Mediterranean Diet Cuts 20-Year Heart Disease Risk by 21% Versus Low-Fat in US Adults
A large target trial emulation finds sustained Mediterranean diet adherence substantially lowers cardiovascular risk compared to low-fat and AHA diets.
Summary
Researchers used a sophisticated statistical method called target trial emulation to compare three diets over 20 years in more than 12,000 high-risk US adults. People who consistently followed a Mediterranean diet had a 28% absolute cardiovascular disease risk, compared to 36% on a low-fat diet and 31% on the AHA-2020 dietary guidelines. That translates to a 21% lower relative risk for Mediterranean versus low-fat eating. Even the AHA diet outperformed low-fat, cutting risk by 13%. These findings are notable because most prior Mediterranean diet research has been conducted in European populations, leaving questions about whether the same benefits apply to Americans. This study suggests they do — and the advantage holds even in a general population without elevated cardiovascular risk.
Detailed Summary
Heart disease remains the leading cause of death in the United States, and diet is one of the most modifiable risk factors. Yet rigorous long-term comparisons of specific dietary patterns in US adults have been lacking, particularly for the Mediterranean diet, whose benefits have been established primarily in European cohorts. This study addresses that gap directly.
Researchers drew on data from the Nurses' Health Study and the Health Professionals Follow-up Study, enrolling 12,197 adults aged 55 to 80 who had diabetes or at least three major cardiovascular risk factors. Diet was assessed every four years using validated food frequency questionnaires. The primary outcome — a composite of nonfatal heart attack, coronary revascularization, stroke, and cardiovascular death — was confirmed by medical records over a 20-year follow-up period, during which 3,469 CVD events occurred.
To estimate what would happen under sustained adherence to each dietary strategy, the team applied the parametric g-formula, a causal inference method that emulates a randomized controlled trial from observational data. This approach adjusts for time-varying confounders and models diet adherence over the full follow-up period — a significant methodological strength.
The results were striking. The Mediterranean diet produced a pooled 20-year CVD risk of 28.2%, compared to 35.9% for the low-fat diet and 31.2% for the AHA-2020 dietary goals. Risk ratios versus low-fat were 0.79 for the Mediterranean diet and 0.87 for the AHA diet. In a broader general population not selected for high risk, Mediterranean diet adherence still showed a 10% relative risk reduction.
For clinicians counseling patients on diet, these findings reinforce the Mediterranean diet as the strongest evidence-based dietary strategy for cardiovascular prevention in American adults. The AHA guidelines are also beneficial, but the Mediterranean diet's advantage is meaningful and consistent.
Key Findings
- Mediterranean diet adherence lowered 20-year CVD risk by 21% relative to a low-fat diet in high-risk US adults.
- Absolute CVD risk was 28.2% (Mediterranean), 31.2% (AHA-2020), and 35.9% (low-fat) over 20 years.
- AHA-2020 dietary goals reduced CVD risk by 13% versus low-fat diet, but underperformed the Mediterranean diet.
- In a general population without elevated risk, Mediterranean diet still reduced CVD risk by 10% vs. low-fat.
- Benefits were consistent across 3,469 CVD events including MI, stroke, revascularization, and CVD death.
Methodology
The study used target trial emulation with the parametric g-formula applied to two large US prospective cohorts (n=12,197) with 20 years of follow-up and repeat dietary assessments every four years. This causal inference approach models sustained dietary adherence and adjusts for time-varying confounders, partially approximating what a long-term randomized trial would show. Diet was measured via validated food frequency questionnaires.
Study Limitations
Summary is based on the abstract only, as the full text was not accessible; specific subgroup analyses and sensitivity analyses cannot be evaluated. Target trial emulation, while methodologically rigorous, cannot fully eliminate residual confounding inherent to observational data. The study population was predominantly health-professional adults, which may limit generalizability to broader or more diverse US populations.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.