Author: Dr. Georgios Kakavas, on behalf of the author team.

Clinicians regularly encounter patients whose post-operative X-rays are unremarkable yet who have satisfactory postoperative results and improved range of motion. However, weeks or months later, these patients are unable to achieve full knee extension, experience difficulty activating the leg during stair ascent, and demonstrate persistent quadriceps weakness. In these patients, though the surgical procedure is successful, functional recovery is incomplete.

This blog summarises our narrative review (Kakavas et al., 2026), which synthesized evidence from across the AMI literature, including randomized controlled trials, cohort studies, and mechanistic investigations, to identify the key underlying factor of this functional deficit and advocate for prioritizing arthrogenic muscle inhibition (AMI) as a primary rehabilitation target rather than a secondary concern.

What AMI is (and isn’t)

AMI is a reflexive inhibition of muscles around an injured, inflamed, swollen, or operated joint, most commonly affecting the quadriceps. It is not due to lack of effort, pain tolerance, or motivation, but to neurophysiological mechanisms: altered joint signaling changes spinal reflexes and motor cortex activity, limiting muscle mobilization even at maximal effort. This distinction is critical; misattributing AMI to insufficient patient effort may lead to prescribing increased loading, which the underlying inhibition actively impedes.

Inhibition at every level: a multi-level problem

AMI can be conceptualized as a multi-level process, with inhibition occurring across three interconnected physiological levels; peripheral, spinal, and supraspinal. Each level contributes to the impaired voluntary muscle activation seen clinically:

• Peripheral: Even minimal joint effusion significantly impedes quadriceps activation. Consequently, swelling control should be regarded as a neurophysiological intervention in place of solely a comfort measure.
• Spinal: Presynaptic inhibition of Ia afferents, a disrupted gamma-loop, and reduced reflex gain blunt alpha-motoneuron excitability, keeping high-threshold motor units offline.
• Supraspinal: Changes in corticospinal excitability and intracortical inhibition appear to limit voluntary drive.

A considerable caveat in our review is that much of the evidence regarding cortical involvement is extrapolated from the anterior cruciate ligament (ACL) and broader neurorehabilitation literature rather than from arthroplasty-specific studies. Nevertheless, the practical implication remains: many patients present for surgery with pre-existing inhibition, which is additionally compounded by postoperative pain and effusion.

Why “standard rehab” can under-deliver

Our review found the interventions such as range-of-motion exercises, progressive strengthening, cycling, and basic functional drills constitute essential components of rehabilitation. However, the issues of specificity and timing are critical. During the initial postoperative weeks, when inhibition is most pronounced, generic strengthening exercises depend on a neurologically restricted system. Neglecting to address this window can result in an ‘activation debt,’ resulting in persistent compensatory patterns and quadriceps deficits of 20–40% that may persist for months or even years.

A staged, mechanism-led rehabilitation framework

Our review reframes rehabilitation around the inhibition itself, corresponding to the intervention to the main mechanism at each stage:

• Days 0–7 — break the effusion–inhibition loop. Cryotherapy, compression, elevation, early activation and exposure, and neuromuscular electrical stimulation (NMES) — a technique that uses electrical current to elicit involuntary muscle contractions — initiated as early as feasible to bypass voluntary activation failure.
• Weeks 1–3 — restore voluntary drive. Higher-intensity NMES (superimposed on contractions), EMG/pressure biofeedback, terminal-extension quality work, and motor imagery/action observation as a plausible cortical-priming adjunct (largely extrapolated from non-arthroplasty literature).
• Weeks 2–8 — bridge with blood flow restriction training (BFRT). BFRT uses a pressurized cuff to partially occlude venous return, enabling meaningful strength gains at low loads (20–30% of 1RM) when high-load exercises are not possible due to pain or AMI. Appropriate vascular screening and supervision are required.
• Weeks 4–10 — sensorimotor retraining. Balance, perturbation training, and dual-task work to rebuild reactive control and coordination — for example, single-leg stance on unstable surfaces, step-response drills, and walking with a concurrent cognitive task.
• Weeks 8–16+ — high-level motor control. Power, eccentric stair-descent capacity, gait symmetry with feedback, and task-specific demands to dismantle compensatory synergies.

Throughout all rehabilitation phases, it is important to educate patients that AMI is a neurophysiological issue rather than a matter of willpower, to reduce under-recruitment.

Effective treatment requires accurate measurement.

A major barrier identified in our review is the lack of routine measurement of voluntary activation. Most clinical trials report strength and functional outcomes yet do not quantify voluntary activation, causing difficulty in determining whether interventions address inhibition or merely compensate for it. We recommend reporting the central activation ratio (CAR) — a measure of the proportion of quadriceps motor units successfully recruited during a maximal voluntary contraction, typically assessed via twitch interpolation — alongside validated EMG-based metrics and functional outcomes. Phenotyping patients (identifying who has AMI, its severity, and pattern) would enable more targeted rehabilitation.

Take-home messages for clinicians

For physiotherapists and rehabilitation clinicians managing patients after knee osteoarthritis or knee arthroplasty, our review suggests the following key actions:

• Treat effusion as a neurological problem — control it aggressively and early.
• Start NMES early; don’t wait for “enough” voluntary activation.
• Use BFRT as a bridge when high loads aren’t tolerated.
• Go beyond strength: retrain coordination, gait extension, and address cortical/psychological layers.
• Measure activation, not just range and reps.
• Consider AMI before surgery, too — prehabilitation may lower the baseline inhibition you start from post-operatively.

Conclusion

Implants do not facilitate motor unit recruitment; this depends on the patient’s nervous system. By reframing AMI as a primary, measurable target, clinicians can adopt a coherent, mechanism-based approach to improve strength, gait, and patient satisfaction following knee osteoarthritis and knee replacement. While the structural issue may be resolved, the focus must shift to restoring neuromuscular connectivity.

Read the full review: Kakavas G, Sasse C, Królikowska A, Wong S, Becker R, Prill R. Rehabilitation of Arthrogenic Muscle Inhibition in Patients with Knee Osteoarthritis and after Knee Arthroplasty. Curr Rev Musculoskelet Med. 2026;19:44. https://doi.org/10.1007/s12178-026-10038-7

Competing interests: None declared.

Author: Georgios Kakavas (Fysiotek Spine and Sports Lab, Athens; Queen Mary University of London, SEMS).