Abstract
IMPORTANCE: Anemia has been associated with increased dementia risk, but its relationship with Alzheimer disease (AD) blood biomarkers remains unclear. OBJECTIVE: To investigate whether there is a cross-sectional association between hemoglobin and AD blood biomarker levels and a longitudinal association of hemoglobin and AD biomarkers with incident dementia. DESIGN, SETTING, AND PARTICIPANTS: This cohort study used data from the Swedish National Study on Aging and Care in Kungsholmen (SNAC-K), a longitudinal population-based study. Randomly selected adults aged 60 years or older were enrolled at baseline (March 21, 2001, to August 30, 2004) and followed up every 3 or 6 years according to age through December 31, 2019. Data analysis was conducted between September 1, 2024, and January 7, 2026. EXPOSURES: Hemoglobin level measured at baseline. Anemia was defined according to World Health Organization criteria. MAIN OUTCOMES AND MEASURES: Incident dementia, diagnosed according to Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria, and serum concentrations of phosphorylated tau 217 (p-tau217), neurofilament light chain (NfL), and glial fibrillary acidic protein (GFAP), measured using Simoa assays. Cox proportional hazards regression and quantile regression models were used to examine the association of hemoglobin with dementia risk and AD blood biomarkers. The joint association of hemoglobin and AD blood biomarkers with dementia development using Cox proportional hazards regression was also explored. RESULTS: A total of 3363 adults were enrolled in SNAC-K (73.3% participation rate). After excluding those with dementia or missing AD biomarkers or hemoglobin measures, 2282 dementia-free participants were included in the current study (median age, 72.2 [IQR, 60.8-81.1] years; 1406 [61.6%] female). During a mean (SD) follow-up of 9.3 (4.3) years, 362 participants (15.9%) developed dementia. Compared with individuals with a normal hemoglobin level, those with anemia had higher baseline levels of p-tau217 (β, 0.22; 95% CI, 0.15-0.30), NfL (β, 0.25; 95% CI, 0.19-0.31), and GFAP (β, 0.08; 95% CI, 0.03-0.12) and showed a higher risk of developing dementia (hazard ratio [HR], 1.66; 95% CI, 1.21-2.28) during follow-up. Participants with both anemia and high p-tau217, NfL, or GFAP had the highest hazard of dementia (eg, adjusted HR of 3.64 [95% CI, 2.39-5.56] among those with anemia and high NfL). CONCLUSIONS AND RELEVANCE: In this cohort study of dementia-free older adults, anemia was associated cross-sectionally with higher levels of AD blood biomarkers and longitudinally with increased dementia risk. The highest dementia risk occurred when low hemoglobin and elevated AD biomarkers coexisted, suggesting a potential interplay between anemia and neuropathology in dementia development.