Kind of interesting you should say that, because the two intervention studies included in this review show that they aren't. The observational studies are anything but consistent, much less causative, and the bench research seems to be virtually worthless. For something as important as obesity...