The foster GD found no relation between insulin and diet assignement so its probably not the insulin (which happens to be a satiety hormone) - it likely has more to do with what people are actually eating as indicated by the fact that foster's study shows a much greater regain in low-carbers (because such a caloric deficiency is more difficult to maintain)