U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES

AGENCY FOR TOXIC SUBSTANCES AND DISEASE REGISTRY

ATLANTA, GEORGIA

DIAGNOSTIC EVALUATION OF CHILDREN WITH RESPIRATORY SYMPTOMS

AND POTENTIAL EXPOSURE TO DIISOCYANATES RELEASED FROM

THE TRINITY AMERICAN CORPORATION

GLENOLA, NORTH CAROLINA

FINAL REPORT

JULY 2001

In 1980, Congress created the Agency for Toxic Substances and Disease Registry (ATSDR) to implement health-related sections of laws that protect the public from hazardous wastes and environmental spills of hazardous substances. The Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA), commonly known as the "Superfund" act, designated ATSDR as the lead agency within the U.S. Department of Health and Human Services to help prevent or reduce further exposure to hazardous substances and the adverse health effects that result from such exposures, and also to expand the knowledge base about such effects.

This publication reports the results and findings of a health study, registry, or other health-related activity supported by ATSDR in accordance with its legislative mandate described above.

Comments regarding this report are welcome. Please address to:

Agency for Toxic Substances and Disease Registry Jeffrey P. Koplan, MD, Administrator

Henry Falk, MD, MPH, Assistant Administrator

Robert Spengler, ScD, Associate Administrator for Science

Division of Health Studies Jeffrey A. Lybarger, MD, MS, Director

Sharon S. Campolucci, MSN, Deputy Director

Lucy A. Peipins, Assistant Director for Science

Nancy A. Whitehead, Editor

Health Investigations Branch Mary C. White, ScD, Chief

Epidemiology and Surveillance Branch Wendy E. Kaye, MHS, PhD, Chief

Exposure and Disease Registry Branch Ginger L. Gist, PhD, Chief

Use of trade names and commercial sources is for identification only and does not imply endorsement by the agency for Toxic Substances and Disease Registry or the U.S. Department of Health and Human Services.

LIST OF TABLES

Table 1-Demographic and study characteristics, by participant selection group

Table 2-Respiratory history of study participants reported during the screening interviews, by residential distance category

Table 3-Respiratory history of study participants reported during the in-person interviews and diagnostic outcomes, by residential distance category

Table 4-Improvement in abnormal pulmonary function test results after administering albuterol

Table 5-Diagnostic outcomes and pulmonary function test results compared to predicted values, by residential distance category

Table 6-Pulmonary function test results compared to predicted values, by diagnostic outcome

FIGURES

Figure 1.- Flowchart
GIF version
PDF version

LIST OF APPENDICES

APPENDICES

ABSTRACT

Previous environmental and biomedical testing in the Glenola community of North Carolina was consistent with human exposure to diisocyanates, a potent group of pulmonary sensitizers. These substances were released from the nearby Trinity American Corporation facility during polyurethane foam production. Airborne releases from the plant increased when the "quick-cure" method was introduced in 1993.

The main goal of this investigation was to identify children with asthma who lived near the plant when the quick-cure method was used. When the parents or guardians of 231 local children were interviewed by telephone, a potential for exposure was confirmed for 204 children; 118 of these children had respiratory symptoms and were offered a clinical evaluation. A diagnosis of asthma was made for 28 of 55 children from the study area who completed a clinical evaluation; asthma was considered possible for another 10 children. Recommendations for medical care were provided as appropriate.

A secondary goal was to characterize the current burden of pediatric asthma in the community. Participation in the telephone screening interviews was excellent, but a limited number of eligible children completed a clinical evaluation. As planned, statistical inferences are avoided and grouped analyses are presented as descriptive and exploratory. Even so, the information collected is most consistent with a high prevalence of asthma among the community's children. Two children had antibodies to diisocyanates, adding to the evidence for exposure.

INTRODUCTION

Background Information

The Trinity American Corporation (Trinity American) produced polyurethane foam and fiber in Glenola, North Carolina, from 1981 until September 1997. Introduction of the "quick-cure" process in 1993 was associated with increasing concern among local residents about odors and the possibility of adverse health effects (1). This process produced foam by reacting a polyoxypropylenetriol resin with water and an excess of toluene diisocyanate (TDI), a well-recognized cause of occupational asthma (2). TDI that remained after the polymerization process was exhausted directly into the air; at times, methylene chloride was used as a blowing agent to produce a higher grade of foam.

Recent Events

In 1995, a number of adult Glenola residents complained to the North Carolina Department of Environment, Health, and Natural Resources (NCDEHNR) that an increase in emissions from the Trinity American facility was adversely affecting their health. The State of North Carolina supported a case series of standardized clinical evaluations for symptomatic residents between June 1997 and March 1998, which was conducted by the Occupational and Environmental Medicine program at Duke University. Six of 33 adults tested (18.2%) had antibodies to one or more diisocyanates, but one of them may have been exposed at work. The report released in 1998 concluded that the results were "...highly suggestive of environmental exposure from the plant." Many residents had symptoms consistent with reactive airway disease, and 61.1% (22 of 36 adults tested) reacted during methacholine challenge testing. The authors concluded that "...a plausible link exists between exposure...and symptoms experienced by community residents" (3).

In the fall of 1996, Glenola residents petitioned the Agency for Toxic Substances and Disease Registry (ATSDR) to determine whether emissions from the Trinity American facility could adversely affect their health. Environmental monitoring provided evidence that diisocyanates were periodically present in the air. When ATSDR established a "call-in line" in 1997, Glenola residents made more than 200 phone calls to report site emissions or neurological and respiratory symptoms. Some personnel from ATSDR and the U.S. Environmental Protection Agency (EPA) who were present during a visible release from the facility confirmed the odor and also experienced some of these symptoms (4).

In order to further assess exposure in this community, ATSDR and the Randolph County Health Department conducted a biological exposure investigation in the fall of 1997. Local residents (n=113) provided blood samples to be tested for immunoglobulin E (IgE) and immunoglobulin G (IgG) antibodies to diisocyanates. The majority of these participants were adults. Ten participants had specific antibodies to one or more diisocyanates, but 1 of the 10 may have been exposed at work (5,6).

The ATSDR Health Consultation (4) found evidence of a completed exposure pathway; accordingly, ATSDR issued a Public Health Advisory recommending additional testing to assess exposure and health effects among Glenola residents (1). Community members offered little support for additional medical testing among adults, but acknowledged their concern for area children. Anecdotally, one of the original petitioners to ATSDR moved away from the area because a preschool child developed respiratory problems.

GOALS AND OBJECTIVES

The primary goal was to identify and make treatment recommendations for children with asthma who were exposed to airborne emissions from the Trinity American facility. The objectives supporting this goal were to

• screen potentially exposed children for respiratory symptoms by interviewing parents or guardians by telephone,

refer symptomatic children for diagnostic evaluations (including pulmonary function tests and total IgE antibodies) by asthma specialists, and test symptomatic children for antibodies to diisocyanates.

A secondary goal was to characterize the current burden of pediatric asthma in this community.

STUDY METHODS

Study Design

This study evaluated respiratory health among children residing near the Trinity American facility in Glenola, North Carolina. A cross-sectional, telephone-screening survey identified children with respiratory symptoms commonly associated with asthma. These symptomatic children were offered a clinical evaluation by a physician specializing in the diagnosis and treatment of pediatric asthma. Clinical participants were asked to provide a blood sample to be tested for antibodies to diisocyanates, a biomarker that occurs in some exposed individuals.

Telephone screening was conducted from December 30, 1998, through January 30, 1999. Diagnostic evaluations were offered on weekends during March 1999 and on April 10, 1999.

Eligibility Criteria

The study area included residences located within a 1-mile radius of the point source of emissions from the Trinity American facility. The study time period was from January 1993 through September 1997.

Children were eligible for the study if they met the following criteria:

• the child's name appeared on the list of Randolph County School System students who registered for the 1998-1999 school year,

• the child's current residence was confirmed to be located in the study area by a geographic information system (GIS), and

information collected during the screening survey confirmed that the child had resided in the study area for at least 2 months during the study time period.

In addition, the Randolph County Health Department distributed ATSDR information sheets about the study to households in the study area. Parents and guardians were offered a toll-free phone number to use for identifying additional school-aged children.

Screening Outcomes

If the parent or guardian reported that the child had one or more respiratory symptoms associated with asthma, the child was considered symptomatic and was offered a medical evaluation. In summary, the category entitled "symptomatic" included children with any of the following:

night cough, wheezing, or shortness of breath during the past 12 months,

ever wheezing, or

• an asthma diagnosis from a doctor after reaching 5 years of age.

Younger children can have episodes of wheezing without bronchospasm, and they often are not able to cooperate during spirometry.

Study Instruments

Screening Interview Questionnaire

The screening questionnaire (Appendix A) was administered by telephone. The residential address(es) of each child during the study time period (1993-1997) was recorded on the interview form and plotted using GIS. If the study's eligibility criteria were met, the child became a study participant. Study participants were assigned to one of the three following residential distance categories on the basis of the distance between their closest residence during the study period and the point source of airborne emissions:

• the near-distance category ( 0.50 mile),

• the middle-distance category (0.50 miles < distance 0.75 mile), or

• the far-distance category (0.75 mile < distance 1.0 mile).

These category labels are meaningful only in relation to each other; in addition, school-aged children are mobile. All children living within 1 mile of the point source were considered at risk for exposure.

The questions used to categorize each child as symptomatic or asymptomatic are found in Appendix A (Questions B1 through B7). These questions identified key indicators

(symptoms or medical history) associated with asthma. The wording was modeled after questionnaires from the following:

• International Study of Asthma and Allergies in Childhood (ISAAC),

National Health and Nutrition Examination Survey (NHANES), and

Guidelines for the Diagnosis and Management of Asthma (7-9).

In-Person Interview Questionnaire

When symptomatic children came for a clinical evaluation, informed consent was obtained from the parent or guardian and assent from the child. Afterwards, the interviewer administered the questionnaire in Appendix B to the parent or guardian; the information collected was then reviewed by the examining physician.

Physician's Checklist

Physicians checked off significant findings related to medical history, respiratory symptoms, and the physical examination on a standardized checklist (Appendix C). This information was later used to dictate the medical record for each child and letters to their parents or guardians.

Diagnosis Form

The diagnosis form used by the study physicians (Appendix D) was designed to encourage consistent recording of medical diagnoses without inappropriately limiting their choices. The form provides discrete outcome categories that are amenable to electronic storage and descriptive analyses. On the basis of each participant's medical history, physical examination, and pulmonary function tests (PFTs), the physician assigned one of the following outcomes:

Outcome 1 -- Asthma is present (supported by the history and abnormal PFTs),

Outcome 2 -- Asthma is present (supported by the history without abnormal PFTs),

Outcome 3 -- Asthma is possible (cannot diagnose or exclude asthma), or

Outcome 4 -- Asthma is not present (the diagnosis of asthma was excluded).

Pulmonary Function Tests

The consensus guidelines published by the National Institutes of Health (NIH) base the diagnosis of asthma on (a) the presence of episodic symptoms of airflow obstruction,

(b) a determination that the airflow obstruction is at least partially reversible, and

(c) the exclusion of alternative diagnoses (9).

Pulmonary function tests (PFTs) in children typically include the forced expiratory volume (liters) in 1 second (FEV1), the forced vital capacity (FVC) in liters, and their calculated ratio (FEV1/FVC). The FEV1 and FEV1/FVC are lower when the airway is obstructed. Generally speaking, airflow obstruction is established if (a) the FEV1 is less than 80% of predicted volume or (b) the FEV1/FVC ratio is less than 65% (or below the lower limit of normal). Reversibility is typically established if the FEV1 increases by 12% (or more) after a short-acting beta2-agonist (e.g., albuterol) is administered. These guidelines for interpreting PFT results were specified in the study protocol.

It is more difficult to obtain consistent spirometry results in younger children, and the airway obstruction in asthma is variable. Because priority was given to the clinical standard of care, the clinicians actually interpreted the results as they would in their routine clinical practice (see Discussion section). This included consideration of the flow rate during the middle of forced expiration (FEF ^25-75), a parameter shown to be clinically relevant in pediatric asthma (10). In descriptive analyses of the information collected during this investigation, "abnormal" PFT results refer to those that the clinician interpreted as "abnormal."

Biomarkers

After obtaining informed consent from the parent or guardian, and assent from the child, a blood specimen was collected and sent to the University of Cincinnati Diagnostic Allergy Laboratory. This specimen was tested for immunoglobulin G (IgG) and immunoglobulin E (IgE) antibodies to toluene diisocyanate (TDI), diphenyl-methane diisocyanate (MDI), and hexamethylene diisocyanate (HDI). The laboratory procedures used were identical to those described by Orloff et al. (6). A blood specimen also was sent to a local laboratory to be tested for total serum IgE antibodies.

Field Procedures and Quality Assurance

Interviewing and Scheduling

The National Opinion Research Center (NORC) was contracted to administer the screening interviews by telephone, schedule eligible children for diagnostic evaluations, obtain informed consent from parents or guardians and assent from children, administer the in-person interview, and to escort participants through the phases of diagnostic evaluation. NORC developed procedure manuals for telephone interviewers and field personnel. After receiving training, NORC interviewers demonstrated their competence during mock interviews before they began to interview study participants.

Medical Personnel

Two physicians, certified by the American Board of Pediatrics and the American Board of Allergy and Immunology, were recruited from the same university-based specialty practice. The two spirometry technicians selected were experienced in administering spirometry with children, and they met the current criteria of the American Thoracic Society (ATS) (11). Phlebotomy technicians with pediatric experience were recruited from a local hospital.

Pulmonary Function Testing

Objective measurements provide better evidence for airway obstruction and reversibility than either symptoms or physical examination. Spirometry was chosen (rather than peak flow measurements) for the pulmonary function tests (PFTs), because more reliable comparison values were available for spirometry. The spirometry equipment used met the standards set by the American Thoracic Society, and the technicians followed established procedures regarding technique, calibration methods, and preventive maintenance.

Data Management and Descriptive Analyses

The information obtained during the telephone survey, the in-person interview, and the clinical evaluations was recorded initially on paper forms. EpiInfo Version 6.04b (12) was used to create a customized data-entry program with error-checking capability. After data entry, the database was rechecked by selected comparisons with the original (paper) forms.

All analyses were performed using Version 6.12 of the Statistical Analysis System (SAS) for Windows (13). The database was translated from EpiInfo to SAS and combined with other electronic data (i.e., PFT results and blood test results) submitted by the clinical contractor. The study parameters were examined for consistency using standard SAS procedures. The prevalences of various respiratory symptoms reported during the screening interviews were examined after grouping children by the distance between their residence and the point source of airborne emissions. Respiratory symptoms and diagnostic outcomes for children who participated in the clinical phase of this study were examined in a similar manner.

RESULTS

Figure 1 tracks study participants through the selection, screening, and clinical phases of the investigation. The reader may wish to refer to Figure 1 periodically while reading this report.

Screening

Selection and Participation in the Screening Interview

GIS plots confirmed that 225 of the 259 children identified by the Board of Education resided in the study area. During the telephone interviews, 24 siblings of study participants were identified and added to the potential study population. Interviews were completed with parents or guardians of 231 (92.8%) of the 249 children who were potentially eligible for the study (Figure 1).

In addition to the children from study-area households, some parents or guardians requested screening for themselves or for children who did not meet the study's eligibility criteria. Those with a plausible potential for exposure (three parents and nine children) were screened and, if symptomatic, were offered a diagnostic evaluation. Because they did not meet the study's eligibility criteria, they did not become study participants, and they were not included in grouped analyses.

Demographics and Other Characteristics

The race of participants was not recorded during the interviews, but virtually all residents of the study area were white. Table 1 groups participants in the telephone screening by the manner each was initially identified; that is, they are grouped as (a) children from the original list provided by the School Board (n=207), (b) siblings added to the original list (n=24), and (c) all participants in the telephone screening combined (n=231). After plotting the addresses reported for the 1993-1997 period, 88.3% of these children met the study's eligibility criteria (Figure 1) and formed the study group (N=204).

Respiratory Symptoms Reported During the Screening Interview

On the basis of the proximity of their residences to the point source of airborne emissions, study participants were assigned to one of three residential distance categories: (1) near (n=77), (2) middle (n=54), or (3) far (n=73). Compared to the other two distance categories, children in the near-distance category had a slightly higher prevalence of the following indicators:

night cough during the past 12 months (42.9%),

wheezing during the past 12 months (31.2%),

ever wheezing (46.8%), and

ever having a diagnosis of asthma (19.5%)(Table 2).

These differences among distance categories were small, and no obvious dose-response relationship was observed. Among children in the near-distance category, prevalences of the following indicators were not increased:

sudden, severe, or recurrent episodes of shortness of breath, and asthma diagnosis (after reaching 5 years of age).

If one or more key indicators of asthma were reported during screening, the child was considered to be symptomatic. The overall prevalence of being symptomatic was 57.8% (118 of 204 children). The prevalence of symptoms among children in the near-distance category was similar to the prevalence among children in the far-distance category.

Diagnostic Evaluations

Selection and Participation

All 118 symptomatic children were offered a medical evaluation, including an in-person interview, a medical history, an examination of the respiratory system, and pulmonary function testing. Fifty-five (46.6%) of the symptomatic children (n=118) completed this evaluation.

Demographics and Smoking Prevalence

The 55 children who completed the study's diagnostic evaluation included 36 boys and 19 girls. The prevalence of "regular smoking in their homes at any time during the study period" was 58.8%. According to data from the Behavioral Risk Factor Surveillance System (BRFSS),

26.1% of children in North Carolina were exposed to environmental tobacco smoke in their homes during 1996 (14).

Respiratory Symptoms Reported During the In-Person Interview

Compared to the other two residential distance categories, children in the middle-distance category had a higher prevalence of "wheezing during the past 12 months" (76%) and "ever wheezing" (84.6%). A higher prevalence of asthma diagnoses after reaching 5 years of age (30.8%) also was reported for children in the middle-distance category.

Biomarkers

Forty-four of the 55 participants provided a blood specimen. One child had IgG antibodies to TDI and to HDI; this child lived approximately two-thirds of a mile from the Trinity American facility. A second child (tested by special request) also had IgG antibodies to TDI. This child was not a study-area resident, but the child had spent time at a residence near the Trinity American facility during the study period.

The total IgE levels of eight children were interpreted as "probably atopic allergy" by the reporting laboratory. Although total IgE levels for children with asthma are frequently increased, the relationship between total IgE and TDI-induced asthma is unknown. The test provided a moderately useful piece of information for clinicians to consider, but total IgE levels are not necessary to either diagnose or exclude asthma.

Pulmonary Function Tests

One 5-year-old participant was unable to successfully complete the pulmonary function tests. Table 4 shows the spirometry results of the 17 children whose PFTs were interpreted as "abnormal." For these subjects, administering the bronchodilator led to an average increase of

(a) 8.8% in FEV1 (range 3.1% to 15.9%) and (b) 31.9% in FEF ^25-75 (range 11.3% to 70.6%). Only 1 of the 17 children with asthma and abnormal PFTs had used a modern medication known to reduce airway inflammation (e.g., an aerosolized steroid) during the past 12 months.

Diagnostic Outcomes

The asthma specialists chose one of four diagnostic outcomes for each child (Appendix D). Because participation in the diagnostic evaluations was low, the outcomes were combined in Table 3 and Table 5. In Table 3, clinical asthma included "Outcome" 1 (asthma with abnormal PFTs) and "Outcome 2" (asthma without abnormal PFTs). The prevalence of clinical asthma was similar for children in the near (53.9%), middle (53.9%), and far (43.8%) distance categories.

In Table 5, Outcomes 1, 2, and 3 were combined to form the outcome "Asthma (present or possible)" and compared to Outcome 4 ("Not Asthma"), by residential distance category. The prevalence of Asthma ("present or possible") was higher among children in the middle-distance category (84.6%) than among children in the other two distance categories.

DISCUSSION

Study Strengths

The participation rate for the screening phase was high, with only one parent (of two children) refusing to participate. The parents of 16 children could not be located, so the screening interview was completed for 231 (92.8%) of the 249 eligible children.

The potential for residential exposure to airborne diisocyanates was supported by previous environmental and biomedical investigations, but symptomatic children had not previously been tested for biomarkers of exposure or disease. The health effect most often associated with TDI is asthma, a chronic disease with well-defined symptoms (2). The airway obstruction in asthma is variable; when obstruction is present, it can be documented with spirometry.

The physicians selected to conduct the clinical evaluations are recognized specialists in the diagnosis and treatment of asthma in children. This enhanced their credibility and provided a level of consultation that was not readily available in the community.

Study Limitations

The participation rate for the clinical phase was low. Although 118 children were considered symptomatic, only 55 (46.6%) participated in the diagnostic evaluations. The clinic at the Randolph County Health Department in Asheboro, North Carolina, was an ideal facility for the clinical evaluations, but participants had to drive up to 30 minutes to reach it. Travel time and distance were undoubtably limiting factors for some parents. The low level of participation also may be related to a progressive decline in community concern after the Trinity American facility closed in September 1997.

The screening interview was designed to screen for symptoms and collect residential history without burdening participants. This investigation was not designed to measure the association

between asthma and any exposure. No comparison group was studied and information about additional contributors to asthma (e.g., household pets) was not collected.

The study protocol described criteria from national consensus guidelines that are useful when a child is followed over time or evaluated during an acute respiratory illness (9). These criteria did not form a useful case definition for study participants, who were identified by screening interviews and evaluated in a one-time clinical encounter. This was predictable, given that airway obstruction and reversibility vary over time and that methacholine challenge testing was not employed.

For children with clearly abnormal PFTs (i.e., well-documented obstruction and reversibility) on the day tested, the diagnosis of asthma is straightforward. For children with normal PFTs on the day tested, categories that reflect some uncertainty are necessary (e.g., "possible asthma") to avoid constraining the examining physician to arbitrary choices (15). The specialists ultimately used their clinical judgment, giving additional weight to respiratory symptoms and accepting evidence for obstruction and reversibility that were not specified in the protocol (e.g., changes in FEF ^25-75). This was consistent with the specialists' usual clinical practice and with the priority given to providing clinical assistance to individual children in this investigation. Of course, clinical judgment might have contributed to the nondifferential misclassification of health outcomes. Although it is not feasible to assess the validity of individual diagnoses, it is possible to partially assess the validity of the diagnostic outcome groups that are formed (see Pulmonary Function Tests and Diagnostic Outcomes section).

Interpretation of Results

Within the study area, no relationship was observed between residential proximity to the Trinity American point source of airborne emissions and adverse respiratory health effects. It was assumed that the likelihood and intensity of exposure to TDI decreased with increasing distance between the residence and the source, but this is not a certainty. School-aged children are mobile, and the location of a child's residence may not accurately reflect the locations frequented during outdoor play.

Biomarkers

Antibodies to diisocyanates are uncommon in the general population (16, 17). Most individuals do not make these antibodies even when exposed, but a few positive results are usually found when an exposed group is tested. One large study of exposed workers identified IgE antibodies to diisocyanates in less than 10% of 1,780 adults tested (18). Among subsets of these workers who were also tested for IgG antibodies, IgG antibodies to diisocyanates were somewhat more common than IgE antibodies. The value for total serum IgE antibodies is sometimes increased in atopic (allergy prone) children, who are at increased risk for asthma. This test is not a sensitive biomarker for asthma, and it need not be elevated to make the diagnosis.

Previous biomarker studies identified adult residents in this community with antibodies to diisocyanates (3,5). Although the immunologic response of children to diisocyanates has not been studied, community children were tested with the expectation that a few of them were likely to have positive test results. Indeed, two children did have diisocyanate-specific antibodies, providing additional evidence of exposure in the community. These two children may have been exposed at two separate residences, both of which were located approximately two-thirds of a mile from the emission point (i.e., in the middle-distance category). Children with diisocyanate antibodies have not previously been reported in the scientific literature.

Respiratory Symptoms

Of the 204 participants who completed screening interviews, 118 (57.8%) were considered to be symptomatic and, thus, eligible for a clinical evaluation. Compared to children in the other two residential distance categories, children in the near-distance category had a slightly higher prevalence of night cough during the past 12 months, wheezing during the past 12 months, ever wheezing, and ever having a diagnosis of asthma. The differences in the prevalences of these symptoms among children in the three residential distance categories were small, however, and no obvious dose-response relationship was observed.

For the 55 participants who completed in-person interviews, certain key indicators were more common among children in the middle-distance category (Table 3). These indicators were wheezing during the past 12 months (76%), ever wheezing (84.6%), and an asthma diagnosis after reaching 5 years of age (30.8%). With such high prevalences of wheezing, one might expect that the prevalence of asthma also would be higher among children in the middle-distance category. In fact, Table 3 shows that the prevalence of clinical asthma (Outcomes 1 and 2 combined) was similar among children in all three residential distance categories. When Outcomes 1, 2, and 3 were combined as asthma (present or possible) and compared to Outcome 4 (not asthma), the prevalence of asthma (present or possible) was highest among children in the middle-distance category (84.6%) (Table 5).

Pulmonary Function Tests and Diagnostic Outcomes

The comparison in Table 5 ("Asthma, present or possible" versus "Not Asthma") leads to outcomes more consistent with the high prevalence of wheezing among children in the middle-distance category. In addition, the two children with diisocyanate-specific antibodies resided in this residential distance category during the study period.

Diagnostic outcomes were based on one clinical interaction and one set of pulmonary function tests. Predicted values for selected PFT parameters were based on the child's height, age, and sex. An actual test result could be (a) equal to 100% of the predicted value, (b) more than 100% of the predicted value, or (c) less than 100% of the predicted value. If a participant's actual test results for FEV1 or FEV1/FVC are less than 80% of the predicted values, the results are typically considered "abnormal." When the results from a number of participants are grouped by diagnostic outcome, the prevalence of actual test results (normal and abnormal) that are less than 100% of predicted values can be compared (Table 6). For example, the prevalence of actual test results that were less-than-predicted values was consistently higher among children categorized as having asthma with abnormal PFTs versus participants who did not have asthma (88.2% versus 41.2% for FEV1, 94.1% versus 23.5% for FEV1/FVC, and 100% versus 23.5% for FEF^25-75).

The distinction between participants categorized as having asthma without abnormal PFTs and those with possible asthma is harder to define. Both groups have respiratory symptoms and normal (or, at least, not abnormal) test results. The prevalence of actual test results that were less than 100% of predicted values is similar for these two groups (Table 6). At least for this characteristic (i.e., the prevalence of actual test results that were less than predicted values), the distinction between these two diagnostic categories does not appear to be meaningful. The spirometry results in Table 6 suggest that a natural break occurs between results for children in the asthma-related categories (Outcomes 1, 2, and 3) and children who do not have asthma (Outcome 4).

One way to summarize this study's findings is that asthma was present or possible in 38 of the 55 children evaluated. Eighteen of the 55 participants had a previous diagnosis of asthma, but only seven of those diagnoses had been made after the child had reached age 5 years or more.

Seventeen of the 55 children evaluated had abnormal PFTs on the day they were tested. During the past 12 months, only 1 of these 17 children had used a medication that reduces airway inflammation. This finding suggests that children with asthma in the Glenola community may not have been diagnosed and treated according to national consensus guidelines (9).

Regular exposure to tobacco smoke in the child's home "at some time during the study period" was common (58%) among participants who completed the in-person interview (n=55). The numbers are too small to make generalizations by either diagnostic or residential distance categories. Also, the contribution of passive smoke exposure to asthma prevalence is unknown for both symptomatic participants (n=118) and asymptomatic participants (n= 86).

There were 118 symptomatic children identified during the screening phase of the study (N=204). Asthma was diagnosed in 28 of the 55 symptomatic children who participated in the diagnostic evaluations, and was considered possible in another 10 participants. The prevalence of asthma among the 55 symptomatic children clinically evaluated cannot be assumed to be representative of all participants (or even of all symptomatic participants), given the low participation rate in the study's clinical evaluations. Still, asthma prevalence can be estimated by making certain explicit assumptions. Consider the following three scenarios:

Scenario 1: Assume that the prevalence of asthma among symptomatic children who were not clinically evaluated is the same as the prevalence found among symptomatic children who were clinically evaluated. Overall, asthma would be present or possible for 82 of 204 participants (40%). This figure could represent an estimated upper boundary for asthma prevalence.

Scenario 2: Assume that the prevalence of asthma among symptomatic children who were not clinically evaluated is half of the prevalence found among symptomatic children who were clinically evaluated. Overall, asthma would be present or possible for 60 of 204 participants (29%). This figure could represent the best estimate of asthma prevalence.

Scenario 3(a): Assume that the prevalence of asthma among symptomatic children who were not clinically evaluated is zero (that is, all symptomatic children who had asthma were clinically evaluated and diagnosed). Overall, asthma would be present or possible for 38 of 204 participants (19%). This figure could represent an estimated lower boundary for asthma prevalence.

Scenario 3(b): Or, assume Scenario 3 (a) is true and that none of the children categorized as "asthma is possible" actually has asthma. Asthma would still be present (diagnosed) in 28 of 204 participants (14%).

The true prevalence of asthma among children in the Glenola community is unknown, but it is higher than the prevalence (6.2%) reported for children (under 18 years of age) in the 1996 National Health Interview Survey (19), and it is higher than the prevalences (<10%) found in other studies similar to the one reported here (15). Indeed, the prevalence of asthma in this rural community could be similar to high-risk, inner-city environments, where the prevalence of asthma (diagnosed and undiagnosed) often exceeds 25%.

CONCLUSIONS

The estimated prevalence of asthma among school-aged children living near the Trinity American facility is higher than expected. Even with conservative assumptions, the prevalence of asthma is unlikely to be less than 15% to 20%.

2. The prevalence of respiratory symptoms reported during telephone screening was high among school-aged children living near the facility. The significance of this finding was confirmed by two asthma specialists, who categorized asthma as present or possible for a majority of the symptomatic children they evaluated in this study.

3. Environmental air samples and biomarker results from participants in the ATSDR biological exposure investigation, the Duke case series, and the present study are consistent with community exposure to diisocyanates.

4. The factors that lead to the apparent increase in respiratory health effects are not known. However, it is possible that poorly controlled releases during the production of polyurethane foam contributed to the respiratory problems identified in this community.

RECOMMENDATIONS

1. Community health education is needed to enable parents to identify common symptoms that suggest asthma, seek medical evaluation when appropriate, and take steps to improve every child's environment (e.g., not smoking in the car or at home).

2. Continuing medical education is needed to enable area physicians to discuss the evidence of exposure and adverse pulmonary health effects with families, and diagnose, treat, follow, and refer asthma patients according to consensus guidelines.

3. Some individuals in this community may have been sensitized to diisocyanates; therefore, diisocyanates should not be released into this community's air during the foreseeable future.

4. The airborne emissions from other foam-producing plants adjacent to residential communities should be monitored to ensure that residents are not being exposed. If diisocyanates or other harmful substances are identified off-site, consideration should be given to testing residents for evidence of exposure and adverse health effects.

ACKNOWLEDGMENTS

ATSDR Collaborators

The principal investigator was Dan Middleton, with support from Mary White, Chief of the Health Investigations Branch, Division of Health Studies, ATSDR. Roberta Hilsdon provided assistance during initial planning and data collection. Ravishankar Rao plotted current addresses, and Bill Henriques plotted residential histories using GIS. Carolyn Harris and Judith Smith provided financial oversight and monitored the contractors' performance and progress.

The original team of ATSDR employees to investigate releases from the Trinity American facility provided the environmental groundwork for this health study. Theresa Kilgus was team leader, with support from John Abraham, Chief of the Exposure Investigations and Consultations Branch, Division of Health Assessment and Consultation. Lynn Wilder (and others) collected many environmental samples, and Greg Zarus developed exposure models. Dahna Batts-Osborne developed a biomarker protocol, and Ken Orloff performed the ATSDR biomarker study.

Susan Metcalf served as in-house pediatric consultant. Steven Kinsler helped to explain diisocyanate toxicology. Cate McKinney, Maria Terán-MacIver, and Dan Holcombe facilitated the agency's community involvement activities. Steve Blackwell planned educational activities for health professionals.

Other Collaborators

Randolph County Health Department

MiMi Cooper, director of the Randolph County Department of Health, provided advice and counsel throughout ATSDR's activities in the Glenola area. Ms. Cooper also participated in numerous public meetings for area residents and graciously allowed the use of the department's clinic for diagnostic evaluations.

Duke University Medical Center

Physicians Larry Williams and Laurie Anne Myers of Duke University commuted from the Durham area to provide clinical services, as did spirometry technicians Virginia Labelle and Debra Sedlak.

National Opinion Research Center

Participant interviews were conducted by the National Opinion Research Center (NORC). Several NORC employees made important contributions, including Missy Koppelman, Ann Cederlund, Bronwyn Nichols, Toby Singer, Heather Ferguson, Tina Dennis, and the telephone interviewers. Cheryl Gilbert, Barbara Watt, and Ken Miller assisted at the clinical evaluation site.

Midwest Research Institute

The clinical evaluations and laboratory testing were provided by the Midwest Research Institute (MRI), through a contractual agreement with ATSDR. Sarah Hatch planned laboratory services and enlisted specialty physicians, spirometry technicians, and phlebotomists. Lee Patterson was at the clinic early and late to accommodate participants' schedules and to address other matters important to the success of this study.

REFERENCES

• 1. Agency for Toxic Substances and Disease Registry. Public Health Advisory for Trinity American Corporation. Atlanta: U.S. Department of Health and Human Services; 1997.
• 2. Banks DF, Rando RJ, Barkman HW. Persistence of toluene diisocyanate- induced asthma despite negligible workplace exposures. Chest 1990;97:121-5.
• 3. Darcey D and Lipscom H. Report of clinical evaluations, Archdale/Glenola area residents, Randolph County, NC: Report from Division of Occupational and Environmental Medicine, Department of Community and Family Medicine, Duke University Medical Center. Durham (NC): Duke University; 1998.
• 4. Agency for Toxic Substances and Disease Registry. Health Consultation for the Trinity American Corporation. Atlanta: U.S. Department of Health and Human Services; 1997.
• 5. Agency for Toxic Substances and Disease Registry. Exposure Investigation for the Trinity American Corporation. Atlanta: U.S. Department of Health and Human Services; 1998.
• 6. Orloff KG, Batts-Osborne D, Kilgus T, Metcalf S, Cooper M. Antibodies to toluene diisocyanate in an environmentally exposed population. Environ Health Perspect 1998, 106(10):665-6.
• 7. Second International Workshop on Monitoring Trends and Determinants of Asthma and Allergies. International study of asthma and allergies in childhood (ISAAC) manual. Bochum (Germany); 1991.
• 8 .Centers for Disease Control and Prevention. National Health and Nutrition Examination Survey (NHANES III). Hyattsville (MD): U.S. Department of Health and Human Services.
• 9. National Heart, Lung, and Blood Institute (NHLBI). Expert Panel Report 2: Guidelines for the Diagnosis and Management of Asthma. Bethesda (MD): National Institutes of Health; 1997. (NIH) Publication No. 97-4051.
• 10. Sole D, Komatsu MK, Carvalho KV, Naspitz CK.. Pulse oximetry in the evaluation of the severity of acute asthma and/or wheezing in children. J Asthma 1999;36(4):327-33.
• 11. American Thoracic Society (ATS). Standardization of spirometry: 1994 update. Am J Respir and Crit Care Med 1995;152:1107-36.
• 12. Centers for Disease Control and Prevention. Epi Info computer program. Version 6.04b. Atlanta: U.S. Department of Health and Human Services; 1997.
• 13. SAS Procedures Guide, Version 6, Third Edition. (NC): SAS Institute; 1995:1-703.
• 14. Centers for Disease Control and Prevention. State-specific prevalence of cigarette smoking among adults, and children's and adolescents' exposure to environmental tobacco smoke- United States, 1996. MMWR 1997; 46(44):1038-47.
• 15. Sly RM. Changing prevalence of allergic rhinitis and asthma. Ann Allergy Asthma Immunol 1999;82:233-52.
• 16. Selden A, Lelin L, Wass U. Isocyanate exosure and hypersensitivity pneumonitis-report of a probable case and prevalence of specific immunoglobulin G antibodies among exposed individuals. Scand J Work Environ Health 1989;15:234-7.
• 17. Keskinen H, Tupasela O, Tikkainen U, Nordman H. Experiences of specific IgE in asthma due to diisocyanates. Clin Allergy 1988;18:597-604.
• 18. Bauer X, Marek W, Ammon J, Czuppon A, Marczynski B, Rault-Heimsoth M, Roemmelt H, Fruhmann G. Respiratory and other hazards of isocyanates. Int Arch Occup Environ Health 1994;66:141-52.
• 19. National Institutes of Health. National Health Interview Survey (NHIS). Current estimates from the National Health Interview Survey, 1996. Vital and Health Statistics, Series 10(200). Bethesda (MD): National Institutes of Health; October 1999.

TABLES

Table 1.--Demographic and study characteristics, by participant selection group

^* Participants are grouped according to the manner in which they were identified as study participants; i.e.,

(a) children identified on the original list provided by the school board, and

(b) children residing in a study household who were added to the original list by parents/guardians.

^† The "Combined" group included children from the original list, plus siblings in study households identified by parents or guardians and added to the study.

^‡Exposure potential was categorized as (a) "Yes" (204 children), or (b) "No" (17 children) on the basis of confirmation that the child resided within 1 mile of the emission source during the study time period, or (c) "Unknown" (10 children), if address information could not be plotted with GIS.

Table 2.--Respiratory history of study participants reported during the screening interviews, by residential distance category

^*Residential distance categories. The distances between participants' residences during the study exposure period and the point source of airborne emissions were grouped into three concentric areas defined by radiuses of 0.50 miles ("near"), 0.75 miles ("center"), and 1.0 miles ("far").

^† A child became a study participant after a parent or guardian completed the screening interview and researchers confirmed that the residential information collected met the study's eligibility criteria.

^‡The interviewer coded responses during the interview as (a) "Yes" (b) "No," (c) "Don't know," or (d) "Refused."

Comments that clarified responses were recorded. See Appendix A for screening interview questionnaire.

^§In the past 12 months.

^¶Patient was previously diagnosed with asthma after reaching the age of 5 years or more.

^**Participants were classified as "Symptomatic" on the basis of symptoms reported during the screening interview

or a previous diagnosis of asthma at the age of 5 years or more.

Table 3.--Respiratory history of study participants reported during the in-person interviews and diagnostic outcomes, by residential distance category

^*The interviewer coded responses during the interview as (a) "yes," (b) " no," (c) "don't know," or (d) "refused." Comments that clarified responses were recorded. See Appendix B for interview questions.

^† Residential distance categories: The distances between participants' residences during the study exposure period and the point source of airborne emissions were grouped into three concentric areas defined by radiuses of 0.50 miles ("near"), 0.75 miles ("center"), and 1.0 miles ("far").

^‡Symptoms of asthma observed in the past 12 months.

^§Patients previously diagnosed with asthma, characterized as "ever" and "after reaching the age of 5 years or more."

^¶The diagnostic outcome categories are "Clinical Asthma" (with or without abnormal PFTs), "Possible Asthma" (further evaluation needed), and "Not Asthma."

Table 4.--Improvement in abnormal pulmonary function test results after administering albuterol

^*FEV1 is the forced expiratory volume (liters) in 1 second.

^† FEF^25-75 is the flow rate in liters per second (l/s) during the middle of forced expiration.

^‡For the 17 subjects whose PFTs were interpreted as "abnormal" by the asthma specialists, the average increase in FEV1 after administering a bronchodilator was 8.8% (range 3.1% through 15.9%).

^§For the 17 subjects whose PFTs were interpreted as "abnormal" by the asthma specialists, the average increase in FEF^25-75 after administering a bronchodilator was 31.9% (range 11.3% through 70.6%).

Table 5--Diagnostic outcomes and pulmonary function test results compared to predicted values, by residential distance category

^*Diagnostic outcomes are combined to form (a) "Asthma, present or possible" or (b) "Not Asthma" (see Appendix D).

^† Pulmonary function test results are expressed as the number (percentage) of test results that were below the predicted value.

^‡Residential distance categories: The distances between participants' residences during the study exposure period and the point source of airborne emissions were grouped into three concentric areas defined by radiuses of 0.50 miles ("near"), 0.75 miles ("center"), and 1.0 miles ("far").

^§"Possible Asthma" means that a diagnosis can not be made or ruled out without additional evaluation.

^¶FEV1 is forced expiratory volume (liters) in 1 second.

**FVC is forced vital capacity (liters). FEV1/FVC is the ratio of FEV1 to FVC.

^†† FEF^25-75 is the flow rate (liters) per second during the middle of an individual's forced expiration of air from the lungs.

Table 6--Pulmonary function test results compared to predicted values, by diagnostic outcome

^* The result of each pulmonary function test (PFT) was compared to the predicted value for each individual participant; the results are expressed, by diagnostic outcome, as the number (percentage) of results below the predicted value.

^† A completed evaluation included (minimally) the screening interview, the in-person interview, medical history, physical examination, and pulmonary function testing.

^‡"Abnormal PFT" refers to the physician's interpretation of a child's spirometry results.

^§"Possible Asthma" means that a diagnosis can not be made or ruled out without additional evaluation.

^¶FEV1 is forced expiratory volume (liters) in 1 second.

^**FVC is forced vital capacity (liters). FEV1/FVC is the ratio of FEV1 to FVC.

^†† FEF^25-75 is the flow rate (liters/second) during the middle of forced expiration of air from the lungs.

FIGURES

Figure1 -- Flowchart
GIF version
PDF version

APPENDICES Click Here

Top of Page