Lung function defects in treated pulmonary tuberculosis patients

We are grateful to Aggarwal and colleagues for their interest in our paper [1]. They are correct to point out the limitations of cross-sectional studies, particularly when they rely on recalled information, and we discussed these at some length in the paper. However, evidence needs to be evaluated in the round. Not all error is biased; the outcomes (lung function) that we used were measured objectively and, as the results were not shared with the participants at the time, are unlikely to have influenced their answers to questions about a past history of tuberculosis. We agree, of course, that objective measures of tuberculosis would have been preferable, but it seems more likely that these would have reduced random error and so strengthened the current findings. We hope to add these to future investigations in the Burden of Obstructive Lung Disease (BOLD) cohort. A history of tuberculosis is an important risk factor for obstructive disease and low lung function