Trial of the MIND Diet for Prevention of Cognitive Decline in Older Persons

**Intervention:** The MIND diet — a hybrid of the Mediterranean and DASH diets, specifically designed to emphasise foods linked to brain health. It promotes 10 "brain-healthy" food groups (green leafy vegetables ≥6 servings/week, other vegetables ≥1/day, berries ≥2 servings/week, nuts ≥5 servings/week, olive oil as primary oil, whole grains ≥3 servings/day, fish ≥1 serving/week, beans ≥4 meals/week, poultry ≥2 servings/week, wine ≤1 glass/day) and limits 5 "unhealthy" groups (red meat, butter/margarine, cheese, pastries/sweets, fried/fast food).

**Comparator:** A "healthy diet control" — a diet also guided by nutritional counselling and designed to be healthy, but without the specific MIND diet emphasis. This is a critical design choice: the control group was *not* eating their usual diet; they received active dietary guidance too.

**Primary outcome:** Change in a global cognitive composite score (average of 12 individual cognitive tests) over 3 years.

**Secondary outcomes:** Changes in individual cognitive domains (episodic memory, semantic memory, working memory, perceptual speed, visuospatial ability), and brain MRI measures in a subset.

---

**Sample size:** 604 participants randomised (301 MIND diet, 303 control diet)

**Population:** Adults aged 65 years or older (mean age ~70) with a family history of dementia (a parent or sibling with Alzheimer's disease or dementia) OR suboptimal diet quality at baseline (scoring below median on a dietary screener)

**Setting:** Two US academic sites — Rush University Medical Center (Chicago) and Harvard University / Brigham and Women's Hospital (Boston)

**Exclusion criteria included:** Existing dementia or mild cognitive impairment, BMI >40, active cancer, recent cardiovascular event, major psychiatric illness, inability to follow a dietary intervention

**Baseline diet:** Participants were deliberately selected for having room to improve — they were not already eating optimally

---

**Cognitive composite score:** Average of 12 neuropsychological tests spanning five domains, converted to z-scores so they could be averaged. This is the primary outcome — a single number capturing overall cognitive function.

**Individual cognitive domains tested:**

- *Episodic memory:* Word list memory, word list recall, word list recognition, story recall

- *Semantic memory:* Boston Naming Test, verbal fluency (category and letter)

- *Working memory:* Digit span forward and backward

- *Perceptual speed:* Symbol Digit Modalities Test, Number Comparison

- *Visuospatial ability:* Standard Paper Folding Test

**Dietary adherence:** MIND diet score (0–15 scale) assessed via multiple 24-hour dietary recalls and food frequency questionnaires at regular intervals

**Body weight:** Monitored as a safety/adherence check

**Brain MRI:** Conducted in a subset at baseline and follow-up — volumes of hippocampus, white matter hyperintensities, and other structural measures

**Blood biomarkers:** Collected to assess nutritional status and metabolic markers

---

### Study design

This is a **parallel-group randomised controlled trial (RCT)** — the gold standard for causal inference. Participants were randomly assigned to either the MIND diet or the healthy control diet and followed for **3 years**.

### Randomisation

Participants were randomised with stratification by site (Chicago vs. Boston) and by baseline cognitive composite score (above vs. below median). This stratification ensures the two groups are balanced on the most important variables at the start.

### Blinding

**Participants:** Could **not** be blinded — they knew which diet they were eating. This is an inherent limitation of dietary intervention trials.

**Cognitive assessors:** Were blinded to group assignment — the people administering the neuropsychological tests did not know which diet the participant was on. This is the critical blinding that protects the primary outcome from observer bias.

**Statisticians:** Analysed data blinded to group assignment.

### Dietary intervention delivery

Both groups received individualised dietary counselling from registered dietitians — including group sessions, individual sessions, and cooking demonstrations. This "active control" design is sophisticated: it means any difference in outcomes between groups is attributable to the *specific components* of the MIND diet, not just "being told to eat better" or "receiving attention from a dietitian." This makes the null result more meaningful — even structured healthy eating guidance didn't separate from the MIND diet specifically.

### Duration

3 years of active intervention — long enough to plausibly detect diet-related cognitive changes, though possibly still short for a condition that develops over decades.

### Statistical approach

Primary analysis: linear mixed-effects models comparing the rate of change in global cognitive score between groups over 3 years

Intention-to-treat (ITT) analysis — participants were analysed in the group they were assigned to, regardless of adherence

Pre-specified subgroup analyses included: baseline diet quality, age, sex, race, site, and APOE ε4 genotype (a major genetic risk factor for Alzheimer's)

### What this design CAN prove

Because of randomisation and the active control group, this study can tell us whether the *specific* MIND diet pattern produces different cognitive outcomes than a generic healthy diet over 3 years in high-risk older adults.

### What this design CANNOT prove

It cannot tell us whether *either* diet prevents dementia compared to an unhealthy diet (there was no "eat whatever you want" control group)

3 years may be insufficient to detect effects on a process that unfolds over 20–30 years

It cannot generalise to people without family history of dementia or without dietary room for improvement

Open-label dietary trials always carry risk of non-specific effects (social engagement, placebo effects on self-reported outcomes — though here the primary outcome was objective testing)

---

### Primary outcome

The MIND diet group did **not** show significantly greater improvement in global cognitive composite score compared to the healthy control diet group over 3 years

Mean change in cognitive composite score: approximately **+0.205 standardised units** in the MIND diet group vs. **+0.170 standardised units** in the control group

Difference: ~0.035 standardised units (95% CI not significantly different from zero; p-value not statistically significant)

**Both groups improved** — likely reflecting practice effects on cognitive tests and/or the benefit of any dietary improvement

### Secondary outcomes (cognitive domains)

No statistically significant differences between groups on any individual cognitive domain

No significant between-group differences in episodic memory, perceptual speed, semantic memory, working memory, or visuospatial ability

### Brain MRI (subset)

No significant between-group differences in hippocampal volume or white matter hyperintensity volume

### Pre-specified subgroup: baseline diet quality

Participants with the **lowest baseline diet quality** showed a larger cognitive benefit from the MIND diet compared to controls — suggesting the diet may help most in people who have the most dietary room for improvement

This was a pre-specified subgroup analysis, lending it more credibility than a post-hoc finding, but subgroup results must still be interpreted cautiously

### Adherence

The MIND diet group did meaningfully increase their MIND diet scores from baseline, and the groups were separated in dietary pattern throughout the trial — confirming the intervention was actually received

---

The difference in cognitive trajectories between groups (~0.035 standardised units over 3 years) is **clinically negligible** — far below the threshold that would be detectable by the person themselves or their family, and smaller than what the trial was powered to detect as meaningful.

To put it in perspective: the within-group improvement in *both* groups over 3 years was roughly 0.17–0.21 standardised units — meaning the benefit of "any structured healthy diet" was real and visible, but the extra benefit of the *specifically MIND-designed* diet on top of that was not.

The sub