

And this is why I linked the article first instead of the paper.
The line you’re quoting is the authors explaining they have a large enough sample size to detect differences. They could detect a 4.2% difference in their sampling of women with 97% confidence, at a 5% significance threshold. They are saying they are extremely confident they would be able to detect a difference, but didn’t.
Oh yes, absolutely, that’s another conclusion to draw from this paper. Exercise is still good for you, but it’s not a means of losing weight.