Short, single-effort durability tests are mostly noise — protocol design matters as much as the result
Interpretation
A systematic review of 48 race-simulation protocols found most don't anchor intensity correctly and almost none report reliability data under genuinely fatigued conditions. Sex differences in durability response (Mateo-March et al. 2025) further mean a single test, once, tells you very little without a consistent, repeated protocol.
Practical implication
A training-plan durability test should specify a fixed preload dose, fixed test duration/intensity anchored to the individual's own thresholds, and be repeated under matched conditions before any trend is trusted — not a one-off effort.
Sources
5 sources, independently verified.
Found most durability-test protocols have real limitations — many anchor intensity incorrectly (%peak power/VO2max rather than relative to each rider's own thresholds), and only 1 of 10 studies reporting reliability data did so under appropriate fatigued conditions. The core evidence behind 'most durability tests were never checked for reliability.'
Found high-intensity efforts (above critical power) produced greater power-output reductions with less accumulated work than low-to-moderate intensity efforts — 10-20% power declines seen after just 2.5-15 kJ/kg of prior high-intensity work. Concludes kJ alone is an insufficient fatigue metric; intensity distribution matters.
Volume 18(1):99-103. Compared a 20-min TT fresh vs. immediately after a ~4-hour, ~40 kJ/kg submaximal ride. Performance was significantly impaired after that dose, unrelated to traditional lab-based fitness markers or training-load metrics — one of the field studies establishing the professional-cyclist durability dose benchmark.
Volume 7, article 1530162. Measured power/HR for 5- and 20-min efforts fresh vs. fatigued (70-80% initial power until 1,000 kJ completed). Successful amateurs showed only a 6.5% power drop in the fatigued 20-min effort vs. 12.5% for less-successful amateurs — the key amateur-population counterpoint to durability research otherwise dominated by elite/pro samples.
Volume 28(8):661-665. Note: the source Bike articles cited this as 'Journal of Science and Medicine in Sport, 2024' — re-verified real publication year is 2025, corrected here.