Effects of heavy industrial pollution on respiratory function in the children of Cubatao, Brazil: a preliminary report.

Under a cooperative agreement between New York University and the Environmental Protection Agency, and in collaboration with the University of Sao Paulo (USP), a study is ongoing in Cubatao, Brazil, to try to establish exposure-response relationships on the impact of specific industrial effluents on respiratory function in school-age children. Cubatao, located on the coast about 44 km from the city of Sao Paulo, is surrounded by U-shaped mountains (approximately 800 m) covered with subtropical forests. Its area is approximately 160 km2, and it has a population of approximately 90,000. The geography is such that it causes a consistent diurnal land-sea breeze pattern and the opposite during the night, with low dispersion of the air pollutants. In a small area (approximately 40 km2) against the mountains there is a concentration of over 20 large plants: oil refinery; iron and steel mill; fertilizer, cement, and gypsum production; coke kilns; and chemical, paint, and many other ancillary plants. During the 1988 school year, March through June, August through November, 600 six-year-old children, attending six different kindergarten schools, underwent monthly spirometry tests. Because the children live within a 500 m radius of their school, pollution monitors were located on each of the six schools. Particles were collected using dichotomous stacked filter units placed on 20 m towers to reduce the influence of dust from unpaved roads. The units use different pore size filters for coarse, 2 to 10 microns, and fine particles, (dp) less than 2 microns, and took separate samples for day and nighttime.(ABSTRACT TRUNCATED AT 250 WORDS)


Introduction
Cubatao, Brazil, has been affected for many years by very serious pollution problems due to a combination ofunfivorable topography, enormous industrial emissions, lack of urban planning, etc. The area ofCubatao, 162 kn2, with 10700 inhabitants, is located about 44 km from Sao Paulo on the Tropic of Capricorn. This region extends parallel to the South Atlantic coastline, and it is surrounded on three sides by hills and mountains covered by a tropical forest of the Atlantic humid type. The mountains run parallel to the coastline (SW-NE), reaching heights of700 to 1000 m above sea level. Inside this region there are some low, isolated hills and rivers and a very irregular distribution of industrial and residential centers. The wind patterns in the Cubatao area are strongly influenced by the local topography. The behavior ofdraining winds depend on the time of day, solar incidence, and angle. The draining winds start after sunset, but strong winds from the Mogi Valley merging with NE winds of about 6 km/hr from the Quilombo Valley tend to carry the emissions back to the Vila Parisi area (Fig. 1). Fertilizer plant emissions from Cubatao Centro (downtown Cubatao) have been observed deep in the Mogi Valley. In winter mornings, there are layers ofthermal inversions ofdifferent depths and varied intensities. Emissions estimates, based on the industrial processes from 21 sources in the area of Cubatao (11 chemical and petrochemical industries, 7 fertilizer plants, 1 nonmetallic mineral plant, 1 paper mill, and 1 cement plant) yielded, for 1988, 25,500 tons ofparticulate matter, 17,800 The residential areas ofCubatao are intermingled with the industrial plants, commonly within 100 to 1000 m. The houses are extremely well ventilated, and cooking is done with liquified gas. While a large percentage ofthe adults smoke, the children's pollution exposure is essentially determined by the emissions from the nearest industrial plants. Children of 6 years of age attend kindergarten schools within a 500 m radius of their home, and in almost all cases have lived at the same address all their lives.
Hofmeister and Fischer (2) conducted a preliminary study in the summer of 1983 on 480 kindergarten children attending 11 schools and again in the winter of 1985 on 630 kindergarten children attending 16 schools in order to investigate the prevalence of impairment in lung function indices after living in the different neighborhoods ofCubatao for most oftheir lives. Each child performed stndard spirometric tests, and their results were compared to Polgar's (3) predicted values. Air pollution levels were monitored continuously, for SO2 and thoracic aerosol mass concentration (PMIo) by the state agency CETESB (Companhia de Technologia de Saneamento Ambiental) at two locations: Cubatao Centro (downtown, which is mostly residential; and Vila Parisi, surrounded by fertilizer, cement, and steel mill plants. The daily maximum and mean values ofPM1o, SO2, and ozone (03) for 1983 to 1988, as released by CETESB, are listed in Table  1. The measured forced vital capacity, forced expiratory volume in 1 sec, peak expiratory flow rate, and forced expiratory flow between 25 and 75 % ofFVC (FVC, FEVI, PEFR, and FEF25-75, respectively) as a percent of their predicted values showed significant differences among schools. Table 2 gives a summary of the mean PEFR values measured in 1983 and 1985 for the kindergarten children that attended the same schools as in the present study. In the 1983 study, the worst values were attained by the children of School No. 6 located near the petrochemical and styrene plants, at Cota 200. Low values were also measured in a school located near an important highway and a refinery, Vila Elizabeth, School No. 5. Other schools, based at greater distances from industries, showed less lung impairment.
As part of a broad cooperative study between New York University Medical Center's Institute of Environmental Medicine and the University ofSao Paulo's School ofPublic Health and Institute ofPhysics, a new cohort of600 6-year-old children attending their neighborhood kindergarten school were recruited in March 1988. Simultaneously, six new air sampling stations were installed by us to gather more comprehensive air pollution data at five kindergarten schools and at a strong source area (Vila

Methods
Schools were selected by location, some close and some far from industries, to have a range and variety of pollution levels and, at the same time having similarities in socioeconomic levels. As shown in Figure 1, Cota 95 (at 95 m) and Cota 200 (at 200 m) are residential areas located in the Atlantic forest relatively far from industries but within the thermal inversion layer. Nova Republica is a new residential area in the south, developed to house the ex-residents of the extremely polluted Vila Parisi. Jardim das Industria, Vila Elizabeth, and Vila Nova are residential areas within the industrial complex in the central area of Cubatao.

Air Sampling Measurements
Aerosols were collected at six sampling stations from May to November. Five sampling sites were located at kindergarten schools attended by the children in the new cohort. The sixth sampling site was located in the highly polluted neighborhood of Vila Parisi next one ofthe CETESB stations to allow for crosscomparisons ofresults. The Vlla Parisi and Jardim das Industrias samplers were located in a tower about 20 m tall to reduce the contribution ofsoil dust aerosols from the unpaved roads in the area; the other sites, Vila Nova, Nova Republica, Cota 95, and Cota 200, were sampled at about 3 m height. The CETESB Cubatao Centro sampler was used to represent exposures at Vila Elizabeth. Due to wind pattern, day and night samples were collected separately; daytime samples from 7 A.M. to 7 P.M. and nighttime samples from 7 P.M. to 7 A.M. Aerosol particles were collected using stacked filter units (SFU) (4) at flow rates between 3 and 6 L/min, coarse particles (2 Mm <dp< 10 Mm; dp = particle diameter) were collected on Nucleopore 8-gm pore size, 47-mm diameter filters, and fine particles (dp< 2 Mm) were sampled on 0.4 itm Nucleopore filters. A total of 710 SFU samples were collected. The fine and coarse aerosol mass concentrations were obtained by gravimetric analysis. The filters were weighed before and after sampling with a Mettler M3 electronic microbalance with a 1-pg sensitivity. Before weighing, the filters were equilibrated for 24 hr at 20°C and 50% relative humidity. Electrostatic charges were controlled by 210o sources.

Lung Function Measurements
The children, 55 % boys and 45 % girls, performed standard spirometric maneuvers on a Vitalograph spirometer until they had three acceptable perforances. Spirograms were read by two independent observers and the results stored in computer files. Testing was performed monthly from March to November, the standard school year in the southern hemisphere, except for the month ofJuly, when schools close for their winter recess; Each child's height and weight was recorded every month. Each child's monthly measurement was performed between 28 and 32 days from the previous one. The children's collaboration was excellent, 456 children had at least 6 measurements, ofa possible 8 yielding a compliance rate of 76%.
All children's homes were visited by an interviewer who completed a questionnaire on the medical history ofthe child and on the habitational characteristics and working, residence, and smoking history of the household members.

Aerosol Concentrations
The average PMlo or thoracic aerosol mass concentration was very high for all sampling sites. Figure

Discussion
The preliminary results ofthis study suggest that cumulative monthly exposure to nonspecific thoracic mass concentration reduces lung airway flow rate. The average yearly PM10 concentrations ranged from 59 ,ug/m3 in Cota 200 to 110 jIg/M3 in the Nova Republica neighborhood, but there was no evidence for a threshold in that the decrement per unit mass concentration was similar for the less exposed and more heavily exposed regions.
The school cohort showing the largest decrements in FEF25-75 is the one located in Nova Republica. This is the school attended by the children that lived in Vila Parisi before being relocated within the previous few years. These children had been exposed in Vila Parisi to PMlo concentrations about twice those in their present location. High exposures in the past may predispose them to greater transient changes in FEF25-75 as PM10 concentrations rise.
Additional analyses are currently in progress to examine the effects of chronic pollution exposure on lung growth rates and to determine whether specific constituents ofthe particulate matter have disproportionate effects on lung function.