The take-away, however, is that the commonly understood models for relating microorganisms to health, such as the ones that guide doctors' decisions and health policy, do not predict that antibiotics would cause animals to grow more muscle and less fat -- or have any similar effect -- and are therefore likely to be incomplete or wrong in important ways. Most plausible amendments to the models that would cause them to predict the effects we observe in farm animals would also cause them to predict that antibiotics given to humans could have similarly complex and large effects.