REF-707 · Zambolin, Vikmoen, Cumming, Skattebo, Ødemark, Øfsteng, Hammarström, Hansen, Rønnestad & Raastad, 2026 · Journal of Applied Physiology, 2026 Mar 1;140 · Paper · PubMed
In well-trained cyclists, effort-matched intervals with BFR gave similar performance gains at 42% lower power
BFR is a tool for substituting mechanical load, not for amplifying adaptation
Protocol
n=17 well-trained cyclists randomised into two groups. Both completed five HIIT sessions in a 6-day microcycle, each 6 × 5 min intervals with 2.5 min recovery, matched by effort rather than by power.
| Group | Intervals |
|---|---|
| BFR | thigh cuffs occluding the legs during the intervals |
| HIIT (no BFR) | the same intervals without cuffs |
Intervals prescribed by RPE 17-19 (Borg 6-20), with power adjusted between series to hold that RPE; cuffs 13 cm wide on the proximal thighs at a constant 180-220 mmHg (≈ 80% of limb occlusion pressure), inflated during each 5-min interval and deflated during the 2.5-min recovery (full text, checked 2026-09-28: Methods, training).
Measures: VO₂max, power at 4 mmol·L⁻¹ lactate (LT4), mean power in a 5-min maximal test, %VO₂max at LT4, haemoglobin mass, blood volume, plus vastus lateralis biopsies — fibre cross-sectional area (CSA), capillaries, citrate synthase, hydroxyacyl-CoA dehydrogenase, COX4.
Key Findings
Mean interval power: BFR 177 ± 3 W vs HIIT 307 ± 8 W (42% lower, P<0.01), with no differences in heart rate or RPE. Cardiovascular strain and effort were preserved; external mechanical work was not.
Both groups improved 5-min maximal power by ~4%, with no changes in LT4, VO₂max, haemoglobin mass or blood volume.
Muscle: the HIIT group showed a significant reduction in type 2 fibre CSA compared with BFR; no changes in the other muscle analyses (capillaries, oxidative enzymes).
BFR added nothing on top of performance. This study does not show that BFR makes intervals better for a trained athlete — it shows a similar short-term result at much lower external load.
Practical Application
In n=17 well-trained cyclists (VO₂max 67), five effort-matched 6 × 5 min HIIT sessions over 6 days with thigh-cuff BFR were done at 177 W vs 307 W without cuffs (42% lower) at the same heart rate and RPE, and both groups gained ~4% in 5-min power — read BFR as load substitution, not as an amplifier: at this calibre, cuffs did not make the intervals work better.
The authors suggest BFR as a tool for periods requiring reduced mechanical stress. Inference, not tested here: e.g. blocks already saturated with mechanical load (heavy strength, low-cadence work) or return from a lower-limb injury, where the tissue will not tolerate full power but the aerobic stimulus should be preserved.
Not from this paper; the base's own practice — load accounting breaks on this kind of session. Power-based TSS/IF will read like an easy ride while heart rate and RPE are near-maximal. Do not treat it as a soft session, do not let it drag the weekly load picture down, and do not estimate thresholds from it. Judge by heart rate and RPE, and tag the session.
Not from this paper; the base's own practice — not a do-it-yourself method. The protocol used measured occlusion pressure with proper cuffs; an elastic band tightened by feel does not reproduce it. Take method and safety from REF-566.
Inference, not tested: REF-102 pooled an extra benefit of adding BFR to interval training, including in trained individuals; this effort-matched 6-day study in well-trained cyclists found no extra gain — the comparisons differ (same intervals plus cuffs vs effort-matched at lower power).
Limitations
n=17, i.e. roughly 8-9 per group. This is a null result in a very small sample. "No difference found" is not "the effects are the same" — it lacks the statistical power to exclude a real difference.
6 days, 5 sessions. Nothing is established about a block spanning weeks or months: not cumulative effect, not tolerance of repeated BFR intervals, not long-term safety.
There was no non-training control group. Both groups gained ~4% in the 5-min test; after a 6-day overload microcycle part of that may be freshening and test familiarisation rather than adaptation (inference).
Matching by effort rather than by work is a design choice. It does not follow that 177 W without cuffs would give the same result — that arm did not exist.
Only well-trained cyclists; novices and masters are not covered.
Near-maximal 5-min intervals under occlusion are hard and require supervision; this is not a self-prescribed protocol.
How this entry was checked and changed
Content correction 2026-09-27
Key Findings said "no significant between-group differences on any measure — performance or muscle": the abstract reports a significant reduction in type 2 fibre CSA in the HIIT group compared with BFR. Added.
Removed values not in the abstract: HR 168/170, ΔVO₂max +0.9/−1.8%, ΔLT4 +2.4/+0.3%, 5-min +3.8/+3.8% ("matched to a tenth of a percent"), recovery at 50% power, "predominantly male"; the abstract gives ~4% for both groups and no change in LT4/VO₂max. RPE target and cuff pressure kept, labelled as full text.
Tier 4 -> Tier 3 (well-trained competitive cyclists, VO₂max 67). Load-accounting and DIY advice labelled as the base's own practice; use-case list labelled as inference beyond the authors' "periods requiring reduced mechanical stress".
Cross-reference blurbs corrected: REF-102 does not say it rests on Tier 1-3 (it reports a benefit in trained individuals); REF-105 speaks of recovery phases and in-season maintenance, not injury; REF-016 is an unsourced summary.
2026-09-28: lines labelled "full text; not checked" read against the open-access full text (repository.mmu.ac.uk, published version): 0 confirmed, 1 corrected ("≈ 80% of individual limb occlusion pressure" -> "≈ 80% of limb occlusion pressure": the paper does not say LOP was measured per athlete), 0 removed.