Ten research investigations

Original notes reproduced from the merged research dossier without converting preliminary interpretations into approved medical recommendations. Source links are provided above. Other limitations in the full text still apply.

**Question:** Do glucosamine, chondroitin sulfate, or the combination reliably improve knee pain beyond placebo in the overall sample?  
**Source:** Clegg et al., 2006; [PMID 16495392](https://pubmed.ncbi.nlm.nih.gov/16495392/), doi:10.1056/NEJMoa052771.  
**Design:** NIH-sponsored multicenter randomized double-blind placebo- and celecoxib-controlled GAIT trial, **1,583** people with symptomatic knee OA; **24 weeks**; primary outcome was at least 20% knee pain reduction.  
**Result:** Overall treatment contrasts were **not significant** versus 60.1% placebo response: +3.9 percentage points for glucosamine (p=.30); +5.3 for chondroitin (p=.17); +6.5 for combination (p=.09). Celecoxib positive control was +10 percentage points (p=.008). In **exploratory** baseline moderate-to-severe pain subgroup, combined therapy response was 79.2% vs 54.3% placebo (p=.002).  
**Evidence boundary:** The exploratory subgroup is *not* proof of reproducible efficacy for all people with more severe disease. The tested glucosamine/chondroitin preparations cannot be interchanged with arbitrary commercial salt/formulations; clinical pain endpoint is not cartilage regrowth or disease modification.  
**Editorial candidate:** Put the overall placebo-controlled null result first; clearly label the subgroup **exploratory**, without “proven joint repair” or “works for everyone.”  
**Disposition:** P1 claim precision, awaiting live-page comparison and independent signoff.

Original notes reproduced from the merged research dossier without converting preliminary interpretations into approved medical recommendations. Source links are provided above. Other limitations in the full text still apply.

**Question:** Can biomarker restoration be mistaken for symptom improvement, and do positive smaller trials settle efficacy?  
**Sources:** Taylor et al., 2015; [PMID 25545331](https://pubmed.ncbi.nlm.nih.gov/25545331/), doi:10.1016/j.atherosclerosis.2014.12.016. Skarlovnik et al., 2014; [PMID 25375075](https://pubmed.ncbi.nlm.nih.gov/25375075/) (contrasting positive 50-person trial).  
**Design:** Taylor double-blind simvastatin-vs-placebo symptom-confirmation crossover in 120 people reporting prior myalgia; only **41** developed symptoms on statin but not placebo and were allocated to CoQ10 vs placebo adjunct in an 8-week trial. The other randomized study enrolled **50** statin users reporting milder symptoms, without the same statin-specific confirmation step.  
**Result:** In the symptom-confirmed Taylor sample, circulating CoQ10 increased substantially, but neither pain severity nor pain interference improved versus placebo (**p=.53**, **p=.56**), nor did strength or VO2max. The smaller, different-population trial reported improved pain scores.  
**Null and scope:** Mechanistic “statins deplete CoQ10” is not by itself a clinical-treatment efficacy result. Mixed trial designs and ascertainment explain why competing findings require qualification; no universal resolution, equivalence to changing prescribed therapy, or causal endpoint claim.  
**Editorial candidate:** Use *conflicting results* and separate measured serum exposure from patient-important symptom outcome.  
**Disposition:** P1 mixed evidence; no medication-change recommendations.

Original notes reproduced from the merged research dossier without converting preliminary interpretations into approved medical recommendations. Source links are provided above. Other limitations in the full text still apply.

**Question:** Is either nutrient effective for prostate cancer prevention in relatively healthy older men?  
**Sources:** Lippman et al., 2009 original randomized report [PMID 19066370](https://pubmed.ncbi.nlm.nih.gov/19066370/), doi:10.1001/jama.2008.864; Klein et al., 2011 extended randomized-cohort follow-up [PMID 21990298](https://pubmed.ncbi.nlm.nih.gov/21990298/), doi:10.1001/jama.2011.1437.  
**Design:** Four-arm phase III RCT, **35,533** randomized men, selenium (L-selenomethionine), vitamin E (all-rac-alpha-tocopheryl acetate), both, or placebo. Prostate cancer occurrence, not antioxidant biomarkers, was the primary outcome.  
**Result:** Initial 2009 report: **no statistically significant prostate cancer reduction** from selenium, vitamin E, or combination. Selenium HR **1.04** (99% CI .87–1.24). Extended 2011 follow-up: **vitamin E-alone group higher prostate cancer incidence**, 620 vs 529 placebo cases; HR **1.17** (99% CI 1.004–1.36). Selenium-alone follow-up HR **1.09** (99% CI .93–1.27), not statistically significant.  
**Important separation:** The vitamin E risk is a **separate randomized arm and formulation**, not evidence that dietary vitamin E or every antioxidant causes cancer. The 2011 paper is follow-up of the same SELECT trial, **not** an independent replication. An initial nonsignificant selenium diabetes signal does not establish a statistically proven diabetes effect in the cited report.  
**Editorial candidate:** Explicitly reject “selenium supplements prevent prostate cancer”; note vitamin E-alone high-exposure arm's later risk signal with population/formulation context.  
**Disposition:** P0 review for cancer-prevention or multivitamin sales assertions; no change to published copy until approved.

Original notes reproduced from the merged research dossier without converting preliminary interpretations into approved medical recommendations. Source links are provided above. Other limitations in the full text still apply.

**Question:** What exactly is supported for people already at high risk for advanced age-related macular degeneration (AMD)?  
**Sources:** AREDS2 Research Group 2013 randomized trial [PMID 23644932](https://pubmed.ncbi.nlm.nih.gov/23644932/), doi:10.1001/jama.2013.4997; Chew et al. 2022 ten-year **post-trial epidemiologic extension** of the same original randomized cohort [PMID 35653117](https://pubmed.ncbi.nlm.nih.gov/35653117/), doi:10.1001/jamaophthalmol.2022.1640.  
**Design:** 4,203 participants aged 50–85 with intermediate AMD / high-risk features, compared lutein+zeaxanthin, omega-3 fatty acids, their combination and placebo **on top of AREDS base formulation** (not no supplements); additional factorial comparisons varied beta carotene and zinc.  
**Original primary result:** Lutein+zeaxanthin as an **add-on** did not significantly reduce 5-year progression to advanced AMD versus placebo add-on (HR **0.90**, 98.7% CI .76–1.07, **p=.12**). Omega-3 add-on also null (HR **0.97**, p=.70); no statistically significant visual acuity benefit in primary comparison.  
**Long-term follow-up:** 2022 cohort extension suggested favorable AMD association for prior lutein/zeaxanthin assignment (late AMD HR **0.91**, 95% CI .84–.99), and increased lung-cancer odds for prior beta-carotene assignment (OR **1.82**, 95% CI 1.06–3.12). These **are not independent new randomizations**; subsequent AREDS2 product use and follow-up design should be disclosed.  
**High-risk boundary:** Lutein/zeaxanthin **replacement** for beta carotene in a defined AREDS2 formulation is different from “taking extra lutein improves eyesight,” “prevents AMD in anyone,” or “restores vision.” Former-smoker/lung-cancer rationale must not be omitted from beta-carotene safety discussions.  
**Disposition:** P0 formulation and beta-carotene safety review; P1 add-on efficacy claim correction.

Original notes reproduced from the merged research dossier without converting preliminary interpretations into approved medical recommendations. Source links are provided above. Other limitations in the full text still apply.

**Question:** Why do promising symptom-scale trials and a larger trial using standardized endpoints differ?  
**Sources:** Cash et al., 2016 randomized trial [PMID 26319955](https://pubmed.ncbi.nlm.nih.gov/26319955/), doi:10.1007/s10620-015-3858-7; Weerts et al., 2020 PERSUADE RCT [PMID 31470006](https://pubmed.ncbi.nlm.nih.gov/31470006/), doi:10.1053/j.gastro.2019.08.026.  
**Design/results:** Cash: double-blind **72-person**, four-week trial of one **small-intestinal targeted sustained-release** peppermint-oil formulation in IBS-M/IBS-D; Total IBS Symptom Score dropped **40%** from baseline versus **24.3%** placebo (p=.0246). Weerts: double-blind **190 randomized** IBS participants, eight weeks, two release formulations vs placebo; FDA-style abdominal pain responder primary endpoint **not significant** (small-intestine-release 46.8% vs 34.4% placebo, p=.170; ileocolonic-release 41.3%, p=.385). Co-primary overall symptom relief also **not significant**. Some **secondary** pain/discomfort/IBS-severity scores favored small-intestinal release; mild adverse events were more common than placebo.  
**Boundary:** Different formulation, endpoint definition, patient mix and trial length prevent pooling nominal percentages as if identical comparisons. An IBS symptom study is not proof for every peppermint tea, essential oil, reflux symptom or all IBS phenotypes.  
**Editorial candidate:** Report the later prespecified primary/co-primary nulls beside small-trial positive signal; report secondary positives as secondary.  
**Disposition:** P1 mixed/route-specific evidence review.

Original notes reproduced from the merged research dossier without converting preliminary interpretations into approved medical recommendations. Source links are provided above. Other limitations in the full text still apply.

**Question:** Do 12-week cognitive-domain outcomes imply acute attention improvement or ADHD treatment?  
**Sources:** Stough et al. 2001 double-blind placebo-controlled trial, branded KeenMind extract, [PMID 11498727](https://pubmed.ncbi.nlm.nih.gov/11498727/), doi:10.1007/s002130100815 (**PubMed flags a 2015 erratum involving a dosage error**; do not reproduce article dosing without checking correction). Morgan and Stevens 2010 randomized double-blind placebo-controlled trial [PMID 20590480](https://pubmed.ncbi.nlm.nih.gov/20590480/), doi:10.1089/acm.2009.0342.  
**Result:** 2001 trial reported gains in selected learning/memory measures following **12 weeks**, alongside state-anxiety signal; maximum effects were not immediate. 2010 trial recruited **98 older adults**, with **81 completing**; gains on selected AVLT learning/recall tasks versus placebo, but complex-figure, subjective memory complaint, and trail-making differences **not statistically significant**. Gastrointestinal adverse effects (including cramps, nausea and increased stool frequency) were reported.  
**Evidence boundary:** Distinct extracts, chronic durations, and older/healthy populations. Selected positive test domains do not prove comprehensive executive enhancement, short-term stimulant-like results, ADHD therapy or dementia prevention.  
**Editorial candidate:** Place delayed-onset memory-domain results beside null task results and tolerability; check flagged erratum before any dose disclosure.  
**Disposition:** P1 correction/precision; two PMIDs should undergo canonical source reuse checks.

Original notes reproduced from the merged research dossier without converting preliminary interpretations into approved medical recommendations. Source links are provided above. Other limitations in the full text still apply.

**Question:** What did the early objective/subjective sleep pilot actually show?  
**Source:** Pigeon et al., 2010, [PMID 20438325](https://pubmed.ncbi.nlm.nih.gov/20438325/), doi:10.1089/jmf.2009.0096.  
**Design:** Randomized, double-blind, cross-over trial of a **proprietary juice beverage** versus matched placebo in **15 older adults with chronic insomnia**; each arm two weeks with a two-week washout. Sleep primarily measured with daily diaries and Insomnia Severity Index—not a large diagnostic/long-term clinical outcome trial.  
**Result:** Within-beverage pre/post sleep variables improved, but the **between-beverage** significant improvements were restricted to insomnia severity and minutes awake after sleep onset; **sleep latency, total sleep time and sleep efficiency did not significantly improve versus placebo**.  
**Boundary:** The comparison is a formulation-specific, very small short-term pilot in older adults; cannot promise increased sleep duration to all users, normalize circadian rhythm as a clinical outcome or claim equivalent benefit for capsules.  
**Editorial candidate:** Promote placebo comparisons above before/after changes; explicitly include total sleep time null.  
**Disposition:** P1 correction. The exact PMID is already present in repo research and existing tart cherry article; **source reuse**, not an automatically new citation.

Original notes reproduced from the merged research dossier without converting preliminary interpretations into approved medical recommendations. Source links are provided above. Other limitations in the full text still apply.

**Question:** Can a positive open/no-prophylaxis comparison survive a larger blinded placebo-controlled community trial?  
**Sources:** Kranjčec et al., 2014 randomized three-group study [PMID 23633128](https://pubmed.ncbi.nlm.nih.gov/23633128/), doi:10.1007/s00345-013-1091-6; Hayward et al., 2024 double-blind placebo-controlled primary-care RCT [PMID 38587819](https://pubmed.ncbi.nlm.nih.gov/38587819/), doi:10.1001/jamainternmed.2024.0264.  
**Design/results:** 2014: **308** women with prior recurrent UTI post-acute treatment, compared D-mannose, prophylactic nitrofurantoin, or **no prophylaxis** over six months; recurrence **14.6%** in D-mannose versus **60.8%** in no-prophylaxis arm. No blinded placebo comparison was established in that design. 2024: **598** eligible women across 99 UK primary-care sites, D-mannose versus matching placebo, six months; medically attended suspected UTI **51.0%** (150/294) versus **55.7%** (161/289); risk difference approximately **−5 percentage points** (95% CI −13% to +3%), **p=.26**.  
**Synthesis:** The large placebo-controlled primary-care trial **did not demonstrate clinical prevention benefit**. Earlier positive data should be retained only with comparator and study-design distinctions. These studies address prevention, **not treatment** of active bacterial infection.  
**Editorial candidate:** Remove “prevents UTIs” universal copy / any substitute-for-antibiotics impression; keep null and population/exposure boundaries.  
**Disposition:** P0 clinical false-reassurance review. A 2024 PMID already appears in audit/cache files; canonical admission still needs dedupe.

Original notes reproduced from the merged research dossier without converting preliminary interpretations into approved medical recommendations. Source links are provided above. Other limitations in the full text still apply.

**Question:** How well do selected market product assays match the labels, and can results be generalized to all products?  
**Sources:** Erland and Saxena 2017 Canadian-market **analytical assay** [PMID 27855744](https://pubmed.ncbi.nlm.nih.gov/27855744/), doi:10.5664/jcsm.6462; Cohen et al. 2023 U.S. melatonin-gummy **analytical assay** [PMID 37097362](https://pubmed.ncbi.nlm.nih.gov/37097362/), doi:10.1001/jama.2023.2296. **Neither is an efficacy RCT.**  
**Results:** 2017: **31 product samples**, melatonin ranged from **83% below** to **478% above** label; serotonin detected in **8/31** samples. 2023: **25 U.S. gummy products** assayed; 22/25 (88%) outside a ±10% tolerance of declared amount; one had undetectable melatonin but detectable **CBD**; among those with melatonin, measured amounts were **74%–347% of stated quantities**. **No serotonin was detected** in the 2023 assay, unlike the 2017 sample set.  
**Critical quality firewall:** Samples differ in country, date, brand, ingredient class, and dosage form. Never interpret the 2017 serotonin count as today's U.S. gummy contamination rate, or the 2023 gummy result as tablets/capsules or every current brand. Historical cross-sectional rates are **not** present-day manufacturing defect estimates. Specific product/lot requires its own assay.  
**Editorial candidate:** Disclose label-dose uncertainty using dates, sample frame and null contrast, avoiding false universal “all melatonin gummies contain serotonin” claim.  
**Disposition:** P0 quality-control review; no automatic affiliate/product assertions.

Original notes reproduced from the merged research dossier without converting preliminary interpretations into approved medical recommendations. Source links are provided above. Other limitations in the full text still apply.

**Question:** What does surveillance support about neurologic adverse reactions, and can an incidence rate or safe threshold be inferred?  
**Source:** van Hunsel et al., 2018 pharmacovigilance **spontaneous-report case series**, [PMID 29737502](https://pubmed.ncbi.nlm.nih.gov/29737502/), doi:10.1007/s40264-018-0664-0.  
**Design:** Netherlands Lareb case-series assessment of **90 neuropathy-related adverse-reaction reports** linked to supplements containing B6, including reported product exposures, some serum concentrations and causality scoring. Products in reports listed **1.4–100 mg/tablet**; 36 cases had serum concentrations; the authors found **no statistically significant correlation between label dose and measured blood B6** in available cases.  
**Synthesis:** Plausible neuropathy risk with higher and prolonged exposures; the authors could not exclude lower reported exposures. But spontaneous reports have **no denominator**, may have incomplete concomitant-exposure/diagnostic details, and **cannot** establish adverse-event rates, a validated lowest-harm dose, or whether a specific product caused any single patient's symptoms.  
**Editorial candidate:** Support clear neurologic risk caution and check stacked supplement exposure; avoid universal “below X mg is guaranteed safe” or fabricated incidence. Do not equate usual dietary B6 with supplement-excess risk.  
**Disposition:** P0 safety/editorial review. No patient-specific medical advice or dosing.

## Integrity receipt and next gates

1. **10 investigation dossiers** cover **10 distinct questions** from prior DE10. The corpus already contains source references for at least AREDS2 (PMID 23644932), Bacopa (PMID 11498727), tart cherry (PMID 20438325), and D-mannose (PMID 38587819). Source references in article/audit/index files do **not** establish net-new evidence rows. Full canonical source-key comparison remains required for every PMID, including those absent in GitHub code search.
2. **Study classes separate:** RCTs vs follow-up of same original RCT vs analytical chemistry vs spontaneous surveillance. No observational or in-vitro work presented as proven RCT clinical benefit.
3. **Null/negative endpoints explicit:** overall GAIT pain null; confirmed statin-myalgia CoQ10 null; selenium prostate prevention null; initial AREDS2 add-on null; later peppermint oil trial prespecified primary/co-primary null; some Bacopa memory tests null; tart cherry total sleep time null; D-mannose primary trial null; 2023 melatonin serotonin assay null; B6 report series cannot estimate incidence.
4. **High-priority checks before any live output:** SELECT vitamin E cancer risk, beta carotene among former smokers in AREDS2, D-mannose false reassurance for active/recurrent UTIs, melatonin product-assay limitations, and pyridoxine neuropathy are **P0 editorial-review flags**, *not* proof that current site copy is wrong until page-level comparison.
5. **Erratum caution:** Stough Bacopa PMID 11498727 has a linked 2015 dose-correction notice. No “recommended dose” may be inferred from the uncorrected article text.
6. **Source verification:** Every PMID link was individually located in PubMed's original citation record in this round; no URL or DOI listed without matching its published record. This does not establish an independent secondary reviewer or current retraction-check certification.
7. **Publication/admission:** No runtime ledger, public site claim, source promotion, indexability, ads/affiliate placements or dosage fields are modified. Require independent review, exact study identity/dedupe, canonical entity/route/strength checks, governance signoff, PR checks on exact head, then monitored published-page QA if a downstream implementation is separately approved.
8. **ROI claim:** Research-only contribution. No supported immediate traffic, conversion, profit, or deployed content metric.

**Change type:** Additive research document on a new branch. Existing research corpus and all prior waves are unchanged.