The Achilles doesn’t need a break, it needs load: how to rebuild a runner’s most temperamental tendon

Morning stiffness in the back of the lower leg, a dull ache in the first minutes of a run that “warms up” and eases off, a thickening a couple of centimeters above the heel — this is how Achilles tendinopathy begins, one of the most common running ailments. The logical first impulse is to stop running and wait for it to pass. And that is exactly the impulse that most often doesn’t work. A tendon isn’t a bruise; it doesn’t “heal at rest,” it remodels in response to load. The only question is which load triggers the remodeling and which one derails it.

What was studied

In 2026 the journal Frontiers in Sports and Active Living published a systematic review with a network meta-analysis (Fu et al.): the authors gathered 49 studies, of which 31 RCTs with 1388 participants entered the comparison network. They compared how different training regimens change the morphology (cross-sectional area, length) and mechanics (stiffness, elongation, strain) of the Achilles tendon. They set stretching, running-technique correction, strength work (plyometrics, isometrics, eccentrics — the classic Alfredson protocol) and combined protocols against one another.

In parallel, PLOS One (2026) published the protocol of a blinded RCT (Moreno et al.): 60 recreational runners aged 18–60 with midportion Achilles tendinopathy (pain 2–6 cm above the insertion, for at least 3 months) will be split into two groups. One works with a constant moderate load (55% of the one-repetition maximum), the other progressively raises it from 55% to 90% over 12 weeks — with equal total volume (about 1050 repetitions). This is still just a protocol, there are no results yet, but it nicely illustrates the very logic of treatment through load.

What works: load, not rest

The main takeaway of the meta-analysis: it is strength work, not stretching, that changes the tendon’s structure and mechanics. For stiffness — the key property that lets a tendon elastically return the energy of each stride — the top spot went to prolonged isometric work of the ankle (SUCRA 89.3%). For cross-sectional area, plyometrics — jump-based work — led the way (SUCRA 95.6%). Stretching programs, meanwhile, consistently landed at the bottom of the rankings on almost every measure.

Why load? The RCT protocol states the principle plainly: contractions of sufficient force cause tendon strain in a range considered favorable for positive remodeling — roughly 4.5–6%. Too weak a load doesn’t deliver that stimulus, while complete rest deprives the tendon of the signal to adapt, so it stays weak and painful. A second observation from the protocol matters too: remodeling is triggered even at moderate loads (55% of maximum), not only at heavy ones. So “heavy to failure” isn’t a required condition; the condition is a regular and sufficient stimulus.

How to load in practice

The practical skeleton that runs through the research is simple.

Start with isometrics. Calf raises (on both legs, using support if needed) with a 30–45 second hold, several sets. They are well tolerated when there’s pain and, per the meta-analysis, have the strongest effect on stiffness.
Move on to heavy slow raises. Weighted calf raises, 3 seconds up and 3 seconds down — a slow tempo matters more than heavy weight. This is exactly the phase (“slow calf raises”) the RCT protocol uses in the first weeks.
Progress gradually. First on both legs, then on one, then — adding weight and, closer to a return to running, light jump work. The principle of progressive loading: a little more than yesterday, but no sudden leaps.
Tolerate moderate pain. The benchmark built into the protocol: pain during and the day after loading should not exceed 4/10. Discomfort up to 3–4 points is acceptable and doesn’t signal harm.
Stock up on patience. Tendon remodeling takes weeks and months (the protocol runs 12 weeks of work plus follow-up of up to six months), not a couple of days.

What you should not do: retreat into complete rest “until the pain fully disappears,” aggressively stretch the tendon through pain, and chase a quick result with an abrupt spike in load.

Limitations

This isn’t a license for self-treatment. The network meta-analysis is honest about its weak points: 88% of the trials had <50 participants, the measurement methods and the protocols themselves varied widely, and there are still no generally accepted thresholds for a “clinically meaningful” change in stiffness and thickness. The PLOS One RCT is only a protocol, and it is limited to male runners with midportion Achilles tendinopathy; the conclusions don’t transfer directly to insertional pain or to women. This is about average trends, not a recipe for a specific case.

The main points

A tendon remodels under load, not at rest — complete rest more often drags the problem out.
Per the 2026 meta-analysis, stiffness is boosted most by isometrics and thickness by plyometrics; stretching barely works.
A working scheme: isometrics → heavy slow calf raises → progression to single-leg and jumps.
Pain up to 3–4/10 during and after is acceptable; the protocol’s benchmark is no higher than 4/10.
Be patient: the timeline runs in weeks and months, not days.
Avoid complete rest and stretching through pain.

With acute, sharp pain, swelling, a sensation of a “snap,” night pain, or a lack of progress over several weeks, be sure to see a sports physician or physiotherapist — this article does not replace an in-person consultation and is not individual medical advice.

Sources: Fu B. et al., Frontiers in Sports and Active Living, 2026. https://doi.org/10.3389/fspor.2026.1782503; Moreno M.J. et al., PLOS One, 2026. https://pmc.ncbi.nlm.nih.gov/articles/PMC12952620/