Take Three: Direct Evidence that Vegans have Lower Cancer Rates
On November 26, 2012, I wrote about a report from Adventist Health Study-2 looking at the cancer rates of vegans, Direct Evidence that Vegans have Lower Cancer Rates.
Two days later, after a reader brought it to my attention, I posted a Follow-Up, pointing out that the authors had not included physical activity as a variable in their multivariable adjustment model. For a number of reasons, I suspected that if they had adjusted for physical activity, the finding that vegans have lower cancer rates would no longer be statistically significant.
I have since corresponded with one of the authors of the paper, Yessenia Tantamango, who told me that when they included physical activity in the multivariable analysis (in the model that did not include BMI), vegans had a cancer rate of .85 (.73-.99) which is statistically significant.
The original finding for the vegan cancer rate was .84 (.72-.99). An upper confidence interval limit of .99 is only of borderline significance; if it were 1.00 it would not be considered significant. In other words, I realize we are splitting hairs here. But I wanted to follow-up to let readers know that after adjusting for physical activity the finding was still statistically significant and so, by conventional nutrition science standards, it is legitimate to say that vegans had a lower rate of cancer in this study and there is reason to think it was due to their diet.
Tantamango-Bartley Y, Jaceldo-Siegl K, Fan J, Fraser G. Vegetarian diets and the incidence of cancer in a low-risk population. Cancer Epidemiol Biomarkers Prev. 2012 Nov 20. | link