Research at a Glance
The AGA found that probiotics work for a few specific conditions with specific strains — but the vast majority of probiotic products lack evidence for their marketed uses.
Americans spend $9 billion a year on probiotics. The American Gastroenterological Association — the medical society of the doctors who actually treat gut disease for a living — published a comprehensive review of the evidence in 2020 and recommended probiotics for exactly three conditions: preventing necrotizing enterocolitis in preterm infants, preventing C. difficile infection in adults on antibiotics, and treating pouchitis. For IBS, Crohn's disease, ulcerative colitis, and acute pediatric gastroenteritis, the AGA explicitly recommends against routine probiotic use or notes insufficient evidence.
That is the most rigorous review of the entire category, and it concludes that the evidence base supports probiotics for three things. Everything else on the label is either marketing or extrapolation from research the manufacturer hopes you will not check. This article walks through what the evidence actually supports, why strain specificity matters more than CFU count (and why that distinction matters for which product you should actually buy), and how to evaluate a probiotic if you decide you want one.
The AGA's three approved use cases
The AGA's clinical practice guidelines on probiotics — developed using GRADE methodology, which is the standard for evidence-based medicine — identified three conditions where the evidence supports probiotic use:
- Preventing necrotizing enterocolitis in preterm, low-birth-weight infants. Multiple specific strain combinations (notably L. reuteri DSM 17938, B. lactis Bb-12, and Lactobacillus + Bifidobacterium combinations) showed significant reductions in NEC incidence in randomized trials.
- Preventing C. difficile infection in adults and children taking antibiotics. The strongest evidence is for Saccharomyces boulardii CNCM I-745 (Florastor) and a multi-strain combination of Lactobacillus acidophilus + L. casei + L. rhamnosus. The Szajewska 2015 meta-analysis showed a 53% relative reduction in antibiotic-associated diarrhea across 21 RCTs and 4,780 participants.
- Treating pouchitis after ileal pouch anal anastomosis surgery in inflammatory bowel disease patients. The VSL#3 multi-strain combination (now sold as Visbiome in the US) is the supported product.
Notice what is on this list and what is not. Specific strains, specific doses, specific use cases, supported by randomized trials. Not 'probiotics for gut health' or 'probiotics for IBS' or 'probiotics for immune support.' The AGA explicitly recommended AGAINST routine probiotic use for IBS, Crohn's disease, ulcerative colitis, and acute pediatric gastroenteritis — the conditions probiotics are most often marketed for.
If you are taking a daily probiotic for general gut health and you are not in one of the three AGA-approved categories, you should know that the AGA's position is that the evidence does not yet support that use. That does not mean you are wasting your money — it means the evidence base is thinner than the marketing implies, and the placebo effect in this category is large.
Strain specificity: why CFU count is the wrong metric
The marketing on probiotic packaging emphasizes one number: CFU count. 'Contains 30 billion CFU!' or '50 billion CFU per capsule!' or, increasingly, '100 billion CFU!' This is a vanity metric. The research evidence is built on specific strains, not on bulk bacteria. A 10 billion CFU dose of Lactobacillus rhamnosus GG (the most studied probiotic strain in human history, with 250+ published clinical trials) is worth more than a 100 billion CFU blend of anonymous bacteria with no individual evidence base.
The distinction that matters: strains are identified by genus, species, AND a specific strain identifier (a code like 'GG' or 'CNCM I-745' or 'BB-12'). A label that lists 'Lactobacillus acidophilus' is telling you the genus and species but not which strain — and there are thousands of L. acidophilus strains, most of which have no clinical evidence at all. A label that lists 'Lactobacillus acidophilus NCFM' is telling you the specific strain (NCFM) that has been studied in published trials. The difference is the difference between buying a researched probiotic and buying random bacteria.
When you read a probiotic label, look for full strain identifiers. If the label gives you genus + species but no strain ID, the marketing is hiding what you actually need to know. Cross-reference any specific strain you find against PubMed or examine.com to see whether the strain has been clinically studied for the condition you are trying to address. Most strains in mass-market probiotics have not been.
The other thing nobody talks about: delivery system
Stomach acid is hostile to most probiotic bacteria. The pH of an empty stomach is around 1.5 to 2.5 — acidic enough to denature most proteins and kill most live cells. A probiotic capsule that dissolves in the stomach delivers its bacteria directly into that acid bath, and the survival rate of the live bacteria through the gastric phase of digestion is the difference between a probiotic that works and a probiotic that is dead by the time it reaches the gut.
The good brands address this with delivery technology. Delayed-release capsules use enteric coatings that resist stomach acid and only break down in the higher pH of the small intestine. Some brands (Seed in particular) use a two-capsule design where an outer capsule contains a prebiotic and protects an inner capsule that contains the probiotic. Saccharomyces boulardii (Florastor) is a yeast rather than a bacterium and naturally survives stomach acid much better than bacterial probiotics — that is one of the reasons S. boulardii has the strongest published evidence for antibiotic-associated diarrhea prevention.
Brands that ship probiotics in standard capsules without delayed-release technology are betting that enough bacteria will survive the stomach to do something useful. Sometimes that is true. Often it is not. The published gastric survival data for delayed-release products is meaningfully better than for standard capsules.
Refrigeration and shelf-stability honesty
Some probiotic strains require refrigeration to remain viable. Others are shelf-stable. Brands that ship room-temperature shelf-stable products containing strains that need cold are misleading customers — by the time you take the capsule, the live bacterial count has dropped significantly from what the label claims at time of manufacture.
The CFU count on the label is also worth scrutinizing. Many brands list CFU count 'at time of manufacture,' which means the number is highest the moment the capsule was filled and drops over the shelf life of the product. The good brands guarantee CFU 'through expiration,' which means the number on the label is what you actually get when you take it. The bad brands let you believe the manufacture-time number applies to you, and it does not.
Six brands, ranked by what actually matters
Seed Daily Synbiotic DS-01 ($49.99/month) is the most scientifically defensible probiotic on the consumer market. 24 strains published at full strain-ID level. Patented two-capsule ViaCap delivery system with published gastric survival data. The only major DTC brand running randomized placebo-controlled trials on its actual finished formula. Expensive, but if you want a probiotic that treats the category like science rather than marketing, this is it.
Pendulum Glucose Control ($99/month membership) is the only consumer probiotic with a published randomized, double-blind, placebo-controlled trial showing a meaningful clinical outcome — a 0.6% A1c reduction in type 2 diabetics on metformin. Contains Akkermansia muciniphila WB-STR-0001, which no other consumer brand has commercialized. Not a general-purpose probiotic — this is targeted medical nutrition for one specific population.
Florastor (~$40/month) is the only probiotic the AGA actively recommends by name for a specific use case: preventing C. difficile and antibiotic-associated diarrhea in adults on antibiotics. Contains Saccharomyces boulardii CNCM I-745, which is a yeast rather than a bacterium and survives stomach acid better than bacterial probiotics. 27+ RCTs and 84% efficacy in antibiotic-associated diarrhea prevention per meta-analysis. Narrow use case; strong evidence within it.
Culturelle Digestive Daily (~$20/month) is the best value in the category. Contains Lactobacillus rhamnosus GG, the single most studied probiotic strain in human history with 250+ clinical trials and strong evidence for preventing antibiotic-associated diarrhea. You won't get the strain diversity of Seed, but you will get one strain that actually has evidence at a quarter of the price.
Ritual Synbiotic+ ($54/month) is Seed's most direct competitor — built around two well-studied branded strains (LGG and BB-12) in a 3-in-1 probiotic + prebiotic + postbiotic format. The honest critique is that until 2024 Ritual had run zero clinical trials on the actual Synbiotic+ product, only cited general strain research. Their first product-level trial is now ongoing, which will be meaningful evidence whenever it publishes.
Garden of Life Dr. Formulated (~$30/month) is the grocery-aisle default. Widely available, reasonably priced, 30 billion CFU shelf-stable. The honest problems: leans heavily on high-CFU marketing without strain-level specificity, no published clinical trial on the finished product, and many strains listed without full strain IDs. Fine if your bar is 'a basic probiotic at Whole Foods.' Doesn't hold up to the strain-specificity test the premium brands pass.
The honest takeaway
Probiotics are low-risk and occasionally helpful. They are also one of the most evidence-mismatched categories in the supplement aisle, and the AGA's review is a useful corrective to the marketing. If you fall into one of the three AGA-supported use cases (NEC prevention in preemies, antibiotic-associated diarrhea prevention, pouchitis), the evidence is strong and you should pick a product whose strains match the published research. If you are taking a daily probiotic for general gut health, the case is weaker — the placebo effect in this category is large, the published evidence for general benefits is thin, and the specific product you pick matters more than the category does. In either case, look for full strain IDs, finished-product clinical trials, delivery systems that survive stomach acid, and refrigeration-honest labeling. Skip anything marketed primarily on CFU count.
Frequently Asked Questions
For three specific use cases, yes — the AGA's 2020 clinical practice guidelines recommend probiotics for preventing necrotizing enterocolitis in preterm infants, preventing C. difficile in adults on antibiotics, and treating pouchitis. For everything else (IBS, Crohn's disease, ulcerative colitis, general 'gut health,' immune support, mood) the AGA's position is that the evidence does not yet support routine use. That does not mean probiotics are useless — it means the evidence base is thinner than the marketing implies, and the placebo effect is large in this category. If you are taking a daily probiotic for general gut health, you are operating outside the evidence-supported use cases.
Less than the marketing wants you to believe. CFU count (colony-forming units, the live bacteria count per dose) is the number probiotic brands emphasize because it is easy to understand and easy to make impressive. But the research evidence is built on specific strains, not on bulk bacteria. A 10 billion CFU dose of Lactobacillus rhamnosus GG — the most studied probiotic strain in human history with 250+ clinical trials — is worth more than a 100 billion CFU blend of anonymous bacteria. Strain specificity beats CFU count. If a brand emphasizes '100 billion CFU!' without telling you which specific strains are in the blend or why, that is a red flag, not a feature.
Strains are identified by genus, species, AND a specific strain code. A label that says 'Lactobacillus acidophilus' tells you the genus and species but not which strain — and there are thousands of L. acidophilus strains with different properties and clinical evidence. A label that says 'Lactobacillus acidophilus NCFM' tells you the specific strain (NCFM) that has been studied in published trials. The difference is the difference between a researched probiotic and random bacteria. When you read a probiotic label, look for full strain identifiers like 'GG' (Lactobacillus rhamnosus GG, the most studied strain), 'BB-12' (Bifidobacterium lactis BB-12), 'CNCM I-745' (Saccharomyces boulardii CNCM I-745, the strain in Florastor), or 'NCFM.' If the label gives you genus + species but no strain code, the brand is hiding what you actually need to know.
Stomach acid is hostile to most probiotic bacteria — the pH of an empty stomach (1.5-2.5) is acidic enough to kill most live cells. A probiotic capsule that dissolves in the stomach delivers its bacteria into that acid bath, and the survival rate through gastric digestion can be very low. The good brands address this with delayed-release capsules (enteric coatings that resist stomach acid and break down in the higher pH of the small intestine), two-capsule designs (Seed's ViaCap), or naturally acid-tolerant species like Saccharomyces boulardii (Florastor) which is a yeast rather than a bacterium. Brands that ship probiotics in standard capsules without delayed-release technology are betting that enough bacteria will survive — sometimes that is true, often it is not.
The AGA's 2020 clinical practice guidelines on probiotics — developed using GRADE methodology, the standard for evidence-based medicine — recommend probiotics for exactly three conditions: (1) preventing necrotizing enterocolitis in preterm infants, (2) preventing C. difficile infection in adults on antibiotics, and (3) treating pouchitis after IBD surgery. For IBS, Crohn's disease, ulcerative colitis, and acute pediatric gastroenteritis, the AGA explicitly recommends AGAINST routine probiotic use or notes insufficient evidence. The AGA's official position is that probiotics are not supported for most of the conditions they are most commonly marketed for. This is the most rigorous review of the category and a useful corrective to the marketing claims you see on probiotic labels.
