[Important references and tables are omitted from this version -- please
follow the link to the original]
The World Trade Center Residents' Respiratory Health Study: New Onset
Respiratory Symptoms and Pulmonary Function
By Joan Reibman, Shao Lin, Syni-An A. Hwang, Mridu Gulati, James A.
Bowers, Linda Rogers, Kenneth I. Berger, Anne Hoerning, Marta Gomez, and
Edward F. Fitzgerald
Environmental Health Perspectives
Online 20 December 2004
http://ehp.niehs.nih.gov/members/2004/7375/7375.pdf
Acknowledgements
We would like to thank Tom Matte, M.D., for his valuable guidance, and
Heidi Lee, Marcy Lopez, and Koji Park for their outreach work. We are
grateful to the residents in the "exposed" and "control" communities, as
well as to the local community boards, tenant's organizations, community
groups, Dr. Lester Blair and The American Lung Association of the City
of New York.
This publication was supported by Cooperative Agreement Number
U1Q/CCU221059 from the Centers for Disease Control and Prevention (CDC).
Its contents are solely the responsibility of the authors and do not
necessarily represent the official views of CDC.
Funding for this study was provided by the Centers for Disease Control
U1Q/CCU221059, NIH NIEHS P30 ES00260 and NIH NCRR MO1RR00096
Abstract
The destruction of the World Trade Center (WTC) on September 11, 2001 in
New York City resulted in the massive release of pulverized dust and
combustion products. The dust and smoke settled in the surrounding area,
which encompassed a large residential community. We hypothesized that
previously normal residents in the community surrounding the former WTC
would have an increased incidence of persistent respiratory symptoms and
abnormalities in screening spirometry. A hybrid cross-sectional and
retrospective cohort study using a symptombased questionnaire and
on-site screening spirometry in residents in an "exposed area" and in a
"control area" was performed 12+4 months after the collapse. Surveys
were analyzed from 2812 residents. New-onset respiratory symptoms were
described by 55.8% of residents in the "exposed area" compared to 20.1%
in the "control area" after the event. "Persistent new-onset symptoms"
were identified in 26.4% vs. 7.5% of residents in the "exposed area" vs.
"control area" respectively. No differences in screening spirometry
between the groups were detected. A small pilot study suggested the
possibility of an increase in bronchial hyperresponsiveness in "exposed"
participants with persistent symptoms. The data demonstrated an
increased rate of new onset and persistent respiratory health effects in
residents near the former WTC compared to a control population.
Introduction
The destruction of the World Trade Center (WTC) in New York City on
September 11, 2001 resulted in the pulverization of two 107-story
buildings and the massive release of combustion products from jet fuel
and burning structures. An initial cloud of dust and smoke enveloped the
area in all directions. Subsequent wind-blown plumes dispersed dust and
smoke throughout lower Manhattan and Brooklyn. Fires in the 16-acre site
continued for more than three months after the event, with the prolonged
release of combustion products. Analyses of the settled dusts have
revealed cement, glass and particulate matter including gypsum, calcium
carbonate, cement dust, and glass fibers. The dusts were alkaline, with
a pH ranging from 9.3 to 11.5 (Lioy et al. 2002; McGee et al. 2003;
Service 2003). Metals including chromium, iron, magnesium, manganese,
aluminum, barium, titanium and lead were also detected (Lioy et al.
2002). Particles were also noted to contain polycyclic aromatic
hydrocarbons, polychlorinated biphenyls, and organochlorine pesticides
(Lioy et al. 2002; Offenberg et al. 2003).
Although often considered a financial district, lower Manhattan contains
a large residential community with approximately 58,000 residents living
south of Canal Street. The residential communities encompass many
socioeconomic levels and residents of diverse race/ethnicity. Housing
stock consist of large housing complexes containing thousands of
residential units as well as smaller residential buildings. Some
residents in the immediate area surrounding the former WTC (Ground Zero;
GZ) were immediately evacuated, however many remained in their
apartments. Residents who were evacuated returned to their apartments
over the ensuing weeks to months. Dusts from the collapse settled on
streets, playgrounds, cars, and buildings. Dusts entered apartments
through open windows, building cracks and ventilation systems. Removal
of these dusts in individual apartments was accomplished in a variety of
ways; some residents used professional cleaners whereas many performed
the operation themselves. No consistent clean-up operation was offered
to the residential community until one year after the event.
Adverse respiratory health effects from exposures to WTC dusts are being
reported. Firefighters exposed to materials generated during the
collapse of the WTC have developed cough and bronchial
hyperresponsiveness (Banauch et al. 2003; Prezant et al. 2002). A
preliminary telephone survey of a small sample of residents in
Manhattan, also suggested the presence of respiratory health effects
eight weeks after the event (Fagan 2002). To examine whether the
destruction of the WTC resulted in adverse respiratory health effects in
the residential community, we developed a collaborative effort between
the New York State Department of Health (NYSDOH), New York University
(NYU) School of Medicine/Bellevue Hospital and numerous community health
programs and local community organizations. The overall study was
designed to test the hypotheses that the rates and severity of new and
previously existing respiratory diseases increased among residents after
9/11/01 in the community surrounding GZ compared to a control community.
We now report results of the first part of the study, which was designed
to test the hypothesis that the destruction of the WTC resulted in an
increased incidence of persistent new-onset respiratory symptoms and
airflow obstruction in previously normal residents in the surrounding
community. Additional studies will address upper respiratory symptoms,
exacerbations of preexisting asthma and medical care utilization.
Materials and Methods
Study participants
Because of the unforeseen nature of the event, the study was by
necessity designed as a hybrid cross-sectional and retrospective cohort
study of residents in an "exposed area" and a "control area" and was
approved by Institutional Review Boards of the New York State Department
of Health and New York University. All participants gave their written
consent. Community residents, advocacy groups, local community boards,
local tenant organizations, local medical organizations and clinics all
actively participated in the design and implementation of the study.
Residents in buildings within a one-mile radius from the former WTC were
considered to be in the "exposed area." Building complexes in the
"exposed area" were identified in each direction of GZ. Residents in
buildings located at a greater than 4.8 mile distance north of the WTC
in Manhattan were considered to live in the "control area." Areas south,
east and west of the WTC that were impacted by the plume were excluded
from selection as a "control area." Building complexes in both the
"exposed area" and "control area" were identified by type of housing
unit e.g. low or high-income rental, cooperative, condominium, or
federally funded housing complex, in order to obtain a distribution of
socioeconomic levels in the survey. Building complexes with similar
characteristics were identified for the "exposed area" and "control
area." Residual socioeconomic differences among the study areas were
controlled for during analysis. We over-sampled the population in the
exposed area in order to obtain a large and representative population.
In addition, at the time that this study was developed and implemented,
this was the only study of health effects of local residents and we
thought that the detection of individuals in this study might provide
the only opportunity for identification of residents for follow-up
studies of health effects. We used the ratio of 9:1 (exposed: control
area) while recruiting study participants. The "exposed area" included
49 buildings in lower Manhattan comprising approximately 9,200
households. The "control area" included approximately 1000 households.
Study procedure
A self-administered questionnaire to identify asthma or asthma-like
symptoms was developed from previously validated questionnaires
(Abramson et al. 1991; Asher et al. 1995; Burney et al. 1989; Ravault
and Kauffmann 2001). Questions were modified to delineate reference
times in relation to 9/11/01, to identify the onset of symptoms and to
determine whether symptoms were present within the four-week period of
responding to the questionnaire. Additional questions were included to
obtain demographic information as well as to identify the presence of
the resident in the apartment during the time of interest and any
preceding or subsequent medical problems and medications. Questionnaires
were available in Spanish and Traditional Chinese.
The study was publicized at community board meetings, tenant meetings,
local health fairs, building luncheons and meetings. Notices of the
study were included in local newspapers and building newsletters.
Postings were also placed in buildings and streets. Outreach workers
were situated in buildings at the time of delivery of the questionnaires
to help distribute questionnaires and respond to questions.
Questionnaires were distributed to all defined buildings 12+4 months
after the collapse of the WTC. Questionnaires were initially distributed
via bulk mail. However, it became apparent that the federal postal
service was not functioning in many of the areas near GZ in a consistent
manner and that many of the questionnaires had not been delivered.
Subsequently questionnaires were hand delivered to every apartment or,
when entry was denied, to every building lobby in areas with
questionable mail delivery. A first class mailing of the questionnaire
was then repeated and all apartments were sent reminder postcards. Up to
four residents (two adults, two oldest children) in each apartment were
requested to complete the questionnaire. Because of a concern about
potential selection bias in our response, two buildings of similar
housing stock were targeted in the "exposed area" and in the "control
area" for more intensive outreach. Residents of these buildings received
a third copy of the packets and outreach workers remained in these
buildings for additional days and evenings to respond to questions and
reinforce participation in the study. These targeted buildings
represented 440 households in the "exposed area" and 240 in the "control
area." These targeted buildings with higher response rates were used to
provide an estimate of selection bias compared to the remaining study
sites.
Case definitions
"Previously normal" residents were considered to be those who did not
have a physician diagnosis of asthma, chronic obstructive pulmonary
disease (COPD) or emphysema prior to 9/11/01.
"Previously normal" residents with "new-onset symptoms" were considered
those that answered positively to any of the questions pertaining to
respiratory symptoms of cough, shortness of breath (SOB), or wheeze, or
were using oral or inhaled medications for asthma at any time after
9/11/01. "Previously normal" residents with "persistent new-onset
symptoms" were defined as participants with symptoms that began after
9/11/01 and who had a frequency of symptoms more than twice each week or
medication use within four weeks of responding to the questionnaire.
Screening spirometry
Participants who were "previously normal" participants and had
"persistent new-onset symptoms" were invited to perform a scheduled
screening spirometry at a local community site. Participants were
excluded for analysis if they were under 6 years of age because of the
potential for technical difficulties. Participants older than age 65, or
with a history of cardiovascular disease were excluded for safety
reasons, because the studies were performed in the field. Participants
with a current or greater than 5 pack-year history of cigarette use, who
lived in the "control area" but worked in the "exposed area," or
returned to the residence after 1/2002 or refused to be recontacted were
also excluded.
Spirometry was performed in the field by trained personnel with a Micro
Direct portable spirometer that complied with American Thoracic Society
(ATS) specifications. Studies with three measurements within 5% of each
other were considered acceptable. Participants on medications were asked
to withhold use of medications for at least 4 hours. Values of forced
expiratory volume in one second (FEV1), forced vital capacity (FVC),
FEV1/FVC, and flows at mid lung volumes (FEF25-75) were obtained.
Analyses were performed using normal predicted values from Hankinson et
al. (Hankinson et al. 1999). Because studies were being performed in the
community, bronchodilator responses were not assessed.
Airway hyperresponsiveness
"Previously normal" participants were invited to perform a methacholine
challenge test (MCT) as a monitor of bronchial hyperresponsiveness.
Participants less than age 55 with an FEV1 > 70% predicted and either
"persistent new-onset symptoms" or absence of symptoms were invited to
undergo a MCT at the New York University/Bellevue Hospital pulmonary
function laboratory. Spirometry was performed to confirm baseline
values. MCT was performed using the 2 minute tidal breathing protocol
with methacholine delivered via a nebulizer up to a maximal dose of 8
mg/ml (Crapo et al. 2000). A test was considered positive if the subject
had a > 20% drop in FEV1.
Statistical methods
The overall response rate was calculated based on the number of
households responding in the "exposed area" and " control area" because
of the variation in the number of individuals residing in each
household. An undetermined number of residents permanently moved out of
the "exposed area" after the event. Packets that were returned unopened
were therefore considered to have come from vacant households and were
considered vacant for this calculation. The rates for each health
outcome were calculated as the number of participants with a specific
outcome, divided by the number of eligible participants. Cumulative
incidence ratios (IR) comparing the "exposed area" and "control area"
were computed and 95% confidence intervals (CI) were used to estimate
the precision of the cumulative incidence ratio. Unconditional logistic
regression analysis was used to compute adjusted odds ratios (OR) while
controlling for potential confounders including age, gender, education,
race and smoking. Because respiratory diseases are not rare events, odds
ratios from logistic regression tended to persistently overestimate
cumulative incidence ratios. Therefore, the crude incidence ratios with
95% CI were presented in result tables, and adjusted OR as well as
confidence intervals were used only to examine if the results were still
statistically significant after controlling for confounders.
The demographic characteristics between the participants in the "exposed
area" and "control areas" were compared using the t test of continuous
variables (e.g. age) or the chi-square test for categorical variables
(e.g. gender). For the analysis of the spirometry data, means and
standard deviations (SD) are presented. The t test was used to compare
the mean in the "exposed area" to the mean in the "control area."
Results
Study participants
A total of 9,168 survey packages were sent to households in the "exposed
area" and 962 to households in the "control area." Responses were
obtained from 2,520 households in the "exposed area" (22.3%) and 295 in
the "control area" (23.3%). Household responses were greater in the
targeted buildings, with 205 of 440 households responding from the
"exposed area (43.8%) and 99 of 240 (41.2%) households responding from
the "control area." A total of 3196 individual responses were returned
for analysis. 384 respondents were excluded from analysis because they
did not reside in the residence on 9/11/01, they returned to the
residence after January 1, 2002, the residence was in the "control area"
but the respondent worked in the "exposed area or the questionnaire was
answered for a person born after 9/11/01. Of the 2812 responses that
were therefore used for analyses, 2520 were returned from residents in
the "exposed area," and 292 from residents in the "control area" (see
fig. 1). The demographic characteristics of the 2812 remaining
respondents are shown in Table 1. In both the "exposed area" and
"control area," there were more women respondents compared to men.
Although most respondents were between the ages of 35 and 64, there was
a higher response rate from older participants in the "control area." A
wide distribution of income levels was detected in both the "exposed
area" and "control area," however more respondents from the "exposed
area" earned less than $25,000 compared to the "control area." More
respondents in the "control area" were Caucasian, whereas more
respondents in the "exposed area" were of Asian and Hispanic/Latino
descent. These differences reflect differences in the underlying
populations according to the 2000 U.S. Census and were considered
potential confounders. As such, they were controlled for in multivariate
analyses.
Respiratory symptoms in residents
A previous diagnosis of respiratory disease was identified in 417
(16.6%) and 41 (13.9%) of the residents in the "exposed area" and
"control area" respectively. These residents were not considered
"previously normal" and were excluded from subsequent analysis. Thus
information from 2103 participants in the "exposed area" and 251
participants in the "control area" was available for analysis.
As shown in Table 2, more than twice as many "previously normal"
residents in the "exposed area" complained of respiratory symptoms at
some time after the collapse of the WTC compared to residents in the
"control area." Cough was the most common symptom, and was noted in
three times as many participants in the "exposed area" compared to the
"control area." Four times as many residents in the "exposed area"
complained of wheeze compared to residents in the "control area."
Approximately three times as many residents in the "exposed area"
complained of SOB. The difference in these symptoms in the residents in
the "exposed area" remained significant even after adjusting for age,
gender, education, smoking and race. To assess whether respiratory
symptoms were persistent, "previously normal" participants were asked
about the presence and frequency of individual symptoms within the four
weeks preceding the survey. Symptoms were considered persistent if they
occurred with a frequency of at least twice each week. As shown in Table
3, symptoms had resolved in many of the residents by the four weeks
preceding the survey. However, almost three times the number of
residents in the "exposed area" continued to have any persistent
respiratory symptom compared to residents in the "control area." The
predominant symptom remained cough. Persistent wheezing was reported in
10.5% of participants in the "exposed area" compared to 1.6% in the
"control area."
Similar results were noted for the targeted population that received
intensive outreach and had a greater response rate (43.8 and 41.2%
response rate for "exposed area" and "control area" respectively)
compared to the total study population. Respondents from the targeted
"exposed area," had a greater risk of new-onset respiratory symptom (IR
3.05, CI 2.12,4.39) and persistent respiratory symptoms (IR 4.63, CI:
2.50,8.57) compared to residents in the targeted "control" area.
Persistent daytime shortness of breath was reported in 13.7% and
wheezing was reported in 13.7% of these "previously normal" residents.
The severity of the reported persistent symptoms, as defined by the
frequency of each individual symptom was assessed in "previously normal"
participants with "persistent new-onset symptoms." This analysis is
shown in Table 4. Almost 24% of participants with a persistent symptom
complained of cough on a daily basis. Daily wheezing was described by
17.5% of the residents in the "exposed area" who had a persistent
symptom. Using frequency of symptoms to characterize severity of asthma
according to the Global Initiative for Asthma guidelines, this symptom
frequency would be compatible with at least moderate persistent asthma
(National Heart 2002).
Screening spirometry in residents
Three hundred sixteen participants were eligible and agreed to screening
spirometry in the field. Many residents did not respond to repeated
attempts at telephone scheduling, failed to come to the scheduled
appointments, or could not complete a successful study. Spirometry was
successfully completed in 117 (37%) of the eligible residents. No
differences were detected between residents with symptoms in the
"exposed area" compared to asymptomatic residents in any parameter of
airflow including FEV1, FVC, FEV1/FVC, or FEF25-75 (Table 5). We failed
to observe a difference in the number of individuals with an FEV1 or
FEV1/FVC below the lower limit of normal in the individuals in the
"exposed" area with new onset persistent symptoms and asymptomatic
individuals, or between individuals in the "control" area. 20.8% of
participants with persistent symptoms in the "exposed area" had used a
controller medication (inhaled corticosteroid, long acting beta agonist,
theophylline compound, leukotriene modifier) in the 4 weeks before
spirometry, compared to none of the participants in the "asymptomatic"
groups. 16.7% of participants with persistent symptoms in the "exposed
area" had used a short-acting beta-adrenergic agonist inhaler for asthma
compared to 1.5% in the asymptomatic "exposed" group, and none in the
"control" residents.
All participants were invited to undergo a methacholine challenge test
(MCT) according to eligibility criteria defined in the methods section.
MCT was performed in 24 volunteer participants including those with
"persistent new-onset respiratory symptoms" (n=12), asymptomatic
participants from the "exposed area" (n=6), and asymptomatic
participants from outside the "exposed area" (n=6). No significant
difference was noted in baseline spirometry between these groups (data
not shown). Many (6/12) participants with "persistent new-onset
symptoms" had a positive MCT compared to asymptomatic participants
(p<0.05). None of the asymptomatic participants in either group had a
positive MCT.
Discussion
The World Trade Center Residents Respiratory Health Study was initiated
in response to questions by residents in the surrounding community of
the disaster site about the respiratory health risk for residents and
was designed to study upper and lower respiratory tract symptoms,
physician diagnoses, unplanned medical visits, and physical condition of
the apartments after the collapse of the WTC. We now report on the
presence and persistence of new respiratory health issues in residents
near GZ. The study was completed 16 months after the destruction of the
WTC. Our study suggested an increased incident rate of new onset
respiratory symptoms in residents near Ground Zero compared to residents
in a "control area." Although these symptoms resolved in many residents,
an increased incident rate of "persistent new-onset" respiratory
symptoms was also detected compared to a control group. These data
suggest that exposure to dust and fumes from the destruction of the WTC
was associated with new-onset respiratory symptoms that persisted in a
subset of residents.
The predominant respiratory symptom detected in symptomatic residents
consisted of cough, with some participants also experiencing dyspnea and
wheezing. These symptoms are consistent with those identified in the
rescue workers and responder populations such as the firefighters and
ironworkers (Feldman et al. 2004). They fit some, but not all criteria
for reactive airways dysfunction (RADS) (Alberts and do Pico 1996;
Bardana 1999; Brooks et al. 1985). We cannot document the exposure level
of the residents to the dusts and fumes and whereas some of these
residents may have had high-level exposure from the initial dust cloud,
others may have only experienced lower level exposure from settled dust
and persistent fires. Descriptions of irritant-induced asthma have
included cases with a history of repeated low intensity exposures, in
which the symptoms have a more delayed expression and this pattern may
be more consistent with the potential exposure history and symptoms of
many of the residents in this study (Brooks 1998; Kipen et al. 1994).
The persistence of symptoms identified in some of the study participants
is also consistent with an irritant-asthma-like picture, in which
symptoms can persist for years (Chang-Yeung et al. 1994; Demeter et al.
2001). The persistence of symptoms is also consistent with the findings
recently described in firefighters exposed to WTC dusts (Banauch et al.
2003).
Only a subset of residents with potential exposure experienced the onset
and persistence of respiratory symptoms. The characteristics of this
susceptible group are unclear. The variation in response may be due to
differences in the intensity or duration of exposure to the WTC dusts in
the population with persistent symptoms compared to those without.
Alternatively, irritantinduced asthma has been described to be more
common in participants with preexisting allergic or atopic disorders
(Brooks et al. 1998). We did not specifically explore whether
participants with persistent symptoms had pre-existing atopic disorders
in this study.
The possibility exists that psychological stress might play a role in
the reported symptoms, as post-traumatic stress disorder has been
reported to be associated with asthma and other respiratory diseases
(Fagan et al. 2003). The current study could not determine whether
environmental factors, psychological distress, or a combination,
contributed to the increase of respiratory symptoms, since psychological
factors were not examined in this part of the study. Despite our
original hypothesis, we were unable to detect a significant difference
in airflow parameters measured by screening spirometry performed in the
field between residents with persistent new-onset respiratory symptoms
and asymptomatic or control residents. We did not have preexisting
medical information available to us for the population of study and as a
result, we performed between subject comparisons. The possibility exists
that our statistical power was not great enough to detect small
differences in airflow measurements between the two populations. The
symptoms detected in the exposed population may also be due to changes
in the small airways or to an increase in bronchial hyperresponsiveness,
both of which can be missed with routine screening spirometry. In
addition, many participants in the group with persistent symptoms in the
"exposed" area were using a controller medication at the time of the
study. Use of these medications may have improved their lung function.
The findings are however consistent with those described in firefighters
exposed to WTC dusts in which no significant differences in spirometry
values were detected between participants with high and low exposure,
suggesting that these parameters are insensitive for between subject
comparisons in these exposed populations (Prezant et al. 2002). MCT
performed in a small pilot study of participants suggested that the
symptoms of some of these residents might be explained by the presence
of bronchial hyperresponsiveness, a finding that would be consistent
with the data reported for firefighters (Prezant et al. 2002).
The predominant compounds detected in the settled dusts collected one
and two days after the WTC explosion included calcium sulfate (gypsum)
and calcium carbonate (calcite) (Lioy et al. 2002; McGee et al. 2003;
Service 2003). The aqueous extracts were extremely alkaline (Lioy et al.
2002). These particle characteristics are associated with mucus membrane
irritation and thus have the potential to elicit airway symptoms
consistent with those detected in this study (Stellman 1998). Biological
plausibility for health effects from WTC dusts is supported by In vitro
and in vitro studies. Primary human lung cells (alveolar macrophages and
epithelial cells) reveal an increase in inflammatory cytokines
Interleukins 8 and 6 in response to WTC dusts (Payne et al. 2004).
Animal studies, using WTC-derived fine particulate matter demonstrate
that very high doses elicit pulmonary inflammation and
hyperresponsiveness (Gavett et al. 2003). Although lower doses of these
particles did not induce inflammation or hyperresponsiveness, the
effects of chronic exposures were not tested in these studies. Despite
the large sample size of this study, there are some potential
limitations to the study. In contrast to the firefighters, in whom a
baseline health and pulmonary function profile was well established and
documented before 9/11/01, no consistent information was available about
the health of the residents in the surrounding GZ community before
9/11/01. Many of these residents were considered normal before 9/11/01,
and therefore do not have documented respiratory health information
preceding 9/11/01. We therefore used self-reported health information.
The possibility of reporting bias or differential recall by persons in
the different study areas exists. To minimize this possibility,
questions about health problems that should be unrelated to WTC events
were also included in the questionnaire. The similar rate of problems
such as disability affecting physical activity in the two areas (14.2%
and 13%, respectively) suggested the absence of significant reporting
bias due to residence area. Participants responding affirmatively about
every symptom may have been affected by recall bias. Ten of the
respondents answered in this way, however minimal changes were observed
when these individuals were excluded from the analysis. We also obtained
information about unplanned medical visits in the months after the WTC
collapse, events that may be more memorable than symptoms. Unplanned
medical visits for respiratory problems were significantly increased in
the affected area (14.7 percent) over the control area (8.4 percent)
(CIR: 1.76, 95% CI: 1.15, 2.68) after controlling for potential
confounders. A significantly higher proportion of affected area
residents started using respiratory medication after 9/11/01 (17.9
percent) compared to controls (6.2 percent) (CIR: 2.88, 95% CI: 1.75,
4.75). We also compared the proportion of respondents reporting a
specific respiratory symptom and unplanned medical visits in both areas.
We found that the proportions were similar in the affected and control
areas for most symptoms. If there had been over-reporting in the
affected area, the proportion of individuals reporting a specific
symptom who also had unplanned medical visits should have been lower in
the affected area than in the control area.
Despite the active involvement of the community in the design and
implementation of this study, the response rate in both the "exposed
area" and "control area" was low. Several possible explanations can be
suggested for this low response rate. First, although we used many means
to deliver the questionnaires, the absence of reliable mail in many of
the "exposed areas" may have reduced our ability to reliably distribute
the surveys. Moreover, because of the well-documented emotional
aftermath of the event, many residents may have been unwilling to answer
questions that may have provoked sensitive emotions even one year after
the event. In addition, at the time of the study, residents were also
receiving forms from many other agencies. Both confusion over which
studies were being completed, as well as study fatigue, may have
occurred. Finally, we were unable to determine a true response rate
because a significant number of residents permanently moved out of the
"exposed area" after 9/11/01. In some buildings, residents estimated
that over 50 per cent of the occupants had moved from the buildings. We
were unable to obtain a listing of residents in the area before and
after the event and for this reason, the denominator for calculating the
household response rate may have been an overestimate, resulting in an
underestimate of the actual response rate. Furthermore, low response
rates are common for studies performed in New York City; the 2000 Census
only recorded a final response rate in New York City of 55% despite
intense advertising and door-todoor follow-up.
The potential for selection bias exists in this self-administered survey
and it is possible that residents with new onset respiratory symptoms
may have been more likely to participate in this study compared to those
without symptoms. Several procedures were used during the study in an
attempt to minimize this potential problem. The importance of
participation for residents with and without breathing problems was
stressed in all announcements of the study. In addition, a target
population that received intensive outreach was studied in both the
"exposed area" and "control area." This target population, which had a
higher response rate compared to the study population as a whole,
demonstrated an even greater increase in persistent symptoms in
residents in the "exposed area" compared to the "control area" with an
increase in individual symptoms ranging from 14% to 63% in the target
population. Had there been a significant selection bias or an
over-estimation of the association, analysis of the target population
should have demonstrated a weaker exposure-disease association compared
to the control population. In contrast, analysis of symptoms in the
target population demonstrated that increases in newonset symptoms were
consistently and significantly higher in the exposed areas compared to
the control area. This finding suggests that if there were selection
bias, it would be in the opposite direction (i.e., the true association
would be underestimated).
A plume dispersion model is not yet complete by the United States
Environmental Protection Agency and is thus was not available to us to
allow a detailed exposure assessment. However, we obtained self-report
information on the condition of the individual households as a possible
surrogate for exposure. Many of the apartments that were in close
proximity to GZ were severely damaged by the event. Apartments that
surrounded GZ in all directions were covered in dusts from the initial
dispersion. The presence of "persistent new onset respiratory symptoms"
was significantly associated with the presence of physical damage of the
apartment, dust on the surfaces, or a long duration of dust or odors
(data not shown). In addition, residents who were south of Canal Street
in lower Manhattan on 9/11/01 i.e. in close proximity to the World Trade
Center were at higher risk of developing "persistent new onset
respiratory symptoms" compared to residents who were not in the area on
the day of the event.
The possibility of exposure misclassification may also exist. To
minimize this bias we excluded individuals who had moved out of their
residence for a prolonged period of time, or who may have had exposure
that was unrelated to their area of residence. Some residents may have
altered their behavior and spent less time at home in the aftermath of
9/11, however we would not be able to identify these residents. In
addition, because of wind, it is also possible that the WTC dust plume
also impacted residents in the "control area".
Conclusion
These data suggest that residents living in the community surrounding
the former WTC experienced a higher rate of adverse respiratory health
effects one year after the event compared to a control population.
Respiratory symptoms consisted of cough, dyspnea and wheeze. Although
most of these symptoms resolved by approximately 12 months after the
event, a significant number of residents continued to have "persistent
new-onset" respiratory symptoms. Abnormalities in screening spirometry
failed to explain the symptoms in these participants and additional
tests, including tests for bronchial hyperresponsiveness may be helpful
to further characterize these symptoms. Biological plausibility for
these complaints is provided by the current chemical analysis of the
settled dusts and animal studies. Long-term health effects remain
unknown and warrant further investigation and follow-up of exposed
residents.
|
|
|
