Walk into any pharmacy and ask for a probiotic. You will be handed something that says “10 billion CFU” or “30 billion CFU” on the front of the box, the assumption being that bigger is better. That framing is one of the most expensive mistakes in supplement merchandising. The colony-forming-unit count tells you almost nothing about whether a product will do what you want for the patient in front of you. The strain does. Two products with identical CFU counts can produce dramatically different clinical outcomes — or no outcome at all — based on which specific bacterial strains they contain.
Genus, species, strain — why all three matter
Probiotic bacteria are identified at three taxonomic levels. The genus is the broad family (e.g. Lactobacillus). The species is the subgroup within the genus (Lactobacillus rhamnosus). The strain is the specific isolate, often denoted by a letter-number code (Lactobacillus rhamnosus GG). The clinical evidence — virtually all of it — is strain-specific. A trial showing L. rhamnosus GG reduces antibiotic-associated diarrhoea does not generalise to other L. rhamnosus strains. The trial of Saccharomyces boulardii CNCM I-745 for traveller’s diarrhoea does not transfer to a different S. boulardii isolate.
This is not a pedantic technicality. Strain-specific gene expression determines whether a bacterium can survive stomach acid, adhere to the intestinal mucosa, produce specific metabolites, or modulate the immune system in the way the trial demonstrated. A product that lists only “Lactobacillus rhamnosus, 10 billion CFU” without a strain identifier is — clinically — a black box. You have no way to know which behaviour the bacteria will exhibit.
The strains worth knowing
A small set of well-studied strains accounts for most of the credible clinical evidence in probiotic therapy. Knowing them by name is one of the easier high-value upgrades a practitioner can make.
Lactobacillus rhamnosus GG (LGG)
The most-studied probiotic strain in the world. Strongest evidence for prevention of antibiotic-associated diarrhoea, acute infectious diarrhoea in children, and mucosal barrier support. Typical clinical dose: 10–20 billion CFU/day.
Saccharomyces boulardii CNCM I-745
A yeast, not a bacterium, which means it is unaffected by concurrent antibiotic use. Robust evidence for traveller’s diarrhoea, prevention of C. difficile recurrence, and adjunctive treatment of H. pylori. Typical dose: 5–10 billion CFU/day.
Bifidobacterium infantis 35624
Notably effective in IBS, particularly the bloating and abdominal discomfort variants, in multiple randomised trials. Typical dose: 1 billion CFU/day — much lower than the headline CFU counts on most products, but specifically what the trials used.
Lactobacillus reuteri DSM 17938
Strong evidence in paediatric infantile colic and functional abdominal pain in children. Typical dose: 100 million CFU/day for paediatric use.
Multi-strain blends (VSL#3 / Visbiome formulation)
An eight-strain formulation with the strongest evidence base in ulcerative colitis remission maintenance and pouchitis. Typical clinical dose: 450 billion CFU/day during active intervention. This is one case where the high CFU count is genuinely required by the clinical trial protocol.
What CFU count actually means (and does not)
The colony-forming-unit count tells you how many viable bacterial cells were in the product at time of manufacture. It does not tell you how many are still alive at the patient’s small intestine — many strains die during stomach transit. It does not tell you how many will adhere and colonise (most do not; most probiotics transit). It does not tell you anything about clinical effect.
The trial doses for the strains above range from 100 million to 450 billion CFU/day — a 4,500-fold difference. There is no universal “more is better” rule. Match the dose the strain was studied at. Recommending 50 billion CFU of B. infantis 35624 because “more is better” wastes the patient’s money and does nothing clinically that 1 billion would not.
How to read a probiotic label
A useful label, the kind worth recommending, includes four things. Strain identification — genus, species, AND the strain code (the letter-number suffix). CFU count at expiration, not at manufacture (the difference can be 50% or more over an 18-month shelf life). Storage requirements — refrigerated vs. shelf-stable, and what happens to the strain count at room temperature. Independent verification — third-party testing or the manufacturer’s stated identity-confirmation protocol. A label that gives you only genus + species + a big CFU number on the front is, in clinical terms, marketing copy. Move on.
The clinical thinking that follows from strain specificity
Once strain matters as much as it does, the questions you ask a patient change. “Are you taking a probiotic?” is the wrong question. “Which strain, at what dose, for what symptom, for how long?” is closer. A patient on an unspecified probiotic for “general gut health” is almost certainly getting placebo at best — and the practitioner can usefully reframe the conversation: either match a specific strain to a specific clinical issue, or take the supplement out of the plan and use the spend on actual food-based gut work (fermented foods, fibre diversity, polyphenols).
This thinking extends naturally to other supplement categories, but probiotics are where it is most visible because the gap between marketing and evidence is widest. We covered the foundational gut-health framework in our earlier piece on IBS, leaky gut, and microbiome testing — that piece pairs well with this one as a paired reference for any practitioner working seriously in the gut-health space.
Probiotic recommendation is a place where practitioner expertise produces enormous patient value — because the patient cannot evaluate strain quality on their own and the marketing universally misleads. A clinician who recommends L. rhamnosus GG at 10 billion CFU/day for 14 days post-antibiotic rather than “a probiotic” is doing the kind of specific, evidence-anchored work that distinguishes a real clinical practice from a wellness recommendation. At Horta Nutrition, we manufacture strain-specific probiotic formulations under our nutraceutical contract programme with full strain identity verification, CFU-at-expiration labelling, and stability data per batch — so practitioners can recommend products whose actual contents they can vouch for. If you want to discuss strain selection for a specific clinical use case, the Horta practitioner network includes a peer channel for protocol consultation.
