Association of asthma symptoms with peak particulate air pollution and effect modification by anti-inflammatory medication use.
Maxima of hourly data from outdoor monitors may capture adverse effects of outdoor particulate matter (PM) exposures in asthmatic children better than do 24-hr PM averages, which form the basis of current regulations in the United States. Also, asthmatic children on anti-inflammatory medications may be protected against the proinflammatory effects of air pollutants and aeroallergens. We examined strengths of pollutant associations with asthma symptoms between subgroups of asthmatic children who were on versus not on regularly scheduled anti-inflammatory medications, and tested associations for different particle averaging times. This is a daily panel study of 22 asthmatic children (9-19 years of age) followed March through April 1996 (1,248 person-days). They lived in nonsmoking households in a semirural area of Southern California within the air inversion mixing zone (range, 1,200-2,100 feet) with transported air pollution from urban areas of Southern California. The dependent variable derived from diary ordinal scores is episodes of asthma symptoms that interfered with daily activities. Minimum to 90th-percentile levels of exposures at the outdoor monitoring site were 12-63 microg/m(3) for 1-hr PM < 10 microm in aerodynamic diameter (PM(10)); 8-46 microg/m(3) for 8-hr PM(10); 7-32 microg/m(3) for 24-hr PM(10); 45-88 ppb for 1-hr O(3); 6-26 ppb for 8-hr NO(2); 70-4,714 particles/m(3) for 12-hr daytime fungi; and 12-744 particles/m(3) for 24-hr pollen. Data were analyzed with generalized estimating equations controlling for autocorrelation. There was no confounding by weather, day of week, or linear time trend. Associations were notably stronger in 12 asthmatic children who were not taking anti-inflammatory medications versus 10 subjects who were. Odds ratios (95% confidence intervals) for asthma episodes in relation to lag 0 minimum to 90th-percentile pollutant changes were, respectively, 1-hr maximum PM(10), 1.92 (1.22-3.02) versus 0.96 (0.25-3.69); 8-hr maximum PM(10), 1.68 (0.91-3.09) versus 0.75 (0.18-3.04); 24-hr average PM(10), 1.35 (0.82-2.22) versus 0.80 (0.24-2.69); 1-hr maximum O(3), 1.28 (0.75-2.17) versus 0.76 (0.24-2.44); 8-hr maximum NO(2), 1.91 (1.07-3.39) versus 1.08 (0.30-3.93); 12-hr fungi, 1.89 (1.24-2.89) versus 0.90 (0.35-2.30); 24-hr pollen, 1.90 (0.99-3.67) versus 0.85 (0.18-3.91). Pollutant associations were stronger during respiratory infections in subjects not on anti-inflammatory medications. Although lag 0 1-hr maximum PM(10) showed the strongest association, the most robust associations were for lag 0 and 3-day moving averages (lags 0-2) of 8-hr maximum and 24-hr mean PM(10) in sensitivity analyses testing for thresholds. Most pollutant effects were largely driven by concentrations in the upper quintile. The divergence of exposure-response relationships by anti-inflammatory medication use is consistent with experimental data on inflammatory mechanisms of airborne pollutants and allergens.