|
Winthrop
Environmental Health Facts Subcommittee (Winthrop
Airport Hazards Committee) Winthrop
Board of Health AIR
Brian
Dumser, PhD, CIH Chair
of the Subcommittee
August
18, 1999
Summary
In many communities located close to major airports,
power generation facilities, or other major industries, there is a strong
perception that pollution generating activities at these facilities result in
a direct negative impact on the health of residents. Statements to this effect have been repeatedly voiced by
representatives of the communities surrounding Logan airport, but, absent hard
data in the existing record, no action has been taken by responsible
authorities to investigate further. Currently,
plans are underway for the construction of additional facilities Logan airport
which will markedly increase operational capacity and the generation of
pollutants. While potent
arguments in favor of this expansion are being presented from an economic
standpoint, once again no consideration is being given to the possible public
health impact.
In light of the failure to address this issue by
Massport, or by Federal or State regulatory authorities, the Winthrop
Environmental Health Facts Subcommittee, a voluntary group made up of
residents of the Town of Winthrop Massachusetts, elected to address the
question directly. A strong
correlation is known to exist between exposure to petrochemical exhaust
emissions and a variety of respiratory and cardiovascular diseases (1-10).
Logan airport estimates its daily production of such pollutants at
approximately 50,000 pounds per day (11). The Subcommittee undertook a survey
to determine whether a correlation also exists between frequency and severity
of respiratory disease and level of exposure to these pollutants as determined
by location in Winthrop relative to the airport.
The
results of this survey demonstrate that a clear increase in several
respiratory diseases and disease symptoms exists between areas of the Town
which are adjacent to the airport, and those more distantly located on Broad
Sound. In fact, for the most common respiratory diseases, asthma and allergy, disease is
twice as common in the most heavily exposed neighborhood as it is in the least
exposed. Finding no other
likely explanation for this effect, the Subcommittee proposes that airport
activities, most likely the generation of airborne pollution from the
combustion of gasoline and kerosene, are indeed negatively affecting the
health of the residents of Winthrop.
The implications of these findings are serious.
While the unique geography and demographics of Winthrop provided a
situation where the effects of airport generated pollution could be studied in
isolation from other pollutant sources, Winthrop is by no means the only
community impacted, nor the community most highly impacted by airport
activity-generated emissions. As sample size determines the sensitivity of the
analysis, only the most frequently occurring respiratory diseases could be
adequately tested. Thus, while
the case can be made strongly for asthma and allergenic disease, effects on
other less common serious or life-threatening respiratory and cardiopulmonary
conditions which are also linked to fuel exhaust exposure remain an unexplored
possibility. Finally, while the study convincingly illustrates the
difference in impact due to relative exposure level, it does not define a
level of exposure where impact is minimal or tolerable. In brief, the study
demonstrates that serious damage
is being done to the health of the residents of Winthrop at current levels of
airport activity, and this damage correlates with location, a measure of
exposure to airport activity-generated pollution. The Subcommittee feels it is incumbent on State regulatory
authorities responsible for the public health to further investigate this
matter, to further define the scope and severity of the problem, and initiate
processes which will return our community to the state of health enjoyed by
the majority of Massachusetts citizens.
Introduction
Winthrop is a peninsula which extends from East
Boston south by south east to form the division between Broad Sound, on its
eastern shore, and Boston Harbor on its western shore.
A portion of the western shore entirely encloses, and closely
approaches Logan airport. Winthrop is subjected to a variety of disturbances from the
airport, including excessive noise and odors from burned and unburned fuel.
Although Logan carries out no air pollution monitoring in the
surrounding communities, their published estimates from modeling studies
indicate approximately 50,000 pounds of airborne pollutants are released
daily, primarily from the combustion of Jet Fuel A.
Elsewhere it has been shown that a strong correlation exists between
exposure to such pollutants and a variety of respiratory and cardiovascular
diseases including lung cancer, chronic obstructive pulmonary disease, asthma
and allergic rhinitis (1-10). Individuals
residing in communities surrounding Logan airport show a considerably higher
incidence of these diseases compared to the statewide average (12-14). It has not been possible to determine whether Logan airport
activities contribute substantially to this health burden however, since the
urban location of these communities presents a complex picture of pollution
sources, including petrochemical pollution from power plants, industries, and
heavy road traffic.
Winthrop, by contrast, is a stable, mature
residential community without significant pollution sources except for the
airport. Despite this fact,
asthma incidence in Winthrop closely mirrors that in the mainland communities
which abut the airport, and lung cancer rates for females is 50% higher than
the statewide average (14). Some
neighborhoods in Winthrop are located within a few hundred feet of major
airport runways, while others are located as much as a mile and a half away.
Residents report a marked difference in perception of chemical odors
from the airport in relation to location in the Town, indicating that
different levels of exposure occur within the Town resulting from distance
from the airport and wind direction. In
consideration of these facts, this study was conducted to determine whether
any correlation exists between the level of exposure to air pollutants
generated by airport activity and the incidence of and frequency of symptoms
to respiratory disease.
Methods
The Town was divided into 10 neighborhoods, primarily
on the basis of natural topography, containing between 1,000
and 2,500 residents each. Two
neighborhoods were selected as likely representing
areas of highest (#1, Court Road, and #2, Cottage Park), and lowest (#5,
Winthrop Beach, and #6, Winthrop Highlands) exposure.
A questionnaire was devised, consisting of 30 questions to obtain
information on the incidence of diagnosed asthma, allergies, chronic
bronchitis, chronic sinusitis, and emphysema, and on the frequency of symptoms
experienced. Standard demographic
information was also obtained on gender, age, and the duration of residence in
the neighborhood. A smoking
history was obtained, and information on the frequency of perception of odors
caused by airport-related activities. Responses
to questions on diagnosed disease incidence were yes/no, followed by a
question on time since onset. Responses
to questions on symptom frequency included none and either 4 or 5 frequency
ranges.
Interviews were conducted by volunteers from the
community who were trained in requirements for objective data collection,
chain-of-custody, and anonymity requirements.
Interviews were conducted 4 weekday evenings per week, between the
hours of 6:30 and 8:30 PM. Team
leaders assigned streets to the interviewers.
Every residence in the neighborhood was approached, one time only,
until the entire neighborhood was canvassed.
All residences, single and multiple family dwellings and apartment
complexes were sampled, with the exception of mechanically ventilated
buildings. No commercial
establishments were encountered in the zones polled. In this manner, a random
sample of residents was polled which averaged approximately 18% of the
population of the selected neighborhood.
The only exception to this was neighborhood 5, the last area sampled. Activity was continued in this area, progressing from north
to south, until the desired quota of 500 interviews each in low and high
exposure areas was obtained. Each
questionnaire was identified only by neighborhood, and no names or addresses
were collected. The
questionnaires were collected each evening and held centrally.
Following data entry, the database was screened to
exclude unsuitable responses. Corrections
were made to the database where possible, for example intelligible but
non-numerical responses. Questionnaires
with critical data missing or internally contradictory responses were
excluded. Data was also discarded
for individuals residing in the identified zone for less than one year, or who
were not in residence for at least four days per week.
All such changes were recorded. Of
the 1000 questionnaires obtained, 838 were admissible, 430 from the
high-exposure zone (Area 1 - 172; Area
2 - 258) and 408 from the low-exposure zone (Area 5 - 197;
Area 6 - 211).
In light of the seriousness of the effects on human
health, and the truncated timetable presented by airport expansion activities,
simplified exploratory statistical analyses were first carried out by
excluding from the data all individuals not smoke-free for the past five
years. Data from high exposure
(areas 1 and 2) and low exposure (areas 5 and 6) zones were pooled, and
symptom frequency compared by chi-squared contingency analysis.
The results of this analysis formed the basis for an earlier report
which was presented by the Caucus on Air Transportation to representatives of
the state government July 1, 1999.
While that approach provided a convincing and
statistically significant demonstration of the differential effect of location
on disease incidence, the dataset contains more information which can be
accessed by more sophisticated analyses.
To this end, the Subcommittee contracted the services of an
epidemiological analytical firm, John Snow Inc., to further analyze the data.
SAS software was employed to re-incorporate smokers into the study,
correcting for smoking history, age and sex by means of the Mantel-Haenszel
Test. Additional statistical
analyses were performed with Epi Info V6 (15).
Further, it was noted that while low-exposure zones 5 and 6 were
essentially equivalent, high exposure zones 1 and 2 showed a differential from
one another which was consistent with position relative to the airport.
Contingency analysis was thus carried out for each of these zones
separately, compared to the joined low-exposure population 5 and 6.
The complete set of statistical analyses, identification and criteria
for data exclusion, complete and amended datasets, and original survey
questionnaires are on file with the Winthrop Board of Health.
Results
Table
1. Frequency
of Odor Perception %
Response on Scale 0 - 100 (Days/Year)
Table
2. Relative
Risk High
Exposure Area 1 vs Pooled Low Exposure Zone (Areas 5 + 6) Total
Sample Size - 580
Table
3. Relative
Risk High
Exposure Area 2 vs Pooled Low Exposure Zone (Areas 5 + 6) Total
Sample Size - 666
** Relative Risk is the
proportionate increase (or decrease) in disease incidence in the high exposure
area compared to the low exposure area, adjusted for influences due to the
age, sex and smoking history as estimated by the Mantel-Haenszel procedure.
** p value is the likelihood
that the values obtained in the high and low exposure zones come from the same
population and differences are due simply to random variation.
The results clearly show that a differential increase
in respiratory disease occurs from the low exposure zones (area 5 and 6)
through the moderately exposed area 2 to the highly exposed Court Road area 1.
The statistical significance is absent for the infrequent conditions
chronic bronchitis and emphysema, though a positive trend is still evident.
Chronic sinusitis shows a strong correlation with the most highly
exposed area. For the more common
diseases, allergies and asthma, statistical significance of the correlation
with location is extremely strong for the most highly exposed area 1; while
less strong for the more moderately exposed area 2, the trend is well
maintained.
Table
4. Disease
Incidence; Clinically Diagnosed, Self-Reported Most
Likely Estimate, 95% Confidence Limits
Table
5. Predicted
Excess Disease in High Exposure Areas
Table
6. Frequency
of Respiratory Symptoms %
Response in Scale 0 - 100
Table
7. Percent
of Respondents Symptomatic At Any Level Restricted
Lung Function (Inhaler Use, Asthma Attack, Wheezing) and Bronchonasal
Irritation (Cough, Rhinitis)
Discussion
The primary goal of this study was to determine
whether spatial location relative to Logan airport, as a determinant of
chemical exposure, has an influence on respiratory disease in the Town of
Winthrop. While the exact
component or mixture of components responsible for the effect is as yet
unclear, it has been well established in the literature that exposure to
pyrolysis products of fossil fuels correlates strongly with both incidence of
and symptomatic response for several important respiratory diseases. In the majority of urban settings, multiple sources of such
pollutants make it difficult or impossible to identify the impact of
individual polluters. Winthrop, a
residential community occupying a peninsula in Massachusetts Bay, has no major
local petrochemical pollution sources with the exception of Logan airport.
While generalized airborne pollution from nearby Boston and its suburbs
no doubt contributes to the burden, such effects are sufficiently distant as
to be well-mixed, affecting the Town equally.
Logan airport by contrast approaches within a few hundred yards of
portions of the Town. Residents
report a very distinct geographical pattern of odor perception of burned and
unburned kerosene (Jet Fuel A) and burning rubber from airplane tires.
Other neighborhoods within the Town are more remote and less plagued by
this problem. We thus conducted a
survey to determine if there existed a correlation between spatial location
and odor perception, as an index of chemical exposure, and both frequency of
diagnosed respiratory disease, and prevalence of symptoms to that disease as
an indicator of negative health impact.
Odor Perception / Exposure Level
A central component of the argument put forward in
this report is that spatial location within the Town of Winthrop relative to
the airport is an adequate determinant of exposure to airport-activity
generated pollutants. While
anecdotal reports regarding the perception of fuel and burnt rubber odors from
residents support the contention, and epicenters of the sampled neighborhoods
are approximately 0.4 miles (area 1), 0.8 miles (area 2) and 1.5 miles (areas
5 and 6) from runways, direct correlation of location/exposure level is
lacking. Actual pollutant
concentration in these areas is unknown, as no monitoring is carried out. In lieu of direct measurement, Massport carries out
mathematical dispersion modeling of
several important components of fuel and fuel exhaust (Carbon Monoxide,
Nitrogen Dioxide, Volatile Organic Compounds, and Particles of diameter 10
�m. or less). Three sites in the
Massport projection grid correspond very closely to the areas sampled in this
study. Exact matches are found
for area 1 (Court Road) and area 2 (Cottage Park), areas in close proximity to
the airport. In addition, area 6
forms its northern border with the Massport projection area Revere Beach.
While such models are useful tools, they are at best approximations of
real conditions and subject to considerable error (16).
Massport�s model predicts uniform particulate concentrations at all
three sites, and an increase in combustion gases of approximately 10% at the
Court Road site, with equivalent concentrations at both Cottage Park and
Revere Beach. Concentrations of
Volatile Organic Compounds, which comprise the fraction responsible for the
noticeable odor, show a wider latitude of dispersion.
Concentrations at Court Road are approximately double that predicted at
Revere Beach. The difference in
concentration between Court Road (area 1) and Cottage Park (area 2) varies
from about 20% (highest peak value in 1 hour) to about 90% (highest peak value
in 24 hours).
Direct evidence of this differential local
concentration was sought in the survey. Frequency
of perception of fuel and rubber odors was sampled in each neighborhood, and
the responses converted to an approximately linear scale from 0 (never) to 100
(two or more times per week). Results
(Table 1) were consistent with spatial location, with mean scores ranging from
approximately 30 in zones 5 and 6 to 60 and 69 in zones 2 and 1.
Median scores were 0 (never) in zone 5, 12 (once per month) in zone 6,
50 (once per week) in zone 2 and 100 (two or more times per week) in zone 1.
While it is clear that only direct monitoring can establish actual and
relative concentrations of these pollutants, sufficient information has been
presented here to justify the classifications of low (areas 5 and 6), moderate
(area 2) and high exposure (area 1).
Disease Incidence.
Ten questions were posed regarding the presence of
each of five respiratory diseases which have been correlated with exposure to
fossil-fuel exhaust , and the date of onset of the disease. The wording of the
questions stressed that the diagnosis had to have been made by a physician (
�Have you ever been told by a Doctor that you have...�), and this fact of
clinical diagnosis reinforced with an approximate date of diagnosis.
Thus, while the replies to these questions are self-reported diagnoses,
and actual incident rates derived from them should be viewed with that
qualification, they are presmed to be reasonably
truthful and at least should not be affected by reporting bias between
different areas sampled. Bias on
the part of the interviewer is also controlled in part by the binary response
(yes/no) recorded. It should be
further noted that the initial sampling strategy presented to the interviewers
who were also members of the Subcommittee which defined the study was a
sampling of highest expected and lowest expected exposure zones.
The initial report presented by the committee was analyzed within that
paradigm, and only further analyzed by individual zones following recognition
of real response difference between zones 1 and 2.
It is very unlikely that the interviewers regarded these two contiguous
neighborhoods as different in terms of exposure level during the course of the
survey, and the existence of substantial difference in response indicates an
absence of interviewer bias.
Tables 3 and 4 show that a very clear increase in
diagnosed disease exists in the neighborhoods in close proximity to the
airport relative to the more remote locations.
Further, while areas 1 and 2 are contiguous, the epicenter of area 2 is
approximately twice the distance from the airport as that of area 1.
Relative risks were calculated, controlling for possible confounding
variables of sex, age and smoking history.
In fact, all four neighborhoods are demographically very similar, and
little effect of these variables was noted.
Estimates of the reliability of the predicted
relative risks, as indicated by the p values, are influenced both by
the magnitude of the difference and the
frequency of the disease in the population.
For the three most prevalent conditions, allergy, asthma, and chronic
sinusitis, the existence of a clear increase in frequency with position closer
to the airport is striking. Further,
the size of the difference is also impressive.
For allergy and asthma, the most highly exposed population experiences
a two-fold increase in disease incidence compared to the least exposed
neighborhoods. As mentioned above, these incident rates (Table 4)
are a reasonable estimate of the level of diagnosed disease in the sample
group, although they should not be compared to other studies which are
primarily based on hospitalization rate or mortality. The rates presented here are consistent throughout the
population under study, and appropriate for analysis of spatially-located
differences in disease rate among the subgroups of that population. They do however include historical cases, and well-controlled
or other asymptomatic conditions which would not appear for example in the
Massachusetts Disease Registry. However,
they do represent negative impacts on the health of the community, and to
place these figures in a more human context, predictions on the effects of
this differential are presented in Table 4.
This estimates that, in areas 1 and 2, contiguous neighborhoods with a
combined population of about 3200 people, there are 220 individuals with
asthma, 435 with allergies, and 131 with chronic sinusitis whose condition is
correlated with their location relative to Logan airport.
Symptom frequency
In contrast to the clear differences demonstrated for
disease incidence, symptom frequency presents a much more complex picture.
Table 6 illustrates symptom frequency for the five diseases sampled in
each zone, as mean values within an approximately linear scale from 0 to 100.
Results are highly variable, and overall scores low due to the high
percentage in each group of asymptomatic respondents. It is probable that the
sample size employed is insufficient to adequately characterize differences in
the much smaller symptomatic subset, and the results should be regarded as
inconclusive. The results
reinforce rather than contradict data presented on disease incidence
distribution however. If the
responses are recast as binary elements (Table 7.
Symptomatic vs Asymptomatic, grouped by functional pathology) a
differential of approximately 50% again emerges between the pooled high
exposure and low exposure zones. References:
1
Abbey DE, Ostro BE, Petersen F, Burchette RJ.
Chronic respiratory symptoms associated with estimated long-term
ambient concentrations of fine particulates less than 2.5 microns in
aerodynamic diameter (PM2.5) and other air pollutants.
J Expo Anal Environ Epidemiol 1995 5: (2) 137-159
2
Bhatia R, Lopipero P, SmithAH Diesel
Exhaust Exposure and Lung Cancer. Epidemiology
1997: 8 : 364
3
Brunekreef B, Janssen NAH, de Hartog J, Harssema
H, Knape M, van Vliet P Air
Pollution from Truck Traffic and Lung Function in Children Living
near Motorways. Epidemiology 1997; 8 : 298
4
Dockery DW, Pope CA III, Xiphing X, Spengler JD, Ware JH, Fay ME,
Ferris BG, Speizer FE. An association between air pollution and mortality in six US
cities. New England Journal of
Medicine, 1993 329: (24)
1753-1760
5
Duhme H, Weiland SK, Keil
U, Kraemer K, Schmid M, Stender
M, Chambless L The
Association between Self-Reported Symptoms of Asthma and Allergic Rhinitis and
Self-Reported Traffic Density on Street of Residence in Adolescents.
Epidemiology 1996;7:578�582
6
LoomisD, Castillejos M, Gold DR, McDonnell W, Borja-Aburto VH Air
Pollution and Infant Mortality in Mexico City.
Epidemiology 1999; 10: 118
7
Moolgavkar SH, Luebeck EG, Anderson EL
Air Pollution and Hospital Admissions for Respiratory Causes in
Minneapolis-St. Paul and Birmingham
Epidemiology 1997: 8: 364
8
Schwartz J Air Pollution and Hospital Admissions for
Cardiovascular Disease in Tucson. Epidemiology 1997; 8 : 371
9
Sheppard L, Levy D, Norris G, Larson
TV, Koenig JQ Effects
of Ambient Air Pollution on Nonelderly Asthma Hospital Admissions in Seattle,
Washington, 1987-1994 Epidemiology 1999; 10: 225
10
Verhoeff A, Hoek G, Schwartz J, van
Wijnen JH Air
Pollution and Daily Mortality in Amsterdam
Epidemiology 1996; 7: 225 �230
11
Logan Airside Improvements Planning Project Volume IV 1999 EOEA #10458
12
Boston Neighborhood Health Status Report: The Health of South Boston.
Boston Department of Health and Hospitals, Division of Public Health,
Office of Research and Health Statistics.
November 1994
13
The Health of Boston 1998. Boston
Public Health Commission, Office of Research, Health Assessment and Data
Systems, Boston Massachusetts 1998
14
Massachusetts Community Information Health Profile, Massachusetts
Department of Public Health, Bureau of Health Statistics and Evaluation,
Boston Masachusetts.
15
Dean AG, Dean JA, Coulombier D, Brendel KA, Smith DC, Burton AH,
Dicker, RC, Sullivan K, Fagan, RF, Arner, TG.
Epi Info Version 6: a word processing, database and statistics program
for public health on IBM-compatible microcomputers.
Centers for Disease Control and Prevention, Atlanta, Georgia, USA, 1996
16
Beychok MR. Error Propagation in Air Dispersion Modleing.
Newport Beach CA
Winthrop
Environmental Health Facts Subcommittee Members
of the Subcommittee Conducting the Survey
Barbara
Bishop Madeline
Burke Brian
Dumser Eleanor
Casey Greg
Curci John
Dowd Arthur
Flavin, Sr. Barbara
Corbett Flavin Connie
Mara John
Macy Harvey
Maibor Bob
Massa Kathleen
Mccauley Ellie
Olivolo Judith
Silck Claire
Sweeney Winthrop Health Study Questionnaire Response Sheet
1.
Sex:
0 Male
1 Female.
2.
Age:
0 1
2 3
4 5
3 Current Residence: 0 1 2 3 4 5 6 7 8 9
C
Number of years:
0 1
2 3
4 5
6 7
8 9
C
Former Residence:
0 1
2 3
4 5
6 7
8 9
C
Number of years:
0 1
2 3
4 5
6 7
8 9
C
Former Residence
0 1
2 3
4 5
6 7
8 9
C
Number of years:
0 1
2 3
4 5
6 7
8 9
9 Smoking 0 Yes 1 No
10 Packs per day 0 - Less than 1 1 - 1 2 - 1 � 3 - 2 4 - More than 2
11 Quit 0 Yes 1 No
12 Smoke-Free 0 1 2 3 4 5 6 7 8 9
13 Days/Week Away 0 1 2 3 4 5 6 7
14 Local employment 0 Yes 1 No
15 Allergies 0 Yes 1 No
16 How Long? 0 Less than 1 year 1 1-5 years 2 6-10 years 3 More than 10 years
17 Asthma 0 Yes 1 No
18 How Long? 0 Less than 1 year 1 1-5 years 2 6-10 years 3 More than 10 years
19 Chronic Bronchitis 0 Yes 1 No
20 How Long? 0 Less than 1 year 1 1-5 years 2 6-10 years 3 More than 10 years
21 Emphysema 0 Yes 1 No
22 How Long? 0 Less than 1 year 1 1-5 years 2 6-10 years 3 More than 10 years
23 Sinusitis 0 Yes 1 No
24 How Long? 0 Less than 1 year 1 1-5 years 2 6-10 years 3 More than 10 years
25 Inhaler Use 0 Once a month 1 Once a week 2 2 or more times per week 3 Once a day 4 2 or 3 times a day 5 More than 3 times a day
26 Asthma Attack 0 Once a month 1 Once a week 2 2 or more times per week 3 Once a day 4 2 or 3 times a day 5 More than 3 times a day
27 Wheezing or Shortness of Breath? 0 Once a month 1 Once a week 2 2 or more times per week 3 Once a day 4 2 or 3 times a day 5 More than 3 times a day
28 Coughing Spells? 0 Once a year 1 Once a month 2 2 or more times per month 3 Once a week 4 More than once a week
29 Runny Nose, Tearing Eyes, Sinus Headache? 0 Once a year 1 Once a month 2 2 or more times per month 3 Once a week 4 More than once a week
30 Exhaust, Chemical or Fuel Odors? 0 Once a year 1 Once a month 2 2 or more times per month 3 Once a week
4
More than once a week
|