REF-960 · Van Hooren et al., 2026 · Current Biology, Vol. · Paper · PubMed

Longitudinal assessment of total daily energy expenditure in professional cyclists supports a maximal sustainable metabolic ceiling

Blimited evidenceMeasured on elite to world-class athletesWritten before October 2026, being re-checked

The question was whether elite athletes can stay above the proposed ceiling on human energy expenditure — put at 2.5 times basal metabolic rate (BMR) — for long periods. Across a full year, eight male UCI-registered professional cyclists had their total daily energy expenditure (TDEE) estimated and expressed as a multiple of BMR, called physical activity level (PAL).

Energy spent cycling came from power data via validated equations for professional cyclists; non-exercise expenditure and BMR came from methods specific to this population. Across the year PAL averaged 2.52 ± 0.11, which did not differ significantly from the proposed 2.5× BMR ceiling.

Counting only cycling training days, PAL averaged 2.70 ± 0.14, significantly above the limit (p < 0.001). The authors conclude that professional cyclists can go above the threshold for a while but their long-term average stays held near 2.5× BMR, and that the higher expenditure on training days shows why rest days matter.

Participants

n=8 male UCI-registered professional cyclists, followed across a full year. Age, body mass and discipline are not given in the abstract.

Protocol

Total daily energy expenditure estimated across a full year and expressed as a multiple of BMR (PAL).
Cycling energy expenditure: estimated from power data using validated equations for professional cyclists.
Non-exercise energy expenditure and BMR: estimated using methods specific to this population.
Comparison: yearly average PAL, and PAL on cycling training days only, against the proposed 2.5× BMR ceiling.

Key Results

PeriodPAL (× BMR)Against the 2.5× BMR ceiling
Whole year (average)2.52 ± 0.11not significantly different
Cycling training days only2.70 ± 0.14significantly above (p < 0.001)
The authors' conclusions: professional cyclists can go over the threshold for a time during intense activity, but their long-term average stays held close to 2.5× BMR; even athletes who take nutritional supplements regularly still fit the theorized limit; and the higher expenditure on training days shows how much rest days matter for restoring energy balance and letting athletes maintain performance.

Practical Application

In eight male UCI-registered professional cyclists followed for a full year (energy expenditure estimated from power data), the yearly average was 2.52 ± 0.11 × BMR — not different from the proposed 2.5× ceiling — while cycling training days alone averaged 2.70 ± 0.14, significantly above it (p < 0.001). To "why do I need rest days if I can handle the training?": in these professionals, days with riding sat above the ceiling and the year came back to it only because of the days without riding.
The authors' own conclusion is that recovery days are necessary for sustainable long-term endurance training; the study observed this pattern and did not test what happens when rest days are removed.
Inference, not tested here: an amateur's yearly expenditure is far below these values, so for them this is an illustration of why load and rest balance over a year, not the reason a rest day is needed this week.
Not from this paper: the ceiling model is REF-958; ultra-endurance athletes against the same ceiling are in REF-959; the base's use of all three is SYNTH-24; energy-availability targets for cyclists are in REF-748.

Limitations

n=8, all men, all professionals; nothing on women or amateurs.
Energy expenditure was estimated, not measured: cycling expenditure from power data, non-exercise expenditure and BMR by population-specific methods. An error in the estimated BMR moves every ratio.
Observational; rest days were not manipulated, so the conclusion about their necessity is the authors' interpretation of the gap between training days and the yearly average.
No energy intake, body mass or performance outcome is reported in the abstract.
Written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

small cohort, n=8; every component of energy expenditure is estimated, none measured directly