Breath-by-breath changes of lung oxygen stores at rest and during exercise in humans

Respiratory Physiology & Neurobiology 164 (2008) 291–299

Contents lists available at ScienceDirect

Respiratory Physiology & Neurobiology journal homepage: www.elsevier.com/locate/resphysiol

Breath-by-breath changes of lung oxygen stores at rest and during exercise in humans Rob C.I. Wüst a,b , Andrea Aliverti c , Carlo Capelli d , Bengt Kayser e,∗ a

Institute for Biomedical Research into Human Movement and Health, Manchester Metropolitan University, Manchester, UK Research Institute MOVE, Faculty of Human Movement Sciences, Vrije University, Amsterdam, The Netherlands c TBM Lab, Dipartimento di Bioingegneria, Politecnico di Milano, IIT unit, Milano, Italy d School of Sport and Exercise Sciences, Department of Neurological and Visual Sciences, University of Verona, Verona, Italy e ISMMS, Faculty of Medicine, University of Geneva, Geneva, Switzerland b

a r t i c l e

i n f o

Article history: Accepted 3 June 2008 Keywords: Oxygen uptake Kinetics Breath-by-breath

a b s t r a c t ˙ O ) as the difference between When measuring breath-by-breath (BbB) oxygen uptake at the mouth (VM 2 the amount of inspired and expired oxygen, BbB variation in lung oxygen stores due to BbB variation in end-expiratory lung volume (VLET) introduces an error leading to a decreased signal-to-noise ratio when ˙ O ). Conventional BbB measurement of compared to oxygen uptake at the alveolo-capillary interface (VA 2 oxygen uptake does not measure BbB changes in lung oxygen stores. Opto-electronic plethysmography (OEP) allows accurate monitoring of absolute lung volume changes and BbB quantification of changes in pulmonary oxygen stores. To quantify BbB variation of lung oxygen stores and to assess variability ˙ O due to BbB variation of lung oxygen stores, we measured, VM ˙ O and VA ˙ O in parallel, at rest, in VM 2 2 2 during transients and during steady state cycling exercise at 60, 90 and 120 W in 7 healthy male subjects. ˙ O and VA ˙ O at steady state were not different (p = 0.328). Direct measurement of VA ˙ O reduced Average VM 2 2 2 ˙ O and VA ˙ O could overall BbB variability by 24% (p < 0.0001) and variance of the difference between VM 2 2 ˙ O was higher than VM ˙ O be explained for 55% by BbB changes in VLET and expiratory oxygen fraction. VA 2

2

15 and 30 s after exercise onset (p < 0.01). We conclude that (1) by taking into account changes in lung oxygen stores BbB variability of oxygen uptake is reduced, (2) alveolar oxygen stores change rapidly during transients to exercise, and (3) changes in alveolar oxygen stores affect BbB oxygen uptake measured at the mouth during the cardio-dynamic phase I. © 2008 Published by Elsevier B.V.

1. Introduction Measuring oxygen uptake (V˙ O2 ) at the mouth is the common way to indirectly assess aerobic metabolic activity at the cellular level. With the advent of modern gas analysis systems it became possible to measure V˙ O2 at the mouth on a BbB basis. This has increased the time resolution sufficiently to study oxygen uptake in the non-steady state, for example during on- or off-transients between rest and exercise when oxygen uptake varies, reflecting the changing rate of oxygen use by the organism (for recent papers see Ozyener et al., 2001; Jones and Poole, 2005; Whipp et al., 2005). However, BbB oxygen uptake measured at the mouth typically shows high variability, even in steady state conditions. At steady state, when averaged over time, BbB oxygen uptake at the mouth

∗ Corresponding author at: ISMMS, Faculté de Médecine, Université de Genève, 1211 Genève 4, Switzerland. Tel.: +41 22 3790028; fax.: +41 22 3790035. E-mail addresses: [email protected] (R.C.I. Wüst), [email protected] (A. Aliverti), [email protected] (C. Capelli), [email protected] (B. Kayser). 1569-9048/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.resp.2008.06.002

faithfully represents oxygen consumption of the whole organism. However, in the non-steady state BbB oxygen uptake at the mouth gives a distorted representation of metabolic cellular activity due to concomitant changes in the various bodily stores of O2 (lung, blood and peripheral tissue) and experimental sources of signal variability. In order to increase the signal-to-noise ratio it has thus become customary in studies of oxygen uptake kinetics to use averaging methods to decrease BbB oxygen uptake variability. Typically multiple tests are first purged of aberrant breaths (sighs, coughs, swallows and unfinished breaths), then ensemble averaged, and finally some smoothing method may be also applied (Lamarra et al., 1987; Ozyener et al., 2001; Rossiter et al., 2002). One of the main reasons for the variability in the BbB oxygen uptake is BbB variation in inspired and expired gas volume. Tidal volume varies both between breaths and within breaths in an unpredictable manner independent of the level of ventilation (Roecker et al., 2005) or metabolic rate (Lamarra et al., 1987, 1989). When inspired volume is higher than expired volume the endexpiratory lung volume increases; when measured at the mouth this will lead to an apparent increase of V˙ O2 for this breath due to

292

R.C.I. Wüst et al. / Respiratory Physiology & Neurobiology 164 (2008) 291–299

an increase in the lung oxygen store and vice versa for an inspired volume lower than expired volume. One way to come closer to measuring actual metabolic O2 consumption, therefore, is to measure oxygen uptake not at the mouth ˙ O ) but at the alveolar-capillary interface (VA ˙ O ). However, this (VM 2 2 is only possible by taking into account BbB changes in pulmonary gas stores (Auchincloss et al., 1966; Giezendanner et al., 1983; Cautero et al., 2002; Aliverti et al., 2004), especially in non-steady state conditions where significant changes in lung gas stores are likely to occur. At the onset of exercise, for example, expiratory muscle recruitment will lead to a sudden decrease in end-expiratory lung volume (Henke et al., 1988; Aliverti et al., 1997), which in turn will induce a decrease in end expiratory lung gas stores and thus to an underestimation of alveolar O2 uptake (di Prampero and Lafortuna, 1989). This problem has been recognized before and several studies have attempted to take into account estimates of changes in pulmonary gas stores (Auchincloss et al., 1966; Wessel et al., 1979; Beaver et al., 1981; Swanson and Sherrill, 1983; Grønlund, 1984; di Prampero and Lafortuna, 1989; Capelli et al., 2001; Cautero et al., 2003). However, all these studies only estimated changes of lung gas stores because until recently it was not possible to accurately measure actual absolute changes in lung volume on a BbB basis. A new method, opto-electronic plethysmography (OEP), now allows the measurement of changes of chest wall volume during breathing (Cala et al., 1996; Aliverti et al., 2000; Dellaca et al., 2001; Aliverti and Pedotti, 2003). By combining chest wall volume changes with measurements of vital capacity (VC) and functional residual capacity (FRC), it is possible to accurately determine absolute lung volumes at any point during the breathing cycle (Dellaca et al., 2001; Aliverti and Pedotti, 2003). This technique was recently adapted to assess changes in lung gas stores for the assessment of BbB oxygen uptake at the gas–blood interface in the lung (Aliverti et al., 2004). In the present study we applied this method for measuring actual BbB alveolar oxygen uptake in order to quantify, on a BbB basis, the variation in oxygen stores remaining in the lung at the end of each breath and to assess differences between classic BbB V˙ O2 measured at the mouth and actual BbB oxygen uptake at the alveolar level. We hypothesized that the variability of alveolar BbB gas exchange would be lower as compared to that measured at the mouth and that most of the BbB variability is due to BbB changes in lung gas stores. In addition, in order to test the robustness of this new method of alveolar oxygen uptake measurement we asked the subjects to voluntarily change their end-inspiratory and endexpiratory lung volumes on a BbB basis to amplify the effects of changes in gas stores in the lung on the difference between BbB ˙ O and VA ˙ O . For a given breath a small inspiration followed VM 2 2 by a big expiration should result in negative oxygen consumption (i.e. oxygen excretion) measured at the mouth while correction for the concomitant decrease in lung oxygen stores should yield a physiological oxygen uptake measured at the alveolar level.

2. Methods 2.1. Participants Seven non-smoking, healthy male subjects with normal lung function, between 21 and 48 years old (mean 35 years), volunteered to participate. The participants were fully informed about the nature of the protocol before signing an informed consent and were free to withdraw from the study at any time. The study conformed to the standards set by the Declaration of Helsinki and was approved by the local institutional ethical committee.

2.2. Equipment To measure chest wall volume (VCW) by OEP (Smart OEP system, BTS, Milano, Italy), 89 reflective markers were placed on the surface of the trunk ventrally and dorsally as previously described (Cala et al., 1996). The subjects were studied seated on an electromagnetically braked cycle ergometer (Excalibur, Lode, Groningen, The Netherlands) with their arms on rests at the level of the shoulder to allow visualization of markers placed in the mid-axillary line. The three-dimensional displacement of each of the reflective markers was measured by means of six video cameras placed in front and behind the subject. VCW was calculated using Gauss’ theorem (Cala et al., 1996). The system was equipped with an analog-to-digital data-acquisition system allowing simultaneous recording of chest wall volume and expiratory gas fractions. The subjects wore a nose clip and breathed through a mouthpiece connected in series with an ultrasonic air flow meter (TUBA, GHG, Zurich, Switzerland, dead space volume 35 mL). Volume and gas calibrations were performed before each test, using a 3 L calibration syringe (Hans Rudolph Inc., Kansas city, Missouri, USA) and gas mixtures of 16% O2 , 4% CO2 and 80% N2 , of 11% O2 and 89% N2 and ambient air. Expired gas samples were obtained through a needle in the mouthpiece to measure O2 and CO2 concentrations (FO2 and FCO2 ) continuously with a quadrupole mass spectrometer (Airspec 2200, Biggin Hill, Kent, UK), which was calibrated before each test. The data sampling frequency of all signals was 60 Hz and the data were stored on computer for later analysis. 2.3. Protocol Prior to the exercise test, functional residual capacity (FRC) was measured by nitrogen washout using a standard protocol (Sensor Medics Vmax metabolic cart, Yorba Linda, California, USA). In short, after several minutes of quiet breathing, while seated on the cycle ergometer, the subjects first performed a vital capacity (VC) maneuver. After some more quiet breathing the subjects then inspired 100% O2 until end-tidal N2 concentration was <1%. Chest wall volume was simultaneously measured by OEP. The volume of N2 expired with each breath was calculated from the expired FO2 and FCO2 , these volumes were summed, and, from the total amount of N2 expired, the volume of gas in the lung at the beginning of the washout period was calculated. With the results of the VC maneuver combined with the measurement of FRC the subdivisions of lung volume were determined (residual volume, expiratory and inspiratory reserve volume). Raw chest wall volume measured with OEP was then converted into absolute lung volume (VL) after subtracting non-gaseous thorax content. After determination of FRC and other lung volume variables, the subjects remained seated on the cycle ergometer and breathing quietly for 6 min. They then performed a 6 min exercise test at 60 W with pedal frequency set at 60–70 rpm. Data acquisition for the measurement of chest wall volume changes and gas fractions began 1 min prior to the onset of exercise and continued during the exercise period with the exception of the 5th minute due to the maximum time window for continuous data acquisition of the OEP system. Data acquisition was started again during the last minute of the exercise and continued for 4 min after the subject had stopped pedaling in order to assess the off-transient. The participants then breathed quietly at rest for 6 min and the protocol was repeated at 90 and 120 W (see Fig. 1). After completing the 120 W level and after 5 min of quiet breathing at rest, the participants were asked to voluntarily vary their end-inspiratory and end-expiratory lung volumes between individual breaths during 1 min to assess the effect of changes in operational lung volumes on the variation in oxygen

R.C.I. Wüst et al. / Respiratory Physiology & Neurobiology 164 (2008) 291–299

293

Fig. 1. Graphical representation of the experimental protocol and actual data of end-expiratory lung volume and O2 fraction in the expired air in one representative subject. Data acquisition was started 1 min before every on- or off-transient and was limited to 5 min periods interspaced with 1 min pauses used to pre-treat and store OEP data. Data were averaged over 1 min periods (A, B, C, D and E) for each of the exercise levels. Periods B were also subdivided in 15 s periods to increase resolution during on- and off-transitions.

uptake. The whole protocol was performed twice for each participant.

tory volume was used to calculate VIO2 and the following expiratory volume to calculate VEO2 , for that particular breath.

2.4. Measurement of BbB gas exchange

2.4.4. Synchronization of O2 and VL signals Large errors in BbB gas analysis can be made in situations where gas sampling and flow measurements are not tightly synchronized. This is especially true at higher breathing frequencies, possibly leading to large errors in VO2 measurements (Proctor and Beck, 1996; Roecker et al., 2005). For the apparatus we have used, delay times of 700 ms were necessary to synchronize the gas fractions and flow signals, following the procedure described previously by Aliverti et al. (2004).

2.4.1. Determination of zero-flow points In order to separate inspiratory from expiratory flow, zero flow points were determined as the nearest point to zero in the flow data when chest wall volume changed direction. 2.4.2. Correction for vapor pressure and temperature The raw flow data were first corrected for differences in vapor pressure during inspiration and expiration taking into account actual barometric pressure, ambient temperature and humidity. Inspiratory flow was converted from ATPS (corrected for ambient humidity) to STPD and expiratory flow from BTPS to STPD. Flow data were then filtered with a low-pass (8 Hz) bi-directional Butterworth 4th order filter before integrating to obtain the volume. 2.4.3. Correction for integrator drift During exercise alveolar pressure varies sufficiently with respect to barometric pressure to lead to gas compression and expansion. Consequently we did not use OEP VCW measurements to calculate inspired oxygen volume (VIO2 ) and expired oxygen volume (VEO2 ) because of the expected differences between VCW and VL over the breathing cycle and at end-expiration due to gas compression and blood shifts. However, this difference can be considered negligible at end-inspiration when, in healthy subjects, alveolar pressure is atmospheric. Therefore, after integrating the flow signal we corrected the difference between end-inspiratory volume, as measured by integrated flow, with that of end-inspiratory VL as measured by OEP (Aliverti et al., 2004). For each breath, the endinspiratory volume, measured by integrating flow, was set to be equal to VL measured by OEP after which the integrated flow was corrected for the drift by setting the next end-inspiratory volume equal to VL. The corrected volume obtained from the flow signal was then used to calculate BbB ventilation, the preceding inspira-

˙ O and VA ˙ O 2.4.5. Calculation of VM 2 2 This section is essentially the same as the theory section of a previous paper (Aliverti et al., 2004). It is summarized for the sake of convenience. The net transfer of oxygen at the alveolar level (VAO2 ) is the difference of the gas exchange at the mouth (VMO2 ) minus the changes of the alveolar O2 stores during breath i (VSO2 ): VAO2 = VMO2 − VSO2

(1)

VSO2 itself is made up of two components: the change in stores due to the inequality of inspired and expired tidal volume at constant alveolar fraction of O2 (FAO2 ), and the change in FAO2 during the course of the breath at constant alveolar gas volume (VL) (Auchincloss et al., 1966): VO2,s = FAO2(i) · (VL(i) − VL(i−1) ) + VL(i−1) · (FAO2(i) − FAO2(i−1) ) (2) and VAO2 = VMO2 − [FAO2(i) · (VL(i) − VL(i−1) ) +VL(i−1) · (FAO2(i) − FAO2(i−1) )]

(3)

where i represents the ith breath and i − 1 the immediate preceding breath. These interrelationships are illustrated in Fig. 2 which is a plot of O2 fraction (FO2 ) at the mouth against absolute VL. Point B

294

R.C.I. Wüst et al. / Respiratory Physiology & Neurobiology 164 (2008) 291–299

and S.D. of the mean of the 4 one-minute-periods of rest. Periods B were subdivided in 15 s intervals to increase the accuracy during the first minute of onset and offset of exercise, respectively. Differences ˙ O were analyzed using a 2 (two trials) × 8 ˙ O and VM between VA 2 2 ˙ O and (periods: A through E) × 3 (intensities) × 2 (methods: VA 2 ˙VMO ) ANOVA repeated measures. Since there was no significant 2 effect between the two trials, data were averaged and further analyzed using an 8 × 3 × 2 ANOVA repeated measures. Other variables, such as oxygen stores and lung volume were analyzed using an 8 × 3 × 2 ANOVA repeated measures. Mauchly’s sphericity test was used to check for homogeneity of variance. Violations of this assumption were corrected with the Greenhouse–Geisser F-value. If there was a significant effect of intensity, a post hoc test was performed to locate the exact difference between the trials. Differences were considered significant using an alpha of 0.05. Results are presented as mean (S.D.). Fig. 2. O2 fraction (FO2 ) absolute lung volume (VL) diagram during a single breath, from which change in the amount of O2 stored in the lung is calculated. For details on how this is done, see text (adapted from Aliverti et al., 2004).

gives FO2 and VL at the end of breath i − 1 when the volume of O2 remaining in the lung is given by the area ABCO, the product of OA (end-expiratory FO2 ) and OC (end expiratory VL). Inspiration of the ith breath proceeds from B to D, so the volume of O2 inspired is given by the area BDEC. Expiration proceeds from D to G; the VL expired is less than that inspired, and the FAO2 at the end of breath i is less than it was at the end of breath i − 1. The expired volume of O2 is given by the area GDEH, and the VMO2 is BDEC − GDEH, or the area BDGHC. The volume of O2 remaining in the lung at the end of the ith breath is given by FGHO. This is less than the amount remaining at the beginning of the breath, because of a decrease in FAO2 by an amount equal to ABKF = VL(i−1) · [FAO2(i) − FAO2(i−1) ] but is greater than at the beginning by an amount equal to KGHC = FAO2(i) · [VL(i) − VL(i−1) ] because the expired volume was less than the inspired volume, thereby adding to alveolar O2 stores. Correcting for changes in alveolar gas stores by subtracting KGHC from VMO2 and adding ABKF, reveals that the desired quantity VAO2 is given by the area ABDGF. The combination of OEP, with an independent measure of the subdivisions of lung volume, allows accurate tracking of absolute lung volume, and, in normal subjects at least, the measurement of endexpiratory O2 concentration is a reasonable estimate of alveolar concentration. Thus all terms on the right-hand side of Eq. (3) can be measured, allowing a direct measurement of VO2A . Dividing the measurement of VAO2 for each breath by the period for that breath ˙ O . gives a BbB measurement of VA 2 2.5. Data analysis After initial data analysis (Matlab, the Mathworks, Natick, Massachusetts, USA), the results for the exercise protocol were divided into 1 min periods (see Fig. 1). Every 5 min data acquisition period was thus subdivided in 5 periods, A, B, C, D and E for each level of exercise intensity (60, 90 and 120 W) and for the on- or off-set of exercise (-on, respectively -off). Periods B, corresponding to the onand off-transitions, were subdivided in 15 s periods to increase resolution (B1 to B4). The data for the last part of the protocol when the subjects voluntarily changed their end inspiratory and expiratory lung volume were analyzed separately. 2.6. Statistics All statistical analysis were carried out with SPSS, version 11.5 (SPSS Inc., Chicago, IL, USA). Mean values of BbB V˙ O2 and standard deviation (S.D.) during each 1-min-period (period A–E) were calculated at 60, 90 and 120 W for each subject as were the mean

3. Results 3.1. Subject characteristics All participants had normal lung function. Mean functional residual capacity and total lung capacity were 4.5 (0.9) and 7.6 (1.0) L, respectively. An average of 588 (73) breaths per subject, per test, were analyzed. In one subject during the second test the position of the OEP cameras was accidentally changed due to an earthquake leading to invalid lung volume data and hence in this subject only one test could be analyzed. 3.2. Oxygen uptake 3.2.1. Rest and steady state Fig. 3 shows individual BbB V˙ O2 data points for one subject illustrating the reduction in variability in BbB V˙ O2 during rest and exercise. Table 1 summarizes the V˙ O2 levels, at the mouth and at alveolar level, at steady state, at rest before each exercise level, during the last minute of exercise at 60, 90 and 120 W (periods A), and during the variable breathing period. Overall, by correction for pulmonary gas stores using OEP, BbB variability (measured using 1 min periods) decreased by 29.0 (14.7)% during steady state (periods A) and 29.6 (10.9)% during transients (periods B1) (Table 1). ANOVA repeated measures (4 intensities and 2 methods) revealed ˙ O compared to VA ˙ O a significant larger standard deviation for VM 2

2

(p < 0.001, Figs. 3 and 4). There was no significant effect of intensity or interaction effect on the standard deviation of the 1 min periods. ˙ O and VA ˙ O Overall, variance (R2 ) of the difference between VM 2

2

could be explained for 46% by BbB changes in VLET and 9% by BbB changes in end-expiratory FO2 . Mean V˙ O2 values during steady-state exercise were only significantly different between the intensities (p < 0.001) and not between the two methods (p = 0.328) (Fig. 5C and D). There was no statistical ˙ O and VA ˙ O during variable breathing difference between mean VM 2 2 and during the resting periods just before exercise started (Table 1 and Fig. 6). During variable breathing, standard deviation of BbB ˙ O was significantly higher than that of V˙ O (p < 0.001; Fig. 6). VM 2 2A 3.2.2. On- and off-transients A significant effect of period was observed in the difference ˙ O and VA ˙ O during on-transients (6 × 8 ANOVA between VM 2 2 ˙ O − VA ˙ O , p < 0.001, Fig. 5D). This effect repeated measures with VM 2 2 ˙ O ˙ O compared to VA was due to a significantly lower value of VM 2

2

during the first 15 s of the on-transients (B1) (Fig. 5C), which persisted during period B2 (p = 0.01). No significant differences ˙ O were found at other periods. Pearson’s ˙ O and VA between VM 2 2

R.C.I. Wüst et al. / Respiratory Physiology & Neurobiology 164 (2008) 291–299

295

˙ O ) and corrected for alveolar gas stores (VA ˙ O ). Note the significant higher variation in VM ˙ O Fig. 3. Representative graph for oxygen uptake as measured at the mouth (VM 2 2 2 ˙ O . during rest, transient to and steady state exercise of 60, 90 and 120 W compared to VA 2

Table 1 ˙ O ) and at the alveolar level (VA ˙ O ) Average values and standard deviations of BbB V˙ O2 during the last minute of the 6-min-period, measured at the mouth (VM 2 2

Rest 60 W Rest 90 W Rest 120 W Variable breathing

˙ O (L min−1 )† Mean VM 2

˙ O (L min−1 )† Mean VA 2

˙ O (L min−1 )* , # S.D. VM 2

˙ O (L min−1 )* S.D. VA 2

Decrease in S.D. (%)

0.251 (0.071) 0.824 (0.051) 0.214 (0.059) 1.077 (0.081) 0.272 (0.051) 1.335 (0.115) 0.306 (0.176)

0.251 (0.068) 0.836 (0.086) 0.249 (0.052) 1.071 (0.093) 0.251 (0.045) 1.343 (0.113) 0.246 (0.049)

0.437 (0.178) 0.358 (0.105) 0.444 (0.181) 0.457 (0.141) 0.473 (0.215) 0.562 (0.354) 2.208 (0.542)

0.220 (0.088) 0.306 (0.165) 0.251 (0.167) 0.353 (0.094) 0.237 (146) 0.356 (0.109) 0.323 (0.178)

44.9 (14.7) 8.1 (45.2) 38.1 (31.5) 18.0 (18.3) 41.6 (27.6) 23.2 (32.5) 84.5 (10.3)

Values are mean (S.D.), n = 7. † Effect of intensity during steady state exercise (p < 0.001). This effect was significant between all intensities (p < 0.001). * Difference between VM ˙ O and VA ˙ O (p = 0.01). 2 2 # Difference between variable breathing and rest (p < 0.001).

˙ O and VA ˙ O and correlation between the difference between VM 2 2 changes in end-expiratory lung volume during B1 on-transients was 0.847 (p < 0.0001), while Pearson’s correlation coefficient ˙ O and VA ˙ O and changes in between the difference between VM 2 2 end expiratory FO2 was −0.08 (NS). During the early phase of exercise onset, end-expiratory FO2 and VL decreased significantly (p < 0.001) (Fig. 5A and B), resulting in a marked decrease in pulmonary oxygen stores (p < 0.001) (Fig. 7). A significant interaction between intensity and period was ˙ O and VA ˙ O during the offfound in the difference between VM 2 2

˙ O transients (p < 0.05). This effect was due to a significant lower VM 2 in period B1 of the 120 W off-transient (p < 0.05; Fig. 5D). 4. Discussion The main purpose of this study was to directly measure variation of pulmonary oxygen stores present in the lungs at the beginning of each breath, at rest, during on- and off-transients and steady state exercise. The main findings were first that pulmonary oxygen stores vary considerably between sequential breaths, especially

˙ O was Fig. 4. Standard deviation of oxygen uptake during 1 min periods (15 s intervals for the B-period). Throughout the whole protocol, the standard deviation of VM 2 ˙ O . Letters A–E represent the mean values for each minute, where B is split up in four periods of 15 s (see Section 2.5 for more details). significantly higher (p < 0.001) than VA 2

296

R.C.I. Wüst et al. / Respiratory Physiology & Neurobiology 164 (2008) 291–299

ble that expiratory FO2 is influenced by respiratory frequency and intrapulmonary gas mixing, independently from changes in VLET. In the past, many different attempts were made to correct for lung oxygen stores (Auchincloss et al., 1966; Wessel et al., 1979; Beaver et al., 1981; Swanson and Sherrill, 1983; Grønlund, 1984; di Prampero and Lafortuna, 1989). However, none of these methods accurately determines the BbB changes in absolute lung volume and therefore can only estimate alveolar oxygen uptake. Choosing a constant value of VL(i − 1) (as was done by Beaver et al., 1981; Giezendanner et al., 1983; Swanson and Sherrill, 1983; di Prampero and Lafortuna, 1989) introduces errors in the calculation of alveolar gas exchange at the onset of exercise, because VL(i − 1) (Linnarsson, 1974; Aliverti et al., 1997, 2004) and end-expiratory oxygen fraction (Suskind et al., 1950; Beaver et al., 1981) rapidly decrease during transients from rest to exercise (Fig. 5A and B). Aliverti et al. (1997) and Henke et al. (1988) showed that this is caused by an increase in expiratory muscle recruitment at the onset of exercise. This leads to a rapid decrease in end-expiratory pulmonary oxygen stores which is, in part, not taken into account using the classical algorithms of BbB oxygen uptake measurement. OEP adequately overcomes this problem and, after correction of pulmonary gas stores, actual oxygen uptake at the alveolar-capillary membrane can now be determined. 4.2. Transients

Fig. 5. Mean values and standard deviation for the change in end-expiratory lung volume (A), end-expiratory oxygen fraction (B), oxygen uptake consumption (C) and the difference between alveolar oxygen uptake and as measured at the mouth (D) during the protocol (on and off-transients at 60, 90 and 120 W) for all participants. Letters A–E represent the mean values for each minute, where B is split up in four ˙ O compared periods of 15 s (see Section 2.5 for more details). Significant higher VA 2 ˙ O were measured at the first 30 s of exercise onset. to VM 2

during exercise transients but also at steady state. Secondly, by taking into account actual changes in pulmonary oxygen stores BbB variability of oxygen uptake is reduced substantially (∼29%). Lastly, that pulmonary oxygen stores change rapidly during transients to and from exercise and that this affects oxygen uptake at the mouth at exercise on- and off-set. 4.1. Measurement of absolute lung volume When oxygen uptake is measured as the difference between inspired volume of oxygen and expired volume of oxygen, BbB ˙ O introduces large errors when interpreting it as VA ˙ O VM 2 2 (Figs. 3 and 4). Even negative oxygen uptake values can be measured at the mouth, when more oxygen is expired than inspired, largely because of differences in expired and inspired volume in one ˙ O and VA ˙ O are caused by breath (Fig. 6). Differences between VM 2 2 factors, such as differences between inspired and expired volume at constant oxygen fraction and changes in oxygen fraction at constant lung volume (see Section 2 and Fig. 5). These factors explain ˙ O and ∼55% of the variance between the difference between VM 2 ˙ O , suggesting that other factors must play a role too. It is possiVA 2

The results from this study, the first based on actual measurements of absolute lung volume, confirm that changes in pulmonary gas stores of oxygen have a major impact on the assessment of oxygen uptake measurements at the alveolar-capillary membrane, especially within the first 30 s after the on-set of an exercise bout. The conventional way to describe oxygen uptake kinetics using BbB ˙ O avoids this problem by neglecting the initial 20–25 s after a VM 2 change in power output (also known as the cardio-dynamic phase, phase I; Lamarra et al., 1987; Ozyener et al., 2001). Instead of assigning this phase to increased pulmonary blood flow or changes in blood distribution among body compartments only, a major component might be the changes of the pulmonary oxygen stores. Our results corroborate these contentions since during the early stages of exercise (periods B1 and B2), oxygen uptake at the alveolarcapillary membrane is significantly higher than as measured at the mouth (Fig. 5C). This was mainly due to a reduction of pulmonary oxygen stores present in the lung at the end of each breath (Fig. 7). This finding is in agreement with Cautero et al. (2002) who demonstrated that, by applying the algorithm of Grønlund (1984) to estimate BbB changes of alveolar gas stores, the time constant of the phase II was shorter than that calculated on the basis of the BbB data obtained by the algorithm of Auchincloss et al. (1966), which assumes a given absolute vale of VL(i − 1) . Moreover, the time constant of oxygen uptake kinetics was linearly related to VL(i − 1) , increasing from about 35 s for VL(i − 1) = 0 to 47 s for VL(i − 1) equal to FRC + 0.5 L. This was essentially due the fact that the absolute value of VL(i − 1) appearing in the equations needed to estimate gas lung stores variations, amplifies the contribution of the alve˙ O . In fact, alveolar olar gas fraction differences to the calculated VA 2 gas fraction differences during the transient phase at the onset of exercise, estimated by means of the corresponding end-tidal surrogates, are not nil, do not conform to a random time series, and do follow a deterministic behavior that is progressively transformed into a stochastic one after about 80 s (Cautero et al., 2002). As such, the multiplying factor VL(i − 1) amplifies the weight of alveolar gas ˙ O . volume differences and changes VA 2 ˙ O as compared An additional explanation for the more rapid VA 2 ˙ O was recently proposed by Lador et al. (2006) who sugto VM 2

R.C.I. Wüst et al. / Respiratory Physiology & Neurobiology 164 (2008) 291–299

297

Fig. 6. Representative graphs for one subject to show the existence of negative values for oxygen ‘uptake’ when measuring oxygen exchange at the mouth. This subject ˙ O . All panels: circles indicate the start of ˙ O , open circles VA voluntarily varied his end-expiratory and inspiratory lung volume during rest. Top panel: closed circles, VM 2 2 inspiration. At t = 14 s, a small inspiration is followed by a bigger expiration, resulting in a decrease in absolute lung volume. This led to a negative oxygen consumption as measured at the mouth, compared to at the alveolar level (top panel). However, this ‘oxygen excretion’ only reflected a decrease in oxygen stores in the lung and not a decrease in oxygen uptake at the alveolar membrane. The same holds true for the breath starting at t = 50 s. VTin is inspired tidal volume and VTout is expired tidal volume.

Fig. 7. Changes in oxygen stores in the lungs during rest and exercise. These changes were due to changes in end-expiratory lung volume and in end-expiratory O2 fraction (see Fig. 5 and Eq. (2)). Letters A–E represent the mean values for each minute, where B is split up in four periods of 15 s (see Section 2.5 for more details).

298

R.C.I. Wüst et al. / Respiratory Physiology & Neurobiology 164 (2008) 291–299

gested that faster kinetics at exercise on-set of cardiac output (Q˙ ) as compared to metabolic O2 consumption kinetics, combined with an initially invariant alveolar-capillary oxygen difference according to ˙ O that increases more rapidly than the Fick principle, leads to a VA 2 metabolic oxygen consumption. This finding implies a redistribution of blood volume between body compartments. It remains to be understood and quantified how this redistribution leads to the ˙ O from metabolic oxygen consumption apparent uncoupling of VA 2 during on-transients. ˙ O and VA ˙ O During off-transients, the difference between VM 2 2 is smaller and only significantly different at higher intensities compared to lower intensities (120 W; Fig. 5D). Indeed, during offtransients absolute lung volume did return to pre-exercise values, although not rapidly (Fig. 5A). This led to a gradual increase in the pulmonary oxygen stores resulting in higher values for oxygen uptake measured at the mouth compared to at the alveolar level. These findings imply that post-exercise excess oxygen uptake is influenced by the gradual increase in pulmonary oxygen stores.

CO2 , however, this assumption is not correct and correction of V˙ CO2 for pulmonary CO2 stores will not lead to a decreased BbB variability in V˙ CO2 (di Prampero and Lafortuna, 1989 and unpublished results from this study). 4.4. Conclusions From the results from this study, it can be concluded that by using OEP for the measurement of BbB gas exchange, BbB V˙ O2 is less ‘noisy’ and better estimates ‘real’ alveolar gas exchange during rest, transients from rest to exercise and back to rest, and steady state exercise, while average steady state values remain the same. Expiratory muscle activation leads to a rapid decrease in lung volume at exercise on-set and thus to a reduction of lung oxygen stores resulting in an underestimation of BbB oxygen uptake during the first 30 s when measured at the mouth. Conversely, at the end of exercise lung oxygen stores increase again thus leading to an overestimation of the early part of post-exercise excess oxygen consumption when measured at the mouth.

4.3. Limitations Acknowledgements Even though our method allows us to reduce BbB oxygen uptake variability, a degree of variability remains. This variability comes from two sources: physiological sources and experimental error. BbB measurement of gas exchange is subject to considerable experimental error. Synchronizing flow with gas sampling is critical, as is the measurement of flow itself. We measured flow by means of a Doppler device and used an OEP ‘optimized’ flow for the ˙ O , which probably already reduced determination of classical VM 2 ˙ O substantially. Moreover, subtle changes in the variability of VM 2 temperature of the probe over time, dead space volume of the mouthpiece with air not going in or out of the subject’s body and differences in air density of inspired and expired air may have introduced experimental error. OEP measures changes in chest wall volume during breathing but cannot distinguish volume changes from air flow in and out of the lungs and displacement of blood into or out of the trunk. At the onset of exercise the muscle pump effect of muscle contraction will increase venous return towards the thorax resulting in an increased thorax blood volume. At the same time it is likely that a combination of increased ventilation and cardiac output will have led to increased alveolar capillary recruitment and increased lung circulation. Zavorsky et al. (2003) reported differences in cardio-pulmonary blood volume between resting and heavy exercise of 0.42 L per square meter of body surface area. Therefore, it is conceivable that our measurements somewhat underestimated the effective reduction in oxygen stores at exercise on-set. Our findings of a more rapid onset of oxygen uptake when measured at the alveolar-capillary membrane may thus be even more pronounced. Without simultaneous measurement of actual changes in lung blood volume this remains to be directly proven. ˙ O and VA ˙ O values are presented in this study. ValOnly VM 2 2 ues for the excretion of CO2 at the alveolar-capillary membrane can also be calculated using the same theory. The reason why these values are not presented is that CO2 elimination across the alveolarcapillary membrane fluctuates greatly with respect to changes in alveolar pressure and ventilation (Ward et al., 1983; Allen et al., 1984), resulting in a large variability in CO2 stores in the bloodstream, both arterial and venous. This is, however, not the case for O2 , because oxygen saturation level in the arterial blood can be assumed to be invariant and close to full saturation, at least at the low altitude of the present experiments (93 m above sealevel). Therefore, oxygen uptake measured at the alveolar level better reflects the processes taking place at the mitochondrial level. For

We gratefully acknowledge the excellent technical help of Michaela Cautero. RCIW was supported by a grant through the GJvIS Young Promising Scientist Award from the Vrije University, Amsterdam, the Netherlands. References Aliverti, A., Cala, S.J., Duranti, R., Ferrigno, G., Kenyon, C.M., Pedotti, A., Scano, G., Sliwinski, P., Macklem, P.T., Yan, S., 1997. Human respiratory muscle actions and control during exercise. J. Appl. Physiol. 83, 1256–1269. Aliverti, A., Dellaca, R., Pelosi, P., Chiumello, D., Pedotti, A., Gattinoni, L., 2000. Optoelectronic plethysmography in intensive care patients. Am. J. Respir. Crit. Care Med. 161, 1546–1552. Aliverti, A., Kayser, B., Macklem, P.T., 2004. Breath-by-breath assessment of alveolar gas stores and exchange. J. Appl. Physiol. 96, 1464–1469. Aliverti, A., Pedotti, A., 2003. Opto-electronic plethysmography. Monaldi Arch. Chest Dis. 59, 12–16. Allen, C.J., Jones, N.L., Killian, K.J., 1984. Alveolar gas exchange during exercise: a single-breath analysis. J. Appl. Physiol. 57, 1704–1709. Auchincloss Jr., J.H., Gilbert, R., Baule, G.H., 1966. Effect of ventilation on oxygen transfer during early exercise. J. Appl. Physiol. 21, 810–818. Beaver, W.L., Lamarra, N., Wasserman, K., 1981. Breath-by-breath measurement of true alveolar gas exchange. J. Appl. Physiol. 51, 1662–1675. Cala, S.J., Kenyon, C.M., Ferrigno, G., Carnevali, P., Aliverti, A., Pedotti, A., Macklem, P.T., Rochester, D.F., 1996. Chest wall and lung volume estimation by optical reflectance motion analysis. J. Appl. Physiol. 81, 2680–2689. Capelli, C., Cautero, M., di Prampero, P.E., 2001. New perspectives in breath-by-breath determination of alveolar gas exchange in humans. Pflugers Arch. 441, 566–577. Cautero, M., Beltrami, A.P., di Prampero, P.E., Capelli, C., 2002. Breath-by-breath alveolar oxygen transfer at the onset of step exercise in humans: methodological implications. Eur. J. Appl. Physiol. 88, 203–213. Cautero, M., di Prampero, P.E., Capelli, C., 2003. New acquisitions in the assessment of breath-by-breath alveolar gas transfer in humans. Eur. J. Appl. Physiol. 90, 231–241. Dellaca, R.L., Aliverti, A., Pelosi, P., Carlesso, E., Chiumello, D., Pedotti, A., Gattinoni, L., 2001. Estimation of end-expiratory lung volume variations by optoelectronic plethysmography. Crit. Care Med. 29, 1807–1811. di Prampero, P.E., Lafortuna, C.L., 1989. Breath-by-breath estimate of alveolar gas transfer variability in man at rest and during exercise. J. Physiol. 415, 459–475. Giezendanner, D., Cerretelli, P., Di Prampero, P.E., 1983. Breath-by-breath alveolar gas exchange. J. Appl. Physiol. 55, 583–590. Grønlund, J., 1984. A new method for breath-to-breath determination of oxygen flux across the alveolar membrane. Eur. J. Appl. Physiol. Occup. Physiol. 52, 167–172. Henke, K.G., Sharratt, M., Pegelow, D., Dempsey, J.A., 1988. Regulation of endexpiratory lung volume during exercise. J. Appl. Physiol. 64, 135–146. Jones, A.M., Poole, D.C., 2005. Oxygen uptake dynamics: from muscle to mouth—an introduction to the symposium. Med. Sci. Sports Exerc. 37, 1542–1550. Lador, F., Azabji Kenfack, M., Moia, C., Cautero, M., Morel, D.R., Capelli, C., Ferretti, G., 2006. Simultaneous determination of the kinetics of cardiac output, systemic O(2) delivery, and lung O(2) uptake at exercise onset in men. Am. J. Physiol. Regul. Integr. Comp. Physiol. 290, R1071–1079. Lamarra, N., Ward, S.A., Whipp, B.J., 1989. Model implications of gas exchange dynamics on blood gases in incremental exercise. J. Appl. Physiol. 66, 1539–1546.

R.C.I. Wüst et al. / Respiratory Physiology & Neurobiology 164 (2008) 291–299 Lamarra, N., Whipp, B.J., Ward, S.A., Wasserman, K., 1987. Effect of interbreath fluctuations on characterizing exercise gas exchange kinetics. J. Appl. Physiol. 62, 2003–2012. Linnarsson, D., 1974. Dynamics of pulmonary gas exchange and heart rate changes at start and end of exercise. Acta Physiol. Scand. Suppl. 415, 1–68. Ozyener, F., Rossiter, H.B., Ward, S.A., Whipp, B.J., 2001. Influence of exercise intensity on the on- and off-transient kinetics of pulmonary oxygen uptake in humans. J. Physiol. 533, 891–902. Proctor, D.N., Beck, K.C., 1996. Delay time adjustments to minimize errors in breathby-breath measurement of Vo2 during exercise. J. Appl. Physiol. 81, 2495–2499. Roecker, K., Prettin, S., Sorichter, S., 2005. Gas exchange measurements with high temporal resolution: the breath-by-breath approach. Int. J. Sports Med. 26 (Suppl. 1), S11–S18. Rossiter, H.B., Ward, S.A., Kowalchuk, J.M., Howe, F.A., Griffiths, J.R., Whipp, B.J., 2002. Dynamic asymmetry of phosphocreatine concentration and O(2) uptake between the on- and off-transients of moderate- and high-intensity exercise in humans. J. Physiol. 541, 991–1002.

299

Suskind, M., Bruce, R.A., McDowell, M.E., Yu, P.N.G., Lovejoy Jr., F.W., Vernarelli, S.J., 1950. Normal variations in end-tidal air and arterial blood carbon dioxide and oxygen tensions during moderate exercise. J. Appl. Physiol. 3, 282–290. Swanson, G.D., Sherrill, D.L., 1983. A model evaluation of estimates of breath-tobreath alveolar gas exchange. J. Appl. Physiol. 55, 1936–1941. Ward, S.A., Whipp, B.J., Koyal, S., Wasserman, K., 1983. Influence of body CO2 stores on ventilatory dynamics during exercise. J. Appl. Physiol. 55, 742–749. Wessel, H.U., Stout, R.L., Bastanier, C.K., Paul, M.H., 1979. Breath-by-breath variation of FRC: effect on VO2 and VCO2 measured at the mouth. J. Appl. Physiol. 46, 1122–1126. Whipp, B.J., Ward, S.A., Rossiter, H.B., 2005. Pulmonary O2 uptake during exercise: conflating muscular and cardiovascular responses. Med. Sci. Sports Exerc. 37, 1574–1585. Zavorsky, G.S., Walley, K.R., Russell, J.A., 2003. Red cell pulmonary transit times through the healthy human lung. Exp. Physiol. 88, 191–200.