Sala-Vila · Nutrients 2022 · prospective cohort study · n=1490

Red Blood Cell DHA Is Inversely Associated with Risk of Incident Alzheimer's Disease and All-Cause Dementia: Framingham Offspring Study.

Level 3 - non-randomized controlled study

Prospective cohort study

PubMed 35745137 · doi:10.3390/nu14122408 · record verified 2026-08-26

What was done

This prospective observational cohort study evaluated 1,490 dementia-free participants aged 65 years or older from the Framingham Offspring Cohort. Baseline red blood cell (RBC) docosahexaenoic acid (DHA) levels were measured as an objective biomarker of long-term DHA intake. Participants were followed for incident Alzheimer's disease (AD) over a median follow-up of 7.2 years, assessing AD risk across quintiles of RBC DHA and testing for interaction with APOE-ε4 carrier status.

What was found

During follow-up, 131 incident AD cases occurred. Participants in the highest RBC DHA quintile (Q5) had a 49% lower risk of incident AD compared to the lowest quintile (Q1) in fully adjusted models (Hazard Ratio [HR]: 0.51, 95% CI: 0.27 to 0.96), which was estimated to provide 4.7 additional years free of AD. A DHA × APOE-ε4 interaction was observed: each standard deviation increase in RBC DHA showed borderline statistical significance for lower AD risk in APOE-ε4 carriers (HR: 0.71, 95% CI: 0.51 to 1.00, p = 0.053), but not in non-carriers (HR: 0.85, 95% CI: 0.65 to 1.11, p = 0.240).

Why it matters

This study provides prospective human evidence that higher objective biomarker levels of DHA are associated with significantly reduced Alzheimer's risk, potentially offering greater benefit to genetically susceptible APOE-ε4 carriers.

Limits

The study is observational and cannot demonstrate causality. Absolute incident case numbers were relatively small (131 cases), and the subgroup association in APOE-ε4 carriers achieved only borderline significance (p = 0.053). Potential residual confounding cannot be ruled out, and detailed cohort demographics were not reported in the abstract.