Drummond and colleagues (1) report that smokers with mild to moderate chronic obstructive pulmonary disease (COPD) and worse initial FEV 1 /FVC have a more rapid decrease of FEV 1 and higher rate of death. Only subjects with lower than 70% FEV 1 /FVC ratio are at risk for excessive lung function decline. The main limitation of their study is the only 5 years of followup, so their conclusion that spirometry predicts long-term outcome is not warranted. The authors refer to the previous study by Burrows and colleagues (2) of 13 smokers, with initial FEV 1 /FVC less than 70%, followed for 10 years, but omit another study of 56 middle-aged smokers, followed for 13 years (3), including besides spirometry, airway resistance, maximal expiratory flow rates, and small-airway disease tests such as closing volume and N 2 slope. Both studies showed that subjects with a low FEV 1 /FVC or a low FEV 1 /VC at the entry had a subsequent accelerated loss of FEV 1 . In the article by Burrows and colleagues (2) on 141 male smokers, with an FEV 1 greater than 60% of predicted, followed by spirometry for at least 6 years, initial normal FEV 1 /FVC explained only about 16% of the subsequent decrease of FEV 1 . Similarly, in our study, 79% of smokers with a normal FEV 1 /VC still had normal values (more than 70%) 13 years later (3). Conversely, a decreased FEV 1 /VC has no necessarily serious implication in itself. Indeed, only a minority of our subjects, with lower than 60% FEV 1 /VC, showed an accelerated loss of FEV 1 , reaching end FEV 1 /VC values lower than 45%. An associated high N 2 slope adds the necessary information to predict a low FEV 1 /VC. In agreement with our data, the analysis of Drummond and colleagues (1) showed that excessive lung function decline occurred only below FEV 1 /FVC of 0.65. Both figures of Drummond and colleagues (1) and ours (3) are lower than current thresholds used to define COPD. Our data suggest that a subgroup of smokers in their 50s with a low FEV 1 /VC and a high N 2 slope are probably the susceptible smokers at high risk to develop COPD.Recently it was reported that in some subjects, the progressive decrease with age of FEV 1 is not the rule. In them, FEV 1 remains stable or even increased (4, 5). However, these results were obtained from follow-ups of only 3 to 5 years, too short a time. In 7 out of 56 subjects, over 13 years, FEV 1 either did not change or even increased (3). We thank Dr. Stanescu for his interest in our article describing the association between baseline lung function measurements and longitudinal lung function decline and mortality in over 5,800 smokers from the Lung Health Study (1). Dr. Stanescu and colleagues' prior publication of 13-year follow-up of 56 smokers and ex-smokers (2) was an important foundation for our analysis. We agree that there is substantial heterogeneity in lung function decline, and a single baseline measure does not necessarily predict excessive decline in all persons. Although the individual risk for excessive lung function decline may not be compl...