Interest in probiotics continues to grow as our understanding of the human gut microbiota deepens. We now know far more about how intestinal microbes interact with host physiology, immunity and metabolism than we did even a decade ago. Scientific evidence increasingly points to meaningful interactions between intestinal microbes and host health. However, transforming this knowledge into safe, effective, and compliant probiotic products remains a demanding process.
In the third webinar of Biosafe’s microbiome series, Professor Emeritus Atte von Wright provided a comprehensive overview of the scientific criteria for probiotics, the methods used to evaluate them, and the regulatory landscape governing probiotic products in the EU. The picture is promising, though far from simple. In the EU especially, probiotics sit at the intersection of microbiology, toxicology, formulation science and one of the strictest health claims regimes in the world. EU food health claims are governed by Regulation (EC) No 1924/2006, while novel strains or new uses may also trigger the Novel Food Regulation, Regulation (EU) 2015/2283.
What defines a probiotic? Fundamental requirements
A probiotic is easy to describe in principle and difficult to prove in practice. It is not enough for a strain to be interesting in the lab or biologically plausible on paper. To be viable as a product, it must fulfil several criteria simultaneously, spanning safety, biological function, technological feasibility, sensory performance, and demonstrated efficacy.
That is one reason the category can look deceptively simple from the outside. The science may begin with gut microbes, but product development quickly becomes multidisciplinary.
Safety comes first, and the burden depends heavily on the strain
For probiotic development, safety is always the first and most critical requirement. In the EU, many well-known microbial species benefit from Qualified Presumption of Safety or QPS, a system EFSA uses to streamline safety assessment for biological agents with an established record of safe use. For QPS organisms, the focus shifts to strain-level identity and any specific qualifications such as acquired antimicrobial resistance. EFSA’s current guidance on microorganism characterisation places strong emphasis on correct taxonomic identification, whole genome sequencing and assessment of traits relevant to safety, including AMR and virulence-related features.
For non‑QPS strains, safety evaluation becomes significantly more demanding. As Atte von Wright explained,this typically includes:
-
Whole genome sequencing (WGS) to confirm species and strain identity
-
Screening for virulence factors, toxin genes, and acquired antibiotic resistance
-
In vitro genotoxicity tests, such as a bacterial gene mutation assay (Ames test) and a mammalian cell micronucleus test
-
A 90‑day repeated‑dose oral toxicity study in rodents
-
Additional toxicological studies where justified
As Professor von Wright noted, these requirements are not without limitations:
“I am not a great friend of either of these tests (Ames & mammalian cell micronucleus), but they are in the guidebooks, they are the law of the Union, and you have to comply with them.”
Despite ongoing debate about how well current tests reflect real biological risk, such studies remain a regulatory necessity for non‑QPS organisms.
Surviving the gut is only part of the story
A probiotic that cannot survive the gastrointestinal tract is unlikely to do much at all. It must survive gastrointestinal conditions to exert any potential effect. Early evaluation typically relies on in vitro stress tests, including tolerance to gastric acidity, bile salts, and digestive enzymes.
Adhesion assays using intestinal epithelial cell lines, such as Caco‑2, are commonly applied to study interaction with the gut surface. However, in vitro data alone is not sufficient.
Human intervention studies remain the most relevant way to evaluate survival and persistence in vivo. Increasingly, advanced gastrointestinal model systems are also used to assess microbial behaviour and interactions with the resident microbiota, although they cannot yet fully replicate host–microbe crosstalk.
A good probiotic also has to behave like a product
The commercial reality of probiotics is that a promising strain is useless if it cannot be made, stabilised and delivered in a form people will actually consume. Beyond biological function, a probiotic must be manufacturable at scale. This requires:
-
Efficient growth in industrial conditions
-
Tolerance to formulation processes such as freezing, freeze‑drying, or spray‑drying
Closely linked to this is product stability. Probiotics are expected to remain viable in declared amounts throughout their shelf life, often beyond the sell‑by date. According to Professor von Wright, insufficient stability remains a frequent issue in commercial products:
“Very often the actual content of live bacteria is considerably lower than the declared content.”
Advanced formulation strategies, such as microencapsulation, are increasingly used to address this challenge, though optimisation remains largely empirical. There is no universal fix.
Sensory quality can end a project faster than biology
This is the part many technically minded teams underestimate. When probiotics are incorporated into foods, organoleptic properties become critical. Taste, aroma, texture, and mouthfeel can ultimately determine whether a product has any future at all.
Professor von Wright shared personal experience from development projects where scientifically promising strains were abandoned because of poor sensory performance:
“Some tasted tolerable. Most tasted horrible. That was the end of that development project.”
That may sound mundane next to genomics and host-microbe interaction, but it is a reminder that successful food innovation is always both biological and practical.
Efficacy is where things get difficult
After safety, efficacy is the decisive hurdle. It is also where many probiotic concepts start to unravel. Developers can build an evidence base in several ways, including:
-
In vitro antimicrobial activity against pathogens
-
Immunological assays measuring cytokines or immunoglobulins
-
Human intervention studies
-
Epidemiological studies in defined populations
-
Clinical trials, especially for pharmaceutical applications
While convincing effects have often been demonstrated in disease conditions, human studies remain the gold standard, particularly from a regulatory point of view. The regulatory value of that evidence depends heavily on the context in which the product is being positioned. This has been one of the central problems for probiotics in Europe. Evidence may look strong in a disease context and still fail to support a food claim.
Regulatory classification of probiotic products in the EU
In regulatory terms, probiotics do not form one neat category. In the EU, probiotic products may be marketed as three main regulatory categories:
-
Foods (including functional foods)
-
Food supplements
-
Pharmaceuticals (also known as live biotherapeutic products)
That classification determines everything that follows. The applicable regulatory pathway depends on intended use, composition, and claims.
Novel Food considerations
Novelty is one key dividing line. If a strain or its use was not established in the EU before 1997, it may be considered a novel food under Regulation (EU) No 2015/2283. Novel foods require pre‑market authorisation based on a detailed safety assessment.
Professor von Wright clarified that novelty often becomes relevant for next‑generation probiotics:
“If you introduce a strain that has never been used in food or as a supplement, it is pretty obvious that it will be considered novel.”
A recent example is Akkermansia muciniphila, authorised as a novel food for special medical purposes only when in pasteurised and non‑viable form, a case commonly referred to as a postbiotic.
Why have probiotic health claims failed at EFSA?
The EU maintains a strict demarcation between foods and medicines. Foods may not claim to prevent, treat, or cure disease.
Under Regulation (EC) No 1924/2006, only:
-
Article 13 functional claims and
-
Article 14 reduction‑of‑risk claims
are permitted for foods. To date, no probiotic‑associated Article 14 claim has been accepted by EFSA, except for yoghurt cultures improving lactose digestion.
Addressing an audience question, Professor von Wright explained:
“The main reason was that there was no really convincing evidence of how a healthy person would benefit from regular use of a probiotic.”
Many early submissions relied on studies in sick populations or on vaguely defined outcomes such as “modulation of the microbiota”, which EFSA did not consider a demonstrated health benefit.
That is also why the term probiotic has become contentious in EU food labelling, as it may be interpreted as an implicit, unsubstantiated health claim, even though some Member States have taken more permissive national approaches.
Probiotics as pharmaceuticals
Once a product is positioned around treatment or prevention of disease, it leaves the food space and enters pharmaceutical territory. Claiming therapeutic effects requires the probiotic to be regulated as a medicinal product. Such products undergo full evaluation of safety and efficacy and can be authorised via national, decentralised, mutual recognition, or centralised procedures.
In practice, this route is much more demanding and expensive, which helps explain why most probiotic innovation has tried to stay on the food side of the line. Most probiotic pharmaceuticals have been authorised at the national level, reflecting both development costs and limited commercial incentives compared to food products.
So is there any real hope for probiotics?
Professor von Wright expressed cautious optimism:
“When we understand the microbiota–host interaction well enough, reduction‑of‑risk type claims should become acceptable.”
The scientific foundations are much stronger than they were during the first wave of probiotic enthusiasm. With improved bioinformatics, mechanistic understanding, and study design, scientific knowledge of microbiota–host interactions continues to advance to new opportunities.
Over time, that may create a more credible basis for probiotic claims tied to clear physiological functions rather than vague microbiome language or overtly therapeutic outcomes. Until then, bringing a probiotic product to market remains a long‑term project, requiring careful coordination of microbiology, technology, toxicology, and regulatory strategy.
The field is moving, but the gap between scientific promise and regulatory success remains real. Support is available from academic institutions and specialised partners to help companies navigate probiotic or microbiome-related product development.
Key questions from the webinar audience
1. Why have probiotic health claims failed at EFSA?
Not due to lack of data, but the wrong type of evidence. Evidence has often been built for the wrong regulatory question.
Atte clarified that:
-
Early probiotic claims focused on effects in sick populations (IBD, IBS, infections)
-
EFSA required evidence of measurable benefits in healthy individuals
-
Strain characterisation was also insufficient at the time, but this is no longer a major hurdle
The core problem was the lack of convincing evidence showing how healthy people benefit from regular probiotic use. This directly explains why strong clinical results still failed in the food context.
2. Where is the line between “novel” and “traditional” probiotics?
History of use before 1997 is the key divider.
From Atte’s answer:
-
If a strain (or its use) was not consumed in the EU before 1997, it is likely a Novel Food
-
QPS‑listed strains used in traditional food matrices are generally not novel
-
Introducing a known strain into a new matrix or application may still trigger novelty
Novelty is assessed case by case, but history of consumption remains the simplest rule of thumb.
3. Are probiotics naturally present in foods automatically safe?
Usually yes but safety is not the bottleneck.
Traditional lactic acid bacteria, bifidobacteria and related groups are typically safe due to a long history of food use and especially where QPS status applies.
Safety concerns arise mainly in rare cases e.g. unexpected antibiotic resistance. However efficacy, not safety, is the main barrier for health claims, even for traditional strains. This is an important nuance many developers misunderstand.
4. Is there real hope for future probiotic health claims?
Cautious optimism with better tools and better targets.
Atte was notably optimistic here:
-
Modern bioinformatics allows much better strain characterisation
-
Improved understanding of microbiota–host interaction opens new endpoints
-
Innovative companies might succeed by focusing on clear physiological functions, not disease outcomes
Reduction‑of‑risk claims may become realistic once mechanistic understanding is strong enough.
