REF-1543 · Strepp et al., 2024 · Medicine & Science in Sports & Exercise · Paper · PubMed · Free full text

Did 10 interval sessions in 7 days help trained runners?

Against usual training, 5-km time improved 2.7% and 2.3% and threshold velocity rose; VO2max showed no interaction. Added easy runs gave no edge. It tests a block, not doubles.

Main caveat: Not a test of doubles. The block's density, volume of running, the doubles and the rest day all come as one package.
moderate evidenceMeasured on trained to highly trained athletes

Thirty-three endurance-trained athletes were randomly assigned after baseline testing to ten running interval sessions in 7 days (HSM), the same plus 30 min of low-intensity running after each session (HSM + LIT), or their regular training (CG).

Tests ran 3, 7 and 14 days after the block, and a self-directed 5-km time trial 3-4 days before baseline and 10-11 days after the block. No interaction was found for VO2max (P = .170), peak power (P = .734) or economy (P = .804).

Velocity at lactate threshold showed an interaction (P = .006): up 3.2-4% in HSM at all three tests and 3.2% in HSM + LIT at the first, with the control unchanged. The 5-km time trial showed an interaction (P = .044): 2.7% and 2.3% faster, control 0.3% slower. Neither block arm beat the other.

The secondary analysis (n = 29, best of three post-tests) shows time in the top zone by speed and by power stable through the block while time in the top heart-rate zone fell from 47% to 28% and from 45% to 16%; it also gives time × group P = .071 for threshold and P = .053 for the time trial.

Our reading: a short overload of intervals in trained runners, including double days, went with a faster threshold and 5 km than ordinary training. What it isolates is the block and the added easy running. It cannot credit the doubles.

Protocol

Design: randomised 1:1:1 from a concealed computer-generated sequence, revealed after baseline testing by phone from a researcher the authors call unbiased; blinding not applicable after allocation. Sample-size calculation on VO2max asked for 27, 36 planned, 35 allocated.
Training: 10-min warm-up, then 5 × 4 min at the speed linked to 90-95% of maximal heart rate, 2.5-min recovery (1 min passive, 1.5 min active). Four sessions on a laboratory treadmill (both sessions on day 2 and both on the last day) and six self-monitored outdoors on a flat course with heart-rate targets. HSM + LIT added 30 min at no more than the intensity of 1.5 mmol/L lactate. No other training allowed.
The doubles: morning sessions 6:00-10:00; on two-session days, afternoon sessions 15:00-19:00, at least 5 h apart. The secondary paper refers to one rest day, which would make 10 sessions on 6 days and so four double days (our arithmetic; the schedule itself is in a figure not read as data).
Fuel: no carbohydrate or energy control is reported in the abstract or the secondary analysis.
Control: regular training, which could include cycling or swimming. Time in the 7 days, secondary sample (n = 29): HSM 7.4 ± 0.1 h, HSM + LIT 12.3 ± 0.1 h, CG 8.4 ± 1.1 h. Running: 90.9 ± 8.6, 144.7 ± 17.3 and 44.6 ± 18.8 km.
Tests: two-phase treadmill test, lactate threshold by a 1 mmol/L rise rule, economy at 11 km/h; self-chosen time and track for the 5 km.
Analysis: two-way repeated-measures ANOVA. Secondary analysis used change from baseline to the best of three post-tests, Spearman correlations on pooled block arms.
Funding and interests: the secondary paper prints funding from the Red Bull Athlete Performance Center, described as scientific cooperation with no commercial interest, and no competing interests. The main paper's own statement was not read.

Key Results

Main paper (abstract, n = 33):

MeasureInteraction P (pη2)HSMHSM + LITControl
VO2max.170 (0.09)
Peak power.734 (0.04)
Economy.804 (0.03)
Velocity at lactate threshold.006 (0.18)+3.2% at 3 d (P = .030), +4% at 7 d (P = .006), +4% at 14 d (P = .003)+3.2% at 3 d (P = .011)no change
5-km time trial.044 (0.19)2.7% faster (P = .003)2.3% faster (P = .010)0.3% slower (P = .821)

Secondary analysis (n = 29; baseline to best post-test, or to the second time trial):

MeasureHSM (n = 9)HSM + LIT (n = 9)Control (n = 11)Time × group P (pη2)
VO2max (mL/kg/min)59.2 ± 6.5 → 62.9 ± 5.4 (+6.5 ± 7.0%)61.5 ± 3.8 → 64.7 ± 6.4 (+5.2 ± 6.2%)59.4 ± 2.5 → 59.8 ± 3.2 (+0.7 ± 3.8%).076 (.18)
Lactate threshold (km/h)12.3 ± 2.0 → 12.9 ± 1.6 (+5.3 ± 4.2%)12.7 ± 1.4 → 13.2 ± 1.2 (+4.2 ± 3.7%)12.5 ± 1.0 → 12.7 ± 1.0 (+1.6 ± 2.6%).071 (.18)
5-km time trial19:20 ± 2:31 → 19:02 ± 2:18 (−1.5 ± 1.6%)18:56 ± 1:23 → 18:27 ± 1:11 (−2.3 ± 2.7%)19:04 ± 1:24 → 19:07 ± 1:38 (+0.1 ± 2.1%).053 (.20)
Time in top heart-rate zone, first → last session (± here appears to be SEM, from the figure; other rows SD)47 ± 5% → 28 ± 5% (p = .034)45 ± 4% → 16 ± 4% (p < .001)
Same trial, two readings. Threshold: P = .006 in the abstract's 33 people across three post-tests, P = .071 in 29 people on the best post-test. Time trial: P = .044 in the abstract's 33, P = .053 in 29, one post-test in both, and the block arm's gain reads 2.7% in one paper and 1.5% in the other. A direction, not a firm size.
Double-day reading, four laboratory sessions (two mornings, two afternoons) on two double days, pooled; morning and afternoon not compared. Interval speed 14.6, 14.6, 14.8 and 14.9 km/h; time in the top heart-rate zone 40, 34, 24 and 22%; time in the top oxygen-uptake zone 11, 10, 11 and 13 min, with no significant change in time by velocity, power or oxygen uptake. No single-session arm to compare with.
Speed held is partly prescribed. Treadmill speed was set by the researchers and raised only as far as completion allowed, and outdoors athletes chased heart-rate targets. Athletes reported that late in the block they could not run faster to hit those targets. It is not a free test of second-session output.
Correlations. Time in the top zone by velocity went with peak power change (rs = 0.52, p = .028) and time-trial change (rs = −0.47, p = .049), pooled across arms; by heart rate, power or oxygen uptake, no such relation.

Mechanisms

Measured: no muscle. Heart rate fell away from speed across the block; the secondary paper cites maximal heart rate 4-6 beats lower 3 days after the block.
Authors' interpretation: heart rate underestimates the work in such a block because of slow kinetics and accumulated fatigue, including autonomic changes, so heart-rate targets are a poor guide there. Autonomic measures are not in what was read here.
Not measured, in what was read: food between sessions, glycogen, perceived exertion through the block (the authors list its absence as a limit), any comparison of morning against afternoon performance as an outcome.

Practical Application

Did 10 interval sessions in 7 days help trained runners? Against usual training, 5-km time improved 2.7% and 2.3% and threshold velocity rose; VO2max showed no interaction. Added easy runs gave no edge. It tests a block, not doubles.
Good for: a trained runner or a coach considering a one-week overload: threshold velocity and a 5-km test 10-11 days later moved the right way against usual training, with the caveat, from the secondary sample of 29, that the block arms ran about two and three times the control's distance (our arithmetic).
Good for: leaving out a routine 30 min easy after every session of such a block. No advantage appeared in the main paper. The secondary paper's steeper fall in top-zone heart-rate time (45% to 16% against 47% to 28%) is two within-arm results, and the authors themselves say heart rate underestimates intensity here, so it is not a cost shown. A null in 7 days, not proof of harm or equality.
Good for: not judging such a block by heart rate alone. The authors advise caution with heart rate and a mix of measures, adding only that speed "might be" more apt: top-zone time by heart rate fell while top-zone time by speed held.
Not good for: two sessions in a day against one. No arm did the ten sessions on single days. Not from this paper: the nearest such comparisons are REF-1216 and REF-1542.
Not good for: a VO2max promise. The interaction was not significant, although the best-post-test analysis shows a rise inside the block arms and a smaller one in the control.
Not good for: cyclists by direct reading, a season, or a repeat block. One week was tested, with tests to 14 days.

Limitations

Not a test of doubles. The block's density, volume of running, the doubles and the rest day all come as one package.
The comparator ran about a half to a third as far in the secondary sample (44.6 km against 90.9 and 144.7 km) and could cross-train, while all tests were running. Part of the gain may be more running rather than this structure (our inference).
Main paper read from the abstract. The two papers differ in sample and analysis; the secondary figures are not simple subsets of the main ones.
The best of three post-tests favours a rise in every arm (our inference); the control also rose in peak power this way.
Self-directed time trial, self-chosen time and track, unblinded, with self-monitored outdoor sessions.
Small arms and a sex imbalance: no women in the block-plus-easy arm of the secondary sample, with two dropouts and four more excluded for data quality.
Authors' framing. "Negligible" for VO2max sits beside within-group rises in the secondary analysis; no clear effect is the honest form.
How this entry was checked and changed
Why this level

parallel RCT with n = 33 (the rubric's 20-39 band), about 11 an arm (our arithmetic). The main paper has no open copy in PubMed Central or Europe PMC and was read from the abstract; the schedule, gap, control training, entry criteria, allocation, test methods and second-session data come from the open full text of the secondary analysis (CC BY), and the main paper's own methods were not read. Low edge: 7 days, unblinded after allocation, a self-directed time trial, and interactions that did not reach significance in the secondary analysis