56 articles
The same long run: you have 32 km, your partner has 29. A look at validation studies of sport watches: up to 9% under-reading in the city and the forest, systematic over-reading on the track, a 12.1% error for the worst model over 4,000 m, and where the extra three kilometres come from when there are no satellites.
Since 1996 it has been assumed that a 1% incline makes the treadmill equivalent to the road. In 2026 that was checked three times over: on the track a kilometre costs 3.6% more oxygen even with the incline, and elite lactate thresholds on the treadmill and on the track differ by 7β8 s/km.
Hybrid races are sold as a test of strength. Laboratory measurement says the opposite: the only variable that predicts the result is VOβmax, while grip strength and muscle mass do not correlate with finishing time at all.
The largest analysis of marathon splits sorted runners into four pacing types. Among those who ran evenly, a collapse in pace in the second half occurred in 0.2% of cases. Among those who ran erratically β in 85.4%. The gap in finish time between the extreme groups is almost an hour.
Marathon runners ran for 90 minutes and then repeated their lab tests. Peak oxygen uptake fell from 56.7 to 53.4, and speed at lactate threshold from 12.8 to 12.1 km/h. This is durability β the quality your testing protocol never measures.
Six wristbands were tested in a sleep lab against PSG: sleeping epochs are recognised over 90% of the time, wakefulness only 29β52%. WHOOP overstates sleep time by 24 minutes, Apple Watch by 20. Here is what follows for your training.
Across 5311 athletes, maximum heart rate on the bike came out just 2 beats lower than in running. But you still can't carry threshold zones from one sport into the other β and here's why the group average is useless.
Greek runners shut down blood flow in their legs before every interval session. The average kilometre in a 5 Γ 1000 m test got 11 seconds faster against 5 in the control group β but the winner here looks like plasma volume, not haemoglobin.
A 2026 systematic review: handheld lactate meters correlate with the laboratory reference at r = 0.95β0.99. And yet the spread of a single measurement at 3 mmol/L is Β±0.4 mmol/L. Here is why that is not a contradiction.
Across 1,121 athletes, maximal fat oxidation occurred on average at 49.3% of VOβmax β with a range from 22.6 to 88.8%. Known factors explain only 47% of the differences. Here is what to do about it in practice.
The critical heat limit a person can tolerate moved only after nine days of adaptation β after five there was no shift at all. Plus one more finding: you need a gap between the block and race day.
Six weeks of leucine supplementation in cross-country skiers improved maximal oxygen uptake and strength. Here is how a single amino acid differs from Β«just eat more proteinΒ» β and who might actually benefit.
Rinsing the mouth with a menthol solution every six minutes raised mean power by 1.67% in the heat. The body does not actually cool β only perception changes. Here is where that helps and where it gets dangerous.
Flying across time zones costs roughly one day of adaptation per zone travelling east β and half that going west. We unpack where the asymmetry comes from and how to plan your arrival around it.
An ordinary optical sensor on the wrist cannot measure glucose β not because the algorithms are bad, but because it never collects the signal in the first place. We unpack the physics, a fresh re-analysis of the data, the FDA position, and what actually works.
A fresh RCT reported enormous effects β so enormous they are hard to believe in groups of ten people. We unpack what percussion therapy reliably delivers, what it doesn't, and when it's better left alone.
Runners who alternated between several pairs got injured noticeably less often. But there is only one study, it is more than ten years old, and it is observational. We unpack what you can take from it and what has been read into it.
A fresh study measured the cost of transport uphill before and after a 31-kilometre race. Poles reduced the load on the legs β but total energy expenditure doesn't drop, and by the end of the race the arms are just as tired.
The useful signal in an optical heart rate sensor is just 1β2% of the reflected light, and every foot strike attacks it from three directions at once. Here is why your heart rate βsticksβ at 180 during intervals β and why no firmware update will fix it.
A fresh meta-analysis in endurance athletes found a shift in favour of compression for muscle soreness, but the confidence interval crosses zero, the risk of bias is high in 10 assessments out of 11, and the effect on performance could not be pooled at all.
108 elite athletes kept a diary across 554 cycles. It turned out that retrospective questionnaires systematically exaggerate symptoms, and that not everything in training suffers β in football, high-speed work dropped off, while in cycling power did not change at all.
ACWR has become a standard metric in sports watches, but the meta-analysis is built on football, the confidence interval for the safe zone is enormous, and planning a season around ACWR did not reduce injuries at all.
A meta-analysis of 23 studies found a significant link with performance for just one pollutant β ozone. PM2.5 particles and the overall air quality index showed no reliable effect. Here is what that changes in practice.
Critical speed is not another heart-rate zone β it is the physiological line between βI can hold thisβ and βthis is about to endβ. Here is how to measure it with two efforts on the track, and what your Dβ² reserve means.
An eight-week study found a marked gain in power and mitochondrial adaptations from a ketone ester taken after training. Yet most of the literature finds no effect β here is why both pictures may be true.
A review of 30 studies found there is no single marker of overtraining, and the diagnosis is made by exclusion. Here is what is actually worth tracking, and when it is time to see a doctor rather than a coach.
A meta-analysis of 17 studies confirmed the superiority of the polarized model for VOβpeak β but only in blocks shorter than 12 weeks and only in well-trained athletes. In time trials there is no difference at all.
In a crossover RCT participants slept 55 minutes longer β and the next morning covered almost 10 metres more in a shuttle test with half the drop-off in speed. Here is what is genuinely proven and what is being read into it.
An ice vest, cold water, and ice slurry before the start create a heat reserve and push back overheating. We break down the 2024 meta-analysis and what an amateur can actually apply at a hot start.
Running power isn't the same thing as cycling power: it isn't measured directly but modeled. We break down where the device genuinely helps a runner and where it's credited with excess precision.
At record-setting events, glucose in elite athletes without diabetes drops to 3.0 and climbs to 11 mmol/L β and that is normal. Here is why an amateur might want a sensor, why it is not a βfuel gauge,β and where self-deception begins.
Sleep problems, exhaustion and joint pain are the most common companions of menopause among female runners, swimmers and cyclists. A new PLOS One study shows how hard the symptoms hit training β and what to do about it.
A cuff partially restricts blood flow, and loads of just 20β30% of your max deliver an effect close to heavy lifting. We break down where this genuinely helps a runner and why βcinching your leg with a beltβ is a bad idea.
Mental fatigue raises your perceived effort and makes you quit sooner β at the same heart rate. Let's unpack how to train your βbrain enduranceβ without burning out.
Breaking down LHTL without the myths: how long you really need to live up high, why it won't work without iron, and what an amateur β not an elite marathoner β can realistically do.
The diaphragm tires too and βstealsβ blood from your legs. We unpack how a handheld resistance trainer cuts breathlessness and perceived effort β and why it isnβt about a VO2max record, but about finishing without that βnot enough airβ feeling.
A fresh 2025 review: low cadence is linked to lower-leg injuries, while too long a stride loads the knees. We break down how to measure your step rate and raise it safely by 5β10%.
Zone 2 is work below your first threshold, not a number on your watch. We break down why heart-rate formulas get it wrong and how much easy running an amateur actually needs.
Cold-water immersion speeds your return to form between races but blunts strength and hypertrophy adaptations. Letβs figure out when ice works for you and when against.
Your morning RMSSD reflects how well you have recovered. Letβs look at what research says about training βby readinessβ and how to measure HRV without fooling yourself.
A good taper adds an average of 2β3% to your race result. We break down what the research says and lay out a ready-made 3-week plan for the marathon and half marathon.
Five weeks of post-run hot baths raised maximal oxygen uptake in trained runners by roughly 4,4%. We unpack how heat triggers the same blood and heart adaptations as altitude β and how to use it without the risk.
The 2026 meta-analyses confirmed it: on the road a carbon plate really does save 2β3% of energy. But all the science on βsuper shoesβ is built on treadmills and steady pacing β and on trails and climbs there is simply no evidence of any benefit.
A systematic review of 7 randomized trials compared running and cycling training. It found no difference in VOβmax gains or running performance β we break down what this means for triathletes and recovering runners.
Averaged βtrain strictly by phaseβ schemes are overrated. We break down what recent field studies show and why a personal approach and symptom tracking matter more than a universal mapping of training loads to cycle phases.
A new RCT in Frontiers in Nutrition: 4 weeks of a low-carb phase followed by a return to carbs improved running economy at the lactate threshold β something neither constant keto nor constantly high carbs achieved.
Two threshold sessions a day and a lactate meter in your pocket β we break down what in the βdouble thresholdβ is backed by science, what stayed pure hype, and how an amateur can try the method cautiously.
A new study of trail runners found that those who regularly run downhill repeats damage their muscles less and better preserve their squat strength. We break down what this means for your training.
A fresh 2026 network meta-analysis and the protocol of a new study suggest that heavy slow work and isometrics make the Achilles stiffer and thicker, while rest and stretching through pain are almost useless. Hereβs how to load the tendon the right way.
A fresh review in the journal Nutrients gathered 38 creatine studies in endurance sports. The verdict: it has little effect on βpureβ aerobic work, but genuinely helps sprints, strength work, and recovery. We break down who needs it and how to take it.
Ten weeks of strength and plyometric work barely changed βfreshβ running economy, but they preserved it through the end of a 90-minute run and boosted performance under fatigue. Here is why endurance runners need heavy squats and how to fit them in.
New research shows: 3 sets of a simple foot exercise significantly improve dynamic stability. We break down the science and proper technique.
Hydration is one of those topics where it's easy to mess up in both directions. Drink too little β you suffer. Drink too much β you suffer even more. Here's how to tell if something's off.
'Drink more water' is advice about as useful as 'run faster.' How much is more? A liter? Three? Let's calculate specifically.
Familiar situation: you're running your usual route, same pace, but heart rate is 15 beats higher. Legs feel heavy, breathing harder. Overtrained? Getting sick? Or is it simpler β and scarier?
How much do you really need to drink during training? Why do elite marathoners lose 10% of body weight and set records, while amateurs fear losing 2%? Let's dive into the science, myths, and formulas.