Headline finding · criterion 4 of 6
Just one of 18 best-selling gut supplements discloses a dose inside the clinical trial range
Just one of the 18 best-selling gut-health supplements audited by Supplement Facts Check discloses a dose inside the range human trials used for the outcome claimed. The other 17 do not: the disclosed dose falls outside the trial range, or no dose is disclosed to check.
The sample. 18 products scored so far from the top 30 of Amazon’s Best
Sellers in Probiotic Nutritional Supplements, captured 09/26/2026. Every inclusion and exclusion decision is
listed on the results page. Read these as pilot results, not a ranking.
Criterion 4: Dose within the trial range
The disclosed dose falls inside the range human trials used for the outcome the product claims. It scores 1 or 0, with no weighting and no partial credit, under methodology v1.14, locked 2026-10-26. Read the full methodology →
All 18 scored products
Every row that went into the figure above, with the reasoning recorded for this criterion and the date the label was read. Each product links to its full scorecard and the dated label captures the score was read from.
| Product | Criterion 4 | Total | Reasoning recorded for this criterion | Label captured |
|---|---|---|---|---|
| Align Probiotic Pro Formula Align | 0 ✗ | 4/6 | Under methodology v1.13. The panel declares 1x10^9 CFU at manufacture but guarantees only 1x10^7 through the best-by date, and the through-expiry figure governs. Whorwell 2006 (PMID 16863564, n=362) found 1x10^8 CFU significantly superior to placebo in capsule form while 1x10^6 and 1x10^10 were not, so the guaranteed dose is ten times below the only capsule dose shown to work. Note that the at-manufacture figure of 1x10^9 would have scored 1 under the benefit-of-the-doubt rule; the ruling and the objection to it are published in the methodology. | 2026-09-27 |
| Digestive Advantage Probiotic Gummies Digestive Advantage | 0 ✗ | 4/6 | Under methodology v1.13. The CFU figure is marked with a double-dagger footnote that is never defined anywhere on the Supplement Facts panel: the panel defines *, ** and the single dagger, and the double dagger appears only against '1 Billion CFUs'. No at-manufacture qualifier and no through-expiry guarantee is therefore established, and the delivered dose cannot be determined. Independently, the only trial identified for this strain at this outcome (Hun 2009, PMID 19332970) states no CFU dose in its abstract, so no reference range exists to compare against even if the delivered dose were established. Vitamin D at 8 mcg is a nutrient at roughly 40 percent of the adult Daily Value, not a dose aimed at the bloating and abdominal discomfort outcome this product claims. | 2026-09-27 |
| MaryRuth Organics Kids Probiotics MaryRuth Organics | 0 ✗ | 4/6 | No positive-outcome human trial for DE111 at a gastrointestinal outcome was identified on 2026-09-27, and no paediatric trial of the strain was identified at any dose, which is the population this product is sold for. The two DE111 trials that measured gut endpoints both reported no effect (Freedman 2021 PMID 33671071; Townsend 2018 PMID 30049931), and under the positive-outcome-trials-only rule a null trial does not set a range. The reference table row records the full search per the negative-findings rule. Note the label itself is not the problem here: it discloses the dose in both mg and CFU and names the strain, so a buyer could check it if the evidence existed to check it against. The failure is in the evidence base, and the objection recorded under v1.11 applies with full force to this row. | 2026-09-27 |
| Culturelle Daily Chewable Probiotics for Kids Culturelle | 0 ✗ | 3/6 | The failure is on dose rather than on any unit or disclosure technicality. The label delivers 5x10^9 CFU per day. Vanderhoof 1999 (PMID 10547243), the pediatric LGG trial, dosed 1x10^10 to 2x10^10 CFU/day in children aged 6 months to 10 years and found a significant reduction in antibiotic-associated diarrhoea. This product provides one quarter to one half of the trial dose. The clinical-doses.csv row for LGG was carrying a 2026-09-07 flag reading 'NOT YET VERIFIED against PubMed... verify before relying on it'; it was verified and replaced with a real range before this score was assigned. | 2026-09-27 |
| Culturelle Pro Strength Probiotics Culturelle | 0 ✗ | 3/6 | Inulin 200 mg against an independent reference range of 12 g/d for functional constipation (Puhlmann 2025, PMID 41233756), roughly sixty times below the studied dose. Probiotics excluded per methodology. | 2026-09-26 |
| Nature's Bounty Probiotic Acidophilus Nature's Bounty | 0 ✗ | 3/6 | Under methodology v1.11: every active in this product is a probiotic, so nothing on the panel can be checked against a trial range. The probiotic exclusion removes the comparison, it does not award the point. | 2026-09-26 |
| Align Probiotic Bloating Relief + Food Digestion Align | 0 ✗ | 2/6 | Under methodology v1.13. The probiotic declares 1x10^9 CFU at manufacture but guarantees only 1x10^7 through the best-by date, ten times below the 1x10^8 capsule dose Whorwell 2006 (PMID 16863564) found superior to placebo. Vitamin B12 at 2.4 mcg is exactly the Daily Value, but the claimed outcome is bloating relief and food digestion rather than B12 repletion, and no adequate human trial was identified for B12 at that outcome. The enzymes sit inside a blend total, so none can be compared to a reference range under the attribution rule. | 2026-09-27 |
| Lemme Debloat Digestive & Gut Health Gummies Lemme | 0 ✗ | 2/6 | XOS at 80 mg falls far below the 1.4 to 8 g/day reference range, reaching under 6 percent of the low end even if the 80 mg were 100 percent XOS, which the label does not establish. The probiotics sit inside a 44 mg blend total with no CFU declared, so neither strain can be compared to any trial dose under the attribution rule, and under v1.13 no delivered potency is established either. | 2026-09-27 |
| NOW Supplements Probiotic-10 NOW Foods | 0 ✗ | 2/6 | This product satisfies the v1.13 disclosure requirement and still fails on a separate ground. The label states 'This product is formulated to deliver a potency of 25 billion CFU through the Best By date', the first through-expiry guarantee found in the sample, so the delivered dose IS establishable. But the 25 billion is an aggregate across ten strains with no split, and the attribution rule bars a blend total from setting or satisfying a range for any single component. No strain can be compared to its own trial dose. The failure is a disclosure gap on the label, not a units problem in the literature. | 2026-09-27 |
| NewRhythm Probiotics 50 Billion CFU NewRhythm | 0 ✗ | 2/6 | The probiotics are an aggregate with no per-organism split and no strain designations, so nothing can be matched to a trial under the attribution rule. On v1.13, the footnote reads 'Contains a minimum of 50 billion live bacteria when manufactured, and provides viable levels until at least the best by date'. The at-manufacture count is a number; the through-expiry assurance is the phrase 'viable levels', which is not a quantity. No delivered dose is established. The contrast with NOW Probiotic-10, whose label commits to a specific figure through the Best By date, is the distinction v1.13 was written to draw. The prebiotic blend discloses no composition. | 2026-09-27 |
| Phillips' Colon Health Daily Probiotic Phillips' | 0 ✗ | 2/6 | The 3 Billion CFU figure is an aggregate across two strains with no split, so neither can be compared to its own trial range under the attribution rule. The point survives even the most generous possible reading: if the entire 3x10^9 CFU were LGG, it would still fall below the 1x10^10 to 2x10^10 CFU/day range verified from Vanderhoof 1999 (PMID 10547243), so no allocation of the blend puts LGG in range. Under v1.13 no through-expiry guarantee and no at-manufacture qualifier appears anywhere on the panel either. | 2026-09-27 |
| Physician's CHOICE Probiotics 60 Billion CFU Physician's Choice | 0 ✗ | 2/6 | Prebiotic Blend totals 150 mg including inulin from Jerusalem artichoke and chicory, an order of magnitude below the 12 g/d inulin reference range even if the blend were entirely inulin; probiotics excluded per methodology. | 2026-09-07 |
| Seed DS-01 Daily Synbiotic Seed | 0 ✗ | 2/6 | Under methodology v1.7. Previously 1, on the DS-01 finished-product trials. Those no longer set the range: they dose a whole multi-ingredient synbiotic, so the GI benefit cannot be attributed to the pomegranate component, and their authors include manufacturer employees while an independent ingredient-level range exists. Scored instead against that independent range — 320-656 mg phenolics — which the label's 400 mg at >40% meets only to a floor of about 160 mg, at or below the dose an independent dose-response trial found null (PMID 27879044). Probiotics excluded per methodology. | 2026-09-07 |
| BIOMA Probiotics BIOMA | 0 ✗ | 1/6 | XOS 100 mg of extract falls below the 1.4-8 g/d reference range even at 100% purity; tributyrin excluded (no positive-outcome trial identified); probiotics excluded per methodology. | 2026-09-07 |
| Doctor's Recipes Women's Probiotic Doctor's Recipes | 0 ✗ | 1/6 | The probiotics are CFU-declared with no strain designations and only an aggregate count, so no organism can be matched to a trial. Under v1.13 there is no through-expiry guarantee anywhere on the panel, and no at-manufacture qualifier either, so the delivered dose cannot be established. The Prebiotic Fiber Blend has no disclosed composition. Cranberry at 50 mg is unstandardized, and its trial literature addresses urinary tract outcomes rather than the digestive outcome this product is scored against. | 2026-09-27 |
| Garden of Life Dr Formulated Probiotics for Men Garden of Life | 0 ✗ | 1/6 | The probiotics are CFU-declared with no strain designations, so no organism can be matched to a trial even before the unit question arises, and the aggregate count cannot be attributed to any single organism. The prebiotic fibers are disclosed only as a 365 mg blend total, so neither resistant starch nor acacia fiber can be compared to a reference range under the attribution rule. Note on v1.13, amended 2026-09-27: the superscript 1 after '50 Billion CFU' is not defined in the panel image captured for this row. On the sister product rank 30 (Garden of Life Dr. Formulated Kids+, B074JLL5KS) the identical footnote marker is defined as 'At Expiration Date under recommended storage conditions', so this label most likely carries the same definition in label text the capture does not show, and should be read as a through-expiry guarantee rather than an undefined marker. The score is 0 either way, on attribution and on the absence of strain designations, but the earlier wording overstated the disclosure failure and is corrected here. | 2026-09-27 |
| Garden of Life Dr. Formulated Kids+ Probiotics Garden of Life | 0 ✗ | 1/6 | This product passes the v1.13 disclosure test while still failing. The footnote reads '1 At Expiration Date under recommended storage conditions', so the 5 Billion CFU is guaranteed through expiry and the delivered dose IS established, only the second label in the sample to do this after NOW Probiotic-10. It fails on attribution instead: the 5 Billion is an aggregate across 14 organisms, split only into Lacto (4 Billion) and Bifido (1 Billion) sub-totals, so no organism can be compared to its own trial range, and none carries a strain designation to look up in any case. The prebiotic blend is likewise an undifferentiated total. Vitamins C and D are nutrients at 33 and 50 percent of Daily Value, not doses aimed at the digestive outcome claimed. | 2026-09-27 |
| Florastor Probiotic Florastor | 1 ✓ | 5/6 | Under methodology v1.12, the first product in the sample to take this point. The panel declares 500 mg of preparation rather than CFU, so the probiotic unit mismatch does not arise, and 500 mg falls at the floor of the 500-1000 mg/day range set by independent adult trials of this strain (Seddik 2019 PMID 30694338; Sjomina 2023 PMID 37942999; Salazar-Parra 2023 PMID 37400812). Maev 2025 excluded from setting the range under the independence rule: it dosed Enterol, a Biocodex product, and Biocodex manufactures Florastor. The outcome objection is published alongside the v1.12 ruling. | 2026-09-26 |
See these rows in the dataset →
Citing this finding
The sentence above is generated from the dataset at build time, so it cannot drift from the rows listed on this page. Quote it as:
Just one of the 18 best-selling gut-health supplements audited by Supplement Facts Check discloses a dose inside the range human trials used for the outcome claimed. Supplement Facts Check, methodology v1.14, labels captured 2026-09-07 to 2026-09-27. https://supplementfactscheck.org/findings/dose-within-clinical-trial-range.html
The underlying data is public domain under CC0 1.0 and needs no permission to reuse. The repository holds the scores, the rubric and the commit history of every ruling. Corrections are welcome as issues.