1) Develop mean and 95% confidence lower limit of normal (LLN) formulae for FEV3/FVC for Black, Latin, and White men and women to supplement the findings of Hankinson et al. 2) Ascertain comparative variability of FEV1/FVC, FEV3/FVC and FEF25-75 in never-smoking adults and evaluate their utility in measuring the effects of smoking on airway obstruction. 3) Develop and use the concept of 1-FEV3/FVC to identify airway obstruction.
We identified 5938 never-smokers and 3570 current smokers with spirometric data meeting American Thoracic Society standards from the NHANES-III nationwide database. In these groups we developed regression formulae for the FEV3/FVC, quantified variability and LLN of the FEV1/FVC, FEV3/FVC and FEF25-75 to identify abnormalities in current-smokers, and evaluated 1–FEV3/FVC as a marker of airflow obstruction.
With normal aging, there were concurrent linear decreases in FEV1/FVC and FEV3/FVC and increases in 1-FEV3/FVC, the latter attributable to slower emptying of acini with longer time constants. By middle age these spirometric measurements worsened, on average, about 20 years earlier in current-smokers. Two-thirds of current-smokers who manifested airway obstruction had both FEV1/FVC and FEV3/FVC abnormal, while 1/6 had only FEV1/FVC abnormal and 1/6 had only FEV3/FVC abnormal. The normal variability of FEF25-75 is greatest and that of the FEV3/FVC is least. If <80% of mean predicted FEF25-75 values were used to identify abnormality, >25% of all never-smokers would have been falsely identified as abnormal. Using 95% confidence limits for FEF25-75, only 1% of never-smokers had isolated abnormal FEF25-75 while 42% of 683 smokers without restriction but with reduced FEV1/FVC and/or FEV3/FVC had normal FEF25-75 values (false negatives).
FEV1/FVC, FEV3/FVC and 1-FEV3/FVC all characterize expiratory obstruction well. In contrast, FEF25-75 has an unacceptably large proportion of false negatives and false positives.
Using <80% of mean predicted to define abnormality is statistically invalid. For valid spirographic assessment of airways obstruction, we recommend requiring statistically correct LLN and replacement of the FEF25-75 with the FEV1/FVC and FEV3/FVC.
James Hansen, None.