INDEPENDENT THINKING. WHOLE-FOOD LIVING.NO ANIMALS / NO OIL
THENANODIET
← Nutrition & Adequacy / Health Desk

Adventist Mortality: The Combined Finding and the Vegan Uncertainty

Orlich’s 2013 cohort linked vegetarian patterns with lower mortality, while the vegan-specific estimate remained less certain.

Whole-food ingredients prepared without added oil

Illustrative food image, not a photograph of the study intervention.

A large cohort, not a diet assignment

Orlich and colleagues analyzed mortality in Adventist Health Study 2, a North American cohort of Seventh-day Adventist adults. The original recruitment included 96,469 men and women between 2002 and 2007; after exclusions, 73,308 participants remained in the mortality analysis. Diet was assessed at baseline using a quantitative food-frequency questionnaire. Researchers classified participants into nonvegetarian, semi-vegetarian, pesco-vegetarian, lacto-ovo-vegetarian and vegan patterns. This was a prospective cohort, not a randomized diet trial. Statistical adjustment addressed important demographic and lifestyle confounders, but the design still establishes associations between reported dietary patterns and subsequent mortality rather than proving that diet caused the differences.

Orlich MJ, Singh PN, Sabaté J, Jaceldo-Siegl K et al.. JAMA internal medicine, 2013. PMID 23836264 · DOI

The combined vegetarian result

There were 2,570 deaths during an average follow-up of 5.79 years. The reported mortality rate for the analytic population was 6.05 deaths per 1,000 person-years, with a 95% confidence interval from 5.82 to 6.29. Comparing all vegetarian groups combined with nonvegetarians, the adjusted hazard ratio for all-cause mortality was 0.88, with a confidence interval from 0.80 to 0.97. This is a relative comparison of hazards, not a 12-percentage-point drop in an individual’s absolute chance of dying. It also combines several dietary patterns and therefore should not be presented as a result exclusive to vegan participants.

Orlich MJ, Singh PN, Sabaté J, Jaceldo-Siegl K et al.. JAMA internal medicine, 2013. PMID 23836264 · DOI

The vegan estimate needs its interval

For vegans compared with nonvegetarians, the adjusted all-cause mortality hazard ratio was 0.85, with a 95% confidence interval from 0.73 to 1.01. The point estimate favored vegans, but the interval included 1.00, so the abstract does not establish a clear vegan-specific reduction. Other reported estimates were 0.91 for lacto-ovo-vegetarians, 0.81 for pesco-vegetarians and 0.92 for semi-vegetarians, each with its own uncertainty. Listing only the most favorable numbers would conceal the differences between groups. The combined vegetarian finding and the vegan-specific finding are related, but they are not statistically interchangeable conclusions from this analysis.

Orlich MJ, Singh PN, Sabaté J, Jaceldo-Siegl K et al.. JAMA internal medicine, 2013. PMID 23836264 · DOI

Limits on a longevity headline

The study reported significant associations for several cause-specific mortality categories, including cardiovascular mortality, and noted that associations in men were larger and more often significant than those in women. Those details caution against claiming an identical benefit for everyone. Mortality was identified through 2009 using the National Death Index, and the mean follow-up was 5.79 years; the analysis did not measure extra lifetime years gained from changing diet. Nor did the dietary categories isolate a no-extracted-oil intervention. The defensible message is that vegetarian patterns were associated with lower mortality in this cohort, with uncertainty remaining for specific groups and for causal interpretation.

Orlich MJ, Singh PN, Sabaté J, Jaceldo-Siegl K et al.. JAMA internal medicine, 2013. PMID 23836264 · DOI