REF-1545 · Mckee et al., 2021 · International Journal of Sports Physiology and Performance · Paper · PubMed

Do intervals late in a 90-min ride lose time near VO2max?

In 16 trained cyclists at fixed power, no clear difference at 13, 36 or 69 min, though they felt harder later. One ride, not a second session or a training effect.

Main caveat: Abstract only. Means, preload intensity, fuelling and the washout could not be read, and the time at or above 90% is known only through its P value.
limited evidenceMeasured on trained athletes

Sixteen trained cyclists rode three 90-min sessions in randomised order, with the same interval set placed at the beginning, middle or end, starting at 13, 36 and 69 min. The set was 3 × 3 min at 90% of the power linked to VO2max, with 3 min of recovery between.

Oxygen uptake, ventilation, respiratory rate and heart rate were recorded through the work intervals; perceived exertion at the end of each interval, and for the session 20 min after it finished. Mean oxygen uptake (P = .479) and time at or above 90% of VO2max (P = .753) showed no significant difference by position.

Against the beginning position, interval perceived exertion was higher for both the middle (effect size 0.83, P < .01) and the end (effect size 0.75, P < .05). Ventilation was higher at the end than at the beginning (P = .015, effect size 0.63); respiratory rate showed no difference (P = .297). Our reading: what it isolates is position within one ride with the power held.

The oxygen response did not clearly change, and the cost showed up in effort and breathing.

Protocol

Design: randomised crossover, three 90-min sessions. Washout and session intensity outside the intervals are not in the abstract.
Who: 16 trained cyclists, 13 men and 3 women.
Intervals: 3 × 3 min at 90% of the power at VO2max, 3-min recoveries; the set starts at 13, 36 or 69 min.
Measures: oxygen uptake, minute ventilation, respiratory rate and heart rate, continuously during work intervals; interval and session perceived exertion.
Not in the abstract: the actual time at or above 90%, the group means, the riding intensity around the set, food or fluid during the ride, the test for VO2max, session perceived exertion results, funding and interests.

Key Results

MeasureBeginning (13 min)Middle (36 min)End (69 min)P
Mean oxygen uptake.479, no clear difference
Time at or above 90% VO2max.753, no clear difference
Interval perceived exertionreferencehigher (effect size 0.83)higher (effect size 0.75)middle < .01; end < .05
Minute ventilationreferencenot reported as differenthigher (effect size 0.63)end vs beginning .015
Respiratory rate.297, no clear difference
No values in the abstract for any of the outcomes, only P values and effect sizes.
No clear difference, not "no impact." The title says position does not impact the time near VO2max; there is no equivalence test and n = 16.
Our arithmetic: the set takes 15 min, so from a 69-min start it ends at 84 min, inside the 90.

Mechanisms

Measured: no clear change in oxygen response, higher perceived exertion in the middle and end positions, and higher ventilation at the end.
Authors' interpretation: the higher ventilation may mean a larger anaerobic contribution to the later intervals. This is a suggestion; the abstract reports no measure of it.
Not measured, as far as the abstract shows: muscle fatigue, glycogen, lactate, power the rider could have produced freely, and any training response.

Practical Application

Do intervals late in a 90-min ride lose time near VO2max? In 16 trained cyclists at fixed power, no clear difference at 13, 36 or 69 min, though they felt harder later. One ride, not a second session or a training effect.
Good for: deciding where a short VO2max set sits in a ride of about 90 min. Late placement did not clearly cut the time near VO2max at this power, and it cost more perceived effort.
Good for: a reason to read oxygen-side targets and perceived effort separately: they did not move together here.
Not good for: stacking after a long ride. A 69-min start is not a 2-h preload, and the intensity of the riding before it is not in the abstract.
Not good for: a same-day double. There was no gap, no meal and no second warm-up.
Not good for: saying output holds when fatigued. Power was prescribed, so it could not fall. Not from this paper, and by inference: this is why it can sit beside REF-1097, where intervals in a later session were lower in power (297 vs 323 W), without contradicting it.
Not good for: any adaptation claim. There was no training period.

Limitations

Abstract only. Means, preload intensity, fuelling and the washout could not be read, and the time at or above 90% is known only through its P value.
Prescribed power. The design removes the outcome that mattered most in self-paced work.
One 90-min session, with intervals only 69 min in; three short 3-min efforts total 9 min of work (our arithmetic).
Few women (3 of 16), no sex analysis reported.
n = 16, no equivalence test; a small real change in time near VO2max could be missed.
Funding and conflicts could not be read.
How this entry was checked and changed
Why this level

a crossover with n = 16 earns B+ by design (rubric: crossover of 15 or more), lowered one step because the outcome is a physiological response at prescribed power in one session, not performance or adaptation. Read from the abstract only; no open copy in PubMed Central or Europe PMC