REF-1440 · Menges et al., 2026 · Frontiers in Sports and Active Living · Paper · PubMed · Free full text

How many minutes on the bike replace an easy run?

Athletes and coaches in this survey said about 1.8× (30 min run ≈ 60 min ride; 3 h ≈ 5 h), and about half the minutes of running for an easy ride. Their opinion, not a tested rule.

Main caveat: Perceptions, not physiology; the authors call the study exploratory.
weak evidence: opinion or lab dataMeasured on highly trained to elite athletes

Menges, Dindorf, Bartaguiz and Fröhlich (Endurance Coach GmbH, RPTU Kaiserslautern-Landau) — the group behind the base's meta-analysis of running-versus-cycling trials (REF-1368) — asked how endurance athletes and coaches perceive and apply cross-training between running and cycling.

Fifty-five high-performance practitioners (24 athletes, 21 coaches, 10 both; 42 from triathlon) completed a 33-item questionnaire on suitability, duration, frequency, intensity and equivalent durations in both directions. Replacing a running session with cycling was rated more suitable than the reverse (median 5 vs 3 on a 5-point scale; p < 0.001).

For easy ("regenerative") sessions, respondents estimated that cycling needs a mean 1.76 times the running minutes and running 0.54 times the cycling minutes; the factors fell with session length (2.01 for a 30-min run), so the authors find them "not adequately represented by a constant conversion factor".

Low-intensity sessions were judged more transferable than threshold or VO2max sessions when running was replaced by cycling; in the other direction the intensity effect was small. Group differences were minor; athletes asked for longer rides than coaches to replace 30- and 60-min runs.

Respondents agreed strongly that cycling should be prescribed from cycling-specific heart-rate zones. The authors stress that the study measures perceived transferability, not physiological equivalence. Our reading: the honest source for the minutes coaches actually use, to be quoted as practice.

On minutes, these practitioners read the asymmetry one way (though they rated cycling the more suitable substitute, median 5 vs 3): they count a minute of running as worth more than a minute of cycling — a cyclist who runs needs fewer minutes than a runner who rides — the reverse of the sentence in REF-537's Practical Application that "a runner adding cycling keeps more than a cyclist adding running".

Neither direction is shown by trials: REF-1368 found small, symmetric deficits.

Protocol

Design: cross-sectional anonymous online survey (LimeSurvey), 10 December 2025 to 20 January 2026; ethics approval from RPTU's social-sciences committee, written consent waived.
Questionnaire: 33 items, developed from the authors' meta-analysis (REF-1368) and exploratory interviews with national triathlon coaches, reviewed for content validity, piloted with five competitive athletes; 5-point Likert items plus free-text equivalent durations for 30, 60, 120 and 180 min of easy running and 45, 60, 120 and 180 min of easy cycling.
Participants: n = 55 complete responses of 97 who started (57%); 42 triathlon, 7 running, 6 cycling; 24 athletes, 21 coaches, 10 athlete-coaches; mean experience 11, 12.3 and 21 years; recruited through networks, federations in Germany, Austria and Switzerland, and Olympic training centres (snowball sampling); a priori sample size 54 for a paired comparison.
Reliability: the running-to-cycling item set α = 0.75 (composite used); the cycling-to-running set α = 0.55 (items analysed singly).
Analysis: Wilcoxon, Kruskal-Wallis, Friedman tests with Holm correction; one-way ANOVA on the composite; logarithmic fits of factor against duration.
Funding and interests (as printed): no financial support received; "Author TM was employed by Endurance Coach GmbH"; the others declare no commercial or financial relationships.

Key Results

Equivalence factor for easy sessions (Table 2)nMean (SD)Median (IQR)
Cycling replaced by running (running min ÷ cycling min)550.54 (0.16)0.51 (0.11)
Running replaced by cycling (cycling min ÷ running min)551.76 (0.36)1.82 (0.47)
Easy running replaced by cycling (Table 3)Mean cycling (SD)MedianEasy cycling replaced by running (Table 4)Mean running (SD)Median
30 min run60.2 (13.8)6045 min ride25.6 (8.94)25
60 min run108 (20.8)12060 min ride39.5 (29.7)30
120 min run201 (62.7)210120 min ride58.7 (25.0)60
180 min run282 (99.7)300180 min ride86.5 (26.8)90
Duration dependence: factors differed across durations in both directions (Friedman χ² 46.90 and 49.19, p < 0.001); running-to-cycling factors fell steadily with longer runs (2.01 for 30 min); cycling-to-running factors were 0.48-0.57 and stabilised from 120 min.
Suitability: cycling as a substitute for easy running median 5, running for easy cycling median 3 (location shift 1.50, 95% CI 1.00-2.00; p < 0.001).
Intensity: replacing running by cycling, low intensity was rated more transferable than threshold (r = 0.84) and VO2max work (r = 0.76), threshold and VO2max alike; replacing cycling by running, a small overall effect (Kendall's W = 0.05) with no significant pairwise difference.
Groups: athletes estimated longer cycling durations than coaches for 30- and 60-min runs; otherwise few group differences.
Heart-rate zones (Item 22): high agreement that cycling intensity should use cycling-specific zones (descriptive only).
Numbers that do not add up: the abstract gives the running-to-cycling factor's SD as 0.12, Table 2 as 0.36; the text gives the cycling-to-running mean as 0.55, Table 2 as 0.54.

Mechanisms

Measured: practitioners' ratings and estimates only.
Authors' interpretation (offered as possible reasons for the perceptions, not tested): running is perceived as the stronger stimulus — more muscle mass and cardiovascular strain, higher delta efficiency, more central fatigue — so longer rides are thought necessary; cycling lets volume be extended with less mechanical load; higher intensities depend more on sport-specific movement and recruitment, so they are thought less transferable.
Inference: a factor that falls with duration fits coaches weighing fatigue and mechanical cost, not only aerobic stimulus; equal adaptation from 1.76 × the minutes has never been tested, and the trials that held running fitness matched minutes and intensity (REF-1428, REF-922, REF-953).

Practical Application

How many minutes on the bike replace an easy run? Athletes and coaches in this survey said about 1.8× (30 min run ≈ 60 min ride; 3 h ≈ 5 h), and about half the minutes of running for an easy ride. Their opinion, not a tested rule.
Quote it as practice, never as a law: the factor changed with session length even in the respondents' own estimates. The tested default is minute-for-minute at matched intensity (REF-1428, REF-922); a longer ride for an easy run is acceptable where time allows (not from this paper; the base's own practice).
A cyclist who must run: the half-the-minutes estimate may also reflect running's higher mechanical load (the authors' possible reading of the perceptions, untested); for a cyclist new to running the tissue limit decides, not the factor (SYNTH-75, PHYS-22, REF-235). Respondents rated running a poorer substitute for cycling (median 3) than the reverse (median 5).
Hard sessions: practitioners judged threshold and VO2max running less replaceable by cycling than easy running — keep race-specific work in the race sport as the race nears (PHYS-40, SYNTH-76).
What it does not show: that any factor yields equal fitness, fatigue or adaptation, or anything about single-sport cyclists (6 respondents).

Limitations

Perceptions, not physiology; the authors call the study exploratory.
Sample: 55 of 97 completed (57%), self-selected, snowball-recruited, mostly triathletes, German-speaking only.
Self-developed questionnaire; the cycling-to-running item set had low reliability (α = 0.55).
Equivalent durations for easy sessions only; skewed free-text answers, means and medians differ.
Reporting slips in the factor's SD and mean between abstract, text and table.
Interest: as printed, the first author was employed by Endurance Coach GmbH; the same group wrote REF-1368. No funding.
Supplementary tables (factors by duration, dispersion, post-hoc tests) not read.
How this entry was checked and changed
Why this level

practitioners' opinion (the rubric's C for expert opinion); it documents practice, not an effect of training, and the authors say so

Full text (checked 2026-10-29)
Of the 97 individuals who initiated the survey, 55 provided complete responses and were included in the final analysis (full text)
Participants represented performance-oriented endurance sports disciplines, primarily triathlon (n = 42), with additional representation from running (n = 7) and cycling (n = 6) (full text)
Cycling vs. Running: n 55, mean factor 0.54, SD 0.16, median 0.51, IQR 0.11; Running vs. Cycling: n 55, mean factor 1.76, SD 0.36, median 1.82, IQR 0.47 (full text)
Across all durations, the running vs. cycling factors were consistently greater than 1 (e.g., mean = 2.01 for 30 min running), whereas the cycling vs. running factors remained consistently below 1 (mean = 0.48–0.57 depending on duration) (full text)
30 min running: mean cycling duration 60.2 min, median 60 min; 180 min running: mean 282 min, median 300 min (full text)
Running-to-cycling substitution received a median rating of 5, whereas cycling-to-running substitution received a median rating of 3; the estimated location shift was 1.50, 95% CI [1.00, 2.00] (full text)
Abstract: replacing a running session with a cycling session (mean factor = 1.76; SD = 0.12); text: for cycling vs. running, the mean factor was 0.55 (full text)
Author TM was employed by Endurance Coach GmbH; the author(s) declared that financial support was not received for this work (full text)