Asthma Aggravation, Combustion, and Stagnant Air
Total Page:16
File Type:pdf, Size:1020Kb
466 Thorax 2000;55:466–470 Asthma aggravation, combustion, and stagnant air Gary Norris, Timothy Larson, Jane Koenig, Candis Claiborn, Lianne Sheppard, Dennis Finn Abstract meteorology with a knowledge of specific air Background—The relationship between pollution point sources. current concentrations of ambient air Other studies have combined meteorology pollution and adverse health eVects is with chemical composition of particulate mat- controversial. We report a meteorological ter. For instance, factor analysis with a varimax index of air stagnation that is associated rotation of the particulate matter composition with daily visits to the emergency depart- collected from 1957 to 1961 from 30 cities ment for asthma in two urban areas. across the USA found seven factors represent- Methods—Data on daily values of a stag- ing heavy industry or steel production, internal nation persistence index and visits to the combustion engines, coal combustion, possible emergency department for asthma were gas production, a zinc-tin factor, plating, and collected for approximately two years in copper.6 Gatz7 included meteorological vari- Spokane, Washington, USA and for 15 ables (mean wind speed, maximum wind months in Seattle, Washington, USA. The speed, ventilation rate, wind direction, rain) stagnation persistence index represents along with the composition of particulate mat- the number of hours during the 24 hour ter in a factor analysis to help identify sources. day when surface wind speeds are less In that study wind direction was the only than the annual hourly median value, an meteorological variable that was correlated index readily available for most urban with particulate matter concentration. 8 areas. Associations between the daily Thurston and Spengler used factor analysis stagnation persistence index and daily with a varimax rotation separately on particu- emergency department visits for asthma late matter composition and meteorological were tested using a generalised additive parameters. The sources representing the Poisson regression model. A factor analy- motor vehicle factor (lead, bromine) were significantly correlated with the meteorological sis of particulate matter (PM2.5) composi- tion was performed to identify the component representing poor dispersion. 9 pollutants associated with increased Lewis and coworkers used carbon monoxide asthma visits. in place of lead as a tracer of motor vehicles in Results—The relative rate of the associ- a study of particulate matter composition in Denver. They found a high correlation between ation between a visit to the emergency 2 department for asthma and the stagnation carbon monoxide and lead (R = 0.91). Similar persistence index was 1. 12 (95% CI 1.05 to motor vehicle apportionment results were seen 1.19) in Spokane and 1.21 (95% CI 1.09 to when either carbon monoxide or lead were used in a multiple linear regression receptor 1.35) in Seattle for an increase of 11 and 10 10 hours, respectively, of low wind speed in a model. given day. The stagnation persistence Recent studies have focused on the associ- index was only correlated with one set of ation between general levels of air pollution in factor loadings; that cluster included the urban areas and visits to the emergency stagnation persistence index, carbon department for asthma. A number of studies University of have found a significant association between Washington, Seattle, monoxide, and organic/elemental carbon. Conclusion—Increased air stagnation was increased levels of particulate matter (PM) and Washington, USA increased visits to the emergency department G Norris shown to be a surrogate for accumulation 11–15 T Larson of the products of incomplete combustion, for asthma. The identity of the specific J Koenig including carbon monoxide and fine par- components of PM responsible for the ob- L Sheppard ticulate levels of organic and elemental served association remains unclear. In the carbon, and was more strongly associated present study we have used our ability to quan- Washington State tify the relevant meteorology associated with University, Pullman, with asthma aggravation than any one of Washington, USA the measured pollutants. increased pollution events to identify more C Claiborn (Thorax 2000;55:466–470) specifically those air pollutants associated with D Finn emergency department visits for asthma in Keywords: asthma; air stagnation; wind speed; air Spokane and Seattle, Washington, USA. Correspondence to: pollution Dr J Q Koenig, Department of Environmental Health, 357234, University of Methods Washington, Seattle, Previous studies have shown that certain meteo- SPOKANE Washington 98195, USA rological conditions are useful for identifying The study was conducted in Spokane, Wash- sources of particulate matter which are respon- ington, a city located in a river valley in an arid Received 27 May 1998 12 Returned to authors sible for outbreaks of asthma. Point sources agricultural region. According to a 1990 census 31 July 1998 of particulate matter from grain3 and soybean the population of the study area including the Revised version received dusts45have been identified as being responsi- city of Spokane and surrounding areas is 21 February 2000 Accepted for publication ble for asthma outbreaks. These studies 300 000. In Spokane the major sources of par- 21 February 2000 demonstrate the usefulness of combining ticulate matter are varied and include wood Asthma aggravation, combustion, and stagnant air 467 hospitals in Spokane. Approval for the use of anonymous hospital data for both cities was 0.2 given by the University of Washington Human Subjects OYce and the participating hospitals. The association between daily emergency 0.0 department visits for asthma and the stagna- tion persistence index were examined. SEATTLE The results found in Spokane were then tested Smoothing spline (df = 4) for comparison on data collected in previous _0.4 research in Seattle, a coastal city located on 0 5 10 15 20 25 Puget Sound. The population of the Seattle Stagnation index (hours) study area contributing cases to the emergency Figure 1 Smooth function of the stagnation index department data was approximately one mil- regressed on visits to the emergency department for asthma lion. In Seattle the major sources of particulate by non-elderly subjects after adjustment for day of week, matter are wood burning, automobiles, heavy time trends, temperature, and dew point temperature. (The duty diesel vehicles, and resuspended road ordinate is the contribution to the additive predictor in units dust, but not grass field burning or dust storms. of (loge(visits/day)). The Seattle study used data collected from burning, grass field burning, automobiles, September 1995 to December 1996 and was heavy duty diesel vehicles, resuspended road confined to asthma cases below the age of 18 dust, and dust storms from nearby arid regions years. Meteorological and air pollution data for of central Washington. Seattle were obtained from the Puget Sound Measurements of atmospheric pollutants Air Pollution Control Agency (now Puget and meteorological variables were obtained Sound Clean Air Agency), Washington State from an ongoing Spokane PM health eVects Department of Ecology, and the National Cli- study, Washington State Department of Ecol- matic Data Center. Wind speed was averaged ogy, and the National Climatic Data Center. A over four sites and emergency department preliminary examination of the meteorology visits for asthma were obtained from six indicated that an increased occurrence of low participating hospitals. Data for factor analysis wind speeds was associated with increases in were available from a previous study conducted the concentrations of combustion related between July 1993 and June 1994 in Seattle. pollutants. To quantify this association we used a simple variable named the “stagnation STATISTICS persistence index”16 defined as the number of The daily emergency department visits for hours in the 24 hour day in which the wind asthma were regressed on predictor and speed is less than the 50th percentile of the confounding variables using a semi-parametric hourly wind speed at that location. The 50th Poisson regression model. As the daily emer- percentile was chosen to provide suYcient gency department asthma visits are count data variation in this index to span the entire and are also rare events, it is reasonable to observed range of values between 0 and 23. We assume (conditional on the covariates) that the averaged this index over three sites. The counts have a Poisson distribution. The eVect stagnation persistence index may reflect rela- of air pollution on the increased number of tive changes in urban pollution exposure levels visits to the emergency department for asthma better than traditional air pollutant measures at is small and confounders such as season either a given site because meteorology is a more obscure or inflate the small eVect estimates. In regional variable. The correlation between the order to evaluate the association between daily stagnation persistence index at any one emergency department visits for asthma and site and the daily average value for all sites in air pollution, base models were created to con- the urban area was high (R varied between 0.90 trol for potential systematic eVects such as day and 0.94 for three Spokane sites). of week, and confounding by seasonal trends, Data on emergency department visits for temperature, and dew point temperature. asthma were obtained from four participating Separate base