Place · Level 3

Probiotics

全球年销 $70B+ · 但广谱益生菌概念几乎不成立 · 真正循证的只有几个菌株 + 几个适应症

Story path

1ISAPP definition · strain mattersISAPP definition · strain matters
2Acid massacre · 0.1-1% surviveAcid massacre · 0.1-1% survive
3Evidence tiers · indication listEvidence tiers · indication list
4Fermented foods · daily probioticsFermented foods · daily probiotics
5FMT · the gold standardFMT · the gold standard
6Practical · should I / howPractical · should I / how

Chapter 1

ISAPP definition · strain matters

The WHO/FAO 2002 + ISAPP 2014 official definition of a probiotic: 'live microorganisms that, when administered in adequate amounts, confer a health benefit on the host'.

Four key words in that definition; each one decides whether 99% of commercial products even qualify.

① 'Adequate live amount': most supplement labels show CFU (colony-forming units); clinically effective doses are typically 10⁹-10¹⁰ CFU/day. Critically, the manufacture-date count is not the end-of-shelf-life count — most strains lose viability over time and require refrigeration, dryness, or enteric coating. Spot checks routinely find labels claiming 10 billion CFU with only 1-10% remaining at end of shelf life.

② 'Strain' ≫ 'species': the species is *Lactobacillus rhamnosus*; the strain is *Lactobacillus rhamnosus* GG (LGG). Clinical effect is strain-specific — LGG has strong evidence for preventing pediatric acute diarrhea, but another strain of the same species may be completely useless. This is the root of the 'lactic acid bacteria' confusion: vendors say 'contains probiotics' without naming the strain, which usually means an inexpensive strain with no clinical data.

③ 'Confers a health benefit on the host': requires human RCT evidence for that strain + that specific indication. Not 'theoretically beneficial', not 'improves a biomarker' — clinically meaningful improvement in a defined disease or symptom.

④ 'Live': dead organisms aren't probiotics by definition, though the *postbiotic* concept is gaining ground. This is why hot water, acidic drinks, and high-heat cooking discount the probiotic value of fermented foods.

So to judge whether the bottle on your shelf is real, check three things:

1. Does it name a specific strain (LGG / BB-12 / DSM xxxxx, etc.)?
2. Does that strain have an RCT for your indication?
3. Is the labeled CFU the minimum at end of shelf life, or just the manufacture-date number?

Star strains cheat sheet

A working list of RCT-supported strains + indications — not exhaustive but it covers 90% of real clinical use.

Lactobacillus genus:

L. rhamnosus GG (LGG): pediatric acute viral diarrhea (cuts duration by ~1 day, Cochrane strong evidence); prevention of eczema in high-allergy-risk infants (*Kalliomäki 2001* *Lancet*)L. rhamnosus GR-1 + L. reuteri RC-14: adjunct for recurrent UTI / vaginosis in women, moderate evidenceL. reuteri DSM 17938: infant colic, some positive RCTsL. casei Shirota (Yakult): general gut function — marketing far outruns clinical evidence
Bifidobacterium genus:

B. lactis BB-12 / DN-173 010: general gut motility, partial symptom improvement in IBSB. infantis 35624 (Align): one of the mainstream IBS recommendations (Ford 2018 AGA)
Saccharomyces (yeast, not bacterium):

S. boulardii CNCM I-745: strong evidence for antibiotic-associated diarrhea (AAD) and traveler's diarrhea (Goldenberg 2017 Cochrane)Crucially it's a yeast — antibiotics don't kill it, so it can be taken concurrently with antibiotics
Multi-strain combinations:

VSL#3 (now renamed Visbiome / Vivomixx): 8-strain combo used for chronic pouchitis and some UC casesPediatric multi-strain combinations (various brands): mixed data
Categories without meaningful evidence:

Most supermarket 'digestive health' blends, labeled only as a *Lactobacillus / Bifidobacterium* blend with no specific strain identityThe 'active probiotics in yogurt' marketing line: most yogurts use *L. bulgaricus* + *S. thermophilus*, which barely survive to the colon — they help lactose tolerance but aren't true probioticsPrebiotics (fiber) feeding the bacteria you already have are often more effective than adding new ones
Before buying, ask three things: which strain (only consider if a strain ID is written); is there a human RCT for this strain + your indication (5 minutes on PubMed); is the labeled CFU end-of-shelf-life or manufacture-date?

Open this on the map

Chapter 2

Acid massacre · 0.1-1% survive

'Swallowing 10 billion live bacteria' and '10 billion reaching the colon' are two completely different orders of magnitude.

Stomach acid at pH 1.5-2.0 is lethal for most lactic-acid and bifido strains:

Typical *Lactobacillus* + *Bifidobacterium* strain: 0.01-1% gastric survival*S. boulardii* (yeast): much more acid-resistant, ~10-30% gastric survivalSpore-formers (*Bacillus coagulans* / *B. subtilis*): nearly 100% gastric survival, but clinical evidence is limited
Protection strategies:

Enteric coating: doesn't dissolve in acid, releases in the small intestine — effective delivered CFU rises 10-100×Take with meals: food buffers acid pH, but also dilutesHeat-killed / postbiotic preparations: some studies show dead cells still work, because cell-wall components (LPS, peptidoglycan) modulate immunity — but the concept is still evolving
Colonization is a separate concept — most probiotics do not establish permanent colonies in your gut:

The classic Sonnenburg + Bhatt 2018 *Cell* study: 11 probiotic strains given after antibiotics — most had vanished 3-4 weeks after stoppingBhatt + Suez 2019 *Cell*: probiotics in some individuals actually *delayed* the natural recovery of the gut microbiomeReality: probiotics are *transient passengers* — they pass through, leave short-term signals (immune modulation, SCFAs, interactions with resident microbes), and then get washed out
Three clinical implications:

1. Probiotics generally need continuous use to maintain an effect — it isn't 'set the gut once and benefit forever'.
2. Individual variation is huge — some people's gut environment lets specific probiotics temporarily settle; others reject them entirely.
3. Fecal microbiota transplant (FMT) is ~100× stronger than probiotics because it delivers the full community + metabolites + the 'home' (mucin / sIgA).

So claims like 'yogurt cures everything' or 'one probiotic capsule rewrites your microbiome' miss the real position of probiotics as 'transient + short-term modulators'.

Fermented food vs supplement

The Sonnenburg lab's 2021 *Cell* study (covered in the digestive story) is a high-quality RCT, N=39, 10 weeks.

Group A: high-fiber dietGroup B: high-fermented-food diet (yogurt, kefir, kimchi, sourdough, kombucha, vinegared vegetables)
Results:

Group B (fermented foods): microbiome diversity ↑; 19 inflammation markers fell across the board (, , IFN-γ, etc.)Group A (high fiber): limited improvement; some subjects' microbiomes couldn't ferment that much fiber (needed adaptation + slow ramp-up)
This result shocked the microbiome field: in theory fiber should win (directly feeds bacteria to produce SCFAs), but fermented foods won.

Mechanism hypothesis: fermented foods simultaneously deliver multiple live cultures, their metabolites (short-chain fatty acids, amino acids, vitamin K2, polyphenol derivatives), and cell-wall components (the postbiotic effect). This is the 'whole-food matrix' effect — not a single strain, but bacteria + metabolites + food matrix together; probiotic capsules deliver a single high-purity strain and lose this matrix.

Practical recommendations:

Rotate fermented foods as the first move: plain unsweetened yogurt, kefir, kimchi, sauerkraut, natto, miso, kombucha, vinegared vegetables1-2 different types daily works better than 'one cup of the same yogurt every day'Pair with fiber to feed your existing bacteria: oats, legumes, whole grains, vegetables, fruitUse supplements only for specific indications — AAD, IBS, recurrent UTI, acute pediatric diarrhea, and other RCT-supported scenariosDon't take probiotic capsules long-term for vague 'gut health' — fermented food + fiber + dietary diversity is stronger for almost everyone
This is why the section is titled 'Gastric massacre': the vast majority of probiotics die before reaching the battlefield. What actually tunes the microbial ecosystem is repeatedly giving the bacteria a good working environment (fiber, diversity, fewer antibiotics) — not dropping a single dose of hundreds of millions of soldiers in.

Open this on the map Watch the mechanism

Chapter 3

Evidence tiers · indication list

Probiotics aren't 'good for everything' — layering by RCT evidence strength makes the picture much clearer.

A-tier (strong RCT + meta-analysis):

Antibiotic-associated diarrhea (AAD) prevention: *S. boulardii* or LGG taken with the antibiotic — risk down ~50% (Goldenberg Cochrane 2017)Recurrent *C. difficile* infection: FMT is the gold standard (80-90% cure); specific probiotics work as adjunctPediatric acute viral diarrhea: LGG or *S. boulardii* shortens duration by ~1 day (Cochrane)Necrotizing enterocolitis (NEC) prevention in preterm infants: multi-strain combos cut risk by ~50% (Sun 2017 and others)
B-tier (RCT evidence but mixed or small effects):

IBS: specific strains (*B. infantis* 35624, multi-strain combos) give modest improvement — weak AGA recommendation (Ford 2018)Infant colic: *L. reuteri* DSM 17938 partially positiveAllergy / eczema *primary* prevention: LGG during pregnancy / infancy in high-risk families cuts incidence ~25%Adjunctive bacterial vaginosis: paired with antibiotics, reduces recurrence
C-tier (mixed, weak real-world impact):

'Immunity' / common-cold prevention: weakly positive small RCTs, effect size smallAllergic rhinitis: weak evidenceTreating (not preventing) atopic dermatitis: limited effectDepression / anxiety ('psychobiotics'): early evidence is interesting but RCTs are few and effects smallAthletic performance: only 1-2 small RCTs
No meaningful evidence but widely marketed: weight loss (no strain has been shown to lower body weight); anti-aging / longevity (no human RCT); autism spectrum (animal studies yes, human RCTs null); beauty / skin glow (marketing); diabetes prevention / glucose control (small signal, weak).

Important warnings:

Critically ill / immunosuppressed / central venous line patients: probiotics can cause bacteremia (rare but real) — use cautiously in ICUSevere pancreatitis: the 2008 PROPATRIA *Lancet* trial showed a multi-strain probiotic doubled mortality — severe illness is not a 'just add bacteria' situationPregnancy / infants: most evidence comes from LGG / BB-12 with long safety records; don't substitute random other strains
To decide whether a probiotic is worth buying, ask three things:

1. Is my indication on the A-B list?
2. Does it contain a specific strain with an RCT — not a generic 'broad-spectrum' blend?
3. Could I substitute fermented foods + fiber?

If none of the three apply, the money is better spent at the produce market.

Label traps · billion bacteria fog

Common label traps.

'10 billion live bacteria': is that the manufacture-date number or the end-of-shelf-life number? Most cheap products overstate at manufacture; 6 months later only 1-10% remains. Good labels write 'At expiration: X CFU' or include refrigeration instructions.

'16-strain blend': sounds more comprehensive, but strains often compete / inhibit each other, and per-strain CFU gets diluted; clinical RCTs are almost all single-strain + single-indication, not 'the full bouquet'. VSL#3 / Visbiome is one of the few multi-strain combos with RCT evidence.

'Now with prebiotic — upgraded version': prebiotics (fiber / FOS / inulin) work on their own, but there's almost never data on whether they synergize with the specific probiotic strain you're buying. Eating more fiber on its own is much cheaper.

'Clinically proven' / 'Doctor recommended': with no FDA or major medical-society evaluation behind it, this is marketing rhetoric. The real consensus that exists: the AGA 2020 guideline recommends probiotics for AAD prevention, NEC prevention, and pouchitis — and explicitly does *not* recommend them for general IBS, IBD, *C. diff* prevention, or generic 'gut health'. WGO and ESPGHAN positions are similar.

'Live bacteria / probiotic / postbiotic' concept upgrades: heat-killed cells or bacterial metabolites — theoretically interesting, but RCTs are scarce; don't pay a premium for a concept upgrade.

Cold-storage vs shelf-stable: cold-storage products typically have higher activity but are harder to travel with; shelf-stable products use spore-formers (*Bacillus*) or lyophilization — convenient, but some strains lose potency relative to their refrigerated equivalents. Pick the form you'll actually keep taking; that matters more than 'theoretically better' but inconvenient.

Operational bottom line:

Have an A-B indication: buy the RCT-validated strain for that indication (LGG, *S. boulardii*, 35624, etc.)'Tune my gut': fermented foods + high fiber + dietary diversity > probiotic supplementSeverely ill / immunosuppressed: discuss with a doctorHealthy person wanting 'insurance' daily supplementation: don't bother — spend the money on vegetables and fruit

Open this on the map

Chapter 4

Fermented foods · daily probiotics

Fermented foods are an 8000-year byproduct of food preservation and the real daily source of probiotics — better evidence and cheaper than capsules.

Main types (by region and culture):

Yogurt: standard *L. bulgaricus* + *S. thermophilus* cultures; real yogurt's ingredient list is just 'milk + active cultures' — no sugar, no flavoring, no thickenersKefir: milk- or water-based, 10-30 bacteria + yeast in mixed fermentation; far more diverse than yogurtKimchi (Korean): vegetables + lactic-acid-bacteria fermentation, plus chili and anti-inflammatory polyphenolsSauerkraut (German): cabbage + salt, spontaneous fermentationNatto (Japanese): *Bacillus subtilis* var. *natto* — the only food source of K2 (MK-7)Miso / soy sauce: *Aspergillus oryzae* koji + later bacterial activity; salt + partial pasteurization leave few live cellsKombucha: tea + SCOBY (symbiotic culture of bacteria + yeast), contains acetic-acid bacteria + yeastSourdough: wild yeast + lactic-acid bacteria; most cells die during baking, but phytate is reduced and glycemic response is gentler
Real fermentation vs fake fermentation:

Real: cultures inoculated naturally or deliberately, refrigerated, label reads 'live cultures'Fake (commercial short-ferment + pasteurized): shelf-stable (no refrigeration), ingredient list reads 'vinegar' rather than 'culture', almost no live cells
Most supermarket pickled cucumbers, vinegar-pickled vegetables, and heat-treated German sauerkraut aren't real fermentation; refrigerated brands like Bubbies, Wildbrine, and homemade kimchi are.

Health evidence for fermented foods:

Sonnenburg 2021 *Cell* (previous page): 10 weeks → 19 inflammation markers down, diversity upKorean cohorts (Park 2017 and others): high kimchi intake correlates with lower BMI / metabolic syndrome (with confounders)Yogurt + CV: multiple meta-analyses link yogurt with lower diabetes / CV risk
Practical:

Rotate 1-2 different fermented foods daily — stronger than 'same brand every day'Breakfast yogurt or kefir 200 g + small kimchi or sauerkraut 50-100 g with meals is a good comboLactose intolerant: try plant-based yogurts, coconut kefir, vinegared vegetables, kombuchaSalt watch: Korean kimchi and German sauerkraut are high-salt — hypertension patients bewareChildren / elderly / immunosuppressed: prioritize commercial pasteurized + culture-added products, avoid home spontaneous fermentation (contamination risk)

Yogurt buying · 1-line ingredient rule

Real yogurt vs the 'milk beverage' trap — the ingredient list tells you in 1 minute.

Real yogurt ingredients: 'raw milk, *Streptococcus thermophilus*, *Lactobacillus bulgaricus*' — that's it, no more than 5 items.

Milk-beverage / flavored-yogurt common ingredients (to avoid):

Raw milk / reconstituted milk / waterWhite sugar / HFCS / sucrose appearing in large quantitiesFlavoring agentsThickeners (pectin / carrageenan / xanthan gum)Food coloringPreservatives (potassium sorbate)
A few common misconceptions:

'Sugar-free' doesn't equal healthy — check grams of carbohydrate. Real yogurt has ~4 g carbs/100 g (lactose); 'sugar-free' flavored yogurts may still contain artificial sweetenersGreek yogurt is strained and concentrated — high protein (10 g/100 g vs 3-4 g in regular), low carb — a good choiceWhole-fat vs non-fat: multiple meta-analyses show whole-fat yogurt associates with *lower* CV risk — don't fall for the 'low-fat' label
For children: under 3, prioritize whole-fat plain yogurt (no sugar) — protein, calcium, and fortified vitamin D for development; sugared 'kids' yogurts' contain sugar comparable to soda — pure marketing.

For lactose intolerance: fermentation partially degrades lactose in yogurt or kefir, and most lactose-intolerant people tolerate 100-200 g; with full intolerance or milk-protein allergy, choose plant-based (soy, coconut, almond) — but plant-based yogurts have lower protein and calcium, so pick fortified versions.

Storage details: refrigerate strictly < 6°C; finish within 3-5 days after opening; whey separation (yellow liquid) is normal — protein-rich, don't pour it off; clumping, off smell, or bulging lid → discard.

'Yogurt cures everything' marketing vs reality:

Gut tuning: real yogurt + dietary diversity has some evidence, don't expect miracles'Boosts immunity': vagueWhitening / weight loss: no evidence
Yogurt's best role: natural, high-protein, moderate-calorie, Ca + D fortified, low-GI breakfast or snack component — treat it as food, not medicine.

Open this on the map

Chapter 5

FMT · the gold standard

Fecal Microbiota Transplantation (FMT) sounds disgusting, but it's one of the rare cases in modern medicine of a 'whole-ecosystem transplant' with an 80-90% cure rate for a single indication.

FDA / NICE-approved indication: recurrent *Clostridioides difficile* infection is the gold standard. After three or more recurrences despite standard antibiotic therapy, FMT achieves an 80-90% cure rate (van Nood 2013 *NEJM*, the classic RCT — stopped early because the effect was so strong). In 2022 the US FDA approved the first capsule-form FMT product, RBX2660 (Rebyota); SER-109 (Vowst) followed in 2023.

Research-stage indications (B-C tier):

IBD (ulcerative colitis): partial positive RCTs for remission, but require multiple high-frequency sessionsIBS-D: small RCTs show symptom improvementMetabolic syndrome / insulin resistance: Vrieze 2012 *Gastroenterology* — transplanting from lean to obese subjects raised insulin sensitivity at 6 weeks, but the effect was transientAutism (ASD): Kang 2017 *Microbiome*, n=18 — small study with symptom improvement, replication pendingDepression / anxiety: early stageMultiple sclerosis: early research
Why FMT is ~100× stronger than a probiotic:

1. Transfers the full microbial community — 1,000+ species including unculturable ones; a probiotic capsule has 1-16 strains
2. Transfers bacterial metabolites — short-chain fatty acids, vitamins, bile-acid metabolism already running
3. Transfers the 'home' — mucin, peptidoglycan, fungal components, phages
4. Transfers a pre-selected community that has already worked in another healthy individual

Delivery methods: colonoscopy / enema (traditional, direct colonization); nasojejunal tube (NJ tube, upper-GI delivery); frozen capsules (oral 'crapsules', now in many hospitals); donor stool banks (OpenBiome in the US; similar in other countries).

Risks: pathogen transmission requires strict donor screening (HIV / HBV / HCV / parasites / multi-drug-resistant organisms). In 2019 the US had two deaths involving multi-drug-resistant *E. coli*, prompting FDA to upgrade screening standards. Long-term effects remain unknown — does FMT transplant metabolic tendency, behavior, or immunity? Active research.

The current consensus is that FMT is not 'general wellness' — it is a specific medical treatment under regulatory oversight. Don't trust any 'DIY FMT' or 'home stool transplant' clinic.

Future of microbiome medicine

The microbiome field's evolution from 2010 to 2025, and the limitations of the probiotic era.

First-generation probiotics (1900-2000s): the concept was to add a single 'good bacterium' to repair the gut; in practice most strains didn't colonize, individual variation was large, and clinical effects were small — represented by yogurt cultures and generic commercial probiotics.

Second-generation probiotics (2010-2020s): the concept shifted to strain-specific + indication RCT validation; in practice products like LGG, *S. boulardii*, and *B. infantis* 35624 became evidence-based options — represented by current AGA / NICE recommended uses.

Third generation — microbiome therapeutics (2022+):

Capsule-form, standardized FMT: Rebyota (2022) / Vowst (2023)Next-generation probiotics (NGP) beyond lactic-acid bacteria and bifido — *Akkermansia muciniphila* (Cani lab work) and *Faecalibacterium prausnitzii* (a major anti-inflammatory strain)Synthetic designed consortia: dozens of strains engineered togetherBacteria + metabolite combinations (the postbiotic concept): a probiotic plus its key metabolites
Personalized microbiome medicine (active research): testing your own microbiome composition (uBiome went under, Viome continues), but clinical utility is very limited — knowing what bacteria you have doesn't tell you what to supplement. Microbiome-based disease prediction: anti-PD-1 immunotherapy response is partly predicted by gut bacteria (Routy 2018 *Science*); drug responses to digoxin and chemotherapy are influenced by gut microbes (Haiser 2013 *Science*).

Most likely directions for the next 5-10 years:

More FDA-approved microbiome therapeutics (live biotherapeutic products, LBPs)NGPs going mainstream (*Akkermansia* already commercialized via the Lacroix team)Microbiome + drug combination therapy (paired with immunotherapy or chemotherapy)Precision probiotics — selection + tracking based on individual microbiome + indication
At the household-daily level, however, the next 5 years are unlikely to bring large change: the best move for healthy people remains a diverse diet + fermented foods + high fiber — not outdated, but a cheap, practical solution repeatedly validated over 30+ years of research.

Final operating psychology: the probiotic industry is $70B+ per year, and massive commercial interest inflates the 'insurance use' concept. The gap between reality and marketing is large; spending a bit more time on information consumption has a far better ROI than supplement consumption.

Open this on the map

Chapter 6

Practical · should I / how

The most practical decision tree.

Q1: Do I have a specific clinical indication?

On antibiotics or just finished: *S. boulardii* 250 mg ×2/day, or LGG 10¹⁰ CFU/day, spaced 2 h from the antibioticTraveling to a developing country: *S. boulardii*, start 2 days before departureRecurrent IBS: try *B. infantis* 35624 (Align) × 4 weeks — stop if it doesn't workInfant acute viral diarrhea: pediatric LGG or *S. boulardii* with oral rehydrationRecurrent vaginosis / UTI in women: *L. rhamnosus* GR-1 + *L. reuteri* RC-14High-risk family allergy prevention during pregnancy / infancy: LGG, used in late pregnancy and infancy
Q2: I just want to 'tune my gut.'

You don't need a probiotic capsule; these three are more effective:

1. 1-2 different fermented foods daily (yogurt, kefir, kimchi, natto)
2. 25-40 g fiber daily, 30+ different plant species per week
3. Fewer antibiotics, less ultra-processed food

Money spent at the produce market has a better ROI than at the pharmacy.

Q3: I've heard probiotics support immunity / anti-aging.

These indications have weak evidence; continuing to take them just means you've been persuaded by marketing rather than science. Energy is better spent on sleep, exercise, waist reduction, and stress management — the things with strong evidence.

How to take — key points:

Take with or after meals: food buffers acid and provides nutrients for the bacteriaSpace 2-4 h from antibiotics — otherwise the antibiotic also kills the probioticKeep refrigerated products refrigerated; keep shelf-stable products cool — don't leave them in a hot carTake continuously for at least 4 weeks — effects usually need time to accumulate; if one strain doesn't work, try anotherDon't mix multiple brands at once — wasteful and you can't tell which is doing whatAfter symptoms improve, taper gradually or switch to fermented foods for maintenance
Red flags (stop and see a doctor):

Bloating or diarrhea worsens rather than improves — not the right fitFever, chills, persistent discomfort: emergency evaluation (rare bacteremia)Critically ill / immunosuppressed / central venous line / severe pancreatitis: pause and don't self-administer
Wellness-grade vs pharmaceutical-grade probiotics: wellness-grade products are loosely regulated by FDA / NMPA and label accuracy varies — choose third-party certified (USP / NSF / ConsumerLab); pharmaceutical-grade probiotics have approved indications, regulation, and RCT data. For everyday use, wellness-grade is usually sufficient; for critical or therapeutic use, look for pharmaceutical-grade.

TL;DR

A few core takeaways:

Probiotics aren't a 'good-bacteria supplement' — they're a precision tool requiring a match of strain + indication + time windowFor most people most of the time, food (fermented + fiber + diversity) beats capsulesFMT is the real microbiome therapy, but only for specific indications — don't fall for 'home stool transplant' scams'Insurance-style daily probiotic capsules' aren't needed for most people; spend the money on higher-ROI things (vegetables and fruit, exercise, sleep)
The core wisdom of this island: a healthy gut isn't the result of 'eating more good bacteria' — it's the result of giving the bacteria that already live in you a good working environment.

Good working environment = enough fiber + diverse plants + fewer antibiotics + less ultra-processed food + adequate sleep + stress management.

These are cheap, sustainable long-term, and side-effect-free — the real 'probiotic alternative'.

Open this on the map