# Berberine hydrochloride plus lifestyle intervention × Reduction of ¹H-MRS liver fat in adults with NAFLD Verdict 3069 · Evidence Check Stored-material review date: 2026-10-05 Original evidence review date: 2026-09-08 Review-criteria version: 1.0 Public page/material preparation date: 2026-10-27 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 NAFLD. Detailed diagnosis, eligibility, diabetes composition, disease stage, concomitant medicines and baseline liver fat in S1 remain unverified. They are not assumed identical to the diabetes-free obese MASLD population in S2. - Ingredient, formulation and route: Oral berberine hydrochloride tablets with standardized lifestyle intervention. Botanical source, product purity, batch, manufacturing specifications and actual lifestyle implementation remain unverified. Extracts, combination products, other derivatives and topical products are not interchangeable. - Dose: The nominal S1 dose is 500 mg 3 times daily (1.5 g/day). Actual intake, adherence, changes and discontinuation remain unverified. It is not a direct dose comparison with 1 g/day in S2 or an instruction for personal use. - Comparator: The selected S1 comparator is lifestyle intervention alone. The overall study is a 3-arm open-label randomized parallel trial; specifications of the other arm, common management and rescue treatment remain unverified. Distinguish the double-blind placebo comparator in S2. - Measured outcome: Relative change in 16-week ¹H-MRS liver fat. Retain reductions of 52.7% versus 36.4% and P=.008 without converting them into a percentage-point difference. Distinguish the original statement that the registry endpoint was secondary from the currently unread registration and analysis plan. - Duration: S1 has a nominal duration of 16 weeks. Individual actual exposure, assessment windows, post-discontinuation measurements and persistence remain unverified. The 6 months in S2 belong to a trial under different conditions, not extension follow-up of the same participants. - Questions outside this scope: Original C, 50 points, Warning and B/S/R0/I2/EX/B2/CX are retained. Checklist statuses are not new efficacy scores or clinical certification.; S1 records 184 randomized participants and completion counts of 53/47/55. Arm assignment of those completion counts, actual 16-week analysis counts, dispersion, and the exact between-arm effect and CI remain unverified. No new sums, attrition rates, effects or intervals are calculated.; The original verdict describes S1 liver fat as secondary in the registry. It is not converted into prespecified primary-endpoint E+ success, and current direct comparison of the original registration, protocol and SAP is not claimed.; S2 is a different comparison: 337 participants, diabetes-free obese MASLD, 1 g/day, placebo, 6 months and MRI liver fat. Retain the original between-group 0.9 percentage points, 97.5% CI −0.4~2.1 and the reported lack of benefit. This is not converted into proof of 0 effect, equivalence or failure of the exact S1 comparison.; S1 and S2 differ in dose, population, comparator, duration and measurement. They are not merged into direct dose-response evidence, an interaction by diabetes status, long-term persistence or independent replication of the exact question, and participant counts are not added.; Weight, HOMA-IR, lipids, liver enzymes, MRI-PDFF, histologic recovery and cirrhosis prevention are distinguished from the selected liver-fat question. The result is not extended to benefit in every person with MASLD or to personal safety.; Retain the original warning about gastrointestinal adverse effects, hypoglycemia/interactions in liver disease or diabetes-medicine use, and pregnancy, lactation and infants. Event counts, safety denominators, exposure time, causation and long-term/rare harms remain unverified.; Retain the distinction between the 2026-09-14 record of wording correction in 2 English fields and the generic English descriptions still present. New reader English corresponds to retained Korean content and is not presented as new clinical verification or a new verdict. ## Checklist items ### N01 · Which liver-fat question is being checked? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The original claim is reduction of ¹H-MRS liver fat in adults with NAFLD. The selected scope is the 16-week comparison of berberine hydrochloride plus lifestyle intervention. - Why unverified: This checks the question boundary in the retained text. Clinical sufficiency and full-text verification are separate. - Reason for missing information: This checks the question boundary in the retained text. Clinical sufficiency and full-text verification are separate. - Required materials and conditions: The current original claim, formulation, population, comparator, measurement and timing - Materials to obtain first: The complete current text and claim identity - Further research: If the question changes, review different interventions, comparators and endpoints as separate comparisons. - Conditions for verification: Link the original claim unchanged and identify 1.5 g/day plus lifestyle, 16 weeks, relative ¹H-MRS change and the lifestyle comparator. - Why this condition is needed: The claim boundary must be fixed to avoid mixing results from other MASLD trials. - Application exceptions: Cirrhosis prevention or histologic recovery is not imposed as a requirement for the 16-week liver-fat question. - Remaining uncertainty: This checks the question boundary in the retained text. Clinical sufficiency and full-text verification are separate. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N02 · Do the actual product and salt form match? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 records purified berberine hydrochloride tablets plus standardized lifestyle intervention. The botanical source is unverified. S2 product details are not established in the retained summary. - Why unverified: Exact specifications, purity, batch and manufacturing records are absent, and whether the product matches S2 is unknown. - Reason for missing information: Exact specifications, purity, batch and manufacturing records are absent, and whether the product matches S2 is unknown. - Required materials and conditions: Hydrochloride tablets, product specifications, purity, batch, manufacturing and botanical source - Materials to obtain first: Methods, product supplements and supply records for each trial - Further research: Distinguish product variants and do not treat an unverified product as interchangeable. - Conditions for verification: Link tablet specifications, manufacturing and composition/batch records for each trial to establish whether they match the selected intervention. - Why this condition is needed: An ingredient name alone does not make different extracts, derivatives or combination products the same intervention. - Application exceptions: The berberine description in S2 does not automatically fill in the exact product or salt form in S1. - Remaining uncertainty: Exact specifications, purity, batch and manufacturing records are absent, and whether the product matches S2 is unknown. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N03 · What are the nominal dose and actual intake? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The nominal S1 dose is 500 mg 3 times daily (1.5 g/day), while S2 describes 1 g/day under different conditions. - Why unverified: Actual S1 intake, adherence criteria, arm-specific changes, rescue treatment and their analysis handling remain unverified. - Reason for missing information: Actual S1 intake, adherence criteria, arm-specific changes, rescue treatment and their analysis handling remain unverified. - Required materials and conditions: Nominal 500 mg TID, actual intake, adherence, discontinuation and exposure - Materials to obtain first: Dosing/return records, arm-specific adherence and discontinuation tables, and the analysis plan - Further research: Retain actual dosing/change information relevant to the effect of interest and do not combine it with other doses. - Conditions for verification: Link the nominal schedule to actual arm-specific intake/adherence, changes and discontinuation to assess actual exposure. - Why this condition is needed: Differences between prescribed dose and actual exposure affect interpretation of the effect of interest and harms. - Application exceptions: The 1 g/day in S2 is not used as actual intake or a dose-reduction record for S1. - Remaining uncertainty: Actual S1 intake, adherence criteria, arm-specific changes, rescue treatment and their analysis handling remain unverified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N04 · Which adults with NAFLD were selected? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 concerns adults with NAFLD and retains an unverified note that detailed diabetes composition depends on the paper. S2 concerns a different population: diabetes-free obese adults with MASLD. - Why unverified: Detailed S1 diagnosis, diabetes proportion, disease stage, concomitant medicines and baseline state cannot be established from the retained summary. - Reason for missing information: Detailed S1 diagnosis, diabetes proportion, disease stage, concomitant medicines and baseline state cannot be established from the retained summary. - Required materials and conditions: Diagnosis, eligibility/exclusions, age, baseline liver fat, diabetes composition and concomitant medicines - Materials to obtain first: Eligibility criteria, baseline characteristics and comorbidity/medicine tables - Further research: Limit interpretation to populations with direct material; application to other populations is a separate question. - Conditions for verification: Check S1 diagnosis, eligibility and baseline state/diabetes composition to explain matches and differences from the selected population. - Why this condition is needed: The labels NAFLD and diabetes-free obese MASLD alone do not establish identical populations. - Application exceptions: The diabetes-free obese eligibility in S2 is not retrospectively assigned to S1. - Remaining uncertainty: Detailed S1 diagnosis, diabetes proportion, disease stage, concomitant medicines and baseline state cannot be established from the retained summary. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N05 · What was the actual comparator strategy? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The selected S1 comparator is lifestyle intervention alone. The overall design is a 3-arm open-label randomized parallel trial. S2 is a double-blind placebo trial. - Why unverified: The summary lacks details of the remaining S1 arm specifications, management, common medicines and arm-specific changes. - Reason for missing information: The summary lacks details of the remaining S1 arm specifications, management, common medicines and arm-specific changes. - Required materials and conditions: Lifestyle alone, medicines/management in each of the 3 arms, and the S2 placebo - Materials to obtain first: Arm-specific treatments/management and rescue-treatment criteria in the methods - Further research: Specify the complete comparator and interpret trials with different common management separately. - Conditions for verification: Check whether actual S1 arm contents, the selected lifestyle comparator and rescue/common management correspond to original material. - Why this condition is needed: Different comparator contents change the question answered by a between-arm difference. - Application exceptions: The open-label lifestyle comparator is not converted into a blinded placebo comparator. - Remaining uncertainty: The summary lacks details of the remaining S1 arm specifications, management, common medicines and arm-specific changes. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N06 · How was background lifestyle intervention applied in each arm? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 records comparison of added berberine with lifestyle intervention alone against a background of standardized lifestyle intervention. - Why unverified: The current summary lacks details of diet/activity prescriptions, actual implementation, counseling, rescue treatment and arm differences. - Reason for missing information: The current summary lacks details of diet/activity prescriptions, actual implementation, counseling, rescue treatment and arm differences. - Required materials and conditions: Diet, activity, visits, education, support, actual implementation and differences - Materials to obtain first: Lifestyle protocol, arm-specific implementation/change records and analysis plan - Further research: Observe common management and actual implementation and analyze them in relation to the add-on effect of interest. - Conditions for verification: Check lifestyle contents, implementation/changes and analysis implications in both arms to interpret the added-drug difference. - Why this condition is needed: In an add-on comparison, differences in lifestyle implementation cannot be merged into the selected drug contribution. - Application exceptions: Identical behavioral change is not imposed, but actual arm differences and their handling must be checked. - Remaining uncertainty: The current summary lacks details of diet/activity prescriptions, actual implementation, counseling, rescue treatment and arm differences. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1 ### N07 · Do the 16-week treatment and observation periods match? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 records 16 weeks. The 6-month MRI result in S2 is a report under different conditions. - Why unverified: Individual S1 exposure, assessment windows, early discontinuation and measurements after discontinuation remain unverified. - Reason for missing information: Individual S1 exposure, assessment windows, early discontinuation and measurements after discontinuation remain unverified. - Required materials and conditions: Start, actual treatment, 16-week assessment, discontinuation and follow-up schedule - Materials to obtain first: Trial schedule, visit/discontinuation tables and assessment-time records - Further research: Retain differences between the specified outcome time and actual observation; persistence requires separate follow-up. - Conditions for verification: Check S1 treatment, assessment, discontinuation and missingness timing and link it to the 16-week result. - Why this condition is needed: Nominal treatment duration cannot be converted into actual exposure or every participant’s observation time. - Application exceptions: The 6 months in S2 are not counted as extension follow-up for S1. - Remaining uncertainty: Individual S1 exposure, assessment windows, early discontinuation and measurements after discontinuation remain unverified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N08 · Which test measured liver fat? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 records relative change in 16-week ¹H-MRS hepatic fat content. The retained description distinguishes it from MRI-PDFF. S2 concerns 6-month MRI liver fat. - Why unverified: Equipment, acquisition/region, reading/quality control and detailed S2 MRI technique are absent from the summary. - Reason for missing information: Equipment, acquisition/region, reading/quality control and detailed S2 MRI technique are absent from the summary. - Required materials and conditions: ¹H-MRS equipment, acquisition, region, quality control and test timing - Materials to obtain first: Protocols for each test and reading/quality-control supplements - Further research: Match measurement methods and conditions; do not substitute different technologies as an identical measure. - Conditions for verification: Check test protocol, assessment region, quality control and identical arm-specific measurement to interpret the selected result. - Why this condition is needed: A matching test name alone does not establish measurement conditions or bias. - Application exceptions: S1 ¹H-MRS is not replaced by MRI-PDFF, and the detailed MRI technique in S2 is not inferred. - Remaining uncertainty: Equipment, acquisition/region, reading/quality control and detailed S2 MRI technique are absent from the summary. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N09 · Can relative reduction be distinguished from percentage points? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 records liver-fat reductions of 52.7% versus 36.4% and P=.008. S2 records a between-group 0.9 percentage points and 97.5% CI −0.4~2.1 under different measures and conditions. - Why unverified: S1 baseline values, original definitions, dispersion and the exact between-arm estimate/CI are absent; the values cannot be compared as the same unit. - Reason for missing information: S1 baseline values, original definitions, dispersion and the exact between-arm estimate/CI are absent; the values cannot be compared as the same unit. - Required materials and conditions: Baseline/final liver fat, relative-change definition, dispersion and between-arm estimate units - Materials to obtain first: Result tables, measure definitions, baseline/final values, dispersion and between-arm estimates - Further research: Retain original units and do not convert, add or generate new effects without original data. - Conditions for verification: Check each result’s measure, reference value, direction, unit and actual between-arm estimate/interval to explain whether they match. - Why this condition is needed: Numbers under different units and times must not be read as the same effect. - Application exceptions: 52.7%/36.4% are not converted into percentage points, and no new difference against S2 is calculated. - Remaining uncertainty: S1 baseline values, original definitions, dispersion and the exact between-arm estimate/CI are absent; the values cannot be compared as the same unit. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N10 · What was the planned hierarchy of the selected liver-fat endpoint? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The claim conditions label 16-week liver fat as the principal endpoint, but the original reader answer/grade rationale record that it was secondary in the registry. - Why unverified: The initial registration, protocol and SAP were not read currently, so the original locations for hierarchy and changes were not compared. - Reason for missing information: The initial registration, protocol and SAP were not read currently, so the original locations for hierarchy and changes were not compared. - Required materials and conditions: Primary/secondary status and change times in the initial registration, protocol and SAP - Materials to obtain first: Initial/amended registrations, protocol, SAP and reported-endpoint table - Further research: Retain the original endpoint-hierarchy description; distinguish changes from added evidence after actual version comparison. - Conditions for verification: Check liver-fat hierarchy, change timing and reasons in relevant registration/plan versions and compare them with actual reporting. - Why this condition is needed: A principal-outcome label and prespecified primary designation are different information. - Application exceptions: Failure to read the registry currently does not establish nonregistration or selective reporting. - Remaining uncertainty: The initial registration, protocol and SAP were not read currently, so the original locations for hierarchy and changes were not compared. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1 ### N11 · Were planned analyses/times compared with actual reporting? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The original research-review field states that the NCT00633282 paper/registration were compared with the NCT05647915 trial. - Why unverified: Only the retained summary was read currently; direct plan/registration items and analysis-change tables are absent. - Reason for missing information: Only the retained summary was read currently; direct plan/registration items and analysis-change tables are absent. - Required materials and conditions: Planned analysis populations, missing-data handling, adjustment, timing and reasons for changes - Materials to obtain first: Plan/SAP, registration history and actual-analysis sections in the paper - Further research: Link material that permits planned-versus-reported comparison; lack of comparison does not establish selective reporting. - Conditions for verification: Compare planned and actual analysis, timing and population for the selected result using original versions, retaining differences and reasons for changes. - Why this condition is needed: Having seen a result does not establish agreement with the planned analysis. - Application exceptions: A historical registration-comparison statement in the public text is not promoted to current SAP reading. - Remaining uncertainty: Only the retained summary was read currently; direct plan/registration items and analysis-change tables are absent. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N12 · What was the randomized design? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 is recorded as a 3-arm open-label randomized parallel trial. - Why unverified: Allocation ratio, sequence generation, implementation responsibilities and allocation after recruitment are absent from the retained summary. - Reason for missing information: Allocation ratio, sequence generation, implementation responsibilities and allocation after recruitment are absent from the retained summary. - Required materials and conditions: 3-arm parallel randomization, allocation ratio, sequence generation and recruitment/allocation timing - Materials to obtain first: Trial methods, randomization/allocation supplements and protocol - Further research: Check actual allocation and implementation; a label alone does not establish absence of bias. - Conditions for verification: Check the actual design, random sequence/allocation ratio and implementation for the selected comparison to explain confounding limitations. - Why this condition is needed: A randomized label alone does not verify the actual allocation process. - Application exceptions: Unverified design is not replaced by nonapplicability of crossover/cluster designs. - Remaining uncertainty: Allocation ratio, sequence generation, implementation responsibilities and allocation after recruitment are absent from the retained summary. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1 ### N13 · Was allocation concealed? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 retains only the design label of open-label randomization. - Why unverified: Direct methods concerning concealment, role separation and predictability are absent. - Reason for missing information: Direct methods concerning concealment, role separation and predictability are absent. - Required materials and conditions: Allocation-concealment method, predictability and enrollment/allocation responsibilities - Materials to obtain first: Allocation-concealment methods, implementation records and protocol - Further research: Check actual allocation procedures relevant to selection bias; do not describe an unverified item as established failure. - Conditions for verification: Check who could see the sequence before allocation and the evidence of actual concealment implementation. - Why this condition is needed: Open-label treatment and allocation concealment are different procedures that require separate checks. - Application exceptions: An open-label design alone does not establish allocation-concealment failure. - Remaining uncertainty: Direct methods concerning concealment, role separation and predictability are absent. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1 ### N14 · How were assessor blinding and arm-specific measurements handled? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 records open-label treatment and liver-fat measurement by ¹H-MRS. - Why unverified: Assessor blinding, identical arm-specific measurement and test-quality exclusion criteria cannot be established from the summary. - Reason for missing information: Assessor blinding, identical arm-specific measurement and test-quality exclusion criteria cannot be established from the summary. - Required materials and conditions: Blinded reading, identical tests, measurement/exclusion decisions and missingness - Materials to obtain first: Reading/acquisition protocols, blinding explanations and test-exclusion tables - Further research: Assess limitations according to observed measurement conditions; inability to blind does not automatically establish efficacy failure. - Conditions for verification: Check assessor access to information and arm-specific differences in testing, reading and quality control to explain measurement limitations. - Why this condition is needed: Objective tests can still be affected by acquisition, reading and exclusion decisions. - Application exceptions: Participant open-label information does not automatically determine assessor-blinding status. - Remaining uncertainty: Assessor blinding, identical arm-specific measurement and test-quality exclusion criteria cannot be established from the summary. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1 ### N15 · Were baseline states balanced in the selected comparison? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 concerns adults with NAFLD, but detailed diabetes composition remains unverified. - Why unverified: Arm-specific baseline liver fat, diabetes/weight, concomitant medicines, balance and adjustment information are absent. - Reason for missing information: Arm-specific baseline liver fat, diabetes/weight, concomitant medicines, balance and adjustment information are absent. - Required materials and conditions: Arm-specific liver fat, diabetes/weight, medicines, sex and baseline risk - Materials to obtain first: Baseline characteristics, selection flow and adjustment model - Further research: Check analysis information relevant to the direct population and arm-specific baseline differences. - Conditions for verification: Check arm-specific baseline tables, selection processes and relevant adjustments to explain important differences. - Why this condition is needed: The results of selection and allocation are needed to interpret actual between-arm differences. - Application exceptions: S2 population composition is not evidence of baseline balance in S1. - Remaining uncertainty: Arm-specific baseline liver fat, diabetes/weight, concomitant medicines, balance and adjustment information are absent. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1 ### N16 · What were the randomized, completed and analyzed counts in each arm? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 retains records of 184 randomized participants and completion counts of 53/47/55. - Why unverified: The summary lacks completion-count arm assignment and actual 16-week analysis n, so denominators cannot be established. - Reason for missing information: The summary lacks completion-count arm assignment and actual 16-week analysis n, so denominators cannot be established. - Required materials and conditions: Arm-specific randomized n, 16-week measured n, completed n and n for each analysis - Materials to obtain first: CONSORT flow and arm-specific randomized/measured/completed/analyzed tables - Further research: Directly check denominator type and arm attribution; do not infer unreported denominators. - Conditions for verification: Link randomized, measured, completed and analyzed counts for the selected arms and the overall 3 arms to actual flow/result tables. - Why this condition is needed: Using 184 or completion counts as the selected effect’s analysis denominator would introduce error. - Application exceptions: Completion counts of 53/47/55 are not assigned to arms, summed or used to calculate new attrition rates. - Remaining uncertainty: The summary lacks completion-count arm assignment and actual 16-week analysis n, so denominators cannot be established. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1 ### N17 · How were missingness, attrition and outcome measurement handled? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The original reader answer identifies analysis-attrition limitations, and the S1 evidence-weight description identifies open-label treatment and attrition. - Why unverified: Arms, timing, reasons, missing-data handling and outcome implications of missingness/attrition are unverified. - Reason for missing information: Arms, timing, reasons, missing-data handling and outcome implications of missingness/attrition are unverified. - Required materials and conditions: Reasons/timing for missingness and attrition, arm differences, missing-data handling and sensitivity analyses - Materials to obtain first: Reasons for attrition/missingness, measurement status and missing-data/sensitivity analyses - Further research: Check missing-data handling and interpretation for the effect of interest; do not calculate new effects from attrition counts. - Conditions for verification: Check arm-specific missingness/attrition, liver-fat measurement, handling models and sensitivity analyses to explain limitations. - Why this condition is needed: Analysis attrition can affect interpretation of both actual effects and harms. - Application exceptions: Completion tables alone do not establish attrition causes, adverse-event discontinuations or random missingness. - Remaining uncertainty: Arms, timing, reasons, missing-data handling and outcome implications of missingness/attrition are unverified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1 ### N18 · Which analysis population and model produced the reported effect? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 is retained as reporting a 52.7% reduction with lifestyle plus berberine versus 36.4% with lifestyle alone, P=.008. - Why unverified: The summary lacks definitions of the result’s analysis population, adjustment model, baseline handling and between-arm change estimate. - Reason for missing information: The summary lacks definitions of the result’s analysis population, adjustment model, baseline handling and between-arm change estimate. - Required materials and conditions: ITT/PP/completers, between-arm change, adjustment, model and effect of interest - Materials to obtain first: Statistical-analysis section, SAP, complete result tables and analysis-population definitions - Further research: Distinguish assignment/adherence effects of interest from actual estimates; do not invent an analysis label. - Conditions for verification: Check the actual analysis population, model, adjustment and effect of interest for the selected between-arm result and explain alignment with the question. - Why this condition is needed: It is necessary to know whether arm-specific relative reductions estimate assignment effects or completer effects. - Application exceptions: Without analysis information, do not label the result as ITT, PP or adjusted effect. - Remaining uncertainty: The summary lacks definitions of the result’s analysis population, adjustment model, baseline handling and between-arm change estimate. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1 ### N19 · Were the 16-week between-arm estimate and interval adequately reported? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 retains reductions of 52.7% versus 36.4%/P=.008 but no exact between-arm effect/CI. S2 retains 0.9 percentage points/97.5% CI −0.4~2.1. - Why unverified: The exact between-arm estimate, interval, dispersion and model for the selected 16-week comparison are absent, preventing a completed precision assessment. - Reason for missing information: The exact between-arm estimate, interval, dispersion and model for the selected 16-week comparison are absent, preventing a completed precision assessment. - Required materials and conditions: Exact between-arm effect, CI level, dispersion, denominator and estimation model - Materials to obtain first: S1 between-arm estimate/interval, dispersion and denominator result tables - Further research: Retain reported numbers only; do not calculate new differences, CIs or meta-analyses. - Conditions for verification: Link the effect estimate, interval and dispersion/denominator for the selected 16-week comparison to interpret the range of effects it permits. - Why this condition is needed: Statistical significance alone cannot establish effect magnitude or precision. - Application exceptions: The 97.5% CI in S2 is not inherited as the S1 interval or changed into a 95% CI. - Remaining uncertainty: The exact between-arm estimate, interval, dispersion and model for the selected 16-week comparison are absent, preventing a completed precision assessment. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N20 · What defines clinical importance for liver-fat change? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The selected result is 16-week relative change in liver fat, a surrogate outcome. The retained summary has no importance threshold. - Why unverified: A justified importance threshold for the measure/population and the selected between-arm CI are lacking. - Reason for missing information: A justified importance threshold for the measure/population and the selected between-arm CI are lacking. - Required materials and conditions: A justified important liver-fat difference, measure, population, duration and interval - Materials to obtain first: Clinical-importance/measurement-change criteria and the S1 between-arm interval - Further research: Explain an evidence-based importance threshold first; do not create new efficacy scores or grades. - Conditions for verification: Check a justified importance threshold for relative ¹H-MRS change in the relevant population and its relationship to the selected effect/interval. - Why this condition is needed: Statistical P values and clinically meaningful changes must be distinguished. - Application exceptions: Do not invent universal minimum sample-size, duration or effect-magnitude thresholds. - Remaining uncertainty: A justified importance threshold for the measure/population and the selected between-arm CI are lacking. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1 ### N21 · Which study family is each report linked to? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: There are identity clues for S1: Yan2015/PMID26252777/PMC4529214/DOI10.1371/journal.pone.0134172/NCT00633282; and for S2: PMID41543854/NCT05647915. - Why unverified: Original-trial/family links, related-report lists and participant overlap for the reports were not currently verified against external originals. - Reason for missing information: Original-trial/family links, related-report lists and participant overlap for the reports were not currently verified against external originals. - Required materials and conditions: Reports, PMID/DOI/PMCID, registration identity, population/arms and overlapping participants - Materials to obtain first: Original papers, registration identity/family tables and related-report lists - Further research: Link reports separately from study families; do not count identifiers as independent trials. - Conditions for verification: Link each report’s identity and same/different/uncertain family relationship using original material, retaining the undetermined total unique-trial count. - Why this condition is needed: Multiple paper, bibliography and registration gateways do not imply multiple trials. - Application exceptions: Identifier strings alone do not establish trial families or total independent-trial counts. - Remaining uncertainty: Original-trial/family links, related-report lists and participant overlap for the reports were not currently verified against external originals. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N22 · Were the conditions of the opposing 2026 result retained? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S2 concerns 337 participants, 1 g/day, 6 months, diabetes-free obese MASLD, placebo and MRI. Retain the between-group 0.9 percentage points/97.5% CI −0.4~2.1 and the original null-result description. - Why unverified: This checks that the related opposing result and condition differences in the retained text are both displayed. It does not mean 0 effect, equivalence or review of all current literature. - Reason for missing information: This checks that the related opposing result and condition differences in the retained text are both displayed. It does not mean 0 effect, equivalence or review of all current literature. - Required materials and conditions: S2 population, 1 g/day, placebo, 6 months, MRI and original effect/interval - Materials to obtain first: Complete S2 paper, population/measure details and registration versions - Further research: Do not hide opposing results under different conditions; review individually whether they answer the same question. - Conditions for verification: Retain the recorded S2 direction, original numbers and condition differences together; do not combine them as the same estimate as S1. - Why this condition is needed: Keeping only the favorable 16-week result would lose related opposing results and applicability differences. - Application exceptions: S2 under different conditions is not converted into direct failure of the exact 16-week comparison or replication at the same dose. - Remaining uncertainty: This checks that the related opposing result and condition differences in the retained text are both displayed. It does not mean 0 effect, equivalence or review of all current literature. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N23 · Were unique trials and participants protected from double counting? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The evidence/citation lists have 2 entries each; the multiple identifiers for S1 are clues to one report. No total unique-trial/participant count was generated currently. - Why unverified: This checks the counting boundary that report-entry counts are not used as clinical independence or study sufficiency. - Reason for missing information: This checks the counting boundary that report-entry counts are not used as clinical independence or study sufficiency. - Required materials and conditions: Relationships among reports, trials, participants and endpoints/times/analyses - Materials to obtain first: Report-family and overlapping-participant linkage material - Further research: Group related reports into families before counting; do not create totals without independence information. - Conditions for verification: Retain established report relationships and uncertainties; do not generate overall unique-trial/participant totals without evidence. - Why this condition is needed: Simply adding Korean/English text, PubMed/PMC/registration gateways or reports under different conditions inflates evidence counts. - Application exceptions: Do not add 184 and 337 or sum completion counts of 53/47/55. - Remaining uncertainty: This checks the counting boundary that report-entry counts are not used as clinical independence or study sufficiency. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N24 · How were funding, product provision and interests checked? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 records Chinese public research support with manufacturer roles requiring separate checking; S2 records academic/public support. - Why unverified: Detailed funder names, product provision, manufacturer roles, author employment/patents, declared COI and design control are absent. - Reason for missing information: Detailed funder names, product provision, manufacturer roles, author employment/patents, declared COI and design control are absent. - Required materials and conditions: Funding, product supply/manufacturer roles, author employment/patents, COI and design control - Materials to obtain first: Original Funding/COI statements, product supply, author affiliations and control descriptions - Further research: Review funding, product provision and author interests separately; unverified information is not independent funding. - Conditions for verification: Link funding, product provision, author/patent/COI information and design-control scope for each report to actual disclosures. - Why this condition is needed: Checking public funding differs from checking product/design control and author interests. - Application exceptions: Academic/public support or the original I2 is not inherited as complete verification of independence. - Remaining uncertainty: Detailed funder names, product provision, manufacturer roles, author employment/patents, declared COI and design control are absent. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N25 · Which materials were acquired and read in this review? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The complete retained 3069 summary was actually read. External PubMed, PMC, DOI and registration originals were not accessed or read in this conversion. - Why unverified: This checks disclosure of actual reading scope and unread external material; it is not clinical full-text certification. - Reason for missing information: This checks disclosure of actual reading scope and unread external material; it is not clinical full-text certification. - Required materials and conditions: Current material type, original bytes, actual reading scope and external access/reading attempts - Materials to obtain first: Current input copies, actual reading events and material-specific access records - Further research: When new material is actually read, link its type, time, scope and original location/SHA separately. - Conditions for verification: Ensure acquisition, actual reading, times/locations/SHA for each material match actual actions, retaining unattempted external access. - Why this condition is needed: Reading a local summary must be distinguished from reading an external paper/registration in full. - Application exceptions: The existence of PMC/DOI/registration links is not converted into full-text acquisition or successful access. - Remaining uncertainty: This checks disclosure of actual reading scope and unread external material; it is not clinical full-text certification. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N26 · Were the evidence date and this review date distinguished? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The original verdict-review date is 2026-09-08. The current date concerns conversion of the retained summary; clinical searches and external accesses in this conversion total 0. - Why unverified: Separate evidence-search cutoff and individual access-date fields are absent. The current scope is the retained summary, not a judgment that no current evidence exists. - Reason for missing information: Separate evidence-search cutoff and individual access-date fields are absent. The current scope is the retained summary, not a judgment that no current evidence exists. - Required materials and conditions: Original verdict-review date, historical search records, current reading date and item-specific access dates - Materials to obtain first: Original search records, individual access/reading dates and evidence-date definition - Further research: Record dates and actual verification scope of additional material separately, updating only the linked rows. - Conditions for verification: Link the meaning of the original date, current reading date, unperformed search/access and the scope of absence statements. - Why this condition is needed: A historical text’s review date cannot be retrospectively assigned as current external access. - Application exceptions: The original 2026-09-08 is not used as a new search date or an individual PubMed/PMC access date. - Remaining uncertainty: Separate evidence-search cutoff and individual access-date fields are absent. The current scope is the retained summary, not a judgment that no current evidence exists. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N27 · How were citation registration identifiers and the structured registration field handled? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The S1/S2 evidence-table registration fields both say unverified. The citations record NCT00633282/NCT05647915 respectively. - Why unverified: The identifiers are retained identity clues rather than results of reading the original registrations currently. Structured-field and actual-registration linkage remain unverified. - Reason for missing information: The identifiers are retained identity clues rather than results of reading the original registrations currently. Structured-field and actual-registration linkage remain unverified. - Required materials and conditions: Bibliographic identifiers, original registration-check values, original-registry links and corrections - Materials to obtain first: Each original registration, identifier identity and report correspondence - Further research: Retain conflicting/unverified identifiers; do not upgrade linkage without original-material comparison. - Conditions for verification: Retain both unverified original registration-check fields and citation identifiers; verify exact trial links against original material. - Why this condition is needed: Unverified structured fields and bibliographic clues cannot overwrite each other. - Application exceptions: S1/S2 identifiers are not interchanged or presented as actual current registration comparison. - Remaining uncertainty: The identifiers are retained identity clues rather than results of reading the original registrations currently. Structured-field and actual-registration linkage remain unverified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N28 · Were generic English descriptions and the existing correction retained? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The existing correction records wording changes in 2 English fields on 2026-09-14, without new research or reassessment. Some English evidence fields retain generic descriptions. - Why unverified: This checks preservation of original values and historical corrections. C50/S/R0/I2/EX/B2/CX is not inherited as new condition statuses or a new verdict. - Reason for missing information: This checks preservation of original values and historical corrections. C50/S/R0/I2/EX/B2/CX is not inherited as new condition statuses or a new verdict. - Required materials and conditions: Original Korean/English values, 2026-09-14 correction scope and original verdict/score - Materials to obtain first: Complete before/after texts, correction event and original assessment axes - Further research: Keep new evidence additions separate from wording/condition changes and preserve complete before/after texts. - Conditions for verification: Link original Korean/English text, correction and score/axes unchanged; do not use generic descriptions to fill specific unknowns. - Why this condition is needed: Generic English descriptions cannot substitute for actual denominator/CI verification. - Application exceptions: The existence of a correction note is not promoted to additional clinical verification in this conversion. - Remaining uncertainty: This checks preservation of original values and historical corrections. C50/S/R0/I2/EX/B2/CX is not inherited as new condition statuses or a new verdict. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N29 · Were persistence and discontinuation effects beyond 16 weeks checked? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 concerns 16-week results; S2 is a 6-month report with different population, dose, comparator and measurement. - Why unverified: Extension/discontinuation follow-up, persistent effects and time-specific denominators under the same conditions are absent from the retained summary. - Reason for missing information: Extension/discontinuation follow-up, persistent effects and time-specific denominators under the same conditions are absent from the retained summary. - Required materials and conditions: Extension/post-discontinuation follow-up under the same conditions, with effects/denominators at additional times - Materials to obtain first: S1 extension/post-discontinuation follow-up and time-specific results/denominators - Further research: When claiming persistence, review a separate question using direct follow-up; do not impose an arbitrary minimum duration. - Conditions for verification: Link direct extension/discontinuation follow-up under the same 16-week conditions, with time-specific estimates and denominators, to the persistence question. - Why this condition is needed: Short-term liver-fat results alone cannot establish post-discontinuation persistence or long-term benefit. - Application exceptions: The 6 months in S2 are not used as extension follow-up of the same S1 participants. - Remaining uncertainty: Extension/discontinuation follow-up, persistent effects and time-specific denominators under the same conditions are absent from the retained summary. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N30 · Were the doses directly compared? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 at 1.5 g/day and S2 at 1 g/day also differ in duration, population, comparator and measurement. - Why unverified: The retained summary has no direct dose comparison under identical conditions, so optimal dose/dose response is not judged. - Reason for missing information: The retained summary has no direct dose comparison under identical conditions, so optimal dose/dose response is not judged. - Required materials and conditions: Randomized dose comparisons in the same population/comparator/test, with adherence and estimates - Materials to obtain first: Direct dose comparisons, actual dosing and dose-specific estimates/intervals - Further research: When extending the dose question, use a direct design that separates differences in other conditions. - Conditions for verification: Link direct dose-comparison design, actual exposure and between-arm estimates to optimal-dose/dose-response questions. - Why this condition is needed: Differences between 1.5 g and 1 g in separate trials do not show a direct dose response. - Application exceptions: A dose-comparison trial is not imposed unconditionally for the narrow 16-week question. - Remaining uncertainty: The retained summary has no direct dose comparison under identical conditions, so optimal dose/dose response is not judged. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N31 · Can effects by diabetes status and disease stage be distinguished? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: Detailed diabetes composition in S1 is unverified, while S2 concerns a different population of diabetes-free obese adults with MASLD. - Why unverified: Prespecified subgroups, disease stage, diabetes composition and direct interactions/denominators are absent. - Reason for missing information: Prespecified subgroups, disease stage, diabetes composition and direct interactions/denominators are absent. - Required materials and conditions: Prespecified subgroups, composition, interactions, estimates/intervals and denominators - Materials to obtain first: Prespecified subgroup plans, baseline composition, interactions and estimates/intervals - Further research: Subgroup application requires direct plans/estimates; different significance across separate trials is not an interaction. - Conditions for verification: Check prespecified relevant subgroups, composition, interactions and subgroup estimates/denominators to assess effect modification. - Why this condition is needed: Overall effects in different populations cannot be converted into effect modification by diabetes status. - Application exceptions: Diabetes-free eligibility in S2 alone does not establish an interaction relative to a diabetic group. - Remaining uncertainty: Prespecified subgroups, disease stage, diabetes composition and direct interactions/denominators are absent. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N32 · Does the liver-fat result answer cirrhosis or histologic-recovery questions? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The original advertising-limit description excludes cirrhosis prevention, histologic recovery and application to all MASLD. Weight, HOMA-IR, lipids and liver enzymes are separate endpoints. - Why unverified: The selected liver-fat material alone cannot establish direct evidence for separate clinical events or histologic recovery. - Reason for missing information: The selected liver-fat material alone cannot establish direct evidence for separate clinical events or histologic recovery. - Required materials and conditions: Direct data for the selected liver-fat measure versus histology, clinical events and quality of life - Materials to obtain first: Direct plans/results for histology, clinical events and quality of life - Further research: When extending to clinical-event questions, review a separate table of direct endpoints. - Conditions for verification: Review extensions to clinical events, histology or quality of life using their direct results, denominators and estimates. - Why this condition is needed: A reduction in a surrogate cannot automatically be extended to cirrhosis prevention or benefit in all MASLD. - Application exceptions: Not every clinical-event trial is imposed as required for a liver-fat numerical question. - Remaining uncertainty: The selected liver-fat material alone cannot establish direct evidence for separate clinical events or histologic recovery. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N33 · How were recorded gastrointestinal adverse effects defined and collected? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The original product-facts/safety-detail descriptions warn that gastrointestinal adverse effects occurred. - Why unverified: Specific events, collection methods, severity, arms, discontinuations and event counts are absent from the retained summary. - Reason for missing information: Specific events, collection methods, severity, arms, discontinuations and event counts are absent from the retained summary. - Required materials and conditions: Symptom/diagnosis definitions, collection methods, severity, harm-related discontinuation and each arm - Materials to obtain first: Arm-specific adverse-event definitions, collection, severity and discontinuation tables - Further research: Check actual harm counts and event definitions separately from efficacy. - Conditions for verification: Check event definitions, ascertainment/collection, severity and arm-specific reporting for the selected exposure to explain the harm-information level. - Why this condition is needed: Warnings or general symptom descriptions differ from complete harm counts. - Application exceptions: Keep liver-fat efficacy status separate from harm-record status. - Remaining uncertainty: Specific events, collection methods, severity, arms, discontinuations and event counts are absent from the retained summary. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N34 · How do harm onset/recovery and dosing timing relate? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: S1 treatment lasts 16 weeks, and the retained description warns of gastrointestinal adverse effects, but individual event timelines are absent. - Why unverified: Onset, recovery, discontinuation, re-exposure and event follow-up duration cannot be established from the retained summary. - Reason for missing information: Onset, recovery, discontinuation, re-exposure and event follow-up duration cannot be established from the retained summary. - Required materials and conditions: Exposure start, discontinuation, onset, recovery, re-exposure and follow-up duration - Materials to obtain first: Event timelines, dosing/discontinuation and recovery/re-exposure records - Further research: Obtain timing material appropriate to the harm question; temporal order alone does not establish cause. - Conditions for verification: Link actual exposure, harm onset/recovery timelines and observation limits to direct material. - Why this condition is needed: A nominal 16-week treatment period alone cannot establish event latency or recovery. - Application exceptions: Without recovery-after-discontinuation or re-exposure information, do not create a causal timeline. - Remaining uncertainty: Onset, recovery, discontinuation, re-exposure and event follow-up duration cannot be established from the retained summary. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1 ### N35 · Against which exposures and concurrent treatments were harms compared? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The original warning calls for checking hypoglycemia and interactions in people with liver disease or diabetes-medicine use. The S1 efficacy comparator is lifestyle alone. - Why unverified: Direct material on actual arm-specific harms, diabetes/concomitant medicines, baseline events and alternative explanations is absent. - Reason for missing information: Direct material on actual arm-specific harms, diabetes/concomitant medicines, baseline events and alternative explanations is absent. - Required materials and conditions: Arm-specific exposure, lifestyle, diabetes/concomitant medicines, baseline disease and event comparisons - Materials to obtain first: Arm-specific adverse events/concurrent exposures, baseline risks and adjustment information - Further research: Review harm comparisons and alternative causes individually; theoretical warnings are not observed incidence. - Conditions for verification: Link arm-specific events, exposure/concomitant medicines, baseline risks, alternative explanations and adjustments to explain comparative-harm limitations. - Why this condition is needed: Without concurrent-treatment, baseline-risk and actual event comparisons, a particular drug contribution cannot be established. - Application exceptions: The existence of an efficacy comparator is not converted into completed harm-event comparison. - Remaining uncertainty: Direct material on actual arm-specific harms, diabetes/concomitant medicines, baseline events and alternative explanations is absent. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N36 · Are event counts and exposure denominators available for harm frequency? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: Retained material records S1 with 184 randomized participants/completion counts of 53/47/55, S2 with 337 participants, and general warnings. - Why unverified: Arm-specific event counts, safety denominators, exposure time and harm intervals are absent, so frequency cannot be displayed. - Reason for missing information: Arm-specific event counts, safety denominators, exposure time and harm intervals are absent, so frequency cannot be displayed. - Required materials and conditions: People/events affected, arm-specific safety denominators, exposure time, discontinuation and CI - Materials to obtain first: Arm-specific safety denominators, event counts, exposure/follow-up and reported estimates - Further research: Link reported harm frequencies only to their denominators/times; do not calculate new incidence. - Conditions for verification: Directly link each event’s counts, arm-specific exposed/safety denominators, observation time and reported estimates. - Why this condition is needed: Randomized totals or warnings alone cannot be used to calculate harm frequency. - Application exceptions: Do not use 184, 337 or completion counts of 53/47/55 as safety denominators or create 0 events. - Remaining uncertainty: Arm-specific event counts, safety denominators, exposure time and harm intervals are absent, so frequency cannot be displayed. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N37 · Were missing harm reports and overlapping related reports checked? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: General warnings and S1 analysis-attrition limitations remain. There is no separate list of event-report families. - Why unverified: Harm collection/detection differences, events among dropouts, the complete disclosure scope and related-event overlap are unverified. - Reason for missing information: Harm collection/detection differences, events among dropouts, the complete disclosure scope and related-event overlap are unverified. - Required materials and conditions: Collection/detection, spontaneous/systematic reporting, attrition, related reports and opposing records - Materials to obtain first: Harm-collection plans/complete tables, attrition follow-up and original related-report/regulatory-citation material - Further research: Retain opposing/unreported records and actual event families; do not infer new replication counts. - Conditions for verification: Link actual event collection/selection, missingness, the full disclosure scope and related-report families/overlap to original material. - Why this condition is needed: Unreported harms differ from absent events, and repeat citations differ from independent events. - Application exceptions: Regulatory warnings or repeated citations of the same case are not counted as independent events. - Remaining uncertainty: Harm collection/detection differences, events among dropouts, the complete disclosure scope and related-event overlap are unverified. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N38 · What is known about long-term/rare harms and vulnerable groups? Verification status: Unverified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: The original warning mentions liver disease, diabetes-medicine use, pregnancy, lactation and infants. - Why unverified: Direct actual exposure/denominator and timing material for vulnerable groups, long-term/rare events and interactions is absent. - Reason for missing information: Direct actual exposure/denominator and timing material for vulnerable groups, long-term/rare events and interactions is absent. - Required materials and conditions: Long-term/rare events, liver disease/diabetes medicines, pregnancy/lactation/infants and interactions - Materials to obtain first: Follow-up, interaction and harm material appropriate to the relevant exposures/vulnerable groups - Further research: Review safety extensions using separate questions, designs and material; unverified information is not personal safety. - Conditions for verification: Explain direct harm material and applicability limits for the duration, vulnerable population or concurrent exposure to which claims are extended. - Why this condition is needed: Short-term results in a restricted population cannot be extended to every person’s safety. - Application exceptions: Not all rare-harm studies are imposed as efficacy requirements for the selected 16-week liver-fat question. - Remaining uncertainty: Direct actual exposure/denominator and timing material for vulnerable groups, long-term/rare events and interactions is absent. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N39 · How far does the current warning support harm statements? Verification status: Partly verified Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: Current retained material preserves the original warning about gastrointestinal adverse effects, hypoglycemia/interactions and vulnerable groups. - Why unverified: Current material does not provide detailed events/exposure, comparisons, timing and alternatives, so causation or incidence is not established. - Reason for missing information: Current material does not provide detailed events/exposure, comparisons, timing and alternatives, so causation or incidence is not established. - Required materials and conditions: Event reports, comparisons, denominators, timing, alternatives and the evidence level behind the warning - Materials to obtain first: Direct warning sources and event/exposure/alternative-explanation material - Further research: Describe only the observable harm-information level; do not generate safety certification or a new verdict. - Conditions for verification: Link the warning’s actual evidence level and distinguish verified scope and limits for signals, comparative harms, causation and frequency. - Why this condition is needed: General warnings differ from actual causation, incidence and personal-safety assessments. - Application exceptions: Do not remove the original warning or convert it into a new safety grade or personal dosing instruction. - Remaining uncertainty: Current material does not provide detailed events/exposure, comparisons, timing and alternatives, so causation or incidence is not established. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ### N40 · Which rows should be updated when new material arrives? Verification status: Met Reading scope: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. - Current findings: Current observations link to actual locations and SHA in the immutable 3069 input summary. Additional external-material acquisition/reading events in this conversion total 0. - Why unverified: This checks update-tracing rules, not new material acquisition, complete clinical review or completed publication. - Reason for missing information: This checks update-tracing rules, not new material acquisition, complete clinical review or completed publication. - Required materials and conditions: Identity, material type, date, actual reading, original location/SHA and condition/observation versions - Materials to obtain first: Additional material identity/version, original location/SHA and actual review event - Further research: Preserve the complete previous version, separate material additions from condition revisions, then recheck linked-row evidence, statuses and reading links. - Conditions for verification: Identify facts, locations, versions and linked rows for each added material; preserve the complete previous material and condition-revision event. - Why this condition is needed: New file collection must be tracked separately from condition revisions and actual review/display changes. - Application exceptions: New API responses or historical verification dates alone do not automatically satisfy a condition. - Remaining uncertainty: This checks update-tracing rules, not new material acquisition, complete clinical review or completed publication. - Referenced methods: Existing method principles distinguish population, formulation, dose, timing, endpoint, comparison, denominators, overlap, funding and harms in the retained reports. Detailed conditions are Chamgap’s own application. - Related sources: S1, S2 ## Evidence and sources ### S1 Yan HM et al. PLoS One. 2015;10:e0134172. PMID: 26252777. PMCID: PMC4529214. DOI: 10.1371/journal.pone.0134172. NCT00633282. [PubMed](https://pubmed.ncbi.nlm.nih.gov/26252777/) Materials actually read: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. Findings recorded in the summary: Lifestyle plus berberine: 52.7% reduction versus 36.4% with lifestyle alone, P=.008. Separately retain 184 randomized participants and completion counts of 53/47/55. Completion-count arm assignment, analysis n and the exact between-arm effect/CI remain unverified. Record level: A retained summary of positive results and open-label/attrition limitations. Preserve the original statement that the registry endpoint was secondary. Current external full-text, registration or SAP verification and new effect calculation were not performed. Independent replication: The retained summary has two report entries, S1/S2. PubMed, PMC, DOI, registration identifiers and Korean/English text for S1 are not independent trials. Family links and participant overlap between reports under different conditions were not newly checked against external originals; total unique trials and participants remain undetermined. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Retained judgment evidence/bibliographic summary | Retained summary acquired/read | Retained copy acquired; summary read | Stored existing review summary read | Retained evidence, bibliography and text summaries were read. Distinguish these from clinical full texts and original binaries. Historical individual access levels/dates are unrecorded. Original review narratives do not establish current external full-text reading. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | External PubMed bibliography/abstract gateway | Public availability unverified | External access not attempted; not acquired | Not read | A retained bibliographic gateway. Current existence, public availability, acquisition, content and reading of the external document were not newly verified. Historical individual access levels/dates are unrecorded. Original review narratives do not establish current external full-text reading. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | External PMC full-text gateway | Public availability unverified | External access not attempted; not acquired | Not read | A retained bibliographic gateway. Current existence, public availability, acquisition, content and reading of the external document were not newly verified. Historical individual access levels/dates are unrecorded. Original review narratives do not establish current external full-text reading. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | External DOI article gateway | Public availability unverified | External access not attempted; not acquired | Not read | A retained bibliographic gateway. Current existence, public availability, acquisition, content and reading of the external document were not newly verified. Historical individual access levels/dates are unrecorded. Original review narratives do not establish current external full-text reading. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | External trial-registration gateway | Public availability unverified | External access not attempted; not acquired | Not read | The registration number is a retained bibliographic clue. Preserve the unconfirmed registration-check field; actual original-registration/plan comparison was not performed. Historical individual access levels/dates are unrecorded. Original review narratives do not establish current external full-text reading. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | ### S2 2026 randomized trial. PMID: 41543854. NCT05647915. [PubMed](https://pubmed.ncbi.nlm.nih.gov/41543854/) Materials actually read: The retained Korean/English summaries and case-specific condition/application records were read. External abstracts, full texts, registration, protocols and analysis plans remain unread, with access not attempted. Findings recorded in the summary: 337 participants. Retain the original between-group 0.9 percentage points, 97.5% CI −0.4~2.1 and the statement of no liver-fat benefit. Analysis denominators, exact product and detailed MRI technique remain unverified. Record level: A related opposing result under different conditions. It is not converted into direct failure of the exact 16-week comparison, 0 effect, equivalence or new independent replication. External full text and registration remain unread. Independent replication: The retained summary has two report entries, S1/S2. PubMed, PMC, DOI, registration identifiers and Korean/English text for S1 are not independent trials. Family links and participant overlap between reports under different conditions were not newly checked against external originals; total unique trials and participants remain undetermined. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Retained judgment evidence/bibliographic summary | Retained summary acquired/read | Retained copy acquired; summary read | Stored existing review summary read | Retained evidence, bibliography and text summaries were read. Distinguish these from clinical full texts and original binaries. Historical individual access levels/dates are unrecorded. Original review narratives do not establish current external full-text reading. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | External PubMed bibliography/abstract gateway | Public availability unverified | External access not attempted; not acquired | Not read | A retained bibliographic gateway. Current existence, public availability, acquisition, content and reading of the external document were not newly verified. Historical individual access levels/dates are unrecorded. Original review narratives do not establish current external full-text reading. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | | External trial-registration gateway | Public availability unverified | External access not attempted; not acquired | Not read | The registration number is a retained bibliographic clue. Preserve the unconfirmed registration-check field; actual original-registration/plan comparison was not performed. Historical individual access levels/dates are unrecorded. Original review narratives do not establish current external full-text reading. The displayed material date is the case-record review date. Original verdict review, correction and case review dates are not historical access dates for individual external materials or search cut-offs. Historical external access dates and reading scope are unrecorded; external access was not attempted in this conversion. | ## 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 record the question, eligibility conditions and reasons for changes; distinguish intervention variants. · Retained method-application records were reused. External originals were not reread and method content was not reassessed. - 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 · Link studies and reports separately; do not omit studies solely because results are unreported. · Retained method-application records were reused. External originals were not reread and method content was not reassessed. - 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 result and effect of interest when examining randomization, deviations, missingness, measurement and selective reporting. · Retained method-application records were reused. External originals were not reread and method content was not reassessed. - 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 estimates, intervals, importance and applicability. Sample size alone does not determine precision. · Retained method-application records were reused. External originals were not reread and method content was not reassessed. - 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. · Retained method-application records were reused. External originals were not reread and method content was not reassessed. ## What the statuses mean - Partly verified: Some retained information is available while required unverified material remains. - Unverified: The criterion cannot be established within the retained scope; this does not mean 0 studies, 0 effect or no harm. - Review incomplete: Original-material comparison/search was not performed within the permitted scope. - Not reported: A scope-limited status meaning the item was not stated in the particular material/section that was read. - Inaccessible: Used only for material for which access was actually attempted and failed. It was not assigned to clinical material in this conversion. - Insufficient evidence matching the conditions: Used only after relevant studies within a defined scope have been reviewed and evidence is insufficient to answer the criterion. It was not applied indiscriminately in this conversion. - Not applicable: A specific exception based on applicability conditions, not a substitute for unread material. - Conflicting data: Used when incompatible information for the same question, time and analysis has been compared. Different review estimates alone do not establish this status. - Met: The specified retained information and reading/attribution/counting boundary were checked; this is not certification of efficacy, full-text reading or reassessment. ## Change history 2026-10-27 · Public page/materials 1.0. The public view and materials were prepared from the existing 40 items. No new clinical search, verification-state change or scoring was performed.