AIR POLLUTION, AIRWAYS HYPERREACTIVITY, AND
PULMONARY FUNCTION MEASUREMENTS IN ASTHMA
by
Sheldon L. Spector and Robert A. Kinsman
Melissa Dunning
James Murphy
Richard Jones
James P. Lodge
Helen Rose
National Jewish Hospital and Research Center
Denver, Colorado 80206
Environmental Protection Agency Contract No.
68-02-2517
Project Officer
Charles Stevens
Office of Air Programs
Environmental Protection Agency
Denver, Colorado
National Jewish Hospital and Research Center
3800 East Colfax Avenue
Denver, Colorado 80206
-------
DISCLAIMER
This report has been reviewed by the G'FFlCB OF f{[R PRO(S^/4/^lS ,
U. S. Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and policies
of the U. S. Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation for use.
-------
ACKNOWLEDGEMENT
We gratefully acknowledge the statistical assistance of Helen Rose, and
the assistance of Sharon Robinson and Jolene Fukuhara during various aspects
of the study. Carolyn Wangaard and Cynthia Westmoreland dedicated consider-
able time assembling large amounts of medical information for the patients
studied.
The Colorado Department of Health Air Pollution Surveillance Section
cooperated consistently throughout this project, providing technical advice
and all daily air pollution data for the study. We particularly appreciate the
cooperation of Mr. Robert D. Siek, Assistant Director for Environmental Affairs,
and Dave Rose and Steve Arnold, Senior Air Pollution Specialists from that
office.
Indoor-outdoor air pollution monitoring was provided by Pedco Environ-
mental Specialists, Inc. of Cincinnati, Ohio working under EPA contract No.
68-02-2294. Larry Elfers of Pedco supervised this work and S. David Shearer,
Director of the Environmental Monitoring and Support Laboratory of EPA faci-
litated cooperation between these projects.
Finally, the Department of Biometrics of the University of Colorado
Medical Center, headed by Dr. Strother Walker, provided invaluable statistical
expertise for the analyses of the data.
ii
-------
CONTENTS
Page
Disclaimer i
Acknowledgement ii
List of Tables iv
1. Abstract 1
2. Introduction 2
3. Methods 3
3.1 Periods of Study 3
3.2 Patient Characteristics 4
3.3 Measures 4
3.3.1 Air Pollution and Meterologic Variables 4
3.3.2 Pulmonary Function Measurements 5
3.3.3 Indices of Airways Hyperreactivity 6
3.4 Statistical Analyses 7
4. Results 8
4.1 Air Pollution and Meterologic Variables 8
4.2 Pulmonary Function Measruements 12
4.3 Pulmonary Function Measurements and Air Pollution/Meterologic
Variables 15
4.3.1 Univariate Regression Analyses IS
4.3.2 Multivariate Regression Analyses 16
5. Discussion 26
6. Recommendations 30
7. References 31
iii
-------
LIST OF TABLES
Page
4.1 Means and Standard Deviations for All Air Pollution and Metero-
logic Variables by Month for the Summer (A) and Winter (B)
Periods ........ 9
4.2 Frequency Distributions of Air Pollution and Meterologic Variables
for Summer and Winter Periods 10
4.3 Intercorrelations Among Air Pollution and Meterologic Variables for
the Summer (A) and Winter (B) Periods 11
4.4 Frequency Distributions of Mean Pulmonary Function Measurements for
Patients During Summer and Winter Periods . 12
4.5 Means and Standard Deviations of the Correlations Among Pulmonary
Function Measurements for the Summer and Winter Periods 13
4.6 Frequency Distributions of Correlations Between Pulmonary Function
Measurements and Day of Study for Summer and Winter Periods .... 14
4.7 Univariate Regression Analyses Summary Table for Individual
Regressions 15
4.8 Variables Included in the Leaps and Bounds Multiple Linear Regres-
sion Analyses 17
4.9 Summary Table for Leaps and Bounds Multiple Linear Regression
Analyses for Individual Patients 18
4.10 Leaps and Bounds Multiple Linear Regression Analyses Summary Table
for Summer Period Summarizing Best Regressions for Airways Hyper-
reactivity Groups . 20
4.11 Leaps and Bounds Multiple Linear Regression Analyses Summary Table
for Winter Period Summarizing Best Regressions for Airways Hyper-
reactivity Groups 23
iv
-------
1. ABSTRACT
This study evaluated the relationship between spirometric pulmonary func-
tion measures (FEVj, FEVjyFVC%, and PEFR) and air pollution and meterologic
variables (CO, SO2, total hydrocarbons, O3, coefficient of haze, minimum
temperature, and barometric pressure)for adult asthmatic inpatients in an
urban, long-term intensive treatment center.
Two groups of 67 patients were studied during a late spring-summer (May 1
through August 12, 1975) and a late fall-winter (November IS, 1975 through
March 31, 1975) period. Patients were classified according to the role of
allergic factors and the degree of airways hyperreactivity.
In all analyses, 24-hour average levels of air pollution preceding pul-
monary function measurement were related to subsequent 8:00 am pulmonary func-
tions. Multiple linear regression analyses by leaps and bounds for individuals
and groups, defined by reaginic factors and airways hyperreactivity, failed
to identify a systematic effect for any air pollution variable upon any pul-
monary function measure. The lack of any relationship does not appear attri-
butable to limited exposure to air pollution variables, but may be affected
by as-needed (PRN) medication usage, type of pulmonary function measurement
employed, and periods selected for analyses. The relative strengths and
limitations of the study are discussed.
This report was submitted in fulfillment of Contract No. 68-02-2517 by
National Jewish Hospital and Research Center under the sponsorship of the U.S.
Environmental Protection Agency. This report describes work completed during
the contract period September 1, 1976 to March 1, 1977.
1
-------
2. INTRODUCTION
Air pollution has often been implicated in aggravating respiratory ill-
ness, but its effect upon asthmatic patients may be particularly significant
due to the hyperreactivity of their airways. Airways hyperreactivity has
been regarded as among the definitive features of asthma, predisposing these
patients to react to a wide range of known irritants with increased airway
obstruction.1,2 Histamine and methacholine inhalation challenges enable dif-
ferentiation of asthmatics from patients who wheeze due to other causes, and
permit classification of patients according to the degree of airways hyper-
reactivity.3"5 The degree of airways hyperreactivity has been related to
exercise-induced asthma,amount of oral corticosteroids prescribed at dis-
charge from long-term treatment,4 and the response to bronchoconstrictive
suggestion.® Likewise, those patients with the most hyperreactive airways may
be affected most by irritant air pollution variables (e.g., SO2 and O3).
There is a considerable literature on the effects of air pollution upon
respiratory illness. Many early studies have been epidemiologic, enumerating
clinic and hospital visits and morbidity during periods of significant urban
air pollution. ~12 These epidemiological studies have often made retrospective
determinations of the incidence of respiratory illness and lack precise infor-
mation about the types of patients affected, but they have generally agreed
that increased clinic and hospital visits and morbidity are associated with
episodes of heightened urban air pollution.9"12 In contrast, Greenberg, et al.
found no increase in difficulties among asthmatics (as indexed by clinic and
hospital visits and morbidity) in New York City for a high pollution period
in the fall of 1952 during which the observed increase in air pollution levels
could not be related to pollens, SO2, COH, or CO.13 Such negative findings may
sometimes result from the mode of analyses. For example, Ribon, Glasser, and
Sudhivoraseth14 observed no increase in medical visits and morbidity during
a period of high air pollution in New York City. However, they related clinic
and office visits to air pollution levels on the same day and thereby may have
failed to allow for the delay in symptom appearance and the subsequent decision
or need to seek treatment. Lawther and Waller13,^ have noted that the physio-
logical response to air pollution may lag, exacerbating bronchitis only after
a period of relatively prolonged exposure to high levels of urban smog.
Increased medical visits have also been associated with periods of low
temperatures or decreases in temperature.1?-1® Laboratory studies with asth-
matic patients have suggested that inhalation of cold air, and not mere cold
exposure, results in increased airway resistance.2® Additionally, Salvaggio
and his colleagues have suggested that seasonal factors, such as low tempera-
tures and airborne pollens, may lead to increased asthma attack rates in New
Orleans, an effect easily misattributable to increased levels of air pollutants. 1
2
-------
More recently, prospective studies in Salt Lake City and New York City have
measured the effect of specific pollutants (notably SO2, particulates, and sus-
pended sulfates) on the reports of asthma attacks within preselected groups of
asthmatic patients,22, 23 in these studies, increased asthma attack rates were
associated with cooler temperatures, increased suspended particulates, and sul-
fates. Surprisingly, cooler temperatures did not increase the impact of air
pollution levels in either study, but instead the effect of sulfates and par-
ticulates were more pronounced on warmer than cooler days. Since both air
pollutants and temperature appear to affect asthma attacks, this apparent
paradox may be due to decreased outdoor exposure on cold days. For both
studies, SO2 levels were not clearly related to attack rates. However, when a
major plant producing SO2 in Salt Lake City was closed during a strike, SO2
levels decreased abruptly and were accompanied by a pronounced decrease in the
asthma attack rates.22 These observations agree with those of others that SO2
appears to contribute to bronchoconstriction among individuals with hyperreac-
tive airways.24, 25
Systematic studies such as these have produced the clearest findings about
asthmatic symptoms and air pollution. However, neither study controlled for
the seasonal effect of airborne pollens, which alone could exacerbate symptoms
for patients with an allergic Ci*e•j reaginic) component to their asthma,21* -5
Furthermore, objective pulmonary functions have not yet been related to air
pollutants for a well-defined group of asthmatics with documented airways
hyperreactivity. The need for objective measurement is emphasized by studies
of subjective symptomatology in asthma,26, 27 which document the variety of
ways asthmatic patients report their distress.
As a preliminary project, this proposed study attempted to contribute use-
ful information by Q-) using a well-defined asthmatic population, C2) classi-
fying patients according to the role of reaginic factors and airways hyper-
reactivity, C3) considering the effect of temperature and barometric pressure
in all analyses, (4) using daily spirometric pulmonary functions as objective
measures of airway obstruction, and C5) minimizing the influence of airborne
pollens by studying separately nonallergic asthmatics as determined by skin
testing and antigen inhalation challenges.^-5
3. METHODS
3.1 PERIODS OF STUDY
Asthmatic inpatients were studied throughout two periods, May 1 through
August 12, 197S (spring-summer) and November 15, 1975 through March 31, 1976
Cfall-winter). For both periods, daily 8:00 a.m. spirometric pulmonary
tions and medications were available for all patients throughout treatment. 8
These periods were selected to maximize the range and variability of air
pollution and meteTologic variables. The principal variable of interest dur-
ing the late spring-summer period was ozone (O3), which achieves its maximum
levels during summer.29 Levels of all other air pollutants are highest
during the late fall-winter period, while lower temperatures also occur dur-
ing this period. During both periods, 67 patients were in residence at the
hospital for continuous periods of not less than two weeks. In addition,
3
-------
these patients included only those without significant gaps in their pulmonary
function data.
3.2 PATIENT CHARACTERISTICS
The patients were well-defined groups of 67 asthmatics for both the summer
and winter periods admitted to the intensive long-term treatment program at
National Jewish Hospital and Research Center (NJHRC) in Denver, Colorado.
Patients within periods were comparable with respect to age (mean +SD; 36
+_ 16 and 40 16, for summer and winter respectively) and sex (46 females and
48 females). Reaginic factors, identified by skin testing and antigen inhala-
tion challenges, were determined to be involved in the asthma of 37 patients
for the summer and 26 patients for the winter period.5,30 Tests for airways
hyperreactivity were available for all patients during the summer and 60
patients during the winter period. Of these, with the exception of 4 patients
during summer and 3 during winter, all had hyperreactive airways as deter-
mined by methacholine and histamine inhalation challenges.3-5 Experience with
these patients has indicated that nearly all (86%) could be expected to have
asthma exacerbated by infection.31
Skin testing and/or antigen inhalation challenges were available for 62
patients during the summer and 63 patients during the winter period. Since
only patients with asthma mediated by allergic (i.e., reaginic) factors should
be affected by increased airborne allergens, all analyses were performed with-
in groups of patients classified according to the role of reaginic factors in
their asthma. Specifically, patients were categorized into three groups:
(A) reaginic asthma, as determined by skin testing and antigen inhalation
challenges (summer: 14 patients, winter: 3 patients); (B) possibly reaginic
asthma, as determined by skin test results alone (summer: 23 patients; winter:
23 patients); and (C) not reaginic asthma (summer: 25 patients, winter: 37
patients). Testing was unavailable for 5 patients during summer and 4 patients
during the winter period; these patients were treated separately in all sub-
sequent analyses.
3.3 MEASURES
3-3.1 Air Pollution and Meterologic Variables
All air pollution variables were monitored at the Colorado Department of
Health Air Pollution Surveillance Substation located on the hospital grounds.
This substation was approximately 100 yards from the Andrew Goodman Building
where the adult asthmatic patients resided during treatment. Two weather
variables were also available on a daily basis: 8:00 am barometric pressure,
obtained from the NJHRC Pulmonary Function Laboratory and temperature (daily
average and minimum) obtained from two substations, Overland Park Golf Course
and Welby, approximately two miles southwest and northwest of NJHRC, respec-
tively. Therefore, average values from both substations were used.
In summary, the air pollution and meterologic variables studied were:
4
-------
Measure
Source
Units
1. Coefficient of haze (COH) NJHRC Substation
2. Carbon Monoxide (CO) NJHRC Substation
3. Sulfur Dioxide (SO2) NJHRC Substation
4. Total Hydrocarbons (THC) NJHRC Substation
5. Ozone (O3) NJHRC Substation
6. Minimum Temperature Welby + Overland
7. Barometric Pressure (BP) NJHRC
(8:00 am)
COHs/lOOO feet
parts per million (ppm)
parts per million (ppm)
parts per million (ppm)
parts per million (ppm)
degrees Celsius (C°)
mm of mercury (mmHg)
For coefficient of haze (COH), measurements were taken every two hours.
For all other air pollution variables, hourly measurements were available. An
initial step in data preparation involved calculating the mean level of the air
pollution variables for the 24-hour period preceding 8:00 am pulmonary func-
tion measurement. Fot temperature, the average and minimum values (Tmin)
for the same 24-hour period were used. The objective was to capture any
cumulative effect of air pollution for the 24-hour period immediately pre-
ceding pulmonary function measurement.
3.3.2 Pulmonary Function Measurements
For all patients, spirometric pulmonary function measurements (Vertek
5000 VF Lung Function Analyzer; Hewlett-Packard, Boston, Mass.; calibrated
biweekly with a 1.5 liter syringe) were obtained twice daily (8:00 am and
4:00 pm) throughout treatment. These measurements, which provide objective
indices of airway obstruction, include Forced Vital Capacity (FVC), measuring
the volume of air (liters) expired forcefully following a full inspiration;
First Second Forced Expiratory Volume (FEVj), measuring the volume of air
(liters) expired during the first second of FVC; and Peak Expiratory Flow
Rate (PEFR), measuring the maximum flow rate (liters/sec) attained during FVC.
Of these, FEV^ and PEFR are used as indices of airway obstruction.
As an additional index, FEVj was expressed as a percentage of FVC
(FEVi/FVC%). Normal values of FEV^/FVC% are generally between 75% to 85%,
indicating that this percentage of FVC is expired during the first second.
As airways obstruction increases, the decrease in the diameter of the airways
serves to reduce FEV]/FVC%, while FVC may remain unaffected. All pulmonary
function measurements, summarized below, were based on at least two reprodu-
cible determinations to insure stable and reliable measures for each occasion:
Measure
1. First Second Forced Expiratory Volume (FEV^)
2. FEVj as a percentage of Forced Expiratory
Volume (FVC) (FEVx/FVC%)
3. Peak Expiratory Flow Rate (PEFR)
Units
Liters
Percent
Liters/Second
On a daily basis, all measurements are verified and added to the computer
files of the Clinical Data Management System (CDMS) and subsequently available
for retrieval.
5
-------
3.3.3 Indices of Airways Hyperreactivity
Airways hyperreactivity was indexed by histamine and methacholine in-
halation challenges, performed following the recommendation of the National
Institute of Allergy and Infectious Disease panel on Standardization of
Bronchial Inhalation Challenge Procedures.^
Patients were asymptomatic at the time of the methacholine challenges as
indicated by baseline FEV^s of 80% or greater of the highest previously ob-
served value. Three minutes following inhalation of 5 breaths of diluent,
delivered by a Bell and Gossett Compressor (Model LCHB, Morton Grove, 111.)
adjusted to a flow rate of 6 liters per minute using a Vaponefrin nebulizer
(Fisons, New Bedford, Mass.), studies were repeated. This diluent value
served as a control, and testing proceeded if FEVi was not reduced by 10% or
more. Diluent was a sterile aqueous solution (pH - 7.0) of 0.036% potassium
phosphate, 0.107% sodium phosphate, 0.40% phenol, and 0.50% sodium chloride
obtained from Greer Laboratories, Lenoir, N.C. Pulmonary functions were per-
formed with a Medistor spirometer (Simlog Instruments, Redmond, Washington)
calibrated with a 1.5 liter syringe.
For the methacholine challenge, 5 breaths of serial dilutions of metha-
choline were given following the diluent measure. Spirometric measures were
made 3 minutes thereafter. A positive test was a greater than 20% reduction
in the 3-minute reading from the diluent value. If the test was negative, 5
breaths of the next dilution was given. Eight dilution increments were used:
0.15, 0.31, 0.62, 1.25, 2.5, 5.0, 10.0, and 25.0 mg/ml. The end point was
that methacholine concentration producing at least a 20% decrease in FEVi from
diluent control.
For histamine, 5 breaths of serial dilutions followed the diluent mea-
sure. Spirometry was performed 3 minutes thereafter. A positive test was
a 3-minute decrease in FEV^ of 20% or greater from the diluent value. If the
test was negative, 5 breaths of the next dilution were given. Eight dilution
increments were used: 0.06, 0.12, 0.25, 0.50, 1.0, 2.5, 5.0, and 10.0 mg/ml.
The end point was that histamine concentration producing at least a 20% de-
crease in FEVi from diluent control.
Methacholine inhalation challenges were available for all patients during
the summer period and 64 patients during the winter period. Using categories
established previously, 32 patients were classified during summer as having
low thresholds CO•IS and 0.31 mg/ml), 27 moderate thresholds CO.62, 1.25, and
2.5 mg/ml), and 17 as high thresholds C5.0, 10.0, and 25.0 mg/ml), with 4
patients having a negative methacholine challenge. During winter, 25 patients
were classified as having low thresholds, 27 moderate, and 7 as high thresholds,
with 5 patients having a negative methacholine challenge.
Histamine inhalation challenges were available for 66 patients during the
summer period and 60 patients during the winter period. During the summer
period, 18 patients had low thresholds (0.06 and 0.12 mg/ml), 27 had moderate
thresholds (0.25, 0.50, and 1.0 mg/ml) and 17 had high thresholds (2.5, 5.0,
6
-------
and 10.0 mg/ml), with 4 patients having a negative histamine challenge. Dur-
ing the winter period, 9 patients were classified as low, 30 as moderate, and
16 as high thresholds with 5 patients having a negative histamine challenge.
A final classification was based on the lowest threshold category of both
methacholine and histamine inhalation challenges. For this classification, 66
patients had complete testing during the summer period and 58 patients during
the winter period. During summer, 36 patients were classified within low,
23 within moderate, and 4 within high threshold categories, with 4 having
negative challenges. During winter, 26 patients were classified within low,
24 within moderate, and 5 within high threshold categories, with 3 patients
having negative challenges.
3.4 STATISTICAL ANALYSES
Separate analyses were performed for patients within both the summer
and winter periods.
(A) Univariate regression and correlational analyses were performed
for each individual within the summer and winter periods, relating
each 24 hour air pollution and meterologic variable to the follow-
ing 8:00 am pulmonary function measurements.
(B) The pattern of significant univariate regressions for individuals
was then examined within groups defined by reaginic factors and
airways hyperreactivity as indexed by methacholine and histamine
challenges.
(C) Multivariate regression analyses using all 7 air pollution and
meterologic variables and four concommittant variables (intra-
veneous bronchodilators given at the time of pulmonary function
measurement, bronchodilators taken within 2 hours prior to pul-
monary function measurement, bronchodilator treatments taken within
one hour prior to pulmonary function measurement, and the previous
day's 8:00 am pulmonary function) were performed next for all
individuals within the summer and winter periods using leaps and
bounds multiple linear regression procedures.32 These analyses
identified the best regression, defined as the regression pro-
viding the greatest gain in information (compared to no regres-
sion) indexed by the Akaike Information Coefficient (AIC).*3
(D) The pattern of best multivariate regressions for individuals were
then examined within groups defined by reaginic factors and airways
hyperreactivity categories.
(E) Finally, multiple linear regression by leaps and bounds was then
performed for all groups of patients defined by airways hyper-
reactivity and reaginic factors, using all 7 air pollution and
meterologic variables and the 4 concommittant variables (see (3)
above). For each group, the best regression was identified by
the AIC criterion.
7
-------
(F) The pattern of best regressions among groups defined by reaginic
factors and airways hyperreactivity was examined to identify the
role of air pollution and meterologic variables upon the pulmonary
function measurements of subgroups of patients.
The final series of leaps and bounds multiple linear regression analyses,
both for individuals (C and D) and groups (E and F), were performed for each
of the 3 spirometric pulmonary function measures: CI) FEVj, (2) FEVi/FVC%,
and (3) PEFR.
4. RESULTS
4.1 AIR POLLUTION AND METEROLOGIC VARIABLES
Average monthly levels of each air pollution and meterologic variable
studied are presented in Table 4.1 on the following page for the summer and
winter periods. Each of the seasonal periods is subdivided into months in
order to portray the consistency of the average levels during each period.
Table 4.2 shown on page 9 presents frequency distributions for each vari-
able during the summer and winter periods, showing a marked increase in vari-
ability and the number of high levels for all air pollution variables during
the winter period, with the exception of O3. Minimum temperature and baro-
metric pressure also showed more variability during winter. In comparison to
the summer, mean daily levels for winter were higher for CO (6.25 +_ 2.02 vs
3.16 +_ 0.98 ppm) > THC C2- 95 +_ 0.55 vs 2.33 +_ .24 ppm), SO2 (0*006 +_ 0.004
vs 0.003 +_ 0.003 ppm), and COH (0.59 +_ 0.22 vs 0.25 +_ 0.07), but winter values
were clearly lower for O3 than in the summer (0.009 +_ 0.003 vs 0.102 +_ 0.059).
SO2 levels were generally low for both periods and showed little variability.
Mean daily minimum temperature was 16C° lower during winter (-4.09C0) than
summer (11.95C0), with much greater variability represented in the winter
values. Mean barometric pressure was slightly lower and varied more widely
during winter (627.3 +_ 4.3) than summer (629 +_ 2.9).
Table 4.3 shown on page 10 presents the intercorrelations among all
variables for each period. For both summer and winter periods, there were
moderate to high correlations among SO2, THC, and CO. For summer, O3 was
nearly independent of all other air pollution and meterologic variables,
while in winter 03 was inversely related to CO, SO2, and THC. The index of
particulates, COH, was positively related to CO, SO2, and THC during the
winter period, but inversely related to O3. In contrast, during summer COH
was essentially independent of SO2 (r = .05) and O3 (r = .07). Finally, as
would be expected, minimum temperature was essentially unrelated to the air
pollution variables during summer, but negatively related to COH during
winter (r = -.30). Scatterplots did not indicate anything other than a
linear trend in any of the data.
8
-------
TABLE 4.1. MEANS AND STANDARD DEVIATIONS FOR ALL AIR POLLUTION AND MOTEROLOGIC VARIABLES BY MONTll FOR THE SUWER (A) AND WINTER (I) PERIODS
Total Barooetric
Carbon Monoxide Sulphur Dioxide Hydrocarbons Ozone Coef. of llaze Pressure
(CO) (S02) (TilC) (Oj) (COfl) Hin Tenp (BP)
ppm ppm ppa ppa COtls/IOOO ft C° oaHg
A. Suaaer Period N Mean SI) Mean SD Mean SD Mean SD Mean SI) Mean SD Mean SD
May 1, l97S-Hay 31, 197S
31
2.960
1.107
0.003
0.005
2.26S
0.284
0.107
0.057
0.230
0.073
6.403
3.562
627.46
4.164
June I, 1975-June 30,
197S
30
J. 25*
1.128
0.002
0.003
2.283
0.165
0.095
0.050
0.226
0.067
11.S83
3.187
629.44
2.533
July 1, 1975-July 31.
197S
31
3.OSS
0.S80
0.003
0.002
2.3S2
0.198
0.100
0.068
0.270
0.062
H.226
1.564
632.05
2.040
Aug. 1, 197S-Aug. 12,
197S
12
3.629
1.289
0.004
0.002
2.384
0.287
0.114
0.063
0.283
O.US
16.167
1.723
631.13
2.45?
B. Winter Period
Nov. IS, 1975-K'ov. 30,
, 1975
16
6.168
1.97S
0.009
0.006
3.093
0.691
0.008
0.002
0.6*6
0.242
-6.187
5.645
62T.3S
s.rsr
Dec. I, 1975-Dec. 31,
197S •
25
7.129
2.407
0.007
0.003
3.300
0.619
0.005
0.002
0.700
0.218
-2.900
S.605
629.12
4.100
Jan. I, 1976-Jan. 31,
1976
30
7.247
2.592
0.008
0.007
3.200
1.187
0.006
0.003
0.719
0.298
-6.900
6.073
629.12
4.124
Feb. 1, 1976-Fcb. 29,
1976
29
S.B02
1.9S0
0.005
0.003
2.712
0.355
0.010
0.004
0.422
0.146
-2.414
S.622
626.39
3.504
Mar. I, 1976-Mar. 31.
1976
31
5.032
1.262
0.004
0.003
2.582
0.365
0.014
0.005
0.487
0.1BS
-2.806
5.927
624.90
4.712
* Six days at the end of December were excluded (Deccaber 25 through December 31).
-------
TABLE 4.2. FREQUENCY DISTRIBUTIONS (DAYS) OF AIR POLLUTION AND HETEROLOG1C VARIABLES FOR SUWER AND WINTER PERIODS
Carbon Monoxide (CO)
Sulfur Dioxide (SO?)
Coefficient
of Haze
(001!)
Total Hydrocarbons (THC)
PPa
Summer
Winter
EE"
Sumter
Winter
COHS/1000 ft
Sumner
Winter
ppm Sumner
Winter
5.6 - 6.0 0
1
1.5 - 1.6
0
2
S.l - 5.5 0
0
IJ.O - 14.0
0
4
.026 - .030
1
1
1.3 - 1.4
0
0
4.6 - 5.0 0
5
U.O - 12.0
0
2
.021 - .025
0
2
1.1 - 1.2
0
8
4.1 - 4.5 0
4
9.0 - 10.0
0
26
.016 - .020
0
S
.9 - 1.0
0
9
3.6 - 4.0 0
9
7.0 - 8.0
2
43
.011 - .01S
0
13
.7 - .a
0
27
3.1 - 3.5 1
25
S.O - 6.0
i:
35
.006 • .010
9
39
.5 - .6
0
42
2.6 - 3.0 9
43
3.0 - 4.0
77
20
.0(11 - .005
82
67
.3 - .4
50
42
2.1 - 2.5 SS
43
0.0 - 2.0
•0
1
Zero
__?
4
.1 - .2
SI
1
1.6 - 2.0 _3
1
No. of days
101
131
No. of days
101
131
Ha. of days
101
131
No. of days 101
131
* + SD
J. 16^.98 6.
25+2.02
X + SD
. 00 J+. 003
006+.004
I ~ SD
2S+.07
.59+.22
X+SD 2.J3+.24
2.95+.S5
Ozone fOt)
Minimum Temperature (T)
Barometric Pressure (BP)
EE™
Stunner
Winter
£
Sumer
Winter
¦nut
Sumer Winter
+16 - +20
32
0
.21 - .26
2
0
<~11 - +15
34
1
.19 - .22
t
0
~ 6 - +10
19
4
.15 - .IS
20
0
~ 1 - ~ 5
15
33
633 -
636
16 14
.11 - .14
IS
0
- 4 - 0
1
44
629 -
632
59 44
.0" - .10
26
0
- 9 - - S
0
19
625 -
628
21 43
•OS - .Oft
IS
3
-14 - -10
0
23
621 -
624
2 20
.00 - .02
13
128
-19 - -IS
0
7
617 -
620
J 10
So. of days
101
131
Ho. of days
101
131
No. of
days
101 131
X + SD
.102*.060
.009*.003
T ~ SD
~11.95+2.
65 -4.09+5.SO
* +
SD
629.8+2.9 627.3+4.3
-------
TABLE 4.3. INTERCORRELATIONS AMONG AIR POLLUTION AND METEROLOGIC
VARIABLES FOR THE SUMMER (A) AND WINTER (B) PERIODS
A. Summer
Period
CO
so2
THC
03
COH
Avtemp
Mintemp
S02
.42**
THC
.69**
.42**
°3
.00
.20*
-.04
COH
.50**
.05
.62**
.07
Avtemp
.10
.10
.08
-.01
.16
Mintemp
.10
.07
.10
-.01
.18
.95**
Barometric
Pressure
-.04
-.16
.16
-.16
.18
.29**
.31**
B. Winter
Period
CO
so2
THC
03
COH
Avtemp
Mintemp
so2
.44**
THC
.71**
.56**
°3
-.58**
-.34**
-.49**
COH
.76**
.64**
.73**
-.54**
Avtemp
.21*
-.17
-.09
.11
-.30**
Mintemp
.20*
-.18
-.11
.12
-.30**
.95**
Barometric
Pressure
.08
-.03
.12
-.24*
.08
-.20*
-.18
*P<. 05.
**P<.01.
11
-------
4.2 PULMONARY FUNCTION MEASUREMENTS
Table 4.4 below presents the distributions of the mean pulmonary functions
for both periods. For the separate groups of 67 patients within both periods,
the pulmonary functions were distributed in a similar way: FEVj (2.04 +_ .57
vs 2.08 +_ .72 liters, for summer and winter, respectively), PEFR (6.22 +_ 1.69
vs 6.17 +_ 1.97 liters/sec), and FEVx/FVC% (73.97 + 9.54 vs 72.25 +_ 9.97 per
cent).
TABLE 4.4. FREQUENCY DISTRIBUTIONS OF MEAN PULMONARY FUNCTION MEASURES
FOR PATIENTS DURING SUMMER AND WINTER PERIODS
First Second Forced Expiratory Volume
FEVi as a Percent of Forced
Vital Capacity
(FEVx)
(FEVy/FVC*.)
Liters
Simmer
Winter
Percent
Summer
Winter
96 - 100
1
0
91 - 35
2
1
4.51 - 5.00
0
1
86 - 90
3
7
4.01 - 4.50
0
0
81 - 85
10
6
3.51 - 4.00
0
2
76 - 80
16
9
3.01 - 3.SO
2
2
71 - 75
11
18
2.SI - 3.00
13
9
66 - 70
10
10
2.01 - 2.50
19
21
61 - 65
8
5
1.51 - 2.00
22
19
56 - 60
5
9
1.01 - 1.50
9
10
51 - 55
1
1
.51 - 1.00
_2
_3
46 - SO
_0
N/Period
67
67
N/Period
67
67
X +_ SD
2.04*.57
2.08^.72
X +_ SD
74.0+9.5
72.2+10.0
Peak Expiratory Flow Rate
(PEFR]
Liters/Sec
Summer
Winter
12.01 -
13.00
0
1
11.01 -
12.00
1
0
10.01 -
11.00
0
3
9.01 -
10.00
1
3
8.01 -
9.00
5
3
7.01 -
8.00
16
5
6.01 -
7.00
15
18
5.01 -
6.00
13
17
4.01 -
S.00
10
10
3.01 -
4.00
5
6
2.01 -
3.00
0
0
1.01 -
2.00
_1_
N/Pcriod
67
67
X + SD
6.22+1.69
6.17+1.97
12
-------
To evaluate the relationships among the three pulmonary function measures,
correlations were performed for each patient using all available days within
the respective periods. The average correlations, and standard deviations,
are presented in Table 4.5 for the summer and winter periods. For patients
TABLE 4.5. MEANS AND STANDARD DEVIATIONS OF THE CORRELATION
COEFFICIENTS AMONG PULMONARY FUNCTION MEASUREMENTS
FOR THE SUMMER AND WINTER PERIODS*
Period Pulmonary Function FEV2 FEVj/FVC% PEFR
FEVj . 54+_. 32
Summer FEVi/FVC%
PEFR
FEVj . 59+. 36
Winter FEVx/FVC*.
PEFR
* Data are expressed as Mean +_ SD.
within both periods, the pattern of the mean correlations is similar: FEV^ and
PEFR are highly related, while the relationships among FEVj/FVC% and the other
measures are only moderate. Since the pulmonary function measurements were not
uniformly highly related and since no single pulmonary function was clearly
preferable, analyses relating pulmonary functions to air pollution and metero-
logic variables were subsequently performed using each measure.
It had been anticipated that patients would show generally improving
pulmonary function levels throughout hospitalization. Preliminary correla-
tional analyses, relating pulmonary functions to day of study were performed
84+.15
48+.31
.78+.25
.50+.33
13
-------
to evaluate this expected trend. Table 4.6 summarizes the distribution of
these correlations for each pulmonary function measure for patients within both
summer and winter periods. Correlations within both periods varied widely and
TABLE 4.6. FREQUENCY DISTRIBUTIONS OF CORRELATIONS BETWEEN PULMONARY
FUNCTION MEASURES AND DAY OF STUDY FOR SUMMER AND WINTER PERIODS
Pulmonary Function Measurements*
Correlation
FEVi
FEVi/FVC%
PEFR
Summer
Winter
Summer
Winter
Summer
Winter
.81 to
1.00
0
4
0
1
1
1
.61 to
.80
9
4
8
3
9
12
.41 to
.60
9
14
13
16
13
12
.21 to
.40
13
7
14
10
15
5
. 01 to
.20
17
7
9
6
14
7
-.19 to
.00
10
6
7
8
10
13
-.39 to
-.20
7
8
9
7
3
5
-.59 to
-.40
2
10
6
9
1
7
-.79 to
-.60
0
6
1
7
1
3
-.99 to
-.80
_0
_1
_0
_0
_0
_1
N/Period
67
67
67
67
67
67
X + SD
.19+.32
.05+.50
.16+.39
.04+.45
.25+.36
.13+.46
* FEV} = First Second Forced Expiratory Volume; FEVi/FVC% = FEVi as a % of
Forced Vital Capacity; PEFR = Peak Expiratory Flow Rate.
were not consistently positive. For example, for PEFR, 15 correlations were
in the range of 0 to -.60 for the summer period, while 29 were within this
range for the winter period. These analyses confirmed visual examination of
individual plots of pulmonary functions made across days of treatment, and
indicated that a wide range of patterns - random, increasing, decreasing, etc.
existed. Specific patterns could not be attributable to patients within groups
defined by airways hyperreactivity or reaginic factors.
14
-------
4.3 PULMONARY FUNCTIONS IN RELATION TO AIR POLLUTION AND METEROLOGIC FACTORS
4.3.1 Univariate Regression Analyses
Simple regressions were performed to provide an initial impression of
the relationship between pulmonary function measures (FEVj/FVC%) and the air
pollution/raeterologic variables. All analyses were performed on individual
patients, thereby providing a matrix of regression equations for the summer
and winter periods defined by 67 patients per period and 7 air pollution and
meterologic variables. Scatterplots indicated no trends beyond the linear in
the data.
Table 4.7 summarizes the significant regressions for each variable. Air
pollution variables failed to show a consistent relationship to FEVj/FVC%
TABLE 4.7. UNIVARIATE REGRESSION ANALYSES SUMMARY TABLE FOR
INDIVIDUAL REGRESSIONS
Air
Pollution -
Meterologic
Variable
Period
Regression
Coefficient
CO
S02
THC
03
COH
T* BP
Positive
1
5
3
1
4
6 13
Summer
Negative
2
3
1
3
1
3 2
Not Significant +
64
59
63
63
62
58 52
N/Period = 67
Positive
5
2
3
5
8
5 2
Winter
Negative
1
2
4
11
6
7 3
Not Significant +
61
63
60
51
53
55 62
N/Period = 67
* Minimum Temperature.
+ P<.05.
15
-------
measurements. For example, during the winter period COH was significantly-
related to FEVj/FVC% foT 14 patients; but of these, 8 regression coefficients
were positive (higher values of COH associated with higher values of FEVi/
FVC%) while only 6 were in the expected negative direction. Surprisingly,
ozone (O3) was significant for 14 individuals during winter; however, 5 of
these 14 individuals had higher FEVi/FVC% associated with higher levels of O3.
There were no differences in age, sex, airways hyperreactivity or reaginic
factors among patients with positive, negative, or nonsignificant regressions
for any air pollution variable.
Barometric pressure was significantly associated with FEVi/FVC% during
the summer period for 15 patients with 13 of these patients having higher
FEVi/FVC% on days of higher barometric pressure. No consistent differences
could be found between patients apparently affected by barometric pressure and
others. During the winter period, barometric pressure exerted virtually no
consistent effect upon FEVi/FVC%. Minimum temperature was unrelated to FEVi/
FVC% either during the summer or winter periods.
4.3.2 Multivariate Regression Analyses
In these final series of analyses, extensive efforts were made to account
for variables which could obscure a real relationship between air pollution/
meterologic variables and pulmonary function measures.
A total of eleven variables were involved in leaps and bounds multiple
linear regression analyses,-^ using all 7 air pollution/meterologic variables
and 4 concomittant variables shown in Table 4.8. The concomittant variables
included (1) intraveneous bronchodilators given at the time of the pulmonary
function measurement, (2) the previous day's pulmonary function value, (3)
bronchodilator medications taken within 2 hours prior to pulmonary function
measurement, and (4) bronchodilator treatments taken within 1 hour prior to
pulmonary function measurement. The previous day's pulmonary function measure-
ment was included to account for the usual variations noted across time in the
level of airway obstruction, while each of the remaining 3 variables could in-
fluence airway calibre directly.
As described in the Methods section, identical analyses were performed
for FEVi, FEVi/FVC%, and for PEFR for (A) individual patients within each
period, (B) patient groups defined by the role of reaginic factors in their
asthma, and (C) patient groups defined by airways hyperreactivity, indexed by
methacholine and histamine challenges.
A summary table for the leaps and bounds multiple linear regression analy-
ses for individual patients is presented in Table 4.9. The table shows the fre-
quency with which each variable appeared within the best regression equation
selected according to the AIC criterion.^3 As expected, prior bronchodilator
treatments clearly influenced subsequent pulmonary function measures, appearing
most often within the best regression for all pulmonary function measures for both
periods. With few exceptions, prior bronchodilator treatment was positively related
to pulmonary function levels. The next strongest variable was the presence of
16
-------
TABLE 4.8. VARIABLES INCLUDED IN THE LEAPS AND BOUNDS MULTIPLE LINEAR REGRESSION ANALYSES
Variable No.
Name
Value
Description
6
7
8
9
10
11
Intraveneous bronchodilators 1,0
Previous day's function Raw value
Prior bronchodilators 1,0
Prior treatment 1,0
Carbon monoxide (CO) ppm
Sulfur dioxide (S02) ppm
Total hydrocarbons (THC) ppm
Ozone (Oj) ppm
Coefficient of haze (COH) COHs/lOOO ft
Minimum temperature (T) C°
Barometric pressure (BP) mmllg
Intraveneous bronchodilators given at time
of pulmonary function measurement
8:00 am pulmonary function for the preceding
day
Bronchodilators taken within two hours pre-
ceding pulmonary function measurement
Inhaled bronchodllator treatment taken within
1 hour preceding pulmonary function measurement
24-hour average preceding 8:00 am pulmonary
function
-------
TABLE 4.9. SU>MARY TABLE FOR LEAPS AND BOUNDS MULTIPLE LINEAR REGRESSION ANALYSES FOR INDIVIDUAL PATIENTS
Sumner Period Winter Period
Variable FEVj FEVj/FVC% PEFR FEVj FEVj/FVC* PEFR
Number
Variable Name
+¦
-
No*
~
-
No
~
-
No
-
No
~
-
No
+
-
No
1
Intraveneous
Bronchodilators
3
IS
49
7
10
SO
3
16
48
3
17
47
6
11
50
4
13
50
2
Previous Function
1
9
57
4
5
58
0
8
59
1
6
60
3
6
58
2
7
SB
3
Prior
Bronchodilators
3
15
49
2
10
55
6
14
47
1
12
54
8
6
S3
2
9
56
4
Prior Treatments
29
3
3S
16
S
46
31
4
32
27
4
36
21
9
37
28
6
33
S
CO
3
5
59
6
7
54
5
4
58
3
5
59
6
3
58
5
S
57
6
so2
4
3
60
3
5
59
4
4
59
4
9
54
2
6
59
3
7
57
7
nic
3
1
63
4
3
60
3
3
61
2
6
59
2
4
61
5
6
56
8
°3
11
2
54
2
6
59
7
2
58
5
9
53
1
6
60
9
8
50
9
con
7
3
57
7
3
57
5
1
61
11
4
52
7
6
54
10
S
52
10
Temperature
14
3
50
14
2
51
13
1
53
8
4
55
2
fi
57
5
7
55
U
Barometric Pressure
8
0
59
7
3
57
7
0
60
2
2
63
5
3
59
4
2
61
* Header labels indicate: + * positive regression coefficient (P<.05); - ¦ negative regression coefficient (P<.05); No « not
significant.
-------
intraveneous bronchodilators at the time of pulmonary function measurement; it
was generally associated with lower pulmonary functions since patients are
typically placed on these intensive regimens when experiencing major breathing
difficulties.
More importantly, the difficulty in constructing a general model is appa-
rent. No consistent relationship between any air pollution/meterologic variables
and any of the pulmonary function measures were found. When air pollution vari-
ables appeared within the best regression, they were nearly evenly distributed
between positive and negative regression coefficients. For the meterologic
variables, minimum temperature appeared with a positive regression coefficient
14 times during the summer period for each pulmonary function measure, and
with a negative coefficient rarely. Unexpectedly, temperature was less con-
sistently related to pulmonary function during the winter period, suggesting
that limited exposure may occur to the outdoors on cold winter days. Baro-
metric pressure was generally positive when it appeared during summer, but not
consistently related to pulmonary functions for the winter period.
Finally, Table 4.9 indicates that there were patients within both seasonal
periods where no best regression could be identified on the basis of the AIC
criterion.
Supplementary analyses indicated that the presence or absence of reaginic
factors in asthma, and degree of airways hyperreactivity (indexed by metha-
choline and histamine challenges) could not differentiate between patients
having or not having significant negative regression coefficients for air
pollution or temperature within the best regressions.
Tables 4.10 and 4.11, shown on the following pages, summarize the final
leaps and bounds multiple linear regression analyses for the summer and winter
periods, performed for groups of patients defined according to airways hyper-
reactivity. Analyses for groups defined according to reaginic factors re-
vealed no systematic differences among these groups, and reaginic factors
were therefore disregarded as a basis for subsequent grouping.
In agreement with the parallel analyses for individuals, prior broncho-
dilator treatments taken within one hour before pulmonary function measurement
(Var. #4) appeared most often within the best regression of these groups, and
invariably with a positive regression coefficient. Intraveneous bronchodilators
(Var. #1), in contrast, invariably had a negative regression coefficient with-
in the best regressions, indicating that patients are generally experiencing
breathing difficulty when placed on these Tegimens.
The 5 air pollution variables appeared rarely within the best regression
for these groups, and typically later than the medication variables. While
the pattern of results is generally similar for all pulmonary function measures,
careful examination of the table reveals inconsistencies in the direction of
the relationships. For example, when O3 (Var. #8) appeared, it was generally
negatively related in winter. The only time it appeared in summer (for FEVi),
O3 was positively related to the pulmonary function measurement. For CO
19
-------
TABLE 4.10. LEAPS AND BOUNDS MULTIPLE LINEAR REGRESSION ANALYSES SUMMARY TABLE FOR SUMMER PERIOD SUMMARIZING
BEST REGRESSIONS FOR AIRWAYS HYPERREACTIVITY GROUPS*
Variable A
Variable B
Variable C
Variable D
Group
Coefficient of
Determination
N R2
Regression
Var. Coef. P
Regression
Var. Coef. P
Regression
Var. Coef. P
Regression
Var. Coef. P
Variable E 6 (F)
Regression
Var. Coef. P
A. FEVj
Methachollne
Missing
0
10 ~ <.001 11 ~ <.01
Low
32
4.365
1
<.001
2
+
<.0S
4
+
<.05
Moderate
27
5.379
1
<.01
4
+
<.001
9
~
<.0S
High
5
2.362
1
<.10
4
~
<.001
Negative
3
6.419
4
<•05
8
~
<¦01
9
+
<.01
Histamine
Missing
1
2S.6S
4
+
<.001
10
*
<.05
low
18
3.749
1
-
<.01
4
~
<.001
B
~
<.05
10
~
<.001
11
~
<.01
Moderate
27
3.917
1
-
<.001
4
~
<.001
10
+
<•001
High
17
4.492
1
-
<.001
4
+
<.001
10
+
<•001
Negative
Lowest
No regression is best
Missing
1
25.65
4
~ <.001
10
~ <.05
Low
36
3.727
1
<.001
2
~ <.05
4
~
<.001 10
~
<•001
Moderate
23
6.237
1
<.001
4
~ <.001
9
~
<.05 10
~
<.05
High
4
1.917
1
<.10
4
~ <.01
Negative
3
6.419
4
<-05
8
* <.01
9
<.01
11
<.01
-------
TABLE 4.10. continued
Group
N
Coefficient of
Determination
R2
Variable
A
Variable
B
Variable
c
Variable
D
Variable E 6 (F)
Var.
Regression
Coef. P
Var.
Regression
Coef. P
Var.
Regression
Coef. P
Var.
Regression
Coef. P
Regression
Var. Coef. P
B. FGVj/FVC*
Methacholine
Missing
0
Low
32
2.S24
3
-
<•01
4
+
<.001
5
-
<.01
6
+
<.10
106(11) ~(~) <.001
Moderate
27
1.239
1
-
<.01
4
+
<.001
10
+
<.05
11
+
<.10
6 (<. 05 )
High
S
2.309
1
-
<.10
4
+
<.01
10
+
<.05
Negative
3
1.725
S
-
<.10
9
~
<.05
Histamine
Missing
1
19.36
4
~
<.01
10
+
<.01
Low
18
3.435
1
-
<.10
4
+
<.001
6
-
<.10
10
+
<.001
11 ~ <.01
Moderate
27
.589
1
-
<.05
4
~
<.01
10
+
< .05
High
17
3.031
1
-
<.001
2
+
<.10
4
+
<.001
5
-
<.04
10 ~ <.001
Negative
4
2.252
4
*
<.01
Lowest t
Missing
1
19.36
4
*
<.01
10
+
<.01
Low
36
2.716
1
-
<.001
2
+
<.05
4
+
<.001
5
-
<.01
10 * <.001
Moderate
23
0.702
1
-
<.05
4
+
<.001
9
4-
<.10
11
~
<.01
High
4
2.275
1
-
<.20
4
+
<.05
10
+
<.05
Negative
3
1.72S
5
-
<.10
9
*
<.05
-------
TABLE 4.10. continued
Group
N
Coefficient of
Determination
R2
Variable
A
Variablo
B
Variable
C
Variable D
Variable E 6 (F)
Var.
Regression
Coef. P
Var.
Regression
Cocf. P
Var.
Regression
Coef. P
Var.
Regression
Coef. P
Regression
Var. Coef. P
C. PEFR
Metli.icholine
Missing
0
Low
32
4.998
1
-
<.001
2
+
<.0S
4
+
<.001
10
~ <.001
11 ~ <.01
Moderate
27
6.086
1
-
<.001
4
~
<.001
10
+
<.001
High
5
1.383
4
~
<•01
Vnnit 2 «/#»
3
U/% roorikcclnn 4c kf>ct
iM ^il i. 1 ¥C
" (111 1 C5IV391UII * 9 Uv •> L
Histamine
Mi ssing
1
45.24
1
-
<.01
4
+
<.001
10
+
<¦05
l.ow
18
4.498
1
-
<-001
4
+
<.001
10
4
<.001
11
~ <.05
Moderate
27
4.863
1
-
<.001
4
<.001
10
+
<.001
High
17
3.786
1
-
<.001
4
+
<.001
10
+
<.01
11
~ <.10
Kegat ive
4
2.918
4
~
<.001
Lowest
Missing t
1
45.24
1
-
<.01
4
+
<.001
10
«¦
<.05
Low
36
4.311
1
-
<.001
2
+
<.05
4
<.001
10
~ <.001
11 ~ <.01
Moderate
23
6.688
1
-
<.001
4
+
<.001
10
*
<.001
High
4
0.928
4
~
<.10
Negative
3
2.596
3
-
<.10
7
+
<.10
10
-
<•10
* Variable Code: 1 « Intravcneous bronchodilators; 2 = previous day's pulmonary function; 3 « prior bronchodilators; 4 » prior treatments;
5 « CO; 6 ¦ SOj; 7 » TJIC; 8 =¦ O3; 9 » COM; 10 « minimum temperature; 11 • barometric pressure.
t Note - This is the same group as missing histamine above.
-------
TABLE 4.11. LEAPS AND BOUNDS MULTIPLE LINEAR REGRESSION ANALYSES SUMMARY TABLE FOR WINTER PERIOD SHOWING
BEST REGRESSIONS FOR AIRWAYS HYPERREACTIVITY GROUPS *
Variable A Variable B Variable C Variable D Variable E 4 (P)
Coefficient of
Determination Regression Regression Regression Regression Regression
Group N R^ Var. Coef. P Var. Coef. P Var. Coef. P Var. Coef. P Var. Coef. P
A. FEX'!
Methactioline
Missing
3
2.323
4 ~ <.10
8
-
<.05
Low
25
1.417
1 - <.001
2
-
<.10
4
<.001
5
<.0S
9
~
Moderate
27
2.706
1 - <.001
4
~
<.001
High
7
8.319
1 - <.001
4
+
<.001
8
-
<.05
Negat ive
5
20.973
3 - <.fl01
4
~
<.001
Histamine
Missing
7
8.880
1 - <.05
2
-
<.01
4
+
<.001
5
<.10
9«(I0)
~(~)
Low
9
2.202
1 - <.01
4
~
<.001
6
-
<.05
8
~ <.10
9
~
Moderate
30
0.882
1 - <.001
4
+
<.001
High
16
5.593
1 - <.001
4
~
<•001
11
+
<.0S
Negative
S
9.424
1 - <.001
4
~
<.001
8
-
<.0S
Lowest
Missing
9
5.673
1 - <.01
4
~
<.001
9
+
<.01
Low
26
.554
I - <.001
4
~
<.001
Moderate
24
3.129
I - <.001
4
~
<.001
High
5
9.379
1 - <.001
4
<.001
8
-
<.0S
Negative
3
22.420
3 - <.001
4
~
<¦001
<.01
<.001
S(< .01)
<.os
-------
TABLE 4.11. continued
Variable A
Variable B
Variable C
Variable D
Coefficient of
Group
N
Determination
R2
Var.
Regression
Coef. P
Var.
Regression
Coef. P
Var.
Regression
Coef. P
B. FEVj/FVC%
Methacholine
Missing
3
0.8BS
10
<.10
Low
25
0.764
1
<.001
4
~ <.001
10
<.0S
Moderate
27
0.980
3
<.001
4
+ <.001
8
<.05
High
7
0.479
11
+ <.05
Negative
S
7.800
1
<.001
4
~ <.001
Regression
Var. Coef. P
Variable E S (F)
Regression
Var. Coef. P
Histamine
Missing
Low
Moderate
High
Negat ive
Lowest
7
9
30
16
5
0.525
0.855
2.323
<.05
<.001
<.001
— No regression is best
~ <.001 5
~ <.001 11 +
-- No regression is best
<.05
<.01
Missing
9
0.736
8
<.01
Low
26
0.736
1
<.001
4
~
<.001
10
<.01
Moderate
24
1.059
3
<.001
4
~
<.001
8
<.10
High
5
0.933
1
<.10
4
~
<.05
Negative
3
9.847
1
<.001
4
+
<•001
<.0S 95(11) ~{-) <.25
«(<-01)
-------
TABLE 4.11. continued
Croup
N
Coefficient of
Determination
n2
Variable
A
Variable
B
Variable
C
Variable D
Variable E ( (F)
Var.
Regression
Coef. P
Regression
Var. Coef. P
Var.
Regression
Coef. P
Var.
Regression
Coef. P
Regression
Var. Coef. P
C. PI-FR
Methacholine
Missing
3
4.748
1
-
<.05
4 ~
<.01
8
-
<.05
Low
25
1.198
1
-
<.001
4 +
<.001
7
-
<•05
8
~ <.05
9 ~ <.001
Moderate
27
1.858
1
-
<.001
4 +
<.001
High
7
6.038
1
~
<.001
4 +
<.001
Negative
5
10.SIS
1
-
<.01
4 +
<•001
8
~
<.01
9
~ <.05
Histamine
Missing
7
8.199
2
-
<.001
4 +
<.001
9
+
<.001
Low
9
1.725
1
-
<.001
4 +
<•001
Moderate
30
0.479
1
-
<.05
<.10
4
~
<.001
High
16
4.512
1
-
<.001
4 ~
<.001
11
+
<.01
Negative
5
7.629
3
-
<.001
4 +
<.001
6
-
<.05
8
<.001
Lowest
Missing
9
7.515
2
-
<.001
4 +
<.001
8
-
<.0S
9
~ <.01
Low
26
0.329
1
-
<•05
4 ~
<.01
8
<.10
Moderate
24
2.138
1
-
<.001
4 ~
<.001
High
5
7.449
1
-
<.001
4 ~
<.001
Negative
3
15.481
1
<.001
4 ~
<.001
8
~
<.05
9
~ <.05
* Variable Code: 1 ¦ intraveneous bronchodilators; 2 = previous day's pulmonary function; 3 = prior bronchodilators; 4 ¦ prior treatments;
5 = CO; 6 « SO2; 7 =» TllC; 8 » Oj; 9 » COII; 10 » minimum temperature; 11 » barometric pressure.
-------
(Var. #5), the pattern was somewhat more consistent, appearing twice as a
negatively related variable in winter (for FEVi/FVC% and FEV^) and several
times during summer (for FEVj/FVC%). SO2 (Var. #6) appeared twice (for FEVj
and PEFR) with a negative regression coefficient in winter and twice in summer
(on FEVi/FVC%), but with a positive and a negative regression coefficient.
THC (Var. #7) appeared only once in winter and with a negative regression co-
efficient (PEFR), and not at all for summer. Finally, COH (Var. #9) which
appeared several times during summer (FEVi and PEFR) and winter (FEVi/FVC%
and FEVi), always had a positive regression coefficient.
Patient groups defined according to airways hyperreactivity were not
consistently influenced by air pollution variables in the same way. The
regression equations could only account for a relatively small amount of the
variability in the pulmonary function scores.
For the meterologic variables, minimum temperature (Var. #10) and baro-
metric pressure (Var. #11), the results are also equivocal. Temperature and
barometric pressure appeared rarely during the winter period, and with nega-
tive and positive regression coefficients. In contrast, during the summer
period, minimum temperature appeared frequently and always with positive
regression coefficiants, but this effect was not attributable to any group
or groups defined by airways hyperreactivity. In summary, the specific in-
fluence of air pollution variables when they appeared, was small, variable in
direction, inconsistent for groups defined by airways hyperreactivity, and
inconsistent across pulmonary function measures (FEV]/FVC%, FEVi, and PEFR).
S. DISCUSSION
The results do not support the hypothesis that the air pollution vari-
ables studied within a metropolitan area affect the pulmonary functions of
asthmatic patients. Neither univariate or multivariate regression analyses
for individuals and groups identified any consistent effect. In addition,
there were no subgroups of asthmatic patients defined according to airways
hyperreactivity or the role of reaginic factors in their asthma who appeared
to be affected more than others. Failure to identify any systematic relation-
ships ruled out specification of a general model. Importantly, this inability
to specify a general model occurred despite elaborate attempts to account for
extraneous factors that could influence pulmonary function measurements, in-
cluding the presence or absence of intraveneous medications at the time of
measurement, bronchodilators or treatments taken before measurement, and the
previous day's pulmonary function measurement. This paucity of clear results
occurred for all spirometric pulmonary function measurements studied, inclu-
ding FEVi, FEVj/FVC*, and PEFR.
The Tesults must, however, be viewed in relation to the strengths and
limitations of the study. There are several notable strengths. First, the
study focused upon objective pulmonary function measurements indexing airway
obstruction, and studied several of these. This focus is in contrast to
26
-------
epidemiologic studies (based on clinic and hospital records of treatment) or
studies dealing exclusively with subjectively reported breathing difficulties.
Second, a well-defined group of asthmatic patients, with moderately severe
illness, were studied. Documentation about their asthma was extensive in re-
gard to history, role of allergic (reaginic) factors, and airways hyperreac-
tivity. At each step, analyses proceeded with regard to subclassification
based on reaginic factors (thereby minimizing the influence of seasonal air-
borne pollens upon pulmonary function) and airways hyperreactivity, the degree
of which could logically affect the response to air pollution variables in
crucial ways. By defining the asthmatic groups in such a detailed way, the
study also avoids problems inherent in grouping diverse respiratory illnesses
arbitrarily. Third, the final series of analyses considered those factors
which could obscure potential relationships between air pollution variables
and pulmonary function measures. Using the Clinical Data Management System,^
it was possible to retrieve not only the pulmonary function data, but per-
mitted verified information about any intraveneous medications or bronchodila-
tor medications and treatments taken before pulmonary function measurement,
and the time at which these were taken. These variables, particularly prior
bronchodilator treatments, were found to influence pulmonary function measure-
ment importantly. This effect was documented and accounted for within the
final series of analyses. Fourth,.two periods of study were selected to maxi-
mize the potential impact of all air pollution and meterologic variables. The
differences in average levels and variability of the air pollution/meterologic
variables contrasted in the expected fashion between the seasonal periods.
Finally, the number of available patients within each period was fairly
large, providing ample opportunity for any systematic effect of air pollution
to be documented.
Nonetheless, the study has a number of real or potential limitations.
One potential limitation, regarding the indoor-outdoor air pollution rela-
tionships, was more apparent than real. Unlike typical hospitalization,
patients at NJHRC are not confined to the building or even the hospital
grounds, but are encouraged to move about the Denver area freely - shopping,
attending sporting and cultural events or movies - as they would at home.
Adolescent patients, for example, continue to walk to school during treatment.
The high school is nearly eight blocks away from the hospital. For most
patients, therefore, exposure to air pollution should be comparable to people
residing at home. Since the study involved retrospective data, no information
was available to document time spent outdoors by various patients. Reduced
exposure to outdoors on cold days may have limited the relationship between
minimum temperature and pulmonary functions in winter. In the Salt Lake City
study by Finklea, et_al_. ,22 there was also indication that reduced exposure
on cold days may have influenced the relationship between temperature and
attack rates for asthmatics living at home.
Additional information was acquired about the indoor-outdoor air pollu-
tion relationships for the patient floors. The patients resided on the
second and third floors of a new 10-story building. The intake for the
ventilation system of the building drew in outside air at a rate of 180,000
cu ft/min. and provided a minimum of six air exchanges per hour for the
building's total volume of 1,800,000 cubic feet. Each patient room had
27
-------
individual input and exhaust ducts, with additional exhausts located in halls.
There are no operating rooms within the hospital and no smoking is permitted
on the patient floors or elevators. Research laboratories are located above
the patient floors, with each laboratory with any potential for generating
fumes provided with hoods over the laboratory areas, and having rooms exhausted
at rates exceeding the rest of the building. None of these laboratories are
located on the patient floors.
For four days during November, 1976 (November 10 through November 14),
a mobile unit from Pedco Environmental Specialists (Cincinnati, Ohio) moni-
tored indoor-outdoor air pollution levels for three patient rooms (two on the
third floor and one on the second floor). Examination of this data indicated
that for NO, N02, CO, THC, and SO2 the indoor-outdoor air pollution levels
are highly comparable, both in regard to time and absolute magnitude. For O3,
which is known to break down rapidly, the indoor levels were consistently near
zero despite changes in the outdoor levels. Due to the rapid air exchange in
the patient building the time lag was small and on the order of 20 minutes.
The unusually high rate of air exchange for the patient building was confirmed
by the timed disappearance of tracer gas released in the patient rooms. Addi-
tional in-place monitoring of NO and CO described a daily pattern associated
with periods of high automobile traffic. Thus, with the exception of O3 and
COH, even those patients remaining exclusively within the building would be
exposed to CO, SO2, and THC. The absence of an effect of these air pollution
variables upon pulmonary functions cannot be attributed to lack of exposure
resulting from patient behavior and low indoor-outdoor air pollution relation-
ships .
There are, however, certain real limitations of the study. First, all of
these patients were inpatients on around-the-clock bronchodilator regimens,
while most were also taking oral corticosteroids either on a daily or alter-
nate day basis. These medication regimens are prescribed to increase the
tolerance of patients' airways to irritants capable of aggravating the asthma.
Therefore, it is possible that considerable protection was afforded by medica-
tions against the adverse affect of air pollution variables. Patients on
more minimal medication schedules may therefore be more appropriate to study.
Second, only spirometric pulmonary function measurements were used. If the
effect of relatively transient changes in air pollution is slight, specific
conductance (SGaw), or airway resistance (Raw) may be preferable indices of
airway obstruction.34 Third", if the effect of air pollution upon asthmatic
airways depends upon more prolonged exposure, or if, conversely, the effect is
very transient, this study would be expected to fail to document such effects.
The study considered only average levels of the air pollution variables for
24 hours preceding pulmonary function measurement. Fourth, these results
were limited to the specific air pollution variables studied, as well as the
levels achieved during the two seasonal periods. In comparison to the Salt
Lake City and New York studies,22>the levels of SO2 were very low: On
only one day during the winter period did SO2 achieve a level near 79 qg/m,3
whereas weekly averages for high exposure patients in the earlier studies
occasionally exceeded this level. The levels of SO2 were consistently below
levels suggested as producing significant harm in asthmatics.In contrast,
28
-------
CO levels are generally high, but its effect is principally hematologic and
would not be expected to trigger airway obstruction directly by an irritant
effect. Related to these considerations, in-place monitoring and indoor-out-
door air pollution monitoring by PEDCO indicated that the highest levels of
THC, NO and CO occurred twice daily, near 9:00 am and 6:00 pm. These times
of maximum appear to be associated with usual traffic patterns in the area.
If the effect is transient, then 8:00 am pulmonary function measurement
represents a poor choice. In future studies, pulmonary function measurement
should follow times of peak air pollution levels closely.
Finally, the full range of patient medication behavior could not be con-
sidered. It is of particular interest that patients often take as-needed
(PRN) treatments (usually nebulized bronchodilators) when experiencing breath-
ing difficulty.^ During periods of high air pollution, more PRNs may be
taken and overusage of aerosolized treatments has been related to death in
asthma.37-39 jf SOj usage may serve to increase pulmonary functions and
obscure any direct relationship between air pollution and pulmonary function
measurement. By monitoring PRN behavior continuously among nonhospitalized
asthmatics, using adapted nebulizers with mini-storage units,40'41 it may be
possible to objectify the impact of air pollution upon this behavior of out-
patients.
29
-------
6. RECOMMENDATIONS
This preliminary project has enabled better delineation of the factors
which could obscure potential relationships between air pollution variables
and airway obstruction in metropolitan, asthmatic populations. Further re-
search should be based on the following recommendations. First, since oral
corticosteroid and around-the-clock bronchodilator medications may afford
significant protection against the effect of air pollutants, periodically
medicated asthmatic patients who are not on these intensive regimens should
be studied. Second, if possible, pulmonary function measurements reflecting
changes in the large airways, such as airway or total lung resistance, should
be used together with spirometric measures. Third, as-needed (PRN) use of
aerosolized bronchodilators should be carefully monitored during any study
of these patients, since most patients tend to titrate their use of these
PRN medications in relation to breathing difficulty. Fourth, in studies of
patients not on around-the-clock medications, it would be most desirable tn
obtain pulmonary function measurements at a time closely following peak daily
levels of the air pollution variables. Fifth, lag time for summarizing the
air pollution levels should be varied in the analyses, permitting, for exam-
ple, 8-hour and 24-hour air pollution scores to be related to changes in
subsequent pulmonary function measurements, with longer lags also considered.
It is not yet known how transient or time dependent any effect of air pollu-
tion is upon changes in airway calibre in asthma. Finally, further studies
might best be conducted on a prospective basis, permitting stringent control
of such factors as medications taken prior to pulmonary function measurement
and exposure to the outdoors. It is hoped that the present study has served
as a useful model for the development of these future studies.
30
-------
7. REFERENCES
1. Nadel, J.: Airways: Autonomic regulation and airway responsiveness, in
Bronchial Asthma: Mechanisms and Therapeutics, E. B. Weiss and M. S.
Segal, eds., Little, Brown and Co., Boston, 1976, p. 155.
2. Nadel, J. A.: Structure-function relationships in the airways: Broncho-
constriction mediated via vagus nerves on bronchial arteries, peripheral
lung constriction mediated via pulmonary arteries, Med Thorac, 1965, 22,
231.
3. Spector, S. L. and Farr, R. S.: Bronchial provocation tests, in Bronchial
Asthma: Mechanisms and Therapeutics, E. B. Weiss and M. S. Segal, eds.,
Little, Brown and Co., Boston, 1976, p. 639.
4. Spector, S. L. and Farr, R. S.: A comparison of methacholine and histamine
inhalations in asthmatics, J Allergy Clin Immunol, 1975, 56, 308.
5. Spector, S. L. and Farr, R. S.: Bronchial inhalation challenge procedures
in asthmatics, Med Clin North Am, 1974, 58, 71.
6. Kiviloog, J.: Bronchial reactivity to exercise and methacholine in bron-
chial asthma, Scand J Respir Dis, 1973, 54, 347.
7. Mellis, C. M., Katten, M., Keens, T. G., Mansell, A., and Levison, H.:
Histamine and exercise provocation compared, Am Rev Respir Dis, 1977,
115, 285.
8. Horton, D. J., Suda, W. L., Kinsman, R. A., Souhrada, J., and Spector,
S. L.: Bronchoconstrictive suggestion in asthma: A role for airways
hyperreactivity and emotions, NJHRC Psychophysiology Research Labora-
tories Report No. 43, Denver, Colorado, July, 1977.
9. Lave, L. B. and Seskin, E. P.: Air pollution and human health, Science,
1970, 169, 723.
10. Heiman, H.: Episodic air pollution in metropolitan Boston: A trial
epidemiological study, Arch Environ Med, 1970, 20, 230.
11. Schrenk, H. H., Heiman, H., Clayton, C. D., Gafafer, IV. M., and Wexler,
H.: Air pollution in Donora, Pa.: Epidemiology of unusual smog episode
of October, 1948, Public Health Bull. 306, Washington, D. C., 1948.
12. Drinker, P. Deaths during the severe fog in London and Environs, December
5 to 9, 1952, Arch Ind Hyg, 1953, 7, 275.
31
-------
13. Greenburg, L., Field, F., Reed, J. I., and Erhardt, C.: Air pollution
and morbidity in New York City, JAMA, 1962, 182, 161.
14. Ribon, A., Glasser, M., and Sudivoraseth, N.: Bronchial asthma in child-
ren and its relationship to weather and air pollution, Ann Allergy, 1972,
30, 276.
15. Lawther, P. J., Brooks, A. G. F., Lord, P. W., and Waller, R. E.: Day
to day changes in ventilatory function in relation to the environment,
Environ Res, 1974, 8, 119.
16. Lawther, P. J., Waller, R. W., and Henderson, M.: Air pollution and
exacerbation of bronchitis, Thorax, 1970, 25, 525.
17. Tromp, S.: Influence of weather and climate on asthma and bronchitis,
Rev Allergy, 1968, 22, 1027.
18. Girsh, L. S., Shubin, E., Dick, C., and Shulanek, R. A.: A study on the
epidemiology of asthma in children in Philadelphia, J Allergy, 1967, 39,
347.
19. Lewis, R., Gilkerson, M. M., Jr., and McCaldin, R. 0.: Air pollution and
New Orleans asthma, Public Health Rep, 1962, 77, 947.
20. Wells, R.: Effects of cold air on respiratory air flow resistance in
patients with respiratory tract disease, N Engl J Med, 1960, 263, 268.
21. Salvaggio, J., Hasselblad, V., Seabury, J., and Heiderscheit, L. T.:
New Orleans asthma. II. Relationship of climatologic and seasonal
factors to outbreaks, J Allergy, 1970, 45, 257.
22. Finklea, J. F., Calafiore, D. C., Nelson, C. J., Riggan, W. B., and
Hayes, C. G.: Aggravation of asthma by air pollutants: 1971 Salt Lake
Basin studies. In Health Consequences of Sulfur Oxides: A Report from
CHESS, 1970-1971. EPA Report 650/1-74-004.
23. Finklea, J. F., Farmer, J. H., Love, C. J., Califiore, D. C., and Sovocool,
G. W.: Aggravation of asthma by air pollutants: 1970-1971 New York
studies. In Health Consequences of Sulfur Oxides: A Report from CHESS,
1970-1971. EPA Report 650/1-74-004.
24. Nadel, J. S., Salem, H., Tamplin, B., and Tokiwa, Y.: Mechanism of
bronchoconstriction during inhalation of sulfur dioxide, J Appl Physiol,
1965, 20, 164.
25. Lopez, M., Wessels, F., and Salvaggio, J. E.: Environmental influences
in asthma, in Bronchial Asthma: Mechanisms and Therapeutics, E. B. Weiss
and M. S. Segal, eds., Little, Brown and Co., Boston, 1976, p. 503.
26. Kinsman, R. A., Luparello, T., 0'Banion, K., Spector, S. L.: Multidi-
mensional analysis of the subjective symptomatology of asthma, Psychosom
Med, 1973, 35, 250.
32
-------
27. Kinsman, R. A., Spector, S. L., Shucard, D. W., and Luparello, T. J.:
Observations on patterns of subjective symptomatology of acute asthma,
Psychosom Med, 1974, 36, 129.
28. Kinsman, R. A., Murphy, J. R., Dunning, M., Spector, S. L., Walker, S.,
and Farr, R. S.: A clinical data management system for residential treat-
ment in asthma. NJHRC Psychophysiology Research Laboratories Report No.
24, Denver, Colorado, 1976.
29. Assessment of air quality: Metro Denver region. Report of Air Pollution
Control Division, Colorado Department of Health, August, 1975.
30. Chai, H., Farr, R. S., Froelich, L. A., Mathison, D. A., McLean, J. A.,
Rosenthal, R. R., Sheffer, A. L., Spector, S. L., and Townley, R. G.:
Standardization of bronchial inhalation challenge procedures, J Allergy
Clin Immunol, 1975, 56, 323.
31. Kinsman, R. A., Dahlem, N. W., Spector, S. L., and Staudenmayer, H. :
Observations on subjective symptomatology, coping behavior, and medical
decisions in asthma, Psychosom Med, 1977, 39, 102.
32. Fumival, G. M., and Wilson, R. W., Jr.: Regression by leaps and bounds,
Technometrics, 1974, 16, 499.
33. Akaike, H.: Information theory and an extension of the maximum likelihood
principal in Second International Symposium on Information Theory, B. N.
Petroy and F. Csaki, eds., Akademia Kaido, Budapest, p. 267.
34. Bouhuys, A. and van de Woestijone, K. P.: Respiratory mechanics and dust
exposure in byssinosis, J Clin Invest, 1970, 49, 106.
35. Love, G. J., Sharp, C. R., Finklea, J. F., Shy, C. M., and Knelson, J. H. :
Atmospheric levels of air pollution producing significant harm. In-
House Technical Report, Division of Health Effects Research, EPA, Research
Triangle Park, N.C., June, 1972.
36. Dahlem, N. W., Kinsman, R. A., and Horton, D.: Panic-Fear in asthma;
Requests for as-needed (PRN) medications in relation to pulmonary func-
tion measurements, J Allergy Clin Immunol, in press.
37. Inman, W. H. M. and Adelstein, A. M.: Rise and fall of asthma mortality
in England and Wales in relation to the use of pressurized aerosols,
Lancet, 1969, 2, 279.
38. Fraser, P. M., Speizer, F. E., Waters, S. D. M., Doll, R.t and Mann, N.
M.: The circumstances preceding death from asthma in young people in
1968 to 1969, Br J Dis Chest, 1971, 65, 71.
39. Segal, N. S.: Death in bronchial asthma, in Bronchial Asthma: Mechanisms
and Therapeutics, E. B. Weiss and M. S. Segal, eds., Little, Brown and Co.,
Boston, 1976, p. 1124.
33
-------
40. Hahn, P. M., and Yee, R. D.: Medication chronolog, Biotelemetry II,
2nd Int. Symp., Davos 1974, pp. 71-73.
41. Yee, R. D., Hahn, P. M., and Christensen, R. E.: Medication monitor foT
opthalmology, Am J Opthalmol, 1974, 78, 774.
34
-------
INSTRUCTIONS
1. REPORT NUMBER
Insert the EPA report number as it appears on the cover of the publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
4. TtTLE AND SUBTITLE
Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitte, if used, in smaller
type or otherwise subordinate it to main title. When a report is prepared in more than one volume, repeat the primary title, add volume
number and include subtitle for the specific title.
5. REPORT DATE
Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of issue, date of
approval, date of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHORISI
Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.). List author's affiliation if it differs from the performing organi-
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hirearchy.
10. PROGRAM ELEMENT NUMBER
Use ihe program element number under which the report was prepared. Subordinate numbers may be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant number under which report was prepared.
72. SPONSORING AGENCY NAME AND AOtiRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGENCY CODE
Insert appropriate code.
15. SUPPLEMENTARY NOTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation with, Translation of, Presented'at conference of,
To be published in, Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (200 words or less) factual summary of the most significant information contained in the report. If the report contains a
significant bibliography or literature survey, mention it here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
concept of the research and are sufficiently specific and precise to be used as index entries tor cataloging.
(b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc. Use open-
ended terms written in descriptor form for those subjects for which no descriptor exists.
(c) COSATI FIELD GROUP - Field and group assignments are to be taken from the 1965 COSAT1 Subject Category List. Since the ma-
jority of documents are multidisciplinary in nature, the Primary Field/Group assignment(s) will be specific discipline, area of human
endeavor, or type of phy sical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
the primary posting(s).
18. DISTRIBUTION STATEMENT
Denote releasability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
the public, with address and price.
19. a 20. SECURITY CLASSIFICATION
DO NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any.
22. PRICE
Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Form 2220—1 (Rev. 4-771 (Reverse)
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completingj
1. REPORT NO. 2.
EPA-90S/1-77-004
3. RECIPIENT'S ACCESSION NO.
4' WAIRWAYS HYPERSENSITIVITY, AND PULMONARY
FUNCTION MEASUREMENTS IN ASTHMA
5. REPORT DATE
October, 1977
6. PERFORMING ORGANIZATION CODE
7'^jTcfon L. Spector, Robert A. Kinsman, Melissa
Dunni ng
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
National Jewish Hospital and Research Center
3800 East Colfax
Denver, CO 80206
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-2517
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Programs
U.S. Environmental Protection Agency
1860 Lincoln Street
Denver, CO 80295
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This study evaluated the relationship between spirometric pulmonary function measures
and air pollution and meterologic variables for adult asthmatic in-patients in an
urban, long-term intensive treatment center.
Two groups of 67 patients were studied during a late spring-summer period and a late
fall-winter period. Patients were classified according to the role of allergic
factors and the degree of airways hyperreactivity. In all analyses, 24-hour average
levels of air pollution preceding pulmonary function measurement were related to sub-
sequent 8:00AM pulmonary functions. Multiple linear regression analyses by leaps and
bounds for individuals and groups, defined by reaginic factors and airways hyperre-
activity, failed to identify a systematic effect for any air pollution variable upon
any pulmonary function measure. The lack of any relationship does not appear attri-
butable to limited exposure to air pollution variables, but may be affected by as-
needed (PRN) medication usage, type of pulmonary function measurement employed, and
periods selected for analyses. The relative strengths and limitations of the study
are discussed.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air pollution
Human health
Lung diseases
Asthma
Denver, CO
18. DISTRIBUTION STATEMENT
Distribution unlimited
19. SECURITY CLASS (This Report)
unclassified
21. NO. OF PAGES
39
20. SECURITY CLASS (Thispage)
unclassified
22. PRICE
EPA Fo»m 2220-1 (R«v. 4-77) previous edition is obsolete
------- |