REF-1200 · Mujika et al., 2002 · International Journal of Sports Medicine · Paper · PubMed

Should I take rest days in race week?

In one tiny trial, training daily through a 6-day taper improved 800 m by 2.4 s and resting every third day did not. 5 vs 4 runners, no direct comparison: a hint, not proof.

Blimited evidenceMeasured on trained to highly trained athletes

Mujika and colleagues tapered nine male middle-distance runners for 6 days after 18 weeks of training. Both groups reduced their high-intensity interval training by 80% in a progressive nonlinear way; one group trained every day (high-frequency taper, HFT, n = 5) and the other rested every third day (moderate-frequency taper, MFT, n = 4).

The 800-m time improved significantly after the daily taper (124.2 ± 4.9 → 121.8 ± 4.7 s) and not after the taper with rest days (127.1 ± 2.1 → 126.6 ± 2.8 s).

Across the taper, neutrophils, granulocytes, haptoglobin, total testosterone and peak post-race blood lactate rose; peak lactate correlated with pre-taper 800-m time, and its change with the changes in cortisol and in the testosterone:cortisol ratio.

The authors conclude that training daily during a 6-day taper brought significant gains whereas resting every third day did not, and that a high peak lactate and an anabolic hormonal milieu seemed necessary for optimum performance.

Our reading: a hint, not a result — with 5 and 4 runners and no between-group test in the abstract, "one group improved significantly and the other did not" is not evidence that the groups differed; it is still the only direct test of taper frequency the base has, and it points the same way as the meta-analyses (frequency kept).

Protocol

Participants: n = 9 male middle-distance runners, after 18 weeks of training; HFT n = 5, MFT n = 4. Age, personal bests and the allocation method are not given in the abstract.
Taper: 6 days; an 80% nonlinear progressive reduction in high-intensity interval training in both groups. HFT trained daily; MFT rested every third day of the taper (on our reading, 2 of the 6 days).
Measures before and after the taper: 800-m performance time; peak blood lactate after the 800 m; blood samples (white-cell counts, haptoglobin, total testosterone, cortisol).
Not in the abstract: what the low-intensity training contained, whether total volume was matched between groups, where the 800 m was run, and any between-group statistics.

Key Results

HFT — trained daily (n = 5)MFT — rested every third day (n = 4)
800 m before the taper124.2 ± 4.9 s127.1 ± 2.1 s
800 m after the taper121.8 ± 4.7 s (p < 0.05)126.6 ± 2.8 s (not significant)
Change−2.4 s, about 1.9% faster (our arithmetic)−0.5 s, about 0.4% (our arithmetic)
With the taper (as the abstract reports it, apparently both groups together), after vs before: neutrophils 2.89 ± 0.68 vs 2.56 ± 0.61 ×10³/mm³; granulocytes 3.08 ± 0.70 vs 2.77 ± 0.66 ×10³/mm³; haptoglobin 79.7 ± 47.9 vs 60.7 ± 33.6 mg/dl; total testosterone 7.39 ± 1.67 vs 5.52 ± 0.88 µg/l; peak lactate 15.5 ± 1.5 vs 14.4 ± 2.0 mmol/l — all significant increases.
Correlations: peak lactate vs pre-taper 800-m time r = −0.76 (p < 0.05); change in peak lactate vs change in serum cortisol r = −0.75 (p < 0.05) and vs change in the testosterone:cortisol ratio r = 0.82 (p < 0.01).
Not reported in the abstract: a test of HFT vs MFT. MFT started 2.9 s slower than HFT (our arithmetic), with no baseline comparison given.

Mechanisms

Measured: peak lactate after the 800 m, total testosterone and several blood-cell and inflammatory markers rose over the taper.
Authors' interpretation: a high peak lactate and a hormonal milieu "propitious to anabolic processes" seemed necessary for optimum performance.
Inference: correlations across 9 runners cannot show necessity; a higher peak lactate after a faster 800 m may be a consequence of running harder rather than a cause of running faster.

Practical Application

Should I take rest days in race week? In one tiny trial, training daily through a 6-day taper improved 800 m by 2.4 s and resting every third day did not. 5 vs 4 runners, no direct comparison: a hint, not proof.
Short days rather than days off is the base's default (SYNTH-34: "keep session frequency by default", "short days instead of rest days"); this trial is the only direct test of it and points the same way. Resting 2 of 6 days is a cut of about a third, beyond the 20% that REF-1196 allows.
A large cut in interval volume did not stop the gain: the daily group improved with its interval training cut by 80%, a bigger cut than the 41-60% pooled optimum of REF-107 (how much the total volume fell is not given in the abstract).
Do not use the blood markers to steer a taper: they were measured, not tested as a guide, in 9 runners.

Limitations

Tiny groups. 5 and 4 runners; the significant within-group change in one and not the other does not establish a difference between them.
From the abstract only. Allocation, the content and total volume of each taper, the test conditions and any between-group analysis are not available.
Unequal starting points. The rest-day group was about 3 s slower before the taper; how this was handled is not stated.
Biomarkers. Pooled over both groups in the abstract; correlations from 9 runners.
Who. Men, 800-m specialists; a 6-day taper; no women, no longer events.
Funding and conflicts: not stated in the abstract.
How this entry was checked and changed
Why this level

parallel trial with n = 9 (5 vs 4), within-group tests only in the abstract; the full text was not available (Thieme, not open access; no repository copy found in OpenAlex)