Your Knee Hurts, But Your Spinal Cord Is Making It Worse

Take a painkiller for your arthritic knee, wait an hour, and — if you're lucky — the edge comes off a little. That's about the best most of the 500 million people living with osteoarthritis can hope for right now. Not relief. Just slightly less bad. A 2013 review out of the University of Maryland, published in Osteoarthritis and Cartilage, lays out in uncomfortable detail why that is: because by the time your knee is screaming, your spinal cord is already involved. And your brainstem. And, in ways researchers are still puzzling over, your brain's own reward system. Osteoarthritis pain is not where you think it is.

The Arthritis Pain Problem Nobody Solved

Most people's mental model of osteoarthritis stops at the cartilage — it wears away, bone grinds on bone, it hurts. And yes, that's part of it. But here's the thing cartilage doesn't have: nerve endings. It genuinely cannot send pain signals. So where is the pain actually coming from? The synovial lining, the bone beneath, the nerve fibers threading through the joint tissue. And once those fibers start firing persistently — which they do, quickly — the story gets a lot more complicated than a worn joint. One detail buried in the Maryland review struck me as particularly telling: patients who had hip replacements and finally felt no pain showed measurable reversal of brain changes that had built up over years of OA. The brain had been silently restructuring itself the whole time. That's not a plumbing problem in your knee. That's a neurological condition that happens to start there.

How Researchers Actually Study Joint Pain

You can't ask a rat how much its knee hurts. So the field got creative. The go-to method is something called the MIA model — researchers inject monoiodoacetate, a chemical that shuts down cartilage cell metabolism, into a rodent's knee. Within days, the tissue starts dying. By day 28, the bone underneath is visibly damaged. By day 56, there's large-scale remodeling going on. None of that sounds pleasant, and the behavior of the animals confirms it: they shift their weight off the affected leg, flinch at light pressure on the paw, slow down on the running wheel. What's newer — and honestly more interesting — is a test called conditioned place preference. You train the animal to associate a specific chamber with an injection of local anesthetic into the sore knee, then see if it actively seeks out that chamber. If it does, the animal isn't just reacting to pain. It wants it to stop. That's aversion. That's the brain. And in OA, it showed up clearly.

Why Does Osteoarthritis Pain Spread Beyond the Joint?

Here's what nobody expected. Inject MIA into a rat's knee — just the knee — and within days the animal is also hypersensitive to touch on the bottom of its paw. A spot nowhere near the joint. That's the first hint that something systemic is happening. Inflammatory proteins, specifically TNF-α and IL-6, flood the synovial lining and start sensitizing the nerve fibers embedded in the joint tissue. Peaks at day 4. Those fibers feed into neurons clustered in the dorsal root ganglia, which ship the signal up to the spinal cord — now running hotter than it should. Spinal immune cells, the microglia and astrocytes, kick in and start releasing their own chemical signals. By this point the pain has developed a kind of momentum. It's partly self-sustaining, no longer strictly tethered to what's happening in the knee. Then there's the serotonin problem, which is genuinely counterintuitive: a descending pathway from the brainstem that ordinarily suppresses pain appears to invert its function during OA. Instead of damping signals, it amplifies them. So the brain is now actively making the pain worse. No ibuprofen in the world touches that mechanism.

What This Means for Treatment

There's a silver lining in all of this, though it comes with caveats. Mapping where the pain machinery lives also maps where drugs could interfere. Cannabinoid receptors — CB1 in the periphery, both CB1 and CB2 in the spinal cord — seem to naturally throttle OA pain when activated. The body actually tries to use them this way; block them with an antagonist in an OA animal and pain gets worse, not better. That's a genuinely useful finding for the growing number of clinicians curious about cannabinoid-based analgesia. MMP-13 inhibitors are another angle: they block an enzyme that destroys cartilage, and in rat models they reduced both the structural damage and the pain simultaneously. One drug, two problems. That's rare. Probably the most practically interesting finding, though, concerns pregabalin and gabapentin — drugs that have been prescribed for nerve pain for years. Turns out they reduce OA pain signals in the spinal cord partly by counteracting that aberrant serotonin pathway. Which means some OA patients already on these drugs for unrelated reasons may be getting joint pain relief without their doctors realizing it.

The Questions Still Unanswered

Credit to Zhang, Ren, and Dubner for not overselling this. The rodent-to-human translation is always dicey, and the MIA model — for all its usefulness — produces a faster, nastier disease than most people actually experience. Brain imaging of OA animals was barely off the ground at the time of publication; one surgical rat study found heightened connectivity between the nucleus accumbens and the thalamus three to five weeks post-injury, but what that actually means for the lived experience of pain is still unclear. TRPV1, a heat-sensitive ion channel that kept appearing across multiple experiments, gave contradictory results depending on whether you were measuring electrical activity in neurons or behavior in the whole animal — and nobody's fully reconciled that yet. The honest takeaway is that the field understands the shape of the problem far better than it did a decade ago. What it doesn't have yet is a drug that targets the spinal and brain components of OA pain without trading one problem for another. That target, at least, is now clearly identified.

• Pain isn't local. — Osteoarthritis triggers a cascade that involves the spinal cord, brainstem, and brain's reward system, which is why treating only the knee rarely eliminates pain completely.
• The body fights back — briefly. — Endogenous opioids and cannabinoids naturally suppress early OA pain, but this defense appears to erode as the disease progresses toward a neuropathic state.
• Serotonin can be a villain. — A brainstem pathway that normally inhibits pain appears to switch roles in OA, actively amplifying pain signals — a counterintuitive finding with real drug-development implications.

"Peripheral and central sensitization contributes to OA pain, involving inflammatory cytokines, neuropeptides, and a variety of chemical mediators. Interestingly, brainstem descending facilitation of 5-HT/5-HT3 receptors plays a role in OA pain." — Zhang RX, Ren K, Dubner R, Osteoarthritis and Cartilage, 2013.