On this page
In PREDIMED (Spain, 7,447 adults at high cardiovascular risk, median follow-up 4.8 years), a Mediterranean diet supplemented with extra-virgin olive oil or nuts reduced the combined risk of heart attack, stroke or cardiovascular death compared with advice to cut fat (hazard ratios 0.69 and 0.72 in the 2018 reanalysis). Those figures describe the combined outcome, not a reduction in each event separately. The 2013 report was retracted for randomization problems and republished with similar results. Cochrane's 2019 review rated the evidence as uncertain for several outcomes, with a clearer result for stroke.
PREDIMED tested a Mediterranean diet in people at high cardiovascular risk who had not yet developed cardiovascular disease. Later trials tested it in people who already had heart disease and as part of a weight-loss program.
What did PREDIMED test?
PREDIMED ran at 11 centers in Spain and enrolled participants from October 2003 to June 2009. Eligible men were 55 to 80 years old and eligible women 60 to 80. Each participant either had type 2 diabetes or at least three major cardiovascular risk factors, and none had cardiovascular disease at the start. The mean age was 67, and the average baseline score on the 14-point Mediterranean diet adherence screener was 8.6.
Participants were assigned to one of three diets:
| Group | What participants received |
|---|---|
| Mediterranean diet plus extra-virgin olive oil | 1 liter of olive oil per week for the participant and household, free of charge, with a target of at least 50 mL per day |
| Mediterranean diet plus mixed nuts | 30 g of nuts per day: 15 g walnuts, 7.5 g hazelnuts and 7.5 g almonds |
| Control diet | Advice to reduce dietary fat and small non-food gifts; quarterly dietitian sessions were added after a 2006 protocol change |
None of the three groups received advice on calorie restriction or physical activity. The primary endpoint was a major cardiovascular event: heart attack, stroke or death from cardiovascular causes.
What did the trial find?
The 2018 paper counted 288 primary events. The olive oil group had 96 (3.8%), the nuts group 83 (3.4%) and the control group 109 (4.4%). After adjusting for baseline characteristics and propensity scores, the hazard ratio against the control diet was 0.69 (95% CI 0.53 to 0.91) for olive oil and 0.72 (95% CI 0.54 to 0.95) for nuts.
The investigators stopped the trial after a prespecified interim analysis, at a median follow-up of 4.8 years. In absolute terms, the event rate was 0.6 percentage points lower with olive oil and 1.0 percentage point lower with nuts than with the control diet (our subtraction of the published percentages). That is roughly one fewer person per 100 having an event over the follow-up, which is a modest gain for the individual and a meaningful one across a population.
The authors concluded that major cardiovascular events were less frequent with a Mediterranean diet supplemented with olive oil or nuts than with a reduced-fat diet. PREDIMED compared the Mediterranean diet with advice to eat less fat. It did not compare it with usual eating habits.
Why was PREDIMED retracted and republished?
The New England Journal of Medicine retracted the original 2013 report in June 2018 and published a corrected version the same month. Some household members had joined without random assignment. One of the 11 study sites also assigned some participants without randomization, and another appeared to use the randomization tables inconsistently.
The authors withdrew the 2013 paper and reported revised estimates from analyses that do not rely only on the assumption that everyone was randomly assigned. Those analyses adjusted for baseline characteristics and propensity scores. A sensitivity analysis that left out 1,588 participants whose assignment was known or suspected to have departed from the protocol gave similar results.
The estimates barely moved. The 2013 hazard ratios were 0.70 (95% CI 0.54 to 0.92) for olive oil and 0.72 (95% CI 0.54 to 0.96) for nuts. The republished values are 0.69 and 0.72. The trial was funded by Instituto de Salud Carlos III, the Spanish Ministry of Health and others.
A reanalysis cannot fully replace randomization. Cochrane's reviewers noted the retraction and reanalysis when they graded the trial, as shown below.
What do other trials add?
Three other studies give context.
The Lyon Diet Heart Study (1994, 1999) enrolled 605 people after a first heart attack. One group received advice to follow an alpha-linolenic acid-rich Mediterranean diet, which Cochrane notes included supplemental canola margarine. The other group ate a usual prudent post-infarct diet. At 27 months, the adjusted risk ratio for cardiac death plus non-fatal heart attack was 0.27 (95% CI 0.12 to 0.59). The 1999 final report, at a mean of 46 months per patient, counted 14 events in the Mediterranean group against 44 in the prudent-diet group.
CORDIOPREV (2022) enrolled 1,002 people with established coronary heart disease in Cordoba, Spain, and followed them for 7 years. Participants ate either a Mediterranean diet or a low-fat diet. The primary composite outcome occurred in 87 people on the Mediterranean diet and 111 on the low-fat diet, with adjusted hazard ratios from 0.719 to 0.753 in favor of the Mediterranean diet. The effect appeared in men (adjusted HR 0.669, 95% CI 0.489 to 0.915). Among 175 women, no difference appeared between the groups. Participants were not masked to their diet, and the funders included olive oil foundations.
PREDIMED-Plus compared an energy-reduced Mediterranean diet with physical activity against a Mediterranean diet without those additions. A 2025 secondary analysis of 4,746 adults looked at type 2 diabetes and found fewer new cases over 6 years in the intervention group. That paper addresses diabetes, not heart disease, and we make no claim here about the trial's cardiovascular result. The authors list the secondary outcome, single-blinded design and self-reported adherence as limitations.
How strong is the evidence overall?
The 2019 Cochrane review (Rees and colleagues) pooled 30 randomized trials with 12,461 randomized participants. In primary prevention it found small improvements in risk factors. Systolic blood pressure fell by 2.99 mmHg (95% CI -3.45 to -2.53), rated moderate-quality evidence. Compared with other diets, LDL cholesterol fell by 0.15 mmol/L (95% CI -0.27 to -0.02), also moderate quality.
For PREDIMED, the reviewers found low-quality evidence of little or no effect on cardiovascular mortality (HR 0.81, 95% CI 0.50 to 1.32) over 4.8 years. They found moderate-quality evidence of fewer strokes (HR 0.60, 95% CI 0.45 to 0.80), a decrease from 24 to 14 per 1,000 people.
Cochrane grades evidence by how much future studies could change the answer. Moderate quality means the true effect is probably close to the estimate but could differ. Low quality means the true effect may be quite different, because the studies are few, small or have design problems. The review found uncertainty about both cardiovascular events and risk factors, in people with and without existing cardiovascular disease. It searched MEDLINE up to 25 September 2018, so it does not include CORDIOPREV or the 2025 PREDIMED-Plus analysis.
What counts as a Mediterranean diet in these studies?
Cochrane required two components: a high ratio of monounsaturated to saturated fat, through olive oil as the main cooking ingredient or foods such as tree nuts, and a high intake of plant foods including fruit, vegetables and legumes.
PREDIMED measured adherence with the 14-point Mediterranean Diet Adherence Screener (MEDAS). A 2011 validation in 7,146 PREDIMED participants found that the screener correlated with a full food-frequency questionnaire at r = 0.52. The authors judged it valid for rapid estimation of adherence. A correlation of 0.52 is moderate, so the screener ranks people reasonably well but does not measure intake precisely.
People with diabetes or on medication should check any change in diet with their doctor.
Frequently asked questions
Was the PREDIMED study retracted?
Yes. The New England Journal of Medicine retracted the 2013 report in 2018 because of three randomization deviations, including 1 of 11 sites that assigned some participants without randomization. The journal republished a corrected analysis the same month, and its hazard ratios (0.69 and 0.72) were close to the originals (0.70 and 0.72).
How much olive oil did PREDIMED participants get?
The olive oil group received 1 liter of extra-virgin olive oil per week for the participant and household, with a target of at least 50 mL per day for the participant. The oil was free of charge. The nuts group received 30 g per day of walnuts, hazelnuts and almonds.
Does the Mediterranean diet lower the risk of stroke?
The best available evidence points that way. Cochrane's 2019 review reported a stroke hazard ratio of 0.60 (95% CI 0.45 to 0.80) for the PREDIMED intervention, a fall from 24 to 14 per 1,000 people, and graded it moderate-quality evidence. That figure comes from one trial, which the review rated more cautiously for cardiovascular mortality.
Is a Mediterranean diet better than a low-fat diet for people who already have heart disease?
CORDIOPREV found fewer major cardiovascular events on a Mediterranean diet than on a low-fat diet in 1,002 people with coronary heart disease (87 against 111 events over 7 years). The result held in men but showed no difference in 175 women. Cochrane's 2019 review had called the evidence for secondary prevention sparse, before CORDIOPREV was published.
Sources
- Estruch R, Ros E, Salas-Salvadó J, et al. Primary Prevention of Cardiovascular Disease with a Mediterranean Diet Supplemented with Extra-Virgin Olive Oil or Nuts. N Engl J Med. 2018;378(25):e34. PubMed 29897866
- Estruch R, Ros E, Salas-Salvadó J, et al. Retraction and Republication: Primary Prevention of Cardiovascular Disease with a Mediterranean Diet. N Engl J Med. 2018;378(25):2441-2442. PubMed 29897867
- Toledo E, Hu FB, Estruch R, et al. Effect of the Mediterranean diet on blood pressure in the PREDIMED trial: results from a randomized controlled trial. BMC Med. 2013;11:207. PMC3849640
- Change to a healthy diet in people over 70 years old: the PREDIMED experience. Eur J Nutr. 2022. PMC8921045
- Schröder H, Fitó M, Estruch R, Martínez-González MA, et al. A short screener is valid for assessing Mediterranean diet adherence among older Spanish men and women. J Nutr. 2011;141(6):1140-5. PubMed 21508208
- de Lorgeril M, Renaud S, Mamelle N, et al. Mediterranean alpha-linolenic acid-rich diet in secondary prevention of coronary heart disease. Lancet. 1994;343(8911):1454-9. PubMed 7911176
- de Lorgeril M, Salen P, Martin JL, Monjaud I, Delaye J, Mamelle N. Mediterranean diet, traditional risk factors, and the rate of cardiovascular complications after myocardial infarction: final report of the Lyon Diet Heart Study. Circulation. 1999;99(6):779-85. PubMed 9989963
- Delgado-Lista J, Alcala-Diaz JF, Torres-Peña JD, et al. Long-term secondary prevention of cardiovascular disease with a Mediterranean diet and a low-fat diet (CORDIOPREV): a randomised controlled trial. Lancet. 2022;399(10338):1876-1885. PubMed 35525255
- Rees K, Takeda A, Martin N, et al. Mediterranean-style diet for the primary and secondary prevention of cardiovascular disease. Cochrane Database Syst Rev. 2019;3(3):CD009825. PubMed 30864165
- ISRCTN registry record ISRCTN89898870 (PREDIMED-Plus). ISRCTN
- Ruiz-Canela M, Corella D, Martínez-González MÁ, et al. 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. Ann Intern Med. 2025;178(10):1378-1389. PubMed 40854218
