Where the evidence is still weak¶
Reference contents · Update workflow · Maintained conclusions
| Unresolved decision | What would materially improve the evidence |
|---|---|
| 55–60 versus 65–75% FTP | Longitudinal trained-cyclist comparisons with individually measured domains, controlled volume/quality and performance/durability outcomes |
| More easy hours at 7–12 versus ≥20 | Adequate baseline-controlled dose comparisons with complete habitual-load and feeding records—not study-level correlation alone |
| Four hours versus 2+2 | Chronic equal-weekly-volume comparisons, standardized fueling and fresh/fatigued outcomes |
| Polarized versus pyramidal | Clear domain/classification methods, matched dose, relevant athlete level and event-specific performance |
| Time ≥90% as a surrogate | Longitudinal manipulated internal-dose contrasts controlling baseline, volume and work, with individual prediction validation |
| 4×4 / 5×5 / 4×8 / 30/15 / 40/20 | Adequately powered trained-cyclist head-to-heads that define effort/power matching and distinguish VO₂, TT and race outcomes |
| Late-ride intervals | Equal-dose fresh-versus-fatigued trials, durability outcomes and subsequent recovery measurement |
| Two versus one/three hard days | Matched weekly work with sufficient duration, strength/race accounting and sex-specific evidence |
| Sweet-spot “efficiency” / over-unders | Direct performance and recovery-cost comparisons, not acute lactate/signaling or coaching labels |
| Routine deload calendar / maintenance dose | Training interventions versus adaptive reductions, rather than extrapolated event tapers |
| 100–120 versus 90 g/h | Relevant single-/multi-day cycling outcomes, gut tolerance and subsequent quality, not oxidation alone |
| Sleep extension and naps | Mesocycle trained-cyclist adaptation trials and individualized sleep need; women underrepresented |
| Sauna / haemoglobin / cool performance | Same-modality dose comparisons, mass versus concentration, clinical safeguards and amateur applicability |
| Injury prevention / activation | Cycling-specific exposure-adjusted injuries and adequate controls, not EMG or pain surrogates alone |
| CTL/HRV automated decisions | Prospective performance/health comparisons against sensible fixed/adaptive plans, with individual error characterized |
Common limits: small predominantly male samples, one-laboratory replications, short interventions, loosely reported habitual training/diet, different test protocols and multiple outcomes. A positive mechanism is not performance proof; a nonsignificant comparison is not equivalence. Source access matters: many records are abstract-only. Publication year is not the search-close date, and newer is not automatically better.
The methods specify how to revise an existing claim rather than accumulate competing advice. The search register records discovery/access history; the platform makes no claim of exhaustive coverage.