EPA-600/1-77-043
September 1977
Environmental Health Effects Research Series
RESPIRATORY DISEASE IN CHILDREN
EXPOSED TO SULPHUR OXIDES AND
PARTICULATES
Health Effects Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and. Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL HEALTH EFFECTS RE-
SEARCH series. This series describes projects and studies relating to the toler-
ances of man for unhealthful substances or conditions. This work is generally
assessed from a medical viewpoint, including physiological or psychological
studies. In addition to toxicology and other medical specialities, study areas in-
clude biomedical instrumentation and health research techniques utilizing ani-
mals — but always with intended application to human health measures.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/1-77-043
September 1977
RESPIRATORY DISEASE IN CHILDREN EXPOSED TO
SULFUR OXIDES AND
PARTICULATES
by
Douglas Ira Hammer (*)
Population Studies Division
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
(*) Present address:
Director, Emergency Department
Rex Hospital
Raleigh, North Carolina 27603
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
HEALTH EFFECTS RESEARCH LABORATORY
RESEARCH TRIANGLE PARK, N.C. 27711
-------
DISCLAIMER
This report has been reviewed by the Health Effects Research
Laboratory, U.S. Environmental Protection Agency, and approved for
publication. -Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
ii
-------
FOREWORD
The many benefits of our modern, developing, industrial society are
accompanied by certain hazards. Careful assessment of the relative risk
of existing and new man-made environmental hazards is necessary for the
establishment of sound regulatory policy. These regulations serve to
enhance the quality of our environment in order to promote the public
health and welfare and the productive capacity of our Nation's population.
The Health Effects Research Laboratory, Research Triangle Park,
conducts a coordinated environmental health research program in toxicology,
epidemiology, and clinical studies using human volunteer subjects. These
Studies address problems in air pollution, non-ionizing radiation,
environmental carcinogenesis and the toxicology of pesticides as well as
other chemical pollutants. The Laboratory develops and revises air quality
criteria documents on pollutants for which national ambient air quality
Standards exist or are proposed, provides the data for registration of new
pesticides or proposed suspension of those already in use, conducts research
on hazardous and toxic materials, and is preparing the health basis for
non-ionizing radiation standards. Direct support to the regulatory function
of the Agency is provided in the form of expert testimony and preparation of
affidavits as well as expert advice to the Administrator to assure the
adequacy of health care and surveillance of persons having suffered imminent
and substantial endangerment of their health.
Pollution of the ambient air is complex. Many substances are emitted
into the air from a variety of sources. They differ physically, being
gases, vapors, droplets and particles of many sizes and shapes. They
also differ chemically, some being very irritating, some odorous, some
neutral. They may change considerably after they are emitted into the
air, sometimes forming other compounds. In order to understand the
relative importance of each kind of pollutant in producing effects on
health it is necessary to perform studies where that pollutant is dominant
in relation to others with which it is commonly associated. This study
attempts to examine the effects of high particulate pollution where sulfur
oxides pollution is at a low level.
Son, M.D.
Director,
Health Effects Research Laboratory
iii
-------
PREFACE
Most of this research was completed while I was a commissioned officer in
the United States Public Health Service assigned to the United States
Environmental Protection Agency. Further, the data used in this report were
collected under the auspices of the United States Environmental Protection
Agency. However, the entire report represents my own thoughts, and is not
intended in any way to represent the policy of the United States Environmental
Protection Agency.
iv
-------
ABSTRACT
Acute lower respiratory disease was surveyed by questionnaire among
parents of 10,000 children aged 1 to 12 years in two Southeastern communities
representing intermediate and high exposures to particulates and low sulfur
dioxide levels. Morbidity reporting patterns with respect to age, parental
education, and history of asthma were similar for blacks and whites, but the
frequency of pneumonia was significantly lower, and the frequencies of croup,
bronchitis, and "any lower respiratory disease" were significantly higher
among whites in both communities. Significant increases of any lower
respiratory diseases and hospitalization were found among children in the
high exposure community.
Asthma rates clustered in families, were higher in male children and
female parents, and were comparable to other studies. Significant increases
of lower respiratory disease were also found among asthmatic children in the
high exposure community.
Differences in parental recall, family size, or parental cigarette
smoking were not likely explanations for the excess morbidity in the high
exposure community. Therefore, these results associate excess acute lower
respiratory disease in children with exposure to elevated particulate levels
and low sulfur dioxide concentrations.
This report covers a period from 1960 to 1971 and work was completed
April 1976.
-------
CONTENTS
Foreword ill
Preface iv
Abstract v
Figures vii
Tables viii
Acknowledgment xiii
1. Introduction 1
2. Conclusions 3
3. Recommendations 14
4. Materials and Methods 15
Community Selection 15
Assessing Air Pollution 15
Collection of Health and Demographic Data 15
Data Analysis and Hypothesis Testing 16
5. Experimental Procedures 23
Monitoring Human Exposure to Air Pollution 23
Location and Description of Monitoring Sites 23
Pollutant Measurement Methods (CHESS) 26
Precision of Measurements 27
Quality Control 27
Long Term Exposure Trends 27
6. Results and Discussion 42
References 78
Bibliography ; 85
Appendices 86
A. Review of the Literature . 86
B. Data Analysis and Hypothesis Testing 96
C. Validity and Reliability of Disease Reporting 109
D. Observed Morbidity Rates (Tables 47-56) 112
E. Questionnaire Used in Study 123
vi
-------
FIGURES
Number Page
1 Birmingham, Alabama: Location of air monitoring
stations 24
2. Charlotte, North Carolina: Location of air monitoring
stations 25
3. Birmingham, Alabama: Total suspended particulate matter,
historical exposure. .... 30
4. Charlotte, North Carolina: Total suspended particulate
matter, historical exposure 31
5. Birminham, Alabama: Respirable suspended particulates,
historical exposure 32
6. Charlotte, North Carolina: Respirable suspended
particulates, historical exposure 33
7. Birmingham, Alabama: Sulfur dioxide, historical exposure. . . 34
8. Charlotte, North Carolina: Sulfur dioxide, historical
exposure 35
9. Birmingham, Alabama: Suspended sulfates, historical
exposure 36
10. Charlotte, North Carolina: Suspended sulfates,
historical exposure 37
vii
-------
TABLES
Number Page
1 Four Year Frequency of One or More Episodes of Each
Morbidity Condition: Model Adjusted Rates for
Children Aged 1 to 12 Years 8
2 Crude Rates for One or More Episodes of "Any Lower
Respiratory Disease" in Relation to a History of
Asthma Diagnosed by a Doctor 9
3 Chi-Square and Significance Level for Factors Affecting
Lower Respiratory Disease as Determined by a Saturated
Linear Model for Categorical Data: Black Children With
Three or More Years of Community Residence 10
4 Chi-Square and Significance Levels for Factors Affecting
Lower Respiratory Disease as Determined by a Saturated
Linear Model for Categorical Data: White Children with
Three or More Years of Community Residence 11
5 Four Year Frequency of "Any Lower Respiratory Disease"
and Bronchitis in White Children by Sex and Education
of the Head of the Household 12
6 Crude Rates for One or More Episodes of all Morbidity
Conditions, by City and Race 13
7 Four Year Reported Rates of One or More Episodes of Each
Morbidity Condition among Black Children, by Community ... 19
8 Four Year Reported Rates of Two or More Episodes of Each
Morbidity Condition Among Black Children, by Community ... 20
9 Four Year Reported Rates of One or More Episodes of Each
Morbidity Condition Among White Children, by Community
Exposure 21
10 Four Year Reported Rates of Two or More Episodes of Each
Morbidity Condition Among White Children, by Community
Exposure 22
11 Birmingham Historical Exposure (1960-1971) 38
viii
-------
12 Charlotte Historical Exposure (1960-1971) ........... 39
13 Total Suspended Particulates
CHESS Equilvalent Exposure
Birmingham, Alabama ..................... 40
14 Total Suspended Particulates
CHESS Equivalent Historical Exposure
Charlotte, North Carolina .................. 41
15 Estimated Pollutant Exposure Levels in Charlotte,
North Carolina and Birmingham, Alabama ........... 52
16 Total Number of Questionnaires Distributed and Response
Rate Among Study Families .................. 53
17a Children Aged One to Twelve Excluded Due to Missing
Information ......................... 54
17b Children Aged 1 to 12 Years Excluded from Analysis
Because of Missing Information ............... 55
18a Duration of Residence, Education of Fathers, Presence of
Parents or Guardians and Parental Smoking Habits by
City and Sex ........................ 56
18b Duration of Residence, Education of Fathers, Presence of
Parents or Guardians and Parental Smoking Habits Among
Study Families With One or More Children with a History
of Asthma .......................... 57
19a Maternal Age and Household Characteristics of Study Females,
by City and Race ...................... 58
19b Maternal Age and Household Characteristics of Study Families
With One or More Children With a History of Asthma ..... 59
20 Children Without a History of Asthma, By City, Race, Sex,
and Age ........................... 60
21 Children With a History of Asthma by City, Race, Sex
and Age ........................... 61
22 History of Asthma Ever Diagnosed by a Doctor: Prevalence
in Children by Age, Sex, Race, and Community ........ 62
23 Htstory of Asthma Active During the Past Two Years:
Prevalence in Children by Age, Sex, Race, and Community. . . 63
24 Percent of Children with Active Asthma Among Children With
Asthma Ever Diagnosed by a Doctor, by Age, Sex, Race,
and Community ........................ 64
ix
-------
25 History of Asthma Diagnosed by a Doctor: Prevalence in
Families of Elementary School Children by Race, Sex,
and Community 65
26 Asthma Ever Diagnosed by a Doctor: Prevalence in Children
in Relation to Prevalence in Their Parents 66
27 Asthma Ever Diagnosed by a Doctor: Prevalence in Children
in Relation to Asthma Activity in Their Parents 67
28 Four Year Frequency of Each Morbidity Condition by
Number of Episodes and Community: Model Adjusted
Rates for Black Children Aged 1 to 12 Years 68
29 Four Year Frequency of Each Morbidity Condition by
Number of Episodes and Community: Model Adjusted
Rates for White Children Aged 1 to 12 Years 69
30 "Any Lower'Respiratory Disease": Four Year Frequency
by History of Asthma Diagnosed by a Doctor 70
31 Croup: Four Year Frequency by History of Asthma
Diagnosed by a Doctor 71
32 Four Year Frequency by History of Asthma Diagnosed
by a Doctor 72
33 Pneumonia: Four Year Frequency by History of Asthma
Diagnosed by a Doctor 73
34 Hospitalization: Four Year Frequency by History of
Asthma Diagnosed by a Doctor 74
35 Four Year Frequency of Morbidity Among Black Asthmatic
Children by History of Asthmatic Activity 75
36 Four Year Frequency of Morbidity Among White Asthmatic
Children by History of Asthmatic Activity 76
37 Chi-Square and Significance Levels: Community Differences
in Lower Respiratory Disease Among Children With a History
of Asthma and Three or More Years Residence Duration 77
B-l Chi-Square and Significance Levels for Factors Affecting Lower
Respiratory Disease as Determined by a Reduced Linear Model
for Categorical Data: Black Children with Three or More
Years of Community Residence 99
B-2 Four Year Frequency of Each Morbidity Condition by Number of
Episodes and Community: Model Adjusted Rates for Black
Children Aged 1 to 12 Years 101
-------
B-3 Four Year Frequence of "Any Lower Respiratory Disease" and
Hospitalization: Model Adjusted Rates for Black Children
Aged 1 to 12 years 103
B-4 Chi-Square and Significance Levels for Factors Affecting Lower
Respiratory Disease as Determined by a Reduced Linear Model
for Categorical Data: White Children with Three or More
Years of Community Residence. 104
B-5 Four Year Frequency of Each Morbidity Condition by Number
of Episodes and Community: Model Adjusted Rates for
White Children Aged 1 to 12 Years 106
B-6 Four Year Frequency of Croup and Pneumonia: Model Adjusted
Rates for White Children Aged 1 to 12 Years 107
a
B-7 Summary of Statistically Significant Interactions in Which
the "City/Pollution" Effect was Involved, by Morbidity
Condition and Race 108
C-l Theoretical Expected Positive and Negative Predictive Values
under Varying Sensitivity and Specificity and a True
Prevalence (Pt) of 10% or 20% Ill
D-l "Any Lower Respiratory Disease": Reported Four Year Frequency
Among Black, Nonasthmatic Children With Three or More Years
of Familial Community Residence 113
D-2 Croup: Reported Four Year Frequency Among Black, Nonasthmatic
Children With Three or More Years of Familial Community
Residence 114
D-3 Bronchitis: Reported Four Year Frequency Among Black,
Nonasthmatic Children With Three or More Years of
Familial Community Residence 115
D-4 Pneumonia: Reported Four Year Frequency Among Black,
Nonasthmatic Children With Three or More Years of
Familial Community Residence 116
D-5 Hospitalization: Reported Four Year Frequency Among Black,
Nonasthmatic Children With Three or More Years of Familial
Community Residence 117
D-6 "Any Lower Respiratory Disease": Reported Four Year Frequency
Among White, Nonasthmatic Children With Three or More Years
of Familial Community Residence 118
D-7 Croup: Reported Four Year Frequency Among White, Nonasthmatic
Children With Three or More Years of Familial Community . . . 119
xi
-------
D-8 Bronchitis: Reported Four Year Frequency Among White,
Nonasthmatic Children With Three or More Years of
Familial Community Residence 120
D-9 Pneumonia: Reported Four Year Frequency Among White,
Nonasthmatic Children With Three or More Years of
Familial Community Residence 121
D-10 Hospitalization: Reported Four Year Frequency Among White,
Nonasthmatic Children With Three or More Years of Familial
Community Residence 122
xii
-------
ACKNOWLEDGMENTS
When I first became interested in research and in epidemiology, I
dreamt of discovering things never before known. Now, years later, I still
dream of discovery, but the times I have been fortunate enough to scale a
peak and glimpse some unknown, I have realized that others before me
oft had looked out from the same place. It doesn't impair the view; it just
keeps you honest. Doctors and scientists and other grownups of that ilk
spend too little of their time thanking others who helped them learn and
grow. Too bad, because we might occasionally remember that the world was
running before we got here. It is never possible to mention everyone —
but thanks to the faculty of the Harvard School of Public Health and par-
ticularly to the members of my thesis committee, Drs. Jacob F. Feldman,
Benjamin G. Ferris, Jr., and George B. Hutchinson; to Jim Stebbings, Fred
Miller, Andy Stead, Dennis House, Kathryn McClain, and Carol Riggs of the
United States Environmental Protection Agency, and lastly to my parents,
who always helped me learn and grow.
xiii
-------
SECTION 1
INTRODUCTION
Acute respiratory diseases are the most common illnesses in children,
and those of the lower respiratory tract may well portend chronic respiratory
disease in later life.1"3 The lower respiratory tract is commonly defined as
the portion of the respiratory tract beginning at the larynx and extending out
to terminal alveoli therefore Including the trachea, bronchi, bronchioles and
lung parenchyma, and stroma. In the Harvard longitudinal studies of child
health and development, 83% of all illnesses experienced from birth to 18 years
of age were respiratory tract infections. ** In the survey of "One Thousand
Families in Newcastle-upon-Tyne," 53% of all illnesses in the first five years
involved the respiratory tract. Unlike adults, elementary school children
are generally not exposed to occupational pollution or self-pollution by
cigarette smoking. Children are also less likely to have experienced a variety
of complicated long term ambient air pollution exposures related to residential
mobility. Hence, they are an excellent group in which to study adverse health
effects associated with community air pollution.
Studies in England, Japan, and Russia have implicated sulfur dioxide
and particulates in the ambient air as a cause of increased respiratory
morbidity in children.6"10 Studies in New York and Chicago found increased
acute respiratory morbidity associated with elevated exposure to combined
sulfur oxide and particulate air pollution among parents and their children.11
Recent retrospective surveys of acute lower respiratory disease in children
living in smelter communities in Utah and the Rocky Mountains found excess
bronchitis and croup, but not pneumonia or hospitalization, associated with
elevated sulfur oxide exposures lasting three or more years.12?13 None of
these studies was able to distinguish the effects of individual air pol-
lutants such as sulfur dioxide or particulate matter from the more complex
urban mixtures. A more detailed review of the relevant literature is appended
(Appendix A).
Since air pollution control technology is often directed towards single
pollutants, it is especially important to disentangle the effects of
exposure to multiple pollutants. For example, none of the above studies
could assess the health effects of exposure to total suspended particulates
in the presence of relatively low levels of sulfur dioxide. Yet such an
ambient pollutant pattern is typical of industrial Birmingham, Alabama, where
low sulfur coal has been used for many years. This report describes a
retrospective survey of frequency of acute 'lower respiratory illnesses among
elementary school children in two Southeastern U.S. communities. The primary
stucjy hypothesis was that reported respiratory morbidity rates would be
higher in Birmingham, the high pollution exposure community, than in
Charlotte, the intermediate exposure community.
1
-------
Children with asthma are known to have a high frequency of lower respi-
ratory tract infections.14 In two community morbidity surveys in the western
United States,12'13 children with a history of asthma reported more than
twice as much bronchitis, croup, pneumonia, and hospitalization for any of
these diseases when compared to nonasthmatic children. Furthermore, croup,
bronchitis, and pneumonia rates were significantly increased among asthmatic
children residing in the communities with higher exposures to sulfur oxides,
suspended sulfates, and particulate matter.
Bronchial asthma is characterized by periodic attacks of obstructive
expiratory dyspnea of variable severity, duration and frequency.15 Pathophysi-
ologically, recent work suggests obstruction of large as well as small airways
and that the disease is not solely immunologic in nature.16 Persons with
asthma were affected much more frequently than nonasthmatics during the acute
smog episodes of Donora and London.17*1® Much evidence since then has
associated increased attacks of asthma with exposure to ambient air pollu-
tion.19'20 Hospitalization for asthma among children under 15 years of age
was found to be related to particulate air pollution exposure in Erie County,
New York.21 A more detailed review of the relevant literature is cited in
Appendix A.
Excessive acute lower respiratory disease in nonasthmatic children has
been shown to be associated with exposure to total suspended particulate matter
in the presence of low sulfur dioxide levels in a recent study in Birmingham,
Alabama, and Charlotte, North Carolina.22 This report presents a detailed
account of the prevalence of asthma in families and the frequency of acute
lower respiratory disease in children with a history of asthma. The primary
hypotheses were (1) reported morbidity rates would be positively related to
a history of asthma in children, and (2) reported morbidity rates among
asthmatic children would be highest in Birmingham, the higher pollution
exposure community.
-------
SECTION 2
CONCLUSIONS
In this study, reported acute lower respiratory disease in children was
found to be related to elevated total suspended particulate exposure, to a
history of asthma, to the education of the head of the household, and to race.
This study has suggested that exposure to elevated concentrations of total
suspended particulate matter and suspended sulfates in the presence of
extremely low sulfur dioxide concentrations does indeed increase the risk
of acute lower respiratory disease in children. Morbidity excesses in
Birmingham were found among both black and white children without a history
of asthma. Among black children, significant increases in Birmingham were
found for one or more episodes of "any lower respiratory disease", croup,
pneumonia, and hospltalization, but not bronchitis (Table 1). In general,
two or more episodes of all morbidity conditions did not differ significantly
by community among black children. Among white children, statistically
significantly increases in Birmingham were found for one or more episodes
of all reported morbidity conditions (Table 1). Similar results were found
for two or more episodes of all morbidity conditions. No statistically
significant differences were found in which morbidity rates in Charlotte
exceeded those in Birmingham among either black or white children without a
history of asthma. These results appear strengthened by the finding of
decreased pulmonary function among black and white school children in Birming-
ham compared to those in Charlotte.23
Although one would certainly expect asthmatic children to be at least as
sensitive to air pollution as nonasthmatic children, the findings with regard
to air pollution were less clear for asthmatic children. Among children with
a history of asthma, morbidity rates were higher in Birmingham in half of the
comparisons studied. Croup rates were significantly increased in Birmingham
among black and white children although those for bronchitis (0.10>p>0.05)
were higher in Charlotte among black children.
Several factors may have been related to the weaker intracommunity
differences found among asthmatic children. First, other environmental
exposure factors than air pollution, such as temperature, dusts, pollens,
and other allergens were not measured in this study. Although these factors
have been associated with increased acute asthmatic attacks, and not increased
acute lower respiratory morbidity, they may also increase the risk of lower
respiratory disease in asthmatic children. If the latter is true, the
children's exposure to these factors was not estimated and may have been
quite different from their estimated exposure to ambient air pollution. A
second and more conjectural reason is related to the relatively high frequency
of respiratory morbidity among all asthmatic children regardless of community
-------
or race, viz., the risk associated with a history of asthma may be a much
stronger determinant for lower respiratory disease than exposure to community
air pollution. If this were true, the effects of air pollution upon lower
respiratory disease would be relatively less, and more difficult to determine
statistically. There is some evidence for this hypothesis in that the rela-
tive black/white differences generally were least among asthmatic children
(Table 2). A third factor is the considerably reduced statistical power to
determine any true intracommunity differences which is due to the small
sample size of the children with a history of asthma.
Respiratory morbidity risk was found to be related to a child's history
of asthma, as expected. In general, all respiratory morbidity was lowest in
children without a history of asthma, intermediate in children with asthma
diagnosed by a doctor, but presently inactive, and highest in children with
a history of presently active asthma for both blacks and whites (Table 2).
When compared to nonasthmatic children, those with active asthma reported
from 2.5 to almost 4 times as much "any lower respiratory disease." Further-
more, a much higher proportion of asthmatic children had repeated episodes
of respiratory moribidty. These results provide quantitative estimates for
common clinical experience. Indeed, they underscore the recognized need for
vigorous and vigilant medical care for asthmatic children.
Education of the head of the household and sex, in addition to age and
air pollution exposure, were found to be determinants of morbidity reporting.
"Any lower respiratory disease" and bronchitis were reported more frequently
among white males, but no statistically significant differences with regard
to sex were found among black children. "Any lower respiratory disease" and
bronchitis were found to be increased in both black and white children from
households with a high school or greater education. Two or more episodes of
croup did not vary significantly by education of the head of the household.
It was not possible to properly interpret the statistical relationships of
one or more episodes of croup, pneumonia, and hospitalization to education
of the head of the household, per se, for blacks and whites. This was
because either a significant "city x SES" (SES, socioeconomic status) or a
"city x age x SES" interaction was found for all three of these conditions in
the saturated linear model for categorical data, and the reduced models were
directed towards examining the "city" effect rather than the "SES" effect
(cf. Tables 3 and 4).
In two other published studies which used this questionnaire, similar
results were reported, viz., statistically significant more frequent "any
lower respiratory disease," and bronchitis (and croup) in children from
households with a high school or greater education and the converse for
1 2 T 3
pneumonia and hospitalization. » However, the specific form of the
linear model for categorical data used for statistcal analyses in both of
these studies was not discussed. Statistical summaries in both papers imply
the following model: city/pollution, age, sex, and SES. However, there is
some question of the validity of the previous findings regarding the relation-
ship of croup, pneumonia, and hospitalization to socioeconomic status, since
this study found several significant interactions involving city/pollution
and SES when a saturated model was used (cf. Appendix 8). Nevertheless, we
should learn more about reported lower respiratory disease morbidity in
relation to medical care availability and medical care utilization. The
-------
British-United States differences with respect to childhood lower respiratory
disease and social class suggest that differences In the availability of
medical care may be involved.
Although both sex and education of the head of the household were found
to be determinants of morbidity reporting, the effect of sex was found only
for "any lower respiratory disease" and bronchitis among white children.
However, reporting of "any lower respiratory disease" and bronchitis was
related to education of the head of the household among children of both
races. For any given morbidity condition in which they were both statis-
tically significant, the relative increase for education of the head of the
household was about the same as the relative increase for sex as evidenced
by the example in Table 5. Age and sex distributions would be most likely
to be comparable between communities when sampling through the elementary
schools. However, failure to ascertain education of the head of the house-
hold (or some other good index of socioeconomic status) could easily lead
to confounding intracommunity differences due to air pollution with those
due to socioeconomic status in a study of this type.
When compared to white children, black children reported more pneumonia,
but less of all four other morbidity conditions (Table 6). This was true
for both communities as well as for all categories of asthma history. A
study in New York City using a similar questionnaire has confirmed these
findings.21* Yet in both this and the New York study, excess respiratory
morbidity in black or white children was associated with exposure to sulfur
oxide and particulate matter despite intracommunity black/white morbidity
differences. In addition to the effects of air pollution, respiratory
morbidity in children of both races showed similar patterns with regard to
age, sex, education of the head of the household, and a history of asthma.
Differences in medical care utilization could explain, in part, the
observed differences in morbidity reporting by race. One may assume that
sicker children are brought to physicians more frequently by their parents
and this would be largely independent of other determinants such as race.
In this study, the facts that black/white morbidity differences were least
among children with active asthma, and that pneumonia (a serious illness and
often a consequence of prior, unattended, milder upper or lower respiratory
disease) was reported most frequently among blacks support this assumption.
Further, if the black/white morbidity differences are greatest among less
seriously ill children, this suggests that the observed racial differences
would be most easily explained by differences in cultural and socioeconomic
factors, rather than genetic factors, per se. At any rate, the reasons for
the observed black/white morbidity differences appears to be a fertile area
of research from both the scientific and public health viewpoint.
The questionnaire used in this study has been used twice previously in
reported studies.12*13 This study has shown that the increased risk of
acute lower respiratory disease in children is related to the history of
asthma activity as well as a history of asthma per se. Children with a
history of asthma should be considered separately when estimating the true
risk of acute lower'respiratory disease in children although their inclusion
would confound intercommunity comparisons only if asthma prevalence was
-------
considerably higher than usually observed, and in addition, higher in one
community than the other. Although several findings require further study,
the agreement of asthma prevalence in families with findings in several
other reports, the increased risk of morbidity among asthmatic children,
and the consistent relationships of lower respiratory disease morbidity
in nonasthmatic children with age, sex, and parental education found in
blacks and whites testify to the utility and reliability of this question-
naire for community surveys. Further discussion on the validity of the
questionnaire is appended (Appendix C).
It appears quite likely that elevated exposure to suspended particulate
matter (total and sulfate fraction) without concomitant sulfur dioxide
exposure is sufficient to increase childhood respiratory morbidity. All
children were affected by exposure to air pollution although the effect of
this exposure varied with race, sex, and socioeconomic status. Much of the
suspended sulfate fraction may represent sulfur dioxide which has sorbed
onto particles. In fact, it would be desirable to further characterize the
physical and chemical aspects of ambient suspended particulate matter to
gain a better understanding of the toxicology of these substances. Obviously,
the use of individual pollutant concentrations, as opposed to some product
or combination thereof, is a relatively simple way to estimate human exposure
to air pollutants. Nevertheless, one cannot deny that these simple indices
have been quite useful epidemiologically in many studies all over the world.
Epidemiologic associations with individual pollutant indices are much to be
preferred for regulatory purposes as most control strategies are directed
toward individual pollutants.
Epidemiologic studies relating health effects to long term exposure to
ambient air pollutant concentrations must make a concerted effort to estimate
past exposures in addition to a current monitoring. If ambient pollutant
levels have remained constant in an area, no error is incurred. If pollutant
concentrations have been decreasing with time, attributing the health effect
to current levels imposes an unnecessary and possibly severe economic penalty.
Conversely, when pollutant concentrations have been increasing with time,
attributing the health effect to current levels fails to fully protect the
public health. Two important questions are of related interest: First, to
what extent is exposure to Intermittent peak levels, as opposed to less
variable lower level exposures, the cause long term health affects? Studies
in smelter communities, may provide some answers since they are exposed to
frequent short term fumigations and yet they often have annual pollutant
averages near Federal standards. Second, how long does it take for excess
morbidity to decrease after pollution is controlled in a community? It will
require some time to obtain answers to both of these questions.
Respiratory diseases were a common cause of morbidity and mortality in
Colonial America and are still the most common childhood illnesses in the
United States today.25 Children experience about five or six acute respira-
tory illnesses yearly and many of these will involve the lower respiratory
tract.1»2'26'27'28 Mortality due to acute lower respiratory disease is a
serious problem in children under five years of age. »29« »31 Recent
evidence has strengthened the hypothesis that frequent episodes of lower
respiratory disease in childhood are associated with the development of
-------
chronic respiratory disease in later life. 3> 1'*»32 Although the relative risk
of childhood acute lower respiratory disease from exposure to sulfur oxides
and particulate air pollutions ranges from about 1.2 to 2.0, the attributable
risk becomes enormous when one considers the acute misery, the interference
with school (and parental work) activities, the family medical costs, the
increased burdens on the medical care system, and the real possibility of an
increased risk of chronic respiratory disease in later life for children so
exposed.
-------
TABLE 1. FOUR YEAR FREQUENCY OF ONE OR MORE EPISODES OF EACH MORBIDITY CONDITION:
MODEL ADJUSTED RATES* FOR CHILDREN AGED 1 TO 12 YEARS
00
Race
Black
Children
White
Children
Community
Charlotte
.
Birmingham
Birmingham
Charlotte
Charlotte
Birmingham
Birmingham
Charlotte
Any LRD
Female Male
16.6% 18.4%
21.0% 19.2%
1.27 1.04
28.72
33.6%
1.17
Croup
HS
9.5% 7.8%
11.8 13.4%
1.24 1.72
Second
Order
Interaction
Hospital ization
Second
Order
Interaction
HS
3.1% 2.5%
4.1% 4.8%
1.32 1.92
*Rates for nonasthmatic children with three or more years residence duration from saturated linear
model for categorical data and adjusted for variable(s) not displayed (age, sex, or education of
head of household).
-------
TABLE 2. CRUDE RATES OF ONE OR MORE EPISODES OF "ANY LOWER
RESPIRATORY DISEASE" IN RELATION TO A HISTORY OF
ASTHMA DIAGNOSED BY A DOCTOR
City
Charlotte
Birmingham
History
of Asthma
Never
Diagnosed
Diagnosed,
Inactive
Diagnosed,
Active
Never
Diagnosed
Diagnosed,
Inactive
Diagnosed,
Active
"Any Lower Respiratory
Disease", Crude Rate*
Black
16.2%
42.4%
62.8%
18.8%
48.6%
60.0%
White
27.1%
55.6%
83.5%
31.8%
47.7%
80.0%
Ratio
Black/White
1.67
1.31
1.33
1.69
0.98
1.35%
*Crude rate of children aged 1 to 12 years with three or more years
residence duration.
-------
TABLE 3. CHI-SQUARE AND SIGNIFICANCE LEVELS FOR FACTORS AFFECTING LOWER RESPIRATORY DISEASE AS
DETERMINED BY A SATURATED LINEAR MODEL FOR CATEGORICAL DATA: BLACK CHILDREN WITH THREE
OR MORE YEARS OF COMMUNITY RESIDENCE
Effect
City/Pollution (P)
Age (A)
Sex (S)
SES (E)
City x Age
City x Sex
City x SES
Age x Sex
Age x SES
SES x Sex
P x S x A
P x E x A
P x E x S
A x S x E
P x A x S x E
a - p$0.001
b - p^O.Ol
c - p^O.05
d - 0.10>p>0.05 -
Degrees
of
Freedom
1
2
1
1
2
1
1
2
2
1
2
2
1
2
2
Any
*1
0.79
35.14a
<0.01
5.53°
3.33
3.56d
0.05
0.91
3.49
0.29
4.22
0.04
<0.01
2.63
2.20
For each term in
significance.
•*. .
LRD
*2
0.09
16.343
0.49
8.55b
7.30C
5.79C
4.94°
1.46
5.95C
0.02
2.72
1.17
0.03
0.46
0.50
the model
Croup
*1
0.02
13.47b
0.12
2.88d
1.11
2.23
2.72d
0.45
1.00
0.31
4.27
0.28
0.02
0.75
1.71
, the
*2
0.01
3.39
0.34
1.08
7.49C
1.43
0.14
1.08
2.21
0.16
0.57
2.12
1.53
0.76
1.05
Bronchitis
*1
0.60
13.81a
0.96
8.76b
1.72
1.63
0.96
1.32
3.71
1.81
1.71
2.12
0.37
0.63
0.16
probability for
X
0.71
8.09C
0.06
3.57d
2.30
0.73
0.04
5.48d
3.02
1.90
1.96
1.31
0.02
2.96
2.33
Pneumonia
*1
8.12b
6.92a
1.09
0.56
1.76
0.96
3.20d
1.56
4.07
0.46
0.85
1.35
0.01
2.48
0.97
a" two- tailed test
#
0.99
1.58
0.01
1.89
3.67
0.73
1.07
0.32
0.77
0.46
0.79
0.18
0.08
0.11
0.81
Hospitalization
2
2.
12.
0.
0.
0.
0.
0.
6.
0.
0.
0.
6.
1.
4.
1.
:1 £2 .
50
00b
53
26
Rate
15 too
48 low
41 to
96C fit
05 model
64
69 -
36C
95
87d
56
of statistical
-------
TABLE 4. CHI-SQUARE AND SIGNIFICANCE LEVELS FOR FACTORS AFFECTING LOWER RESPIRATORY DISEASE AS
DETERMINED BY A SATURATED LINEAR MODEL FOR CATEGORICAL DATA: WHITE CHILDREN WITH THREE
OR MORE YEARS OF COMMUNITY RESIDENCE
-
Degree
Effect of
Freedo
City/Pollution (P)
Age (A)
Sex (S)
SES (E)
City x Age
City x Sex
City x SES
Age x Sex
Age x SES
SES x Sex
P x S x A
P x E x A
P x E x S
A x S x E
P x A x S x E
1
2
1
1
2
1
1
2
2
1
2
2
1
2
2
s Any
m ^1
9.91b
56.52a
4.40C
8.43b
0.61
1.21
<0.01
4.18
2.32
0.27
0.72
2.08
0.24
0.71
4.74d
LRD
»2
10.99a
31.593
4.83°
4.44C
0.06
1.51
0.01
2.41
0.43
1.77
0.53
1.56
0.75
4.28
3.56
Croup
*1
0.74
22.83a
0.65
3.57d
2.96
0.32
0.14
4.55
4.24
2.23
0.89
7.78C
0.24
2.13
5.40d
*2
4.59C
7.08C
0.09
0.71
4.09
0.43
0.09
4.40
2.18
0.21
1.03
1.96
2.12
1.80
5.70d
Bronchitis
*1
11.99a
39.69a
6.24b
12.543
0.71
3.21d
2.11
1.90
0.81
0.04
0.47
4.16
1.11
1.29
2.60
*2
24.18a
18.103
8.40b
7.22b
1.37
2.04
0.03
3.03
0.47
0.35
2.33
0.76
0.58
3.51
2.32
3.
10.
0.
0.
0.
0.
0.
1.
1.
2.
4.
8.
1.
3.
2.
Pneumonia
1 *2
43d
95b
01
97
Rate
32 too
91 low
09 to
74 fit
24 model
12
48
37C
25
07
41
Hospitalization
*
5.53C
36.61a
0.77
8.15b
3.48
0.71
3.88C
1.22
3.44
0.07
2.54
2.59
0.49
0.39
1.59
*
Rate
too
low
to
fit
model
a - p<0.001
b - .p<0.01
c - p<0.05
d - 0.10>p>0.05
For each term in the model, the probability for a two-tailed test of statistical
significance.
-------
TABLE 5. FOUR YEAR FREQUENCY OF "ANY LOWER RESPIRATORY
DISEASE" AND BRONCHITIS IN WHITE CHILDREN BY
SEX AND EDUCATION OF THE HEAD OF THE HOUSEHOLD
Morbidity Determinant
Sex
Education, Head
of the Household
Female
Male
Male
Female
HS
>HS
Model Adjusted
$2 "Any LRD" ^
17.3%
19.1%
1.10
16.8%
19.6%
1.17
Rates*
I Bronchitis
9.9%
12.1%
1.22
9.2%
12.7%
1.38
*Age-city-education head of household (for sex) or age-city-sex
adjusted rates for nonasthmatic children with three or more
years residence duration, from saturated linear model for
categorical data (Table 4).
12
-------
TABLE 6. CRUDE RATES FOR ONE OR MORE EPISODES OF ALL MORBIDITY CONDITIONS,
BY CITY AND RACE*
Morbidity Condition
"Any lower respiratory
disease"
Croup
Bronchitis
Pneumonia
Hospital ization
City
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Crude
Black
16.2%
18.8%
6.5%
6.9%
7.9%
7.8%
8.4%
12.2%
1.9%
3.2%
Rates
White
27.1%
31.8%
12.4%
14.4%
18.4%
23.5%
6.3%
7.7%
2.3%
4.5%
Ratio
White/Black
1.67
1.69%
1.91
2.09
2.33
3.01
0.75
0.63
1.21
1.41
*Restricted to children without a history of asthma and with three or more years
of community residence.
-------
SECTION 3
RECOMMENDATIONS
Very little research has been done which makes fine distinctions of the
effects of single pollutants on the lower respiratory tract. This study was
to gather data on lower respiratory infection in children exposed to suspended
particulates in the presence of low levels of sulfur oxides.
The reasons for Increased morbidity rates for whites, except for
pneumonia (as shown in Table 6) should receive further study to determine
valid explanations. Also, the chemical, physical and toxicological properties
of the ambient suspended particulate matter should receive further study.
14
-------
SECTION 4
MATERIALS AND METHODS
Community Selection
Two communities in the southeastern United States were selected on the
basis of past air quality data and historical information regarding pollutant
emissions. The communities were ranked intermediate or high on the basis
of the estimated exposure to total suspended particulates during the period
covered by the study and for the prior decade. The exposure rankings were
relative for the two cities and not quantitatively related to the current
U.S. National primary standard for particulate matter. Charlotte, the inter-
mediate exposure community with a population of over 240,000, is a growing
commercial and light industrial center in the Piedmont region of North
Carolina. Located in Mecklenberg County, its major industries include
chemicals, textiles, fabricated metals, machinery, wood, foundries, cement,
and asphalt. When compared to Birmingham, Alabama, Charlotte had both
lower emissions of particulates and less frequent temperature inversions.
Birmingham, Alabama, the high exposure community, with a population of over
300,000, is one of the nation's leading industrial centers. It is, perhaps,
best known for its steel production and manufacturing. However, it also
produces fabricated metals, transportation equipment, machinery, stone, clay,
glass, and wood products. Industrial plants, located throughout the city,
are surrounded by commercial and residential areas.
Assessing Air Pollution Exposure
At the time of this study, air monitoring stations were located within
each community within 1 1/2 to 2 miles of the study population. All stations
had their air inlet six feet above the ground except for one which was 16
feet above the ground. At each station, 24-hour integrated samples of sulfur
dioxide (modified West-Gaeke method), total suspended particulates (high-
volume samplers), suspended sulfates, and suspended nitrates were monitored
on a daily basis. Dustfall was determined from monthly samples. Estimates
of past exposure were derived from previous aerometric data and emissions
data. A full description of the location of monitoring stations and the
methodology for current data collection and estimation of past exposures has
been presented elsewhere (available from EPA upon request).33 Section 5
provides further detail regarding air pollution exposure.
Collection of Health and Demographic Data
In November 1971, elementary school children (grades 1 to 6) in selected
study communities in each city were asked to take explanatory letters and
questionnaires home to their parents. Socioeconomically similar school
15
-------
districts in each city were selected by consultation with the Superintendent
of Schools and school principals. Children in eleven schools in Birmingham
and twelve schools in Charlotte were surveyed. The frequency of acute lower
respiratory disease was ascertained by means of a School and Family Health
Questionnaire, which was completed by the mother or female guardian in each
family when possible and returned to the school (Appendix E). Mothers
were asked to answer the questions about lower respiratory disease for all
children twelve years of age or younger who lived in the household.
The questionnaire inquired about the frequency of treatment by a physi-
cian for pneumonia, croup, or bronchitis (including bronchiolitis or deep
chest infections other than pneumonia or croup) during the period beginning
in September 1967 and continuing through the time of questionnaire completion.
Hence, all morbidity rates presented encompass a time period of just over
four years. Other information ascertained included hospitalizations for
lower respiratory illnesses, name of children's current physician, history
of asthma diagnosed by a doctor, length of residence in the community, number
of living quarter changes, education of head of household as an index of
socioeconomic status, parents' smoking statuses, race, family census, number
of rooms per household and presence and type of air conditioning per house-
hold. In Birmingham, respondents with telephones were called if any informa-
tion was missing from their questionnaire. In Charlotte, respondents with
telephones were not called if the following information was missing: educa-
tion of head of household, race, number of living quarter changes, number
of rooms per household, presence and type of air conditioning, history of
asthma diagnosed by a doctor and age of mother or female guardian; they
were called if any of the other information was missing.
Data Analysis and Hypothesis Testing
As planned in the first part of this protocol, children with a history
of asthma, and all children with less than three years residence in the
community, were excluded from the analysis. Children with asthma are known
to have a higher risk of respiratory illnesses than nonasthmatic children.
Asthma history was determined by "yes" or "no" answers to the following two
questions: (1) "Has this person ever had asthma diagnosed by a doctor?"
and (2) "Has this asthma been active in the past two years?" As a matter
of scientific interest, these differences were studied and the results are
shown separately. 22»3lt'35 Children with less than three years residence were
excluded for several reasons. They had been exposed to the community air
for only a short time and their previous residences and pollutant exposures
were not known. Moreover, illness itself has been associated with migration,
and recently migrated families would be less likely to have established
patterns of medical care within a community, possibly obscuring the effect
of pollution on reported doctor-diagnosed illnesses within a community. (For
blacks, morbidity rates were generally higher among recent migrants; for
whites morbidity rates were generally higher among residentially stable chil-
dren. This study was not designed to explain these differences.) Children
under one year of age were excluded because of an error in coding instructions
which made children under one year indistinguishable from children with
missing information on age.
16
-------
Actual or direct-adjusted rates are presented for descriptive purposes,
for comparison to other published data and for comparison to the model-
adjusted rates derived from the linear categorical model described below.
Descriptive morbidity rates (as in Tables 7-10) were direct-adjusted
according to standard methods, the reference populations being all non-
asthmatic children with three or more years residence.36 City question-
naire return rates in Table 2 were tested by standard contingency table
techniques.3?
The primary test hypothesis was that the frequency of reported physician
treatment for "any lower respiratory disease," croup, bronchitis, pneumonia
and hospltalization, appropriately adjusted, would correspond to the gradient
in pollutant exposures, viz. that respiratory morbidity and related hospitali-
zation rates would be higher in Birmingham than in Charlotte. The combined
disease category, "any lower respiratory disease," was constructed to include
a child if he had either pneumonia, croup, or bronchitis or any other deep
chest infection. Hence, the category "any lower respiratory disease (any
LRD)" is an index of overall frequency of acute lower respiratory disease
in the community without regard to specific diagnostic categories. For each
of the five reported conditions, two sets of analyses were done. In the
first, the dependent variable was the percent of children reporting one or
more episodes of each morbidity condition. In the second, the dependent
variable was the percent of children reporting two or more episodes of each
condition.
For nonasthmatic children, each specific hypothesis was tested statis-
tically in a general linear model for categorical data.38 An alpha proba-
bility of p<0.05 was chosen as statistically "significant." All statistical
tests were "two-tailed" tests, i.e., for any given term in the model, such
as "city/pollution," the resulting chi-square and associated probability
were for excesses in either city as large or larger than those observed,
given that there is no difference. The general linear model for categorical
data technique utilizes weighted regression on categorical data and allows
estimation of each individual factor adjusted for all other factors in the
model.
For asthmatic children, each specific hypothesis was tested by standard
contingency table techniques.37*39 An alpha probability of p < 0.05 was
chosen as statistically "significant." All statistical tests were "two-
tailed" tests, i.e. the resulting chi-square and associated probability were
for excesses in either city as large or larger than those observed, given
that there is no difference. Morbidity rates in children with a history of
asthma active within the past two years ("active") were generally higher
than those with a history of asthma, but not active within the past two years
("inactive"). Therefore rates for each group were compared separately between
cities, e.g. morbidity rate for Inactive asthmatic children in Charlotte to
the comparable rate for inactive asthmatic children in Birmingham. Because
or the considerably smaller sample sizes of both categories of asthmatic
children it was not possible to adjust their morbidity rates for age, sex,
and education of the head of the household (SES) while testing for "city"
differences. Hence the crude rates of each morbidity condition among children
aged 1-12 years were used for statistical testing. This was not unreasonable
17
-------
since active and inactive children were comparable with regard to age, sex,
and SES distributions. Likewise, the crude morbidity rates of nonasthmatics
were used for descriptive purposes unless otherwise noted.
Morbidity conditions were analyzed in a saturated analysis of variance
(ANOVA) form of the linear model, namely, four main effects: city/pollution
(P), age (A), sex (S), education of head of household (E), and all possible
interactions (six first-order, four second-order, and one third-order). For
each morbidity condition analyzed in the model, age (1 to 4, 5 to 8, 9 to 12),
sex (F-female; M-male), and education of the head of the household (
-------
TABLE 7. FOUR YEAR REPORTED RATES OF ONE OR MORE EPISODES OF EACH
MORBIDITY CONDITION AMONG BLACK CHILDREN, BY COMMUNITY
Morbidity Condition Community
Direct Adjusted* Age-Specific Rates, %
1-4 Years 5-8 Years 9-12 Years
"Any Lower
Respiratory Disease"
Croup
Bronchitis
Pneumonia
Hospitalization
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
27.8
24.0
10.8
8.6
13.7
10.6
14.0
15.1
4.0
5.7
16.4
20.6
7.0
7.9
7.8
7.9
8.3
13.7
1.8
2.9
12.7
15.9
4.7
5.9
6.3
6.9
8.9
10.2
1.3
2.5
*Direct adjusted for sex and education of head of household. Because age
distributions within the 1-4 year old age groups were similar in both
communities, these rates were not adjusted for differences in the number
of years at risk among 1 to 3 year olds.
19
-------
TABLE 8. FOUR YEAR REPORTED RATES OF TWO OR MORE EPISODES OF EACH
MORBIDITY CONDITION AMONG BLACK CHILDREN, BY COMMUNITY
Morbidity Condition Conmunity
Direct Adjusted* Age-Specific Rates, /
1-4 Years 5-8 Years 9-12 Years
"Any Lower
Respiratory Disease"
Croup
Bronchitis
Pneumonia
Hospital ization
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
15.9
12.3
4.8
2.1
6.2
3.8
6.1
4.5
0.3
0.6
7.3
11.6
1.1
3.0
2.5
3.0
3.0
5.6
0.3
0.5
7.4
7.9
1.2
2.4
2.2
2.4
3.4
4.6
0.4
0.1
*Direct adjusted for sex and education of head of household. Because age
distributions within the 1-4 year old age groups were similar in both
communities, these rates were not adjusted for differences in the number
of years at risk among 1 to 3 year olds.
20
-------
TABLE 9. FOUR YEAR REPORTED RATES OF ONE OR MORE EPISODES OF EACH
MORBIDITY CONDITION AMONG WHITE CHILDREN, BY COMMUNITY EXPOSURE
Morbidity Condition Community
Direct Adjusted* Age-Specific Rates, %
1-4 Years 5-8 Years 9-12 Years
"Any Lower
Respiratory Disease"
Croup
Bronchitis
Pneumonia
Hospital ization
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
35.0
38.9
17.5
16.3
23.1
28.7
9.0
10.3
5.5
5.8
29.9
36.3
14.2
16.3
20.4
26.2
6.3
8.5
2.7
6.1
22.0
26.4
9.3
12.2
14.9
18.6
5.2
6.3
0.9
2.4
*Direct adjusted for sex and education of head of household. Because age
distributions within the 1-4 year old age groups were similar in both
communities, these rates were not adjusted for differences in the number
of years at risk among 1 to 3 year olds.
21
-------
TABLE 10. FOUR YEAR REPORTED RATES OF TWO OR MORE EPISODES OF EACH MORBIDITY
CONDITION AMONG WHITE CHILDREN, BY COMMUNITY EXPOSURE
Morbidity Condition Community
Direct Adjusted* Age-Specific Rates, /
1-4 Years 5-8 Years 9-12 Years
"Any Lower
Respiratory Disease"
Croup
Bronchitis
Pneumonia
Hospital ization
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
19.0
23.5
7.1
4.3
8.8
16.4
2.4
2.8
1.0
0.6
17.9
22.3
6.4
8.9
10.1
14.5
2.1
2.8
0.2
1.5
11.8
16.5
3.8
7.4
6.7
12.6
1.5
2.3
0.5
0.2
*Direct adjusted for sex and education of head of household. Because age
distributions within the 1-4 year old age groups were similar in both
communities, these rates were not adjusted for differences in the number
of years at risk among 1 to 3 year olds.
22
-------
SECTION 5
EXPERIMENTAL PROCEDURES
Monitoring Human Exposure to Air Pollutants
Communities were selected within the metropolitan areas of both
Birmingham and Charlotte to document exposure within each of three study
sectors in each city. Within each study sector, monitoring stations were
located within 1 1/2 to 2 miles of the study population. This study was
part of the U. S. Environmental Protection Agency's CHESS (Community Health
and Environmental Surveillance) Program. CHESS monitoring began in Birmingham
in November 1969 and in Charlotte in February 1970. In each of the three
study sectors selected in each city, monitoring sites were chosen to be most
representative of the immediate area in which the study population resided.
Each monitoring station was positioned in an area free of major obstructions
to air flow and removed from the biasing effect of point sources insofar as
possible. Except for Sector III in Charlotte (station located on a flat roof
16 feet above ground level) air sample inlets were positioned at head level
(6 feet) to further represent respirable exposure.
Topography, prevailing weather patterns, and local or point pollutant
sources determine the representativeness of any single sampling site as an
indicator of air pollution exposure for a given area surrounding it. CHESS
monitoring sites were placed in relatively flat areas free from point sources
of pollution, e.g., a cement factory. Hence they are representative, as
intended, of the air pollution exposure within a 1 1/2 to 2 mile radius of
their location. The assumption is made that the study subjects spent most
of their time within this radius, which is a reasonable assumption for
children through age twelve (the assumption is less reasonable for children
not attending neighborhood schools or for adults who may be more likely to
travel out of the neighborhood during the day). Historical exposure estimates
for both cities were usually based on more than one station from 1964 onward
and provide a reasonably integrated estimate of long-term pollution exposure
for residents of each city.
Location and Description of Monitoring Sites
The location of CHESS monitoring stations are given in Figure 1 for
Birmingham and Figure 2 for Charlotte. These sites are prefaced by "C" to
distinguish them from other sites which existed prior to CHESS monitoring
but which were used for historical exposure estimates. National Air Sampling
Network (NASN) sites are prefaced with "N", Jefferson County stations with
"J", and Mecklenburg County stations with "M" to identify the respective
Birmingham and Charlotte sites.
23
-------
to
N.11
A
N. 12
S. 1-
J.2
SECTOR I
BIRMINGHAM, ALABAMA'
LL
1 2 3
I I I
MILES
J(A)8
N.9
>
<^3-<
A
N.2
J.3
S.3
A^i—
/
,• _, •
/S—
Vx
Q SECTOR II
S.4
J.4
SECTOR III
o<
N.4
F.5 *•
A N. IS
"'*.."
•A
J. 11
N. 13
S.8
C.I
SITE IDENTIFICATION
S. • FEDERAL STUDY
N. A NASN
C. Q CHESS
J. • JEFFERSON
COUNTY
Figure 1, Birmingham, Alabama: location of air monitoring stations.
-------
CHARLOTTE. NORTH CAROLINA
N8
O
CHESS
MECKLENBURG COUNTY
NASN
Figure 2. Charlotte, North Carolina: location of air monitoring stations.
25
-------
CHESS measurements for respirable suspended particulate (fine particu-
lates), total suspended particulates, suspended sulfates, suspended nitrates,
and sulfur dioxide are taken daily for each 24-hour period. The 24-hour
period normally begins in the morning between 8:00 a.m. and 12 noon.
Pollutant Measurement Methods (CHESS)
Total Suspended Particulate Matter (TSP) - The high volume (hi vol) air
sampler mounted in a shelter of standard design was employed for the collec-
tion of atmospheric particulate matter for gravimetric and chemical analysis.
Samples were collected on a 20.3 X 25.4 cm glass fiber filter. At an air
flow rate of approximately 1.7 m3 per minute, particles ranging in sizes from
0.1 to 90 microns were collected. The high volume assembly was routinely
inspected, overhauled, and calibrated (with its attached rotometer) every 25
calendar days. Weight of particulate was determined by the EPA standard
reference method. The results are given in micrograms per cubic meter
(yg/m3). A portion of the particulate matter collected was used to deter-
mine suspended sulfate using turbidimetric methods and suspended nitrate
was determined by a reduction diazo coupling reaction and detecting the
color change by automated analysis.
Fine Particulates (Cyclone Separator Method) - For estimating the con-
centration of fine particulates that correspond to respirable suspended
particulates (RSP), a small stainless steel cyclone, 1.27 cm in diameter,
was used. The cyclone operation is such that two fractions are produced:
The heavier, larger particles fall into a plastic cup, the smaller respirable
fraction is drawn up into the air vortex and impinges on a 37 mm glass fiber
filter mounted in a plastic cassette between the cyclone and the vacuum pump.
In a similar fashion, an open-faced cassette is used in parallel to collect
the total suspended particulate (TSP). The cyclone and open-face cassette
each has a critical orifice placed between the vacuum pump and the filter
cassette for sample flow control.
Fine Particulates (Five Stage Cascade Impactor) - A five-stage cascade
impactor was used for detailed particulate monitoring. Particles were
separated from the airstream by inertial impaction on glass fiber "filters"
into five aerodynamically sized fractions: above 5.50 microns, 2.40 to
5.50 microns, 1.75 to 2.40 microns, 0.93 to 1.75 microns, and 0.01 to 0.93
microns. The impactor was mounted in a standard high volume shelter and
operated at a sampling flow rate of 566 liters per minute.
Suspended Sulfate (SS) - Twenty-four suspended sulfate measurements were
made from CHESS high volume particulate samples. A 1.9 cm x 20.3 cm strip of
the exposed high volume filter was refluxed and the sulfate ion concentrations
determined by spectrometric methods.
Suspended Nitrate (SN) - Twenty-four hour suspended nitrate measurements
were made from CHESS high volume particulate samples using a strip of the
exposed filter and spectrometric methods similar to the sulfate analysis above,
26
-------
Sulfur Dioxide (SOz) - Twenty-four S02 measurements were made according
to the West-Gaeke reference method as published in the Federal Register with
minor changes in air flow rate and absorbing solution volume.
NASN Data - The CHESS program supplements its data collection with data
from the National Air Sampling Network (NASN) whenever possible.
Precision of Measurements
The CHESS program conducted tests to determine the precision of environ-
mental measurements. These tests were done in Birmingham, Charlotte and
Greensboro, North Carolina. Duplicate sensors were installed at air monitoring
sites within these cities on a daily basis for 8 months in Birmingham, and
for 6 months in Charlotte. The comparison of the regular samples to duplicates
was used to determine statistically the precision of CHESS measurements for
TSP and S02. The arithmetic mean errors for TSP and S02, were 6.1% and 27.1%
respectively.
Quality Control
An effective quality control program is an essential element of any
environmental surveillance system, since the data outputs of an air sampling
program are subject to'any sources of error. Effective, real time controls
are essential to minimize field errors, systematic drift, chemical labpratory
errors, data transfer errors, computer punch card errors, analysis errors,
etc. A quality control program was used for the CHESS data in this study,
but not for the earlier data.
Long Term Exposure Trends
Because CHESS monitoring was designed to be coupled with epidemiological
health studies and because CHESS sites were specifically selected to repre-
sent the study area, CHESS data were used as the standard for this report.
Equivalency adjustment of non-CHESS data obtained from different locations
used in prior years has been made only to present an appropriately scaled
trend of past exposure. In many instances historical data were unavailable
for some pollutants of interst, or were available in different forms, or
represented an area outside the CHESS study area. The annual standard for
TSP is a geometric mean rather than an arithmetic mean like those for sulfur
dioxide and nitrogen dioxide. This decision was based on the distribution
of TSP values and not health data. Since virtually all ambient pollutant
concentrations are positively skewed, the use of a geometric mean only for
TSP is inconsistent. Certainly, the use of a geometric mean to estimate
human exposure, minimizes differences in peak episodic exposures between
communities. Nevertheless, this is the reason for the use of a geometric
mean of TSP as well as an arithmetic mean in this report.
Relationships between RSP and TSP were analyzed using current data and
have been applied to historical data. On the basis of the correlations
presented, this approach provided reasonable estimates for specific pollutants
for the years during which measured data were not available. Since respirable
suspended partlculates were not collected routinely prior to CHESS studies,
27'
-------
all RSP values prior to 1969 are estimates based on collections of TSP.
Because of the special complexities of historically estimating RSP exposure
and the resulting uncertainties these estimates are presented here, but not
in the body of the report. Although measured RSP data are not available prior
to 1969, it is possible to estimate RSP directly from TSP under certain
conditions which are discussed fully in the report of Hinton, et al.
Comparisons of measured data from cyclone RSP and high volume TSP were
made for both Birmingham and Charlotte during the period in which the cascade
impactor was being evaluated. It was found that as TSP concentrations increase,
RSP concentrations do not increase proportionately but rather at constantly
reducing proportions of the total, i.e. the relationship of RSP to TSP was
that of a tnonotonically increasing function with decreasing slope.
Graphical analysis of the concentrations within each size range as a
function of relative humidity showed that while the slopes vary, concen-
trations decrease with increased humidity. It is postulated that this is
due primarily to the aerodynamic separation technique. As explained pre-
viously, the highly humid particle would be collected as though it were a
larger particle with a smaller percentage reaching either the filter or the
human lung. (Although laboratory equilibration of the filter may remove a
portion of the moisture further reducing the weighed mass, this effect would
apply equally to both RSP and TSP.) This partial explanation indicates the
need for further study. Other factors such as either vertical or horizontal
wind dispersion, and the composition of total particulates would slightly
affect the ratio of RSP to TSP.
Cyclone RSP is the monitoring method for the respirable fraction upon
which the CHESS study was based. Impactor data are available for only a
short period during the CHESS study; therefore, cyclone RSP to hi vol TSP
ratios were used to provide a basis for estimating prior RSP concentrations
from hi vol TSP exposure.
Within the accuracies of past high volume measurements, reasonable
estimates of past RSP exposure may be obtained by multiplying the TSP value
by the following percentage factors:
TSP Range in yg/m3 Percentage RSP
Birmingham 80-100 59.0
100-120 46.0
120-140 43.0
140-160 37.5
Charlotte 60-80 55.0
80-100 47.5
Estimates of RSP exposure obtained in this manner are given in Tables 11
and 12 for 1960 through 1971.
28
-------
Simultaneous collections of data from the cascade impactor sampler, the
RSP cyclone and the high volume sampler were analyzed and it was found that
when the appropriate percentage RSP was applied to the TSP data of the hi vol,
the resultant value obtained was between the impactor value and the cyclone
value, thus lending further credance to the above percentages. Because of the
complexity of pollution sources in both cities, a predictive approach was
considered invalid for estimating pollutant values for years in which measure-
ments for that pollutant were not available.
To develop estimates of past exposure, before the CHESS stations were
installed, available local, state and federal data were obtained (primarily
for the years 1968 through 1972). These data were taken by several different
agencies, under different conditions and for different purposes. Annual
averages were available as arithmetic means, and/or geometric means in some
cases, and in frequency distributions (without means) in others. The number
of samples per year ranged from less than 20 to a maximum of 26 for years
when samples were taken at two-week intervals. Equipment malfunction and
other data losses were potential sources of bias. This bias was minimized
by comparing valid quarterly averages to corresponding quarters in other
years and adjusting the annual average where indicated.
Prior to 1964, only one station in each city was in operation and sampled
one day every two weeks. To approximate annual exposure, geometric means
were used supplemented by 50th percentiles when geometric means were not
available.
Annual geometric means from county data were also used for the more
recent years. A trend line for particulates drawn on the basis of recent
data indicates higher past exposure for Birmingham (Figure 3) than one
based soley upon the values of one station. TSP exposure in Birmingham
prior to 1964 is represented by the solid line in Figure 3 and represents
a "best-judgment" decision based upon the relative accuracy and validity
of both sets of data. Values are given in Table 11. This trend is in
general agreement with a NASN single station analysis for the years 1958-
1964.
A similar approach was used on the Charlotte particulate data. Independent
analysis of both recent and past exposure data in this case, however, resulted
in almost identical trend lines (Figure 4). Exposure values are given in
Table 12. As shown in Figures 5 and 6, CHESS sector RSP values for
Birmingham vary from 13 to 36 percent lower than the historical line calculated
from the TSP values, while those for Charlotte are 6 to 30 percent lower. Less
data were available for other pollutants. These data are summarized in Tables
11 and 12 and displayed for sulfur dioxide and suspended sulfates in Figures 7
through 10. Complete historical estimates for TSP by sector are summarized in
Tables J3 and 14^J (Further details and the original figures may be found in
"Human Exposure to Fine Particulates in the Southeastern CHESS Area: Birmingham,
Alabama and Charlotte, North Carolina," Final Draft, November 1974 by David
0. Hinton, et al., Human Studies Laboratory, EPA-NERC, Research Triangle
Park, North Carolina 27711 This report is available upon request and will
be published in a forthcoming EPA Particulate Monograph.)
29"
-------
220
200
1 180
z
UJ
H GO
< z
si <
=3 UJ
H S
S cs
a.
OJ
3
CO
_J
<£
o
160
140
100
BO
EXPOSURE TREND LINE
(HISTORICAL DATA)
ESTIMATED EXPOSURE
SECTOR III
SINGLE-STATION TREND LINE
ESTIMATED EXPOSURE SECTOR
ESTIMATED EXPOSURE
60
_
f
nl
• SINGLE-STATION DATA SECTOR 1
A MULTIPLE-STATION DATA
B CHESS DATA
1 1 1 1 1 1 1 1 1 1
^_
f
196D 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970
YEAR
Figure 3. Birmingham, Alabama: total suspended particulate matter, historical
exposure.
1971
30
-------
140
120
o
<
oc
100
o n
1|
UJ 3-
I- 00
< z
= uj 80
60
40
i i i i
i i i i r
/N
/ \
SINGLE-STATION TREND LINE
EXPOSURE TREND LINE
(HISTORICAL DATA)
ESTIMATED EXPOSURE
SECTOR II AND III
ESTIMATED EXPOSURE
SECTOR I
• SINGLE-STATION DATA
• MULTIPLE-STATION DATA
A CHESS DATA
I I I I
''
1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971
YEARS
Figure 4. Charlotte, North Carolina: total suspended particulate matter, historical
exposure.
31
-------
70
<
-------
-------
30
£ 3
«S
e
-------
5"
o t-
x% \lf
= tn
u- <
30
20
10
ESTIMATED EXPOSURE. CHESS SECTOR III
ESTIMATED EXPOSURE. CHESS SECTOR II
EXPOSURE TREND LINE (HISTORICAL DATA)
ESTIMATED EXPOSURE, CHESS SECTOR I
MULTIPLE-STATION DATA
CHESS DATA
I I I I L
1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971
.YEAR
Figure 8. Charlotte. North Carolina: sulfur dioxide, historical exposure.
35
-------
20
v>
z
o
o
<=> t-
Z <
LU
O.
CO
=>
CO
i—r
ESTIMATED EXPOSURE. CHESS SECTOR
15
ESTIMATED EXPOSURE, CHESS SECTOR II
• ESTIMATED EXPOSURE, CHESS SECTOR I
10 —
EXPOSURE TREND LINE (HISTORICAL DATA) —
..
0
A MULTIPLE-STATION DATA
• SINGLE-STATION DATA
• CHESS DATA
I I I
J I I
I960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971
YEAR
Figure 9. Birmingham, Alabama: suspended sulfates, historical exposure.
36
-------
12
10
»- E
o .
Z ««
o ==
o <
< o
a oc
z <
UJ
o.
CO
to
i i r
i i i i i r
i i i i i i i i i i
I960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971
YEARS
Figure 10. Charlotte, North Carolina: suspended sulfates, historical exposure.
37
-------
TABLE 11. BIRMINGHAM HISTORICAL EXPOSURE (1960-1971)
(Yearly Averages)**
YEAR
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
TSPa
166.5
167
168
168.5
166
161.5
155.5
153
145.5
142.5
136
133
RSPb
62.4
62.6
63.0
63.2
62.3
60.6
58.3
57.2
54.6
53.4
58.5
57.2
S0x
9.2C
10. 0C
11. 1C
10. 9C
11. 9g
10. 5g
8.7q
11.4f
10.7*
10.7*
16. 0C
H
11.8°
N0x
2.2°
2.1C
2.1C
2.1C
2.4C
2.8C
3.0C
3.0*
3.0*
3.1*
2.0C
H
2.5°
so2
<25.0*
<25.0*
<25.0*
<25.0*
<25.0*
5.5f
6.8f
9.5f
10.7*
14. 6C
14. 9C
H
12. r
a - Values obtained from trend line Figure 3
b - TSP x 0.375 (thru 1968). TSP x 0.43 (1969-71).
c - One site only
d - CHESS data
e - NASN data
f - Jefferson County data
g - Federal study data
* - Estimated value
**Twenty-four hour integrated estimates of concentrations were
measured, expressed in micrograms per cubic meter. For each year
the average of all daily estimates was completed.
38
-------
TABLE 12. CHARLOTTE HISTORICAL EXPOSURE (1960-1971)
(Yearly Averages)**
YEAR
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
TSP3
112
109
105
102
98
94.5
91.5
87.5
84.5
80
77
74
RSPb
61.6
59.9
57.8
56.1
53.9
51.9
50.3
48.1
46.5
44.0
42.4
40.7
S0x
4.3C
5.8C
6.1*
C AC
6.4
7.5*
8.6*
9.8C
8.8C
8.2C
10. 2C
10.4d'e
9.6d
N0x
1.6C
2.1C
2.0*
1.9C
2.2*
2.3*
2.5C
1.8C
1.6C
1.3C
0.7d
1.7d
so2
17.0*
16.9*
16.8*
16.7*
16.6*
16.5*
16.4*
13. 6f
19. 4f
19. 9f
12. 8d
16.3d'f
a - Values obtained from trend line Figure 4.
b - TSP x 0.55
c - One site only
d - CHESS data
e - NASN data
f - Mecklenburg County data
* - Estimated value
**Twenty-four hour integrated estimates of concentrations were
measured, expressed 1n micrograms per cubic meter. For each year,
the average of all daily estimates was completed.
39
-------
TABLE 13. TOTAL SUSPENDED PARTICULATES CHESS EQUIVALENT EXPOSURE
BIRMINGHAM, ALABAMA
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
TSP
History
166.5
167.0
168.0
168.5
166.0
161.5
155.5
153.0
145.5
142.5
136.0
133.0
Sector I
122.0
122.5
123.0
123.0
120.0
115.0
110.0
105.0
100.0
94.5
88.5
84.5
Sector II
125.0
126.0
127.0
128.0
126.0
123.5
117.5
111.5
105.5
100.0
94.0
88.0
Sector III
148.5
149.0
150.0
151.0
149.5
145.0
140.0
135.0
129.0
124.0
120.0
114.5
Sector
Average
131.8
132.5
133.3
134.0
395.5
127.8
122.5
117.1
111.5
106.1
100.8
95.6
-------
TABLE 14. TOTAL SUSPENDED PARTICIPATES CHESS EQUIVALENT HISTORICAL EXPOSURE
CHARLOTTE, NORTH CAROLINA
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
TSP
History
112.0
109.0
105.0
102.0
98.0
94.5
91.5
87.5
84.5
80.0
77.0
74.0
Sector I
80.0
76.5
73.5
69.5
65.5
63.5
60.0
56.5
53.5
50.0
48.0
43.5
Sector II
108.0
105.0
102.0
98.5
95.0
90.5
87.0
84.0
82.0
78.5
75.0
71.5
Sector III
108.0
105.0
102.0
98.5
95.0
90.5
87.0
84.0
82.0
78.5
75.0
71.5
Sector
Average
98.6
95.5
92.5
88.8
85.1
81.5
78.0
74.8
72.5
69.0
66.0
62.1
41
-------
SECTION 6
RESULTS AND DISCUSSION
Environmental Exposure
Estimates of ambient air pollutant exposure levels for the entire lives
of the oldest study children are summarized in Table 15. Youngest children,
aged 1 to 4 years, were exposed to pollutant levels only from 1968 through
1971. Children aged 12,. on the other hand, were exposed to ambient air
pollutant levels as early as 1960. Particulate levels decreased from 1960
through 1971 in both communities. Estimates suggested that annual total
suspended particulates ranged from 133 to 169 yg/m3 in Birmingham, and from
74 to 112 yg/m in Charlotte for the twelve years preceding the study. These
values confirmed the expected exposure gradient.
Annual sulfur dioxide levels were less than 25 yg/m3 in both communities.
Suspended sulfate levels in Birmingham (9 to 16 yg/m3) were higher than
those in Charlotte (4 to 10 yg/m3). An important observation is that such
suspended sulfate levels can occur in the presence of low sulfur dioxide
levels. Annual suspended nitrate levels were low in both communities. Thus,
in comparison to Charlotte, residents of Birmingham were exposed to elevated
levels of particulate matter in the presence of low levels of sulfur dioxide.
Furthermore, these exposure differences appeared to be relatively constant
for the lifetime of even the oldest group of children in the study.
Questionnaire Response
Response rates were excellent in both communities, although significantly
fewer families returned the questionnaires in Charlotte (Table 16). In
Birmingham, there was no difference in return rates by race; in Charlotte,
the return rates were significantly lower for blacks, but the absolute
difference was small (84 vs. 89%). For those questionnaires which were
returned, 90 to 94% were completed for all city-race categories. Only a
small proportion of children had to be excluded from the analysis because
of missing information (Table 17a, 17b). Missing information about a child's
lower respiratory disease morbidity was the major reason for exclusion. It
was not possible to determine return rates among families by the presence or
absence of one or more children with a history of asthma because asthma
prevalence was not determined among the nonrespondents.
Family Characteristics
In both communities, 87 to 95% of the families had resided there for
three or more years (Table 18a, 18b). Educational attainment of fathers was
42
-------
highest among younger persons and whites. Black parents in Birmingham smoked
somewhat less than the other three parental groups, which were quite comparable,
The median age of female parents or guardians ranged from 34 to 37 years.
Blacks in both communities had a higher proportion of mothers or female
guardians, aged 50 years or more, than did whites (Table 19a, 19b). When
compared to blacks within each community, whites tended to have fewer persons
per household, but more rooms and more air conditioning in their households
(Table 19a, 19b). When communities were compared with one another within
each racial group, families were generally similar in regard to these attri-
butes. Families with one or more children with a history of asthma were
comparable to families with children without a history of asthma.
Prevalence of Asthma in Families
A total of about 75 black and 150 white children were reported with some
history of asthma in each community (Table 21). The age distribution of
white asthmatic children was comparable to white nonasthmatic children, but
among black asthmatic children, 1 to 4 year olds were less frequent than
their nonasthmatic counterparts. The reasons for this are not clear.
"Asthma ever diagnosed by a doctor" was more prevalent among Birmingham
white children than among the other three groups ranging from 3 to 14% with
an excess of males to females for all four city-race groups (Table 22). The
prevalence of "asthma active during the past two years" for all ages ranged
from 2 to 5% and also showed a male excess for both races (Table 23).
Almost all of the white asthmatic children in the youngest age group were
two or more years old. The proportion of active to all asthmatic children
generally decreased with age for both boys and girls although it was some-
what higher in boys when compared to girls (Table 24).
The prevalence of asthma in parents differed from that of children
(Table 25). For parents of both races, asthma prevalence was higher in
Charlotte than in Birmingham. Prevalence rates ranged from 2 to 9% overall,
but did not vary with age or parental education. Whereas "asthma ever
diagnosed" did not differ by sex in whites, it was higher among black females
than among black males in both cities. Females reported a significantly
higher proportion of active asthma for all city-race groups when compared
to males. Asthma prevalence in children and adults clustered in families.
One or more children with asthma were reported three and eight times more
frequently in families with one or both parents with asthma, respectively,
when compared to families in which neither parent had asthma (Table 26).
Asthma prevalence in children was also related to the activity of parental
asthma, but the increases were statistically significant in Charlotte only
(Table 27). Parental cigarette smoking habits, education, duration of
residence, household size, and air conditioning were the same for families
with and without any parents with asthma.
Lower Respiratory Illness in Children
Sox distributions were generally comparable among all children and
about 15%. of all children without asthma were in the 1 to 4 years of age
proup (Table 20). Morbidity rates, direct-adjusted for sex and education of
43
-------
the head of household, were tabled by four-year age intervals for blacks
(Tables 7, 8) and whites (Tables 9, 10). Respiratory disease rates decreased
with age among children of both races in both communities. When compared to
whites, blacks reported higher rates of pneumonia and lower rates of "any
lower respiratory disease," croup, bronchitis, and hospitalization. Rates
for two or more episodes were lower, of course, but showed the same pattern
with respect to age and race. Among blacks, the direct-adjusted rates of
both one or more and two or more episodes of "any lower respiratory disease",
croup, and bronchitis among 5 to 12 year olds and pneumonia among all ages
were generally higher in Birmingham (Table 7, 8). Nevertheless, the rates
for one or more episodes of hospitalization and two or more episodes of all
conditions except "any lower respiratory disease" and one or more and two or
more episodes of croup among 1 to 4 year olds were higher in Birmingham
(Tables 9, 10). For whites, rates of two or more episodes of pneumonia and
hospitalization were quite low.
Because of the black/white differences in family characteristics and
in reported morbidity patterns, the relationship of reported morbidity to
air pollutioi* exposure, age, sex, and education of the head of household
was tested statistically in a linear categorical analysis of variance model
separately for each race. Chi-square and statistical significance levels
for morbidity relationships in black children are summarized in Table 3.
Morbidity rates generally decreased significantly with age but did not
vary by sex. Morbidity reporting was significantly increased in children
from families with a high school or greater education for one or more
episodes of bronchitis and the increase approached significance (0.10>p>0.05)
for two or more episodes of bronchitis. Only one or more episodes of
pneumonia was significantly higher in Birmingham but significant "city"
interaction terms were observed for this as well as five other of the nine
morbidity by number of episode models. (Rates for two or more episodes of
hospitalization were too low to be tested in the model.) Significant
"city" differences were not observed for both episodes of bronchitis and
two or more episodes of pneumonia, i.e., for those morbidity conditions in
which significant "city" interactions did not occur in the saturated model.
Reduced statistical models were used to account for the morbidity
conditions reported for black children in which one or more significant
"city" interaction terms were found (These models are presented in detail
in Appendix B). For one or more episodes of "any lower respiratory disease,"
the city within sex comparison showed that for females the rate was
significantly increased in Birmingham as compared to Charlotte, but no
city differences were detected among males. Rates also decreased signi-
ficantly with age and the increased rate in children from better educated
families when compared to less educated families approached significance
(0.10>p>0.05). The city within sex comparisons did not differ signifi-
cantly for two or more episodes of "any lower respiratory disease;" rates
declined significantly with age, but only among Charlotte males, and rates
were significantly increased in children from better educated families
in Birmingham, but not in Charlotte.
44
-------
One or more episodes of croup was significantly higher in Birmingham,
when compared to Charlotte, in children from households with a high school
or greater education, but the city difference was not significant for
children from less educated families. One or more episodes of croup
decreased significantly with age but did not vary significantly by sex.
For two or more episodes of croup, the city within age comparison was
significantly increased only in Birmingham among 5-8 year olds, and rates
did not vary significantly with either sex or education of the head of the
household. For one or more episodes of pneumonia, the city within SES com-
parisons detected that the Birmingham rate was significantly greater than
the Charlotte rate in children from better educated families. A similar
pattern was observed in children from less educated families, although
the test for city differences only approached significance (0.10>p>0.05).
Rates decreased significantly with age but did not vary significantly by
sex. One or more episodes of hospitalization was significantly increased
in Birmingham when compared to Charlotte, and these rates also did not
vary significantly by sex.
In summary, for black children, when the city (pollution) effect was
distinguished from effects of age, sex, and socioeconomic status in the
analysis of variance, rates of one or more episodes of "any lower respiratory
disease," croup, pneumonia, and hospitalization as well as two or more
episodes of croup were significantly higher in Birmingham when compared
to Charlotte. In no case were city differences found in which morbidity
rates in Charlotte were significantly higher than those in Birmingham.
When significant, morbidity rates decreased with age and were higher in
children from households with a high school or greater education. No
significant differences by sex were found.
Chi-square and significance levels for morbidity relationships in
white children are summarized in Table 4. When compared to Charlotte,
rates in Birmingham were significantly higher for one and two or more
episodes of "any lower respiratory disease," two or more episodes of
croup and two or more episodes of bronchitis, i.e., for those morbidity
conditions in which significant "city" interactions did not occur in the
saturated model. Rates decreased significantly with age for all four of
the above morbidity conditions. Rates were significantly increased in
males and in children from households with a high school or greater education
in all conditions except for two or more episodes of croup.
As for black children, reduced statistical models were used to account
for the morbidity conditions reported for white children in which one or
more significant "city" interaction terms were found (These models are
presented in detail in Appendix B). One or more episodes of croup was
significantly increased in Birmingham when compared to Charlotte, but the
rates did not vary significantly by sex. For one or more episodes of
bronchitis, the city within sex comparisons indicated significantly increased
rates in Birmingham males when compared to-those in Charlotte. Similar
results (0.10>p>0.05) were observed for females. Bronchitis rates decreased
significantly with age and were significantly increased in children from
45
-------
better educated families when compared to children from less educated
families.
One or more episodes of pneumonia was significantly increased in
Birmingham when compared to Charlotte, but these rates did not vary
significantly by sex. For one or more episodes of hospitalization, the
city with SES comparison showed hospitalization rates were significantly
increased in Birmingham as compared to Charlotte among children from
better educated families, but no significant city differences were found
in children from less educated families. Hospitalization rates decreased
significantly with age, but did not vary significantly by sex.
*
In summary for white children, rates of all morbidity conditions were
significantly higher in Birmingham when compared to Charlotte. For white
children, as for black children, in no case were city differences found in
which morbidity rates in Charlotte were significantly higher than those in
Birmingham. When significant, morbidity rates decreased with age and were
higher in males and in children from households with a high school or greater
education.
Model adjusted rates for "city" effect in the saturated linear
model for categorical data were computed. For blacks, rates were higher
in Birmingham except for bronchitis (Table 28). For whites, model-
adjusted morbidity rates were higher in Birmingham for all conditions with
relative increases ranging from 1.07 for one or more episodes of croup (in
which there was a second order interaction involving "city") to 1.64 for
two or more episodes of bronchitis (Table 29). Statistically significant
differences in morbidity reporting by sex and by education of the head of
the household were in the same order of magnitude as those for city. For
example, for one or more episodes of "any lower respiratory disease,"
the model adjusted (for age, city, and SES) rates were 20.5% for females and
32.8% for males and model adjusted rates for children from households with
less than a high school education were 28.9% compared to 33.4% for children
from households with a high school or greater education.
As expected, children with a history of asthma had higher rates of all
morbidity conditions than did their nonasthmatic counterparts (Tables 30-34).
This pattern was observed for one or more and two or more episodes of each
condition. Children with active asthma generally had higher morbidity rates
than children with asthma diagnosed but not active; exceptions to this were
croup among whites in Charlotte and blacks in Birmingham, pneumonia among
whites in Birmingham, and hospitalization among blacks in both cities and
whites in Birmingham.
Differences in morbidity patterns by race were evident regardless of
asthma history (Tables 30-34). Compared to white children within each
community, for any asthma category, black children had significantly
higher rates of pneumonia and significantly lower rates for all four other
conditions. These differences were consistent for sex and parental educa-
tion. Among children with a history of asthma, occasional decreases of
46
-------
morbidity rates with age were observed, but no consistent patterns with
regard to sex or parental education were evident (these data are not pre-
sented) .
Morbidity rates among Birmingham black children with "inactive" asthma
were generally higher than those in Charlotte whereas the differences were
not that consistent for children with "active" asthma (Table 35). For
blacks, the increase of one or more episodes of croup among Birmingham
"inactive" asthmatic children approached significance (0.10>p>0.05) as
did the increase of two or more episodes of bronchitis among the similar
group in Charlotte (Table 35). Among white children, morbidity rates for
"active" asthmatics tended to be higher in Birmingham whereas those among
"inactive" asthmatics were higher in Birmingham for only 4 of 10 possible
morbidity by number of episode categories (Table 36). Croup rates were
significantly higher in Birmingham among "active" asthmatics (Table 37).
No other community differences approached significance among white children
with a history of asthma. In summary, descriptive city differences in
morbidity rates favored the hypothesis among "inactive" black asthmatics and
"active" white asthmatics. Statistically significant differences were
infrequent, with croup being higher in Birmingham among children of both
races and bronchitis in Charlotte approaching significance among black,
inactive asthmatic children.
Morbidity rates were significantly increased among residentially stable
children of both races in Birmingham, the high particulate exposure community,
confirming the study hypothesis. Reported excesses were significant for one
or more episodes of "any lower respiratory disease," croup, pneumonia, and
hospitalization among black children. Among whites, significant increases in
Birmingham were reported for all five morbidity conditions. Model-adjusted
morbidity rates were almost always higher in Birmingham than Charlotte. In
no case were morbidity rates significantly higher in Charlotte. With regard
to age, sex, and education of the head of the household, when significant
differences were found, morbidity rates decreased with age, were higher in
males and in children from better educated families.
Morbidity reporting was generally comparable to other reported studies
for white children. Children with a history of asthma diagnosed by a doctor
had two to three times the morbidity rates of their counterparts without such
a history.12*13 Morbidity rates decreased markedly with age, and significant
sex differences were due to male excesses.40'1*1 Bronchitis and croup were
reported much more frequently than pneumonia and hospitalization in white
children. Our results differ from British studies in that we found bron-
chitis to be more frequent among children from families whose parents had
at least a high school education, whereas in England bronchitis prevalence
was more frequent among lower social class children. 6 >7> 8»1|2»'t3 There are
several differences in our study methodology which could explain this seeming
disparity. We did not clinically examine our population, limited the recall
period to four years, restricted reported illnesses to those diagnosed by a
doctor, and used parental education rather than parental occupation as an
index of socioeconomic status. Thus, in each community our questionnaire
estimated lower respiratory disease morbidity in terms of use of medical care
47
-------
rather than in terms of symptoms, physical findings, or other indices which
might be expected to identify illness not coming to the attention of medical
care services. Nevertheless, U.S. National Health Survey data show that
although chronic bronchitis prevalence varies inversely with the education
of the head of the family for persons aged 17 years and over, chronic bron-
chitis prevalence actually increases with parental education for children
under seventeen years of age."*1*
Patterns of lower respiratory disease morbidity reporting among blacks
and whites within both communities were similar with regard to age, sex,
parental education, and a history of asthma. Such internal consistency
suggests that the increased frequency of pneumonia and decreased frequency
of bronchitis, croup, and any lower respiratory disease in blacks when
compared to whites, was not merely a statistical artifact. Pulmonary
function (FEVo-7s) of black children was significantly lower than white
children in a recent survey of Cincinnati elementary schools and for adult
males, blacks have a lower FEVi and FVC than whites. **"*>l*5 U.S. adult blacks
appear to be less susceptible to the effects of cigarette smoke than whites
on the basis of mortality, morbidity or decline in pulmonary function with
age.**6 However, many other factors besides a true inherent black-white
difference were likely to have caused the observed black-white pattern of
morbidity found in this study.
For both communities, when compared to white families, black families
in this study were less educated, had a lower proportion of two parent
families, had a higher proportion of mothers or female guardians aged 50 or
above, had larger families, lived in more crowded living conditions, and
enjoyed less air conditioning. Differences in medical care utilization would
in part explain the differences in morbidity reporting between races. For
example, although the number of physician visits per family increased with
family size, nonwhites reported fewer visits for each size category than did
whites in a recent U.S. survey.1* Thus, black parents may have utilized a
doctor only for more serious illnesses which could explain the higher rate
of pneumonia than bronchitis and croup among blacks. Differences in recall,
access to care, parental perception of illness, as well as cultural or genetic
differences all may be associated with the observed black-white disparity in
reported morbidity. Although these questions are of clinical and public
health interest, this study was not designed to answer them.
Several other known derminants of childhood respiratory disease morbidity
besides ambient air pollution exposures seemed reasonably comparable between
the two communities for each race. Family size and composition and household
size were similar in both communities. Exposure to sidestream cigarette smoke
has been associated with increased acute respiratory morbidity in children.1*8
Parental cigarette smoking was slightly more prevalent in Charlotte than in
Birmingham which would only minimize any true difference associated with air
pollution. Children's personal smoking habits were not ascertained, but there
is no a priori reason to believe they would differ markedly in the two cities.
Likewise, both communities suffered the impact of the 1968-1969 influenza
epidemic. Age, sex, and education of the head of the household, of course,
were accounted for in the analysis.
48
-------
Two factors which certainly could have affected reported morbidity rates
in this study are parental recall and the reliability of the questionnaire.
Historical morbidity information was restricted to four years for each child
rather than his entire life which minimized memory loss regarding older
children. Children less than 4 years old were, of course, not at risk for
the full four year period, but the age distribution of children aged 1 to 4
years was the same for all city-race groups. In a follow-up study of res-
piratory illnesses in Sheffield, England, only 52% (81/157) of children with
a history of bronchitis or pneumonia at age 5 were reported to have had either
illness four years later (illnesses were not restricted to "diagnosed by a
doctor").1*9 Whereas the effect of decreased parental recall would tend to
underestimate the true prevalence of childhood respiratory disease, it would
not alter the pattern of our results except in the unlikely circumstances of
intercommunity differences in recall.
Morbidity data obtained from household surveys are known to vary with
regard to whether the mother or father is interviewed, but in this study,
the mother or female guardian was requested to complete the questionnaire
whenever possible.50' Our results with regard to respiratory morbidity
and age, sex, education of the head of the household, and history of childhood
asthma agree quite well with the findings in two previous U.S. surveys
using the same questionnaire.12'13 British studies showing decreased
ventilatory function in children with a past history of bronchitis or
pneumonia give some confidence in the validity of the questionnaire in
assessing the frequency of lower respiratory tract disease.*** »lf3 >49 Further
discussion on the validity of the questionnaire is appended (Appendix C).
In this study, excess acute lower respiratory disease morbidity has been
associated with exposure to estimated average annual total suspended particu-
late concentrations of about 133 to 169 yg/m3 compared to about 74 to 112
Vg/m3 among children aged 1 to 12 years. These results do not suggest that
the present U.S. Federal primary standard for particulate matter (75 yg/m3,
annual average, geometric mean) is too stringent because total suspended
particulate concentrations in Charlotte were near this standard for several
years. Nevertheless, the data cannot tell us if morbidity rates in Charlotte
would have been higher compared to a community with even lower total suspended
particulate exposure. However, the results clearly associate exposure to
elevated particulate matter, in the presence of low sulfur dioxide concentra-
tions, with excess acute lower respiratory morbidity in children.
Children with a history of asthma diagnosed by a doctor were found to
have significantly higher rates of "any lower respiratory disease," croup,
bronchitis, pneumonia, and hospitalization than children without such a
history; morbidity rates were highest among children with a history of
asthma active during the past two years and intermediate among children
with a history of asthma ever diagnosed by a doctor, but not active during
the past two years. Relationships of lower respiratory disease morbidity
to age, sex, parental education, race, and ambient air pollution among
nonasthmatic children have been reported elsewhere.22 Morbidity rates of
children with asthma.showed exactly the same pattern with regard to race,
similar, but less striking decreases with age, and no consistent trends in
49
-------
relation to sex or parental education. Lower respiratory disease morbidity
for all asthmatic children was comparable to those in two other studies.12'13
With regard to air pollution, the findings were not as clear as those among
nonasthmatic children. Among children with a history of asthma, morbidity
rates were higher in Birmingham in half the comparisons studied. Rates
for croup were significantly increased in Birmingham among black and white
children, although those for bronchitis (0.10>p>0.05) were higher in
Charlotte among black children only.
Missing information on asthma history was rare (<1% of whites and 1-4%
of blacks). In this study, asthma prevalence in children was comparable with
published data from Australia, England, The Netherlands, Switzerland, and
the United States. 31" "*9»52~59 The higher proportion of male children with
both diagnosed and active asthma was also in accord with expectation.
Asthma prevalence among persons aged 17 to 44 years in the U.S. National
Health Survey was higher in females and did not vary by race or education
of head of family.58 .Our data for asthma among parents were quite similar
in nature. The clustering of asthma in families was also expected and would
probably have been even stronger if we had asked about other allergic
conditions as well as surveying near relatives.1S »60~6'* Thus, no major bias
regarding the reporting of asthma seems to have been present in the study
although no attempts were made to verify positive asthma histories by
clinical examination or medical record reviews.
Other factors besides a true increased frequency of acute lower
respiratory diseases could have tended to magnify the relative increase
we found for children with a history of asthma. Children with a history
of asthma would have been more likely to be seen by a physician for an
episode of respiratory illness and this possibility would seem most likely
for the children with active asthma at the time of the study. When examining
an asthmatic child with an acute respiratory illness, a physician probably
would be more vigorous in seeking evidence of lower respiratory tract
involvement and would be more likely to assume lower respiratory involvement
if there were any doubt in his mind. Parental recall would probably be
better for children with asthma. Recent evidence suggests a higher frequency
of viral infection associated with wheezing in asthmatic children than
previously thought.65'66 Conversely, children who wheeze in association
with viral infection appear to have inherited or acquired bronchial hyper-
activity.67 Since there is still question as to whether asthma and wheezy
bronchitis in childhood are different diseases, this would increase the
association of lower respiratory disease and asthma in children as
ascertained in this study.31"61'62'68"70 Hence, the relative increases
of lower respiratory disease in asthmatic children compared to nonasthmatic
children from this study, if anything, might be overestimates of the true
relative risk.
Croup alone, was significantly increased among asthmatic children
in Birmingham. Unlike bronchitis or bronchiolitis with wheezing, croup
is not likely to be confused with an attack of asthma by the clinician.26'71
It is well known that para-influenza viruses Types 1 and 3 are associated
with croup with Type 1 being more frequent.27' Infections with
50
-------
parainfluenza Type 3 viruses produce a more variable clinical picture with
bronchiolitis or pneumonia predominating in infants, croup in children two
to three years of age, and tracheobronchitis above this age. Biannual
epidemics of parainfluenza Type 1 virus have been found and at least one
longitudinal study has noticed the pattern of parainfluenza Type 3 infections
change toward discrete waves from an initial endemic nature.29 It is possible
that an epidemic of parainfluenza Type 1 virus occurring in Birmingham, but
not in Charlotte, could have caused the observed increase. Against this is
that epidemics of parainfluenza Type 1 virus have occurred simultaneously
in places as far apart as North Carolina and Washington whereas other studies
have not demonstrated distinct seasonal patterns for this virus.29 Community
differences in diagnostic custom also seems an unlikely explanation.
Measurements of ambient sulfur dioxide and suspended particulates
obviously are an index of complex and varied pollution sources.73 In the
two studies of western smelter communities where excess croup was found
in exposed asthmatic children, they were more likely to be exposed to
frequent fumigations, acid aerosols, and airborne trace metals. These
exposure differences might have accounted for the increase of bronchitis
and pneumonia observed among exposed asthmatic children in those studies,
but not in this one. Although the statistically significant increase in
croup could simply be a chance finding, one cannot totally exclude a causal
association with particulate air pollution in light of the previous studies.
Against this are the generally inconsistent city differences in
morbidity rates along with the significantly (0.10>p>0.05) increased
rates of bronchitis in Charlotte black children. One could explain the
differences in bronchitis morbidity as a chance deviation with a probability
of between 0.05 and 0.10 in a population sample of this size. If they are
not due to chance, they may reflect some "protective" effect of particulate
air pollution exposure, an explanation which seems even more unlikely.
Another possible explanation for the inconsistent city differences in
morbidity rates among asthmatic children is that other factors than air
pollution such as temperature, dusts, pollens, and other allergens which
are known to be related to asthmatic attacks were not measured in this
study.?l* Although these factors have been associated with increased
asthmatic attacks, and not increased lower respiratory morbidity, in
asthmatic children, they, too, may increase the risk of lower respiratory
disease in asthmatic children. Alternately, some of the reported acute
lower respiratory disease in these asthmatic children may, in fact have
been episodes of asthma. Further studies will be necessary to fully
clarify the relationship of air pollution exposure and morbidity reporting
among children with a history of asthma.
51
-------
TABLE 15. ESTIMATED3 POLLUTANT EXPOSURE LEVELS IN CHARLOTTE, NORTH CAROLINA (INTERMEDIATE EXPOSURE) AND
BIRMINGHAM, ALABAMA (HIGH EXPOSURE): 1960-1971
N5
Pollutant
Total
Suspended
Parti culates
Sulfur
Dioxide
Suspended
Sul fates
Suspended
Nitrates
Community
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Estimated
1960-63
Average
107
168
17
<25
6
10
2
2
Pollutant
1964-67
Average
93
159
16
11
9
11
2
3
Concentrati
1968-71
Average
79
139
17
13
10
14
1
3
ons, yg/m3
1960-71
Average
93(74-112)b
155(133-169)
17(13-20)
15(6-<25)
8(4-10)
10(9-16)
2(1-3)
2(2-3)
National Air
Quality Std.
Annual Average
75 yg/m3
(geometric
mean)
80 yg/m3
(0.03 ppm)
None
None
All values obtained from reference 14 and based on both measured and estimated values. Twenty-four
hour integrated estimates of concentrations were measured expressed in micrograms per cubic meter.
For each year, the average of all daily estimates was computed. For periods of several years the
average of the individual years is tabulated. For the period 1960 to 1971 the average for the
12-year period is shown together with the range of the 12 individual years.
}Range in parentheses
-------
TABLE 16. TOTAL NUMBER OF QUESTIONNAIRES DISTRIBUTED
AND RESPONSE RATE AMONG STUDY FAMILIES*
Community
Charlotte
Birmingham
Race
Total
Black
White
Total
Black
White
Questionnaires
Returned, %
88
84
89
95
95
94
Distributed
Total
Number
3448
1083
2365
2941
1382
1559
*Refers to number of families (households) of children aged one to twelve
years (of whom only children aged six to twelve received questionnaires at
school).
53
-------
TABLE 17a. CHILDREN AGED ONE TO TWELVE EXCLUDED FROM ANALYSIS BECAUSE OF MISSING INFORMATION, BY CITY
AND RACE*
Percent Excluded by Specific Category
Total Children
Excluded
Community
Charlotte
Birmingham
Race
Black
White
Black
White
Number
294/2208
175/3964
278/3010
113/2611
Percent
13
4
9
4
.3
.4
.2
.3
Morbidity
Only
6.7
2.8
7.2
3.2
Asthma**
History
Only
4
0
1
0
.2
.8
.0
.3
Age, Sex, Residence
Duration or
Education of Head
of Household Only
1.
0.
1.
0.
7
3
0
8
Two or
More Missing
Categories
0.
0.
0.
0.
7
5
0
0
*Refers to number of children aged one to twelve (of whom only children aged six to twelve
received questionnaires at school).
**Missing information about history of asthma was obtained from respondents by telephone
only in Birmingham.
-------
TABLE 17b. CHILDREN AGED 1 TO 12 YEARS EXCLUDED FROM ANALYSIS
BECAUSE OF MISSING* INFORMATION
Community
Charlotte
Birmingham
Race
Black
White
Black
White
Percent Excluded
Asthma,
(ever diagnosed)
11.1
5.4
13.2
4.7
By Asthma History
Not Asthma
(never diagnosed)
12.9
4.3
9.0
4.3
*Missing information on morbidity, age, sex, education of the
head of household, or duration of residence. About 4% of black
children in Charlotte and <1% of all other children were excluded
because of missing information on history of asthma.
55
-------
TABLE 18a. DURATION OF RESIDENCE, EDUCATION OF FATHERS, PRESENCE OF PARENTS OR GUARDIANS AND PARENTAL
SMOKING HABITS AMONG STUDY FAMILIES, BY CITY AND SEX
Ui
Community
Charlotte
Race
Black(744)a
White(1731)
Birmingham Black (1094)
White(1144)
Three or
More Years
Residence in
Community, %
92
89
95
87
Fathers Completing
High School, *
< 39 years
55
65
58
65
> 40 years
26
57
33
49
Presence of Parents
or Guardians, %
Both
63
85
71
83
Mother
Only
33
13
27
15
Current Cigarette
Smokers, %
Mothers
43
48
34
43
Fathers
63
64
56
66
Number of families with three or more years residence in community
-------
TABLE 185. DURATION OF RESIDENCE, EDUCATION OF FATHERS, PRESENCE OF PARENTS OR GUARDIANS AND PARENTAL
SMOKING HABITS AMONG STUDY FAMILIES WITH ONE OR MORE CHILDREN WITH A HISTORY OF ASTHMA
Ul
"vj
Comnunity
Charlotte
Birmingham
Race
Black(89)a
White(209)
Black(lOO)
White(184)
Three or
More Years
Residence in
Community
87%
89%
92%
87%
Fathers Completing
High School
<39 years
70%
71%
44%
65%
>40 years
41%
48%
27%
45%
Presence of Parents
or Guardians
Both
63%
85%
77%
85%
Mother
Only
32%
13%
22%
13%
Current Cigarette
Smokers
Mothers
46%
51%
38%
48%
Fathers
56%
65%
45%
71%
Number of families with three or more years residence in community.
-------
TABLE 19a. MATERNAL AGE AND HOUSEHOLD CHARACTERISTICS OF STUDY FAMILIES, BY CITY AND RACE
CO
Community
Charlotte
Birmingham
Race
Black
White
Black
White
Age of
Female
Median
34 yrs
34 yrs
37 yrs
34 yrs
Mother or
Guardian
>50 years
10%
•*•*
4%
13%
4%
Persons per
Household
Median
4
3
4
3
>5
35%
9%
39%
10%
Number of Rooms
per Household
Median
5
5
6
6
>5
69%
83%
73%
81%
Air Conditioning
in Households
Window Only
28%
56%
37%
60%
Central
6%
7%
6%
8%
Excluding parents or guardians
-------
TABLE 19b. MATERNAL AGE AND HOUSEHOLD CHARACTERISTICS OF STUDY FAMILY WITH ONE OR MORE CHILDREN
WITH A HISTORY OF ASTHMA
VO
Community
Charlotte
Birmingham
' f '
Race
Black
White
Black
White
Age of
Female
Median
33 yrs
35 yrs
37 yrs
34 yrs
Mother or
Guardian
>50 years
15%
4%
9%
2%
Persons per
Household
Median
4
3
4
3
>5
45%
15%
46%
14%
Number of Rooms
per Household
Median
5
5
6
6
>5
76%
85%
78%
85%
Air Conditioning
in Households
Window Only
26%
60%
35%
62%
Central
14%
4%
8%
6%
Excluding parents or guardians
-------
TABLE 20. CHILDREN WITHOUT A HISTORY OF ASTHMA, BY CITY, RACE, SEX AND AGE*
Number of Children
Community
Charlotte
Birmingham
All Children
Race and Sex
Black Female
Male
White Female
Male
Black Female
Male
White Female
Mai e A
1-4 Years
134
120
206
243
206
205
143
148
1405
(15%)
5-8 Years
331
296
583
599
480
503
396
388
3576
(38%)
9-12 Years
398
378
753
777
553
549
454
456
4318
(47%)
1-12 Years
863
794
1542
1619
1239
1257
993
992
9299
(100%)
*Excludes children with a history of asthma, children with less than three
years residence duration and children with missing information on morbidity,
history of asthma, age, race, sex, education of head of household or duration
of residence.
60
-------
TABLE 21. CHILDREN WITH A HISTORY OF ASTHMA BY CITY, RACE, SEX AND AGE*
Number of Children
Community
Charlotte
Birmingham
All Children
Race and Sex
Black Female
Male
White Female
Male
Black Female
Male
White Female
Male
1-4 Years
0
2
7
15
1
3
9
24
61
(12%)
5-8 Years
17
16
33
36
17
21
14
38
192
(39%)
9-12 Years
16
25
28
53
16
22
30
51
241
(49%)
1-12 Years
33
43
68
104
34
46
53
113
494
(100%)
*Excludes children without a history of asthma, children with less than
three years residence duration, and children with missing information on
morbidity, history of asthma, age, race, sex, education of head of household
or duration of residence.
61
-------
TABLE 22. HISTORY OF ASTHMA EVER DIAGNOSED BY A DOCTOR: PREVALENCE IN
CHILDREN BY AGE, SEX, RACE AND COMMUNITY
Asthma Ever
Age
in
Years
1-4
5-8
9-12
All Ages
Diagnosed, %
Charlotte
Black
Femal e
0.0
4.9
3.9
3.7
Male
1.6
5.6
6.2
5.2
White
Female
3.3
5.3
3.7
4.3
Male
5.9
5.7
6.6
6.1
Birmingham
Black
Female
0.5
3.4
2.8
2.7
Male
1.5
4.0
4.0
3.6
White
Femal e
5.9
3.4
6.2
5.1
Male
14.0
9.1
10.0
10.3
62
-------
TABLE 23. HISTORY OF ASTHMA ACTIVE DURING THE PAST TWO YEARS:
PREVALENCE IN CHILDREN BY AGE, SEX, RACE, AND COMMUNITY
Asthma Active in
Age
in
Years
1-4
5-8
9-12
All Ages
Past Two Years, %
Charlotte
Black
Female
0.0
2.3
1.2
1.5
Male
0.8
4.8
3.5
3.6
White
Female
2.8
2.1
1.7
2.0
Male
4.3
2.7
3.7
3.4
Birmingham
Black
Femal e
0.0
2.2
1.6
1.6
Male
0.5
2.3
2.1
1.9
White
Female
3.9
1.2
1.7
1.8
Male
9.9
4.7
4.3
5.3
63
-------
TABLE 24. PERCENT OF CHILDREN WITH ACTIVE ASTHMA AMONG CHILDREN WITH ASTHMA
EVER DIAGNOSED BY A DOCTOR, BY AGE, SEX, RACE, AND COMMUNITY
Percent of Active to
Age
in
Years
1-4
5-8
9-12
All Ages
all Asthmatic Children
Charlotte
Black
Female
_*
47
31
39
Male
50
98
56
70
White
Female
86
39
33
46
Male
73
47
56
56
Birmingham
Black
Female
0
65
56
59
Male
33
57
52
53
White
Female
67
36
27
36
Male
71
51
43
52
*No children in this cell, therefore rate is indeterminate.
64
-------
TABLE 25. HISTORY OF ASTHMA DIAGNOSED BY A DOCTOR: PREVALENCE IN
FAMILIES* OF ELEMENTARY SCHOOL CHILDREN BY RACE, SEX
AND COMMUNITY
History of Asthma
Community
Exposure
Charlotte
Birmingham
Race Sex
Black Mother
Father
White Mother
Father
Black Mother
Father
White Mother
Father
Sampl e
Size
342
342
1004
1002
792
788
1053
1046
Yes, Ever
Diagnosed
7.6%
4.7%
8.5%
9.1%
3.9%
2.0%
5.6%
4.9%
Yes, Active
Past 2 Years
4.7%
1.5%
4.4%
3.5%
2.3%
0.5%
3.5%
2.0%
Active
Ever Diagnosed
61.5%
31.3%
51.8%
38.5%
58.1%
25.0%
62.7%
41.2%
*0nly families with both parents present included.
65
-------
TABLE 26. ASTHMA EVER DIAGNOSED BY A DOCTOR: PREVALENCE IN CHILDREN IN
RELATION TO PREVALENCE IN THEIR PARENTS*
Asthma Ever
Diagnosed
in Parents
Neither
One
Both
One or More Children With Asthma Ever Diagnosed, %
Charlotte Birmingham All Families, %
Black White Black White , (Total Families)
10.9 10.2 8.7 14.5 11.1 (3398)
38.2 28.9 35.7 32.6 32.0 (300)
66.7 77.8 100.0 83.0 80.0 (20)
*0nly families with both parents present included.
66
-------
Table 27. ASTHMA EVER DIAGNOSED BY A DOCTOR: PREVALENCE IN CHILDREN IN
RELATION TO ASTHMA ACTIVITY IN THEIR PARENTS*
Asthma Activity**
of Parents
Never diagnosed
Diagnosed, not active
Active, past two years
One or More
Children With Asthma Diagnosed, %
Charlotte
Black
10.9
13.3
57.9
White
10.2
24.7
33.9
Birmingham
Black
8.7
39.1
31.6
White
14.5
31.0
34.0
*0nly families with both parents present, but not with both parents
having asthma included.
**Never diagnosed - never diagnosed in either parent
Diagnosed, not active - diagnosed in one or both parents, not active in
either
Active, past two years - active in one or both parents in the past two
years
67
-------
TABLE 28. FOUR YEAR FREQUENCY OF EACH MORBIDITY CONDITION BY NUMBER OF EPISODES AND COMMUNITY:
MODEL ADJUSTED RATES FOR BLACK CHILDREN AGED 1 TO 12 YEARS
CO
Number
Morbidity Condition of
Episodes
"Any Lower Respiratory jJ.
Disease" ^2
Croup £l
Bronchitis >1
Pneumonia £l
Hospital ization ^1
Model Adjusted
Charlotte
18.7%
10.1%
7.0%
2.4%
9.2%
3.6%
9.5%
4.1%
2.6%
Rate too low to fit
Rates*
Birmingham
20.0%
10.4% .
7.1%
2.5%
8.4%
3.0%
12.8%
4.9%
3.6%
model .
Birmingham
Charlotte
1.07
1.03
1.01
1.04
0.91
0.83
1.35
1.20
1.38
*Age-sex-education head of household adjusted rates for non-asthmatic children with three or
more years residence duration from saturated linear model for categorical data (cf. Table 11).
-------
TABLE 29. FOUR YEAR "FREQUENCY OF EACH MORBIDITY CONDITION BY NUMBER OF EPISODES AND COMMUNITY:
MODEL ADJUSTED RATES 'FOR WHITE CHILDREN AGED 1 TO 12 YEARS
Morbidity Condition
"Any Lower Respiratory
Disease"
Croup
Bronchitis
Number
of
Episodes
g
>2
22
Model
Charlotte
28.7%
16.2%
13.5%
5.6%
19.3%
8.3%
Adjusted Rates*
Birmingham
33.6%
20.3%
14.5%
7.4%
24.2%
13.6%
Birmingham
Charlotte
1.17
1.25
1.07
1.32
1.25
1.64
Pneumonia
Hospitalization
6.8% 8.5%
Rate too low to fit model
3.0% 4.6%
Rate too low to fit model.
1.25
1.53
*Age-sex-education head of household adjusted rates for non-asthmatic children with three or
more years residence duration from saturated linear model for categorical data (cf. Table 4).
-------
TABLE 30. "ANY LOWER RESPIRATORY DISEASE": FOUR YEAR FREQUENCY BY HISTORY
OF ASTHMA DIAGNOSED BY A DOCTOR
Race
Black
White
History -
of
Asthma -
Never diagnosed
Diagnosed, not
active
Active, past
two years
Never diagnosed
Diagnosed, not
active
Active, past
two years
"Any Lower Respiratory Disease*", %
One or More
Charlotte
16.2
42.4
62.8
27.1
55.6
83.5
Episodes
Birmingham
18.8
48.6
60.0
31.8
47.7
80.8
Two or More
Charlotte
8.6
24.2
44.2
15.4
43.2
70.3
Episodes
Birmingham
9.7
31.4
44.4
19.9
33.0
71.8
*Crude rates of children aged 1 to 12 years.
70
-------
TABLE 31. CROUP: FOUR YEAR FREQUENCY BY HISTORY OF ASTHMA DIAGNOSED BY A
DOCTOR
Race
Black
White
History
of
Asthma
Never diagnosed
Diagnosed, not
active
Active, past
two years
Never diagnosed
Diagnosed, not
active
Active, past
two years
One or More
Charlotte
6.5
15.1
23.3
12.4
28.4
23.1
Croup
Episodes
Birmingham
6.9
37.2
21.9
14.4
32.9
51.4
*, %
Two or More
Charlotte
1.8
3.0
14.0
5.3
11.1
19.8
Episodes
Birmingham
2.5
8.9
8.6
7.5
18.1
42.4
*Crude rates of children aged 1 to 12 years.
71
-------
TABLE 32. FOUR YEAR FREQUENCY BY HISTORY OF ASTHMA DIAGNOSED BY A DOCTOR
Race
Black
White
History
of
Asthma
Never diagnosed
Diagnosed, not
active
Active, past
two years
Never diagnosed
Diagnosed, not
active
Active, past
two years ^
Bronchitis*,^
One or More
Charlotte
7.9
24.2
53.5
18.4
48.1
76.9
Episodes
Birmingham
7.8
14.3
48.8
23.5
37.5
78.2
Two or More
Charlotte
2.9
15.1
32.6
8.5
30.8
60.4
Episodes
Birmingham
2.8
2.9
24.4
8.6
22.7
64.1
*Crude rates of children aged 1 to 12 years.
72
-------
TABLE 33. PNEUMONIA: FOUR YEAR FREQUENCY BY HISTORY OF ASTHMA DIAGNOSED
BY A DOCTOR
Pneumonia*, %
Race
Black
White
History
of
Asthma
Never diagnosed
Diagnosed, not
active
Active, past
two years
Never diagnosed
Diagnosed, not
active
Active, past
two years
One or More
Charlotte
8.4
15.1
32.6
6.3
12.3
23.1
Episodes
Birmingham
12.2
25.7
31.2
7.7
6.9
24.3
Two or More
Charlotte
3.7
6.0
7.0
1.9
2.4
9.9
Episodes
Birmingham
4.8
8.6
15.6
2.4
4.6
8.9
*Crude rates of children aged 1 to 12 years.
73
-------
TABLE 34. HOSPITALIZATION: FOUR YEAR FREQUENCY BY HISTORY OF ASTHMA
DIAGNOSED BY A DOCTOR
Hospital ization*, %
Hi story
Race of
Asthma
Black Never diagnosed
Diagnosed, not
active
Active, past
two years
White Never diagnosed
Diagnosed, not
active
Active, past
two years
One or More Episodes
Charlotte
1.9
6.1
2.3
2.3
3.7
8.8
Birmingham
3.2
14.3
6.7
4.5
3.4
16.7
Two or More Episodes
Charl otte
0.4
0.0
0.0
0.3
0.0
0.0
Birmingham
0.3
5.7
4.4
0.8
1.1
1.3
*Crude rates of children aged 1 to 12 years.
74
-------
TABLE 35. FOUR YEAR FREQUENCY OF MORBIDITY AMONG BLACK ASTHMATIC CHILDREN BY HISTORY OF ASTHMATIC
ACTIVITY
Ui
Number
Morbidity Condition of
Epi sodes
"Any Lower Respiratory £l
Disease" 22
Croup £l
Bronchitis £l
Pneumonia £l
Hospitalization ^1
Morbidity Rates*,
"Inactive"
Charlotte
42.4
24.2
15.1
3.0
24.2
15.1
15.1
8.0
6.1
0.0
Asthma
Birmingham
48.6
31.4
37. 2d
8.9
14.3.
2.9d
25.7
8.6
14.3
5.7
by Asthma Activity, %
"Active"
Charlotte
62.8
44.2
23.3
14.0
53.5
32.6
32.6
7.0
2.3
0.0
Asthma
Birmingham
60.0
44.4
21.9
8.6
48.8
24.4
31.2
15.6
6.7
4.4
*Crude rates of children aged 1 to 12 years
a - p<0.001; b - p<0.01; c - p<0.05; d - 0.10>p>0.05
Probabilities for a two-tailed test of significance, viz. the probability for morbidity excesses in
either city as large or larger than those-observed given that there is no difference.
-------
TABLE 36. FOUR YEAR FREQUENCY OF MORBIDITY AMONG WHITE ASTHMATIC CHILDREN BY HISTORY OF
ASTHMATIC ACTIVITY
Number Morbidity Rates*,
Morbidity Condition
"Any Lower Respiratory
Disease"
Croup
Bronchitis
Pneumonia
Hospital ization
of "Inactive"
Episodes Charlotte
£l 55.6
>2 .> 43.2
£l 28.4
$2 11.1
21 48.1
£2 30.8
£l 12.3
*2 2.4
*1 3.7
^2 0.0
Asthma
Birmingham
47.7
33.0
32.9
18.1
37.5
22.7
6.9
4.6
3.4
1.1
by Asthma Activity, %
"Active"
Charlotte
83.5
70.3
23.1
19.8
76.9
60.0
23.1
9.9
8.8
0.0
Asthma
Birmingham
80.8
71.8
51.4?
42. 4b
78.2
64.1
24.3
8.9
16.7
1.3
*Crude rates of children aged 1 to 12 years.
a - p^O.OOl; b - p^O.Ol; c - p$0.05; d - 0.10>p>0.05
Probabilities for a two-tailed test of significance, viz. the probability for morbidity excesses in
either city as large or larger than those observed given that there is no difference (cf. Table 37).
-------
TABLE 37. CHI-SQUARE AND SIGNIFICANCE LEVELS: COMMUNITY DIFFERENCES IN LOWER RESPIRATORY DISEASE
AMONG CHILDREN WITH A HISTORY OF ASTHMA AND THREE OR MORE YEARS RESIDENCE DURATION
Any LRD
Race Asthma Cateqory
Croup Bronchitis Pneumonia Hospital ization
* * *1 * >,l ,2 >,1 ,2 >1 ,2
Black Inactive (33,35)* 0.70 0.15 3.17d 0.95 0.54 3.00d 0.60 0.40 0.52 1.26
Active (43,45) <0.01 0.04 0.01 0.60 0.05 0.36 0.01 0.87 0.97 1.27
White Inactive(81,88) 0.75 1.47 0.22 1.16 1.54 0.85 0.93 0.54 0.25 <0.01
Active(91,78) 0.07 <0.01 13.299 9.07b <0.01 0.11 <0.01 <0.01 1.72 <0.01
*Sample sizes for Charlotte and Birmingham in parentheses.
a - p<0.001
b - psO.Ol
c - p^O.05 For each term in the model, the probability for a two-tailed test of statistical significance,
d - 0.10>p>0.05
-------
REFERENCES
1. Acute conditions (Incidence and Associated Disability, United States—
June 1965-June 1966). National Center for Health Statistics, Public
Health Service, U.S. Department of Health, Education and Welfare.
Washington, D.C. PHS Publication No. 1000, Series 10, No. 38.
June 1967- pp. 61.
2. Dingle, J.H., G.F. Badger, and W.S. Jordan. Patterns of Illness.
In: Illness in the Home. Cleveland, The Press of Western Reserve
University, 1964. pp. 33-37.
3. Reid, D.D. The Beginnings of Bronchitis. Proc. R. Soc. Med.,
62:311-316, 1969.
4. Valadian, I., H.C. Stuart, R.B. Reed. Contribution of Respiratory
Infections to the Total Illness Experiences of Healthy Children from
Birth to 18 Years. Am. J. Public Health, 51:1320-1328, 1961.
5. Annotation: Lancet, 1:865, 1960.
6. Douglas, J.W.B., and R.E. Waller. Air Pollution and Respiratory
Infection in Children. Br. J. Prev. Soc. Med., 20:1-8, 1966.
7- Lunn, J.E., J. Knowelden, and H.A. Handyside. Patterns of Respiratory
Illness in Sheffield Infant School Chidren. Br. J. Prev. Soc. Med.,
21:7-16, 1967.
8. Holland, W.W., T. Halil, A.E. Bennett, and A. Elliot. Factors Influencing
the Onset of Chronic Respiratory Disease. Br. Med. J., 2:205-208, 1969.
9. Toyama, T. Air Pollution and Its Health Effects in Japan. Arch. Environ.
Health, 8:153-173, 1964.
10. Manzhenko, E.G. The Effect of Atmospheric Pollution on the Health of
Children. Hyg. and Sanitation (Moscow), 31:126-128, 1966.
11. French, J.G., G. Lowritnore, W.C. Nelson, J.F. Finklea, T. English, and
M. Hertz. The Effect of Sulfur Dioxide and Suspended Sulfates on Acute
Respiratory Disease. Arch. Environ. Health, 27:129-133, 1973.
12. Nelson, W.C., J.F. Finklea, D.E. House, D.C. Calafiore, M.B. Hertz, and
D.H. Swanson. Frequency of Acute Lower Respiratory Disease in Children:
Retrospective Survey of Salt Lake Basin Communities, 1967-1970.
In: Health Consequences of Sulfur Oxides: A Report from CHESS, 1970-71.
EPA 650/1-74-004, May 1974.
78
-------
13. Finklea, J.F., D.I. Hammer, D.E. House, C.R. Sharp, W.C. Nelson and
G.R. Lowrimore. Frequency of Acute Lower Respiratory Disease in
Children: Retrospective Survey of Five Rocky Mountain Communities,
1967-1970. In: Health Consequences of Sulfur Oxides: A Report from
CHESS, 1970-71. EPA 650/1-74-004, May 1974.
14. Colley, J.R.T. Respiratory Disease in Childhood. Br. Med. Bull.
27(1):9-14, 1971.
15. Gordis, L. Chapter 2. Bronchial Asthma, in Epidemiology of Chronic
Lung Disease in Children. The Johns Hopkins University Press.
Baltimore, Maryland. 1973.
16. Wilson, A.F. and S.P. Galant. Recent Advances in the Pathophysiology
of Asthma. West. J. of Med. 120(6):463-470, 1974.
17. Air Pollution in Donora, Pennsylvania: Epidemiology of the Unusual smog
episode of October, 1948, Public Health Bull. 306, 1949. pp. 173 & ff, ix.
18. Mortality and Morbidity During the London Fog of December, 1952.
Reports on Public Health and Related Subjects, No. 95. Ministry
of Health, London, 1954.
19. Zweiman, B., Slavin, R.G., Feinberg, R.J., Falliers, C.J., and T.H. Aaron.
Effects of Air Pollution on Asthma: A Review. J. Allergy Clin. Immunol.,
50(5):305-314, 1972.
20. Cohen, A.A., Bromberg, S., Buechley, R.W., Heiderscheit, L.T., and
C.M. Shy. Asthma and Air Pollution from a Coal-Fueled Power Plant.
AJPH, 62(9):1181-1188, 1972.
21. Sultz, H.A., Feldman, J.G. Schlesinger, F.R., and W.E. Mosher.
An Effect of Continued Exposure to Air Pollution on the Incidence
of Chronic Childhood Allergic Disease. AJPH, 60(5):891-900, 1970.
22. Hammer, D.I. Acute Lower Respiratory Disease in Children in Relation to
Ambient Sulfur Oxides and Total Suspended Particulate Exposure. Part I.
Frequency of Acute Lower Respiratory Disease in Children: Retrospective
Survey of Two Southeastern Communities, 1968-1971. Thesis, Harvard
School of Public Health, 1976.
23. Chapman, R.S., V. Hasselblad, C.G. Hayes, J. Williams and D.I. Hammer.
Air Pollution and Childhood Ventilatory Function I. Exposure to
Particulate Matter in Two Southeastern Cities, 1971-72. Clinical
Implications of Air Pollution Research. American Medical Association,
Chicago, Illinois, 1976, pp. 285-303.
24. Hammer, D.I., F.J. Miller, A.G. Stead and C.G. Hayes. Air Pollution and
Childhood Lower Respiratory Disaese I. Exposure to Sulfur Oxides and
Particulate Matter in New York 1972. Clinical Implications of Air
Pollution Research. American Medical Association, Chicago, Illinois,
1976, pp. 321-337.
79
-------
25. Duffy, J. Chapter VI. Respiratory Diseases, in Epidemics in Colonial
America. Louisiana State University Press. Baton Rouge, 1953 (1971
printing), pp. 274 + ff. ix.
26. Loda, F.A. Clyde, Jr., W.A., Glezen, W.P., Senior R.J., Sheaffer, C.I.,
and F.W. Denny, Jr. Studies on the Role of Viruses, Bacteria, and M.
Pneumoniae as Causes of Lower Respiratory Tract Infections in Children.
J. of Pediatr., 72(2):16l-176, 1968.
27. Glezen, W.P., Loda, F.A., Clyde, Jr., W.A., Senior, R.J., Shaeffer, C.I.,
Conley, W.G., and F.W. Denny. Epidemiologic Patterns of Acute Lower
Respiratory Disease of Children in a Pediatric Group Practice. J. of
Pediatr., 78(3):397-406, 1971.
28. Hope-Simpson, R.E. and P.G. Higgins. A Respiratory Virus Study in Great
Britain: Review and Evaluation. Prog. Med. Virol. 11:354-407, 1969.
29. Glezen, W.P. and F.W. Denny. Epidemiology of Acute Lower Respiratory
Disease in Children, NEJM., 288:498-505, 1973.
30. Child Health in the European Region. WHO Chron. 25:319-325, 1971.
31. World Health Statistics Report 24:258-263, 1971.
32. Holland, W.W. Air Pollution and Respiratory Disease. Technomic
Publishing Co., Inc., Westport, Connecticut, 1972, pp. 164 + ff. vii.
33. Hinton, D., et al. Human Exposure to Air Pollutants in Birmingham,
Alabama, and Charlotte, North Carolina: 1958-1972. EPA Intramural
Technical Report, November 1974.
34. Williams, H. and K.N. McNicol. Prevalence, Natural History, and
Relationship of Wheezy Bronchitis and Asthma in Children. An
Epidemiological Study. Br. Med. J., 4:321-325, 1969.
35. Hammer, D.I. Acute Lower Respiratory Disease in Children in Relation to
Ambient Sulfur Oxides and Total Suspended Particulate Exposure. Part II.
Acute Lower Respiratory Disease in Children with a History of Asthma:
Survey of Two Southeastern Communities, 1968-1971. Thesis, Harvard
School of Public Health, 1976.
'36. Hill, A.B. Principles of Medical Statistics, Seventh Edition. Oxford
University Press, New York, 1967, pp. 367 + ff. ix.
37. Snedecor, G.W. and Cochran, W.G. Statistical Methods, Sixth Edition.
Iowa State University Press, 1967, pp. 593 + ff. xiv.
38. Grizzle, J.E., C.F. Starmer, and G.G. Koch. Analysis of Categorical
Data by Linear Models. Biometrics, 24:489-504, 1969.
39. Siegel, S. Nonparametric Statistics for the Behavioral Sciences.
McGraw-Hill Book Company, Inc., New York, 1956, pp. 312 + ff. xvii.
80
-------
40. Tucher, D., J.E. Coulter, and J. Downes. Incidence of Acute Respiratory
Illness Among Males and Females at Specific Ages. Study No. 5. Milbank
Memorial Fund Quarterly, xxx, No. 1:42-60 (January) 1952.
41. Goble, F.C. and E.H. Konopka. Sex as a Factor in Infectious Disease.
Trans. N.Y. Acad. Sci., pp. 325-346, 1973.
42. Colley, J.R.T., J.W.B. Douglas and D.D. Reid. Respiratory Disease in
Young Adults: Influence of Early Childhood Lower Respiratory Tract
Illness, Social Class, Air Pollution and Smoking. Br. Med. J.,
3:195-198, 1973.
43. Colley, J.R.T. and D.D. Reid. Urban and Social Origins of Childhood
Bronchitis in England and Wales. Br. Med. J., 2:213-217, 1970.
44. Shy, C.M., V. Hasselblad, R.M. Burton, C.J. Nelson, and A.A. Cohen.
Air Pollution Effects of Ventilatory Function of U.S. Schoolchildren:
Results of Studies in Cincinnati, Chattanooga, and New York. Arch.
Environ. Health, 27:124-128, 1973.
45. A. Damon. Negro-White Differences in Pulmonary Function (vital capacity,
timed vital capacity, and expiratory flow rate). Hum. Biol., 38:380-393,
1966.
46. J.H. Stebbings, Jr. A Survey of Respiratory Disease Among New York City
Postal and Transit Workers. IV. Racial Differences in the FEV .
Environ. Res., 6:147-158, 1973.
47. Acute Conditions, Incidence and Associated Disability, United States,
July 1964-June 1965. National Health Survey, National Center for Health
Statistics, Public Health Service Publication No. 1000 Series 10, No. 26,
1965, pp. 18.
48. Cameron, P., J.S. Kostin, J.M. Zaks, J.H. Wolfe, G. Tighe, B. Oselett,
R. Stocker, and J. Winton. The Health of Smokers' and Nonsmokers1
Children. J. Allergy, 43(6):336-341, 1969.
49. Lunn, J.E., J. Knowelden, and J.W. Roe. Patterns of Respiratory Illness
In Sheffield Junior Schoolchildren, A Followup Study. Br. J. Prev. Soc.
Med., 24:223-228, 1970.
50. Feldman, J.J. The Household Interview Survey as a Technique for the
Collection of Morbidity Data. J.Chronic. Dis., 11:535-557, 1960.
51. Holland, W.W., H.S. Kasap, J.R.T. Colley, and W. Cormack. Respiratory
Symptoms and Ventilatory Function: A Family Study. Br. J. Prev. Soc.
Med., 23:77-84, 1969.
52. Pollard, J.A. Troy, V.G., Shanahan, T.A., and J.B. Hobday. The Prevalence
of Wheezing and Other Respiratory Symptoms in School Children Aged Six to
Eleven Years from Perth, Western Australia. Med. J. Aust., 2:521-523,
1971.
81
-------
53. Smith, .I.M. Prevalence and Natural History of Asthma in School Children.
Br. Med. J., 1:711-713, 1961.
54. Biersteker, K. and P. van Leeuwen. Air Pollution and Peak Flow Rates
Of Schoolchildren in Two Districts of Rotterdam. Arch. Environ. Health,
20:382-384, 1970.
55. Varonier, J.S. Prevalence of Respiratory Allergy Among Children and
Adolescents, in Geneva, Switzerland. Respiration 27, Suppl. 115-120,
1970.
56. Broder, I., Barlow, P.P. and R.J.M. Horton. Epidemiology of Asthma
and Hay Fever in a Total Community, Tecumseh, Michigan. J. Allerg.,
33:513-524, 1962.
57. Smith, J.M. and L.A. Knowler. Epidemiology of Asthma and Allergic
Rhinitis. I. In a Rural Area. Amer. Rev. Resp. Dis., 92:16-30, 1965.
58. Smith, J.M. and L.A. Knowler. Epidemiology of Asthma and Allergic
Rhinitis. II. Tn a University-Centered Community. Amer. Rev. Resp.
Dis., 92:31-38, 1965.
59. Prevalence of Selected Chronic Respiratory Conditions, United States-
1970. National Health Survey, National Center for Health Statistics,
Vital and Health Statistics - Series 10-No. 84. DREW Publication
No. (HRS) 74-1511, Rockville, 1973.
60. McKee, W.D. The Incidence and Familial Occurrence of Allergy.
J. of Allergy, (now J. Allergy Clin. Immun.) 38(4):226-235, 1966.
>
c*
61. Stur, O.B. and H. Grabner. Constitutional Factors in Children with
Asthma and Recurrent Bronchitis at School Age. Respiration 27,
Suppl. 121-126, 1970.
62. Gregg, I. A Study of Recurrent Bronchitis in Childhood. Respiration
27, Suppl. 133-138, 1970.
63. Hagy, G.W. and H.A. Settipane. Bronchial Asthma, Allergic Rhinitis,
and Allergy Skin Tests Among College Students. J. of Allergy,
44(6):323-332, 1969.
64. Lubs, M.E. Empiric Risks for Genetic Counseling in Families with
Allergy. J. of Pediatrics, 80(1):26-31, 1972.
65. Lambert, II.P. and H. Stern. Infective Factors in Exacerbations of
Bronchitis and Asthma. Brit. Med. J., 3:323-327, 1972.
66. Minor, R.E., Dick, E.C., DeMeo, A.N., Ouellette, J.J., Cohen M. and
C.E. Reed. Viruses as Precipitants of Asthmatic Attacks in Children.
JAMA, 227:292-298, 1974.
82
-------
67. Gregg, I. Viral Infections and Asthma (letter to the editor).
Br. Med. J., 3:824-825, 1972.
68. Rooney, J.C. and H.E. Williams. The Relationship Between Proved Viral
Bronchitis and Subsequent Wheezing. J. of Pediatrics, 79(5):744-747,
1971.
69. McNicol, K.N. and H.E. Williams. Spectrum of Asthma in Children -
I. Clinical and Physiological Components. Br. Med. J., 4:7-11, 1973.
70. Editorial. Asthma and Wheezy Bronchitis in Childhood. Br. Med. J.,
4:749-750, 1973.
71. Smith, C.B., and J.C. Overall. Clinical and Epidemiologic Clues to
the Diagnosis of Respiratory Infections. Rad. Clin. North. Amer.
XI(2):26l-278, 1973.
72. Chanock, R.M. and R.H. Parrot. Acute Respiratory Disease in Infancy
and Childhood. Present Understanding and Prospects for Prevention.
Pediatrics, 36:21-39, 1965.
73. Ferris, Jr., B.C. and J.L. Whittenberger. Environmental Hazards.
Effects of Community Air Pollution on Prevalence of Respiratory
Disease. NEJM, 275:1413-1419, 1966.
74. Gross, N.J. Bronchial Asthma. Current Immunologic, Pathophysiologic
and Management Concepts. Harper and Row, Hagerstown, Maryland 1974.
75. Reid, D.D. Air Pollution and Respiratory Disease in Children. In:
Bronchitis, Second International Symposium, Groningen, The Netherlands,
April 22-24, 1974.
76. Watanabe, H. Air Pollution and Its Health Effects in Osaka, Japan.
Preprint. Presented at the 58th Annual Meeting, Air Pollution Control
Association, Toronto, Canada, June 20-24, 1965.
77. Anderson, D.O. and A.A. Larsen. The Incidence of Illness Among Young
Children in Two Communities of Different Air Quality. A Pilot Study.
Canad. Med. Assn. J., 95(18):893-904, 1966.
78. Anderson, D.O. and C. Kinnis. An Epidemiologic Assessment of a Pediatric
Peak Flowmeter. Amer. Rev. Resp. Dis., 95:73-80, 1967.
79. Ferris, B.C. Effects of Air Pollution on School Absences and Differences
in Lung Function in First and Second Graders in Berlin, New Hampshire,
January 1966 to June 1967. Amer. Rev. Resp. Dis., 102:591-606, 1970.
80. Haynes, Jr., W.F., V.J. Krstulovic, and A.L. Loomis Bell, Jr. Smoking
Habit and Incidence of Respiratory Tract Infections in a Group of
Adolescent Males. Am. Rev. Resp. Dis., 93(5):730-735, 1966.
831
-------
81. Leader, S.R., A.J. Woolcock, J.K. Peat and C.R.B. Blackburn. Assessment
of Ventilatory Function in an Epidemiological Study of Sydney School-
children. Bull. Physio-path. Resp,, 10:635-641, 1974.
82. Finklea, J.F., J.G. French, G.R. Lowrimore, J. Goldberg, C.M. Shy, and
W.C. Nelson. 4.3 Prospective Surveys of Acute Lower Respiratory Disease
in Volunteer Families: Chicago Nursery School Study, 1969-1970. Ibid.
pp. 4-37-55.
83. Love, G.J., A.A. Cohen, J.F. Finklea, J.G. French, G.R. Lowrimore,
W.C. Nelson, and P.B. Ramsey. 5.3 Prospective Surveys of Acute
Respiratory Disease in Volunteer Families: 1970-71 New York Studies.
Ibid. pp. 5-49-69.
84. Glasser, M., L. Greenburg, and F. Field: Mortality and Morbidity During
a Period of High Levels of Air Pollution, New York, November 23-25, 1966.
Arch. Environ. Health, 15:684, 1967.
85. Chiaramonte, L.E., J.R. Bougiorno, R. Brown, and M.E. Laano: Air Pollu-
tion and Obstructive Respiratory Disease in Children, N.Y. State. J. Med.,
70:394, 1970.
86. Zeidberg, L.D., R.A. Prindle, and E. Landau. The Nashville Air Pollution
Study. I. Sulfur Dioxide and Bronchial Asthma (A Preliminary Report).
Am. Rev. Resp. Dis., 84:489-503, 1961.
87. Lewis, R., M.M. Gilkeson and R.O. McCaldin. Air Pollution and New Orleans
Asthma. A preliminary report. Public Health Rep., 77:947-954, 1962.
88. Girsh, L.S., E. Shubin^ C. Dick and F.A. Shulaner. A Study on the Epide-
miology of Asthma in Children in Philadelphia. The Relation of Weather
and Pollution to Peak Incidence of Asthmatic Attacks. J. of Allergy,
39:347-357, 1967.
89. Finklea, J.F., D.C. Calafiore, C.J. Nelson, W.B. Riggan, and C.G. Hayes.
2.4 Aggravation of Asthma by Air Pollutants: 1971 Salt Lake Basin
Studies. In: Health Consequences of Sulfur Oxides: A Report from CHESS,
1970-71. EPA-650/1-74-004, May 1974. pp. 2-75-91.
90. Finklea, J.F., J.H. Farmer, G.J. Love, D.C. Calafiore, and G.W. Sovocool.
5.4 Aggravation of Asthma by Air Pollutants: 1970-1971 New York Studies.
Ibid.
91. Goldstein, I.F. and G. Block. Asthma and Air Pollution in Two Inner City
Areas in New York City. J. Air Poll. Cont. Assoc., 24(7):665-670, 1974.
92. Yoshida, R. Clinical and Epidemiological Studies on Childhood Asthma in
Air Polluted Areas in Japan. Clinical Implications of Air Pollution
Research. American Medical Association, Chicago, Illinois 1976.
93. Vecchio, T.J. Predictive Value of a Single Diagnostic Test in Unselected
Populations. NEJM, 274:1171-1173, 1966.
84
-------
BIBLIOGRAPHY
Air Quality and Automobile Emission Control. A Report by the Coordinating
Committee on Public Works, United States Senate. Serial No. 93-15. Stock
No. 5270-02105. Washington, D.C. November 1973.
Air Quality Criteria for Particulate Matter. U.S. DHEW, NAPCA. No. AP-49,
Washington, D.C. 1969.
Air Quality Criteria for Sulfur Oxides. U.S. DHEW, NAPCA. No. AP-50,
Washington, D.C., 1969.
Muskie, E. Air Pollution and Public Health. Congressional Record Senate.
Stock No. 6625-6644. April 4, 1973.
Proceedings of the Conference on Health Effects of Air Pollutants, NAS-NRC.
Committee on Public Works, United States Senate. Serial No. 93-15. Stock
No. 5270-02105. Washington, D.C. November 1973.
Rail, D.P. A Review of the Health Effects of Sulfur Oxides. NIEHS, NIH,
Research Triangle Park, N.C. October 9, 1973.
Washburn, T.C., N.M. Medearis and B. Childs. Sex Differences in Suscepti-
bility to Infections. Pediatrics, 35 (No. 1, Part l):57-64, 1965.
85
-------
APPENDIX A
REVIEW OF THE LITERATURE
Unlike adults, children are generally not subjected to occupational
pollution exposures or to "self-pollution" by smoking tobacco. Children
also are less likely to have experienced several different long term air
pollution exposures related to residential mobility. Surveys of acute
morbidity in children would be least likely to be confounded with symptoms
of chronic disease because of the relatively low frequency of chronic
disease among them. These advantages, first noted a decade ago75 make
children an excellent group in which to study adverse health effects
associated with exposure to ambient air pollutants.
Bronchial asthma, a disorder of the lower respiratory tract, is char-
acterized by periodic attacks of obstructive expiratory dyspnea resulting
in wheezing, cough, and dyspnea. There is not one pathognomonic clinical
or laboratory picture for bronchial asthma, but the American Thoracic
Society has proposed the following definition: "A disease characterized by
an increased responsiveness of the trachea and bronchi to various stimuli
and manifested by widespread narrowing of the airways which changes in
severity either spontaneously or as a result of therapy." The definition
goes on to point out that the term "asthma," as so defined, may occur in
subjects with other bronchotfulmonary or cardiovascular diseases, but in
these instances the airway obstruction is not causally related to these
diseases.15 As defined in this sense, asthma is a chronic lower respiratory
disease which varies in duration, severity, and frequency of attacks. This
review, therefore, is organized as follows: the effects of air pollution on
acute lower respiratory disease in nonasthmatic children, the effects of air
pollution on acute lower respiratory disease in asthmatic children, and the
effects of air pollution on the incidence of asthma and frequency of asthma
attacks in afflicted children. The major portion of this review will deal
with the first topic, since the bulk of the published literature to date
concerns air pollution and nonasthmatic children. Pertinent studies of air
pollution and pulmonary function, per se, in children, will be discussed in
relation to the above topics.
Air Pollution and Acute Lower Respiratory Disease: Nonasthmatic Children
Toyama studied two groups of about 100 school chilren, aged 10 and 11
in Kawasaki, Japan.9 Children in the more polluted area had a higher
frequency of occasional nonproductive cough, of a sense of mucous membrane
irritation, frequent mucous secretion by medical examination, and lower peak
expiratory flow rates (PFR). No statistically significant differences in
school abscences or total vital capacity were observed. Watanabe studied
about 150 fourth grade children in each of three schools (one with low and
86
-------
two with moderate pollution) in Osaka City.76 He found that PFR decreased
more in winter for children in polluted areas, but no difference was found
in the prevalence of cold symptoms measured as: "I have a cold, cough,
sore throat, runny nose, etc."
Anderson and Larsen studied the health effects of kraft pulping mills
upon children in Canada. They studied a total of about 750 first graders
in two polluted and one control community.77 They found no conclusive
differences between communities for school absences or the incidence of
respiratory illnesses, but did find tonsillectomy, inflamed eyes, headache,
feverishness and nausea to be more frequent in the polluted area. Anderson
and Kinnis did find the PFR in the polluted community to be significantly
lower than expected when compared to the less polluted community.78 Ferris
studied a total of about 700 first and second graders in Berlin, New Hamp-
shire (about 60 to 150 in each of seven schools in two successive school
years).79 He concluded school absences did not differ among schools but
that measurements of pulmonary function, PFR and FEV forced vital capacity
in one second (FEVi.o) did show significant differences which could have
been due to air pollution.
Biersteker and van Leeuwen studied about 500 children in a "wealthy"
and a "poor" part of Rotterdam.51* In the wealthy district, the mean smoke
concentration in the winter was 50 yg/m3 and the mean sulfur dioxide level
was 200 yg/m3 while these values were about 50% higher in the poor district.
They found that PFR, height, and weight were all lower in the poor district
and concluded that lower peak flow rates were not due solely to differences
in air pollution. Not only were their samples poorly matched by social
class, but they included no estimates of social class in their study. Also,
one would not expect to see large differences in PFR with the differences
in mean pollutant levels that they estimated for only one season.
Shy, et al., studied the FEVo-ys of schoolchildren living in neighbor-
hoods with high and low pollution in Cincinnati, Ohio and New York City,
New York.1*1* Sulfur dioxide, total suspended particulate matter and sus-
pended sulfates were measured daily. Past exposures were estimated for
the three New York communities. In the Cincinnati study, performance of
children in polluted neighborhoods improved during seasons of low pollution,
but not to the level of their counterparts in low pollution neighborhoods.
In New York, children aged 9-13 years in the high pollution neighborhood
had a poorer FEVo.75 performance whereas those aged 5-8 years did not.
Since pollution levels had decreased in the two high pollution exposure
communities, the authors concluded that the higher past exposures had
caused persistent FEVo.ys performance decrements in the older children.
Chapman, et al., have recently reported a study of FEVo.ys in black and
white schoolchildren living in Birmingham, Alabama and Charlotte, N.C.,
the former being a community with a higher particulate exposure than the
latter.23 Both black and white schoolchildren of both sexes living in
Birmingham had a poorer average FEVo.ys performance than their counterparts
in Charlotte with exposure to less pollution. All studies included current
87
-------
air pollution monitoring (of varying types) but only those of Anderson,
et al. and Ferris noted "mobility" of children.
Thus alterations in pulmonary function (PFR and FEV:.0) did appear to
be related to differences in both short term and long term pollution exposure.
None of these studies, except Shy's, ascertained the children's asthma
history nor did they look at PFR or FWi.o as a function of past respiratory
illness histories in children. Pulmonary function in older children and
young adults has been shown to vary inversely with cigarette smoking.51'80'81
Almost none of the published studies of air pollution and pulmonary function
in children have ascertained their smoking habits (due in great part to
the difficulty of obtaining an accurate smoking history). Theoretically,
the effects of air pollution upon pulmonary function in older children (about
10 or more years of age) could be confounded with those of cigarette smoking
if a much higher proportion of children in one area were cigarette smokers
than in another area. This possibility seems unlikely, especially since
the children in most of the studies were comparable by social class.
Douglas and Waller studied about 4000 children born during the first
week of March 1946 through 1961 and whose families did not move (80% of
total sample) classifying them into four residence areas of increasing
pollution.6 Pollutant levels were estimated from domestic coal consumption
on four areas for the year ending May 1952. Mothers were interviewed about
colds, lower respiratory infection (bronchitis, bronchopneumonia, or
pneumonia), and recorded hospitalizations when their children were two years
old; and colds and recorded hospitalizations when their children were four
years old. School doctors examined the children at ages 6, 7, 11 and 15
years. At these times the mothers were asked about the children's colds,
coughs, and hospital admissipns. Doctors recorded rales, rhonchi, or other
abnormal chest sounds and described the upper respiratory passages and the
tonsils during these examinations. Health interviewers and physicians were
not specially trained for this study. No significant differences among these
areas were found for the following: first cold before age of 10 months; more
than two colds between 21 and 23 months, frequent or continual colds between
46 and 51 months; mucopurulent discharge at 6 or 7 years of age; ears dis-
charging at or before 2 years, 4 years, 6 years; tonsils removed or needing
removal at 6 years, 11 years and 15 years.
Lower respiratory tract illness followed the pollution gradient during
the first and second year of life for boys and girls of middle class and
manual working class families. Hospital admissions for lower respiratory
Infections during the first five years of life, but not acute upper
respiratory infections or tonsillitis and tonsillectomy, also followed the
pollution gradient. The prevalence of rales or rhonchi recorded on one or
more and two or more occasions as well as at age 15 also followed the
pollution gradient. With regard to school absences, excess (1.5 standard
deviations above the average for all schoolchildren) episodes significantly
followed the pollution gradient in the highest pollution community. Among
the three causes for the longer absences, I.e. more than one week's dura-
tion, colds and influenza were equal in the four areas, but absences due to
bronchitis increased with the pollution gradient (0.2>p>0.1). Douglas and
88
-------
Waller interpreted the school absence findings as suggesting that the level
of air pollution is related to the number of short term absences from school
rather than to the total amount of absence. Unfortunately, in the studies
of Toyama (1964), Anderson and Larsen (1966), Anderson and Kinnis (1967).
and that of Ferris (1970), the data were not analyzed in this manner.9*7
Hence school absences appear to be an indicator of pollution, but must be
analyzed with regard to specific causes for absences and number of episodes
as well as total time away from school.
Lunn, Knoweldon, and Handyside (1967) studied a total of 819 children
with an average age of 5 years, 4 months, in four areas of increasing pollu-
tion exposure in Sheffield, England.7 Sample sizes were 413, 194, 130 and
82 in each area from low to high pollution and comprised the whole first
year infant intake of eight local authority schools. Atmospheric pollution
(smoke and sulfur dioxide) was monitored from 1964 to 1966 in all study
areas. Duration of residence in the community was not ascertained. Parents
of school children completed a questionnaire sent out and returned via the
school and were asked whether the child suffered from a persistent or frequent
cough, more than three colds a year, earache or ear discharge, sore throats
or tonsillitis, and whether colds usually went to the child's chest. A
history of asthma was not obtained. Two or three weeks later, children were
examined for palpable tonsillar lymph glands, mucopurulent nasal discharge,
scarring or perforation of the eardrums, tonsillar enlargement and a doctor
recorded the FEVo.75 and FVC (forced vital capacity). Mucopurulent nasal
discharge and a history of three or more colds going to the chest were
significantly higher in the polluted communities. A history of persistent
or frequent cough significantly followed the pollution gradient. The
history of lower respiratory tract disease (pneumonia or bronchitis) also
followed the pollution gradient.
The average height adjusted FEVo.ys and FVC varied inversely with the
pollution gradient. This study also found that the mean FEVo-75 and FVC
were significantly lowered in children with a history of persistent or
frequent cough, a history of colds going to the chest or a history of one
of two episodes of lower respiratory tract illness. Children with a history
of three or more episodes had the lowest FEVo.vs and FVC values. Lunn, et al.,
concluded that both chronic upper respiratory infections and lower respiratory
infections were related to air pollution exposure.7 They suggested that one
reason Douglas and Waller did not show a similar relationship between air
pollution and upper respiratory disease was the less precise estimates of
air pollution exposure. A much more obvious reason is that the studies
looked at different-aged people and measured different things. Lunn, et al.,
obtained a history of "more than three colds a year" in children aged 5 1/2
years (no time period was given), whereas Douglas and Waller only obtained
"more than two colds between 21 and 23 months, frequent or continual colds
between 46 and 51 months, or mucopurulent discharge at age 6 and 7 years."6'7
Thus, the methods of ascertainment are not comparable, and this seems to be
the most likely reason for the difference between the two studies with regard
to upper respiratory'infections. Lunn, et al., stressed that their study was
done in the summertime when pollution levels were low. Because of this, they
89
-------
also concluded that a persistent pattern of respiratory disability had
appeared at an early age.
Lunn, Knowelden and Roe reported the followup findings on the original
group of 5 year olds reexamined four years later. This paper also con-
tained previously unpublished data on a group of 11 year old children who
were examined at the same time as the original 5 year old cohort. A
history of lower respiratory tract illnesses, pneumonia, and bronchitis at
some time in the past was given less frequently by the 11 year old than by
the 5 year old children even though the former had twice as long to suffer
these illnesses. Lunn, et al. were surprised at these results, but they
were most likely due to memory loss. Nevertheless, the 11 year olds also
showed an excess of respiratory illness in the more polluted areas, but to
a lesser degree than the 5 year olds. The children at 9 years of age also
gave a less frequent history of three or more colds, persistent cough or
colds going to the chest than at 5 years. Atmospheric pollution levels had
also fallen during this time period. When the original 5 year olds were
seen four years later, no significant differences between pollution exposure
and respiratory morbidity were found at age 9 suggesting that the absence
of differences in pollution exposure was associated with this.
Holland, et al. studied almost 11,000 children aged 5-14 plus years
in four areas of increased air pollution exposure.8 No quantitative
estimates of air pollution were given. Parents of the children completed
a questionnaire on the respiratory disease history of their children and
were examined by one of the trained medical officers. They found that
area of residence, social class, family size and a past history of pneumonia,
bronchitis, or asthma were related to the childrens1 PFR in an independent
and additive fashion. Coll£y and Reid studied over 10,000 children aged
6-10 years in contrasting urban and rural areas in England and Wales.
Winter mean sulfur dioxide levels were given, but monitoring sites were
present in only two out of five rural areas. Parents completed a question-
naire on respiratory symptoms and illnesses for their children at home, and
the children were subsequently examined by one of 30 school medical officers
following a uniform protocol. Chronic cough and a past history of bronchitis
increased from less polluted rural to more polluted urban areas only in
children of families of social classes IV and V. Upper respiratory tract
infections measured as nasal obstruction or ear discharge, perforation or
scarring did not follow the pollutant gradient although the rates for more
serious ear disease were highest in the two high pollution cities. Little
difference was found between the PFR of the five areas. Unfortunately, the
poor documentation of air pollution exposure weakens the utility of the
otherwise excellent papers of Holland, et al. and Colley and Reid.8'1*3
Manzhenko studied about 3000 children who had resided in two school
districts in Irkutsk, Russia with high and low pollution levels for 5 years
or more (750 children lived in the low community).10 Sulfur dioxide, dust,
and tarry substances were measured during the year 1960-61. The incidence of
upper respiratory tract infections in both communities was determined by
reviewing the records of school medical examinations (carried out by the
district pediatrician). He found upper respiratory tract infections including
/
90
-------
chronic tonsillitis, chronic rhinitis, chronic sinusitis and upper respiratory
tract catarrh all to be significantly increased in the higher pollution dis-
trict. He also found more abnormal X-rays (13% versus 2% for 948 and 250
children, respectively) manifested as hilar changes only, hilar changes plus
findings in the lungs, and marked hilar and pulmonary findings. No informa-
tion on past medical history, social class, or technique of examination was
given. However, monthly mean sulfur dioxide concentrations were quite high,
ranging from 0.11 to 1.99 mg/m3 during the year of the study. Sulfur dioxide
was done by the "aspiration" method, so it is not clear if it was a gaseous
determination or estimated from sulfation fallout. Also, the nature of the
"tarry substances" was not described.
Recently, four studies of acute respiratory disease in U.S. children
have been published, initially in a summary article and subsequently in four
separate papers in a monograph.11'12*13'82'83 Two retrospective surveys of
acute lower respiratory disease were done in several smelting communities
in the Salt Lake Basin and the Rocky Mountains by Nelson, et al. and Finklea,
et al.12'13 Current exposure to sulfur dioxide, total suspended particulate
matter and suspended sulfates was monitored and past exposures were estimated
from smelter production or emissions and previous aerometric data. In Utah,
four communities representing one "Low", two "Intermediate" and one "High"
air pollution exposure were studied. In the Rocky Mountain study, five
communities were studied and pooled into "High" exposure (two communities)
and "Low" exposure (three communities) for statistical hypothesis testing.
Parents completed the questionnaire and returned it via the schools answering
questions about a history of pneumonia, croup, bronchitis, bronchiolitis or
other deep chest infections in children aged 1 to 12 years during the three
years prior to the study. Duration of residence in the community, parental
smoking habits, occupational exposure and education of the head of household
were also obtained. Analyses were restricted to children without a history
of asthma from families with three or more years residence duration (7763
children in Utah and 4305 children in the Rocky Mountain study). Significant
differences with respect to pollution were found in both studies. In the
Utah study, one or more and two or more episodes of "any lower respiratory
disease" (any LRD, a combined category), croup, and bronchitis were
significantly increased. In the Rocky Mountain study, two or more episodes
of "any LRD" and croup were significantly increased in the high pollution
communities. Differences for one or more episodes of "any LRD" and two or
more episodes of bronchitis approached statistical significance (0.10>p>0.05).
Rates for pneumonia and hospitalization did not differ among the communities.
Respiratory morbidity decreased with age and tended to be increased in males.
Any LRD, croup and bronchitis tended to be more frequent in children from
households with a high school or better education; the converse was true for
pneumonia and hospitalization.
Finklea and French, et al. and Love, et al. did prospective surveys of
acute upper and lower respiratory disease in families residing in different
areas of pollution exposure in Chicago and New York City.82'8 Participating
families (about 600 in Chicago and about 1000 in New York City) were called
biweekly by trained interviewers using a standardized questionnaire asking
about the presence of illness, fever, respiratory symptoms, restricted
91
-------
activity, ear infection diagnosed by a physician, and other physician con-
sultation visits. Upper respiratory disease was classified as any one of
the following: cough (dry, nonproductive), head cold, sore throat, sinus
or postnasal drip, or runny nose. Lower respiratory illnesses were limited
to chest colds with a persistent cough, croup, bronchitis or pneumonia.
Excess upper respiratory disease was found in all residentially stable family
segments in Chicago and in school and preschool children in New York. All
family segments in the high pollution communities in both New York and
Chicago, except for children below nursery school age in Chicago, had in-
creased acute lower respiratory disease. Morbidity varied with age, sex,
and socioeconomic status in the expected fashion. No consistent differences
were found with regard to ear infections.
Hammer, et al. studied acute lower respiratory disease retrospectively
in children with differing exposures to sulfur dioxide total suspended
particulates, and suspended sulfates in New York City in 1972 using a
questionnaire and methodology similar to the two Western smelter studies. 2lt
Analyses were restricted to nonasthmatic children aged 1 to 12 years from
families with three or more years residence duration in the community (1134
black children and 6625 white children). Children living in Queens, the
Bronx, and Sheepshead Bay experienced higher air pollution levels than
those living in Riverside. Rates of "any lower respiratory disease" (a
combined category), croup, bronchitis, and chest infections other than croup,
bronchitis and pneumonia were significantly higher among black and white
children residing in the higher pollution exposure communities. Conversely,
pneumonia and hospitalization were significantly higher only among white
children in the low exposure community but the absolute rates were low for
both conditions in all communities. Morbidity excesses in the high exposure
communities could not be explained by differences in family size and com-
position, crowding, nor indoor air pollution from parental cigarette smoking
habits or gas stoves or gas space heaters. Furthermore, morbidity rates
within the three higher exposure communities were comparable and showed only
infrequent and inconsistent statistical differences. This is the first
reported study of the effects of air pollution on exposed black children and
they too showed morbidity excesses. This fact strengthens the link between
air pollution and lower respiratory disease in children since one would
expect socioeconomically similar children to be affected by air pollution
independent of their color or ethnic group. Most recently, Hammer studied
about 4200 black and 5200 white children in two southeastern cities with
differing exposures to total suspended particulate matter and suspended
sulfates, but low exposures to sulfur dioxide.22 Significant increases in
"any lower respiratory disease", croup, bronchitis, pneumonia, and hospitali-
zation for any of these illnesses were found for both black and white
children living in Birmingham, the higher particulate exposure community.
Air Pollution and Acute Lower Respiratory Disease: Asthmatic Children
Only three studies have been able to look at the frequency of acute
lower respiratory disease in children with a history of asthma exposed to
different levels of air pollution. This is in part due to the relatively
low prevalence of asthma in the population (about 1-5%). In a study of
92
-------
four Utah communities involving 7763 nonasthmatic children and 475 asthmatic
children, asthmatic children in the high pollution communities experienced
significantly higher rates of croup. Asthmatics exposed to higher sulfate
levels tended to report more total respiratory illness, pneumonia, bronchitis,
and hospitalizations.12 In a study of 4305 nonasthmatic and 290 asthmatic
children in five Rocky Mountain communities, the exposed asthmatic children
reported more total respiratory illness, croup and bronchitis than their
unexposed counterparts. 3
In a study of over 10,000 children living in two southeastern U.S.
cities and exposed to elevated levels of total suspended particulate matter
and suspended sulfates, but rather low levels of sulfur dioxide, morbidity
reporting in black (156) and white (338) asthmatic children was inconsistent
with regard to pollution exposure.35 Morbidity rates were higher in
Birmingham, the higher particulate exposure city, in half of the comparisons
studied. Croup was significantly increased in Birmingham, although the con-
verse was true for bronchitis (0.10>p>0.05) among Charlotte blacks. Exposed
children in the Utah and Rocky Mountain studies were more likely to have
been exposed to frequent fumigations, acid aerosols and airborne trace metals
which could explain why excess morbidity was found in those two studies, but
not the latter. Although all three studies suggest an increased risk of
acute lower respiratory disease for exposed asthmatic children further studies
will be required to fully clarify this relationship.
Air Pollution and the Incidence and Frequency of Childhood Asthma
Zweiman, et al. reviewed the effects of air pollution on asthma in 1972
without regard to the age of the study subjects.1 In the classical smog
episodes of Donora, Pennsylvania and London, England, a significantly higher
proportion of asthmatics were affected when compared to nonasthmatics.l7'l8
Glasser, et al. and Chiramonte, et al. reported increased asthmatic episodes
in both older adults and children during an air pollution episode in New
York City with the peak flare-ups treated in emergency rooms on the third
day of the episode. »32
Zeidberg, et al. followed 84 asthmatic patients clinically for one year
in Nashville, Tennessee.86 Of these, 35 were children, 25 of whom were male
and 21 of whom were white. Monthly sulfur dioxide and particulates were
measured. In adults, but not children, the attack rates varied directly
with the level of sulfation on their residential environment. Lewis, et al.
studied Charity Hospital emergency clinic admissions for asthma (1960-61) and
between 50 and 85 asthmatic residents in one census tract in New Orleans.87
Virtually all subjects were over twelve years of age and the analyses were
only for blacks since they comprised about 90% of the Charity Hospital
emergency clinic population. Hi-vol air monitoring for particulate matter
was done in one station near census tract 130 in 1969 and 1961. Microscopy
was also done on the particulate matter. The daily number•of asthmatic
patients admitted to the clinic correlated .with "poor combustion particles
with associated silica." No statistically significant relationship between
measured particles and asthma attacks in census tract 130 was observed. This
93
-------
study used crude air pollution measurements and was restricted to adults, but
it did not find a positive relationship between the air pollution and the
frequency of asthma attacks.
Girsh, et al. studied weather, air pollution, and bronchial asthma in
about 1400 asthmatic children at St. Christopher's Hospital for Children in
Philadelphia.88 They found a threefold increase of bronchial asthma during
days of "noteworthy high air pollution." No specific pollutants or concen-
trations thereof are given in this report. Sultz, et al. studied the relation
of air pollution exposure to asthma and eczema in hospitalized children under
15 years of age in Erie County, New York State.21 Air pollutant measurements
were obtained from several monitoring stations. A striking positive relation-
ship between standardized morbidity ratios for hospitalization for asthma and
eczema and four levels of -air pollution exposure was found. The relationship
for air pollution was most striking for hospitalized males under 5 years of
age with asthma or eczema. Understandably, no community incidence rates of
asthma or eczema in children were obtained. The sex ratio for asthma was
almost 2 to 1 for males and hospitalization for asthma was more than three
times higher in children under 5 when compared to those age 5 to 15 years.
It is noteworthy that the affected children in the study of Chiaramonte,
et al. had a high incidence of extrinsic allergic manifestations.85
Two studies using similar methodologies have been reported recently.
Persons with known asthma reported daily attack rates via weekly diaries in
Utah (211 persons) and New York City (148 persons) for about 6 months.89*90
In both studies, about half the study subjects were 16 years old or less.
Asthma attack rates in both studies correlated negatively with ambient
temperature and to a less or extent, positively with ambient total suspended
particulate matter and suspended sulfates. Because the panelists were not
analyzed by age, it was not possible to tell if the relationships for asthma
in children to temperature and air pollution were less than, equal to, or
greater than those in the adults.
Goldstein and Block91 studied emergency room visits for asthma and air
pollution in two inner city areas in New York City, viz. Harlem in Manhatten
and Bedford-Stuyvesant in Brooklyn. Daily emergency room visits for asthma
averaged 22 at Harlem Hospital and 59 for the two Brooklyn hospitals combined.
The percent of pediatric visits in each hospital was as follows: Harlem
Hospital Center-18%, Kings County Hospital-46%, Cumberland Hospital-not given.
There was a strong relationship between daily visits for asthma and the first
cold spells of the fall season in both areas. Daily visits for asthma corre-
lated with daily sulfur dioxide levels in Brooklyn but not in Harlem. No data
are given on the socioeconomic status, race, or the ethnic group of the
persons making the visits. The relationship between daily asthma visits and
dally sulfur dioxide was consistent for those under 13 years of age and those
13 years and over, but was more pronounced in the younger group.
Little information is available on the prevalence of childhood asthma in
pommunitles with high and low air pollution levels. No consistent trends
of asthma prevalence and air pollution were found in the three United States
surveys even though the expected excesses by sex among male children and female
adults were observed.12'13'35 A recent report from Japan did find an increased
94
-------
prevalence of asthma and related allergic conditions in a high pollution
community which apparently had exposure to petrochemical wastes as well as
sulfur dioxide and total suspended particulate matter.92
Summary and Conclusions
Studies from several countries clearly associate excess lower respiratory
tract morbidity in nonasthmatic children with exposure to sulfur oxides and
particulate matter. Most studies were retrospective, using a questionnaire and
several clinically examined children at some point in time. The more recent
studies noted duration of residence, and a history of asthma in the children,
as well. Different investigations looked at different aged children, and the
period of recall for the questionnaire, as well as its specific format, varied
somewhat from study to study. The use of slightly different methods to find
the association between acute lower respiratory disease and air pollution
strengthens, rather than weakens, the credibility of the relationship.
The picture is relatively clear for air pollution and acute upper
respiratory tract disease, as well. Two U.S. studies found a relationship
between air pollution and acute upper respiratory tract disease, whereas a
British and a Russian study actually found the association for chronic upper
•flAOOOq
respiratory tract illnesses. »'' None of the studies obtained an exten-
sive allergy history from the children, and of course the pollutant types and
concentrations varied in the four studies. Nevertheless, an attractive
hypothesis is that air pollution increases acute upper respiratory infections
in all children and it also increases chronic upper respiratory tract disease
and symptoms in children with a history of asthma or allergic conditions.
This hypothesis is still conjectural, but biologically plausible.
Relatively few studies have been concerned with air pollution and
asthmatic children, per se. Asthma prevalence in childhood is low and it
is epidemiclogically distinct from adult asthma.15 Moreover, the disease
"asthma" is markedly heterogeneous in its manifestations in children ranging
from a few attacks for a short time to chronic severe, crippling, respiratory
symptoms. 3<*»52»68~70 Aside from the difficulties in assessing air pollution
exposure, almost none of the studies of asthma and air pollution classified
their study subjects by severity of asthma. Moreover, although several studies
measured daily temperature, other known determinants of asthmatic attacks such
as pollens, dusts, grasses, and other allergens were not measured (largely due
to the technical difficulties of monitoring such exposures in the field). If
one assumes that only more severe cases of childhood asthma are hospitalized,
then the work of Sultz, et al. showed a clear gradient with increasing pollu-
tion exposure.21 What is needed are well-planned studies using good air-
pollution measurements and careful selection and classification of asthmatic
children. However, the literature to date suggests both acute and chronic
effects of air pollution in asthmatic children as well as those without a
history of asthma.
95
-------
APPENDIX B
DATA ANALYSIS AND HYPOTHESIS TESTING
Morbidity conditions were analyzed in a saturated analysis of variance
(ANOVA) form of a linear model for categorical data, viz, four main effects
[city/pollution (P), age (A), sex (S), education of head of household (E)],
and all possible interactions (six first-order, four second-order, and one
third-order). For each morbidity condition analyzed in the model, age (three
categories: 1-14 years,.5-8 years, 9-12 years), sex (two categories: F -
female; M - male), and education of the head of the household (two categories:
HS - high school or more) as an index of socio-
economic status (SES) were considered as intervening variables, and city/
pollution (two categories: C - Charlotte, B - Birmingham) as the independent
variable of primary interst. When statistically significant interaction terms
involving the independent variable, "city/pollution" were found, they were
explored further to distinguish the "city/pollution" effect from those of
age, sex, and socioeconomic status.
Subsequent models to account for the significant "city/pollution" inter-
action were sought in a simple, logically consistent fashion. In brief, for
the morbidity conditions in which only one first order interaction term
involving "city/pollution" was significant, viz. city x age, city x sex, or
city x SES, a subsequent reduced model was fit with the four main effects,
the city effect being fit within the levels of the non-city variable (age,
sex, or SES) causing the interaction. For example, to account for a signi-
ficant city x sex interaction in the saturated model, the terms in the
subsequent reduced model would be as follows: city for females, city for
males, age, sex, and SES.
In one case all three first order "city" interaction terms were signi-
ficant. When more than one significant first order interaction involving
"city" occurs in the saturated model, it is not possible to fit the "city"
effect within the levels of more than one of the non-city variables because
it would create a singularity in the design matrix. Therefore, city, age,
and SES were tested separately by sex. Sex, rather than age or SES, was
chosen because of the well-known biological differences in infectious disease
experiences between males and females. A significant city x age x SES inter-
action occurred three times and this was accounted for by the following
model: city, age, sex, and SES for age within city. The resulting reduced
ANOVA models and their corresponding model adjusted rates are presented in
detail in the following section. However, the verbal summary of the findings
is brief as they are discussed in detail in the main text of the report.
96
-------
For both races, a total of seventeen reported morbidity conditions were
analyzed (two races x five morbidity conditions x two number of episode cate-
gories less the three following conditions which were not statistically tested)
Two or more episodes of hospitalization among both races and two or more
episodes of pneumonia among whites were not tested statistically because their
rates were too low to obtain reliable estimates from the model. Statistically
significant interactions involving the "city/pollution" term were found in
ten of the seventeen morbidity conditions tested, (this includes significance
level d (0.10>p>0.05) but excludes third order interactions).
Chi-square and significance levels for the reduced ANOVA models used to
account for the significant "city" interactions in black children are sum-
marized in Table B-l along with those for the "city" term from the saturated
model for comparison (cf. Table 3). No significant interactions involving
the "city/pollution" term were found for three of the nine morbidity condi-
tions tested. A significant first order interaction term involving "city/
pollution" was found in four cases. All three first order "city" interaction
terms were significant in one case as was a city x age x SES interaction in
another. For black children, model adjusted rates from the reduced models
are presented in Table B-2 for those morbidity conditions with one significant
first order "city" interaction and in Table B-3 for the two morbidity condi-
tions with either three significant first order "city" interactions or a
significant second order interaction.
For white children, chi-square and significance levels for reduced
models used to account for the significant "city" interactions are summarized
in Table B-4 along with those for the "city" term from the saturated model
for comparison (cf. Table 4). No significant "city" interactions were
found for four of the eight morbidity conditions tested. (Third order inter-
actions did approach statistical significance (0.10>p>0.05) in three cases
but were ignored because of the general difficulty of interpreting them.)
A significant first order interaction term involving "city/pollution" was
found in two cases and a significant city x age x SES interaction was found
in the two others. Model adjusted rates from the reduced models for white
children are presented in Table B-5 for the two morbidity conditions with
a significant first order "city" interaction and in Table B-6 for the two
morbidity conditions with a significant second order "city" interaction.
Morbidity conditions were analyzed initially in a saturated analysis of
variance form (four main effects and all possible interactions) of a linear
model for categorical data except for three conditions in which the reported
rates were too low. Statistically significant interactions involving the
independent variable, "city/pollution", were found in 6 of 9 morbidity con-
ditions in black children and 4 of 8 conditions in white children (Table B-7).
Thus statistically significant "city/pollution" interactions were found in
just over half (10/17) of all tested morbidity conditions. Models to
account for the interactions were sought by reducing the saturated model to
the four main effects (city/pollution, age, sex, SES) and fitting the city
effect within the non-city variable whenever possible. Six of the ten
cases involved only one significant first order "city/pollution" interaction
each and reduced models were fit with the four main effects, the city effect
97
-------
being fit within the levels of the non-city variable (age, sex, or SES).
Significant city x age x SES interactions were found in three of the 10 cases
and were accounted for by the following reduced model: city, age, sex, SES
for age within city. In one case, all three first order "city/pollution"
interactions were significant and a reduced model which analyzed city, age,
and SES for each sex was used to account for the interaction. When the city/
pollution effect was distinguished from the effects of age, sex, and socio-
economic status in the analysis of variance, morbidity rates often were
significantly higher in Birmingham and in no case was the converse true.
98'
-------
TABLE B-l. CHI-SQUARE AND SIGNIFICANCE LEVELS FOR FACTORS AFFECTING LOWER RESPIRATORY DISEASE
AS DETERMINED BY A REDUCED LINEAR MODEL FOR CATEGORICAL DATA: BLACK CHILDREN WITH
THREE OR MORE YEARS OF COMMUNITY RESIDENCE
VO
SO
Effect (d.f.)
Saturated Model
City/Pollution(l)
Reduced Model
City/Pol lution(l)
Age (2)
Sex(l)
SES
Charlotte
Age 1-4(1)
Age 5-8(1)
Age 9-12(1)
Birmingham
Age 1-4(1)
Age 5-8(1)
Age 9-12(1)
Fit of Model
ANY LRD CROUP
0.79 0.09 0.02 0.01
. Femal e Mai e
F-7.15(1)D 2.39(1) 2.43(1) HS-4.52(1)C 5-8 7.40(1)D
9-12 0.53(1)
30.013 C-2.93 10.69b 18.42a 3.92
B-2.71 4.21
0.73 - - 0.13 0.51
3.11(l)d C-0.02(l). 0.03(1)K 3.99(l)c 0.02(1)
B-5.07(l)b 5.44(1 )D
16.87(17) 10.38(9) 15.22(17) 14.56(16)
a-p$0.001; b-p^0.01; c-p$0.05; d-0.10>p>0.05 For each term in the model, the probability for a
two-tailed test of statistical significance
(continued)
-------
TABLE B-l. (continued)
Effect (d.f.)
Saturated Model
City/Pol lutlon(l)
Reduced Model
City/Pol lution(l)
Age(2)
Sex (1)
M SES
o
0,
Charl otte
Age 1-4(1)
Age 5-8(1)
Age 9-12(1)
Birmingham
Age 1-4(1)
Age 5-8(1)
Age 9-12(1)
Fit of Model
BRONCHITIS PNEUMONIA HOSPITALIZATION
>1 £2 >1 ^2 >1 ^2
0.60 0.71 8.12b 0.99 2.50
* * HS-10.59(l)a Rate
12.88b 12.05b too
0.18 0.76 low
0.01(1) to
7.56b
1.43 model.
0.12
1.78K
9.05P
10.24°
19.98(17) 18.24(13)
Saturated model adequate
b-p^O.Ol
c-p^O.05
d-0.10>p>0.05
For each term in the model, the probability for a two-tailed test of
statistical significance.
-------
TABLE B-2. FOUR YEAR FREQUENCY OF EACH MORBIDITY CONDITION BY NUMBER OF
EPISODES AND COMMUNITY: MODEL ADJUSTED RATES FOR BLACK CHILDREN
AGED 1 TO 12 YEARS
Effect
Saturated Model
Charlotte
Birmingham
Reduced Model
Charlotte
Birmingham
ANY LRD
>1 >2 >1
18.7% 10.1% 7.0%
20.0% 10.4% 7.1%
Female Male HS
16.6% 18.4% -a 7.4% 5.7%
21.0% 19.2% 6.5% 8.2%
CROUP
>2
2.4%
2.5%
1-4 5-8
3.6% 1.6%
1.4% 3.5%
9-12
2.3%
2.7%
a-Three first order "city" Interactions, cf. Table B-3 for model-adjusted rates
(continued)
101
-------
TABLE B-2. (continued)
Effect
Saturated Model
Charlotte
Birmingham
Reduced Model
Charlotte
Birmingham
BRONCHITIS PNEUMONIA HOSPITALIZATION
>1 >2 >1 >2 >1 >2
9.2% 3.6% 9.5% 4.1% 2.6% Rate
8.4% 3.0% 12.8% 4.9% 3.6% t00
low
to
>HS_ >HS_ fit
9.5% 7.8% * -b mode1'
11.8% 13.4%
*Saturated model adequate, i.e. no significant interactions involving the
"city/pollution"term.
b-City x age x SES Interaction, cf. Table B-3 for model-adjusted rates from
reduced model.
102
-------
TABLE B-3. FOUR YEAR FREQUENCY OF "ANY LOWER RESPIRATORY DISEASE" AND HOSPITALIZATION: MODEL
ADJUSTED RATES FOR BLACK CHILDREN AGED 1 TO 12 YEARS
o
u>
Two or More Episodes of
"Any Lower Respiratory Disease"
Age
1-4 Years
5-8 Years
9-12 Years
SES
HS
HS
HS
Femal
Charlotte
11.7%
12.0%
7.0%
7.2%
7.0%
7.3%
es
Birmingham
9.4%
13.4%
10.3%
14.2%
7.3%
11.2%
Mai
Charlotte
18.6%
18.3%
7.3%
7.0%
7.8%
7.5%
es
Birmingham
6.4%
10.1%
8.7%
12.4%
6.4%
10.1%
One or More
of Hospital
Charlotte
6.9%
1.3%
1.9%
0.9%
1.0%
1.3%
Episodes
ization
Birmingham
3.5%
6.5%
3.6%
0.9%
3.3%
0.7%
-------
TABLE B-4. CHI-SQUARE AND SIGNIFICANCE LEVELS FOR FACTORS
AFFECTING LOWER RESPIRATORY DISEASE AS DETERMINED
BY A REDUCED LINEAR MODEL FOR CATEGORICAL DATA:
WHITE CHILDREN WITH THREE OR MORE YEARS OF
COMMUNITY RESIDENCE
Any LRD Croup
Effect (d.f.) .
Saturated Model
City/Pollution(l) 9.91b 10.993 0.74 4.59C
Reduced Model
C1ty/Po1lution(l) * * 3.97C *
Age(2) 27.033
Sex(l) 0.09
SES
Charlotte *
Age 1-4(1) 0.18.
Age 5-8(1) 4.30C
Age 9-12(1) 0.22
Birmingham
Age 1-4(1) 6.80b
Age 5-8(1) 0.96
Age 9-12(1) 0.07
Fit of Model 18.81(13)
Continued
^Saturated model adequate
a-ps0.001
b-p$0.01
c-pgO.05
d-0.10>0.05
104
-------
TABLE B-4. (Continued)
Effect (d.f.)
Saturated Model
City/Pol lution(l)
Reduced Model
City/Poll ution(l)
Age(2)
Sex(l)
SES
Charlotte
Age 1-4(1)
Age 5-8(1)
Age 9-12(1)
Birmingham
Age 1-4(1)
Age 5-8(1)
Age 9-12(1)
Fit of Model
Bronchitis Pneumonia
>1
11.99a
F- 2.76(1 )d
M-16.81(l)a
40.86a
4.17C
18.23(l)a
13.12(17)
*2 >1 *2
24.18a 3.43d
* 5.05C Rate
h to°
9.66°
0.37 low
to
0.32h f1t
6.34b
<0.01 model
0.05
1.20b
6.21b
13.96(13)
Hospital ization
»1
5.53C
0.05
105
-------
TABLE B-5. FOUR YEAR FREQUENCY OF EACH MORBIDITY CONDITION BY NUMBER OF EPISODES AND COMMUNITY:
MODEL ADJUSTED RATES FOR WHITE CHILDREN AGED 1 to 12 YEARS
Fffppt
1. 1 i eu i
Saturated Model
Charlotte
Birmingham
Reduced Model
Charlotte
Birmingham
Any LRD Croup Bronchitis
>1 £2 ^1 £2 £l ^2
28.7% 16.2% 13.5% -.5.6% 19.3% 8.3%
33.6% 20.3% 14.5% 7.4% 24.2% 13.6%
Female Male
20.3% 18.2%
* * ** 22.9% 25/0% *
Pneumonia
*1 *2
6.8% Rate
8'5* too
low
to
fit
**
model
Hospitalization
>1
3.0%
4.6%
HS
3.1% 2.5%
4.1% 4.8%
*
X
Rate
too
low
to
fit
model
^Saturated model adequate, i;e. no significant Interactions involving the "city/pollution" term.
**City x age x SES interaction.
-------
TABLE B-6.* FOUR YEAR FREQUENCY OF CROUP AND PNEUMONIA: MODEL ADJUSTED RATES
FOR WHITE CHILDREN AGED 1 TO 12 YEARS
Age
1-4 Years
5-8 Years
9-12 Years
SES
HS •
HS
HS
One or More Episodes
of Croup
Charlotte
14.2%
15.2%
T2.0%
16.2%
9.2%
9.9%
Birmingham
11.8%
22.1%
17.3%
14.7%
11.8%
11.2%
One or More Episodes
of Pneumonia
Charlotte
7.4%
8.8%
9.1%
5.5%
4.7%
4.9%
Birmingham
10.2%
9.4%
7.1%
9.2%
8.6%
4.6%
-------
TABLE B-7. SUMMARY OF STATISTICALLY SIGNIFICANT9 INTERACTIONS IN WHICH
THE "CITY/POLLUTION" EFFECT WAS INVOLVED, BY MORBIDITY
CONDITION AND RACE /
Morbidity Condition
"Any Lower Respiratory
Disease"
Croup
Bronchitis
Pneumonia
Hospitalizatlon
Number of
Episodes
*1
*2
51
*2
»1
52
51
52
*1
52
Black
Children
C* x Sb
C x A,C x S,C x E
CxEb
C x A
None
None
Cx Eb
None
C x A x E
_c
White
Children
None
None
C x A x E
None
CxSb
None
C x A x E
c
C x E
_c
b-0.10>p>0.05
c-Rate too low to fit model
*Explanat1on of abbreviations:
C-C1ty/Pollut1on
A-Age
S-Sex
E-Education, head of household
108
-------
APPENDIX C
\
VALIDITY AND RELIABILITY OF DISEASE REPORTING
The validity and reliability of disease reporting by means of the
questionnaire was not determined in this study. However, there is an abundant
amount of direct and Indirect evidence from this and other studies which lend
credence to the observed results. In two other studies (Utah and Rocky
Mountain) using this questionnaire, 15% samples of children whose parents
reported them sick or well for bronchitis were taken, and the parents' reports
were compared with information in physicians' records of the same children.12* 3
Rather than sensitivity and specificity, this is akin to what Vecchio called
the "positive" and "negative" predictive values (PVpos, PVneg) of a single
diagnostic test, i.e. the proportion of true positives among those who test
positive and the proportion of true negatives among those who test negative,
respectively.93 The PVpos and PVneg vary with the true prevalence of the
attribute as well as the specificity and sensitivity of the test (or question-
naire). Theoretical expected PVpos and PVneg have been calculated for varying
sensitivities, specificities and true prevalence (Table C-l). For example,
for a true prevalence of 10% and a sensitivity and specificity of 50%, the ex-
pected PVpos and PVneg are 10% and 90% respectively; they are 20% and 80% for
a true prevalence of 20%. If the sensitivity and specificity are both 90%,
the PVpos and PVneg become 50% and 99% for a true prevalence of 10%, and 69%
and 97% for a true prevalence of 20%. Alterations in true prevalence, sensi-
tivity, and specificity have a greater effect upon PVpos than upon PVneg.
In the Utah study, the overall frequency of bronchitis was about 20%,
the PVpos about 75% and the PVneg about 90%. In the Rocky Mountain study,
the overall frequency of bronchitis was about 15%, the PVpos about 75% and
the PVneg about 85%. Hence, the PVpos and PVneg of the two studies were
quite similar. From the expected values in Table C-l, it appears that the
negative predictive values were in accord with expectation or a bit low and
that the positive predictive values were too high, if anything, unless the
sensitivity of the questionnaire is 90% or better. No further conclusions
can be made regarding the sensitivity and specificity of the questionnaire
from these data until further information is available. Sampling for PVpos
and PVneg in the Utah and Rocky Mountain studies was not restricted by age,
education of the head of the household, history of asthma or duration of
residence in the community (so that in several cases, the physician listed
on the questionnaire was not the child's current physician).
The specificity and sensitivity of this questionnaire could be determined
by querying a sample of parents whose children have been cared for by a known
group of physicians or health clinic for some length of time. Of course,
109
-------
such results would not directly apply to this study. However, they would
provide some quantitative estimates and would be especially useful if families
from a broad range of social classes were included and children were also
sampled with regard to age, history of asthma and duration of residence in
the community (to assure that the illness was in fact treated in their current
community of residence).
Our rates for "any lower respiratory disease" were about 286 and 336/
1000 white, nonasthmatic children per four years in Charlotte and Birmingham,
figures remarkably in accord with the rate of 85/1000 children (aged 1-12
years) per year found by Glezen and Denny in a prospective study with
relatively frequent medical attention for participating families.29 Our
data are in accord with U.S. National Health Survey data which show that
although chronic bronchitis prevalence varies inversely with the education
of the head of the family for persons aged 17 years and over, chronic bron-
chitis prevalence actually increases with parental education for children
under seventeen years of age.59 Children with a history of asthma would be
expected to have an increased risk of lower respiratory disease. Not only
was an increased risk found in this study, but it generally was intermediate
among children with Inactive asthma and highest among those with active asthma.
Asthma prevalence differed by sex among children and adults as expected.
The relation of cigarette smoking to education of the head of the household,
age, and sex, and the relationship of chronic respiratory disease symptom
prevalenc to smoking and a history of occupational exposure in parents of
these children also conformed to expectation (results not reported in this
report). Return rates were excellent in both cities and missing information
was minimal. Black-white morbidity differences were found in both communities
and were confirmed in a recent report which utilized virtually the same
questionnaire.2** Yet similar patterns of morbidity with respect to age,
education of the head of the household, and a history of asthma were found
in both black and white children.
In summary, the evidence for the validity and reliability of the question-
naire is as follows: empirical PVpos and FVneg in accord with expectation (in
two studies) given the uncertainties of these two statistics; the comparability
of the rates in this study to those found in a prospective study, the agree-
ment with U.S. National Health Survey data regarding bronchitis in children
and education of the head of the household; the increased risk of lower
respiratory disease following a gradient among nonasthmatic, inactive, and
asthmatic children; the expected relationship of asthma prevalence to sex
among children and adults of both races; the expected relationships among
cigarette smoking, age, sex, history of occupational exposure, and chronic
bronchitis symptom prevalence in the parents of these children; the excellent
return rates and low missing information rates, and the consistent black-white
morbidity differences which have been confirmed in another study. All of
these facts argue strongly for the reliability and the validity of the ques-
tionnaire.
110
-------
TABLE C-l. THEORETICAL EXPECTED POSITIVE AND NEGATIVE PREDICTIVE
VALUES UNDER VARYING SENSITIVITY AND SPECIFICITY AND A
TRUE PREVALENCE (pt) OF 10% or 20%.
a. Sensitivity and Specificity Variable
Sensitivity -
Specificity
50%
75%
90%
PV
pt = 10%
10%
25%
50%
b. Sensitivity = 75%
Sensitivity = 75%
Specificity
50%
75%
90%
PV
pt = 10%
14%
25%
45%
Positive
Pt = 20% Pl
20%
43%
69%
PV
t =
90%
96%
99%
Negative
10% pt = 20%
80%
92%
97%
, Specificity Variable
Positive
Pt - 20% PI
27%
43%
65%
c. Sensitivity Variable, Specificity =
Specificity = 75%
Sensitivity
50%
75%
90%
PV Positive
pt = 10%
18%
25%
29%
Pt - 20% PI
33%
43%
47%
PV Negative
t =
95%
96%
97%
75%
10% pt = 20%
89%
92%
94%
PV Negative
t =
93%
96%
99%
10% pt = 20%
86%
92%
97%
111
-------
APPENDIX D
OBSERVED MORBIDITY RATES
Actual reported morbidity rates by age, sex, and education of the head
of the household are given here for reference (Tables D-l through D-10).
112
-------
TABLE D-l. "ANY LOWER RESPIRATORY DISEASE": REPORTED FOUR YEAR FREQUENCY AMONG BLACK, NONASTHMATIC
CHILDREN WITH THREE OR MORE YEARS OF FAMILIAL COMMUNITY RESIDENCE
a. One or More Episodes
Education,
Sex Head of
Household
HS
Education,
Sex Head of
Household
H$
HS
City
Charlotte
Birmingham
Charlotte
Birmingham
Charl otte
Birmingham
Charlotte
Birmingham
City
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
1-4 Years
/
21.7 13/60)
20.5(24/117)
24.3(18/74)
31.5(28/89)
25.4(18/71)
20.0(23/115)
38.8(19/49)
23.3(21/90)
b. Two or More
1-4 Years
10.0(6/60)
6.8(8/117)
12.2(9/74)
20.2(18/89)
18.3(13/71)
5.2(6/115)
22.4(11/49)
15.6(14/90)
5-8 Years
9-12 Years
16.8(32/190) 11.7(28/239)
20.8(52/250) 17.5(57/325)
17.7(25/141) 11.9(19/159)
19.6(45/230) 18.9(43/228)
13.1(21/160) 14.3(29/203)
16.7(46/276) 13.3(43/323)
17.6(24/136) 13.1(23/175)
24.7(56/227) 13.7(31/226)
Episodes
5-8 Years
6.3(12/190
10.8(27/250
7.8(11/141
13.0(30/230
9-12 Years
7.5(18/239)
7.7(25/325)
6.3(10/159)
) 10.1(23/228)
6.9 11/160) 8.9(18/203)
8.3 23/276) 6.2(20/323)
8.1 11/136) 6.9(12/175)
13.7(31/227) 7.5(17/226)
All Ages
14.9(73/489)
19.2(133/692)
16.6(62/374)
21.2(116/547)
15.7(68/434)
15.7(112/714)
18.3(66/360)
19.9(108/543)
All Ages
7.4(36/489)
8.7(60/692)
8.0(30/374)
13.0(71/547)
9.7(42/434)
6.9(49/714
9.4(34/360
11.4(62/543
-------
TABLE D-2. CROUP: REPORTED FOUR YEAR FREQUENCY AMONG BLACK, NONASTHMATIC CHILDREN WITH THREE OR MORE
YEARS OF FAMILIAL COMMUNITY RESIDENCE
a. One or More Episodes
Sex
Female
Male
Sex
Female
Male
Education,
Head of
Household
-------
TABLE D-3. BRONCHITIS: REPORTED FOUR YEAR FREQUENCY AMONG BLACK, NONASTHMATIC CHILDREN WITH THREE
OR MORE YEARS OF FAMILIAL COMMUNITY RESIDENCE
a. One or More Episodes
Ul
Sex
Female
Male
Sex
Female
Male
Education,
Head of
Household
HS
-------
TABLE D-4. PNEUMONIA:
REPORTED FOUR YEAR FREQUENCY AMONG BLACK, NONASTHMATIC CHILDREN WITH THREE
OR MORE YEARS OF FAMILIAL COMMUNITY RESIDENCE
a. One or More Episodes
Sex
Female
Male
Education,
Head of
Ho use hoi d
HS
City
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
Charlotte
Birmingham
1-4
13.3
8.5
10.8
*,18.0
15.5
13.9
16.3
18.9i
Years
(8/60)
10/117)
8/74)
16/89)
11/71)
16/115)
8/49)
[17/90)
b. Two or More
Sex
Education,
Head of
Household
City
1-4
Years
5-8 Years
8.9(17/190)
14.8(37/250)
7.1(10/141)
13.5(31/230)
7.5 12/160)
9.8 27/276)
9.6 13/136)
16.3(37/227)
Episodes
5-8 Years
9-12 Years
6.7(16/239)
11.7(38/325)
3.8(6/159)
11.0(25/228)
8.9(18/203)
9.9(32/323)
6.9(12/175)
8.4(19/226)
9-12 Years
All Ages
8.4 41/489)
12.3 85/692)
6.4 24/374)
13.2(72/547)
9.4(41/434)
10.5(75/714)
9.2(33/360)
13.4(73/543)
All Ages
H$ Charlotte
Birmingham
5.0(3/60)
4.3
5.4
6.7
5/117)
4/74)
6/89)
3.2(6/190)
5.2
3.5
6.1
13/250)
5/141)
14/230)
4.2(10/239)
4.0(13/325)
1.9(3/159)
5.3(12/228)
3.9(19/489)
4.5(31/692)
3.2(12/374)
5.9(32/547)
Male
-------
TABLE D-5. HOSPITALIZATION: REPORTED FOUR YEAR FREQUENCY AMONG BLACK, NONASTHMATIC CHILDREN WITH
THREE OR MORE YEARS OF FAMILIAL COMMUNITY RESIDENCE
a. One or More Episodes
Sex
Female
Male
Sex
Femal e
Male
Education,
Head of
Household
-------
TABLE D-6.
00:
"ANY LOWER RESPIRATORY DISEASE": REPORTED FOUR YEAR FREQUENCY AMONG WHITE, NONASTHMATIC
CHILDREN WITH THREE OR MORE YEARS OF FAMILIAL COMMUNITY RESIDENCE
a. One or More Episodes
Sex
Female
Male
Sex
Femal e
Male
Education,
Head of
Household
HS
-------
TABLE D-7. CROUP: REPORTED FOUR YEAR FREQUENCY AMONG WHITE, NONASTHMATIC CHILDREN WITH THREE OR MORE
YEARS OF FAMILIAL COMMUNITY
a. One or More Episodes
NO
Sex
Female
Male
Sex
Female
Education,
Head of
Househol d
-------
TABLE D-8. BRONCHITIS: REPORTED FOUR YEAR FREQUENCY AMONG WHITE, NONASTHMATIC CHILDREN WITH THREE OR
MORE YEARS OF FAMILIAL COMMUNITY RESIDENCE
a. One or More Episodes
Sex
Female
Male
Sex
Female
Male
Education,
Head of
Household
-------
TABLE D-9. PNEUMONIA: REPORTED FOUR YEAR FREQUENCY AMONG WHITE, NONASTHMATIC CHILDREN WITH THREE OR
MORE YEARS OF FAMILIAL COMMUNITY RESIDENCE
a. One or More Episodes
Sex
Female
Male
Sex
Female
Male
Education,
Head of
Household
HS
-------
TABLE D-10.
HOSPITALIZATION: REPORTED FOUR YEAR FREQUENCY AMONG WHITE, NONASTHMATIC CHILDREN WITH
THREE OR MORE YEARS OF FAMILIAL COMMUNITY RESIDENCE
a. One or More Episodes
Sex
Education,
Head of
Household
City
1-4 Years 5-8 Years 9-12 Years All Ages
N>
Male
HS
Charlotte
Birmingham
Charlotte
Birmingham
4.4(4/90)
4.6(3/65)
5.9(9/153)
9.6(8/83)
3.0
3.8
2.7
9.6i
7/230)
6/158)
10/369)
22/230)
1.0(3/298)
1.6(3/193)
1.3(6/479)
2.3(6/263)
2.31
2.9
2.5
6.3
[14/618)
12/416)
25/1001)
36/576)
b. Two or More Episodes
Sex
Education,
Head of
Household
City
1-4 Years
5-8 Years
9-12 Years
All Ages
-------
APPENDIX E
QUESTIONNAIRE USED IN STUDY
123
-------
FORM APPROVED
Budget Bureau
(CARD „
SCHOOL AND FAMILY HEALTH QUESTIONNAIRE
FAMILY SURNAME:
(COL. 9-26)
ADDRESS:
(COL. 29-59)
TELEPHONE:
(COL. 60-66)
Please write on the lines above
your family's surname (last name),
address, and telephone number.
The Information requested In this
Questionnaire will be held in strict
confidence. It will be averaged for
groups of people only.
(COL. 79-60)
I Q I tj
124
-------
I. HEALTH QUESTIONS CONCERNING FATHER AND MOTHER ONLY
(CARD 2)
NOTE: The questions in this section are to be answered for both
parents by the mother or female guardian. Answer the
questions in this section only for parents (or guardians)
living In the home.
1. Do you usually cough first thing in the morning in
winter? (Count two or more coughs upon arising, or
when you first go out of doors, or when you smoke
the first cigarette of the day. Do not count clearing
of throat.)
2. Do you usually cough during the day or night in
winter? (Do not count an occasional cough.)
IF YOU ANSWERED "YES" TO QUESTION 1 OR 2,
PLEASE ANSWER QUESTION 3.
3. Do you cough like this on most days or nights for
as much as three months each year?
MOTHER (or
female guardian)
(COU. 91
Yes No
a a
I 2
(COL. 11)
Yes No
a a
1 2
(COL. 13)
Yes No
i :.i r:..]
1 2
FATHER (or
male guardian)
(COL. 10)
Yes No
n n
i
(COL. 12)
Yes No
n n
1 2
(COL. 14)
Yes No
LT.I a
1 2
125
-------
U. Do you usually bring up phlegm (thick fluid) from
your chest first thing in the morning in winter?
(Count phlegm whether swallowed or expelled, upon
arising, or when you first go out of doors, or when
you smoke the first cigarette of the day. Do not
count phlegm from nose.)
5. Do you usually bring up phlegm from your chest
during the day or night in winter?
IF YOU ANSWERED "YES" TO QUESTION 4 OR 5,
PLEASE ANSWER QUESTION 6.
.
6. Do you bring up phlegm like this on most days
for as much as three months each y-ear?
7. Do you get short of breath walking on level
ground at an ordinary pace?
8. How many cigarettes do you usually smoke now?
Never smoked.
Ex-smoker
Less than 1/2 pack per day (1-5 cigarettes)
About 1/2 pack per day (6-14 cigarettes)
About 1 pack per day (15-25 cigarettes)
" About 1-1/2 packs per day (26-34 cigarettes).
About 2 or more packs per day (35 or more claarettes)
MOTHER (or
female guardian)
(COL. 15)
Yes No
L! [.'!
1 2
(COL. 17)
Yes No
n IT
1 2
(COL. 19)
Yes No
111 LJ
1 2
(COL. 21)
Yes No
r:: t: i
1 2
(COL. 23)
n
n
2
n
3
n
4
n
s
r i
6
r
7
FATHER (or
male guardian)
(COL. 16)
Yes No
i: ; • a
1 2
(COL. 18)
Yes No
a r1
1 2
(COL. 20)
Yes No
LJ n
1 2
(COL. 22)
Yes No
D a
1 2
(COL. 241
n
in
2
n
3
r !
4
n
s
\-\
6
n
7
126
-------
FATHER-MOTHER HEALTH QUESTIONS CONTINUED
9. At your job, are you now or have you been
frequently exposed to irritating smoke, dust,
or fumes? (Do not include neighborhood or
home exposures.)
IF YOU ANSWERED "NO" TO QUESTION 9, SKIP
THE NEXT THREE QUESTIONS BELOW.
9a. If the answer to question 9 is "yes," what
kind of irritant were you exposed to? (For
example: coal dust, cutting oils, asbestos,
mine dust, smelter fumes, raw cotton dust.)
9b. If the answer to question 9 is "yes," what
kind of work did you perform in this |ob?
(For example: miner, maintenance, assembly
line, supervisor.)
9c. If the answer to question 9 is "yes," how
long were you exposed?
Less than 1 year.
1 to 5 years
6 to 10 years..
More than 10 years.
MOTHER (or
female guardian)
(COL. 25)
Yes No
n n
FATHER (or
male guardian)
(COL. 26)
Yes No
n n
(COL. 27)
(COL. 2Q.)
3
.C)
4
3
d
4
(COL. 29)
(COL. 30)
DO NOT MARK THESE BOXES
Yes
CD
No
CJ
Yes
CJ
No
ci
127
-------
(COL. 79-80)
rsm
IQ^Where did you live the longest when- you were
Wb to 20 years of age? (Check only one choice
for each parent.)
Present city, present neighborhood
Present city, different neighborhood
Different city or area of more than 200,000 people
Different city or area of about 50,000 to 200,000
people
Dif'eienl city or area - A small city or town
Different city or area - A rural area or farm •-
11. Where did you live the longest when you were
21 to 30 years of age? (Check only one choice
for each parent . )
Not yet 91 years old
Present city, present neighborhood .
Present city, different neighborhood
Different city or area of more than 200,000 people
Different city or area of about 50,000 to 200,000
people
Different city or area - A small city or town _ ...
Different city or area - A rural area or farm
12. Where did you live the longest after you were 31
years of age? (Check only one choice for each
parent.)
Not yet 31 years old
Preaent city, present neighborhood _
Present city, different neighborhood
Different city or oroa of more than 200,000 people
Different city or area of about 50,000 to 200.000
people
Different city or area - A small city or town
nUfprfint rlt\t nr fima A rural flrAfl ni* farm
MOTHER (or
female guardian)
(COL. 31)
1 i
1
r1
2
r^
3
n
4
ri
5
1 1
6
(COL. 33)
n
1
i-i
2
•n
3
n
4
n
5
M
6
ri
7
(COL. 35)
rn
i
1-1
2
1 1
3
r i
4
,-.
5
fl
6
r
7
FATHER (or
male guardian)
(COL. 32)
r •
i
i"
2
r
3
ri
4
r-
i
,--i
6
(COL. 34)
r.1
i
n
2
n
3
n
4
n
9
n
6
n
7
(COL. 39)
o
1
c
2
n
3
n
4
n
s
ri
6
ri
7
l?ft
-------
•RITE AT THE HEM) OF EACH »
COLUMN THE NAME OF EACH
CWLO 12 YEARS OF AGE OR
YOUNGER.
THEN. M EACH CHILD'S COLUMN
CHECK ONE BOX TO ANtVEH EACH
OF THE FOLLOVWe QUESTIONS
1. Hatt ^w cMM beMei tvMtad by
•orjanininrila?
2» Has yow child bean tooated by
tor an attack of crevp?
3. Has your child been treated by
a doctor since September 1967 tor
oHUs. or otier daap chart In-
fcciioa?
4. Has yaw child been bUhe
hospital since September 1967
tor one ol the Illnesses Mention-
ed In questions 1. 2. or 3 above?
9. MM la the full name and address
•1 OM doctor who takes care ol
yow CMM or. If you take your
cMM to a clinic for medical car*
•hat I* the name and address of
*• airier
M_^_ ol
latcMM
lent, in
QNo
O Yes. once
2
a Yes. twice
i
D Yea. more then
• twice
ICOL. ID
DNo
CD Yes. once
z
D Yes. twice
i
0 Yea. more Van
« twice
ON."01"1"
i
O Yes. once
2
O Yes. twice
i
d) Yes, more than
4 twice
ICOL. 14)
D*»
t
D Yes. once
2
D Yes. twice
i
CD Yes, more than
4 twice
ICOL. 1S-M)
Name of
2nd child
ICOL. 1*
D No
CD Yes, once
i
D Yes. twice
9
D Yes, more than
4 twice
ICOL. an
DNo
DYes. once
2
D Yes. twice
i
D Yes. more then
4 twice
(COL. 21)
DNo
t
DYes. once
2
D Yes, twice
j
D Yes, more than
• twice
(COL. 221
CD No
t
O Yes, once
2
D Yes, twice
i •-
D Yes, more than
« twice
ICOL, 29-24)
ICOt. 2MB fJQ
Name of
3rd child
ICOL. 271
CD"0
D Yes. once
2
CD Yea, twice
t
D Yes. more than
4 twice
ICOL. ast
o*>
Q Yes. once
2
OYes. twice
9
O Yes. more than
4 twice
(COL. 29)
DNo
i
QJ Yes. once
2
C3 Yes. twice
9
Q) Yes. more than
4 twice
ICOL. 90)
(UNO
i.
Q Yes, once
2
Q Yes. twice
9
Q Yes, atom than
4 twice
ICOL.H.HI
•"•<
Name of
4Hi child
(COL. 95)
(D No
O YeV, once
OY«^! twice
9
Q Yes, more than
4 twice
ICOL. 9E)
CD No
CD Yes. once
2
CD Yes, twice
9
CD Yes, more than
4 twice
ICOL. 97)
nn°
j_] Yes. onco
2
CD Yes. twice
9
CD Yes, more than
4 twice
ICOL. 99)
CD No
i
CD Yes, once
2
CD Yes, twice
9
CD Yes, more than
4 twice
ICVL. 9t-4»
Name of
Sth child
ICOL. 49)
CD No
CD Yes, once
2
CD Yos. twice
9
CD Yes. more than
4 twice
ICOL. 44)
CD No
CD Yes. once
2
CD Yes. twice
9
CD Yes, more than
4 twice
(COL. 49)
CD No
i
CD Yes. once
2
CD Yes, twice
9
CD Yes. more than
4 twice
ICOL. «)
CD No
1
CD Yes, once
2
•Q Yes, twice
9
CD Yes. more than
4 twice
ICOL. 47-48)
ICOL.4MQ3
Im^^BBl ttf
ethchlM
ICOL. 911
CD NO
CD Yes. once
2
CD Ye*, twice
9
CD Yes. more than
4 twice
ICOL. 52)
CNo
CD Yes. once
2
CD Yes. twice
9
CD Yes. more than
4 twice
(COL. 99)
CD NO
1
CD Yes, once
2
CD Yes. twice
9
[~) Yes. more than
< twice
ICOL. 94)
CD NO
1
CD Yes. once
i
CD Yes. twice
9
CD Yes. more than
4 twice
ICOL. 3S-SS)
ICOL. S7-99 1[_J_]
Name of
7lh child
ICOL. XH
(UNO
CD Yes, once
2
Q Yea, twice
i
CD Yes, more than
4 twice
(COL. 60)
CD NO
CD Yes. once
2
CD Yes, twice
9
CD Yes, more than
4 twice
(COL. 81)
CD No
i
CD Yes. once
2
CD Yes, twice
9
I ; Yes. more than
« twice
ICOL. 82)
CD No
t
C3 Yes. once
2
CD Yes. twice
9
CD Yet, more than
4 twice
ICOL. «M4>
Name of
8th child
(COL. 67)
DNo
D Yes, once
2
D Yes, twice
9
D Yes, more than
4 twice
ICOL. 60)
CD No
D Yes. once
2
D Yes. twice
9
CD Yes. more than
4 twice
ICOL. e»i
DNo
i
D Yes, once
2
D Yes. twice
9
D Yes, more than
4 twice
ICOL. 701
DNo
i
D Yes. once
2
D Yes. twice
9
D Yes, more than
4 twice
ICOL. 7>-7»
PLEASE CHECK
BOX BELOW IF
THERE ARE MORE
THAN 8 CHILDREN
12 YEARS OF AGE
OR YOUNGER.
ICOL. 79)
CD
ICOL. TWO)
rem
ro
10
-------
III. GENERAL QUESTIONS CONCERNING THE HOUSEHOLD
(CARD 41
1. What educational level has been completed
by the head of the household? (Check one
box only.)
(COL. 9)
D Elementary school
i
CD Part of high school
2
C High school graduate
3
LU Part of college
4
C~.' College graduate
s
L. Graduate school
6
d'.l Other
(please specify)
2. What is the present employment status of
the mother (or female guardian)? (Check
one box only.)
(COL. 10)
£2 Housewife or employed at home
t
d] Employed outside the home (full or
2 part time)
Q"! Currently unemployed but usually
3 employed outside the home.
L7J Other
1 (please specify)
(COL. 11)
3. What is the race of the family?
CD Indian
CD Mexican-American or Spanish American
2
CD Negro
3
CD Oriental
4
CD Wnite
5
CD Other
6
4. How many times have you and your
family changed living quarters during
the last 5 years? (Check one box
only.)
(COL. 12)
073 Zer°
i
L71 One
2
D Two
3
Q ' Three
4
Four
Five or more
130
-------
I COL. 79-801
I 0| 4|
(COL. 13-14)
5. How long has the family lived in
your present city or town? (Check
one box only.)
[~ i Less than
01 1 year
Q: 1 year
02
[3] 2 years
03
[7] 3 years
04
4 years
OS
5 years
06
07
08
| 6 years
7 years
8 years
09
i 9 years
"
1
11
12
10 years.
11 years
[H 12 years or
13 more
6. How many rooms are there in .your
living quarters? (Do not count
bathrooms, porches, balconies,
foyers, halls or half-rooms.)
(Check one box only.)
(COL. IE)
CD o°e
i
Q'J Two
2
d Three
3
P
4
d Six
6
Q] Seven
7
D Eight
e
Q] Nine or more
9
(COL. 16)
7. Do you have air conditioning in
your living quarters?
(Check one box only.)
n N°
i
Yes, window only
Yes, central
131
-------
IV. CENSUS OF HOUSEHOLD
VI* . OKC CMO PCH PEH9OM
<*>
ro
NSMS of persons livina
»• •» household
(•rile Hi the column below
me first and middle name
ol each person living In
VkV MBOSGnoML)
ICOL. »-•» ICOL. 11-39
01.
02.
03.
•*
OS.
OB.
0».
OB.
01.
10. '
II.
•t.
IS.
14.
'*
Asthma among persons living
In the household
(Check the appropriate box
or boxes for each person.)
Has this person
ever had asthma
diagnosed
by a doc tor?
Yea No
ICOL. 3U
0 0
1 X
CJ CJ
CJ d
1 2
CJ £D
1 2
tl CI
1 2
CJ CJ
1 2
CJ Ci
1 2
CJ CI
1 2
rj o
1 2
c: a
1 2
c: a
1 2
CJ CJ
1 2
17 C)
1 2
CJ CJ
1 2
C. CJ,
1 2
Has this asthma
been active In the
past two years?
Yes No
ICOL. 371
n a
1 2
CJ CJ
1 2
cj a
1 2
CJ CJ
1 2
C CJ
1 2
cr a
1 2
CJ Ci
1 2
CD en
I 2
CJ CJ
1 2
C CJ
1 2
cj n-
1 2
C? C!
1 2
c: a
1 2
L7J C"
1 2
IT r;
t 2
Chronic heart or lung disease
among persons living In the
household
(Check "yes" (or each person
only If chronic heart or chronic
lung disease was ever diagnosed
by a doctor).
Chronic
heart
disease
Yes No
ICOL. 301
1 2
c? n
' 2
c- a
1 2
CJ CJ
' 2
C7 CJ
' 2
C' CJ
1 2
rr c;
1 2
cj a
1 2
C" CJ
1 2
17 c;
1 2
L" c
| 2
C C1
1 2
L" c:
1 2
<: n
rr c.i
1 2
Chronic
Lung
disease
Yes No
ICOL. 3*1
1 2
"CJ CJ
• 2
CJ CJ
1 2
C? CJ
1 2
CJ CJ
1 2
C.' C?
1 2
LI' CJ
1 2
C1 CJ
1 2
L' C?
1 2
c: c1
1 2
cr c:
1 2
c: i_
1 2
C: CJ
t 1
[--• 1-
1 2
IV Ci
1 2
Sex ' ol persons living
^ jn ihe>-Hisehold
(Check b elow one appropriate
box for e ach person.)
MaleL Female
ICOL. 401
c: n
1 2
CJ 0
1 2
CJ CJ
1 2
CJ CJ
1 2
n D
1 2
1 n • ' cn
1 2
Li; a
1 2
d1 CJ
1 2
CJ CJ
' 2
r1 cr
1 2
' i 2
c: j cj
' 1 2
'"- » LTJ
< 2
1 ' 2
n c?
1 2
Age of persons living
In Ihe household
Complete yew of birth and age
for each person as Indicated
below.
Age In years
Year of Mrlh at last birthday
ICOL. 41-421
1 1
1 1
1 1
1 1
1 1
II
1 1
II
1 1
II
II
II
II
II
II
ID LJ
II II
ID EU
ID EH
I] CJ
ID D
nj cu
Zl CJ
II II
JJ CJ
JJ EU
JJ CJ
JJ CJ
JJ CJ
JJ CJ
Position In family ol persons
living In the household
(Check below one appropriate box for each person)
Head
of
household Spouse Child Other lcoc
tCOL. 4* *»-«»
CJ CJ CJ CJ E3
t 234
cj a CD a i»)
1 234
c: cj n cj [og
1 234
CJ CJ CD CJ 13
1 234
IT1 CJ CJ CJ i^J
' 234
D CJ CJ C! LaJ
1 234
CJ CJ CJ CJ Q3
1 214
n cj a cj on
1 234
CJ O CJ D 03
1 234
a n cj a Q3
1 234
CJ CJ CJ CJ 03
1 234
CJ CJ CJ CI LU
1 234
CJ CJ CJ CD 03
1 234
D CJ CJ CJ OH
1 234
r:.:i c..i cj CJ OH
1 234
THANK YOU FOR YOUR COOPERATION
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/1-77-043
2.
4. TITLE AND SUBTITLE
RESPIRATORY DISEASE IN CHILDREN EXPOSED TO SULFUR
OXIDES AND PARTICULATES
7. AUTHOR(S)
Douglas Ira Hammer
3. RECIPIENT'S ACCESSION-NO.
6. REPORT DATE
September 1977
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAMF AND ADDRESS
Population Studies Division
Health Effects Research Laboratory
Research Triangle Park, N.C. 27711
10. PROGRAM ELEMENT NO.
1AA601
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Health Effects Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park. N.C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
HERL-RTP
14. SPONSORING AGENCY CODE
EPA 600/11
16. SUPPLEMENTARY NOTES
Originally submitted to the Harvard School of Public Health in partial fulfillment of
the requirements for the degree of Doctor of Public Health in Epidemiology
16. ABSTRACT
Acute lower respiratory disease was surveyed by questionnaire among parents of 10,000
children aged 1 to 12 years in two Southeastern communities representing intermediate
and high exposures to particulates and low sulfur dioxide levels. Morbidity reporting
patterns with respect to age, parental education, and history of asthma were similar
for blacks and whites, but the frequency of pneumonia was significantly lower, and
the frequencies of croup, bronchitis, and "any lower respiratory disease" were
significantly higher among whites in both communities. Significant increases of any
lower respiratory diseases and hospitalization were found among children in the high
exposure community.
Asthma rates clustered in families, were higher in male children and female parents,
and were comparable to other studies. Significant increases of lower respiratory
disease were also found among asthmatic children in the high exposure community.
Difference in parental recall, family size, or parental cigarette smoking were not
likely explanations for the excess morbidity in the high exposure community.
Therefore, these results associate excess acute lower respiratory disease in children
with exposure to elevated particulate levels and low sulfur dioxide concentrations.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Te~spTrsntbry diseases
sulfur oxides
particles
air pollution
b.lDENTIFIERS/OPEN ENDED TERMS
c. cos AT I Field/Group
06, t
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
148
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
133
------- |