Recovery weeks turn training into fitness
High confidence
Easing off every third or fourth week is not lost training. It is when the work you have done turns into fitness. In one trial, recreational runners who took a lighter week absorbed their training better and stayed healthier than those who just kept piling on miles.
Why it works
Hard training builds two things at once: fitness and fatigue. While fatigue lingers it hides the fitness underneath. If it keeps piling up, it tips into overreaching. That is when you push harder than the body can recover from, and injury risk rises. A lighter week lets the tiredness fade faster than the fitness does. The gains you already earned become usable, and the next block starts from a fresher, stronger base.
What it means in practice
Build the plan in cycles instead of one long climb. After a few weeks of rising or steady volume, schedule a recovery week that trims weekly mileage by about 20 percent. In base building this is why the generated ramp drops in a cutback week roughly every fourth week rather than climbing every week. It eases load below the recent peak, but never below the volume the runner has already adapted to.
The evidence
-
Runners whose 8-week program ended with a reduced-load week gained more VO2max than those training straight through: about +16% and +22% versus roughly +11%. The reduced-load groups also recovered better, with lower muscle-damage markers (CK, LDH) and a more favorable free-testosterone-to-cortisol ratio. The continuous linear group that skipped a down week showed rising muscle-damage markers, which the authors read as inadequate recovery.
n=88
-
Progress comes from overload paired with enough recovery: a bout of acute fatigue plus adequate rest drives positive adaptation. When recovery falls short, the athlete drifts toward non-functional overreaching or overtraining, and climbing back out can take weeks to months. The panel is candid about a limit in its own case: direct evidence for a clean supercompensation effect after deliberately intensified training is 'not abundant,' so it stresses securing recovery over timing a precise supercompensation window.
-
Cutting training before a race pays off, and this meta-analysis of 27 endurance studies pins down how. The best results came from a roughly 2-week taper that dropped training volume by 41-60% on an exponential curve while holding intensity and frequency steady. The gain traced to the lower volume, not to easing off pace. The pooled effect was small but real.
-
The load itself doesn't injure athletes; inappropriate progression does. Gabbett argues that high chronic workloads lower injury risk, that clearing 18 weeks of training before a first injury reduces the risk of a second, and that under-training can raise risk, so cutting volume is not always the answer. He proposes the acute-to-chronic workload ratio (acute load over chronic load) as a practical injury-risk predictor, monitored up to twice daily across weeks and months, and concludes that appropriately graded high loads build the fitness that protects against injury and improves competitive performance. This is a single-author narrative review drawn mostly from team-sport data, and the ACWR method was later contested for endurance-running use.
Why we call confidence high
The core idea — that planned cutbacks let fatigue fall faster than fitness and lock in recent training — holds up well. It is backed by an ECSS/ACSM consensus statement, a randomized trial in recreational runners, and taper meta-analyses. What is shakier is the promise of a precise "supercompensation window" to hit perfectly. The same consensus notes that direct evidence for a clean supercompensation effect is "not abundant." So the benefit is solid, while the exact timing stays flexible.
Where it applies
Runners at any level building volume over a multi-week block. The direct evidence is strongest in recreational runners and endurance athletes, though the underlying overload-and-recovery pattern applies broadly.
Does not apply to: very short training blocks of a few weeks at moderate volume, where continuous progression can work without a planned down week; prolonged load reduction beyond about three to four weeks, which causes detraining rather than consolidation.
Last reviewed Jul 1, 2026. See how we score.