Adipose tissue fatty acid composition and cognitive impairment

Objective: The aim of this study was to examine the association among adipose tissue eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), arachidonic acid (AA), and the ratios of EPA to AA and DHA to AA with impaired cognitive function. Methods: This cross-sectional analysis comprised 481 men participating in the Cohort of Swedish Men Clinical and for whom adipose tissue fatty acid composition and results from a telephone-based cognitive test were available. Impaired cognitive function was defined using a predefined cutoff on the cognitive test. Binomial log-linear regression models were used to estimate prevalence ratios. In secondary analyses, Cox proportional hazards models were used to estimate relative risk for incident dementia ascertained by linkage with population-based registers. Results: We observed a graded reduction in the prevalence of impaired cognitive function across tertiles of adipose tissue EPA/AA- ratio (P-trend = 0.01); compared with the lowest tertile, the multivariable-adjusted prevalence ratios were, respectively, 0.89 (95% confidence interval [CI], 0.67-1.17) and 0.64 (95% CI, 0.45-0.91) for the second and third tertiles. EPA, DHA, and the DHA/AA ratio showed similar patterns of association; however, the CIs included the null. AA alone was not associated with impaired cognitive function. Although with lower precision, estimates obtained from the prospective analysis were broadly consistent with the main analysis. Conclusions: Findings from this study suggest that a high ratio of EPA to AA in adipose tissue may be associated with better cognitive function. A similar association was observed with EPA, DHA, and the ratio of DHA to AA, but the results did not exclude a null association. (C) 2018 Elsevier Inc. All rights reserved.