Durability: Endurance’s Fourth Dimension

For most of its history, endurance physiology has rested on three pillars: maximal oxygen uptake (VO₂max), the metabolic thresholds that sit beneath it, and running economy — how much energy a given pace costs. Measure those three fresh, in a lab, and you can predict performance reasonably well.

Except every distance runner knows the catch. The pace that felt controlled at kilometer five can feel impossible at kilometer thirty-five, at the same heart rate, with the same fuel. The lab numbers were real — they just described a version of you that no longer exists two hours in. What decayed is not captured by any single fresh measurement. Exercise physiologists have started calling that missing quality durability.

Durability is defined as the resistance to deterioration of physiological characteristics over prolonged exercise — how much your thresholds, economy and efficiency hold up as fatigue accumulates. Maunder and colleagues made the case that it is a distinct trait: two athletes can share the same fresh VO₂max and threshold and still differ substantially in how far those qualities drift late in a long effort.

That matters because the deciding kilometers of a marathon or an ultra happen in exactly the fatigued state a fresh lab test never sees. A runner who is 3% more durable doesn’t look better on paper on Monday. They look better at the finish line on Sunday.

Naming a construct is easier than measuring it in the wild, and this is where the honest picture gets complicated. Several candidate markers exist, and the evidence behind them genuinely conflicts.

The most intuitive is aerobic decoupling — the drift between pace (or power) and heart rate across a long run. Smyth and colleagues, working with a very large dataset of real marathons, found decoupling related to performance. Then Hunter and colleagues turned the lens on the method itself: how you measure decoupling — a single long run versus a multi-day protocol — changes what you find, and without careful standardization the number is easy to over-read. The signal is real; pinning it down reliably is not a solved problem. Anyone who tells you decoupling is a finished durability metric hasn’t read the second paper.

A second marker comes from heart rate variability: the within-session change in DFA-α1 across a long effort. In recreational runners, Nuuttila and colleagues found that this drift was the one signal that tracked the loss of lactate-threshold speed over a 90-minute run — a moderate relationship (r ≈ 0.46), promising but not decisive.

The most interesting recent thread looks at respiration. Rothschild and Maunder examined which signals track the loss of the first ventilatory threshold over prolonged cycling, and found that respiratory frequency — how fast you breathe, not how much — decoupled with the drift, where minute ventilation did not. In their data, combining heart rate, respiratory frequency and aerobic capacity explained the large majority of the variance in threshold drift.

That is a strong result, but it comes with the caveat every honest reader should apply: it was measured in cyclists, on power. Whether respiratory frequency carries the same weight for runners, on messy outdoor data, is exactly the kind of question that needs replication before anyone builds a verdict on it.

The reason durability is worth taking seriously is that it changes the shape of the problem. Fitness, in the classic sense, tells you what you can do rested. Durability tells you how much of that survives contact with a long race — and it responds to different training than raw fitness does.

But taking it seriously and pretending it’s settled are different things. The construct is real; the best single marker is not yet agreed on; and every candidate has been validated in a narrow population that may not generalize to your Sunday long run. So durability belongs in the conversation as a distinct, trackable dimension — and it belongs there with its uncertainty stated plainly, not sanded off. That is the honest position, and for a physiology worth trusting, it’s the only one.