Eating closer to the MIND diet was linked to slightly younger DNA clocks three years later

The MIND diet was designed with the brain in mind. It blends the Mediterranean diet with DASH, a diet built to lower blood pressure, and scores people on 15 food groups. Points go for eating plenty of leafy greens and other vegetables, berries, beans, nuts, whole grains, fish, chicken and olive oil, and for about a glass of wine a day; more points go for eating less of five foods: red meat, cheese, butter or margarine, sweets and pastries, and fried or fast food.

Studies that follow people over time have linked this kind of eating to slower memory decline. But a three-year randomized trial, mentioned by the authors, found no clear difference in thinking skills between people put on the MIND diet and a comparison group that also cut calories. So researchers are looking for in-between signs that a diet might be changing how the body ages.

One such sign is DNA methylation: small chemical tags on DNA that shift with age. Epigenetic clocks read these tags in a blood sample and estimate a "biological age". If the clock says a person is older than their age in years, that gap is called age acceleration. Earlier work, including a large German study we covered, found that healthier diets went with slightly younger clocks, but diet and blood were measured at the same time. This study measured diet first and the clocks about three years later. It is a preprint, so it is not yet peer-reviewed.

What they did

The team used the Health and Retirement Study (HRS), a long-running survey of Americans over 50 that the University of Michigan runs. In 2013–14, some participants filled in a 164-item questionnaire about how often they had eaten different foods over the past year. In 2016, some gave a blood sample from which HRS computed several epigenetic clocks. 1,740 people had both.

They were 47 to 93 years old, 66 on average, and 59% were women. Their average MIND score was 7.2 out of 15. The researchers looked at three measures: PhenoAge and GrimAge acceleration, two clocks trained to predict health and death, and a "pace of aging" measure (DunedinPoAm38) that estimates how fast the body is wearing out, with 1.0 meaning one biological year per calendar year.

People who eat well also tend to be richer, better educated, more active and less likely to smoke. So the main analysis took into account age, sex, ethnicity, education, wealth, marital status, smoking, physical activity and total calories. The team also checked the results with extra adjustment for body weight, with statistical weights for who ended up giving blood, and with newer machine-learning methods.

What they found

• A small link on all three measures. Each step of about 1.9 points up the MIND score went with about 0.4 fewer years of PhenoAge acceleration, 0.3 fewer years of GrimAge acceleration, and a pace of aging lower by 0.0054. All three held up after correcting for testing several outcomes.
• Top third versus bottom third. People in the top third of MIND scores (average 9.5) had about 0.7 fewer years of GrimAge acceleration than those in the bottom third (average 5.4). For PhenoAge the gap was about 0.8 years, but the statistical uncertainty was wide enough that it could have been zero.
• Robust to the checks. Adjusting for body mass index barely changed the numbers, and weighting for who gave blood made them, if anything, larger. The machine-learning estimates pointed the same way.
• No special group. The link looked similar in men and women and at different ages, and the machine-learning search found no reliable subgroup of people who benefited more than others.

What to keep in mind

• Observational. The authors say plainly that the study "cannot establish a causal effect". Even with many adjustments, people who choose this diet may differ in other ways, such as health awareness or access to good food. Taking wealth and smoking into account shrank the links noticeably.
• One snapshot each. Diet was reported once, by questionnaire, and the clocks were measured once. So the study shows that better eaters had younger clocks later, not that their clocks slowed down.
• Who took part. Only some HRS participants did both the diet survey and the blood test. The weighting can only correct for differences that were measured.
• Wine is in the score. Alcohol affects epigenetic aging on its own, and the authors note its part is hard to separate from the rest of the diet.
• Older clocks. The data contained the original GrimAge and DunedinPoAm38, not the newer versions. The authors call for repeat measurements, newer clocks and other cohorts.
• Small differences. A third of a year on a clock is modest, and it is not known how such shifts translate into health for an individual.

Why it matters

Several studies now point the same way: people who eat closer to healthy diet patterns tend to have slightly younger DNA clocks. This one adds a time order, with diet measured before the blood test, and finds that the link survives a long list of checks. The authors suggest that clocks like these could be tried as shorter-term measures in diet trials, where waiting for disease or death takes decades. Whether changing one's diet actually moves these clocks is a question only such trials can answer.