Functional threshold power (FTP) has been defined as the highest average power output (PO) that can be maintained for 1 hour (1). The use of FTP has become popular in cycling because it can be easily and frequently estimated from user-based, individual power meters during relatively short field-based testing (1,5). However, the physiological concepts underlying FTP are not understood.
In practitioner publications, it is has been suggested that FTP estimated from the average PO during field-based tests of either 8-minute (5) or 20-minute duration (1) is representative of PO at lactate threshold (LT). The LT represents the point during progressive exercise at which blood lactate accumulates (14). Given the recent prevalent use of FTP to define exercise intensity, understanding if FTP represents PO at LT would provide reference to a significant body of scientific literature because it is well established that running speed or cycling PO at LT is a strong predictor of endurance performance (8,12). In addition, if FTP is equivalent to PO at LT, this would be of particular interest to athletes because endurance exercise training is known to increase work rate at LT (11,21).
We have worked with several regionally based cycling teams and have measured PO at LT in our laboratory. Anecdotally, we have observed that when queried, the vast majority of these cyclists report training and racing sessions during which they were able to average a higher PO for 1 hour than our measurement of PO at LT (as determined by a 1 mmol·L−1 or greater increase in blood lactate above baseline) (10). This anecdotal evidence suggests that FTP and PO at LT may not be equivalent. Therefore, we hypothesized that the FTP estimated from average PO during a common field test (FT) is not equivalent to PO at LT.
Experimental Approach to the Problem
There can be significant variation in testing protocols and methodology that can make it difficult to interpret and compare data from different laboratories (2). Consistent with this, McGehee et al. found that running speed at LT can be significantly different based on the methodology used to determine LT (16). Therefore, to test our hypothesis, we completed 2 studies: (a) to identify the appropriate methodology to calculate PO at LT between several different methodologies (study A) and (b) to compare PO at LT with estimated FTP (study B). In study A, blood lactate was measured at each PO throughout a maximal cycle ergometry test to evaluate the potential differences in PO at LT calculated using several different methodologies (see below). In study B, the PO at LT was determined and compared with the estimated FTP for each participant.
Six trained male cyclists participated in the study in accordance with the University and Medical Center Institutional Review Board, and appropriate consent was obtained. Subject characteristics are listed in Table 1. Of these trained cyclists, 2 were of USA Cycling category 1, 2 were of category 2, and 2 were of category 3.
Seven trained male cyclists participated in the study in accordance with the University and Medical Center Institutional Review Board, and appropriate consent was obtained. Subject characteristics are listed in Table 1. Of these trained cyclists, 2 were Professional on Continental teams, 2 were of USA Cycling category 1, one was of category 2, and 2 were of category 3.
The subjects completed a ramped protocol on an electronically braked cycle ergometer (Velotron Dynafit Pro, Racermate, Seattle, WA, USA) as described by Padilla et al. (20). The Velotron is fully adjustable to replicate the position of the cyclists on their road bicycles and permits cyclists to ride at their own cadence while keeping PO constant. Lactate was measured at each stage, and oxygen consumption (V̇O2) was measured continuously. The test began at 110 W for 4 minutes, with further increments of 35 W every 4 minutes interspersed with 1 minute of active recovery at 100 W.
The subjects completed a ramped protocol on an electronically braked cycle ergometer (Velotron Dynafit Pro) as described by Klika et al. with slight modifications (15). The test began with 3 minutes at 150 W and was increased 25 W every 3 minutes until blood lactate was ≥4.0 mmol·L−1 after which the workload was decreased to 100 W for 6 minutes of active recovery. After the recovery, the workload was increased over the course of 1 minute until it was set at the final workload attained during the lactate portion of the test. Thereafter, the workload was increased 25 W every minute until volitional fatigue. Oxygen consumption (V̇O2) was measured at rest and during stages 150, 175, and 200 W for the calculation of gross and delta efficiency. After the completion of the stage at 200 W, the mouthpiece was removed and remained out until the start of the fourth of 6 minutes of the active recovery at 100 W. At the fourth minute of the 6-minute active recovery, the mouthpiece was reinserted into the mouth of the subject, and oxygen consumption was continuously measured for the remainder of the test.
V̇O2max and Lactate in Studies A and B
Minute ventilation (V̇E), oxygen uptake (V̇O2), and carbon dioxide production (V̇O2V) were monitored via open circuit spirometry (True Max 2400, Parvo Medics, Salt Lake City, UT, USA). The heart rate was measured continuously (Accurex Plus, Polar Electro Inc., Woodbury, NY, USA). The subjects were verbally encouraged to continue until volitional fatigue. The criterion used to assess maximal oxygen consumption (V̇O2max) included (a) a heart rate >90% of age-predicted max (220 − age); (b) a respiratory exchange ratio ≥1.10; and (c) identification of a plateau (≤150-ml increase) in V̇O2 despite a further increase in workload. In all tests, at least 2 of 3 criteria were met. Blood was obtained from the piercing of a finger or ear lobe and analyzed for lactate (Lactate Plus, Nova Biomedical, Waltham, MA, USA).
Dual energy x-ray absorptiometry (DEXA) was used to determine body composition (General Electric Lunar Prodigy Advanced) as previously described (17). The value for Region (%Fat) from the DEXA analysis was used for determining the body composition.
Determination of Lactate Threshold
In study A, the LT was calculated using 4 common methods to identify if PO at the LT was different during cycle ergometry testing using different methods to calculate LT, as has been previously shown for running (16). Blood lactate at each stage was plotted vs. the PO during each stage of the cycle ergometer test. The PO was then interpolated for the given measure of LT. The 4 measures of LT used were (a) LT+1, a 1 mmol·L−1 or greater increase above baseline (10); (b) LTΔ1, the point at which blood lactate values increase 1 mmol·L−1 or greater above that of the previous stage (22); (c) LTVisual, the visual determination of the point of exponential elevation by 2 independent investigators, with the average value reported (6); and (d) LT4.0, onset of blood lactate accumulation at a blood lactate concentration of 4.0 mmol·L−1 (13).
In study B, mechanical efficiency was calculated according to Coyle with slight modification (7). Briefly, gross efficiency was calculated as the average ratio of work per minute determined from PO converted to kilocalories and metabolic energy from V̇O2 and respiratory quotient converted to kilocalories from exercise at 150, 175, and 200 W. Delta efficiency is the ratio of the change in work per minute and the change in metabolic energy per minute. Delta efficiency was identified from linear regression (y = mx + b) of the relationship between metabolic energy per minute (y) vs. work per minute (x) and where metabolic rate at 0 W (b) was calculated as 3.5 ml O2·kg−1·min−1× subject body mass. Delta efficiency was calculated from the slope of the relationship and was equal to the reciprocal of m. The values for gross and delta efficiency were then converted into a percentage (%).
Within 1 week before the laboratory-based exercise testing, the subjects were instructed to perform an 8-minute FT on their own. The subjects were instructed to refrain from strenuous activity the day before the test and to complete the test generating the highest PO possible (Cycleops Powertap, Saris Cycling, Madison, WI, USA). The average PO was subsequently reported to the investigators. The subjects were not asked to control their diet before their test or the environmental conditions of their test. This approach was used because this is one common method through which FTP is currently estimated in cycling training (5). The FTP was estimated by using equation 1 (5):
A 1-way repeated measures analysis of variance (ANOVA) was used to compare PO at LT+1, LTΔ1, LTVisual, and LT4.0 in study A and to compare PO at LTΔ1, LT4.0, and FTP in study B. Following a significant F ratio, a Bonferroni post hoc analysis was used. Linear regression was performed to identify the relationships between variables. In study A, the statistical power for the repeated measures ANOVA was 1.0. In study B, statistical power for the repeated measures ANOVA was 0.917 and for the linear regression was 0.974 and 0.600 for Figure 3A, B, respectively. The interclass correlation (ICC) and coefficient of variation (CV) for LT, V̇O2max, and FTP are ICC: 92, 93, and 96%; and CV: 8, 9, and 10%, respectively. The significance was established at p ≤ 0.05 for all statistical sets, and data reported are mean ± SEM.
The PO at LT4.0 (303 ± 23 W) was significantly greater than PO at LTVisual (280 ± 15 W), LTΔ1 (268 ± 18 W), and LT+1 (250 ± 24 W) (Figure 1). The PO at LT+1 was significantly lower than PO at LTVisual. There was no difference between PO at LTΔ1 and LT+1 or between PO at LTΔ1 and LTVisual. Based on these results, the LT was calculated using the LTΔ1 and LT4.0 methods in study B as LTΔ1 and LT4.0 are significantly different from each other, whereas LTΔ1 is similar to both LT+1 and LTVisual in study A. Linear regression between the different measures of PO at LT (LT+1, LTΔ1, LTVisual, and LT4.0) found strong correlations (0.94–0.99; p < 0.01) between all the methods of LT measurement.
The PO at estimated FTP from the 8-minute FT (301 ± 13 W) was significantly greater than the PO at LTΔ1 (264 ± 9 W) but not different from the PO at LT4.0 (293 ± 9 W) (Figure 2A). There was a trend for a positive relationship between the estimated FTP and PO at LT4.0 (Figure 2B).
Next, it was questioned as to what might be the determinants of exercise performance as assessed by PO during the 8-minute FT. This initial analysis was performed using 8-minute FT PO relative to body weight to permit comparisons between cyclists of different sizes. It was reasoned that cycling performance is more related to PO relative to body mass than absolute PO. Linear regression analysis revealed a significant relationship between relative V̇O2max and relative PO during the 8-minute FT (Figure 3A). Relative V̇O2max explained 93% of the variability in PO during the 8-minute FT.
It was next questioned as to what determines the percentage of maximum that is maintained during an 8-minute FT by taking the ratio of 8-minute FT PO and PO at V̇O2max (POmax) expressed as a percentage (%POmax during the 8-minute FT). Relative V̇O2max explained 64% of the variance in the percentage of maximal PO maintained during the 8-minute FT (Figure 3B). There was no correlation between either measure of LT and %POmax during the 8-minute FT. Stepwise linear regression revealed that the addition of either measure of LT (LT4.0 or LTΔ1) did not significantly improve the explained variance in the percentage of 8-minute FT PO to PO at V̇O2max.
The principal finding of this study is that the estimated FTP derived from the average PO during an 8-minute FT is equivalent to PO at the onset of blood lactate 4.0 mmol·L−1. However, the estimated FTP is greater than the PO at LT determined using an increase in lactate of 1 mmol·L−1 or greater in response to an increase in workload. In popular publications, it has been proposed that field testing to estimate FTP provides a PO similar to PO at LT (1,5). Our results found that if the LT of LT4.0 is used, then the estimated FTP appears to approximate the PO at LT. However, if POs are obtained at an LT determined using LTΔ1 or similar criteria, then the estimated FTP appears to overestimate LT.
The FTP has been defined as the highest average PO that can be maintained for 1 hour (1). Coyle et al. found that well-trained cyclists are able to complete a 1-hour performance test at a workload of approximately 110% of workload at LT+1 (8,9). Interestingly, during a successful 1-hour world record attempt, the athlete maintained an estimated PO of 101% of LT4.0 (18). In time trials (TTs) lasting approximately 1 hour, professional cyclists on average maintain an estimated PO of 90% of LT4.0 (19). By comparison, trained competitive time trialists and triathletes can maintain 100% of LT4.0 for 90 minutes (4). Thus, the current finding that FTP is equivalent to PO at LT4.0 (3% difference) but greater than LTΔ1 (14% difference) is consistent with the findings of previous literature.
One popular coaching publication recommends their athletes regularly perform an 8-minute FT to assess current fitness: The test results are used to assess current training status and to design future training programs (5). In this study, the average PO during the self-paced 8-minute FT was used as an index of exercise performance. It was found that relative V̇O2max explained 93% of the variance in PO per kilogram body mass during the 8-minute FT. It should be noted that in this study, relative PO from the FT was used to permit the evaluation between athletes of different body mass, which we believe is the most appropriate approach to performance evaluation. Consistent with this underlying assumption, Balmer et al. found that absolute maximal PO from a maximal exercise test explains only 21% of the variance in the performance of an outdoor 16.1-km TT (average time to completion 22:34 minutes:seconds) (3). Yet, our reanalysis of the data from Balmer et al. (3) dividing absolute PO by body mass to determine relative PO found that 41% of the variance in TT performance could be explained by relative PO, supporting the use of relative POs in the current report.
Several reports including the current report have found close positive associations between PO during a cycling TT effort and V̇O2max (3,4,8,9). Although adenosine triphosphate (ATP) can be generated from anaerobic sources, it should come as no surprise that V̇O2max is closely associated with relatively short duration maximal aerobic exercise because oxygen is the final electron acceptor in mitochondrial ATP production. However, in the current report, we also questioned as to what determines individual submaximal exercise performance, that is, what determines the percentage of maximum that an individual maintains during short-term submaximal aerobic exercise. Individual exercise performance was calculated by taking PO from the 8-minute FT and dividing by maximal PO from the V̇O2max test (POmax) to derive the percentage of maximum PO at which the subjects completed the 8-minute FT (%POmax during the 8-minute FT). Interestingly, relative V̇O2max still explained 64% of the variance in %POmax during the 8-minute FT (Figure 3B) suggesting that the factors associated with a high relative V̇O2max also contribute to the ability of an individual to maintain a high PO relative to maximum during a short duration FT.
It has been proposed that the LT is an important determinant of individual endurance performance (8,9). Interestingly, the addition of either measure of LT (LT4.0 or LTΔ1) did not significantly improve the explained variance in the %POmax during the 8-minute FT suggesting that V̇O2max is a better predictor of the percentage of maximum PO that can be maintained for an 8-minute maximal effort than is LT. Consistent with this, V̇O2max is a better predictor of PO during a 20 minute TT, whereas LT is a better predictor of PO during a 90-minute TT (4). One potential reason for the discrepancy is the duration of the individual submaximal aerobic exercise tests performed: 8 minutes vs. 30–60 minutes. It might be questioned if estimating 60-minute PO from an 8-minute FT is equivalent to PO during a 60-minute effort. Additional work is required to answer the question as to what physiological factor(s) limit individual submaximal exercise performance of different durations.
In summary, we have demonstrated in trained cyclists that the FTP estimated from an 8-minute FT (90% of 8-minute FT PO) is equivalent to PO at LT when onset blood lactate accumulation 4.0 mmol·L−1 is used to determine the LT. When measures of LT from which PO is less than onset blood lactate accumulation are used (LTΔ1), this is not true. It was also found that V̇O2max is an important predictor of both PO during an 8-minute maximal effort test and the percentage of maximal PO maintained during the 8-minute FT. Whether the estimates of FTP based on field based testing are equivalent to measured FTP from a 1 hour maximal effort test remain to be determined.
We investigated claims that FTP, a new training term that can be estimated from a field-based test, is equivalent to PO at LT. We found that estimated FTP was equivalent to LT when LT was determined as the onset of blood lactate at 4.0 mmol·L−1. This finding provides scientific evidence to support popular claims and the framework with which to compare field-based testing with scientific physiological concepts. Previous research has identified LT as a key determinant of endurance athletic performance and identified training methods that improve LT. Knowing that FTP is equivalent to power at LT allows for the application of this body of knowledge to FTP and makes it convenient and cost effective for coaches and athletes to confidently evaluate racing and training performance throughout the season.
The authors wish to thank Gabriel Geyer and Kathryn Wilson for their assistance.
1. Allen, H and Coggan, A. Training and Racing with a Power Meter
. Boulder, CO: VeloPress, 2010.
2. Astrand, P-O and Rodahl, K. Textbook of Work Physiology
. New York, NY: McGraw-Hill, 1986.
3. Balmer, J, Davison, RC, and Bird, SR. Peak power predicts performance power during an outdoor 16.1-km cycling
time trial. Med Sci Sports Exerc
32: 1485–1490, 2000.
4. Bentley, DJ, McNaughton, LR, Thompson, D, Vleck, VE, and Batterham, AM. Peak power output, the lactate threshold
, and time trial performance in cyclists. Med Sci Sports Exerc
33: 2077–2081, 2001.
5. Carmichael, C and Rutberg, J. The Time-Crunched Cyclist: Fit, Fast, and Powerful in 6 Hours a Week
. Boulder, CO: VeloPress, 2009.
6. Cheng, B, Kuipers, H, Snyder, AC, Keizer, HA, Jeukendrup, A, and Hesselink, M. A new approach for the determination of ventilatory and lactate thresholds. Int J Sports Med
13: 518–522, 1992.
7. Coyle, EF. Improved muscular efficiency displayed as Tour de France champion matures. J Appl Physiol
98: 2191–2196, 2005.
8. Coyle, EF, Coggan, AR, Hopper, MK, and Walters, TJ. Determinants of endurance in well-trained cyclists. J Appl Physiol
64: 2622–2630, 1988.
9. Coyle, EF, Feltner, ME, Kautz, SA, Hamilton, MT, Montain, SJ, Baylor, AM, Abraham, LD, and Petrek, GW. Physiological and biomechanical factors associated with elite endurance cycling
performance. Med Sci Sports Exerc
23: 93–107, 1991.
10. Coyle, EF, Martin, WH, Ehsani, AA, Hagberg, JM, Bloomfield, SA, Sinacore, DR, and Holloszy, JO. Blood lactate threshold
in some well-trained ischemic heart disease patients. J Appl Physiol
54: 18–23, 1983.
11. Donovan, CM and Brooks, GA. Endurance training affects lactate clearance, not lactate production. Am J Physiol
244: E83–E92, 1983.
12. Farrell, PA, Wilmore, JH, Coyle, EF, Billing, JE, and Costill, DL. Plasma lactate accumulation and distance running performance. Med Sci Sports
11: 338–344, 1979.
13. Heck, H, Mader, A, Hess, G, Mucke, S, Muller, R, and Hollmann, W. Justification of the 4-mmol/l lactate threshold
. Int J Sports Med
6: 117–130, 1985.
14. Karlsson, J and Jacobs, I. Onset of blood lactate accumulation during muscular exercise as a threshold concept. I. Theoretical considerations. Int J Sports Med
3: 190–201, 1982.
15. Klika, RJ, Alderdice, MS, Kvale, JJ, and Kearney, JT. Efficacy of cycling
training based on a power field test. J Strength Cond Res
21: 265–269, 2007.
16. McGehee, JC, Tanner, CJ, and Houmard, JA. A comparison of methods for estimating the lactate threshold
. J Strength Cond Res
19: 553–558, 2005.
17. Ormsbee, MJ, Choi, MD, Medlin, JK, Geyer, GH, Trantham, LH, Dubis, GS, and Hickner, RC. Regulation of fat metabolism during resistance exercise in sedentary lean and obese men. J Appl Physiol
106: 1529–1537, 2009.
18. Padilla, S, Mujika, I, Angulo, F, and Goiriena, JJ. Scientific approach to the 1-h cycling
world record: A case study. J Appl Physiol
89: 1522–1527, 2000.
19. Padilla, S, Mujika, I, Orbananos, J, and Angulo, F. Exercise intensity during competition time trials in professional road cycling
. Med Sci Sports Exerc
32: 850–856, 2000.
20. Padilla, S, Mujika, I, Orbananos, J, Santisteban, J, Angulo, F, and Jose Goiriena, J. Exercise intensity and load during mass-start stage races in professional road cycling
. Med Sci Sports Exerc
33: 796–802, 2001.
21. Sjodin, B, Jacobs, I, and Svedenhag, J. Changes in onset of blood lactate accumulation (OBLA) and muscle enzymes after training at OBLA. Eur J Appl Physiol Occup Physiol
49: 45–57, 1982.
22. Thoden, J. Testing aerobic power. In: Physiological Testing of the High-Performance Athlete
. J. MacDougal, J. Wenger, and H. Green, eds. Champaign, IL: Human Kinetics Books, 1991. pp. 107–174.