REF-1542 · Atakan et al., 2021 · Journal of Sport and Health Science · Paper · PubMed · Free full text
Can six interval sessions go into 5 days as doubles?
In 28 active men, 10 × 60 s twice a day, 3 h apart, gave VO2max +7.7% against +6.1% over 2 weeks: no significant difference. Not athletes, not a test of equality.
Main caveat: Confounded by design. Two sessions in a day, a shorter calendar and fewer days of adaptation before the post test all change together.
Twenty-eight recreationally active men were randomly assigned to six high-intensity sessions either as two sessions a day, 3 h apart, every other day for 5 days (HIIT-D), or as one session a day on three weekdays a week for 2 weeks (HIIT-S). Every session was 10 × 60 s at the peak power of the VO2max test (264 ± 48 W), with 75 s at 60 W between.
The double arm ate a 10 kcal/kg meal in the first hour of the 3-h gap. VO2max rose 7.7 ± 4.1% in the double arm and 6.1 ± 5.9% in the single arm (the abstract prints 6.0%), and time to exhaustion at about 80% of VO2max rose 80.1% and 79.2%, in both cases significantly over time with no significant difference between arms.
Total fat oxidation in a 60-min ride rose 15.5% in the double arm (p = 0.048) and 9.3% in the single arm (p = 0.290); the authors call this "more effective", but it rests on two within-group tests. Our reading: in recreationally active men the same work over 5 days showed no clear loss against 2 weeks. What it isolates is the compression of a fixed dose.
It does not separate two sessions in a day from a shorter calendar, and it does not measure the fuel or the quality of the second session.
Protocol
Design: randomised, two parallel arms. Sample size 13 per group from a power calculation (effect size 0.25, alpha 0.05, power 0.85). No trial registration is stated.
Who: healthy recreationally active men aged 20-30 y, non-smokers, not in a systematic endurance, sprint or resistance programme. Familiarised at least 2 weeks before baseline.
Training: 3 min at 60 W, then 10 × 60 s at peak power with 75 s at 60 W, then 3 min at 60 W; all participants completed all 10 intervals in every session. Double arm: first session 9:00 to noon, second noon to 15:00, 3 h apart, every other day over 5 days. Single arm: mornings only, three weekdays a week for 2 weeks. No strenuous activity for 48 h before each test, and no alcohol or exercise for 24 h before each training day.
Between sessions: a meal of 10 kcal/kg (58% carbohydrate, 27% fat, 15% protein) in the first hour of the 3-h gap, plus water. Glycogen was not measured; the authors suspect the meal did not replace much of it.
Tests: VO2max (30-s highest value), 60-min steady ride at about 67% of VO2max with stoichiometric fat and carbohydrate oxidation, and time to exhaustion at about 80% of VO2max with no feedback. Post tests began 72 h after the last session, in the same order, at least 48 h apart. Reproducibility in 12 other people: coefficient of variation 1.3% for VO2max and 9.2% for time to exhaustion.
Analysis: two-way repeated-measures ANOVA (time × group), Bonferroni post hoc; mean ± SD.
Funding and interests: Hacettepe University Scientific Research Projects Coordination Unit; a PhD scholarship from the Scientific and Technological Research Council of Turkey. The authors declare no competing interests.
Key Results
| Measure | Double, 5 d: pre → post | Single, 2 wk: pre → post | Between arms |
|---|---|---|---|
| VO2max (mL/kg/min) | 45.2 ± 6.6 → 48.7 ± 6.1 (+7.7 ± 4.1%, p < 0.001) | 41.8 ± 6.9 → 44.5 ± 6.0 (+6.1 ± 5.9%, p = 0.001) | no significant difference (d = 0.68 as printed; it roughly matches the post-training gap, baseline difference included, not the gain; our arithmetic) |
| Peak power (W) | 264.9 ± 50.2 → 292.5 ± 53.3 (+10.4 ± 4.4%) | 259.8 ± 26.8 → 294.9 ± 24.6 (+13.4 ± 4.7%) | no between-arm test printed (authors: similar extent); p < 0.001 over time |
| Time to exhaustion at ~80% VO2max (min) | 17.6 ± 6.9 → 31.7 ± 17.9 (+80.1%, p < 0.001) | 13.2 ± 5.1 → 23.8 ± 10.7 (+79.2%, p < 0.003) | no significant difference |
| Total fat oxidation in 60 min (g) | 19.6 ± 7.5 → 22.6 ± 7.1 (+15.5%, p = 0.048) | 18.6 ± 6.8 → 20.4 ± 6.8 (+9.3%, p = 0.290) | no between-arm test printed (d = 0.32 as printed; it roughly matches the post-training gap, not the gain; our arithmetic) |
| Total carbohydrate oxidation in 60 min (g) | 118.4 ± 18.6 → 107.4 ± 23.2 (−9.2%) | 132.6 ± 33.5 → 107.6 ± 30.7 (−18.8%) | no significant difference |
| Heart rate across sessions (% of maximum) | 90.8 ± 1.6 | 91.5 ± 1.0 | p > 0.05 |
No clear difference, not "as effective." The paper's title and conclusion claim parity, but no equivalence test was run, and the design was powered with 13 per arm. A 9.2% test noise in time to exhaustion leaves room for real differences.
The fat-oxidation claim is within-group. One arm crossed p < 0.05 at p = 0.048 and the other did not. That is not a between-group result; the carbohydrate fall was numerically larger in the single arm.
Baseline differed numerically. VO2max was 45.2 and 41.8 mL/kg/min (p = 0.19) and time to exhaustion 17.6 and 13.2 min.
Our arithmetic: 10 × 60 s is 10 min of hard work a session, 60 min over the six sessions, which the paper also states.
Mechanisms
Measured: only whole-body responses. No muscle was taken.
Authors' interpretation: lower muscle glycogen in the second session 3 h after the first may speed adaptation, and the higher cellular stress of two sessions may do the same. This is a proposal drawn from other studies; glycogen and mitochondria were not measured here, and the authors call the study a proof of concept.
Not measured: power, speed or perceived effort in the first against the second session of a day; rate of perceived exertion in training; muscle biopsy; cardiac output; anything beyond the post tests, so nothing on whether the gain lasts.
Practical Application
Can six interval sessions go into 5 days as doubles? In 28 active men, 10 × 60 s twice a day, 3 h apart, gave VO2max +7.7% against +6.1% over 2 weeks: no significant difference. Not athletes, not a test of equality.
Good for: saying that a compressed week of short doubles showed no clear loss in recreationally active men, with a meal between the sessions and at least a day off between double days.
Good for: a time-pressed non-athlete who cannot spread six sessions out, as a description of what happened here. Not from this paper: the base's skiers on nine sessions in one week (REF-1216) leaned toward the even spread, so this is not a general licence.
Not good for: deciding that a double beats a single. The two sessions and the short calendar cannot be told apart, and no arm did one session a day on a matched 5-day calendar.
Not good for: threshold work, 5 × 4-min or longer intervals as doubles. The protocol is 10 min of hard work a session, chosen because it looked repeatable.
Not good for: fat-oxidation claims against the single schedule, because the comparison is two within-group tests.
Not good for: women, or any competitive cyclist or runner, by direct reading. Neither was in the study.
Limitations
Confounded by design. Two sessions in a day, a shorter calendar and fewer days of adaptation before the post test all change together. Post tests started 72 h after the last session in both arms, about 7 days after the first session in the double arm and about 2 weeks in the single arm (our arithmetic for the single arm).
Non-athletes, men only, 25 ± 4 y, with very large changes in time to exhaustion that the same dose would not produce in a trained rider (our inference).
Six sessions, no follow-up. Nothing on repeating the pattern or on how long the rise lasted.
No muscle, no glycogen, no second-session quality data. The proposed mechanism was not tested.
Reporting. The abstract gives 6.0% and the results 6.1% for VO2max in the single arm; 15.5% and 15.4%, and 9.3% and 9.2%, for fat oxidation; and the two carbohydrate p values (0.014 and 0.012) are attached to opposite arms in the abstract and the results. Baseline SD for VO2max in the double arm is 6.1 in the table and 6.6 in the text.
Authors' reach. They suggest coaches and athletes may apply the model, while saying that athlete data are absent. The second statement is the one the data support.
How this entry was checked and changed
Why this level
parallel RCT with n = 28 (the rubric's 20-39 band), read from the full text (CC BY-NC-ND). It sits at the low edge: the allocation method is not described, no blinding is reported, and the between-group results are printed only as "p > 0.05"