Twenty-two trials (877 patients in total) were included.
Exercise.
1. Intensity of training: 2 studies in patients with mild or moderate COPD were identified, which suggested that patients could achieve maximal physiological training by establishing individual training intensities at or above the anaerobic threshold. However, there was no evidence that this influenced relevant health outcomes.
2. Upper limb training: 1 trial compared upper-limb with lower-limb training and with a combination of both. The treatment effect due to upper-limb training was small and of unknown clinical significance. Upper-extremity training improved arm-specific measures of exercise capacity, though there is no evidence that this has a different effect on HRQL to leg training alone.
3. Inspiratory muscle training as an adjunct to exercise training: the results from the 7 trials examining this intervention were mixed. The authors conclude that the evidence that inspiratory muscle training confers any additional benefit is equivocal.
4. Other types of breathing exercises: the identified trials have presented few details of the intervention, or have had methodological problems. There is, therefore, only weak evidence that patients with COPD may gain from specific breathing exercises such as those used in yoga or t'ai chi.
Education.
There is an absence of valid RCTs, though a non-randomised study was found that showed no association between health education programmes and any health status measure.
Psychological interventions alone, or as an adjunct to exercise training.
The trials assessed had either methodological problems, or showed non significant (statistically or clinically) effects on health-related outcomes such as anxiety or dyspnea. While relaxation was shown to relieve dyspnea in the short term, it is unknown whether there are any long-term benefits. Cognitive and behavioural modification techniques are effective in improving exercise tolerance and HRQL.