Mihrshahi · Preventive medicine 2017 · prospective cohort study · n=243,096

Vegetarian diet and all-cause mortality: Evidence from a large population-based Australian cohort - the 45 and Up Study.

Cited 83 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective population-based cohort study

PubMed 28040519 · doi:10.1016/j.ypmed.2016.12.044 · record verified 2026-08-29

What was done

Data from the 45 and Up Study in New South Wales, Australia, were analyzed for 243,096 men and women aged 45 years and older (mean age 62.3 years, 46.7% men). Participants were categorized at baseline via self-report questionnaire into complete vegetarians, semi-vegetarians (eating meat at most once per week), pesco-vegetarians, and regular meat eaters. All-cause mortality was determined through linked registry data over a mean follow-up of 6.1 years (16,836 total deaths). Cox proportional hazards models quantified associations between diet categories and all-cause mortality with adjustment for potential confounding factors.

What was found

Following extensive adjustment for potential confounders, there was no significant difference in all-cause mortality for vegetarians versus non-vegetarians (HR = 1.16, 95% CI: 0.93–1.45). Similarly, no statistically significant difference was observed for pesco-vegetarians (HR = 0.79, 95% CI: 0.59–1.06) or semi-vegetarians (HR = 1.12, 95% CI: 0.96–1.31) versus regular meat eaters.

Why it matters

Previous reports of mortality benefits among vegetarians may have been influenced by healthy lifestyle confounding in non-population-based cohorts. In this large population-based cohort, vegetarian and related dietary patterns showed no independent protective association with all-cause mortality.

Limits

Diet was assessed only via baseline self-report questionnaire without accounting for dietary changes or overall diet quality. Follow-up was relatively short (mean 6.1 years), the cohort was restricted to adults aged 45 and older in Australia, and residual confounding cannot be ruled out.

Cited by