They measured body composition. Is that your goal? Remember this is for 70 year olds (and a small sample size), even if you got to the ideal 70 year old body, you won't look that good naked - which is all body composition is directly good for. My guess is you care more about health (you should break this down into more details of what health means) and long life.
Body composition is a factor in health and long life. However there are many confounding factors if that is your goal and so you cannot draw any conclusions unless the sample size is very large, and the study runs over a very long time. Thus we get a lot of small studies that study something easy and hope that this is a good proxy for what we really care about. Sometimes science eventually figures the proxy is good, sometimes not, but often we still don't know. (meta studies have been really helpful here)
Large sample sizes are very expensive to study, a grad student without large grants can study 50-100[1] people alone, which makes the study cheap enough that they can do it. This was a 6 month study, again making it something a grad student could do leaving plenty of time to then write the paper and get it published. (Each subject was studied for 6 months, I'm not clear if they were all studied at once, or if different subjects had different start/end dates). All respect for the grad students who do this - despite all the problems I've pointed out[2], they still did a lot of work.
[1] I've never been a grad student, much less one in a field where you would study this. The 50-100 number feels right in my uneducated opinion, but if someone with more knowledge says something I accept their correction in advance.
[2] I wonder what other problems someone in this field could point out.