Introduction
Chronic obstructive pulmonary disease (COPD) is characterized by a usually progressive airflow limitation [
1]. Extrapulmonary comorbid conditions, like cachexia and muscle atrophy, are frequently observed [
2]. Pulmonary rehabilitation has been demonstrated to improve exercise capacity, muscle metabolism, and quality of life [
3,
4].
Detailed instructions on pulmonary rehabilitation in COPD have been published [
5‐
7]. However, only a limited number of studies compared the impact of various forms and intensities of outpatient training programs in a randomized fashion [
8,
9].
The role of exercise effects on the peripheral muscle and its myokines has been acknowledged [
10]. The underlying mechanism of exercise training includes the expression of the transcription factor peroxisome-proliferator-activated receptor-γ coactivator 1α (PGC1-α) [
11‐
16]. PGC1-α stimulates the expression of FNDC5, a membrane protein that is cleaved and secreted as a newly identified hormone called irisin [
17]. The upregulation of PGC1-α has been shown to attenuate inflammation [
18].
The most effective frequency of pulmonary rehabilitation is not known [
6]. Most pulmonary rehabilitation studies demonstrating benefits are based on at least two sessions per week [
19]. However, a recent review article pointed out that only a small proportion of enrolled patients demonstrated continued commitment raising the question whether a once-weekly training program would be associated with better compliance [
20].
We conducted a randomized controlled trial to compare the effects of two different training approaches on exercise capacity, QoL, muscle mass, myokines and serum inflammatory markers. Furthermore, screening failure rates and drop-out rates were assessed.
Discussion
To our knowledge this is the first randomized trial comparing a low frequency individualized (IT) vs. non-individualized exercise training (NT) in stable outpatient COPD patients. We found that only IT significantly improved 6-MWT and muscle rectus femoris cross sectional area in a group of patients that exercised once weekly.
Baumann et al. randomized 100 patients with moderate to severe COPD to a continuous outpatient interdisciplinary rehabilitation program or standard care [
24]. After 26 weeks, the intergroup difference of the 6-MWT was 59 m in favour of rehabilitation. While the individualized training intervention was similar, the higher intergroup difference compared to our study (32 m) is explained by the different comparator (standard care and non-individualized group training, respectively).
Behnke et al. were able to demonstrate a significant effect of a supervised walking training at home in preserving the hospital-achieved improvement in six-minute walk test and quality of life in patients with severe COPD. 30 out of 46 completed the program and walked 2308 m on 157 days. Thus, the effect was seen in a group of highly compliant patients [
25]. No randomization was done in this study. Significant effects have also been observed by du Moulin et al. in patients with moderate COPD. In this randomized trial, ten patients performed home-based exercise training and 10 patients served as controls. After six months the training group had better results than the control group in exercise capacity and lung function [
26].
Göhl et al. randomized 34 patients to participate in a multimodular 12 months training program [
27]. The training group demonstrated increases in a variety of parameters including the 6-MWT (79 m) and SGRQ (>4 units) whereas in the control group no significant changes were observed. In contrast to our NT group, the intervention included modules of increasing intensity and time, resulting in an increase of 2.4 to 4.2 hours of training per week [
27]. The higher intensity and the longer period of time may very well explain larger effects.
The cross-sectional area of the M. rectus femoris rose by 0.57 cm
2 in the IT-group. Seymour et al. described the difference of 115 mm
2 between healthy subjects and COPD patients [
22]. In this regard, an increase of 57 mm
2 (as found in our study) would roughly bisect the difference, which might be considered as relevant. This is further confirmed by earlier data that also demonstrated an increase of cross-sectional area (0.57 cm
2) of the M. rectus femoris after eight weeks of bilateral high intensity isokinetic knee extensor resistance training [
28]. To the best of our knowledge there are no data the correlate the rectus femoris cross sectional area to clinically relevant outcome parameter.
We could not detect any differences in QoL using SGRQ and CAT. This is most likely explained by the low-frequency and the low-intensity nature of both programs. In contrast, higher intensity programs have demonstrated positive effects on health-related quality of life [
9,
27,
29]. It may be speculated that a low intensity training program does not result in effects large enough to measure. In a randomized study of two exercise training programs of different intensity Camillo et al. observed a significant improvement in heart rate variability only after the high-intensity protocol [
8]. Effing et al. demonstrated statistically significant between-group differences in exercise capacity and daily activity in an evaluation of the “COPE-active program” [
29]. The interventional exercise program consisted of a 6-month “compulsory” period (3 sessions/week) and subsequently a 5-month “optional” period (2 sessions/week). One session/week (control group) consisted of unsupervised home-based exercise training. Of 153 patients, 74 intervention and 68 control patients completed the one-year follow-up. Again, significant effects were seen in a relatively intense program (2 – 3 sessions/week) [
29]. Finally, Probst et al. compared the effects of a high-intensity whole-body endurance-and-strength program and a low-intensity calisthenics-and-breathing-exercises program on different outcome parameters [
9]. Both groups underwent 3 sessions per week for 12 weeks. Exercise capacity and muscle force significantly improved only in the endurance-and-strength group. Health-related quality of life and functional status improved significantly in both groups. Even the “low-intensity” exercise program included 3 sessions per week [
9]. In summary, most “positive studies” published were of higher intensity than the two programs we conducted.
A major problem of all these studies is the inclusion criteria and the compliance of patients. One recent review pointed out that the majority of positive studies did not clarify which patients were included [
20]. Only 12% of studies included in this review reported the number of contacted patients. In these studies only 28% completed the program. Altogether 75% of the patients suitable for exercise programs were omitted due to sampling exclusion and dropout. The authors concluded that most of the study populations were not representative of the target population. In general, adherence is a common problem in rehabilitation studies with COPD patients. Drop-out-rates up to 50% are not unusual [
5,
28‐
30]. The main causes are often difficult to clarify. By telephone interviews the most often mentioned reasons were disease-related drop-out, disagreement with group assignment, and missing motivation [
20]. Missing motivation may be a sign of depression which has been reported to be a frequent comorbidity in COPD [
30,
31]. No clear recommendation exists how to deal with these frequently occurring problems.
When considering only patients that entered the training period we observed a differential dropout, with significantly more subjects stopping NT than IT (13/27 vs. 5/25; p = 0.04). Assessing possible reasons for this phenomenon, the main causes were not significantly different between both groups (Table
3). To our knowledge, published literature does not provide plausible data to sufficiently explain this issue.
On a closer view it becomes clear that the patients that dropped out had a lower quality of life, worse lung function and elevated inflammatory markers like AAT and CRP (Table
5). In summary, patients with a worse baseline condition had a higher probability to drop out. This is in line with the observation that a higher FEV
1, CRQ-Score or a shorter distance to the location where training takes place would increase the adherence [
28].
We could demonstrate an increase of PGC1-α in the IT group. Since the Irisin values did not show a subsequent increase, the relevance of this result remains unclear. We could not find significant changes in all other measured inflammatory markers and myokines which strengthenes the assumption that low intensity training of 1 hr /week regardless of the modality is not sufficient.
The study has significant limitations. First, we included a relatively small number of patients from a wide range of the disease (GOLD stage I – IV). As a potential training effect might be achievable in some stages of the disease and not in others, this may have influenced the results. On the other hand, as neither the mean FEV1 nor the GOLD stages differed significantly between both groups we do not think that the results have biased on a (potentially missed) between-group differences. Secondly, we observed a high dropout rate (as other groups before). This may have attenuated the effect of training. Thirdly, we did not assess depression via standardized questionnaires.
Taken together, this is to our knowledge the first study comparing different low-intensity training approaches in stable COPD patients in a randomized way. The data seem to favor the individualized low frequency training program but do not result in significant improvements of quality of life.
At the moment, it remains unclear how to resolve the discrepancies between guideline recommendations and existing structures. In many countries, we do not have the infrastructure to train our patients 3–5 times per week (as recommended in recent international guidelines [
32]). Furthermore a significant proportion of patients would not attend more frequent training opportunities. This results in the need to optimize once weekly training. We believe, that - beside the training approach - the training intensity is an important trigger of success. We conclude, that if low intensity training was chosen and only once weekly training can be proffered, we would suggest to offer an individualized training. The significance of exercise-intensity increase has to be evaluated in further studies.
Competing interests
The project was supported by the German Centre for Lung Research (DZL). The study has been funded partially by GSK. No further conflict of interest has to be acknowledged.
Authors’ contributions
JK, K Kehr, DH, SF, UK and K Koehler performed experiments, measurements and included patients to the study. ARK, TG, CV, CN and K Kenn contributed to the design, statistics and conception of the study, and contributed to drafting the manuscript. ARK contributed to the design and conception of the study. He included patients, analysed and interpreted the data and drafted the manuscript. All authors read and approved the final manuscript.