Long-term oral folic-acid exposure and gait ataxia: causal limits in a B12-deficient case

Useful facts when choosing a product

• The original case lists folic acid as a separate medication.
• The reported oral exposure of about 5 mg/day is medication history, not a product trial with a verified complete ingredient list.
• The UK 5 mg-tablet SmPC supplies cautionary context, not product identification for the case.

What the research actually shows

# Long-term oral folic-acid exposure and gait ataxia: causal limits in a B12-deficient case

TASK-1048 / R01-088 · New independent safety report

Preserved input snapshot: 2026-09-18T01:36:44+09:00 Actual content preparation/review: 2026-09-18T02:02:36+09:00

## Thirty-second answer

A 56-year-old woman reporting oral folic acid about 5 mg/day for 30 years had gait ataxia progressing over five months and B12 deficiency. This single uncontrolled case, with dietary/possible absorptive contributors, the course of B12 deficiency and concurrent treatment, cannot quantify risk or isolate direct folic neurotoxicity, exacerbation or actual diagnostic delay. Relevant human evidence exists, but this is neither proof of harm in B12-replete people nor a declaration of zero risk. [S01]

This is a completed unscored safety report. Grade, score and seven axes are null; content is completed_with_uncertainty and publication mapping is needs_format_mapping.

## The actual selected boundary

| Item | Verified value and limit | |---|---| | Main clinical-harm outcome | Progressive gait ataxia clinically documented at presentation | | Main observation | Presentation in May 2023 | | Actual exposure window | Retrospectively reported oral folic acid about 5 mg/day, almost regularly for 30 years; symptom progression over the preceding five months | | Actual population | One 56-year-old woman with B12 deficiency at presentation, a vegan diet and possible absorptive contributors | | Molecule/route | folic acid / Oral context in the article; administration records not independently verified | | Formulation, added actives and control | null; full product composition unconfirmed and no control group. | | Selection rationale | Selected because the original report describes actual folic-acid exposure, clinical gait dysfunction, B12 status, an exposure window and examination/course. Selection does not confer causal validity or representativeness; full product composition and a comparator are unavailable. |

[S01]

## Separating exposure, diagnosis and causality

Temporal ordering is limited to the reported medication history. B12 84 pg/mL and folate >24 ng/mL at presentation describe diagnostic context, not a causal folic-acid effect or nutritional status at exposure initiation. One reported patient is not a 1/1 incidence estimate. [S01]

A vegan diet and gastrointestinal/possible absorptive findings provide alternative explanations. A grade-A duodenal lesion is not upgraded to confirmed coeliac disease, nor ANA to confirmed pernicious anemia. The natural course of B12 deficiency and initial neurological status are unknown, preventing separation of direct folic toxicity from exacerbation. [S01]

Gait ataxia and positive Romberg are the main clinical observations. The case/figure caption report no significant MRI signal change, but the discussion mentions brain white-matter abnormality. A final clinical diagnosis of combined degeneration is not upgraded to imaging confirmation. Source ambiguities in MRC 4/4, the MCV unit f/l and later B12 reference-range wording are retained. [S01]

Improvement followed B12 treatment, folic cessation, rehabilitation and fluoxetine adjustment. This course is not an isolated folic-acid dechallenge/rechallenge, and nine-/ten-day hospital observations are not a risk estimate over 30 years. [S01]

Transient correction of anemia and direct exacerbation of B12-related neurological injury are different causal links. EFSA distinguishes an established former link from low certainty in the latter. Its certainty language is neither a patient event probability nor this report’s score. The selected case lacks before-and-after evidence establishing actual masking or diagnostic delay. [S01,S02]

Associations of blood folate with cognitive performance differ from an effect of a verified single oral product on gait-ataxia incidence. UMFA, mechanisms, low cognitive scores and failure to improve cognition are not substituted for clinical neurotoxicity. Other folate forms, dietary/fortified exposure, B-vitamin mixtures and MTX/anticonvulsant interactions are outside the main claim. [S03,S05]

Risk ratio, odds ratio, hazard ratio, absolute risk, person-time, CI, P value and MCID are null. There is no valid denominator/comparison for statistical execution. Unscored is not zero effect, equivalence or a safety guarantee. [S01]

Before presentation, reported co-medications were bisoprolol fumarate 2.5 mg/day, olmesartan medoxomil 20 mg/day, fluoxetine 20 mg/day and aspirin 100 mg/day. They were listed separately from folic acid; this was not a comparison ruling out each drug contribution. Fluoxetine was later increased to 40 mg/day. [S01]

Contrary and differently bounded safety information remains separate. CDC guidance dated 15 July 2026 states that harm has not been shown at the recommended 400 micrograms/day and addresses B12 detection in recommended-intake/fortification contexts. This differs from the approximately 5 mg/day, 30-year case with B12 deficiency at presentation. It does not justify claims that every dose is safe or that every exposed person is harmed. [S08]

The official 2004 RIVM report found limited direct human evidence of oral neurotoxicity and could not exclude unrecognized B12 deficiency in some cases. It also noted studies not designed to ascertain adverse events and limitations of nonreporting. Its tables were not reanalyzed as if every molecular form or combined intervention were single folic acid. This dated context does not verify 30-year safety today. [S09]

## Evidence table

These seven rows are factual evidence blocks, not seven independent trials. E01/E02 concern one case, and the EFSA material in E03/E04 is the same opinion. E03/E06 also identify separately bounded official contextual documents. Access-limited leads S06/S07 are not numerical effect evidence.

| Evidence | Role/family | Observed information | Limitation | Sources | |---|---|---|---|---| | E01 | Primary clinical signal / CICCHINI_2025_SINGLE_CASE | Gait ataxia and positive Romberg; one reported patient, no control group. | About 5 mg/day for 30 years is retrospective exposure history, distinct from five months of symptoms. | S01 | | E02 | Causal limits in the same case / CICCHINI_2025_SINGLE_CASE | B12 deficiency and vegan diet coexisted; improvement followed multiple interventions. | This is the same person as E01. B12 replacement, folic cessation and rehabilitation cannot isolate a folic-acid dechallenge effect. | S01 | | E03 | Official causal-certainty assessment / EFSA_2023_OPINION + RIVM_2004_NEUROTOXICITY | EFSA rates causal exacerbation of B12-related neurological manifestations with low certainty. The 2004 RIVM report also found limited direct oral-toxicity evidence in humans. | Uncontrolled historical cases cannot separate the natural course of untreated B12 deficiency. The opinion adds no new patients. | S02,S09 | | E04 | Separate anemia-correction/masking link / EFSA_2023_OPINION | Transient correction of B12-related megaloblastic anemia differs from treatment of neurological damage. | This is the same opinion as E03. Actual masking or diagnostic delay is not demonstrated in the selected case. | S02 | | E05 | Related human observation, excluded from main outcome / MOORE_2014_THREE_COHORTS | The primary abstract reports associations between blood B12/red-cell folate and cognitive performance. | It is not an effect of verified single-product exposure on gait-ataxia incidence and cannot exclude confounding or reverse causation. | S05 | | E06 | Official safety context / ODS_FOLATE_OFFICIAL + CDC_2026_SAFETY | ODS distinguishes correction of B12-related anemia from persistent neurological injury. CDC 2026 states that harm has not been shown at the recommended 400 micrograms/day. | Official context, not a control group for the case. Recommended 400 micrograms/day is not the same condition as about 5 mg/day for 30 years with B12 deficiency; an upper limit is not a toxicity threshold. | S03,S08 | | E07 | Product-label caution / SMPC_14604 | Warns against treating B12 deficiency with folic acid without adequate B12 treatment. | This UK product is not established as the Italian case product. Label duration and categorical short-course wording are not adopted as causal risk or a no-harm guarantee. | S04 |

## Seven assessment explanations

| Field | Assessable scope for this claim | |---|---| | claim_type | This is a causal clinical-harm question, not efficacy. An unsupported A–F safety grade is not invented. | | endpoint | The main outcome is clinical gait ataxia at presentation in May 2023. The 30-year exposure history and five-month symptom progression are distinct. | | replication | One new original case family was directly extracted. PDF/index copies and treatment follow-up are not independent replications, and historical cases mapped by an opinion are not double-counted. | | independence | Case authors declare no competing interests, but funding, product support and full independence are unverified. Official contextual sources do not certify the case’s independence. | | effect | Exposure and clinical harm co-occurred, but B12 deficiency, diet/possible absorptive contributors and lack of control prevent estimation of a causal folic-acid harm effect. | | bias | Retrospective long-term exposure, absent initial B12 data, case selection, inconsistent diagnostic/imaging descriptions and concurrent treatments limit causal inference. | | precision | No exposed denominator, comparator, person-time, effect interval or verified MCID supports a numerical risk estimate. Unscored is neither zero risk nor a safety guarantee. |

## Explicitly unscored original seven axes

This is a causal clinical-harm/risk question, not efficacy. Common instruction line 94 and task-contract lines 312–314 prohibit forcing safety into efficacy A–F. The supplied rubric and calculator contain no approved mapping of this uncontrolled harm question to a numeric grade. Grade, score and seven axes are explicitly null; the calculator was not executed.

| Original axis | Code | Explanation in this report | |---|---|---| | claim_type | null | No supported original scoring-axis mapping for this safety question. This is a causal clinical-harm question, not efficacy. An unsupported A–F safety grade is not invented. | | endpoint | null | No supported original scoring-axis mapping for this safety question. The main outcome is clinical gait ataxia at presentation in May 2023. The 30-year exposure history and five-month symptom progression are distinct. | | replication | null | No supported original scoring-axis mapping for this safety question. One new original case family was directly extracted. PDF/index copies and treatment follow-up are not independent replications, and historical cases mapped by an opinion are not double-counted. | | independence | null | No supported original scoring-axis mapping for this safety question. Case authors declare no competing interests, but funding, product support and full independence are unverified. Official contextual sources do not certify the case’s independence. | | effect | null | No supported original scoring-axis mapping for this safety question. Exposure and clinical harm co-occurred, but B12 deficiency, diet/possible absorptive contributors and lack of control prevent estimation of a causal folic-acid harm effect. | | bias | null | No supported original scoring-axis mapping for this safety question. Retrospective long-term exposure, absent initial B12 data, case selection, inconsistent diagnostic/imaging descriptions and concurrent treatments limit causal inference. | | precision | null | No supported original scoring-axis mapping for this safety question. No exposed denominator, comparator, person-time, effect interval or verified MCID supports a numerical risk estimate. Unscored is neither zero risk nor a safety guarantee. |

Function/CLI/check_verdict calls, error count, exit code, raw grade and numeric anchor are null because grading was not performed. Four receipts link the final whole report SHA. Nonexecution is not labeled PASS.

## Safety label and detail

**Caution**

Caution is an information label concerning B12-related neurological injury and changes in anemia findings during folic-acid exposure. It is not a grade establishing folic acid as the sole cause or harm in every exposed person. Clinical interpretation depends on actual prescriptions, B12 status and other medicines. Described exposure and treatment are not instructions to take, alter or stop medication. Unreported adverse events do not establish safety. [S03,S04]

Long-/short-course wording in the SmPC is product-label information. It is not used as the case product identity, an individual safe period, a neurotoxicity threshold or a no-harm claim for everyone. An intake upper limit is not an incidence effect derived from this case. [S02,S03,S04]

## Unknown values and revision conditions

| Field | Current value/status | Reason | |---|---|---| | causal_effect | null / not_estimable | An uncontrolled case cannot isolate direct neurotoxicity or exacerbation caused by folic acid. | | comparator | null / not_applicable | The selected case has no unexposed, lower-exposure or active control group. | | at_risk_denominator | null / not_reported | One is the number of selected reported patients, not an at-risk population denominator. | | RR_OR_HR_absolute_risk_CI | null / not_estimable | No valid comparison group, exposed denominator or person-time supports a risk ratio, incidence or confidence interval. | | baseline_B12 | null / not_reported | The value of 84 pg/mL is at presentation, not B12 status at exposure initiation 30 years earlier. | | baseline_neurology | null / not_reported | No neurological examination at exposure initiation or intervening time series is reported. | | dose_verification | null / not_reported | About 5 mg/day almost regularly for 30 years is reported history, without continuous dispensing records, pill counts or cumulative exposure. | | formulation | null / not_reported | Oral folic acid is described, but tablet/capsule, salt, manufacturer, complete composition and absence of added actives are not independently verified. The UK SmPC product is not substituted. | | source_part | null / not_reported | The product’s manufacturing origin and process are unreported; no botanical part is invented. | | additional_active_ingredients | null / not_reported | Folic acid is listed as a separate medication, but there is no complete product ingredient list establishing a chemically single-active formulation. | | diet_total_fortification | null / not_reported | A vegan diet is reported, but dietary folate, fortified intake, total intake and diet duration are not quantified. | | renal_function | null / not_reported | General biochemical descriptions are not converted into confirmed normal eGFR or creatinine values. | | exact_latency | null / not_estimable | Five months of symptoms after a 30-year exposure history do not establish a causal folic-acid latency. | | masking_or_delay_demonstrated | null / not_established | The case lacks before-and-after anemia correction or evidence that folate caused diagnostic delay. B12-related neurological harm does not establish folic acid as its cause. | | MCID_threshold | null / not_reported | No gait-ataxia MCID or individual harm threshold was verified; upper limits and label durations are not substitutes. | | imaging | null / conflicting | The case/caption report no significant MRI signal change, while the discussion mentions a brain white-matter abnormality. Imaging-confirmed cord pathology is not claimed. | | strength_scale | null / conflicting | The source MRC 4/4 is not silently corrected to a conventional five-point scale and is not used as the primary endpoint. | | laboratory_units | null / conflicting | The source MCV unit f/l and its description of follow-up B12 890 as upper reference range are ambiguous; neither is converted into a harm effect. | | funding_independence | null / not_reported | Authors declare no competing interests, but full funding/product support and complete independence remain unverified. | | registration_primary | null / not_applicable | This is not a registered comparative trial. The report’s main outcome is an editorial selection, not a registered primary endpoint. | | longterm_followup_missingness | null / not_reported | Neurological follow-up beyond the described roughly two-month treatment course, losses and an ascertainment denominator are not known. | | dechallenge_rechallenge | null / not_estimable | B12 treatment, folic cessation, rehabilitation and fluoxetine adjustment co-occurred; this is not an isolated folic-acid dechallenge/rechallenge. | | correction_retraction_completeness | null / incomplete_access | No notice was seen in the accessed publisher material, but noisy DOI searches and some failed index access prevent guaranteeing the absence of all corrections or retractions. | | special_populations | null / not_applicable | This single case does not estimate risk for B12-replete people, pregnancy, children, dialysis, other neurological diseases or other folate forms. |

Revise under the eventual existing public ID if a case correction, full product/exposure/B12 trajectory, valid comparison, new direct neurological-harm study or approved safety scoring policy changes this boundary. These are revision conditions, not a recipient research to-do list.

## Reuse, classification and editorial review

This is a new independent claim after full comparison of the 14 candidates and the 3162-record index. Existing safety sources and family exclusions were reused, but efficacy grades, MTX-nausea values and manuscript quality were not inherited. Reused IDs: 121, 486, 1043, 1059, 1133, 1690, 3107, 3108, 3109, 3110, 3159, 3160, 3161, 3162.

Kind M reflects the medication context in the case, not verification of a new product authorization. Category cognition is the existing neurological-health navigation category. The representative list tag is folic acid, and unknowns in the twelve facets remain null with reasons.

Manuscript quality A is a same-model completeness/self-review assessment, not an efficacy/safety grade, independent external review or journal certification. Advertising is uncollected/null, and initial corrections=[]. Separate pre-submission auditing records revisions, source inconsistencies and access failures.

New ID, slug, URL, semantic code and first publication remain null. All 234 input files and historical unpublished/FULL manifest states are unchanged. Latest recipient FULL4 is separate from this content completion. Clinical TODO=[]; only technical format mapping remains downstream. This task ends at 88 without starting 89, deploying or rotating the chat.

## Sources

- [S01] High-folate-low-vitamin B12 interaction syndrome. DOI 10.12890/2025_005398. https://www.ejcrim.com/index.php/EJCRIM/article/download/5398/4633/52747 - [S02] Scientific opinion on the tolerable upper intake level for folate. DOI 10.2903/j.efsa.2023.8353. https://www.efsa.europa.eu/en/efsajournal/pub/8353 - [S03] Folate — Health Professional Fact Sheet. https://ods.od.nih.gov/factsheets/Folate-HealthProfessional/ - [S04] Folic Acid Tablets BP 5mg — Summary of Product Characteristics. https://www.medicines.org.uk/emc/product/14604/smpc - [S05] Among Vitamin B12 Deficient Older People, High Folate Levels are Associated with Worse Cognitive Function: Combined Data from Three Cohorts. DOI 10.3233/JAD-131265. https://journals.sagepub.com/doi/10.3233/JAD-131265 - [S06] Lack of historical evidence to support folic acid exacerbation of the neuropathy caused by vitamin B12 deficiency. DOI 10.1093/ajcn/nqz089. https://pmc.ncbi.nlm.nih.gov/articles/PMC6785032/ - [S07] Pernicious Anemia with Neuropsychiatric Dysfunction in a Patient with Sickle Cell Anemia Treated with Folate Supplementation. DOI 10.1056/NEJMoa022639. https://www.nejm.org/doi/abs/10.1056/NEJMoa022639 - [S08] Folic Acid Safety, Interactions, and Health Outcomes. https://www.cdc.gov/folic-acid/about/safety.html - [S09] Neurotoxicity of folic acid. https://www.rivm.nl/bibliotheek/rapporten/340230001.pdf

Why the efficacy grade and score are unscored

This is a causal clinical-harm/risk question, not efficacy. Common instruction line 94 and task-contract lines 312–314 prohibit forcing safety into efficacy A–F. The supplied rubric and calculator contain no approved mapping of this uncontrolled harm question to a numeric grade. Grade, score and seven axes are explicitly null; the calculator was not executed.

Review performed and remaining limitations

Original product, realized exposure, initial B12/neurological status, control/denominator, causal risk, MRI wording, funding independence and correction-access limits are stated in the body and unresolved.json.

Original evidence table

Receipt — 9 References

Evidence access cutoff: 2026-09-18. Each citation states its actual access level (full-text transcription, abstract, index, or other) and remaining limitations. Bibliographic checking does not certify the study results or treatment efficacy.

What this document does and does not do

Chamgap is an information source. It reports what research has and has not confirmed; it does not tell readers what to take or buy. That decision belongs to readers and, when needed, medical or legal professionals. This verdict reflects literature available up to the search date and may change as new research appears. Nothing here is medical advice.