REF-823 · Rapoport, 2010 · PLoS Computational Biology, Vol. · Paper · PubMed

Metabolic Factors Limiting Performance in Marathon Runners

Cweak evidence: opinion or lab dataMeasured on a mixed groupWritten before October 2026, being re-checked

A computational study of the energetic limits on marathon runners. The author notes that more than two-fifths of those who attempt a marathon experience severe, performance-limiting depletion of carbohydrate reserves ("hitting the wall"), and that approximately 1-2% of starters drop out.

Rather than glycogen being simply too small to fuel a marathon, or maximal loading being required to finish, the model shows that the constraint depends on several individual variables — muscle mass distribution, liver and muscle glycogen densities, and running speed as a fraction of aerobic capacity — in personalised but quantifiable ways.

It estimates the distance at which a runner will exhaust glycogen as a function of running intensity, and offers a basis for setting personally appropriate paces and prescribing mid-race fuelling to avoid the wall. It also applies the analysis to the Boston Marathon qualifying times.

Participants

None (computational model). The prevalence figures in the abstract (more than two-fifths hitting the wall; approximately 1-2% of starters dropping out) are the author's framing, not data collected in this study.

Protocol

Analytic model of carbohydrate stores and their use during marathon running (42.195 kilometers).
Inputs: muscle mass distribution, liver and muscle glycogen densities, and running speed expressed as exercise intensity (a fraction of aerobic capacity).
Output: the distance at which glycogen stores are exhausted, as a function of running intensity; applied to pacing, mid-race fuelling and the Boston Marathon qualifying times.

Key Results

The wall is not inevitable and not universal: whether and where a runner exhausts glycogen depends on individual physiology and on running intensity.
Neither extreme view holds in the model: glycogen reserves are not simply insufficient for a marathon, and maximal glycogen loading is not simply required to finish.
The model gives the distance to glycogen exhaustion as a function of intensity; the abstract does not report the numbers.
The abstract gives no validation of the model's predictions against real race data.

Practical Application

This model of marathon energetics (no participants) predicts that the wall depends on how fast you run as a fraction of your aerobic capacity, how much glycogen your muscles and liver hold and how your muscle mass is distributed — so pace, loading and in-race fuel can each move it, and a pace that is safe for one runner empties another's stores. If you hit the wall in your last marathon, the first fix is a slightly lower pace relative to your fitness, then more carbohydrate on board before and during the race.
The abstract's framing, not measured here: more than two-fifths of marathon starters experience the wall. Race-record data in REF-775 put the figure lower, so treat this as a high estimate.
Not from this paper; the base's own practice: fuel during the race from the start rather than waiting for the wall (SYNTH-02), and load without a depletion phase (REF-774, REF-819).
The model is a C-level explanation; use it for the "why", not as a calculator, since the abstract reports no validation.

Limitations

Mathematical model; no validation against measured glycogen or race outcomes is reported in the abstract.
Individual inputs (muscle mass distribution, glycogen densities) are not routinely measurable for an athlete.
The abstract gives no numerical outputs (distances, paces, fuelling rates).
The prevalence figures are background framing, not results of this study.
Written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

unvalidated modelling: the abstract reports no test of the model's predictions against measured runners