# Ordinary thiamine (vitamin B1) × Improvement of cognition in mild cognitive impairment Verdict 3087 · Evidence Check Stored-material review date: 2026-10-05 Original evidence date: 2026-09-15 Review-criteria version: NGM-3087-ORDINARY-B1-MCI-1.0 Public page/material preparation date: 2026-10-31 Clinical papers were not read directly; existing review summaries were checked. No new clinical-literature search was performed. Unverified does not mean that research is absent or that there is no effect. ## Assessment scope - Population: Adults with mild cognitive impairment (MCI); established Alzheimer dementia and confirmed B1-deficiency encephalopathy excluded - Ingredient, formulation and route: The question concerns single-ingredient oral ordinary thiamine; the exact salt and formulation of an eligible MCI result are unverified - Dose: Dose in an eligible MCI trial unverified - Comparator: Question: placebo, no addition, or the same usual care; eligible between-group MCI results unverified - Measured outcome: Change on a validated cognitive test; the specific scale in an eligible MCI trial is unverified - Duration: Duration of an eligible MCI trial unverified - Questions outside this scope: Benfotiamine; Established Alzheimer dementia; Confirmed B1-deficiency encephalopathy; Effects of ingredient combinations; Dementia prevention, survival and optimal dose ## Checklist items ### N01 · Are this table’s question about ordinary B1 alone and cognitive change in MCI, and its exclusions, clear? Verification status: Verified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The original claim concerns the comparative effect of single-ingredient oral ordinary thiamine on validated cognitive-test change in adults with MCI. The subclaim list is an actual empty array; one independent question was retained without inventing a specific scale, dose or duration. - Why unverified: The selected question and exclusions are explicit in the preserved source. This verification does not establish a direct clinical effect. - Reason for missing information: Specific eligible-trial conditions remain unverified in the individual rows below. - Required materials and conditions: A boundary record linking the current claim’s wording, question identity, population, endpoint and comparison. - Materials to obtain first: A boundary record linking the current claim’s wording, question identity, population, endpoint and comparison. - Further research: Before searching for eligible materials or designing a study later, specify this PICO and separate other molecular forms and diagnostic groups into different questions. - Conditions for verification: Link single-ingredient oral ordinary thiamine, adults with MCI, change on validated cognitive tests and comparison with placebo, no addition or the same usual care; exclude other molecular forms, dementia and deficiency encephalopathy. - Why this condition is needed: This boundary is required to avoid combining cognitive improvement in MCI and dementia treatment into one effect status. - Application exceptions: Unverified scales, doses or durations do not justify inventing combinations as new claims. - Remaining uncertainty: Specific eligible-trial conditions remain unverified in the individual rows below. - Referenced methods: Common conditions C01, C04, C05, C06 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N02 · Can MCI diagnosis, selection and baseline cognition be verified for an actual comparative result on ordinary B1? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The question concerns adults with MCI and excludes established Alzheimer dementia and confirmed B1-deficiency encephalopathy. The preserved summaries do not establish an isolated MCI result for ordinary B1; S26 involved established Alzheimer dementia. - Why unverified: The question’s population is fixed, but actual diagnosis, selection and baseline characteristics for an eligible MCI result have not been verified. - Reason for missing information: Diagnostic criteria, composition, selection and isolated results for an eligible MCI population are unverified. - Required materials and conditions: Eligibility criteria, diagnostic procedures, baseline characteristics and isolated MCI results from an eligible MCI trial. - Materials to obtain first: Eligibility criteria, diagnostic procedures, baseline characteristics and isolated MCI results from an eligible MCI trial. - Further research: Report population flow and results together in a comparative study that clearly selects clinical MCI and records baseline cognition and deficiency status. - Conditions for verification: Verify clinical MCI diagnostic criteria, age, baseline cognition, eligibility and MCI-only results; distinguish established dementia and deficiency encephalopathy. - Why this condition is needed: Relabeling an established Alzheimer disease population as MCI because both concern cognition would misapply the population. - Application exceptions: Retain mixed diagnostic populations and examine whether separate MCI results exist. Unverified information does not mean that there is no effect in MCI. - Remaining uncertainty: Diagnostic criteria, composition, selection and isolated results for an eligible MCI population are unverified. - Referenced methods: Common conditions C04, E08 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: S26 ### N03 · Was benfotiamine distinguished from direct MCI evidence for ordinary thiamine? Verification status: Verified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: S09 used benfotiamine in its cognitive trial. The preserved summary classified it as a molecular form different from ordinary thiamine and did not adopt it as a direct result for the current ordinary-B1–MCI question. - Why unverified: The preserved source explicitly states the intervention identity and exclusion reason, confirming the boundary for attributing direct evidence. - Reason for missing information: The identity of an actual eligible ordinary-B1 intervention in MCI remains separately unverified. - Required materials and conditions: Current source locations supporting S09’s intervention identity and its distinction from ordinary B1. - Materials to obtain first: Current source locations supporting S09’s intervention identity and its distinction from ordinary B1. - Further research: A future comparison of ordinary B1 and benfotiamine would need separately specified molecular interventions and direct comparative results in the same population. - Conditions for verification: Distinguish the original record that S09 used benfotiamine from ordinary thiamine; do not use its cognitive results as the current effect of B1 alone. - Why this condition is needed: This is required to avoid transferring results from a different molecular form to ordinary B1 simply because it is a related search term. - Application exceptions: This exclusion neither denies benfotiamine efficacy nor establishes equivalence among all B1 derivatives. - Remaining uncertainty: The identity of an actual eligible ordinary-B1 intervention in MCI remains separately unverified. - Referenced methods: Common conditions C05, C08, E02 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: S09 ### N04 · Can the actual ordinary-B1 salt, formulation and route in an eligible MCI result be verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The question concerns single-ingredient oral ordinary thiamine. Molecular form, route and chemical-form values for an actual eligible result are null, and the linked-source list is empty; the exact salt and formulation are unverified. - Why unverified: Once the question’s identity is distinguished from actual observations, there are no linked materials establishing an actual eligible preparation or route. - Reason for missing information: Actual ingredient, salt, formulation, product, administration route and linked sources for an eligible MCI result are unverified. - Required materials and conditions: Treatment composition, salt/formulation, product specifications, actual administration route and the original report for a comparative ordinary-thiamine MCI study. - Materials to obtain first: Treatment composition, salt/formulation, product specifications, actual administration route and the original report for a comparative ordinary-thiamine MCI study. - Further research: Specify the single ordinary-B1 product and oral administration procedure; distinguish molecular-form and route effects using a comparator with the same concomitant ingredients. - Conditions for verification: Distinguish the question’s single oral ingredient from the actual result’s ingredient, salt, product, formulation and route; link the source identity of the actual report. - Why this condition is needed: This is needed to avoid treating the question’s oral route as verification of an actual trial route. - Application exceptions: Do not infer hydrochloride from the name ordinary thiamine or transfer results from another route to oral administration. - Remaining uncertainty: Actual ingredient, salt, formulation, product, administration route and linked sources for an eligible MCI result are unverified. - Referenced methods: Common conditions C05, E02 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N05 · Can the actual ordinary-B1 dose and adherence in MCI be verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The dose in an eligible MCI trial is unverified. Actual-dose and supplement-dose records have null values; doses and tolerability from other molecular forms or dementia trials are not presented as equivalent. - Why unverified: Directly eligible doses and actual use or adherence records could not be verified from the preserved inputs. - Reason for missing information: Dose, frequency, adherence and changes or discontinuations for single-ingredient ordinary B1 are unverified. - Required materials and conditions: Dose, frequency, actual exposure, adherence and changes or discontinuations in an eligible MCI trial. - Materials to obtain first: Dose, frequency, actual exposure, adherence and changes or discontinuations in an eligible MCI trial. - Further research: Collect prescribed and actual doses together and prespecify how adherence and discontinuation will be handled in the analysis. Do not invent an optimal dose or participant count. - Conditions for verification: Verify nominal dose, frequency, actual use, adherence and changes; do not copy doses from other molecular forms or dementia trials into MCI. - Why this condition is needed: This condition is needed to attribute cognitive results to actual exposure without substituting doses from other trials. - Application exceptions: Missing adherence information does not prove nonadherence. This table provides no individual dosing instructions. - Remaining uncertainty: Dose, frequency, adherence and changes or discontinuations for single-ingredient ordinary B1 are unverified. - Referenced methods: Common conditions C05, E02, E08 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N06 · Can actual MCI treatment duration and cognitive-assessment timing be verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: Duration in an eligible MCI trial is unverified. The duration value is null and the linked-source list is empty; actual dose, duration and safety denominators for the current question are also recorded as unverified. - Why unverified: No materials were identified linking administration and measurement periods for a directly eligible result. - Reason for missing information: Actual treatment, assessment, discontinuation and follow-up schedules are unverified. - Required materials and conditions: Treatment, discontinuation and cognitive-assessment schedules and outcome-specific timepoints from an eligible MCI trial. - Materials to obtain first: Treatment, discontinuation and cognitive-assessment schedules and outcome-specific timepoints from an eligible MCI trial. - Further research: Plan cognitive-assessment schedules separately from treatment-exposure duration, and report outcomes and attrition at the relevant timepoints. - Conditions for verification: Link treatment start and end, assessments and follow-up to the relevant cognitive result; do not interpret unknown duration as a long-term effect. - Why this condition is needed: This is needed to avoid using the duration of related dementia research as the observation window for the current MCI cognitive-change question. - Application exceptions: Do not prescribe a universal numerical minimum treatment duration or long-term maintenance period. - Remaining uncertainty: Actual treatment, assessment, discontinuation and follow-up schedules are unverified. - Referenced methods: Common conditions C05, C06, E08 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N07 · Can baseline B1 status and the method for diagnosing deficiency in the MCI population be verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: Detailed baseline vitamin status and deficiency definitions are unverified and the value is null. Although the question excludes confirmed B1-deficiency encephalopathy, no verified value establishes that an actual eligible MCI population was nondeficient. - Why unverified: An explicit exclusion boundary does not verify actual baseline B1 status or its assessment method in an eligible population. - Reason for missing information: Baseline tests, deficiency definitions, exclusion procedures and results separated by baseline status are unverified. - Required materials and conditions: Baseline B1-status tables, testing/deficiency definitions, exclusion of deficiency encephalopathy and relevant subgroup results for an eligible MCI population. - Materials to obtain first: Baseline B1-status tables, testing/deficiency definitions, exclusion of deficiency encephalopathy and relevant subgroup results for an eligible MCI population. - Further research: Measure baseline nutritional status and prespecify hypotheses about effect modification by deficiency, reporting these separately from the main effect. - Conditions for verification: Verify nutritional status, deficiency definitions and baseline B1 values; distinguish treatment results for confirmed deficiency encephalopathy from cognitive effects in general MCI. - Why this condition is needed: This condition avoids converting responses to treatment of confirmed deficiency into supplementation efficacy in MCI with unknown deficiency status. - Application exceptions: Do not describe nutritional status as normal when deficiency tests are missing. Do not create new effect subgroups based on current deficiency status. - Remaining uncertainty: Baseline tests, deficiency definitions, exclusion procedures and results separated by baseline status are unverified. - Referenced methods: Common conditions C04, E08 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N08 · Can total dietary intake and other supplement exposure in each group be verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: Total dietary intake, all supplement exposure and group-specific details of all concomitant vitamins in an eligible comparison are unverified. Both values are null and the linked-source lists are empty. - Why unverified: No actual materials establish the distribution or changes in background exposure for a directly eligible comparison. - Reason for missing information: Total exposure to dietary B1 and other supplements, and changes by group, are unverified. - Required materials and conditions: Dietary B1, supplement composition, intake, exposure changes and their analytical handling by group. - Materials to obtain first: Dietary B1, supplement composition, intake, exposure changes and their analytical handling by group. - Further research: Record diet and supplement use in both groups using the same methods; explicitly address differences and changes in background exposure in the analysis. - Conditions for verification: Verify total B1 exposure and group distributions or changes in diet and supplements; do not attribute other nutritional differences to ordinary B1 alone. - Why this condition is needed: Even when asking about adding a single ingredient, differences in background nutritional exposure matter for interpreting its contribution. - Application exceptions: Do not convert unverified total intake into 0 intake or absence of all supplement use. - Remaining uncertainty: Total exposure to dietary B1 and other supplements, and changes by group, are unverified. - Referenced methods: Common conditions C05, E02, H05 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N09 · Can an actual comparator differing only in ordinary B1 be verified for an eligible MCI result? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The question specifies comparison with placebo, no addition or the same usual care. Eligible between-group MCI results and the actual comparison are unverified; the record relating the study’s primary endpoint to this table’s endpoint has a null value. - Why unverified: The question defines a comparator, but actual group treatments and usual care for an eligible result could not be verified. - Reason for missing information: Actual comparator composition and materials separating the between-group contribution of B1 alone are unverified. - Required materials and conditions: Treatment/control composition, shared usual care, background treatment by group and actual results for an eligible MCI comparison. - Materials to obtain first: Treatment/control composition, shared usual care, background treatment by group and actual results for an eligible MCI comparison. - Further research: Construct a comparison differing only in the addition of ordinary B1; record usual care and treatment changes in both groups to estimate the between-group effect. - Conditions for verification: Verify the actual placebo, no-addition or identical-usual-care comparator and explain whether ordinary B1 is the only difference between groups. - Why this condition is needed: This is needed to avoid presenting the comparison planned in the question as a comparison that was actually performed. - Application exceptions: Absence of placebo alone does not automatically invalidate a result; record the actual comparison and confounding limitations. - Remaining uncertainty: Actual comparator composition and materials separating the between-group contribution of B1 alone are unverified. - Referenced methods: Common conditions C05, E01, E02 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N10 · Can concomitant medicines, rescue treatment and treatment changes that may affect cognition be verified by group? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: Group-specific details of concomitant vitamins in an eligible comparison are unverified. No separate comprehensive search of all drug interactions was performed, and the current inputs lack materials establishing actual concomitant medicines, rescue treatment or treatment changes. - Why unverified: Concomitant-treatment and change information for a direct MCI comparison has not been verified, preventing an assessment of attribution to ordinary B1 using actual data. - Reason for missing information: Concomitant medicines, rescue treatment, treatment changes and handling of the effect of interest by group are unverified. - Required materials and conditions: Concomitant medicines and vitamins, rescue treatment, adherence/treatment changes and their analytical handling for an eligible MCI result. - Materials to obtain first: Concomitant medicines and vitamins, rescue treatment, adherence/treatment changes and their analytical handling for an eligible MCI result. - Further research: Collect concomitant and rescue treatments in both groups and prespecify how assignment effects and actual-use effects will be distinguished. - Conditions for verification: Verify concomitant vitamins, cognitive-disease treatment, rescue treatment and discontinuations or changes; distinguish their contribution from that of ordinary B1. - Why this condition is needed: This is needed to protect attribution to the selected ingredient without assuming shared treatment in the usual-care question. - Application exceptions: Unknown concomitant details do not mean there was no concomitant treatment; not every change must be identical across groups. - Remaining uncertainty: Concomitant medicines, rescue treatment, treatment changes and handling of the effect of interest by group are unverified. - Referenced methods: Common conditions E02, E04, H05 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N11 · Was the cognitive-function question distinguished from blood vitamers and mechanistic markers? Verification status: Verified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The preserved summary distinguished the patient-function question of validated cognitive-test change from blood vitamers. The biochemical/clinical distinction record has a null value but a verified information state; advertising and mechanistic statements were not adopted as direct clinical evidence. - Why unverified: The endpoint’s meaning and the boundary against substituting mechanisms or advertising are explicit in the current source. Actual effect values have not been verified. - Reason for missing information: Eligible cognitive-test results remain separately unverified. - Required materials and conditions: Current source and nutritional-extraction records showing the distinction between the question’s endpoint and biochemical markers. - Materials to obtain first: Current source and nutritional-extraction records showing the distinction between the question’s endpoint and biochemical markers. - Further research: Even if future work collects mechanistic and biochemical markers, report cognitive-test results as a separate endpoint and analyze relationships between markers separately. - Conditions for verification: Retain the selected endpoint as change on a patient’s cognitive test; do not substitute biochemical, mechanistic or advertising statements for its direct effect. - Why this condition is needed: This boundary prevents automatically extending improvements in blood vitamers to improvements in patient cognition. - Application exceptions: Verification that this is a patient-function question does not verify an actual cognitive effect. - Remaining uncertainty: Eligible cognitive-test results remain separately unverified. - Referenced methods: Common conditions C06, E03, E09 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N12 · Can the cognitive scale, direction and measurement procedures for an eligible MCI result be verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The question of validated cognitive-test change is fixed, but the specific scale in an eligible MCI trial is unverified. The value linking actual endpoint results is also null, so scale direction, measurement procedures and blinding could not be verified. - Why unverified: No results or methods materials establish a directly eligible cognitive scale and assessment conditions. - Reason for missing information: Scale name, unit, direction, measurement schedule, assessor, blinding/repeat-testing handling and results are unverified. - Required materials and conditions: Cognitive-scale specifications, assessment procedures, assessors/blinding and timepoint-specific results in a comparative MCI study. - Materials to obtain first: Cognitive-scale specifications, assessment procedures, assessors/blinding and timepoint-specific results in a comparative MCI study. - Further research: Specify a validated scale suitable for MCI and report results using consistent assessment schedules and procedures while addressing repeat-testing effects. - Conditions for verification: Link a validated scale’s name, unit, direction of improvement, assessment timing, assessors/blinding and handling of practice effects to its results. - Why this condition is needed: This is needed to avoid combining scores and measurement differences from different cognitive tests into one effect. - Application exceptions: When the specific scale is unverified, do not invent ADAS-Cog, CDR or another scale from adjacent research. Unknown blinding does not establish blinding failure. - Remaining uncertainty: Scale name, unit, direction, measurement schedule, assessor, blinding/repeat-testing handling and results are unverified. - Referenced methods: Common conditions C06, E03 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N13 · Can registration and the planned primary endpoint of an eligible MCI trial be compared with the actual report? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: Independent comparison of registration and the primary endpoint is unverified. The registered-primary-endpoint record and study/table-endpoint relationship have null values; clinical-trial registration numbers for both adjacent reports are also null. - Why unverified: No materials were identified to compare registration and plans with cognitive results for a directly eligible trial. A null clinical-trial registration number is not evidence of nonregistration. - Reason for missing information: Actual registration identity, primary endpoint, analysis plan, consistency with reporting and reasons for changes are unverified. - Required materials and conditions: Registration, protocol, analysis plan and original report of the relevant cognitive result for an eligible MCI trial. - Materials to obtain first: Registration, protocol, analysis plan and original report of the relevant cognitive result for an eligible MCI trial. - Further research: Specify cognitive endpoints, primary analyses and timepoints before actual data collection; link registration/plans to all results and report reasons for changes. - Conditions for verification: Compare primary endpoints, timepoints and analyses in the registration/protocol and analysis plan with actual reports; verify reasons for changes. - Why this condition is needed: This is needed to avoid treating an unavailable plan as proof of selective reporting or absence of an eligible trial. - Application exceptions: Freezing this table’s conditions with knowledge of the current materials does not replace clinical-trial preregistration. - Remaining uncertainty: Actual registration identity, primary endpoint, analysis plan, consistency with reporting and reasons for changes are unverified. - Referenced methods: Common conditions C06, C09, E06 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: S09, S26 ### N14 · Can the actual design, allocation and concealment of an eligible ordinary-B1–MCI comparison be verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: No directly eligible human result has been verified. Design records for the two adjacent studies are strings summarizing exclusion reasons, not actual MCI-trial allocation methods. The original bias axis also states that unknown methods must not be invented and labeled as defects. - Why unverified: No methods materials establish actual design, allocation, concealment or baseline balance for a result on ordinary B1 alone in MCI. - Reason for missing information: Directly eligible design, allocation/concealment, baseline balance and relevant blinding information are unverified. - Required materials and conditions: Methods materials on design, randomization/concealment, baseline characteristics by group and blinding for a directly eligible study. - Materials to obtain first: Methods materials on design, randomization/concealment, baseline characteristics by group and blinding for a directly eligible study. - Further research: Design a controlled study suited to the cognitive effect of interest and report allocation, concealment, baseline characteristics and measurement blinding. Do not assign a design from currently unverified materials. - Conditions for verification: Link parallel/crossover/cluster and randomized/nonrandomized design, allocation/concealment and baseline balance to the selected cognitive result. - Why this condition is needed: This is needed to avoid claiming verified control of confounding in an MCI comparison based only on a trial title or the design of adjacent research. - Application exceptions: Unknown design does not justify marking crossover or cluster conditions not applicable. Retain any nonrandomized evidence with its design and limitations. - Remaining uncertainty: Directly eligible design, allocation/concealment, baseline balance and relevant blinding information are unverified. - Referenced methods: Common conditions E01, E03, E04 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: S09, S26 ### N15 · Can actual analysis denominators, attrition and missing data for MCI cognitive results be verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The actual analysis denominator for the relevant primary endpoint is unverified and its value is null. Both n fields are statements requiring only verified values from historical numerical records and prohibiting substitution of allocated participant counts. That separate file was not opened in this work, and the existing arithmetic-record list is empty. - Why unverified: The current inputs do not establish actual analysis counts, flow or missing-data handling for a direct cognitive result. References to separate records were not used as numerical denominators. - Reason for missing information: Allocation/completion/analysis denominators, measurement losses, dropout reasons/handling and effect uncertainty are unverified. - Required materials and conditions: Participant flow, numbers assessed/analyzed at each cognitive timepoint, dropout reasons and missing-data handling in direct MCI research. - Materials to obtain first: Participant flow, numbers assessed/analyzed at each cognitive timepoint, dropout reasons and missing-data handling in direct MCI research. - Further research: Disclose flow by group and analysis denominators by outcome; report reasons for missingness, handling and result sensitivity. Do not invent a universal minimum n. - Conditions for verification: Distinguish selection, allocation, completion, measurement and analysis denominators, and missingness reasons/handling; link these to the relevant cognitive result. - Why this condition is needed: This is needed to avoid substituting references to separate files or randomized participant counts for actual cognitive-analysis denominators. - Application exceptions: When an adjacent paper’s n field provides no number, do not infer a participant count or 0 dropouts. - Remaining uncertainty: Allocation/completion/analysis denominators, measurement losses, dropout reasons/handling and effect uncertainty are unverified. - Referenced methods: Common conditions C07, E04, H06 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: S09, S26 ### N16 · Can the definition and analysis of an actual between-group MCI effect be verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: Directly eligible MCI results and their analysis are unverified. The original effect-size review records only state that eligible observations were not verified and that required adjusted/crossover effects of interest were not replaced with unadjusted values. - Why unverified: Actual between-group MCI estimates, analysis populations and models could not be verified. Calculation-attempt records were not used as actual results. - Reason for missing information: Effect of interest, analysis population, within-/between-group definitions, adjustment, design handling and results are unverified. - Required materials and conditions: Effect-of-interest definition, analysis population, model, group results and between-group estimates for an eligible MCI comparison. - Materials to obtain first: Effect-of-interest definition, analysis population, model, group results and between-group estimates for an eligible MCI comparison. - Further research: Specify the effect of interest and primary analysis; report between-group differences at the same timepoint separately from within-group improvement. Include crossover/cluster handling required by the actual design. - Conditions for verification: Link assignment/adherence effects, ITT/PP, final values/change scores, between-/within-group definitions, adjustment models and design handling to actual results. - Why this condition is needed: This is needed to avoid transferring cognitive results from related research or within-group improvement to a between-group effect of ordinary B1 in MCI. - Application exceptions: Do not fill an unknown analysis population with ITT; the manuscript’s analysis-attempt records are not actual cognitive effects. - Remaining uncertainty: Effect of interest, analysis population, within-/between-group definitions, adjustment, design handling and results are unverified. - Referenced methods: Common conditions C06, E04 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N17 · Can direct cognitive-change effects, variability, CIs and an important difference be verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The original record explicitly states that the effect estimate and an eligible effect interval for the exact claim could not be verified. Effects were not approximated by converting P statistics, and a specific MCI scale is also unverified. - Why unverified: Materials establishing direct cognitive effects, variability/CIs and importance criteria are unavailable, preventing interpretation of magnitude, precision and importance. - Reason for missing information: Between-group effect, SD/SE, CI, denominator and evidence for scale-specific important differences are unverified. - Required materials and conditions: Scale-specific between-group effects, variability types, denominators, CIs and evidence for clinical-importance criteria for an eligible MCI result. - Materials to obtain first: Scale-specific between-group effects, variability types, denominators, CIs and evidence for clinical-importance criteria for an eligible MCI result. - Further research: Report the main cognitive effect and CI and specify evidence for an important difference, separating uncertainty from clinical meaning. Do not determine sufficiency from significance or study count alone. - Conditions for verification: Distinguish actual scales, between-group effects, SD/SE, denominators and CIs; compare with an evidence-based important difference or record why interpretation is unavailable. - Why this condition is needed: This is needed to avoid writing an unknown effect as 0 or interpreting ? and null as numerical small-effect or low-precision values. - Application exceptions: Do not estimate new effects from P values or invent universal numerical minimums for duration, sample size, study count or importance. - Remaining uncertainty: Between-group effect, SD/SE, CI, denominator and evidence for scale-specific important differences are unverified. - Referenced methods: Common conditions C07, E05 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N18 · Was S09’s cognitive-trial summary separated from a direct effect of ordinary B1 in MCI? Verification status: Verified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: S09 is preserved as Gibson and colleagues’ 2020 cognitive trial of benfotiamine in early Alzheimer disease. The current summary records the molecular-form difference and does not adopt it as a direct ordinary-B1–MCI result. - Why unverified: The adjacent report’s identity, endpoint, exclusion reason and nontransfer to direct evidence were verified within the source’s scope. - Reason for missing information: S09 numerical results, CIs and denominators absent from the current summary, and direct ordinary-B1–MCI effects, remain separately unverified. - Required materials and conditions: Preserved bibliography, intervention, population/endpoint, exclusion role and result-summary locations for S09. - Materials to obtain first: Preserved bibliography, intervention, population/endpoint, exclusion role and result-summary locations for S09. - Further research: If ordinary B1 in MCI is newly studied, link benfotiamine/dementia findings only as hypothesis background and perform a separate direct comparison with matching intervention and population. - Conditions for verification: Explain how S09’s benfotiamine/early-Alzheimer-disease endpoint differs from the current question; do not delete the adjacent report or count it as direct replication. - Why this condition is needed: This is needed to preserve the existence of human cognitive trials without inheriting results from other molecular forms or diseases as the current effect. - Application exceptions: This source provides no numerical results, CIs or denominators for S09. No new assessment of the whole paper’s effect direction or precision is made here. - Remaining uncertainty: S09 numerical results, CIs and denominators absent from the current summary, and direct ordinary-B1–MCI effects, remain separately unverified. - Referenced methods: Common conditions C04, C05, C06, C08, E07 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: S09 ### N19 · Was the negative ordinary-thiamine dementia result kept from becoming a conclusion of no effect in MCI? Verification status: Verified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: S26 is a cognitive trial of ordinary thiamine in Alzheimer dementia. The original body text mentions its negative result; because it is not an isolated MCI result, it was not transferred to MCI as no effect or repeated refutation. - Why unverified: The unfavorable adjacent finding, population difference and reason for nontransfer were preserved together. This verification neither establishes an MCI effect nor revalidates S26’s numerical results. - Reason for missing information: S26 statistical details and directly eligible MCI effects are unverified in the current inputs. - Required materials and conditions: S26 population/endpoint, ordinary-B1 identity, preserved results and original body-text locations addressing the negative result. - Materials to obtain first: S26 population/endpoint, ordinary-B1 identity, preserved results and original body-text locations addressing the negative result. - Further research: Obtain results by directly comparing ordinary B1 in MCI; interpret differences from dementia evidence separately by population and disease stage. - Conditions for verification: Preserve both S26’s established-Alzheimer-dementia population and the manuscript’s negative-result statement; do not transfer these to an MCI effect of 0, equivalence or repeated refutation. - Why this condition is needed: This is needed to retain unfavorable adjacent findings while respecting applicability limits arising from differences in the actual question. - Application exceptions: The source provides no actual S26 numerical results, P values/CIs or denominators; do not invent statistical no-difference findings or effect sizes. - Remaining uncertainty: S26 statistical details and directly eligible MCI effects are unverified in the current inputs. - Referenced methods: Common conditions C04, C08, E07 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: S26 ### N20 · Were the benfotiamine paper’s identifier correction and both reports’ bibliographic links preserved accurately? Verification status: Verified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The original correction identifies S09 as PMID 33074237 and DOI 10.3233/JAD-200896 and records that a potentially mislinked PMID was not adopted. Both reports’ bibliographic/publisher URLs were preserved unchanged; no external recheck was performed in this work. - Why unverified: The current manuscript’s identifier correction and actual bibliographic links were checked. This does not mean the original external paper’s identity was independently verified in this work. - Reason for missing information: Current publisher-material status, full text and any additional corrections or retractions were not rechecked. - Required materials and conditions: Exact original wording of current citation/evidence records and the PMID/DOI linkage correction. - Materials to obtain first: Exact original wording of current citation/evidence records and the PMID/DOI linkage correction. - Further research: When obtaining materials later, compare PMID, DOI, authors/year and intervention identity; preserve before/after identifiers and the grounds for any correction. - Conditions for verification: Link the original correction’s PMID/DOI to citation/evidence identity without labeling this as a current external identity check or full-text verification. - Why this condition is needed: This is needed to avoid mixing a potentially mislinked paper identifier with different effect evidence. - Application exceptions: Do not interpret a publisher URL’s existence or its /fullarticle path as successful full-text reading. - Remaining uncertainty: Current publisher-material status, full text and any additional corrections or retractions were not rechecked. - Referenced methods: Common conditions C02, C03, C09 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: S09, S26 ### N21 · Were the two adjacent reports kept from being counted as independent replications of direct MCI trials? Verification status: Verified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The two adjacent reports are distinguished as Benfotiamine-earlyAD-2020 and Nolan-AD-1991 and excluded for intervention and population differences, respectively. The original record likewise did not count them as replication/refutation of the exact MCI claim. 2 is the cited-report count; unique direct-trial and participant counts are unverified. - Why unverified: The identities, exclusion roles and nonattribution to direct replication of both current reports were checked. No unique direct-trial count or effect of 0 was established. - Reason for missing information: Actual unique direct-trial and participant counts, and complete family relationships requiring comparison of registrations/full texts and related reports, are unverified. - Required materials and conditions: Study identifiers, bibliography and roles for both reports, plus family-linkage materials for additional registrations or related reports. - Materials to obtain first: Study identifiers, bibliography and roles for both reports, plus family-linkage materials for additional registrations or related reports. - Further research: When collecting materials, link reports to unique trial families and manage follow-ups, repeat citations and subgroup analyses of the same trial as one family. - Conditions for verification: Link bibliographic/study identifiers and exclusion roles; distinguish report counts from unique directly eligible trial and participant counts. - Why this condition is needed: This is needed to avoid counting two different citations as 2 direct MCI replications or 2 independent refutations. - Application exceptions: Identifiers alone do not fully verify institutional or participant independence across studies. Leave undetermined direct-trial and participant counts null. - Remaining uncertainty: Actual unique direct-trial and participant counts, and complete family relationships requiring comparison of registrations/full texts and related reports, are unverified. - Referenced methods: Common conditions C02, E07 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: S09, S26 ### N22 · Were current local-summary reading and historical publisher-abstract access distinguished for each material? Verification status: Verified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The entire preserved manuscript and exact quotations for each relevant row were read in this work. Historical access declarations for both reports are publisher abstracts. External abstracts and full texts were unattempted and unread in this work; individual access dates were left null because the original citation records lack them. The manuscript cut-off of 2026-09-15 was preserved separately. - Why unverified: The difference between actual current reading and historical abstract declarations was recorded by material. Historical self-checks are not reused as current clinical full-text reading or independent review. - Reason for missing information: Direct verification of current existence, availability, access and results of external abstracts/full texts was not performed. - Required materials and conditions: Exact locations, values and hashes in the preserved materials, original access levels and source locations for both reports, and material-specific reading records. - Materials to obtain first: Exact locations, values and hashes in the preserved materials, original access levels and source locations for both reports, and material-specific reading records. - Further research: If external materials are actually obtained later, separately record timing, access/acquisition/reading locations by material type and compare full-text results with the current summaries. - Conditions for verification: Separate actual reading of the current preserved manuscript summaries from both external abstracts/full texts being unattempted and unread; preserve historical publisher-abstract access, source locations and null individual dates. - Why this condition is needed: This display condition is required to avoid upgrading a link or historical abstract access to current full-text reading. - Application exceptions: Not attempted does not mean access failed. When a citation record lacks an individual access date, leave it null; do not copy the original review date into the access date. - Remaining uncertainty: Direct verification of current existence, availability, access and results of external abstracts/full texts was not performed. - Referenced methods: Common conditions C03, C09 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: S09, S26 ### N23 · Can the original search’s databases, strategies and selection scope be traced in the current inputs? Verification status: Partly verified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The original search cut-off is 2026-09-15 and a historical-search-record path is recorded. The limitation that not every primary database worldwide was exhaustively searched, and the bounded conclusion that no directly eligible result was identified, are preserved. The separate historical search record was not opened here, so the actual complete search strategies and selection flow were not verified. - Why unverified: The cut-off, log reference and bounded scope were verified, but actual databases, search strategies and selection flow cannot be reproduced from the current inputs alone. - Reason for missing information: Actual search sources, strategies and selection flow in the original log, and the subsequent current literature status, are unverified. - Required materials and conditions: Actual search sources, strategies, timing, selection/exclusion flow and coverage of unpublished/registered materials and corrections in the historical search record. - Materials to obtain first: Actual search sources, strategies, timing, selection/exclusion flow and coverage of unpublished/registered materials and corrections in the historical search record. - Further research: When a subsequent update is authorized, reproducibly record question-specific search sources, strategies, selection and links to registered/unpublished materials; state whether the search was exhaustive. - Conditions for verification: Preserve the evidence cut-off and bounded search conclusion; distinguish whether actual search strategies, databases and exclusion flow have been obtained. - Why this condition is needed: This is needed to avoid extending absence statements in the preserved summaries to absence from all current literature. - Application exceptions: No new clinical search is performed here, and the separate historical-search-record path is not claimed to have been opened. Incomplete searching is not displayed as 0 studies. - Remaining uncertainty: Actual search sources, strategies and selection flow in the original log, and the subsequent current literature status, are unverified. - Referenced methods: Common conditions C09, E06 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N24 · Were the original ?, absent score, unused axes and bounded no-eligible-human-result marker preserved unchanged? Verification status: Verified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The original verdict is ? with no score; replication, independence, effect, bias and precision other than B/P are null. The no-eligible-human-result marker is true, human-research existence is null, and eligible-human-result identification is false. The original interpretation is limited to failure to identify a directly eligible result in the documented search, distinct from no effect or worldwide absence. - Why unverified: Original types, values and the bounded interpretation were preserved precisely without conversion into a new clinical score, an effect of 0 or equivalence. - Reason for missing information: Direct effects and the worldwide existence of human research not established by the original status remain separately unverified. - Required materials and conditions: Exact wording of the original grade, score, assessment axes, human-research existence and eligible-result flags, calculation records and bounded unverified interpretation. - Materials to obtain first: Exact wording of the original grade, score, assessment axes, human-research existence and eligible-result flags, calculation records and bounded unverified interpretation. - Further research: If eligible materials are verified later, record previous status, dates and evidence separately from the new decision event. Do not derive clinical scores from row counts or verification rates in this table. - Conditions for verification: Preserve the original grade/score, B/P, null unused axes, false and empty arrays, and the bounded no-eligible-human-result meaning; do not convert these into new efficacy/safety scores or an effect of 0. - Why this condition is needed: This meaning boundary is required to prevent reading the original ? as total absence of human research, no effect or E0/RX/F. - Application exceptions: The original absent score and null unused axes are not 0. Passing current structural checks is separate from assessing clinical efficacy. - Remaining uncertainty: Direct effects and the worldwide existence of human research not established by the original status remain separately unverified. - Referenced methods: Common conditions C01, C08, C09, E05 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N25 · Can research funding, product provision and conflicts of interest for direct MCI evidence be verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: Funding fields for both adjacent reports are strings stating that funding is unverified or not used for this page’s grade. The original independence axis is null; independence was not assigned to a question without a directly eligible result. - Why unverified: No materials establish actual funding, provision, conflicts of interest or independent replication for direct MCI results. - Reason for missing information: Research funding, product provision, author/institution relationships, declared interests and independence of direct evidence are unverified. - Required materials and conditions: Funding, conflicts of interest, product provision, author/institution materials and related trial-family records for direct MCI reports. - Materials to obtain first: Funding, conflicts of interest, product provision, author/institution materials and related trial-family records for direct MCI reports. - Further research: Future research should disclose funding, product provision and author relationships; link families of replication studies and conflicts separately for interpretation. - Conditions for verification: Verify actual funding, product provision, author employment/patents, declared interests and study families; do not interpret unknown information as independent funding. - Why this condition is needed: This is needed to avoid treating unknown funding in adjacent reports or the original null independence axis as evidence of independent funding or replication. - Application exceptions: Do not automatically exclude evidence because of conflicts or infer absence of bias because no conflict is reported. Do not assign a new independence axis when direct results are unavailable. - Remaining uncertainty: Research funding, product provision, author/institution relationships, declared interests and independence of direct evidence are unverified. - Referenced methods: Common conditions C10, C02 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: S09, S26 ### N26 · Can adverse-event definitions, collection and denominators be verified under the same ordinary-B1–MCI exposure? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: Actual dose, duration and safety denominators for the exact MCI question are unverified. The adverse-event value is null; the source explicitly avoids presenting tolerability from other molecular forms or dementia trials as equivalent. - Why unverified: Actual event collection, denominators and event values in the same population/exposure could not be verified. Unknown information is not evidence of safety or 0 adverse events. - Reason for missing information: Event definitions, collection methods, severity, affected-person/event counts, exposure denominators, duration and discontinuations are unverified. - Required materials and conditions: Adverse-event definitions, collection methods, events/exposure denominators by group, severity and discontinuations in ordinary-B1 MCI research. - Materials to obtain first: Adverse-event definitions, collection methods, events/exposure denominators by group, severity and discontinuations in ordinary-B1 MCI research. - Further research: Systematically collect harms in the same population/exposure separately from efficacy outcomes; disclose affected-person counts, event counts, denominators and duration together. - Conditions for verification: Link actual dose, duration, event definitions/severity, collection methods, affected-person/event counts and exposure denominators. - Why this condition is needed: This is needed to avoid turning the existence of a cognitive trial or summary into a claim of safety or 0 adverse events. - Application exceptions: Do not transfer tolerability from other molecular forms or dementia trials as equivalent. Do not calculate incidence without denominators. - Remaining uncertainty: Event definitions, collection methods, severity, affected-person/event counts, exposure denominators, duration and discontinuations are unverified. - Referenced methods: Common conditions E10, H01, H04, H09, C07 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N27 · Can exposure timing, comparisons and follow-up losses for harms be verified in the same MCI materials? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: Duration and safety denominators for direct MCI exposure are unverified, and the current summary contains no verified materials on harm timelines, comparisons or follow-up losses in the same population. Tolerability from other molecular forms or dementia was not substituted. - Why unverified: Materials linking event timing, comparisons and follow-up scope under the same population/exposure are unavailable, preventing a judgment about risk or causality. - Reason for missing information: Exposure/onset/recovery, discontinuation/re-exposure, between-group event comparisons, detection differences and long-term follow-up/losses are unverified. - Required materials and conditions: Event timelines, discontinuation/recovery, group comparisons, follow-up losses and completeness of event disclosure under the same ordinary-B1 MCI exposure. - Materials to obtain first: Event timelines, discontinuation/recovery, group comparisons, follow-up losses and completeness of event disclosure under the same ordinary-B1 MCI exposure. - Further research: Use the same safety follow-up procedures in both groups; collect exposure, onset/recovery and attrition, and report limitations of comparative risk. - Conditions for verification: Link exposure start/stop, event onset/recovery, follow-up, comparators and attrition/detection by group; temporal sequence alone does not establish causality. - Why this condition is needed: This is needed to avoid upgrading harms unseen in short or incomplete summaries to safety, causality or total incidence. - Application exceptions: Do not infer improvement after discontinuation or re-exposure when these are absent. Unverified information also does not mean a harm event occurred. - Remaining uncertainty: Exposure/onset/recovery, discontinuation/re-exposure, between-group event comparisons, detection differences and long-term follow-up/losses are unverified. - Referenced methods: Common conditions H02, H03, H06, H08, H09, E10 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N28 · Was drug-interaction information for ordinary B1 in MCI sufficiently verified? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The original record explicitly states that no separate comprehensive search of all drug interactions was performed. Actual group-specific concomitant-vitamin details and safety information for the exact MCI exposure are also unverified. - Why unverified: No separate search and unknown actual concomitant-treatment details cannot be converted into safety or absence of interactions. - Reason for missing information: Actual concomitant medicines/exposure, direct evidence of specific interactions and comprehensive verification coverage are unverified. - Required materials and conditions: Actual concomitant medicines, dose, timing and materials/search coverage and results for specific interactions. - Materials to obtain first: Actual concomitant medicines, dose, timing and materials/search coverage and results for specific interactions. - Further research: Subsequent safety assessments should record actual concomitant medicines and review suitable exposure/outcome materials for each specified interaction question. - Conditions for verification: Distinguish drug-interaction search/verification coverage from actual concomitant exposure; do not interpret absence of a separate search as absence of interactions. - Why this condition is needed: This is needed to avoid extending the efficacy manuscript’s scope into a guarantee of individual safety with all medicines. - Application exceptions: Failure to perform a comprehensive search alone establishes neither presence nor absence of interactions. Do not create individual dosing instructions. - Remaining uncertainty: Actual concomitant medicines/exposure, direct evidence of specific interactions and comprehensive verification coverage are unverified. - Referenced methods: Common conditions C09, E10, H05, H08 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N29 · Does the current table claim an optimal dose, persistence after discontinuation or long-term cognitive effects? Verification status: Not applicable Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The current table addresses only comparative cognitive change with ordinary B1 in MCI. With direct dose, duration and effect unverified, it makes no claim about optimal dose, persistence after discontinuation or long-term efficacy/safety. - Why unverified: Because optimal-dose and long-term-maintenance extensions are not added to the current original claim, verification of those extensions is not required here. - Reason for missing information: Direct results for extension questions are unverified. Not applicable does not mean no long-term effect. - Required materials and conditions: Direct dose comparisons, discontinuation/long-term follow-up and exposure/timepoint-specific results if an extension question is newly addressed. - Materials to obtain first: Direct dose comparisons, discontinuation/long-term follow-up and exposure/timepoint-specific results if an extension question is newly addressed. - Further research: If an extension is needed, separately plan dose comparisons and discontinuation/long-term follow-up; report direct estimates and safety information. - Conditions for verification: Separate the fixed comparative cognitive-change question from optimal-dose or long-term-maintenance extensions, and do not claim those extensions. - Why this condition is needed: This supplementary boundary prevents extending an unverified main effect to a specific regimen or long-term effect. - Application exceptions: Do not require all dose and long-term studies for the current narrow cognitive-change question. Original dose/duration uncertainty remains in other rows. - Remaining uncertainty: Direct results for extension questions are unverified. Not applicable does not mean no long-term effect. - Referenced methods: Common conditions E08, H08 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N30 · Is cognitive-test change extended to dementia prevention, quality of life or mortality effects? Verification status: Not applicable Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The selected question concerns change on validated cognitive tests; dementia incidence, quality of life and mortality were not adopted as independent endpoints. Established-Alzheimer-dementia results are likewise not transferred as MCI effects. - Why unverified: Because no separate clinical-event extension is claimed, event results are not required for the current cognitive-change table. - Reason for missing information: Direct dementia-prevention, quality-of-life and survival results were not verified within the current question. - Required materials and conditions: Population, denominators, duration and comparative results directly measuring dementia incidence/clinical function, quality of life or survival if an extension question is addressed. - Materials to obtain first: Population, denominators, duration and comparative results directly measuring dementia incidence/clinical function, quality of life or survival if an extension question is addressed. - Further research: To evaluate clinical events or quality of life, plan separate research with clear event definitions and appropriate follow-up/denominators, and report independently of cognitive scores. - Conditions for verification: Separate cognitive-test change itself from clinical events, quality of life and survival; do not use the current results as evidence for those separate questions. - Why this condition is needed: This condition avoids automatically extending a cognitive-test question to dementia prevention or every improvement in patients’ lives. - Application exceptions: Do not require an event trial for a numerical-change question. Unverified direct cognitive effects do not establish an event effect of 0. - Remaining uncertainty: Direct dementia-prevention, quality-of-life and survival results were not verified within the current question. - Referenced methods: Common conditions C06, E09 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N31 · Can presence or absence of interference be judged within the original assay-interference verification scope? Verification status: Unverified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: The original assay-interference value is null and its information state is no separate search performed. The statement that no separate systematic search was performed and that this must not be read as absence of interference was preserved. The selected endpoint is cognitive-test change, distinct from blood vitamers. - Why unverified: No separately verified interference materials support a judgment about presence or absence. This is neither a cognitive-effect verdict nor a finding of measurement failure. - Reason for missing information: Interference materials for specific assays/exposures and separate verification coverage are unverified. - Required materials and conditions: Materials on specific assays, actual B1 exposure and interference/errors, plus separate search/verification coverage. - Materials to obtain first: Materials on specific assays, actual B1 exposure and interference/errors, plus separate search/verification coverage. - Further research: For a subsequent question, specify the assay actually used, compare exposure and sampling/assessment timing with assay principles, and perform the necessary validation. - Conditions for verification: Do not convert absence of a separate assay-interference search into absence of interference; limit relevance to cognitive assessment or separate biochemical tests to the verified scope. - Why this condition is needed: This supplementary condition avoids extending an unsearched interference question to error-free assessment tools or biochemical-assay safety. - Application exceptions: The current cognitive-change question does not itself claim assay interference. Absence of a separate search does not establish that cognitive-trial measurement errors occurred either. - Remaining uncertainty: Interference materials for specific assays/exposures and separate verification coverage are unverified. - Referenced methods: Common conditions C06, C09, E03 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ### N32 · Were the original update conditions and the separate records not opened here clearly documented? Verification status: Verified Reading scope: The relevant locations and context of the preserved review summaries were read. External clinical abstracts and full texts were not read in this work. - Current findings: New eligible evidence, full-text/registration results, numerical/unit changes, source corrections and retractions were preserved as subsequent update triggers. Separate historical search records, historical verification reports, historical numerical records, the calculator, earlier verdict 162 and examples were not opened here. The original verdict remains a publishable record of current uncertainty. - Why unverified: The original update policy was distinguished from the actual unread-reference scope. Acquisition of subsequent materials or public deployment is not claimed to have been completed. - Reason for missing information: The current contents, existence and additional results of each referenced file, and the latest external status, were not checked within this scope. - Required materials and conditions: Original update policy, editorial references and identities, dates and scope of subsequent materials/corrections, with before/after records. - Materials to obtain first: Original update policy, editorial references and identities, dates and scope of subsequent materials/corrections, with before/after records. - Further research: In later updates, preserve actual new materials, dates and reading levels together with previous values and change reasons; separately review linked rows and whether conditions need revision. - Conditions for verification: Distinguish new-evidence, full-text/registration acquisition, numerical/identity correction and retraction events from manuscript/condition revisions; do not upgrade unread referenced files into the actual reading scope. - Why this condition is needed: This is needed to avoid presenting reference paths and historical self-checks as current reinvestigation or automatic verification. - Application exceptions: Only reference strings for historical search records, historical numerical records, the calculator, earlier 162 and supporting examples were read; their individual sources were not opened. The possibility of an update does not establish that a result currently exists. - Remaining uncertainty: The current contents, existence and additional results of each referenced file, and the latest external status, were not checked within this scope. - Referenced methods: Common conditions C03, C09, E06 were individually adapted to the ordinary-B1-alone MCI question. These are case-specific conditions referencing preserved methodological principles, not official methodological certification. - Related sources: ## Evidence and sources ### S09 Gibson et al. — benfotiamine and cognitive decline in early Alzheimer disease (2020) [Official source](https://journals.sagepub.com/doi/abs/10.3233/JAD-200896) Materials actually read: The relevant evidence, citations and full context in the preserved Korean and English manuscript were read. External abstracts and full texts were unattempted, unacquired and unread in this work. Findings recorded in the summary: The cognitive trial used benfotiamine, a different molecular intervention from ordinary thiamine. Record level: Preserved adjacent-study summaries and their attribution in the body text were retained. These are neither direct ordinary-thiamine–MCI results nor a new clinical assessment. Independent replication: This benfotiamine/early-Alzheimer-disease report is not a direct result for ordinary B1 in MCI. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Publisher abstract | Current public availability unverified | External access not attempted in this work | Unread in this work | The historically recorded access level is publisher abstract. Individual access dates are unverified. An abs or fullarticle URL path does not establish the actual reading level; no external access, acquisition or reading occurred in this work. | | Clinical-paper full text | Direct full-text URL and current public availability unverified | External access not attempted in this work | Unread in this work | The direct full-text URL and actual reading are unverified. The report link is a preserved bibliographic connection; historical abstract access is not copied to the full text. External access, acquisition and reading were not performed in this work. | | Preserved verdict summary | Preserved material | Preserved summary accessed and read | Preserved summary read | This is a preserved summary read in the current work. It is distinct from the external clinical abstract/full text; historical abstract access is not upgraded to full-text reading. | ### S26 Nolan et al. — a trial of thiamine in Alzheimer disease (1991) [Official source](https://jamanetwork.com/journals/jamaneurology/fullarticle/590593) Materials actually read: The relevant evidence, citations and full context in the preserved Korean and English manuscript were read. External abstracts and full texts were unattempted, unacquired and unread in this work. Findings recorded in the summary: Human cognitive research on ordinary thiamine does exist, but this study involved Alzheimer dementia. It does not provide an isolated MCI result for the present claim. The Korean and English body text records a negative ordinary-thiamine dementia result. This is not transferred to MCI to declare no effect or repeated refutation. Record level: Preserved adjacent-study summaries and their attribution in the body text were retained. These are neither direct ordinary-thiamine–MCI results nor a new clinical assessment. Independent replication: This ordinary-thiamine report involved Alzheimer dementia and provides no isolated MCI result. The body text’s negative-result statement is not transferred to an MCI effect of 0. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Publisher abstract | Current public availability unverified | External access not attempted in this work | Unread in this work | The historically recorded access level is publisher abstract. Individual access dates are unverified. An abs or fullarticle URL path does not establish the actual reading level; no external access, acquisition or reading occurred in this work. | | Clinical-paper full text | Direct full-text URL and current public availability unverified | External access not attempted in this work | Unread in this work | The direct full-text URL and actual reading are unverified. The report link is a preserved bibliographic connection; historical abstract access is not copied to the full text. External access, acquisition and reading were not performed in this work. | | Preserved verdict summary | Preserved material | Preserved summary accessed and read | Preserved summary read | This is a preserved summary read in the current work. It is distinct from the external clinical abstract/full text; historical abstract access is not upgraded to full-text reading. | ## Review criteria and methods These are Chamgap's own application conditions, informed by the relevant sections of official methodological guidance. They do not constitute official assessment-tool certification or a new efficacy score. - M03 [Cochrane Handbook 6.5 (2024), chapter updated 2023-08](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-03) · §3.1, §3.2.1–3.2.3, MECIR C5/C7/C13 · Clearly document the question, eligibility conditions and reasons for changes, and distinguish intervention variants. · Preserved methodological principles, application records and copy hashes were compared. No new external methodology documents were queried and no complete official assessment tool was applied. - M04 [Cochrane Handbook 6.5.1 (2025), chapter updated 2025-03](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-04) · §4.5, §4.6.1–4.6.3, MECIR C42 · Distinguish and link studies and reports; do not omit studies solely because results are unreported. · Preserved methodological principles, application records and copy hashes were compared. No new external methodology documents were queried and no complete official assessment tool was applied. - M08 [Cochrane Handbook 6.5 (2024), chapter updated 2019-10](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-08) · §8.1–8.2.3, Table 8.2.a · Fix the outcome and effect of interest when reviewing randomization, attrition, missing data, measurement and selective reporting. · Preserved methodological principles, application records and copy hashes were compared. No new external methodology documents were queried and no complete official assessment tool was applied. - M15 [Cochrane Handbook 6.5 (2024), chapter updated 2023-08](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-15) · §15.2.4, §15.3.1–15.3.3 · Distinguish effect values, intervals, importance and applicability. Sample size alone does not determine precision. · Preserved methodological principles, application records and copy hashes were compared. No new external methodology documents were queried and no complete official assessment tool was applied. - M19 [Cochrane Handbook 6.5 (2024), chapter updated 2019-10](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-19) · §19.1.2, §19.2.1–19.2.4 · Harm questions may require different designs, populations and follow-up from efficacy questions. Case reports do not establish incidence. · Preserved methodological principles, application records and copy hashes were compared. No new external methodology documents were queried and no complete official assessment tool was applied. ## What the statuses mean - Verified: Verification of the specified question, identity, attribution and current work scope; separate from establishing clinical efficacy - Partly verified: Some actual information about the question is verified, while remaining data for that condition are unverified - Unverified: Direct information about the actual condition could not be verified in the current preserved inputs; not 0, no effect or worldwide absence - Unread: The external material was not actually read in this work; local-summary reading is separate - Not applicable: Nonapplicability limited to extension questions not required by the current fixed claim ## Change history 2026-10-31 · Public page/materials 1.0. Public display and data were created from the existing 32 items. No new clinical search, verification-status change or scoring was performed.