BackgroundComparisons between endurance training (ET) and resistance training (RT) have produced equivocal findings in chronic obstructive pulmonary disease (COPD) patients. The purpose of our study is to investigate the effectiveness and long-term outcomes of adding ET and RT to conventional medical treatment in patients with COPD. A secondary objective is to investigate the clinical improvements resulting from exercise training in patients with different disease severities.MethodsThe study was a multicenter, prospective trial in people with stable COPD. The cohort was randomized to three groups: individualized medical treatment group (MT), MT + endurance training group (MT + ET) and MT + resistance training group (MT + RT). Exercise was performed 3 times weekly over a 12-week period. The endpoints of exercise capacity, health-related quality of life, COPD symptoms, lung function, and anxiety and depression questionnaires were re-evaluated at baseline, at the completion of the intervention and at 6 and 12-month follow-up. According to the COPD assessment tool offered by GOLD guidelines, patients were stratified into GOLD A and B groups and GOLD C and D groups for further subgroup analysis.ResultsThe intention-to-treat (ITT) population included 366 patients, 328 of them completed the study protocol over 12 months (the PP-population). There were no significant differences in the primary outcome, quality of life, between patients who underwent medical treatment (MT) alone, MT + endurance training (MT + ET), or MT + resistance training (MT + RT) at the completion of the intervention, 6-, or 12-month follow-up. Additionally, no significant differences were observed between MT, MT + RT, or MT + ET groups concerning the primary outcome, exercise capacity (3MWD), after initial 3 months of intervention. However, a small statistically significant difference was noted in favor of MT + ET compared to MT + RT at 12 months (ITT: Δ3MWD in ET vs RT = 5.53 m, 95% confidence interval: 0.87 to 13.84 m, P = 0.03) (PP: Δ3MWD in ET vs RT = 7.67 m, 95% confidence interval: 0.93 to 16.27 m, P = 0.04). For patients in the GOLD C and D groups, improvement in quality of life following ET or RT was significantly superior to medical intervention alone. Furthermore, upon completion of the exercise regimen, RT exhibited a greater improvement in anxiety compared to ET in these patients (ITT: ΔHAD-A at 3-month: RT = -1.63 ± 0.31 vs ET = -0.61 ± 0.33, p < 0.01) (PP: ΔHAD-A at 3-month: RT = -1.80 ± 0.36 vs ET = -0.75 ± 0.37, p < 0.01).ConclusionsOur study presents evidence of the beneficial effects of ET and RT in combination with standard medical treatment, as well as the long-term effects over time after the intervention. While the statistically significant effect favoring ET over RT in terms of exercise capacity was observed, it should be interpreted cautiously. Patients in severe stages of COPD may derive greater benefits from either ET or RT and should be encouraged accordingly. These findings have implications for exercise prescription in patients with COPD.Trial registrationChiCTR-INR-16009892 (17, Nov, 2016).