Using the lower limit of normal for the FEV1/FVC ratio reduces the misclassification of airway obstruction

Maureen P Swanney ^1*, Gregg Ruppel ², Paul L Enright ³, Ole F Pedersen ⁴, Robert O Crapo ⁵, Martin R Miller ⁶, Robert L Jensen ⁵, Emanuela Falaschetti ⁷, Jan P Schouten ⁸, John L Hankinson ⁹, Janet Stocks ¹⁰ and Philip H Quanjer ¹¹

¹ Respiratory Physiology Laboratory, Christchurch Hospital, Christchurch, New Zealand
² Pulmonary Function Laboratory, Saint Louis University Hospital, St Louis, MO, United States
³ College of Public Health, The University of Arizona, Tucson, AZ, United States
⁴ Institute of Public Health, Aarhus University, Aarhus, Denmark
⁵ Pulmonary Division, LDS Hospital and University of Utah, Salt Lake City, Utah, United States
⁶ Department of Medicine, University Hospital Birmingham NHS Trust, Birmingham, United Kingdom
⁷ Department of Epidemiology and Public Health, University College London, London, United Kingdom
⁸ Department of Epidemiology, University Medical Center Groningen, University of Groningen, Groningen, Netherlands
⁹ Valdosta, Georgia, United States
¹⁰ Portex Anaesthesia, Intensive Therapy and Respiratory Unit, UCL, Institute of Child Health, London, United Kingdom
¹¹ Department of Pulmonary Diseases, Erasmus Medical Centre, Erasmus University, Rotterdam, Netherlands

^* To whom correspondence should be addressed. E-mail: maureen.swanney{at}cdhb.govt.nz.

Accepted 5 July 2008

Abstract

Rationale: The prevalence of airway obstruction varies widely with the definition used.

Objectives: To study differences in the prevalence of airway obstruction when applying four international guidelines to three population samples using four regression equations.

Methods: We collected predicted values for FEV1/FVC and its lower limit of normal (LLN) from the literature. FEV1/FVC from 40,646 adults (including 13,136 asymptomatic, never-smokers) aged 17-90+ years were available from American, English, and Dutch population-based surveys. The prevalence of airway obstruction was determined by the LLN for FEV1/FVC, and by using the GOLD, ATS/ERS, or BTS guidelines, initially in the healthy subgroup and then the entire population.

Results: The LLN for FEV1/FVC varied between prediction equations (57 available for men and 55 for women), and demonstrated marked negative age dependency. The median age at which the LLN fell below 0.70 in healthy subjects was 42 and 48 years in men and women respectively. When applying the reference equations (Health Survey for England 1995-1996, NHANESIII, ECCS/ERS and a Dutch population study) to the selected population samples, the preva-lence of airway obstruction in healthy never-smokers aged over 60 years varied for each guideline: 17-45% of men and 7-26% of women for GOLD; 0-18% of men and 0-16% of women for ATS/ERS; 0-9% of men and 0-11% of women for BTS. GOLD guidelines caused false positive rates of up to 60% when applied to entire populations.

Conclusions: Airway obstruction should be defined by FEV1/FVC and FEV1 being below the LLN using appropriate reference equations.