Translating Ramp Test Peak Power into FTP and Structured Training Zones

Ramp Testing Mechanics and the Derivation of FTP

Incremental ramp protocols are widely used across modern cycling platforms to estimate Functional Threshold Power (FTP) without the pacing demands of extended time trials [1, 8]. Standard testing protocols typically initiate cycling resistance at 100 W following a warm-up and apply linear step increases of 20 W each minute until volitional exhaustion [1]. Across commercial platforms and physiological testing frameworks, step rate increments vary based on sex and performance classification, commonly utilizing 15 W/min for women, 20 W/min for elite males, and 25 W/min for non-elite males [6].

In standard commercial implementations, FTP is calculated by applying a fixed multiplier to the peak power output (PPO) achieved in the final completed stage. The standard multiplier of 0.75 (75% of maximum ramp power) was originally derived by Ric Stern [1, 6]. However, mathematical definitions vary across software systems and coaching protocols: Joe Friel utilizes 85% of final-minute power following 20 W/min steps, alternative coaching formulations use 82.5%, and conventional commercial platforms apply 75% across 20–25 W/min steps [6]. While the mean conversion factor across large populations (~1,800 tests) centers around 0.753 to 0.755 (deviating by just 1.4 W from the 0.75 standard), the individual 5th-to-95th percentile spread spans from 67.0% to 83.1% [20]. Approximately 40% of tested individuals deviate from the fixed 0.75 estimate by more than 5%, yielding absolute discrepancies between −58 W and +60 W relative to true sustained capacity [20].

Physiological Divergence: Anaerobic Capacity and Glycolytic Rate

The primary driver of error when translating ramp-derived peak power into sustained threshold power is the athlete's anaerobic capacity ($W'$) and maximum glycolytic rate ($V\text{La}_{\text{max}}$) [1, 17, 20]. Because a standard ramp test reaches terminal fatigue over short, supra-threshold step durations, athletes with substantial fast-twitch motor unit recruitment and high anaerobic work capacity can produce elevated final-stage power outputs that do not reflect their steady-state oxidative capacity [1, 18].

Empirical modelling highlights that the true conversion factor between terminal ramp power and 1-hour sustained power is heavily regulated by $V\text{La}_{\text{max}}$ [20]:

• Low Glycolytic Rate ($V\text{La}_{\text{max}} < 0.30\text{ mmol}\cdot\text{L}^{-1}\cdot\text{s}^{-1}$): Athletes average a true conversion factor of 0.832. Applying a generic 0.75 multiplier underestimates their threshold, causing prescribed 95% FTP intervals to land at roughly 86% of actual physiological threshold [20].
• High Glycolytic Rate ($V\text{La}_{\text{max}} \ge 0.65\text{ mmol}\cdot\text{L}^{-1}\cdot\text{s}^{-1}$): Athletes average a true conversion factor of 0.696. Applying a generic 0.75 multiplier overestimates their threshold, pushing a nominally sub-threshold 95% FTP interval up to 102% of true threshold [20].

This divergence is clearly demonstrated in matched-power case comparisons. Two athletes achieving identical final ramp outputs (~397 W and 399 W, each predicting an FTP of ~298–299 W under the 0.75 rule) exhibited an actual 1-hour sustained power difference of 37 W (322 W vs. 285 W) driven by differences in $V\text{La}_{\text{max}}$ (0.32 vs. 0.75 $\text{mmol}\cdot\text{L}^{-1}\cdot\text{s}^{-1}$) and $W'$ (188 vs. 346 J/kg) [20]. Athletes with strong anaerobic backgrounds who rely on standard ramp multipliers frequently encounter elevated heart rates and premature exhaustion during prescribed sweet-spot or threshold interval sessions [18].

Comparison with Time Trials, Critical Power, and Lactate Metrics

Alternative field testing formats include the 20-minute time trial (utilizing a 0.95 multiplier to estimate 60-minute power and lactate threshold) and Carmichael Training Systems (CTS) 8-minute time trials (utilizing a 0.90 multiplier on the higher or average of two bouts separated by 10 minutes of recovery) [1]. However, these short-duration formats also risk overestimating sustained thresholds in athletes with large anaerobic reserves [1].

Discrepancies between testing modalities and direct physiological markers are consistently observed in scientific literature:

• $FTP_{20}$ vs. Critical Power (CP): In trained cyclists, CP (256 ± 50 W) is significantly higher than 20-minute time-trial FTP (249 ± 44 W; $P = 0.041$), with 95% limits of agreement ranging from −19 W to +33 W, confirming that CP and FTP are not physiologically interchangeable [2]. Similarly, Pringle and Jones demonstrated that CP was significantly higher than power at maximal lactate steady state (P-MLSS: 242 W vs. 222 W; $r = 0.95$) [4].
• Time-to-Exhaustion at FTP: While $FTP_{20}$ shows strong test-retest reliability ($r = 0.94$), trained cyclists exercising at 95% of $FTP_{20}$ demonstrated a time-to-exhaustion of only 42 ± 17 minutes, with only a small minority lasting within 10 minutes of the theoretical 60-minute threshold [5].
• Individual Threshold Concordance: Borszcz et al. compared $FTP_{60}$ (231 W), $FTP_{20}$ (236 W), and individual anaerobic threshold (IAT: 237 W) in 23 competitive road cyclists; 11 of the 23 riders exhibited a 20-minute estimate that differed by 30–50 W from their lactate test [4].
• 8-Minute Protocols vs. Lactate Thresholds: In trained cyclists, 8-minute test estimations showed moderate to very large effect size differences when compared to blood lactate metrics, including modified $D_{\text{max}}$ (ES = 0.77), fixed 4.0 mmol/L blood lactate (ES = 0.83), initial 1.0 mmol/L rise (ES = 1.37), and standard $D_{\text{max}}$ (ES = 2.42) [4].

Zone Prescription and Physiological Domain Anchors

Structured endurance training models divide exercise intensity into metabolic domains bounded by the first and second physiological thresholds [9, 10]:

1. Moderate Domain (Low Intensity / Zone 2): Resides below the first ventilatory (VT1) and lactate (LT1) thresholds, where oxygen consumption ($\text{VO}_2$) and blood lactate achieve steady-state dynamics within 2–3 minutes with minimal metabolic perturbation [10].
2. Heavy Domain (Medium Intensity): Located between VT1/LT1 and the second threshold (VT2/CP/MLSS), characterized by a $\text{VO}_2$ slow component that delays steady-state attainment to 10–20 minutes [10].
3. Severe Domain (High Intensity): Positioned above CP/MLSS/VT2, where metabolic steady state cannot be attained, driving blood lactate and $\text{VO}_2$ continuously upward until reaching exhaustion or $\text{VO}_2\text{max}$ [10, 17].

Laboratory ramp tests accurately delineate these boundaries using gas exchange kinetics [9]. VT1 is established via the V-slope method (the breakpoint of $\dot{V}\text{CO}_2/\dot{V}\text{O}_2$), the first exponential rise in minute ventilation ($\dot{V}\text{E}$), and an increase in $\dot{V}\text{E}/\dot{V}\text{O}_2$ without an increase in $\dot{V}\text{E}/\dot{V}\text{CO}_2$ [9]. VT2 is identified by the second exponential increase in $\dot{V}\text{E}$, concurrent rises in both $\dot{V}\text{E}/\dot{V}\text{O}2$ and $\dot{V}\text{E}/\dot{V}\text{CO}2$, and a rise in end-tidal oxygen tension ($,P{\text{ET}}\text{O}2$) with a decrease in end-tidal carbon dioxide tension ($,P{\text{ET}}\text{CO}2$) [9]. In graded exercise tests, VT1 typically aligns between 65% and 75% of maximum heart rate ($,\text{HR}{\text{max}}$), whereas VT2 falls between 82% and 93% $,\text{HR}{\text{max}}$ [12].

Anchoring training zones to individual threshold markers prevents the maladaptation associated with generic formulas [13]. Prescribing intensity from heart-rate-reserve formulas carries an estimated error margin of 29%; in a 12-week controlled intervention, 60% of individuals prescribed training via heart rate reserve failed to improve $\text{VO}2\text{max}$, whereas 100% of individuals improved when intensities were anchored to individual ventilatory thresholds [13]. When performing threshold-level interventions, interval targeting just above VT2 (such as $4\times8$ min at ~90% $,\text{HR}{\text{peak}}$) has been shown to produce superior adaptations in $\text{VO}_2\text{peak}$, power at $\text{VO}_2\text{peak}$, and power at 4.0 mmol/L blood lactate compared to shorter or longer interval durations [13].

Reconstitution and Modeling in Severe-Intensity Intervals

When training intensities exceed critical power, interval prescription depends on the depletion and reconstitution of finite anaerobic work capacity ($W'$) [17, 22]. The two-parameter critical power model assumes a fixed work capacity above CP, though during extended cycling (>40–80 minutes), $W'$ degrades due to non-glycogen fatigue processes and cannot be preserved by carbohydrate intake alone [17].

During intermittent severe-intensity cycling, $W'$ recovery exhibits non-linear kinetics [22]:

• In trained cyclists, recovery kinetics are best characterized by a bi-exponential model ($R^2 = 0.999$), comprising a fast component (amplitude ~50.7%, time constant $\tau_{\text{FC}} = 21.5\text{ s}$) and a slow component (amplitude ~49.3%, time constant $\tau_{\text{SC}} = 388\text{ s}$) [22].
• Across repeated recovery bouts, the reconstitution of $W'$ slows substantially (decreasing by 8–9% at 3–4 minutes), which extends the slow component time constant $\tau_{\text{SC}}$ to 716 seconds while the fast component remains unchanged [22].
• For short work intervals in the severe domain (e.g., at 95% of peak aerobic velocity with a 2:1 work-to-rest ratio), passive recovery provides greater time to exhaustion than active recovery (1523 ± 411 s vs. 902 ± 239 s), whereas active recovery differences attenuate across long-duration intervals (984 ± 260 s vs. 886 ± 254 s) [21].