Running got a bad rep
Cartilage Needs a Job: Why Full-Range Loading (Not Rest) Protects Your Knees
If you've ever told someone you run for exercise, you've probably heard it: "Running is so bad for your knees." It's one of the most durable pieces of movement folklore out there — repeated by well-meaning relatives, personal trainers, and even some healthcare providers. It sounds intuitive. Cartilage is a finite tissue, running is repetitive impact, so surely all those footstrikes are grinding it away like tire tread.
The problem is that this idea gets the biology backwards. Articular cartilage isn't a passive wear surface that only depletes with use — it's a living, mechanically responsive tissue that depends on movement and loading, through a full range of motion, to stay healthy. Understretched, underloaded cartilage doesn't stay pristine. It thins.
Cartilage needs load
Articular cartilage is made almost entirely of chondrocytes sitting inside a dense matrix of collagen and proteoglycans. Those chondrocytes are mechanosensitive — they have receptors that detect deformation, fluid shifts, and shear as the joint moves and bears load, and translate that into a cellular signal. Under normal, physiologic loading, this signaling drives matrix maintenance and synthesis — an anabolic, tissue-preserving response. Under a repeated compression-and-release cycle, chondrocytes physically deform and rebound, which is part of how they sense mechanical input and regulate their metabolic activity in the first place.
In other words: cartilage doesn't just tolerate load. It requires it to function normally. This is the same logic we already accept for bone (Wolff's Law) and tendon (mechanotransduction-driven remodeling) — connective tissue adapts to the demands placed on it. Cartilage plays by the same rules; it's just easy to forget because we can't feel it working the way we feel a muscle burn.
when cartilage doesn't get load
This is the part that tends to surprise people: the clearest evidence that cartilage needs mechanical input comes from what happens when you take load away. Immobilization and unloading studies — in casted or externally fixated joints, non-weight-bearing protocols, and even short-term partial weight-bearing after fracture — consistently show cartilage thinning, softening, and reduced proteoglycan content within weeks, not years. One frequently cited clinical example found measurable cartilage thinning across all compartments of the knee after just seven weeks of partial weight-bearing. Populations with chronically reduced joint loading, such as people with spinal cord injury, lose cartilage at a rate that outpaces typical age-related degeneration. And because cartilage is avascular and has very limited capacity for spontaneous repair, that disuse-driven thinning doesn't bounce back quickly once normal loading resumes.
This flips the "protect the cartilage by resting it" instinct on its head. Rest isn't neutral for cartilage — chronic underloading is itself a degenerative stimulus.
So what does the running research actually say?
Given all that, the running-and-knees question becomes an empirical one rather than a hunch, and it's been studied extensively. The pattern that keeps showing up is a U-shaped risk curve: sedentary behavior and elite/high-volume running both carry more risk of knee and hip osteoarthritis than moderate recreational running.
A widely cited systematic review and meta-analysis pooling data on over 100,000 people found that only about 3.5% of recreational runners developed hip or knee arthritis, compared with roughly 10% of sedentary, non-running individuals and 13% of competitive/elite runners. Being sedentary was a bigger risk factor than being a recreational runner. Follow-up reviews looking specifically at running volume and years of exposure have generally reinforced this: recreational-level running is not associated with increased knee OA risk, and some evidence even points toward biological markers of better joint lubrication and cartilage health in trained runners compared with non-runners. A 2023 systematic review of MRI studies — the imaging method sensitive enough to catch real-time cartilage changes — concluded that a history of recreational running does not accelerate cartilage degeneration or increase incident OA risk, despite widespread public and even clinical belief to the contrary (survey data shows over 40% of the public, and nearly one in ten healthcare providers, still assume running damages joints).
Where the picture gets murkier is at the extremes — high-volume competitive running sustained over 15+ years, where the evidence is more mixed and harder to disentangle from confounders like prior injury. But that's a very different exposure than the recreational runner logging a few miles several times a week, and it's not what the old adage is usually warning people about.
range matters as much as volume
This is the piece that gets left out of the running-and-knees debate entirely, and it's arguably more clinically relevant than the volume question. Articular cartilage isn't loaded uniformly across the joint surface — different regions of the tibiofemoral and patellofemoral cartilage make contact at different points in the flexion-extension arc. A joint that only ever moves through a narrow, partial range — because of pain avoidance, a stiff post-surgical knee, deconditioning, or someone who simply never squats, lunges, or bends past 90 degrees — is only ever mechanically stimulating a fraction of its cartilage surface. The rest goes chronically underloaded, and by the disuse mechanisms above, that's exactly the environment associated with thinning.
This is also part of why immobilization research is so instructive clinically: it isn't just the absence of load overall, it's the absence of movement through range that restricts which cartilage surfaces ever get stimulated, alongside the well-documented muscular and capsular changes that come with immobilization. For rehab and training purposes, this argues for programming that actually takes joints through meaningful flexion and extension ranges — deep squats, full-range lunges, step-downs, running itself, closed-chain work through end ranges — rather than treating knee flexion past 90 degrees as inherently risky.
The takeaway
None of this means every load is automatically good, or that pain and reactive symptoms should be pushed through blindly — dosage, tissue irritability, and progressive loading principles still apply, especially post-surgically or in a joint with active symptoms. But the underlying premise that cartilage is a depleting resource best preserved by minimizing movement doesn't hold up. Cartilage health tracks with appropriately dosed mechanical stimulation across a full range of motion far more than it tracks with total steps taken or miles run.
For most people asking whether they should keep running for their knees, the evidence-backed answer is reassuring: recreational running is, if anything, associated with better joint outcomes than sitting it out. And for the range-of-motion question that too often gets ignored in that conversation, the message is just as clear — the cartilage that never gets loaded is the cartilage most at risk.
Selected references
Alentorn-Geli E, et al. The Association of Recreational and Competitive Running With Hip and Knee Osteoarthritis: A Systematic Review and Meta-analysis. J Orthop Sports Phys Ther. 2017;47(6):373-390.
Timmins KA, et al. Running and Knee Osteoarthritis: A Systematic Review and Meta-analysis. Am J Sports Med. 2017;45(6):1447-1457.
Hartwell MJ, et al. Does Running Increase the Risk of Hip and Knee Arthritis? A Survey of 3804 Marathon Runners. Sports Health. 2024;16(4):622-629.
Systematic review/meta-analysis: Is running good or bad for your knees? Cartilage morphology and composition changes. Osteoarthritis and Cartilage. 2023;31(2).
Mechanotransduction pathways in cartilage chondrocyte homeostasis (Piezo1/2, TRPV4, integrins) — PMC review.
Hinterwimmer S, et al. Cartilage atrophy in the knees of patients after seven weeks of partial load bearing. Arthritis Rheum. 2004;50:2516-2520.
Review on immobilization-induced cartilage thinning, tissue softening, and reduced proteoglycan content (Ann NY Acad Sci).
