This is based off a meta analysis and the problem with those is you can filter the data to prove almost anything you want. Many other studies have found a j-cu

This is based off a meta analysis and the problem with those is you can filter the data to prove almost anything you want. Many other studies have found a j-curve effect. So there is ambiguity, and there likely will be until someone does a long-term randomized controlled trial which is incredibly difficult.