By Jack Norris, RD
Contents
- Introduction
- Cancer incidence among vegans
- Incidence of specific cancers among vegans
- Cancer mortality among vegans
- Cancer progression on a vegan diet
- Footnotes
- Bibliography
Summary: Vegans appear to have ~20% lower cancer incidence than other dietary groups. No sufficiently large or long studies have assessed cancer mortality among vegans. Lower average body mass index (BMI) explains part of the reduction; other dietary factors likely contribute. Whole‑food plant‑based diets show some promise for slowing progression, but evidence is limited; the plant‑based component may matter more than the whole‑food component.
Introduction
Do vegans have lower cancer incidence and mortality? If diet is responsible, is lower average BMI (Key et al., 2014; Fraser et al., 2025) and lower weight gain over time (Rosell, 2006) most important? Can a vegan diet slow cancer progression? If so, must it be whole‑foods?
Cancer incidence among vegans
Two large cohorts have reported cancer rates for vegans versus other diet groups. The Adventist Health Study‑2 (AHS‑2) includes over 5,000 vegans;A participants are Seventh‑day Adventists, who have low smoking and drinking rates. The other is the British Oxford Vegetarian Study/EPIC‑Oxford cohort, which recruited participants across the United Kingdom, including 2,246 vegans.
Table 1 shows cancer incidence compared with regular meat‑eaters (those who eat meat and are not semi‑vegetarian or pescatarian).
| Table 1. Cancer incidence for vegans versus regular meat-eaters | ||||
|---|---|---|---|---|
| Study | Authors | Year | Follow-up | Incidence (95% CI) |
| British cohort | Key et al. | 2014 | 14.9 yrs | RR 0.81 (0.66, 0.98) |
| adjusted for BMI | RR: 0.82 (0.68, 1.00) | |||
| AHS-2 | Fraser et al. | 2025 | 7.9 yrs | HR 0.76 (0.68, 0.85) |
| adjusted for BMI | HR 0.82 (0.73, 0.92) | |||
| CI – confidence interval • HR – hazard ratio • RR – relative risk | ||||
These two studies found lower cancer incidence in vegans than meat‑eaters, with lower BMI explaining part of the difference.
Both studies adjusted for non‑dietary cancer risk factors. The British cohort adjusted for age, sex, smoking, alcohol, physical activity, parity, and oral contraceptive use. AHS‑2 adjusted for age, sex, race, education, height, physical activity, family history of breast cancer, menopausal factors, parity, oral contraceptive use, hormone replacement therapy, and major cancer‑screening behaviors.
In both cohorts, lacto‑ovo‑vegetarians and pescatarians had ~10% lower cancer risk than meat‑eaters; in AHS‑2, these differences lost statistical significance after adjusting for BMI.
The AHS-2 researchers note that “The incidence of total cancers is 30% lower in United States Adventists than in a United States census population. Given that vegetarians are approximately half of our analytic population and have a 12% lower rate than nonvegetarian Adventists, this suggests that nonvegetarian Adventists experience 25.5% and vegetarians 34.5% less cancer than average Americans.”
AHS‑2’s 7.9‑year follow‑up is short for cancer outcomes. A strength is that, although vegans screened for cancer less often than other diet groups (Oh, 2021; Oh, 2022), the analysis adjusted for screening.
Observational studies show associations, not causation, and long randomized diet–and-cancer trials are impractical. In practice, prospective observational evidence informs causal inference, even if correlation isn’t causation; stronger statistics increase, but don’t guarantee, causal likelihood. Overall, the studies in Table 1 suggest a vegan diet may modestly reduce cancer risk; longer‑term, broader studies are needed.
Incidence of specific cancers among vegans
In the British cohort, site-specific cancer incidence did not differ significantly between vegans and meat‑eaters, whereas AHS‑2 reported some significant differences (Table 2).
| Table 2. AHS-2 cancer types: vegans vs meat-eaters (Significant findings only) |
|
|---|---|
| Cancer type | HR (95% CI) |
| Breast cancer by age 65 | 0.69 (0.51, 0.92) |
| Adjusted for BMI | 0.73 (0.54, 0.98) |
| Prostate cancer by age 65 | 0.57 (0.41, 0.80) |
| Adjusted for BMI | 0.58 (0.41, 0.81) |
| Medium frequency cancers | 0.77 (0.66, 0.89) |
| Adjusted for BMI | 0.85 (0.73, 1.00) |
| CI – confidence interval • HR – hazards ratio • Medium frequency cancers: melanoma, endometrial, renal, urothelial, thyroid, ovarian, central nervous system, lung, rectal, pancreatic, primary liver, stomach, esophageal, mouth-pharyngeal-laryngeal, lymphoma, lymphoid, and myeloid leukemias. Source: Fraser et al., 2025. |
|
Cancer mortality among vegans
The only report including cancer deaths among vegans is Key et al. (1999): among 753 vegans, only 5 cancer deaths were observed—insufficient for statistical significance.
Cancer progression on a vegan diet
There are no observational studies of cancer progression among vegans, but two clinical trials have tested whole‑food plant‑based diets.
Ornish et al. (2005) ran a randomized controlled trial of 93 men with low‑risk, biopsy‑proven prostate cancer on active surveillance, assigning them to an intensive lifestyle intervention or usual care. The intervention combined a very‑low‑fat (10% of calories) vegan diet supplemented with soy, fish oil, vitamins C and E, and selenium; moderate aerobic exercise (walking); daily stress management (including yoga and meditation); and weekly support groups. After 12 months, no men in the intervention group required conventional treatment, versus six in the control group due to rising PSA or progression. Mean PSA fell 4% in the intervention group but rose 6% in controls (p=0.016). After 5 years, telomere length—a marker of disease risk and premature death—was greater in a subset of intervention participants than controls, and adherence over 5 years correlated with telomere length (Ornish, 2013). These results are notable, but the multimodal intervention makes the diet’s specific contribution unclear.
Campbell et al. (2024) randomized women with metastatic breast cancer to an 8‑week, calorically unrestricted, whole‑food plant‑based diet (n=21) or control (n=11). The intervention included weekly assessments and provided meals of fruits, vegetables, whole grains, legumes, nuts, and seeds, with no animal products or added oils/solid fats. Assessments occurred at baseline, 4 weeks, and 8 weeks. Within the intervention group, fasting insulin, insulin resistance, total cholesterol, and LDL cholesterol decreased. The authors report no significant changes in cancer markers at 8 weeks, though trends were more stable in the intervention group, and conclude the diet is well tolerated; longer trials are needed to assess effects on progression.
It’s often assumed that avoiding processed foods yields better outcomes than regularly eating them, but cancer‑progression data don’t clearly support that. Liu et al. (2024) divided men with prostate cancer into quintiles across 18 food groups: a plant‑based dietary pattern was inversely associated with progression risk, but the “healthy” plant‑based pattern was not—possibly because some “unhealthful” plant foods still outperformed animal foods.
Footnotes
A. Most AHS-2 reports do not provide participant counts by diet group. The “over 5,000” vegans cited above comes from Orlich et al. (2013; n=5,548) and Dunneram et al. (2024; n=5,225). Participant counts are usually reported at baseline; Dunneram et al.’s lower figure may exclude individuals who had died by the time it was compiled.