It seems to be the calories. At least in animal models they aren't comparing against obesity. I believe for a non-restricted diet, then it's best to exercise rather than not, but for caloric restriction, I think the exercise effect isn't as significant.
There are a lot of caveats with these studies, however it would be terrible design if they exercised one group and not the other. Here's a study where they found that exercise increased healthspan, but not lifespan in mice: https://www.ncbi.nlm.nih.gov/pubmed/24472376 It doesn't necessarily follow that if caloric restriction is good for longevity, that extensive exercise is bad. Not all metabolism is equal, and it's not even clear that the CR effect is primarily due to metabolism. It might primarily be due to an up-regulation of genes that promote longevity.If the caloric restriction advocates really are correct, then marathon runners, bodybuilders, and powerlifters are putting themselves at risk of an early death much like morbidly obese people or smokers.
Exercise is bunk. If you are healthy, you don't need it: if you are sick you should not take it. -Henry Ford
Yeah, good luck training for anything of consequential length and difficulty if you're under caloric restrictions. I lost 3 lbs yesterday from one difficult hike alone and would have lost more / not easily be able to recover under a restriction. Balancing a diet in this situation is, of course, still very important.
I second Odder somewhat on this... Every study of starvation shows that as calories are reduced, voluntary motion reduces too. We know exercise is correlated with calories consumed in animals. This study does not control for both total calories and total energy expenditure, so it isn't correct to say the result is due to one and not the other. Diet studies are super hard because it's impossible to change only one thing and the sheer number of variables is massive.it would be terrible design if they exercised one group and not the other