US Army veterans do have a higher rate of arthritis but their days are quite different from the "run 3-5 days a week" that most people think of when talking about recreational runners.
And the pacemaker comment stood out so I did a bit of digging and found a study [1] you might be referring to. Again, the effects were strong only in the heavy-duty-exercisers/pro/semi-pro cross-country skiier group. Additionally, this didn't offset the gains to cardiovascular or mortality risk - that group was still "healthier."
Humans are evolved to run, but not to have heavy frames and not carrying material for fighting wars.
I'm also part of the barefoot running army and tend to think that the braking forces from shoes have a role to play in knee problems (I personally stopped having them when I started running barefoot so that's where my bias comes from.)
After surviving early years, people still used to get old. Infant mortality was just way higher bringing the average down. (And that those metrics often compare to poor peasants and our paleo ancestors were not peasants)
> One quirk to these model tables I want to note, because it sometimes confuses folks, is that they express ‘life expectancy’ not as a total expected age, but as average years of life expectancy from a given age, so a 25-year-old with 26.6 years of life expectancy is expecting to life to age 51, not just a year and a half.
> What we see in these models is that life expectancy (female:male) at birth is very low, 25:22.8, but after the first year rises dramatically to 34.9:34.1 (note the gender gap narrowing) and by age 5 to 40.1:39.0 (remember to add the five years they’ve already lived). So life expectancy goes up quite a lot over the first several years of life, which is not, intuitively, the pattern we expect: we normally assume the more you’ve lived, the less years you have left.