A retrospective analysis of data from the Cardiovascular Health Study (CHS) performed by Mannino et al1 concludes that older adults with borderline abnormal spirometry results have an increased risk of death and hospitalisations related to chronic obstructive pulmonary disease (COPD). The all-cause mortality result from the analysis by Mannino et al merely confirms many previous reports.2 However, the mechanism is not obstructive lung disease since the vital capacity (FVC), aptly named by British surgeon John Hutchinson more than 150 years ago, is a stronger predictor than forced expiratory volume in 1 s (FEV₁),3 and respiratory muscle weakness accounts for much of the relationship.4

It is intuitive that patients with COPD must make a transition from normal spirometry to clinically relevant airway obstruction. However, only a small minority of adults with borderline abnormal spirometric results will ever develop COPD, regardless of their smoking status. For example, in the Lung Health Study, only 10% of the variability in the subsequent decline in lung function in continuing smokers was predicted by baseline spirometric results, even when bronchodilator responsiveness and airway reactivity were included in the predictive model.5 Therefore, very few participants in the CHS whose FEV₁/FVC was below the middle quintile (falling into the arbitrary GOLD stage I) have ever developed COPD. In fact, chronic bronchitis, emphysema, asthma and dyspnoea (as reported at study entry) were not independent predictors of a subsequently more rapid decline in lung function in the CHS cohort.6

The paper by Mannino et al suggests that GOLD stage I predicted “COPD-related” hospitalisations. However, the medical records were not examined for spirometric test results. I believe that the vast majority of these hospitalisations were not due to an exacerbation of COPD. In one recent study only one-third of 800 consecutive patients with a discharge diagnosis of COPD had a spirometric test.7 A history of chronic cough, dyspnoea on exertion and cigarette smoking in hospitalised patients is more likely to be associated with heart failure or pneumonia than COPD.

In CHS participants with a normal baseline FEV₁ (GOLD stage I), 11 years of follow-up was not sufficiently long for even the continuing smokers to have lost enough lung function to have developed a COPD exacerbation. Current smokers in the CHS lost an average of only 48 ml/year during the first 7 years of follow-up,6 which extrapolates to a loss of only 0.5 litres after 11 years. The average baseline FEV₁ was 2.4 litres in men and 1.8 litres in women.

In summary, as the CHS investigator responsible for spirometric testing in this study, I am not convinced by the analyses of Mannino and colleagues that GOLD stage I spirometry was a risk factor for COPD morbidity or mortality.