration of a polyvalent mechanical bacterial lysate (PMBL) in patients with moderate, severe, or very severe chronic obstructive pulmonary disease (COPD) according to the GOLD spirometric classification: a multicentre, double-blind, randomised, controlled, phase IV study (AIACE study: Advanced Immunological Approach in COPD Exacerbation). Pulm. Pharmacol. Ther. 2015;33:75-80.
• Bunyavanich S., Shen N., Grishin A., Wood R., Burks W., Dawson P., Jones S.M., Leung D.Y.M., Sampson H., Sicherer S., et al. Early-life gut microbiome composition and milk allergy resolution. J. Allergy Clin. Immunol. 2016;138:1122-1130.
• Chiron C, Tompkins TA, Burguiere P. Flow cytometry: a versatile technology for specific quantification and viability assessment of micro-organisms in multistrain probiotic products. J. Appl. Microbiol. 2018;124:572-584.
• Depommier C, Everard A, Druart C, Plovier H, Van Hul M, Vieira-Silva S, et al. Supplementation with Akkermansia muciniphila in overweight and obese human volunteers: a proof-of-concept exploratory study. Nat Med. 2019;25(7):1096-1103.
• EFSA NDA Panel. Scientific opinion on the safety of pasteurised Akkermansia muciniphila as a novel food pursuant to Regulation (EU) 2015/2283. EFSA J. (2021) 19:6780.
• Engevik MA, et al. Human-derived Bifidobacterium dentium modulates the mammalian serotonergic system and gut-brain axis. Cell Mol. Gastroenterol. Hepatol. 2021;11:221-248.
• Esposito S, et al. Nonspecific immunomodulators for recurrent respiratory tract infections, wheezing and asthma in children: a systematic review of mechanistic and clinical evidence. Curr. Opin. Allergy Clin. Immunol. 2018;18:198-209.
• European Food Safety Authority. Scientific Opinion on the safety of heat-treated milk products fermented with Bacteroides xylanisolvens DSM 23964 as a novel food. EFSA J. 2015;13:3956.
• European Medicines Agency. Assessment report. Referral under Article 31 of Directive 2001/83/EC. Bacterial lysates-containing medicinal products for respiratory conditions. EMA (2019).
• Feng Y, Wang Y, Wang P, Huang Y, Wang F. Short-chain fatty acids manifest stimulative and protective effects on intestinal barrier function through the inhibition of NLRP3 inflammasome and autophagy. Cell Physiol. Biochem. 2018;49:190-205.
• Fiore W., Arioli S., Guglielmetti S. (2020). The neglected microbial components of commercial probiotic formulations. Microorganisms 8:1177.
• Food and Drug Administration. What does FDA regulate?
• Gibson GR, et al. Expert consensus document: The International Scientific Association for Probiotics and Prebiotics (ISAPP) consensus statement on the definition and scope of prebiotics. Nat. Rev. Gastroenterol. Hepatol. 2017;14:491-502.
• He X., Zhao S., Li Y. Faecalibacterium prausnitzii: A Next-Generation Probiotic in Gut Disease Improvement. Can. J. Infect. Dis. Med. Microbiol. 2021;2021:6666114.
• Hill C, et al. Expert consensus document. The International Scientific Association for Probiotics and Prebiotics consensus statement on the scope and appropriate use of the term probiotic. Nat. Rev. Gastroenterol. Hepatol. 2014;11:506-514.
• Hishiki H, Kawashima T, Tsuji NM, Ikari N, Takemura R, Kido H, et al. A double-blind, randomized, placebo-controlled trial of heat-killed Pediococcus acidilactici K15 for prevention of respiratory tract infections among preschool children. Nutrients. (2020) 12:1989.
• Imperial I.C.V.J., Ibana J.A. Addressing the Antibiotic Resistance Problem with Probiotics: Reducing the Risk of Its Double-Edged Sword Effect. Front. Microbiol. 2016;7:1983.
• Kaczynska A., Klosinska M., Janeczek K., Zarobkiewicz M., Emeryk A. (2022). Promising immunomodulatory effects of bacterial lysates in allergic diseases. Front. Immunol. 13:907149.
• Kiekens S, et al. Impact of spray-drying on the pili of Lactobacillus rhamnosus GG. Microb. Biotechnol. 2019;12:849-855.
• Kirjavainen PV, Salminen SJ, Isolauri E. Probiotic bacteria in the management of atopic disease: underscoring the importance of viability. J. Pediatr. Gastroenterol. Nutr. 2003;36:223-227.
• Kothari D., Patel S., Kim S.-K. Probiotic supplements might not be universally-effective and safe: A review. Biomed. Pharmacother. 2019;111:537-547.
• Lebeer S, Vanderleyden J, De Keersmaecker SC. Host interactions of probiotic bacterial surface molecules: comparison with commensals and pathogens. Nat. Rev. Microbiol. 2010;8:171-184.
• Long SL, Gahan CGM, Joyce SA. Interactions between gut bacteria and bile in health and disease. Mol. Asp. Med. 2017;56:54-65.
• Lukasik J., Salminen S., Szajewska H. Rapid review shows that probiotics and fermented infant formulas do not cause d-lactic acidosis in healthy children. Acta Paediatr. 2018;107:1322-1326.
• Mack I, Schwille-Kiuntke J, Mazurak N, Niesler B, Zimmermann K, Monnikes H, et al. A nonviable probiotic in irritable bowel syndrome: a randomized, double-blind, placebo-controlled, multicenter study. Clin Gastroenterol Hepatol. 2022;20(5):1039-47.
• Malagon-Rojas JN, Mantziari A, Salminen S, Szajewska H. Postbiotics for preventing and treating common infectious diseases in children: a systematic review. Nutrients. (2020) 12:389.
• Marco ML, et al. The International Scientific Association for Probiotics and Prebiotics (ISAPP) consensus statement on fermented foods. Nat. Rev. Gastroenterol. Hepatol. 2021;18:196-208.
• Mitsuoka, T. Intestinal Flora and Probiotics (Gakkai Shuppan Center, 1998).
• Park S-A, Lee G-H, Hoang T-H, Lee H-Y, Kang I-Y, Chung M-J, et al. Heat-inactivated Lactobacillus plantarum nF1 promotes intestinal health in loperamide-induced constipation rats. PLoS One. (2021) 16:e0250354.
• Periti P, Mazzei T. Antibiotic-induced release of bacterial cell wall components in the pathogenesis of sepsis and septic shock: a review. J. Chemother. 1998;10:427-448.
• Pique N, Berlanga M, Minana-Galbis D. Health benefits of heat-killed (tyndallized) probiotics: an overview. Int J Mol Sci. 2019;20(10):2534.
• Plovier H, Everard A, Druart C, Depommier C, Van Hul M, Geurts L, Chilloux J, Ottman N, Duparc T, Lichtenstein L, et al. A purified membrane protein from akkermansia muciniphila or the pasteurized bacterium improves metabolism in obese and diabetic mice. Nat Med. 2017;23(1):107-113.
• Salminen S., Collado M.C., Endo A., Hill C., Lebeer S., Quigley E.M.M., Sanders M.E., Shamir R., Swann J.R., Szajewska H., et al. The International Scientific Association of Probiotics and Prebiotics (ISAPP) consensus statement on the definition and scope of postbiotics. Nat. Rev. Gastroenterol. Hepatol. 2021;18:649-667.
• Sanlier N., Gokcen B.B., Sezgin A.C. Health benefits of fermented foods. Crit. Rev. Food Sci. Nutr. 2019;59:506-527.
• Sharieff W, Bhutta Z, Schauer C, Tomlinson G, Zlotkin S. Micronutrients (including zinc) reduce diarrhoea in children: the Pakistan Sprinkles Diarrhoea Study. Arch. Dis. Child. 2006;91:573-579.
• Shinkai S, Toba M, Saito T, Sato I, Tsubouchi M, Taira K, et al. Immunoprotective effects of oral intake of heat-killed Lactobacillus pentosus strain b240 in elderly adults: a randomised, double-blind, placebo-controlled trial. Br J Nutr. (2013) 109:1856-65.
• Steinert R.E., Lee Y.K., Sybesma W. Vitamins for the Gut Microbiome. Trends Mol. Med. 2020;26:137-140.
• Swanson KS, et al. Expert consensus document: The International Scientific Association for Probiotics and Prebiotics (ISAPP) consensus statement on the definition and scope of synbiotics. Nat. Rev. Gastroenterol. Hepatol. 2020;17:687-701.
• Taverniti V, Guglielmetti S. The immunomodulatory properties of probiotic microorganisms beyond their viability (ghost probiotics: proposal of paraprobiotic concept). Genes Nutr. 2011;6:261-274.
• Thibault H., Aubert-Jacquin C., Goulet O. Effects of Long-term Consumption of a Fermented Infant Formula (with Bifidobacterium breve c50 and Streptococcus thermophilus 065) on Acute Diarrhea in Healthy Infants. J. Pediatric Gastroenterol. Nutr. 2004;39:147-152.
• Thorakkattu P., Khanashyam A. C., Shah K., Babu K. S., Mundanat A. S., Deliephan A., et al. (2022). Postbiotics: current trends in food and pharmaceutical industry. Foods. 11:3094.
• Tian L., Wang X.-W., Wu A.-K., Fan Y., Friedman J., Dahlin A., Waldor M.K., Weinstock G.M., Weiss S.T., Liu Y.-Y., et al. Deciphering functional redundancy in the human microbiome. Nat. Commun. 2020;11:6217.
• Vale A.D.S., Pereira G.V.D.M., de Oliveira A.C., Neto D.P.D.C., Herrmann L.W., Karp S.G., Soccol C.R. Production, formulation, and application of postbiotics in the treatment of skin conditions. Fermentation-Basel. 2023;9(3):264.
• Wang W., Zhang L., Li Y. Production of Volatile Compounds in Reconstituted Milk Reduced-Fat Cheese and the Physicochemical Properties as Affected by Exopolysaccharide-Producing Strain. Molecules. 2012;17:14393-14408.
• Wen J.-J., Li M.-Z., Gao H., Hu J.-L., Nie Q.-X., Chen H.-H., Zhang Y.-L., Xie M.-Y., Nie S.-P. Polysaccharides from fermented Momordica charantia L. with Lactobacillus plantarum NCU116 ameliorate metabolic disorders and gut microbiota change in obese rats. Food & Function. 2021;12(6):2617-2630.
• World Health Organization. Carbohydrate intake for adults and children: WHO guideline. 2023.
• Xiao SD, Zhang DZ, Lu H, Jiang SH, Liu HY, Wang GS, Xu GM, Zhang ZB, Lin GJ, Wang GL. Multicenter, randomized, controlled trial of heat-killed lactobacillus acidophilus lb in patients with chronic diarrhea. Adv Ther. 2003;20(5):253-260.
• Yelin I, et al. Genomic and epidemiological evidence of bacterial transmission from probiotic capsule to blood in ICU patients. Nat. Med. 2019;25:1728-1732.
• Zhou X., Qi W., Hong T., Xiong T., Gong D., Xie M., Nie S.-P. Exopolysaccharides from Lactobacillus plantarum NCU116 Regulate Intestinal Barrier Function via STAT3 Signaling Pathway. J. Agric. Food Chem. 2018;66:9719-9727.
• Zhou X., Zhang D., Qi W., Hong T., Xiong T., Wu T., Geng F., Xie M., Nie S. Exopolysaccharides from Lactobacillus plantarum NCU116 facilitate intestinal homeostasis by modulating intestinal epithelial regeneration and microbiota. Journal of Agricultural and Food Chemistry. 2021;69(28):7863-7873.