The MIND diet's brain link may depend on your genes.

Cohort study. 745 people. diet assessed once at baseline (first FFQ, 2004-2013); 2 to 23 annual cognitive assessments per person (77% had 5 or more); mean 9.1 years from diet questionnaire to death in the autopsy subset.

People were watched for years. Nobody was told what to do. A clue, not proof.

Who: Older adults in the Chicago area, dementia-free at enrolment, mean age 80.0 (SD 6.7), 74% women; genetic data only from participants self-reporting as non-Hispanic white; Met/Met n = 198, Met/Val n = 383, Val/Val n = 164

• Everyone togetherno clearAlzheimer's risk
• Met/Met gene type-67%best vs worst diet
• Other 2 gene typesno clearAlzheimer's risk

More to know, from our read of the paper:

• Observational: linked, not proven. Diet self-reported once by food-frequency questionnaire; authors note residual confounding and dietary measurement error
• Met/Val and Val/Val: no clear link (HR 0.94 and 0.71, CIs cross 1). Say 'no clear link', never 'the diet doesn't work for them'
• All participants with genetic data were white, older (mean 80) Chicago volunteers who agreed to brain donation; not representative (authors' own limitation)
• Domains differ from the MIND RCT (perceptual speed there; global, episodic, visuospatial here); no COMT x MIND interaction for Alzheimer's brain pathology at autopsy; protein (plasma/brain COMT) findings are exploratory
• Background: the 3-year MIND RCT (Barnes 2023 NEJM) found no significant difference vs the control diet; this paper suggests genotype might explain mixed findings, but does not prove it

Researchers in Chicago followed 745 older adults (average age 80) for years, scoring their diet and testing their thinking every year. Overall, eating closest to the MIND diet was linked to slower decline, but not to a clearly lower dementia risk. In the roughly 1 in 4 with the Met/Met version of the COMT gene, the top third of MIND eaters had 67% lower Alzheimer's risk than the bottom third. In the other two gene types, there was no clear link. Everyone with gene data was white, and this is observational: linked, not proven.

We read the abstract and the full text of this paper and checked every number on this page against it (17 facts, checked on 2026-10-29). The study design is named above. Cohort and other observational studies are described as linked, not proven. Found an error? Email theis@onestudyaday.com.