A clock in Vienna once stopped at 1:59:40, and the mark went into no record book.
About six and a half years later, a runner crossed a London finish line in 1 hour, 59 minutes and 30 seconds, in an ordinary race under ordinary rules.
Sabastian Sawe became the first person to run a marathon under two hours in a competition that counts.
The barrier had long been described as beyond human physiology.
So what let the body hold together when the models said it should fall apart?
The fuel problem the body was never supposed to solve
At the most basic level, a marathon is a test of energy supply. Running the full distance at sub two hour pace costs roughly 2,500 kilocalories, while the body’s stored glycogen typically supplies something closer to 2,000. That shortfall is the gap physiologists pointed to whenever the barrier was called impossible.
But the body is not a simple combustion engine, and the shortfall can be narrowed from both ends. Modern elite fueling pushes carbohydrate intake during the race far beyond what was once thought tolerable, and the cheaper each stride becomes, the less there is to pay for. The question is less whether human physiology can support the pace than how much of that capacity survives the last hour.
Three things separate athletes who can hold this speed: a very high maximal oxygen uptake, the ability to run at a large fraction of it without accumulating lactate, and running economy well beyond the national class standard. Economy is the quiet variable, the one that decides how much oxygen each stride actually costs.
Get it lean enough and the fuel arithmetic changes entirely.
What the London course revealed about the body under pressure
Sawe’s race was shaped like nothing before it. Pacemakers took the lead group through halfway in 60:29, and after they stepped away Sawe ran the second half in just under 59 minutes and 01 seconds, per the splits, closing the final 1.4 miles in 5 minutes and 51 seconds. Most runners slow after 20 miles; he accelerated from about 19 miles to the line.
That shape points at how much aerobic capacity he still had available when the pace should have been eating into it. He was not alone out there either. Yomif Kejelcha, making his marathon debut, stayed in contact past 25 miles and finished second in 1:59:41, while Jacob Kiplimo took third in 2:00:28. All three beat the previous official record of 2:00:35.
The physiology of running in company is real: a pack cuts air resistance and lets an athlete hold a steadier metabolic rhythm. The barrier fell inside a collective performance environment rather than through an isolated effort.
Why the same time did not count six years earlier
In Vienna, a 1:59:40 finish was recorded in a staged exhibition built around a rotating formation of pacemakers, with drinks handed over from a bicycle and no open field to race against. Those conditions put the run outside the rules for ratification. The time was real. The race was not a race in the sense the record requires.
London used pacemakers too, but standard ones entered under normal rules and dropping out mid race rather than cycling in and out around the leader. That is the whole difference between a demonstration and a record.
There is a connection here that runs deeper than most readers expect. The limits on oxygen delivery and economy that govern elite marathon running also shape performance in freediving, where the body must stretch a single breath across minutes of effort. Both rest on the same cardiovascular architecture and the same oxygen economy problem.
The record that could not be built in a laboratory
Sawe trains at altitude in Kenya’s Nandi County, which sits at roughly 6,560 feet, and his marathon blocks under coach Claudio Berardelli regularly run past 120 miles a week. Permanent life at altitude is the norm among Kenya’s best marathoners, though researchers still argue over how much of the benefit comes from the elevation itself rather than the volume, the terrain and the group dynamic.
Every stride in London was built on that base. The carbon plate and resilient foam in a modern racing shoe measurably lower the metabolic cost of running, even if biomechanists still disagree about which part of the shoe does most of the work. Across 26.2 miles, a few percent adds up fast.
Technology only explains part of it. The more useful frame now is fatigue resistance and race execution: preserving economy, fuel availability and neuromuscular output deep into the final hour.
What the number leaves still unresolved
The record is real, but the ceiling is not visible. Projections of the eventual limit vary widely and none of them is settled science, which is roughly where the mile stood after four minutes fell.
Even the data needs care. One widely shared split from mile 24 was later withdrawn by London’s official timing provider after analysis indicated it had been mismeasured. The 5K splits stood, and the finishing time was never in doubt.
What altitude does to a climber and what pace does to a marathoner at mile 20 are different problems with the same engine underneath. That arithmetic sits at the heart of Everest summit physiology just as it sat in Sawe’s final mile.
Sawe’s laboratory profile has not been published, so the exact numbers behind the run remain out of view. What the clock recorded is certain; what the body did to produce it is still being worked out. That is not a weakness in the story. It is where the science goes next.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
