Abstract
In order to investigate underlying mechanisms, the present authors studied the effect of pulmonary rehabilitation on the regulation of total chest wall and compartmental (ribcage, abdominal) volumes during exercise in patients with chronic obstructive pulmonary disease.
In total, 20 patients (forced expiratory volume in one second, mean±sem 39±3% predicted) undertook high-intensity exercise 3 days·week-1 for 12 weeks. Before and after rehabilitation, the changes in chest wall (cw) volumes at the end of expiration (EEV) and inspiration (EIV) were computed by optoelectronic plethysmography during incremental exercise to the limit of tolerance (Wpeak).
Rehabilitation significantly improved Wpeak (57±7 versus 47±5 W). In the post-rehabilitation period and at identical work rates, significant reductions were observed in minute ventilation (35.1±2.7 versus 38.4±2.7 L·min-1), breathing frequency (26±1 versus 29±1 breaths·min-1) and EEVcw and EIVcw (by 182±79 and 136±37 mL, respectively). Inspiratory reserve volume was significantly increased (by 148±70 mL). Volume reductions were attributed to significant changes in abdominal EEV and EIV (by 163±59 and 125±27 mL, respectively). The improvement in Wpeak was similar in patients who progressively hyperinflated during exercise and those who did not (24 and 26%, respectively).
In conclusion, pulmonary rehabilitation lowers chest wall volumes during exercise by decreasing the abdominal volumes. The improvement in exercise capacity following rehabilitation is independent of the pattern of exercise-induced dynamic hyperinflation.
Therapeutic interventions, such as the use of bronchodilators 1, 2, oxygen supplementation 3, 4, lung transplantation 5 and lung volume reduction surgery 6, 7, have been documented to increase exercise tolerance in patients with chronic obstructive pulmonary disease (COPD) by reducing operational lung volumes during sub-maximal exercise. Furthermore, exercise training in COPD has also been shown to enhance exercise performance by reducing the ventilatory drive and respiratory discomfort at a given level of exercise 8. Recently, Porszasz et al. 9 showed that enhanced exercise tolerance after rehabilitation was mediated by the improved dynamic hyperinflation, only in part, because a significant number of patients improved exercise capacity without reducing end-expiratory dynamic hyperinflation. Similarly, Puente-Maestu et al. 10 reported that the reduction in end-expiratory lung volume accounted for 50% of the improvement in exercise tolerance in patients with severe COPD. These findings were attributed to the suggestion that, besides the improved dynamic hyperinflation, other factors might contribute to the enhanced exercise tolerance after rehabilitation 9, 10. Such factors may involve the reduction of the mechanical restriction on tidal volume (VT) expansion, an improved cardiac output, reduced competition between respiratory and locomotor muscles for the available oxygen supply and locomotor muscle reconditioning 9–11.
In addition, it is known that not all patients with COPD progressively hyperinflate during exercise. There have been reports demonstrating that, besides patients who dynamically hyperinflate from the onset of exercise, there are also patients who either reduce end-expiratory volume (EEV) during exercise 12 or do not change EEV until they approach the maximal exercise work rate (WR), also known as late hyperinflators (LH) 13. Therefore, exercise limitation in the latter two patient categories is not associated with end-expiratory dynamic hyperinflation and, as such, simply tracking changes in EEV during exercise is not informative of all the respiratory factors that could mediate improvements in exercise performance following an intervention. Assessment of all operational volumes is, therefore, important for understanding the respiratory mechanisms that contribute to the improved exercise capacity after rehabilitation in patients adopting different breathing strategies during exercise.
Accordingly, the main aim of the present study was to assess, by optoelectronic plethysmography (OEP), the effect of exercise training on the course of all operational chest wall volumes during exercise in patients with COPD. It is hypothesised that since exercise training is known to reduce the ventilatory requirement and the breathing frequency (fB) at a given level of sub-maximal exercise 8–10, this would reduce both the chest wall end-expiratory (EEVcw) and end-inspiratory volume (EIVcw) so that the required alveolar ventilation would be achieved at lower operating chest wall volumes. Providing that the chest wall volume at total lung capacity (TLCVcw) would not be affected by training, the adjustments in the EEVcw would be expected to increase the inspiratory reserve chest wall volume (IRVcw), thus alleviating dyspnoea sensation and enhancing exercise tolerance. In addition, in order to better understand the potential mechanisms of improvement in chest wall volume modulation after exercise training, the effect of rehabilitation on the regulation of the ribcage and abdominal chest wall volumes was also investigated.
METHODS
Subjects
In total, 20 patients with clinically stable COPD met the following entry criteria: 1) post-bronchodilator forced expiratory volume in one second (FEV1) <50% predicted and FEV1/forced vital capacity (FVC) <65% without significant reversibility (<12% change of the initial FEV1 value); 2) optimised medical therapy according to Global Initiative for Chronic Obstructive Lung Disease 14; and 3) absence of other significant diseases that could contribute to exercise limitation.
Study design
Patients participated in a pulmonary rehabilitation programme consisting predominantly of exercise training. Initially, patients were assessed for pulmonary function. Thereafter, they underwent an incremental cycle exercise test to the limit of tolerance (Wpeak) prior to entering the programme. The same procedure was followed upon completion of the programme.
Rehabilitation programme
The rehabilitation programme consisted of supervised exercise training and comprised of three sessions per week over a 12-week period. Patients enrolled in the rehabilitation programme performed either constant-load (initially at 60% Wpeak for 30 min; n = 10) or interval (initially at 100% Wpeak, 30 s of work alternated by 30 s of rest for 45 min; n = 10) cycling exercise on electromagnetically braked cycle ergometers (EC-1600 CatEye Ergociser; CatEye Co., Osaka, Japan) as previously described elsewhere 15–17. The WR was increased on a weekly basis as detailed later.
Pulmonary function assessment
Spirometry, lung transfer factor for carbon monoxide (Masterlab; Jaeger, Wurzburg, Germany) and subdivisions of lung volumes by body plethysmography (Medgraphics Autolink 1085D; Medical Graphics, St Paul, MN, USA) were performed according to recommended techniques 18.
Exercise testing protocol
Prior to entering the rehabilitation programme, all patients performed an incremental test on an electromagnetically braked cycle ergometer (Ergoline 800; Sensor Medics, Anaheim, CA, USA). The test consisted of measurements during 3 min of rest (R), followed by 3 min of unloaded pedalling and a ramp increase of load (increments of 5 or 10 W·min-1) to Wpeak. During the test, flow rate at the mouth and gas exchange variables were recorded breath-by-breath (Vmax 229; Sensor Medics). Cardiac frequency (fC) and arterial oxygen saturation measured by pulse oximetry were determined using the R-R interval from a 12-lead online ECG (Marquette Max; Marquette Hellige GmbH, Freiburg, Germany) and a pulse oximeter (Nonin 8600; Nonin Medical, Plymouth, MN, USA), respectively. Symptom ratings were monitored every 2 min throughout exercise using the 1–10 Borg scale 19. The V-slope technique was used to detect the oxygen uptake (V’O2) at which the anaerobic threshold (AT) occurred, whereas the change in V’O2 as a function of WR (ΔV’O2/ΔWR) was calculated as an index of aerobic work efficiency 20. In addition, the oxygen pulse, expressing the volume of oxygen extracted by the peripheral tissues with each beat (an index of stroke volume) was calculated by dividing V’O2 by fC 20.
Operational chest wall volume measurements
End-inspiratory volume (EIV) and EEV of the total chest wall and its compartments, the ribcage and the abdomen, were measured by OEP as previously described elsewhere 13, 21–23. To ensure that TLCVcw remained unchanged during exercise pre- and post-rehabilitation, patients were asked to perform inspiratory capacity manoeuvres during rest, unloaded cycling and every 2 min throughout the incremental test, according to previously described methods 23, 24. IRVcw was calculated by OEP as the difference between TLCVcw and EIVcw. Reproducibility of chest wall volume variations during exercise was established in a subgroup of four patients studied on two different occasions without pulmonary rehabilitation in the interim.
Statistical analysis
Data are presented as mean±sem. Two-way ANOVA with repeated measures was used to identify statistically significant differences in chest wall volumes across different time-points within and between groups, followed by the Tukey test for post hoc analyses. Baseline differences between groups concerning pulmonary function and exercise characteristics were tested by independent t-tests, whereas differences within each group in pre- and post-rehabilitation exercise characteristics were tested by paired t-tests. Linear regression analysis was performed using the least square method. Statistical significance was set at p<0.05.
RESULTS
Baseline patient characteristics
At baseline, patients were characterised by severe airflow limitation and moderately reduced diffusion capacity, as shown in table 1⇓. Patients were hyperinflated at rest as indicated by the increased total lung capacity, functional residual capacity and residual volume. Their exercise capacity was severely compromised (table 2⇓). Post-rehabilitation there were no significant changes in lung function characteristics compared with measurements at baseline.
Demographic and post-bronchodilator data of the study population and subgroups
Peak exercise data of study population#
Exercise training programme
The mean intensity sustained during the two training modalities increased progressively throughout the rehabilitation programme such that weeks 4, 8 and 12 corresponded to 105±13, 115±13 and 130±16% of baseline Wpeak during interval training, and to 63±3, 72±5 and 81±5% of baseline Wpeak during constant-load training, respectively. Examination of the mean training intensities (117±15 versus 72±5% of baseline Wpeak for interval versus constant-load training, respectively) revealed that the total amount of work sustained during the two training regimes was not significantly different.
Post-rehabilitation exercise tolerance
Following rehabilitation there were significant improvements in Wpeak (p = 0.001), total exercise endurance time (p = 0.001), peak V’O2 (p = 0.001) and AT (p = 0.011; table 2⇑). At an identical WR (i.e. that of peak pre-rehabilitation), there was a significant (p = 0.044) reduction in minute ventilation (V’E) during the post-training exercise test that was attributed to a significant (p = 0.045) reduction in fB (table 3⇓). This allowed for a significant increase in the expiratory time (tE; p = 0.039), since the inspiratory time did not change significantly (p = 0.33; table 3⇓). At an identical WR, post-training dyspnoea sensation was significantly (p = 0.042) lower compared with the pre-training exercise test (table 3⇓). The magnitude of change in all the aforementioned variables was not significantly different between the two training modalities.
Exercise data at an identical work rate(i.e. that of peak pre-rehabilitation) of the study population#
Operational chest wall volumes
The reproducibility of chest wall volume measurements was evaluated by testing four patients on two different occasions without rehabilitation in the interim. Close correlations were found between changes from rest (Δ) in the EIVcw, EEVcw and IRVcw recorded throughout all exercise stages (fig. 1⇓). The linear regression analyses between the two measurements (i.e. first and second) provided the following equations:
Regressions between changes from rest (Δ) in a) the end-expiratory chest wall volume (ΔEEVcw), b) the end-inspiratory chest wall volume (ΔEIVcw) and c) the inspiratory reserve chest wall volume (ΔIRVcw) calculated by optoelectronic plethysmography during unloaded cycling (○) at 33% incremental exercise to the limit of tolerance (Wpeak; •), at 66% Wpeak (□) and at 100% Wpeak (▪). Points were obtained from four patients studied on two occasions without rehabilitation in the interim. Line of identity (-----) is also shown.
ΔEIVcw2nd = 1.112 ΔEIVcw1st −0.0085 (r = 0.97, p<0.001) (1)
ΔEEVcw2nd = 0.983 ΔEEVcw1st −0.0074 (r = 0.93, p<0.001) (2)
ΔIRVcw2nd = 1.188 ΔIRVcw1st +0.127 (r = 0.92, p<0.001) (3)
The effect of rehabilitation on operational chest wall volumes is shown in figure 2⇓. Changes in EEVcw and EIVcw during exercise, both expressed as differences from rest (ΔEEVcw and ΔEIVcw, respectively) were significantly lower (p = 0.028 and p = 0.042, respectively) following rehabilitation compared with those before rehabilitation (fig. 2a⇓). TLCVcw during incremental exercise did not change following rehabilitation (fig. 2⇓).
Volume changes of a) the total chest wall, b) the ribcage and c) the abdomen before (▪, • and ▴) and after (□, ○ and ▵) the pulmonary rehabilitation programme during rest (R), unloaded cycling (0) and incremental exercise expressed as a percentage of peak pre-rehabilitation work rate. • and ○: end of inspiration; ▪ and □: end of expiration; ▴ and ▵: total chest wall volume at total lung capacity. All volumes are referred to the EEVcw at rest (-----). *: p<0.05 between pre- and post-rehabilitation.
At an identical WR after rehabilitation (i.e. that of peak pre-rehabilitation), ΔEEVcw decreased significantly by 182±79 mL (p = 0.028; table 3⇑), as did ΔEIVcw by 136±37 mL (p = 0.043). There was a significant increase (p = 0.011) in the IRVcw by 148±70 mL after rehabilitation at an identical WR, which was accompanied by a significant increase in the ratio of IRVcw over the TLCVcw (p = 0.045; table 3⇑). These changes in operational volumes during incremental exercise were attributed to significant reductions in mean abdominal EIV (by 125±27 mL, p = 0.020) and EEV (by 163±59 mL, p = 0.046; fig. 2c⇑). Accordingly, changes in abdominal EEV and EIV were correlated to ΔEEVcw (r = 0.92, p = 0.009) and ΔEIVcw (r = 0.91, p = 0.025), respectively. In addition, post-rehabilitation changes in abdominal EEV and EIV for each individual subject were significantly correlated with the changes in the individual subject's exercise endurance time (r = -0.68, p = 0.006 and r = -0.51, p = 0.041, respectively). The ribcage compartment EIV and EEV remained unchanged after rehabilitation (fig. 2b⇑). Operational total and compartmental chest wall volumes did not change significantly before and after rehabilitation at Wpeak (fig. 2⇑).
Patterns of dynamic hyperinflation
Prior to rehabilitation, 11 patients were characterised by progressive development of dynamic hyperinflation (EH), as indicated by the significant increase (p = 0.026) in ΔEEVcw from the first stage of exercise (unloaded pedalling, 0 W) reaching 605±128 mL at Wpeak (fig. 3a⇓); whereas the remaining patients were characterised by late development of dynamic hyperinflation (LH) as ΔEEVcw was significantly increased (p = 0.042; fig. 3a⇓) only at Wpeak by 161±78 mL (table 4⇓).
Volume changes of a, b) the total chest wall, c, d) the ribcage and e, f) the abdomen before (a, c, e) and after (b, d, f) the pulmonary rehabilitation programme during rest (R), unloaded cycling (0) and incremental exercise expressed as a percentage of peak pre-rehabilitation work rate in early hyperinflators (EH; □ and ○) and late hyperinflators (LH; ▪ and •). □ and ▪: end of inspiration; ○ and •: end of expiration; ▵: total chest wall volume at total lung capacity in EH; ▴: total chest wall volume at total lung capacity in LH. All volumes are referred to the end-expiratory chest wall volume at rest (----). ----: total chest wall volume during exercise. *: p<0.05 between EH and LH.
Peak exercise data of the subgroups pre- and post-rehabilitation
Comparisons of the spirometric and lung volume characteristics between EH and LH revealed significantly higher absolute and predicted FEV1 and FVC (table 1⇑), and a significantly lower degree of resting hyperinflation in LH compared with EH (table 1⇑). During exercise prior to rehabilitation, Wpeak was significantly higher in the EH group even though their baseline lung function was more impaired. Furthermore, besides the aforementioned difference in ΔEEVcw between the EH and the LH groups at baseline, the ratios of ΔV’O2/ΔWR and that of V’E in relation to the external WR at Wpeak, (i.e. peak V’E/peak WR), were significantly different between EH and LH, as were V’E, VT and IRVcw at Wpeak prior to rehabilitation (table 4⇑).
Following rehabilitation, both EH and LH significantly improved Wpeak (by 24 and 26%, respectively), peak V’O2 (by 21 and 18%, respectively), the oxygen pulse (by 18 and 17%, respectively), AT (by 25%) and the ratio of peak V’E/peak WR (by 17 and 14%, respectively; table 4⇑). At an identical WR, both EH and LH significantly (p<0.05) reduced V’E (by 3.2±1.5 and 3.5±1.8 L·min-1, respectively) and dyspnoea sensation (by 1.0±0.3 and 0.9±0.2, respectively). As shown in table 4⇑, EH and LH reached similar post-rehabilitation at Wpeak and pre-rehabilitation IRV and dyspnoea symptoms. The magnitude of change in these variables was not significantly different between EH and LH, and between interval and constant-load training modalities.
Total and compartmental chest wall volumes before and after rehabilitation in the two groups are shown in figure 3⇑. Prior to rehabilitation ΔEEVcw and the abdominal EEV were significantly different between EH and LH (fig. 3a⇑ and e). These significant differences between EH and LH persisted after rehabilitation (fig. 3b⇑ and f). No differences were found in the ribcage compartmental volumes among EH and LH.
DISCUSSION
The major findings of the present study are as follows: 1) rehabilitation reduced V’E and fB at a given level of exercise, as well as operational chest wall volumes; 2) the reductions of EEVcw and EIVcw were attributed to changes of the abdominal, and not the ribcage, compartment; and 3) the improvement in exercise capacity was similar in patients who progressively hyperinflated during exercise (EH) and those who did not (LH).
In accordance with previous studies 8–10, the present rehabilitative exercise training programme induced significant improvements in peak exercise capacity, peak V’O2 and the AT in patients with COPD. Such improvements have been shown to be independent of the exercise training modality implemented during rehabilitation 16, 17. Similarly, in the present study both constant-load and interval exercise training were equally effective in terms of enhancing exercise tolerance. Moreover, at an identical WR there were significant reductions in the ventilatory demand, fB and, hence, dyspnoea sensation. Reductions of similar magnitude in V’E have also been previously reported following exercise training 8–10. The reduced fB yielded a significant increase in tE, a response that allowed for greater lung emptying with each breath 8–10. This adaptation after rehabilitative exercise training was shown to be associated with smaller EEVcw and EIVcw at a given level of exercise and, since TLCVcw remained unchanged, IRVcw improved significantly, thereby yielding a considerably lower sensation of dyspnoea.
The magnitude of change in end-expiratory dynamic hyperinflation at an identical WR (by 182±79 mL) after exercise training is comparable to that reported for other therapeutic interventions, including rehabilitation 1–10. Bronchodilators 1 can lower the regulated chest wall volumes, however, retaining the same pattern of increase during exercise but shifting them towards lower operational points. Since the patients of the present study were under optimal pharmacological treatment and their resting lung function remained unchanged post-rehabilitation, improvements in chest wall volumes during exercise should be mainly attributed to the effect of rehabilitation.
The reductions in the operational chest wall volumes during exercise after rehabilitation were mainly reflected in the abdominal and not in the ribcage compartment (fig. 2⇑). Changes in the abdominal compartment reflect changes in diaphragmatic activity 25, which has been shown to be impaired by end-expiratory dynamic hyperinflation 26. The decrease in abdominal EEV post-rehabilitation compared with pre-rehabilitation could be the result of either increased abdominal muscle activity or the combination of prolonged tE with increased abdominal muscle contraction. Indeed, due to the lower fB, tE was significantly prolonged post-rehabilitation at an identical WR (table 3⇑). Abdominal muscle contraction during expiration is common at rest in COPD patients 12 and, in some cases, becomes much greater during exercise 13. This contraction may help partition the work of breathing between the main inspiratory and expiratory muscles 27. In healthy subjects, inward abdominal displacement during expiration stores elastic energy, which, when released, aids inspiration by allowing the diaphragm to descend 27, as could also be the case post-rehabilitation.
The present study is the first to examine the effect of rehabilitative exercise training by stratifying patients according to the pattern of EEV changes exhibited during exercise. The patients who exhibited progressive end-expiratory dynamic hyperinflation during exercise resembled those described as EH by Vogiatzis et al. 13. The patients who exhibited end-expiratory dynamic hyperinflation only at Wpeak, described by Vogiatzis et al. 13 as LH, also resembled to some degree those described by Aliverti et al. 12 in that they were able to reduce abdominal EEV throughout exercise. In the study by Aliverti et al. 12, as well as in the present study, the EH exhibited better exercise capacity at baseline compared with LH, despite worse baseline lung function characteristics (tables 1⇑ and 4⇑). It is necessary, therefore, to recognise important differences between EH and LH in terms of aerobic work efficiency.
Not allowing a progressive rise in lung EEV during exercise, as in the case of LH, has been shown to substantially increase the work of breathing in patients with COPD, mainly due to the greater recruitment of the expiratory (abdominal) muscles during expiration 12. In addition, at Wpeak the LH had a significantly higher VT and V’E compared with EH, despite similar peak V’O2 (table 4⇑). Hence, it is reasonable to suggest that the increase in V’O2 by the respiratory muscles in LH, consequent of the increase in the work of breathing, could effectively compete for a larger share of the whole body V’O2, thereby reducing that available to leg muscles 28, 29. Inadequate oxygen supply to meet metabolic demand of the limb muscles may play a more dominant role in limiting exercise capacity in some patients with COPD than impaired lung function or dynamic hyperinflation 29, 30. This may explain why the LH group exercised for less time at Wpeak (table 4⇑), reached lower peak WR and had a higher ratio of ΔV’O2/ΔWR compared with EH.
Regardless of the pattern of end-expiratory dynamic hyperinflation, both EH and LH improved peak work capacity and maximal ability to transport and extract oxygen in the peripheral muscles during exercise to a similar degree. As such, the degree of improvement in whole body V’O2, oxygen pulse (representing stroke volume) and AT was nearly identical between EH and LH (table 4⇑), as it was the reduction in the ventilatory requirement at an identical WR during the incremental test. This finding, reported by Porszasz et al. 9, may explain why patients who did not reduce end-expiratory dynamic hyperinflation after training still benefited from pulmonary rehabilitation.
In conclusion, the results of the present study demonstrate that rehabilitative exercise training is effective in reducing the operational chest wall volumes during exercise in patients with chronic obstructive pulmonary disease. These reductions are mainly attributed to decreases in the end-expiratory and end-inspiratory volumes of the abdominal compartment, reflecting prolongation of expiratory time and possibly training-induced adaptations of the respiratory muscles. Shifting operational volumes towards a lower operating point after rehabilitation resulting in an increased inspiratory reserve volume, thus allowing exercise tolerance to be prolonged before tidal volume expansion was severely restricted. The improvement in exercise capacity following rehabilitation was independent of the pattern of end-expiratory dynamic hyperinflation during exercise, since both early and late hyperinflators improved their work and oxygen utilisation capacities to a similar degree.
- Received September 15, 2006.
- Accepted October 30, 2006.
- © ERS Journals Ltd