That olive oil is «good for the heart» appears on bottles, in newspapers and in doctors' waiting rooms. This article examines that one question only — heart and blood vessels — and leaves everything else aside: the nutrition facts and authorised health claims, the chemistry of the polyphenols, the practice of daily consumption and the country comparison on life expectancy are covered elsewhere.
Cardiovascular disease is the only area where olive oil has been tested in a large randomised trial with hard endpoints — heart attack, stroke, death. That makes the evidence here stronger than for any other health question about olive oil. It does not make it proof. That distinction is what this article is actually about.
PREDIMED: the one trial with hard endpoints
PREDIMED was a Spanish multicentre trial in primary prevention. It enrolled 7447 people aged 55 to 80 (57 per cent women) who were at high cardiovascular risk but had no cardiovascular disease at enrolment. They were assigned to one of three diets:
- Mediterranean diet with extra virgin olive oil (2543 people). This group received the oil free of charge — 15 litres every three months, roughly one litre per week per household — with the recommendation to consume at least 4 tablespoons (about 50 grams) a day per person.
- Mediterranean diet with mixed nuts (2454 people), 30 grams a day.
- Control group (2450 people): advice to reduce dietary fat, plus small non-food gifts.
The primary endpoint was a composite event: myocardial infarction, stroke, or death from cardiovascular causes. After a median follow-up of 4.8 years the trial was stopped on the basis of a prespecified interim analysis. By then 288 participants had had an event.
| Group | Events | 5-year absolute risk | Hazard ratio (95% CI) |
|---|---|---|---|
| Mediterranean diet + olive oil | 96 (3.8%) | 3.6% (2.8–4.5) | 0.69 (0.53–0.91) |
| Mediterranean diet + nuts | 83 (3.4%) | 4.0% (3.1–5.0) | 0.72 (0.54–0.95) |
| Control (reduced fat) | 109 (4.4%) | 5.7% (4.6–6.9) | 1.00 (reference) |
Read in relative terms: about 31 per cent fewer events in the olive oil group. Read in absolute terms: 3.6 instead of 5.7 per cent over five years, a difference of roughly two percentage points. Both numbers describe the same result; the second is the more honest one if you want to know what it means for one person.
Which part of the effect actually carried it
A composite endpoint hides where the difference came from. The trial analysed the components separately, and there the picture narrows considerably:
| Endpoint | Events, oil vs. control | Hazard ratio (95% CI) | Statistically significant? |
|---|---|---|---|
| Stroke | 49 vs. 58 | 0.65 (0.44–0.95) | yes |
| Myocardial infarction | 37 vs. 38 | 0.82 (0.52–1.30) | no |
| Death from cardiovascular causes | 26 vs. 30 | 0.62 (0.36–1.06) | no |
| Death from any cause | 118 vs. 114 | 0.90 (0.69–1.18) | no |
This is the most important and least quoted finding in the whole study: the effect was carried by stroke. For myocardial infarction the upper confidence limit sat well above 1, and for all-cause mortality there was essentially no difference. The authors themselves note that event rates were lower than anticipated, which reduced the statistical power to assess the components separately; they speak of «favorable trends» for stroke and myocardial infarction, not of two proven effects.
So anyone reading «olive oil lowers your heart attack risk» is reading something this trial did not show. What it showed is that a Mediterranean diet with generous olive oil was followed by fewer major vascular events in this population, mostly fewer strokes.
The 2018 retraction — and why it matters
PREDIMED was published in 2013 in the «New England Journal of Medicine». The authors then identified protocol deviations themselves: at one of eleven study sites some participants were assigned to a group without being randomised; at another, randomisation tables were apparently used inconsistently; and household members of people already enrolled were taken in without randomisation. The paper was withdrawn in 2018 and republished with revised effect estimates — analyses that no longer rely exclusively on the assumption that every participant was randomly assigned.
The headline result held: the hazard ratio for the primary endpoint moved from 0.70 to 0.69. Results also remained similar after omitting 1588 participants whose group assignment was known or suspected to have departed from the protocol. Even so, PREDIMED can no longer be read as a clean randomised trial, but as a trial with partly non-randomised components and a retrospective statistical correction. Anyone who rates the olive-oil-and-heart evidence highly has to say this part out loud too.
PREDIMED-Plus: the sequel whose main result is missing
PREDIMED-Plus is the follow-up trial and on paper the stronger one: 6874 participants — men aged 55 to 75, women aged 60 to 75, all with a body mass index between 27 and 40 and metabolic syndrome — across 23 Spanish centres, with six years of intervention. The primary endpoint is again a combined cardiovascular outcome: myocardial infarction (acute coronary syndrome with a positive troponin test), stroke, or cardiovascular mortality. It compares an energy-restricted Mediterranean diet with physical activity promotion and behavioural support against a Mediterranean diet without energy restriction.
What has been published so far concerns risk factors, not events. In a subsample of 626 people, the intervention group lost an average of 3.2 kg after twelve months against 0.7 kg in the control group; weight loss of at least five per cent was reached by 33.7 versus 11.9 per cent. Waist circumference, fasting glucose, triglycerides and HDL cholesterol improved significantly more.
As far as our search of the indexed literature goes, no results for the primary cardiovascular endpoint have been published to date (as of October 2026). That is not a footnote: the second major randomised test of the question is still open. Everything said today about olive oil and the heart still rests on PREDIMED and on observational data.
The observational studies: consistent, but observational
Meta-analyses of prospective cohorts are strikingly uniform. A review in «Clinical Nutrition» (2022) pooled 36 studies; for cardiovascular disease it included 806,203 participants with 49,223 events. For every additional 25 grams of olive oil a day the relative risk was 0.84 (95% CI 0.76–0.94), that is 16 per cent lower. A meta-analysis in «Frontiers in Nutrition» (2022) covering 13 cohorts found a relative risk of 0.85 (0.77–0.93) for high versus low consumption, and 0.96 (0.93–0.99) per additional 5 grams a day — with a non-linear shape: above roughly 20 grams a day, little or nothing was added.
These figures are reassuringly consistent, and they remain observational. People who use olive oil smoke less, eat more vegetables and move more; the models adjust for that, but never completely. The authors of the «Clinical Nutrition» analysis point out themselves that only the type 2 diabetes results were homogeneous and that the sources of heterogeneity for the other outcomes were not always apparent.
The mechanism, as far as it has been measured
Between «events are rarer» and «this is how it works» lies a gap filled with intermediate markers. A meta-analysis of 26 controlled trials compared high-polyphenol with low-polyphenol olive oil. The high-polyphenol oil improved markers of lipid oxidation — oxidised LDL (standardised mean difference −0.44; 95% CI −0.78 to −0.10) and malondialdehyde (−0.07 µmol/L; −0.12 to −0.02) — and slightly raised HDL cholesterol (+2.37 mg/dL; 0.41 to 5.04). The authors state that longer trials and studies in non-Mediterranean populations are needed.
Those are laboratory values, not heart attacks. A better oxidation marker points to a plausible pathway; it does not demonstrate an event that was prevented.
What may officially be claimed — and what may not
In the EU — and through food law in practice in Switzerland too — health claims are authorised one by one. For olive oil itself, Regulation (EU) No 432/2012 authorises exactly one: olive oil polyphenols contribute to the protection of blood lipids from oxidative stress. It may only be used for oil containing at least 5 milligrams of hydroxytyrosol and its derivatives — including the oleuropein complex and tyrosol — per 20 grams of oil, and the statement must tell the consumer that the effect applies to a daily intake of 20 grams of olive oil.
Read the sentence closely: it is about protecting blood lipids from oxidation. It does not say olive oil prevents heart attacks or strokes. No authorised claim of that kind exists. There is also a practical problem: the regulation prescribes no measurement method for those 5 milligrams, and which of the more than 30 known olive oil phenolics may be counted is disputed among specialists — which is why the industry uses the claim only cautiously.
What this means for a Swiss household
Soberly summarised:
- The evidence concerns a dietary pattern, not a product. In PREDIMED olive oil was the most conspicuous difference between the groups, but not the only one: fish and legumes were eaten differently too, and on the 14-item Mediterranean diet questionnaire the groups differed on 12 of the 14 items.
- The quantity was high. The target of at least 50 grams a day per person is two and a half times the roughly 20 grams beyond which the observational data show little additional benefit — and a multiple of what is usual in Switzerland.
- Participants were at high risk. The authors write explicitly that whether the results generalise to people at lower risk requires further research.
- The control group was not a straw man, but it was not nothing either. It received advice on a healthy, reduced-fat diet. The comparison was good against decent, not good against bad.
This article is journalism, not medical advice. Anyone with high blood pressure, raised blood lipids or existing heart disease should discuss their diet with a doctor, and should not stop taking a medicine because there is a good oil in the kitchen cupboard.
Frequently asked questions
Does olive oil lower the risk of a heart attack?
The evidence does not show that. In PREDIMED the hazard ratio for myocardial infarction in the olive oil group was 0.82, with a confidence interval from 0.52 to 1.30 — not distinguishable from chance. The only significant component in that trial was stroke.
What exactly did PREDIMED prove?
That in a Spanish high-risk population over almost five years, a Mediterranean diet with added extra virgin olive oil went together with fewer major vascular events than a reduced-fat comparison diet: 3.6 instead of 5.7 per cent absolute risk over five years. Because of the protocol deviations disclosed in 2018, the trial can no longer be read as fully randomised.
How much olive oil was consumed in PREDIMED?
The olive oil group received 15 litres of extra virgin olive oil free of charge every three months, about one litre per week per household, with a target of at least 4 tablespoons — roughly 50 grams — a day per person. That is considerably more than the amounts at which observational studies see their effect.
May a bottle say that olive oil is good for the heart?
No. The only authorised claim for olive oil concerns the protection of blood lipids from oxidative stress by olive oil polyphenols, and only for oils with at least 5 milligrams of hydroxytyrosol and derivatives per 20 grams. There is no authorised claim about heart attack or stroke.
Is olive oil better than rapeseed oil on this question?
There is no randomised comparison with cardiovascular events as the endpoint. PREDIMED compared a Mediterranean diet with olive oil against a reduced-fat diet, not olive oil against another good vegetable oil.
When will the PREDIMED-Plus results appear?
No date has been announced publicly. The six-year intervention is complete; what has been published so far concerns weight and risk factors, not the primary cardiovascular endpoint.
Sources
- Estruch R 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:e34 (PREDIMED, republished version)
- Full text at the «New England Journal of Medicine»
- Martínez-González MA et al.: Cohort Profile – Design and methods of the PREDIMED-Plus randomized trial. Int J Epidemiol 2019;48(2):387–388o
- Salas-Salvadó J et al.: One-Year Results of the PREDIMED-Plus Trial. Diabetes Care 2019;42(5):777–788
- Martínez-González MA et al.: Effect of olive oil consumption on cardiovascular disease, cancer, type 2 diabetes, and all-cause mortality. Clin Nutr 2022;41(12):2659–2682
- Xia M et al.: Olive oil consumption and risk of cardiovascular disease and all-cause mortality: a meta-analysis of prospective cohort studies. Front Nutr 2022;9:1041203
- George ES et al.: The effect of high-polyphenol extra virgin olive oil on cardiovascular risk factors. Crit Rev Food Sci Nutr 2019;59(17):2772–2795
- Regulation (EU) No 432/2012 establishing a list of permitted health claims made on foods
- Tsimidou MZ et al.: In House Validated UHPLC Protocol for the Determination of the Total Hydroxytyrosol and Tyrosol Content in Virgin Olive Oil Fit for the Purpose of the Health Claim Introduced by the EC Regulation 432/2012. Molecules 2019
Note: This article is for general information only and does not replace advice from a doctor or dietitian. If you have health questions, please consult a professional.
