Vitamin B ₁₂ as a Predictor of Mortality in Elderly Patients

To the Editor: We read with great interest the recently published article by Salles et al.1 concerning high vitamin B12 levels as a predictor of long-term mortality. We would like to report similar results in a study that focuses not on long-term but intrahospital mortality. These were obtained from a larger project about mortality predictors of hospitalized nonagenarians that were related to a control group aged 65 to 75. The setting was the internal medicine department of a third-level hospital; data were obtained between July 1999 and December 2004. Vitamin B12 was determined within the first 24 hours of admission to the hospital using the luminescent system technique (Modular Analitic, Roche, Barcelona, Spain) and related to the intrahospital mortality. The Mann-Whitney U test was used to compare the cobalamin values of those who survived with those who died. Vitamin B12 levels were available for 366 patients: 187 nonagenarians and 179 controls aged 65 to 75 (132 men and 234 women). Forty patients died: 31 nonagenarians and nine controls. Median serum vitamin B12 levels of the patients who died were significantly higher than those of the patients who survived (696.5 pg/mL vs 442.1 pg/mL (interquartile range=434.8–1,298.3 and 299.8–778.2, respectively, P=.001). Data related to age groups are shown in Table 1. If the 68 patients who had a neoplasm were excluded, a statistically significant difference was still observed, with cobalamin levels higher in those who died (total number of patients, as well as the subgroups). Albumin and creatinine levels were also associated with mortality. A multiple logistic regression model was used that included albumin, creatinine, and vitamin B12 levels to evaluate the independent prognosis value of vitamin B12. All three (high cobalamin levels (P=.02), creatinine levels (P=.02) and low albumin levels (P<.001)) were associated with mortality. Although we agree with Salles et al. that high vitamin B12 levels might be a marker to assess high risk of mortality, we would also point out that other factors (e.g., diseases that generate liver damage) might influence cobalamin levels.2 Although another study3 also found that elevated creatinine and low serum albumin levels could influence their values, we found that the prognostic value of cobalamin was not dependent on albumin and creatinine levels. In our series, it definitely was associated with intrahospital mortality. Cobalamin has been reported to be an acute phase reactant, which could explain that there are probably confounding factors that might alter the results presented by Salles et al. as well as ours. We therefore think that further research is needed do define the meaning and the implications of high cobalamin levels in the geriatric population. Conflict of interests and funding sources: None. This manuscript had no financial support. Author Contributions: Dr. Marion Hemmersbach-Miller, Dr. Alicia Conde-Martel, and Dr. Pedro Betancor-León have full authorship criteria in preparation of this manuscript.