My guess is the control group wasn't made of people with no history of antibiotics. They just used the model to predict the risk difference between zero and on

My guess is the control group wasn't made of people with no history of antibiotics. They just used the model to predict the risk difference between zero and one time of taking antibiotics. The control was people from the general population, and the "test" was people who had already developed IBD. Though there are people, like me, who are allergic to many common antibiotics. Between some bad reactions in childhood and an emergency three years ago (over two decades), I never had antibiotics.