Dietary patterns

The diets with the best longevity evidence — Mediterranean, DASH, MIND, Nordic, and what the long-lived "Blue Zone" populations eat — all land on much the same plate: plenty of vegetables, legumes, nuts, fish, whole grains and plant oils, and not much else. Get that pattern roughly right and the differences between the named diets are small; what is oversold is the idea that any one label, or any one brain-boosting tweak, is the secret.

Make a Mediterranean-style plate your default: vegetables and fruit daily, legumes at least three times a week (ideally most days), fish two to three times a week, nuts and olive oil daily, whole grains over refined, unprocessed red meat at most about three portions a week (350–500 g cooked), and processed meat and ultra-processed food rarely. You don't need perfect compliance — the pattern dominating your eating is what counts, and improving it in mid-life still registers.

Why the Mediterranean diet leads the field

A Mediterranean-style diet probably lowers heart attacks and strokes, and people who eat this way live longer. It has the deepest evidence of any named diet: one large randomised trial whose result held up on re-analysis despite flawed randomisation, and a body of observational studies. Probably true, not settled.

The core of the pattern: vegetables and fruit (5+ servings a day), whole grains, legumes (3+ servings a week), a daily handful of nuts, extra-virgin olive oil as the main fat, fish 2+ times a week, modest yogurt and cheese, poultry and eggs a few times a week, and red meat only once or twice a week. Eating that little is fine, but the health-based ceiling is about three portions of unprocessed red meat a week, roughly 350–500 g cooked (see Red and processed meat). Traditional versions include wine; the modern evidence on alcohol has shifted against it — see Alcohol.

The trial. PREDIMED, a Spanish trial of 7,447 adults at high cardiovascular risk, found roughly 30% fewer heart attacks, strokes and cardiovascular deaths on a Mediterranean diet supplemented with olive oil (at least four tablespoons, about 50 g, a day) or nuts than on a low-fat control diet.[1] Two caveats travel with it. The original 2013 paper was retracted and republished in 2018 after the investigators found randomisation problems affecting about 21% of participants; the ~30% survived re-analysis, but it is no longer a clean randomised trial. And while the funding was public, the olive oil and nuts were donated by producers, who the paper says had no role in the study.

A follow-up trial added a hard outcome. Its successor tested a calorie-reduced Mediterranean diet plus exercise against Mediterranean-diet advice alone, and over a median six years the calorie-reduced arm had 31% fewer new cases of type 2 diabetes.[2] Because both arms were advised to eat Mediterranean-style, this shows the value of cutting calories and moving more on top of the diet, not the diet itself. Its main cardiovascular result has not yet been published.

The observational evidence. A review of 28 studies in older adults found high adherence linked to about 23% lower death from any cause and 27% lower cardiovascular death. Non-fatal heart attacks and strokes pointed the same way but did not reach statistical significance, and the review's two randomised studies were both the Spanish trial above.[3] Pooling 54 cohorts and 1.8 million people, each one-point step up a Mediterranean adherence score tracked with about 4% lower mortality — roughly 15–19% between typical low and high adherers.[4]

Much of the raw association is the lifestyle that comes with the diet. In a study of 25,315 women followed for about 25 years, the best Mediterranean eaters had 23% lower mortality before adjusting for smoking, exercise, alcohol and menopause — and 11% after. Measured blood markers explained only about a fifth of the link.[5]

Cancer is more selective than "lower cancer risk" suggests. Pooling 117 studies, adherence tracked with lower colorectal, stomach, liver and breast cancer, but not blood, oesophageal, pancreatic or prostate cancer. The steadier signal is about 13% lower cancer death.[6]

What is probably doing the work

No single nutrient explains the benefit. Three ideas have the most support; none is proven.

Polyphenols, not just "healthy fat". Extra-virgin olive oil carries plant compounds called polyphenols alongside its fat, and these plausibly drive much of its benefit to blood vessels. Within the Spanish trial, an observational analysis of 327 deaths found people in the top fifth of total polyphenol intake had 37% lower mortality than the bottom fifth — but there was no clear dose-response trend, and oddly the link held only in the low-fat control arm, not in either Mediterranean arm.[7] Pooled cohorts of nearly 180,000 adults find a smaller 7% lower mortality at higher polyphenol intake.[8]

Plant protein in place of red meat. Legumes, nuts and fish supply the protein to maintain muscle without the strong growth signalling that comes with a lot of red and processed meat. See Protein.

The gut. In small pilot studies, switching from a Western to a Mediterranean diet changed gut microbes within four days; fibre and polyphenols may damp chronic low-grade inflammation — plausible, but early.

The brain: where the hype has outrun the evidence

That a Mediterranean diet protects against dementia might be true, but the evidence is thin — the weakest part of the story: observational links, a null trial, and an effect that may partly run backwards.

Mediterranean eating and dementia. A 2025 pooling found strong adherence linked to 18% lower risk of cognitive impairment, 11% lower dementia and 30% lower Alzheimer's — three separate outcomes from three separate groups of studies.[9] A 2026 pooling of an overlapping literature found the Mediterranean–dementia link not statistically significant overall, and clearest in studies with ten years or less of follow-up — the signature of early cognitive decline changing what people eat. Oddly, the same pooling did find a significant link for the MIND score.[10] The leading international commission on dementia prevention lists 14 modifiable risk factors, and no dietary pattern is among them.[11]

The MIND diet is a Mediterranean–DASH hybrid built for the brain, adding targets for green leafy vegetables, berries and nuts and limits on butter, cheese, pastries and fried food. Its famous "53% lower Alzheimer's risk" came from the smallest study, 923 older adults;[12] pooling 11 cohorts and 224,049 people puts it at about 17%, a third of the headline.[13]

Then it was tested, and failed. A three-year trial randomised 604 cognitively healthy older adults — chosen for a family history of dementia, excess weight and a poor diet — to the MIND diet or their usual diet, with both groups cutting about 250 kcal a day. Both improved, the difference between them was well within chance, and brain scans showed no difference either.[14] The observational picture is too muddled to say which part of a healthy pattern, if any, protects the brain.

DASH and Nordic: real effects, smaller than quoted

In randomised trials, DASH probably lowers blood pressure, by less than the famous number. DASH (Dietary Approaches to Stop Hypertension) — fruit, vegetables, whole grains, low-fat dairy, less saturated fat, red meat and sugary drinks — is the pattern built and tested for blood pressure. The original eight-week feeding trial lowered systolic pressure by about 11 mmHg in people with hypertension but only about 3.5 mmHg in those without; the widely quoted ~6 mmHg is the whole-trial average.[15] Across 30 later trials the typical effect is about 3 mmHg systolic, much the same with or without hypertension — real but modest, a few points, not a drug.[16]

Cutting salt adds more, but the two overlap rather than stack: going from high to low sodium was worth about 6.7 mmHg on an ordinary diet and about 3.0 mmHg on top of DASH.[17] In cohorts, each 5-point step up a DASH score tracks with about 5% lower mortality, but the studies vary widely and are observational.[18]

The Nordic diet is a cold-climate cousin of the Mediterranean one: rapeseed (canola) oil instead of olive oil, rye, barley and oats instead of wheat, and local berries. Pooling 15 trials, it improved blood pressure, blood sugar, insulin and blood fats, but not inflammation.[19] Scored side by side in 38,428 Swedish women, Mediterranean eating carried 24% lower mortality against 11% for Nordic eating.[20]

The label matters less than what replaces what

Every healthy pattern helps, and the winners share the same foods. The most direct comparison scored eight patterns in the same 105,015 health professionals followed for up to 30 years, and asked who reached 70 free of major chronic disease with mind and body intact — 9.3% did. The best performer, the Alternative Healthy Eating Index, gave up to 86% higher odds for top versus bottom scorers. More ultra-processed food meant 32% lower odds.[21] The spread between patterns was real but modest, and even among high scorers only about 8–12% reached 70 in full health. The participants were mostly white health professionals.

"Plant-based" is not enough. A healthful plant-based diet — whole grains, legumes, fruit, vegetables, nuts — is linked to lower mortality. An unhealthful one built on refined grains, juice, sugary drinks and sweets is linked to higher mortality, though both are technically plant-based.[22] Swapping meat for cola and biscuits is not a longevity strategy. Plant quality is a floor to clear, not what ranks the good diets: in the eight-pattern comparison, the healthful plant-based score came last.[23]

Carbohydrate follows a U-shape, and the substitute probably decides it. In a 25-year cohort of 15,428 people plus a pooling of 432,179, the lowest mortality fell at 50–55% of energy from carbohydrate; below 40% and above 70% both carried higher risk. Mortality rose when carbohydrate was replaced by animal fat and protein, but fell when it was replaced by plant sources.[24] Moderate low-carb looks best for heart deaths; very-low-carb and ketogenic diets less so, and ketogenic diets raise LDL ("bad") cholesterol in trials with no shown long-term benefit.[25]

Olive oil is good; it is not uniquely good. Against butter, rapeseed/canola and soybean oil do as well, so the claim that seed oils are toxic doesn't survive the evidence: per 5 g a day, about 15% lower mortality for canola, 8% for olive, 6% for soybean.[26] Olive oil earns its place through the pattern and its polyphenols; the main cohort analysis finds it no better than other plant oils.[27] Seed oils and the rest are covered under Dietary fats.

It's never too late. In a large observational study tracking about 74,000 adults over 12 years, those who improved their diet quality the most had 9–16% lower mortality than people whose diet stayed the same, and those whose diet got worse had higher mortality.[28]

Vegetarian and vegan: a trade-off, not a free win

In observational studies, dropping meat shifts which diseases people get rather than simply adding years, and the evidence for a lifespan gain is thin.

A large study of Seventh-day Adventists first found vegetarians had 12% lower mortality than meat-eaters.[29] The same cohort's 2024 update of 88,400 people found the advantage was about 11% at 65 but gone by 85, and clearest in men; in late old age, vegetarians were more likely to die of stroke (17% higher) and of dementia (13% higher, borderline).[30]

Two British cohorts of 60,310 people found no overall mortality difference between vegetarians and regular meat-eaters.[31] Underneath that is a trade: pooled studies find 13% less cancer in vegetarians and 23% less in vegans,[32] while over 18 years British vegetarians had 22% less heart disease but 20% more strokes — about 3 extra strokes per 1,000 people over 10 years, mostly bleeding strokes.[33] The full weighing is under Vegetarian diet.

Blue Zones: the plate is credible, some of the ages are not

The long-lived regions — Sardinia, Okinawa, Nicoya, Ikaria and Loma Linda — eat in ways that line up with the Mediterranean pattern: mostly plants, beans as a daily protein, starchy staples (the traditional Okinawan diet got over half its calories from sweet potato), fermented foods, and modest meat. The famous "95% plant" figure is a popular framing, not a measurement; the one hard number is a reconstructed 1949 Okinawan food record in which animal foods supplied only about 1–2% of calories.[34] Other shared habits: eating to about 80% full, the smallest meal in the late afternoon or early evening, and moderate drinking in four of the five zones — the commonality that has aged worst.

Some centenarian counts have been challenged as products of record errors and possibly welfare fraud. That critique is serious but remains an unreviewed preprint,[35] and demographers who validated the Sardinian records have answered it, noting it concerns people aged 110 and over while the Blue Zones claim rests on the share reaching 90.[36] Their diets remain a reasonable template either way. See Blue Zones.

Newer claims to hold loosely

The "planetary health" diet. In a review of 227 studies, high adherence was linked to about 20% lower mortality and 17% less cardiovascular disease, with no link to dementia or overall cancer incidence — but the review rated its own certainty very low, and no trial exists.[37] It may need attention to iron and vitamin B12, and the commission behind it — whose 2025 edition claims around 15 million premature deaths averted a year — is advocating as much as measuring.[38]

Calorie restriction and biological aging. In a two-year randomised trial of 220 healthy non-obese adults, aiming for 25% fewer calories slowed one DNA-based "pace of aging" measure by roughly 2–3%. Three deflators: people actually managed about 12% restriction, the analysis was done after the fact, and two other aging clocks did not move. The investigators' estimate of 10–15% lower mortality is an extrapolation, not something the trial measured.[39] See epigenetic alterations.

Very short eating windows: a caution. In a US nutrition survey followed for about 8 years, people who reported eating within under 8 hours a day had more than double the cardiovascular mortality of those eating across 12–14 hours.[40] The estimate rests on 383 people and 33 deaths, who were more often smokers, food-insecure and on low incomes and were never shown to be fasting on purpose, so confounding is likely. The full weighing is under Short eating windows and heart deaths.

What to actually eat

Limit:

  • Ultra-processed food — the most consistently replicated dose-response signal in modern nutrition, though not a high-certainty one: about 15% higher mortality at the highest intakes, and somewhere between 3% and 10% more for each extra tenth of the diet, depending on the analysis.[41] [42] See Ultra-processed food.
  • Sugary drinks, processed meat (charcuterie, bacon, sausages, hot dogs), excess red meat, refined grains and added sugar, and excess salt, mostly from processed food.

A simple week:

  • Daily: salad or leafy greens, vegetables and fruit in many colours (especially berries), olive or another plant oil, nuts and seeds, water or unsweetened drinks
  • Most days: whole grains over refined, vegetables in volume
  • 3+ times a week, most days if you can: legumes, as the default protein
  • 2–3 times a week: fish, especially oily fish
  • Occasionally: poultry, eggs
  • Up to about three times a week (350–500 g cooked in all), less is fine: unprocessed red meat
  • Rarely: processed meat, fried fast food, sweets

Cook from whole ingredients and build meals around vegetables.

If you are over 70, two things change. The goal shifts from living longer toward staying independent, and here the pattern still earns its place: pooling 94,072 older adults, each step up a Mediterranean score tracked with about 5% lower risk of becoming frail, though not with less disability.[43] The second change is protein. Older muscle needs more protein to respond, so a plant-forward plate needs a deliberate protein target — Protein sets 1.0–1.5 g/kg/day past 65, spread as about 0.4 g/kg per meal (25–30 g at 65–75 kg, 30–40 g if heavier). Past 70, adequate protein no longer takes care of itself.

Further reading

  • Estruch R et al. Primary Prevention of Cardiovascular Disease with a Mediterranean Diet (PREDIMED). NEJM 2018.[44]
  • Furbatto M et al. Mediterranean diet and cardiovascular outcomes in older adults — meta-analysis of 28 studies. Nutrients 2024.[45]
  • Nucci D et al. Mediterranean diet adherence and all-cause mortality — dose-response meta-analysis of 54 cohorts. Nutrition 2026.[46]
  • Ahmad S, Moorthy MV, Mora S et al. Mediterranean diet adherence and risk of all-cause mortality in women. JAMA Network Open 2024;7(5):e2414322.[47]
  • Chen H et al. Association of the Mediterranean-DASH Intervention for Neurodegenerative Delay diet with dementia — meta-analysis of 11 cohorts. JAMA Psychiatry 2023.[48]
  • Barnes LL et al. Trial of the MIND Diet for Prevention of Cognitive Decline in Older Persons. NEJM 2023.[49]
  • Appel LJ et al. A clinical trial of the effects of dietary patterns on blood pressure (DASH). NEJM 1997.[50]
  • Filippou CD et al. Dietary Approaches to Stop Hypertension (DASH) diet and blood pressure — meta-analysis of 30 randomized controlled trials. Adv Nutr 2020.[51]
  • Tessier AJ, Wang F, Guasch-Ferré M et al. Optimal dietary patterns for healthy aging. Nature Medicine 2025;31:1644–1652.[52]
  • Seidelmann SB et al. Dietary carbohydrate intake and mortality: a prospective cohort study and meta-analysis. Lancet Public Health 2018;3(9):e419–e428.[53]
  • Sotos-Prieto M, Bhupathiraju SN, Hu FB et al. Association of Changes in Diet Quality with Total and Cause-Specific Mortality. NEJM 2017;377:143–153.[54]
  • Abris GP et al. Cause-specific and all-cause mortalities in vegetarian compared with those in nonvegetarian participants from the Adventist Health Study-2 cohort. Am J Clin Nutr 2024.[55]
  • Tong TYN et al. Risks of ischaemic heart disease and stroke in meat eaters, fish eaters, and vegetarians (EPIC-Oxford). BMJ 2019.[56]
  • Waziry R, Ryan CP, Belsky DW et al. Effect of long-term caloric restriction on DNA-methylation measures of biological aging (CALERIE). Nature Aging 2023.[57]

— § —