Is running bad for your knees?

What the largest review found

In 2017 a team pooled 17 studies covering 114,829 people and asked one question: who ends up with hip or knee osteoarthritis? ^[1]

Recreational runners came in at 3.5 percent. Controls, meaning people who did not run, came in at 10.2 percent. Competitive runners, the elite and professional end of the sport, came in at 13.3 percent. ^[1]

So the honest answer is not a flat no. It is a U. Recreational running sits at the bottom of that curve, with lower rates than sitting still. Running at competitive volume for years sits at the top, higher than both.

The authors concluded that running at a recreational level for up to 15 years, and possibly longer, may be safely recommended as general health exercise. ^[1]

What happens to knees over eighteen years

A meta-analysis pools snapshots. It is fair to ask what happens when you follow the same knees forward, and one study did exactly that.

Starting in 1984, researchers tracked 45 long-distance runners and 53 controls, average age 58 and ranging from 50 to 72, with serial knee X-rays through 2002. Nearly two decades later, they found no acceleration of radiographic knee osteoarthritis in the runners. ^[2]

These were healthy older adults, not a random sample, and people who keep running for eighteen years are people whose knees let them. That caveat is real. It is also true that the alarming version of this story, where the running itself grinds joints down, has now failed to show up in both the pooled data and the long follow-up.

What if your knees already hurt

This is the fairer question, because plenty of people asking it already have a diagnosis.

A study drawing on the Osteoarthritis Initiative looked at adults aged 50 and over who already had knee osteoarthritis and chose to run anyway. Self-selected running was not associated with worsening knee pain or with faster structural progression. It was associated with improved knee pain. ^[3]

The authors were careful about why, and so should you be. People choose their own running when their knee hurts, which usually means less of it and slower. The finding is that running need not be discouraged in people with knee osteoarthritis, not that running is a treatment for it. ^[3]

The thing that actually hurts beginners

It is not running. It is the rate at which you add it.

In 2023 researchers followed 110 people through beginner running programs in the UK, average age 47 and 82 percent women. 27.3 percent finished. ^[4]

Of the people who dropped out, 73 percent reported a musculoskeletal injury. When the researchers grouped the reasons people gave for stopping, one of the three themes that emerged was how fast the program progressed. ^[4]

That is the failure mode, and it is a design flaw, not a personal one. A calendar-driven plan tells you to run three minutes in week three whether or not week two left you limping, because a calendar cannot see you. It moves at the speed of the average person in a spreadsheet, and almost nobody is that person.

A heart-rate ceiling can see you. When the work is costing you more than it should, the ceiling holds you back that day, and the progression waits. That is a slower-looking plan that more people finish.

See how the heart-rate ceiling works