What heritability actually shows is percent of variance is accounted for by genetics/inborn factors. How much variance there is can be environmental - that is, environment that is shared among all individuals in the study population, not just family environment. E.g. if weight was 100% heritable then genetic profile A would be 20th percentile and genetic profile B would be 90th percentile no matter what, but in 1920 that might have meant 110 pounds and 200 pounds, whereas today it might mean 130 pounds and 500 pounds. The actual weight of an individual still cannot be determined only by genetics.

Height is a good example to understand the counterintuitive math of heritability: it has become much more heritable over time because it's much more rare for people to experience malnutrition that impedes them from reaching their maximum genetic potential. The genetic pool hasn't changed significantly, and it has become more "genetically determined," but the increase in average height or the increase of height in subsequent generations of a family is very much due to environment.

Type 2 diabetes gives you a higher risk for Alzheimers, yes.

Exercise (especially cardio) is so effective for mood and appetite regulation for me that I'm experimenting with running 6-7 days a week just to get "booted up" correctly at the beginning of the day. I've always had scheduled rest days before, but they were always just less nice, and 3 km at easy pace really isn't a lot of stress on the body.

I dunno, there's significant ground between regular running and struggling to walk. You'd think if the running wasn't helping (or an index of something else that helps, like general health motivation) then he'd know plenty of people who don't run but haven't lost much ability.

For whole organisms, it was hamsters. But they also demonstrated it in human organs, and better COVID outcomes among humans like your wife already taking it for other reasons. There's no way to know what her individual reaction to the virus would have been if she wasn't on the medication.

It's not a full double blind trial yet, but it's a lot more steps done than just rodents.

Risk applies to the whole group that was measured, some were hospitalized and some weren't, so being in that group gives you a certain risk. In this context it's essentially equivalent to "rate."

It's hard to parse from just the abstract if they did their analysis correctly, but they did say they were looking for an "effect modifier" which I would interpret as an interaction effect. That is to say, vaccination is designed to reduce your hospitalization risk, and physical activity may also do this even in unvaccinated people, but the effect of having both on your side is more than just adding up the two factors individually.

Again, I can't tell from what I'm able to access if they actually showed that, but it sounds like that is what they are claiming.