A clinical observation worth opening with
The runner who walks in with a thickened, painful Achilles at six weeks of symptoms has been told three things by three clinicians: rest it, stretch it, and ice it. None of those are correct. The 2024 JOSPT Clinical Practice Guideline (Chimenti et al., 2024) and the 2025 BJSM Delphi consensus on exercise parameters (Demangeot et al., 2025) make this unambiguous. The tendon needs progressive mechanical load, calibrated to symptoms, sustained for months. And yet the same runner will return in three months no better, because somewhere in the loading prescription, the soleus has been quietly ignored.
That is the through-line of this piece. The current evidence on Achilles tendinopathy is converging — but the residual gap between guideline and practice is largest at the soleus. And the second gap — narrower, but as costly — is what happens to the tendon between the loading sessions, in the ambulatory hours of an ordinary day. We will come back to both.
Midportion and insertional are not the same condition
Treat them as two pathologies that happen to share a tendon. The 2024 JOSPT CPG is restricted to midportion tendinopathy for a reason: the management diverges sharply at the insertion.
Midportion disease is a tensile-overload problem in a relatively avascular zone two to six centimetres proximal to the calcaneus. Insertional disease is a compressive-and-tensile problem at the bone-tendon junction, frequently accompanied by retrocalcaneal bursitis, Haglund’s morphology, and intratendinous calcification. The mechanism that drives the pathology — tensile overload versus compression against the calcaneal tuberosity in end-range dorsiflexion — dictates the exercise.
In practice this means the Alfredson protocol (heel drops off a step into end-range dorsiflexion, three sets of fifteen, twice daily) is appropriate for midportion disease and frequently counterproductive at the insertion. For insertional pathology, work the calf raise on the floor or to neutral, not into the compressive zone. The 2024 CPG explicitly addresses midportion disease; clinicians who apply its loading prescription indiscriminately to insertional cases are not following the guideline, they are misreading it.
What the 2024 JOSPT CPG actually says
Chimenti and colleagues (JOSPT 2024) revise the 2018 guideline with literature through March 2024. Three points matter for practice.
First, the diagnostic recommendations are conservative. Localised midportion pain on palpation, pain reproduction with loading (the hop, the heel raise, the single-leg decline), and morning stiffness remain the diagnostic backbone. Imaging is not required for diagnosis. This matters because clinicians still order ultrasound to “rule in” tendinopathy in patients who already meet the clinical criteria — adding cost, anxiety, and the well-documented risk of detecting clinically irrelevant changes.
Second, the loading recommendation is strong: progressive resistance exercise targeting the plantarflexors, sustained for at least twelve weeks. The CPG does not prescribe Alfredson specifically. It treats heavy slow resistance, eccentric, and combined eccentric-concentric protocols as broadly equivalent at the group level — consistent with Beyer et al. (Am J Sports Med 2015) showing comparable outcomes between Alfredson eccentrics and Kongsgaard-style heavy slow resistance at twelve months.
Third, the CPG acknowledges that pain during loading is acceptable and probably necessary. The Silbernagel pain-monitoring model — pain up to about 5/10 during loading, settling by the next morning — is endorsed. Pain-free rehabilitation is a misreading of the evidence.
The 2025 BJSM Delphi: what experts actually do
Demangeot, O’Neill, Degache and colleagues (Br J Sports Med 2025;59:1337–1349) ran a three-round modified Delphi with international Achilles experts on heel-raise exercise parameters. The consensus is more granular than the JOSPT CPG and worth knowing.
For midportion disease, four parameters reached agreement: contraction intensity (ranked most important), time under tension, sets and repetitions, and contraction type. The intensity target the panel converged on is 70–90% of one-rep maximum, which delivers the 4.5–6.5% tendon strain window associated with collagen synthesis upregulation. This is heavier than most patients are loaded in clinical practice, and considerably heavier than the original Alfredson prescription, which was never an intensity prescription — Alfredson asked patients to add load when the exercise stopped being painful, which is not the same thing.
For insertional disease, three parameters reached agreement, with range of ankle dorsiflexion ranked most important. The recommendation is to restrict dorsiflexion to neutral or below, at least early in rehabilitation, to avoid compressive loading at the enthesis.
Concordance between the CPG and the Delphi is high on the principle of progressive load. Where they diverge — and the Delphi is more useful here — is on the prescription detail. The Delphi gives you numbers.
The soleus problem
Now to the part of the literature that gets read but not internalised. The Achilles tendon is the common tendon of two architecturally distinct muscles. Lieber and Ward’s work on human muscle architecture (Ward et al., Clin Orthop Relat Res 2009; Lieber & Ward, Philos Trans R Soc B 2011) puts the soleus at roughly three times the physiological cross-sectional area of the medial gastrocnemius, with shorter fascicles, higher pennation, and a fibre composition that is predominantly slow-twitch. The soleus is built for force, not for excursion.
The functional consequence is well established in the biomechanics literature. Neptune, Sasaki and colleagues’ musculoskeletal simulations, and Arnold et al.’s fibre-length work in walking and running (J Exp Biol 2013), place the soleus as the dominant generator of Achilles tendon force during walking and slow-to-moderate running. The gastrocnemius contributes more at the push-off of faster running and in jumping, where its biarticular geometry helps it operate near optimal length. During the activities most tendinopathic patients actually do — walking, jogging, returning to running — it is the soleus that dominates the load delivered to the tendon.
This is the issue. Standard calf-raise protocols are performed with the knee extended, which places the gastrocnemius near optimal sarcomere length. With the knee flexed to around 70–90°, the biarticular gastrocnemius shortens past its active force-length plateau and contributes substantially less tension — the relationship has been characterised across the triceps surae force-length literature (Maganaris on triceps surae in vivo length-tension; Hamill & Knutzen biomechanics texts). What remains as the prime mover is the soleus.
Bent-knee calf raises are not an accessory. They are how you load the muscle that actually generates the force you are trying to rehabilitate. The clinical literature has begun to reflect this: O’Neill and colleagues’ work on soleus strength deficits in midportion Achilles tendinopathy demonstrated meaningful soleus weakness that is invisible to a straight-knee heel raise. The 2024 CPG mentions soleus loading; the 2025 Delphi includes knee angle as a parameter of interest. Neither has fully reckoned with what the architecture demands.
A defensible Tuesday morning prescription: at least half of the plantarflexor loading volume should be performed with the knee flexed to around 70°, at a heavy slow tempo (3-second concentric, 3-second eccentric), 70–80% 1RM, three to four sessions per week. The remaining volume targets the gastrocnemius with straight-knee work. This is not a fringe position. It is what the muscle architecture and the activity-specific load distribution call for.
Imaging: Docking and the donut
Sean Docking’s ultrasound tissue characterisation work (Docking & Cook, Scand J Med Sci Sports 2016) reframed how the tendon should be read. Pathological tendons retain a near-normal volume of aligned fibrillar structure surrounding the disorganised area — what Docking called “the donut around the hole.” Total aligned fibrillar structure, in his data, was preserved or even increased in symptomatic tendons compared to controls.
The clinical translation is this: imaging findings do not track symptoms. Asymptomatic tendons show pathology. Symptomatic tendons show pathology that often does not change over a course of successful treatment. Reassuring a patient with a thickened, hypoechoic Achilles that the surrounding tendon is structurally intact is both accurate and clinically useful. Re-scanning at three months to “see if it has healed” is not.
Pain: Rio and the central component
Rio and colleagues (Br J Sports Med 2015) demonstrated that a single bout of heavy isometric loading (five sets of forty-five seconds at roughly 70% MVIC) produced immediate and sustained reductions in patellar tendon pain, accompanied by reduced cortical inhibition measured by transcranial magnetic stimulation. The mechanism is not purely peripheral — there is a measurable central component to tendon pain modulation.
The Achilles literature on isometrics is less clean than the patellar literature; replication in the Achilles has been inconsistent. The practical translation: isometric calf holds — heavy, sustained, in mid-range — are a reasonable tool early in rehab for symptomatic relief and to maintain loading capacity, but they are not a substitute for progressive heavy resistance over months.
Rio’s broader contribution is to take tendon pain out of the purely structural frame. The brain participates. Cortical inhibition matters. This explains why some patients with grim imaging respond beautifully to load, and others with cleaner scans linger.
Prognosis and time
Silbernagel and Crossley’s return-to-sport framework (JOSPT 2015) remains the most honest summary. Expect twelve weeks before meaningful change. Expect three to six months to meaningful return to sport. Expect ongoing vigilance for at least twelve months — recurrence is highest in the first year post-return, particularly during the early reload phase. Tell patients this at the first visit. The injuries that go badly are not usually the ones with the worst initial findings; they are the ones where the patient was sold a six-week recovery and lost faith in the process at week eight.
Prognostic indicators worth tracking: symptom duration at presentation (longer is worse), bilateral symptoms (worse), BMI and metabolic comorbidities (worse), and capacity on the single-leg heel-raise endurance test (lower is worse, and a useful objective measure to retest monthly).
It is worth keeping the public cases in mind when sketching prognosis for a patient. Nic Naitanui’s Achilles had been troubling him from February 2023 onward; by June he was in surgery; in 2024 he retired without playing another senior game. Will Skelton stepped, ruptured, and finished his European season in March 2026. These are the silent-pathology-to-acute-failure trajectories Cook and Purdam’s continuum predicts — tendons that had been giving warning signs the system did not adequately address. Most of our patients will not rupture. But the lesson — that an undertreated chronic tendinopathy is not a benign holding pattern — is one to convey honestly at the first visit.
Clinical translation: three patterns to take into practice on Tuesday
One. Load the soleus deliberately. Half the plantarflexor volume with the knee bent around 70°, heavy slow resistance, 70–80% 1RM, sustained for at least twelve weeks. This is not a fad. The architecture says so, the load distribution during walking and slow running says so, and the emerging RCT evidence says so. If the patient is doing only straight-knee Alfredsons, you are loading the wrong muscle for the activity they want to return to.
Two. Separate insertional from midportion at intake and load differently. Midportion gets full-range or step-edge work. Insertional gets neutral-to-floor work, at least for the first six to eight weeks. Adding a temporary heel lift at the insertion is not capitulation; it is biomechanically appropriate offload.
Three. Use temporary offload tools during the reactive flare and the early return-to-running window. Heel lifts, training-load modification, footwear changes, and external compression are all legitimate. This is the bucket the Orthopaedic Sleeve sits in, and it warrants its own treatment because the dose problem it addresses is the variable most rehab plans miss.
The ambulatory dose problem, and the Sleeve as a clinical adjunct
Graded mechanical loading is the cornerstone of the rehabilitation. The CPG and the Delphi agree on that, and the loading session you prescribe is the lever you have most control over. But the loading session is the smaller share of the patient’s day. The larger share — the unmeasured, uncalibrated, ambulatory load between sessions — is the variable that most rehab plans do not address. The tendon does not differentiate between the heavy slow resistance you prescribed at 70% 1RM and the cumulative micro-loading of three thousand steps to and from the office. From the tissue’s perspective it is all signal. The rehab dose you wrote down is one part of the total dose. The rest is whatever the patient’s day looks like.
This is the gap clinicians have historically lacked a measurable tool for. Heel lifts modify dose at the insertion. Training-load modification modifies the higher-intensity dose. Neither addresses the ordinary walking that fills the day. The Orthopaedic Sleeve was developed to address that specific gap — a daily-wear external device that reduces the share of Achilles tendon force the tissue is exposed to during ambulation, characterised in independent biomechanical testing by the University of Queensland (UQ Final Report, June 2025).
The relevant numbers from that testing are worth getting right, because the temptation to overstate them is real and the temptation to dismiss them is real, and both should be resisted.
The group-level primary finding: medial gastrocnemius EMG during eyes-open standing balance was reduced by 32% with the Sleeve on, statistically significant (p=0.002). That is a standing-balance result on a small sample (n=11), and it should be framed as such — a robust group-level reduction in baseline plantarflexor activation under quiet stance conditions.
The walking results are mixed in their statistical power but mechanistically coherent. Heel contact time during walking was reduced by 5.1% (p=0.009) — a small change, but consistent with a more efficient stance phase. On a Hill-type musculoskeletal model of Achilles tendon force during walking, peak tendon force reduced by a group-average 1.7%, with individual reductions reaching 8.1% in the participants who responded most. And during the late-stance push-off phase of walking, individual EMG reductions across the triceps surae reached 47.8% in the lateral gastrocnemius, 21.9% in the medial gastrocnemius, and 20.4% in the soleus, though these are individual late-stance walking maxima from an n=11 dataset and the group-level walking EMG effects are underpowered. The standing-balance 32% should not be conflated with the larger individual late-stance walking maxima. They are different measures of different things, both real, both worth understanding.
The clinical translation: the Sleeve is a defensible adjunct for the days the patient is walking around carrying load the tendon does not need to carry. Pair it with the loading prescription, not in place of it. The progressive heavy slow resistance, the soleus work, the staged return — that is the rehabilitation. The Sleeve is the tool that reduces the ambient micro-loading dose between sessions, so that when the patient turns up for the next loading session the tendon is not already at its reactive threshold from a thousand uncalibrated steps. Used that way, with intellectual honesty about what the data does and does not say, it gives a measurable load-modification lever for the variable your rehab plan was previously silent on.
The reason most Achilles tendinopathy fails to resolve is not exotic. It is underdosed soleus, undermanaged ambulatory load between sessions, mistimed return to running, and a clinician who promised six weeks when the tissue needs six months. Fix those four and most of the rest takes care of itself.
References
Chimenti RL, Neville C, Houck J, Cuddeford T, Carreira D, Martin RL. Achilles Pain, Stiffness, and Muscle Power Deficits: Midportion Achilles Tendinopathy Revision – 2024. J Orthop Sports Phys Ther. 2024;54(12):CPG1–CPG56.
Cook JL, Purdam CR. Is tendon pathology a continuum? A pathology model to explain the clinical presentation of load-induced tendinopathy. Br J Sports Med. 2009;43(6):409–416.
Demangeot Y, O’Neill S, Degache F, et al. Exercise parameters to consider for Achilles tendinopathy: a modified Delphi study with international experts. Br J Sports Med. 2025;59(10):1337–1349.
Docking SI, Cook J. Pathological tendons maintain sufficient aligned fibrillar structure on ultrasound tissue characterization (UTC). Scand J Med Sci Sports. 2016;26(6):675–683.
Lieber RL, Ward SR. Skeletal muscle design to meet functional demands. Philos Trans R Soc Lond B Biol Sci. 2011;366(1570):1466–1476.
Rio E, Kidgell D, Purdam C, et al. Isometric exercise induces analgesia and reduces inhibition in patellar tendinopathy. Br J Sports Med. 2015;49(19):1277–1283.
Silbernagel KG, Crossley KM. A Proposed Return-to-Sport Program for Patients With Midportion Achilles Tendinopathy: Rationale and Implementation. J Orthop Sports Phys Ther. 2015;45(11):876–886.