# Folic acid — sole added active ingredient on common enalapril therapy × Reduction of first symptomatic stroke in Chinese adults with hypertension Verdict 3107 · Evidence Check Stored-material review date: 2026-10-05 Original evidence date: 2026-09-15 Review-criteria version: NGM-v3107-FOLIC-ADDON-CSPPT-FIRST-STROKE-1.0 Public page/material preparation date: 2026-11-03 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: Chinese adults aged 45–75 with hypertension, no prior stroke or myocardial infarction, and tolerability and adherence during run-in - Ingredient, formulation and route: folic acid — added to common enalapril maleate therapy · tablets placed in matching capsules for blinding; release type, batch and purity unverified · oral - Dose: Folic acid 0.8 mg/day + enalapril maleate 10 mg/day; 1 time daily - Comparator: The same enalapril maleate 10 mg/day + no added folic acid; a common policy permitting additional antihypertensives - Measured outcome: Time from randomization to the first fatal or nonfatal symptomatic stroke; subarachnoid hemorrhage and silent stroke excluded - Duration: Planned duration 5 years; median follow-up 4.5 years; 3 weeks of run-in - Questions outside this scope: Recurrent stroke, TIA, silent stroke and subarachnoid hemorrhage; Homocysteine levels, composite cardiovascular events and reduction in all-cause mortality; Correction of diagnosed folate deficiency, optimal doses by genotype and universal prevention in other countries/fortification settings; Safety of B-vitamin combinations that also change other active ingredients, and products with different doses ## Checklist items ### N01 · Which comparison concerning first stroke is being asked about? Verification status: Met Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The original /ko/effect is reduction of first symptomatic stroke in Chinese adults with hypertension. The actual comparison is added folic acid versus the same enalapril, and the primary endpoint is time to the first fatal or nonfatal symptomatic stroke. subclaims is an empty list. - Why unverified: The exact original claim is linked to the scope of the current single table. This status did not establish benefit for other events. - Reason for missing information: The full original SOP was not read in this review. Eligibility criteria concerning prior TIA remain separately unverified in N07. - Required materials and conditions: The authoritative versions of the current claim, endpoint SOP and eligibility criteria, and a mapping to the original ID/URL. - Materials to obtain first: The authoritative versions of the current claim, endpoint SOP and eligibility criteria, and a mapping to the original ID/URL. - Further research: Use an eligibility table for each question to classify recurrence, TIA, silent events and composite cardiovascular events separately, and compare only reports of first events. - Conditions for verification: Link the selected Chinese adults with hypertension, folic acid as the sole added ingredient, identical background therapy and time to the first symptomatic event to the original /ko/effect. - Why this condition is needed: Required to avoid merging other outcomes on the page into a single prevention claim. - Application exceptions: Recurrence, TIA, subarachnoid hemorrhage, silent stroke and death do not substitute for the primary endpoint. - Remaining uncertainty: The full original SOP was not read in this review. Eligibility criteria concerning prior TIA remain separately unverified in N07. - Referenced methods: Question clarification principles in M03 §3.1, §3.2.1–3.2.4; detailed separation rules are internally designed; M03 §3.2.4, M15 §15.3.1–15.3.3; adaptations of outcomes to each question are internally designed; M03 §3.2.4; internally defined limits on extrapolation — an internal adaptation for the separate question. Not official certification of the method. - Related sources: S01, S02 ### N02 · Was folic acid the only difference between randomized groups? Verification status: Met Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The folic acid arm receives folic acid 0.8 mg + enalapril maleate 10 mg, and the control arm receives the same enalapril 10 mg without added folic acid. The direct comparison was not excluded because of the common antihypertensive medication. - Why unverified: The retained difference between the actual intervention and control is linked to the addition of folic acid. The overall effect of the combined prescription was not recast as the effect of folic acid alone versus no treatment. - Reason for missing information: Actual total exposure and medication changes for all participants are separate gaps in N09. - Required materials and conditions: Actual prescriptions by arm, the policy permitting background medication and a table of the common product composition. - Materials to obtain first: Actual prescriptions by arm, the policy permitting background medication and a table of the common product composition. - Further research: Use a table of intervention components by arm to compare the sole added active ingredient and common therapy, and keep results from other combination products separate. - Conditions for verification: Distinguish the addition of folic acid 0.8 mg from enalapril common to both arms, and indicate whether other active ingredients changed. - Why this condition is needed: Required to distinguish the contributions of common therapy and the selected intervention. - Application exceptions: Do not extrapolate to B-vitamin combinations that also differ in active ingredients other than folic acid, or to folic acid monotherapy without background treatment. - Remaining uncertainty: Actual total exposure and medication changes for all participants are separate gaps in N09. - Referenced methods: M03 §3.2.2–3.2.3; reverse-direction flagging and adaptations of checks are internally designed; M03 §3.2.3, M08 §8.2.2–8.2.3; internal application — an internal adaptation for the added active ingredient. Not official certification of the method. - Related sources: S01, S02 ### N03 · Have the salt, release type, batch and raw material been verified? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The verified ingredients are folic acid and common enalapril maleate. A separate folic acid salt designation is not_reported, while manufacturing raw material, release type, batch, purity and complete excipients are unverified. The wording of the AUSA supplier name varies by document. - Why unverified: The ingredients and supplier are linked, but detailed data on trial product identity remain incomplete. - Reason for missing information: Raw-material source, separate folic acid salt, release type, brand/batch, purity and complete excipients. - Required materials and conditions: Manufacturing specifications for the trial product, batch certificates, raw material and salt/release type, and the complete excipient list. - Materials to obtain first: Manufacturing specifications for the trial product, batch certificates, raw material and salt/release type, and the complete excipient list. - Further research: Link historical trial product documents to the actual formulations in each arm, and separately review whether the trial products are identical to current products with different doses. - Conditions for verification: Separate the verified folic acid and common enalapril maleate from the unverified raw material, separate folic acid salt, release type, batch/purity and excipients. - Why this condition is needed: Needed to prevent automatic extrapolation of evidence from different formulations. - Application exceptions: Do not substitute folinic acid or 5-MTHF. Do not conflate an unreported salt designation with an unverified active ingredient when the ingredient has been verified. - Remaining uncertainty: Raw-material source, separate folic acid salt, release type, brand/batch, purity and complete excipients. - Referenced methods: M03 §3.2.2–3.2.3; reverse-direction flagging and adaptations of checks are internally designed — an internal adaptation for manufacturing identity. Not official certification of the method. - Related sources: S01, S02 ### N04 · What are the daily dose and capsule blinding method? Verification status: Met Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The oral regimen is administered every 1 day in 1 dose: folic acid 0.8 mg + enalapril maleate 10 mg versus the same enalapril 10 mg. Tablets were placed in matching capsules for blinding. - Why unverified: Nominal dose, route, frequency and blinding by appearance are linked in the retained content. This does not mean actual individual intake or release type has been fully verified. - Reason for missing information: Original data on individual medication amounts, the complete capsule composition and release type are unverified. - Required materials and conditions: Actual trial medication administration instructions, tablet/capsule appearance specifications and individual medication records. - Materials to obtain first: Actual trial medication administration instructions, tablet/capsule appearance specifications and individual medication records. - Further research: Compare protocol administration instructions with dispensing records to distinguish nominal dose, administration frequency and blinding method from actual medication use. - Conditions for verification: Link the nominal oral dose given 1 time per 1 day to blinding of tablets in matching capsules, without recasting it as actual individual intake. - Why this condition is needed: Required to interpret the identity of actual exposure and blinding of measurement. - Application exceptions: Do not replace 0.8 mg with the dose in the warning document for the 5 mg product. Do not confuse capsule blinding with the active preparation itself being a capsule. - Remaining uncertainty: Original data on individual medication amounts, the complete capsule composition and release type are unverified. - Referenced methods: M03 §3.2.2–3.2.3; reverse-direction flagging and adaptations of checks are internally designed; M08 Table 8.2.a, M03 §3.2.4; internal application — an internal adaptation for dose and formulation. Not official certification of the method. - Related sources: S01, S02 ### N05 · Are the planned 5 years and median follow-up the same? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The planned duration was 5 years, median follow-up was 4.5 years (IQR 4.2–4.7 years), and run-in was 3 weeks. Early stopping for benefit is recorded; this does not mean everyone was followed for the same 4.5 years or 5 years. - Why unverified: The different planned and observed durations were distinguished. Individual exposure and censoring distributions and details of the stopping decision remain unresolved. - Reason for missing information: Individual follow-up times, discontinuation/censoring dates, and details of data lock and trial termination. - Required materials and conditions: Planned follow-up duration, actual follow-up distribution and individual randomization/censoring/end dates. - Materials to obtain first: Planned follow-up duration, actual follow-up distribution and individual randomization/censoring/end dates. - Further research: Compare the plan with the actual follow-up flow over time, without asserting that this was a study of participants completing a fixed 4.5 years or exposure for 5 years. - Conditions for verification: Retain the planned 5 years, median follow-up of 4.5 years/IQR and 3 weeks of run-in as separate time scales. - Why this condition is needed: Required to interpret time-to-event estimates and the observation period underlying absolute event values. - Application exceptions: Do not arbitrarily establish a new minimum required duration. Do not automatically classify a difference between planned duration and the median as Conflicting information. - Remaining uncertainty: Individual follow-up times, discontinuation/censoring dates, and details of data lock and trial termination. - Referenced methods: M03 §3.2.4, M15 §15.3.1–15.3.3; adaptations of outcomes to each question are internally designed; M03 §3.2.1–3.2.4; details of dose, duration and effect modification are internally designed — an internal adaptation for planned and observed duration. Not official certification of the method. - Related sources: S01, S02 ### N06 · Who were the participants who passed run-in? Verification status: Met Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: Men and women aged 45–75 with hypertension in Jiangsu/Anhui, China, with no prior stroke or myocardial infarction and with tolerability and adherence during the enalapril run-in were randomized. The findings cannot be said to represent safety among people who could not tolerate treatment. - Why unverified: The population in the current table is linked to the selected population in the original trial. Universal representativeness or a preventive effect for all adults with hypertension is not claimed. - Reason for missing information: Detailed participant counts at each run-in stage and the distribution of reasons for exclusion are absent from the currently retained summary. - Required materials and conditions: Inclusion and exclusion criteria, recruitment areas, selection flow before and after run-in, and reasons for exclusion. - Materials to obtain first: Inclusion and exclusion criteria, recruitment areas, selection flow before and after run-in, and reasons for exclusion. - Further research: Reconstruct the flow from recruitment to randomization to document differences between the population selected for tolerability/adherence and the general population with hypertension. - Conditions for verification: Verify Chinese adults aged 45–75 with hypertension, no history of stroke/MI, and run-in selection for tolerability/adherence. - Why this condition is needed: Required to reveal differences in efficacy applicability and safety arising from selection. - Application exceptions: Do not recast run-in selection as proof of no effect or a definitive magnitude of bias. Children and patients with prior stroke or MI are outside the included population. - Remaining uncertainty: Detailed participant counts at each run-in stage and the distribution of reasons for exclusion are absent from the currently retained summary. - Referenced methods: M03 §3.2.1; detailed population criteria are internally designed; design considerations in M19 §19.2.2–19.2.4; detailed recording of confounding is internally designed — an internal adaptation for participant selection. Not official certification of the method. - Related sources: S01, S02 ### N07 · Was prior TIA separately excluded? Verification status: Unverified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: Exclusion of prior stroke/MI is linked, but prior_TIA_status is an empty value/not_verified. TIA outcomes are not this page's primary endpoint. - Why unverified: The currently retained manuscript explicitly states that a separate exclusion criterion for prior TIA could not be verified. - Reason for missing information: Criteria permitting/excluding prior TIA, baseline numbers with TIA and diagnostic scope. - Required materials and conditions: The original eligibility wording on TIA history, diagnosis and timing, and the baseline history table. - Materials to obtain first: The original eligibility wording on TIA history, diagnosis and timing, and the baseline history table. - Further research: Compare event definitions and medical-history criteria in the original protocol to separate TIA eligibility criteria from its exclusion as an outcome. - Conditions for verification: Distinguish exclusion of prior stroke from the unverified separate TIA history criterion. - Why this condition is needed: Needed to avoid mixing selection based on medical history with outcome definitions. - Application exceptions: Do not infer that patients with prior TIA must have been excluded merely because TIA was excluded from the primary endpoint. - Remaining uncertainty: Criteria permitting/excluding prior TIA, baseline numbers with TIA and diagnostic scope. - Referenced methods: M03 §3.2.1; detailed population criteria are internally designed; question clarification principles in M03 §3.1, §3.2.1–3.2.4; detailed separation rules are internally designed — an internal adaptation for prior TIA. Not official certification of the method. - Related sources: S01, S02 ### N08 · Were a setting without mandatory fortification and a diagnosis of deficiency distinguished? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: In the trial-era Chinese setting without mandatory folic acid fortification, baseline median serum folate was 8.1 ng/mL in both arms. This does not mean every participant had diagnosed deficiency; individual dietary, fortification and concurrent supplement amounts are unverified. - Why unverified: The setting and aggregate nutritional values are linked, but individual total exposure and diagnoses of deficiency have not been verified. - Reason for missing information: Individual dietary/fortification and concurrent supplement amounts, diagnosed deficiency status and diagnostic criteria. - Required materials and conditions: Baseline nutritional distributions, diagnostic criteria, individual dietary/fortification/concurrent supplement records and regional policies. - Materials to obtain first: Baseline nutritional distributions, diagnostic criteria, individual dietary/fortification/concurrent supplement records and regional policies. - Further research: Record baseline nutritional status and setting by stratum, and predefine and compare diagnosed deficiency status and differences in applicability across fortification settings. - Conditions for verification: Separate the trial-era fortification setting and baseline folate distribution from correction of diagnosed deficiency or individual dietary intake. - Why this condition is needed: Needed to distinguish the applicability limits of the nutritional setting from interpretation of dose/population. - Application exceptions: Do not use a single median to assert deficiency in everyone or establish the same effect in other countries with fortification. - Remaining uncertainty: Individual dietary/fortification and concurrent supplement amounts, diagnosed deficiency status and diagnostic criteria. - Referenced methods: M03 §3.2.1; detailed population criteria are internally designed; M03 §3.2.1–3.2.4; details of dose, duration and effect modification are internally designed — an internal adaptation for nutrition and setting. Not official certification of the method. - Related sources: S01, S02 ### N09 · Do aggregate blood pressure and medication use imply that they were identical for each individual? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: Mean on-treatment blood pressure was 139.7/83.0 versus 139.8/83.1 mmHg, calcium-channel blocker use was 48.8% versus 48.9%, and diuretic use was 24.0% versus 24.2%. The policy permitting other antihypertensives was common to both arms, and the original report records that additional B vitamins were prohibited. - Why unverified: Similar aggregates by arm and a common policy are verified, but there is no evidence that changes and doses were identical for every individual. - Reason for missing information: Individual medication changes, actual doses taken, and use and handling of prohibited supplements. - Required materials and conditions: Background medication policies by arm, medication changes and doses at each visit, and records of additional B-vitamin/supplement use. - Materials to obtain first: Background medication policies by arm, medication changes and doses at each visit, and records of additional B-vitamin/supplement use. - Further research: While retaining aggregate blood pressure and concomitant medication rates, compare medication and total exposure at each visit to distinguish attribution to assignment effects from attribution to adherence effects. - Conditions for verification: Indicate separately the common policy, aggregates by arm and unverified individual medication changes. - Why this condition is needed: Required to distinguish the additional effect of the selected intervention from concurrent treatment. - Application exceptions: Do not require identical changes in every individual as a condition of validity. Do not use similar mean blood pressure as evidence that confounding has been completely removed. - Remaining uncertainty: Individual medication changes, actual doses taken, and use and handling of prohibited supplements. - Referenced methods: M03 §3.2.2–3.2.3; reverse-direction flagging and adaptations of checks are internally designed; M03 §3.2.3, M08 §8.2.2–8.2.3; internal application; design considerations in M19 §19.2.2–19.2.4; detailed recording of confounding is internally designed — an internal adaptation for common background therapy. Not official certification of the method. - Related sources: S01, S02 ### N10 · What level of detail is available on the actual design and allocation procedures? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The original evidence records a multicenter randomized double-blind trial and analysis according to randomized assignment. Blinding with matching capsules is retained, but the currently retained summary lacks detailed wording on random sequence generation and implementation of concealment. - Why unverified: The design label and blinding by appearance are linked, but the label alone cannot establish that the entire allocation process has been verified. - Reason for missing information: Sequence generation, implementation of concealment, stratification methods and detailed records of maintenance of blinding. - Required materials and conditions: Original documentation of random sequence generation, stratification/concealment, implementation of allocation and blinding procedures. - Materials to obtain first: Original documentation of random sequence generation, stratification/concealment, implementation of allocation and blinding procedures. - Further research: Compare the original protocol/SAP with allocation implementation records, and document allocation, concealment and blinding relevant to the actual outcome item by item. - Conditions for verification: Distinguish the report of a multicenter parallel randomized double-blind trial from the level of detail retained on specific allocation/concealment and assessment procedures. - Why this condition is needed: Required to verify the actual design for the causal efficacy question. - Application exceptions: Do not impose a crossover/cluster design on the question addressed by the verified parallel comparison. Do not recast an unverified detail as a confirmed allocation error. - Remaining uncertainty: Sequence generation, implementation of concealment, stratification methods and detailed records of maintenance of blinding. - Referenced methods: M08 §8.1–8.2.3; internal application; M08 Table 8.2.a, M03 §3.2.4; internal application — an internal adaptation for allocation and blinding. Not official certification of the method. - Related sources: S01, S02 ### N11 · What are the first symptomatic event and the excluded events? Verification status: Met Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The primary endpoint is time from randomization to the first fatal or nonfatal symptomatic stroke. The original SOP excludes subarachnoid hemorrhage and silent stroke. Homocysteine, recurrence, TIA and composite cardiovascular events do not substitute for it. - Why unverified: The direct clinical-event question is linked to the primary endpoint definition actually reported. The measure is neither a change in numerical values nor a simple risk ratio at a fixed time point. - Reason for missing information: The original SOP and individual adjudicated cases were not read in this review. - Required materials and conditions: Endpoint SOP, event classification, case adjudication documents and definitions of other endpoints. - Materials to obtain first: Endpoint SOP, event classification, case adjudication documents and definitions of other endpoints. - Further research: Use an event definition table to compare included/excluded types and whether an event was the first, and prevent automatic substitution of biomarkers or composite events. - Conditions for verification: Indicate time to the first fatal/nonfatal symptomatic event and exclusion of subarachnoid hemorrhage and silent events within the scope of the original SOP. - Why this condition is needed: Required to define the measurement boundaries of the selected outcome and prevent clinical extrapolation. - Application exceptions: Do not extend benefit to all-cause mortality, every stroke subtype or silent lesions merely because this is a hard endpoint. - Remaining uncertainty: The original SOP and individual adjudicated cases were not read in this review. - Referenced methods: M03 §3.2.4, M15 §15.3.1–15.3.3; adaptations of outcomes to each question are internally designed; M08 Table 8.2.a, M03 §3.2.4; internal application; M03 §3.2.4; internally defined limits on extrapolation — an internal adaptation for the clinical endpoint. Not official certification of the method. - Related sources: S01, S02 ### N12 · How were symptomatic events assessed? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: actual_trial_fields records that an independent event adjudication committee used clinical records and imaging/SOP. Follow-up visits were every 3 months. The currently retained manuscript lacks details of imaging acquisition rates by arm, adjudication disagreements and committee blinding. - Why unverified: The adjudication system and visit frequency are linked, but materials needed to review all measurement limitations remain outstanding. - Reason for missing information: Diagnostic/imaging acquisition rates by arm, assessor blinding, resolution of disagreements and records of missed events. - Required materials and conditions: Assessor blinding, committee operations, imaging/medical records and information on disagreements in event adjudication. - Materials to obtain first: Assessor blinding, committee operations, imaging/medical records and information on disagreements in event adjudication. - Further research: Compare event adjudication procedures and diagnostic/verification pathways by arm, and record detection differences and the scope of unverified information. - Conditions for verification: Link the independent event adjudication committee, clinical records/imaging and SOP at the level verified. - Why this condition is needed: Needed to distinguish differences in clinical event detection from the actual effect. - Application exceptions: Do not recast the description of an independent committee as independent clinical certification in this review. - Remaining uncertainty: Diagnostic/imaging acquisition rates by arm, assessor blinding, resolution of disagreements and records of missed events. - Referenced methods: M08 Table 8.2.a, M03 §3.2.4; internal application; M03 §3.2.4, M15 §15.3.1–15.3.3; adaptations of outcomes to each question are internally designed — an internal adaptation for event adjudication. Not official certification of the method. - Related sources: S01, S02 ### N13 · Do primary outcome event counts correspond to randomized denominators? Verification status: Met Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: A total of 20,702 participants were randomized: 10,348 to the folic acid arm and 10,354 to the control arm. First stroke is retained as 282/10,348 versus 355/10,354 participants. Safety analysis denominators and PP sets are separate. - Why unverified: Event counts for the original primary outcome are linked to randomized denominators. Safety and PP denominators were not substituted for the denominator of the primary effect. - Reason for missing information: Individual event/censoring source data and corrections to detailed denominators by genotype remain separate outstanding items. - Required materials and conditions: CONSORT allocation flow, the primary outcome table, numbers analyzed/experiencing events by arm and censoring records. - Materials to obtain first: CONSORT allocation flow, the primary outcome table, numbers analyzed/experiencing events by arm and censoring records. - Further research: Link randomized, analysis, event and safety sets separately, and compare the unit and denominator for each outcome. - Conditions for verification: Link the total of 20,702 participants, folic acid/control counts of 10,348/10,354 and first events of 282/355 to the same primary analysis. - Why this condition is needed: Required to avoid confusing events, people and analysis populations. - Application exceptions: Do not arbitrarily change the verified total number randomized because of conflicting genotype totals. - Remaining uncertainty: Individual event/censoring source data and corrections to detailed denominators by genotype remain separate outstanding items. - Referenced methods: M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; adaptations by study design are internally designed; M08 §8.2.2, M15 §15.3.1; internal application — an internal adaptation for randomized denominators. Not official certification of the method. - Related sources: S01, S03 ### N14 · What do ITT and censoring at the last observation estimate? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The primary analysis followed randomized assignment. Loss to follow-up was 32 versus 35 participants, censored at the last confirmed observation. This is an analysis of time to the first event, not an incidence-rate analysis with newly calculated total follow-up person-years. - Why unverified: The original analysis and the documented handling of loss to follow-up are linked, but detailed censoring assumptions, handling of competing risks and the complete sensitivity analyses are not currently retained. - Reason for missing information: Individual event/censoring times, handling of deaths, the complete SAP and sensitivity analysis results. - Required materials and conditions: Primary analysis SAP, individual event/censoring data, handling of deaths and loss to follow-up, and sensitivity analyses. - Materials to obtain first: Primary analysis SAP, individual event/censoring data, handling of deaths and loss to follow-up, and sensitivity analyses. - Further research: Compare the effect of interest and handling of censoring in the original SAP, and distinguish assignment effects, adherence effects and competing-risk analyses. - Conditions for verification: Indicate together the primary analysis according to randomized assignment, time to the first event and censoring of those lost to follow-up at the last confirmed observation. - Why this condition is needed: Required to distinguish the effect of assignment from time-to-event interpretation. - Application exceptions: Do not assume that the term ITT resolves all missing-data issues. Do not newly calculate person-years or fixed-time risks. - Remaining uncertainty: Individual event/censoring times, handling of deaths, the complete SAP and sensitivity analysis results. - Referenced methods: M08 §8.2.2, M15 §15.3.1; internal application; M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; adaptations by study design are internally designed — an internal adaptation for assignment effect and censoring. Not official certification of the method. - Related sources: S01, S02 ### N15 · Were PP exclusions and medication discontinuations combined as missing data? Verification status: Met Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The PP set, defined by at least 70% medication adherence and no major violations, contains 7,159 versus 7,152 participants. Loss to follow-up was 32 versus 35 participants. The original text did not count all PP exclusions and medication discontinuations as missing outcomes. - Why unverified: The known PP criteria/counts and loss to follow-up were retained separately. Current verification concerns the consistency of this distinction; it does not mean all reasons for deviations have been fully documented. - Reason for missing information: Detailed reasons for PP exclusion and medication discontinuation, and individual overlap between the sets. - Required materials and conditions: Medication use by arm, major violations, reasons for discontinuation, missed follow-up visits and PP/ITT flow. - Materials to obtain first: Medication use by arm, major violations, reasons for discontinuation, missed follow-up visits and PP/ITT flow. - Further research: Link medication adherence violations, medication discontinuation and loss to follow-up as different events in the participant flow, and compare them with their handling in the original analysis. - Conditions for verification: Distinguish the PP criteria of ≥70% medication adherence/no major violations from PP participant counts and loss to follow-up. - Why this condition is needed: Needed to avoid mixing adherence effects with loss of reported outcome data. - Application exceptions: Do not substitute PP results for the primary ITT effect or newly calculate the number with missing data from the number of PP exclusions. - Remaining uncertainty: Detailed reasons for PP exclusion and medication discontinuation, and individual overlap between the sets. - Referenced methods: M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; adaptations by study design are internally designed; M03 §3.2.3, M08 §8.2.2–8.2.3; internal application; M08 §8.2.2, M15 §15.3.1; internal application; M19 §19.1.2–19.2.4, M04 §4.6.3; internal application — an internal adaptation for PP and loss to follow-up. Not official certification of the method. - Related sources: S01, S02 ### N16 · Which analyses produced HR 0.79 and the P value? Verification status: Met Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The primary analysis reports HR 0.79 (95% CI 0.68–0.93), P=0.003. The original retained text distinguishes these as the unadjusted Cox hazard ratio and log-rank P value. This is not a simple RR or an absolute risk reduction of 21%p. - Why unverified: The original primary effect, interval, P and analytical attribution are linked. Statistical significance was not converted into a new judgment of clinical sufficiency. - Reason for missing information: Proportional-hazards diagnostics, the full adjusted analyses and reproduction using individual-level data are currently unverified. - Required materials and conditions: Original Cox/log-rank results tables, SAP, adjusted analyses and model diagnostic documents. - Materials to obtain first: Original Cox/log-rank results tables, SAP, adjusted analyses and model diagnostic documents. - Further research: Cross-check the analysis name, adjustment, censoring and measure for the reported values, and display the HR separately from cumulative risk ratios/absolute differences. - Conditions for verification: Attribute the unadjusted Cox HR, 95% CI and log-rank P to the relevant first-event question. - Why this condition is needed: Required to interpret the reported measure and level of precision. - Application exceptions: Do not convert the HR to an RR or extrapolate it to the same absolute effect over every time period. - Remaining uncertainty: Proportional-hazards diagnostics, the full adjusted analyses and reproduction using individual-level data are currently unverified. - Referenced methods: M03 §3.2.4, M15 §15.3.1–15.3.3; question-specific outcome adaptations are internally designed; M08 §8.2.2, M15 §15.3.1; internal application; M15 §15.3.1–15.3.3; internal application — internal adaptation for the primary HR and interval. This is not official methodological certification. - Related sources: S01 ### N17 · Were the earlier simple absolute difference and the paper's NNT distinguished? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The existing simple difference is 0.703%p: per 1,000 observed participants, approximately 7.0 fewer events. The paper reports a 4.5-year NNT of 141 (95% CI 85–426), distinguished from the internally calculated reciprocal of the crude event-proportion difference, 142.154. No new calculation was performed in this review. - Why unverified: Each value's source and derivation were distinguished. The paper's NNT calculation method was not reproduced using individual-level data, and the old calculation sheet itself was not read in this review. - Reason for missing information: The original NNT calculation method, fixed-time/censoring-based risk estimation, the old calculation sheet and individual-level data. - Required materials and conditions: Original event tables, the paper's NNT calculation method, censoring-adjusted risks and historical calculation records. - Materials to obtain first: Original event tables, the paper's NNT calculation method, censoring-adjusted risks and historical calculation records. - Further research: Link the paper's reported values and retained historical descriptive calculations to their respective sources, and specify whether they were reproduced using individual-level data and what censoring assumptions were used. - Conditions for verification: Preserve the existing simple difference in event proportions and its reciprocal separately from the paper's 4.5-year NNT as values derived by different methods. - Why this condition is needed: Required for relative/absolute interpretation of effect size and preservation of historical calculations. - Application exceptions: A simple difference in proportions is not a fixed-time risk difference recalculated with adjustment for censoring. Do not correct 142.154 to the paper's 141 or automatically classify the numerical difference as conflicting information. - Remaining uncertainty: The original NNT calculation method, fixed-time/censoring-based risk estimation, the old calculation sheet and individual-level data. - Referenced methods: M03 §3.2.4, M15 §15.3.1–15.3.3; question-specific outcome adaptations are internally designed; M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; design-specific adaptations are internally designed; M15 §15.3.1–15.3.3; internal application — internal adaptation for absolute difference and NNT. This is not official methodological certification. - Related sources: S01 ### N18 · Was a universal MCID or the planned effect adopted as a new criterion? Verification status: Unverified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The original assessment interpreted HR 0.79, 95% CI 0.68–0.93 and the existing absolute difference of about 0.703%p together. The universal first-stroke MCID is an empty value/not_verified. Preserve the original explanation that the planned HR 0.80 was not used as an MCID. - Why unverified: A retained judgment of importance exists, but a validated universal MCID cannot be said to have been verified. The original E+, C0, C and score of 56 were not reassessed. - Reason for missing information: Direct evidence for a justified importance criterion, whether universal or specific to a population/time period. - Required materials and conditions: Original evidence for clinical importance, assumptions about the planned effect, and documents interpreting the reported CI and absolute event difference. - Materials to obtain first: Original evidence for clinical importance, assumptions about the planned effect, and documents interpreting the reported CI and absolute event difference. - Further research: Collect endpoint-specific evidence of importance and link it to the original interpretation, separating design assumptions, statistical significance and importance for decision-making. - Conditions for verification: Distinguish the original interpretation of importance from an unverified universal MCID, and do not create new effect grades or numerical criteria. - Why this condition is needed: Needed to distinguish statistical significance and intervals from clinical importance. - Application exceptions: Do not newly assign C1 or sufficiency solely because the CI excludes 1. Do not arbitrarily require a minimum n, duration or number of trials. - Remaining uncertainty: Direct evidence for a justified importance criterion, whether universal or specific to a population/time period. - Referenced methods: M15 §15.3.1–15.3.3; internal application; M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; design-specific adaptations are internally designed — internal adaptation for clinical importance. This is not official methodological certification. - Related sources: S01, S02 ### N19 · Were ischemic and hemorrhagic values counted as independent replications? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: First ischemic strokes numbered 223 versus 292, HR 0.76 (95% CI 0.64–0.91), P=0.002. Hemorrhagic strokes numbered 58 versus 62, HR 0.93 (95% CI 0.65–1.34), P=0.71. Total first strokes include 2 events with an unverified subtype. - Why unverified: Positive and nonsignificant subtype results and unverified subtypes were preserved together. The full original subtype adjudication/denominator tables and attribution of unclassified events remain to be checked. - Reason for missing information: Arm attribution of the 2 events with an unverified subtype, detailed denominators, original adjudication tables and the prespecified subgroup analysis plan. - Required materials and conditions: Original subtype event tables, definitions, denominators, analysis plans and records of unclassified subtypes. - Materials to obtain first: Original subtype event tables, definitions, denominators, analysis plans and records of unclassified subtypes. - Further research: Link total first events with their subtypes, indicate participant/event overlap and the analytical hierarchy, and do not extend them to separate clinical questions. - Conditions for verification: Preserve subtypes, findings of no detected difference and unverified subtypes within the same first-event family, regardless of direction. - Why this condition is needed: Needed to prevent selective use of favorable subtypes from obscuring the full set of gaps. - Application exceptions: Do not add subtypes as independent trials or new participants. Do not interpret nonsignificance for hemorrhagic stroke as an effect of 0 or equivalence. - Remaining uncertainty: Arm attribution of the 2 events with an unverified subtype, detailed denominators, original adjudication tables and the prespecified subgroup analysis plan. - Referenced methods: M04 §4.6.3, M15 §15.3.1–15.3.3; separation of statuses is internally designed; the distinction between studies and reports in M04 §4.6.1–4.6.2; cohort/outcome linkage is an internal application; M03 §3.2.4, M15 §15.3.1–15.3.3; question-specific outcome adaptations are internally designed — internal adaptation for stroke subtypes. This is not official methodological certification. - Related sources: S01 ### N20 · Were the 18 versus 10 events and nonsignificance for mortality preserved? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: Fatal strokes numbered 18 in the folic acid arm and 10 in the control arm. The HR and CI for this suboutcome are unverified. All-cause deaths are preserved as 302 versus 320, HR 0.94 (95% CI 0.81–1.10), P=0.47. - Why unverified: The favorable total first-event result, fatal events in the opposite direction and nonsignificant mortality were linked together. Precise estimates for fatal events have not been verified. - Reason for missing information: Fatal stroke HR/CI, detailed adjudication/denominators and details of the mortality analysis. - Required materials and conditions: Fatal stroke definitions, HR/CI, original event tables and all-cause mortality results tables. - Materials to obtain first: Fatal stroke definitions, HR/CI, original event tables and all-cause mortality results tables. - Further research: Link the primary endpoint, fatal stroke suboutcome and all-cause mortality by their hierarchy, and record unfavorable findings/nonsignificance as reported in the sources. - Conditions for verification: Separate the unfavorable fatal stroke event counts, unverified precision and nonsignificant mortality results from total first events. - Why this condition is needed: Required to preserve unfavorable values and boundaries on interpretation of different clinical events. - Application exceptions: Do not extend the reduction in total first strokes to a reduction in fatal stroke or all-cause mortality. Do not convert nonsignificance into equivalence. - Remaining uncertainty: Fatal stroke HR/CI, detailed adjudication/denominators and details of the mortality analysis. - Referenced methods: M04 §4.6.3, M15 §15.3.1–15.3.3; separation of statuses is internally designed; M03 §3.2.4; internally defined limits on extension; M15 §15.3.1–15.3.3; internal application — internal adaptation for fatal stroke and mortality outcomes. This is not official methodological certification. - Related sources: S01 ### N21 · Do the genotype participant counts in Figure 1 and Table 1 agree? Verification status: Conflicting information Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The current manuscript explicitly states that the pooled genotype participant counts in Figure 1 do not agree with the total in Table 1. The current retained summaries do not contain the individual conflicting counts themselves, and the verified total of 20,702 randomized participants was not changed. - Why unverified: A discrepancy for the same item is explicitly stated in the original retained explanation. No new discrepancy was calculated, and neither value was selected. - Reason for missing information: The actual participant-count strings in the two original tables, counting definitions and whether an official correction exists. - Required materials and conditions: The original Figure 1 image, original genotype tables in Table 1, errata and definitions for participant counts. - Materials to obtain first: The original Figure 1 image, original genotype tables in Table 1, errata and definitions for participant counts. - Further research: Compare the original figure/table and counting definitions side by side, recording where discrepancies occur, differences between recruited/randomized sets and official corrections. - Conditions for verification: Preserve the discrepancy in counts for the same genotype specified in the manuscript, and do not arbitrarily replace the total randomized denominator. - Why this condition is needed: Required to preserve the original scope without concealing conflicting subgroup denominators. - Application exceptions: The conflicting values are absent from the current retained summaries, so do not use old numerical IDs to fill in arbitrary participant counts. The overall randomized denominator and genotype denominators are distinct. - Remaining uncertainty: The actual participant-count strings in the two original tables, counting definitions and whether an official correction exists. - Referenced methods: M08 Table 8.2.a, M15 §15.3, M19 §19.1.2–19.2.4; design-specific adaptations are internally designed; M04 §4.6.3, M15 §15.3.1–15.3.3; separation of statuses is internally designed; M03 §3.2.1–3.2.4; details on dose, duration and effect modification are internally designed — internal adaptation for conflicting genotype denominators. This is not official methodological certification. - Related sources: S01 ### N22 · Does interaction P=0.16 establish genotype-specific optimal doses? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The MTHFR and folate interaction P values are each 0.16. The manuscript did not establish genotype-specific optimal doses or definitive effect differences from subgroups alone. Baseline median folate is 8.1 ng/mL in both arms. - Why unverified: The interaction P and interpretive boundaries are linked, but detailed stratum-specific estimates, prespecification and complete denominator comparisons remain to be checked. - Reason for missing information: Each stratum's denominator, HR/CI, plans and handling of multiple testing, and original sources resolving the discrepancy. - Required materials and conditions: Prespecified subgroup plans, denominators by genotype/folate stratum, separate HR/CI estimates and interaction analysis tables. - Materials to obtain first: Prespecified subgroup plans, denominators by genotype/folate stratum, separate HR/CI estimates and interaction analysis tables. - Further research: Cross-check interactions together with stratum-specific estimates, and record prespecification, multiple testing and denominator conflicts separately from claims about optimal doses. - Conditions for verification: Distinguish the verification levels for MTHFR/folate interactions, subgroup denominators, prespecified plans and separate effect estimates. - Why this condition is needed: Needed to prevent extension of the primary effect to effect modification by population/dose. - Application exceptions: A nonsignificant interaction does not prove equivalent effects in all groups. Do not determine genotype-specific optimal doses from a trial of one dose. - Remaining uncertainty: Each stratum's denominator, HR/CI, plans and handling of multiple testing, and original sources resolving the discrepancy. - Referenced methods: M03 §3.2.1; detailed population criteria are internally designed; M03 §3.2.1–3.2.4; details on dose, duration and effect modification are internally designed; M15 §15.3.1–15.3.3; internal application — internal adaptation for effect modification. This is not official methodological certification. - Related sources: S01 ### N23 · Are the protocol date and public registration the same event? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The initial protocol date is 2008-04-12, and recruitment began on 2008-05-19. Public registration on 2008-11-20 is stated in the investigators' subsequent paper S08. The full original S04 registry history was historically inaccessible and has not been read in this review either. - Why unverified: Retrospective public registration and the existence of an initial protocol were linked together. Absence of prior planning or outcome switching was not established. - Reason for missing information: The complete original registry change history, details of the initial plan/SAP and comparison of endpoints/analyses across versions. - Required materials and conditions: Initial/revised protocols, SAP, documents on the start of recruitment and the full original registry history. - Materials to obtain first: Initial/revised protocols, SAP, documents on the start of recruitment and the full original registry history. - Further research: Place dates and document versions on a timeline, and compare endpoints, analyses and reasons for changes to distinguish prior planning from public registration. - Conditions for verification: Link the initial protocol, the start of recruitment and the public registration date stated in the subsequent paper as events attributed to their respective sources. - Why this condition is needed: Required to distinguish the scope of verification of selective reporting from documented events. - Application exceptions: Do not elevate the date statement in S08 to a finding based on reading the full original S04 registry record. Do not equate original sources not having been obtained with nonregistration. - Remaining uncertainty: The complete original registry change history, details of the initial plan/SAP and comparison of endpoints/analyses across versions. - Referenced methods: M08 Table 8.2.a, M04 §4.6.3; internal application; M04 §4.5, §4.6.3; limits on the current transformation of materials follow the user's scope and internal design — internal adaptation for planning and public registration. This is not official methodological certification. - Related sources: S01, S02, S04, S08 ### N24 · Has the scope of decision-making on early stopping been verified? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The manuscript records that the trial stopped early for benefit. This limitation and retrospective public registration were preserved as the two items underlying the original B2, without counting funding, a single trial or applicability again as deductions. - Why unverified: The fact of early stopping is linked, but stopping boundaries, details of interim analyses and actual decision records are unverified in the current retained summaries. - Reason for missing information: Planned/actual interim analyses, stopping criteria, decision meetings, termination date and handling of the full set of estimates. - Required materials and conditions: Interim analysis plans, stopping boundaries, data monitoring committee records and final estimates/SAP. - Materials to obtain first: Interim analysis plans, stopping boundaries, data monitoring committee records and final estimates/SAP. - Further research: Compare prespecified stopping rules with actual interim analyses and termination decisions, recording implications for estimation/reporting and unverified items. - Conditions for verification: Separate the reported early stopping for benefit from unverified planned interim analyses, stopping criteria and actual decisions. - Why this condition is needed: Required to distinguish the scope of planned analyses from reports of actual termination. - Application exceptions: Do not conclude that the benefit is false because the trial stopped early, or apply a new deduction to the original grade/effect. - Remaining uncertainty: Planned/actual interim analyses, stopping criteria, decision meetings, termination date and handling of the full set of estimates. - Referenced methods: M08 Table 8.2.a, M04 §4.6.3; internal application; M08 §8.2.2, M15 §15.3.1; internal application; M04 §4.6.3, M15 §15.3.1–15.3.3; separation of statuses is internally designed — internal adaptation for early stopping for benefit. This is not official methodological certification. - Related sources: S01, S02 ### N25 · Have the full registry history and data lock been verified? Verification status: Unverified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The full registry history and data-lock date are unverified in the original summary_note/editorial_review. Reading in this review is limited to retained summaries; the original registry record, ZIP and separate logs were not additionally checked. - Why unverified: The necessary full history and original data-lock materials are absent from the current inputs, and external materials have not been read in this review. - Reason for missing information: Complete change history, lock time, initial/final SAP and reasons for changes. - Required materials and conditions: Original registry change history, initial/final SAP, final data lock and change approval records. - Materials to obtain first: Original registry change history, initial/final SAP, final data lock and change approval records. - Further research: Link registry, plan and data-lock versions, and record actual primary-endpoint/analysis changes and unchecked portions for each document. - Conditions for verification: Distinguish the retained existence of initial planning from the unverified full history/data lock. - Why this condition is needed: Needed to retain the scope of unchecked prior planning and analytical choices. - Application exceptions: Do not convert missing materials into established endpoint changes or selective reporting. - Remaining uncertainty: Complete change history, lock time, initial/final SAP and reasons for changes. - Referenced methods: M08 Table 8.2.a, M04 §4.6.3; internal application; M04 §4.5, §4.6.3; limits on the current transformation of materials follow the user's scope and internal design — internal adaptation for data lock and change history. This is not official methodological certification. - Related sources: S02, S04 ### N26 · Were the original trial, protocol and safety tables grouped into one family? Verification status: Met Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The original evidence F01 links S01, S02 and S03. The S04 registration number is NCT00794885, and S08 provides the registration statement in a subsequent CSPPT analysis. S12 is retained only as the discovered title of a subsequent lifestyle report from the same CSPPT. - Why unverified: The one verified CSPPT family was distinguished from the number of reports. Only the original scope of 1 directly confirmatory family was preserved; the overall unique trial/participant counts are empty values. - Reason for missing information: Checks of individual-level data/outcome overlap across subsequent reports and of all trials included in the reviews remain incomplete. - Required materials and conditions: DOI/PMID, registry and report links, and family tables by cohort/participants and outcome/time point. - Materials to obtain first: DOI/PMID, registry and report links, and family tables by cohort/participants and outcome/time point. - Further research: Create a crosswalk of report identifiers and trial/participant links so that repeat citations of one CSPPT are not counted as independent replications or independent harm events. - Conditions for verification: Distinguish the original CSPPT trial, related documents and subsequent analyses from separate candidates/reviews. - Why this condition is needed: Required to prevent double-counting reports, studies and replications of harm findings. - Application exceptions: Do not add safety tables, the original paper, registry records and subsequent reports from the same family as separate replications or unique events. - Remaining uncertainty: Checks of individual-level data/outcome overlap across subsequent reports and of all trials included in the reviews remain incomplete. - Referenced methods: The distinction between studies and reports in M04 §4.6.1–4.6.2; cohort/outcome linkage is an internal application; M04 §4.6.1–4.6.2, M15 §15.3.1; internal application; M04 §4.6.1–4.6.2; adaptation for harm data is internally designed — internal adaptation for the CSPPT family. This is not official methodological certification. - Related sources: S01, S02, S03, S04, S08, S12 ### N27 · Were the review's 21 RCTs and dietary observations counted as direct replications? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The S06 title describes a meta-analysis of 21 RCTs, but the original evidence distinguishes it as an evidence map including CSPPT and heterogeneous combinations/populations. S07 is a review mixing dietary observational and intervention cohorts. Neither was included in the count of direct replications. - Why unverified: The identities and mixed scope of the reviews are linked, but the included trials have not all been individually cross-checked. Do not use 21 as the number of independent eligible trials in this table. - Reason for missing information: Individual included trials, overlap involving the same participants/reports, eligibility for folic acid as the sole add-on, and outcomes. - Required materials and conditions: Trials included in reviews, interventions/populations/endpoints, source-specific effect tables and links to repeat citations of CSPPT. - Materials to obtain first: Trials included in reviews, interventions/populations/endpoints, source-specific effect tables and links to repeat citations of CSPPT. - Further research: Cross-check included trials in individual eligibility tables, separating B-vitamin combinations, different populations and dietary observations, and flag CSPPT overlap. - Conditions for verification: Display the composition/mixed scope of reviews separately from eligible families for the direct single-add-on intervention. - Why this condition is needed: Needed to preserve differences and overlap between studies at the level verified. - Application exceptions: Do not repurpose a review average or the number of trials in its title as direct replications or a new participant denominator. - Remaining uncertainty: Individual included trials, overlap involving the same participants/reports, eligibility for folic acid as the sole add-on, and outcomes. - Referenced methods: The distinction between studies and reports in M04 §4.6.1–4.6.2; cohort/outcome linkage is an internal application; M04 §4.6.1–4.6.2, M15 §15.3.1; internal application; M04 §4.6.3, M15 §15.3.1–15.3.3; separation of statuses is internally designed — internal adaptation for the review evidence map. This is not official methodological certification. - Related sources: S06, S07 ### N28 · Have the final direct results of the two CSPPT2 candidates been verified? Verification status: Unverified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: S09 is the CSPPT2 CC/CT NCT04974138 candidate, and S10 is the TT NCT04974151 candidate. The original evidence F02/F03 records that the full records current at that time were inaccessible. Direct results remain unverified or of mixed scope; no new access occurred in this review, and the candidates were not included in the count of direct replications. - Why unverified: Registered candidate identities exist, but complete materials to verify current final first-stroke results and progress status are absent. - Reason for missing information: Complete registry records, final results, actual doses/populations/comparisons, completion denominators and funding sources. - Required materials and conditions: Complete registry records/change histories for NCT04974138 and NCT04974151, final first-stroke results and actual populations/interventions/denominators. - Materials to obtain first: Complete registry records/change histories for NCT04974138 and NCT04974151, final first-stroke results and actual populations/interventions/denominators. - Further research: Track the two candidates by their registration identities, but link them to the original family only after cross-checking the final reports' eligible questions, overlap and actual results. - Conditions for verification: Distinguish the CC/CT and TT registration identities from unverified final direct results/current progress status. - Why this condition is needed: Needed to prevent unpublished/registered candidates from being automatically counted as replications. - Application exceptions: The identities of two candidates do not mean 2 new eligible completed trials. Do not upgrade the original access_level metadata_shell_only to reading of the full record. - Remaining uncertainty: Complete registry records, final results, actual doses/populations/comparisons, completion denominators and funding sources. - Referenced methods: M04 §4.5, §4.6.3; limits on the current transformation of materials follow the user's scope and internal design; the distinction between studies and reports in M04 §4.6.1–4.6.2; cohort/outcome linkage is an internal application; M04 §4.6.1–4.6.2, M15 §15.3.1; internal application — internal adaptation for registered candidate results. This is not official methodological certification. - Related sources: S09, S10 ### N29 · Does the discovered S12 title establish a new direct effect? Verification status: Unverified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: S12 consists of the title of a subsequent CSPPT lifestyle report discovered at the cutoff (DOI 10.1038/s41371-026-01144-7). Its historical access level is search_metadata_only, and the current retained summaries contain no detailed results. - Why unverified: The identity was preserved, but the original report's detailed results/eligibility are absent from the current inputs, and there was no external lookup in this review. - Reason for missing information: The original document's outcomes, population/time points, analyses, cohort linkage and actual publication status. - Required materials and conditions: The original S12 report, publication identity, cohort/time points, outcomes and materials linking it to CSPPT. - Materials to obtain first: The original S12 report, publication identity, cohort/time points, outcomes and materials linking it to CSPPT. - Further research: Cross-check the subsequent report's question, data level and family so that it is not mistaken for independent replication of the original trial or a new primary effect. - Conditions for verification: Distinguish the subsequent lifestyle report's identity in search metadata from its unverified detailed results. - Why this condition is needed: Needed to distinguish the existence of an additional report from verification of its actual contents. - Application exceptions: Do not elevate discovery of a title to reading of the full text, a new direct effect or an independent confirmatory trial. - Remaining uncertainty: The original document's outcomes, population/time points, analyses, cohort linkage and actual publication status. - Referenced methods: M04 §4.5, §4.6.3; limits on the current transformation of materials follow the user's scope and internal design; the distinction between studies and reports in M04 §4.6.1–4.6.2; cohort/outcome linkage is an internal application; data collection/selection principles in M04 §4.5–4.6; separation of access and reading, and hyperlink conventions, are internally designed — internal adaptation for subsequent report identity. This is not official methodological certification. - Related sources: S12 ### N30 · Were reading retained summaries and reading original sources distinguished? Verification status: Met Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: In this review, the complete Korean text, evidence, bibliographic records and structure of the locally retained manuscript were read. The access_level entries for the original S01 HTML/original table images, S02/S03 PDFs and other materials are historical records. Access to the external documents for the 12 reports was not attempted in this review; they were not obtained or read. - Why unverified: Current actual actions were separated from historical access levels. A summary-reading link does not mark the original external document as read in this review. - Reason for missing information: Current public availability/bytes of original external materials, actual reading locations and report-specific access dates. - Required materials and conditions: Each report's original citation, actual access records/original bytes, and locations and dates of reading. - Materials to obtain first: Each report's original citation, actual access records/original bytes, and locations and dates of reading. - Further research: Separate summaries from original papers/registry records in the materials registry, and update a material's current reading record only when that original source has been read. - Conditions for verification: Record actual reading of the current retained summaries separately from each report's historical access_level and the current status that access to original sources was not attempted and those sources were not read. - Why this condition is needed: Required to accurately connect the user's summary-reading links with the verification status of the materials. - Application exceptions: Do not retrospectively count historical full-text reading as a success in this review, or label an unattempted access as an access failure. - Remaining uncertainty: Current public availability/bytes of original external materials, actual reading locations and report-specific access dates. - Referenced methods: Data collection/selection principles in M04 §4.5–4.6; separation of access and reading, and hyperlink conventions, are internally designed — internal adaptation for current and historical reading. This is not official methodological certification. - Related sources: S01, S02, S03, S04, S05, S06, S07, S08, S09, S10, S11, S12 ### N31 · Were the 2026-09-15 cutoff and this review date distinguished? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The original evidence_cutoff and verified_date are 2026-09-15. The review date for this table is 2026-10-05, and no new clinical search was performed. The manuscript's references to historical search records and selection_log.json are historical references; the separate files were not read in this review. - Why unverified: The cutoff, current review date and scope limitations are linked, but the complete historical search/selection logs and the latest unpublished studies/corrections/retractions were not cross-checked. - Reason for missing information: The actual historical search strategies, selection/exclusion logs, scope of completeness and current materials verifying corrections/retractions. - Required materials and conditions: Historical search/selection logs, the cutoff date and records of the currently authorized search/correction scope. - Materials to obtain first: Historical search/selection logs, the cutoff date and records of the currently authorized search/correction scope. - Further research: Link the scope and dates of search/selection logs to publication, registration and unverified items, distinguishing transformation of retained materials from a complete up-to-date search. - Conditions for verification: Distinguish the original evidence cutoff, original review date, this table's review date and the fact that separate search logs were not read. - Why this condition is needed: Required to retain the scope to which statements of absence or insufficiency apply. - Application exceptions: Do not label the absence of a new search as 0 human studies, absence of all unpublished studies or complete verification of the latest evidence. - Remaining uncertainty: The actual historical search strategies, selection/exclusion logs, scope of completeness and current materials verifying corrections/retractions. - Referenced methods: M04 §4.5, §4.6.3; limits on the current transformation of materials follow the user's scope and internal design; data collection/selection principles in M04 §4.5–4.6; separation of access and reading, and hyperlink conventions, are internally designed — internal adaptation for search and verification scope. This is not official methodological certification. - Related sources: S04, S05, S06, S07, S09, S10, S12 ### N32 · Was AUSA support double-counted under independence and bias? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The original trial's public/private funding, AUSA support, product provision and author consulting relationships were classified as mixed I1. The manuscript records that funding, a single trial and applicability were not counted again as deductions under B2. - Why unverified: The funding/relationship level is linked, but complete disclosures and roles in the research process for each subsequent report have not all been verified in the current retained summaries. - Reason for missing information: Complete disclosures for each report; specific roles in support/product provision; employment/patents; and whether analysis was controlled. - Required materials and conditions: Original documents on study funding, product provision, author disclosures/consulting, institutions/roles and competing interests in related reports. - Materials to obtain first: Original documents on study funding, product provision, author disclosures/consulting, institutions/roles and competing interests in related reports. - Further research: Link funding, product/analytical roles and author relationships for each report, separating verified and unverified aspects of independence. - Conditions for verification: Preserve mixed public, private and manufacturer funding, product provision, author relationships and the attribution of the original I1. - Why this condition is needed: Needed to assess claims and attribution of independent funding and replication. - Application exceptions: Do not automatically exclude results because competing interests exist, or assign I1 as a new judgment in this review. - Remaining uncertainty: Complete disclosures for each report; specific roles in support/product provision; employment/patents; and whether analysis was controlled. - Referenced methods: Chamgap's own criteria for independence and attribution. The official methods sections used in this pilot do not warrant all of these detailed items; the distinction between studies and reports in M04 §4.6.1–4.6.2; cohort/outcome linkage is an internal application — internal adaptation for funding and competing interests. This is not official methodological certification. - Related sources: S01, S02, S08 ### N33 · What are the safety denominators and the scope of event collection? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The safety analysis denominators are 10,124 people in the folic acid arm and 10,119 people in the control arm. They differ from the original randomized denominators for the primary effect, and event counts were not combined with numbers of people. Do not substitute the original median follow-up for individual exposure time for each harm. - Why unverified: The safety analysis population is linked, but details of collection/follow-up and exposure time for each event and inclusion/exclusion from the denominators remain unresolved. - Reason for missing information: Safety population definition; AE collection methods; exposure/timelines at each visit; and data on discontinuation, nonreporting and loss. - Required materials and conditions: Safety tables; definitions of analysis populations; collection at each visit; exposure time; and the flow of discontinuation/follow-up. - Materials to obtain first: Safety tables; definitions of analysis populations; collection at each visit; exposure time; and the flow of discontinuation/follow-up. - Further research: Compare the safety population and event collection system against the original tables, and separately link affected people/event counts, exposure time and loss to follow-up. - Conditions for verification: Distinguish the safety analysis population, people/events, duration and collection/reporting limitations from the population randomized for efficacy analysis. - Why this condition is needed: Needed to distinguish denominators and the scope of collection for efficacy and harms. - Application exceptions: Do not classify everyone omitted from the safety population as event-free or as having missing outcomes. Do not remove the efficacy effect. - Remaining uncertainty: Safety population definition; AE collection methods; exposure/timelines at each visit; and data on discontinuation, nonreporting and loss. - Referenced methods: M19 §19.1.2–19.2.4; internal application; M19 §19.1.2–19.2.4; detailed distinctions internally designed; M19 §19.2 on harm questions by design; details of temporal relationships designed here; M19 §19.1.2–19.2.4, M04 §4.6.3; internal application — internally designed adaptation for the safety population. Not official methodological certification. - Related sources: S01, S03 ### N34 · Does nonsignificance for people with gastrointestinal events and cough mean no risk? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: People with gastrointestinal adverse events numbered 610 versus 579 (P=0.40), and those with drug-related cough numbered 2,009 versus 2,011 (P=0.98). The original safety denominators are 10,124 versus 10,119 people; the report of a nonsignificant difference in overall AEs was not converted into a claim of no risk. - Why unverified: The numbers of people compared and P values are linked. Detailed CIs, causality assessment/collection and data on individual changes in concomitant medication have not been verified. - Reason for missing information: Original definitions, CIs and collection methods for each event; individual causation/concomitant medication/discontinuation; and overlap in affected people. - Required materials and conditions: Original AE tables; event/person definitions; background medication/collection by arm; CIs and discontinuation data. - Materials to obtain first: Original AE tables; event/person definitions; background medication/collection by arm; CIs and discontinuation data. - Further research: Attribute the comparison values in the original table to distinguish background enalapril from added folic acid, and describe nonsignificance, equivalence and individual causation separately. - Conditions for verification: Preserve the reported numbers of people, P values and interpretive limits of the comparison on the same common treatment. - Why this condition is needed: Needed to distinguish the scope of statements about harms actually compared, signals and causation. - Application exceptions: The comparison is on common enalapril therapy; do not establish folic acid as the cause of all reported cough. Do not interpret nonsignificance as equivalent safety. - Remaining uncertainty: Original definitions, CIs and collection methods for each event; individual causation/concomitant medication/discontinuation; and overlap in affected people. - Referenced methods: M19 §19.1.2–19.2.4; internal application; M19 §19.1.2–19.2.4; detailed distinctions internally designed; M19 §19.2.2–19.2.4, M03 §3.2.3; internal application; design considerations in M19 §19.2.2–19.2.4; detailed recording of confounding designed here; Chamgap’s own interpretive boundaries. References M19 principles, without applying an official causality tool — internally designed adaptation for GI and cough comparisons. Not official methodological certification. - Related sources: S01, S03 ### N35 · Have the unique number of people with SAEs and discontinuation been verified? Verification status: Unverified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The original safety description explicitly states that the total number of unique people with serious adverse events is unverified. The currently stored summary does not provide the total deduplicated number of people with serious events or timelines of discontinuation/recovery for each AE. - Why unverified: The unaggregated unique person counts and detailed collection/discontinuation data are absent from the currently retained scope. This does not mean no events or no harm. - Reason for missing information: Total unique people with SAEs; events/overlap; discontinuation/recovery for each AE; nonreporting and loss to follow-up. - Required materials and conditions: Unique people with serious AEs; event counts; overlap; exposure/reasons for discontinuation; and data on unreported events. - Materials to obtain first: Unique people with serious AEs; event counts; overlap; exposure/reasons for discontinuation; and data on unreported events. - Further research: Link the original safety tables and participant flow, and separately compare serious events, affected people, discontinuation/medication violations and loss to follow-up. - Conditions for verification: Distinguish unverified totals of unique people with serious AEs and unverified discontinuation/missing reports from 0. - Why this condition is needed: Needed to avoid converting missing reports into absence of events. - Application exceptions: Do not combine PP exclusions, drug discontinuation, loss to follow-up and SAEs into the same count or calculate unavailable frequencies. - Remaining uncertainty: Total unique people with SAEs; events/overlap; discontinuation/recovery for each AE; nonreporting and loss to follow-up. - Referenced methods: M19 §19.1.2–19.2.4; detailed distinctions internally designed; M19 §19.1.2–19.2.4; internal application; M19 §19.1.2–19.2.4, M04 §4.6.3; internal application — internally designed adaptation for SAEs and missing reports. Not official methodological certification. - Related sources: S01, S03 ### N36 · Was the 5 mg SmPC used as the incidence estimate for the 0.8 mg trial? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: S11 is the 2026-05-20 revision of the folic acid 5 mg SmPC. The original safety text preserves the context of masking anemia due to B12 deficiency and delaying recognition of neurological injury, tumor history, antiepileptics and methotrexate. The document for a different-dose product was not repurposed as trial incidence or proof of product identity. - Why unverified: The identity of the product document and the scope of its original warnings are linked, but trial batch/product identity and direct incidence in the relevant vulnerable populations remain unverified. - Reason for missing information: Identity of the trial product/batch with the SmPC product; direct safety data for the relevant medications, conditions and doses. - Required materials and conditions: Original SmPC version; product/dose specifications; warning text; and evidence of identity with the trial product. - Materials to obtain first: Original SmPC version; product/dose specifications; warning text; and evidence of identity with the trial product. - Further research: Compare the population/dose and medication context in the product document with the trial to separate official warnings from trial incidence/individual safety. - Conditions for verification: Distinguish the separate 5 mg product warnings, the current trial formulation/dose and the actual incidence of harm. - Why this condition is needed: Needed to maintain the safety boundaries of a document for a different product. - Application exceptions: Do not convert the partial search extract from S05 ODS into full-text verification of all S11 warnings. Do not use the 5 mg warnings as personal instructions for taking 0.8 mg. - Remaining uncertainty: Identity of the trial product/batch with the SmPC product; direct safety data for the relevant medications, conditions and doses. - Referenced methods: M19 §19.1.2–19.2.4; internal application; Chamgap’s own interpretive boundaries. References M19 principles, without applying an official causality tool; M03 §3.2.2–3.2.3; reverse-direction flagging and adaptations of checks are internally designed — internally designed adaptation for separate product warnings. Not official methodological certification. - Related sources: S11, S05 ### N37 · Does this also answer questions about pregnancy, severe disease and higher doses? Verification status: Partly verified Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: Pregnancy and lactation were excluded from the trial. The original page does not establish results in children or severe hepatic/renal disease, at higher doses or over longer periods. kidney_liver preserves the explanation of exclusions based on tests/severe disease, but the exact eGFR exclusion criterion remains unverified. - Why unverified: Verified exclusions and applicability limits are linked, but direct results for each vulnerable population and long-term exposure, as well as exact severity criteria, are lacking. - Reason for missing information: eGFR/hepatic disease exclusion thresholds; direct results for each vulnerable population; long-term/high-dose exposure; and rare harms. - Required materials and conditions: Original inclusion/exclusion criteria; hepatic/renal measures; and data for pregnancy/lactation, children and high-dose/long-term follow-up. - Materials to obtain first: Original inclusion/exclusion criteria; hepatic/renal measures; and data for pregnancy/lactation, children and high-dose/long-term follow-up. - Further research: Compare the selected population with excluded and vulnerable populations, and record the data needed for each separate question. - Conditions for verification: Link the populations excluded from the trial or unverified to the scope of missing direct information on longer-term/higher-dose use. - Why this condition is needed: Needed to avoid extending results from a selected population to overall safety/prevention. - Application exceptions: Do not require studies of every rare event/vulnerable population for the current narrow first-stroke question. Do not use unverified status to deny an effect in that population. - Remaining uncertainty: eGFR/hepatic disease exclusion thresholds; direct results for each vulnerable population; long-term/high-dose exposure; and rare harms. - Referenced methods: M03 §3.2.1; detailed population criteria designed here; M19 §19.1.2–19.2.4; internal application; M03 §3.2.1–3.2.4; details of dose, duration and effect modification designed here — internally designed adaptation for vulnerable populations and long-term scope. Not official methodological certification. - Related sources: S02, S11 ### N38 · Are studies of optimal dose and effects after discontinuation required for the original question? Verification status: Not applicable Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The current assessment concerns time to the first event with add-on therapy at one dose. The current input contains no optimal-dose comparison or results on persistence after discontinuation. These two extension studies were not added as requirements for the primary question. - Why unverified: The selected original question makes no claim about optimal dose or effects after discontinuation. Nonapplicability does not mean those effects have been verified. - Reason for missing information: Direct comparisons and follow-up after discontinuation for the expanded questions remain unverified. - Required materials and conditions: If an expanded question is requested: direct comparisons of active doses, follow-up after discontinuation and prespecified subgroup data. - Materials to obtain first: If an expanded question is requested: direct comparisons of active doses, follow-up after discontinuation and prespecified subgroup data. - Further research: Define the expanded question separately, then examine actual results and the scope of estimates through direct dose comparisons or follow-up designs after discontinuation. - Conditions for verification: Distinguish the current question about the observation period for 0.8 mg add-on therapy from dose optimization and persistence after discontinuation. - Why this condition is needed: A supplementary boundary to avoid requiring unnecessary extension studies for the narrow original question. - Application exceptions: Even with the original dose/observed effect, do not claim an optimal dose or lifelong effects. Future extension claims require newly adapted criteria. - Remaining uncertainty: Direct comparisons and follow-up after discontinuation for the expanded questions remain unverified. - Referenced methods: M03 §3.2.1–3.2.4; details of dose, duration and effect modification designed here; question-clarification principles in M03 §3.1, §3.2.1–3.2.4; detailed separation rules designed here — internally designed adaptation for optimal dose and effects after discontinuation. Not official methodological certification. - Related sources: S01, S02 ### N39 · Are comparisons across studies required for the current single direct trial family? Verification status: Not applicable Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The current direct confirmatory evidence is the one CSPPT family recorded in the manuscript. Subanalyses, safety tables and subsequent reports from the same family are not independent replications. The full families of registered candidates and reviews remain unconfirmed under N27–N29. - Why unverified: Within the one direct trial family currently verified, comparisons of consistency/heterogeneity across studies are not applicable. Do not substitute this status for unverified candidate families/results. - Reason for missing information: Other independent direct trial families; new unique harm events; and distributions of results by condition. - Required materials and conditions: Families, conditions and results of other independent direct trials/safety events, and data for comparing included trials. - Materials to obtain first: Families, conditions and results of other independent direct trials/safety events, and data for comparing included trials. - Further research: When additional independent direct trial families are verified, compare distributions by population/intervention/endpoint/analysis and remove repeated citations of the same events. - Conditions for verification: Distinguish results from the one verified direct CSPPT family from unconfirmed candidates/reviews, and do not automatically generate heterogeneity or replication counts. - Why this condition is needed: A supplementary boundary to avoid arbitrarily requiring comparisons across multiple studies within one verified family. - Application exceptions: Unverified registered candidates/reviews are not NA. Distinguish nonapplicability of comparisons across studies within the current scope from separate gaps in replication data. - Remaining uncertainty: Other independent direct trial families; new unique harm events; and distributions of results by condition. - Referenced methods: M04 §4.6.1–4.6.2, M15 §15.3.1; internal application; distinction between studies and reports in M04 §4.6.1–4.6.2; cohort/outcome linkage applied here; M04 §4.6.1–4.6.2; adaptation of harm data designed here — internally designed adaptation for replication across studies. Not official methodological certification. - Related sources: S01, S02, S03, S06, S07, S09, S10 ### N40 · Does conversion into a table change the original grade or deployment events? Verification status: Met Reading scope: Summary read: the currently retained manuscript was actually read. The original papers, registry records and separate historical materials were not directly read in this review. - Current findings: The original grade C, score 56, 7 axes B/H/R1/I1/E+/B2/C0 and safety rating Caution remain unchanged. The original first_published_at/deployed_at 2026-09-15 and server_deployed=true are retained metadata, not evidence of public deployment of this table. corrections and subclaims are empty lists. - Why unverified: The work performed to convert the table has been distinguished from the original assessment/publication metadata. The original verdict, score, numerical values and historical access have not been changed. - Reason for missing information: The old separate original deployment/audit files have not been read in this review, and no new external evidence has been verified. - Required materials and conditions: Original verdict/score; revision history; original canonical URL; and publication/server records for each event. - Materials to obtain first: Original verdict/score; revision history; original canonical URL; and publication/server records for each event. - Further research: Preserve the entire pre-change version, and separately record criteria revisions, evidence updates, display changes and deployment events, linking them to the same ID/URL. - Conditions for verification: Preserve the original C, 56, 7 axes and publication/server metadata as historical values, separate from the current table creation/nondeployment. - Why this condition is needed: Required to avoid confusing conversion of retained content with reassessment/publication events. - Application exceptions: Do not newly calculate the number of criteria met, probabilities or grades for this table, or present the original self-review as independent clinical certification. - Remaining uncertainty: The old separate original deployment/audit files have not been read in this review, and no new external evidence has been verified. - Referenced methods: Question-clarification principles in M03 §3.1, §3.2.1–3.2.4; detailed separation rules designed here; M04 §4.5, §4.6.3; the limits of the current data conversion follow the user's scope and the design developed here — internally designed adaptation for the original verdict and updates. Not official methodological certification. - Related sources: ## Evidence and sources ### S01 Huo et al. 2015 CSPPT primary randomized trial [Source](https://jamanetwork.com/journals/jama/fullarticle/2205876) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: 20,702 people (10,348 versus 10,354 people) were randomized. First events occurred in 282 versus 355 people, HR 0.79 (95% CI 0.68–0.93), log-rank P=0.003, with median follow-up of 4.5 years (IQR 4.2–4.7 years). The existing crude absolute difference of 0.703%p/about 7.0 events per 1,000 people, the paper's NNT of 141 (95% CI 85–426) and the existing reciprocal of 142.154 are distinguished. Ischemic/hemorrhagic stroke, fatal strokes of 18 versus 10 events, nonsignificant mortality differences, PP/loss to follow-up, conflicting genotype denominators, early stopping and funding scope were also preserved. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: One CSPPT F01 family. The original trial, supplementary materials and subsequent reports are not added as separate replications. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Full-text article on the web | Unverified (no lookup this time) | Not attempted this time | Not read | Direct primary evidence on time to first symptomatic stroke with folic acid as the sole added active ingredient. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: full_text_html_and_original_table_images. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | | Original article table images (individual URLs/file formats unverified) | Unverified (no lookup this time) | Not attempted this time | Not read | Only the historical access_level for reading original table images is preserved in the currently stored summary. Individual image URLs, original bytes and image reading in this review are unverified. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: full_text_html_and_original_table_images. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ### S02 CSPPT protocol version 3 and SAP, JAMA Supplement 1 [Source](https://cdn.jamanetwork.com/ama/content_public/journal/jama/933696/joi150028supp1_prod.pdf) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: S02 is identified as CSPPT protocol version 3 and SAP, JAMA Supplement 1. Methodological details such as blinding of tablets in matching capsules, common enalapril, the primary endpoint/selection, the planned 5 years and the 3-week run-in are linked to the original content_blocks/actual_trial_fields. The initial protocol date of 2008-04-12 is a statement retained in the manuscript; the current v3 link alone does not establish that the entire initial document has been checked. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: A methods document from the same CSPPT F01, not an independent replication or results from new participants. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Full PDF of the article's supplementary protocol/SAP | Unverified (no lookup this time) | Not attempted this time | Not read | Links methods for the endpoint SOP, intervention/population, plan/follow-up and adjudication/analysis; it is not a new effect report. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: full_text_pdf_with_screenshots. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ### S03 CSPPT supplementary tables including safety sets [Source](https://cdn.jamanetwork.com/ama/content_public/journal/jama/933696/joi150028supp2_prod.pdf) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: The safety populations comprise 10,124 versus 10,119 people; people with GI events number 610 versus 579 (P=0.40), and those with drug-related cough number 2,009 versus 2,011 (P=0.98). The total unique number of people with serious AEs is unverified. Randomized denominators/PP, event counts/people and nonsignificance/no risk are distinguished. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: Results tables from the same CSPPT F01. Repeated citations are not counted as independent harm events or trials. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Full PDF of the article's supplementary results tables | Unverified (no lookup this time) | Not attempted this time | Not read | Links the safety population and the originally reported comparison of affected people. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: full_text_pdf_with_screenshots. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ### S04 ClinicalTrials.gov CSPPT record [Source](https://clinicaltrials.gov/study/NCT00794885) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: The original CSPPT registry candidate/identity is NCT00794885. The access_level is metadata_shell_only; the full original registry history and data lock are unverified. The retained source for public registration on 2008-11-20 is the statement in the subsequent S08 paper; it is not upgraded to direct verification of the full S04 history. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: Registry identity for the same CSPPT F01. The registration number is not an additional independent confirmatory trial. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Full clinical trial registry record on the web | Unverified (no lookup this time) | Not attempted this time | Not read | Preserves the NCT00794885 identity and incomplete access to the full history. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: metadata_shell_only. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ### S05 NIH ODS Folate health-professional fact sheet [Source](https://ods.od.nih.gov/factsheets/Folate-HealthProfessional/) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: The Folate HealthProfessional fact sheet URL is preserved. The historical search_extract_partial_then_access_failure is retained, and the original text has not been read directly in this review. Do not automatically substitute it as the original source for S11 product warnings or trial incidence. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: An official information document, not a direct efficacy replication of CSPPT or new harm events. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Full official NIH ODS web document | Unverified (no lookup this time) | Not attempted this time | Not read | Official document identity and limited historical access level. This stored summary contains no detailed results specific to that document. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: search_extract_partial_then_access_failure. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ### S06 Zhang et al. 2024 folic acid and stroke meta-analysis of 21 RCTs [Source](https://pubmed.ncbi.nlm.nih.gov/38824900/) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: The title mentions 21 RCTs, but the original evidence R01 distinguishes it as a review including CSPPT and heterogeneous combinations/populations. Historical access is abstract_and_metadata. It was not included in the count of directly eligible replications, and this stored summary contains no complete table of direct effects for every trial in the review. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: R01 review map. The full overlap with repeated CSPPT citations, combinations and other populations remains unconfirmed. No new count of unique trials is generated. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | PubMed abstract and bibliographic record | Unverified (no lookup this time) | Not attempted this time | Not read | Literature map of mixed RCTs; direct comparison of included trials is incomplete. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: abstract_and_metadata. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | | Publisher’s full-text article (DOI entry point) | Unverified (no lookup this time) | Not attempted this time | Not read | The DOI identifier can lead to the actual original article, but external accesses, downloads and full-text readings in this review number 0. This is an identifier entry point linking the retained DOI through a standard DOI link. The original URL is PubMed; this is not a direct publisher full-text URL or evidence of successful current access. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: abstract_and_metadata. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ### S07 Yang et al. 2024 primary stroke prevention review [Source](https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2024.1288417/full) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: The original evidence R02 is retained as a review including dietary observational and intervention cohorts. The historical full-text article HTML for S07 does not establish access in this review. It was not included in the count of direct replications, and a full comparison of included trials for eligibility/family remains incomplete. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: R02 review map. Not added as an independent completed trial or a new participant denominator. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Full-text review article on the web | Unverified (no lookup this time) | Not attempted this time | Not read | Literature map combining dietary observational and intervention cohorts. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: full-text article HTML. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ### S08 Zhang et al. 2020 CSPPT secondary observational analysis: registration statement [Source](https://link.springer.com/article/10.1186/s12933-020-01161-x) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: The manuscript records that the investigators' subsequent paper S08 specifies public registration on 2008-11-20. This is later than recruitment beginning on 2008-05-19 and is distinguished from the existence of the initial protocol dated 2008-04-12. Separate efficacy results from the subsequent observational analysis are not preserved in the currently stored summary. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: Subsequent observational report from CSPPT F01. Distinguishes the registration statement from reading the full original registry record; not a new independent randomized replication. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Full-text article on the web | Unverified (no lookup this time) | Not attempted this time | Not read | Links the subsequent paper's statement of the public registration date. Not a separate randomized replication of the primary effect. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: full-text article HTML. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ### S09 CSPPT2 CC/CT registered study candidate [Source](https://clinicaltrials.gov/study/NCT04974138) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: For candidate NCT04974138, the original evidence F02 records the historical description "full current record inaccessible", with direct results unverified or of mixed scope. Current progress status, final first-event results and direct randomization/funding are unverified; it was not included in the count of direct replications. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: F02 is the original candidate identity. It is not automatically accepted as a completed direct efficacy trial. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Full clinical trial registry record on the web | Unverified (no lookup this time) | Not attempted this time | Not read | CC/CT candidate identity; direct results unverified. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: metadata_shell_only. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ### S10 CSPPT2 TT registered study candidate [Source](https://clinicaltrials.gov/study/NCT04974151) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: For candidate NCT04974151, the original evidence F03 records the historical description "full current record inaccessible", with direct results unverified or of mixed scope. Current progress status, final first-event results and direct randomization/funding are unverified; it was not included in the count of direct replications. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: F03 is the original candidate identity. It is not automatically accepted as a completed direct efficacy trial. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Full clinical trial registry record on the web | Unverified (no lookup this time) | Not attempted this time | Not read | TT candidate identity; direct results unverified. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: metadata_shell_only. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ### S11 Flamingo Pharma folic acid 5 mg tablet SmPC, text revised 20 May 2026 [Source](https://www.medicines.org.uk/emc/product/14604/smpc) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: The 2026-05-20 revision of the Flamingo Pharma folic acid 5 mg SmPC is identified. The original safety text preserves the context of masking anemia due to B12 deficiency/delaying recognition of neurological injury, tumor history and antiepileptics/methotrexate. This document does not establish identity with the trial's 0.8 mg product, incidence or individual safety. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: Official product warnings, not an independent efficacy replication of CSPPT or a new incidence report. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | Manufacturer's full official SmPC on the web | Unverified (no lookup this time) | Not attempted this time | Not read | Warnings for a product at a different dose; separate from trial product identity/incidence. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: full_text_official_manufacturer_label. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ### S12 CSPPT follow-up lifestyle report discovered at cutoff [Source](https://doi.org/10.1038/s41371-026-01144-7) Materials actually read: The relevant Korean and English summaries and bibliographic records in the retained manuscript were read. The linked external original sources have not been read in this review. Findings recorded in the summary: The subsequent CSPPT lifestyle report discovered at cutoff has DOI 10.1038/s41371-026-01144-7 and historical access level search_metadata_only. This stored summary contains no separate results/denominators or details of actual publication status. It is not added as a new independent direct replication of the original trial. Record level: Information recorded in the retained manuscript. It is not the result of a new comparison against external original sources in this review. Independent replication: Identity of a discovered subsequent CSPPT report. Details of outcome/participant overlap are unverified; no independent replication is generated automatically. | Material | Public availability | Access status | Actually read | Notes | |---|---|---|---|---| | DOI identifier and publication search metadata | Unverified (no lookup this time) | Not attempted this time | Not read | Only the identity of a discovered subsequent report; detailed results unverified. URL identity preserved in the original citation. The existence of a link is separate from current acquisition/reading. No attempt was made to access this external original source or read its original text in this review. What has been read is the stored summary for each report. Even if the original source address is unverified, the external material remains unread; this does not mean a local full text has been acquired. Historical access level in the retained bibliography: search_metadata_only. The actual historical access date and location of the access record are unverified. The retained manuscript's review date, 2026-09-15, is not the access date of the original source. | ## 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 criteria and reasons for changes, and distinguish variations in the intervention. · Reuses the original HTML bytes/SHA retained for 1585 and the principles from that time. No new full methodological text was read, no official tool was applied, and no certification was obtained in this review. - M04 [Cochrane Handbook 6.5.1 (2025), chapter updated 2025-03](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-04) · §4.5, §4.6.1–4.6.3, MECIR C42 · Distinguish and link studies and reports, and do not omit a study solely because its outcomes are not reported. · Reuses the original HTML bytes/SHA retained for 1585 and the principles from that time. No new full methodological text was read, no official tool was applied, and no certification was obtained in this review. - 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 · Hold the outcome and effect of interest fixed, then assess randomization, deviations, missing data, measurement and selective reporting. · Reuses the original HTML bytes/SHA retained for 1585 and the principles from that time. No new full methodological text was read, no official tool was applied, and no certification was obtained in this review. - 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. · Reuses the original HTML bytes/SHA retained for 1585 and the principles from that time. No new full methodological text was read, no official tool was applied, and no certification was obtained in this review. - 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 · Questions about harms may require designs, populations and follow-up different from efficacy questions. Case reports do not establish incidence. · Reuses the original HTML bytes/SHA retained for 1585 and the principles from that time. No new full methodological text was read, no official tool was applied, and no certification was obtained in this review. ## What the statuses mean - Met: The retained identity, values and interpretive boundaries of this criterion have been verified. This does not signify sufficiency of clinical efficacy/safety or certification. - Partly verified: Some dimensions of the data have been linked, but specific original sources, scope and comparisons remain unresolved. - Unverified: Required values, criteria or materials are lacking within the retained scope currently read. This does not mean an effect or event count of 0 or incomplete review of the table. - Conflicting information: A discrepancy for the same item is explicitly stated in the original retained manuscript. No arbitrary correction or substitution of denominators. - Not applicable: Not applicable as a requirement for the current narrow question. Not a substitute for external materials being unread or study families being unverified. - Not applicable (spaced label): Display alias for Not applicable. - Not read: The relevant external original source has not been read in this review. Distinguished from reading the retained manuscript summary. - Review incomplete: Individual reviews for table creation remain to be completed. Distinguished from insufficient data. ## Change history 2026-11-03 · Public page/materials 1.0. Public-facing display and data were created from the existing 40 items. No new clinical search, change in verification status or scoring was performed. ## Material-specific source addresses - [S01-M1 · Full-text article on the web](https://jamanetwork.com/journals/jama/fullarticle/2205876) - S01-M2 · Original article table images (individual URLs/file formats unverified): Material URL unverified; external material unread. - [S02-M1 · Full PDF of the article's supplementary protocol/SAP](https://cdn.jamanetwork.com/ama/content_public/journal/jama/933696/joi150028supp1_prod.pdf) - [S03-M1 · Full PDF of the article's supplementary results tables](https://cdn.jamanetwork.com/ama/content_public/journal/jama/933696/joi150028supp2_prod.pdf) - [S04-M1 · Full clinical trial registry record on the web](https://clinicaltrials.gov/study/NCT00794885) - [S05-M1 · Full official NIH ODS web document](https://ods.od.nih.gov/factsheets/Folate-HealthProfessional/) - [S06-M1 · PubMed abstract and bibliographic record](https://pubmed.ncbi.nlm.nih.gov/38824900/) - [S06-M2 · Publisher’s full-text article (DOI entry point)](https://doi.org/10.1016/j.clnu.2024.05.034) - [S07-M1 · Full-text review article on the web](https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2024.1288417/full) - [S08-M1 · Full-text article on the web](https://link.springer.com/article/10.1186/s12933-020-01161-x) - [S09-M1 · Full clinical trial registry record on the web](https://clinicaltrials.gov/study/NCT04974138) - [S10-M1 · Full clinical trial registry record on the web](https://clinicaltrials.gov/study/NCT04974151) - [S11-M1 · Manufacturer's full official SmPC on the web](https://www.medicines.org.uk/emc/product/14604/smpc) - [S12-M1 · DOI identifier and publication search metadata](https://doi.org/10.1038/s41371-026-01144-7)