High B12 as a marker of underlying disease: analysis of observational data

In recent years, a disturbing idea has taken hold in medicine: “high blood vitamin B12 predicts premature death.” A recent systematic review and meta-analysis in Nutrients tested this hypothesis in large observational data sets and came to a more conservative conclusion: Overall, hypervitaminosis B12 did not reliably increase the risk of all-cause mortality in adults. There were hints of risk in specific subgroups (patients with chronic diseases and hospitalizations), but the statistical strength of these signals was diminished under rigorous testing.

Background of the study

Vitamin B12 is a water-soluble cofactor of methionine synthase and methylmalonyl-CoA mutase; its deficiency leads to megaloblastic anemia and neuropathy. In routine practice, we almost always measure total serum B12, whereas the “biologically active” fraction is holo-transcobalamin, and functional status is more accurately reflected by methylmalonic acid (MMA) and homocysteine. However, vitamin B12 does not have a clearly established upper tolerable intake level and classical toxicity, so unexpectedly high values in the blood are often interpreted as a marker of concomitant pathology, and not as an “overdose” of the active vitamin.

Where to get "high B12" without taking megadoses:

• liver disease (release of stores and decreased clearance),
• myeloproliferative and some solid tumors (growth of binding proteins - transcobalamin I/III),
• renal failure (accumulation),
• systemic inflammation/infection (acute phase changes in transport proteins),
• less often - massive supplementation or analyzer artifacts.

Against this background, observational studies have emerged in recent years where elevated B12 was associated with higher overall mortality. These signals quickly made their way into clinical reports and the media, giving rise to the thesis that “too much B12 is dangerous.” But such studies have a vulnerability: they do a poor job of separating cause from effect. High B12 in a seriously ill patient may be an epiphenomenon of the disease (liver, cancer, inflammation), which determines the risk of death, rather than an independent “toxic level” of the vitamin.

This is where the research request arose: a rigorous summary of prospective data was needed, analyzing different populations (general, hospital, patients with chronic diseases), taking into account the heterogeneity of the thresholds for “high” B12, and checking the stability of the relationship after statistical adjustments. The systematic review and meta-analysis to which your news item relates is precisely testing whether there is an independent prognostic value of hyperB12 for overall mortality, or whether it is mainly a proxy marker of underlying pathology, requiring clinical clarification of the causes.

What exactly was studied?

• Type of work: systematic review + meta-analysis of longitudinal observational studies with mortality registration.
• Search: PubMed, Scopus, Web of Science, Google Scholar, ProQuest - until June 30, 2024; protocol registered in PROSPERO (CRD42022361655).
• Size: 28 studies, 69,610 participants and 15,815 deaths; follow-up length ranged from approximately 0.9 to 132 months.
• Methods: frequentist and Bayesian approach, subanalyses by disease type and by setting (hospital/general population), meta-regressions, screening of “small studies”, network meta-analysis with ranking of B12 levels.

The results show that there is no strong overall signal in the pooled analysis. In the frequency models, there were “marginal” increases in risk in patients with chronic diseases (RR≈1.40) and in hospitalized patients (RR≈1.57), but in the meta-regression, these effects statistically “spread out”. Bayesian network analysis supported the direction of risk for these groups, but was limited by the number of eligible studies. The authors’ overall conclusion is that hypervitaminosis B12 per se has not been proven to be a reliable predictor of overall mortality.

Why is there so much fuss around B12?

• In a number of observational studies, high B12 has been found in severely ill patients and associated with poor outcomes - hence the temptation to "blame" the vitamin itself.
• But B12 is part of a complex transport-metabolic network; its elevation may reflect liver disease, kidney disease, cancer, inflammation, or laboratory/classification features - and not an excess of the active form of the vitamin.
• The overall mortality outcome is often dominated by causes unrelated to B12, blurring causality.

Key figures and findings of the meta-analysis

• Included: 28 studies / 69,610 participants / 15,815 deaths.
• Risk signals: chronic diseases (RR≈1.40, 95% CI 1.05-1.85) and hospital sample (RR≈1.57, 95% CI 1.19-2.07) - but without stability in meta-regression.
• Network analysis (Bayesian): Risk distribution across B12 'thirds' differed between groups, but conclusions are limited by lack of direct comparisons and heterogeneity.
• Conclusion: No convincing increase in overall mortality was shown with high B12 in adults.

The authors specifically highlight methodological pitfalls: inconsistency in thresholds for “high” B12, single measurements, different analysis platforms, population heterogeneity, and confusion between causality and prognosis studies. All of this can generate “artifacts” when high B12 is only a marker of a severe background, and not an independent risk factor.

What does this mean for practice today?

• Don't panic over one high B12. Repeat the test, clarify the method and units, evaluate the clinical picture and context.
• Look for the cause of secondary increase. Check liver/kidney functions, inflammatory markers, drug history; if indicated - oncosearch.
• Don't confuse prognosis with etiology. High B12 may be an indicator of a condition, not an active "poisonous" substance.

What should science do next?

• Standardize B12 thresholds and assays (including forms/carriers of the vitamin).
• Conduct well-powered, well-designed prospective studies that test B12 as a component of a prognostic model rather than an abstract “risk factor.”
• Use model validation and avoid mixing causal and predictive questions.

Summary

High B12 is a reason to investigate, not a ready-made sentence. The connection with general mortality in the general population appears weak and unstable; it is more logical for clinicians to interpret hypervitaminosis B12 as a diagnostic clue to search for background pathology, and not as an independent “black marker”.

Source: Valdez-Martínez E., Márquez-González H., Ramírez-Aldana R., Bedolla M. The Controversial Issue of Hypervitaminosis B12 as a Prognostic Factor of Mortality: Global Lessons from a Systematic Review and Meta-Analysis. Nutrients. 2025;17(13):2184. https://doi.org/10.3390/nu17132184