## Abstract

The authors investigated changes in respiratory health after September 11, 2001 (“9/11”) among residents of the area near the World Trade Center (WTC) site in New York City as compared with residents of a control area. In 2002, self-administered questionnaires requesting information on the presence and persistence of respiratory symptoms, unplanned medical visits, and medication use were sent to 9,200 households (22.3% responded) within 1.5 km of the WTC site (affected area) and approximately 1,000 residences (23.3% responded) in Upper Manhattan, more than 9 km from the site (control area). Residents of the affected area reported higher rates of new-onset upper respiratory symptoms after 9/11 (cumulative incidence ratio = 2.22, 95% confidence interval (CI): 1.88, 2.63). Most of these symptoms persisted 1 year after 9/11 in the affected area. Previously healthy residents of the affected area had more respiratory-related unplanned medical visits (prevalence ratio = 1.73, 95% CI: 1.13, 2.64) and more new medication use (prevalence ratio = 2.89, 95% CI: 1.75, 4.76) after 9/11. Greater impacts on respiratory functional limitations were also found in the affected area. Although bias may have contributed to these increases, other analyses of WTC-related pollutants support their biologic plausibility. Further analyses are needed to examine whether these increases were related to environmental exposures and to monitor long-term health effects.

The destruction of the World Trade Center (WTC) on September 11, 2001 (“9/11”) resulted in the release of large amounts of pollutants into the surrounding areas. These pollutants included particulate matter, asbestos, metals, and organic compounds (1). Initially, smoke and debris came from the fires that erupted when the planes struck the buildings, followed by the release of airborne dust and debris from the collapse of the buildings. Subsequently, the fires that burned at the site for several months and clean-up activities released particulates. In many cases, dust from the collapse infiltrated homes and apartments at a depth of up to several inches (2). As people returned to their residences, settled dust was redispersed into the air.

One of the immediate public health concerns was the effect of this intensive exposure to these air pollutants on the health of local residents. Analyses of settled dust samples collected 5–6 days after the disaster indicated that 1–4 percent by weight were particles that can be inhaled deeply into the lungs (3) and are associated with respiratory diseases. Adverse health effects could have arisen from both acute high-level exposures and prolonged low-level exposures. Moreover, it is unknown whether the adverse respiratory effects, if they existed, were transient or persistent. Studies of asthmatic Lower Manhattan residents found worse symptoms and increases in medical care utilization and asthma medication prescriptions after 9/11 (4, 5). To our knowledge, these studies of persons with asthma are the only published studies of the respiratory health of residents near the site of the former WTC (“Ground Zero”). The pollution from the WTC disaster may have also caused new disease among previously healthy residents of New York City. Additionally, local residents complained about upper respiratory and other symptoms consistent with exposure to irritants. Since there are large residential communities around Ground Zero, the potential for respiratory health effects from exposure to these agents deserves investigation.

The goals of the present study included 1) determining whether there was an increase in the incidence of new-onset and persistent upper and lower respiratory symptoms in residents living near Ground Zero as compared with residents of a control area and 2) investigating whether there was an increase in symptom exacerbation among asthmatic residents living near Ground Zero as compared with a control area. Additionally, subgroups of residents with new-onset persistent symptoms and asymptomatic persons were identified and followed for assessment of chronic respiratory health effects, including symptom persistence and physiologic abnormalities as measured by spirometry. In this paper, we discuss the results pertaining to upper respiratory symptoms.

## MATERIALS AND METHODS

### Study design and study population

This retrospective cohort study was started 8 months after 9/11 (May 2002). The cumulative incidence of reported new-onset and persistent new-onset upper respiratory tract symptoms and unplanned medical visits/medication use among residents living near Ground Zero (the affected area) was compared with the incidence in a control population. Potential study buildings were selected so as to include major population areas below Canal Street in Lower Manhattan. These buildings were then stratified by housing characteristics (e.g., low- or high-income rentals, cooperatives or condominiums, public housing complexes) and selected to include a range of these characteristics. Finally, 2000 US Census data were examined to ensure that the final study buildings were representative of the range of socioeconomic characteristics present in the underlying population of Lower Manhattan.

The area defined as the affected area is located within 1.5 km of the former WTC site and includes 49 buildings in Lower Manhattan with approximately 9,200 households. A control area was used for comparison, because the health histories of residents living near the WTC prior to 9/11 were not available and respiratory diseases usually have a strong seasonal component. The prevailing wind direction was considered in selecting the control area. Therefore, areas south, east, and west of the WTC that were impacted by the plume, including Brooklyn, New Jersey, and Staten Island, were excluded from the control area. Efforts were made to identify control buildings in census blocks with similar characteristics as the affected area. The control area consisted of approximately 1,000 households in five Upper Manhattan apartment buildings more than 9 km from the WTC site. To obtain a large, representative sample in the affected area, we oversampled the population of the affected area at a 9:1 ratio (affected area:control area). Figure 1 shows the study areas and the prevailing wind directions at John F. Kennedy International Airport for September 11–30, 2001.

FIGURE 1.

Locations of study buildings in the affected and control areas, World Trade Center Health Survey, New York City, 2002. (1 mile = 1.61 km).

FIGURE 1.

Locations of study buildings in the affected and control areas, World Trade Center Health Survey, New York City, 2002. (1 mile = 1.61 km).

All residents of the identified affected and control buildings were eligible to participate in the study. Up to four residents in each household were asked to complete the questionnaire. To eliminate confounding effects due to residential mobility and to minimize potential misclassification due to occupational exposure, we excluded persons meeting any of the following criteria: 1) born after 9/11; 2) temporarily moved out of the residence after 9/11 and returned on or after January 1, 2002; 3) did not reside at the current address on 9/11; or 4) lived in the control area but worked in the affected area.

### Study materials and procedures

A study packet containing a cover letter, consent forms, questionnaires, and a stamped, addressed envelope was mailed to all apartments in the study buildings. The study packets also contained information about Project Liberty (a federally sponsored program providing free crisis counseling services to those affected by the WTC disaster) for persons who experienced anxiety while reviewing the materials. To accommodate the needs of the large populations of non-English speakers in the study areas, translations of the study materials were available for Spanish- and Chinese-speaking residents.

The questionnaires included a household questionnaire and four individual questionnaires. An adult resident was asked to complete the household questionnaire by providing information on the age, gender, and asthma status of all household members. The household questionnaire also asked about the condition of the apartment immediately after 9/11, the duration and frequency of odors or dust in the apartment, and any cleaning, sampling, or inspections that were performed.

In households with more than four persons, two adult residents and the two oldest residents under age 18 years were asked to complete the individual questionnaire. For children younger than age 12, a parent or legal guardian completed the questionnaire. The individual questionnaire was primarily designed to estimate the prevalence and incidence of asthma and respiratory symptoms. This questionnaire was derived from the International Union Against Tuberculosis and Lung Diseases questionnaire and the International Study of Asthma and Allergies in Childhood questionnaire, both of which have been validated and used in epidemiologic studies to detect symptoms associated with asthma and bronchial hyperresponsiveness (69).

For questions about upper and lower respiratory symptoms and irritation symptoms, the resident was asked whether the problem had occurred in the past 12 months, whether it started after 9/11, whether it worsened after 9/11, and, depending on the type of symptom, either the average frequency or the perceived severity of the symptom during the past 4 weeks. Additional questions assessed unplanned medical visits (outpatient visits, emergency department visits, and hospitalizations), physician diagnoses of asthma and other respiratory disorders, use of asthma medication, and respiratory functional status. The questionnaire also included questions related to sociodemographic factors, smoking history, temporary residence changes after 9/11, and employment location. After receipt of a completed questionnaire, a Metro Card with a value of $6 was mailed to the participant to acknowledge participation. The study packets were initially distributed 1 year after 9/11 (±4 months) via bulk mail. Because of inconsistencies in the handling of this material, additional deliveries were made to each residence by hand. Where access could not be gained to make hand deliveries, the packets were left in building lobbies. Finally, in addition to the bulk mailing, a first-class mailing of the packets was made to all households, followed by a reminder postcard. After distribution of the packets, field-workers spent time in the buildings to encourage participation, provide additional copies of the study materials, and answer questions. The days and times of these outreach activities were varied to maximize the numbers and types of persons encountered. Posters advertising the study were placed in and around the buildings. Further publicity about the study was generated through notices in local newspapers and building newsletters, as well as by staff in attendance at meetings of community boards and tenant organizations and local health fairs. To estimate potential selection bias, we selected one building in the affected area (440 apartments) and two buildings in the control area (240 apartments) to receive additional outreach. These intensive outreach activities included additional mailings, advertisements, and time spent in the buildings by field-workers. ### Outcome definitions Health outcomes were defined on the basis of reported respiratory symptoms, unplanned medical visits, physician diagnoses, medication use, respiratory functional limitation, and the time period in which symptoms occurred. “New-onset” symptoms were defined as upper respiratory symptoms that began after 9/11. A “persistent new-onset” symptom was a new-onset symptom that had bothered the respondent “some” or “a lot” in the 4 weeks prior to completing the survey. ### Statistical analysis Because of variations in the number of persons residing in each apartment and the lack of information about the number of persons in nonresponding households, the response rate was calculated using the number of responding households as a numerator. Packets that were returned marked “vacant” were omitted from this calculation. The demographic characteristics of participants in the affected and control areas were compared using the χ2 test. For new-onset respiratory health outcomes, we computed cumulative incidence by dividing the number of participants with a new-onset outcome after 9/11 by the total number of participants. However, the denominators for unplanned medical visits, new diagnoses of asthma, and medication use were based on the subgroup of participants who were “previously healthy” (i.e., free of a physician's diagnosis of asthma, emphysema, chronic obstructive pulmonary disease, and chronic bronchitis prior to 9/11). Cumulative incidence ratios (CIRs) comparing the affected and control areas were computed, and 95 percent confidence intervals were used to estimate the precision of the CIRs. For respiratory functional status (before and after 9/11), prevalence rates (the number of persons in a disease status category divided by the total number of participants in each area) and prevalence ratios and their 95 percent confidence intervals were computed. Finally, the χ2 test was used to compare data for the self-described breathing statements. Unconditional logistic regression analysis was used to compute adjusted odds ratios while controlling for potential confounders, including age, gender, education, race, and smoking. Education was used as a surrogate for socioeconomic status, because information about education was more complete (11 percent missing data) than information for income (25.3 percent missing data), and education and income were highly correlated. Because respiratory diseases are not rare events, adjusted odds ratios from logistic regression tend to consistently overestimate the CIRs. Therefore, the crude CIRs and 95 percent confidence intervals are presented in the tables, and adjusted results were used only to determine whether the results were still statistically significant after controlling for confounders. ## RESULTS A total of 9,168 survey packets were sent to residences in the affected area, and 962 packets were sent to residences in the control area. After exclusion of apartments that were definitely vacant, household response rates were 22.3 percent in the affected area and 23.3 percent in the control area. Among the buildings targeted for increased outreach, the response rates were 43.8 percent and 40.3 percent in the affected and control areas, respectively. A total of 553 respondents (17.3 percent) were excluded from the analysis on the basis of the four criteria described in Materials and Methods, giving us a total of 2,362 participants in the affected area and 291 in the control area. Although we attempted to make the residents of affected and control areas demographically comparable, differences remained. The affected area had distributions of age and household income that were significantly different from those of the comparison area (table 1). In paired χ2 tests, there were significantly higher proportions of Hispanics, Asians, and residents with less education (i.e., not high school graduates) but lower proportions of African Americans and Caucasians from the affected area as compared with the control area. In general, these demographic differences are similar to differences in the underlying populations according to 2000 US Census data. Since low socioeconomic status is associated with asthma, these variables were considered potential confounders and were controlled for in the multivariate analyses. TABLE 1. Demographic characteristics (%) of residents of the affected area (n = 2,362) and the control area (n = 291), World Trade Center Health Survey, New York City, 2002 Characteristic Affected area Control area p value (χ2 test) Gender 0.35 Male 38.0 41.0 Female 62.0 59.0 Age (years) <0.0001 0–34 23.4 23.8 35–64 51.0 35.3 ≥65 25.6 40.9 Annual household income <0.0001 <$24,999 34.7 19.9
$25,000–$49,999 18.8 19.9
$50,000–$99,999 23.6 30.5
≥$100,000 22.9 29.6 Race/ethnicity* Hispanic 14.1 7.6 0.003 Asian 16.3 3.3 <0.0001 African-American 8.4 11.6 0.08 White 61.0 79.4 <0.0001 Other 4.8 4.7 0.95 Education 0.0002 Not a high school graduate 20.3 11.0 Characteristic Affected area Control area p value (χ2 test) Gender 0.35 Male 38.0 41.0 Female 62.0 59.0 Age (years) <0.0001 0–34 23.4 23.8 35–64 51.0 35.3 ≥65 25.6 40.9 Annual household income <0.0001 <$24,999 34.7 19.9
$25,000–$49,999 18.8 19.9
$50,000–$99,999 23.6 30.5
≥\$100,000 22.9 29.6
Race/ethnicity*
Hispanic 14.1 7.6 0.003
Asian 16.3 3.3 <0.0001
African-American 8.4 11.6 0.08
White 61.0 79.4 <0.0001
Other 4.8 4.7 0.95
Education   0.0002

20.3

11.0

*

Race/ethnicity groups were not mutually exclusive; therefore, percentages do not add up to 100%. For calculation of p values, instead of an overall χ2 test, paired χ2 tests were performed for each race/ethnicity group versus the other groups combined.

Table 2 describes the relation between residence in the affected area and upper respiratory and irritation symptoms. Rates of all new-onset symptoms were significantly higher in the affected area after we controlled for potential confounders (CIRs were 3.00–4.23). Residents of the affected area reported a significantly higher rate of at least one of these new-onset symptoms (71.8 percent) than did controls (32.3 percent), an increase of 121 percent (CIR = 2.22, 95 percent confidence interval (CI): 1.88, 2.63). Furthermore, the persistence of these new-onset symptoms was significantly higher in the affected area. CIRs ranged from 2.49 to 5.03, with the highest CIRs being for congestion symptoms and recurring headaches. Affected-area residents also reported a significantly higher incidence of at least one persistent new-onset symptom (43.7 percent) than did controls (14.4 percent)—a 200 percent elevation (CIR = 3.02, 95 percent CI: 2.28, 4.02).

TABLE 2.

Incidence of upper respiratory symptoms after September 1, 2001, and associated cumulative incidence ratios among residents of the affected area (n = 2,362) versus the control area (n = 291), World Trade Center Health Survey, New York City, 2002

Symptom(s)

New-onset

New-onset persistent

No.

%

CIR*

95% CI*

No.

%

CIR

95% CI

Eye irritation or burning 1,143 52.9 3.22 2.45, 4.25 501 21.2 3.43 2.18, 5.40
Nose irritation or burning 896 41.3 4.23 2.92, 6.11 432 18.3 3.80 2.26, 6.38
Nasal congestion 864 40.7 3.12 2.26, 4.31 486 20.6 4.28 2.55, 7.17
Hoarse throat or other throat irritation 1,142 53.1 3.41 2.56, 4.55 543 23.0 3.34 2.18, 5.14
Sinus congestion 719 34.2 3.11 2.18, 4.44 425 18.0 4.76 2.65, 8.55
Nose bleeding 310 13.7 3.47 1.92, 6.24 101 4.3 2.49 1.02, 6.06
Recurring headaches 720 33.3 3.00 2.12, 4.26 449 19.0 5.03 2.80, 9.03
One or more of the above symptoms

1,696

71.8

2.22

1.88, 2.63

1,031

43.7

3.02

2.28, 4.02

Symptom(s)

New-onset

New-onset persistent

No.

%

CIR*

95% CI*

No.

%

CIR

95% CI

Eye irritation or burning 1,143 52.9 3.22 2.45, 4.25 501 21.2 3.43 2.18, 5.40
Nose irritation or burning 896 41.3 4.23 2.92, 6.11 432 18.3 3.80 2.26, 6.38
Nasal congestion 864 40.7 3.12 2.26, 4.31 486 20.6 4.28 2.55, 7.17
Hoarse throat or other throat irritation 1,142 53.1 3.41 2.56, 4.55 543 23.0 3.34 2.18, 5.14
Sinus congestion 719 34.2 3.11 2.18, 4.44 425 18.0 4.76 2.65, 8.55
Nose bleeding 310 13.7 3.47 1.92, 6.24 101 4.3 2.49 1.02, 6.06
Recurring headaches 720 33.3 3.00 2.12, 4.26 449 19.0 5.03 2.80, 9.03
One or more of the above symptoms

1,696

71.8

2.22

1.88, 2.63

1,031

43.7

3.02

2.28, 4.02

*

CIR, cumulative incidence ratio; CI, confidence interval.

The effect was still statistically significant (p < 0.05) after adjustment for age, gender, education, race, and smoking.

Data on medical visits and medication use after 9/11 among previously healthy participants are presented in table 3. The incidence of unplanned medical visits for respiratory problems was significantly increased in the affected area (14.5 percent) over the control area (8.4 percent; CIR = 1.73, 95 percent CI: 1.13, 2.64) after we controlled for potential confounders. A significantly higher proportion of affected-area residents started using respiratory medication after 9/11 (18.0 percent) in comparison with controls (6.2 percent) (CIR = 2.89, 95 percent CI: 1.75, 4.76). By examining medication use in the past 4 weeks as an indication of disease persistence, we found that affected-area residents reported significantly higher rates (15.1 percent) than controls (6.2 percent) (prevalence ratio = 2.44, 95 percent CI: 1.48, 4.02). In particular, use of fast-relief (9.7 percent) and controller (10.4 percent) asthma medications was significantly higher in the affected area. The rates of new diagnosis of asthma and use of more medication after 9/11 were not statistically significantly different between two areas.

TABLE 3.

Incidence of medical consultations, asthma diagnoses, and use of respiratory medication after September 11, 2001, among previously healthy* residents of the affected area (n = 2,362) versus the control area (n = 291), World Trade Center Health Survey, New York City, 2002

Affected area

Control area

Cumulative incidence ratio

95% confidence interval

No.

%

No.

%

Unplanned medical visit(s) for respiratory problems in past 12 months 286 14.5 21 8.4 1.73 1.13, 2.64
Physician diagnosis of asthma after September 11 101 18.0 12.0 1.50 0.69, 3.24
Started using respiratory medication after September 11 340 18.0 15 6.2 2.89 1.75, 4.76
Used more respiratory medication after September 11 62 3.3 1.3 2.64 0.83, 8.34
Used respiratory medication in past 4 weeks 285 15.1 15 6.2 2.44 1.48, 4.02
Used fast-relief asthma medicine in past 4 weeks 178 9.7 3.8 2.55 1.32, 4.91
Used controller§ asthma medication in past 4 weeks

191

10.4

10

4.3

2.44

1.31, 4.55

Affected area

Control area

Cumulative incidence ratio

95% confidence interval

No.

%

No.

%

Unplanned medical visit(s) for respiratory problems in past 12 months 286 14.5 21 8.4 1.73 1.13, 2.64
Physician diagnosis of asthma after September 11 101 18.0 12.0 1.50 0.69, 3.24
Started using respiratory medication after September 11 340 18.0 15 6.2 2.89 1.75, 4.76
Used more respiratory medication after September 11 62 3.3 1.3 2.64 0.83, 8.34
Used respiratory medication in past 4 weeks 285 15.1 15 6.2 2.44 1.48, 4.02
Used fast-relief asthma medicine in past 4 weeks 178 9.7 3.8 2.55 1.32, 4.91
Used controller§ asthma medication in past 4 weeks

191

10.4

10

4.3

2.44

1.31, 4.55

*

No diagnosis of asthma, chronic obstructive pulmonary disease, chronic bronchitis, or other lung disease before September 11, 2001.

The effect was still statistically significant (p < 0.05) after adjustment for age, gender, education, race, and smoking.

Asthma medication that induces rapid bronchodilation of the airways.

§

Asthma medication that can either prevent inflammation or maintain bronchodilation.

We examined three indicators characterizing different degrees of shortness of breath with exertion before and after 9/11 (table 4). Before 9/11, data for all three indicators were similar in the two areas. However, after 9/11, these indicators pointed to greater increases in the degree of shortness of breath in the affected area (prevalence ratios were 1.51–1.83). When the participants were asked to describe their breathing in the past 4 weeks (data not shown), approximately 16 percent of respondents in the affected area reported that they had “regular trouble with breathing, but it always got completely better” as compared with 10 percent in the control area (p < 0.05). Furthermore, significantly more residents of the affected area reported that their “breathing was never quite right” (21 percent) as compared with residents of the control area (9 percent) (p < 0.05).

TABLE 4.

Self-reported prevalence of shortness of breath with varying levels of exertion before and after September 11, 2001, among residents of the affected area (n = 2,362) versus the control area (n = 291), World Trade Center Health Survey, New York City, 2002

Level of exertion

Affected area

Control area

Prevalence ratio

95% confidence interval

No.

%

No.

%

Shortness of breath when hurrying on level ground or walking up a slight hill
Before September 11 541 26.1 61 24.0 1.09 0.86, 1.37
After September 11 1,075 53.8 71 32.0 1.68 1.38, 2.05*
Shortness of breath when walking with other people of one's own age on level ground
Before September 11 325 15.9 38 14.8 1.08 0.79, 1.47
After September 11 730 38.6 47 21.1 1.83 1.41, 2.38*
Having to stop for breath when walking at one's own pace on level ground
Before September 11 303 14.4 39 15.1 0.95 0.70, 1.30
After September 11

653

32.9

50

21.7

1.51

1.18, 1.95*

Level of exertion

Affected area

Control area

Prevalence ratio

95% confidence interval

No.

%

No.

%

Shortness of breath when hurrying on level ground or walking up a slight hill
Before September 11 541 26.1 61 24.0 1.09 0.86, 1.37
After September 11 1,075 53.8 71 32.0 1.68 1.38, 2.05*
Shortness of breath when walking with other people of one's own age on level ground
Before September 11 325 15.9 38 14.8 1.08 0.79, 1.47
After September 11 730 38.6 47 21.1 1.83 1.41, 2.38*
Having to stop for breath when walking at one's own pace on level ground
Before September 11 303 14.4 39 15.1 0.95 0.70, 1.30
After September 11

653

32.9

50

21.7

1.51

1.18, 1.95*

*

The effect was still statistically significant (p < 0.05) after adjustment for age, gender, education, race, and smoking.

## DISCUSSION

### Respiratory outcomes

In the current study, we found that incidence rates of at least one new-onset upper respiratory symptom and all individual symptoms were significantly increased by 122 percent and over 200 percent, respectively, among affected-area residents. Additionally, almost half of residents in the affected area reported persistence of these symptoms, and the rate of persistent new-onset symptoms was increased by 200 percent. Prezant et al. (10) reported that among firefighters with high or moderate levels of exposure, 54 percent had nasal congestion and 41 percent had nasal drip after 9/11. Eighty-two percent of the firefighters who had high levels of exposure experienced sore throat after 9/11. Banauch et al. (11) reported persistence of symptoms and bronchial hyperreactivity among these firefighters 6 months after 9/11.

Among previously healthy residents in the affected area, we found a 73 percent increase in unplanned medical visits and a 189 percent increase in new medication use for respiratory problems after 9/11. Moreover, there was a 144 percent increase in the use of respiratory medication in the past 4 weeks, including use of fast-relief and controller medicines, among previously healthy affected-area residents. Additionally, the self-described respiratory functional status of affected-area residents was impacted more than that of control-area residents. Shortness of breath with varying levels of exertion was significantly higher in affected-area residents than in the controls. Self-reported descriptions of breathing during the past 4 weeks also indicated significantly higher levels of breathing trouble in the affected area. Szema et al. (5) found that visits to a health clinic for asthma and prescriptions for asthma medication both increased among pediatric asthma patients after 9/11. All of these findings suggest that residents near Ground Zero experienced a significant increase in respiratory diseases related to the WTC disaster and that these symptoms were still persistent in a significant portion of the residents after 1 year.

Our study was one of the earliest of the few studies to estimate the incidence of respiratory disease among residents of Lower Manhattan after 9/11. Both Fagan et al. (4) and Szema et al. (5) studied residents of Lower Manhattan; however, their populations were restricted to persons with asthma. Although residents near Ground Zero were probably not exposed to air pollution levels as high as those of the firefighters involved in WTC rescue, recovery, and clean-up activities, we have shown that residents of the affected area did report significantly more upper respiratory symptoms than residents of the control area.

### Ambient air quality after 9/11

The New York City Department of Health and Mental Hygiene (12) measured the levels and composition of outdoor and indoor surface and airborne dust from November 4 to December 11, 2001, in residential areas near Ground Zero and in a comparison area. That study found a greater percentage of synthetic vitreous fibers, asbestos, quartz, calcite, portlandite, and gypsum in settled dust in Lower Manhattan than in the comparison area. The Environmental Protection Agency collected dust samples at various locations in the immediate vicinity of the WTC site 1–2 days after 9/11 (2). The WTC samples of particulate matter less than 2.5 μm in diameter were alkaline and composed primarily of calcium-based compounds such as calcium sulfate (gypsum) and calcium carbonate (calcite). Gypsum and calcite can irritate the mucus membranes of the eyes, nose, throat, and upper airways (13), and calcium carbonate dust can cause coughing, sneezing, and nasal irritation (14).

Although smoke or debris might have contributed to the increase in adverse respiratory health outcomes in this study, psychological stress might also have played an important role in these effects (15). In the current study, we could not determine whether environmental factors, psychological distress, or both contributed to the increase in respiratory symptoms, since psychological factors were not examined.

### Strengths and limitations

This study is an important first step in identifying the acute and chronic respiratory health impact of the WTC disaster. Of the few studies that have investigated respiratory health among residents of Lower Manhattan after 9/11, it is one of the largest. This study responded to local residents by examining specific symptoms of concern to the community. The design and analysis used in this study allowed for the control for seasonal and socioeconomic confounding effects. In addition, the use of a cohort design allowed for examination of multiple health outcomes.

Although intensive outreach activities were implemented as described in Materials and Methods, we obtained low response rates. This may have been due to the emotional aftermath of the disaster—residents might not have been willing to answer questions that would provoke an emotional reaction. In addition, at the time of this study, the residents of Lower Manhattan were inundated with forms from government agencies and other organizations. The amount of information requested during this time was probably overwhelming. In addition, residents may have thought they had already completed a questionnaire when in fact they had not. New York City also has a history of lower response rates. The 2000 Census only recorded a final response rate of 55 percent in New York City, despite intense advertising and door-to-door follow-up. More importantly, a significant number of residents moved out of the affected area after 9/11. For this reason, if the denominator for calculating the household response rate was overestimated (despite attempts to identify vacant households), the actual response rate would have been underestimated.

The low response rates, although similar between the two study areas, may have introduced selection bias. That is, residents who experienced symptoms, especially those who lived in the affected area, might have been more likely to participate than those who did not. This could have caused the incidence of new-onset symptoms to be overestimated, particularly in the affected area. To minimize this bias, we emphasized the importance of participation for people with and without breathing problems during recruitment activities. In addition, general terms such as “breathing or lung problems” rather than specific terms like “asthma” were used.

To examine possible selection bias due to low response rates, we compared results for the buildings targeted for increased outreach and the nontargeted buildings. Results from the targeted buildings, in which higher response rates were achieved, are assumed to be more accurate and representative. If there was selection bias, we would expect to find a weaker exposure-disease association in the targeted buildings. Instead, we found that the risk estimates for new-onset and new-onset persistent symptoms were consistently higher in the targeted buildings than in the nontargeted buildings (see appendix table 1). These results suggest that any selection bias was in the opposite direction than we would have expected (i.e., the true association may have been underestimated).

Another potential problem with this study is reporting bias. Affected-area participants may have recalled or reported more symptoms than the controls. To prevent such reporting bias, we asked symptom questions not only qualitatively but also quantitatively, by including questions on specific time frames, severity, and frequency, which are less prone to recall bias. To estimate potential reporting bias, we compared rates of self-reported physical disability (which should not have been related to WTC exposures) between the affected area and the control area. The similar rates in the two areas (14.7 percent and 13.1 percent, respectively) indicate no significant reporting bias due to residence area. A participant responding affirmatively about every symptom may have been affected by recall bias (n = 10). Minimal changes were observed when these persons were excluded from the analysis. We believe recall of unplanned medical visits, including emergency department visits and hospitalizations, is more likely to be accurate than recall of symptoms, since such events are more likely to be memorable, and we solicited information on the reason for and exact month and year of the visit. Among respondents reporting a specific respiratory symptom, we compared the proportions who had unplanned medical visits. We found that the proportions were similar in the affected and control areas for most symptoms. If there was overreporting in the affected area, the proportion of persons reporting a specific symptom who also had unplanned medical visits should have been lower in the affected area than in the control area. Therefore, there is no clear evidence of reporting bias on the basis of our limited assessment. In general, reporting bias can be minimized by using objective indicators (e.g., medical records) rather than self-reported information. In this study, it was not feasible to review medical records. Additionally, an analysis of medical records would probably have underestimated or completely missed the less severe symptoms included in our survey.

One final area of concern is the possibility of exposure misclassification. As described above, we excluded persons with evidence of residential mobility and exposures unrelated to their area of residence in order to minimize this bias. However, an unidentified group of affected-area residents may have altered their behavior, spending less time at home in the aftermath of 9/11. Thus, their actual exposure may have been overestimated. On the other hand, it is also possible that control-area residents were impacted by the WTC plume in unforeseen ways.

### Conclusion

This study suggests that residents who lived near Ground Zero on 9/11 reported significantly more upper respiratory and irritation symptoms, unplanned medical visits, and use of respiratory medications and decreased respiratory functional status after 9/11. In a significant portion of the residents, these symptoms persisted 1 year after 9/11. Although we cannot rule out the possibility that selection and reporting bias may have contributed to these increases, chemical analyses of WTC-related pollutants by other researchers support the biologic plausibility of these findings. Further analyses are needed to examine whether increases in reported respiratory disease can be related to differences in exposure and to monitor the potential long-term health effects of the 9/11 disaster in this population.

APPENDIX TABLE 1.

Incidence of new-onset and persistent new-onset upper respiratory symptoms after September 11, 2001, among residents of the affected area versus the control area, by level of outreach (targeted areas and nontargeted areas), World Trade Center Health Survey, New York City, 2002

Symptom(s)

New-onset symptoms

Persistent new-onset symptoms

Targeted area

Nontargeted area

Targeted area

Nontargeted area

CIR*

95% CI*

CIR

95% CI

CIR

95% CI

CIR

95% CI

Eye irritation or burning 4.50 2.72, 7.45 2.76 1.99, 3.85 4.28 2.02, 9.10 3.22 1.81, 5.74
Nose irritation or burning 8.96 4.07, 19.73 3.19 2.11, 4.83 5.68 2.33, 13.82 3.35 1.76, 6.37
Nasal congestion 5.47 2.89, 10.38 2.50 1.72, 3.62 9.24 3.45, 24.69 3.32 1.81, 6.09
Hoarse throat or other throat irritation 4.62 2.81, 7.61 3.06 2.15, 4.36 4.22 2.19, 8.15 3.42 1.92, 6.08
Sinus congestion 6.42 3.09, 13.35 2.40 1.61, 3.59 6.53 2.70, 15.82 4.83 2.19, 10.65
Nose bleeding 10.56 2.60, 42.89 2.41 1.26, 4.59 5.09 1.20, 21.47 2.18 0.69, 6.81
Recurring headaches 5.19 2.73, 9.89 2.48 1.64, 3.75 8.16 3.04, 21.91 4.41 2.12, 9.16
One or more of the above symptoms

3.46

2.49, 4.82

1.83

1.51, 2.21

3.69

2.37, 5.74

3.01

2.07, 4.37

Symptom(s)

New-onset symptoms

Persistent new-onset symptoms

Targeted area

Nontargeted area

Targeted area

Nontargeted area

CIR*

95% CI*

CIR

95% CI

CIR

95% CI

CIR

95% CI

Eye irritation or burning 4.50 2.72, 7.45 2.76 1.99, 3.85 4.28 2.02, 9.10 3.22 1.81, 5.74
Nose irritation or burning 8.96 4.07, 19.73 3.19 2.11, 4.83 5.68 2.33, 13.82 3.35 1.76, 6.37
Nasal congestion 5.47 2.89, 10.38 2.50 1.72, 3.62 9.24 3.45, 24.69 3.32 1.81, 6.09
Hoarse throat or other throat irritation 4.62 2.81, 7.61 3.06 2.15, 4.36 4.22 2.19, 8.15 3.42 1.92, 6.08
Sinus congestion 6.42 3.09, 13.35 2.40 1.61, 3.59 6.53 2.70, 15.82 4.83 2.19, 10.65
Nose bleeding 10.56 2.60, 42.89 2.41 1.26, 4.59 5.09 1.20, 21.47 2.18 0.69, 6.81
Recurring headaches 5.19 2.73, 9.89 2.48 1.64, 3.75 8.16 3.04, 21.91 4.41 2.12, 9.16
One or more of the above symptoms

3.46

2.49, 4.82

1.83

1.51, 2.21

3.69

2.37, 5.74

3.01

2.07, 4.37

*

CIR, cumulative incidence ratio; CI, confidence interval.

Editor's note: An invited commentary on this article appears on page 508, and the authors' response appears on page 511.

This study was supported by Cooperative Agreement U1Q/CCU221059 from the Centers for Disease Control and Prevention.

The authors thank Dr. Thomas Matte for his valuable guidance in the design of the study and Heidi Lee, Marcy Lopez, and Koji Park for their outreach work. They also thank the local community boards, tenants' organizations, and downtown New York City health organizations for their cooperation and Dr. Lester Blair of the American Lung Association for his assistance.

The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention.

Conflict of interest: none declared.

## References

1.
Environmental Protection Agency. EPA response to September 11. Daily environmental monitoring summary—New York (September 2001–July 2002). Washington, DC: Environmental Protection Agency,
2002
. (http://www.epa.gov/wtc/data_summary.htm). (Accessed October 28, 2002).
2.
McGee JK, Chen LC, Cohen MD, et al. Chemical analysis of World Trade Center fine particulate matter for use in toxicologic assessment.
Environ Health Perspect

2003
;
111
:
972
–80.
3.
Lioy PJ, Weisel CP, Millette JR, et al. Characterization of the dust/smoke aerosol that settled east of the World Trade Center (WTC) in Lower Manhattan after the collapse of the WTC 11 September 2001.
Environ Health Perspect

2002
;
110
:
703
–14.
4.
Fagan J, Galea S, Ahern J, et al. Self-reported increase in asthma severity after the September 11 attacks on the World Trade Center—Manhattan, New York, 2001.
MMWR Morb Mortal Wkly Rep

2002
;
51
:
781
–4.
5.
Szema AM, Khedkar M, Maloney PF, et al. Clinical deterioration in pediatric asthmatic patients after September 11, 2001.
J Allergy Clin Immunol

2004
;
113
:
420
–6.
6.
Abramson MJ, Hensley MJ, Saunders NA, et al. Evaluation of a new asthma questionnaire.
J Asthma

1991
;
28
:
129
–39.
7.
Asher MI, Keil U, Anderson HR, et al. International Study of Asthma and Allergies in Childhood (ISAAC): rationale and methods.
Eur Respir J

1995
;
8
:
483
–91.
8.
Burney PG, Laitinen LA, Perdrizet S, et al. Validity and repeatability of the IUATLD (1984) Bronchial Symptoms Questionnaire: an international comparison.
Eur Respir J

1989
;
2
:
940
–5.
9.
Ravault C, Kauffmann F. Validity of the IUATLD (1986) questionnaire in the EGEA study. International Union Against Tuberculosis and Lung Disease. Epidemiological Study on the Genetics and Environment of Asthma, Bronchial Hyperresponsiveness and Atopy.
Int J Tuberc Lung Dis

2001
;
5
:
191
–6.
10.
Prezant DJ, Weiden M, Banauch GI, et al. Cough and bronchial responsiveness in firefighters at the World Trade Center site.
N Engl J Med

2002
;
347
:
806
–15.
11.
Banauch GI, Alleyne D, Sanchez R, et al. Persistent hyperreactivity and reactive airway dysfunction in firefighters at the World Trade Center.
Am J Respir Crit Care Med

2003
;
168
:
54
–62.
12.
Jeffery NL, D'Andrea C, Leighton J, et al. Potential exposure to airborne and settled surface dust in residential areas of Lower Manhattan following the collapse of the World Trade Center—New York City, November 4–December 11, 2001.
MMWR Morb Mortal Wkly Rep

2003
;
52
:
131
–6.
13.
Stellman JM, ed. Encyclopaedia of occupational health and safety. 4th ed. Geneva, Switzerland: International Labour Office,
1998
.
14.
National Library of Medicine. Hazardous Substances Data Bank. Bethesda, MD: National Library of Medicine,
2002
. (http://toxnet.nlm.nih.gov). (Accessed October 28, 2002).
15.
Fagan J, Galea S, Ahern J, et al. Relationship of self-reported asthma severity and urgent health care utilization to psychological sequelae of the September 11, 2001 terrorist attacks on the World Trade Center among New York City area residents.
Psychosom Med

2003
;
65
:
993
–6.