The researchers at Harvard couldn't run a randomised controlled trial where half the participants ate burgers for two decades and the other half didn't, but they could use statistical techniques as the next best thing.
BBC: Red meat: What is a 13% increase in the risk of death?