Probiotics have solid evidence for a short list of uses and thin evidence for most others: a 2017 randomized trial in the Lancet found that a multi-strain probiotic did not prevent antibiotic-associated diarrhea any better than placebo in older hospitalized adults, yet systematic reviews of other formulations and populations — including a 2019 Cochrane-family review cited by the American Gastroenterological Association — concluded that certain strains, notably Lactobacillus rhamnosus GG and Saccharomyces boulardii, modestly reduce the risk of antibiotic-associated and C. difficile diarrhea in children and adults. The gap between those two findings is the probiotic story in miniature: effects are strain-specific, condition-specific, and often smaller than advertising suggests.
Newspaper Daily publishes information, not medical advice. Probiotic products in the United States are regulated as foods or dietary supplements rather than as drugs, meaning they are not reviewed for effectiveness before sale, and anyone with a weakened immune system, a central line, or a serious illness should speak with a physician before using them.
What exactly are probiotics?
The internationally accepted definition, adopted by a FAO-WHO expert consultation and updated by the International Scientific Association for Probiotics and Prebiotics in 2014, describes probiotics as live microorganisms that, when administered in adequate amounts, confer a health benefit on the host. Commonly used genera include Lactobacillus, Bifidobacterium, Saccharomyces (a yeast), Streptococcus thermophilus, and certain Bacillus strains.
Three consequences of that definition matter for consumers. First, "adequate amounts" is dose-specific: benefits demonstrated in trials apply to the doses used in those trials, and billions of colony-forming units in one product do not transfer to a different strain count in another. Second, strain matters more than species — Lactobacillus rhamnosus GG is not interchangeable with other L. rhamnosus strains. Third, the definition requires live organisms, which excludes most dead-cell products and raises storage questions for shelf stability.
What does evidence support using probiotics for?
Based on systematic reviews and guideline statements through early 2026, the uses with the strongest support include:
- Prevention of antibiotic-associated diarrhea and reduction of C. difficile risk, particularly with L. rhamnosus GG and S. boulardii in children, per reviews summarized by the American Gastroenterological Association
- Shortening the course of acute infectious diarrhea in children by roughly one day, per a 2019 review in JAMA Pediatrics of 82 randomized trials
- Reducing the risk of necrotizing enterocolitis in preterm infants, an in-hospital use supported by multiple randomized trials and endorsed in neonatal guidance abroad
- Management of infantile colic with certain strains, with modest and mixed evidence
- Ulcerative colitis pouchitis maintenance with the multi-strain preparation VSL#3-class products, per gastroenterology literature
Even in supported categories, reviewers consistently grade effect sizes as modest and heterogeneity as high — meaning results vary by strain, dose, and population, and no effect is guaranteed for an individual.
Where does the evidence fail or fall short?
The cautionary record is instructive. The 2017 Lancet trial mentioned above found a multi-strain product no better than placebo for preventing antibiotic-associated diarrhea in elderly inpatients. In irritable bowel syndrome, practice guidance from the American College of Gastroenterology in 2018 and 2021 notes that individual probiotic trials show modest or no benefit and stops short of recommending specific products. In healthy adults, a 2018 review of evidence for "gut health" benefits in people without a condition found the case weak, and a well-publicized 2018 Cell pair of studies found that colonization after a course of probiotics varied dramatically between individuals and, in one arm, actually delayed the return of the native microbiome after antibiotics.
For the marketing claims that drive most purchases — general immune boosting, weight loss, mood improvement, skin health — the consensus among reviewers, including a 2020 assessment in JAMA, is that evidence is preliminary, strain- and study-specific, and insufficient for broad claims. The microbiome-gut-brain field has produced genuinely interesting early trials, but as of 2026, no probiotic has drug-level evidence for treating depression or anxiety.
Related stories: Creatine: What the Supplement Evidence Supports · Mediterranean Diet: What the Evidence Actually Shows.
How are probiotics regulated, and why do labels differ from drugs?
Most probiotics sold to consumers in the United States are foods or dietary supplements under the Dietary Supplement Health and Education Act, so manufacturers cannot legally claim to treat or cure disease but face no pre-market efficacy review. The FDA has approved only a handful of biological-drug probiotics for specific uses — the fecal-microbiota products Rebyota (2022) and Vowst (2023) for recurrent C. difficile sit adjacent to the category — and import alerts have targeted preterm-infant probiotics marketed without authorization after infections in 2023 prompted an FDA safety alert. Consumers can check whether a product carries third-party certification, but certification verifies contents, not benefits.
Are probiotics safe?
For healthy people, trials report a generally benign profile dominated by mild gas and bloating. The exceptions define the safety boundary: the FDA's 2023 warning to clinicians followed reports of invasive disease and one infant death linked to a probiotic product in preterm infants; fungal and bacterial infections from probiotic organisms have been reported mainly in immunocompromised, catheterized, or critically ill patients; and guidance in hospital settings increasingly reserves probiotics for patients who are not critically ill. S. boulardii, a yeast, carries theoretical concern in people with central venous catheters.
Do fermented foods count as probiotics?
Often not technically. Yogurt with live cultures, kefir, kimchi, and sauerkraut contain live microbes, and some yogurts carrying the L. rhamnosus GG strain meet the probiotic definition; but many fermented foods contain uncharacterized microbial communities at unknown doses, and the 2014 ISAPP consensus distinguishes them from probiotics unless strains and doses are documented. Interest in fermented foods grew after a 2021 Stanford randomized trial found a fermented-food diet increased microbiome diversity and lowered inflammatory markers — an intriguing small-trial result, distinct from commercial probiotic capsules.
What about prebiotics and synbiotics?
Prebiotics — fibers such as inulin that feed resident bacteria — occupy a parallel evidence track, with documented effects on bowel function and calcium absorption at studied doses and bloating as the main side effect. "Synbiotic" products combine the two. ISAPP's 2017 and 2021 consensus papers formalized these definitions; the products' evidence base, again, is formulation-specific rather than general.
When to see a doctor or registered dietitian
Professional evaluation should precede probiotic use when diarrhea is severe, bloody, or accompanied by fever, since C. difficile and invasive infections require medical treatment rather than supplements; when symptoms persist more than a few days despite self-care; when immune suppression, recent surgery, a central line, or critical illness is present; in preterm or very young infants, outside of physician-directed care; and whenever medication interactions are possible, since probiotic organisms can interact with antifungal and antibiotic timing. Persistent digestive symptoms also warrant screening for celiac disease and other conditions that probiotics will not address.
The bottom line on what probiotics can and cannot do
The trial record supports probiotics for specific, narrow uses — chiefly preventing antibiotic-associated diarrhea with particular strains, shortening childhood infectious diarrhea, and hospital-level use in preterm infants — and does not support the broad immunity, mood, and weight claims common on packaging. Effects are strain- and dose-specific, regulation treats most products as foods rather than drugs, and safety has real limits in vulnerable patients. As of April 2026, the most defensible summary remains the one researchers have repeated for years: strong claims require the strain, dose, and condition to match the studies.