Vegetarian and vegan diets and risks of total and site-specific fractures: results from the prospective EPIC-Oxford study

TLDR

Limited prospective evidence exists on fracture risk differences among vegetarians, vegans, and non‑vegetarians. This study aimed to evaluate fracture risks in a large prospective cohort of EPIC‑Oxford participants, including substantial numbers of non‑meat eaters. Participants were classified into four diet groups, fracture outcomes were linked to hospital records and death certificates, and multivariable Cox regression was used to estimate risks over an average 17.6‑year follow‑up. Compared with meat eaters, fish eaters, vegetarians, and vegans had higher hip fracture risks (HR 1.26, 1.25, 2.31), and vegans also had increased total, leg, and other main site fracture risks (HR 1.43, 2.05, 1.59), while wrist, ankle, and arm fracture risks were unchanged.

Abstract

There is limited prospective evidence on possible differences in fracture risks between vegetarians, vegans, and non-vegetarians. We aimed to study this in a prospective cohort with a large proportion of non-meat eaters.In EPIC-Oxford, dietary information was collected at baseline (1993-2001) and at follow-up (≈ 2010). Participants were categorised into four diet groups at both time points (with 29,380 meat eaters, 8037 fish eaters, 15,499 vegetarians, and 1982 vegans at baseline in analyses of total fractures). Outcomes were identified through linkage to hospital records or death certificates until mid-2016. Using multivariable Cox regression, we estimated the risks of total (n = 3941) and site-specific fractures (arm, n = 566; wrist, n = 889; hip, n = 945; leg, n = 366; ankle, n = 520; other main sites, i.e. clavicle, rib, and vertebra, n = 467) by diet group over an average of 17.6 years of follow-up.Compared with meat eaters and after adjustment for socio-economic factors, lifestyle confounders, and body mass index (BMI), the risks of hip fracture were higher in fish eaters (hazard ratio 1.26; 95% CI 1.02-1.54), vegetarians (1.25; 1.04-1.50), and vegans (2.31; 1.66-3.22), equivalent to rate differences of 2.9 (0.6-5.7), 2.9 (0.9-5.2), and 14.9 (7.9-24.5) more cases for every 1000 people over 10 years, respectively. The vegans also had higher risks of total (1.43; 1.20-1.70), leg (2.05; 1.23-3.41), and other main site fractures (1.59; 1.02-2.50) than meat eaters. Overall, the significant associations appeared to be stronger without adjustment for BMI and were slightly attenuated but remained significant with additional adjustment for dietary calcium and/or total protein. No significant differences were observed in risks of wrist or ankle fractures by diet group with or without BMI adjustment, nor for arm fractures after BMI adjustment.Non-meat eaters, especially vegans, had higher risks of either total or some site-specific fractures, particularly hip fractures. This is the first prospective study of diet group with both total and multiple specific fracture sites in vegetarians and vegans, and the findings suggest that bone health in vegans requires further research.