All volumes are expressed as percentage of normal predicted values based on reference values for the patients age, sex, and height (Knudson et al for spirometry and flow volume curves and Cotes for lung volumes and Deo).
The presence of pulmonary involvement was confirmed when either FVC or TLC or Deo was <80 percent of predicted, when FEVj/FVC was 70 percent or less, or when MEF25 was <60 percent of predicted, according to Intermountain and American Thoracic societies. Four patterns of abnormality were defined as follows: (1) restrictive: reduced TLC with normal FEV,/FVC; (2) obstructive: reduced FEV,/FVC irrespective of other values; (3) isolated diffusing impairment: reduced Deo with normal TLC, FEV,/FVC, and MEF25; and (4) isolated MEF25 reduction: reduced MEF25 with normal TLC, Deo, and FEVj/FVC.
Concerning the SAD, we used the MEF25 reduction as the single criterion. Thus, we considered a MEF25<60 percent of predicted, because of the large variability of the values reported by Knudson et al, as the marker of SAD. All the obtained values for MEF25 have been normalized for the TLC decrease from 100 percent.
Serologic studies included the detection of antinuclear antibodies (ANA) and anticentromere antibodies (ACA) by the indirect immunofluorescence technique (positive ANA titer >:160) and anti-SCL by counterimmunoelectrophoresis.