Determinants of Percent Predicted FEVi in Current Asthmatic Subjects: Results
Initial characteristics of the 261 subjects in this analysis are as follows: No., 261; age, 40.8 ± 21.7 (SD); percentage of male subjects, 42.5; % FEVi, 84.5 ± 20.6 (absolute FEVi2.55L. ± 1.05); percent age of current smokers, 29.1 (pack-years of smoking, 18.8 ±18.8); percentage of ex-smokers, 24.1 (pack-years of smoking, 14.4 ±21.3); Z-scored IgE (n=204), 0.82 ±0.94 (percentage of Z-scored IgE>0.0, 83.3); and percentage of skin tests that were positive (n=222), 65.3. Z-scored IgE is expressed as standard deviations from the mean of the age-sex specific group.
We looked at the relationship of wheezing to %FEVi in this group. Figure 1 shows the %FEVj by wheeze scale within three age groups. As seen in the illustration, the %FEVj progressively worsens with age and also with severity of wheeze. Analysis of variance showed that not only were the wheeze scale and age groups significant in relation to the %FEVi, but also that the interaction term of wheeze scale and age group was also significant (p<0.01). A wheeze scale-age group (WS-AG) interaction term was calculated by multiplying the linear values for wheeze scale (1 through 5) and age group (] through 3). Thus, subjects age 6 to 34 years with mild wheezing had a WS-AG value of 1 and the subjects age 55 years or more with severe wheezing had a value of 15. In the youngest age group, low %FEVj appears only in the most severe grade of wheeze, while in the oldest age group any wheeze greater than the mildest is significantly lower (p<0.0001) than that in the younger age group. flovent inhalers
Multiple, regression results are as follows: for the wheeze scale-age group, the multiple regression was 0.526 (p<0.0001); for the ATS dyspnea scale, 0.564 (p<0.0001); for the diagnosed chronic bronchitis, 0.585 (p<0.0001). The interaction of wheeze and age (WS-AG) was significant and replaced the separate independent terms. In fact, forcing the main effects of wheeze scale and age group as well as the interaction term of WS • AG into a multiple regression with %FEV] as the dependent variable revealed insignificant correlations with %FEVi for both of the main effects (p=0.34 and p=0.41).
Figure 1. The %FEV[ by age groups and wheeze scale. The wheeze scale is derived from the severity of wheezing and the frequency of attacks of shortness of breath with wheezing. See text for wheeze scale definition.