Don't take our word for it. Read the evidence.

Cook and Purdam's 2009 paper proposed three overlapping tendon states: reactive tendinopathy, tendon dysrepair, and degenerative tendinopathy.^[6] In the early reactive state, increased proteoglycans can make a tendon swell after an abrupt load change while much of the collagen remains intact. Later stages include progressively greater matrix disorganisation.

The authors' 2016 update is just as important as the original model: it says that classifying a pain condition from structure alone is incomplete. Pain, function, and structure must be considered together, and their relationship is still not fully understood.^[7]

Docking and Cook then used ultrasound tissue characterisation in Achilles and patellar tendons. Pathological tendons contained more disorganised tissue, but because they were thicker, they also contained a greater absolute cross-sectional area of aligned fibrillar structure than normal tendons.^[8] That finding supports a capacity-focused question: what can the whole tendon currently tolerate?

Exercise adaptations are specific to the demand that is trained. In the Campos study, the training zone that best improved maximal strength was not the same zone that best improved repeated-effort performance.^[9] The tendon literature also shows that loading conditions influence adaptation.^[11]

A wrist curl can improve a useful physical quality, but it does not fully reproduce the duration, speed, coordination, posture, attention, and recovery demands of an eight-hour workday or a long gaming session. Conversely, simply repeating an irritable task may not provide a well-controlled enough stimulus to build general capacity.

That is why our framework uses both. General exercise creates room; structured task practice teaches the person to use that room in the context that matters.

Rio and colleagues compared one bout of isometric and isotonic quadriceps exercise in six athletes with patellar tendinopathy. Isometrics reduced mean pain during a test from 7.0 to 0.17 out of 10, the effect persisted at 45 minutes, and the change occurred alongside reduced cortical inhibition and increased force.^[14]

Their 2016 narrative review proposed “tendon neuroplastic training”: heavy strength work paired with external pacing to target both the tendon and motor system.^[15] Preliminary participants moved toward more typical inhibition values, but the paper did not establish that the metronome caused the change or that self-paced repetitions cannot work.

A 2025 randomized trial is the first controlled test of this idea in lateral epicondylitis. Thirty-four adults trained three times a week for four weeks. One group did externally paced wrist flexion and extension. The other group received extensor stretching and myofascial release. Paced training reduced pain and tennis-elbow disability more than that comparison, with effect sizes of 0.37 and 0.449. Grip strength improved in both groups, without a clear difference between them.^[16]

The defensible takeaway is that chronic tendon pain is not only a local-tissue story, and that paced exercise can make loading consistent. This elbow trial does not show that external pacing outperforms the same strengthening done without a metronome.

Central sensitization describes increased responsiveness in central pain-processing neurons, including amplification and reduced inhibitory control. These are measurable neuroplastic processes, not evidence that pain is imagined.^[18]

Louw and colleagues reviewed eight studies with 401 participants and concluded that pain neuroscience education may improve pain, function, catastrophizing, and movement in chronic musculoskeletal pain.^[19] The studies were too heterogeneous for a meta-analysis, and most involved chronic low back pain.

A 2025 systematic review and meta-analysis included 19 randomized trials and 693 adults receiving pain neuroscience education with a physical therapy intervention. Across the intervention groups, mean pain scores decreased from 5.89 to 3.03, which is nearly half and can be stated conservatively as roughly a 45% reduction. Disability outcomes also improved across the included measures. The controlled random-effects analysis estimated a 0.784-point advantage on a 0–10 pain scale in favor of combined care, with moderate certainty but high heterogeneity.^[20]

Moseley and Butler's later review likewise frames education as part of a broader clinical approach.^[21] Information can change the meaning of pain, but meaningful recovery still requires active practice and restored function.

A broad review of training and immobilization described reductions in muscle size, capillary density, and tendon or ligament properties with unloading, while noting that much of the tendon evidence available at the time came from animal work.^[24] In one experiment from that literature, the rat gastrocnemius muscle-tendon unit was cast for up to three weeks. Force to failure fell by 21% when the muscle was held lengthened and 31% when it was held shortened after one week, and by about 26% and 48% after three weeks. Energy absorbed before failure fell 49% in the lengthened position and 64% in the shortened position. The unit lost 12% of its weight when lengthened and 36% when shortened. Elastic stiffness fell in both positions, but that report did not give a separate percentage for the stiffness loss.^[25]

In eight healthy men, 20 days of bed rest reduced knee-extensor tendon stiffness from 70.3 to 50.1 N/mm, about 29% lower, and increased hysteresis. The plantar-flexor stiffness change did not meet statistical significance.^[26] This was a lower-limb model, not a wrist study.

More directly relevant to the upper limb, Motobe and colleagues placed 14 healthy men in forearm casts for 21 days. Muscle oxidative recovery worsened in the immobilization group, and endurance training twice weekly prevented some of the decline; grip strength still fell in both groups.^[27]