Added vitamin B12 injections × depressive symptoms during antidepressant treatment

Useful facts when choosing a product

• The paper reports intramuscular B12 1,000 μg once weekly for six weeks.
• The molecular form, brand, manufacturer and complete historical active composition are unverified.
• TCAs or SSRIs were background treatment, and the control group received no sham injection.
• The primary assessment is Table 2’s three-month follow-up; research exposure is not personal prescribing advice.

What the research actually shows

# Added vitamin B12 injections × depressive symptoms during antidepressant treatment

TASK-1050 / R01-090 · Evidence cutoff 2026-09-18 · New independent claim · C / 50 · Safety **Caution** · Manuscript quality **A**

## Thirty-second answer

A small randomized trial in antidepressant-treated patients with low-normal B12 found more participants reaching a 20% HAM-D reduction at 3 months after adding the reported B12 injection. No sham, absent allocation concealment, unequal baseline scores and missing adjusted estimates/CIs prevent a definitive claim about pure B12 or general oral supplementation. The current rule-based assessment is C, 50; safety is Caution and manuscript quality is A.

## Question and verified scope

The original question concerns B12 and depressive symptoms in adults, with diagnosis and severity separated. Oral single-agent treatment, PHQ-9 and placebo were proposed search targets, not inherited study facts. The main claim is therefore **addition of the reported B12 intramuscular injection to antidepressant treatment**. It was selected because the accessible paper provides symptomatic treatment patients, an actual comparator, a declared primary endpoint and both arm denominators. Full product composition is unverified; this is not called proof of pure single-active B12 efficacy. [S01]

The one primary page endpoint is **achievement of at least a 20% reduction from baseline HAM-D**, and the one primary time is **3 months**. The Methods describe follow-up during 12 weeks; Table 2 labels it 3 months. This is the paper-defined primary outcome, not a verified original registered primary endpoint for NCT00939718. The Urdu 20-item HAM-D improves downward. Its exact possible total range and a validated MCID in this setting are unverified. The 20% criterion is a study-defined improvement threshold, not remission. [S01, S02]

PHQ-9 was not selected because it was not the main trial’s actual measure. A 50% reduction, final score and change score remain secondary. Serum B12, homocysteine, fatigue and memory cannot substitute for a depressive-symptom outcome. Field’s oral DASS evidence is separate context because route/population/design differ; B-VITAGE and original 1059 combinations cannot isolate the contribution of B12.

## Boundary against original 1059 and reuse

Original 1059 concerns **oral folate+B6+B12 for primary prevention of depression**, with original **D / 22** retained. The WAFACS/Okereke and Walker families and the Markun review map were reused. Walker also tested a folate+B12 combination; its symptom context is not an identical completed injection-augmentation claim. Original grades, scores, bilingual text, tags, sources and unknowns remain unchanged. Its English equivalence wording is not adopted here, and the historical file was not edited. [R1059]

The semantic fields of the full 3,164-record index and all nine supplied native candidates were checked. 110, 1340, 1440, 3111–3114 and 3164 address other fatigue/anemia, metabolic/patch, neuropathy/cognition/memory outcomes. The already completed 089/3164 memory claim and D28 were not researched or graded again. The input provides **221 selected original files** from 033–036 and 089, not the historical whole ZIPs. The latest recipient record is FULL1/5; it does not mean this new item is published.

## Actual evidence table

| Evidence and access | Intervention, population and time | Actual endpoint, denominator and result | Role here | |---|---|---|---| | S01 Syed 2013; full text, Tables 1/2 | B12 190–300 pg/mL, HAM-D≥16; reported B12 1,000 μg IM weekly for 6 weeks + antidepressants versus antidepressants alone | ≥20% reduction at 3 months: 34/34 versus 27/39; original P<0.001, baseline-adjusted P=0.001 | Main trial; molecular form/full composition unverified; no sham | | S03 Misal 2024; publisher index/reposted abstract only | Adjacent positive report of injection augmentation in late-life depression, 4 weeks | Publisher full-text arm denominators, remission definition and CI unavailable | Positive evidence lead retained, not pooled or counted as verified matching replication | | S04/S05 Field 2022/2026; full text | Same participant family; oral methylcobalamin 1,000 μg/day versus lactose placebo for 30–40 days; symptom-defined rather than diagnosed subgroup | DASS-D≥10: B12 n21, 15.7±6.0→13.9±8.6; placebo n18, 13.9±3.6→10.6±7.7 | Separate oral context; B6 is a separate arm. Isolated adjusted B12 contrast/CI absent | | S06 B-VITAGE; targeted full-text sections | Oral folate+B6+B12 + citalopram; adults≥50, n153 | No significant 12-week remission difference; positive 52-week model kept distinct | Combination-treatment context, not an isolated B12 estimate | | R1059 WAFACS/Walker and review map | Previously completed prevention/oral-combination material | Original values, sources and uncertainty retained | Duplicate-boundary/search reuse only; no new grading |

## Main-trial denominators, scale and numbers

Of 199 screened patients, the 73 with low-normal B12 were allocated 34/39. The 44 with deficiency below 190 pg/mL were outside randomization; the 82 in the normal category are not added to the main denominator. Despite the diagnostic title, the Methods provide a HAM-D≥16 definition; a structured diagnostic interview is not invented. TCAs or SSRIs constituted background treatment, without proof that every participant received an identical drug, dose or stable-treatment period. [S01, Methods/Results]

| Table 1/2 item | Added injection n34 | Control n39 | Original unadjusted P | Original baseline-adjusted P | |---|---:|---:|---:|---:| | Baseline HAM-D, mean±SD | 23.21±5.85 | 19.38±5.70 | 0.006 | Not applicable | | Baseline B12, pg/mL mean±SD | 238.49±33.21 | 245.16±27.82 | 0.36 | Not applicable | | ≥20% reduction at 3 months | 34 (100%) | 27 (69.2%) | <0.001 | 0.001 | | ≥50% reduction, secondary | 15 (44.1%) | 2 (5.1%) | <0.001 | <0.001 | | 3-month HAM-D, mean±SD | 12.12±5.12 | 14.38±4.73 | 0.053 | <0.001 | | Baseline minus follow-up, mean±SD | 11.09±4.58 | 5.00±3.38 | <0.001 | <0.001 |

SD is not SE, and a within-arm change is not an adjusted between-arm effect. The adjusted results provide P values but no adjusted OR/mean difference, SE or CI. Table denominators matching allocation do not prove zero attrition or a complete ITT analysis. No sham and no investigator allocation concealment remain limitations despite assessor masking. Recruitment reached 73 rather than the planned 292, with grant-extension constraints reported. [S01]

## Actual calculations and analyses not performed

Simple arithmetic gives an unadjusted threshold-achievement difference of **30.7692 percentage points** and RR **1.44444**. A separate aggregate-data analysis yields two-sided Fisher P=0.0002425354 and Pearson P=0.0004027167 without Yates correction, consistent with the direction of the paper’s P<0.001. The newly computed unadjusted Newcombe–Wilson 95% interval for the proportion difference is **0.148956–0.464213**. This is neither a published CI nor a baseline-adjusted CI or a probability of treatment success. [CALC-STAT: statistics/summary_reanalysis_execution_receipt.json]

Using rounded final-score means/SDs, the pooled t-test gives P=0.0539596 rather than the paper’s exact display of 0.053. Both are preserved. Baseline-adjusted logistic regression/ANCOVA was **not performed** because linked individual data are unavailable; output, error count and exit code are null with reasons. The all-success intervention cell makes logistic separation handling relevant, without establishing that the original model actually failed. No ARR/NNT or personal treatment instruction is derived.

## Opposing and indirect evidence; study families

In Field 2026’s symptomatic oral B12 subgroup, simple mean improvement was 1.8 points versus 3.3 for placebo. Without an adjusted contrast and change SD, this does not establish harm, exact null effect or equivalence. The paper’s B6-versus-combined-B12/placebo results are not relabeled as B12 effects. The 2026 and 2022 publications share participants, not two independent replications. Conflicting 267/268/254 totals and the 13 exclusions are disclosed. [S04, S05]

B-VITAGE’s 12- and 52-week results concern different time contracts and a combination product. Primary prevention, augmentation, symptom severity, remission and relapse are not pooled into one number. The positive Misal abstract lead is not interpreted beyond its access level. Observational serum B12 associations do not establish the effect of prescribing B12 to an individual and do not replace the randomized primary endpoint. [S03, S06, R1059]

## Safety: Caution

The main trial states that no adverse effects or complications were observed in either group. It does not provide a detailed event table, event-specific surveillance denominators, severity or longer-term exposure, so safety is not established merely by that statement. The MHRA cobalt-sensitivity warning for hydroxocobalamin/cyanocobalamin is separate safety context, without assigning either molecular form to the main trial. Severe allergic symptoms require prompt medical assessment; this is not an instruction to stop all treatment preventively. [S01 Discussion; S07]

Study doses and injection schedules describe research exposure, not personal dosing instructions. B12 is not recommended as a substitute for stopping or changing antidepressants. The low-normal-status trial is not generalized to confirmed deficiency replacement, pregnancy/lactation, children, severe liver/kidney disease or patients with suicidal ideation.

## Seven assessment explanations ### Effect

A small positive signal is present, but its adjusted clinical magnitude is not established. The 20% threshold is neither remission nor a validated MCID, and the original adjusted CI is absent. Magnitude and certainty must be read separately. ### Certainty

The absence of sham/adequate allocation concealment, small sample and baseline imbalance limit confidence. C is the supplied Chamgap rule-based grade, not an official GRADE rating or a probability of treatment success. ### Applicability

The confirmed scope is injection augmentation in antidepressant-treated patients with low-normal serum B12. It is not generalized to oral/sublingual/intravenous use, sufficient status, deficiency replacement, primary prevention, B-vitamin combinations or B12 antidepressant monotherapy. ### Safety

The separate safety label is Caution. The trial’s statement of no observed adverse effects is displayed alongside the absence of detailed event tables. Form-specific regulatory warnings are a separate evidence layer, not reported events in the main trial. ### Verification

The supplied full index, nine native candidates and selected prior originals were examined. Accessible primary publications, official indexes and numeric summaries were checked. Aggregate tests and arithmetic were actually run; original adjusted models, independent external review and server checks were not. ### Limitations

Registration, product composition, diagnostic detail, MCID, adjusted CI, missing-data handling, special groups and longer-term safety remain uncertain. Field’s participant accounting conflict and the rounded-data P discrepancy are disclosed. All reasons appear in this manuscript’s unresolved section. ### Score interpretation

For B/P/R1/I2/EX/B2/CX, the original calculator’s C is adopted. The number 50 is the separate C/one-strength (I2) anchor in the supplied table. Manuscript quality A and safety Caution are independent of C50.

## Seven-axis reasons and separate numeric anchor ### claim_type = B

This is an incremental clinical depressive-symptom claim, not a claim about raising serum B12 or lowering homocysteine. ### endpoint = P

HAM-D improvement is a clinician-rated symptom target placed in policy category P. It is neither a hard event such as death/hospitalization (H) nor a biomarker (S). ### replication = R1

Syed 2013 is the fully accessible main trial for this injection-addition/3-month primary outcome. A 4-week abstract, a different oral arm and combinations are not counted as verified matching replications. R1 does not certify a large confirmatory trial. ### independence = I2

S01 explicitly identifies the Aga Khan University Research Council grant G&C 70210 and declares no conflicts. Unreported product supply remains unresolved. This axis concerns evidence funding, not independent external review of this manuscript. ### effect = EX

The positive P and 34/34 versus 27/39 primary counts are retained. Baseline-adjusted estimates/SEs/CIs are absent and the original model cannot be reconstructed. Policy case 30 supports EX(a); unadjusted proportions or within-arm changes do not establish E+. ### bias = B2

No sham/participant masking and no investigator allocation concealment are key limitations. Baseline imbalance, under-recruitment and incomplete missing-data/antidepressant-stratification reporting further limit confidence. Assessor masking does not remove them. ### precision = CX

The original adjusted confidence interval is unavailable. A separately computed unadjusted proportion-difference CI is not its substitute. The C1/C0 lower-bound branch for E0/E− is not activated for EX. The supplied policy’s clinical-claim classification, case 29 (funding), case 30 (adjusted estimates), EX/B2 caps and within-band anchor apply. The letter function does not return the number 50. I2 is the only strength; H/R2/E+/C1/B0/a large hard-endpoint trial are not met. C with one strength maps to 50. Four separate original execution receipts bind the full final-verdict SHA.

## Twelve classification facets

| Facet | Value | |---|---| | Kind | M | | Canonical entity | Vitamin B12 | | Source/part | Manufacturing origin and molecular form unreported; not a plant-part intervention | | Formulation | Reported B12-addition injection; brand, molecular form and complete active composition unverified | | Route | Intramuscular | | Dose | Reported 1,000 μg per injection, once weekly; not a personal prescribing instruction | | Duration | Injections for 6 weeks; primary assessment at 3 months (Methods: follow-up during 12 weeks) | | Population | Antidepressant-treated outpatients with Urdu 20-item HAM-D≥16 and serum B12 190–300 pg/mL | | Claim | Depressive-symptom improvement when the reported B12 injection is added to antidepressants | | Primary endpoint | At least 20% HAM-D decrease at 3 months; paper primary, original registered primary unverified | | Comparator | Antidepressants alone, without sham injection; arm-specific medication distribution unreported | | Duplicate key | reported-B12-injection\|IM\|antidepressant-treated-depression\|serum-B12-190-300pg-mL\|HAMD-ge20pct-improvement\|3months\|AD-alone-no-sham |

## Complete disclosure of unresolved information ### U01 · not_reported

The paper does not identify the molecular form, brand, manufacturer or complete historical injection composition. It reports B12 addition; it does not independently verify a pure single-active product. ### U02 · inaccessible

The original primary outcome, first registration date and change history of NCT00939718 were not verified from the official page/API. The paper-defined primary outcome is kept separate. ### U03 · conflicting

A registry mirror mentions methylcobalamin/Bevidox, 17 items and masking details that were not verified in the original record and partly differ from the paper. A current retail tablet is not evidence of the historical injection composition. ### U04 · not_reported

The paper defines depression using Urdu 20-item HAM-D≥16. A structured diagnostic interview, DSM criteria, exact age limits and a uniform diagnostic severity category were not verified. ### U05 · not_reported

The possible total range of the particular 20-item scale and a validated MCID for this population/outcome were not verified. Neither a 17-item range nor the trial’s 20% threshold is substituted. ### U06 · not_reported

Baseline MMA, homocysteine, functional deficiency, malabsorption classification and final B12 concentrations were not verified. Low-normal serum values do not establish full nutritional sufficiency. ### U07 · not_reported

Total dietary B12 intake, other supplements/folate/B6 use, and product excipients/additional active ingredients were not adequately reported. ### U08 · not_reported

Arm-specific antidepressant distributions, stable-dose duration, medication changes and prior treatment-failure severity are absent. The participants are not assumed to have all received the same SSRI. ### U09 · not_reported

Adjusted logistic/ANCOVA effect estimates, SEs, CIs, design matrices and individual data are absent. Marginal summaries and P values cannot reconstruct the original adjusted models. ### U10 · not_reported

No participant in the injection arm failed the primary threshold, making separation handling relevant to ordinary logistic regression; its handling is not reported. An actual model error is not asserted. ### U11 · not_reported

Tables retain the allocated 34/39 denominators, but a separate completion/attrition flow, ITT definition and missing-data/imputation procedure are insufficiently described. Zero attrition is not invented. ### U12 · not_reported

Multiplicity adjustment for secondary endpoints and most exact P values are absent. <0.001 is not converted into an exact value, and nonsignificance is not equivalence. ### U13 · not_reported

The 20% and 50% reductions are improvement criteria in this paper. A remission definition and remission rate for the main trial were not verified. ### U14 · not_reported

The authors state that no adverse effects or complications were noted, but do not tabulate event-specific counts, surveillance denominators, severity, causality or person-time. Safety tables of 0/34 or 0/39 are not invented. ### U15 · not_reported

This trial does not establish depressive-symptom benefit or safety for longer-term injections, pregnancy/lactation, children, severe liver/kidney disease or patients with suicidal ideation. ### U16 · not_reported

Syed’s university grant and conflict declaration were read, but product supply/manufacturer and every possible patent/employment relationship were not established. ### U17 · inaccessible

For Misal 2024, publisher-index information and a reposted abstract were available. Arm denominators, remission definition, adjusted estimates, product, registration, funding and safety were not established, so it was not pooled or used for a replication upgrade. ### U18 · conflicting

Field 2026 reports 267 in the abstract, 268 in Methods and 254 after 13 exclusions in Results. The arithmetic gives 255. Table 1’s B12 21/placebo 18 subgroup is retained without inventing an extra excluded person. ### U19 · not_reported

Field 2026 does not provide the isolated B12-versus-placebo baseline-adjusted contrast, CI, change SD or response/remission results. B6-versus-pooled-control statistics are not assigned to B12. ### U20 · not_reported

Field’s biochemical B12 status, standardized antidepressant background and B12-specific adverse-event table were not established. A vivid-dream withdrawal in the B6 arm is not attributed to B12. ### U21 · not_searched

Embase, subscription-only specialist databases and private study data were not searched. The review is bounded by the web tool, accessible primary publications/official indexes and supplied material. ### U22 · not_applicable

This is same-model self-verification, not independent external review or journal certification. Unassigned ID/URL/semantic code/first publication is a technical state, not a clinical deferral. ### U23 · not_searched

Advertising claims were not systematically collected, so what_ads is null. Research descriptions are not relabeled as advertising. ### U24 · not_reported

Completed injection counts/adherence by participant, assay interference, kidney-function values and depression response by malabsorption cause were insufficiently reported. ### U25 · conflicting

The reported unadjusted follow-up HAM-D P=0.053 is not identical to the pooled t-test P=0.0539596 reconstructed from rounded means/SDs. Individual/unrounded data were unavailable; neither value overwrites the other.

## Editorial and publication status

This is a current-value manuscript with same-model self-verification, not independent clinical review or journal certification; independent=false. Efficacy C50, safety Caution and manuscript quality A are separate values. result_status=completed_with_uncertainty, editorial_review.status=ready_with_uncertainty, and publication_status=needs_id_assignment. New ID, slug, URL, semantic code and first publication are null; initial corrections=[]. Pre-submission findings are retained in a separate audit. clinical_todo=[]: recipients need no clinical re-investigation, regrading, statistical rerun or calculator rerun, only technical identity/display mapping. No server access or deployment was performed.

## Sources and locations

**S01.** Vitamin B12 Supplementation in Treating Major Depressive Disorder: A Randomized Controlled Trial. DOI: 10.2174/1874205X01307010044. [publisher_full_html](https://openneurologyjournal.com/VOLUME/7/PAGE/44/FULLTEXT/). Methods: Randomization and Masking; Statistical Considerations; Results; Table 1; Table 2; Discussion; Conflict of Interest; Funding

**S01B.** PubMed bibliographic record for Syed et al.. DOI: 10.2174/1874205X01307010044. [official_index_abstract](https://pubmed.ncbi.nlm.nih.gov/24339839/).

**S02.** NCT00939718 official trial registration. [inaccessible_fields](https://clinicaltrials.gov/study/NCT00939718). Official page returned shell only; API retrieval failed

**S03.** Effect of Vitamin B12 supplementation in late-life depression: A randomized controlled trial. DOI: 10.4103/jgmh.jgmh_11_24. [publisher_search_index_only_fulltext_inaccessible](https://www.ovid.com/jnls/jgmh/fulltext/10.4103/jgmh.jgmh_11_24~effect-of-vitamin-b12-supplementation-in-late-life).

**S04.** High-dose Vitamin B6 supplementation reduces anxiety and strengthens visual surround suppression. DOI: 10.1002/hup.2852. [publisher_full_html](https://onlinelibrary.wiley.com/doi/full/10.1002/hup.2852). Methods; Depression results; Conflicts/Funding

**S05.** High Dose Vitamin B6 Reduces Depression Symptoms. DOI: 10.1155/da/8926512. [publisher_full_html](https://onlinelibrary.wiley.com/doi/10.1155/da/8926512). Introduction last paragraphs; 2.1 Participants and Design; 2.2 DASS Questionnaire; 2.3 statistical analysis; Table 1; 3.1 Outliers; 3.3 depression analyses; Funding/Conflicts

**S06.** B vitamins to enhance treatment response to antidepressants in middle-aged and older adults: results from the B-VITAGE randomised, double-blind, placebo-controlled trial. DOI: 10.1192/bjp.bp.114.145177. [publisher_full_html_targeted_sections](https://www.cambridge.org/core/journals/the-british-journal-of-psychiatry/article/b-vitamins-to-enhance-treatment-response-to-antidepressants-in-middleaged-and-older-adults-results-from-the-bvitage-randomised-doubleblind-placebocontrolled-trial/1E778971A157175B13BBA6084E581F2C). Abstract; Methods; Randomisation and masking; Statistical analyses; Results

**S07.** Vitamin B12 (hydroxocobalamin, cyanocobalamin): advise patients with known cobalt allergy to be vigilant for sensitivity reactions. [official_regulator_full_html](https://www.gov.uk/drug-safety-update/vitamin-b12-hydroxocobalamin-cyanocobalamin-advise-patients-with-known-cobalt-allergy-to-be-vigilant-for-sensitivity-reactions). Advice for healthcare professionals; Advice for patients; Cobalt sensitivity reactions

**R1059 / R110 / R1340 / R1440 / R3111–R3114 / R3164.** Supplied native JSON and selected completed originals, used for boundaries/reuse rather than repeated clinical validation. Exact original paths and hashes are in sources.json and intake/input_mapping.json.

Why this is classified as C (50)

For B/P/R1/I2/EX/B2/CX, the original calculator’s C is adopted. The number 50 is the separate C/one-strength (I2) anchor in the supplied table. Manuscript quality A and safety Caution are independent of C50.

Counterpoint. The confirmed scope is injection augmentation in antidepressant-treated patients with low-normal serum B12. It is not generalized to oral/sublingual/intravenous use, sufficient status, deficiency replacement, primary prevention, B-vitamin combinations or B12 antidepressant monotherapy.

Rejudgment record. Original-calculator letter C plus separate original-table anchor 50, linked to actual pre-submission execution receipts — Supplied policy cases 29/30 and C/one-strength anchor

Review performed and remaining limitations

Registration, product composition, diagnostic detail, MCID, adjusted CI, missing-data handling, special groups and longer-term safety remain uncertain. Field’s participant accounting conflict and the rounded-data P discrepancy are disclosed. All reasons appear in this manuscript’s unresolved section.

Evidence Table

Receipt — 17 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.