669

EPIDEMIOLOGICAL STUDIES OF LUNG DISEASE IN

.

URBAN AND RURAL COMMUNITIES

A. BOUHUYS, C. A. MITCHELL, H. R. HOSEIN,

J. B. SCHOENBERG, R. E. BINDER, R. S. F. SCHILLING

Yale University Lung Research Center, New Haven, Connecticut, USA

ABSTRACT

We have studied over 7000 residents of one urban and two rural communities in the United States, aged 7 years and older, residing in geographically determined areas in which a private census was obtained. The 7000 persons seen comprise about 75% of the defined populations; additional door-to-door surveys provided information about those not seen for validation of the resulta in terms of the total community populations. A mobile laboratory with an on-line computerized data collection system was used to record answers to a questionnaire on respiratory symptoms, residential and family history, smoking habits and other environmental exposures, and to record maximum expiratory flow-volume curves. Outdoor air quality is being determined over a period of more than one year in each community, using standard methods for particulates, SO, NO2, ozone, sulfate and nitrate. In addition, studies with portable air samplers for respirable particulates, 50, and NO2, were carried out by selected subjects, monitoring their personal environment over 24-hour periods. The results in the urban and in one rural population, both in Connecticut, indicate: (1) there is no significant excess of respiratory symptoms among lifetime urban dwellers, living in an area where particulate pollution exceeds the primary AQS in 39% of 102 samples; (2) sensitive lung function tests do not show differences attributable to an increased prevalence of

2

670

airway obstruction among urban adults when compared to rural dwellers; (3) indoor particulate pollution may add considerably to the total respirable particulate load; (4) occupational or domestic, toxic and allergenic inhalants, including tobacco smoke, are important etiological factors in many persons with airway obstruction;

role.

outdoor air pollution may only play a minor

671

1. Introduction

Epidemiological studies on chronic bronchitis and other lung diseases in communities have generally used population samples. Such samples often contain too small numbers of persons in defined, homogeneous subgroups to allow study of more than a few variables. Computerized methods now allow us to broaden the scope of such surveys to total community populations, to a wider range of questions concerning lung disease, and to more sophisticated lung function tests. We have developed a mobile computerized laboratory, in which standardized, computer-prompted interviews on respiratory symptoms and environmental exposures (for text of questionnaire, see Bouhuys [1]) as well as lung function testing are conducted simultaneously on 2 terminals connected to a PDP-8E computer, using a time-sharing system. With this facility, we have recorded symptoms and lung function data in about 2500 residents of a rural town. (Lebanon) and 1500 residents of an urban area (Ansonia), both in Connecticut, U.S.A. Similar data have recently been obtained in about 4000 residents of a rural town (Winnsboro) in South Carolina (Feb.-May 1974). In each community, a total population census as well as door-to-door surveys in selected areas provide information on persons not seen in the mobile laboratory, thus enabling us to establish the validity of our findings in terms of the total population. An air sampling network with 3-5 stations is used to characterize outdoor air quality, using standard methods recommended by the Environmental Protection Agency. 24-hour samples are analyzed for suspended particulates, sulfate and nitrate, as well as for SO2, NO2 and ozone. In the two Connecticut

towns, samples were obtained at least weekly; in Winnsboro, S.C., daily samples are obtained during four 3-week periods in the fall, winter, spring and summer. This schedule of sampling defines the outdoor pollution load

over a year.

This paper describes initial results of the surveys in Lebanon and Ansonia, Conn., and of air quality measurements in these communities. In addition, preliminary data on 24-hour personal environmental sampling to compare individual pollutant exposure and outdoor air quality are included. 2. Lung function tests. We have selected the maximum expiratory flowvolume (MEFV) curve as a simple, objective test of ventilatory function which provides: (a) standard spirometric values, (b) measurements of flow rates at low lung volumes, which are sensitive indices of airway obstruction, and (c) visual recognition of obstructive and restrictive function loss (see review in Bouhuys [1]).

672

1.0

The mobile laboratory is equipped for simultaneous recording of MEFV curves on two terminals, using Fleisch pneumotachographs to measure expiratory flow rates. Flow signals are digitized and stored by the computer, which also integrates flow to volume, presents each MEFV curve on the graphic terminal, and calculates forced vital capacity (FVC), forced expiratory volumes (FEV, and FEV2) and peak expiratory flow rate (PEP) from the 2 blows with the highest FEV in a series of 5 blows. Average maximum expiratory flow rates at 50% and at 25% of the FVC (1.e., total lung capacity minus 50 or 75% of the FVC) were measured from the same two blows. In contrast to PEF, these flow rates (MEF50% and MEF25%) are largely independent of muscular effort; while PEF reflects tracheal size and effort, MEF50% and MEF25% depend primarily on the caliber of small airways. Quality-control features of the software include rejection of too short blows, and calibration and zero drift correction subroutines. Precision flowmeters, as well as a newly designed motor-driven 5-liter syringe which delivers air at varying rates simulating an MEFV curve, are used to calibrate the volume and flow measurements and to check the computation subroutines. For a more detailed description of the system, see Bouhuys (1).

Output format. After the questionnaire and MEFV curves have been recorded, data output is obtained (a) in a condensed hard copy format for immediate review and checking, (b) on punched paper tape, for editing and transfer to magnetic tape on a home-based PDP-11 computer which classifies the results in tabular form and performs statistical analyses.

3. Personal environment sampling. To obtain air quality data that may more adequately reflect an individual's changing environment, we have developed a portable light-weight and low-noise personal environment sampler (PES) that allows individuals to monitor their environment during 24 hours (Hosein et al. [2]). The design is a modification of an instrument developed by Burgess et al. [3]. The unit is housed in a suitcase, which coatains a blower drawing in air at 250 liters/min. From the mainstream, iadividual sampling pumps draw air through collection vessels for SO2 and W and through a membrane filter for respirable particulates (approx. <7pa), using a cyclone.

4. Results. Symptoms and lung function. Analysis of data obtained in door-to-door surveys in selected areas of each town has shown similar sy tom prevalences among those seen in the mobile laboratory and those not seen, if men and women of similar age are compared. These results suggest that our results can be extrapolated to the total community population.

673

There were no significant differences between the prevalences of persistent cough and phlegm, frequent wheezing and dyspnea on exertion among rural and urban males (age 7-80 yrs); their smoking habits were similar. The prevalence of persistent cough and phlegm was similar among rural and urban nonsmoking men and women (25-64 yrs), except that older urban women (45-64 yrs) more frequently reported persistent phlegm (x2 5.26, p <0.05). The prevalence of these symptoms was higher among current smokers (men and women, 25-64 yrs), without significant differences between rural and urban dwellers.

The lung function data showed similar differences. For instance, FEV and MEF50% were lower in smokers than in nonsmokers but there were 1.0 no significant differences between nonsmoking or smoking rural and urban men or women, 25-64 yrs. In this analysis, stature was taken into account by including only data from 150-169 cm tall women and 160-179 cm tall men. The interpretation of lung function in younger persons (7-24 yrs) requires a more complex analysis of the simultaneous effects of growth, smoking and urban or rural residence; this has not yet been completed. Thus, we have found no evidence of increased airway obstruction among urban adult men and women when compared to rural residents, using a sensitive lung function test (MEF50%) which, e.g., detects significant airway obstruction in teenaged smokers (Seely et al. [4]).

Outdoor environment sampling.

In Lebanon (rural), weekly SO2 and ozone levels were always less than about 50% of the primary air quality standard (AQS), while total particulates and NO2 exceeded the AQS in, respectively, 2 and 10% of all samples taken over a 42-week period. In Ansonia (urban), ozone was always less than 50% of the AQs, while particulates, NO2 and SO2 exceeded the AQS in, respectively, 39, 35 and 1% of all samples taken over a 42-week period. Ansonia, situated in the heavily industrialized Naugatuck valley, is generally considered a polluted town, and our measurements confirm that particulate and NO2 levels frequently exceed those in rural

Connecticut.

Personal environment sampling. The highest outdoor total particulate level measured in Ansonia was 144 μg/m3. The personal environment samplers (PES) frequently indicated higher respirable particulate levels (up to 303 μg/m3) when they were carried about by selected residents for 24 hours, covering their normal daily indoor and outdoor activities, within 2 miles

from the outdoor samplers. In general, SO2 and NO2 levels with the PES

were less than those in outdoor air. Thus, indoor environments may add