And this is absolutely something one needs to consider when reading medical studies -- if they only use animal (usually mice) models, there's a decent chance the conclusions are not directly transferable to humans.