Epidemiological studies of humans typically include plants in the diet, making it difficult to observe any signal of plant-induced harm.

“if you look at groups of people which is mostly epidemiology there really are no studies that have a group without plants and a group of plants everybody's eating plants some people eat more meat some people eat less meat but everyone is eating plants how are we going to see even an epidemiologic observational signal that plants are harmful for humans in that sort of a design study”