Mediterranean Diet

Diet Reference Archive: dietary pattern

Mediterranean Diet

A way of eating built on vegetables, fruit, legumes, whole grains, nuts and fish, with olive oil as the main fat. It is among the most studied eating patterns in nutrition research, and it is most often recommended for heart health.

Family: Guideline-endorsed and evidence-based dietary patterns Promoted by: no single owner; described by researchers since the 1960s First described: 1960s GRADE certainty of evidence against a low-fat diet: Moderate for stroke; Low for heart attack and cardiovascular death

ON THIS PAGE

1. In brief
2. Overview
3. In practice
4. For the coach
5. The evidence (opens on request)
6. Mechanisms (opens on request)
7. Risks, contraindications and myths (opens on request)
8. Key terms
9. Review questions
10. References
11. Evidence rating and review record

The summary and sections 1 to 3 are the short version and are open when the page loads. Sections 4 to 6 hold the research detail, the biology, and the cautions. Each of those opens when you select its heading, and you can read the page without opening them, though 2 of the review questions draw on them. Back to the Diet Reference Archive index.

In brief

What is it? A way of eating based on the traditional foods of countries around the Mediterranean Sea. Plenty of vegetables, fruit, beans, whole grains and nuts, fish a few times a week, olive oil instead of butter, and small amounts of red meat and sweets.

Does it work? It has been studied more than most ways of eating, and the research points in a good direction, but it is less settled than the headlines suggest. The largest trial found fewer heart attacks, strokes and cardiovascular deaths, counted together, in people at high risk who ate this way instead of following low-fat advice. Taking those outcomes one at a time against that low-fat comparison, independent reviewers rated the evidence on stroke moderate in certainty, and the evidence on heart attack and on death from cardiovascular causes low, because of a problem with how people in that trial were assigned to groups and because those results are imprecise. Blood pressure and blood sugar tend to improve modestly, most clearly against minimal advice. Trials of Mediterranean eating programs measured against minimal advice do show fewer deaths, rated moderate in certainty, though those programs bundled the eating together with exercise and support. Measured against an active low-fat diet, the effect on how long people live is unsettled.

Is it safe? Yes for most people. It removes no food group, and a well-planned version of it can meet nutrient needs without supplements, though that depends on the individual and is not a promise that nobody needs any. Two cautions come up most often. Wine is often described as part of it, and the case for any health benefit from drinking has weakened considerably. And a version heavy on olive oil and nuts costs more than the way many people eat now.

Who should not do it without talking to a doctor or dietitian first? Anyone taking warfarin, a blood thinner affected by the vitamin K in leafy greens. Anyone with chronic kidney disease, because how much potassium and phosphorus is right for them depends on their stage of kidney disease, their lab results and whether they are on dialysis. Anyone with a fish or tree nut allergy. Anyone on blood pressure or diabetes medication, because the pattern can lower both, and the prescriber should know about a substantial change in eating.

Related pages: DASH Diet, MIND Diet and Anti-Inflammatory Diet. Full list on the archive index.

1. Overview

What you eat

Vegetables and fruit at most meals. Beans, lentils and chickpeas several times a week. Bread, pasta, rice and other grains, in their less refined forms where possible. Nuts and seeds. Fish and shellfish 2 or 3 times a week. Olive oil as the main cooking fat and the main dressing. Modest amounts of cheese and yogurt. Eggs and poultry in small amounts. Fresh fruit as the usual dessert. Herbs, garlic, onion and lemon rather than salt for flavor.

The proportions matter more than any single food. The questionnaire researchers used to score how closely people followed the pattern in the largest trial asked for at least 4 tablespoons of olive oil a day, 2 or more servings of vegetables a day, 3 or more pieces of fruit a day, and 3 or more servings each of beans, fish and nuts a week. It also gave a point for drinking 7 or more glasses of wine a week, which is discussed below.

What you avoid or limit

Red meat and processed meat are kept low, at less than one serving a day and often much less. Butter, margarine and cream are limited. Sugar-sweetened and carbonated drinks are limited to less than one a day. Commercial cakes, cookies and pastries are limited to fewer than 3 times a week. None of these are forbidden. The pattern works by proportion rather than by rule.

Where it came from

Nobody designed this pattern. Researchers observed it in places where people were already eating that way, and then described it. In the 1950s and 1960s the physiologist Ancel Keys and colleagues compared the diets and death rates of groups of middle-aged men in 7 countries, and found that the men in Crete and southern Italy had strikingly low rates of coronary heart disease. Their report on 15 cohorts, covering 11,579 men followed for 15 years, found that death rates rose with the share of calories from saturated fat and fell with the share from monounsaturated fat, and that both all-cause and coronary death rates were lowest in the groups whose main fat was olive oil. The authors were careful to write that they were not claiming cause (Keys et al., 1986).

What turned that observation into a named diet was a conference in 1993, where a group including Walter Willett of Harvard and the nonprofit Oldways produced the first Mediterranean diet pyramid, describing the eating of Crete, much of Greece and southern Italy in the early 1960s (Willett et al., 1995). That description helped formalize the pattern and carry it to a wide audience, and it anchors the pattern to a particular place at a particular time, before large-scale food production changed what people in those countries eat.

Cultural and philosophical context

The Mediterranean diet is a living cultural practice as well as a research category, and the difference matters when a client from one of these cultures brings it up. In 2013 UNESCO added it to the Representative List of the Intangible Cultural Heritage of Humanity, in a listing shared by Cyprus, Croatia, Spain, Greece, Italy, Morocco and Portugal. The listing describes it as a set of skills, knowledge, rituals, symbols and traditions covering crops, harvesting, fishing, animal husbandry, conservation, processing and cooking, and above all the sharing and consumption of food. Eating together, in that description, is the foundation of cultural identity across the region.

A client who grew up with this food may hear a coach describe it as a diet and feel that something has been taken from it. There is no single Mediterranean diet in any case. The eating of coastal Greece, inland Spain, southern Italy and Morocco differ from one another, and the research versions are a composite that no one household ever ate exactly.

What it claims to do

Because the pattern has no owner and nothing to sell, the claims made for it come from health organizations, researchers and the press rather than from a promoter. The claims a client is most likely to have met are that it prevents heart attacks and strokes, that it lowers blood pressure and blood sugar, that it protects memory and lowers the risk of dementia, that it reduces the risk of several cancers, and that it lengthens life. Section 4 takes each of these in turn. They are not all equally supported, and 2 of them are considerably weaker than the others.

A client may also cite a ranking. U.S. News and World Report rated the Mediterranean diet its best overall diet every year from 2018 through 2025. In the 2025 edition that rating came from a panel of 69 experts judging 38 diets across 21 categories. The panel size and the number of categories change from year to year, so those figures describe the 2025 edition and should not be attached to the earlier years. A ranking is a summary of expert opinion, not a study, and it is not the reason for anything a coach or a client does.

What it gets right

Almost everything it asks for is supported on its own terms. Vegetables, fruit, legumes, whole grains, nuts and fish each carry independent evidence of benefit. Replacing saturated fat with unsaturated fat lowers cholesterol. The pattern lines up closely with the American Heart Association's 2026 description of a heart-healthy eating pattern, which asks for plenty of vegetables and fruit, mostly whole grains, healthy protein sources, unsaturated fats in place of saturated fats, minimally processed foods, little added sugar and low sodium (Lichtenstein et al., 2026). It also asks a person to add foods rather than remove them, which is unusual among named diets.

How clients describe it, and what it gets confused with

Clients say "I'm doing Mediterranean," "my cardiologist put me on the Mediterranean diet," "we cook with olive oil now," or simply "we eat like my grandmother did." Any of these can mean anything from a careful pattern to olive oil poured over an otherwise unchanged diet. Ask before assuming.

It is most often confused with the DASH diet, which was designed in a laboratory to lower blood pressure and puts low-fat dairy at the center; with the MIND diet, which is a deliberate blend of this pattern and DASH aimed at the brain; and with a general plant-based or vegetarian pattern, which this is not, since fish, poultry, eggs and dairy all appear in it. The DASH Diet and the MIND Diet have their own pages in this archive. A client who says "Mediterranean" and means "mostly plants" is describing something real but different.

2. In practice

A typical day of eating

Breakfast is bread or oats with yogurt and fruit, or eggs with tomatoes, and coffee. Lunch is the larger meal in much of the region: a bean or lentil dish, or a grain salad with vegetables, cheese and olive oil, with fruit after. An afternoon snack is a handful of almonds or walnuts, or fruit. Dinner is lighter, often fish or a vegetable dish with bread, dressed with olive oil and lemon. Water is the usual drink. Nothing here is unusual food, and most of it is cheaper in season.

Cost, products, and who profits

There is no program to buy, no membership, no app and no supplement. That is a real advantage over most of the diets in this archive. What costs money is the food itself, and the honest answer on cost is mixed. In a United States trial, researchers priced the foods participants recorded eating rather than tracking their actual grocery receipts, and found a difference of about 60 cents a day on the Mediterranean pattern that did not reach statistical significance (Kahleova et al., 2024). Read that as a small estimated increase that the study could not distinguish from no change, which is not the same as a finding of no difference. In a study of 2,833 adults in a lower-income region of Spain, the monthly cost of food rose with how closely people followed the pattern, from about 11 percent to about 17 percent of income; the denominator there is the mean disposable income for the region rather than each participant's own income, so the percentages describe the region and not any individual household (Rubini et al., 2022).

The expensive parts are fish, nuts and good olive oil. The inexpensive parts are beans, lentils, cabbage, carrots, onions, frozen vegetables, canned tomatoes, oats and seasonal fruit, and those parts carry much of the benefit. A client with limited money or store access can move a long way into this pattern through the cheap end of it, and a coach who assumes the expensive version is the only version has narrowed the conversation for no reason. Time is the other cost. This is a pattern of cooking, and a client working 2 jobs is being asked for something real.

Nobody owns the pattern, but money does attach to it. Olive oil and nut industry bodies have funded research on it, including part of the largest trial, and olive oil marketing leans on the name heavily. That is worth knowing without treating it as disqualifying, because the funding is disclosed in the papers and the findings line up with evidence from studies funded differently.

Environmental footprint

A diet's environmental footprint depends mostly on how much of it comes from animals, and this pattern uses meat sparingly. A systematic review of studies that modeled populations moving from their current diets to one of 14 more sustainable patterns, the Mediterranean pattern among them, found median reductions of 22 percent in greenhouse gas emissions, 28 percent in land use and 18 percent in water use, and the size of the reduction generally tracked how much animal food was cut (Aleksandrowicz et al., 2016). A Spanish modeling study looked at the Mediterranean pattern on its own. It estimated that moving the country's current diet to the traditional Mediterranean pattern would cut greenhouse gas emissions by 72 percent and agricultural land use by 58 percent, while moving it toward a Western pattern like that of the United States would raise every measure it tracked (Saez-Almendros et al., 2013).

The closest look at how this works out in the United States comes from a national survey of what 17,079 adults reported eating. The people whose eating most resembled the pattern had diets with about 25 percent lower greenhouse gas emissions and about half the land use of the people whose eating resembled it least, largely because they ate fewer beef dishes, meat sandwiches, dairy foods, refined grain dishes and soft drinks. Their diets also had a water-scarcity footprint about 35 percent higher, driven mostly by more fruit, nuts and seeds; the authors call the result a set of sustainability trade-offs (Conrad et al., 2024).

None of these studies measured what happens when one person changes how they eat. The first 2 are models of whole populations, and the survey compares different people on a single day of reported eating. A person's own footprint depends on the version they build. Fish is the complication. It was the third-largest source of emissions in every pattern the Spanish study modeled, and the review points out that eating more fish has to be reconciled with the fragility of many wild fisheries (Aleksandrowicz et al., 2016; Saez-Almendros et al., 2013). A client who came to this pattern for environmental reasons has a reason worth working with.

Cultural fit

The pattern adapts unusually well, because what defines it is a shape rather than a menu. Plenty of vegetables, beans as a protein, a plant oil as the main fat, whole grains, fish, and meat as a flavoring rather than a center. Many food traditions already have that shape or can take it: much of South Asian, Middle Eastern, North and West African, Mexican and Southeast Asian home cooking fits with no loss of identity. Swapping olive oil for another unsaturated oil the household already cooks with keeps the point intact.

Where it conflicts is with food cultures built around meat at every meal, with households where one person cooks for others who will not eat this way, and with religious or cultural practices around particular foods. Wine is the clearest example, and the pattern does not depend on it. The traditional eating of Muslim communities around the Mediterranean has always excluded it.

What makes people stop

Adherence in the long trials was unusually good, which is itself a finding. A review of 18 studies of adults' experiences with this pattern grouped what helps and hinders them as cost, knowing what to do, social and cultural fit, motivation, the shape of daily life, whether the food is available nearby, taste and enjoyment, and a person's circumstances. Its authors noted that most of these are the same things people meet with any healthy eating, and that the ones specific to this pattern were concerns about particular foods, fit with culture and tradition, and a cooler climate (Tsofliou et al., 2022). Taste and enjoyment are on that list, so a coach should ask how the food sits with this client rather than assume people like it.

What happens when you stop

There is more information here than for most diets, because several trials ran for years rather than months. Blood pressure, blood sugar and blood fats respond within weeks and drift back when the eating changes back, in the same way they do for any dietary change. One further piece of evidence is often brought into this discussion and is worth placing carefully. In a network comparison of 121 trials covering 21,942 people, improvements in cardiovascular risk factors seen at 6 months had largely faded by 12 months for every named diet the reviewers examined except this one, where they persisted (Ge et al., 2020). That compares 6-month results with 12-month results in people who were still assigned to their diets. It says nothing about what happens after someone stops, which is the question this section asks.

Effects on heart attacks and strokes are a different kind of question, because they accumulate over years of eating a certain way. No trial has tested what happens when a person follows this pattern for several years and then stops.

Common variations

The MIND diet is a deliberate cross between this pattern and DASH, built around foods thought to protect the brain; it has its own page, MIND Diet, and its own trial evidence, which is weaker than people expect. The green Mediterranean diet is a research version that adds green tea and a plant protein shake and cuts meat further; a 6-month trial of 294 people, 88 percent of them men, in which both Mediterranean versions were calorie-restricted and paired with physical activity, found a somewhat larger reduction in waist measurement on the green version than on the standard one, 8.6 cm against 6.8 cm, a difference that held only among the men. Its larger fall in LDL cholesterol was shown against general healthy-eating advice, not against the standard Mediterranean version (Tsaban et al., 2021). An energy-reduced version paired with exercise has been tested for diabetes prevention. Clients may also describe a "Mediterranean-style" or "Mediterranean-inspired" way of eating, which usually means olive oil and more vegetables without the rest of the pattern.

3. For the coach

SCOPE OF PRACTICE

A health and wellness coach does not prescribe, adjust, endorse or discourage this or any diet, and does not diagnose or treat. The coach helps the client understand what they have chosen, notice how it is going, and decide what to do next. This pattern needs a physician or a registered dietitian at 3 specific points: when the client takes warfarin, because the vitamin K in leafy greens changes how that drug works; when the client has chronic kidney disease, because potassium and phosphorus needs depend on the stage of the disease, lab results and whether the client is on dialysis; and when the client takes medication for blood pressure or diabetes, because the pattern can lower both, and whether anything about the medication needs to change is the prescriber's call. A client who wants an individualized eating plan built for them is describing the work of a registered dietitian, not a coach.

Who does what

The physician

Decides whether this pattern is safe for this person and adjusts the medications it affects. The relevant ones are warfarin, where the clotting test is monitored and the dose adjusted around a steady vitamin K intake; blood pressure medication, where a falling blood pressure may mean too much drug; and diabetes medication, particularly insulin and the older sulfonylurea drugs, where improving blood sugar raises the risk of a low. The labs that matter are kidney function and potassium in anyone with kidney disease, and blood lipids and hemoglobin A1c to follow the pattern's effect.

The registered dietitian

Builds and monitors the individual eating plan. For this pattern that usually means adapting it to a diagnosed condition, working out how to hold vitamin K steady rather than avoiding greens, fitting it to a food budget, adapting it to a client's own food culture, and handling a fish or nut allergy without losing the pattern's shape.

The health and wellness coach

Finds out what drew the client to this pattern and what they are hoping for. Tests it against the life they are living: who cooks, what the budget is, who else eats at the table, how much time there is. Helps them pick a starting place small enough to hold. Notices what is working and what is quietly not. And holds the referral line when a question turns out to belong to someone else.

What a client bringing this diet needs

Most clients arrive with this pattern for one of 3 reasons: a clinician recommended it after a test result, they read that it is the best diet, or someone in the family has had a heart attack. The reason shapes the conversation. A client sent by a cardiologist needs help turning a recommendation into something they can cook. A client who read a ranking may not yet have a reason of their own. A frightened client may need the fear named before any plan will hold.

What the coaching conversation offers that a search engine cannot is the fit between this pattern and this life. The information is everywhere. What is not everywhere is a clear look at whether a person who works nights, cooks for a family of 5 and has never cooked a lentil can make this their own, and where they would start.

What to ask

• When you say Mediterranean, what does that look like on your plate?
• What made this the thing you wanted to try?
• Who else eats what you cook, and what would they make of it?

Referral red flags

A client on warfarin who mentions eating more salad or greens goes back to the prescriber before the change, not after. A client on insulin or a sulfonylurea who reports shakiness, sweating or confusion may be having a low blood sugar episode right now, and the first step is the one in their own diabetes plan for treating it, with emergency help if the symptoms are severe or they cannot manage the episode themselves. Telling the prescriber is the separate step that follows, so the medication can be reviewed. A client with kidney disease who is changing what they eat should keep their kidney care team informed, especially if they are on dialysis or their lab results have shown high potassium or phosphorus. A client on blood pressure medication who reports dizziness or lightheadedness should tell their prescriber, who can work out the cause; the coach does not guess at it. A client whose questions have moved from what to eat to exact grams, or who has started turning down meals out because nothing fits, is worth asking about it: what the precision is for, whether it causes distress, whether it is narrowing what they eat, and whether it is getting in the way of their life. There are ordinary reasons for all of it, from an allergy to a dietitian's plan. Where the answers raise a real concern about disordered eating, the coach follows their own referral process.

Special populations

Pregnancy and breastfeeding

Studied directly and favorably. A trial of 1,221 pregnancies at high risk in Barcelona found fewer small-for-gestational-age babies in the group supported to eat this pattern, 14.0 percent against 21.9 percent with usual care, although the authors called the findings preliminary and asked for replication (Crovetto et al., 2021). That trial looked at pregnancy outcomes and says nothing directly about breastfeeding. Fish during pregnancy raises the question of mercury. With the client's permission, a coach can share the federal advice on which fish are safe to eat and how often, which the FDA and EPA publish as a simple chart for people who are or might become pregnant or are breastfeeding (U.S. Food and Drug Administration & U.S. Environmental Protection Agency, 2024), and can explore how the client wants to use it. The individual plan, and any question about the client's own pregnancy, belongs to the obstetric clinician or dietitian.

Children and adolescents

Short trials in children have measured cardiometabolic markers such as blood lipids and blood pressure, but no trial has followed children long enough to test disease outcomes, which is a different and much harder question. The pattern is ordinary family food and carries no particular risk, but a coach works with the adult in front of them, and a child's eating belongs to the family and their pediatrician.

Older adults

The trial evidence comes largely from this group, since the participants were 55 to 80. Following the pattern more closely is also associated with a lower risk of becoming frail, across 4 studies of 5,789 older people (Kojima et al., 2018). Chewing difficulty, a smaller appetite and a fixed income are the practical obstacles, and protein needs rise with age, which is a dietitian's question.

Athletes and people with high activity

No particular problem, and the pattern supplies carbohydrate and unsaturated fat readily. A person training hard may need more total food and more protein than the traditional pattern provided, which is a matter of quantity rather than of changing the pattern.

People taking medication for diabetes, blood pressure, or blood thinning

This is the group where the referral is not optional. Warfarin, insulin, sulfonylureas and blood pressure medication can all be affected by a real change in eating. The standing rule is that a client on any of these lets the prescriber know before a substantial change in eating, not after a problem appears. That is the client keeping the prescriber informed, not asking permission for every food choice, and whether anything about the medication needs to change is the prescriber's call.

People with a history of disordered eating

This pattern is flexible: it removes no food group, counts nothing and forbids nothing. That does not make it safe for everyone, because any way of eating can be turned into a set of rules, and this one can be too. Rules about olive oil grams, or turning down food someone else cooked, are worth noticing and asking about, because the same behavior can come from an allergy, a religious practice or a clinician's plan. Distress, narrowing and interference with daily life are what raise a concern, and a concern goes to the client's clinician under the existing care plan.

Client misconceptions

Three come up most. That it is proven to prevent heart attacks, when the trial evidence on heart attack specifically is rated low in certainty and the largest trial had a serious flaw, even though the evidence on stroke is stronger than that. That wine is the healthy part, when the case for a benefit from drinking has weakened considerably. And that olive oil is the active ingredient, so pouring more of it on an unchanged diet will do the job, when the trials tested a whole pattern and the oil arrived alongside vegetables, beans, fish and much less red meat. Each is addressed in section 6.

🤝 COACHING CONNECTION

Dana comes in with a printout from her doctor and says she is supposed to be on the Mediterranean diet now. Her cholesterol came back high. She has bought olive oil and a bag of walnuts and is waiting to be told what else. The coach asks what the doctor said, and then what Dana makes of it. It turns out she cooks for her husband and her mother, both of whom expect meat at dinner, and she has been eating the walnuts in the car so nobody asks about them. The coach asks what a first change might look like that the 3 of them could eat. Dana says her mother likes the white bean soup she used to make and she had forgotten about it. The coach asks how that would fit in a week. Nothing was recommended and nothing was prescribed.

4. The evidence

(select the heading to open or close this section)

This pattern has been studied more than most, and the certainty of the evidence differs sharply from one outcome to the next and from one comparison to the next. Trials in people at high cardiovascular risk found fewer strokes and heart attacks, and the certainty is higher for stroke than for anything else on this page. Improvements in blood pressure, blood sugar and blood fats are small but consistent. On how long people live, the answer depends on what the pattern was compared against: measured against minimal advice, pooled trials of Mediterranean eating programs show fewer deaths; measured against an active low-fat diet, the question is unsettled. The observational research is far larger and points the same way throughout, but it cannot separate the eating from everything else that travels with it.

Evidence from randomized controlled trials

A trial assigns people to an eating pattern by chance and compares them with people assigned to something else, which is the strongest way to test whether the eating itself causes a result. Three trials carry most of the weight here, and a 2019 Cochrane systematic review pooled 30 trials covering 12,461 people (Rees et al., 2019). Nobody can be blinded to what they are eating, so every trial in this field is open-label, and that limitation runs through all of them.

Cardiometabolic outcomes

The largest trial enrolled 7,447 adults in Spain, aged 55 to 80, who had no cardiovascular disease but were at high risk of it. They were assigned to this pattern supplemented with free extra-virgin olive oil, to the pattern supplemented with free mixed nuts, or to advice to eat a low-fat diet. Over a median of 4.8 years, 288 people had a heart attack, a stroke or died of cardiovascular causes: 96 in the olive oil group, 83 in the nut group and 109 in the control group. Compared with the control diet, the hazard ratio was 0.69 for the olive oil group and 0.72 for the nut group, a reduction of roughly 30 percent in relative terms. The absolute picture comes in 2 forms and this trial reports both. Counting events against everyone enrolled, the raw proportions were 3.8 percent in the olive oil group, 3.4 percent in the nut group and 4.4 percent in the control group, a crude difference of about 1 event per 100 people across the median 4.8 years of follow-up. The trial's own 5-year absolute risks, estimated with survival methods that account for people entering and leaving the study at different times, were 3.6 percent, 4.0 percent and 5.7 percent, and the paper states the absolute difference as 1.7 to 2.1 percentage points, which is about 2 fewer events per 100 people over 5 years (Estruch et al., 2018). The published 5-year figure is the one to quote.

Both of those numbers are true and a coach should be able to say both. This trial is also the source of the most important caution on this page, and it is set out in the highlighted studies table below and again in section 6.

A second Spanish trial tested the pattern in people who already had coronary heart disease. It enrolled 1,002 patients at a single hospital and followed them for 7 years against a low-fat diet. There were 198 major cardiovascular events, 87 in the Mediterranean group and 111 in the low-fat group, a rate of 28.1 against 37.7 per 1,000 person-years. The hazard ratios across the statistical models ranged from 0.72 to 0.75 in favor of the Mediterranean pattern. The effect was clear among the 827 men. Among the 175 women the trial did not detect a difference, but a subgroup that size could not have detected a modest difference either, so that is an inconclusive result rather than evidence of no benefit in women (Delgado-Lista et al., 2022). Note also that this trial counted a broader set of events than the primary prevention trial above, adding revascularization and peripheral artery disease to heart attack, stroke and cardiovascular death, so the 2 trials' headline numbers are not measuring quite the same thing.

The oldest of the 3 is the Lyon Diet Heart Study, which enrolled 605 people in France who had survived a heart attack. Over a mean follow-up of 46 months, cardiac death and non-fatal heart attack occurred 14 times in the Mediterranean group and 44 times in the control group (de Lorgeril et al., 1999).

Set against those results, the 2019 Cochrane review rated every outcome separately, and its ratings differ. For prevention in people without heart disease, Cochrane's results on clinical events come from the Spanish primary prevention trial on its own, compared against an active low-fat diet. They are not pooled across trials. On that comparison Cochrane rated stroke moderate certainty, a hazard ratio of 0.60 from 0.45 to 0.80, which in their terms is a fall from about 24 strokes per 1,000 people to about 14. It rated peripheral artery disease moderate as well, 0.42 from 0.28 to 0.61, while noting that those figures come from earlier reports of the trial that were not reanalyzed after the randomization problem was found, which makes them less certain. It rated heart attack low, 0.79 from 0.57 to 1.10; death from cardiovascular causes low, 0.81 from 0.50 to 1.32; and death from any cause low, 1.00 from 0.81 to 1.24. The outcomes rated low were lowered both for the randomization problem in that trial and for imprecision, meaning their intervals are wide enough to include a worthwhile benefit and a worthwhile harm at the same time (Rees et al., 2019). An inconclusive result of that kind is not a finding of no effect, and reading it as one is the most common error made with this review.

For blood pressure the picture is steadier, and here too the rating depends on the comparison. Against no advice or minimal advice, Cochrane rated the evidence moderate: systolic pressure 2.99 mmHg lower, from 3.45 to 2.53 lower, in 269 people across 2 trials, lowered one level only because that is a small number of participants. Against another active diet it rated the evidence low: systolic 1.5 mmHg lower in 448 people across 4 trials, with an interval from 3.92 lower to 0.92 higher that includes no difference (Rees et al., 2019). A larger synthesis outside Cochrane, covering 19 trials and 4,137 people, found reductions of 1.4 mmHg systolic and 1.5 mmHg diastolic against control diets, with larger reductions in longer trials and in people whose blood pressure started higher; that synthesis assigned no certainty rating of its own (Cowell et al., 2021). All 3 estimates favor the pattern and all describe a change of a couple of mmHg, but of Cochrane's 2 comparisons only the one against minimal advice excludes no difference. For blood sugar, a synthesis of 9 trials in 1,178 people with type 2 diabetes found hemoglobin A1c 0.30 percentage points lower than on control diets, along with small reductions in fasting glucose, blood fats and blood pressure (Huo et al., 2015).

Fewer people developed type 2 diabetes on the pattern in the one trial that measured it, but that result is not conclusive and its certainty is low. The trial evidence comes from the 3,541 participants in the Spanish primary prevention trial who did not have diabetes at the start (Salas-Salvado et al., 2014). A 2025 review took the 2 Mediterranean groups together against the low-fat control and reported a hazard ratio of 0.75, from 0.56 to 1.01, which is about 21 fewer cases per 1,000 people with a range running from 38 fewer to 1 more. It rated that low certainty, lowered once for risk of bias and once because the interval crosses no effect (Wallerer et al., 2025). The same review pooled 24 cohort studies covering 988,337 people and 68,052 new cases of diabetes and rated that evidence moderate: 8 percent lower risk for every 2 points of adherence, and 21 percent lower comparing an adherence score of 8 with a score of 1. A later Spanish trial of 4,746 adults with metabolic syndrome found that adding calorie reduction and physical activity to the pattern lowered diabetes incidence further over 6 years. The paper's absolute risk figures were corrected after publication, so they are not quoted here. That trial tested a different intervention and should not be read as evidence about the eating pattern by itself (Ruiz-Canela et al., 2025).

Cancer

Trial evidence on cancer is thin and should be treated carefully. Among the 4,152 women in the Spanish primary prevention trial without previous breast cancer, 35 cases of breast cancer occurred over a median of 4.8 years, at 1.1, 1.8 and 2.9 cases per 1,000 person-years in the olive oil, nut and control groups. In the corrected version of that analysis, the hazard ratio for the olive oil group was 0.31 against control, from 0.13 to 0.77 (Toledo et al., 2015). Cancer was a secondary outcome there, the whole finding rests on 35 events, and the authors wrote that the result needs confirmation in longer and larger studies. At least 1 randomized trial has taken a cancer outcome as its main question. DIANA-5 assigned 1,542 women at high risk of breast cancer recurrence to a Mediterranean and macrobiotic eating program or to a control group and followed them for 5 years. Recurrence occurred in 95 women in the intervention group and 98 in the control group, a hazard ratio of 0.99 from 0.69 to 1.40, so the trial did not demonstrate a reduction, though an interval that wide cannot rule one out either (Berrino et al., 2024). Its diet excluded dairy products and alcohol and is not the pattern described on this page, which limits what it settles, but it is the closest thing to a direct test found for this page.

Cognitive outcomes

This outcome is included because clients and students meet the memory claim constantly, and because the MIND Diet page in this archive links here. Within the Spanish trial, 447 people in Barcelona took cognitive tests at the start and 334 of them again after a median of 4.1 years. Those on the pattern with olive oil scored better on 2 of the composite measures than the control group, whose scores declined (Valls-Pedret et al., 2015). Against that, a trial designed specifically to test a brain-focused version of this pattern randomized 604 older adults with a family history of dementia to the MIND diet or a control diet for 3 years. Both groups ate slightly fewer calories and both received counseling, so the comparison was against an active diet, not against doing nothing. Cognition improved in both groups, the difference between them was small and not statistically significant, and brain imaging showed no difference either (Barnes et al., 2023). Anyone repeating the memory claim should know about both.

Body weight, as measured

This archive reports body weight as something a study measured, in that study's own numbers, and not as a goal or as a measure of whether an eating pattern worked. See the standing rule in the review record at the foot of this page.

In the Spanish primary prevention trial, where people were not asked to eat less and physical activity was not promoted, all 3 groups showed slightly lower body weight and slightly larger waist measurements after 5 years. Compared with the control group, the adjusted difference in body weight was 0.410 kg lower in the olive oil group, from 0.830 lower to 0.010 higher, and 0.016 kg lower in the nut group, from 0.453 lower to 0.421 higher. Neither reached statistical significance. The adjusted difference in waist circumference was 0.466 cm smaller in the olive oil group, from 1.109 smaller to 0.176 larger, and 0.923 cm smaller in the nut group, from 1.604 to 0.241 smaller, and only the nut result cleared significance (Estruch et al., 2019). That paper is the 2019 reanalysis, and the change from its withdrawn 2016 version is instructive: the same weight comparison was reported then as 0.43 kg lower with an interval that excluded zero, and the reanalysis moved it to an interval that includes zero. In a separate 2-year trial of 322 adults comparing 3 assigned diets, mean weight change was 2.9 kg lower on the low-fat diet, 4.4 kg lower on the Mediterranean diet and 4.7 kg lower on the low-carbohydrate diet (Shai et al., 2008). That trial and the one above are answering different questions, and their numbers should not be merged into a single judgment.

Mortality

Mortality is 2 questions, not 1, and they have different answers.

The first question is what happens compared with minimal advice, meaning a pamphlet, usual care or limited dietary advice. Here the trials show fewer deaths. A 2023 network analysis of 40 trials covering 35,548 people at increased cardiovascular risk found an odds ratio of 0.72, from 0.56 to 0.92, for death from any cause on Mediterranean dietary programs, and rated that moderate certainty. For a person at intermediate risk that works out to about 17 fewer deaths per 1,000 people over 5 years. Death from cardiovascular causes came out at 0.55, from 0.39 to 0.78, about 13 fewer per 1,000 (Karam et al., 2023). One qualification travels with that result and must not be dropped: these were programs, not food alone. They could include exercise, behavioral support and, in some trials, medication, so the benefit belongs to the program rather than to the groceries.

The second question is what happens compared with an active low-fat diet, and there the answer is unsettled. In the Spanish primary prevention trial, death from any cause came out at a hazard ratio of 1.00 with an interval from 0.81 to 1.24, which Cochrane rated low certainty (Rees et al., 2019). The 2023 analysis likewise found no convincing difference between Mediterranean and low-fat programs for death or for non-fatal heart attack. Both did better than minimal advice; neither clearly did better than the other.

One individual trial did find fewer deaths. In the Lyon Diet Heart Study, 14 people in the Mediterranean group died against 24 in the control group, an adjusted risk ratio of 0.44 from 0.21 to 0.94, and there were 6 cardiac deaths against 19, a risk ratio of 0.35 from 0.15 to 0.83 (de Lorgeril et al., 1999). Cochrane rated that evidence low certainty, lowering it 2 levels for risk of bias, because the trial's method of randomization is unclear and its modified Zelen design may have introduced other biases (Rees et al., 2019). It is also a single trial in 605 people who had already had a heart attack, and its diet used a canola oil margarine rather than olive oil, both of which bear on how far the result carries. All of that is worth saying. What is not correct is the flat claim that no trial has shown this pattern to lengthen life, because this one did.

Evidence from long-term prospective cohort studies

A cohort study follows people who already eat a certain way and records what happens to them over years. It can show that a pattern goes with an outcome. It cannot by itself show that the pattern caused it, because people who eat this way also tend to differ in income, education, smoking, exercise and much else. In this literature the cohorts are very large, the follow-up is long, and how closely a person follows the pattern is scored with a questionnaire, usually out of 9 or 14 points.

Cardiometabolic outcomes

A synthesis commissioned for European diabetes guidelines pooled 38 cohort studies covering 1,209,956 people free of cardiovascular disease at the start, with follow-up from 2 to 26 years. Comparing the people who followed the pattern most closely with those who followed it least, cardiovascular mortality was 21 percent lower, coronary heart disease incidence 27 percent lower, stroke incidence 20 percent lower and heart attack incidence 27 percent lower (Becerra-Tomas et al., 2020). These are associations, and the size of them is consistent across many separate populations, which is part of why the pattern is taken seriously despite the limits of the trials.

Cancer

A synthesis of 117 studies covering 3,202,496 people, searched to April 2020, reports its results in 2 forms, and the difference matters. From cohort studies, the people following the pattern most closely had 13 percent lower cancer mortality, from 0.82 to 0.92, across 18 studies; 8 percent lower colorectal cancer risk, from 0.87 to 0.99, across 10 studies; and 16 percent lower respiratory cancer risk. Among people who had already had cancer, all-cause mortality was 25 percent lower, from 0.66 to 0.86, across 8 studies. For breast cancer the cohort estimate was 0.97, from 0.94 to 1.00, across 12 studies, which is close to no difference. The same paper also pools cohort studies together with case-control studies, which is a weaker basis because case-control studies ask people to recall how they used to eat after they are already ill, and those mixed pools give larger numbers: 17 percent lower for colorectal cancer, 6 percent for breast, and lower risks for head and neck, gastric, liver and bladder cancer. When a larger figure from this paper is quoted, it is usually one of those mixed pools. No clear association was found for prostate, pancreatic, esophageal or blood cancers. The reviewers rated their certainty with NutriGrade, which is a scoring system built for nutrition research and is not the same tool as GRADE, and they rated the cohort evidence moderate for cancer mortality and for colorectal cancer and low or very low for most of the rest, including breast cancer (Morze et al., 2021).

Body weight, as measured

No cohort synthesis of body weight was retrieved and read for this page, so none is reported here. The trial evidence above is the better source for this question in any case, because it compares people assigned to an eating pattern rather than people who already differ from one another in many ways.

Mortality

This is where the cohort evidence is largest and where it diverges most sharply from the trials. A dose-response synthesis of 29 studies covering 1,676,901 people and 221,603 deaths found that each 2-point rise in an adherence score went with 10 percent lower death from any cause, in a steady line rather than at a threshold. The association was stronger in people living in Mediterranean countries than elsewhere, and the reviewers rated the evidence low quality (Soltani et al., 2019). An earlier synthesis found an 8 percent lower risk of death for each 2-point rise, much the same size of association; its often-quoted total of 4,172,412 people covers every health outcome it examined, not mortality alone (Sofi et al., 2014).

One study is worth reading closely because of what happened when the researchers adjusted their numbers. Following 25,315 initially healthy women in the United States for a mean of 24.7 years, during which 3,879 died, those in the highest adherence group had 23 percent lower mortality than those in the lowest. When the analysis was further adjusted for other lifestyle factors, that gap narrowed to 11 percent, still statistically significant but much smaller (Ahmad et al., 2024). The shrinkage is the lesson. A good part of what looks like an effect of the eating is the company the eating keeps.

Highlighted studies

These 4 sources carry the most weight for this pattern. Each is described so that a reader can judge it without opening the paper. The limitation column gives the one weakness that most affects how far the finding can be trusted.

Highlighted studies for the Mediterranean Diet

Study	Design	Sample size	Duration	Finding	The limitation that matters
Estruch et al., 2018, the Spanish primary prevention trial known as PREDIMED	Randomized controlled trial, primary prevention, open-label	7,447 adults aged 55 to 80 at high cardiovascular risk	Median 4.8 years, stopped early for benefit	288 heart attacks, strokes or cardiovascular deaths. About 30 percent fewer in each Mediterranean group than in the control group. The trial's 5-year absolute risks were 3.6, 4.0 and 5.7 percent, an absolute difference of 1.7 to 2.1 percentage points, or about 2 fewer events per 100 people over 5 years	Two problems. The first is widely reported and the second rarely is. The assignment to groups failed at some clinics, the first report was withdrawn in 2018, and the data were re-examined and published again without assuming everyone had been assigned by chance. Beyond that, the groups' diets differed less than the labels suggest: the main difference was the free olive oil and nuts, in a population already eating close to this pattern, while the control group was also advised to eat healthily. The authors themselves question how far the result carries to high-risk people in other countries
Delgado-Lista et al., 2022, the CORDIOPREV trial	Randomized controlled trial, secondary prevention, open-label	1,002 adults with established coronary heart disease	7 years	198 major cardiovascular events, 87 on the Mediterranean diet against 111 on a low-fat diet, a reduction of roughly a quarter to a third	One hospital in one Spanish city, and participants knew which diet they were on. 82.5 percent of the participants were men, and while the benefit was clear among them, no difference was found among the 175 women. Whether the pattern helps women after a cardiac event is not answered by this trial
de Lorgeril et al., 1999, the Lyon Diet Heart Study	Randomized controlled trial, secondary prevention	605 adults who had survived a heart attack	Mean 46 months	14 cardiac deaths or non-fatal heart attacks in the Mediterranean group against 44 in the control group	The two groups were not treated alike beyond the food. By the trial's own account, participants were not fully told that 2 diets were being compared, and only the Mediterranean group received dietary counseling from the investigators, while the control group was left to advice from their own physicians. The method of assigning people to groups is also unclear. The intervention supplied a canola margarine, so it was not purely a change of eating pattern
Rees et al., 2019, the Cochrane systematic review	Systematic review and meta-analysis of randomized trials	30 trials, 12,461 people	Trials of 3 months or longer, searched to September 2018	Certainty rated one outcome at a time rather than as a single verdict. For prevention in people without heart disease, compared against another diet: stroke and peripheral artery disease moderate, heart attack and death from cardiovascular causes low, death from any cause low and inconclusive. Blood pressure moderate against minimal advice and low against another diet	The review is only as good as the trials in it, and it says so. Two secondary prevention trials covering 1,406 people were left out of the main analysis because the reliability of their data had been publicly questioned, which thinned the secondary prevention evidence considerably. Definitions of the pattern also vary between trials, so the reviewers had to apply their own definition to decide what counted

Head to head

This pattern has been tested directly against other diets more often than most. Both large Spanish trials used a low-fat diet as the comparison, which means their results say this pattern beat low-fat advice, not that it beat every alternative. A 2-year trial of 322 adults compared it head to head with a low-fat diet and a low-carbohydrate diet and found the low-carbohydrate group lost the most weight and the Mediterranean group the most improvement in blood sugar among the participants who had diabetes (Shai et al., 2008). A network comparison of 121 trials covering 21,942 adults found that most named diets produce similar modest improvements at 6 months, and that by 12 months the benefits for cardiovascular risk factors had largely disappeared for every diet examined except this one (Ge et al., 2020). The DASH Diet page in this archive carries the comparison from the blood pressure side.

Where the evidence is absent

Against an active low-fat diet, no single trial has shown an effect on how long people live, and none was designed to answer that question. The pooled comparison against minimal advice does show an effect, but it tests whole programs rather than food alone. The trial evidence comes almost entirely from Spain, France and Italy, in people at high cardiovascular risk and mostly over 55, which leaves the effect in younger people, in healthy people and in populations outside the Mediterranean largely untested. Women are underrepresented in the secondary prevention trials, and the secondary prevention trial that reported a female subgroup had too few women in it to detect a modest difference in either direction. The one randomized trial found for this page that took a cancer outcome as its main question tested a diet that is not the pattern described here. No trial has followed people who adopted the pattern and then stopped. The contributions of the individual parts, the olive oil against the vegetables against the fish against the absence of red meat, have not been separated, so a claim about any one component does not follow from a trial of the whole pattern. And there is no agreed definition of the pattern itself, which means 2 studies using the name may be testing meaningfully different things.

5. Mechanisms

(optional reading; select the heading to open or close this section)

This section explains how the diet is thought to work in the body. You can skip it. Nothing in the coaching sections depends on it.

Established mechanisms

The clearest one is a change in the kind of fat eaten rather than the amount. This pattern is not low in fat. What it does is replace the hard fats of meat, butter and cream with the softer fats of olive oil, nuts and fish, which lowers low-density lipoprotein cholesterol, the form of cholesterol that builds up in artery walls. Replacing saturated fat with unsaturated fat is one of the 9 features of a heart-healthy pattern named by the American Heart Association, and it is among the best established relationships in nutrition (Lichtenstein et al., 2026).

The second is fiber, the part of plant food the body cannot digest. Vegetables, fruit, beans and whole grains supply a great deal of it. Fiber slows how quickly sugar enters the blood after a meal, feeds the bacteria in the large intestine, and lowers cholesterol. Eating patterns rich in fiber are associated with lower risk of cardiovascular disease, type 2 diabetes and colorectal cancer (Lichtenstein et al., 2026).

The third is blood pressure, which falls modestly and consistently on this pattern in trials, as reported in section 4. Eating more vegetables and fruit raises potassium intake and cooking from raw ingredients lowers sodium intake, and both of those move blood pressure in the same direction.

A fourth mechanism is simply nutritional adequacy. A review of the evidence on whether this pattern supplies enough of what people need found that closer adherence went with a better nutrient profile in both children and adults, and with fewer people falling short on micronutrients (Castro-Quezada et al., 2014). Adequacy and adherence go together in that review, though it does not show that one causes the other. Whether a particular person needs a supplement is a question for their clinician or dietitian, not something the pattern settles.

Proposed mechanisms

Several further explanations are plausible and widely repeated, and none of them supports a claim in section 4. The most discussed is inflammation. Olive oil, vegetables, fruit, nuts and whole grains contain compounds called polyphenols, and the proposal is that these lower the low-level inflammation involved in artery disease. That is the proposal, and it is all this page claims. No measured result on inflammatory markers is reported here, because the sub-studies that report them were not examined for this page, and a mechanism labeled proposed cannot carry a number.

A second is the gut microbiome, the community of bacteria in the large intestine. Fiber that reaches the colon is fermented by those bacteria into short-chain fatty acids, and the proposal is that these affect appetite signaling, the gut lining and inflammation. Much of this work is in animals or in short human studies.

A third concerns the lining of the blood vessels and the tendency of blood to clot. A fourth, drawn mostly from animal work, concerns the lower intake of animal protein on this pattern and its effect on growth signaling pathways that are relevant to cancer. Each of these is a reasonable idea under investigation. None is a reason to tell a client that this pattern will do something for them.

6. Risks, contraindications and myths

(select the heading to open or close this section)

Nutrient gaps

Because it excludes no food group, a well-planned version of this pattern can supply a broad range of nutrients. The review of nutritional adequacy cited in section 5 found that closer adherence went with better nutrient intake rather than worse (Castro-Quezada et al., 2014). Three qualifications are worth knowing. Calcium intake can be low in versions that keep dairy to a minimum, since the traditional pattern used cheese and yogurt in modest amounts rather than milk by the glass. Iron from plant sources is absorbed less readily than iron from meat, which matters most for people who menstruate. And a person who drops red meat without adding beans, fish or eggs in its place can end up short on protein, which is a construction error rather than a property of the pattern.

Who should not follow it without medical supervision

Anyone taking warfarin, because the pattern raises intake of the leafy greens that supply vitamin K, and warfarin works by opposing vitamin K. The issue is not that greens are forbidden but that the amount has to stay steady, and the dose is set against it (NIH Office of Dietary Supplements, 2021). Anyone with chronic kidney disease, because how much potassium and phosphorus is safe depends on the stage of the disease, lab results and treatment. That does not make the foods at the center of this pattern off limits. The National Kidney Foundation says most people with chronic kidney disease who are not on dialysis, and most transplant recipients, do not need to limit beans because of potassium or phosphorus, partly because the body absorbs these minerals less well from plant foods than from animal foods or food additives. People on dialysis work out their intake with a kidney dietitian (National Kidney Foundation, 2023). Anyone with a fish or tree nut allergy, since both sit near the center of the pattern and a safe version has to be built deliberately. And anyone taking medication for blood pressure or for diabetes, because the pattern can lower both, and the prescriber is the one to judge whether the medication still fits.

Medication interactions

A coach never adjusts medication. The point of knowing these is to know when to send a client back to the prescriber. Warfarin is the clearest case and the interaction is well documented. Insulin and the sulfonylurea drugs used for type 2 diabetes can cause blood sugar to fall too low when eating improves, which is a reason to keep the prescriber informed rather than a reason to eat worse. Blood pressure medication raises a similar question as blood pressure falls. If a client on it reports feeling lightheaded, the coach's job is to encourage them to tell the prescriber, not to guess the cause. Alcohol also interacts with a long list of medications. That does not put the wine question off limits to a coach. With the client's permission, a coach can share what current guidance says, as set out below, and explore what the client makes of it. What goes to the prescriber is any question about how alcohol fits with that client's own medications, conditions or drinking.

Signs it is going wrong

Physical signs a coach or a family member might notice are dizziness or lightheadedness on standing in someone on blood pressure medication; shakiness, sweating or confusion in someone on diabetes medication; and unintended weight loss in an older adult. A coach notices these and encourages the client to tell their clinician; working out the cause is the clinician's job. Behavior matters as much, but it has to be read with care. Weighing olive oil or turning down food other people cooked can come from an allergy, celiac disease, a religious practice, a cultural preference or a dietitian's plan, so on its own it tells a coach little. Talking about foods as clean and dirty, or treating a meal out as a failure, carries more meaning and is worth asking about directly. In every case the questions are the same: what the behavior is for, whether it causes distress, whether it is narrowing what the person eats, and whether it is getting in the way of daily life. None of this is a diagnosis. Where the answers raise a real concern about disordered eating, the coach follows their own referral process.

Wine, and what the evidence now says

Wine is part of how the Mediterranean pattern has been described since it was first written down. The 1995 food pyramid that introduced the pattern to an American audience lists wine in low to moderate amounts, normally taken with meals, and the 14-item questionnaire the Spanish trial used to score adherence gave participants a point for drinking 7 or more glasses of wine a week. Leaving wine out would misdescribe what was studied.

What has changed is the health claim attached to it. For years, large observational studies appeared to show that people who drank a little lived longer than people who drank nothing at all. That finding has weakened considerably under closer examination, for 3 separate reasons. The comparison groups were part of the problem, because the people counted as non-drinkers often included those who had quit drinking after becoming ill, which made the drinkers look healthier by comparison. When researchers corrected for that and for study quality across 87 studies, the apparent benefit disappeared (Stockwell et al., 2016). Income, education and other habits that travel with moderate drinking explain more of the remaining difference than the alcohol itself does. And studies that compare people by inherited differences in how they process alcohol, an approach called mendelian randomization, are not distorted by lifestyle in the same way. Those studies looked for a protective effect on coronary artery disease and did not find one (Lichtenstein et al., 2026).

The harms, meanwhile, are better established than they used to be. Blood pressure rises with alcohol along a steady line that starts at the lowest intakes, with no threshold below which it stays flat. The American Heart Association's 2026 dietary guidance names alcohol as one of the 9 features of a heart-healthy eating pattern, in these words: if alcohol is not consumed, do not start; if alcohol is consumed, limit intake. The World Health Organization and the US Department of Health and Human Services both advise avoiding alcohol to lower the risk of several cancers, among them cancers of the mouth, esophagus, breast, liver and colon (Lichtenstein et al., 2026). The Surgeon General's advisory on alcohol and cancer risk reports that alcohol contributes to nearly 100,000 cancer cases and about 20,000 cancer deaths a year in the United States, and that for breast, mouth and throat cancers the risk may begin to rise at around 1 drink a day or fewer (Office of the Surgeon General, 2025).

Clients may also quote a federal number, and which number they quote depends on which edition they met. The 2025-2030 Dietary Guidelines for Americans state no numeric drink limit at all. They say to consume less alcohol for better overall health, and they keep a list of circumstances in which alcohol should be avoided completely (U.S. Department of Agriculture & U.S. Department of Health and Human Services, 2026). Earlier editions did carry specific daily limits, and federal web pages, printed handouts and textbooks built on those editions may still show them. When the federal alcohol guidance page at dietaryguidelines.gov was checked in September 2026 it still displayed the older figures of 2 drinks a day for a man and 1 for a woman, so a client can easily arrive holding a number the current edition no longer states. This program does not rest a health claim on either version. They are named here only so a coach recognizes what a client is working from, and if a specific limit matters for a client, the current edition is the thing to open rather than a remembered figure.

Among the current guidance documents this page examined, none recommends that a person who does not drink should start drinking for their health, and the American Heart Association says so in as many words.

Myths and where they came from

Myth: it is proven to prevent heart attacks

Where it came from: the headline result of the large Spanish trial, reported worldwide in 2013 as a roughly 30 percent reduction, repeated ever since in press coverage and in textbooks. What is true: that trial's 2013 report was withdrawn in 2018 because people had not all been assigned to groups by chance, and the data were re-examined and published again, with a similar result. Note that the 2013 report was withdrawn, not the trial itself, and the republished paper is the one to cite. Independent reviewers then rated that trial's outcomes one at a time: stroke moderate certainty, heart attack low, death from cardiovascular causes low and inconclusive. The pattern is well supported as a sensible way to eat, and the stroke finding is the strongest single result behind it. Proven to prevent heart attacks is still more than the evidence says.

Myth: the wine is the healthy part

Where it came from: decades of observational studies that appeared to show lower mortality among moderate drinkers, amplified by wine industry marketing and by the popular story of the French paradox. What is true: as set out above, the apparent benefit largely disappears when the comparison groups are corrected and other advantages are accounted for. The American Heart Association's 2026 dietary guidance states that if alcohol is not consumed, a person should not start, and that if it is consumed, intake should be limited (Lichtenstein et al., 2026). The pattern works without it.

Myth: olive oil is the active ingredient

Where it came from: the largest trial supplied free extra-virgin olive oil and reported its best result in that group, and olive oil marketing has leaned on this ever since. What is true: the trials tested a whole pattern in which the oil arrived alongside more vegetables, more beans, more fish and much less red meat. The separate contributions have never been untangled, and the trial authors themselves noted that their groups differed mainly in the supplied oil and nuts while eating otherwise similar diets. Adding olive oil to an unchanged diet has not been tested and is not what these trials showed.

7. Key terms

Dietary pattern

The whole of what a person eats, rather than a single nutrient or food. Nutrition guidance moved toward patterns because foods are eaten together and interact, and because people eat meals rather than nutrients.

Adherence score

A questionnaire that converts what a person eats into a number showing how closely they follow a pattern. Several exist for this pattern, scored out of 9, 14 or 18 points, and they do not all measure the same thing, which is one reason studies are hard to compare.

Primary prevention and secondary prevention

Primary prevention means preventing a first heart attack or stroke in someone who has not had one. Secondary prevention means preventing another one in someone who has. The two questions need different trials and the answers can differ.

Hazard ratio

A way of comparing how quickly an event happens in 2 groups. A hazard ratio of 0.70 means events occurred at about 70 percent of the rate in the comparison group, which is a 30 percent relative reduction. It says nothing on its own about how common the event was, which is why absolute numbers belong beside it.

Relative and absolute risk

A 30 percent relative reduction sounds larger than about 2 fewer events per 100 people over 5 years, and in the largest trial here they describe the same result. Both belong in any honest description. Absolute numbers also come in 2 kinds, and they are not interchangeable: a crude count divides the events by everyone enrolled, while a survival estimate fixes a time point, such as 5 years, and accounts for people who entered late or left early. In that trial the crude count gives about 1 per 100 and the published 5-year estimate gives about 2 per 100. Say which kind you are quoting, and prefer the published one.

Retraction and republication

When a serious problem is found in a published paper, the journal withdraws it. If the underlying work can be salvaged by a corrected analysis, the authors may publish a new version. That happened to the largest trial of this pattern in 2018, and it is why the year on the citation matters.

GRADE

The 4-level scale used in clinical guidelines to say how certain the evidence for a conclusion is: high, moderate, low or very low. It rates the evidence, not the diet, and it rates each outcome separately.

Mendelian randomization

A study design that compares people according to inherited genetic differences, which are settled before birth and so are not shaped by income, education or habits. It is used to test whether an association seen in ordinary observational studies reflects cause.

8. Review questions

1. Describe this eating pattern to someone who has never heard of it, in 4 sentences, without using the word healthy. Then name the 2 parts of it that cost the most money and the 2 that cost the least.
2. This question draws on section 4, The evidence. The largest trial of this pattern reported a 30 percent relative reduction and an absolute difference of 1.7 to 2.1 percentage points over 5 years. Explain how both can be true, who was in that trial, how long it ran, and why the crude event count from the same trial gives a smaller absolute number than its published 5-year estimate.
3. A client says her cardiologist put her on the Mediterranean diet and she wants to know if she is doing it right. Write what you would say and what you would ask. Then say where the referral line is, and name one thing about her medications you would want to know.
4. This question draws on section 4 and on the evidence rating table near the end of the page. Independent reviewers rated this pattern's effect on stroke moderate in certainty and its effect on heart attack low, using the same trial and the same comparison. Explain to a classmate how one trial can support 2 different ratings, and why that is not a sign of sloppy reviewing.
5. Where did the belief that wine is the healthy part of this pattern come from, and what has changed? Write the 2 sentences you would use with a client who raises it.

9. References and Resources

Ahmad, S., Moorthy, M. V., Lee, I. M., Ridker, P. M., Manson, J. E., Buring, J. E., Demler, O. V., & Mora, S. (2024). Mediterranean diet adherence and risk of all-cause mortality in women. JAMA Network Open, 7(5), e2414322. https://doi.org/10.1001/jamanetworkopen.2024.14322

Aleksandrowicz, L., Green, R., Joy, E. J. M., Smith, P., & Haines, A. (2016). The impacts of dietary change on greenhouse gas emissions, land use, water use, and health: A systematic review. PLOS ONE, 11(11), e0165797. https://doi.org/10.1371/journal.pone.0165797

Barnes, L. L., Dhana, K., Liu, X., Carey, V. J., Ventrelle, J., Johnson, K., Hollings, C. S., Bishop, L., Laranjo, N., Stubbs, B. J., Reilly, X., Agarwal, P., Zhang, S., Grodstein, F., Tangney, C. C., Holland, T. M., Aggarwal, N. T., Arfanakis, K., Morris, M. C., & Sacks, F. M. (2023). Trial of the MIND diet for prevention of cognitive decline in older persons. New England Journal of Medicine, 389(7), 602-611. https://doi.org/10.1056/NEJMoa2302368

Becerra-Tomas, N., Blanco Mejia, S., Viguiliouk, E., Khan, T., Kendall, C. W. C., Kahleova, H., Rahelic, D., Sievenpiper, J. L., & Salas-Salvado, J. (2020). Mediterranean diet, cardiovascular disease and mortality in diabetes: A systematic review and meta-analysis of prospective cohort studies and randomized clinical trials. Critical Reviews in Food Science and Nutrition, 60(7), 1207-1227. https://doi.org/10.1080/10408398.2019.1565281

Berrino, F., Villarini, A., Gargano, G., Krogh, V., Grioni, S., Bellegotti, M., Venturelli, E., Raimondi, M., Traina, A., Zarcone, M., Amodio, R., Mano, M. P., Johansson, H., Panico, S., Santucci de Magistris, M., Barbero, M., Gavazza, C., Mercandino, A., Consolaro, E., ... Bruno, E. (2024). The effect of diet on breast cancer recurrence: The DIANA-5 randomized trial. Clinical Cancer Research, 30(5), 965-974. https://doi.org/10.1158/1078-0432.CCR-23-1615

Castro-Quezada, I., Roman-Vinas, B., & Serra-Majem, L. (2014). The Mediterranean diet and nutritional adequacy: A review. Nutrients, 6(1), 231-248. https://doi.org/10.3390/nu6010231

Conrad, Z., Korol, M., DiStaso, C., & Wu, S. (2024). Greater adherence to the Mediterranean diet pattern in the United States is associated with sustainability trade-offs. Nutrition Journal, 23, Article 159. https://doi.org/10.1186/s12937-024-01050-z

Cowell, O. R., Mistry, N., Deighton, K., Matu, J., Griffiths, A., Minihane, A. M., Mathers, J. C., Shannon, O. M., & Siervo, M. (2021). Effects of a Mediterranean diet on blood pressure: A systematic review and meta-analysis of randomized controlled trials and observational studies. Journal of Hypertension, 39(4), 729-739. https://doi.org/10.1097/HJH.0000000000002667

Crovetto, F., Crispi, F., Casas, R., Martin-Asuero, A., Borras, R., Vieta, E., Estruch, R., & Gratacos, E. (2021). Effects of Mediterranean diet or mindfulness-based stress reduction on prevention of small-for-gestational age birth weights in newborns born to at-risk pregnant individuals: The IMPACT BCN randomized clinical trial. JAMA, 326(21), 2150-2160. https://doi.org/10.1001/jama.2021.20178

de Lorgeril, M., Salen, P., Martin, J. L., Monjaud, I., Delaye, J., & Mamelle, N. (1999). Mediterranean diet, traditional risk factors, and the rate of cardiovascular complications after myocardial infarction: Final report of the Lyon Diet Heart Study. Circulation, 99(6), 779-785. https://doi.org/10.1161/01.CIR.99.6.779

Delgado-Lista, J., Alcala-Diaz, J. F., Torres-Pena, J. D., Quintana-Navarro, G. M., Fuentes, F., Garcia-Rios, A., Ortiz-Morales, A. M., Gonzalez-Requero, A. I., Perez-Caballero, A. I., Yubero-Serrano, E. M., Rangel-Zuniga, O. A., Camargo, A., Rodriguez-Cantalejo, F., Lopez-Segura, F., Badimon, L., Ordovas, J. M., Perez-Jimenez, F., & Lopez-Miranda, J. (2022). Long-term secondary prevention of cardiovascular disease with a Mediterranean diet and a low-fat diet (CORDIOPREV): A randomised controlled trial. The Lancet, 399(10338), 1876-1885. https://doi.org/10.1016/S0140-6736(22)00122-2

Estruch, R., Ros, E., Salas-Salvado, J., Covas, M. I., Corella, D., Aros, F., Gomez-Gracia, E., Ruiz-Gutierrez, V., Fiol, M., Lapetra, J., Lamuela-Raventos, R. M., Serra-Majem, L., Pinto, X., Basora, J., Munoz, M. A., Sorli, J. V., Martinez, J. A., Fito, M., Gea, A., ... Martinez-Gonzalez, M. A. (2018). Primary prevention of cardiovascular disease with a Mediterranean diet supplemented with extra-virgin olive oil or nuts. New England Journal of Medicine, 378(25), e34. https://doi.org/10.1056/NEJMoa1800389

Estruch, R., Martinez-Gonzalez, M. A., Corella, D., Salas-Salvado, J., Fito, M., Chiva-Blanch, G., Fiol, M., Gomez-Gracia, E., Aros, F., Lapetra, J., Serra-Majem, L., Pinto, X., Buil-Cosiales, P., Sorli, J. V., Munoz, M. A., Basora-Gallisa, J., Lamuela-Raventos, R. M., Serra-Mir, M., & Ros, E. (2019). Effect of a high-fat Mediterranean diet on bodyweight and waist circumference: A prespecified secondary outcomes analysis of the PREDIMED randomised controlled trial. The Lancet Diabetes & Endocrinology, 7(5), e6-e17. https://doi.org/10.1016/S2213-8587(19)30074-9

Ge, L., Sadeghirad, B., Ball, G. D. C., da Costa, B. R., Hitchcock, C. L., Svendrovski, A., Kiflen, R., Quadri, K., Kwon, H. Y., Karamouzian, M., Adams-Webber, T., Ahmed, W., Damanhoury, S., Zeraatkar, D., Nikolakopoulou, A., Tsuyuki, R. T., Tian, J., Yang, K., Guyatt, G. H., & Johnston, B. C. (2020). Comparison of dietary macronutrient patterns of 14 popular named dietary programmes for weight and cardiovascular risk factor reduction in adults: Systematic review and network meta-analysis of randomised trials. BMJ, 369, m696. https://doi.org/10.1136/bmj.m696

Healio. (2025, January 3). Mediterranean diet named best overall for 8th straight year. healio.com/news/primary-care

Huo, R., Du, T., Xu, Y., Xu, W., Chen, X., Sun, K., & Yu, X. (2015). Effects of Mediterranean-style diet on glycemic control, weight loss and cardiovascular risk factors among type 2 diabetes individuals: A meta-analysis. European Journal of Clinical Nutrition, 69(11), 1200-1208. https://doi.org/10.1038/ejcn.2014.243

Kahleova, H., Sutton, M., Maracine, C., Nichols, D., Monsivais, P., Holubkov, R., & Barnard, N. D. (2024). Food costs of a low-fat vegan diet vs a Mediterranean diet: A secondary analysis of a randomized clinical trial. JAMA Network Open, 7(11), e2445784. https://doi.org/10.1001/jamanetworkopen.2024.45784

Karam, G., Agarwal, A., Sadeghirad, B., Jalink, M., Hitchcock, C. L., Ge, L., Kiflen, R., Ahmed, W., Zea, A. M., Milenkovic, J., Chedrawe, M. A., Rabassa, M., El Dib, R., Goldenberg, J. Z., Guyatt, G. H., Boyce, E., & Johnston, B. C. (2023). Comparison of seven popular structured dietary programmes and risk of mortality and major cardiovascular events in patients at increased cardiovascular risk: Systematic review and network meta-analysis. BMJ, 380, e072003. https://doi.org/10.1136/bmj-2022-072003

Keys, A., Menotti, A., Karvonen, M. J., Aravanis, C., Blackburn, H., Buzina, R., Djordjevic, B. S., Dontas, A. S., Fidanza, F., Keys, M. H., Kromhout, D., Nedeljkovic, S., Punsar, S., Seccareccia, F., & Toshima, H. (1986). The diet and 15-year death rate in the Seven Countries Study. American Journal of Epidemiology, 124(6), 903-915. https://doi.org/10.1093/oxfordjournals.aje.a114480

Kojima, G., Avgerinou, C., Iliffe, S., & Walters, K. (2018). Adherence to Mediterranean diet reduces incident frailty risk: Systematic review and meta-analysis. Journal of the American Geriatrics Society, 66(4), 783-788. https://doi.org/10.1111/jgs.15251

Lichtenstein, A. H., Khera, A., Anderson, C. A. M., Appel, L. J., DeSilva, D. M., Gardner, C., Hu, F. B., Jones, D. W., Petersen, K. S., & American Heart Association. (2026). 2026 dietary guidance to improve cardiovascular health: A scientific statement from the American Heart Association. Circulation, 153(18), e1285-e1295. https://doi.org/10.1161/CIR.0000000000001435

Martinez-Gonzalez, M. A., Garcia-Arellano, A., Toledo, E., Salas-Salvado, J., Buil-Cosiales, P., Corella, D., Covas, M. I., Schroder, H., Aros, F., Gomez-Gracia, E., Fiol, M., Ruiz-Gutierrez, V., Lapetra, J., Lamuela-Raventos, R. M., Serra-Majem, L., Pinto, X., Munoz, M. A., Warnberg, J., Ros, E., & Estruch, R. (2012). A 14-item Mediterranean diet assessment tool and obesity indexes among high-risk subjects: The PREDIMED trial. PLoS One, 7(8), e43134. https://doi.org/10.1371/journal.pone.0043134

Morze, J., Danielewicz, A., Przybylowicz, K., Zeng, H., Hoffmann, G., & Schwingshackl, L. (2021). An updated systematic review and meta-analysis on adherence to Mediterranean diet and risk of cancer. European Journal of Nutrition, 60(3), 1561-1586. https://doi.org/10.1007/s00394-020-02346-6

National Kidney Foundation. (2023, January 2). Beans. https://www.kidney.org/kidney-topics/beans

NIH Office of Dietary Supplements. (2021, March 29). Vitamin K: Fact sheet for health professionals. ods.od.nih.gov/factsheets/VitaminK-HealthProfessional

Office of the Surgeon General. (2025). Alcohol and cancer risk: The U.S. Surgeon General's advisory. U.S. Department of Health and Human Services. hhs.gov/sites/default/files/oash-alcohol-cancer-risk.pdf

Oldways. (n.d.). Mediterranean diet. Oldways Preservation and Exchange Trust. oldwayspt.org/traditional-diets/mediterranean-diet

Rees, K., Takeda, A., Martin, N., Ellis, L., Wijesekara, D., Vepa, A., Das, A., Hartley, L., & Stranges, S. (2019). Mediterranean-style diet for the primary and secondary prevention of cardiovascular disease. Cochrane Database of Systematic Reviews, 3, CD009825. https://doi.org/10.1002/14651858.CD009825.pub3

Rubini, A., Vilaplana-Prieto, C., Flor-Alemany, M., Yeguas-Rosa, L., Hernandez-Gonzalez, M., Felix-Garcia, F. J., Felix-Redondo, F. J., & Fernandez-Berges, D. (2022). Assessment of the cost of the Mediterranean diet in a low-income region: Adherence and relationship with available incomes. BMC Public Health, 22(1), 58. https://doi.org/10.1186/s12889-021-12433-w

Ruiz-Canela, M., Corella, D., Martinez-Gonzalez, M. A., Babio, N., Martinez, J. A., Forga, L., Alonso-Gomez, A. M., Warnberg, J., Vioque, J., Romaguera, D., Lopez-Miranda, J., Estruch, R., Santos-Lozano, J. M., Serra-Majem, L., Bueno-Cavanillas, A., Tur, J. A., Martin-Sanchez, V., Riera-Mestre, A., Delgado-Rodriguez, M., ... Salas-Salvado, J. (2025). Comparison of an energy-reduced Mediterranean diet and physical activity versus an ad libitum Mediterranean diet in the prevention of type 2 diabetes: A secondary analysis of a randomized controlled trial. Annals of Internal Medicine, 178(10), 1378-1389. https://doi.org/10.7326/ANNALS-25-00388

Saez-Almendros, S., Obrador, B., Bach-Faig, A., & Serra-Majem, L. (2013). Environmental footprints of Mediterranean versus Western dietary patterns: Beyond the health benefits of the Mediterranean diet. Environmental Health, 12, Article 118. https://doi.org/10.1186/1476-069X-12-118

Salas-Salvado, J., Bullo, M., Estruch, R., Ros, E., Covas, M. I., Ibarrola-Jurado, N., Corella, D., Aros, F., Gomez-Gracia, E., Ruiz-Gutierrez, V., Romaguera, D., Lapetra, J., Lamuela-Raventos, R. M., Serra-Majem, L., Pinto, X., Basora, J., Munoz, M. A., Sorli, J. V., & Martinez-Gonzalez, M. A. (2014). Prevention of diabetes with Mediterranean diets: A subgroup analysis of a randomized trial. Annals of Internal Medicine, 160(1), 1-10. https://doi.org/10.7326/M13-1725

Shai, I., Schwarzfuchs, D., Henkin, Y., Shahar, D. R., Witkow, S., Greenberg, I., Golan, R., Fraser, D., Bolotin, A., Vardi, H., Tangi-Rozental, O., Zuk-Ramot, R., Sarusi, B., Brickner, D., Schwartz, Z., Sheiner, E., Marko, R., Katorza, E., Thiery, J., ... Stampfer, M. J. (2008). Weight loss with a low-carbohydrate, Mediterranean, or low-fat diet. New England Journal of Medicine, 359(3), 229-241. https://doi.org/10.1056/NEJMoa0708681

Sofi, F., Macchi, C., Abbate, R., Gensini, G. F., & Casini, A. (2014). Mediterranean diet and health status: An updated meta-analysis and a proposal for a literature-based adherence score. Public Health Nutrition, 17(12), 2769-2782. https://doi.org/10.1017/S1368980013003169

Soltani, S., Jayedi, A., Shab-Bidar, S., Becerra-Tomas, N., & Salas-Salvado, J. (2019). Adherence to the Mediterranean diet in relation to all-cause mortality: A systematic review and dose-response meta-analysis of prospective cohort studies. Advances in Nutrition, 10(6), 1029-1039. https://doi.org/10.1093/advances/nmz041

Stockwell, T., Zhao, J., Panwar, S., Roemer, A., Naimi, T., & Chikritzhs, T. (2016). Do "moderate" drinkers have reduced mortality risk? A systematic review and meta-analysis of alcohol consumption and all-cause mortality. Journal of Studies on Alcohol and Drugs, 77(2), 185-198. https://doi.org/10.15288/jsad.2016.77.185

Toledo, E., Salas-Salvado, J., Donat-Vargas, C., Buil-Cosiales, P., Estruch, R., Ros, E., Corella, D., Fito, M., Hu, F. B., Aros, F., Gomez-Gracia, E., Romaguera, D., Ortega-Calvo, M., Serra-Majem, L., Pinto, X., Schroder, H., Basora, J., Sorli, J. V., Bullo, M., ... Martinez-Gonzalez, M. A. (2015). Mediterranean diet and invasive breast cancer risk among women at high cardiovascular risk in the PREDIMED trial: A randomized clinical trial. JAMA Internal Medicine, 175(11), 1752-1760. https://doi.org/10.1001/jamainternmed.2015.4838

Tsaban, G., Yaskolka Meir, A., Rinott, E., Zelicha, H., Kaplan, A., Shalev, A., Katz, A., Rudich, A., Tirosh, A., Shelef, I., Youngster, I., Lebovitz, S., Israeli, N., Shabat, M., Brikner, D., Pupkin, E., Stumvoll, M., Thiery, J., Ceglarek, U., ... Shai, I. (2021). The effect of green Mediterranean diet on cardiometabolic risk; a randomised controlled trial. Heart, 107(13), 1054-1061. https://doi.org/10.1136/heartjnl-2020-317802

Tsofliou, F., Vlachos, D., Hughes, C., & Appleton, K. M. (2022). Barriers and facilitators associated with the adoption of and adherence to a Mediterranean style diet in adults: A systematic review of published observational and qualitative studies. Nutrients, 14(20), 4314. https://doi.org/10.3390/nu14204314

UNESCO. (2013). Mediterranean diet: Representative List of the Intangible Cultural Heritage of Humanity. ich.unesco.org/en/RL/mediterranean-diet-00884

U.S. Department of Agriculture, & U.S. Department of Health and Human Services. (2026). Dietary guidelines for Americans, 2025-2030. https://cdn.realfood.gov/DGA.pdf

U.S. Food and Drug Administration, & U.S. Environmental Protection Agency. (2024, March 5). Advice about eating fish: For those who might become or are pregnant or breastfeeding and children ages 1 to 11 years. https://www.fda.gov/food/consumers/advice-about-eating-fish

Valls-Pedret, C., Sala-Vila, A., Serra-Mir, M., Corella, D., de la Torre, R., Martinez-Gonzalez, M. A., Martinez-Lapiscina, E. H., Fito, M., Perez-Heras, A., Salas-Salvado, J., Estruch, R., & Ros, E. (2015). Mediterranean diet and age-related cognitive decline: A randomized clinical trial. JAMA Internal Medicine, 175(7), 1094-1103. https://doi.org/10.1001/jamainternmed.2015.1668

Wallerer, S., Stadelmaier, J., Floegel, E., Kiesswetter, E., Bantle, G., Hoffmann, G., & Schwingshackl, L. (2025). Adherence to Mediterranean diet and risk of type 2 diabetes: An updated systematic review and dose-response meta-analysis. Advances in Nutrition, 16(12), 100562. https://doi.org/10.1016/j.advnut.2025.100562

Willett, W. C., Sacks, F., Trichopoulou, A., Drescher, G., Ferro-Luzzi, A., Helsing, E., & Trichopoulos, D. (1995). Mediterranean diet pyramid: A cultural model for healthy eating. American Journal of Clinical Nutrition, 61(6 Suppl), 1402S-1406S. https://doi.org/10.1093/ajcn/61.6.1402S

Evidence rating and review record

Every page in this archive rates its evidence outcome by outcome, using GRADE wherever a published GRADE rating exists. GRADE is the same 4-level scale used in the clinical guidelines you will meet in practice. Where a source used a different tool, or no published rating exists, the table below says so row by row. High means further research is very unlikely to change the conclusion. Moderate means further research may change it. Low means further research is likely to change it. Very low means the conclusion is uncertain. GRADE rates each outcome separately, so a diet can be high for one outcome and very low for another. A high rating means the research is solid; it does not mean the diet is right for everyone, and it does not mean a coach should recommend it.

Three things about the table below matter before reading it. A rating belongs to an outcome and to a comparison together: the same trial supports moderate for stroke and low for heart attack, and the same outcome can be moderate against a pamphlet and low against an active low-fat diet, so a rating quoted without its comparison is a fragment rather than a finding. The tool is named in every row, because the cancer rows use NutriGrade, a different scoring system whose labels share GRADE's words without being interchangeable with them. And one row is this page's own judgment rather than a published rating, and says so. The banner at the top of the page carries the stroke rating, because that is the best supported single outcome here, not a verdict on the pattern as a whole.

Certainty of evidence by outcome and by comparison, with the source of each rating and the reasons for it

Outcome	Certainty	Rated by	Why it is rated this way
Stroke	Moderate against an active low-fat diet. Moderate against minimal advice	Rees et al., 2019 (Cochrane); Karam et al., 2023	Randomized trials start at high certainty. Against an active low-fat diet the evidence is 1 trial of 7,447 people, hazard ratio 0.60 from 0.45 to 0.80, lowered a single level because assignment to groups failed at some sites in that trial. It is not lowered for imprecision, because the interval stays well clear of no effect. Against minimal advice a network analysis of 40 trials gives an odds ratio of 0.65 from 0.46 to 0.93, also rated moderate. This is the best supported result on the page and the rating in the banner.
Peripheral artery disease	Moderate against an active low-fat diet	Rees et al., 2019 (Cochrane)	Same single trial and the same single downgrade as stroke. Hazard ratio 0.42 from 0.28 to 0.61, which is about 18 events per 1,000 people falling to about 8. Cochrane adds that these figures come from earlier reports of the trial and were not reanalyzed after the randomization problem was found, which makes them less certain than the stroke result.
Heart attack	Low against an active low-fat diet. Moderate against minimal advice	Rees et al., 2019 (Cochrane); Karam et al., 2023	Against an active low-fat diet, hazard ratio 0.79 from 0.57 to 1.10, lowered once for the assignment problem in that trial and once for imprecision, because the interval includes both a worthwhile benefit and a worthwhile harm. Against minimal advice, non-fatal heart attack comes out at an odds ratio of 0.48 from 0.36 to 0.65 and is rated moderate. The conclusion depends on which comparison is asked about.
Death from cardiovascular causes	Low against an active low-fat diet. Moderate against minimal advice	Rees et al., 2019 (Cochrane); Karam et al., 2023	Against an active low-fat diet, hazard ratio 0.81 from 0.50 to 1.32, lowered for the assignment problem and for imprecision. That result is inconclusive, which is not the same as a finding of no effect. Against minimal advice, odds ratio 0.55 from 0.39 to 0.78, about 13 fewer deaths per 1,000 people over 5 years, rated moderate.
Death from any cause	Low against an active low-fat diet. Moderate against minimal advice	Rees et al., 2019 (Cochrane); Karam et al., 2023; de Lorgeril et al., 1999	Against an active low-fat diet, hazard ratio 1.00 from 0.81 to 1.24 in 1 trial of 7,447 people, lowered for the assignment problem and for imprecision, and the network analysis likewise found no convincing difference between Mediterranean and low-fat programs. Against minimal advice, odds ratio 0.72 from 0.56 to 0.92, about 17 fewer deaths per 1,000 people over 5 years at intermediate risk, rated moderate; those were programs that could also include exercise, behavioral support and medication, so the benefit belongs to the program. Separately, the Lyon trial in 605 people who had already had a heart attack reported an adjusted risk ratio for death of 0.44 from 0.21 to 0.94 in its own paper. Cochrane rated that evidence low, lowering it 2 levels for risk of bias because the randomization method is unclear and the modified Zelen design may have introduced other biases. The canola oil margarine in its diet bears on how far the result applies to this pattern, but it is not the reason Cochrane gave for the rating.
Blood pressure	Moderate against no advice or minimal advice. Low against another active diet	Rees et al., 2019 (Cochrane)	Against no or minimal advice, systolic pressure 2.99 mmHg lower from 3.45 to 2.53 lower, in 269 people across 2 trials, lowered a single level for imprecision because that is a small number of participants. Against another active diet, systolic 1.5 mmHg lower in 448 people across 4 trials, with an interval from 3.92 lower to 0.92 higher that includes no difference, rated low. A larger synthesis of 19 trials and 4,137 people outside Cochrane finds about 1.4 mmHg systolic and assigns no rating of its own (Cowell et al., 2021). Moderate certainty here is a statement about the reliability of a change of a few mmHg, not about its importance for one person.
New cases of type 2 diabetes	Low from the one trial, against a low-fat diet. Moderate from cohort studies of adherence	Wallerer et al., 2025	The trial evidence is the Spanish primary prevention trial with its 2 Mediterranean groups taken together, hazard ratio 0.75 from 0.56 to 1.01, about 21 fewer cases per 1,000 people with a range from 38 fewer to 1 more. Lowered once because that trial carries risk-of-bias concerns arising from randomization, deviations from the intervention and missing outcome data, and once because the interval crosses no effect against a stated threshold of 5 per 1,000. The cohort evidence is 24 studies, 988,337 people and 68,052 cases, 8 percent lower risk per 2 points of adherence, lowered once for risk of bias and not lowered for imprecision.
Cancer mortality, and colorectal cancer	Moderate, on NutriGrade rather than GRADE	Morze et al., 2021	This synthesis rates its own certainty with NutriGrade, a scoring system built for nutrition research, and its labels are not interchangeable with GRADE labels of the same name. Cancer mortality 13 percent lower from 0.82 to 0.92 across 18 cohort studies. Colorectal cancer 8 percent lower from 0.87 to 0.99 across 10 cohort studies. Those cohort estimates are what the rating applies to. The larger colorectal figure of 17 percent from the same paper adds case-control studies and does not carry the moderate rating. Both rated results rest on observational data, so confounding is the limit that matters.
Other cancer sites, including breast	Low or very low, on NutriGrade rather than GRADE	Morze et al., 2021	The same synthesis rates most other sites low or very low. Breast cancer is rated low, and its cohort estimate is 0.97 from 0.94 to 1.00 across 12 cohorts, which is close to no difference. The 6 percent lower figure sometimes quoted for breast cancer adds case-control studies, which ask people to recall how they used to eat after they are already ill, and does not carry the rating. No clear association was found for prostate, pancreatic, esophageal or blood cancers.
Breast cancer recurrence	Not rated on this page. One trial, no reduction found	Berrino et al., 2024	DIANA-5 is a randomized trial that took a cancer outcome as its main question, and the only one found for this page. It assigned 1,542 women at high risk of recurrence to a Mediterranean and macrobiotic eating program or to a control group for 5 years and found recurrence in 95 against 98, hazard ratio 0.99 from 0.69 to 1.40. It did not demonstrate a reduction, though an interval that wide cannot rule one out either. No certainty rating is offered here, because the program it tested excluded dairy products and alcohol and is not the pattern this page describes.
Body weight over 5 years, on a version with no calorie restriction, against low-fat advice	Moderate	This page's own assessment, not a published rating	There is no published certainty rating for this outcome, so this row is an author judgment and is labeled as one. The question is deliberately narrow: what happens to body weight over 5 years in older adults at high cardiovascular risk who follow this pattern without being asked to eat less, compared with low-fat advice. On that question 1 trial of 7,447 people found differences of 0.410 kg and 0.016 kg against control, with intervals of 0.830 lower to 0.010 higher and 0.453 lower to 0.421 higher. Randomized trials start at high certainty and this is lowered once for risk of bias, from the unblinded design and the assignment problem. It is not lowered for imprecision, because both intervals rule out an average difference between the groups larger than about 0.8 kg in either direction, and this page chose 1 kg across 5 years as the smallest average difference worth caring about. That threshold is a judgment, not a finding, and the intervals describe group averages, not how any one person responded. Read as a wider claim about how much weight any Mediterranean program changes, this rating would not hold.

How this archive talks about weight. This page reports what studies measured, including body weight, and does not treat weight change as a goal or as a measure of a diet's success. People are described the way the study described them.

Status: reviewed and approved by James Peth, MS, MPH, PhD, on September 27, 2026. What changed: first version, revised after an independent adversarial review and a final verification review. The certainty table was rebuilt so that every rating names its outcome, its comparison, its scoring tool and its source; the account of mortality, cancer and absolute risk was corrected; and 3 sources missing from the first version were added. The section that sorted the pattern by dimensions of wellness was removed because it repeated other sections, and the environmental footprint now has its own section, with sources, under In practice.

Updated October 21, 2026: The American Heart Association is now named as an author of its 2026 dietary guidance statement (Lichtenstein et al., 2026), as the publisher's citation credits it. No content changed.

Next scheduled review: August 2027. Pages built on annual guidance are rechecked each year. The alcohol material on this page and the federal guidance it refers to are in that category and should be rechecked first.

---

Diet Reference Archive | One James

Prepared by James Peth, MS, MPH, PhD

© 2026–2027 One James. All rights reserved.

This page is for education and coaching practice. It is not medical or dietary advice.