Probiotics Are Not Interchangeable
The term "probiotic" covers thousands of bacterial and yeast strains, yet marketing treats them as a single category. In reality, probiotic effects are profoundly strain-specific: Lactobacillus rhamnosus GG has different clinical evidence than Lactobacillus rhamnosus ATCC 7469, even though they share a species name. A 2018 systematic review in Frontiers in Medicine confirmed that swapping one strain for another, even within the same species, cannot be assumed to produce equivalent outcomes.
What Has Robust Evidence
A relatively small number of probiotic strains have strong evidence for specific, narrow indications. Saccharomyces boulardii CSMCN I-745 reduces antibiotic-associated diarrhoea risk. Lactobacillus rhamnosus GG shortens acute infectious diarrhoea in children. VSL#3 (a multi-strain formulation) has evidence in maintaining remission of pouchitis in post-surgical ulcerative colitis patients. Bifidobacterium infantis 35624 showed modest improvement in global IBS symptom scores in well-designed RCTs.
Where the Evidence Falls Short
For the vast majority of commercial probiotic products — the capsules, yoghurts, and fermented drinks lining pharmacy shelves — rigorous clinical trial evidence is either absent, inconclusive, or limited to small, poorly controlled studies. The European Food Safety Authority (EFSA) has not approved a single health claim for any probiotic product, and the FDA classifies most probiotics as dietary supplements rather than drugs, exempting them from pre-market efficacy testing.
Colony-Forming Units and Survival
Probiotic labels advertise colony-forming units (CFUs) in the billions, implying that more is better. However, most ingested probiotics do not colonise the gut. A landmark 2018 study in Cell showed that probiotic strains were detectable in stool but largely failed to engraft in the mucosal microbiome, and that their impact on the resident community varied enormously between individuals — some were "persisters" and others were "resisters" with no measurable effect.
Personalisation Is the Frontier
Emerging research suggests that baseline microbiome composition predicts whether a given probiotic will engraft or be rejected. This insight points toward personalised probiotic prescribing — matching strains to individual microbial ecosystems — but this approach remains experimental. Until validated prediction tools exist, the most honest advice is: choose strains with evidence for your specific condition, at the dose and duration tested, and be sceptical of broad health claims.
When Probiotics Can Cause Harm
In immunocompromised patients, critically ill individuals, or those with central venous catheters, probiotics carry a small but real risk of bacteraemia or fungaemia. Saccharomyces boulardii fungaemia and Lactobacillus bacteraemia have been reported in ICU settings. Additionally, some individuals experience worsened bloating and gas on probiotics, particularly those with SIBO, where adding more bacteria to an already overpopulated small intestine can amplify fermentation symptoms.