Postbiotics are scientifically relevant because some microbial structures and associated biological activities can interact with a host after the microorganisms have lost their ability to replicate. Colonization is not required for cell envelopes, surface structures or other constituents in a preparation to influence mucus, epithelial junctions or immune responses. However, the absence of live cells describes a material state; a health benefit must be established for the tested preparation in a defined host. Understanding postbiotics therefore requires a connected account of material identity, processing and measured outcomes. [1, 2]

What a postbiotic preparation contains

The 2021 ISAPP consensus defines a postbiotic as a preparation of inanimate microorganisms and/or their components that confers a health benefit on the host. The defining elements are a microbial origin, inanimate cells or cellular components, and a demonstrated health benefit. The source microorganism does not first have to qualify as a probiotic. Purified lactate, butyrate or vitamins, and filtrates without relevant cellular constituents, do not automatically qualify simply because they originate from fermentation. [1, 2]

This distinction is useful when reading earlier literature. Some studies used the term postbiotics for secreted proteins, culture supernatants and purified metabolites. These studies can still explain microbial effector molecules, but the identity of the tested material must remain explicit. Inactivated biomass, fermentation preparations containing cells, cell-free supernatants and purified proteins have different compositions; a shared label does not make their effects interchangeable. [1, 2, 3]

Table 1. Identifying microbial materials

Why biological activity can remain after inactivation

Hosts recognize molecular structures, not only whether a microorganism can reproduce. Inanimate cells may retain surface features that interact with host receptors, while a preparation may also contain cell fragments and co-present metabolites. Effects may involve direct contact with the host or indirect changes mediated by the intestinal environment and resident microbiota. Neither route requires the inactivated organism to resume growth. [1, 2]

Cellular structures and soluble molecules provide different mechanistic clues. Gao and colleagues observed increased mucus production when Caco-2 cells were exposed to a Lactobacillus rhamnosus GG culture supernatant. Protease treatment weakened this response, whereas nuclease and carbohydrate-digesting enzyme treatments did not show the same pattern. The researchers then identified and produced the HM0539 protein and examined mucin and barrier function, moving from an active fermentation preparation toward a more specific source of activity. [3]

The HM0539 study predates the 2021 consensus. Here, its purified protein is classified as a microbial effector molecule rather than equated with a complete postbiotic preparation. The enzyme comparisons in the original Figure 1 support a protein contribution to the mucus response induced by the supernatant. They do not establish that all lactobacillus supernatants contain the same protein or that every inactivation process preserves the effect. [1, 2, 3]

Barrier protection requires functional measurements

The intestinal barrier includes mucus, epithelial cells and intercellular junctions. Increased MUC2 or tight-junction protein expression provides information about structure and regulation. Transepithelial electrical resistance (TEER) and tracer permeability add functional evidence of how effectively a cell layer restricts passage. Reading expression and functional measurements together gives a fuller account of barrier protection than interpreting gene upregulation alone as repair. [3, 5]

The JY062 material studied by Guo and colleagues was an inactivated, dried preparation containing a fermented skimmed-milk matrix, rather than isolated biomass or a purified acid. In a cell co-culture model, TEER fell by 238.2 ± 0.35 Ω·cm² over 72 hours in the LPS group and by 107.93 ± 3.64 Ω·cm² in the JY062 group. The paper reported p < 0.001 between these groups, with FITC-dextran flux results providing additional support. This is in vitro barrier protection; it cannot be converted into a percentage improvement in human nutrient absorption. [5]

In a separate study involving 24 male mice, inactivated Akkermansia muciniphila AKK PROBIO was associated with a lower disease activity index than the DSS model group, alongside colonic tissue and mucus-related findings. The results show that an intact inactivated microbial preparation can affect host outcomes in a defined colitis model. DSS causes chemical injury, and the strain and material also differ from those used in the human studies. [4]

Immune and microbiota findings depend on context

Immune modulation does not mean suppressing every cytokine or increasing every immune response. Cellular constituents, host status and the inflammatory stimulus can alter the direction of a measured response. In colitis models, reduced inflammation often accompanies preservation of the mucosa. Microbiota and fecal short-chain fatty acid measurements in the AKK PROBIO study can be read alongside tissue findings, but correlations and predicted functions do not establish that a particular genus or pathway is a necessary mediator. [1, 4]

For this reason, increasing beneficial bacteria is not a sufficiently precise mechanistic description. A higher relative abundance does not necessarily mean a higher absolute cell count, and predicted metabolic functions are not measured metabolic fluxes. A more informative question is which preparation changes which measurable physiological outcome in which host context. [2, 4]

Human outcomes depend on the preparation and endpoint

In an eight-week randomized, double-blind trial, Andresen and colleagues assigned 443 adults with irritable bowel syndrome to heat-inactivated Bifidobacterium bifidum MIMBb75 or placebo. The composite primary endpoint required at least 30% improvement in abdominal pain together with adequate relief of overall symptoms in at least four of eight treatment weeks. It was reached by 74/221 participants versus 43/222, reported as rounded rates of 34% and 19%. The risk ratio was 1.7 (95% CI 1.3–2.4; p = 0.0007). This supports a symptom benefit for the specified inactivated preparation, rather than a shared therapeutic effect of all postbiotics. [6]

Metabolic findings also depend on the population and endpoint. A 2026 trial in 90 adults with overweight or obesity began with an eight-week low-energy diet and at least 8% weight loss, followed by a 24-week maintenance period. Weight regain was 1.2 kg with pasteurized A. muciniphila MucT and 3.2 kg with placebo (p = 0.012). The result concerns maintenance after diet-induced weight loss; it is not a direct estimate of weight loss without the preceding dietary intervention. [7]

A separate multicenter trial randomized 143 participants, of whom 142 received at least one administration of study product. After four months, the primary endpoint, whole-body insulin sensitivity assessed by the Matsuda index, did not differ significantly between groups. Signals in subgroups such as participants with low baseline Akkermansia levels were subsequent exploratory findings and cannot replace the primary result. A reported body-weight p value of 0.06 should not be described as a significant reduction. These trials address different questions and must retain their respective conditions. [8]

Processing defines material identity and application value

Inactivation is more than reducing a viable count to zero. The strain, fermentation matrix, inactivation process and drying method jointly shape the cellular structures and soluble constituents in the final preparation. Materials made from the same named species may have different functions. Changing the matrix, process or strain of a clinically studied preparation requires a renewed connection between material characteristics and evidence of benefit. [1, 2, 5]

Quality assessment can be organized into three levels: establish strain and matrix provenance; verify inactivation status and compositional consistency using suitable methods; and determine whether health benefits are supported for the same preparation, relevant population and intended context. Inactivation may reduce the burden of maintaining cell viability, but it does not guarantee room-temperature stability, processing tolerance or a long shelf life. These are questions for product specifications and stability data. [1, 2]

The absence of live cells also does not establish suitability for every population. The source organism, cellular constituents, fermentation matrix and process residues remain part of safety assessment. Good short-term tolerability cannot substitute for long-term or population-specific data. The practical value of postbiotics lies in connecting an identifiable non-viable microbial preparation with reproducible quality and a measured host benefit, so that each claim has a defined scope. [1, 2, 6]

Table 2. What the endpoints measure