There are many controlled intervention studies into diet and exercise. It is done. They provide the only results worth any consideration. More often than not

There are many controlled intervention studies into diet and exercise. It is done. They provide the only results worth any consideration. More often than not regression analyses and "risk-adjustment" models are invalidated.