MIND diet trial brain MRI (AGMP_MIND)

The MIND trial was a three-year randomized controlled trial run at Rush University Medical Center in Chicago and at Harvard in Boston. It tested whether the MIND diet, a blend of the Mediterranean and DASH diets aimed at brain health, slows cognitive decline compared with a control diet; both groups followed mild caloric restriction. A subset of the participants had brain MRI before and after the intervention. On the AD Knowledge Portal the trial appears as the AGMP_MIND study, which shares serum metabolomics but no images. This entry records what the publications say about the MRI.

Composition

604 people were enrolled, 301 in the MIND diet group and 303 in the control group. Participants were 65 years or older, cognitively unimpaired (Montreal Cognitive Assessment of at least 22), had a body-mass index of 25 or more, a first-degree relative with dementia and a suboptimal diet by a 14-item MIND diet score. Mean age was 70.4 years in both groups and about 35% were men. MRI was offered to all participants at first and later to a nonrandom subgroup, aiming for 300 scanned people. 267 had a baseline scan and 201 of them a second scan at year 3.

Acquisition

Scans were acquired only at baseline and at month 36. The Chicago site used a 3 T Philips Achieva and the Boston site a Siemens Skyra. The sequence protocol is described in the supplementary appendix of the primary paper, which was not reachable for this entry, so no contrasts are listed.

Annotations

The trial derived total gray and white matter volume, hippocampal volume and white matter hyperintensity volume, each normalized to intracranial volume. One baseline scan was excluded for poor quality, leaving 200 people in the imaging analysis. Changes over three years did not differ between the diet groups.

Known limitations

• No license or data use terms cover the images. Trial papers say limited data are available on an approved request, without naming the MRI.
• The MRI subgroup was not randomly selected from the trial population.
• Sex and age are reported only for the whole trial, not for the scanned subset.