Deep divefrom Episode 5Training6 October 202618 min

Tired Calves, Busy Knees, and a Super Shoe That Didn't Help

A deep dive into one paper: what was known before, what the researchers did, what the numbers say and what they don't.

The paper: Nahan, Willwacher, Kerr, Héroux & McDonald — Positive Joint Work Redistribution in Running: The Role of Plantar Flexor Fatigue and the Effect of Advanced Footwear Technology (Medicine & Science in Sports & Exercise)

MP3

Chapters

Also linked

Preprint full text (bioRxiv)

Transcript

The script the two AI voices read from. The times are worked out from the script, not heard, so a line can start a few seconds early or late.

Read the whole episode

Cold open

Host 1You heard the trailer — this is the whole story. Welcome to the Hastrup Review of Science — where artificial intelligence reads real science, so you don't have to pretend you understood the paper. One paper today: when your calves get tired late in a race, does your knee quietly take over the work, and can a carbon-plated super shoe stop that from happening?

Host 2The paper is by Key Nahan, Steffen Willwacher, John Kerr, Martin Héroux and Kirsty McDonald, from the University of New South Wales and Neuroscience Research Australia in Sydney, the University of Massachusetts Amherst, Offenburg University in Germany and Edith Cowan University in Perth. It appeared in Medicine and Science in Sports and Exercise in September twenty twenty-six, and the full text we worked from is the preprint the authors posted on bioRxiv in October twenty twenty-five, so small details may differ from the final journal version. The short answer: tired calves did shift work from the ankle to the knee, and the super shoe made no detectable difference to that shift.

Host 1So the shoe lost.

Host 2In this specific test, yes, and the authors themselves say the result surprised them. But the way the test was built matters a great deal for what that loss means, and we will spend some time on exactly that.

Why joint work matters

Host 1Start with the basics. What is joint work, and why would anyone measure it in runners?

Host 2Every stride, your leg absorbs energy when you land and generates energy when you push off, and biomechanists split that across the ankle, the knee and the hip. Positive work is the pushing part: muscles shortening and tendons recoiling to drive the body up and forward. Measuring it tells you which joint is doing the job.

Host 1And why should a runner care which joint does it?

Host 2Because the joints are not equally cheap. As the paper sets out, mechanical work explains roughly seventy-six percent of the metabolic cost of running across a range of submaximal speeds, according to a twenty twenty-two study by Riddick and Kuo. And the ankle is the economical engine: the Achilles tendon stores and returns elastic energy, and the calf muscles have short fibres and small volumes, so each unit of work costs less fuel than the same work done by the big muscles around the knee and hip.

Host 1How much of the work does the ankle normally do?

Host 2In fresh running at three and a half meters per second, earlier work cited in the paper found the ankle supplies about fifty-four percent of the positive work, with the knee and the hip at about twenty-three percent each. That distribution is thought to be part of why trained runners are efficient.

What was known before

Host 1What was known before this study?

Host 2Several studies, including one from twenty eighteen by Sanno, Willwacher and colleagues, found that over an exhaustive run of three to ten kilometers, positive work drifts from the ankle up to the knee. That has been seen in recreational and trained runners, both men and women. Running economy is also known to get worse late in hard efforts, so the suspicion is that this shift is part of the reason.

Host 1And the suspected culprit is the calf.

Host 2The plantar flexors, meaning the calf muscles that point the foot down and push you off the ground. The circumstantial case is decent: the shift appears at ten-kilometer race pace or faster, but a study by Melaro and colleagues found no shift after a long, easy twenty-kilometer run, in rearfoot strikers. Hard running fatigues the calves, easy running mostly does not, so the pattern fits.

Host 1Where do the shoes come in?

Host 2As established background, advanced footwear technology, often called super shoes, arrived with Nike's Vaporfly around twenty seventeen and combines a thick, very resilient foam with a stiff, curved carbon plate. Lab studies found they lower the energy cost of running by a few percent on average, and road racing times dropped across the board. The paper adds that super shoes reduce both negative and positive ankle work, and that a shoe with a carbon plate under the insole delayed, but did not prevent, the shift of work towards the knee during a fatiguing run.

The open question

Host 1So what was the open question?

Host 2Two of them. First, nobody had shown that calf fatigue on its own actually causes the shift, because in a real race everything gets tired at once: calves, thighs, hips, breathing and the brain. Second, if super shoes help partly by protecting the calves, does a runner who already has tired calves still keep a better distribution of work in them?

Host 1How do you isolate the calf from everything else?

Host 2You tire the calves directly, without the run, and then make the person run. That is the core trick of this study.

Methods: who and how

Host 1Who were the runners?

Host 2Twelve healthy trained men, average age twenty-six, with ten-kilometer race times between thirty-five and forty-five minutes. Their power calculation called for thirteen participants, so they landed one short. Only men were recruited, because female fatiguability differs and varies with the menstrual cycle, and the study was originally planned as part of a larger multi-day project.

Host 1And the two shoes?

Host 2The traditional shoe was the Brooks Hyperion Tempo, a light trainer with conventional foam and no plate. The super shoe was the ASICS Metaspeed Sky Plus, with a rocker shape, a thicker stack of a more resilient foam and an embedded carbon plate. Both weighed exactly two hundred and four grams in a men's size nine, which removes weight as an explanation, and each runner's shoe order was randomized.

Host 1Walk me through the day.

Host 2First, maximal calf strength: seated, knees straight, foot strapped to a plate on a custom dynamometer, pushing as hard as possible three times in each shoe. Then a thirty-second run on an instrumented treadmill at each runner's own ten-kilometer race pace, which averaged four meters per second, about fourteen and a half kilometers per hour, in each shoe. Eight infrared cameras tracked reflective markers on the body, and force plates under the belts recorded every footstrike.

Host 1Then the fatigue part.

Host 2Two sets of bodyweight calf raises on the edge of a twenty-centimeter step, to a metronome at sixty beats per minute, each set to failure. Then another maximal test to confirm the calves had dropped to between seventy and eighty percent of their peak torque, which matches what earlier studies measured after a near-maximal ten-kilometer run. If they were not in that window, they did more raises or rested briefly, and then went straight onto the treadmill within about ten seconds for an identical thirty-second race-pace run.

Host 1And then the same thing again in the other shoe.

Host 2Yes, with another maximal test after each fatigued run, and a top-up of calf raises in the second shoe until they were back in the target range. Everything happened in a single session.

Methods: the analysis

Host 1How do you turn markers and force plates into joint work?

Host 2With inverse dynamics, which works backwards from the ground reaction force and the motion of each body segment to the net moment and power at each joint. They analyzed the last twenty stance phases of each run and integrated the power curves, adding up the positive parts as positive work and the negative parts as negative work. Everything was scaled to body mass, leg length and stride, so the numbers are unitless, and the most useful output is each joint's share of the total.

Host 1Anything unusual in how they treated the ankle?

Host 2They used what is called the ankle joint complex approach, treating the foot as one rigid block from the heel, to avoid overestimating ankle power. The consequence is that the small joints of the foot, including the toe joint where a carbon plate does much of its work, were not modeled separately. We will come back to that.

Host 1And the statistics?

Host 2Linear mixed models with each runner as a random effect, testing run, shoe and the interaction between them. The key question for the shoes was that interaction: did fatigue change joint work differently in one shoe than in the other.

Results: the fatigue

Host 1Did the calf-raise protocol work?

Host 2It did. Pooled across shoes, peak calf torque went from about eighty-six newton meters at baseline to sixty-six after the calf raises, a twenty-four percent drop, right inside the target. The unfatigued run itself caused no measurable fatigue, and baseline strength was similar in the two shoes.

Host 1Did the fatigue hold through the run?

Host 2Partly. After the thirty-second fatigued run, torque was back up to about seventy-two newton meters, so the calves recovered by about seven percent during the run itself. They were still around seventeen percent below fresh, which the authors argue is enough to matter.

Results: joint work

Host 1Now the headline. What happened to the work?

Host 2With tired calves, positive ankle work went down and positive knee work went up, with p values below zero point zero zero one for the ankle and of zero point zero zero one for the knee. Hip work did not change detectably. The ankle's share of total positive work fell by about four percentage points and the knee's share rose by about three, with the hip picking up roughly one.

Host 1Four percent sounds small.

Host 2It is four percentage points of the total, which hides something. Working from the paper's table, the ankle's absolute positive work fell by roughly eight percent, and the knee's rose by roughly a fifth, because the knee starts from a much smaller base. Total positive work across the three joints stayed almost exactly the same, so the body did the same job, with a different mix of joints.

Host 1What were the actual shares?

Host 2From the table, in the fresh run the ankle did a little over half the positive work, the hip about a third and the knee about a sixth. After calf fatigue, the ankle was down to under half and the knee was close to a fifth. Note that in these runners the hip already carried more and the knee less than in the older reference values I mentioned earlier.

A word from our sponsor

Host 1Are your calves tired of holding everything together under load? Then you are in luck.

Host 1The Hastrup Review of Science is brought to you by clothes pegs. A small steel spring stores elastic energy and returns it on release, every single time.

Host 2I use mine every week. Under heavy load, none of them shifts the work to the peg next to it, although I have not run a mixed-effects model on that.

Host 1They come in wood, plastic and stainless steel, in packs from twenty to two hundred. Widely available and very reasonably priced.

Host 2And unlike a fatigued calf, they recover fully between sets.

Host 1Clothes pegs. Our referral link is in the show notes.

Results: the shoe question

Host 1Back to the runners. Did the super shoe change any of that?

Host 2No. For ankle, knee and hip positive work, adding the shoe-by-fatigue interaction did not improve the model, with p values of zero point eight two, zero point two five and zero point eight five. In plain words, the shift towards the knee looked the same in the Brooks and in the ASICS.

Host 1Did the super shoe change anything at all?

Host 2Looking at the table, in both fresh and fatigued runs the super shoe shows somewhat lower positive ankle and knee work than the traditional shoe, which fits earlier findings. But the paper does not report or discuss a test of that overall shoe difference, so we should not lean on it. What it does test is whether fatigue moved work differently in the two shoes, and it did not.

Host 1What about the landing side, the negative work?

Host 2Negative work at the ankle, knee and hip did not change with fatigue, and again no shoe interaction. Stride length, stride time and flight time were unchanged too. Contact time was slightly shorter in the fatigued runs, but with a p value of zero point zero seven, which is not a finding.

Host 1So outwardly the stride looked the same.

Host 2Same stride length, same cadence, same time on the ground, within what this study could detect. The change was inside the leg: which joint powered the push.

Why it matters

Host 1Put it together. What is the principle here?

Host 2That calf fatigue on its own is enough to produce the proximal shift seen in exhaustive races, without the rest of the body being tired. That moves the calf from suspect to a demonstrated contributor, at least in trained men and at race pace. It does not mean the calf is the only cause in a real race.

Host 1And for the shoes?

Host 2Once the calves are substantially fatigued, the shoe does not seem to rescue the distribution of work. The authors are careful that this is not the same as saying super shoes do not help with fatigue, because a super shoe might delay the calves getting tired in the first place. By imposing the fatigue with calf raises, this design deliberately removed any chance for the shoe to show that.

Host 1What changes for a runner listening to this?

Host 2The authors suggest that calf endurance, for example from calf-raise training, matters for holding form late in a race, and that shoes may not be enough to offset heavy calf fatigue. That is a reasonable inference, but this study did not test any training program. It also did not measure oxygen cost, so the idea that the shift makes running more expensive comes from earlier research, not from this experiment.

Host 1Is there any direct evidence that the shift actually costs energy?

Host 2The paper is candid that it is not settled. The authors mention their own unpublished work finding no effect of calf fatigue on running economy at slower speeds of ten to twelve kilometers per hour, and argue that race pace probably amplifies the effect. They also note that at these speeds standard breath-gas measurements cannot capture the anaerobic part of energy use, so the real cost is hard to pin down.

Honest caveats

Host 1Let's do the caveats properly.

Host 2Start with size: twelve men, one fewer than their own power calculation asked for. That is enough to detect a fatigue effect that every runner showed in the same direction, but it is weak for showing that two shoes behave the same. A non-significant interaction means no difference was detected, not that the shoes are proven equivalent, and a modest shoe effect could easily hide in a sample this small.

Host 1What about the fatigue itself?

Host 2Calf raises are not running. In a race the calves fatigue through thousands of rapid stretch-and-recoil cycles, and the authors acknowledge that fatigue from a slow single-joint exercise may differ. Strength was also measured isometrically, with a straight knee, which matches earlier ten-kilometer studies but may miss dynamic changes.

Host 1And the recovery during the run?

Host 2The calves recovered by about seven percent within thirty seconds, so the fatigued runs were somewhat less fatigued than intended by the end. The authors suggest this may be why the shift here was smaller than in studies using a real exhaustive run. Thirty seconds is also a very short window compared with the final kilometers of a race.

Host 1The foot was not modeled separately. Why does that matter for the shoe question in particular?

Host 2Because the carbon plate and the rocker shape work largely around the toe joint, and the paper treats the foot as one block from the heel. The authors note the foot can supply up to about two point six percent of positive power and that its work may change with fatigue, so a shoe effect located there could be missed. The foot markers were also fixed on the shoes, not the skin, which is standard but adds uncertainty with thick-soled shoes.

Host 1Any design issues beyond that?

Host 2The fatigued run always came after the fresh run, which is unavoidable, and the second shoe's fatigued run came after a second dose of calf raises on already tired legs, with shoe order randomized to spread that out. Everything was in a single session with men only, so the authors say the results cannot be directly applied to female runners. And the full text we read is the preprint, where a couple of confidence intervals in the knee row of the work table look like typos.

Host 1Anything you would flag in how the result is framed?

Host 2The conclusion calls the shifted pattern a suboptimal gait strategy. That is plausible from earlier work, but this study did not measure economy or performance, so whether the shift is a cost, or simply the best available response to tired calves, is still an open question. The authors' own suggested next step is to find out how big a shift has to be before performance actually suffers.

What to watch next

Host 1What should we watch for next?

Host 2A study that runs people to genuine exhaustion in super shoes and traditional shoes, measuring calf strength before and after, would test the delay idea directly. The paper notes that a half-marathon study found more calf fatigue with flat carbon plates inserted into ordinary shoes, which is the opposite direction, so the fully integrated super shoe design needs its own test. Beyond that: women, foot and toe-joint modeling, and a direct link between the size of the shift and the race clock.

Host 1And for training?

Host 2A randomized trial of calf endurance training that checks whether it preserves ankle work late in a race, and whether that shows up in finishing times. Until then, calf raises are a sensible idea supported by mechanism, not a proven performance intervention.

Sign-off

Host 1That is the deep dive: positive joint work redistribution in running, by Nahan, Willwacher, Kerr, Héroux and McDonald, in Medicine and Science in Sports and Exercise. The link is in the show notes.

Host 2Every Friday this same feed carries the roundup episode: five new papers as trailers, and you pick which become deep dives. This has been the Hastrup Review of Science.