Is cramming the week's hard sessions into one block week better than spreading them?
Not here: 9 HIT in one week vs 3 a week for 3 weeks made no detectable difference in 20 junior skiers; recovery dipped in the block week.
McGawley and colleagues asked whether concentrating high-intensity interval training (HIT) into one week beats spreading it, inside the same polarized 3-week cycle.
Twenty well-trained junior cross-country skiers (10 women, 10 men) did both: EVEN, 3 HIT sessions a week for 3 weeks with easy training (LIT) around them, and BLOCK, all 9 HIT sessions in week 2 (twice a day on three days) with only LIT in weeks 1 and 3.
Every HIT session was 5 × 4 min of uphill diagonal-stride roller-skiing at the maximal sustainable intensity, with 6 min of active recovery; LIT time and strength sessions were matched (742 vs 754 min of LIT). The skiers covered the same average distance per interval in both (740 vs 736 m).
Changes in VO2max, skiing economy, 600-m time trial, salivary hormones, capillary density, fibre area and type, enzyme activities and VEGF and PGC-1α protein did not differ between EVEN and BLOCK; the 600-m time improved over the study regardless of intervention.
Perceived recovery before HIT sessions was lower in BLOCK (median 4 vs 6) and the recovery-stress state was lower after the loading week, and during BLOCK the female skiers' recovery-stress state was lower than the males' after week 2 and still after week 3 (15 vs 22, p = 0.014); the BLOCK-vs-EVEN differences were gone by the end (overall pREC and recovery-stress state not different).
Our reading: an independent, female-inclusive trial finding no evidence that blocking HIT is better, against a block literature that is otherwise mostly from one laboratory; it is small, short, end-of-season, and its BLOCK arm had no hard session in the week before testing — the point estimates for VO2max and economy leaned toward EVEN (effect sizes 0.67 and 0.79, not significant).
Protocol
Design: crossover of two 3-week interventions over 53 days; EVEN-BLOCK group 13 (6 male, 7 female), BLOCK-EVEN group 7 (4 male, 3 female); tests before, between (days 25-28) and after. The abstract calls the design randomized; the Methods do not describe how order was allocated and explain the unequal groups by logistics. Timing: end of the racing season, April to June.
Participants: 20 junior cross-country skiers from two Swedish specialist ski schools; males 72.3 ± 4.8 kg, females 61.1 ± 7.5 kg; 60-80 h a year of functional strength training on top of endurance training. Regional ethics approval (Umeå); parental consent for those under 18.
EVEN: LIT 2, 3 and 2 sessions, HIT 3, 3 and 3, strength 2, 2 and 2 in weeks 1-3.
BLOCK: LIT 4, 0 and 5 sessions, HIT 0, 9 and 0, strength 3, 0 and 3. In week 2, two HIT sessions on days 1, 2 and 5 (at least 5 h and a meal between), one on days 3, 6 and 7, day 4 rest.
HIT session: 75 min: 20-min standardized warm-up, 5 × 4 min uphill (~12%/7°) diagonal-stride roller-skiing at maximal sustainable intensity (as far as possible, as evenly as possible), 6 min active recovery (jogging back down), 15-min cool-down; groups of 6-8 supervised by at least two researchers and one coach.
LIT: 58 ± 4 to 127 ± 11 min at about 60-80% of HRmax; about 60% skiing, 35% running, 5% cycling, similar in both interventions.
Load measures: heart rate (summated heart-rate-zone score, five zones of %HRmax), session RPE (CR-10 × minutes), perceived recovery (0-10) before every session, RESTQ-Sport weekly.
Tests: submaximal and maximal incremental roller-skiing treadmill tests (economy at the second submaximal stage, VO2max as the highest 30-s average), a 600-m uphill (7°) time trial after a familiarization, resting saliva (cortisol, testosterone, IgA); vastus lateralis biopsies in 11 skiers aged 18 or over (capillaries, fibre type and area, citrate synthase, HAD, PFK, VEGF, PGC-1α).
Statistics: two-way repeated-measures ANOVA (intervention × time); effect sizes with Hopkins' thresholds (< 0.2 trivial, < 0.6 small, < 1.2 moderate); non-parametric tests for ordinal data; p < 0.05.
Adherence: HIT 97 ± 6% in EVEN and 100% in BLOCK; LIT 95 ± 4% and 98 ± 4%.
Funding and interests: no funding statement in the article; the authors declare no commercial or financial conflict of interest.
Key Results
Pre-to-post change (Δ, mean ± SD) in each 3-week intervention (Table 3); the p value is for the difference in change, and the effect sizes carry no sign:
| Outcome | EVEN Δ | BLOCK Δ | p (EVEN vs BLOCK) | ES |
|---|---|---|---|---|
| VO2max (mL/kg/min) | +0.9 ± 2.6 | −0.8 ± 2.4 | 0.071 | 0.67 |
| Skiing economy (O2 L/min; lower is better) | −0.06 ± 0.10 (significant vs pre) | +0.02 ± 0.16 | 0.117 | 0.79 |
| 600-m time trial (s) | −3 ± 5 (significant vs pre) | −1 ± 6 | 0.280 | 0.44 |
| Resting IgA (μg/mL) | +40 ± 96 | −26 ± 69 | 0.081 | 0.68 |
| Capillary density (per mm², biopsy group) | +3 ± 34 | −20 ± 39 | 0.253 | 0.66 |
| HAD activity (μmol/min/g, biopsy group) | 0.0 ± 1.1 | −0.7 ± 1.0 (significant vs pre) | 0.317 | 0.60 |
| PGC-1α protein (AU, biopsy group) | +0.02 ± 0.02 (significant vs pre) | +0.03 ± 0.04 (significant vs pre) | 0.862 | 0.11 |
No interaction for any performance, saliva or biopsy variable. Time effects regardless of intervention: 600-m time trial (p = 0.003), resting testosterone and testosterone:cortisol ratio, mean fibre area, HAD and PFK activity and PGC-1α.
Direction (our reading of the Δ columns): for VO2max, economy, the 600-m time and capillary density the larger favourable change was in EVEN; the paper does not comment on this.
Work and load were matched: distance per 4-min interval 740 ± 71 m (EVEN) vs 736 ± 75 m (BLOCK); total summated heart-rate-zone score 3,739 ± 440 vs 3,684 ± 449 (p = 0.329); total session-RPE load 7,751 vs 8,127 (p = 0.286); LIT 742 ± 33 vs 754 ± 28 min (p = 0.218). About 16, 37, 27, 10 and 10% of total time fell in heart-rate zones 1-5 in both.
Recovery: perceived recovery before HIT sessions median 6 [3-10] in EVEN vs 4 [1-8] in BLOCK (p < 0.001), in both sexes; RESTQ-Sport recovery-stress state after week 2 was 14 [1-28] in BLOCK vs 18 [5-35] in EVEN (p = 0.033), overall not different (p = 0.510).
Sex: during BLOCK the female skiers had a lower recovery-stress state than the males after week 2 (11 [1-20] vs 21 [13-28], p < 0.001) and week 3 (15 [0-23] vs 22 [11-30], p = 0.014), and a higher heart-rate load during EVEN (4,012 ± 392 vs 3,466 ± 298, p = 0.003); in LIT sessions the females spent 40% of the time at 70-80% of HRmax against the males' 18%.
Abstract vs body: the abstract says the skiers completed 9 HIT sessions in 1 week "without experiencing greater stress or reduced recovery over a 3-week polarized microcycle", which matches the body's overall figures (overall pREC and recovery-stress state not different, p = 0.510); what it leaves out is the transient drop inside the loading week (pREC before HIT 4 vs 6, p < 0.001; recovery-stress state after week 2, 14 vs 18, p = 0.033).An omission, not a contradiction; the Discussion's conclusion repeats the same wording. "Randomized" appears only in the abstract; the Methods describe no allocation procedure.
Mechanisms
Measured: nine sessions in seven days, two a day on three days, were completed at the same average interval distance as three a week, with the same total heart-rate load over the 3 weeks; the subjective cost appeared during the loading week and was gone by the end of the period.
Authors' interpretation: the null result may come from the end-of-season timing (skiers already do more HIT late in the season), the lack of any HIT in the final BLOCK week (skiers reported feeling lethargic at testing; Rønnestad's designs kept one HIT a week after the block), the short tests-to-block interval, the 3-week window and the long 6-min recoveries.Same-day sessions were sustained thanks to 4-5 h of rest and a meal between them and the whole-body nature of skiing.
Authors' interpretation: the female skiers may have worked relatively harder in mixed-sex group sessions, especially in LIT, raising their internal load; the authors suggest female athletes may be more vulnerable to the stress of a blocked week.
Inference: the contrast is not only "blocked vs spread" but also "a week without intensity before the test vs three hard sessions in it". With VO2max and economy leaning toward EVEN, the content of the last week before the retest may matter as much as the block; the study cannot separate the two.
Inference: by heart rate the "polarized" cycle was not very polarized (27% of time at 70-80% of HRmax overall, 40% of LIT time in the female skiers), so the easy days were partly moderate.
Practical Application
Is cramming the week's hard sessions into one block week better than spreading them? Not here: 9 HIT in one week vs 3 a week for 3 weeks made no detectable difference in 20 junior skiers; recovery dipped in the block week.
Blocking is an option, not an upgrade: at equal HIT and equal easy time, the timing gave no evidence of a gain over 3 weeks (B+, crossover, juniors). Choose a loading week for logistics (a camp, a free week) or variety, not for a promised extra adaptation; REF-098 (B+, 12 weeks, load-matched) found the same in trained cyclists.
Expect a dip: perceived recovery before hard sessions was lower in the loading week (median 4 vs 6 on a 0-10 scale) and the recovery-stress state fell after it, more in the female skiers; read a low readiness score during a loading week as the block's expected cost and judge it by how fast it recovers in the following easy days (inference).
Two hard sessions a day can hold quality when separated by 4-5 h and a meal (measured: the same average interval distance across a block week that included three double days); this was whole-body skiing with 6-min recoveries, so a cyclist's or runner's second session may sag more (inference).
Limitations
Small and short: 20 skiers, two 3-week periods; a 4-day test period was the only gap, so a carry-over from the first period cannot be excluded, and no period or order effect is reported.
Allocation: randomization stated in the abstract but not described; order groups of 13 and 7.
Testing after BLOCK followed a week with no HIT, which the authors themselves name as a possible reason for the null result; the study cannot say whether the blocked week or the empty week decided it.
Timing: end of season (April to June), unlike the pre-season block trials; training status at that point may blunt any stimulus.
Biopsies: 11 skiers aged 18 or over; underpowered for muscle outcomes.
Power: non-significant moderate effect sizes (0.66-0.79) for VO2max, economy and capillaries mean the study cannot rule out a difference of that size in either direction; "no evidence of a difference", not "no difference".
Reporting: the abstract's "without ... reduced recovery" understates the measured, transient drop in perceived recovery.
Funding: not stated; no conflicts declared.
How this entry was checked and changed
Why this level
controlled crossover with n = 20, read from the full text (open access, CC BY). The rubric places a crossover of 15 or more at B+; within B+ a weak one: the Methods do not describe the randomization and the order groups were unequal (13 vs 7, "for the logistical reason" of one training hill), there is no washout beyond the 4-day test period, each period lasted only 3 weeks, and the biopsy markers come from 11 skiers
Full text (checked 2026-10-26)
EVEN-BLOCK (n = 13) and BLOCK-EVEN (n = 7); 53-day crossover study design incorporating two, 3-week training interventions (full text)
VO2max: EVEN 60.3 ± 7.2 to 61.2 ± 7.9, Δ 0.9 ± 2.6; BLOCK 61.4 ± 8.1 to 60.6 ± 8.2, Δ −0.8 ± 2.4; P-value 0.071, ES 0.67 (full text)
Skiing economy Δ −0.06 ± 0.10 vs 0.02 ± 0.16, P-value 0.117, ES 0.79; 600-m time-trial Δ −3 ± 5 vs −1 ± 6 s, P-value 0.280, ES 0.44 (full text)
600-m TT performance improved from pre- to post-training, irrespective of intervention (P = 0.003) (full text)
Average distance per 4-min interval: EVEN 740 ± 71, BLOCK 736 ± 75 m (full text)
The pREC prior to each of the HIT sessions: 6 [3–10] vs. 4 [1–8], P < 0.001 (full text)
Recovery-stress state after week two for BLOCK compared with EVEN: 14 [1–28] vs. 18 [5–35], P = 0.033 (full text)
Females vs males during BLOCK after week two: 11 [1–20] vs. 21 [13–28], P < 0.001 (full text)
The number of athletes in each group was not equal for the logistical reason that all HIT sessions needed to be completed on the same hill (full text)