Cardiovascular disease and plant-based diets

Cardiovascular disease is the world’s leading cause of death, and dietary pattern is among the most consistently identified modifiable risk factors. Plant-based diets — particularly those built on whole foods — show some of the strongest and most replicated associations with reduced risk in the nutrition literature. The evidence spans prospective cohorts tracking hundreds of thousands of participants, randomized trials of intensive dietary intervention, and mechanistic work on lipids, blood pressure, and vascular function.

The finding is robust. The interpretation requires care.

What the cohorts show

The largest body of evidence comes from prospective cohort studies that follow participants for years or decades and record cardiovascular events.

EPIC-Oxford followed roughly 44,500 British adults, of whom about 34% were vegetarian. Over an average 11.6 years of follow-up, vegetarians had a 32% lower risk of hospitalization or death from ischemic heart disease compared with meat and fish eaters, with the association largely mediated by lower LDL cholesterol and systolic blood pressure (Crowe et al., 2013).

Adventist Health Study-2 tracked about 96,000 Seventh-day Adventists in the US and Canada. Compared with non-vegetarians, vegetarian dietary patterns were associated with a 12% lower all-cause mortality hazard, with vegans and lacto-ovo vegetarians showing the strongest cardiovascular mortality reductions among men (Orlich et al., 2013). Adventists are a particularly useful cohort because the non-vegetarian reference group shares most other health behaviors — low smoking, low alcohol, similar exercise levels — which narrows the residual confounding.

Kim et al. (2019) analyzed roughly 12,000 middle-aged adults in the US Atherosclerosis Risk in Communities (ARIC) cohort over a median 25 years of follow-up. Adherence to an overall plant-based diet index was associated with a 16% lower risk of incident cardiovascular disease and a 25% lower risk of all-cause mortality comparing highest to lowest quintiles.

Not all plant-based diets are equal

The single most important refinement in this literature came from Satija et al. (2017), who analyzed three large US cohorts — Nurses’ Health Study, Nurses’ Health Study 2, and Health Professionals Follow-up Study — totaling over 200,000 participants and more than 8,600 coronary events. They split plant-based diets into three indices:

• Overall plant-based diet index (PDI) — scores all plant foods positively and animal foods negatively.
• Healthful plant-based diet index (hPDI) — scores whole grains, fruits, vegetables, legumes, nuts, and vegetable oils positively, but scores refined grains, sugar-sweetened beverages, fruit juices, potatoes, and sweets negatively.
• Unhealthful plant-based diet index (uPDI) — reverses the hPDI: it rewards refined and sugary plant foods and penalizes whole plant foods.

The hPDI showed a 25% lower coronary heart disease risk comparing highest to lowest deciles. The uPDI showed a 32% higher risk. The overall PDI fell in between at about 8% lower risk (Satija et al., 2017). A diet technically composed of plants — white bread, fries, soda, vegan junk food — does not confer cardiovascular benefit and may do the opposite.

This is the finding that reframed the field. “Plant-based” as a label is not protective. Whole-food plant-based patterns are.

Mechanisms: LDL, blood pressure, endothelial function

Three mechanisms do most of the work.

LDL cholesterol reduction. Meta-analyses of controlled feeding and cohort studies find consistent reductions in total and LDL cholesterol on vegetarian and vegan diets. Yokoyama et al. (2017) pooled 49 studies and reported mean reductions of roughly 13 mg/dL (0.34 mmol/L) in LDL on plant-based compared with omnivorous diets. Wang et al. (2015), in a meta-analysis of randomized controlled trials, found similar LDL reductions of around 10–12 mg/dL. The mechanism is multifactorial: lower intake of saturated fat and dietary cholesterol, higher intake of soluble fiber (which binds bile acids and upregulates hepatic LDL receptors), and higher intake of plant sterols.

Blood pressure. Yokoyama et al. (2014) meta-analyzed 32 observational studies and seven clinical trials and found that vegetarian diets were associated with mean reductions of about 7 mmHg systolic and 5 mmHg diastolic in observational data, and smaller but significant reductions in the trials. Mechanisms include lower sodium-to-potassium ratios, lower body weight, higher magnesium intake, and favorable effects on arterial stiffness. A sustained 5 mmHg systolic reduction at the population level meaningfully shifts cardiovascular risk.

Endothelial function. The endothelium — the single-cell lining of blood vessels — regulates vasodilation, inflammation, and clotting. Diets high in saturated fat and refined carbohydrate impair endothelial function, typically measured by flow-mediated dilation. Whole-food plant-based diets improve it, likely through higher nitric oxide availability (from dietary nitrates in leafy greens and beets), reduced oxidative stress, and lower postprandial lipemia.

The intensive intervention trials

Two small but influential trials tested whether aggressive whole-food plant-based intervention could do more than slow progression.

Ornish et al. (1990) — the Lifestyle Heart Trial — randomized 48 patients with angiographically documented coronary disease to either usual care or an intensive lifestyle program: a low-fat vegetarian diet (roughly 10% of calories from fat), stress management, moderate exercise, smoking cessation, and group support. At one year, 82% of the lifestyle group showed angiographic regression of coronary stenosis compared with progression in the control group. Follow-up at five years showed continued divergence. The trial is small and the intervention bundled — lifestyle, not diet alone — but it demonstrated the principle that established atherosclerosis is not necessarily a one-way process.

Esselstyn et al. (2014) reported a case series of 198 patients with established cardiovascular disease who were counseled on a whole-food plant-based diet with essentially no added oils. Of the 177 adherent patients, 112 reported angina improvement and the major cardiac event rate was 0.6% over an average 3.7 years of follow-up, compared with 62% among the 21 non-adherent patients. The study is observational and self-selected, but the effect size is striking.

Neither trial is definitive on its own. Together with the mechanistic and cohort evidence, they support the view that a sufficiently strict whole-food plant-based pattern can not only prevent but, in some patients, arrest or reverse established disease.

Caveats

Healthy-user effect. People who adopt vegetarian or vegan diets in Western populations also tend to smoke less, exercise more, drink less alcohol, and seek out medical care more proactively. Cohort studies attempt to adjust for these factors, but residual confounding cannot be fully eliminated. The AHS-2 cohort is valuable precisely because its non-vegetarian comparators share most of these behaviors.

Ultra-processed vegan foods. The Satija finding is a persistent reminder: vegan cookies, refined-grain pasta, and sugary beverages are plant-based and unhelpful. The proliferation of ultra-processed meat analogues raises the possibility that a vegan diet composed primarily of packaged substitutes and refined grains may not replicate the benefits seen in whole-food cohorts. The mechanism is plausible; longer-term cohort data on heavy substitute users are still accumulating.

Reverse causation and ascertainment. Some people adopt plant-based diets because of existing cardiovascular concerns, which can bias observational estimates in either direction. Newer cohorts with baseline screening partially address this.

Heterogeneity. “Vegetarian” in EPIC-Oxford and “plant-based” in Kim et al. do not describe the same diets. Pooled estimates blur real differences between whole-food and processed patterns — which is exactly what Satija’s indices were designed to separate.

The bottom line

The cardiovascular case for whole-food plant-based eating is among the most solidly supported in clinical nutrition. The effect converges from cohorts, trials, and mechanistic studies; it is mediated through LDL, blood pressure, and vascular function; and it is large enough to matter clinically. The key caveat is that the label “plant-based” does not guarantee the diet. A pattern built on legumes, whole grains, vegetables, fruits, nuts, and seeds is what the literature describes. A pattern built on refined flour, added sugar, and packaged substitutes is not — and the data suggest it behaves differently.

For prevention, the intervention is unglamorous and well-established. For established disease, the Ornish and Esselstyn work suggests that more aggressive dietary change is worth discussing with a cardiologist, particularly alongside standard medical therapy rather than in place of it.