The correlations between higher weight and greater health risk are weak except at statistical extremes. The extent to which those correlations are causal is poorly established. There is literally not a shred of evidence that turning fat people into thin people improves their health. And the reason there’s no evidence is that there’s no way to do it.