THS THEBlS 300 I LIBRARY Michigan State University w ' This is to certify that the thesis entitled PROTOCOL DEPENDENCY OF THE BREAKPOINT IN THE VOz-WATT RELATIONSHIP presented by Peter J. Osmond has been accepted towards fulfillment of the requirements for M.S. degree in Kinesiology Major professor Date 2/9/0700] 0-7 639 MS U is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requeSted. DATE DUE DATE DUE DATE DUE 6/01 cJClRC/DateDuepG5-p. 15 PROTOCOL DEPENDENCY OF THE BREAKPOINT IN THE VOz-WATT RELATIONSHIP By Peter J. Osmond A THESIS Submitted to Michigan State University In partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of IGnesiology 2001 ABSTRACT PROTOCOL DEPENDENCY OF THE BREAKPOINT IN THE VOz-WATT RELATIONSHIP By Peter J. Osmond The purpose of this study was to determine whether the breakpoint in the VOz-WBtt relationship (V02ap) observed during incremental exercise testing is protocol dependent. Specifically, we attempted to determine if alterations in stage duration engendered quantifiable changes on the V02». In addition, we assessed lactate thresholds (Us) and ventilatory thresholds (VT s) to determine values of these variables relative to the occurrence of the V023p. Sixteen subjects (mean age (SD) 26.8 :I: 4.2 years, mean peak 02 uptake (SD) ~3.1 :l: .526 Uminute) who had no known cardiovascular disease participated in this study. All subjects performed three maximal exercise tests. These tests included incremental protocols of 1-, 3-, and 5-minute stages. All testing was performed on a cycle ergometer. Potential differences in the V023p, LT, and VT between the three protocols were determined using a two-way, repeated measures analysis of variance. The main findings in this study were threefold: ( 1) there was a disparity in the relationship between the V023.» and the LT, (2) there was a significant difference between LT and VT for all protocols, and (3) the power outputs at V023p, LT, and VT were higher during the 1-minute stage protocols than the 3- and 5-minute stage protocols. TABLE OF CONTENTS KEY TO ABBREVIATIONS...... .. INTRODUCTION... Overview of the Problem Need for the Study... Statement ofthe proaem...31:33::::::::::::::::::1::1: 1.1.3.: I: I. I. I. .12.. vi .vii Research Questions...... Hypothesis Statement... REVIEW OF LITERATURE... Characteristics etmevoz-watt Relationship fflffiffiffiff'...’.fflffifflfflffiff.. Significance of the V02 Slow Component... Protocol Effect on the LT and VT... Assessing Maximal Steady State (MSS) Summary... tomNmVN (batman-*4 Subjects... Testing... Lactate Sampling and Measurement Methods... Gas Exchange Measurements. Heart Rate Analysis ofVT............ Analysis ofLT.............. Coefficient ofVariation......... Analysis OfVOzBp...... Statistical Analysis......... iii sseee 8833338 wwN 0k)“ LIST OF TABLES Table1. DescriptiveDatabyGenderforAllSubjects.................................21 Table2. CoefficientsofVariation...........................................................25 Table 3. POM“, Peak Stage, and Time to Exhaustion for Each Subject by Protocol. . .. .. ... .....27 Table 4. Mean PC at Thresholds and Subject Characteristics... ...28 Table 5. Comparison of mean PC at LT, “VO233.W and POM..." for sow/3min Stage Protocols... .... . . ... ... .33 LIST OF FIGURES Figure 1. An example of the V0233, indicated by the arrow. The linear regression is based on data from the below MV0233" data (Zoladz Szkutnik, Majerczak & Duda, 1998)... . ... ......2 Figure 2. Critical power, represented by” the dashed line. (Gaesser, Camevale, Garfinkel, Walter, &Womack, 1995)... . ... .......17 Figure 3. The V02 response to different constant-power exercise bouts. The shaded regions represent V02, above which is predicted from extrapolation of the below LT V02” response (i. e. HVan) LT Is indicated ”by”. the arrow. ”(Gaesser & Poole, 1996). . . .. . . .. .......20 Figure 4. Blood lactate analysis. LT was defined as a 0.5 mmollL increase. In the case of subject 8, 5-minute stage protocol LT was determined to be at 120 Figure 5. A plot of the rolling sum of the error of V02 (actual vs. estimated) against power output The abrupt change In ”V02“ at 150 watts represents the V0233... . ......25 Figure 6. Mean peak V02 for all subjects by protocol.... ....29 Figure 7. Mean P033... for all subjects by protocol... ......29 Figure 8. Mean power at V0233, LT, and VTfor all subjects... .....30 Figure 9. Mean V02at V0233, LT, and VTfor all subjects... ....30 BTPS HR LT M38 P ETC02 P5702 POmax STPD VC02 VE V02 V0299 V0233»: VOW V033 H+ KEY TO ABBREVIATIONS Body temperature and pressure saturated Heart rate Lactate threshold Maximal steady state End tidal carbon dioxide End tidal oxygen Maximum power output Respiratory Exchange Ratio Standard temperature and pressure dry Volume of carbon dioxide expired Minute ventilation Volume of oxygen consumed Breakpoint in the V02-watt relationship Maximal oxygen consumption Peak oxygen consumption Slow component of V02 kinetics Ventilatory threshold Hydrogen ion DEFINITIONS Critical power. The theoretical power output for an individual that could be maintained indefinitely, as it represents the upper limit for sustainable exercise. Lactate threshold: The power output, during incremental exercise, at which the first abrupt change in blood lactate concentration occurs (minimum 0.5 mmollL) and after which continues to increase with increasing power output. Maximal steady state: The highest power output at which an individual can perform exercise while maintaining stable V02. Slow component of V02 kinetics: The second non-linear component of the V02- watt relationship. It is calculated by the difference between actual V02 and that which is expected based upon linear regression analysis of the below-lactate threshold V02 response. Ventilatory threshold: The power output at which there is an increase in V3NO2 without an increase in V3NC02. V02—watt breakpoint: The power output just prior to where the VO2-watt relationship becomes non-linear. This point, which occurs above the lactate threshold, is believed to be indicative of the slow component of V02 kinetics. INTRODUCTION Qvgfligw of the Problem The V02-watt relationship reflects the pattern of oxygen utilization during exercise relative to specific power outputs on a cycle ergometer. It is therefore also reflective of the energy requirements necessary for an individual to perform at a given work-rate. During continuous incremental exercise on a cycle ergometer, this relationship can be observed through an entire range of intensities up to an individual’s maximum exercise capacity. In healthy individuals, the VO2-watt relationship dun’ng incremental exercise has historically been perceived as linear (Astrand & Rodahl, 1986). This notion, however, only holds true for exercise intensities below the lactate threshold (LT). Exercising at an intensity that induces a sustained increase in blood lactate concentration is accompanied by a second, slower rise in V02 (Zoladz, Rademaker, 8. Sergeant, 1995). This second component is termed the slow component of V02 kinetics (VON), which can be defined as, and calculated by, the difference between actual V02 and that which is expected based on linear regression of the below LT V02 response. The use of linear regression techniques reveal a linear slope in the V02-watt relationship below the LT, followed by a non-linear component at intensities that are above the LT. However, it has not been determined whether the slow component itself conforms to an exponential or linear model. In fact, some evidence suggests that V02BC kinetics follow both exponential and linear trajectories, although there is no agreement as to the exact process (Barstow & Mole, 1991). V02 kinetics above the LT demonstrate a diminution of the step pattern that is typically observed during incremental exercise. This departure from linearity reflects an increase in oxygen utilization, and thus an additional energy requirement that is above that predicted from an extrapolation of the below-lactate threshold VO2-watt relationship. The power output just prior to where the VO2-watt relationship goes from being linear to being non-linear is defined as the V02 breakpoint (V0233)(see Figure 1). 5000 4500 4000 'Eesoo 33000 .ggsoo «2000 c>’1500 1000 500 0 Power Output (watts) Figure 1. An example of the V033, indicated by the arrow. The linear regression Is based on data firm the below V0233 data. Exercise that elicits a significant V02.0 will be limited in duration, because it represents non-steady-state oxygen consumption (Poole, Ward, Gardner, & Whipp, 1988). The VO2-watt breakpoint observed during constant load exercise Whipp, 1988). The V02-watt breakpoint observed during constant load exercise is presumed to be a result of the slow component of V02 (Zoladz, Szkutnik, Majerczak, & Duda, 1998). This is important since it can be used to predict the maximal V02 steady state (MSS)—the point above which the slow component occurs. Despite some research in this area and the fact that the slow component is only detectable above the LT, investigators have been unable to demonstrate a causal relationship between blood lactate concentration and V02 kinetics. Roston, Whipp, Cunningham, Effros, and Wassennan (1987), for example, showed the magnitude of the V02” to be highly correlated with increases in blood lactate levels. Similarly Zoladz et al. (1998) found that the LT occurred just prior to or at the same power output as the V0233 in 23 of 26 subjects performing on a cycle ergometer. These studies suggest that the V0233 may be concurrent with the LT. However, data from Poole, Gladden, Kurdak, and Hogan (1994) suggest that although there is a relationship between LT and the slow component it is not causal. Support for this also comes from the work of Womack et al. (1995), who found that the diminution of the V02.c that occurred as a result of a 6-week cycle ergometry training program was coincident with reductions in blood lactate concentration. However, subsequent epinephrine infusion, which increased blood lactate concentration, did not increase exercise V02. Previous research suggests that both the LT and the VT are altered in response to stage increment size and duration during exercise testing (Davis et al., 1982; Stockhausen, Grathwohl, Burklin, Spranz, & Keul, 1997). Based on these studies, and assuming that the relationship between blood lactate and V02 is causal, one would expect that as LT and VT respond to changes in stage duration, the V0233 would respond In a similar manner. Furthermore, the influence of the V02ac itself has been reported as being dependent upon stage duration. That is, as stage duration increases, the more apparent the V02,, becomes (Zhang, Johnson, Chow, & Wesserman, 1991 ), and it may be undetectable for exercise involving fest work-rate increments (Whipp, 1994). A possible reason for this may be that the fast work-rate increments do not allow enough time at a particular power output for a steady state in V02 to occur, and therefore the VOz-watt relationship may appear linear in the domain of heavy to severe exercise intensity. Therefore, long stage durations may be more appropriate for elucidating the V0233. To determine the most appropriate stage duration for detecting the V0233, subjects must perform exercise protocols of different stage durations so to commre the V02 and the power output at which it occurs for each test. Also, LT should be assessed to determine if changes in stage duration affect its occurrence relative to the V0233, V02, end/or power output. This will provide insight into the relationship between lactate concentration and V02. N f r t The significance of the slow component is evidenced by its practical and clinical applications. It has been shown that a reduction in the V02.c as a result of exercise training was concurrent with improvements in walking economy in patients with peripheral arterial occlusive disease (Womack, Sieminski, Ketzel, Yetaco, & Gardner, 1997). This reduction in the magnitude of the V02” is regarded as a positive adaptation to exercise training. Therefore, the V022 can be used as e submaximel indicator of improvements in exercise tolerance. This is important since many clinical populations have difficulty in reaching the V023” criterion due to their pathologies. Furthermore, since the V0233 is believed to indicate the point at which the V023, occurs, it can be used as a marker for assessing improvements in exercise tolerance based on the power output at which it occurs (Zoladz et al., 1998). Despite this apparent significance, the specific mechanisms that cause the V02» are largely unknown. Typically, researchers have used the LT as a critical value, above which it is assumed the slow component occurs. From the data of Poole et al. (1994), evidence now exists that the slow component and blood lactate concentration are not causally related. They found that infusing lactate directly into a working dog muscle did not increase V02. Based on this, LT may not be the most appropriate indicator of the V02», and therefore it is important to determine the best methodology for elucidating the V0233. In addition, it is also important to investigate how the relationships between the V0233 and LT and VT are impacted by protocol design. Currently, no study has been performed specifically for the purpose of investigating the influence of stage duration on the V0233. e r em The purpose of this study was to determine if the breakpoint in the V02- watt relationship is affected by stage duration during incremental exercise on a cycle ergometer. e r uestion This study was designed to answer the following questions: (a) will the V02-watt breakpoint be affected by alterations in stage duration, and (b) do these changes in protocol affect the lactate and/or ventilatory thresholds relative to the V0233? H h sis ement We hypothesized that the V0233 would change with stage duration. More specifically, 1) the power output at V0233 would be lowest for 5-minute stage protocol, 2) the power output at V0233 would be higher for the 3-minute than the 5-minute stage protocol, and 3) V0233 would be detected at the highest power output in the case of the 1-minute stage protocol. REVIEW OF LITERATURE h ra 'stics oft a ti elati hi The pattern of oxygen uptake in response to increasing workloads during cycle ergometry exercise has historically been described as linear and has been presented as such in exercise physiology textbooks (Astrand & Rodahl, 1986; DeVries & Housh, 1994; McArdie, Ketch, & Ketch, 1996). Despite this, there is a substantial amount of evidence that this linearity only occurs below LT and that the relationship is non-linear in the domain of exercise intensities above LT. In fact, this non-linear V02 response in the domain of heavy to severe exercise has been validated and accepted, and the kinetics of 02 uptake relative to exercise intensity have been well documented (Geesser & Poole, 1996; Whipp, 1994). Whipp (1994) characterizes V02 kinetics as being intensity dependent and describes three phases of the response to constant-load exercise. For moderate constant-load exercise, that is for all intensities below the LT, these phases are as follows: (a) an initial rapid response, (b) a slower, exponential response, and (c) a steady state. During light to moderate exercise intensity the steady state is typically reached in 3 to 5 minutes. An analysis of the VO2-watt relationship during incremental cycle ergometry reveals a step pattern, which is reflective of this response at each stage. However, above the LT the kinetics of V02 become progressively slower at each stage, and the characteristic step pattern is increasingly damped as lactate concentration increases, which reflects an increase in V02 and a delayed steady state at each stage (Zhang et al., 1991). The use of linear regression techniques will then reveal a linear slope below the LT, followed by a non-linear component at intensities above the LT, which demonstrates the diminished step pattern. The dampening of the step pattern is a result of the slow component of V02 kinetics and reflects a disproportionate oxygen uptake above the LT compared to that of moderate exercise. The magnitude of the V02” during prolonged, constant-load exercise increases as a function of time and power output and can account for as much as 1 - 1.5 L 02 [minute during heavy exercise (Geesser & Poole, 1996). Although the V023, response occurs only above the LT, and despite the fact that studies have reported high correlations between lactate levels and V02, no research has been able to find a cause and effect relationship between them. Womack et al. (1995) found that infusing epinephrine could cause blood lactate levels to rise without subsequent increases in oxygen consumption. Similarly, Poole et al. (1994) found that direct lactate infusion did not effect the magnitude of the V02». These studies suggest a relationship exists between the LT and the V02»; however, the nature of this relationship does not appear to be causal. i nificance of the V low om n nt The V02.c can delay the attainment of a V02 steady state or it can drive V02 to a maximum value that is higher than would be expected from an observation of the below LT linean’ty (Zoladz et al., 1998). This is the fundamental importance of the slow component that must be examined and understood when exercise testing and prescription are considered. It is common practice for exercise physiologists to use maximal exercise testing techniques to define and prescribe exercise training as percentages of V023“. This practice is flawed if the augmented V02 response above the LT is not recognized, by assuming linearity up to maximum exercise capacity. This may be especially important when considering different exercise-testing modalities. It has been reported in research involving untrained subjects that the amplitude of the V02» is greater for cycle ergometry than for treadmill running (Jones & McConnell, 1999). This has also been found to be the case in trained athletes. Billet, Binsse, Haouzi, and Koralsztein (1998) found in elite triathletes, presumed to be trained equally in both modes of exercise, that the magnitude of the slow component was higher for cycling. These studies suggest that regardless of fitness level or familiarity with exercise modality, the V02» generally has a greater influence during cycling ergometry. Therefore if the V02» is not considered when exercise testing, especially during cycling exercise, then the metabolic cost of an exercise bout may be underestimated. This may then lead directly to inappropriate prescription, particularly for individuals who maintain low levels of fitness. Despite the fact that both trained and untrained subjects have greater V02» responses during cycling when compared to running, each individual’s condition will directly influence the magnitude of the response. Jacobsen, Coast, and Donnelly (1998) reported that in comparing fit and unfit groups performing cycle ergometry, the unfit subjects consistently had greater absolute V02» than did the fitter subjects. This suggests there is a training effect, at least for cycle exercise, in that the result is a diminished V02». In fact, Womack et al. (1995) found a 50% reduction in the slow component with only 2 weeks of cycle ergometry training. Since the V02» represents an additional energy requirement at a particular power output, its delay or diminution signifies an improved economy and therefore should be considered a positive training response. Furthermore, exercise at an intensity that elicits the slow component will be temporally limited, so that attenuating the slow component may result in improved endurance performance. Although research has not described the specific benefits of a diminished slow component to perfonnence, some studies have successfully applied this rationale to the designing of exercise training techniques as well as to exercise testing methodology. For example, Stoudemire et al. (1996) demonstrated that subjects could remain at V02...» for longer than previously thought by the systematic reduction of treadmill velocity after V02...» had been reached. It is likely that the specificity of training at V02...» is essential for its improvement. The longer period of time an individual can exercise at V02,“.x during a particular training session, the more profound the effect that training session may have on the improvement of V023». Therefore this method of systematically reducing treadmill velocity in order for V02,...x to be maintained could prove to be a useful training technique. This rationale could also be applied successfully to designing maximal exercise-testing protocols. Gaesser and Poole (1996) pointed out that there are a variety of power outputs that will, over time, allow for V02,.m to be achieved. - In the severe exercise intensity domain (which typically begins at a power output that is about 50% of the difference between LT and V023») neither lactate concentration nor V02 can stabilize. Therefore, constant-load exercise at 10 any power output performed in this domain will drive V02 to its maximum value. This allows for versatility in exercise-testing protocols, due to the variety of power outputs that may be used, as well as provides an altemative to incremental testing. However, LT cannot be determined using such a protocol since it is performed entirely above that intensity. Incremental exercise testing designed to determine LT and V023.“ can provide coaches and their athletes with additional information regarding fitness if the slow component is considered. Since the V0233 represents the boundary between steady-state and non-steady-state V02, one can determine the maximal steady-state power output. It has been suggested that the V0233 be used as an additional indicator of exercise capacity by determining the power output at which the V0233 occurs (Zoladz et al., 1998). Research has not determined the long- terrn effects of training at a power output that corresponds to the M88, nor has it determined the best methodology for clarifying its occurrence. The significance of the SIQll! ggmmnent for clinical gggulations. The kinetics of the V02 response to exercise above the LT for cardiac and respiratory patients also reflects the delayed attainment of the steady state and may even be more pronounced in these populations (Zhang et al., 1993). In such cases this occurs at very low work-rates, as patients are limited by their specific pathology. However, the training effect described above (which is marked by a diminished slow component and an elevated LT) is applicable to these populations (Gaesser & Poole, 1996). For patients with cardiopathology, training may not increase stroke volume and subsequent V023”, but the diminution of the slow component 11 will allow them to exercise at slightly higher work-rates and therefore improve exercise capacity. Womack et al. (1997) studied the effect of a 4-month exercise rehabilitation program involving patients with peripheral arterial occlusive disease. They found that walking economy was significantly improved with training, and the improvement was due to the diminution of the slow component. For respiratory patients, whose maximum exercise capacity is limited by their maximum ventilatory capacity, improvements in lung function may not occur from exercise training, due to irreversible damage. However, the lowering of the ventilatory work at a particular work-rate may occur by the attenuation of the V02» and the subsequent rise in the LT (Gaesser & Poole, 1996). The V0233 has promise as a marker for exercise tolerance or training adaptation for clinical p0puletions. The V02,.um criterion is often not achieved by patients during exercise testing (Simonton, Higginbothem, 8. Cobb, 1988). Therefore, the V0233 may be a more appropriate criterion on which to determine initial exercise capacity and subsequent improvements. Pgfll Eflegt 99 the LT and VT The lactate and ventilatory thresholds are often used as predictors of endurance performance and as criteria on which exercise intensity is based. Historically these variables have been collectively termed the anaerobic threshold, which is based on the concept that measurement of gas exchange variables (i.e., VT) indicates the occurrence of lactic acidosis (i.e., LT). In recent years, however, research has shown a disparity in the occurrence of the LT and VT, providing evidence that these variables are not causally related. Chicharro, 12 Perez, Vaquero, Lucia, and Legido (1997) found that the VT occurred at a lower power output in 33 of 39 subjects performing a ramp protocol (25 watts/minute) on a cycle ergometer, suggesting that the increase in ventilatory response may not be caused by lactate concentration as had been previously suggested (Wasserman, Beaver, & Whipp, 1990). This dichotomy has also been observed during incremental treadmill exercise (Dickhuth et al., 1999), which suggests that this is not a phenomenon unique to a specific protocol or exercise modality. VT may be affected by protocol design. Davis et al. (1982) compared four ramp protocols of varying work rates using a cycle ergometer. They found that the VT occurred at the same V02 for all protocols, but that the power output that corresponded with the VT increased as the work rate increased. Other research has reported no protocol effect on VT; however, these results either are based on units of mllminute, rather than power output, or used stage durations of longer than 1 minute (Zhang et al., 1991). Wasserman and Whipp (1975) explained that when considering ramp protocol data, it is more appropriate to describe VT in units of mllminute rather than in watts, due to the fact that VT always occurs at the same V02 but not necessarily the same power output. This may be explained by the pattern of responses in the AV02-Awatt relationship observed at different work-rate increments. Hansen, Casaburi, Cooper, and Wasserman (1988) reported that as work-rate increased during ramp protocols from 15 to 60 watts/minute, the AV02-Awatt slope progressively decreased. Based on Wessennan and Whipp’s (1975) contention that for ramp exercise the VT 13 corresponds to the same V02 regardless of protocol, the VT must then necessarily occur at a higher power output. In contrast, Wasserman and Whipp also reported for protocols with stage durations of 1 minute or more VT can be expressed as either V02 or power output. Carta, Am, Barbieri, and Mele (1991) found that the VT occurred at the same V02 and power output for a 30-wattI3-minute and e 10-wattlminute cycle protocol. However, the power output at VT was much higher for a 30-wattlminute protocol, while the V02 at VT was not significantly different from the other protocols. This suggests that this effect may be more a function of work-rate increment than of stage duration for ramp protocols. For incremental protocols of greater than 1 minute, there is no evidence to suggest that VT is affected by work-rate or stage duration. Weltman et al. (1990) found that the V02 and velocity at LT did not change for stage durations of 3- and 10—minutes during treadmill running. However, Stockhausen et al. (1997) defined minimum stage durations for a range of increment sizes necessary for a steady lactate concentration to be achieved. They reported that if the stage duration is too short for a particular workload increase, then a lactate steady state will not occur, and the power output at the LT may be overestimated. They determined that increment sizes of 20-, 30-, and 50-watts are required for stage durations of 3-, 4-, and 5-minutes, respectively. An objective of the endurance athlete is to perform at the highest speed or power output, while still maintaining a constant lactate concentration. Often training intensity is based on the speed or power output that corresponds with the 14 maximum lactate steady-state. Therefore, accuracy in determining the LT is critical to coaches and athletes for designing this type of training. The stage duration and/or work-rate used in identifying the LT may affect the speed or power output associated with its occurrence. Therefore, protocol design should be considered when accurate determination of the LT or VT is essential. The protocol effect on parameters of aerobic fiJnction during cycle ergometry has also been extensively studied. Zhang et al. (1991) compared a ramp protocol and 1-, 2-, and 3-minute stage protocols on a cycle ergometer. Workload size was altered so that all tests were performed at the same work- rate. The authors found no significant differences between these protocols for V023», HR, V3, R, or V002. The only notable difference was a slightly shorter total running time for the 3-minute stage test Similar findings were reported by Davis et al. (1982), who found that various ramp slopes of 20-, 30-, 50-, and 100- wattslminute elicited consistent results for V02,...“ and exercise economy. It is interesting that in each of these studies discussed in the preceding paragraph, a linear V02 response was observed both above and below the LT in the ramp protocols (protocols of continuously increasing power output), but a non-linear response was apparent for the incremental protocols (protocols with temporally defined stages of increasing power output) above LT. In fact, as stage duration increased and work-rate subsequently decreased, the V02 step pattern became increasingly more apparent, but was virtually undetectable for the ramp and 1 -minute protocols. Therefore, it is only during slow work-rates that evidence of the slow component of V02 kinetics can be observed (Whipp, 1994). 15 Parameters of aerobic function such as V023“, V3, V002, and economy are not affected by stage duration or changes in work-rate. However, as with LT, researchers who use exercise testing to identify the V0233 should consider the length of stage as well as the rate of work increment. However, the most appropriate stage length end/or work-rate has not yet been determined. Asse in Maximal teed ta e M The LT has been used as a demarcation point for the upper limit of steady-state exercise as well as a critical indicator upon which to prescribe exercise intensity (Casaburi, Storer, Sullivan, 8. Wasserman, 1995). Due to the relationship that is often believed to exist between lactate concentration and V02, it has been assumed that the LT is coincident with the onset of the V02» and therefore can be used for its detection. However, as described above, the dissociation that has been found regarding these variables has brought this hypothesis into question (Poole, Gladden, Kurdak, & Hogan, 1994; Womack et al., 1 995). It has been suggested that the maximal steady state can be determined by assessing the hyperbolic relationship between power output and endurance time to fatigue (Morton, Green, Bishop, 8. Jenkins, 1996; Poole, Ward, Gardner, 8. Whipp, 1988). Based on this relationship it is possible to determine a power output for an individual that could in theory be maintained indefinitely. This power output represents the upper limit for sustainable exercise and is termed critical mar. This power output is represented in Figure 2 as the vertical asymptote. l6 Time. Power Figure 2. Critical power, represented by the dashed line. Gaesser, Camevale, Garfinkel, Walter, 8. Womack (1995). Specifically, determining critical power involves performing a series of exercise tests to fatigue, each set at a different power output. Critical power is calculated using the formula T = AWCI (P-CP), where T is in seconds, AWC (anaerobic work capacity) represents the total work performed, P is a constant power output, and CP is critical power. The theoretical significance of critical power is that it represents a power output that one could maintain indefinitely. Poole et al. (1988) explained that for power outputs below the LT, steady-state V02 kinetics can be maintained, and consequently the duration of exercise may be prolonged. Also, exercise performed at a power output above the LT but below the critical power is marked 17 by delayed steady-state V02 kinetics. It is not until one reaches a power output above calculated critical power that exercise duration is limited by a disproportionate increase in V02. Due to its theoretical nature, it is difficult to define critical power mathematically. The inability of subjects in various studies to maintain a calculated critical power for prolonged periods exemplifies the apparent difficulty of this calculation (McLeIlan & Cheung, 1992; Pepper, Housh, 8. Johnson, 1992). Another problem in determining critical power is that several bouts of exhaustive exercise at different power outputs are necessary, making this technique time-consuming. Therefore, a method that requires the use of only one trial to determine MSS would be advantageous to coaches and trainers who have time constraints. Zoladz et al. (1998) suggest that the breakpoint in the V02-watt relationship be used as an indicator of exercise tolerance. They postulate that the V0233 has the same mechanistic basis as the slow component of V02 kinetics and therefore could be used as a critical point for assessing exercise capacity. That is, since the slow component represents non-steady- state V02, and the V0233 marks the power output above which the slow component occurs, one is then able to use the V0233 to define the M88. They were able to detect the V0233 in 25 of 26 subjects using a single cycle protocol of 3-minute stages, with 30-watt increments. Therefore, this protocol may be appropriate for the detection of the V0233 and only requires an individual to perform one test. ‘18 Summary The primary objective of this project is to find out whether the V0233 will be affected by alterations in stage duration. It is important to find this answer in order to provide researchers who are studying the V0233 with the most appropriate method for elucidating its occurrence. Previous research suggests that cycle ergometry may be the most appropriate mode of exercise for observing the slow component (Jones & McConnell, 1999). However, researchers have not determined which protocol may be best in elucidating the V0233. Using stage durations of longer than 1 minute may prove to be more appropriate when attempting to identify the V0233 since the V02 step pattern tends to be more visible than with shorter stages. Also, stages of longer than 1 minute have been shown to have this step pattern visibly diminished above the LT and no variability in the LT and VT relative to power output or V02 (Zhang et al., 1991). Furthermore, due to the fact that the increase in V02 is a function of both time and power output in the heavy exercise domain, stages of longer than 3 minutes may engender changes in the V0233. That is, a 3-minute stage may not be of sufficient duration for the V02 to rise to a detectable level. However, a 5-minute stage may be long enough so that the breakpoint is detected at the power output in which it initially occurs and not at the subsequent stage. Figure 3 shows that V02 continues to rise after minute 3 of exercise so that the magnitude of the V02» is greater by minute 5. The graph also shows that 1 minute, even at the highest power output (135 watts), is not long enough to detect the V02». 19 2500 2000 + 30 watts E 1500 + 90 watts E + 150 watts E1000 +213 watts 9' 500 O 1 2 3 4 5 6 7 8 9 10 Time (minutes) Figure 3. The V02 response to different constant-power exercise bouts. LT is indicated by the arrow. Above LT, V02 continues to rise throughout the exercise period. 20 METHODS m Sixteen (8 males and 8 females) subjects (age 26.8 :I: 4.2 years, height 168.9 1 8.1cm, weight 67.4 t 13.3 kg), who were physically active and had no known cardiovascular disease, participated in this study. Each subject signed a consent form prior to his or her participation. This study and the consent form were approved by the University Committee on Research Involving Human Subjects. Table 1 Descri tive a r for All 'ects Gender Age (years) (SD) Height (cm) (SD) Weight (kg) (SD) Male 27.9 (15.3) 173.9 (:I:5.8) 75.2 (113.2) Female 25.7 (12.8) 164.9 (17.0) 61.5 ($8.3) All 26.8 ($4.2) 168.9 (1:8. 1) 67.4 (118.3) 19.31.1093 All subjects performed three maximal exercise tests on a SensorMedics 800$ electrically braked cycle ergometer. These tests included three continuous incremental protocols of 1-, 3-, and 5-minute stages. All tests began at 30 watts and had 30-watt increments. This work-rate increment was chosen since it had been successfully used in conjunction with a 3-minute protocol to determine the V0233 (Zoladz et al., 1998). This study also showed that 30-watt increments are large enough so that subjects are fatigued in a reasonable amount of time. Each exercise protocol continued until the subjects reached volitional exhaustion (i.e. less than 30 minutes). Blood samples were 21 taken prior to each test, at the end of each stage, and immediately upon completion of each test to determine lactate concentration. All samples were assessed by e YSI 2700 Select Biochemistry Analyzer. The tests were performed on three non-consecutive days and the order of the tests was randomized for each subject. LT and VT were determined via blood sampling and gas exchange data, respectively (Davis, 1985; Gaesser & Poole, 1986). Legete sampling end meeeuLemerjt methode. Blood samples were taken during the last 30 seconds of each stage, for each protocol, while the subject continued the exercise. The samples were taken via fingerstick and a very small amount of blood was placed into a heparinized capillary tube. The blood was then transferred into a micro-centrifuge tube before being presented to the lactate analyzer. Samples were assessed by a YSI 2700 Select Biochemistry Analyzer, which provided a lactate concentration value in mmolIL blood. The lactate analyzer was calibrated prior to each test. Due to the limited amount of time available to obtain blood samples during the 1-minute protocols, subjects were asked to hold a heparinized gauze pad. This prevented clotting and allowed for blood to be collected into the capillary tubes during subsequent stages. This was also the procedure for the 3- and 5—minute stage protocols. W The subjects breathed room air through a mouthpiece connected to a one-way valve. During exercise, expired air was analyzed breath-by-breath using a SensorMedics Metabolic Measurement Cart, which provided the following data: V02 (STPD), V002 (STPD), V3 (BTPS), R 22 (V002N02), and and tidal P002 and P02 (P3002, P3T02). The metabolic cart was calibrated prior to each test. Heart rate. Heart rate was measured continuously by a Polar wireless heart rate monitor (Oulu, F inland) and was recorded prior to every test and at the end of each stage for all protocols. Analysis of ventiletery threshold, Wasserman, Whipp, Koyal, and Beaver (1973) defined the VT as the point where one or all of the following occurred, as determined by visual inspection: (1) a distinct non-linear increase In V3; (2) an increase in V3N02 without an increase in V3N002; and (3) an increase in end- tidal partial pressure of oxygen (P3702) without a simultaneous decrease in end- tidal partial pressure of carbon dioxide (P31002). Davis (1985) reported that the increase in V3N02 with no increase in V3NCO2 is the most specific criterion for detecting the VT. For consistency among subjects and protocols, this was the method of detecting VT used in this study. VT was independently determined by two investigators. Research has demonstrated that menstrual cycle phase can affect the value of V02 at the ventilatory threshold (Bemben, Salm, & Salm, 1995). For this reason all female subjects were tested during the follicular phase of the menstrual cycle. Follicular phase was defined as the 7 days following menstruation and was verified by the subjects reporting the end of menstruation. Anelysis of lactate threshold. Lactate concentrations were plotted against power output (watts) and visually inspected for gradual versus abnrpt changes. The LT was defined as the power output prior to the first abrupt increase in 23 lactate concentration (minimum 0.5 mmollL), as agreed upon independently by two investigators (Zoladz at. al. 1998). Figure 4 represents an example of the visual inspection of LT. Subject #8-LT [Lactate] (mmollL) rest 30 60 90 120 150 180 210 240 270 Power Output (watts) Figure 4. Blood lactate analysis. LT was defined as a 0.5 mmollL increase. In the case of subject 8, 5-minute stage protocol LT was determined to be at 120 watts. mm. Predicted V02 for each power output was determined using linear regression of all the measured V02 data below LT. The difference between actual and predicted V02 was determined using a rolling sum of the error, and this was plotted against power output An abnrpt, sustained increase in the rolling sum was visually detected and was determined to be the V0233 (see Figure 5). Wm Three subjects performed each protocol twice for the purpose of determining variability of the V0233, LT, and VT. A coefficient of 24 variation, which is the ratio of the standard deviation to the mean expressed as a percentage. was determined for power output and V02 (see Table 2). . Table 2 Mounts of ngetj' ‘93 V02 Power Output v02... 9.2% 135% LT 5.7% 14.2% VT 4.3% 5.7% RSE (ml/minute) sssasss Subject s-vom, FFT‘N Power Output (watts) Figure 5. A plot of the rolling sum of the error of V02 (actual vs. estimated) against power output. The abrupt change in V02 at 150 watts represents the V0233. 25 Statietjeel agelyeis. Differences in the power output and V02 corresponding to V0233, LT, and VT between the three protocols were determined using a two-factor (protocol and threshold) repeated measures analysis of variance. Statistical significance was set at p < 0.05. 26 RESULTS The LT, VT and V0233 were detected in all of the 48 tests performed. All subjects who participated in this study completed at least four stages of all tests (reached 150-watt stage) and achieved an average V023,... of ~3.1Uminute. TIme to fatigue and Pop»... on average, decreased as stage duration increased (see Table 3). ANOVA showed no significant gender x threshold or gender x stage duration interaction. Therefore the data for all 16 subjects were pooled. Table 3. P a sta and ti e to exhaustion for each s 'ect b tocol 1-Minute 3~Minute 5-Minute Subject watts/stage#1time watts/stageriltime watts/stage#1time 1 270191848 24018122204 21017133200 2 330/11111z14 27019125216 24018137251 3 330111110216 21017119236 24018135237 4 30011019213 27019125204 27019140222 5 2401818209 24018121236 21017130259 6 2701918236 24018122209 21017130:59 7 2401817245 18016117241 18016127226 8 30011019259 27019126211 27019141:08 9 240181731 18016116252 18016125243 10 2401817218 18016117259 18016128253 11 2101716226 1801611739 18016127239 12 2101716239 18016116214 15015124220 13 1801615245 18016115215 15015121246 14 2401817243 21017118259 18016129259 15 240181758 18016117259 18016127242 16 2401817212 15015115200 15015124258 Mean 255l8.517:47 21017119242 19916.6130231 Mean V023,... for each subject is listed in Table 4. V023»). for all subjects ranged from 34.7 ml O2lkglminute to 61.8 ml O2lkglminute, with an average for 27 male subjects of 47.9, and an average for female subjects of 43.5 ml O2lkglminute. Table 4. Mean P at Threshold an ub' haracterist'cs Subject Age Weight Height V02,»... P 03.... V0233 LT VT (kg) (cm) (ml/kg) (watts) (watts) (watts) (watts) 1 29 59.6 166 55.6 250 1 10 90 120 2 22 87.6 175 46.9 280 120 90 170 3 34 79.1 183 44.2 250 100 80 120 4 30 69.9 179 55.9 280 100 80 130 5 31 90.6 170.5 34.7 230 110 100 100 6 28 67.3 173 48.3 230 110 50 110 7 31 89.1 168 35.9 200 90 50 90 8 18 58.5 177.4 61.8 270 150 140 150 9 26 50.7 165 51.6 190 120 40 100 10 24 53.3 152 49.2 190 110 60 120 11 26 53.2 163 49.1 180 110 80 100 12 27 53 159 43.9 170 100 50 90 13 31 69.6 176.1 36.9 210 100 60 120 14 26 70.9 168 37.6 200 100 70 100 15 22 66.9 166 41 .4 200 90 60 90 16 23 60.5 161.1 38.5 180 90 60 90 Mean 27 67.5 168.9 45.7 219 107 72.5 112.5 These values fall within the ranges of age and gender norms, occurring in the 80‘" percentile for men and the 90th percentile for women (ACSM, 1995). Shvartz & Reibold (1990) reviewed average V02...“ values from various studies performed in various countries, including the US, over many years. They found for subjects aged 18-30 that males averaged 48-50 ml Ozlkglminute while females averaged 41-44 ml O2lkglminute. Table 4 lists mean V023,... by subject, as well as, mean PC at V0233, LT, and VT. There was no significant difference between protocols for V023,“; however, mean P0,»... was significantly greater for the 1-minute stage protocol compared to the 3- and 5-minute stages (see Figures 8 8. 9). The average POM attained for the 3- and 5-minute stage protocols were 82.2% and 76.7% of that reached during the 1-minute stage protocol respectively. Mean Peak V02 v02 (Uminute) N N 00 0) O) m 00 N -h~ 1 2 3 Stage Duration (minutes) Figure 6. Mean peak V02 for all subjects by protocol. Mean Peak Power Output 270 1 Watts a 1Stage Duratlaon (minutes) Figure 7. Mean P0,»... for all subjects by protocol. ANOVA for V02 showed a main effect for the thresholds (V0233, LT, VT) in which the mean V02 at V0233 and VT were significantly greater than the V02 29 at LT. The V0233 and VT occurred at 48.7% and 49.8% of V023... respectively, whereas the LT occurred at 35.3% of V023»... There was no overall effect for stage duration and no interaction between stage duration and threshold. mm 13)“ } } 12°“ r - r - ovorzep 5123+ I : + I I + 3'}: on £21 ill I ; 2 WT 23: l I VII I ii ; m ,n' ..l_n 1 _.I__ 1 3 5 mulflaflm) Figure 8. Mean power at V0233, LT, and VT for all subjects. i i V02 (Uminute) .09 r‘r‘r‘r‘r‘r‘r‘r‘ (DID-I‘Mw-fimmfim I I - C - . P—l—I 'l I-—|—-I é :E" l l I mist-1... Figure 9. Mean V02 at V0233, LT and VT for all subjects. ANOVA for power output showed a main effect for threshold and stage duration, but no threshold x stage duration interaction. The mean power output 30 at V0233 (48% P03...) and VT (51.4% P03...) were significantly greater than the power output at LT (32.8% POM“). Power outputs at the three thresholds for the 1-minute stage protocol were significantly greater than for the 3- and 5-minute stage protocols. Figures 8 and 9 show the mean power output and V02 for the thresholds for each protocol. 31 DISCUSSION The purpose of the present study was to determine if the power output and V02 at the V0233, LT, and VT are affected by stage duration during incremental cycle ergometry. In addition, we investigated potential interactions between the parameters V0233, LT, and VT. The main findings were that the PC at V0233. LT, and VT were higher for the 1-minute stage protocol compared to the 3- and 5—minute stage protocol. Also, there was a disparity in the relationship between the V0233 and the LT, as well as between VT and LT. We hypothesized that the V0233 would be detectable in both 3- and 5- minute stage protocols and that it would be detectable at a lower P0 in the case of the 5-minute stage protocol. We also thought it likely that the V0233 would occur at the highest P0 in the 1-minute stage protocol due to the fast work-rate increment. This hypothesis was partially supported by the data in that it showed a general trend for stage duration for all thresholds (LT, Vl', V0239) with the highest P0 occurring for the 1-minute stages. However there was no difference between the 3- and 5-minute stages for our group of subjects. It is possible that for a group of less fit subjects, the time delay in reaching a V02 steady-state may require a longer stage dwation (i.e. 5-minute) for the accurate determination of V0233. The mean V0233 for all protocols occurred at ~49% of V02,»... This is consistent with the findings of Zoladz et al. (1998), who also found that the mean of what they termed the 'V02 change point” occurred at 49% V023,... In that study, the investigators used an incremental cycle protocol consisting of 30 32 wattsI3-minute stages. They found no statistically significant difference between the power output at LT and that at the V0233, suggesting that these parameters are concurrent. However, in the present study, there was a significant difference between the power output, as well as the V02 at which these parameters occurred. A potential explanation for this contrast in results may be explained by subject fitness level. The most striking difference between these studies is the fact that the former study used subjects that were considerably fitter than the subjects in our study, based on the data in Table 5. Table 5 shows data for the P0 at LT, V0233, and PC»... for the 30-watts13-minute protocols for each study. The mean PC at the LT and V0233 were higher for the Zoladz et al. group than for the subjects in this study. Also, the Pop... occurred on average >2 stages higher for the subjects in the previous study. How fitness affects the relationship between LT and V0233 is not clear, and may require further investigation. Table 5. Mean P0 for LT, V0233 and P03»... for 30-watts/3-minute stage protocols. StUdY LT (W) V02eP (W) P03». (W) Zoladz (1998) 125.8 136.8 279.2 Osmond 67.5 106.9 212 The results of this study support previous evidence that the relationship between V02 and power output is not linear in the domain of heavy to intense exercise (i.e., >LT) (Gaesser 8. Poole, 1996). In all of the 48 incremental exercise tests performed for this study the V0233 was detected. In each case the 33 V0233 occurred at, or at least one stage above, the LT. That we were able to detect the V0233 in all of the 1-minute stage tests performed is an interesting finding, since others have reported that the slow component of V02 kinetics may not be detectable during fast work-rate increments (Whipp, 1994). For well- trained individuals the initial fast component of V02 kinetics has been shown to be of shorter duration than for less trained subjects, and therefore a V02 steady- stete is attained in less time (Jacobean, Coast, 8. Donnelly, 1998). It is thus reasonable to assume then that a protocol of fast work rate increments (Le. 30- watts/minute) may be sufficient to allow for a steady-state to be reached for these individuals, and thus may result in a detectable V0233. Since our subjects were of average fitness, as pointed out in the previous section (see Results), this does not explain the fact that we were able to detect V0233 for all subjects for the 1- minute stage protocols. However, the 1-minute stage duration did tend to overestimate the PO and V02 at LT, VT, and V0233 compared to the longer stage durations. The dissociation between the LT and the V0233 found in this study supplements the growing body of evidence that although there may be a mechanistic link between [lactate] and V02, the relationship is not causal (Womack et al., 1995). Womack et al. found that the diminution of the V02» that occurred as a result of a 6-week cycle ergometry training program was coincident with reductions in blood lactate concentration. However, subsequent epinephrine infusion, which increased blood lactate concentration, did not increase exercise V02. In our study we also observed that although the power output and V02 at 34 LT decreased from the 1-minute stage protocol compared to the longer stage durations, the V0233 did not follow a similar pattern, as one would expect from the purported relationship. The fact that the V0233 was detected in every case, either above or at LT, suggests that the two are in some way related; however, the exact nature of this relationship remains unclear. Based on the aforementioned observations by Womack et al. (1995), it is unlikely that increased [lactate] alone causes this nonlinear increase in V02. The power output at LT is apparently altered by stage duration and work- rate increment (Stockhausen et al., 1997). Stockhausen’s group found that 30- wett increments necessitate the use of 4-minute stage durations in order to reach a steady state [lactate] or risk the likelihood of overestimating the power output at LT. Our results support this finding as the mean thresholds occurred at a substantially higher power output for the shortest stage duration compared to the 5-minute stage tests. Our results suggest that 3-minute stages may also be of sufficient length for accurate determination of LT (see Figures 6 and 7). The results of our study show a disparity between LT and VT in terms of theV02andpoweroutputatwhichtheyocwrred, withtheVl'occurringata substantially higher V02 and power output for all protocols (see Figures 6 and 7). Furthermore, the power output at LT was impacted by stage duration, whereas VT was unaffected. This may be due to the fact that the appearance of lactate in the blood is a time dependent variable, whereas VT is not. The fact that there was no interaction between stage duration and VT has been previously reported 35 using 1-, 2-, and 3-minute stages (Zhang et al., 1991). The results of the present study extend this observation to 5-minute stages. Ventilatory threshold is believed to be related to initial increments in blood lactate concentration and the subsequent buffering of lactate by increased carbon dioxide production. It should therefore occur after relatively small increases in lactate. Since lactate appearance in the blood is time dependent, whereas VT has been shown to be unaffected by stage duration, incremental exercise testing must account for both stage duration and work-rate or risk overestimating LT (Stockhausen et al., 1997). Previous research using fast work- rate increments has shown a disparity between VT and LT during cycle ergometry (Chicharro et al., 1997). This study, which utilized a 1-minute stage protocol similar to the one used in our experiment, found that the VT occurred prior to the LT. Therefore, the mean VT was lower than the LT in both power output and V02. Based on this relationship, other mechanisms have been suggested to cause VT rather than lactate, such as H+. Bangsbo, Johansen, Graham, and Saltin (1993) postulated that the H+ enters the blood prior the lactate and causes the rise in VENoz prior to the observed increase in lactate during fast work-rate incremental protocols. This, however, does not explain why we observed the LT significamly preceding the VT both in terms of P0 and V02. In fact our data suggest that the VT and V0235: are more closely related to one another than LT. However, past investigations have shown that increases in ventilation, or the increased work of the respiratory muscles at VT, only account for a small portion 36 of the slow component and thus does not explain this relationship (Womack et al., 1995). The cause of the delay in VT relative to the LT observed in the present study needs further investigation. P0,...“ was significantly higher for the 1-minute stage protocol compared to the 3- and 5-minute protocols, whereas VOzpeak was unaffected by stage duration. This finding is consistent with previous research involving incremental cycle ergometry (Hansen et al., 1988; Zhang et al., 1991). The LT has been used as a primary identifier for the upper limit of steady- state exercise as well as a critical indicator that is often used for exercise training. Due to the relationship that is believed by some to exist between [lactate] and oxygen consumption, it has been assumed that the LT and the onset of the VOzsc are concurrent events and therefore the LT can be used to detect the V02”. Our observation regarding these variables brings this theory into question. The V023p is believed to be the point above which the slow component of V02 kinetics occurs. V029.» is therefore a potential indicator of reduced exercise efficiency, and therefore may be used as a marker for exercise tolerance. Therefore, due to the purported relationship between lactate and V02, the V029,]: and the LT should occur at a similar power output However, our data show that these variables are not concurrent, as the LT occurred at a significantly lower PO and V02 than the V0239 for all protocols. Based on this, we conclude that it is not valid to use the LT to demarmte the occurrence of the V02.c during cycle ergometry exercise. Therefore, researchers who want to study the V02» 37 need to determine the V023p and LT independently and not regard the two as the same phenomenon. The overestimation of LT for the 1-minute stage protocols compared to the 3- and 5-minute stage protocols suggests that this is not long enough for accurate LT determination. This should be taken into consideration when designing protocols; however, the 5—minute stage protocol causes unnecessary subject discomfort as it is not completed in as timely a fashion as the 3-minute stage test, yet yields similar results. It is our suggestion that for 30-watt increments, 3-minutes is the most appropriate stage duration for the determination of these parameters. From this protocol, accurate determination of LT, VT, V0299, V033“, and Pop.“ is possible. 38 REFERENCES AQSM’s guidelines for exerg‘se tpgtipg and prescription (5th ed). (1995). Baltimore, MD: Williams 8. Wilkins. Astrand, P., & Rodahl, K (1986). Textpook of work QhVSiOIW basis of expgcise. New York: McGraw-Hill. Bangsbo, J., Johansen, L., Graham, & T, Saltin, B. (1993). Lactate and H+ effluxes from human skeletal muscle during intense, dynamic exercise. Journal of Physiolpgy, 462, 1 15-133. Barstow, T. J., & Mole, P. A. (1991). Linear and nonlinear characteristics of oxygen uptake kinetics during heavy exercise. qumal of Applied Physiolpgy, 71, 2099-2106. Bemben, D. A., Salm, P. C., & Salm, A. J. (1995). Ventilatory and blood lactate responses to maximal treadmill exercise during the menstrual cycle. Journal of §mrts Medicine and Physipal F itn_;__ess, pp, 257-262. Billat, V., Binsse, V., Haouzi, P., & Koralsztein, J. P. (1998). High level runners are able to maintain a V02 steady-state below VOW in an all-out run over their critical velocity. Archives of Physiolpgical Biochemism, 107, 1 -8. Billat, V. L., Slawinski, J., Bocquet, V., Demarle, A, Lafitte, L., Chassaning, P., & Koralsztein, J. (2000). Inten'nittent runs at the velocity associated with maximal oxygen uptake enables subjects to remain at maximal oxygen uptake for a longer time than intense but submaximal runs. Eggpppgp Journal of Applied Physiolpgy, §1, 188-196. Carta, P., Am, G., Barbieri, M. T., & Mele, M. (1991). Bicycle ergometry tests: a comparison between 3 protocols with an increasing load. Med L_av, Q2, 56-64. Casaburi, R., Storer, T. W., Sullivan, C. S., & Wasserman, K (1995). Evaluation of blood lactate as an intensity criterion for exercise training. Mppicinp and §cience in Sm and gecisp, 27, 852-836. Chicharro, J. L., Perez, M., Vaquero, A. F., Lucia, A., & Legido, J. C. (1997). Lactate threshold vs ventilatory threshold during a ramp test on a cycle ergometer. Journal of §pprts Medipine and Phfiical Fitness, g7, 117-121. 39 Davis, J. A. (1985). Anaerobic threshold: review of the concept and directions for future research. Medig'ne and §pience in Sage and Expgcise, 17, 6-18. Davis, J. A., Whipp, B. J., Lamarra, N.. Huntsman, D. J., Frank, M. H., & Wasserman, K (1982). Effect of ramp slope on the determination of aerobic parameters from the ramp exercise test. Medicine and Scieng in §pprts app Exercise, 14, 339-343. Devries, H. A & Housh, T. J. (1994). Physiology of exercise: for physical pgu jSpn n,pt_l1]etics a_r_ig pxerpjg scipnce. Madison,le Brown 8. Benchmark. Dickhuth, H. H., Yin, L., Niess, A, Rocker, K, Mayer, F., Heitkamp, H. C., & Horstmann, T. (1999). Ventilatory, lactate-derived and catecholamine thresholds during incremental treadmill ninning: relationship and reproducibility. WW1 22-127. Gaesser G. A, Camevale, T. J., Garfinkel, A, Walter, D. 0., &Womack, C J. (1995). Estimation of critical power with nonlinear and linear models. Mpgicinp egg §9_ien pain §ppps and Ma, 27, 1430-1438. Gaesser, G. A, & Poole, D. C. (1986). Lactate and ventilatory thresholds: disparity in time course of training adaptation in humans. qumalg Applied Physiolppy, 61, 999—1004. Gaesser, G. A, 8. Poole, D. C. (1996). The slow component of oxygen uptake kinetics In humans W 35-70 Hansen, J. E., Casaburi, R., Cooper, D. M., & Wasserman, K (1988). Oxygen uptake as related to work rate increment during cycle ergometer exercise. 521%" Journal of Apflifl Physiplmy, §7, 140-145. Jacobean, D. J., Coast, R., & Donnelly, J. E. (1998). The effect of exercise intensity on the slow component of V02 in persons of different fitness levels. a rts 'ci Ph ' lFi 124-131. Jones, A. M., & McConnell, A M. (1999). Effect of exercise modality on oxygen uptake kinetics during heavy exercise, 4‘“ edition. W Appljgg Phy§io olpgy, 80.,213-219 McArdle, W. D., Katch, F. l., 8. Katch, V. L. (1996). er ‘ iol : nptp’tipp, and human gnmpppp. Baltimore, MD: Williams & Wilkins. McLellan, T. M., & Cheung, K S. Y. (1992). A comparative evaluation of the individual anaerobic threshold and the critical power. Mpgjg;n_e_g_n_d ci rice in rts Exe ci 24 543-550. 40 Morton, R. H., Green, 3., Bishop, 0., & Jenkins, D. G. Ramp and constant power trials produce equivalent critical power estimates. Medicine and Science in Sports and Exercise, 29, 833-836. Pepper, M. L., Housh, T. J., 8 Johnson, 6.0. (1992). The accuracy of the critical velocity test for predicting time to exhaustion during treadmill running. lntemational Journal of §ports Medicine. 13, 121-124. Poole, D. C., Gladden, L. B., Kurdak, 8., & Hogan, MC. (1994). L-(+)-Lactate infusion into a working dog gastrocnemius: no evidence lactate per se mediates V02 slow component. Jopmgl g Applied Physiolpgy, 76, 787- 792. Poole, D. C., Ward, S. A, Gardner, G. W., 8. Whipp, B. J. (1988). Metabolic and respiratory profile of the upper limit for prolonged exercise in man. grgpnomipg, 31, 1265-1279. Roston, W. L., Whipp, J. A, Cunningham, D. A, Effros, R M., 8 Wasserman, K (1987). Oxygen uptake kinetics and lactate concentration during exercise in humans. W1W1W. Simonton, C. A, Higgirbotham, M. B., & Cobb, F. R. (1988). The ventilatory threshold: quantitative analysis of reproducibility and relation to arterial lactate concentration in normal subjects and in patients with chronic congestive heart failure WW 100-107 Stockhausen, W., Grathwhol, C., Bi'irklin, C., Spranz, P., 8. Keul, J. (1997). Stage duration and increase of work load in incremental testing on a cycle ergometer. uro a i Ph ‘01 76 295-301. Stoudemire, N. M., Wideman, L., Pass, KA, McGinnes, C. L., Gaesser, G. A. 8. Weltman, A (1996). The validity of regulating blood lactate concentration during mnning by ratings of perceived exertion. WW §ppps ang Exercise, 2p, 490-495. Wasserman, K, Beaver, W. L., & Whipp, B. J. (1990). Gas exchange theory and the lactic acidosis (anaerobic) threshold. Qirculgtipp, 81, 14-30. Wasserman, K, 8. Whipp, B.J. ( 1975). Exercise physiology in health and disease. Merigp vaiew of Bespirptpg pigagp, 112, 219-249. Wasserman, K, Whipp, B. J., Koyal, S. N., & Beaver, W. L. (1973). Anaerobic threshold and respiratory gas exchange during exercise. Jpprmjpt Applied Physiolpgy, 35, 236-243. 41 Weltman, A, Snead, D., Stein, P., Seip, R., Schurrer, R., Rutt, R., & Weltman, J. (1990). Reliability and validity of a continuous incremental treadmill protocol for the determination of lactate threshold, fixed blood lactate concentrations, and vozmx, lntemational oumal of rts icine 1 26-32. Whipp, B. J. (1994). The slow component of 02 kinetics during heavy exercise. Medicine and Science in §Ms and Exercise, 26, 1319—1326. Womack, C. J., Davis, 8. E., Blumer, J. L., Barrett, E., Weltman, A. L., & Gaesser, G. A (1995). Slow component of 02 uptake during heavy exercise: adaptation to endurance training. Journal pf &plied Physiolpgy, 19, 838-845. Womack, C. J., Sieminski, D. J., Katzel, L. l., Yetaco, A, 8: Gardner, A W. (1997). Improved walking economy in patients with peripheral arterial occlusive disease. edici and ciencei ' 1286-1290. Zhang, Y., Johnson, M. C., Chow, N., & Wasserman, K (1991). Effect of exercise testing protocol on parameters of aerobic function. Midne and Science in Smrts and Exercig, 23, 625-630. Zhang, Y., Wasserman, K, Sietsema, K E.. Ben-Dov, l., Barstow, T. J., Mizumoto, G., & Sullivan, C. S. (1993). 02 uptake in response to exercise: a measure of tissue anaerobiosis in heart failure. Chest, 103, 735—741. Zoladz, J. A, Rademaker, A, & Sergeant, A J. (1995). Non-linear relationship between 02 uptake and power output at high intensities of exercise in humans. Joumal of Physiojpgy, 488, 21 1-217. Zoladz, J. A, Szkutnik, Z, Majerczak, J., & Duda, K (1998). Detection of the point in oxygen uptake during an incremental exercise test using recursive residuals: relationship to the plasma accumulation and blood acid base balance. uro an ou l of ied Ph ' l 8 369-377. 42 T ' liliiljilliljjlj T illijjlljljjjml ' 2