The MIND Diet and Brain Ageing: What ‘2.5 Years’ Really Measured

For a few days this month, a diet slowed brain ageing by two and a half years. That was the headline, from a press release on the first of October to the evening news a week later. The diet was MIND, a cross between the Mediterranean diet and the DASH eating plan, and the study behind it was large, long and careful.

I read the paper because I wanted the number to be true. Berries and leafy greens buying back years of brain tissue is the kind of finding that makes a grocery list feel like a plan. The paper is better than most of the coverage. But the 2.5 years is not something anyone measured. It is a conversion, and it rests on one of twelve brain measures the team looked at.

None of this means the diet is useless. It means the honest version is quieter than the headline, and more useful, because it tells you what to actually put on the plate.

Where the 2.5 years comes from

The study, led by Hui Chen, Changzheng Yuan and colleagues and published in the Journal of Neurology, Neurosurgery & Psychiatry, used the Framingham Offspring cohort in Massachusetts. These are 1,647 adults, average age 61 at the start, who filled in food questionnaires in the 1990s and then had brain scans, usually three, over a median of 12.3 years.

Every brain loses some grey matter with age. In this group the average loss was about 1.4 cubic centimetres a year. For each three-point step up the 15-point MIND score, that loss was 0.28 cubic centimetres a year slower. That is roughly a fifth of the typical rate.

The authors then translated that fifth into time: a fifth of a 12.3-year follow-up is about 2.5 years. It is a fair piece of arithmetic. It is also the only reason the press release could say “years”. In physical terms, the difference is about a seventeenth of a teaspoon of tissue a year.

Small is not the same as meaningless. Brain changes add up slowly over decades, and a fifth of a slope is not nothing. But a headline written in years invites you to picture a scan that looks younger. Nobody in the study had one.

Twelve measurements, three movers

The team measured twelve features of the brain on every scan. Total brain volume, grey matter, white matter, the fluid around the brain, four measures of the fluid-filled spaces called ventricles, three measures of the hippocampus, and the small bright spots on scans known as white matter hyperintensities.

After a standard correction for testing many things at once, three of those twelve held up. Grey matter declined more slowly, and two ventricle measures grew more slowly. Total brain volume did not move. The hippocampus, the structure most people have heard of because it matters for memory, did not move at all. White matter, before correction, leaned the other way.

Twelve brain measures, one study

What a three-point higher MIND score was linked to over 12 years

Three ways to read “2.5 years of slower brain ageing”

As written: grey matter shrank about a fifth more slowly per three-point step on the score, and a fifth of 12.3 years is 2.5 years.

As measured: about 0.28 cubic centimetres a year of grey matter, in one of twelve measures, with total brain volume and the hippocampus unchanged.

As evidence: an association in a mostly white, mostly college-educated group, with diet recorded by questionnaire. It suggests a direction. It cannot show that the food caused the difference.

Cheese helped and whole grains did not

The MIND score is a checklist of fifteen food groups. Ten are “brain-healthy”: leafy greens, other vegetables, nuts, berries, beans, whole grains, fish, poultry, olive oil and a little wine. Five are to be limited: red meat, butter and margarine, cheese, pastries and sweets, and fried or fast food.

When the team looked at the foods one by one, the list partly broke. Berries and poultry carried much of the favourable signal. Sweets and fried food went the expected way. But whole grains, one of the ten recommended groups, were linked to faster grey matter and hippocampal loss. And cheese, one of the five to limit, was linked to slower loss on almost every measure. The authors call these results unexpected, and they are right to.

I do not think this means you should swap your oats for cheddar. Food-by-food results in a single cohort are noisy, and with fifteen foods and twelve outcomes, some odd pairs are likely to appear by chance alone. What it does show is that the score works as a pattern, not as a menu of medicines. If you want to know why berries keep turning up in this kind of work, our piece on flavonoids and colourful plants follows that thread.

Who scores high on a diet score

The baseline table is where I would look first in any diet study, and this one is worth reading. People in the top third of MIND scores were not only eating more berries. Eighty-three percent had a college degree, against 62% in the bottom third. About 5% smoked, against 15%. They were more often women, less often obese, and had less diabetes and heart disease.

The team adjusted for all of this, and they did it carefully. But adjustment works with the categories you can record: degree or no degree, smoker or not. It cannot fully capture a life that has more money, more time, more walking and more doctor visits. Nutrition researchers call this the healthy-user problem, and a food questionnaire is the instrument most exposed to it. It is also the reason we spend so much time on what a food category really measures.

There is a second limit, and the authors name it. The scans began in 1999, and some people may already have had small changes in memory or thinking that the study could not exclude. Early changes in the brain can change appetite and eating habits. In that case the arrow runs partly backwards.

The trial that already ran

Most coverage this month left out one fact. The MIND diet has been tested in a randomised trial, and the results were published in 2023.

Lisa Barnes and colleagues enrolled 604 older adults with overweight and a less-than-ideal diet, and followed them for three years. Half were coached to eat the MIND way with a mild cut in calories. The other half were coached to cut the same calories on their usual diet. Both groups lost about five kilograms. Both groups improved on memory and thinking tests. The difference between them was small and not statistically clear.

The participants who had brain scans told the same story. Grey matter, white matter, hippocampus and white matter spots changed the same way in both groups.

Three years is short for a brain, and the control group was not a junk-food group. It was a group of people paying close attention to what they ate, with a dietitian, for three years. You can read the trial as a failure of MIND. I read it as a result about attention. When people with a mediocre diet start eating with care and lose a little weight, both arms move in the same direction, and the specific checklist adds little on top in the time available.

What three points looks like on a plate

Here is the useful part. On the original scoring by Martha Clare Morris and colleagues, each food group earns up to one point, and the median person in the Framingham study scored under seven of fifteen. Three points is roughly the gap between the bottom third and the top third of the cohort. It is not a new diet. It is about three habits:

• Berries at least twice a week. Frozen counts.
• Leafy greens on six or more days a week. A side salad or a handful of spinach in whatever you are cooking is enough.
• Fried or fast food less than once a week, and pastries and sweets fewer than five times.

These are the moves this study and the earlier ones point to, and none of them asks you to trust a single food. They also overlap heavily with what we wrote about a four-week diet trial and biological age: eat more plants, less of the processed end, and stay consistent for years rather than weeks.

Two things a diet score does not include are worth keeping beside it. In this study, one ventricle result was stronger in the more physically active participants. And fitness has its own evidence for the ageing brain, which we cover in cardiorespiratory fitness and the brain. No wristband measures grey matter. Steps, sleep timing and resting heart rate are trends you can watch, and they sit on the same side of the ledger as the berries.

The bottom line

The new study is real, long and well run. It found that people who ate closer to the MIND pattern lost grey matter about a fifth more slowly over twelve years, with no difference in total brain volume or the hippocampus. The 2.5 years is a translation of that slope, not a scan that looked younger. The one randomised trial found that eating with care helped both groups about equally.

So eat the berries, the greens and less fried food, because the pattern is sound and the cost is low. Do not buy the years. If you have concerns about your memory or thinking, talk to your doctor rather than a food list.