LAUSR

To investigate if dairy, meat, and fish intakes associate with dementia and cognitive performance. We included 2497 dementia-free men from Eastern Finland, aged 42–60 years in 1984–1989 at the baseline examinations. Data on cognitive tests [Mini Mental State Exam (MMSE), trail making test (TMT), verbal fluency test (VFL), selective reminding test (SRT), and Russell’s adaptation of the visual reproduction test (VRT)] at the 4-year re-examinations were available for 482 men and on the ApoE phenotype for 1259 men. Data on dementia events were obtained by linkage to national health registers. Diet was assessed with baseline 4-day food records. Cox regression and analysis of covariance were used for analyses. During a mean 22-year follow-up, 337 men had a dementia diagnosis. Among the foods, only cheese intake associated with dementia risk (hazard ratio in the highest vs. the lowest quartile = 0.72, 95% confidence interval = 0.52–0.99, P-trend = 0.05). In the cognitive tests, higher non-fermented dairy and milk intakes associated with worse verbal fluency (VFT). Higher processed red meat intake associated with worse verbal (SRT) and visual memory (VRT), whereas higher unprocessed red meat intake associated with better general cognitive functioning (MMSE) and processing speed and executive functioning (TMT). Higher fish intake associated with better verbal memory (SRT). Among APOE-ε4 carriers, especially non-fermented dairy intake associated with higher risk of dementia outcomes, and higher fish intake indicated better cognitive performance. Although higher intake of some food groups associated with cognitive performance, we found little evidence for associations with dementia risk.