Large diet study links vegetarian patterns to lower cancer risk

A pooled analysis of about 1.8 million people followed for an average of 16 years reports that vegetarians had lower incidence of five cancers than meat-eaters, while vegans showed a higher incidence of colorectal cancer. The study, published in the British Journal of Cancer and reported by The Guardian and Business Insider, combined cohorts from the UK, US, India and Taiwan, including 1,645,555 meat-eaters, 57,016 poultry-eaters (no red meat), 42,910 pescatarians, 63,147 vegetarians and 8,849 vegans.

The headline results are relative risks across 17 cancer outcomes, not a controlled dietary trial. The “vegetarian” category is a broad umbrella that can include dairy and eggs; “vegan” is a much smaller group, and therefore more sensitive to misclassification and to differences in screening, supplement use, and lifestyle clustering. Participants in several cohorts were also unusually health-conscious: Business Insider notes the inclusion of Seventh-day Adventists in the US, a population with distinctive norms around smoking, alcohol and diet—factors that can move cancer risk independent of meat intake.

The reported associations cut both ways. Compared with meat-eaters, vegetarians had lower risk of multiple myeloma (31%), kidney cancer (28%), pancreatic cancer (21%), prostate cancer (12%) and breast cancer (9%), according to The Guardian. But the same analysis found nearly double the risk of oesophageal squamous cell carcinoma among vegetarians, and a 40% higher risk of bowel (colorectal) cancer among vegans. Researchers floated nutrient pathways rather than a single “plant-based is better” mechanism: low calcium intake among vegans (around 590 mg/day versus a UK recommendation of 700 mg/day) is one candidate explanation for colorectal risk, while B-vitamin or riboflavin shortfalls were suggested as possible contributors for the oesophageal finding.

What the study can adjust for—BMI, smoking and other measured covariates—does not eliminate the core difficulty of diet epidemiology: people self-select into diets, and the most important differences between groups are often the ones that are easiest to mis-measure. Screening behaviour is a recurring blind spot. If one group is more likely to undergo colonoscopy or PSA testing, incidence can rise even as mortality falls, simply because cancers are found earlier and more often. Socioeconomic status, alcohol patterns, physical activity, and long-run adherence also matter; a single baseline dietary questionnaire can turn decades of eating into a noisy label.

The authors themselves caution that the work cannot tell whether meat is causal, whether vegetarian patterns are protective, or whether the answer varies by cancer type. In the meantime, the cleanest “signal” in the paper is not a universal dietary prescription but the way different risk directions appear once the dataset is large enough to look beyond the usual red-meat-and-bowel-cancer pairing.

The analysis includes 8,849 vegans—small enough that a modest shift in who counts as vegan, who gets screened, or who supplements calcium can move the colorectal result by a lot.