Pulmonary function tests were performed at the initial visit, and at each of the follow-up visits. The initial test values were used as the baseline data. When follow-up data were available > 1 year after the initial test, dFEV1 was obtained by subtracting the last follow-up FEV1 value from the baseline FEV1 value, and dividing by the time in years between the two tests (milliliters per year). At each clinic, all the lung function measurements were performed with consistency using the same type of electrospirometer; the Autospiro AS-600 (Minato Medical Science; Osaka, Japan) was used at the COPD clinic, and Chestac-65V (Chest; Tokyo, Japan) was used at the smokers’ clinic. In patients with COPD, the measurements were performed in their stable state, ie, there were no exacerbations of the conditions from the preceding 6 weeks. The use of bronchodilators was prohibited for at least 12 h before the tests were performed. The predicted values for pulmonary function were calculated based on the formula proposed by the Japan Society of Chest Diseases. Based on the previous studies, subjects who had a dFEV1 > 90 mL were defined as rapid decliners (RDs). more
Measurement of Chest HRCT Parameters in Patients With COPD
The CT scans were taken in the supine position using a HRCT scanner (X-Vigor; Toshiba; Tokyo, Japan) with 2-mm collimation, scanning time of 1.0 s, 120-kilovolt electrical voltage, 200-mA electrical current, and 35-cm field of view. A high-resolution reconstruction algorithm for the lung (FC83; Toshiba) was employed. During scanning, the patients held their breath after a deep inspiration. Contrast medium was not used. Three slices (upper, middle, and lower lung) from each patient were ana-lyzed. Each CT image was composed of 512 X 512 matrices with numerical data (CT numbers) in Hounsfield units (HU). The lung fields were automatically identified in each image. Using the previously published method, the LAA% ratio in both lung fields in three slices was calculated automatically. The cut-off level between the normal lung density area and the low-attenuation area was defined as — 960 HU. Considering that the LAA% of normal subjects was 60% as having severe emphysema at the outset of this study. The mean CT score was calculated as the average of the CT numbers in HU of all pixels in both lung fields in three slices. Each continuous low-attenuation area (CLA) was recognized automatically, and the total number of CLA and the average size of CLA units (sCLA) of the three slices were calculated.