Peak and Taper: the phase that reveals what was built

The Training Phases Series, No.05 . Finale . Premium preview

Peak and Taper is the final phase of the endurance build, and it is the one most often executed on folklore. The earlier phases constructed an engine: a rebuilt aerobic base, accumulated volume, a raised threshold and ceiling, and the race-specific fidelity that teaches fitness to express itself under competition. This phase does something different again. It does not build anything new. It applies a precise, controlled reduction of training so that the fatigue masking months of work falls away faster than the fitness underneath it, placing a physiological peak on race day. The science here is unusually well-developed for endurance training, anchored by two meta-analyses and a coherent mathematical model, which is exactly why the gap between evidence and folk practice is worth closing with care. This preview opens the high-level findings of the full edition. The complete reference sits behind membership. Built on evidence, not affirmations.

Who this is written for

The returning masters athlete, not the young or continuously trained

Most taper guidance is written for continuously trained athletes chasing a peak across a season. The athlete in focus here sits apart: the returning masters endurance athlete, typically aged 40 or older, with a meaningful prior history in long-course sport, who has been away from structured training for an extended period and now intends to rebuild toward competition. That framing matters acutely in a taper, because the comeback master carries two things at once: a recovery clock that runs slower than a younger athlete's, and a narrower margin before sharply reduced training tips into detraining of a base that has only recently been rebuilt. The taper is the place where the entire rebuild is either protected or quietly eroded.

A note on method comes before the science. The taper attracts more durable folklore than almost any other phase of training: the belief that the final week should be near-total rest, that more taper is always safer, the elaborate depletion-then-load carbohydrate ritual, the conviction that a bad night's sleep before the race ruins it, and the slogan that caffeine only helps elites. Each is examined honestly, and where a popular claim outruns its evidence, the correction is stated openly rather than repeated. Every substantive claim is anchored to peer-reviewed primary literature, graded for strength, and kept separate by category, so model predictions, mechanisms, measured outcomes, and reasoned masters extrapolations are never silently merged.

The mechanics of the controlled reduction

Two weeks, roughly half the volume, intensity untouched

The pooled evidence points to a taper of roughly one to two weeks. A meta-analysis of fourteen tapering studies in endurance athletes found that a duration of eight to fourteen days produced the largest time-trial effect of any window tested, with tapers up to twenty-one days still effective, and an earlier and broader meta-analysis of twenty-seven studies reached a compatible conclusion with a two-week taper as the typical optimum. The magnitude of the volume cut has a defined sweet spot: reducing training volume by forty-one to sixty percent produced a significant performance benefit, while a cut of sixty percent or more produced no further benefit, because beyond that point the reduction crosses into detraining and erases the adaptation the taper exists to protect. The single most violated rule is intensity. Maintaining intensity during the taper produced a significant improvement, while reducing it produced a non-significant and slightly negative outcome, which is why two or three short, sharp quality sessions a week are not optional extras but the mechanism that keeps neuromuscular sharpness intact while volume falls away.

Sources: taper duration and the volume and intensity findings, Wang et al. 2023 (10.1371/journal.pone.0282838) and Bosquet et al. 2007 (10.1249/MSS.0B013E31806010E0); the two-week practice optimum, Mujika 1998 (10.1055/s-2007-971942); detraining timeline, Mujika & Padilla 2000, Detraining Part I (10.2165/00007256-200030020-00002); taper shape, Bagley & Murach 2015 (10.3390/SPORTS3030209).

The model behind the method

Fatigue clears faster than fitness, and arriving tired can help

The taper works because fitness and fatigue rise and fall on different clocks. The governing framework treats performance as the gap between a slow-growing, slow-decaying fitness response and a fast-growing, fast-decaying fatigue response. Because fatigue decays roughly three times faster than fitness, cutting training load lets the negative component fall away while the positive one holds near its trained value, and the gap between them, which is performance, opens up. That same arithmetic places the peak roughly ten to fourteen days after the taper begins, which is the physiological reason the empirically optimal duration and the model-predicted peak land in the same window. One of the most counterintuitive and best-supported findings follows from this: arriving at the taper deliberately tired, in a brief and monitored state of functional overreaching, amplifies the gain, because the supercompensation a taper unlocks is most reliable when the preceding block has pushed just into overreaching and then released. The gains themselves are neuromuscular and hormonal, not aerobic, which is precisely why intensity must be preserved while volume falls.

Sources: the fitness-fatigue model and peak timing, Morton, Fitz-Clarke & Banister 1990 (10.1152/jappl.1990.69.3.1171); the model's contested precision, Clarke & Skiba 2013 (10.1152/advan.00078.2011); deliberate overload, Aubry et al. 2014 (10.1249/MSS.0000000000000301) and Wang et al. 2023 (10.1371/journal.pone.0282838); overreaching definitions and markers, Meeusen et al. 2013 (10.1080/17461391.2012.730061), Le Meur et al. 2013 (10.1249/MSS.0b013e3182980125), Hausswirth et al. 2014 (10.1249/MSS.0000000000000177); individualised timing, Pyne, Mujika & Reilly 2009 (10.1080/02640410802509136).

Race week, where the peak is banked or squandered

Loading is two days, sleep is a week, and the day before is for priming

Race week is where a well-built taper is either banked or quietly squandered, and it carries more outdated tradition than any other domain. The week-long depletion-and-load ritual has been superseded: trained athletes reach full glycogen supercompensation within about twenty-four hours of high-carbohydrate intake, so current consensus for events longer than ninety minutes is ten to twelve grams per kilogram per day across thirty-six to forty-eight hours, with no depletion phase, favouring low-fibre sources because higher fibre during loading tracks strongly with gastrointestinal distress. A loading-related gain of one to two kilograms is fuel and water bound to glycogen, not fat, and should prompt no corrective restriction. Sleep is best treated as a multi-day asset: extending sleep across the week before improves performance, while a single poor race-eve night causes far less impairment than sustained restriction, because competitive arousal largely compensates for one disrupted night. The belief that the day before should be total rest is contradicted by the priming literature, which favours a brief activation session with a handful of short race-pace pick-ups to preserve the taper's intensity stimulus and prime the legs.

Sources: rapid loading without depletion, Bussau et al. 2002 (10.1007/s00421-002-0621-5); fibre and gut distress, and the race-morning meal, Fortis et al. 2026 (10.1002/ejsc.70193); sleep extension and a single bad night, Mah et al. 2011 (10.5665/SLEEP.1132), Vitale et al. 2019 (10.1055/a-0905-3103), Bonnar et al. 2018 (10.1007/s40279-017-0832-x); jet lag and melatonin, Janse van Rensburg et al. 2020 (10.1136/bjsports-2020-103163) and Benito et al. 2026 (10.3390/sports14030093); pre-race priming, Marrier et al. 2019 (10.1123/ijspp.2018-0505); caffeine, Wang et al. 2023 (10.3390/nu15010148) and Xue et al. 2025 (10.3390/nu17233792).

The masters reality, reasoned across an evidence gap

A slower clock, a slightly longer taper, intensity still preserved

The honest starting point for the masters athlete is that the direct evidence does not exist. No randomised trial has optimised the taper for athletes over forty, and the foundational studies drew participants aged roughly eighteen to thirty-five. Everything here is therefore either a Moderate-grade finding about recovery physiology or a Heuristic extrapolation built on top of it, and the two are kept apart. The physiology is reasonably clear: masters athletes show lower integrated muscle protein synthesis after damaging exercise and generally require longer to restore function, which supports a coherent set of reasoned adjustments. A masters athlete may reasonably run a slightly longer taper toward the fourteen-to-twenty-one-day end, complete the final long, mechanically demanding session about three weeks out rather than two, allow more recovery between quality sessions, and take a higher post-session protein dose. None of these touch the one rule that becomes, if anything, more important with age: intensity is still preserved, because the neuromuscular sharpness it maintains is exactly what an older athlete can least afford to lose.

Sources: slower masters recovery and protein synthesis, Doering et al. 2016 (10.1249/MSS.0000000000000935) and post-exercise protein (10.1123/ijsnem.2015-0102); recovery kinetics with age, Hottenrott et al. 2021 (10.1123/ijspp.2020-0604); cardiac recovery, Frandsen et al. 2025 (10.1111/sms.70104); modified overload, Meeusen et al. 2013 (10.1080/17461391.2012.730061); detraining timeline, Mujika & Padilla 2000, Detraining Part I (10.2165/00007256-200030020-00002).

Cut volume by roughly half while holding intensity, engineer and time the peak rather than merely resting, bank it through race week with modern loading and protected sleep, and respect an age-adjusted recovery clock without ever relaxing the one rule that matters most: preserve intensity. The master who tapers this way arrives not merely recovered but sharp. The engine was built across four phases; the taper is where it is finally allowed to run. The fittest founders win.

The track, drawn together

Five editions, one continuous argument

This edition closes a track, and the five editions are best read as one continuous argument rather than five separate topics. No.01 rebuilt the aerobic base, the floor on which everything else stands. No.02 loaded volume onto that base, expanding the engine's capacity through patient accumulation. No.03 raised the threshold and the ceiling, building the high end that sets sustainable race pace. No.04 taught the built engine to express its fitness under the exact demands of competition. No.05 has now pointed all of it at a single day: clearing the fatigue that masks the fitness, timing the peak, and banking it through race week. The through-line of the whole track is one discipline, the refusal to confuse training stress with training adaptation, and the returning master, recovering more slowly and with less margin for error, has the strongest reason of any athlete to respect that distinction.

Colophon and method

This preview summarizes the high-level findings of IronPreneur Training Phases No.05, the fifth and final edition of the track. It is a synthesis of peer-reviewed literature that maintains a strict separation between mechanistic and model-based results and measured human outcomes, between general-population and athlete-specific data, and between findings in young or elite athletes and reasoned extrapolations to the returning master. Nothing here is medical advice. Implementation, and any underlying condition, belongs with a qualified clinician. The Training Phases track is now complete: No.01 Adaptation, No.02 Volume Build, No.03 Capacity and Threshold, No.04 Race-Specific, and No.05 Peak and Taper.

Selected sources: Wang et al. 2023 (10.1371/journal.pone.0282838); Bosquet et al. 2007 (10.1249/MSS.0B013E31806010E0); Mujika 1998 (10.1055/s-2007-971942); Morton, Fitz-Clarke & Banister 1990 (10.1152/jappl.1990.69.3.1171); Clarke & Skiba 2013 (10.1152/advan.00078.2011); Aubry et al. 2014 (10.1249/MSS.0000000000000301); Meeusen et al. 2013 (10.1080/17461391.2012.730061); Bussau et al. 2002 (10.1007/s00421-002-0621-5); Fortis et al. 2026 (10.1002/ejsc.70193); Mah et al. 2011 (10.5665/SLEEP.1132); Marrier et al. 2019 (10.1123/ijspp.2018-0505); Doering et al. 2016 (10.1249/MSS.0000000000000935); Mujika & Padilla 2000, Detraining Part I (10.2165/00007256-200030020-00002). The full reference carries the complete citation set. Prepared June 2026. Built on evidence, not affirmations.