The blow-up after 30km is almost always filed away as a sugar problem. It’s a convenient reading, and partly true, but incomplete. A recent review puts fuel back in its rightful place: essential, but not the sole culprit.
The wall is a crossroads, not a switch
The review covered by Marathon Handbook isn’t a new experiment: it pulls together the existing literature on glycogen depletion, heat and dehydration, gastrointestinal issues, neuromuscular fatigue, deteriorating running mechanics, pacing errors and perceived exertion.
The concept that holds the whole picture together is durability: the ability to resist physiological and performance decline as fatigue accumulates. In this model, low glycogen doesn’t act alone. Heat raises the cardiovascular demand. Dehydration amplifies it. A protesting stomach limits how much you can take on board. Neuromuscular fatigue makes every stride less economical. And an aggressive start speeds all of it up.
Between 30km and 35km — what the authors call the critical zone — these pressures converge and push perceived effort beyond what you can sustain. It isn’t a single system shutting down. It’s the sum that becomes unmanageable.
Fuelling remains the foundation: carb loading in the days beforehand, taking on fuel early and regularly during the race, roughly 60–90 grams an hour using multiple types of carbohydrate. Intakes closer to 120 grams an hour can work for very well-trained athletes, but only after specific gut training. These are reference figures, to be tested and adapted in your long runs, not copied wholesale.
The real point is something else: a perfect nutrition plan won’t make up for bad pacing, legs unprepared for eccentric loading, poor durability in the long runs, or a stomach that has never tried those gels at that pace.
The legs give out before the tank does
There’s one component runners almost always underestimate: accumulated muscle damage. In a marathon it isn’t only a lack of substrate that slows you down, it’s a quadriceps that no longer responds the way it did at 10km.
A recent controlled experiment illustrates this well, albeit under extreme conditions. Eight sedentary or moderately active men ran for 60 minutes on a treadmill at a gradient of –14%, at 70–85% of their theoretical maximum heart rate. Seven acted as a control group with 30 minutes of level walking.
- In the downhill group, force production capacity fell from 214.5 to 160.2 newton metres immediately afterwards: around a quarter less.
- Two days later, recovery was only partial: 177.1 newton metres, still below baseline.
- The control group showed no meaningful changes.
- The vastus lateralis suffered a greater deficit than the vastus medialis, with more soreness at 48 hours.
The sample is small and the subjects weren’t trained runners, so we’re not transferring the numbers to your legs. We’re transferring the concept: soreness is not a good indicator of readiness. Strength can remain compromised once the pain has gone, or nearly so.
Apply that to the marathon. Even a course described as “fast” has downhill sections, and in the early hours you run those descents on fresh legs, often harder than you’d planned. The bill arrives later, when the quads still have to brake on every footstrike but no longer have the torque available. Stride length shortens, energy cost rises, and the pace slips even though the fuel is there.
Cardiac drift: same pace, more beats
The third piece of the puzzle, and the one you can watch unfold in real time on your watch. Sixteen kilometres into a long run, flat road, identical pace to half an hour earlier, and your heart rate has gone from 144 to 152. You’re not running harder. It’s costing you more.
The most intuitive mechanism is haemodynamic: the heat generated by running is dissipated by diverting part of the blood flow to the skin, and if you’re sweating heavily while replacing little, plasma volume drops too. Less blood reaches the heart to be ejected with each contraction, and the difference is made up by increasing the rate. There’s also an alternative hypothesis involving sympathetic activation, but for runners it makes little difference: the phenomenon is consistent and predictable.
Two situations need separating. Physiological drift is slow and evenly spread — a few beats every twenty or thirty minutes — with pace and breathing holding steady: you can ignore it. It becomes a warning sign when it arrives early, climbs steeply, and above all when perceived effort follows it and the pace starts to slip without you having decided to ease off.
There are three typical causes: going out too fast, a fluid deficit, and environmental conditions. A fourth, less common one, is unabsorbed training load or an infection brewing — and in that case the heart rate is already high from the first kilometre.
The practical takeaway for the long run: steer by pace and breathing, and treat heart rate as a ceiling, not a target. Slowing down to get back into the zone means finishing the session at an intensity that no longer trains much. The pace of those first kilometres should be chosen with an eye on how you’ll feel at 25km, not on how you feel at 3km.
After the session, a useful check: split the file into two equal halves and compare the ratio between average pace and average heart rate. A 5% shift circulates among coaches as a practical threshold — it’s a field convention, not a validated value — and the comparison only holds between sessions that are similar in route, temperature and time of day.
How to train the last twelve kilometres
If the wall is the sum of several failures, the long run needs to be built to train each one. Here are a few lines of work, with indicative volumes to be adapted to your background and weekly mileage.
- Progressive long run. Final 8–12km faster than the middle section, never starting hard. You train the ability to hold on when mechanics start to deteriorate.
- Marathon pace under fatigue. Not race-pace blocks on fresh legs, but 15–20km easy followed by 8–10km at marathon pace. That’s where durability is built.
- Gradual eccentric loading. Short, controlled descents, introduced weeks before the race, never right before it. Add lower-limb strength work, which protects the quads better than any gel.
- Genuine fuelling rehearsals. Gels and electrolytes in the long run, at the same frequency and the same pace you’ll use on race day. The stomach trains like everything else.
- Rehearsed pacing. Race pace is tested in the long runs, not decided at kilometre one.
And if the course has an undulating profile, study it in advance: runners preparing for a marathon abroad with us at KiRun work on the specific descents of that route during the middle weeks, not in the final month.
In practice
- Treat the wall as a multiple problem: fuel, legs, heat, head, pacing. Fixing just one isn’t enough.
- Rehearse your real fuelling strategy in your long runs, aiming for 60–90g of carbohydrate an hour and adjusting for tolerance and duration.
- Introduce downhills progressively and well away from race day: strength can still be reduced 48 hours later, once the soreness has gone.
- In the long run, set your pace by feel and breathing, with heart rate as a guardrail. Eight or ten extra beats in the second hour are normal.
- At the end of the month, repeat the same long run in the same conditions: if the decoupling between pace and heart rate shrinks, the adaptation is coming.
Sources: Marathon Handbook — The Marathon Wall Is More Than A Fueling Problem; Marathon Handbook — The Hidden Cost Of A Single Downhill Run; RunLovers — Deriva cardiaca: la frequenza sale a passo costante.
