Vegetarians have lower rates of several common and rare cancers

An analysis of the Adventist Health Study-2 (AHS-2) cohort was published in The American Journal of Clinical Nutrition: among 79,468 residents of the United States and Canada without cancer at the start, vegetarians (all types together) had a 12% lower risk of all cancers ( HR 0.88; 95% CI 0.83-0.93), and for “moderately rare” tumors (e.g. stomach, lymphomas) - by 18% ( HR 0.82; 0.76-0.89). Separately, the risk of colorectal cancer (HR 0.79; 0.66-0.95), stomach cancer ( HR 0.55; 0.32-0.93) and lymphoproliferative tumors ( HR 0.75; 0.60-0.93) was statistically significantly lower.

Background

Dietetics in cancer prevention has long gone beyond individual products and increasingly looks at dietary patterns. The starting points are clear: back in 2015, the International Agency for Research on Cancer (IARC) classified processed meat as a group 1 carcinogen (reliably causes colorectal cancer), and red meat as a “probable carcinogen”; every 50 g of processed meat per day is associated with an increase in the risk of CRC by about 18%. In parallel, the World Cancer Research Fund/American Institute for Cancer Research recommend limiting red and processed meat, emphasizing the “convincing” evidence for CRC. Against this background, there is growing interest in plant-based diets as a “portfolio” of habits that simultaneously reduces exposure to meat carcinogens and adds fiber, polyphenols and other protective factors.

However, large population cohorts have methodological difficulties: they have few strict vegans/vegetarians, and "non-vegetarians" are often very heterogeneous in lifestyle, which makes clean comparison difficult. This niche has historically been filled by the Adventist Health Study-2 (AHS-2) - a North American cohort of Adventist church members with a high proportion of people on plant-based diets and relatively homogeneous habits (little smoking and alcohol). Early publications of AHS-2 showed the advantages of "plant-based" patterns in relation to body weight, diabetes and a number of outcomes, and also hinted at a decrease in the risks of individual tumors, but required longer follow-up and detailed analysis by localization and subtypes of vegetarianism. In fact, the key task of the new work in AJCN is to fill these gaps by assessing the relationship of different branches of plant-based nutrition (veganism, lacto-ovo, pesco-, semi-) with the risk of both common and "moderately rare" cancers.

The current analysis of AHS-2 included 79,468 US and Canadian adults without cancer at baseline, with a median/mean follow-up of about eight years. Diet was recorded with a validated frequency questionnaire, outcomes were cross-checked with cancer registries, and proportional hazards models were adjusted for a wide range of covariates (age, sex, ethnicity, education, smoking, alcohol, physical activity, etc.). This design allows us to simultaneously see the “total” cancer risk and lift the veil on less common locations where power is usually limited - and to take into account that even non-vegetarians in this cohort eat “healthier” than the average population, making the comparison rather conservative.

The biological plausibility of the expected effect is also supported by accumulating mechanistic evidence: reducing red/processed meat reduces exposure to nitrosamines and cooked foods, and shifting to whole plant foods increases fiber and short-chain fatty acid intake, modulates the microbiota, and reduces chronic inflammation—key links in gut carcinogenesis and beyond. These mechanisms are not directly tested in the epidemiological paper, but provide a context in which to logically place the AHS-2 results.

What kind of work is this?

• Design: Prospective cohort AHS-2 (inclusion 2002-2007), mean follow-up 7.9 years; cancer cases were linked to US and Canadian registries. Diet was assessed with a validated frequency questionnaire, participants were categorized as vegans, lacto-ovo-, pesco-, semi-vegetarians, and non-vegetarians; hazards were calculated as proportional hazards with multiple imputation of missing items.
• Sample: 79,468 people were included in the analysis (from an initial ~96,000), about 26% were black participants; ~half followed vegetarian regimens.

Main results with figures

• All-cause cancer: HR 0.88 for all vegetarians vs. non-vegetarians.
• "Moderately rare" cancers in general: HR 0.82 (includes stomach, lymphomas, etc.).
• By localization (significant signals):
  • Colorectal cancer: HR 0.79.
  • Gastric cancer: HR 0.55.
  • Lymphoproliferative tumors (including lymphomas): HR 0.75.
• Vegetarianism subtypes: According to the pooled test, differences from non-vegetarians were noted for breast, prostate, colorectal, lymphoma, and all cancers (p-values for pooled comparisons <0.05). The Loma Linda press release clarifies: vegans have the most pronounced risk reduction for common sites (breast/prostate), and pesco-vegetarians - especially for colorectal cancer; however, some signals for lung, ovary, and pancreas are still suggestive and require strength.
• Role of body weight: Adjustment for BMI slightly attenuated the effects - part of the benefit may be due to lower weight in vegetarians (mediator).
• Important about the comparison: even non-vegetarians in AHS-2 are generally healthier than the average population (less meat and alcohol). Therefore, the real differences when compared with the "usual" Western diet may be even greater, the authors emphasize.

Why is this important?

• The study closes a long-standing gap: associations with plant-based regimens have long been described for "common" cancers (breast/prostate/colorectal), but there was little reliable data on stomach and lymphomas. Here is a large cohort, long-term observation, and accurate registries.
• No increased risk was found with vegetarian diets for any type of cancer, which has important implications for the debate about the safety of plant-based diets.

How does this compare to previous data?

The results fit into the line of evidence on the harm of red and processed meat for colorectal cancer and the benefits of plant-based dietary patterns. The novelty is the systematic look at less common localizations and the comparison of subtypes of vegetarianism in one model.

Restrictions

• Observational design: links diet and risk but does not prove causation; possible residual confounders (screening, income, lifestyle).
• Diet was measured at baseline; changes over the years were not taken into account.
• For some "rare" cancers, power is still limited; signals require confirmation in other cohorts.

What does this mean for the reader (practical angle)

• You don't have to be "100% vegan" to benefit: even a shift toward plant-based foods (whole grains, legumes, nuts, fruits/vegetables) and less red/processed meat is a realistic risk-reduction strategy.
• Watch your BMI: Part of the effect is mediated by body weight - a plant-based plate helps maintain weight.
• Remember about completeness: iron, B12, iodine, omega-3 - plan your diet or supplements with a doctor/nutritionist (especially on a strict vegan diet). This is a general rule, not a conclusion of a specific article.

Source: AJCN article abstract (August 2025): Fraser GE et al. Longitudinal associations between vegetarian dietary habits and site-specific cancers in the Adventist Health Study-2 North American cohort - main risk estimates and methodology. doi: 10.1016/j.ajcnut.2025.06.006