Association of red cell omega-3 fatty acids, particularly EPA, with cognitive performance in atrial fibrillation

Abstract

Abstract Introduction Atrial fibrillation (AF) is associated with cognitive impairment and dementia. Whether or not omega-3 Fatty Acids (n-3 FAs) have neuroprotective effects is controversial. We therefore aimed to determine the correlation of n-3 FAs with cognitive function in a large cohort of AF patients. Methods Red cell n-3 FAs levels were determined by gas chromatography and mass spectrometry in 2359 patients from the prospective ongoing multicenter Swiss Atrial Fibrillation study (Swiss-AF). To assess the overall cognitive performance of patients, we used several validated neurocognitive assessment tools, including the Montreal Cognitive Assessment (MoCA), Semantic Fluency Tests (SF), Trail Making Test (TMT A&B), and the Digit Symbol Substitution Test (DST). We also used a summary score that we have previously developed (the cognitive construct (CoCo) score). We performed mixed-effect linear regression analyses. Two adjustment models were applied; model 1 (adjusted for age, sex, and educational level) and model 2 (adjusted for age, sex, educational level, history of stroke/TIA, AF type, heart rate, arterial hypertension, diabetes, coronary artery disease, BMI, chronic kidney disease, smoking status, alcohol consumption, geriatric depression scale, use of oral anticoagulation and antiplatelet medication). We also included a random intercept to control for center-effect in both models. Results The mean age of our study population was 73 years (standard deviation [SD] 8.4), and 662 patients were females (28%), 318 had a history of stroke (13%), and 716 suffered from coronary artery disease (30%). We observed statistically significant associations of total n-3 FAs levels (EPA + DHA + DPA + ALA) with all neurocognitive assessments, as seen in figure a. Individual n-3 FAs levels (EPA and ALA) were positively associated with MoCA score in both models. EPA was significantly associated with multiple neurocognitive assessments including MoCA, SF, TMT-B, DST, and overall CoCo score in both models (figure b). We found that a 1% red cell EPA level increase was associated with 0.19 (Model 1, 95%CI 0.06-0.31, p=0.003) and 0.14 (Model 2, 95% CI 0.01-0.26, p=0.02) units increase in the standardized MoCA score. Conclusions Red cell n-3 FAs levels are associated with better cognitive performance in AF patients. We observed a particularly strong association of EPA levels with all neurocognitive assessments.