[Daishiman]

Those findings are pretty robust as there are hundreds of experiments with an aggregate of tens to hundreds of thousands of subjects on the effects of moderate exercise and eating fiber. Fiber has pretty much linear benefits in all-cause mortality up until something like 50g per day and this shows up again and again. Likewise with the effects of moderate exercise.

More than the positive effects, the list does not list things that account for negative effects, like obesity (very strongly correlated with bad health outcomes), tobacco, alcohol, high non-HDL cholesterol thorough whatever eating habits enable it, etc.

The thing about the negatives is that they depend on dose across your lifetime, so being obese for several years, or having high levels of cholesterol has long-term impact even if you reverse the trend.

If you look at all that in the aggregate I think those factors can explain differences in life span quite well. Then again you still need to look at an individual's genetics and the randomness of cancer and various diseases that, while still affected by the previous factors, have enormous variance.