Abstract
Dietary omega-3 fatty acids are promising nutrients in dementia. Several prospective cohort studies have examined the relationships between circulating omega-3 (an objective biomarker of dietary intake) and incident dementia, the largest to date being a report from the UK Biobank (n = 102,722). Given the recent release of new metabolomics data from baseline samples from the UK Biobank, we re-examined the association in a much larger sample (n = 267,312) and also focused on associations with total omega-3, docosahexaenoic acid (DHA), and non-DHA omega-3. Using Cox regression models, we observed that the total omega-3 status was inversely related to the risk of Alzheimer's (Q5 vs. Q1, hazard ratio [95% confidence interval] = 0.87 [0.76; 1.00]) and all-cause dementia (Q5 vs. Q1, 0.79 [0.72; 0.87]). The strongest associations were observed for total omega-3 (and non-DHA omega-3) and all-cause dementia. In prespecified strata, we found stronger associations in men, and in those aged ≥60 years at baseline (vs. those aged 50-59). Thus, in the largest study to date on this topic, we confirmed the favorable relationships between DHA and risk for dementia, and we also found evidence that non-DHA omega-3 may be beneficial. Finally, we have better defined the populations most likely to benefit from omega-3-based interventions.</p>