High Vitamin B12 Levels and In‐hospital Mortality

To the Editor: The relationship between vitamin B12 deficiency and hematological and neurological morbidity has been established,1 but the effects of vitamin B12 deficiency as a marker of cardiovascular risk is unclear (although hyperhomocysteinanemia is an established risk factor for cardiovascular disease).2 The association between high vitamin B12 levels and mortality has been less studied. Some articles have described an association between high cobalamin levels and intrahospital and 90-day postdischarge mortality3,4 in patients aged 65 and older. Recently, this association was confirmed in patients without cancer, chronic liver disease, gastric surgery, or small bowel disease.5 Nevertheless, these reports do not clearly describe the causes of this association, which could be biased by uncontrolled risk factors for mortality such as functional status, comorbidity, and admission medical diagnosis. The aim of the current study was to analyze the association between vitamin B12 level and intrahospital mortality in all patients aged 65 and older admitted during 2009 for acute medical disorders to the Acute Geriatric Unit at Red Cross Central Hospital, Madrid, Spain, excluding those with a diagnosis of neoplasia (International Classification of Diseases, Ninth Revision, (ICD-9) codes 140–239) and liver disease (ICD-9 code 573.8). Laboratory screening of blood vitamin B12 levels was routinely performed the day after admission, using the electrochemiluminescent immunoassay system technique (Cobas, Roche Mannheim, Germany). Demographic (age and sex), clinical (admission diagnosis grouped according to diagnosis-related groups (DRGs) and Charlson comorbidity Index6), and laboratory (serum albumin and creatinine) data were obtained from medical records on admission. Previous functional (Barthel Index)7 and cognitive (Red Cross Mental Scale) status8 and functional impairment on admission (previous Barthel Index – Barthel index on admission) were recorded. Vitamin B12 levels were categorized in quartiles, and risk of intrahospital mortality was evaluated using a logistic regression analysis (using odds ratios and 95% confidence intervals) including all baseline variables at admission. After excluding 147 patients with neoplasia and 16 with liver disease, 1,160 patients aged 86.8±6.7 were analyzed (67.5% female) with an average length of stay of 10.3±7.9 days and a hospital mortality rate of 10.1%. Serum vitamin B12 levels were measured in 1,136 patients; median serum level was 337 pmol/L (interquartile range 243–481 pmol/L). Mortality was 7.7%, 6.0%, 7.4%, and 16.5% for the first, second, third, and fourth (vitamin B12≥481 pmol/L) quartiles, respectively. The greater risk of death associated with the fourth quartile of vitamin B12 was also found with the multivariant analysis (Table 1). Similarly to the previous reports, the current study found that higher vitamin B12 levels were independently associated with greater risk of death during hospitalization in older patients with acute medical disorders after adjustment for other known risk factors for mortality such as age, comorbidity, previous functional and cognitive status, functional impairment on admission, and other biomarkers such as serum albumin and creatinine. Regardless of the consistency of this relationship, the causes and physiopathology mechanisms of the association between high levels of vitamin B12 and intrahospital mortality are not clear. Further research is needed to answer questions. For instance, what is the vitamin B12 cutoff value that predicts risk of death? (≥481 pmol/L in our study vs >351 pmol/L in a previous study5) Is this risk related to lower levels of homocysteine, or is this an independent association, as previous reports have described?9 Because other studies have found a greater risk of cancer and death in patients supplemented with vitamin B12 and folic acid,10 what are the optimal serum levels of vitamin B12? Author Contributions: All authors contributed to study concept and design, acquisition of subjects and/or data, analysis and interpretation of data, and preparation of manuscript. Sponsor's Role: No sponsor's role in the design, methods, subject recruitment, data collections, analysis, and preparation of paper.