Postoperative pulmonary complications (PPCs) are the most frequent complications occurring in patients undergoing thoracic surgery and they are associated with prolonged hospital stay, decreased survival and expanding medical costs. Implementation of structured and supervised exercise programs including endurance training (ET), respiratory muscle training (RMT) or a combination of both, within the short waiting period before surgery, has been shown to enhance patients' physical fitness, to provide protective effects against PPCs and therefore to spare health care resources by shortening intensive care unit (ICU) and hospital lengths of stay. More recently, a simple intervention consisting in patient's instruction and education about modifiable risk factors, optimal breathing pattern and the impact of physical exercise has emerged as a simple alternative intervention prevent PPCs, although the evidence is inconclusive. Therefore, the investigators propose a multicentre randomized, open, blinded end point controlled trial testing the hypothesis that preoperative education and instruction focused on breathing exercise and endurance training reduce the occurrence of PPCs in patients undergoing thoracic or abdominal surgery. Patients with Intermediate-to-high risks factors for PPCs will be randomized on a 1:1 basis into an intervention arm and a usual care arm (Control group). In the Education group, patients will be asked to use a flow resistive device (One set of 30 repetitions, two times a day and to increase their daily physical activities (\> 5'000 steps or equivalent) until surgery. Primary study endpoint will be the incidence of PPCs (e.g., atelectasis, pneumonia, respiratory failure) according to the European Perioperative Clinical Outcome definitions. Secondary outcomes will include non-respiratory complications, utilization of hospital resources (e.g., hospital length of stay, ICU admission),and preoperative changes in maximal inspiratory pressure \[MIP\]. Assuming a rate of 39% PPCs in the controls and a possible reduction to 26% in the intervention group, enrollment of 203 patients per group will provide 80% power with an alpha value of 0.05. Taking into account dropouts (5%) and in-hospital mortality rate (2%), a total of 436 surgical patients will be enrolled.