Skip to main content
. 2021 Dec 22;601(7892):263–267. doi: 10.1038/s41586-021-04224-5

Extended Data Table 2.

Association between cancer risk and diet animal content

graphic file with name 41586_2021_4224_Tab2_ESM.jpg

Models explain variance in a, CMR or b, ICM in function of body mass, life expectancy and diet animal content. Diet animal content is characterized by variables on three taxonomic level: 1. animal content (including any vertebrate or invertebrate prey); 2. invertebrate or vertebrate prey; and 3. within vertebrates fish, reptiles, birds and mammals. Each diet item is coded as a two-level factor: rarely/never occurring in diet and representing the primary/secondary food item of the species. Each diet variable is added one by one to a base model containing the two significant predictors of cancer risk: body mass and life expectancy. Akaike’s Information Criteria (AIC) are directly comparable among models with the same independent variable. All models are PGLS regressions.