
Live microorganisms in a bottle — the category whose honest unit is not the product but the strain?
Probiotics are live microorganisms dosed for health benefit — the FAO/WHO definition, and the last word of it matters most: health benefit is claimed per STRAIN, never per category. The strain is the product: L. rhamnosus GG (the most-published strain in history — the antibiotic-diarrhea flagship), B. longum 35624 (the IBS strain our UTI and IBS pages reference), the GR-1/RC-14 pair (the urogenital strains), the De Simone 8-strain formulation (pouchitis maintenance — probiotics' hardest clinical result). Around this short list of proven performers sits a billion-dollar shelf of unnamed blends whose labels borrow the category's vocabulary and none of the strains' evidence — the borrowing rule this page exists to teach. Add the field's two honest complications: most strains are transient visitors (they work while passing through, and stopping stops the effect), and they are live organisms — safe for the healthy, a genuine caution in the immunocompromised and critically ill.
The definition with teeth (FAO/WHO, 2001): 'live microorganisms which, when administered in adequate amounts, confer a health benefit on the host.' Every clause is a consumer skill: LIVE (storage and expiration matter), ADEQUATE AMOUNTS (the studied dose, not an arms race), and HEALTH BENEFIT (demonstrated per strain, per indication — never by category). Your gut holds roughly 38 trillion bacteria; a capsule adds a few billion visitors to that nation. The visitors' value is in what they do while passing through.
The category's best general-population evidence: probiotics — L. rhamnosus GG and Saccharomyces boulardii most-supported — reduce the risk of diarrhea during antibiotic courses (Cochrane-backed), with the practical protocol: dose during the course and a week past it.
The 35624 file: randomized trials of B. longum 35624 report modest improvements in IBS composite scores, pain, and bloating — the evidence behind the ACG's conditional recommendation and the reason the strain's name appears on this estate's UTI and IBS pages.
Probiotics' strongest clinical result: the De Simone 8-strain formulation prevents pouchitis relapse after ileal pouch surgery — guideline-used, gastroenterology-partnered, the high-dose medical tier this page stocks as Visbiome.
The core mechanism: live strains occupying adhesion sites and producing lactic acid and bacteriocins — crowding and acidifying against pathogens during transit. The mechanism behind the prevention effects, honest about its scope: a passing influence, not a permanent garrison.
The 70% connection our immune page teaches: gut-associated lymphoid tissue samples the passing flora; specific strains modulate IgA output and regulatory T-cell signaling — the mechanism tier behind the modest immune claims.
The mood-strain frontier: specific pairs (L. helveticus R0052 + B. longum R0175 — the strains our mood pages reference) show early trial signals for stress-adjacent measures, and the vagus-nerve traffic is established physiology — an early-tier, honestly-sized field.
The mechanism's honest headline is the transit fact: <strong>most commercial strains do not permanently colonize</strong>. They adhere, act, and pass — which explains everything practical about the category: why dosing is daily, why stopping stops the effect, why benefits measured in trials fade in washout weeks, and why 'rebuild your gut' is the aisle's most misleading phrase. The Weizmann studies sharpened the point further: even after antibiotics, a standard multi-strain product's colonization of the gut wall was limited and highly individual — the visitor model with personalization on top. What the visitors DO while passing is real: competitive exclusion, acid and bacteriocin production, immune signaling, and — for the strains with trials — the indication-specific benefits the evidence chapter grades. The strain is the product because every step of this mechanism is strain-specific: survival, adhesion, chemistry, and signaling all differ strain to strain.
Shelf-stable, strain-named — LGG, 35624
The medical tier — De Simone 8-strain
Antibiotic-compatible — the antibiotic can't kill it
The candy-math category this chapter retires
L. rhamnosus GG. B. longum 35624. L. rhamnosus GR-1 + L. reuteri RC-14. The letters and numbers after the species name ARE the product — the specific organism with the specific trials. The deep check: registry designations (ATCC numbers) on the most rigorous labels. This page's three products are strain-named because that is the category's only honest currency.
'Lactobacillus blend for digestive health' carries the category's name and none of its evidence — the same borrowing rule our UTI page teaches for urogenital strains and our IBS story teaches for 35624. If the label won't name the strain, the trials can't either. Put the bottle down.
Start with the shape of the field, because the shape IS the lesson. Probiotics are a $70-billion global market whose products number in the thousands — and whose evidence list is short: a handful of strains with genuine human trials for specific indications (LGG for antibiotic-associated diarrhea, 35624 for IBS, GR-1/RC-14 for urogenital health, the De Simone formulation for pouchitis, S. boulardii for the antibiotic course) — surrounded by an enormous periphery of products that reference the category's reputation without carrying any of its named strains. The consumer skill this page teaches is the strain-first read: before price, before CFU count, before any marketing sentence, find the strain name — because the evidence lives there or nowhere.
Then the regulatory oddity that explains the label chaos: probiotics are live organisms regulated as dietary supplements — no efficacy pre-approval required, no standardization enforced, the word 'probiotic' itself carrying no legal threshold. A drug-like product (a live biotherapeutic, in FDA's own terminology when the same organisms are developed for medicine) sold in a supplement frame — the gap that the strain rule exists to bridge. The named-strain products on this page are honest because their manufacturers chose to tie themselves to specific organisms with specific trials; the unnamed-blend shelf exploits the gap the same choice leaves open.
And the safety frame, stated once and completely: live organisms behave like live organisms. For the healthy population, the studied strains are among the safest supplements on earth — decades of use, benign tolerability. For the immunocompromised, the critically ill, and central-line patients, the calculus changes: rare bloodstream infections with commercial strains are documented, and the PROPATRIA trial in predicted severe pancreatitis — where a multi-strain product increased mortality — stands as the category's harm card. 'Generally safe' is a population statement, not a universal one; the physician conversation is the immunocompromised reader's standing rule.
The honest negative space — the probiotic aisle's five favorite inflations: