REF-1211 · Houmard et al., 1994 · Medicine & Science in Sports & Exercise · Paper · PubMed

Can an 85% volume cut still work in a taper?

In distance runners, 7 days at 85% less volume with intervals kept took 3% off 5-km time and improved economy 6% (within-group); controls did not change significantly. Abstract only.

Blimited evidenceMeasured on trained to highly trained athletes

Houmard et al. tested whether a 7-day systematic reduction in training volume improves distance running performance. Three groups of eight runners were studied: a run taper of high-intensity intervals with an 85% reduction in training volume, a cycle taper with an equivalent amount of interval training on a cycle ergometer, and a control group that continued normal training.

Performance was a self-paced 5-km time trial. The run-taper group ran 3% faster (1036.2 ± 30.6 to 1006.8 ± 28.2 s, P < 0.005) and its submaximal oxygen consumption (6%) and calculated caloric expenditure (7%) at a speed eliciting 80% of VO2max fell significantly (P < 0.01). Five-km performance and running economy did not change in the cycle-taper or control groups.

The authors conclude that 7 days of tapered running improved distance running performance and running economy, and that a cycle-training taper of the same length maintained performance.

Our reading: a small trial in which a volume cut far larger than the 41-60% pooled optimum improved a 5 km within group when the intervals were kept, though Wang's pooled ≥60% band, which includes this study (entered at n = 8, SMD −0.33, −1.32 to 0.66), was not significant (SMD −0.21, −0.58 to 0.16; REF-3492 full text); and the clearest direct evidence in the base that running economy can improve over a taper; with eight per group and within-group tests, it is a direction, not a dose.

Protocol

Who: n = 24 distance runners in three groups of 8 (from the abstract); sex, age, VO2max and normal training volume are not given in the abstract.
Run taper: high-intensity intervals with an 85% reduction in training volume, 7 days.
Cycle taper: an equivalent amount of interval training to the run taper, done on a cycle ergometer, 7 days.
Control: normal training continued.
Outcomes: self-paced 5-km time trial; submaximal oxygen consumption and calculated caloric expenditure at a running speed eliciting 80% of VO2max (running economy).
Not given in the abstract: how runners were allocated, interval content and intensity, baseline weekly distance, between-group statistics, the timing of the tests, and funding.

Key Results

Group (7 days)Training5-km time trialRunning economy at 80% VO2max speed
Run taper (n = 8)high-intensity intervals, volume −85%1036.2 ± 30.6 → 1006.8 ± 28.2 s, −3% (P < 0.005)submaximal VO2 −6%, calculated caloric expenditure −7% (P < 0.01)
Cycle taper (n = 8)the same intervals on a cycle ergometernot alterednot altered
Control (n = 8)normal trainingnot alterednot altered
The run-taper gain is 29.4 s, 2.8% of the starting time (our arithmetic; the abstract rounds to 3%).
Between-group comparisons are not reported in the abstract; the contrasts above are each group against its own baseline.
In a pooled analysis: the Wang 2023 full-text table (REF-3492, Table 1) lists this study's outcomes as time trial, time to exhaustion and VO2max, not movement economy, so the meta-analysis's "economy of movement did not change" does not include this 6% (our reading of that table).

Mechanisms

Measured: after the run taper, oxygen consumption at a fixed submaximal speed fell 6%, alongside the faster 5 km; neither changed with the cycle taper or normal training.
Authors' interpretation: tapered running improved performance and running economy; equivalent cycle intervals were enough to maintain, not to improve, running performance over a week.
Inference: that economy improved only when the intervals were run suggests a running-specific component (mechanics or neuromuscular freshness in the running muscles), but the abstract cannot separate it from the reduced running volume itself.

Practical Application

Can an 85% volume cut still work in a taper? In distance runners, 7 days at 85% less volume with intervals kept took 3% off 5-km time and improved economy 6% (within-group); controls did not change significantly. Abstract only.
Large cuts work if the hard reps stay: this sits with Mujika & Padilla's 60-90% range (REF-1196, B+) and the 80% cut with daily training in REF-1200; the pooled 41-60% optimum (REF-107, REF-3492) stays the default; larger cuts rest on small within-group results, and Wang's pooled ≥60% band was not significant. In cyclists, an 80% one-step cut was not significant in 6 riders (REF-1210); the evidence on cuts above 60% is mixed.
Economy can improve over a taper: "economy did not change" in the pooled data (REF-3492) means not shown, not ruled out; this study measured +6%. A fresher, more economical runner on race day is a plausible gain, not a promise (SYNTH-34, "Promise a better race").
Short races, short tapers: 7 days for a 5-km fits what elite runners do (about 6 days in REF-818) and the race-week evidence in REF-1201 and REF-1202; the 5-10 km plan is SYNTH-47.
If running must stop for a niggle: doing the same intervals on a bike for the taper week held 5-km performance (maintained, not improved). Inference: a reasonable fallback; any pain decision follows SYNTH-73.

Limitations

Abstract only. The full text is not open access (the publisher's page shows the abstract); allocation, interval content, baseline volume and between-group tests could not be read.
Small: eight per group; within-group significance only.
Population not described in the abstract; a later meta-analysis's Table 1 reports a mixed-sex sample of 16 (12 men, 4 women), which does not match three groups of 8; its forest plots enter the run-taper group only (n = 8), with SDs equal to the abstract's ± values × √8, i.e. it read them as standard errors (REF-3492 full text, Figs 2 and 4; our reading).
One outcome race: a self-paced 5-km; economy measured at one submaximal speed; caloric expenditure was calculated, not measured separately.
Funding and conflicts: not given in the abstract.
How this entry was checked and changed
Why this level

from the abstract; n = 24 (three groups of 8). How runners were allocated is not stated in the abstract, so it is not graded as a randomised trial; significance is reported within groups