Why knee OA starts before you feel it

Knee osteoarthritis can begin dismantling cartilage years before it announces itself with pain. Cartilage contains no blood vessels and no nerve endings, which means early cell-level breakdown produces no signal the joint can send to the brain. The deterioration is, at first, entirely silent.

This is not a theoretical concern. A seven-year longitudinal MRI study of 3,710 knees — the MOST cohort — found measurable cartilage loss in the central femoral condyle at Kellgren–Lawrence grade 0, meaning knees that showed no abnormality whatsoever on standard X-ray. The structural damage had already begun before any imaging test used in routine clinical practice could detect it. Tibial osteophytes, the first visible radiographic event, appear only after that earlier, invisible phase has passed.

By the time an X-ray flags early OA, meaningful cartilage change has typically already occurred. Compounding this, the relationship between imaging findings and symptoms is unreliable in both directions: some patients with notable radiographic changes report only mild discomfort, while others with limited X-ray findings experience considerable functional impairment. A scan, in other words, is not a dependable measure of either disease stage or patient burden.

The practical implication is straightforward: waiting for pain, or waiting for imaging to confirm a problem, is already a delayed response. Recognising the earliest clinical signals — before standard imaging catches up — is where meaningful intervention is still possible.

What 'early OA' actually means

The term 'early OA' does not mean mild symptoms or a low X-ray grade. It describes the phase — often entirely pain-free — when cartilage matrix begins to degrade and subchondral bone starts to remodel in response. The 2022 Early OA textbook and the First International Early Knee OA Workshop (Tokyo, 2014) are both explicit on this: a patient may be asymptomatic while these tissue-level changes are already under way, with painless crepitus or joint line tenderness sometimes the only detectable clinical signal.

The formal classification system does not capture this window. The ACR's 1986 criteria — still widely referenced in clinical practice — require radiologically detectable signs such as osteophytes and joint space narrowing before a diagnosis of knee OA can be made, effectively excluding the earliest disease phase from any formal label.

Even within the radiographic timeline, sequence matters. Tibial osteophytes are the first event visible on X-ray; by the time they appear, cartilage loss in the central femoral condyle has already been progressing for some time. Once osteophytes have formed, medial meniscal extrusion becomes the strongest MRI predictor of further structural deterioration — advancing from roughly 1.4 mm at KL grade 0 to 6.0 mm at KL grade 3 as the disease takes hold.

Chondromalacia patella fits within this same pre-diagnostic territory: a softening of the patellar cartilage that, left unmanaged, is a recognised precursor to secondary knee OA. Each of these presentations — chondromalacia included — illustrates the same diagnostic gap: structural change is present and progressing before the classification system is built to formally name it.

Six early signs worth taking seriously

Taken individually, none of the following signs proves cartilage damage is occurring. They are non-specific, and several have entirely benign explanations in a fit, younger person. What matters is the pattern across signs, and whether that pattern persists.

• Post-rest stiffness lasting under 30 minutes. A sense of tightness or reluctance in the knee after sitting, overnight, or on waking — which loosens as you move — is a recognised early feature of knee OA. The duration is a useful marker: the prolonged morning stiffness of rheumatoid arthritis typically exceeds 45–60 minutes, whereas OA stiffness tends to resolve quickly once the joint is loaded.
• Crepitus — audible or felt during movement. A grating, clicking, or grinding sensation on bending or straightening the knee reflects roughening of the cartilage surfaces. Crepitus alone in an otherwise well young person is rarely significant. It becomes more meaningful when it is consistent, associated with other signs, or new.
• Activity-related aching that fully eases with rest. Discomfort during stairs, squatting, or sustained walking that settles completely once the joint is offloaded — rather than lingering background pain — fits the early OA pattern. Constant or night pain suggests a different or more advanced process.
• Mild swelling or tenderness around the kneecap. Low-grade puffiness after heavier use, or tenderness on direct pressure along the joint line or around the patella, may reflect early synovial irritation rather than major structural change.
• Reduced range of motion compared with the other knee. A subtle loss of full flexion — easier to notice when kneeling or squatting — often develops gradually and without obvious pain.
• Quadriceps weakness or a sense of instability. Thigh-muscle weakness is both an early consequence and a driver of knee OA. A feeling that the knee may give way on stairs or uneven ground, without a locking or catching sensation, warrants attention.

If two or more of these features are present together, particularly if they have persisted beyond a few weeks or are affecting daily tasks, that combination is worth a structured clinical assessment rather than continued self-management.

When self-management stops being enough

Most episodes of knee discomfort settle within a few weeks with simple measures — rest, activity adjustment, over-the-counter analgesia, and gentle exercise. The question is when that reasonable expectation no longer holds.

Two signals consistently point toward professional assessment rather than continued self-management. The first is duration: symptoms persisting beyond six weeks without meaningful improvement warrant clinical evaluation, even if pain remains mild. NHS Inform and converging clinical guidance agree on this threshold — not because six weeks marks a biological turning point, but because it is a practical signal that the joint is not self-resolving. The second signal is functional impact: any symptom that begins to limit a routine task — climbing stairs, walking a familiar distance, sitting through a working day — warrants assessment regardless of how long it has been present.

Pain intensity alone is an unreliable guide. Some patients with significant cartilage change have modest symptoms; the mismatch between structural findings and subjective pain in knee OA is well established. Anchoring the decision to act on functional limitation rather than pain severity is the more dependable approach.

Self-management — quadriceps strengthening, weight management, activity pacing — is the appropriate first response and forms the foundation of any subsequent plan. It does not, however, arrest underlying tissue change. Evidence indicates that the pre-radiographic phase is precisely when intervention carries the greatest potential to influence disease course, which makes erring toward earlier assessment rational rather than overcautious.

A GP referral is a sensible starting point. Specialist input becomes appropriate when imaging findings or persistent functional limitation suggest more than simple soft-tissue aching — the point at which structural questions require a more detailed clinical answer.

How early cartilage change is assessed

Standard X-rays measure bone — joint space, osteophyte formation, and density changes that signal established disease. Cartilage itself is not directly visible on a plain radiograph, which means the pre-radiographic phase of tissue change sits outside what this tool can detect. That is not a reason to dismiss X-ray; it remains appropriate and cost-effective for ruling out fracture, assessing alignment, and staging disease once progression is visible. It does, however, explain why specialist assessment of suspected early OA often requires additional imaging.

Advanced MRI sequences

T2 mapping and T1ρ (T1 rho) MRI sequences assess the biochemical state and water content of cartilage, not just its thickness. Early degeneration disrupts the collagen network and depletes proteoglycans before any visible thinning occurs; these sequences detect that process as a measurable shift in relaxation time. T1ρ values above 33 ms and T2 values above 35.04 ms are each independently associated with early-stage knee OA. A nomogram combining these measures with infrapatellar fat pad volume achieved a C-index of 0.975 in a 199-patient validation cohort — a level of discrimination well beyond what standard radiographic grading can provide.

Radiomics extends the detection window further still. A model integrating MRI texture features from cartilage and subchondral bone predicted OA incidence in knees classified as KL grade 0–1 — with no X-ray abnormality at baseline — at an area under the curve of 0.985, identifying structural risk years before radiographic change appeared.

T2 mapping and T1ρ sequences are currently available through specialist private musculoskeletal imaging centres; they are not part of the standard NHS referral pathway. Whether advanced MRI is warranted depends on clinical history, symptom duration, and whether the result is likely to change management — a judgement that requires consultant input rather than self-referral.

Imaging as one input

Even with high-resolution sequences, findings are interpreted alongside clinical assessment — range of motion, quadriceps strength, and symptom pattern — not as a standalone verdict. A signal change on MRI in an otherwise fully functional, symptom-free knee does not automatically indicate treatment. The imaging adds tissue-level resolution to the clinical picture; it does not replace it.

Acting in the early window

Quadriceps strengthening reduces compressive load across the medial compartment with every step — making it the best-evidenced first-line measure, and one that targets a genuinely modifiable mechanical stressor from the earliest stage of cartilage change. Weight management and activity pacing work through the same logic: reducing cumulative joint stress during the period when tissue remains capable of adapting, rather than simply deteriorating. Both are well supported; neither reverses established cartilage loss, and framing them honestly as progression-slowing rather than curative is what makes sustained commitment to them realistic.

The therapeutic argument for early action runs deeper than symptom management. A narrow window exists during which intervention may genuinely alter disease course rather than merely manage its consequences. That window corresponds to the pre-radiographic phase: GAG depletion, collagen triple-helix disruption, and subchondral bone texture changes are each detectable before joint space narrowing appears on X-ray, and evidence suggests this is when mechanical and structural signals remain most responsive to change. Once damage consolidates across cartilage and bone, the range of viable options contracts.

Disease-modifying strategies specifically targeting this window — approaches addressing early molecular cartilage changes and subchondral bone remodelling — remain an active research frontier rather than standard care. They are not yet ready for routine prescription, but they are the basis for specialist-level clinical evaluation rather than simple watchful waiting. For patients in London at this stage, that evaluation is available through the London Cartilage Clinic at londoncartilage.com.

1. [1] Longitudinal correlation between X-ray and MRI findings in medial compartment knee OA (MOST dataset). (2025). https://doi.org/10.1002/ksa.70016 https://doi.org/10.1002/ksa.70016
2. [2] MRI-based radiomics framework for early identification and progression stratification in knee OA (OAI dataset). (2025). https://doi.org/10.1186/s12891-025-09234-2 https://doi.org/10.1186/s12891-025-09234-2
3. [3] MRI-based patient-specific nomogram for diagnostic risk stratification of early knee OA. (2025). https://doi.org/10.1093/rheumatology/keaf319 https://doi.org/10.1093/rheumatology/keaf319
4. [4] Cationic Contrast Agents for CT Imaging of Cartilage for Early OA Diagnosis. (2023). https://doi.org/10.1021/acsnano.2c12376 https://doi.org/10.1021/acsnano.2c12376