The biological rationale for why B vitamins might matter in cancer is compelling. These water-soluble compounds function as coenzymes in one-carbon metabolism, the biochemical network that supplies methyl groups for DNA methylation and nucleotides for DNA synthesis and repair. Folate is converted to S-adenosylmethionine, the universal methyl group donor for DNA methylation, while vitamin B12 acts as a cofactor in that reaction and vitamin B6 supports several key enzymes in methyl group metabolism. A deficiency in these nutrients could, in theory, disrupt DNA integrity, impair repair mechanisms, and increase genetic mutations. Yet the same pathways raise a darker possibility: in cells with pre-existing neoplastic lesions, abundant folate might also fuel the proliferation of damaged tissue, a concern that has shadowed folic acid supplementation trials for decades.

The review team conducted a structured literature search in PubMed up to 2 September 2025, prioritizing meta-analyses—including dose-response meta-analyses—of observational studies, and incorporating randomized controlled trials of B-vitamin supplementation when clinically relevant. The initial search yielded 512 records; after screening, 120 articles were retained for full-text assessment and 45 studies made it into the final narrative synthesis. Although the work was not designed as a systematic review, the selection process followed a structured and transparent approach inspired by PRISMA recommendations, with three reviewers resolving disagreements through discussion. The team also appraised methodological quality narratively, focusing on exposure assessment, confounding, selection bias, and outcome assessment.

The clearest signal emerged for vitamin B6. In a meta-analysis of 28 observational studies, Lai and colleagues reported an inverse association between dietary pyridoxine intake and colorectal cancer risk (OR 0.80; 95% CI 0.68–0.94), while circulating pyridoxal 5′-phosphate—the active form of the vitamin—showed an even stronger protective association (OR 0.54; 95% CI 0.35–0.84). Favorable associations were particularly evident among women and for colon cancer specifically. The review’s summary table reports a relative risk of 0.56 (95% CI 0.46–0.67, p < 0.001) for the highest versus lowest dietary intake categories, with biomarker-based analyses suggesting risk reductions of 30 to 50 percent. In a large field synopsis of 121 observational studies, Mocellin and colleagues found that higher dietary B6 intake was associated with reduced risk of gastrointestinal cancers overall (RR 0.68; 95% CI 0.61–0.75), although site-specific findings for gastric cancer were less consistent.

Vitamin B6 also stood out for pancreatic cancer, one of the deadliest malignancies. Wei and Mao observed strong inverse associations, with relative risks of 0.63 (95% CI 0.48–0.79) for dietary intake and 0.65 (95% CI 0.52–0.79) for circulating pyridoxal 5′-phosphate. Their dose-response analysis indicated a 9 percent risk reduction per 10 nmol/L increment in circulating levels. Beyond the digestive tract, a dose-response meta-analysis by Qiang and colleagues reported that higher dietary intake of vitamin B6 and folate was associated with reduced risk of esophageal cancer, with relative risks of 0.57 (95% CI 0.47–0.69) for B6 and odds ratios of 0.62 (95% CI 0.56–0.68) for folate.

Riboflavin showed a more modest but still favorable pattern. An observational meta-analysis by Liu and colleagues found a protective association between vitamin B2 intake and colorectal cancer risk (OR 0.83; 95% CI 0.75–0.91). A Korean case-control study by Lu and colleagues involving 377 individuals with gastric cancer and 756 healthy controls suggested that regular riboflavin intake offers protection against gastric cancer, particularly in carriers of specific genetic variants of the MTRR gene involved in cobalamin metabolism. However, the picture is not uniformly rosy: a recent retrospective case-control study exploring blood riboflavin levels and colorectal cancer onset reported an increased risk proportional to circulating concentration, a finding that underscores how the source, dose, and timing of exposure may critically modify any apparent association.

Folate delivered the most context-dependent results. Moazzen and colleagues reported that randomized controlled trials of folic acid supplementation showed no significant effect on colorectal cancer risk (RR 1.07; 95% CI 0.86–1.43), whereas prospective cohort studies of dietary folate intake indicated a significant risk reduction (RR 0.71; 95% CI 0.59–0.86). This divergence between observational and interventional evidence suggests that dose, timing of exposure, baseline folate status, and background fortification policies may critically influence observed associations. Differences between naturally occurring folates and synthetic folic acid may also partly explain the lack of concordance. For breast cancer, a dose-response meta-analysis of prospective studies by Zhang and colleagues found no overall association (RR 0.97; 95% CI 0.90–1.05), but a J-shaped relationship emerged, with the lowest risk between 200 and 320 μg/day and an increased risk above 400 μg/day. For ovarian cancer, Wang and colleagues found no significant association with either dietary or total folate intake. Regarding melanoma, one meta-analysis of randomized controlled trials suggested a reduced risk with folic acid supplementation (RR 0.47; 95% CI 0.23–0.94), but larger pooled analyses involving more than 50,000 participants did not confirm this finding.

Vitamin B12 proved the most inconsistent of the four. A dose-response meta-analysis by Sun and colleagues indicated a modest inverse association with colorectal adenoma and cancer risk (RR 0.963 per +4.5 μg/day), with stronger effects for dietary intake compared with total intake. Yet for esophageal adenocarcinoma, the same dose-response meta-analyses revealed a positive association, with each additional 1 μg/day of cobalamin intake associated with a 2 percent increase in risk (OR 1.30; 95% CI 1.05–1.62). The review authors caution that this finding should not be overlooked, as it suggests cobalamin may have site-specific and context-dependent associations with cancer risk rather than uniformly neutral or protective effects. Possible effect modifiers include obesity, gastroesophageal reflux disease, alcohol consumption, and overall dietary patterns, alongside residual confounding inherent to observational research. For pancreatic cancer, no significant associations were identified for vitamin B12 (RR 0.99; 95% CI 0.59–1.38).

The review also contextualizes these findings within a broader landscape of B-vitamin oncology research. Long-term high-dose supplementation with vitamins B6 and B12 has been associated with increased lung cancer risk in specific subgroups such as male smokers, according to the VITAL cohort study by Brasky and colleagues. Randomized trials of folic acid supplementation have generally shown no significant protective effect on cancer incidence and, in some cases, have raised concerns about potential increases in cancer risk or progression, particularly in populations with pre-existing neoplastic lesions or high baseline exposure. Emerging evidence from Mendelian randomization studies provides additional support for a potential protective role of genetically predicted vitamin B6 across multiple cancer sites, though findings remain limited in number. Meanwhile, high-dose pyridoxine supplementation has been associated with neurotoxicity, particularly sensory neuropathy, a risk that becomes clinically relevant when supplementation exceeds physiological requirements or is used long term.

The authors conclude that the current body of evidence does not support the use of high-dose B-vitamin supplementation for primary cancer prevention in the general population. Rather, the findings are consistent with the importance of achieving adequate B-vitamin intake through balanced dietary patterns. The review acknowledges several limitations: the observational nature of most underlying evidence, susceptibility to residual confounding and dietary misclassification through food frequency questionnaires, considerable variability in baseline dietary patterns and fortification policies across geographic regions, and the frequent use of pharmacological doses in randomized trials that may not reflect habitual dietary intake. Future research should conduct well-designed prospective cohort studies with repeated dietary and biomarker assessments, integrate genetic and epigenetic markers to clarify mechanisms, and target high-risk subgroups such as individuals with folate deficiency, high alcohol consumption, pre-existing preneoplastic lesions, or genetic variants affecting one-carbon metabolism. Until then, the message remains clear: B vitamins are essential, but more is not necessarily better when it comes to cancer prevention.

Subject of Research: B vitamins intake and cancer risk: a structured narrative review of evidence on riboflavin, pyridoxine, cobalamin and folate

Article Title: B vitamins intake and cancer risk: a structured narrative review of evidence on riboflavin, pyridoxine, cobalamin and folate

Article References: Bertuccioli, A., Palazzi, C. M., Massaro, I. A., Cazzaniga, M., Cavecchia, I., Matera, M. R., Biagioli, V., Imperatori, L., Di Pierro, F., & Belli, A. (2026). B vitamins intake and cancer risk: a structured narrative review of evidence on riboflavin, pyridoxine, cobalamin and folate. Pathology & Oncology Research, 32, Article 1612380. https://doi.org/10.3389/pore.2026.1612380

Image Credits: AI Generated

DOI: 10.3389/pore.2026.1612380

Keywords: vitamins, intake, cancer, risk, structured, narrative, review, evidence, riboflavin, pyridoxine, cobalamin, folate

The biochemistry of cobalamin explains why deficiency matters so much. Vitamin B12 is a water-soluble vitamin synthesized only by microorganisms, and it is found almost exclusively in animal-derived foods such as red meat, fish, eggs, and milk, with a very small contribution from foods or water exposed to B12-producing bacteria. Within the body, cobalamin serves as a critical cofactor in two reactions: the conversion of homocysteine to methionine via methionine synthase, and the conversion of methylmalonyl-CoA to succinyl-CoA via methylmalonyl-CoA mutase. These reactions make the vitamin essential for DNA synthesis, red cell production, and the maintenance of nerve myelin. Deficiency can therefore manifest as megaloblastic anemia and glossitis, peripheral neuropathy, subacute combined degeneration of the spinal cord, and cognitive impairment. In pregnancy, low maternal B12 has been associated with neural tube defects and other adverse birth outcomes, as well as longer-term metabolic risk in the offspring.

The life-course pattern reported in India is striking. According to the CNNS, the first nationally representative estimate of pediatric B12 status, deficiency affects 13.8 percent of pre-school children aged 1 to 4 years and 17.3 percent of school-age children aged 5 to 9 years, with little difference between boys and girls at these ages. The rate roughly doubles in adolescence: 31.0 percent of adolescents aged 10 to 19 years were deficient in the CNNS, and a pooled meta-analytic estimate for the 0 to 17-year age band was higher still, at 57 percent, though based on fewer studies and a wider confidence interval. Among adults, the pooled prevalence across the largest meta-analysis was 48 percent, with a 95 percent confidence interval of 35 to 62 percent, and one Indian study reported deficiency as high as 78.5 percent in healthy adults depending on the serum cut-off used. Among the elderly, a review of Indian adults reported 61.7 percent deficiency using varied cut-offs, and hospital series show diagnosed cases increasing with age and peaking around 50 to 65 years.

Pregnant women represent a particularly vulnerable group. A synthesis of 46 studies on B12 status among pregnant and lactating women in India found deficiency prevalence ranging from 40 to 70 percent, and a meta-analysis estimated that approximately 67 percent of pregnant women are affected. International meta-analyses have similarly found insufficient B12 during pregnancy to be common globally and associated with lower birth weight. In Indian cohorts, low maternal B12 alongside adequate-to-high folate has been associated with neural tube defects, recurrent pregnancy loss, gestational diabetes, pre-eclampsia, intrauterine growth restriction, and adverse metabolic and cognitive outcomes in the child, although the review emphasizes that these data derive from observational cohorts and should be interpreted as associations rather than demonstrated causal pathways. The risk persists into lactation, since low maternal stores lower breast-milk B12 and place exclusively breastfed infants at risk, yet current Indian antenatal policy supplements only iron and folate, not vitamin B12.

Diet is the single most important determinant of B12 status in India. Individuals consuming a lacto-vegetarian diet, a substantial share of the Indian population, have significantly lower cobalamin levels than those consuming a mixed diet, and reviews consistently show several-fold higher deficiency among lacto-vegetarians than non-vegetarians. In one study of healthy young vegetarian adults, approximately 50 percent had serum B12 below the deficiency threshold, 70 percent of men and 50 percent of women had low holotranscobalamin, and 92 percent of men and 50 percent of women had hyperhomocysteinaemia. Dietary insufficiency, not pernicious anemia, is the dominant cause of B12 deficiency in India, reflecting the combined effects of a largely cereal-based lacto-vegetarian diet, low consumption of animal-based foods, and high rates of poverty. Beyond diet, the most consistently reported associated factors are low socioeconomic status, advancing age, and female sex, with age-related malabsorption from atrophic gastritis, reduced gastric acid secretion, and Helicobacter pylori infection adding to the burden among older adults.

Regional heterogeneity within India is substantial. In the CNNS, B12 deficiency prevalence was highest in the central region of India across every pediatric age band assessed: 21.0 percent in pre-schoolers, 29.1 percent in school-age children, and 42.7 percent in adolescents, each substantially above the corresponding national estimates. Despite this marked regional gradient, the same survey found no consistent difference in B12 deficiency between rural and urban locations at the national level, a pattern that sits alongside inconsistent rural/urban findings from individual state-level studies. Studies from Pune comparing middle-aged men in rural, slum, and middle-class urban settings found a high frequency of low B12 across all three settings, with rural and slum residents faring worse than urban middle-class residents, suggesting that dietary intake and socioeconomic status appear to be stronger determinants of B12 status than geographic setting per se.

Outside India, B12 deficiency follows clear geographic and economic lines. High-income populations with adequate access to animal-source foods have lower deficiency prevalence, while low- and middle-income populations are more affected owing to both scarcity of animal-source foods and a higher burden of malabsorptive conditions. A widely cited compilation of national and regional datasets found serum B12 deficiency in 11 to 15 percent of women in Germany and the United Kingdom, 31 percent of elderly individuals in the UK, 49 percent of adults in India, 40 percent of children in Kenya, 28 percent of pregnant women in Nepal, and 20 percent of the same demographic group in New Zealand. In Latin America and the Caribbean, an analysis of 41 studies suggested up to 61 percent of individuals in Mexico and Central/South America had inadequate B12 stores. In sub-Saharan Africa, available national micronutrient surveys indicate 54 percent, 35 percent, and 15 percent B12 deficiency among reproductive-age women in Malawi, Kenya, and Ethiopia respectively, with a lower prevalence of 7 percent in Ghana. The Indian subcontinent and neighboring South Asian countries report the highest deficiency prevalence worldwide, 70 to 80 percent in some series, primarily reflecting a predominantly vegetarian, carbohydrate-based diet.

A key constraint on reconstructing a 50-year trend is that no single longitudinal, nationally representative cobalamin survey exists in India, or indeed anywhere else, spanning this period. Most available estimates are cross-sectional snapshots using different biomarkers, cut-offs, and populations. Evidence from the 1970s and 1980s was largely clinical and descriptive, with early recognition that asymptomatic Indian lacto-vegetarians had lower B12 than non-vegetarian counterparts and observations of anemia among Asian immigrants in London raising the possibility of a population-level problem. Systematic biochemical assessment began at the turn of the millennium with the Pune studies, which documented a high prevalence of low cobalamin with concomitantly elevated homocysteine across sampled Indian populations. Since 2000, the evidence base has grown substantially, with clinic and community studies through the 2010s consistently showing an adult prevalence of approximately 47 percent, the CNNS providing the first national pediatric and adolescent estimates, and two independent meta-analyses synthesizing studies to 2020 and 2023 reporting pooled adult prevalence of 51 to 53 percent. A large 2024 health-check dataset of over 15,000 asymptomatic adults across seven northern Indian states reported a considerably lower figure of 24.2 percent, which most plausibly reflects a higher serum cut-off and a more affluent, urban, self-selected sample rather than a genuine decline in burden.

A limitation running through every prevalence estimate cited in the review is the absence of biomarker and cut-off harmonization. Total serum/plasma B12 thresholds used across the cited studies range from below 150 pmol/L to below 211 pg/mL, and most studies rely on total B12 rather than on functional markers such as holotranscobalamin or methylmalonic acid, which more directly reflect tissue cobalamin status. Total serum B12 is itself an indirect marker that can be influenced by recent dietary intake, by pregnancy-related haemodilution, and by inter-laboratory assay variability. A concrete illustration of the resulting inconsistency comes from a single cohort of healthy young Indian vegetarians, in which roughly 50 percent were deficient by total serum B12, but 70 percent of men and 50 percent of women had low holotranscobalamin, and 92 percent of men and 50 percent of women had hyperhomocysteinaemia, meaning a 20-percentage-point-or-greater spread in deficiency depending solely on which biomarker was used. Consequently, differences between surveys, such as the 24.2 percent from the 2024 health-check sample versus 51 to 53 percent from pooled meta-analyses, are driven substantially by differences in cut-points, biomarkers, and sampling frames in addition to any true temporal change.

The review concludes that vitamin B12 deficiency is a largely preventable, straightforward-to-treat disorder whose continued neglect in India’s public health programming is difficult to attribute to anything other than insufficient prioritization relative to its documented scale. From a public health perspective, the authors call for harmonized diagnostic criteria for B12 deficiency, ideally combining total circulating B12 with a functional marker such as holotranscobalamin or methylmalonic acid, and routine screening particularly for pregnant and lactating women, adolescents, the elderly, and strict vegans and vegetarians. Evidence-based strategies to reduce the burden include dietary diversification, food fortification, and the addition of B12 to existing antenatal and adolescent supplementation programmes, which currently contain only iron and folic acid. On fortification specifically, the evidence is directionally supportive but implementation-dependent: India’s current 2018 FSSAI fortification standards for folic acid and vitamin B12 are set at roughly 90 percent below WHO-recommended levels for Indian consumption patterns, having been lowered from stronger 2016 standards, and a recent review identifies concrete barriers to more effective scale-up, including uneven regulatory enforcement, continued reliance on voluntary fortification of open-market atta, variable miller-level capacity and quality assurance, and insufficient long-term monitoring of nutritional outcomes among beneficiaries. Large-scale staple fortification has also drawn published concerns from Indian public-health researchers regarding the strength of the underlying evidence base and potential safety considerations at higher fortification levels, so any revision of national standards should be accompanied by the monitoring and evaluation infrastructure needed to track both benefit and safety at population scale. Regional data gaps, particularly the absence of CNNS-scale state-level surveillance in adults, should also be prioritized in future national nutrition surveys.

Subject of Research: Vitamin B12 deficiency epidemiology in India and worldwide

Article Title: Vitamin B12 deficiency: epidemiological trends and risk factors

Article References: Toteja, G. S., & Toteja, N. (2026). Vitamin B12 deficiency: epidemiological trends and risk factors. Frontiers in Nutrition, 13, Article 1930037. https://doi.org/10.3389/fnut.2026.1930037

Image Credits: AI Generated

DOI: 10.3389/fnut.2026.1930037

Keywords: vitamin B12, cobalamin, micronutrient deficiency, India, epidemiology, vegetarian diet, pregnancy, food fortification, Vitamin, deficiency, epidemiological, trends

The scale of the potential problem is considerable. When the researchers pooled studies that used consistent diagnostic thresholds for vitamin B12 deficiency, they found that approximately 32.1 percent of metformin-treated patients were deficient — nearly one in three. That figure, derived with strikingly low statistical heterogeneity (an I² of 0.0 percent, meaning the included studies agreed with each other to an unusual degree), suggests the association is not an artifact of any single population or laboratory method. In a drug taken by well over one hundred million people worldwide, a deficiency prevalence of one third translates into an enormous population-level burden.

The link between metformin and nerve damage held up across study designs. In the pooled analysis of six cross-sectional studies, metformin users had 84 percent higher odds of peripheral neuropathy than non-users (pooled odds ratio of 1.84, with a 95 percent confidence interval of 1.18 to 2.88, and a p-value of 0.007). Two cohort studies pointed in the same direction: one reported an adjusted odds ratio of 1.57, while another found a hazard ratio of 1.84, meaning that over time, metformin users developed neuropathy at nearly twice the rate of comparable patients not taking the drug. Consistency across cross-sectional snapshots and longitudinal follow-up strengthens the case that the association is not merely a statistical fluke.

The mechanistic story centers on vitamin B12, or cobalamin, an essential micronutrient the human body cannot synthesize. B12 is a cofactor for two enzymatic reactions: one converts methylmalonic acid to succinyl-CoA, and the other regenerates methionine from homocysteine via methylation. When B12 runs short, methylmalonic acid accumulates and impairs myelin synthesis and maintenance — the fatty sheath that insulates nerve fibers and allows electrical signals to propagate rapidly. The resulting damage manifests as the numbness, tingling, burning pain, and balance problems characteristic of peripheral neuropathy. Metformin is thought to interfere with B12 absorption in the ileum, likely through effects on calcium-dependent membrane processes and alterations in intestinal motility and the gut microbiome, though the precise molecular pathway remains an active area of investigation.

Crucially, the meta-analysis showed that B12 deficiency itself is an independent predictor of neuropathy in people with type 2 diabetes. In a diabetes-specific pooling of four studies, B12 deficiency more than doubled the odds of peripheral neuropathy (odds ratio of 2.09, 95 percent confidence interval 1.03 to 4.22, p = 0.041). This is an important distinction, because diabetic peripheral neuropathy is usually attributed to chronic hyperglycemia, oxidative stress, and microvascular disease. The new data indicate that a correctable nutritional deficit contributes a measurable share of the risk — a share that has historically been misattributed to diabetes itself and therefore left untreated.

Perhaps the most striking finding is the evidence of synergy. In one included cohort study, patients who took metformin and were simultaneously B12 deficient faced a hazard ratio of 5.76 for peripheral neuropathy — a nearly six-fold elevation in risk compared with patients who had neither exposure. That multiplicative pattern is biologically plausible: metformin depletes the nutrient, the nutrient’s absence damages myelin, and diabetes independently stresses the same peripheral nerves. When all three forces converge, the nerve’s reserve capacity is overwhelmed. For clinicians, the implication is that the two risk factors cannot be evaluated in isolation; a metformin user with borderline B12 levels may be in a far more dangerous position than either number alone suggests.

The rigor of the review lends weight to its conclusions. The protocol was prospectively registered in PROSPERO before the literature search began, a safeguard against outcome switching and selective reporting. The authors searched PubMed, Google Scholar, Europe PMC, and the Cochrane Central Register from January 2009 to April 2026, and two reviewers independently screened records with strong agreement (Cohen’s kappa of 0.82, well above the conventional threshold for substantial agreement). Study quality was assessed with the Newcastle-Ottawa Scale, and the certainty of evidence was graded using the GRADE framework. The meta-analyses employed random-effects models using the DerSimonian–Laird estimator, the standard approach when true effect sizes may vary across populations.

There are, inevitably, caveats. The headline neuropathy estimate came from cross-sectional pooling with moderate heterogeneity (I² of 68.9 percent), which means the included studies varied considerably in their populations, neuropathy diagnostic tools, and definitions of B12 deficiency. Cross-sectional designs cannot establish temporality — it is possible, though biologically less likely, that patients with neuropathy were simply more likely to be prescribed metformin. The observational nature of the evidence means that randomized trials of B12 supplementation in metformin users remain the gold standard for proving that treatment can prevent or reverse the nerve damage. The authors themselves frame their findings as a call for routine monitoring rather than a verdict against the drug.

Indeed, the clinical calculus is not a simple one. Metformin is among the most effective, best tolerated, and most affordable glucose-lowering agents available, with established cardiovascular benefits and a safety record spanning decades. The researchers are emphatic that the answer is not to abandon the drug but to close the surveillance gap around it. Current practice varies widely: some guidelines recommend periodic B12 testing in long-term metformin users, while others are silent on the matter, and in routine care, B12 levels are often checked only after symptoms appear — by which point neurological damage may be partially irreversible. The finding that deficiency affects roughly a third of treated patients at uniform diagnostic thresholds argues for proactive screening, particularly for those on higher doses or longer durations of therapy.

The practical response, the authors argue, is straightforward: routine B12 monitoring and supplementation for long-term metformin users. Oral B12 supplements are inexpensive, safe, and effective in most cases, and early correction of deficiency offers the best chance of preventing permanent nerve injury. With type 2 diabetes affecting more than half a billion people globally and metformin remaining the first-line therapy almost everywhere, even modest reductions in neuropathy risk through a simple blood test and a cheap vitamin could avert enormous suffering. The study, registered prospectively under PROSPERO number CRD420261363992, transforms a long-suspected side effect into a quantified, actionable clinical target — and reminds the medical world that even the most trusted drugs deserve continued scrutiny.

Subject of Research: Association of long-term metformin use with vitamin B12 deficiency and peripheral neuropathy in type 2 diabetes

Article Title: Association of metformin use with vitamin B12 deficiency and peripheral neuropathy in type 2 diabetes mellitus: a systematic review and meta-analysis

Article References: Mohammed, N., Taha, A. S. O., MohamedSalih, N., Mohammed, M., Rahama, T., Abdelrahman, N., Idris, M. O. A., Saeed, N., & Mustafa, N. A. A. (2026). Association of metformin use with vitamin B12 deficiency and peripheral neuropathy in type 2 diabetes mellitus: a systematic review and meta-analysis. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02520-7

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02520-7

Keywords: metformin, vitamin B12 deficiency, peripheral neuropathy, type 2 diabetes, meta-analysis, systematic review, diabetic neuropathy, cobalamin, drug safety, B12 monitoring, endocrinology, myelin

The review, led by Daniela Debone and colleagues and prospectively registered with PROSPERO, searched PubMed, Embase, and Cochrane CENTRAL from database inception to July 2025, with an update in July 2026, imposing no language or date restrictions. The team screened 4,543 records, removed 1,499 duplicates, and ultimately included 13 studies: seven randomized controlled trials, three observational cohorts, two conference abstracts, and one research letter. The studies came from Japan, Israel, Turkey, the United States, Switzerland, Brazil, Spain, India, Austria, and New Zealand, with sample sizes ranging from just 16 participants to 4,281. Because the studies varied so widely in design, dosing, populations, and outcome reporting, the authors performed a descriptive synthesis rather than a statistical meta-analysis.

The physiological stakes are considerable. Vitamin B12, or cobalamin, is indispensable for DNA synthesis, methylation reactions, and the formation of myelin, the insulating sheath around nerve fibers. It also serves as a cofactor in mitochondrial energy metabolism and in the maturation of blood cells. Because the human body cannot synthesize it, cobalamin must come from animal-derived foods or fortified products. When deficiency sets in, the consequences can include megaloblastic anemia, peripheral neuropathy, cognitive impairment, and elevated cardiovascular risk. Populations at particular risk include children, vegetarians and vegans, people with pernicious anemia, patients on long-term metformin, and those who have lost part or all of their stomach.

Intramuscular injection has traditionally been favored because it bypasses the gastrointestinal tract entirely, guaranteeing systemic delivery even when gut absorption is impaired. Yet injections carry real costs: pain, the need for a healthcare professional, and repeated clinic visits. Oral supplementation has long faced skepticism, particularly in pernicious anemia, where the absence of intrinsic factor cripples active absorption. The counterargument rests on passive diffusion: roughly 1% of orally ingested cobalamin can be absorbed independent of intrinsic factor, which is why high-dose oral regimens of 1 to 2 milligrams per day can work even in malabsorptive conditions. Sublingual formulations, meanwhile, are absorbed through the highly vascularized oral mucosa, entering systemic circulation directly and avoiding hepatic first-pass metabolism.

The biochemical evidence assembled in the review is strikingly consistent in one direction: all three routes work. In the landmark trial by Kuzminski and colleagues, patients receiving 2 milligrams of oral cyanocobalamin daily actually achieved higher serum levels after four months (1,005 pg/mL) than those receiving monthly 1-milligram injections (325 pg/mL), a statistically significant difference. In contrast, the Swiss trial by Metaxas and colleagues found the opposite pattern over 28 days, with 100% of intramuscular patients normalizing their levels compared with 84.2% of oral patients, and mean concentrations of 2,796 pmol/L versus 354 pmol/L. The Spanish OB12 non-inferiority trial in older adults found essentially equivalent normalization rates at eight weeks, with a slight per-protocol edge for injections at 52 weeks that did not hold up under intention-to-treat analysis.

The sublingual route, though underrepresented with only five studies, produced some of the review’s most intriguing findings. In a large Israeli retrospective cohort of 4,281 patients, sublingual tablets produced significantly larger mean increases in serum B12 than injections (252 versus 218 ng/L) and higher odds of achieving increased levels. In a New Zealand trial of metformin-treated patients with type 2 diabetes, sublingual methylcobalamin outperformed a single hydroxocobalamin injection at three months, though the advantage vanished by six months. Head-to-head comparisons of sublingual versus oral supplementation, in Israel and India, found no meaningful differences: in the Indian trial, normalization rates were 68.4% for sublingual and 66.7% for oral after three months of 1,500 micrograms daily, with compliance above 90% in both arms.

Yet the review’s authors are careful to sound a note of caution that cuts through the numbers: higher serum B12 does not automatically mean better clinical outcomes. The magnitude of a biochemical response does not necessarily translate into greater symptom relief, neurological recovery, or quality-of-life improvement. Indeed, no study in the review demonstrated superiority of any route for long-term clinical outcomes, largely because clinical endpoints were rarely measured at all. The certainty of the evidence was further constrained by open-label designs, unadjusted confounding in the observational studies, which were judged to be at serious risk of bias, and inconsistent outcome reporting, with some trials lacking formal statistical comparisons altogether.

What the evidence does support is a shift in how route selection might be framed. When biochemical responses are broadly similar, the decision can reasonably be guided by patient preference, tolerability, and access rather than presumed efficacy differences. Patient preference data from the review are telling: in the Spanish trial, 83.4% of participants preferred oral administration, including more than two-thirds of those actually receiving injections. In the Swiss trial, 45.9% initially favored oral treatment versus 21.6% for injections, though preferences shifted after experiencing treatment. Non-invasive options also reduce reliance on healthcare infrastructure, a meaningful advantage for adherence and for resource-limited settings where repeated clinic visits impose real burdens on patients and systems alike.

The review also refines, rather than overturns, a 2024 network meta-analysis that had ranked intramuscular administration highest for biochemical efficacy while finding no statistically or clinically significant differences between routes. The new synthesis improves on that work in several ways: it was prospectively registered, searched four years more recently, applied no language restriction, focused exclusively on adults rather than pooling pediatric and adult populations, and incorporated conference abstracts and a research letter that the earlier review excluded. The authors note the review was funded by a pharmaceutical manufacturer and was undertaken to establish a shared evidence base for a subsequent Delphi consensus on B12 replacement strategies.

The gaps that remain are as instructive as the findings. Safety data were sparse across all routes, and none of the included studies reported adverse events for sublingual supplementation. Follow-up periods were mostly short to medium term, ranging from four weeks to 24 months. Future research, the authors argue, should include adequately powered, long-term randomized trials using standardized, patient-centered outcomes such as symptom improvement, adherence, and preference, particularly for the sublingual route. Until such trials arrive, the practical message for patients and clinicians is quietly radical: the shot in the arm that has defined B12 replacement for generations may be no better than a tablet dissolved under the tongue, and the choice may ultimately belong to the patient rather than the syringe.

Subject of Research: Comparative efficacy of oral, sublingual, and intramuscular vitamin B12 supplementation routes in adults

Article Title: Comparative efficacy of vitamin B12 supplementation routes in adults with replacement indication: a systematic review

Article References: Debone, D., Azeredo da Silva, A. F., Cordenonssi, J. T., & Cohen, R. V. (2026). Comparative efficacy of vitamin B12 supplementation routes in adults with replacement indication: a systematic review. eClinicalMedicine, 100, Article 104214. https://doi.org/10.1016/j.eclinm.2026.104214

Image Credits: AI Generated

DOI: 10.1016/j.eclinm.2026.104214

Keywords: vitamin B12, cobalamin, sublingual supplementation, intramuscular injection, oral supplementation, systematic review, vitamin B12 deficiency, pernicious anemia, metformin, patient preference, eClinicalMedicine, biochemical outcomes