The short answer
- In a study of 110 experienced runners aged 18 to 60, running speed declined with age because stride length shortened. Step rate did not change.
- Peak ankle torque and ankle power declined with age. Hip and knee measures did not correlate with age at all.
- That makes calf and ankle strength the single most targeted thing an older runner can train.
- Recreational running is associated with less hip and knee arthritis than not running, so the joints are not the reason to stop.
- The things that change most are recovery time between hard sessions and tolerance for sudden increases, rather than the total you can handle.
Most advice for older runners is a list of things to do less of. The research points somewhere more specific and more useful: one joint does most of the changing, and it is not the one people worry about.
What actually declines
That last sentence is the useful one. The hip and the knee, the two joints older runners are usually warned about, did not track age in this sample. The ankle did.
So the picture of ageing in running is not a general fading. It is a specific loss of push from the calf and ankle, which shortens the stride, which slows you down.
What to do about it
Train the thing that is changing.
Calf and ankle strength is the most directly targeted work an older runner can do, and it is also the thing most commonly missing from a running week. Heavy calf raises, done properly, through full range, progressed over months. Then, once that base is there, springier work: skipping, low hops, short fast ground contacts, which train the recoil the tendon provides rather than only the muscle force.
The general case for strength training holds here too. Two to three sessions a week improved running economy across the studies in Blagrove's review2, and strength training had the strongest injury prevention effect in Lauersen's meta-analysis3. Neither of those was about older runners specifically, but neither has any reason to stop applying.
What does not need to change
The joints are not the reason to stop.
Recreational running is associated with less hip and knee arthritis than not running at all, in the largest analysis of the question4. Our article on running and arthritis goes through that in full, including the finding that people over 50 who already had knee arthritis did not get worse over four years of running.
What does change
Two things, in our experience, and neither is a total.
Recovery between hard sessions. The session you could do twice a week at thirty may need three days between it at fifty-five. That is a scheduling change rather than a reduction in what you can do.
Tolerance for sudden increases. The jump that would have caused a niggle at thirty causes an injury at fifty-five. The same rules about gradual progression apply, with less margin for error.
Neither of those means running less overall. Both mean arranging the week differently: the same volume, spread more evenly, with the hard days further apart and the easy days genuinely easy.
The honest limits
The DeVita study ran from 18 to 60, so it does not describe what happens after that. It was also cross-sectional, comparing runners of different ages rather than following the same runners over decades, so it shows how older runners differ from younger ones rather than proving how any one runner will change.
And people who are still running at sixty are a selected group. Some of what looks like graceful ageing is that the people who could not keep going are no longer in the sample.