Vegan, vegetarian, omnivore, carnivore: what the labels actually buy you

In 200,000 adults followed for over two decades, a plant-based diet heavy in refined grains, potatoes, and sweetened drinks carried 32% higher coronary heart disease risk, while one built on whole grains, vegetables, nuts, and legumes carried 25% lower risk. Both count as "plant-based" (Satija, J Am Coll Cardiol 2017). The same split runs through every camp: the pattern name tells you what someone excludes, not what they actually eat.

Across 86 cross-sectional and 10 prospective studies, vegetarians had about 25% lower death from ischemic heart disease and 8% lower cancer incidence, but no significant advantage in all-cause mortality (Dinu, Crit Rev Food Sci Nutr 2017). The sharpest test is EPIC-Oxford, which recruited health-conscious meat eaters as the comparison group: there, vegetarians and omnivores died at the same rate (Appleby, Am J Clin Nutr 2016).

In the Adventist Health Study-2, following 73,000 adults, pesco-vegetarians had the lowest all-cause mortality of any group: 19% below omnivores, ahead of vegans at 15% and lacto-ovo vegetarians at 9% (Orlich, JAMA Intern Med 2013). A mostly-plant diet plus fish also describes the traditional diets of most of the longest-lived populations studied.

Vegans did well on mortality in Adventist data (Orlich, JAMA Intern Med 2013), but the diet has documented failure modes. B12 exists only in animal foods and fortified products, and deficiency rates among unsupplemented vegans are high in every population studied (Pawlak, Nutr Rev 2013). In EPIC-Oxford, vegans had 43% more total fractures and more than double the hip-fracture rate, largely tracking low calcium and protein intake (Tong, BMC Medicine 2020).

The omnivore average in cohort studies is dragged down by its processed tail. Remove that tail and omnivores match the plant camps: EPIC-Oxford's health-conscious meat eaters equalled vegetarians (Appleby, Am J Clin Nutr 2016), and in the PREDIMED trial a Mediterranean diet including fish, poultry, and dairy cut major cardiovascular events by about 30% against a low-fat control (Estruch, NEJM 2018).

Processed meat is classified as a Group 1 carcinogen, with each daily 50g portion (roughly two slices of bacon) linked to 18% higher colorectal cancer risk (Bouvard, Lancet Oncol 2015). In pooled US cohorts, higher processed meat intake was associated with more cardiovascular disease and earlier death (Zhong, JAMA Intern Med 2020). Unprocessed red meat carries a smaller signal; processed is the clear problem.

There are no cohort or trial data following all-meat eaters for years; the best available is a self-selected survey reporting short-term satisfaction, with no mortality follow-up (Lennerz, Curr Dev Nutr 2021). The nearest studied pattern is an animal-based low-carb diet, and in 25 years of follow-up it was associated with higher all-cause mortality, while the plant-based low-carb version was associated with lower (Seidelmann, Lancet Public Health 2018). An all-meat diet also removes fiber entirely, and every 8g of daily fiber tracks with 5 to 27% lower mortality and disease risk (Reynolds, Lancet 2019).

Modeling of the global cohort data puts the biggest life-expectancy gains not in avoiding meat but in adding foods: legumes, whole grains, and nuts top the list, worth up to 10 years if the shift starts in early adulthood (Fadnes, PLoS Med 2022). Every one of those foods is available to every camp; the camps differ mainly in how easy they make the additions.