VO2WR relationship Chronic Obstructive 2

VO2WR relationship

Thu, 26 Jan 2012 21:18:12 | Chronic Obstructive 2

There are several quantifiable features of the VO2 response to incremental exercise that provide important clues to ascertaining why a subject may manifest exercise intolerance. The most obvious of these is the maximum Vo2 (most usually inferred from VO2,peak), but additional submaximal indices may also be informative. Interpreting abnormalities in the Vo2 response to incremental exercise requires consideration of its determining variables, as characterised by the Fick Principle. For the body as a whole, these are cardiac output (Q'), cardiac frequency fc), stroke volume (SV) and the arterio-mixed venous oxygen content difference or "extraction" (Ca,02-Cv,02):

and for muscle oxygen consumption (Q'02), these are muscle blood flow (Q'm) and muscle oxygen extraction (Ca,02-Cv,M02):

Baseline V'O2. It is most usual to initiate an incremental test from a background of "unloaded" or very light exercise. This, unlike a test performed from rest, provides a known baseline WR upon which the actual incremental response is superimposed. For cycle ergometry, baseline Vo2 (Vo2,bl) values in the region of ~500 mL-min-1 are to be expected in a normal healthy young adult exercising at or close to true unloaded pedalling (i.e. 0 W) [5].

Vo2,bl is made up of two components that are required to: 1) support resting metabolism (Vo2,rest); and 2) move the legs (for example, in the case of cycling) at the required cadence against no applied load to the flywheel of the cycle ergometer (V02,0). Vo2,rest is influenced by body mass and, more specifically, fat-free mass (FFM) [6] and, therefore, is reduced with muscle wasting (e.g. chronic deconditioning). Resting energy expenditure has been widely reported to be increased by 15-20% both in chronic obstructive pulmonary disease (COPD) [e.g. 7, 8] and in interstitial lung disease (ILD) [9], a phenomenon attributable to the increased oxygen costs of ventilation.

Accurate measurement of V02,0 requires a steady state to have been achieved following the onset of unloaded pedalling. A 3-min period is typically advocated by many investigators [e.g. 4, 10-12]; this is likely to be adequate in most subjects. However, if visual inspection shows Vo2 not to have stabilised by this time, clearly the phase should be extended to allow a steady state to be confidently achieved. An additional consideration specific to patients with severe chronic heart failure (CHF) is the high probability of ventilatory and gas exchange oscillations at rest and moderate exercise [12-14]. As these can have an appreciable magnitude and periodicity (fig. 1), it is important to ensure that the measurement period includes an integral number of cycles rather than relying on an arbitrarily fixed period. Naturally, the oscillatory magnitude will be influenced by the particular data-averaging strategy employed.

Alterations in V02,0 influence the position of the Vo2-WR relationship. V02,0 is influenced by the mass of the involved muscles, i.e. the legs (largely) for cycle ergometry. As a result, higher-than-normal values of Vo2,bl (e.g. ~>700 mL-min-1) result with elevated body, and especially leg, mass, as in obese subjects and in athletes with highly

Time min

Fig. 1. - Periodic oscillations in carbon dioxide output (Vco2;----), oxygen uptake (Vo2;----) and minute ventilation (VE; -), during incremental exercise in a patient with heart failure. Reproduced from [14] with permission.

Time min

muscled lower limbs (e.g. body builders, speed skaters); values increase by ~5.8 mL-min-kg-1 body mass at constant pedalling frequency (fig. 2) [5, 15]. This would be expected to impose an upward shift of the Vo2,-Wr relationship, resulting in an increased oxygen cost at a given WR (fig. 3).

Cycle ergometers with a substantial ''unloaded'' pedalling setting will also induce higher-than-normal Vo2,bl responses. It is rarely the case that ergometers are truly unloaded; i.e. requiring only 0 W [reviewed in 3, 10-12]. Although there are some electromagnetically braked models of cycle ergometer that are loadless (with the resistances provided by the pedals and flywheel being overcome by a driver motor), it is not uncommon to encounter ''unloaded'' settings as high as 30 W. The baseline conditions associated with treadmill exercise can be even more variable, reflecting differences in baseline grade, speed and body mass, and the mechanical efficiency of

Fig. 2. - Oxygen cost of unloaded pedalling as a function of body mass (pedalling frequency 5 60 rpm). •: lean; O: moderately obese; ■: obese. Vo2: oxygen uptake. y=5.8x+151; r=0.94. Reproduced from [5] with permission.

Body weight kg

Fig. 3. - Schematic diagram of the oxygen uptake (Vo2)-work rate (WR) relationship for ramp-incremental exercise showing the influence of altered baseline Vo2 consequent to increased body mass (e.g. obesity;-----), and decreased body mass (e.g. cachexia;----) relative to normal (----). Influence of the obese subject pedalling with a progressively increasing cadence ...... Vertical solid line: the start of the ramp phase of the test, from a baseline of unloaded pedalling. Horizontal solid line: peak Vo2. The arrows represent peak WRs for each condition. See text for further details.

walking (even when the patient does not hold on to the handrail of the treadmill or is partially supported by the investigator) [reviewed in 3, 10-12]. Such factors can elevate Fo2,bl and by amounts that are not readily predictable. Elevated baseline oxygen costs can impose significant ''performance'' implications, as they impose a necessarily high oxygen cost at all WRs and, therefore, impose a limitation on the operational WR range that can be accommodated (see below).

In contrast, appreciably lower values of Fo2,bl (e.g. <300 mL-min-1) can be found in conditions where there is a low body mass, FFM and/or leg mass; for example in children, the frail elderly, anorexic subjects and patients with cachexia (e.g. some COPD patients). This would be expected to shift the Fo2-WR relationship downwards relative to normal, with the Vo2 at any particular WR being lower (fig. 3).

Incremental gain ( AVo2IAWR). As discussed previously [16], for ramp-type cycle ergometry of the kind commonly used in CPET applications, VO2 increases as a linear function of time and, therefore, WR (after a small lag-phase, consequent to the system response kinetics; fig. 3). The actual value of VO2 at any WR is, therefore, lower for the incremental (or ramp) test than for the steady-state condition, by an amount that reflects the Vo2 ''mean response time'' (MRT). This amount will be a close approximation to tVOi.; specifically, the VO2 equivalent of the increment in WR that occurs over a period equal to the MRT. However, the slope of the FO2-WR relationship, or "incremental gain'' (AFo2/AWR), over the linear response range of the incremental test has been shown not to differ from that of the steady state in healthy young adults, being of the order of ~9-12 mL-min-W-1. The incremental gain is often used as an index of the work efficiency (gw) but, importantly, efficiency cannot be validly estimated unless both VO2 and VCO2 have achieved their respective steady states; this, of course, is never the case during incremental exercise.

The profile of the V'O 2-WR relationship has the potential to be influenced by the components of gw, namely the efficiency of phosphorylative coupling (gp), the efficiency of contraction coupling (the utilisation of the phosphate bond energy for the contraction process itself) and the motor efficiency (or skill) with which the task is accomplished [e.g. 16]. A Vo2/AWR is not appreciably influenced by age [17, 18] or fitness (see below).

The dietary substrate mixture undergoing oxidation has a modest effect on gp [e.g. 16]. For example, the preferential oxidation of free fatty acids over that of carbohydrate, as occurs in highly fit endurance athletes [e.g. 19] and poorly regulated diabetics [e.g. 20,21] is associated with a modest reduction in gp [e.g. 16]. With reference to the former, it is of interest that there are reports of slightly greater A KO2/AWR values in subjects with a high VO2,peak [12, 22].

A reduction in gp consequent to a defect at some point(s) within the energy transduction mechanisms linking adenosine triphosphate production to oxygen utilisation, as might be expected for hyperthyroidism, would predispose towards increased AVo2/AWR values [23]. This has also been reported in patients with a deficiency of muscle succinate dehydrogenase and mitochondrial aconitase [24].

A further factor that has the potential to influence A PO2/AWR is the profile of muscle fibre-type recruitment with increasing WR. Based on single-fibre glycogen depletion profiles, force generation requirements for modest WRs are met from the type I slow-twitch fibre pool, with an involvement of type IIA and type IIX fast-twitch fibres occurring only at relatively high WRs [25,26]. Type II (fast-twitch) muscle fibres have been reported to have a lower gp than type I fibres [27-29, for an alternative view see 22, 30, 31]. It has been widely argued that fast-twitch fibre recruitment is a significant contributor to the ''slow'' and "excess" component of Vo2 response [reviewed in 32-35], through an increased high-energy phosphate (~P) cost of force generation rather than an increased oxygen cost of ~P production [36], which is expressed as a steepening of the Vo2-WR relationship in the later stages of longer-duration incremental tests [16, 37-39].

It has also been suggested that fibre type might contribute to the increased oxygen cost of exercise that has been reported in COPD patients, relative to normal [e.g. 5, 40, 41]. Thus, while A Vo2/AWR is often within normal limits in patients with COPD [e.g. 10-12,42], Vo2 at any particular WR can be elevated above normal (fig. 4). It has been proposed [43] that this might reflect, in part at least, the bio-energetic consequences of selective atrophy of type I (slow-twitch, oxidative) muscle fibres (in m. quadriceps), especially in severe COPD [reviewed in 44, 45]. Whether there is also a similarly increased oxygen cost in other conditions of selective type I fibre atrophy, such as CHF [reviewed in 44, 46], or of reduced type I fibre oxidative capacity consequent to mutations in the p myosin heavy-chain gene in hypertrophic cardiomyopathy [47] is unclear. Neither is it known whether the selective atrophy of type II fibres characteristically found in ageing [48], which has been reported in patients with moderate COPD [reviewed in 45], might have the opposite effect on oxygen cost.

The superimposition of a progressively larger contribution from extraneous "work" on the external work requirements of the task can also lead to a steepening of the Vo2-WR relationship. This can occur during cycle ergometry if there is an increasing reliance on the upper-body musculature, as a result of subjects ''pulling'' forcefully on the handlebars in an attempt to sustain the required power output near the tolerable limit. Likewise, an exacerbated oxygen cost might be expected in clinical conditions characterised by spasms and poor motor coordination. Furthermore, an increased oxygen cost of moving the legs (see above) could also influence AVo2/AWR. For example, while AVo2/AWR can be relatively normal in obesity [49, 50], were an obese subject to elect to cycle with a progressively increasing cadence as the exercise test proceeds, the associated increased oxygen cost would cause AVo2/AWR to increase progressively (fig. 3). It would be prudent to ensure that such subjects exercise at a constant cadence throughout the test (fig. 3). Even so, the elevated Vo2,bl constrains the effective WR range for the obese subject: for example, an obese subject who is physically active with a reasonably normal Vo2,peak would achieve a lower peak WR (WRpeak) than a lean subject with the same Vo2,peak and WR incrementation rate (fig. 3).

0.00

90 WRW

Fig. 4. - Whole-body and one-leg oxygen uptake (Vo2)-work rate (WR) relationship (mean + se) for ramp-incremental exercise pre-training (O and □) and post-training (• and ■) for healthy subjects (O and •) and patients with chronic obstructive pulmonary disease (□ and ■). Reproduced from [41] with permission.

90 WRW

An increased oxygen cost of breathing consequent to an exacerbated ventilatory response to exercise (caused by an increased physiological dead space fraction of the breath (dead space volume (Vd)/ tidal volume (VT)) and/or arterial hypoxaemia, for example) in conditions having abnormally high respiratory impedance such as COPD, restrictive disease and obesity [e.g. 42, 51-53] also has the potential to increase AVo2/ AWR. Despite this, AFo2/AWR is often normal in COPD. It has been suggested [10] that this might reflect the offsetting influence of the abnormally slow Ko2 kinetics typically found in these patients [54-56]; i.e. there is evidence of an increased steady-state oxygen cost at submaximal WRs [5, 12].

In contrast, it is relatively infrequent that an increase in gp is encountered, which would be expected to lead to a shallower Vo2-WR relationship, especially at higher WRs. Some investigators have suggested this occurs for tests having a very fast WR incrementation rate [e.g. 12]. This is certainly the case for exercise-related abnormalities of intramuscular oxygen utilisation, such as have been reported in mitochondrial myopathies [57, 58], CHF [59-61], COPD [e.g. 62-65] and cystic fibrosis [66].

Impaired oxygen delivery to the exercising muscles (e.g. inappropriately low muscle perfusion; compromised oxygen carriage by haemoglobin) is the more usual cause of a flatter VO2-WR relationship. A compromised muscle perfusion response to exercise is a major cause of the reduced A Vo2/AWr commonly seen in many cardiovascular disease states, either over the entire WR range or over the higher reaches as symptom limitations are approached [reviewed in 11, 12, 67-69]. This latter scenario is exemplified by ischaemic heart disease, for which the VO2-WR relationship initially may be reasonably normal, but becomes flatter as the tolerable limit is approached, often with a rather abrupt transition that coincides with the onset of electrocardiographical abnormalities [12, 69, 70] and with a premature plateau in the oxygen pulse (O2-P) response (see below). In such conditions, A VO2/AWR would be expected to be low for a ramp-incremental test, but not for a series of constant-WR tests conducted to steady state.

Sluggish, and even abnormal, Vo2 response kinetics can also influence the form of the VO2-WR relationship, recognising that the nonsteady-state nature of the incremental (ramp) test format. For example, when t Vo2 is abnormally long (see below), the Vo2-WR relationship is shifted to the right (for simplicity, shown here with no change of slope), reflecting the increased prominence of the kinetic phase. When this is accompanied by a reduced Vo2,peak (as is commonly the case), establishing a linear phase with confidence may be hampered because of the compressed tolerable WR range (fig. 5).

In addition, AVo2/AWR can be lower than normal were there to be a progressive slowing of the VO2 kinetics as the WR increases, with t VO2 lengthening progressively as WR increases during the ramp [16, 33]. Such a scenario could occur were the regional distribution of perfusion to metabolic rate within the exercising muscles to be highly heterogeneous [c.f. 33, 71, 72]. This has been suggested to occur in COPD [62] and CHF [69], for example. Again, one would not expect a priori the steady-state gain to be low as well, as the effects of these kinetic heterogeneities would have dissipated.

All of these influences will be compounded in patients with severe CHF who exhibit periodic breathing (fig. 1). Particular care is then needed in slope estimation. Whatever the aetiology of these various AVO2/AWR abnormalities, it cannot be emphasised enough that meaningful interpretation of purported changes in AVo2/AWR requires that the slope estimation be conducted over a WR region for which the VO2-WR

Fig. 5. - Schematic diagram of the oxygen uptake (Vo2)-work rate (WR) relationship for ramp-incremental exercise showing the response for a subject of poor fitness (slow kinetics, low peak Vo2:----) with a restricted linear phase, relative to normal (-). The solid line represents the steady-state relationship. The dotted horizontal arrows indicate

Vo2 mean response time. Vertical arrows indicate peak WR.----: peak Vo2. See text for further details.

relationship is convincingly linear (fig. 5). This certainly requires exclusion of the initial nonlinear or "kinetic" phase at the beginning of the ramp response. Furthermore, it may not necessarily be the case that a normal AVO2/AWR is indicative of normal function; there is the potential for offsetting influences to disguise underlying abnormality. Finally, in this and subsequent contexts, the manner in which breath-by-breath data are analysed deserves consideration. In particular, the practice of subjecting the data series to a moving- rather than a stationary-average is not to be recommended, especially when departures from linearity are present. This practice predisposes to distortion of the actual response profile and the time-point at which a break-point occurs.

In contrast to cycle ergometry, discriminating abnormalities of AVo2/AWR for treadmill exercise is more problematic. Apart from the considerations raised earlier, there is also the difficulty of providing a metabolic equivalent of the grade and speed profile because of variable and unquantifiable contributions from factors such as technique and vertical displacement of the body [reviewed in 3].

Peak V'O2. The maximum Vo2 (Vo2,max) reflects the attainment of a limitation at some point(s) in the oxygen conductance pathway from the lungs to the site of the mitochondrial oxygen consumption at the cytochrome-oxidase terminus of the electron transport chain [reviewed in 73, 74]. Thus, dysfunction in the responses of the convective oxygen flows into the lungs and through the vasculature, and the diffusive oxygen flows across pulmonary and muscle capillary beds will be reflected in abnormally low values of Vo2,max.

The criterion for Vo2,max is that Vo2 be demonstrated to no longer increase despite further increases in WR [75]. The classical approach for determining Vo2,max requires the subject to complete several discrete exhausting constant-WR tests [76, 77]; this is both taxing and time-consuming for the subject and the investigator(s). However, plateauing of the Vo2-WR relationship is not commonly seen with rapid incremental tests of the kind used in clinical exercise testing (i.e. having a duration of ~10-12 min); although, this may be more likely for tests which are of substantially longer duration because of a relatively low WR incrementation rate and/or because subjects are highly fit. Therefore, it is standard practice to refer to the highest value attained on an incremental test as Vo2,peak; i.e. the highest value achieved with good subject effort. This corresponds well with the more traditionally determined Vo2,max in well-motivated healthy individuals [e.g. 78, 79]. However, Vo2,max cannot be determined in conditions where an exercise test is terminated prematurely because of symptom limiting perceptions, such as dyspnoea, angina or claudicating pain [e.g. 10-12].

A recently published useful expedient allows confirmatory evidence for Vo 2,max on the same testing session by imposing a supra-maximal constant-WR test to the limit of tolerance after a short recovery period (e.g. ~5 min or so) immediately following the incremental test (fig. 6) [80]. This test is likely to be useful in normal subjects and especially athletes, for whom an accurate measure of VO2,peak is important. However, its utility has not yet been demonstrated in patients with cardiac or pulmonary disease. Indeed, whether this test proves to be realistic in such patients is uncertain, given the additional time and compliance demands.

With good subject effort, Vo2,max is independent of the WR incrementation rate [e.g. 37]. Even so, slow incrementation rates are not to be recommended, as there is the risk of underestimating Vo2,max because of problems with maintaining motivation for prolonged periods. Tests with fast WR incrementation rates also predispose to underestimation of Vo2,max, because the subject may experience difficulties in being able to generate the necessary muscular force to support the high WRs attained. However, reflective of the underlying response kinetics, the WRpeak attained is progressively greater the faster the ramp (fig. 7) [e.g. 37]. Therefore, WRpeak is not a

Time s

Fig. 6. - Representative response of oxygen uptake ( Vo2) to a rapid incremental test, followed by a 5-min period of 20 W exercise and then step exercise at 95% peak work rate, both performed to the limit of tolerance. Note that peak Vo2 on the ramp test was similar to maximum Vo2 on the step test. Reproduced from [80] with permission.

Time s

reliable surrogate for Vo2,peak. This should be borne in mind when interpreting any intervention-related change in WRpeak, where an improved exercise tolerance might lead the investigator to impose a slight larger WR incrementation rate post-intervention.

Inadequacies of cardiovascular function, whether the result of chronic inactivity, disease or both, are associated with abnormally low values of Vo2,peak (e.g. in the elderly and patients with cardio-circulatory diseases). Discrimination between these two causes requires levels of habitual physical activity between patient and control groups to be matched. The exacerbated exertional dyspnoea resulting from the respiratory mechanical impairments in obstructive and restrictive pulmonary disease, likewise, leads to a low Vo2,peak. However, the scenario should also be borne in mind that a "normal'' Vo2,peak could actually represent an abnormality. The case in point being that of the previously "highly fit'' subject who became impaired, resulting in a VO2,peak that has plummeted but is still in the normal range for sedentary subjects.

The issue of the period over which Vo2,peak is best measured deserves some attention. Due to the typical breath-by-breath variability inherent in Po2 (and other gas exchange variables), it is usual practice to select a short interval of time back from the end-exercise point in order to allow the effects of this "noise" to be minimised by breath averaging. However, several considerations should be kept in mind. First, it is preferable that the averaging period is chosen to include an integral number of actual breaths, rather than relying on a fixed interval that would variously truncate the initial breath in the period, i.e. depending on breathing frequency fR) characteristics. This effect will become less important with the use of longer averaging periods. However, as the duration of the averaging period is increased, so will the resulting average Vo2,peak value progressively fall short of the actual value. The magnitude of this shortfall will also depend on the WR incrementation rate being used. For example, assuming an incremental gain of 10 mL-min-W-1, with a 5 W-min-1 incrementation rate, the Vo2 increment for a 20-s period would be trivially small (i.e. 17 mL-min-1). However, for a AWR/At of 30 W-min-1, which some investigators would advocate for a more highly fit subject, the corresponding Vo2 increment would be 100 mL-min-1 and 150 mL-min-1 for a 20-and 30-s period, respectively. It has recently been stated that errors of the order of 20%

30 60

Incrementduration s

90 240

30 60

Incrementduration s

90 240

Time

Fig. 7. - a) Influence of work rate (WR) incrementation on oxygen uptake ( Vo2) and WR at peak exercise (O) and at the lactate threshold (•) for a series of incremental tests performed to the limit of tolerance. b) Schematic diagram of the Vo2 (-) and WR (----) responses to ramp-incremental exercise as a function of time for three different WR

incrementation rates. Note the lack of effect on peak VO2 (........) and VO2 at the lactate threshold. However, the influence of the Vo2 mean response time (horizontal solid lines) results in peak WR (----) being progressively higher, the faster the incrementation rate. (See text for details). Modified from [37] with permission.

Time

can accrue from this source [81]. There are no formal recommendations in this regard, although an interval of 30 s has been recommended [81].

Johnny Silva Cancer Research

» Chronic Obstructive 2