Prolonged Cycling Reduces Power Output at the Moderate-to-Heavy Intensity Transition
Participants
n=14 endurance-trained cyclists and triathletes (13 males, 1 female; VO2peak 59.9 ± 6.8 mL/kg/min)
Protocol
Characterisation trial, then a five-stage incremental step test to locate the moderate-to-heavy transition (power and heart rate) by expired gases (VT1) and by blood lactate (LoglogLT)
Step test repeated before and after 2 h of cycling at 90% of the estimated VT1 power
Key Results
| Parameter | PRE | POST (2h) | Change | p |
|---|---|---|---|---|
| VT1 power (gas exchange) | 217 ± 42 W | 196 ± 42 W | −21 W (≈ −10%) | < 0.0001 |
| LoglogLT power (lactate) | 212 ± 47 W | 190 ± 47 W | −22 W (≈ −10%) | 0.004 |
| HR at VT1 | 142 ± 9 bpm | 151 ± 12 bpm | +9 bpm | < 0.001 |
| HR at LoglogLT | 140 ± 13 bpm | 150 ± 15 bpm | +10 bpm | 0.006 |
The power decrease was attributable to loss of efficiency (VT1 −8 ± 10 W; LoglogLT −7 ± 9 W) and to a lower rate of metabolic energy expenditure at the transition (VT1 −14 ± 11 W; LoglogLT −15 ± 22 W)
Authors' conclusion: external work at the moderate-to-heavy transition falls during prolonged exercise because efficiency and the metabolic rate at the transition both fall, while the associated heart rate rises; individual durability assessments are warranted
Mechanisms
Measured in this study: the split of the power loss into an efficiency component (more oxygen per watt) and a metabolic-rate component (the transition itself occurs at a lower energy turnover)
Not measured here: glycogen depletion, fibre-type recruitment shifts and central fatigue are plausible upstream causes but the paper did not test them (inference, not tested)
Practical Application
In n=14 endurance-trained cyclists/triathletes (VO2peak 59.9), 2 h of cycling at 90% of VT1 power lowered the power at the moderate-to-heavy transition by about 10% (VT1 217 → 196 W, P < 0.0001; lactate-derived 212 → 190 W, P = 0.004) while the heart rate at that transition rose 9-10 bpm (142 → 151 bpm) — late in a long ride the top of zone 2 sits at fewer watts but a higher heart rate than when fresh.
Do not cap late-ride zone-2 work by the fresh heart-rate ceiling: the transition HR moved up, so holding a fresh VT1 heart rate after 2 h means riding below the boundary, and holding fresh VT1 power means riding above it. Power falls, HR rises — the two drift in opposite directions.
Pacing for long events: the fresh VT1 is not the boundary that applies after 2 h; plan the steady-state target with a margin below fresh VT1 power. The study gives no percentage prescription; the decline beyond 2 h was not measured, so do not project a fixed % onto a 5-7 h event.
Not from this paper: carbohydrate intake during the preload modifies this drift — see REF-181 (Dudley-Rode 2025).
Durability is individual (the authors' own recommendation): the ±10 W spreads on the efficiency and metabolic components mean some athletes lost far more than others; test it rather than assume the group mean.
Limitations
n = 14 trained cyclists/triathletes, single 2 h preload in the laboratory — a controlled model of durability rather than a race.
One preload duration and intensity; the shape of the decline over 3-5 h, which is what matters for long events, is extrapolated.
Fuelling status during the preload strongly modifies this effect (REF-181) — a decline measured under one nutrition condition is not a general constant.
The outcome is the moderate-to-heavy transition, a physiological boundary; no race or time-trial performance was measured.
13 males and 1 female at this sample size cannot resolve sex differences in durability. A later two-group comparison (16 women, 16 men) found smaller threshold losses in women after 90 min of heavy cycling (REF-1132), under a different protocol (110% of GET, not 90% of VT1 for 2 h).
How this entry was checked and changed
Why this level
regraded 2026-09-26 from A-: within-subject pre/post, n=14
Content correction 2026-09-27
"VO2 at VT1 stable, NS" -> not in the abstract; removed. The decline is attributed to lower efficiency (−8/−7 W) and a lower metabolic rate at the transition (−14/−15 W).
Added: heart rate at the transition ROSE after 2 h (VT1 142 → 151 bpm, P < 0.001; LoglogLT 140 → 150 bpm, P = 0.006) and the lactate-derived transition power (212 → 190 W).
Glycogen / Type II / central-fatigue mechanisms -> labelled unmeasured; "70-75% of fresh VT1" and "4-5 h durability rides" prescriptions removed; CHO note moved under "Not from this paper" → REF-181.
Full text (checked 2026-10-12)
Participants were aged 34 ± 10 years and trained 9 ± 3 h.week−1; both visits were done after a 10-h overnight fast, and during the 2-h phase participants drank plain water ad libitum (full text).
The step test used five 4-min stages starting 50 W below the previously estimated VT1 power and rising 25 W per increment; the 2-h phase was ridden at 194 ± 41 W (full text).
From 15 to 120 min of the 2-h phase heart rate rose 8.2 ± 2.7% (P < 0.0001), while energy expenditure (3.4 ± 4.3%, P = 0.07) and VO2 (3.4 ± 4.9%, P = 0.08) did not significantly increase and RER did not change (P = 0.61) (full text).
VT1 power fell by 21 ± 12 W, or 10.0 ± 5.8% (full text).
VO2 at VT1 decreased from 2.89 ± 0.55 to 2.69 ± 0.51 L.min−1 (68 ± 7 vs 63 ± 8% of VO2peak, P < 0.0001), and energy expenditure at VT1 from 14.4 ± 2.7 to 13.5 ± 2.7 kcal.min−1 (P = 0.0002) (full text).
The drop in VT1 power was not significantly associated with peak fat oxidation, VO2peak, sweat loss or dehydration; its association with PRE VT1 as %VO2peak was rs = 0.54 (P = 0.05) (full text, Table 1).
Whole-body fat oxidation was greater POST than PRE in the first and second steps, and the efficiency-related power loss correlated with the change in fat oxidation (step 1 r = − 0.72, P = 0.004) (full text, Table 2).
Revision 2026-10-13: Limitations — pointer to REF-1132 (sex comparison of threshold durability) added.