Dose reference table
Criterion 4 checks a label’s dose against the range human trials actually used for the outcome the product claims. This is that table: 32 rows, 18 with a range set. Rows without one record why, including the date the literature was searched.
| Ingredient | Form | Daily range | Outcome measured, and basis | PubMed ID |
|---|---|---|---|---|
| L-glutamine | free-form | 15 g | post-infectious IBS-D symptom severity and permeability; 5g TID for 8 weeks (Zhou 2019); PubMed verified 2026-09-07 — confirmed: 5 g three times daily (15 g/d) for 8 wk, n=106 completing, primary endpoint met by 79.6% versus 5.8% on placebo, with lactulose/mannitol ratios normalised (Zhou 2019) | 30108163 |
| L-glutamine | free-form | ≥ 30 g | intestinal permeability; meta-analysis found no overall effect, significant reduction only in subgroup >30g/day; PubMed verified 2026-09-07 — the overall null result is confirmed (WMD -0.00, 95% CI -0.04 to 0.03, 10 studies, 352 participants). WARNING on the subgroup this range rests on: the source abstract is internally inconsistent. Its results section reports the significant subgroup as doses over 30 g/day while its conclusion states 30 mg/day, a thousandfold discrepancy; and the subgroup statistic is printed as WMD -0.01 with 95% CI -0.10 to -0.08, an interval that excludes its own point estimate. This row follows the 30 g/day reading as the only physiologically plausible one, but the figure should be checked against the full text before it is used to score any product | 39397201 |
| Zinc carnosine | polaprezinc | 75 mg | prevention of NSAID-induced intestinal permeability increase; 37.5mg BID for 5 days, crossover n=10 (Playford 2006); tested prevention not repair; PubMed verified 2026-09-07 — confirmed: 37.5 mg twice daily (75 mg/d) for 5 d, randomised crossover, n=10 healthy volunteers, preventing the threefold indomethacin-induced rise in lactulose/rhamnose ratio. Formal citation is Gut 2007;56:168-75, epub 2006 | 16777920 |
| Slippery elm | inner bark powder | No range set | no adequate human trial identified; RCT-filtered search returns only multi-herb proprietary blends (PMID 29958034, null for GI symptoms) and other Ulmus species; PubMed verified 2026-09-07 — 10 records, none isolating slippery elm on a gut outcome. The null multi-herb 'detox' RCT already cited is PMID 29958034; a second multi-herb formula containing slippery elm alongside curcumin, aloe, glutamine, pectin, guar gum and peppermint reported large improvements but was an uncontrolled pre-post study, not randomized (Ried 2020, PMID 32151878), and cannot attribute effect to any one component | 29958034 |
| Marshmallow root | root extract | No range set | no adequate human trial identified for gut outcomes; RCT-filtered search returns candidiasis, cough syrup and topical studies only; remaining evidence is animal and in vitro; PubMed verified 2026-09-07 — 4 records, none testing Althaea root extract on a gut outcome. Trap to avoid: PMID 25951410 is a genuine randomized crossover reporting reduced ileostomy output (n=28), but it tested confectionery marshmallows and attributes the effect to gelatine. It is indexed under Althaea and must not be read as evidence for marshmallow root extract | — |
| DGL licorice | GutGard branded extract | 150 mg | GERD symptoms and quality of life; 75mg BID post-meals for 4 weeks, n=200 (Raj 2025); branded flavonoid-rich DGL standardized to glabridin >=3.5%, does not transfer to generic DGL; PubMed verified 2026-09-07 — n=200, 28 d, and the glabridin >=3.5% w/w standardisation are all confirmed, as is the benefit on GERD quality of life (p=0.014), heartburn and regurgitation. However the 150 mg dose is NOT stated in the abstract; like the peppermint row, the figure rests on a source this verification could not reach | 39929150 |
| Collagen peptides | hydrolyzed type I/III | No range set | no positive-outcome controlled trial identified; 10g/day 7 days null for GI integrity and symptoms, n=20 crossover (PMID 36370176); 20g/day 8wk reported benefit but open-label single-arm, 14/40 completed (PMID 35639457); both studies include authors affiliated with the collagen manufacturer; PubMed verified 2026-09-07 — cited PMID 36370176 confirmed: 10 g/d for 7 d, n=20 crossover, no effect on lactulose/rhamnose ratio, I-FABP, inflammatory markers or subjective GI symptoms; two authors are employed by the collagen manufacturer (Rousselot BV) | 36370176 |
| Butyrate | sodium butyrate | No range set | 300mg/day null in pediatric IBD 12wk (Pietrzak 2022); no positive-outcome human trial verified; PubMed verified 2026-09-07 — cited PMID 36014789 confirmed: 150 mg twice daily (300 mg/d) for 12 wk added to standard therapy in newly diagnosed paediatric IBD, n=72, no difference in remission rate or disease activity versus placebo | 36014789 |
| Butyrate | tributyrin | No range set | no PubMed-indexed positive-outcome trial identified; a 100-200mg 21-day pilot published outside PubMed reported null results for butyrate levels and microbiome with a triglyceride increase at 200mg; 4g/day depression trial is a protocol with no results; remaining evidence in vitro and animal; PubMed verified 2026-09-07 — only 3 RCT-indexed records exist. The one testing isolated tributyrin (6.53 g, acute meal, n=12) was null for GLP-1, GIP, PYY and neurotensin, the authors concluding dietary butyrate did not stimulate gut hormone secretion (PMID 26178726); the other two embed tributyrin in multi-component immunonutrition feeds and found no benefit (PMID 16832133, 21044933). A 2025 Parkinson's study at 1500 mg/d is open-label, uncontrolled and has no gut outcome (PMID 41271518); the 4 g/d depression protocol is PMID 41248397 | 26178726 |
| Berberine | berberine HCl | 400 mg | IBS-D diarrhea frequency, abdominal pain, urgency; 8 weeks (Chen 2015; 400mg/day given as 200mg BID); PubMed verified 2026-09-07 — confirmed: 400 mg/d as 200 mg twice daily for 8 wk, n=132 randomised of 196 recruited, significant reductions in diarrhoea frequency (P=0.032), abdominal pain and urgency (both P<0.01) | 26400188 |
| Quercetin | quercetin dihydrate | ≥ 500 mg | systemic inflammation (CRP reduction) per meta-analysis at >=500 mg/day (PMID 28537580); no RCTs identified for intestinal permeability — gut-barrier evidence is preclinical only; PubMed verified 2026-09-07 — a clinical-trial-filtered search for quercetin with intestinal permeability, gut barrier, leaky gut or irritable bowel returns zero records. The gut-barrier claim rests entirely on preclinical work; PubMed verified 2026-09-07 — confirmed: 7 RCTs across 10 treatment arms, CRP reduced by 0.33 mg/l overall and by 0.34 mg/l in the subgroup at or above 500 mg/d. Caveat: the same meta-analysis reports that meta-regression found no association between CRP change and dose, which weakens 500 mg/d as a genuine threshold rather than a subgroup artefact | 28537580 |
| Aloe vera | inner leaf gel | No range set | no usable reference range; IBS extract trial null vs inulin control, n=160, dose not stated in abstract (Ahluwalia 2020); positive UC trial used 200mL/day liquid gel (Langmead 2004), not convertible to capsule mg; PubMed verified 2026-09-07 — 8 RCT-indexed records, none yielding a convertible capsule dose. The positive UC trial dosed 100 mL twice daily of liquid gel (Langmead 2004, PMID 15043514); a second IBS trial of AVH200 extract, n=68 for 4 wk, missed its primary endpoint (p=0.09) with positive secondary signals and no dose in the abstract (Storsrud 2015, PMID 26405698); the remaining positives are multi-ingredient formulas where aloe is combined with inulin and lactitol (PMID 31686768) or with senna (PMID 39951927) | 32314514 |
| N-acetyl glucosamine | NAG | No range set | tested and null; no significant effect on faecal biomarkers of mucosal damage in randomized multi-arm trial, n=24 arm, 14 days (Chandwe 2024, P=0.67); dose not stated in abstract; open-label IBD pilots at 3-6 g/day; Phase 3 IBS-D trial registered at 300mg/day (NCT02504060), no published results located; PubMed verified 2026-09-07 — cited PMID 38632262 confirmed: NAG arm of a multi-arm phase II trial in children with severe acute malnutrition, n=24, 14 days, effect on the composite faecal mucosal-damage biomarker -0.20 (90% CI -1.01 to 0.60), P=0.67, while teduglutide was the only arm to reach the trial's pre-specified threshold | 38632262 |
| Ashwagandha | root extract standardized | 600 mg | stress and serum cortisol reduction; 300mg BID for 60 days, n=64 (Chandrasekhar 2012); extract described as high-concentration full-spectrum, brand not named in paper; no gut-outcome trials identified; NOT YET VERIFIED against PubMed as of 2026-09-07. This row sets no gut reference range (it is a stress and cortisol outcome with no gut trial identified), so it cannot decide a criterion 4 score, and it was deprioritised in the verification pass. Verify before relying on it. | 23439798 |
| L-theanine | free-form | 200–400 mg | stress and anxiety reduction; 200-400mg/day per systematic review of 9 RCTs (Williams 2020); no gut-outcome trials identified; NOT YET VERIFIED against PubMed as of 2026-09-07. This row sets no gut reference range (it is a stress and anxiety outcome with no gut trial identified), so it cannot decide a criterion 4 score, and it was deprioritised in the verification pass. Verify before relying on it. | 31758301 |
| Magnesium | glycinate | No range set | no adequate human trial identified for gut outcomes; constipation evidence is for magnesium oxide (PMID 32969946), a different salt; PubMed verified 2026-09-07 — search returns 6 records and none is a human gut-outcome trial: a 1994 bioavailability crossover in ileal-resection patients measuring absorption, not a gut outcome (PMID 7815675); rat ileum motility (PMID 37636381); two Caco-2 cell studies (PMID 32098378, 41599937); a marine-multimineral tolerability trial using bisglycinate only as an in vitro comparator (PMID 38608850); and a single hypomagnesemia case report (PMID 42567449) | — |
| Magnesium | citrate | No range set | no adequate human trial identified for supplement-dose use; available trials are colonoscopy bowel-prep protocols at purgative doses, usually combined with sodium picosulfate; PubMed verified 2026-09-07 — all 8 RCT-indexed records are colonoscopy or surgical bowel preparation at purgative doses, typically combined with sodium picosulfate or sennosides (e.g. PMID 25019969, 27782976, 20799286, 9496494). None tests supplement-dose magnesium citrate for a gut outcome | — |
| Astaxanthin | natural haematococcus | No range set | tested and null for gut outcomes. CORRECTION 2026-09-07: this row previously read 'no gut-outcome trial identified', which was wrong — two randomized placebo-controlled trials in functional dyspepsia exist. 16 and 40 mg/d for 4 wk, n=132 across three arms, found no difference between groups on the primary GSRS endpoint of abdominal pain, indigestion and reflux, with a secondary signal on reflux at 40 mg only (Kupcinskas 2008, PMID 18467083); 40 mg/d, n=44, found gastric inflammation fell in both arms with no between-group difference and no change in H. pylori density or interleukins (Andersen 2007, PMID 17521392). Both are null on their primary endpoints, so under the positive-outcome rule no reference range is set and the criterion 4 exclusion is unchanged — but the reason is 'tested and did not work', not 'never studied'. Remaining human RCTs are for skin ageing and oral submucous fibrosis (5mg BID) | 18467083 |
| Piperine | black pepper extract | 10–15 mg | dosed as absorption enhancer, not for an independent outcome; 10mg/day with 1g curcuminoids (PMID 25618800) and 15mg/day (PMID 25688638); excluded from criterion 4 as no trial tests piperine against a gut outcome of its own; NOT YET VERIFIED against PubMed as of 2026-09-07. This row sets no gut reference range (it is an absorption enhancer excluded from criterion 4), so it cannot decide a criterion 4 score, and it was deprioritised in the verification pass. Verify before relying on it. | 25618800 |
| Curcumin | standardized extract | 1500–2000 mg | UC disease activity and relapse prevention; 1500mg/day 8wk (Sadeghi 2020, PMID 31802559) and 2000mg/day 6mo (Hanai 2006); both as add-on to mesalamine/sulfasalazine; PubMed verified 2026-09-07 — both trials confirmed, but they are not the same outcome and the row should not be read as one range. Sadeghi 2020 gave 1500 mg/d for 8 wk to n=70 with ACTIVE mild-to-moderate UC, improving disease activity, hs-CRP and ESR. Hanai 2006 gave 2000 mg/d (1 g twice daily) for 6 mo to n=89 with QUIESCENT UC as maintenance, cutting relapse from 20.5% to 4.7% (P=0.040). Under the outcome-specific rule, a product claiming symptom relief scores against 1500 mg and one claiming remission maintenance against 2000 mg | 17101300 |
| Curcumin | Curcugen branded extract | 500 mg | general digestive symptoms (GSRS) and anxiety; 500mg once daily 8wk, n=79 (Lopresti 2021); no effect on microbiota or SIBO; branded enhanced-bioavailability extract, does not transfer to standard 95% extract; PubMed verified 2026-09-07 — confirmed: 500 mg once daily for 8 wk, n=79 randomised and 77 analysed, significant reduction in GSRS total and in DASS-21 anxiety, with no effect on microbiota or SIBO. Not previously noted: one author is affiliated with DolCas Biotech, which manufactures Curcugen | 33478482 |
| Saccharomyces boulardii | live strain | No range set | excluded from criterion 4 per methodology; antibiotic-associated diarrhoea prevention in children, 250mg BID, n=269, RR 0.3 (Kotowska 2005); dose stated in mg of preparation, not convertible to label CFU; PubMed verified 2026-09-07 — confirmed: 250 mg twice daily, n=269 enrolled and 246 analysed, RR 0.3 (95% CI 0.2-0.7). Dose is milligrams of preparation with no CFU stated anywhere in the abstract, which is the unit mismatch the probiotic exclusion rests on | 15740542 |
| Lactobacillus rhamnosus GG | live strain | No range set | excluded from criterion 4 per methodology; probiotic trial dosing is strain-specific and reported inconsistently as mg or CFU, not convertible to label CFU; NOT YET VERIFIED against PubMed as of 2026-09-07. This row sets no gut reference range (it is a probiotic excluded from criterion 4, carrying no PMID), so it cannot decide a criterion 4 score, and it was deprioritised in the verification pass. Verify before relying on it. | — |
| Bifidobacterium longum | live strain | No range set | excluded from criterion 4 per methodology; probiotic trial dosing is strain-specific and reported inconsistently as mg or CFU, not convertible to label CFU; NOT YET VERIFIED against PubMed as of 2026-09-07. This row sets no gut reference range (it is a probiotic excluded from criterion 4, carrying no PMID), so it cannot decide a criterion 4 score, and it was deprioritised in the verification pass. Verify before relying on it. | — |
| Psyllium husk | powder | ≥ 10 g | chronic constipation; >10 g/day for >=4 weeks optimal per systematic review and meta-analysis (2022); PubMed verified 2026-09-07 — confirmed: 16 RCTs, 1251 participants, psyllium and pectin the fibres with significant effects, optimal above 10 g/d for at least 4 wk. Caveat the authors state and this row should carry: heterogeneity was considerable (I2 86% for stool frequency), and flatulence was significantly higher on fibre than control | 35816465 |
| Peppermint oil | enteric-coated | 561 mg | IBS abdominal pain; 187mg TID for 8 weeks (Merat 2010; dose inferred from Colpermin formulation, not stated in abstract); largest trials null at 182mg TID; PubMed verified 2026-09-07 — the inference this row already flags is confirmed as unresolved: Merat 2010 states only 'one capsule of Colpermin three times daily' for 8 wk in n=90. No milligram dose appears in the abstract, so 561 mg/d rests entirely on the Colpermin formulation being 187 mg per capsule | 19507027 |
| Zinc | picolinate | No range set | no trial identified for this form; permeability normalization reported with zinc sulfate 110mg TID for 8 weeks in Crohn's (Sturniolo 2001); other gut trials use gluconate or elemental zinc in diarrhoea and HIV; doses inconsistently reported as compound vs elemental; PubMed verified 2026-09-07 — 4 RCTs indexed for this form, none with a gut outcome: a 1987 absorption comparison at 50 mg elemental zinc (PMID 3630857), a null multi-nutraceutical depression trial (PMID 30699842), COPD antioxidant status at 22 mg (PMID 18295467), and taste disorder at 29 mg TID (PMID 12132610) | — |
| Magnesium | oxide | 1500 mg | chronic idiopathic constipation; 1.5g MgO daily for 28 days, n=90, 68.3% response vs 11.7% placebo (Morishita 2021); dose stated as MgO compound weight not elemental; PubMed verified 2026-09-07 — confirmed on dose, duration and response rates. Precision correction: n=90 is the whole three-arm trial (senna 1.0 g, MgO 1.5 g, placebo), so the MgO arm is roughly 30 participants, not 90 | 32969946 |
| Xylooligosaccharide | XOS from corn cob | 1.4–8 g | bifidobacteria increase in healthy adults; 1.4 and 2.8 g/d for 8 wk n=32 (Finegold 2014); 5 g/d for 4 wk n=60 (Lecerf 2012, PMID 22264499); 8 g/d for 21 d (Childs 2014, PMID 24661576); 4 g/d T2DM trial (Sheu 2008) not a gut outcome; INU-XOS arm in Lecerf used 1 g XOS + 3 g inulin and does not set the range; PubMed verified 2026-09-07 — all three cited trials confirmed on dose, duration and sample size: Finegold 2014 n=32 at 1.4 and 2.8 g/d for 8 wk with bifidobacteria up in both arms versus placebo and 2.8 g significantly greater; Lecerf 2012 n=60 at 5 g/d for 4 wk; Childs 2014 at 8 g/d for 21 d, which also raised bowel movement frequency. The INU-XOS arm is confirmed as 3 g inulin + 1 g XOS | 24513849 |
| Inulin | chicory root | 12 g | functional constipation; 12 g/d for 4 wk, randomized double-blind placebo-controlled crossover, n=39 Rome III, improved stool frequency and constipation-related quality of life (Puhlmann 2025, BMC Gastroenterol, PMID 41233756); 12 g/d as 3x4 g for 4 wk, randomized double-blind placebo-controlled crossover, n=44, increased stool frequency vs placebo (Micka 2017, Int J Food Sci Nutr 68:82-89, PMID 27492975); both trials used chicory-derived inulin. Not a probiotic: assessed under criterion 4 normally; PubMed verified 2026-09-07 — both trials confirmed on dose, duration, design and sample size | 41233756 |
| Pomegranate extract | ellagitannin-standardized fruit extract | 320–656 mg phenolics | gut microbiota modulation and reduced endotoxemia in overweight-obese adults; 656 mg phenolics/d for 3 wk crossover n=49 reduced plasma LBP and increased Faecalibacterium, Odoribacter and Butyricicoccus (Gonzalez-Sarrias 2018, PMID 29665619); 320 mg phenolics/d for 1 mo crossover n=50 modulated microbiota in metabolic syndrome but the effect depended on concurrent drug therapy (Cortes-Martin 2021, PMID 33458928); a dose-response trial found 160 mg phenolics/d null, with effects only at 640 mg/d and only in urolithin metabotype B subjects (Gonzalez-Sarrias 2017, PMID 27879044), so 160 mg does not set the range. Doses are phenolic weight, not extract weight: comparing a label to this range requires both the extract weight and a stated standardization percentage; PubMed verified 2026-09-07 — all cited trials confirmed; METHODOLOGY v1.7 — this is the governing range for pomegranate extract. It is independent of any manufacturer and tests the ingredient itself, so under the evidence hierarchy it takes precedence over the finished-product row below | 29665619 |
| Pomegranate extract | MAPP, as formulated in the DS-01 multi-species synbiotic | 400 mg extract | GI quality of life, bloating, gas, abdominal discomfort, constipation and regularity; 400 mg Indian pomegranate extract given with 53.6 billion AFU multi-species probiotic for 6 wk, randomized placebo-controlled n=350 (Allegretti 2026, PMID 41599868); same formulation for 91 d n=32 increased faecal butyrate, urinary urolithin A and microbial diversity (Napier 2025, PMID 40944126). Finished-product trials: the pomegranate component is not tested in isolation, so this range does not transfer to pomegranate extract generally. Several authors of both trials are employees of the manufacturer (Seed Health); PubMed verified 2026-09-07 — both trials confirmed; METHODOLOGY v1.7 — this row NO LONGER SETS A RANGE. It fails both hierarchy rules: the trials dose a whole multi-ingredient synbiotic, so the effect cannot be attributed to the pomegranate component (the same rule that bars the multi-herb formulas elsewhere in this table), and their authors include employees of the manufacturer while an independent ingredient-level range exists. Retained as evidence of the finished product, not as a dose reference | 41599868 |