Carbon “super shoes”, checked

Prefer to watch? The whole check in 4:45, narrated, subtitles in the picture — it has its own page too. Every number in it is traced to a source below.

1 What the number actually is now

Every article about these shoes quotes 4%. That number is real, and you can point at where it came from: a 2018 laboratory study by Hoogkamer and colleagues at the University of Colorado — eighteen very fast men on a treadmill, in Nike's Vaporfly prototype, with the shoes weighted so the comparison was not simply about mass. In that study the prototype cut the energy cost of running by an average of 4.16%.

One thing about that study belongs in the same breath as its number, and is usually left out: it was funded by Nike, on a contract to the university, and two of its authors were Nike employees while a third was a paid Nike consultant. That does not make the measurement wrong — the methods are careful and the mass-matching was unusually rigorous. It does mean the single most-quoted figure in this entire field comes from a study paid for by the company selling the shoe, and it is a reasonable partial explanation for why it sits at the top of the range rather than the middle.

Eight years of studies later, someone put them all together. That is what a meta-analysis is: you gather every study that asked the same question and work out what they say as a group, which is far more reliable than any single one of them. A 2026 meta-analysis gathered fourteen studies and 271 runners, and across its four ways of measuring the same thing — running economy, metabolic cost, oxygen uptake, energy cost of transport — the saving came out between 2.62% and 2.88%, averaging 2.75% (figure 2). The authors graded their own confidence in it as moderate — in plain words: fairly sure, and the number could still move a little, but probably not far. It is a real effect, a little smaller than the famous number.

Two things are worth being precise about, because they are easy to garble. The four estimates are remarkably consistent with each other, but the confidence interval around them is wide: averaged across the outcomes it runs from roughly 1% to 4.5%. That range describes how sure we are of the average — it is not a spread between studies, so "as little as 1%" describes the uncertainty around the average, not a study that found almost nothing. And the headline number is an average across four different measures; running economy on its own rests on four of the studies, not all thirteen that contributed data.

There is a second thing hiding in that comparison. The 2018 study measured men who had recently run 10 km under 31 minutes. The pooled studies include recreational runners. If a benefit shrinks as your sample gets more ordinary, that is worth knowing before you decide it applies to you.

And there is a third thing, which almost nobody mentions and which we only added here in August: every one of these numbers is measured on the flat. A 2026 study ran twelve competitive runners at four gradients in a Vaporfly against a light racing flat, and the saving shrank the moment the road tilted: 4.22% on the level, 2.42% at 3%, 1.05% at 6%, and 0.52% at 9% — roughly a fifth of the benefit gone for every extra percent of gradient. Above a moderate climb there is essentially nothing left. The mechanism makes sense: at a steep gradient you are lifting your body rather than bouncing it forward, and a shoe that returns energy to a forward bounce has less to give. So if your race is hilly, or you run trails, the headline number is not your number. On a rolling course it is somewhere between the two, and nobody has measured that yet. (What a gradient costs you in time is exactly what the climb calculator works out.)

And there is a problem running underneath every study in this field that almost nobody mentions: you cannot blind a shoe. A runner knows immediately whether they have a stiff plated racer or a flat trainer on their feet, and they know which one the sport says is fast. Expectation moves performance measurably, and no trial here controls for it. The one study that tried — giving runners two identical pairs and telling them one was a cheap knock-off — found belief changed how good the shoes felt without changing the physiology. That is reassuring for the metabolic numbers and not at all reassuring for anything measured by racing.

Two more design details worth knowing, since they shape everything above: all fourteen studies were crossover trials, where each runner tries both shoes — good, because each person is their own comparison. But only eight of the fourteen randomised the order, and a first trial is almost always slower than a second. Where randomisation is missing, some of the benefit could be familiarity rather than footwear.

Energy saved is not time saved

This is the easiest step in the whole story to miss, because the two numbers sound interchangeable and are not. Running economy is how much energy it costs you to hold a given speed. Finishing time is what the clock says. They are related, but they are not the same quantity, and the exchange rate between them is not one-to-one.

A 2019 modelling paper worked out the conversion. At around world-class marathon pace, a percentage improvement in economy buys you roughly two-thirds of that percentage in speed. So a 2.75% saving works out around 1.8% faster for an elite — although the meta-analysis's own authors put the expected marathon-time gain closer to 1%, which is a useful reminder that this step is a model and not a measurement. There is a wrinkle here that is easy to get backwards: at paces slower than roughly nine minutes per mile — which is where most people reading this actually run their marathons — the model predicts the speed gain is slightly larger than the economy gain, not smaller.

But modelling is modelling. When researchers pooled the studies that actually timed a trial instead of measuring oxygen, the effect on actual performance came out at roughly half the size of the effect on energy. Careful: that is half in the statisticians' way of comparing two different measurements to each other — it does not mean half of 2.75%. Somewhere between the treadmill and the finish line, most of the saving goes missing — figure 3 shows how little survives each step.

The plate and the foam work together

The category is named after the plate. The evidence points at the combination instead.

The largest pooled analysis in this field — 48 studies and 878 runners — set out to separate the two ingredients, the stiff plate and the springy foam, and found that neither does anything measurable on its own. The saving appears only when they are combined, along with the rockered shape that comes with them (figure 4). One care in reading that: "nothing measurable" means the studies could not pin an effect down — it is not proof of exactly zero. And the authors rated the quality of that evidence as low to very low — meaning they are not very sure at all, and new studies could change it — which is itself worth sitting with: the headline mechanism of a technology that has reshaped road racing rests on a shaky evidence base.

The clearest demonstration of that came in 2026, when researchers isolated one ingredient properly. Not the plate, and not the foam's springiness, but the midsole's compliance — how far it squashes underfoot. On its own, more compliance improved running economy by 3.90%. And the mechanism was not where everyone looks: the saving showed up at the knee, where runners used nearly ten percent less positive joint power, not at the ankle where the plate is supposed to do its spring impression. That does not contradict the pooled analysis above — that one tested the foam's energy return, which is a different property from how far it gives way. Together they say something more interesting than either alone: the soft, deep midsole is carrying more of the benefit than the part the category is named after.

There is an apparent counterexample on the track, and it is worth walking through because it turns out to make the same point. A 2025 study put thirteen middle-distance runners in spikes with and without a carbon plate. In the national-level group, the plated spike was measurably worse — they used more oxygen wearing it, and the difference was too large to be chance. The authors' reading is that stiffness has an optimum that differs per athlete, and past that point a stiffer shoe costs you.

Then in 2026 a second track study reached the opposite-sounding result. Seventeen competitive runners — seven of them women — ran in two models of complete advanced-technology spikes, and their economy improved by 2.1% and 2.3% against traditional spikes. Both findings are real, and the difference between them is exactly the theme of this section. The 2025 study added a plate to a spike and made things worse; the 2026 study compared a whole shoe, designed around compliant foam and geometry, against an old-fashioned one and made things better. Adding the famous ingredient is not the same as building the shoe around it.

2 What we still do not know

Two gaps matter more than the rest, and both are large enough that confident numbers outrun the evidence.

Almost everything was measured on a treadmill

Thirteen of the fourteen studies in the 2026 meta-analysis were treadmill work. In late 2025 a group took the question outdoors with thirty-six national-to-international runners — the largest sample in this literature, and the only one with a roughly even split of women and men. Running outdoors at threshold pace — roughly the fastest pace you can hold for an hour — they found no reliable difference between shoes at all: compare the two panels in figure 6.

That is not proof the shoes do nothing outdoors, and it would be dishonest to present it that way. The authors' own explanation is measurement noise: with a single trial per shoe and limited familiarisation, the method simply cannot resolve an effect of two or three percent. A separate 2026 study showed how bad this gets: with three repeat trials the effect was large and unmistakable, and using only the first day's data from the same runners it nearly vanished — the same shoes, the same people, a different answer depending on how many times you measured.

This has a direct consequence for a claim you will see everywhere: that some fixed percentage of runners are “non-responders”. Those figures come from single-session studies, and single sessions cannot separate a person who genuinely does not benefit from a person who simply had a bad morning. The best-documented spread makes the point: in a study of world-class Kenyan runners and European amateurs, individual results ranged from an 11.3% drawback to an 11.4% benefit in the world-class group alone. Numbers that wide are exactly what you get when one measurement carries all the noise of one day. We could not trace the most-quoted non-responder percentages to a primary study.

What 2026 did add is a run of failed explanations, and that is its own kind of finding. If some runners really do gain more than others, something ought to predict it — so researchers went looking. It is not your sex, and it is not your build: height, mass, leg length and Achilles tendon length all came up empty. It is not your calf strength either; a separate study fatigued runners' plantarflexors and found that neither the fatigue nor their push-off power was related to how much the shoes helped them. Three plausible candidates, three misses.

One thing did track, and it changes the shape of the question. When researchers looked at what distinguished a good response from a poor one, it was not a fact about the runner but about the pairing: the shoe models that shortened a given runner's ground contact time were the ones that improved that runner's economy — and no single model did that for everybody. So the honest position is not that responsiveness is random. It is that the thing worth knowing is how a particular shoe changes your stride, which nobody can tell from your height, your sex or your calf strength — and which no shop can measure while you stand on their treadmill for ninety seconds.

The largest reported gains are in women, and the evidence in women is thinnest

Of the fourteen studies in the meta-analysis, three included any women. Several of the best-known studies are entirely male. Meanwhile, when researchers compare marathon times before and after the technology arrived, the women's improvement is more than double the men's (figure 7).

Both of those things cannot be comfortably true at once. The era comparison watches what happened rather than testing anything, and everything else changed in the same decade — field depth, prize money, pacing rules, training — so it should be read as an upper bound on all of it together, not as a shoe effect. But it does point at the biggest hole in the literature.

Since we first wrote this, someone has started filling it. A 2026 study put fifteen women and fifteen men, all highly trained, through the same shoes at three speeds — the balanced design this field keeps not doing. Women used less energy in absolute terms, as you would expect from lighter runners, but the improvement the shoes gave them was no different from the men's (4.2% against 4.7%, a gap far too small to call real). The individual spread was large in both groups: 1.1% to 6.4% among the women, 0.2% to 8.7% among the men. And no body measurement they took — height, mass, leg or Achilles tendon length — predicted who would benefit. That cuts against the simple reading of the race data: if the shoes do not treat women differently in a lab, the doubled improvement in marathon times is even more likely to be about everything else that changed for women's distance running in the same decade. The honest answer to do these work better for women? has moved from nobody has looked to one balanced study looked and found no difference — which is progress, and still one study.

And the size of that upper bound depends on who is counting. The 4.0% for women in figure 7 comes from a 2026 analysis of marathon-majors results. The governing body's own scientists ran the same kind of before-and-after comparison in 2021, over 2016–2019, and found women's marathon times down 2.0% and men's down 1.2% — for the marathon, half the size. Two teams, the same question, a factor of two apart: that is how sensitive these era comparisons are to the years, races and fields you choose, and one more reason to treat figure 7's right-hand panel as a ceiling rather than a measurement.

3 What is actually in the shops

The rest of this section leans on sources outside the research literature — manufacturer pages, a corporate press release, independent teardowns and practitioners talking on the record. Those can establish what is being sold and said. They cannot establish that something works, and we have not used them that way.

The market itself has quietly moved on. The “super trainer” — the everyday version of a race shoe — has become the fastest-growing category in running. And a large share of them have no plate in them at all.

The most praised super trainer of 2026, the ASICS Superblast 3, is plateless. ASICS' own press release sells it entirely on foam and rebound and never mentions a plate, because there is not one. Others in the category use nylon, plastic or fibreglass rather than carbon (figure 8). The name stuck; the hardware moved on.

That matters for a practical reason. If the benefit comes from the combination of foam, geometry and stiffness rather than from carbon specifically, then “carbon-plated” is a marketing category, not a performance one. Judging a shoe by whether it contains carbon is judging it by the ingredient the evidence is least able to credit.

What practitioners say, and who is paying them

Coaches and physiotherapists converge on rotation rather than daily use — typically saving the plated shoes for the sessions where they matter, with one coach putting a rough ceiling of about a fifth of weekly mileage. Their reasoning is consistent: there is little to gain on an easy run, the foam degrades quickly, and the shoes are expensive.

Two things are worth knowing about that advice. The first is that the most widely repeated durability figure — that these shoes are finished at around 100 km, about 60 miles — appears most prominently in material published by a shoe retailer. Worth weighing, because advice that shoes wear out fast, and that you also need separate daily trainers, also sells shoes. It may still be right; it is just not disinterested. The second is more honest, from a coach who put it plainly: nobody has actually tested whether a gradual transition protects you. It is recommended because the doctors who reported the injury cases recommended it, not because there is a trial.

The clearest voice against daily use came from a product lead at one of the manufacturers, who suggested runners keep a range of plated and unplated shoes rather than living in the plated ones. When someone selling plated shoes tells you to spend less time in them, that is worth noting.

4 How long they last, and the rule nobody tested

Ask around and you will be told these shoes are finished somewhere between 100 and 250 miles — roughly 160 to 400 km, a range that itself spans a factor of four — and that you should save them for race day. That advice is repeated everywhere. It rests on a single experiment.

In 2024 a group put 22 runners through four conditions: a new and a worn shoe in each of two foams. The fast foam lost 2.28% of its running economy after 450 km — which is to say it lost roughly the entire benefit a fresh super shoe gives you in the first place. The ordinary foam lost nothing measurable (figure 9). It is a careful study — and it is one study, on prototypes that were never sold, worn by a single researcher, measuring how the shoe bends rather than how it compresses. Two years on, nobody has repeated it.

That last detail deserves more attention than it gets. The worn shoes scored 73.9 out of 100 for comfort when new and 73.8 when worn out. Perceived exertion and soreness did not shift either. If a shoe can lose nearly forty percent of its stiffness without anyone noticing, then “they feel dead” is not a reliable instrument, and neither is your sense of when to replace them.

There is counter-evidence, but we are not going to put numbers on it here: the two tests that point the other way — an independent physiologist's own pair at 400 km, and a nine-pair mechanical study at roughly 850 km — are a self-published single-subject test and a conference abstract respectively, and we could not read either at source. They are enough to say the question is open. They are not enough to quote. The honest position is that the popular mileage numbers were never measured. They are convention.

A related story, worth telling because it shows how these numbers drift. The adidas Adizero Adios Pro Evo 1 was widely reported as designed to survive a single marathon. In a long-mileage test that Trail Runner published in May 2024, Patrick Nava, adidas' vice-president of product for running, called the single-use reading a misconception — a nice, sensational headline, in the reporter's rendering — and put the intended claim in his own words: “What we said is that it is optimized for one race. […] Optimized means you get the best performance in your first race.” The tester in that piece ran the shoe for nearly 200 miles before it stopped feeling like a marathon shoe, and past 300 was still happy to race 5Ks and 10Ks in it. The claim drifted from optimised for to designed for to only good for as it passed through headlines.

5 Injuries: what changed since the last version

When we first wrote about these shoes, the honest summary was that nobody had counted injuries. That is no longer true, and the update runs against the popular story.

The concern started with a 2023 paper describing five navicular bone stress injuries in competitive athletes who trained in plated shoes. It is worth being precise about what that paper is: the authors labelled it a current-opinion piece illustrated by a case series, not a study. There were five cases and no comparison group, two of the five had suffered the same injury before — the strongest known risk factor for it recurring — and the authors stated plainly that the mechanism in each case could not be determined.

Since then, three separate groups have actually counted — though it matters what kind of sources these are: one is a preprint that has not been peer-reviewed, and two are conference abstracts whose full texts we could not obtain. A six-month trial in 414 recreational runners found no reliable difference. A twelve-week trial in 195 half-marathon runners reported injury risk halved in the plated group — though one of its authors works at the manufacturer whose shoe was tested, which is worth weighing accordingly. And a group of 271 marathon entrants, followed over time, showed that people using these shoes ran 37% more distance and were not injured more often.

The mechanism has moved too. A 2026 pooled analysis of fifteen biomechanical studies found these shoes do not consistently shift load to the knee or the hip — the change everyone assumed was happening. The only thing that came close to a real finding was less push-off power at the ankle — and even that sat right on the edge of what counts as a result at all, with the range of possible values just touching zero. And in a study from the same research group that raised the original alarm, the plated shoe reduced the ankle push-off moment, which their own institution described as a possible protective effect.

What remains genuinely unresolved is smaller and more specific than the popular story: the measured shifts are small and mostly too weak to count as findings — a trend towards lower cadence (p = 0.06, a whisker outside what counts as statistically clear) and, in one elite sample, more movement at the arch. Both are associated with bone stress injuries, and nobody has followed runners long enough to know whether shifts this small add up to anything. Note also that a separate study measuring joint loading found these shoes reduced the forces on the ankle and calf without raising them anywhere else. The one hint of a countercurrent comes from a 2026 study that tired runners out first: its authors suggest that once the calves fatigue, a very stiff shoe may push load up towards the knee and hip. They put it as a maybe, it is a single small experiment, and it is the sort of thing that will either be replicated or quietly disappear — but it is the most specific version of the worry anyone has managed to state. And the universal advice to transition gradually has never been tested at all — it is recommended because the doctors who described the cases recommended it.

6 The rule that quietly disappeared

In January 2020, after the first wave of these shoes, the governing body wrote a fairness guarantee into its rules: a shoe had to have been on open sale to any athlete for four months before it could be raced. Anything else counted as a prototype and was banned. The stated purpose was to stop elite results being decided by who had access to unreleased equipment.

That rule no longer exists. We read the current regulations — the version in force today was approved by the World Athletics Council on 19 March 2026 and took effect on 20 April 2026; we re-read and archived it on 10 August — and the requirement is now only that a shoe be purchasable “at any time prior to the start date” of the race — explicitly including pre-orders, and with “no obligation on a shoe manufacturer to re-stock” a shoe that has sold out. There is no minimum window at all. A single pair, sold once, minutes before the gun, satisfies it.

What that looks like in practice arrived in April 2026, when Sabastian Sawe ran 1:59:30 at the London Marathon on 26 April — the first sub-two-hour marathon in a record-eligible race. The shoe Sabastian Sawe raced in — the adidas Adizero Adios Pro Evo 3, $500 — went on sale in a limited drop two days after the race, sold out almost immediately, and turned up on resale sites at several times its price. Under the 2020 rule that shoe could not have been raced at all. Under the 2026 rules it did not have to be: sponsored athletes can compete in a “Development Shoe” — defined in the regulations as one that has never been available for purchase — for up to twelve months, provided it is not a World Championship or the Olympics.

So the honest reading is not that a rule was broken, and not even simply that a rule was deleted. It is that the sport now has two doors: the availability requirement was loosened to almost nothing, and alongside it there is a formal route for racing a shoe nobody can buy. Both are written down, both were followed. Whether that is the right trade is a fair argument to have — but it should be had about what the rules actually say.

A caution about how this is being reported. Several pieces have described this as the governing body breaking its own rule, and cite a requirement that shoes be available a month before competition. We went to the current regulation text and that requirement is not in it; nor is the older “reasonably available to all” wording. What the text supports is simpler: nothing was broken — the requirement had been removed first.

Meanwhile the price collapsed at the other end

The affordability complaint is the one part of this story that has genuinely improved, and it gets almost no coverage. Nike's original Vaporfly launched at $250 in 2017; the current mainstream racers — Vaporfly 4, Alphafly 3, adidas Adios Pro 4, ASICS Metaspeed — sit between $250 and $305 — roughly flat in nominal terms, and a real-terms fall over nine years. What is new is a halo tier above them at $500, which is a new category rather than inflation of the old one.

And on 9 March 2026 Lidl put a carbon-plated shoe, the Crivit CarbonLite 1.0, on sale for €49.99. Whatever that shoe does or does not do — nobody has tested it — the technology has commoditised to a tenth of the elite price in under a decade. The access problem that dominated commentary in 2020 was real, and it is largely solving itself.

The unfairness that elite athletes actually describe was never the retail price. It was contractual: being locked by a sponsorship deal into a shoe that did not suit them. That is a different problem, it has named casualties, and it is not fixed by anything getting cheaper.

7 So what does this mean for you

If you race road distances and you are chasing a time, the shoes are worth having, and it is worth being as concrete about the upside as about the caveats. On this article's own chain — a 2.75% saving in the lab, roughly two-thirds of it surviving into speed at fast paces — a three-hour marathon runner is looking at something in the order of two to three and a half minutes. That is not a rounding error. It is more than most people gain from a training block.

The honest caveat sits right next to it: that estimate is built from a lab measure and a model, and the one pooled analysis that timed actual trials found an effect about half as large when the two are put on the same footing. Treat two to three minutes as the optimistic end of a real range, not a promise.

If you are choosing between models, the evidence gives you less guidance than the marketing implies. The presence of carbon is not the thing to optimise for. Fit and comfort are measurable and personal; the plate material is neither.

If you run slower than about nine-minute miles, the honest position is that almost nobody has measured you. The one study that went down to genuinely recreational speeds found the benefit held. The modelling suggests the translation from economy to time may actually work slightly in your favour. Neither of those is the same as evidence in runners like you at race distances.

And if you are a woman, the gap is starker still: the biggest reported gains in race data are on your side of the sport, and the laboratory evidence behind them is the thinnest in the whole field.

The shoes are real. The famous number was an average of elite men, the plate works only in company, and much of what happens between a treadmill and a finish line has yet to be measured.

Sources

Sources we could read in full were read at the place they are cited from; where we could only reach an abstract, or no text at all, the entry or the article says so. Since 10 August 2026 every source we could reach is also archived in our own files, so the numbers on this page can be checked against the documents after publication. Research sources are marked T1; practitioner, manufacturer and product sources are marked T3/T4 — those establish what is claimed and sold, never that something works.

How we searched. This check leans on three recent meta-analyses rather than re-screening the primary literature ourselves — a legitimate shortcut, and one we should have stated here from the start. On 13 August 2026 we ran our own search anyway, to see what those reviews could not have covered: PubMed, five strings across economy, performance, mechanism, injury and durability, 2016 onwards (the year the first of these shoes appeared). 250 records after de-duplication, each screened on title and abstract — 246 had an abstract, four did not and were judged on the title alone. Of those, 13 are sources on this page, 66 fall inside the reviews' own search window, and 25 are newer than the reviews — 17 of those we could read in full, and five of them are now in the text above. Nothing found so far contradicts what stands here; if that changes, it goes on the corrections page. Databases we cannot reach — SPORTDiscus, Web of Science, Scopus — were not searched, and we say that rather than calling them unavailable.

1. T1 Kobayashi EN, de Toledo RRF, de Almeida MO, Sprey JWC, Jorge PB. Metabolic effects of carbon-plated running shoes: a systematic review and meta-analysis. Front Sports Act Living. 2026;7:1710224. doi:10.3389/fspor.2025.1710224
2. T1 Xiao Y, Hu X, Tian D, Qiu A. Effects of Advanced Footwear Technology on Running Economy and Endurance Performance: A Meta-Analysis. Int J Sports Med. 2026;47(2):81–94. doi:10.1055/a-2637-7283
3. T1 Stephen CHN, Kelly LA, Schuster RW, Diamond LE. The effects of running shoe longitudinal bending stiffness and midsole energy return on oxygen consumption and ankle mechanics and energetics: a systematic review and meta-analysis. J Sport Health Sci. 2025;14:101069. doi:10.1016/j.jshs.2025.101069
4. T1 Hoogkamer W, Kipp S, Frank JH, Farina EM, Luo G, Kram R. A Comparison of the Energetic Cost of Running in Marathon Racing Shoes. Sports Med. 2018;48(4):1009–1019. doi:10.1007/s40279-017-0811-2
5. T1 Kipp S, Kram R, Hoogkamer W. Extrapolating Metabolic Savings in Running: Implications for Performance Predictions. Front Physiol. 2019;10:79. doi:10.3389/fphys.2019.00079
6. T1 Muniz-Pardos B, Angeloudis K, Zelenkova I, et al. Advanced footwear technology in well-trained athletes: methodological insights from outdoor running. Front Physiol. 2025;16:1713902. doi:10.3389/fphys.2025.1713902
7. T1 Bolliger A, Spengler CM, Beltrami FG. Impact of Advanced Footwear Technology on Running Economy at Slower Running Speeds. Sports Med Open. 2026;12:12. doi:10.1186/s40798-026-00977-3
8. T1 Wu Y, Zhang H, Wang S, et al. Comparative analysis of foam-only versus carbon-plated advanced footwear technology spikes in distance runners. Front Physiol. 2025;16:1703854. doi:10.3389/fphys.2025.1703854
9. T1 Tenforde A, Hoenig T, Saxena A, Hollander K. Bone stress injuries in runners using carbon fiber plate footwear. Sports Med. 2023;53(8):1499–1505. doi:10.1007/s40279-023-01818-z
10. T1 Maruo Y, Takezawa K. The impact of advanced footwear technology on running performance and pacing in world marathon majors. Front Physiol. 2026;17:1800107. doi:10.3389/fphys.2026.1800107
11. T1 Rodrigo-Carranza V, Hoogkamer W, González-Ravé JM, et al. Influence of different midsole foam in advanced footwear technology use on running economy and biomechanics in trained runners. Scand J Med Sci Sports. 2024;34(1):e14526. doi:10.1111/sms.14526
12. T1 Giachetti Martin S, Kobayashi EN, et al. Carbon plates in running shoes biomechanics: a systematic review and meta-analysis. Front Sports Act Living. 2026;8:1764338. doi:10.3389/fspor.2026.1764338
13. T1 Kim H, Ahn J. Technologically advanced running shoes reduce biomechanical factors of running related injury risk. Sci Rep. 2025. doi:10.1038/s41598-025-03029-0
14. T1 Bruneau MM, Gaudette LW, et al. Biomechanics associated with bone stress injuries while using advanced footwear technology in elite distance runners. PM R. 2026;18(S2):S143–S150. doi:10.1002/pmrj.70153 — shoes donated by a manufacturer; noted in the text.
15. T2 Brüggemann G-P, Hirschhaeuser E, Rehorst S. How conventional or technologically advanced running shoes influence the risk and location of injuries: a prospective interventional study of 414 amateur runners. SportRxiv preprint, 25 June 2025 — not peer-reviewed.
16. T2 Injury-incidence work presented in Footwear Science 2025 supplements (a 195-runner randomised comparison with a manufacturer co-author, and a 271-runner Stockholm cohort). Conference-proceedings tier; full texts were not reachable to us.
17. T3 World Athletics. Book C — C2.1A Athletic Shoe Regulations. Version approved by Council on 19 March 2026, in force from 20 April 2026 — re-read and archived by us on 10 August 2026. Stack-height limits: 40 mm road, 20 mm track.
18. T1 Bermon S, Garrandes F, Szabo A, Berkovics I, Adami PE. Effect of Advanced Shoe Technology on the Evolution of Road Race Times in Male and Female Elite Runners. Front Sports Act Living. 2021;3:653173. doi:10.3389/fspor.2021.653173 — the governing body's own scientists: race times down 0.7–1.2% (men) and 1.7–2.0% (women).
19. T1 Knopp M, Muñiz-Pardos B, Wackerhage H, et al. Variability in Running Economy of Kenyan World-Class and European Amateur Male Runners with Advanced Footwear Running Technology. Sports Med. 2023;53(6):1255–1271. doi:10.1007/s40279-023-01816-1 — individual responses from +9.7% to −1.1% (amateurs) and +11.4% to −11.3% (world-class).
20. T3 World Athletics press release, 31 January 2020 — the original four-month availability requirement, since removed.
21. T3 ASICS corporate press release, 13 February 2026 — launch of the Superblast 3. Marketed on foam and rebound; no plate mentioned.
22. T4 Doctors of Running — independent reviews and category definition, by licensed physiotherapists. Discloses receiving product free of charge.
23. T4 Practitioner commentary on rotation and transition (coaches and physiotherapists, 2025–2026). Several of the outlets carrying this advice sell shoes or coaching; noted in the text where it matters.
24. T1 Askew GN, Hill S, Payne M, Stewart TD, Charles D, Brown AP. Running Economy Benefits of Shoes Incorporating Advanced Footwear Technology Decrease With Increasing Incline and Are Negligible Above Moderate Gradients. Scand J Med Sci Sports. 2026;36:e70298. doi:10.1111/sms.70298 — n=12 men, four gradients; the level-ground figure is 4.22%.
25. T1 Sex and Isolated Anthropometric Measures Do Not Explain Individual Differences in Responsiveness to Advanced Footwear Technology in Highly Trained Runners. Scand J Med Sci Sports. 2026;36:e70234. doi:10.1111/sms.70234 — 15 women and 15 men; no sex difference in the benefit.
26. T1 Petrella D, Swinnen W, Tam N, Lane B, Vanwanseele B. The Isolated Effect of Midsole Compliance on Running Economy and Biomechanics in Highly Trained Runners. Med Sci Sports Exerc. 2026. doi:10.1249/MSS.0000000000003864 — compliance alone: +3.90% economy, via the knee. One author is employed by a shoe manufacturer (On); noted here for the same reason we note it elsewhere.
27. T1 The running economy and biomechanics of competitive distance runners wearing advanced footwear technology spikes. Eur J Appl Physiol. 2026;126(8):4507–4519. doi:10.1007/s00421-026-06257-y — n=17 (7 women); AFT spikes 2.1% and 2.3% better than traditional spikes.
28. T1 Economy benefits of running in advanced footwear technology shoes remain with plantarflexion fatigue. Eur J Appl Physiol. 2026;126(7):3843–3857. doi:10.1007/s00421-026-06190-0 — neither calf fatigue nor push-off power explains who responds.
29. T1 Biomechanical Factors Associated with Intraindividual Differences in Running Economy Across Advanced Footwear Technology. Scand J Med Sci Sports. 2026. PMID 41860638 — the AFT models that shortened an individual's ground contact time were the ones that improved their economy; no model was best for everyone.
30. T2 Influence of carbon-plated running shoes and fatigue on lower limb biomechanics. J Biomech. 2026;113102. doi:10.1016/j.jbiomech.2025.113102 — n=16; the authors suggest stiffness may shift load proximally once fatigued. Read in the SSRN preprint version; conclusion is hedged.
31. T3 Smith C. “We Ran in Adidas's $500 Super Shoe Until It Gave Out.” Trail Runner / RUN (Outside), 1 May 2024 — the long-mileage Evo 1 test and the Patrick Nava interview quoted in §4.
32. T3 runABC, March 2026 — launch coverage of the Lidl Crivit CarbonLite 1.0 (€49.99, on sale from 9 March 2026; £39.99 in the UK from 29 March).

Found a mistake in this piece? Tell us at info@enduranceproof.com — you do not need to be a scientist, and “this number looks off” is a perfectly good message. Anything we correct, and when, goes on our corrections page.

This is educational material about published evidence, not training or medical advice. Nothing here is a recommendation about what any individual should do.