US study of 130,000 finds closest Mediterranean diet followers had 21% lower dementia risk; trials mixed

A study in the United States followed about 130,000 nurses and health professionals for up to more than 40 years. It found that the group whose diet was closest to a Mediterranean diet had a 21% lower risk of dementia than the group furthest from it.

The study, by Liu and colleagues, looked at long-term adherence to, and changes in, the Mediterranean and MIND diets in relation to dementia risk and cognitive function. It appeared in Alzheimer's & Dementia, volume 22, issue 3, published in March 2026 (DOI: 10.1002/alz.71324). It also compared a second diet score called MIND: the group with the highest adherence had a 14% lower risk.

The numbers are appealing, but do not rush to rewrite your shopping list. Set against the wider evidence, the picture is far more complicated.

130,000 people, two scores: how the study worked

This was a prospective cohort study using two long-running US cohorts: 86,740 women from the Nurses' Health Study and 43,500 men from the Health Professionals Follow-up Study, about 130,240 people in total. Follow-up ran from 1980 to 2023 for the former and from 1986 to 2023 for the latter. Researchers assessed participants' diets repeatedly with food frequency questionnaires rather than asking only once.

According to the published record, the hazard ratio (HR) for dementia in the highest versus lowest Mediterranean diet adherence group was 0.79 (95% confidence interval 0.73–0.84); for the MIND diet it was 0.86 (95% confidence interval 0.81–0.91).

Psychiatric News reported that each one-point increase in adherence score corresponded to a 5% lower dementia risk (Mediterranean diet) and a 3% lower risk (MIND diet). The study abstract also states that higher adherence was associated with 0.75 to 1.59 fewer years of cognitive ageing, and with a 40% to 42% lower risk of subjective cognitive decline.

The point most useful to most readers: it was reported that improving adherence over 4 or 8 years showed a similar association, even for those who started from a low level.

What observational studies can and cannot tell us

The authors point out that the results are observational associations and cannot establish causation. The report also notes that people with good eating habits may also have other healthy behaviours, and that such confounding cannot be ruled out.

On measurement, dementia cases were identified from participants' self-reported physician diagnoses and death records; cognitive function was tested by telephone; and subjective cognitive decline was self-reported by participants.

A further note: we could not check the full text of the paper, so the adjustment variables and the authors' full list of limitations have not been verified. The figures above rely on the published record, the abstract and press reports.

Randomised trials give mixed answers

To ask whether changing diet directly protects the brain, randomised controlled trials are better placed to test causation. The results so far are one positive and one negative.

PREDIMED Barcelona sub-study: the olive oil group did better

This randomised clinical trial, by Valls-Pedret and colleagues, was published in JAMA Internal Medicine in 2015. Between 2003 and 2009 it recruited 447 adults who were cognitively healthy but at high cardiovascular risk, with a mean age of 66.9. Participants were split into three groups: a Mediterranean diet with extra-virgin olive oil, a Mediterranean diet with mixed nuts, and a control diet advising reduced fat.

In the end 334 people completed follow-up testing, with a median follow-up of 4.1 years. The olive oil group outperformed the control group on some tests and on composite scores, while all composite scores in the control group fell from baseline. The authors concluded that a Mediterranean diet supplemented with olive oil or nuts was associated with improved cognitive function.

However, according to a secondary review, the participants were older adults at high cardiovascular risk rather than the general population, so the results should not be applied directly to everyone.

MIND diet trial: no significant difference after three years

The trial, published online in the New England Journal of Medicine in July 2023, enrolled 604 older adults who were cognitively healthy, had a family history of dementia, a body mass index (BMI) above 25 and a suboptimal diet. 301 were assigned to the MIND diet and 303 to a control diet, with both groups mildly restricting calories, for 3 years.

Cognition improved in both groups, and the difference between groups was only 0.035 standardised units (95% confidence interval -0.022 to 0.092, P=0.23), which was not statistically significant. Brain MRI measures were also similar in the two groups. The trial completion rate was 93.4%.

With longer follow-up, the association disappeared

The Rotterdam Study, published in Alzheimer's Research & Therapy in 2022, analysed the MIND diet and dementia. In the first baseline cohort of about 5,375 people, followed for an average of 15.6 years, 1,188 developed dementia. A higher diet score predicted lower risk only in the first 7 years; with longer follow-up, the association disappeared.

The authors suggested the association may partly reflect reverse causation and lifestyle confounding, and that more research is needed to determine the actual effect of diet.

An earlier overview made the same point. A 2016 systematic review by Hardman and colleagues in Frontiers in Nutrition included 18 studies, 13 with favourable results and 5 without; because study designs differed too much, the data could not be pooled and only a qualitative synthesis was possible. The review noted that studies assessed adherence in different ways, that there were few intervention trials in healthy older adults, and called for more randomised controlled trials using standardised dietary assessment. Notably, PREDIMED was not among those 18 studies.

The strength of the evidence at a glance

• Large observational study (US, about 130,000 people): higher adherence is associated with lower risk, but this cannot prove causation, and some case identification relied on self-reported data.
• PREDIMED Barcelona sub-study (randomised trial, 447 people): the olive oil group beat the low-fat control group on some cognitive tests, but participants were older adults at high cardiovascular risk.
• MIND diet randomised trial (604 people, 3 years): no significant difference compared with the control diet.
• Rotterdam Study (observational): the association appeared only in the first 7 years and vanished with longer follow-up.
• 2016 systematic review (18 studies): 13 had favourable results, but methods varied and the data could not be pooled.

Back at the table: a few editorial observations

The following are general directions drawn by this publication from the studies above. They are not medical advice and do not involve any supplements. This article does not cite specific meal guidance from health authorities; for a personalised plan, consult a doctor or registered dietitian.

• Look at long-term patterns, not single meals: the studies assessed participants' diets repeatedly over many years, so they reflect habits rather than any one "brain-boosting dish".
• It is not too late to start: it was reported that improving Mediterranean diet adherence was associated with lower risk even from a low starting point.
• Each one-point rise in score was associated with lower risk (an observational association, not proof of cause), which suggests the studies looked at degrees of overall dietary pattern rather than all or nothing.
• The two Mediterranean diet groups in PREDIMED added extra-virgin olive oil or mixed nuts respectively; the olive oil group did better than the control group on some cognitive tests, but this is not a specific shopping recommendation.

The safest reading is to treat a good dietary pattern as a long-term habit, not a guarantee against dementia. The evidence so far is mixed, and it still cannot show that diet itself is the cause.

Prices and details correct as of October 2026

This article is general health information, not medical advice. For any health concern, consult a doctor or registered health professional.