Serum vitamin B12 levels and 90-day outcomes in hospitalized patients with dementia: A cohort study

Elevated serum vitamin B12 levels have emerged as a paradoxical prognostic marker associated with increased mortality across various clinical settings, yet their significance in hospitalized dementia patients remains unexplored. Given the high vulnerability of this population, we aimed to investigate the association between elevated serum vitamin B12 levels and all-cause mortality. This retrospective cohort study evaluated 90-day outcomes in hospitalized adults with dementia who had serum vitamin B12 levels measured within 7 days of admission (2010-2024). Patients were categorized into exposure (vitamin B12 ≥ 900 pg/mL) and control (300-900 pg/mL) groups. Propensity score matching balanced baseline characteristics, including demographics, comorbidities, and laboratory parameters. The primary endpoint was all-cause mortality within 90 days. Secondary endpoints comprised sepsis, pneumonia, urinary tract infection, and admission to the intensive care unit (ICU). Sensitivity analyses restricted the cohort to patients with unspecified dementia and to patients with a dementia diagnosis established at least 1 year before the index admission. After propensity score matching, 16,513 patients were included in each cohort. Elevated vitamin B12 was significantly associated with increased 90-day mortality (10.9% vs 8.3%; odds ratio [OR] 1.36, 95% confidence interval [CI] 1.26-1.46, P < .001), sepsis (4.2% vs 3.0%; OR 1.43, 95% CI 1.27-1.61, P < .001), pneumonia (8.3% vs 6.6%; OR 1.28, 95% CI 1.18-1.39, P < .001), and ICU admission (3.9% vs 3.0%; OR 1.31, 95% CI 1.16-1.47, P < .001). No association was observed with urinary tract infection (OR 0.97, 95% CI 0.90-1.04, P = .377). The findings remained robust across sensitivity and subgroup analyses. Elevated vitamin B12 levels are associated with increased short-term mortality, infectious complications, and ICU admission in hospitalized patients with dementia. This readily available biomarker may serve as a potential indicator for risk stratification in this vulnerable population. However, given the observational nature of this study, future prospective studies are warranted to confirm these findings and elucidate the underlying mechanisms.