Association of the Mediterranean diet with cognition in a Western population

Abstract

Background The Mediterranean‐style diet (MD) is promoted as one of the healthiest dietary patterns and has been shown to reduce cognitive decline in some randomised controlled trials (RCTs). This pattern is plant‐based with a focus on vegetables, legumes, fish, and olive oil, with low intake of red meats and foods high in sugar and saturated fats. The Maintain Your Brain (MYB) study is an online multi‐domain RCT targeting modifiable risk factors for dementia, including diet. This study investigated cross‐sectional associations between adherence to a MD pattern and cognition. Methods Participants enrolled in MYB were recruited from the Sax Institute’s larger 45 and Up Study, and included those with all baseline MYB assessments (n=6236). Diet was assessed using the validated Mediterranean Diet and Culinary Index tool. This 50‐item tool assesses intake of nine desirable and four undesirable features of the MD diet and allows derivation of the Mediterranean Diet Adherence Screener (MEDAS). Computerised cognitive tests, administered through the MYB digital platform, were used to assess domains of executive function, complex attention, learning and memory and global cognition. Results Participants were 46% male, with mean (SD) age 65.0 (5.8) years, BMI 26.6 (4.9) kg/m2 and well educated (46% with tertiary education). Higher MEDAS scores were associated with being female, younger, better educated and having a lower BMI (all p<0.001). Overall participant adherence to a MD pattern was low; MEDAS score 6.1 (2.1)/14. After covariate adjustment and correction for multiple analyses in hierarchical linear regression models, better MEDAS scores were significantly associated with worse (lower) z‐scores for executive function (β ‐0.018, p=0.003). For every 1‐point higher MEDAS score, executive function z‐scores were 0.018 lower. This unexpected negative association was clinically very small. There were no significant associations with other cognitive domains. Conclusion Adherence to a MD was sub‐optimal in this well‐educated Australian sample. The one unexpected finding is likely not clinically meaningful given its size. Additionally, as this was a cross‐sectional study, reverse causality cannot be ruled out. Results from the main MYB RCT will provide data on MD acceptance and impact in a Western dietary environment.