2011 Boston Marathon. Ryan Hall was running 4:47/mile through halfway — on pace to break 2:06. When he reached the Newton Hills at km 32, everything changed. His pace dropped from 4:50/mile to over 6:00/mile — a 25% collapse. His arms dropped, his shoulders rounded, his metronome cadence became first labored, then desperate. Hall finished fourth in 2:04:58 — but the face he wore over the final 10 km wasn’t that of someone digging deep, it was that of someone who had crossed a metabolic threshold from which there is no return. What marathoners call “hitting the wall” is not a metaphor. It is a metabolic crisis that plays out in real time, in a real body.
Table of Contents
- The Energy Arithmetic: Why the Marathon Is an Impossible Equation
- The Three-Pool Model: Where Does Glycogen Actually Run Out?
- The Multi-System Collapse: Why the Wall Is Sudden, Not Gradual
- The Fat-Store Paradox: Why Don't 100,000 Calories Solve It?
- Carbohydrate Loading: From Bergström to Modern Protocols
- The Three Shifts of the ATP Factory: Energy Systems
- In-Race Nutrition: Delaying the Wall
- Dean Karnazes and the Limits of Fatigue
- The Central Governor: Does the Brain Pull the Brake Before the Body Collapses?
- Conclusion: The Wall Is Not Inevitable. It Is Manageable.
Figure: Glycogen stores deplete over the course of a marathon. In-race carbohydrate intake slows depletion and pushes the “wall” further out.
The Energy Arithmetic: Why the Marathon Is an Impossible Equation
The metabolic arithmetic of the marathon is brutal. An elite male runner burns approximately 2,600 kcal over 42.195 km. Maximum glycogen stores — even with carbohydrate loading — hold only 1,500–2,000 kcal. There is a deficit of 600–1,000 kcal in between, and that deficit must be covered by fat oxidation. The problem is this: marathon race pace sits at 80–85% of VO₂max; at this intensity, carbohydrate supplies 85–90% of aerobic fuel. Fat oxidation is capped at 1.0–1.5 grams per minute, even in the most fat-adapted athletes — not enough to sustain race pace on its own [1]. The machine doesn’t stop until the fuel runs out; but when the type of fuel powering the machine changes, performance collapses.
George Brooks’s crossover concept explains this balance. At low intensities (25–30% VO₂max), fat provides 60–80% of total energy. As intensity rises, carbohydrate’s share climbs; at ~65% VO₂max the two are equal. Above 80%, carbohydrate reaches 85–90%. According to Van Loon and colleagues, fat oxidation peaks at 55% of maximal work rate (~32 kJ/min) and begins declining at 75% (~19 kJ/min). The harder you run, the less fat you burn — counterintuitive, but physiologically inevitable.
The Three-Pool Model: Where Does Glycogen Actually Run Out?
For years the assumption was simple: glycogen runs out, the athlete collapses. A fuel-gauge model — tank empties, engine stops. Ørtenblad and colleagues’ 2011 electron microscopy study shattered that model [2]. Muscle glycogen is stored in three separate pools:
Subsarcolemmal pool: Below the cell membrane; fuels the ion pumps (Na⁺/K⁺-ATPase).
Intermyofibrillar pool: Between contractile filaments; fuels sarcoplasmic reticulum calcium pumps and cross-bridge cycling.
Intramyofibrillar pool: Inside the myofibrils, embedded between actin and myosin filaments. The smallest pool — but the most critical for contractile function. It is physically adjacent to the calcium release channels (ryanodine receptors) of the sarcoplasmic reticulum.
Ørtenblad’s finding was this: athletes who hit the wall still had 40–60% of total muscle glycogen remaining. But the intramyofibrillar pool — the smallest, fastest-depleting one — was completely empty. Local depletion impaired calcium release from the sarcoplasmic reticulum even though total glycogen was adequate. The fuel gauge read half-full; but the fuel in the combustion chamber was gone. What causes the wall is not quantity. It is location.
The practical implication is clear: loading strategies that maximize the intramyofibrillar pool, and nutrition protocols that slow depletion of this specific pool, may outperform approaches focused only on raising total glycogen. Location matters more than amount.
The metabolic arithmetic of the marathon is fixed. Glycogen stores hold approximately 1,500–2,000 kcal. The marathon requires approximately 2,600 kcal. The only variables the athlete controls are how full the tank is at the start and how aggressively they fuel during the race.
<em>The Science of Human Endurance, Chapter 5</em>The Multi-System Collapse: Why the Wall Is Sudden, Not Gradual
The wall is not a slow decline — it is a qualitative transition. Multiple failure mechanisms engage simultaneously:
At the muscle level: Critically low intramyofibrillar glycogen impairs calcium release before ATP supply becomes rate-limiting. Excitation-contraction (E-C) coupling fails — the muscle receives the command but cannot contract.
At the systemic level: As liver glycogen depletes, blood glucose falls. The brain, almost entirely glucose-dependent, generates central fatigue signals: motor drive drops, perceived effort rises, cognitive function impairs, the “heavy legs” sensation deepens. Falling blood glucose also reduces core temperature regulation capacity and pain tolerance.
Performance collapses sharply when muscle glycogen falls below approximately 250 mmol/kg dry weight — the point at which most Type I and Type II fibers approach critical depletion. In athletes not taking in fuel, this threshold is typically crossed at km 30–35 (mile 20).
The Fat-Store Paradox: Why Don’t 100,000 Calories Solve It?
Even a lean endurance athlete carries 60,000–120,000 kcal of fat — 40–60 times glycogen. So why doesn’t this enormous energy reserve solve the marathon dilemma? The answer is the rate limit on oxidation. Fat oxidation, as Van Loon and colleagues demonstrated, is capped at 1.0–1.5 g/min in trained athletes (500–700 kcal/hour). In athletes “fat-adapted” through 5–10 days of high-fat dieting, this can reach 1.5–1.8 g/min; in elite ultra-endurance veterans with a decade of background, 1.5–2.0 g/min. But fat oxidation drops at high intensity — because rising calcium activates pyruvate dehydrogenase (PDH), directing pyruvate into the Krebs cycle, and falling free carnitine plus accumulating malonyl-CoA inhibit CPT1, shutting fat oxidation off on the supply side. At marathon race pace (80–85% VO₂max), the body prefers carbohydrate — no amount of fat adaptation can break this physiological constraint.
The sex difference becomes visible here. According to Tarnopolsky and colleagues’ 1990 work, women — via estrogen-mediated increases in fat mobilization and oxidation — burn more fat than men at the same relative intensity. The crossover point occurs at a higher intensity in women. The result: women deplete glycogen more slowly, hold a more even marathon pace, and the male-female gap narrows dramatically over ultra distances.
Carbohydrate Loading: From Bergström to Modern Protocols
In 1966–67 Swedish physiologists Jonas Bergström and Eric Hultman laid the foundation of muscle-glycogen research with the needle biopsy technique [3]. The key experiment: one leg was exercised to complete glycogen depletion, the other rested. Both received three days of high-carbohydrate diet. The depleted leg stored up to twice the glycogen of the rested leg — supercompensation. Mechanism: upregulation of the glycogen synthase enzyme in previously depleted fibers.
The modern protocol is simpler: 8–10 g of carbohydrate per kg of body weight for 2–3 days before the race — no depletion phase required. Muscle glycogen rises 15–25% above normal training levels. Some studies have documented 50–100% increases. Untrained individuals store 300–400 g, trained endurance athletes 400–500 g, and post-loading athletes 600–700 g or more. Liver glycogen ranges from 80 to 120 g.
The Three Shifts of the ATP Factory: Energy Systems
To understand the wall, you must first understand the energy production mechanisms. The body uses three energy systems simultaneously — parallel, not sequential. The ATP-PCr (phosphagen) system is the near-instant fuel source of the first 5–10 seconds: creatine phosphate transfers a phosphate to ADP, regenerating ATP. Total capacity ~500 kJ — 6–10 seconds of maximal sprinting. The body’s total ATP store is only ~250 g, enough for 3–8 seconds of maximal exercise. Anaerobic glycolysis is dominant from ~10 seconds to 2 minutes; it produces 2–3 ATP per glucose molecule, but hydrogen ion accumulation (acidosis) quickly limits performance. A 400 m sprint (~45–50 seconds at elite level) relies primarily on this system.
Aerobic oxidative phosphorylation begins contributing within 10–15 seconds but takes 2–3 minutes to reach full capacity. It yields ~30–32 net ATP per glucose, and ~129 ATP per palmitate (16-carbon fatty acid) — fat is 4× more energy-efficient than carbohydrate, but slower to oxidize. The energy yield per unit oxygen is ~2.73 ATP for carbohydrate and ~2.33 ATP for fat — carbohydrate delivers more energy per liter of oxygen. At high intensity, with oxygen limited, the body prefers carbohydrate. This is the marathon’s dilemma: the most efficient fuel is also the scarcest.
In-Race Nutrition: Delaying the Wall
The physiology of gut absorption sets the limits on in-race feeding. Glucose and maltodextrin are absorbed via the SGLT1 transporter at a maximum of ~60 g/hour. Fructose moves through the independent GLUT5 transporter. A 2:1 glucose-fructose mix saturates both transporters, raising total absorption to 80–90 g/hour — a 40–50% increase over glucose alone. Aggressive in-race fueling (60–90 g/hour of mixed carbohydrate starting in the first 30 minutes) can delay glycogen depletion by 30–45 minutes — pushing the wall at km 32 past km 38 [4].
Tarnopolsky and colleagues’ sex comparison adds an important nuance: women burn proportionally more fat and less carbohydrate than men at the same relative intensity. Estrogen-driven enhancement of fat mobilization and oxidation slows glycogen consumption in female runners. This is one of the physiological foundations of women’s more even ultra-endurance pacing — and of the lower second-half/first-half split ratio.
Dean Karnazes and the Limits of Fatigue
October 2005, California Pacific Coast Highway. Dean Karnazes ran 350 miles (563 km) without sleep — more than 80 hours. He consumed an estimated 8,000–10,000 calories per day: burritos, cheesecake, pizza, sports drinks. His glycogen depleted and partially restored dozens of times. He had hallucinations, heard voices, his speech slowed, his coordination broke down. By every fatigue model of 2005 he should have stopped 200 miles earlier. The explanation for why he didn’t was neither in his muscles nor his metabolism — it was that his brain had recalibrated the relationship between afferent fatigue signals and the will to sustain motor output. For Karnazes, fatigue was no longer a wall to be broken — it had become a geography he had learned to live inside.
The Central Governor: Does the Brain Pull the Brake Before the Body Collapses?
Tim Noakes’s Central Governor model, proposed in 2001, takes a radical view of fatigue: the brain continuously monitors signals from muscles, the cardiovascular system, and temperature sensors, and proactively reduces motor output — to prevent catastrophic failure, before it can occur. The claim is bold: skeletal muscle is never truly “depleted” during prolonged exercise; reserves always remain.
The evidence is convincing. The marathon-sprint paradox: a runner who has hit the wall can sprint the final 400 m — so true muscle exhaustion cannot be the cause. Deception studies: athletes told they are “almost done” at the halfway point of a 5 km maintain or accelerate; when the deception is revealed, they slow down. Samuele Marcora’s 2009 experiment is complementary: cyclists who completed 90 minutes of cognitive work exhausted 15% earlier with zero change in physiological parameters (VO₂, heart rate, blood lactate). The only thing different was perceived effort [5].
Caffeine works through this mechanism: as an adenosine receptor antagonist, it reduces perceived effort. At 3–6 mg/kg, dozens of studies have documented 2–5% performance gains. Motivational self-talk (“I can do this,” “relax”) has increased time-to-exhaustion by up to 18% in controlled studies — with zero change in physiology.
Conclusion: The Wall Is Not Inevitable. It Is Manageable.
Ryan Hall’s Boston collapse was a failure of fuel management — not of talent. Had glycogen stores been maximized before the race (3 days at 8–10 g/kg carbohydrate), had he taken 60–90 g/hour of mixed carbohydrate from the first 30 minutes, the depletion point could have shifted from km 32 to beyond km 38. As Hall’s physiologists put it: “What stopped the machine wasn’t the machine. It was the fuel line.”
The wall is not the alarm of an empty tank. It is the local depletion of a specific glycogen pool — the intramyofibrillar pool at the heart of the contractile machinery. Total fuel can be adequate while location is critical. You cannot change the metabolic arithmetic of the marathon; but you can control the numbers you enter into the equation — how full you start and how aggressively you refuel.
Recovery physiology after glycogen depletion matters just as much. Duhamel and colleagues’ experiment showed that carbohydrate feeding after exhaustive exercise restored calcium release rates and power output within 4 hours; in the water-only group, calcium release remained depressed for 24+ hours. This is direct evidence that the glycogen-calcium link is causal. The first 2 hours after exercise are the fastest phase of glycogen replenishment — 5–10% of pre-exercise stores can be replaced per hour, provided 1.0–1.2 g/kg/hour of carbohydrate is consumed. Full recovery takes 20–24 hours (with a total intake of 7–10 g/kg). In half-time and full-time muscle biopsies of soccer players, more than 75% of both slow- and fast-twitch fibers showed very low glycogen — matching the patterns of late-game pace drop, sprint loss, and defensive positioning errors.
The wall is not inevitable. It is manageable.
Sources: – Karlsson, J., & Saltin, B. (1971). Diet, muscle glycogen, and endurance performance. Journal of Applied Physiology. – Costill, D. L., & Hargreaves, M. (1992). Carbohydrate nutrition and fatigue. Sports Medicine. – Rapoport, B. I. (2010). Metabolic factors limiting performance in marathon runners. PLoS Computational Biology. – Coyle, E. F. (2007). Physiological regulation of marathon performance. Sports Medicine.
Don't miss the latest in sports science
Join our reader community for evidence-based sports science insights, training tips, and early access to new content.
References
- Jeukendrup, A. (2011). Nutrition for endurance sports. Journal of Sports Sciences. ↩
- Ørtenblad, N. et al. (2011). Muscle glycogen stores and fatigue. The Journal of Physiology. ↩
- Bergström, J. & Hultman, E. (1966). Muscle glycogen synthesis after exercise. Nature. ↩
- Jeukendrup, A. (2011). Nutrition for endurance sports: marathon, triathlon, and road cycling. Journal of Sports Sciences. ↩
- Marcora, S., Staiano, W. & Manning, V. (2009). Mental fatigue impairs physical performance in humans. Journal of Applied Physiology. ↩
The Energy Arithmetic: Why the Marathon Is an Impossible Equation
The metabolic arithmetic of the marathon is brutal. An elite male runner burns approximately 2,600 kcal over 42.195 km. Maximum glycogen stores — even with carbohydrate loading — hold only 1,500–2,000 kcal. There is a deficit of 600–1,000 kcal in between, and that deficit must…
The Three-Pool Model: Where Does Glycogen Actually Run Out?
For years the assumption was simple: glycogen runs out, the athlete collapses. A fuel-gauge model — tank empties, engine stops. Ørtenblad and colleagues' 2011 electron microscopy study shattered that model [kaynak id="2" author="Ørtenblad, N. et al." year="2011" title="Muscle glycogen stores and fatigue" journal="The Journal of…
The Multi-System Collapse: Why the Wall Is Sudden, Not Gradual
The wall is not a slow decline — it is a qualitative transition. Multiple failure mechanisms engage simultaneously:
The Fat-Store Paradox: Why Don't 100,000 Calories Solve It?
Even a lean endurance athlete carries 60,000–120,000 kcal of fat — 40–60 times glycogen. So why doesn't this enormous energy reserve solve the marathon dilemma? The answer is the rate limit on oxidation. Fat oxidation, as Van Loon and colleagues demonstrated, is capped at 1.0–1.5…
Carbohydrate Loading: From Bergström to Modern Protocols
In 1966–67 Swedish physiologists Jonas Bergström and Eric Hultman laid the foundation of muscle-glycogen research with the needle biopsy technique [kaynak id="3" author="Bergström, J. & Hultman, E." year="1966" title="Muscle glycogen synthesis after exercise" journal="Nature"]. The key experiment: one leg was exercised to complete glycogen depletion,…
