Postbronchodilator Reference Values: Should They Be the Norm?

When spirometry is used to identify the presence and severity of chronic obstructive pulmonary disease (COPD), pre- and postbronchodilator testing is typically performed. In normal populations, there is at most only a small bronchodilator response (1). As a result, prebronchodilator reference values have been used and compared with postbronchodilator results. The Global Lung Function Initiative (GLI) noted that lung function test results can be interpreted using only robust, relevant, and reliable reference values (2). Using data from nearly 100,000 nonsmokers (55% female subjects) aged 3–95 years from 72 centers and 33 countries, the GLI study showed that the FEV₁:FVC ratio is essentially independent of ethnic group and serves as a reliable, although imperfect, indicator of lung function. However, 9-year follow-up of another multicenter study, PLATINO (Projeto Latino-Americano de Investigação em Obstrução Pulmonar), which used prebronchodilator reference values, identified a substantial underdiagnosis of COPD (3). The subjects in that study who did not receive diagnoses of COPD until they returned after 9 years presented with similar clinical characteristics as those who were not diagnosed in the initial phase. For some time, there has been debate as to whether postbronchodilator reference values would be better than prebronchodilator reference values to identify individuals with COPD or at risk for developing COPD. The landmark GLI study did not explore the effect of using postbronchodilator reference values for interpreting spirometry results. Furthermore, there is limited information as to whether the use of postbronchodilator reference values has a significant impact in the clinical interpretation of spirometry results (4, 5).

In this issue of the Journal, Malinovschi and colleagues (pp. 461–471) report the results of the study (6). They set out to determine whether postbronchodilator reference values were more successful than prebronchodilator reference values to identify individuals with mild COPD. To do this, they used pre- and postbronchodilator reference values generated from SCAPIS (Swedish Cardiopulmonary Bioimage Study) (7) and assessed the value of using postbronchodilator reference values as opposed to prebronchodilator reference values in a random population sample. The authors also compared pre- and postbronchodilator reference values from SCAPIS with reference values from the GLI and PLATINO studies.

The study included 30,154 middle-aged adults, 50–64 years of age, evenly distributed between men and women, at six academic medical centers in Sweden. The SCAPIS postbronchodilator reference values were obtained from 10,156 nonsmokers without respiratory symptoms or respiratory disease, after inhaling 400 μg of salbutamol. The SCAPIS prebronchodilator reference values were derived from 1,498 individuals at one of the six centers. In addition to pulmonary function testing, participants underwent high-resolution computed tomography to identify emphysema and completed several questionnaires. Participants were divided into three groups: 1) postbronchodilator FEV₁:FVC ratio greater than the lower limit of normal (LLN) using both the prebronchodilator and postbronchodilator reference values; 2) postbronchodilator FEV₁:FVC ratio greater than the LLN using the prebronchodilator reference value but lower than the LLN using the postbronchodilator reference values; and 3) postbronchodilator FEV₁:FVC ratio less than the LLN using both pre- and postbronchodilator reference values.

The authors report that using postbronchodilator reference values resulted in higher predicted median values and LLN values for the FEV₁:FVC ratio. When the higher predicted median and LLN values were used, the prevalence of postbronchodilator FEV₁:FVC ratio less than the prebronchodilator LLN was 4.8%, while that of postbronchodilator FEV₁:FVC ratio less than the postbronchodilator LLN was 9.9% in this general population. Importantly, the additional 5.1% of participants who had abnormal postbronchodilator FEV₁:FVC ratios only when using postbronchodilator reference values had more respiratory symptoms, more computed tomography–diagnosed emphysema (13.5% vs. 4.1%; P < 0.001), and more instances of self-reported physician-diagnosed COPD (2.8% vs. 0.5%; P < 0.001) than subjects with postbronchodilator FEV₁:FVC ratio greater than the LLN for both pre- and postbronchodilator reference values. Similar results were seen for FEV₁, but not for FVC.

The study has both strengths and weaknesses. Strengths include the large number of individuals recruited from SCAPIS, the age range of 50–64 years (a population most likely to have early or mild COPD), the even distribution of men and women, and the accompanying clinical data, including respiratory symptoms, diagnosis of emphysema, and physician diagnosis of COPD. These clinical findings demonstrate the clinical relevance of postbronchodilator reference values in this population. Weaknesses include the relatively homogeneous population regarding race, ethnicity, and geographic location (although this may also be a strength by increasing the ability to identify an effect in this population) and the fact that prebronchodilator reference values were obtained from only one of the six research sites. In addition, there were some differences with data compared with other studies, such as PLATINO. However, as noted by the authors, the lower dose of salbutamol (200 μg) used for postbronchodilator testing (8), or the different geographic location (South America), may explain some of the differences.

Overall, this study provides valuable new information, as well as the opportunity for the authors and others to initiate additional studies to examine diverse populations and settings and to determine if the findings in this study are relevant beyond Sweden. Then we will know with greater certainty whether postbronchodilator reference values should be the new norm.