SYNTH-50 · Guide
Gran fondo, sportive, gravel or marathon MTB (3-6 h, mass start, climbing)
Build the rider who still has watts at hour three: a long ride that grows with hard work late in it, two quality sessions, heavy lifting, a gut trained to the race dose, and pacing below fresh thresholds
The question it answers
A 3-6 h mass-start event with climbing is 12-16 weeks away: what is the build, which two key sessions, how does the long ride grow, how is the day paced and fuelled? Physiology is not repeated here: fuelling is SYNTH-02, taper SYNTH-34, multi-hour pacing SYNTH-10, durability SYNTH-32, hours and intensity SYNTH-26, the block's lever SYNTH-42, the gym SYNTH-43.
Default (what to prescribe first)
Not from a study; the base's own practice: this build shape. The only sourced parts are the taper parameters and the strength dose.
| Phase (weeks to event) | Quality (2/week) | Long ride | Gym |
|---|---|---|---|
| Build 1 (−16 to −9; 4 weeks if the build is 12) | One shaped VO2max session (SYNTH-33) + one threshold session (SYNTH-27) | Grows from the athlete's current longest ride | 2×/week heavy (in a 12-week build this is 4 weeks, shorter than the 5-25-week blocks pooled in REF-743 and the 8-12 weeks of SYNTH-43) |
| Build 2 (−8 to −3) | Climb reps at race-climb effort + the long ride carries late efforts (durability) | Reaches its longest 3-4 weeks out; race fuelling rehearsed | About 1 session every 7-10 days, never zero |
| Taper (−2 to 0) | Kept, shortened | Shortened | One light session early, none in race week |
On order (the table's order is the base's own practice; this trial tested a different sequence, more threshold work first and more high intensity later): in 60 well-trained male runners, pyramidal-then-polarized improved most, though all four arms improved (~1.5% 5-km TT; REF-509, A-; transfer to cyclists).In trained road cyclists, neither block nor traditional periodisation was favoured over 8-12 weeks (REF-103, B+, systematic review of seven small studies), so the shape that gets executed is the right one.
Deload on signals, not by calendar (SYNTH-29, house practice).
Decision (who → what)
| Situation | What changes (the base's own practice unless the source cell says otherwise) | Source and certainty |
|---|---|---|
| Good fresh FTP, fades after ~3 h | Check fuelling first, then make durability the block's lever | Carbohydrate vs none after 150 min: VT1 power −3% vs −6% (REF-181, B, n=12 trained); after 1000 kJ, amateurs lost ~10% of 5- and 20-min power, unrelated to VO2max or FTP (REF-427, B, n=20) |
| Course dominated by long climbs | Climb reps; body mass matters more | In male pros, uphill ability scales with power/mass¹ and level ground with power/mass^0.32 (REF-174, B+, 2001 narrative review); in 18 elite amateurs a 6.7-km, 6% climb tracked maximal power per kg (r = −0.66), a 14-km flat TT absolute maximal power (r = −0.90) (REF-850, B, cross-sectional) |
| Flat or rolling course, big bunches | Absolute power and wheel skills over W/kg | REF-174 for the exponent; bunch skills are house practice |
| Gravel or XCM with technical, surging terrain | Add a repeatability dose inside the week's quality budget (SYNTH-46) | Transfer from XCO: short surges at 130-140% MAP, but XCO lasts 80-100 min (REF-742, A, systematic review, XCO only); in an 8-day MTB marathon stage race, 7 amateurs spent 36% of race time in the "high" and "very high" of four lactate-anchored HR zones (presumably the top two; the labels are not defined in the abstract) (REF-859, B, HR only, multi-day) |
| Under ~6 h/week | Two quality sessions + the longest ride life allows; say full durability work needs hours they lack | House practice (SYNTH-26, SYNTH-32) |
| Masters | Keep the gym; widen gaps on what monitoring shows | SYNTH-31, SYNTH-43 |
| Hot event | Heat acclimation weeks ahead; pace by RPE | At 32 °C vs 19 °C a TT after 2 h was 15% slower and exogenous glucose oxidation 14-15% lower (REF-461, B, n=10 trained men); protocol in SYNTH-04 |
How to progress / periodise — the long ride
Not from a study; the base's own practice for all of the following numbers:
Start from the athlete's current longest ride, add about 20-30 min a week, and step back on signals. Keep week-to-week total jumps well under 20-30% (SYNTH-29 convention).
Peak at roughly 75-100% of expected event time for events up to ~4 h, and at 4-5 h for 5-6 h events. The peak comes 3-4 weeks out; SYNTH-32's gran fondo row (practice; SYNTH-12 is superseded) is one 4-5 h ride a week with late-ride intervals.
Keep it one ride, not two: in 14 well-trained male runners one continuous session drifted (HR, lactate, RPE rose) while the same work split 6.5 h apart reset — acute only, no adaptation measured (REF-159, B). The session was 6 × 10 min at moderate intensity, not a long easy ride, and the entry says it does not apply to splitting long low-intensity work; the transfer to the long ride is inference.
The two key sessions and climbing work
Threshold / climb reps (Build 1 → 2): a one-week block of moderate-intensity 10-14-min intervals raised power at 4 mmol/L more than a HIT block (+4.5% vs +2.1%) (REF-421, B+, crossover, n=22 well-trained, 1 woman). Not from a study; the base's own practice: in Build 2, ride these seated on real climbs, or on the trainer, for the length of the event's key climbs.
Low-cadence work: intervals ~20% below preferred cadence improved a preloaded 15-min TT more than high-cadence intervals (REF-259, B+, RCT, n=16 trained men). Moderate-intensity 40-rpm reps changed nothing in well-trained veteran cyclists aged ~47 (REF-118, B+, RCT, n=22). If used, make low cadence part of real interval sessions, not moderate-intensity filler (inference from the two; REF-259's interval intensity is not stated in the entry).
Durability session (Build 2): Not from a study; the base's own practice: climb reps or threshold efforts in the last third of the long ride, at targets below fresh numbers: after 2 h at 90% VT1, VT1 power fell 217 → 196 W while HR at VT1 rose 142 → 151 bpm (REF-179, B, n=14 trained). No entry compares late with fresh efforts as training (SYNTH-32).
Why the gym: 12 weeks of 2×/week heavy lifting raised 5-min all-out power after a 185-min ride 371 → 400 W; endurance-only riders did not change (REF-146, B+, well-trained cyclists).
Race day — pacing (label: practice unless a refId is given)
No trial sets the optimal pacing beyond 4 h (REF-010, B+, 2008 narrative review: even pacing helps in events over 2 min; well-trained athletes tend to pace positively).The one gran fondo study points the same way: in 6,589 finishers of the 2025 Quebrantahuesos, the fastest (under 6 h) paced most evenly, started relatively conservatively and gained most on the major climbs and in the final segment, while the most variable pacers slowed mid-race and lost most at the end (REF-899, B+, split times only; fitter riders may simply pace more evenly).
First hour, mass start: Not from a study; the base's own practice: let the start go and do not chase surges above threshold (consistent with REF-899). Sit in wheels on flat and rolling roads.Continuous drafting vs rotating the lead every 500 m cut bike VO2 49.9 vs 59.8 mL/min/kg and HR 155 vs 173 bpm (REF-657, B, 10 national-level male triathletes; a transfer to amateur bunches); a shielded cyclist needed over 30% less external power at racing speeds in an early measurement (REF-842, C+, 1979). The base has no data on the positioning surge itself.
Climbs: a model favours a power surplus uphill (keeping it moderate is the entry's inference) and less on flats and descents, with bigger gains for long climbs and heavier riders (REF-126, C, a model that ignores physiological cost).Not from a study; the base's own practice: early in the event, cap long climbs at or just below fresh threshold, and lower the cap as the hours pass. The boundary falls: after 2 h at 90% VT1, VT1 power was ~10% lower (REF-179), and in trained riders it fell on a non-linear curve, with the time to a 5% drop averaging 139 ± 78 min (REF-143, B, n=12).
Descents: at the Ironman World Championship, athletes who did best relative to their goal times held faster relative downhill bike speeds and steadier HR from climb to descent (REF-659, B, observational, triathletes; a transfer). House practice: ride descents at a controlled, steady effort; do not coast and then restart cold.
Heart rate: do not cap late riding at the fresh HR ceiling; HR at the boundary rises as power falls (REF-179).
Caffeine: 3-6 mg/kg improved time trials by 3.2 ± 4.3% on average, with a wide range (−0.3 to +17.3%) (REF-083, A, systematic review of 21 TT studies). Rehearse it in training.
Race day — fuelling per hour (numbers from [SYNTH-02], which labels them house practice)
Target by duration and level: for 3-6 h, amateur 80-100 g/h, sub-elite 100-120 g/h, elite 110-120 g/h; for 90 min-3 h, amateur 60-90 g/h. The UCI anchors the day: a 4-h ride is a 90 g/h day, and above 60 g/h the plan is glucose + fructose plus gut training (REF-748, A, position statement written for elite riders). 120 g/h is a proposed ceiling for trained athletes, and 120-200 g/h is not substantiated (REF-3490, B+, perspective review).
When NOT to use it / red flags
Do not stop lifting for the season: elite men who stopped for 8 weeks lost squat jump, Wmax and Wingate power relative to controls (REF-744, B, n=7 per group).
Do not taper longer than three weeks: in 14 studies the ≥22-day band was the only duration band that did not improve performance (REF-3492, A).
House rule: fever or its company (chills, aches, malaise), a chesty cough, breathlessness out of proportion or vomiting override the plan and the event — no start, no finishing; chest pain, palpitations or dizziness mean stop and see a doctor. A head cold with no fever allows easy and short riding only, not the event at target effort (SYNTH-36).
What the evidence does and does not show
Does not show: any gravel or single-day XCM outcome, or gran fondo data beyond split-time pacing (REF-899); declines beyond ~4 h; power pacing on real climbs (models only, C); a long-ride dose; whether late-ride efforts beat fresh ones. Women are few. The build table, long-ride increments and pacing caps are coaching reasoning.
Evidence gaps
Gran fondo power and demand data in amateurs: pacing exists only as split times from one race (REF-899); no power, heart rate or fuelling data.
Whether durability predicts amateur mass-start results (candidate: Barsumyan et al. 2025, Front Sports Act Living).
The first hour of a mass start (positioning surges), drafting magnitude in modern amateur bunches (only a 1979 measurement, REF-842), and descending skill.
Gravel: demands, tyres and pressure, rolling resistance, vibration — only a lab hand-numbness method exists (REF-248, C, n=10 men); equipment advice is house practice.
XCM (marathon MTB): heart rate in one multi-day stage race only (REF-859); no single-day XCM demand or training entry.
Long-ride dose-response, and durability training trials, in amateurs and in women.
How this entry was checked and changed
Why this level
(durability and fuelling rest on small B/B+ lab studies, taper and strength on A meta-analyses; build shape, long-ride progression and pacing numbers are house practice; event data are observational only, one B+ split-time and one B heart-rate study)
Content correction 2026-09-27
Revision 2026-09-29: REF-899 (Sánchez-Jiménez 2026, B+, 6,589 Quebrantahuesos finishers: fastest paced evenly, started conservatively, gained on climbs and at the end; association) added to race-day pacing and the head line — it agrees with the existing pacing default; REF-859 (MTB stage-race HR) to the XCM row, REF-850 to the long-climb row, REF-884 to climb cadence, REF-842 to drafting. Gaps narrowed; level unchanged.
Verification 2026-10-01: all 38 cited REF entries opened in the KB.md snapshot. Every number checked matches its entry (REF-509, 103, 743, 107, 181, 427, 174, 850, 742, 859, 461, 046, 421, 262, 884, 259, 118, 179, 146, 010, 899, 657, 842, 126, 143, 145, 131, 659, 083, 748, 3490, 144, 744, 3492, 248); every level/design/n next to a refId matches the head block (no NUMBER, WRONG or LEVEL findings).
Verification 2026-10-01: no SOURCE flag on REF entries by the spec's rule (all have a DOI). For information: REF-899 has no PMID and is from the Crossref abstract only; REF-842 has no sample size and a partly garbled abstract (the card already calls it 'an early measurement', C+); REF-850, REF-884, REF-859 are abstract-only.
Verification 2026-10-01: the fuelling targets (80-100 / 100-120 / 110-120 g/h; 60-90 g/h) match SYNTH-02 and are labelled house practice in the section header; the long-ride numbers are all under one 'Not from a study' header; pacing caps match SYNTH-10 Default 2 word for word.
Verification 2026-10-01: the pointers to superseded SYNTH-05 and SYNTH-12 also sit in 'The question it answers' and in Cross-References (not edited by the pairs, which fix the two body sentences that rely on them).
2026-10-01: superseded pointers SYNTH-05 and SYNTH-12 removed from the 'not repeated here' line; SYNTH-34 and SYNTH-32 carry the content.
2026-10-01: the illness line now carries the house rule in full (fever or its company = stop; above the neck and no fever = easy and short only; chest pain, palpitations, dizziness = a doctor), consistent with SYNTH-36 and SYNTH-30.