REF-759 · Lamberts RP, Maskell S, Borresen J, Lambert MI — 2011 · International Journal of Sports Medicine, · Paper · PubMed
Adjusting a standard submaximal test so it ends at 86-93% of HRmax roughly halved the error of heart-rate recovery (typical error 6-7 → 3 bpm, CV ~11% → 4-5%) without changing HRR itself
Protocol
Thirty-one subjects (n = 31) performed four submaximal running tests. Based on the heart rate reached in the first, the protocol was either kept identical, made faster, or made slower for the last two tests, and the reliability of HRR was compared.
Key Findings
HRR itself did not change when the protocol was adapted.
Measurement error fell in the adapted groups: CV 11 ± 7% → 5 ± 3% (slower) and → 4 ± 3% (faster); typical error 6-7 → 3 bpm.
The authors recommend protocols that target 86-93% of HRmax.
Practical Application
Heart-rate recovery is about twice as noisy when the effort before it ends outside roughly 86-93% of HRmax — read a small HRR change after an easy effort as noise.
On a graded or heart-rate-anchored protocol (the LSCT), that band is where the top stage should end. It does NOT set the watts of the fixed-power fatigue test, which is about 110% of FTP for 3 minutes by protocol with no HR target (SYNTH-41); there this study only explains why HRR is the least reliable of the test's numbers.
Limitations
Running protocol; small groups (9-12); reliability over a short window.