In continuous exercise there is a hyperbolic relationship between power output and time to exhaustion. The power asymptote of this relationship is called critical power (CP) and defined as the maximum rate that a muscle “can keep up for a very long time without fatigue” (10). Theoretically, if exercise is sustained for a long time without fatigue (i.e., at the CP intensity), rating of perceived exertion (RPE) and physiological variables such as oxygen consumption (V̇o2), blood lactate concentration ([Lac]), heart rate (HR) should stabilize. In fact, Poole et al. (11) described CP as the uppermost continuous exercise intensity associated with V̇o2, [Lac] and blood pH stabilization throughout exercise. However, some studies have reported that these physiological variables do not stabilize during continuous exercises performed at the CP (6,12).
Despite this observation, it is not known whether exercising at the intermittent critical power (CPi) is associated with the uppermost physiological steady state during high-intensity intermittent exercise. It has been reported that intermittent exercise can induce lower [Lac] and greater exercise tolerance compared to a continuous protocol when performed at a similar exercise intensity (3). Furthermore, Turner et al. (14) observed that physiological responses in intermittent exercise are similar to those reported for continuous exercise but show different exercise domain characteristics (i.e., moderate, heavy, and severe exercise domains). Thus, we hypothesized that exercising at CPi may allow greater muscular power engagement to be sustained for a long time compared with continuous exercise, while maintaining a physiological steady state. Determining these responses when exercising at CPi is important considering that intermittent exercise has many variables that can be manipulated, such as the length of the periods, intensity of effort, and active or passive recovery. Astrand et al. (1) showed that with the same average power output during intermittent exercise ([effort work + pause work]/total time) but with different effort and pause durations (10:20 seconds and 60:120 seconds) induced different [Lac] response. These results indicated that average power output in intermittent exercise is not a reliable parameter for exercise prescription, and depending on how they are handled, the physiological responses can be different. Thus, if the physiological steady state at CPi is confirmed, this aerobic indexes may be used as a valuable tool for intermittent exercise evaluation and prescription.
Therefore, this study aimed to compare (a) the power output of CPi with both the CP derived from continuous predictive trials (CPc) and with the intermittent maximal lactate steady state (MLSSi) and (b) the physiological (V̇o2, [Lac] and HR) and perceptual (RPE) responses to exercise performed at CPi and MLSSi intensities.
Experimental Approach to the Problem
The study was divided into 5 phases in the following order: (a) 2 familiarization trials; (b) 1 graded test to determine V̇o2max; (c) 4 continuous predictive trials to determine CPc; (d) 4 intermittent predictive trials to determine CPi; and (e) 2 or 3 tests for MLSSi determination, starting at CPi intensity. The mean power outputs corresponding to CPc, CPi, and MLSSi were compared and correlated among them and correlated with aerobic parameters (V̇o2max, MAP). Additionally, the V̇O2, [Lac] and HR and RPE responses throughout exercise at CPi and MLSSi were compared. These analyses can give indications that CPi is an aerobic parameter specific to intermittent exercise.
Ten male college students (24.4 ± 3.7 years; 76.5 ± 11.7 kg; 1.77 ± 0.04 m) volunteered for this study. The sample size was determined previously by G*Power software (v.3.0.10) assuming α = 0.05 and β = 0.20 and based on a similar study (4). Subjects were asked to refrain from severe physical activity 24 hours before the tests. This study was approved by the local Ethics Committee of Human Studies, and subjects were informed about the procedures and risks before giving written consent.
The subjects performed 2 familiarization trials of high-intensity exercise on a cycle ergometer (Biotec 2100, Cefise, Campinas, SP, Brazil) until exhaustion to avoid a learning effect and thus increase the validity of the predictive trials. Preceding every trial, subjects warmed up by cycling at 30 W for 5 minutes, followed by 3 minutes of passive recovery. The pedal cadence was fixed at 60 rpm, and the power outputs were individually chosen to induce exhaustion within 2-15 minutes (12). All tests were performed on different days (24-72 hours), at approximately the same time of the day (±2 hours), at least 3 hours postprandial, and at a room temperature ranging from 20 to 24°C.
Maximal Graded Test
This session was performed to determine both V̇o2max and maximal aerobic power (MAP). Pulmonary gas exchange was assessed in the breath-by-breath mode (MetaLyzer 3B, Cortex, Leipzig, Germany), with calibration performed using ambient air and gas of known O2 (16%) and CO2 (5%) concentrations. The turbine flowmeter was calibrated using a 3-L syringe. The HR was recorded using a Polar S810i (Electro Oy, Kempele, Finland). The subjects started the graded test cycling at 30 W, and the power output was increased by 30 W each minute until the subject could no longer maintain the pedal cadence of 60 rpm despite verbal encouragement. The highest 30-second average for V̇o2 data during the last stage was taken as the V̇o2max. At least 2 of the following criteria were required to guarantee that V̇o2max had been attained: (a) increase of V̇o2 <150 ml·min−1 between 2 successive stages, (b) a respiratory exchange ratio value >1.10, and/or (c) an HR in excess of 90% of age-predicted maximum. The MAP was derived from the following equation:
Predictive Continuous and Intermittent Trials
The subjects performed 4 continuous predictive trials in a random order to estimate CPc. The power outputs led to exhaustion between 2 and 15 minutes of exercise. Exhaustion was considered when the subject could no longer maintain the pedal cadence of 60 rpm despite verbal encouragement. Subsequently, on different days, 4 intermittent predictive tests with different workloads were performed with 30-second exercise bouts and 30-second passive recovery periods, in which time to exhaustion occurred between 2 and 15 minutes of exercise. These workloads were chosen based on the results of an earlier pilot study (∼110, 120, 130, and 140% of MAP), to estimate CPi. Subjects did not access information about the power output or elapsed time during the tests. For modeling purposes, recovery periods were not included (4).
Continuous and Intermittent Power-Time Modeling
The continuous and intermittent performance data were fitted by means of nonlinear regression using the 2-parameter hyperbolic power-time equation:
where AWC represents anaerobic work capacity, and CP represents critical power. These parameters were estimated for both continuous (CPc and AWCc) and intermittent (CPi and AWCi) trials.
Intermittent Maximal Lactate Steady State
The subjects performed tests starting at CPi intensity, alternating 30 seconds of exercise bouts with the same period of passive recovery. The V̇o2 was measured throughout the session. Additionally, [Lac], RPE (using a 15-point Borg scale) (5), and HR were measured every 5 minutes. The uppermost duration of each test was 30 minutes. When [Lac] steady state was observed, subsequent tests were performed on different days with 10% higher power outputs until no stabilization was observed (2). The MLSSi was considered as an increase of [Lac] <1 mM between the 10th and the 30th minutes (2). When participants exercising at CPi (i.e., the first intensity used to determine the MLSSi) presented a nonsteady blood lactate profile, power outputs were 10% decreased to MLSSi occurrence. If at CPi the [Lac] stabilization did not occur, the power output was subsequently reduced by 10% until MLSSi criteria was fulfilled. Blood samples from the earlobe were collected using heparinized capillary tubes and subsequently placed in tubes containing NaF (1%). Blood lactate concentration was determined using an electrochemical device (YSI 2300 Select,Yellow Springs, OH, USA).
The Gaussian distribution of data was attested by Kolmogorov-Smirnov's test (with Lilliefor's correction). A Student t test was used to compare the AWCc and AWCi. Analysis of variance (ANOVA) with repeated measures, followed by a Scheffé post hoc test, was used to verify the differences among the power output of CPc, CPi, and MLSSi. Two-way ANOVAs were adopted to compare V̇o2, [Lac] and HR and RPE responses throughout the test at the CPi and MLSSi intensity. The repeated measures data were checked for sphericity using Mauchly's test, and whenever the test was violated, we performed the necessary technical corrections through the Greenhouse-Geisser test. Pearson correlation was performed to assess the relationship between the variables. The significance level was set at 5% (p ≤ 0.05). Data are presented as mean ± SD.
During the graded test, mean MAP was 284 ± 29 W, with a corresponding V̇o2max of 3.12 ± 0.37 L·min−1 and a maximal HR of 184 ± 11 b·min−1. The average power output and time to exhaustion of the 4 predictive trials performed during continuous and intermittent settings are presented in Figure 1. The coefficient of determination was 0.981 ± 0.018 for the CPc and AWCc parameters and 0.962 ± 0.055 for the CPi and AWCi parameters.
The CPc (151 ± 30 W) was significantly lower than CPi (267 ± 45 W) and MLSSi (254 ± 39 W) (p < 0.001). However, there was no significant difference between CPi and MLSSi (p > 0.05). The CPi was significantly correlated with both MLSSi (r = 0.88; p < 0.001) and CPc (r = 0.79; p < 0.05). Moreover, CPi and MAP (r = 0.75; p < 0.05), and CPi and V̇o2max (r = 0.76; p < 0.05) pairs also presented significant correlations. The AWCc (23.9 ± 3.0 kJ) and AWCi (27.3 ± 8.6 kJ) were similar (p > 0.05), and there was significant correlation between them (r = 0.62; p = 0.05).
The V̇o2 stabilized throughout the CPi and MLSSi exercises. There was no significant increase in V̇o2 from the fifth (CPi = 2.36 ± 0.44 L·min−1; MLSSi = 2.29 ± 0.42 L·min−1) to the 30th minute (CPi = 2.52 ± 0.52 L·min−1; MLSSi = 2.41 ± 0.32 L·min−1) (Figure 2A). Blood lactate concentration at the end of exercise at CPi (6.9 ± 2.6 mM) was not statistically higher than at the 10th minute (5.7 ± 1.0 mM), but the difference was >1 mM. Also, during the MLSSi test, the [Lac] was similar at the 10th (4.8 ± 0.7 mM) and the 30th minutes (5.1 ± 0.9 mM), with a difference <1 mM, as required by the protocol. When comparing the [Lac] responses between CPi and MLSSi test, there were significant differences after the 15th minute (Figure 2B). The individual curves at CPi and MLSSi are presented in Figure 3. The HR at the 30th minute (156 ± 8 b·min−1) was significantly higher than the 10th minute (146 ± 11 b·min−1) at CPi. During the MLSSi, the HR at the 30th minutes of exercise (152 ± 10 b·min−1) was significantly higher than the value at the fifth minute only (141 ± 13 b·min−1) (Figure 2C). The mean RPE value during the CPi and MLSSi was significantly higher at the 30th minute (CPi = 17.1 ± 2.1 a.u.; MLSSi = 15.7 ± 1.8 a.u.) compared with the 10th minute (CPi = 15.3 ± 1.6 a.u.; MLSSi = 14.3 ± 1.6 a.u.). Furthermore, there were significant differences for the RPE between CPi and MLSSi after the 20th minute (Figure 2D).
This study aimed to compare CPi with MLSSi intensities, and the physiological and perceived exertion responses during cycling at their respective power outputs. The results indicated that CPi and MLSSi were similar. Despite the absence of significant differences from the 15th to 30th minutes of exercise, the [Lac] and RPE were higher in the second half of CPi exercise duration when compared to MLSSi, whereas V̇o2 and HR were similar between them.
The CPi was higher than CPc, as previously reported (3). Additionally, CPi was significantly correlated with MAP (r = 0.75), V̇o2max (r = 0.76) and CPc (r = 0.79). Then, CPi seems to be an aerobic fitness index, as CPc is related to aerobic performances (13). The higher CPi when compared with CPc was probably observed because of the mathematical procedures adopted to estimate the CPi parameters, which excludes the recovery duration from the modeling (4).
Based on the idea that CP represents an intensity that the muscles can sustain for a “long time without fatigue” (10), we hypothesized that physiological responses would stabilize exercising at CPi. In fact, we observed that CPi was similar to the power output at MLSSi, with a high correlation (r = 0.88) between them. Additionally, the physiological responses did not increase significantly after 10-15 minutes of exercise for both CPi and MLSSi intensities (Figure 2). The [Lac] rose rapidly between 5 and 10 minutes of exercise at CPi and thereafter did not increase significantly during the last 20 minutes of exercise. However, despite a similar [Lac] response in both trials, after the 15th minute of exercise, it was greater during exercise at CPi when compared with exercise at the MLSSi. This can be attributed to the observation that 5 of 10 subjects did not stabilize the [Lac] at CPi, according to Beneke's criteria (1) (Figure 3A). Therefore, our results suggest that although CPi and MLSSi did not differ significantly, CPi overestimates the MLSSi in some individuals.
During both the CPi and MLSSi trials, there was a V̇o2 steady state throughout the exercise (80 and 78% of V̇o2max, respectively). Poole et al. (11) reported a similar steady-state value (79% of V̇o2max) during 24 minutes of exercise at the CPc. In both trials, HR steady state was achieved (with no significant increase in HR after the 10th and 15th minutes in the MLSSi and CPi trials, respectively), and there was no significant difference between trials whatever the instant (Figure 2). The HR increase after V̇o2 stabilization can probably be attributed to a decreased venous return, with a subsequent decrease in ventricular filling and stroke volume, in response to skin vasodilation and dehydration (8). A steady state for the RPE was also achieved after the 15th minutes during both trials. However, despite the plateau in both trials, the RPE after the 20th minutes was significantly higher exercising at CPi when compared with exercising at MLSSi. This may reflect the small difference in the power outputs of CPi (267 ± 45 W) and MLSSi (254 ± 39 W), which despite being nonsignificant, can be perceived as a result of corollary discharges associated with central motor drive (9). The similarities of physiological profiles between CP and maximal lactate steady state were only observed previously in continuous exercise (6,11,12). To our knowledge, this is the first study to observe this close correspondence in intermittent exercise.
In intermittent exercise, there are many variables that can be manipulated and the range of combination affects the physiological responses differently. Thus, understanding the physiological responses to intermittent exercise performed at CPi can be a tool for the training prescription. Furthermore, it is important to consider that intermittent exercise may promote different adaptations when compared to continuous one (7). More studies are necessary to verify the effects of training at CPi, or at exercise intensities relative to CPi.
According to our results, CPi is the upper limit of blood lactate and possibly V̇o2 steady state. Hence, it is not suitable to prescribe training to V̇o2max improvement, but rather to increase aerobic capacity with higher muscular engagement compared with CPc. On the other hand, the intensities above CPi allows for maintenance of V̇o2max with predictable tolerance time as estimated by the individual power-time curves. Thus, it seems an important reference for scheduling interval training, with already attested applicability in the field (4).
Based on the present results, we suggest that the CPi may allow physical trainers to determine performance in intermittent protocols and the subsequent individualization for training prescription. Furthermore, the CPi may be used as the training intensity regarded to physiological steady state in intermittent exercise sessions. Because scheduling training sessions based on continuous MAP or maximal aerobic velocity fails to induce homogeneous physiological responses during intermittent training sessions, we consider the CPi as a useful index to training evaluation and prescription.
The authors are grateful to Fundação Araucária de Apoio ao Desenvolvimento Científico e Tecnológico do Paraná and Conselho Nacional de Desenvolvimento Científico e Tecnológico for the financial support.
1. Astrand, I, Astrand, PO, Christensen, EH, and Hedman, R. Intermittent muscular work. Acta Physiol Scand
48: 448-453, 1960.
2. Beneke, R. Maximal lactate steady state concentration (MLSS): Experimental and modelling approaches. Eur J Appl Physiol
88: 361-369, 2003.
3. Beneke, R, Hutler, M, Von Duvillard, SP, Sellens, M, and Leithauser, RM. Effect of test interruptions on blood lactate during constant workload testing. Med Sci Sports Exerc
35: 1626-1630, 2003.
4. Berthoin, S, Baquet, G, Dupont, G, and Van Praagh, E. Critical velocity during continuous and intermittent exercise in children. Eur J Appl Physiol
98: 132-138, 2006.
5. Borg, GA. Psychophysical bases of perceived exertion. Med Sci Sports Exerc
14: 377-381, 1982.
6. Brickley, G, Doust, J, and Williams, CA. Physiological responses during exercise to exhaustion at critical power. Eur J Appl Physiol
88: 146-151, 2002.
7. Helgerud, J, Hoydal, K, Wang, E, Karlsen, T, Berg, P, Bjerkaas, M, Simonsen, T, Helgesen, C, Hjorth, N, Bach, R, and Hoff, J. Aerobic high-intensity intervals improve V̇O2
max more than moderate training. Med Sci Sports Exerc
39: 665-671, 2007.
8. Lonsdorfer-Wolf, E, Richard, R, Doutreleau, S, Billat, VL, Oswald-Mammosser, M, and Lonsdorfer, J. Pulmonary hemodynamics during a strenuous intermittent exercise in healthy subjects. Med Sci Sports Exerc
35: 1866-1874, 2003.
9. Marcora, S. Is peripheral locomotor muscle fatigue during endurance exercise a variable carefully regulated by a negative feedback system? J Physiol
586: 2027-2028, 2008.
10. Monod, H and Scherrer, J. The work capacity of a synergic muscular group. Ergonomics
8: 329-338, 1965.
11. Poole, DC, Ward, SA, Gardner, GW, and Whipp, BJ. Metabolic and respiratory profile of the upper limit for prolonged exercise in man. Ergonomics
31: 1265-1279, 1988.
12. Pringle, JS and Jones, AM. Maximal lactate steady state, critical power and EMG during cycling. Eur J Appl Physiol
88: 214-226, 2002.
13. Smith, JC, Dangelmaier, BS, and Hill, DW. Critical power is related to cycling time trial performance. Int J Sports Med
20: 374-388, 1999.
14. Turner, AP, Cathcart, AJ, Parker, ME, Butterworth, C, Wilson, J, and Ward, SA. Oxygen uptake and muscle desaturation kinetics during intermittent cycling. Med Sci Sports Exerc
38: 492-503, 2006.