Eleven clinically stable patients with stage II–IV disease according to the Global Initiative for Chronic Obstructive Lung Disease guidelines were consecutively recruited on admission to a pulmonary rehabilitation center (Asthma Center Hornerheide, Horn, The Netherlands) and participated in an inpatient pulmonary rehabilitation program on weekdays over 8 wk (FEV₁ < 60% predicted and FEV₁/FVC < 70% and < 10% predicted improvement in FEV₁ after β₂-agonist inhalation) (10, 11). Additional information on the rehabilitation program is available in the online supplement. All patients were ex-smokers, were not depleted (12), did not use oxygen supplementation, and had not experienced respiratory tract infection or exacerbation of their disease for at least 4 wk before the study. Exclusion criteria were no other chronic diseases, such as rheumatoid arthritis and chronic colitis. Also, patients with diabetes, cardiovascular diseases, renal diseases, liver diseases, or mental diseases were excluded from the study. Eleven healthy, age-matched, nonsmoking participants were recruited as the control group. In addition, all participants were questioned on their dietary habits to ensure that none were taking antioxidants or vitamin supplements. Furthermore, all patients received anticholinergic and β₂-agonists as bronchodilator therapy as well as inhaled corticosteroids. Three patients also received theophylline and one patient was on oral corticosteroid therapy. This maintenance medication remained unchanged during the study. Written, informed consent was obtained from all participants, and the study was approved by the medical ethics committee of the University Hospital Maastricht.

The participants were instructed to abstain from strenuous physical activity on the test day. Moreover, patients with COPD were not allowed to take antioxidants or vitamin supplements during the rehabilitation program. At the beginning of the study, all participants underwent pulmonary function testing and anthropometric measurements, and they had to wear a physical activity monitor during 9 consecutive d. First, they completed an incremental cycle ergometry exercise test until exhaustion and, after 4 d, a submaximal constant work-rate exercise test was performed at 60% of the maximal power output achieved on the incremental exercise test. Peripheral venous blood, urine samples, and samples for measurements of breath condensate H₂O₂ were obtained at baseline, immediately and 4 h after both exercise tests. All tests were performed and samples were collected at the same time point of the day. The same procedure was strictly followed at the end of the pulmonary rehabilitation program for patients with COPD.

All participants underwent flow volume tests, including measurements of FEV₁ and FVC, with the highest value from at least three properly performed measurements being used for analysis. In addition, lung volumes and diffusion capacity were obtained in patients with COPD. Residual volume and total lung capacity (TLC) were assessed by whole-body plethysmography, and diffusion capacity for carbon monoxide (Dl_CO) was measured by using the single-breath method (Masterlab; Jaeger, Würzburg, Germany). The values obtained were expressed as a percentage of the reference value (13).

The daily physical activity level was assessed using the Physical Activity Monitor (Pam) accelerometer. Pam (type AM 100; Pam B.V., The Netherlands) is a small-sized, lightweight, unidirectional accelerometer. This accelerometer provides a valid and reliable method to estimate the energy expenditure of the whole body during walking patterns (see the online supplement for additional information).

Participants performed an incremental (10 W/min) cycle ergometry test as described previously (14). Expired gases were analyzed using breath-by-breath analysis through a breathing mask (Oxygen-β; Jaeger, Würzburg, Germany). Simultaneously, heart rate was registered every min (Polar Electro Cy, Kempele, Finland).

The submaximal exercise test was performed at 60% of the peak workload achieved during the incremental exercise test. The patients were instructed to cycle as long as possible but for a maximum of 30 min. For the control group, the endurance time was standardized at 12 min, which was the mean endurance time of the patients with COPD at baseline. In addition, oxygen consumption and heart rate were monitored as described in the incremental exercise test. Antecubital venous blood samples were taken before and at the end of both exercise tests to analyze lactate. Plasma lactate was determined enzymatically using an automated system (Cobas Mira; Roche, Basel, Switzerland).

Venous blood samples (10 ml) were drawn into ethylenediaminetetraacetic acid–containing tubes (Venoject; Terumo Corporation, Leuven, Belgium). All blood samples were immediately put on ice, and kept on ice during sample preparation. Plasma was obtained by centrifugation (800 × g for 10 min at 4°C) and stored at −80°C until analysis. Urine samples were collected in sterile containers and two 10-ml aliquots of urine were stored at −20°C until further analysis. Plasma, urine, and breath samples of one patient, gathered before and after rehabilitation, were processed in the same run for (bio)chemical analyses.

The comet assay (single-cell gel electrophoresis) was adapted from the method of Singh and colleagues (15) with minor modifications. After removal of the plasma (800 × g for 10 min at 4°C), the volume of the remaining blood was brought to 15 ml with cold phosphate-buffered saline (PBS) and layered on an equal volume of Lymphoprep medium in a 50-ml tube with a filter (Bio-One GmbH; Greiner, Frickenhausen, Germany) of PBMCs. After centrifugation at 800 × g for 30 min at 4°C, gradient-separated PBMCs were recovered, resuspended in 10-ml cold PBS, and centrifuged again at 250 × g for 10 min at 4°C. The cells were resuspended in 1 ml cold PBS. For the analysis of DNA damage, 10 μl of the cell suspension was mixed with 90 μl of 1% low–melting-point agarose of 37°C and layered onto prechilled agarose-coated (1.5% agarose in PBS) microscope slides. The slides were covered with a coverslip and kept at 4°C for 5 min to allow the low–melting-point agarose to solidify. Then, the coverslips were removed and the slides were immersed in a jar containing cold lysing solution (2.5 M NaCl, 100 mM Na₂ ethylenediaminetetraacetic acid, 10 mM Tris [pH 10], 1% Triton X-100, 10% dimethyl sulfoxide). For reasons of stability of the DNA during lysis, the PBMCs were kept at 4°C for a maximum of 6 wk (16). All samples of one patient taken before rehabilitation were processed in one run, and samples taken after rehabilitation in a separate run. After lysis, the slides were drained and placed in a horizontal ice-cold electrophoresis box filled with fresh alkaline electrophoresis buffer of 4°C (10 N NaOH, 200 mM Na₂ ethylenediaminetetraacetic acid [pH 12.5–13.5]) for 20 min to allow for DNA unwinding and expression of alkali-labile sites. The electrophoresis was subsequently conducted at 4°C, for 20 min, at 25 V and 300 mA. After electrophoresis, the slides were washed three times with cold neutralization buffer (0.4 M Tris, pH 7.5) and stored at 4°C until further analysis. All of these steps were conducted under dim light to prevent the occurrence of additional DNA damage. For microscopic analysis, 50 μl ethidium bromide (20 μg/ml) was added to each slide and 50 randomly selected cells were analyzed from each sample. Comet measurements were made by image analysis using a fluorescence microscope and the Comet Assay III software (Perceptive Instruments, Suffolk, UK). ROS-induced DNA damage was measured as tail moment (based on the product of the percentage of DNA in the tail and tail length). A representative comet image is shown in Figure E1 in the online supplement. Median, mean, and SE of the mean of tail moment were calculated from each sample. The median of each sample was taken because data had a nonnormal distribution.

The determination of MDA was assayed by measuring thiobarbituric acid–reactive substances according to the HPCL method described by Lepage and coworkers (17). The urinary MDA concentration was corrected for creatinine content. Creatinine in urine samples was measured in triplicate using the Jaffé reaction as previously described (18). The values of MDA in urine are expressed per mmol creatinine. Additional details for these measurements are provided in the online supplement.

An aliquot of 150 μl of plasma was deproteinized with 10% (wt/vol) trichloroacetic acid and centrifuged, and uric acid was assayed by an HPLC method described by Lux and colleagues (19) using a Hypersil BDS C18 end-capped column (125 × 4 mm, particle size μm; Agilent, Palo Alto, CA). Additional details on this method are available in the online supplement.

The collection of EBC was performed using a commercially available condenser as previously described (20). Participants were asked to breathe at a normal frequency and tidal volume for 10 min. The EBC samples were immediately divided into aliquots, stored at −80°C, and thawed at room temperature just before analysis.

EBC H₂O₂ was measured by means of horseradish peroxidase–catalyzed oxidation of tetramethylbenzidine according to the method previously described by Gallati and Pracht (21) and modified for microtiter plate-based analysis. All samples were measured in duplicate, and a separate standard curve for H₂O₂ was constructed for each assay. Mean values were used for subsequent statistical analysis. Additional detail on the method for making these measurements is provided in the online supplement.

All data are expressed as mean ± SEM. Between-group comparisons were analyzed using the Mann-Whitney U test. The Wilcoxon signed rank test was used to evaluate the effect of the maximal and submaximal exercise-induced oxidative stress response within the groups and the effect of pulmonary rehabilitation on oxidative stress in patients with COPD. Nonparametric tests were used because the normality assumption was not obtained. A difference with p < 0.05 was considered statistically significant. Statistical analyses were analyzed with SPSS for Windows (version 11.0; SPSS, Inc., Chicago, IL).

As shown in Table 1

The results of both exercise tests are presented in Table 2

A significant increase in ROS-induced DNA damage was found immediately after the maximal exercise test (p < 0.05; Figure 1), which was similar to values observed before rehabilitation. In contrast to the data obtained before rehabilitation, we found no significant increase of ROS-induced DNA damage after the submaximal exercise test (Figure 1). Moreover, the differences with baseline values immediately after the submaximal exercise were significantly lower compared with values obtained before rehabilitation (p < 0.02; Figure 1).

After rehabilitation, we found no significant increase in plasma MDA after the maximal and submaximal exercise test (data not shown). For the maximal exercise test, we only observed a significant effect of urinary MDA 4 h after exercise, whereas immediately after the submaximal exercise test, a significant effect was found (p < 0.05 and p < 0.02, respectively; Figure 2). After rehabilitation, changes in plasma uric acid found after the maximal exercise test were similar to those observed before rehabilitation. In contrast to data obtained before rehabilitation, we detected no significant increase in plasma uric acid after the submaximal exercise test (Figure 3).

In addition, we observed no significant increase in pulmonary H₂O₂ production after the maximal and submaximal exercise test. Remarkably, baseline values of the submaximal exercise test were significantly higher when compared with baseline values of the maximal exercise test (p < 0.01; Figure 4).

After rehabilitation, we observed a similar increase in DNA damage and plasma uric acid levels after the maximal exercise test as we did before rehabilitation. An additional finding was that the concentration of plasma uric acid slightly, but significantly decreased immediately after the maximal exercise test. The most striking finding in the present study was that, after rehabilitation, we did not detect a significant increase in ROS-induced DNA damage after the submaximal exercise test. This decrease in ROS-induced DNA damage may be attributable to an improved training status of the patients with COPD, because the patients had an increased exercise capacity and physical activity level, as indicated by the Pam score, after rehabilitation. This is in line with a previous study (31), which reported that the extent of DNA damage was dependent on the training status of the healthy participants, because trained participants had less DNA damage compared with untrained participants after exercise.

On the basis of these results, we can conclude that intensive supervised pulmonary rehabilitation reduces exercise-induced DNA damage, when performed at moderate intensity. Moreover, the submaximal exercise test seems to be more discriminative to evaluate the effect of pulmonary rehabilitation.

After rehabilitation, there were no statistically significant changes in plasma MDA after both exercise tests. For urinary MDA, results were similar to those observed before rehabilitation, with the exception of the maximal exercise test, for which we no longer found a significant increase immediately after the exercise test.

In general, we observed a tendency for a decrease in systemic exercise-induced oxidative stress after rehabilitation, when also taking into account that the intensity and duration were higher after rehabilitation. This could be a consequence of an improved oxidative metabolism or an increased capacity of endogenous antioxidative systems. Previously, Rabinovich and colleagues (47) reported that patients with COPD had a reduced ability to adapt to endurance training, as reflected by a lower capacity to synthesize reduced glutathione. The reduced exercise-induced oxidative stress response may be attributable to an improved oxidative metabolism rather than to an upregulation of antioxidant defenses, although we cannot rule out the possibility that the effects were caused by the induction of antioxidant adaptations. Further studies are needed to clarify the mechanisms involved in the exercise-induced oxidative stress response and possible adaptations after pulmonary rehabilitation.

After rehabilitation, we observed no significant increase in exhaled H₂O₂ after both exercise tests, despite a significant increase in intensity and duration. This effect would probably be more pronounced if we had measured at similar intensity and duration before and after rehabilitation. Another striking effect deriving from the data presented is that baseline pulmonary H₂O₂ production at the submaximal exercise test was significantly higher compared with data obtained from the maximal exercise test. This effect was also found before rehabilitation, but was not significant. It can be speculated that the increased baseline H₂O₂ production might be due to the fact that the patients performed the submaximal exercise test after the weekend. The patients spent the weekend at home, and lifestyle factors, especially passive-smoking behavior, may have been different from those in the pulmonary rehabilitation center.

In this study, the maximal and submaximal exercise tests were not performed at similar workload and duration time (isotime) before and after rehabilitation. However, measurements of oxidative stress at isotime would probably have led to even more pronounced differences between measurements before and after rehabilitation, because it has been suggested that oxidative stress increases with the intensity and duration of the exercise.

In conclusion, the present study clearly indicates that patients with COPD, when compared with age-matched healthy control subjects, are characterized by increased systemic and pulmonary oxidative stress markers both at rest as well as induced by exercise. This suggests that healthy control subjects are able to tolerate exercise more effectively than patients COPD. Moreover, this study showed for the first time that intensive supervised pulmonary rehabilitation was associated with decreased systemic exercise-induced oxidative stress, especially after submaximal exercise. This decrease was accompanied by a significantly improved exercise capacity in patients with COPD after rehabilitation. The observed improvement may be attributable to adaptive responses involving a more efficient oxidative metabolism or an increased capacity of endogenous antioxidative systems. The possible mechanisms underlying improved exercise capacity and decreased exercise-induced oxidative stress after intensive supervised pulmonary rehabilitation, and the outcome of specific nutritional or pharmacologic modulation on exercise-induced oxidative stress, will be the subject of further investigations.