If a test on normal mice showed a statistically significant increase in cachexia for those on keto, then it might actually mean something. The fact that the mice were already predisposed to it means that the data says very little of use, not to mention all the standard disclaimers that come with animal trials.
All basic science comes with limitations, just like human clinical trials (heterogeneity and limited scope of control, let’s not even talk about observational). But I don’t think it’s of very little use. I’m just talking generally, I didn’t go through the paper thoroughly to give a technical comment, but if a treatment exacerbates a pre-existing condition or triggers a condition in an at-risk model, I’d think it can be quite useful because such circumstances do occur plenty in humans.