Achilles Tendon Load Tolerance: Heel-Rise Work Benchmarks for Return to Running

A 34-year-old recreational marathoner ruptures his right Achilles landing awkwardly during a pickup basketball game. Surgery, six weeks in a boot, then eight weeks of physical therapy built almost entirely around one number: single-leg heel-rise reps, climbing from three to fifteen. At his twenty-week checkup he manages fifteen reps on the surgical leg against thirteen on the healthy one — more reps on the repaired side. His surgeon clears him for a run-walk program. He starts a ten-minute protocol in week 21. By week 23 the tendon is swollen and tender two centimeters above the heel, and he's back in a boot for six more weeks.

What the fifteen-rep count never asked is the question running actually asks. It counted repetitions, not the work each one produced. Fifteen reps on a leg that's quietly cutting the range short and speeding up the tempo by rep eight isn't the same fifteen reps as a full, controlled rise on the healthy side — and a bare count can't tell the two apart. Running is several thousand ground contacts an hour, each one asking the calf-Achilles complex to absorb load eccentrically and return it concentrically, again and again, without the muscle-tendon unit's output degrading partway through. A test built to stop counting at an arbitrary rep target was never built to expose that, and the same clinical-clearance-versus-mechanical-readiness gap shows up outside the ankle in hamstring return-to-sprint force-velocity criteria — different tissue, same lesson.

Silbernagel, Nilsson-Helander, Thomeé, Eriksson, and Karlsson (2010), in Knee Surgery, Sports Traumatology, Arthroscopy, built and validated the test this article is built around: a standardized single-leg heel-rise protocol paced to a metronome, scored not by rep count but by total work — the height reached on every rep multiplied by body weight, summed until the leg can no longer hit the required height for two consecutive reps. Tested against patients recovering from Achilles tendon rupture, the total-work score picked up meaningful side-to-side deficits that a simple rep count missed outright, and the protocol's test-retest reliability came back high, with intraclass correlation coefficients in the 0.83 to 0.97 range across height, work, and fatigue-decline measures. The limitation the authors flagged themselves: the original apparatus is purpose-built equipment most clinics don't own, and the validation cohort was a single group of surgically repaired ruptures, so the cutoffs don't automatically transfer to tendinopathy managed without surgery.

Olsson, Petzold, Brorsson, Karlsson, Eriksson, and Silbernagel (2014), in the American Journal of Sports Medicine, followed a cohort of acute Achilles rupture patients out to two years using that same heel-rise protocol alongside patient-reported outcome scores. Heel-rise height and total-work deficits on the injured side — running roughly 10-19% below the uninvolved leg depending on the specific measure — persisted at the two-year mark even in patients who reported good subjective function and had returned to their prior activity level. That's the detail worth sitting with: patients felt recovered and functioned well enough day to day that the deficit went unnoticed, while the objective load-tolerance gap was still there on testing. It's an observational cohort without a comparison group managed under different return criteria, so it shows the deficit persists under standard care rather than proving a criterion-based approach would close it faster — but it's a direct warning against assuming symmetric rep counts or comfortable daily function mean the tendon caught up.

A third study addresses the timing side of the same question. Silbernagel, Thomeé, Eriksson, and Karlsson (2007), also in the American Journal of Sports Medicine, ran a randomized controlled trial comparing a pain-monitoring model — patients kept training and running through tendinopathy symptoms as long as pain stayed at or under 5 out of 10 on a VAS scale and settled back down by the next morning — against a more conventional rest-then-progress approach. The pain-monitoring group returned to full sport participation sooner with no worse outcome on the VISA-A tendon symptom score at one year. It's a small, single-blind trial in tendinopathy patients rather than post-rupture patients, and it doesn't measure heel-rise work directly, but it's the evidence behind treating a symptom response, not just a calendar date, as part of any return threshold.

A maximal-effort isokinetic or handheld dynamometer strength test captures peak force at a single joint angle, held for a second or two, maybe repeated three to five times. Running never asks the calf-Achilles complex for that kind of effort. It asks for a submaximal, repeated stretch-shortening cycle — eccentric loading absorbing impact, then a rapid concentric push-off — happening 150 to 180 times a minute, for however long the run lasts. The tendon has to store and return elastic energy efficiently across thousands of cycles without the muscle-tendon unit's output dropping off partway through. That's a fatigue-resistance and total-work question, not a peak-force question, and a peak-force test simply isn't built to expose a deficit that only shows up on the fortieth or hundredth repetition.

This is also why a pass-fail rep-count check — can the patient do fifteen single-leg heel raises, yes or no — misses so much of what matters. Two people can both hit fifteen reps from meaningfully different places: one completing every rep through a full range at a controlled tempo, the other cutting the range short and speeding up by rep eight to keep the count going. The total-work calculation catches this because it multiplies height by body weight for every rep instead of crediting a rep just for having happened — a shortened, rushed rep contributes less to the total, and the score for that side drops even when the raw rep count looks identical. This lines up with why isometric training rebuilds tendon capacity: sustained, controlled loading, not just moving the joint through a range, is what drives the adaptation a return-to-run decision is actually trying to measure.

There's a compensation pattern layered on top, too. An athlete protecting a healing Achilles will often unconsciously shift load toward the forefoot flexors or lean on the opposite leg during any double-limb task, which is why single-leg testing carried through to fatigue — not a double-leg calf raise, and not a single early rep — is the piece that catches what a quick check misses. The same principle underlies limb symmetry index cutoffs after injury in ACL and other lower-limb return-to-sport testing: a test done once, early, and read on its own undercounts a deficit that a repeated-loading task later exposes in full.

The equipment is minimal: a step or low box roughly 5-10 cm tall so the heel can drop slightly below toe level, a metronome (a phone app is fine), a body weight scale, and a way to capture heel-rise height per rep — a wearable position sensor or an app using the phone camera works in a field setting; a linear encoder or motion-capture system gives more precision if it's available. Protocol: (1) a five-minute light warm-up plus five bilateral submaximal heel raises to prime the calf; (2) single-leg stance on the step, knee kept straight, one or two fingers on a wall or rail for balance only, not for weight-bearing assistance; (3) set the metronome to a 2-1-2 tempo (two seconds up, one-second pause at the top, two seconds down) — about 20 reps per minute — and begin; (4) continue until either two consecutive reps fall below 50% of that leg's own first-five-rep average height, the athlete can no longer hold the tempo, or pain crosses the athlete's pre-set stop threshold; (5) record height for every rep and calculate total work as the sum, across all valid reps, of height in meters multiplied by body weight in kilograms multiplied by 9.81, giving a result in joules.

Run the uninvolved leg the same way after 5-10 minutes of rest, then calculate a Limb Symmetry Index: injured-leg total work divided by uninvolved-leg total work, times 100. The number that matters for a return-to-run decision isn't a population average — it's this side-to-side comparison, repeated over the course of rehab, ideally against the athlete's own pre-injury baseline where one exists. Most atraumatic Achilles cases won't have a preseason baseline on file, which is exactly why banking one on high-mileage runners and jumping-sport athletes before anything goes wrong is worth the ten minutes it takes.

Sequence the heel-rise total-work test across the return-to-run timeline rather than running it once at the clearance appointment. Stage 1, before starting run-walk intervals: total-work LSI at 75-80%, pain during the test at 3/10 or under, and no morning stiffness lasting past the first ten minutes out of bed. Stage 2, progressing run-walk ratios toward longer run bouts: total-work LSI climbing toward 85%, using the pain-monitoring model from Silbernagel et al. (2007) as the symptom gate — running-related pain up to 5/10 is acceptable as long as it settles back to baseline by the next morning; pain still elevated the next day means holding the current stage rather than advancing it. Stage 3, continuous running for 20-30 minutes at an easy pace: total-work LSI at 90% or higher, height LSI at 85% or higher, rep-count LSI at 85% or higher — all three, not just one. Stage 4, return to cutting, plyometric work, or sprinting: LSI at 95% or higher across all three heel-rise metrics, confirmed alongside a hop-test battery.

Retest every one to two weeks through the progression, because height and work tend to normalize before pain fully settles, and pain can feel fine while total work still lags — either mismatch is a reason to hold the stage rather than push forward on one measure alone. Once running clearance is reached, that's not the finish line for capacity building: eccentric overload training layered in afterward can push total-work capacity past baseline rather than just back to it, the margin that actually protects against a second injury under real training load.

Olsson et al.'s two-year data is the reason none of this stops at Stage 4. Deficits in that cohort were still measurable at two years in patients who felt fine and were back to full activity — a clean pass at twelve or sixteen weeks earns a return to running, not a permanent pass on ever checking again. Recheck at three months and at a full year in a high-mileage or plyometric sport, because tolerating today's load isn't proof of tolerating next season's.