LAUSR

and further work to explain this is needed. Possibly, the inclusion of oscillometric variables, which seem to identify impaired exercise performance in smokers, would be useful here (6). Whatever its limitations, the CanCOLD exercise data set is a remarkable achievement that takes our understanding of exercise limitation in COPD to a new level. Clearly, mechanical constraints associated with dynamic increase in end-expiratory lung volume during exercise restrict performance in patients with well-established COPD. However, there is now evidence that ventilatory inefficiency plays a role in those with relatively preserved lung function and even in those for whom airflow obstruction has yet to develop. This study broadens our understanding of impaired exercise performance, suggests new approaches to its understanding, and will no doubt be succeeded by more exciting data. It is a fine addition to the long and honorable history of Canadian research into the pathophysiology of COPD.