Gut Health · Supplement Engine
Probiotics
Strain-specific — what actually works, what makes things worse, and how to pick the right one
ConditionalSummary: Probiotics only work when you use the exact right strain for the exact right problem. Taking a generic "multi-strain" supplement for IBS is like taking a random antibiotic for an unknown infection — it might do nothing, or it might make things worse. Two strains — B. coagulans MTCC 5856 and
Think of the bacteria in probiotics like a set of keys. Your gut's immune system has very specific locks — and only certain keys fit. The strange part? The key doesn't even need to be attached to a living bacterium. Scientists ran a 443-person Lancet trial where they killed all the bacteria with heat, and the dead cells worked just as well for IBS — because it's the shape of the key on the bacterial cell wall, not whether the bacterium is alive, that triggers the anti-inflammatory response.
That's the general answer. Your stack is different.
Check your whole stack
Gut Health · Supplement Engine
Strain-specific — what actually works, what makes things worse, and how to pick the right one
ConditionalThe Plain English Version
Probiotics only work with the exact right strain — the wrong one can make your gut worse.
Think of the bacteria in probiotics like a set of keys. Your gut's immune system has very specific locks — and only certain keys fit. The strange part? The key doesn't even need to be attached to a living bacterium. Scientists ran a 443-person Lancet trial where they killed all the bacteria with heat, and the dead cells worked just as well for IBS — because it's the shape of the key on the bacterial cell wall, not whether the bacterium is alive, that unlocks the anti-inflammatory response.
Want the full evidence? Keep scrolling
What People Claim
Athletes are told probiotics reduce muscle soreness and give a performance edge. IBS sufferers are told probiotics will calm flare-ups. Anyone taking antibiotics is told to replenish their microbiome. The implicit message across all of it: all probiotics are roughly equivalent, and the decision comes down to brand and CFU count.
This framing is scientifically incorrect in every material way. The evidence shows that efficacy is locked to specific strains for specific conditions — a principle so fundamental that the world's leading probiotic researchers compare pooling different strains in meta-analyses to pooling all antibiotics together in an infection trial.
McFarland et al. (2018) analysed 228 trials and found that pooled strain analyses showed heterogeneity I²>58% — indicating the studies couldn't be legitimately combined. When strain-specific analysis was applied, heterogeneity collapsed to 0%. The signal was always there. The methodology was obscuring it.
What the Evidence Shows
| Claimed Benefit | Conviction | Verdict |
|---|---|---|
|
IBS symptom relief — B. coagulans MTCC 5856 Majeed et al. 2016, N=36, p<0.01 What would change this: A large independent RCT failing to replicate in a low-FODMAP IBS population. |
STRONG | Works — strain-specific |
|
IBS all subtypes — B. bifidum MIMBb75 (incl. dead cells) Andresen et al. 2020, Lancet, N=443 — 34% vs 19% placebo (p=0.0007) What would change this: Biopsy evidence showing heat-killed cells cannot restore gut barrier integrity (only immune signalling). |
STRONG | Works — paradigm-shifting |
|
Antibiotic-associated diarrhoea (AAD) prevention Guo et al. 2019, Cochrane, N=6,352 — RR 0.49, NNT=11 What would change this: A blinded RCT showing consistent null result when taken correctly 2–3h from antibiotic dose. |
STRONG | Works — standard of care |
|
C. difficile prophylaxis (high-risk patients) Goldenberg et al. 2017, Cochrane, N=8,672 — RR 0.25–0.30 What would change this: Evidence that adherence <50% fully explains the effect (execution confound, not mechanism). |
STRONG | Conditional — high-risk only |
|
Athletic immune support (URTI reduction) ISSN Position Stand 2019 — Jäger et al. What would change this: A large double-blind RCT (>200 athletes) with no sponsor conflict showing null URTI effect. |
MODERATE | Promising — small cohorts |
|
Muscle recovery (DOMS attenuation) Jäger et al. 2019, N=15 — reduced CK, preserved ROM What would change this: Independent replication with N>100 and performance (not just biomarker) primary endpoint. |
MODERATE | Secondary — not ergogenic |
|
Direct ergogenic performance (strength, VO2max) ISSN 2019 Position Stand What would change this: A mechanism-specific trial showing a strain directly modulates mitochondrial biogenesis or oxygen-carrying capacity. |
DEBUNKED | Doesn't work |
|
Generic "gut health" for healthy adults (no specific indication) Multiple class-level pooled reviews — highly variable by microbiome baseline What would change this: A pre-registered RCT in healthy adults with validated microbiome diversity endpoints. |
WEAK | Unproven — category claim |
L. plantarum 299v alleviates IBS abdominal pain. L. plantarum MF1298 — the same species, different strain — actively worsens IBS. Same species name, polar opposite clinical outcome. STRONG
Heat-inactivated (dead) B. bifidum MIMBb75 works as well as live cells for IBS via structural immune signalling — the "live bacteria" dogma is outdated. STRONG
LGG is highly effective for paediatric AAD but loses significance in generic adult populations — microbiome baseline dictates response. STRONG
S. boulardii is a yeast, not a bacterium — it is impervious to antibiotics and requires no temporal spacing from antibiotic doses. STRONG
Up to 33% of commercial probiotics fail third-party label testing — missing declared strains, zero viable CFU before expiry, or contaminating species. STRONG
No probiotic strain demonstrates permanent gut colonisation. Benefits stop within 1–2 weeks of discontinuing supplementation. STRONG
How It Works
Probiotics work through four distinct, well-mapped biochemical pathways. Which pathway is active determines which conditions a given strain can treat — and why strain identity is non-negotiable.
Most commercial probiotics (Lactobacilli, Bifidobacteria) don't produce butyrate directly. Instead, they produce acetate and lactate that feed endogenous butyrate-producing bacteria (Faecalibacterium prausnitzii, Roseburia intestinalis). This cross-feeding cascade diverts intestinal energy, doubles SCFA absorption, and drives systemic anti-inflammatory effects. Up to 37% of adults lack the primary butyrate-producing taxa entirely — for these individuals, probiotic-assisted cross-feeding may be the only viable route to butyrate production without a complete dietary overhaul.
Spore-forming bacteria like B. coagulans create an acidic luminal environment via lactic acid synthesis that suppresses pathogenic overgrowth (E. coli, Salmonella) and prevents C. difficile spore germination. Physical spatial displacement compounds this through bacteriocin secretion. This is the primary mechanism behind antibiotic-associated diarrhoea prevention.
This is the paradigm-shifting pathway. Pathogen-Associated Molecular Patterns (PAMPs) — structural proteins on bacterial cell walls like lipoteichoic acid and peptidoglycans — interact with Toll-Like Receptors on intestinal immune cells even when the bacterium is dead. This shifts immune phenotype from pro-inflammatory toward regulatory, reducing visceral hypersensitivity. This is precisely why heat-inactivated B. bifidum MIMBb75 (every cell killed) works just as well as live cells for IBS. The immune system reads the shape of the key, not whether the bacterium is alive.
Disrupted gut microbiome composition degrades tight junction proteins (ZO-1, occludin, claudin-1), increasing intestinal permeability. B. bifidum MIMBb75 upregulates mucosal mucin (MUC2) production and directly reinforces epithelial barrier integrity — reducing the visceral hypersensitivity responsible for IBS pain across all subtypes, not just the diarrhoea-predominant form.
The Debate
Guo et al. 2019, Cochrane (children)
LGG prevents antibiotic-associated diarrhoea — RR 0.49, NNT=11 in paediatric populations.
Adult subgroup meta-analyses
LGG fails to show statistical significance in generic adult AAD prevention trials.
Why they disagree: The adult gut has higher colonisation resistance and different immune receptor distribution than the paediatric gut. Paediatric benefit doesn't automatically translate. Age and microbiome baseline are the confounding variables.
Goldenberg et al. 2017 (adherence ≥75%)
Multi-strain protocol reduces C. difficile infection by 75% when adherence to timing and dosing is maintained.
Same protocol, adherence 17%
Null CDI result. Under-dosing and delayed administration completely nullify the prophylactic window.
Why they disagree: This isn't a mechanism debate — it's an implementation failure. The efficacy is real; the execution variable is everything. A probiotic that consistently isn't taken 2–3 hours after the antibiotic dose doesn't reach the intestine alive.
L. plantarum 299v (validated strain)
Significantly reduces abdominal pain in IBS patients vs placebo in controlled trials.
L. plantarum MF1298 (same species)
Actively worsens IBS symptoms in the same patient population in a separate controlled trial.
Why they disagree: Strain-level genetic differences create polar-opposite clinical outcomes within the same species. If these were pooled as "L. plantarum" in a meta-analysis, the result would be statistically null — erasing real efficacy and obscuring real harm. The era of genus/species-level probiotic recommendations is scientifically indefensible.
Real World vs Lab
In the lab
Strain identity and CFU counts are verified before every trial begins — you know exactly what's in the capsule.
In the real world
Up to 33% of commercial probiotic products fail third-party label testing — absent declared strains, zero viable CFU before expiry, or undisclosed contaminating species.
In the lab
Clinical benefits are measured during active supplementation. The therapeutic window is clearly defined.
In the real world
Most consumers believe they are "reseeding" the gut permanently. Exogenous strains face massive colonisation resistance from billions of entrenched native microbes. They pass through, exert their effects, then disappear within 1–2 weeks of stopping.
In the lab
Trials enrol populations with known conditions — IBS diagnosis, antibiotic co-administration — where the mechanism-to-benefit pathway is clear.
In the real world
Healthy adults without dysbiosis may see little to no measurable benefit. Individual non-responders — particularly those lacking butyrate-producing taxa (37% of adults) — may not respond even to correctly chosen strains.
Exactly How to Use It
| Population & Strain | Dose | Timing | Duration |
|---|---|---|---|
| IBS-D — B. coagulans MTCC 5856 (spore-forming) | 2 billion spores/day (1 capsule) | With food | 90 days |
| IBS all subtypes — B. bifidum MIMBb75 (live or heat-inactivated) | 1 billion CFU/day (1 capsule) | Any time | 4–8 weeks |
| AAD prevention — L. rhamnosus GG | 10–40 billion CFU/day | 2–3 hours after antibiotic | Duration of antibiotics + 14 days |
| C. diff prophylaxis (high-risk) — S. boulardii CNCM I-745 | 10–40 billion CFU/day | Any time (it's a yeast — antibiotics can't kill it) | Duration of antibiotics + 14 days |
| Athletes (immune) — L. fermentum VRI-003 PCC | 12 billion CFU/day | Daily | Ongoing during training block |
| Athletes (DOMS) — B. breve BR03 + S. thermophilus FP4 | 5 billion CFU each (2 capsules) | Daily | 21 days minimum |
Safety & Interactions
Antibiotics kill bacterial probiotic payload before intestinal delivery. Dose probiotics 2–3 hours apart from antibiotic dose. S. boulardii (a yeast) is fully exempt — take at any time.
Antifungals are lethal to S. boulardii (it is a yeast). If antifungals are prescribed, use a bacterial probiotic (LGG, B. coagulans) instead. Do not use S. boulardii concurrently with antifungals.
Live probiotics risk systemic infection in immunocompromised hosts. Avoid live probiotics. Heat-inactivated postbiotics (e.g. dead B. bifidum MIMBb75) may be acceptable — always consult prescribing physician first.
S. boulardii is associated with 1.70 fungemia cases per 10,000 patient-days (40.8% crude mortality). Live Lactobacillus is associated with 1.62 bacteremia cases per 10,000 patient-days. CRITICAL contraindication — live probiotics are absolutely contraindicated in this population.
Risk of Lactobacillus bacteremia and opportunistic systemic infection. Avoid all live probiotics without explicit physician approval. Postbiotics only.
Elevated bacterial translocation risk from live organisms crossing a compromised epithelial barrier. Medical supervision required.
The Nuance
' width='800' height='400'/%3E%3C/svg%3E)
| Strain | Monthly Cost | Food Alternative | Value Verdict |
|---|---|---|---|
| B. coagulans MTCC 5856 | £10–20 | No food equivalent — spore-forming bacteria are rare in food | Worth it (IBS) |
| B. bifidum MIMBb75 live | £20–35 | Yogurt with B. bifidum (non-validated strain, insufficient dose) | Worth it (IBS) |
| B. bifidum MIMBb75 postbiotic | £25–40 | No food equivalent — postbiotics are supplement-only | Worth it (IBS) |
| L. rhamnosus GG (AAD) | £15–30 | Kefir (contains LGG in some cultures — unverified CFU) | Worth it (antibiotics) |
| S. boulardii CNCM I-745 | £15–25 | No food equivalent — it's a yeast | Worth it (C. diff risk) |
| Generic multi-strain blend | £15–40 | Yogurt, kefir, kimchi, sauerkraut | Skip it |
Conviction
IBS conviction downgrade: A large, independently funded multi-centre RCT demonstrating that B. coagulans MTCC 5856 fails in a low-FODMAP IBS population would force a conviction review. Biopsy evidence proving that heat-inactivated cells strictly produce immune effects but cannot restore gut barrier integrity would reclassify postbiotics as immunomodulatory only (not barrier-restorative).
AAD/C.diff conviction downgrade: A pre-registered RCT with high adherence (≥80%) showing null CDI reduction from S. boulardii would downgrade this from HIGH to MODERATE. Current evidence is Cochrane-level with multiple independent replication.
Sources
1. Goldenberg et al. (2017) — Cochrane Database Syst Rev — N=8,672
Probiotics for the prevention of Clostridium difficile-associated diarrhoea. 70–75% CDI risk reduction with S. boulardii CNCM I-745.
2. Guo et al. (2019) — Cochrane Database Syst Rev — N=6,352
Probiotics for the prevention of paediatric antibiotic-associated diarrhoea. LGG: RR=0.49, NNT=11 for AAD prevention.
3. Andresen et al. (2020) — Lancet Gastroenterol Hepatol — N=443
Heat-inactivated B. bifidum MIMBb75 for IBS (all subtypes). 34% vs 19% placebo primary endpoint (p=0.0007). Dead cells as effective as live.
4. Majeed et al. (2016) — Nutr J — N=36
Bacillus coagulans MTCC 5856 for IBS-D. Significant symptom reduction vs placebo (p<0.01) at 2 billion spores/day × 90 days.
5. Jäger et al. (2019) — J Int Soc Sports Nutr (ISSN Position Stand)
Comprehensive review of probiotics in sport. Athletic immune support + DOMS evidence for B. breve BR03 + S. thermophilus FP4. No ergogenic performance effect found.
6. McFarland et al. (2018) — Front Med — 228 trials
Strain/species specificity review. Heterogeneity collapsed from I²>58% to 0% when strain-specific analysis was applied vs pooled genus/species analysis.
7. Corbin et al. (2023) — Metabolic ward RCT — N=17
SCFA cross-feeding mechanism in humans. 217 kcal/day fecal energy diversion; doubled SCFA absorption (P<0.00001).
8. Baxter et al. (2019)
37% of adults lack primary butyrate-degrading microbes required for resistant starch fermentation. Directly identifies the probiotic non-responder mechanism.
How strong is the evidence for the claims in this review? Higher = more confidence the claims are supported. This does not measure how large the effect is or how important it is compared with other levers.
Is this worth your time, money, effort, risk, and trust for this goal? Different from Verdict Score (evidence strength) and Leverage Map (relative importance) — Action ROI is the worth-it call once friction is priced in.
Evidence-scored dosing, timing, forms, and who should skip it. One page, no fluff.
Get the protocolConviction-scored verdicts on supplements, nutrition, training, physio, and recovery.