Strength of the Association of Elevated Vitamin B12 and Solid Cancers: An Adjusted Case-Control Study

Abstract

The association between elevated plasma vitamin B12 (B12) level and solid cancers has been documented by two national registries. However, their design did not allow for the adjustment for other conditions associated with elevated B12. The objectives of this study were to confirm this association after the adjustment for all causes of elevated B12, and to study the variations according to the increasing B12 level, the type of cancers, and the presence of metastases. We compared 785 patients with B12 ≥ 1000 ng/L with 785 controls matched for sex and age with B12 < 1000 ng/L. Analyses were adjusted for the causes of elevated B12: myeloid blood malignancies, acute or chronic liver diseases, chronic kidney failure, autoimmune or inflammatory diseases, and excessive B12 supplementation. A B12 ≥ 1000 ng/L was associated with the presence of solid cancer without metastases (OR 1.96 [95%CI: 1.18 to 3.25]) and with metastases (OR 4.21 [95%CI: 2.67 to 6.64]) after adjustment for all elevated B12-related causes. The strength of the association rose with the increasing B12 level, in particular in cases of metastases. No association between liver cancers and elevated B12 level was found after adjustment for chronic liver diseases. In conclusion, unexplained elevated B12 levels should be examined as a possible marker of solid cancer.

Keywords: biomarkers; case-control study; neoplasm metastasis; neoplasms; vitamin B12.

Figure 1

Strength of the association between solid cancers with and without metastases and elevated vitamin B12 levels. Notes: Graphics present the adjusted OR illustrating the association between solid cancer and elevated vitamin B12 levels for all values of vitamin B12 between 300 and 2000 ng/L. For each tested threshold, the OR represented the association between elevated B12 and solid cancers by comparing patients with B12 concentration above this threshold with those having B12 concentration below this threshold (and not with a single reference group of patients with normal B12). These two analyses are independent. For each analysis, cancer without metastases in the left panel and with metastasis in the right panel was defined as a dichotomous variable (presence versus absence). The odds ratios are adjusted for sex, age, chronic and acute liver diseases, severe chronic kidney failure, autoimmune and inflammatory diseases, excessive vitamin B12 supplementation, myeloid, and lymphoid blood malignancies.