The authors base their hypothesis on connections between three distinct lines of evidence. First, they demonstrate that people with a wide range of conditions, from obesity and inflammatory bowel disease to type 2 diabetes and cardiovascular disease, have altered microbiomes. Next, they show that altered microbiomes, when taken from diseased people and put into animal models, cause disease. Finally, they provide evidence that the microbiome is naturally transmissible, for example: Spouses who share a house have more similar microbiomes than twins who live separately.