Only virgin type of olive oil consumption reduces the risk of mortality. Results from a Mediterranean population-based cohort

The study investigates whether the association between olive oil consumption and mortality differs by olive oil variety—virgin (unprocessed, rich in phenolic compounds) versus common (largely refined, with fewer bioactive compounds). The Mediterranean diet, characterized by high olive oil intake, has been linked to healthy aging and reduced risks of mortality and chronic diseases. Prior observational and trial evidence suggests olive oil may drive some cardiometabolic benefits of the Mediterranean diet, particularly via virgin olive oil’s anti-inflammatory, antioxidant, and vascular effects. However, most epidemiologic studies have not distinguished olive oil varieties, potentially obscuring differential health effects due to the loss of bioactives during refining. This study aims to clarify the association of total, common, and virgin olive oil consumption with all-cause, cardiovascular, and cancer mortality in a large, nationally representative cohort of Spanish adults, addressing a key evidence gap with implications for dietary guidance.

Prior evidence includes the PREDIMED randomized trial showing lower cardiovascular and total mortality with a Mediterranean diet supplemented with virgin olive oil versus a reduced-fat diet. Observational analyses in PREDIMED linked higher total olive oil intake with reduced total and cardiovascular mortality, but not cancer mortality. In EPIC-Spain, both common and virgin olive oil were associated with decreased total and cardiovascular mortality, with no association for cancer mortality; data were from the 1990s when virgin olive oil consumption was lower. A large U.S. cohort study (Nurses’ Health Study and Health Professionals Follow-up Study) observed inverse associations between total olive oil intake and total, cardiovascular, and cancer mortality, despite not distinguishing varieties. Other European and Middle Eastern cohorts reported mixed or null results, and most did not differentiate virgin from refined olive oil. Mechanistically, virgin olive oil contains higher levels of bioactive polyphenols and other minor compounds with anti-inflammatory and cardioprotective effects, whereas refining substantially reduces these compounds in common olive oil.

Design and population: Data came from the ENRICA study, a nationally representative cohort of non-institutionalized Spanish adults ≥18 years, recruited 2008–2010 using random stratified cluster sampling by province and municipality size, with random selection of municipalities, census sections, and households (random digit dialing). Participants were selected proportional to national sex and age distributions; baseline consent and detailed information were obtained. Response rate was 51.5%. Ethical approvals were obtained, and written informed consent was collected. Baseline data collection: Sociodemographic factors, health behaviors (including smoking, physical activity, TV viewing), morbidity, medication use, anthropometrics (measured weight and height to compute BMI), and clinical risk factors were collected. Definitions: hypertriglyceridemia (triglycerides ≥150 mg/dL), hypercholesterolemia (total cholesterol ≥200 mg/dL or lipid-lowering medication), hypertension (≥140/90 mmHg or antihypertensive medication), diabetes (self-reported or medication). Chronic conditions (COPD, coronary heart disease, stroke, heart failure, osteoarthritis, cancer, depression requiring treatment) were self-reported as physician-diagnosed. Dietary assessment: During a second home visit, habitual diet was assessed with the validated computer-based Dietary History ENRICA (DH-ENRICA), derived from EPIC-Spain, capturing weekday/weekend and seasonal variation across 880 foods and 184 recipes. Participants reported frequency and types of oils/fats used for cooking, dressings, and in recipes, with detailed quantification of common (refined/mixed) and virgin olive oil intake. Nutrient intakes were computed from Spanish food composition tables. Mediterranean Diet Score (modified 0–7) was calculated per Trichopoulou’s definition, assigning 1 point for consumption above sex-specific medians of beneficial components (vegetables, legumes, fruits and nuts, cereals, fish) and 0 for detrimental components (red/processed meats and poultry, dairy). Alcohol and MUFA/SFA ratio were excluded to avoid overadjustment and collinearity with olive oil. Exposure: Total olive oil intake (g/day) was the sum of common and virgin varieties. Intakes were energy-adjusted via the residual method and categorized into sex-specific tertiles; analyses also considered continuous intake per 10 g/day (~1 tablespoon). Outcomes: Mortality status was ascertained via the Spanish National Death Index; cause-specific deaths (cardiovascular and cancer) were obtained from the National Statistics Institute. All-cause mortality follow-up extended to January 31, 2020; cardiovascular and cancer mortality to January 31, 2017. Follow-up time was from baseline until death or end of follow-up. Statistical analysis: Cox proportional hazards models with attained age as the timescale estimated hazard ratios (HR) and 95% confidence intervals (CI) for all-cause, cardiovascular, and cancer mortality across olive oil tertiles and per 10 g/day. Linear trends used median values per tertile. Three models with progressive adjustment: Model 1 adjusted for sex, age, and total energy intake; Model 2 additionally for education, smoking, BMI category, total physical activity (household and leisure, MET-h/week), TV viewing (h/day), alcohol (g/day), fiber (g/day), Mediterranean Diet Score (0–7), number of medications (0, 1–3, >3); Model 3 further for potential mediators: hypertriglyceridemia, hypercholesterolemia, hypertension, diabetes, and number of chronic conditions (0, 1, ≥2). When analyzing common and virgin olive oil separately, models were mutually adjusted for the other variety. Missing covariate values <1% were imputed via stochastic regression; results were checked against complete-case analyses. Restricted cubic splines (knots at 10th, 50th, 90th percentiles) depicted dose–response. Subgroup analyses for all-cause mortality stratified by age (≤/>60 years), sex, BMI (≤/>26.3 kg/m²), physical activity (≤/>61.5 MET-h/week), and Mediterranean Diet adherence (≤/>3), with interaction tested by likelihood ratio tests. Sensitivity analyses excluded the first 2 years of follow-up. Analyses used STATA/SE 16 with survey weights (svy) to account for complex sampling; two-tailed p<0.05 was significant. Sample for analysis: Of 13,105 participants, exclusions for implausible energy intake (n=884) and incomplete diet data (n=60) yielded 12,161 participants (5708 men, 6346 women; mean age 47±17 years).

- Cohort and events: 12,161 adults followed for a mean 10.7 years (129,272 person-years) for all-cause mortality and 8.8 years for cause-specific mortality; deaths: 739 all-cause (6.1%), 143 cardiovascular (1.2%), 146 cancer (1.2%). - All-cause mortality: Total olive oil intake showed a nonsignificant reduction when comparing highest vs lowest tertile after full adjustment: HR 0.88 (95% CI 0.72–1.08; P-trend 0.306). Per 10 g/day increase in total olive oil, HR 0.94 (0.87–1.01). Common olive oil: no association (highest vs lowest tertile HR 0.96; 0.75–1.23; P-trend 0.891; per 10 g/day HR 0.97; 0.90–1.04). Virgin olive oil: significant inverse association (highest vs lowest tertile HR 0.66; 0.49–0.89; P-trend 0.040; per 10 g/day HR 0.91; 0.83–1.00). - Cardiovascular mortality: Total olive oil associated with lower risk (highest vs lowest tertile HR 0.55; 0.35–0.85; P-trend 0.009; per 10 g/day HR 0.87; 0.73–1.04). Common olive oil: not associated (highest vs lowest HR 0.88; 0.49–1.60; P-trend 0.242; per 10 g/day HR 0.90; 0.76–1.06). Virgin olive oil: significantly lower risk (highest vs lowest HR 0.43; 0.20–0.91; P-trend 0.017; per 10 g/day HR 0.78; 0.59–1.03). - Cancer mortality: No significant associations for total, common, or virgin olive oil (e.g., total olive oil highest vs lowest HR 1.03; 0.61–1.74; P-trend 0.924; virgin olive oil highest vs lowest HR 1.19; 0.73–1.93; P-trend 0.322). - Subgroups and sensitivity: The inverse association of virgin olive oil with all-cause mortality appeared stronger among those with higher physical activity (interaction p=0.045) and possibly among those with lower adherence to the Mediterranean Diet; results were robust after excluding the first 2 years of follow-up. In participants without baseline CVD/diabetes, higher virgin olive oil was linked to greater reductions in cardiovascular mortality; among those with prevalent CVD/diabetes, reductions in all-cause mortality were greater.

The study directly addresses whether health effects of olive oil on mortality differ by variety. Findings demonstrate that virgin olive oil, but not common (refined/mixed) olive oil, is associated with substantially lower all-cause and cardiovascular mortality in a representative Spanish population. These results align with biological plausibility: virgin olive oil retains polyphenols and other minor bioactive compounds with anti-inflammatory, antioxidant, endothelial, and anti-atherosclerotic effects, whereas refining markedly diminishes these compounds in common olive oil. The lack of association with cancer mortality is consistent with several prior Mediterranean cohorts, though not with a large U.S. cohort that did not distinguish varieties. Compared with EPIC-Spain and PREDIMED observational analyses, the present study uniquely shows clearer benefits restricted to the virgin variety, potentially reflecting more precise exposure assessment by variety and contemporary consumption patterns with greater virgin olive oil use. Subgroup analyses suggest potential synergy between higher physical activity and virgin olive oil consumption in reducing total mortality, and possible stronger benefits among those with lower overall Mediterranean Diet adherence, indicating virgin olive oil may confer protective effects beyond general diet quality. The findings support the relevance of olive oil quality (virgin vs refined) for mortality risk and may inform dietary recommendations emphasizing virgin olive oil specifically.

In a large, nationally representative cohort of Spanish adults, higher consumption of virgin olive oil was associated with markedly lower risks of all-cause and cardiovascular mortality, whereas common olive oil showed no such associations. No associations were observed with cancer mortality for any olive oil variety. These results underscore the importance of olive oil quality, supporting dietary guidance that favors virgin olive oil. Future research should further delineate dose–response relationships for virgin olive oil, investigate mechanisms linked to its bioactive components, and continue to analyze olive oil varieties separately in diverse populations.