IRRITATION SYMPTOMS DURING THE
NOVEMBER-1971 AIP POLLUTION
EPISODE IN BIRMINGHAM, ALABAMA
(INTERIM ANALYSIS)
C. J. Nelson, C. G. Hayes, V. Hasselblad,
W. B. Riggan, Jose M. Sune, J. E. Gerdina
Intramural Technical Reoort
Division of Health Effects Desearch
National Environmental Research Center
Environmental Protection Aqency
Research Trianale Park, North Carolina
November 24, 1971
-------
Introduction
During November, 1971, the-Birmingham metropolitan area experienced
1 • .
an acute aip pollution episode characterized by elevated levels of sus-
pended particulate matter. The impact of this episode uoon irritation
Symptoms of the eye and respiratory tract among panels of previously
selected volunteers composed of families with school children was ascer-
tained by telephone.
« •
Methods
The Community Health and Environmental Surveillance System (CHESS)
has been previously described by Shy &t..,ja1, J\cu±e. irritation svmotoms
constitute one of the raoid response health effects indicators in CHESS.
Since communities in three southeastern cities, Birminaham, Alabama,
Greensboro, North Carolina and Charlotte, North Carolina comprise a
"CHESS set" designed to measure the imoact of susnended oarticulate air
pollution, it was. possible to corrioare symntom freouency across an exoosure
gradient. Moreover, the same symptoms had been investigated in Birminq-
ham during April, 1971, when the levels of suspended oarticulates were
roughly the sane as the November 1971 enisode. The s.ymotoms were also
monitored during a control period in May, 1971 when oarticulate levels
were considerably lover than either episode, but still well above the
n^f'TO'^^1 r ^ h 1 o n f* 3 •*•••* ^* • -a"1 T ^s-/ c+^nHpvH rt-^\»-sr'f»v*vr, n +• -inv^-sr-'ci^'VT^ /•* ^ 4- a i.*n v»n
1 1 £h U I U . ,-. J C.. . . .J i W « I U — • I _.*..: 1 w_ , J U V* J I *-t «^ . — • V u . i w b* I t Will- h*..l V . I s~ W I 1 — M-^VV« ..... u
collected from monitoring stations situated within each of the three
residential neighborhoods under study. CHESS also permits examinations
of the effects of imoortarit covariates uoon the health indicators in
-------
.each study. These covariates include age, sex, socioeconomic class,
•
smoking .history, concurrent acute infectious disease, meteoroloqic
observations and reporting bias.
Results and Discussion
o
Air Monitoring. Suspended particulate levels (212 yq/m ) in Birmingham
residential neighborhoods during the current enisode were higher than the
o
April '71 episode (183 yg/m ) as shown in Table 1. During the control
period, suspended oarticulate levels in Birmingham (120 yq/m ) were
approximately one half the value of the current enisode. Substantial
variation among the three study communities was noted and the cleanest
o
community (103 yg/m ) had levels one third to one half those of the
dirtier communities. During the April episode suspended oarticulate
3
levels in Greensboro and Charlotte were 59 and 75 yq/n respectively
lower than the Birmingham level. During November 1971 the comparable
participate levels were higher than in April, 116 yq/'nf for Greensboro
and 139 yg/m for Charlotte. The effects of dose rate response shorter
than 24 hours could have been investigated if the CHESS-CKA'!? aerometric
stations had been deployed. Average minimum tenoerature levels were
roughly the s?^? in all rit'ips HuHnn thp <;turHo<;.
Symptom Reporting. Age, sex, smoking status and resnondent bias usually
exert significant effects "upon the prevalence of irritation symptoms.
Thus, symoton.prevalence for adult females, the usual questionnaire
-------
repondent, was partitioned into" smokers and non-smokers (Table 2) as were
• .
adult males-(Table 3). All non-smoking children of these families were
grouped together since no difference between male and female children v/as
^.detected (Table 4). Too few childhood smokers, less than 5 percent, were
available for a good analysis of pollution effects.
Over 1600 families participated in the study. To facilitate com-
parison of symptom rates across cities in Tables 2 through 4, all rates
.*
were standardized to the Charlotte rate to preserve comparably with the
April analysis.
Severity of Symotoms. Two measures of symptom .severity, restricted
activity and medical visits, are tabulated in Tables 2 throuqh 4. Medical
visits were highest in Greensboro among all population segments except
male smokers. Other CHESS studies provided evidence of a concurrent
respiratory illness peak in Greensboro which may well account for this.
Restricted activity v/as highest in Charlotte. Thus, there was no clear
evidence of either excess restricted activity or excess physcian visits
in Birmingham."
Symptom Bias. Evidence of reporting bias in Birmingham can be seen by
ccTT.pzirir.g prevalence ratios for joint pain, a control symptom not likely
to be influenced by pollution. During April, 1971, Birminaham clearly
over-reported this symptom, apparently reflectina an anti-nollution bias.
In November 1971, Birmingham residents grossly under-renorted this
symptom, probably reflecting concern over their economic well being in
-------
the face of industrial closings-. A number, of families had reservations
about answering the questions and some indicated they would no longer
cooperate in the studies because their jobs were threatened by anti-
pollution measures. Despite this bias, an excess in .uo6er-respiratory
irritation symptoms was still evident. There is also evidence of over-
reporting in Greensboro which may be explained by interviewer bias
introduced by DHER teams who could have elicited symptom reporting.
Effect of Time. A comparison of Charlotte rates in April, 1971 with
these of November, 1973 ^I^le.Jj^Howa/J.j-.bai^sjanpiapj *freouency decreased
among all females and among male smokers. -Symptrsn rates, however, in-
creased among male non-smokers. Children showed little change except
for an increase in physician visits. Thus, "time effects" demonstrate
the need for a temprospatial approach to acute episode evaluation.
There is also more than cursory evidence to suooort a decline in
reporting that is a function of the number of times a volunteer is ques-
tioned. Since there was no differential between cities with regard to
length of study participation, time and repeated renortinq effects
should not unduly influence inferences'made regarding air pollution.
Effect: cr Air Psll-jtlcn Eri:cdo or. SyiT-.stc.-.s. Throat discomfort, chest
discomfort, cough and shortness of breath symptom frequency was increased
in all "Birmingham groups. The largest increases were observed amonq the
male smokers. These symptom frequencies are nearly doubled and would be
more than doubled if we adjust for the under-reporting previously mentioned.
-------
There was evidence of intercurrent respira-tory illness in Greensboro as
shown by increased reporting rates for restricted activity and medical
visits among adult females and children. Eye discomfort and headache
were much more variable than the other symptoms. Increasing particulate
3 3
levels from 116 yg/m in Greensboro to 139 yg/m in Charlotte did not
detectably increase the frequency of irritation synotoms.
An independent analysis comparing the Birmingham May control oeriod
«
with the November episode also indicated there was increased reoorting
frequency among adults for throat discomfort, chest discomfort, cough and
shortness of breath. " -
Summary
Acute irritation symptoms of throat discomfort, chest discomfort,
cough and shortness of breath v/ere substantially increased in panels of
normal well families with school age children durino an acute air pollu-
tion episode (212 yg/m particulates) in Birmingham, Alabama whether
compared to a cleaner period earlier in the year or compared to cleaner
cities (116 and 139 yg/m ) during the same period of time. There v/as
clear evidence of under-reporting in Birminaham reflectinci concern over
their economic well being.
-------
TABLE 1. Average Total Suspended Particulate Levels and Minimum Averaqe
Temperatures in Selected Resideatial Communities of Three
i Southeastern Cities During and After Air Pollution Eoisodes.
City
Birmingham
-
Charlotte
Greensboro
Dates
4/18/71-4/21/71
(Episode 1)
11/12/71-11/16/71
(Episode. 2)
* iff i v/ 'i a—iirj: '^L/'f^ji -i
(Control)
~4f23/ 71 -4/28/71
11/12/71-11/17/71
4/25/71-4/28/71
11/1 2/71-11/16/71
24-hour
Average TSP
183
212
• ' "^ ; - -
76
139
59
i
116
Minimum Averaqe
Temo. in °F
54 -
43
,-54
48
43
44
43
-------
TABLE 2. Irritation Symptom Prevalence Ratios Among Females in I"hr;.e
Southeastern Cities Distributed by Snnking and Episode St. tus
Eye Discomfort
Throat DigCQjnfort
Chest DlsfOtyfort
Cough
Shortness of
Breath
Headache
Joint Pains
Medical Visits
Restricted Activity
Adult Female Smokers
Birmingham
(N=193)
UO
1.8
1.1
1,4
0.7
1.6
0.2
0.4
0.0
Grnsboro
(Less P(
(N=230)
1.8
2.4
3.4
1.7
1.0
1.2
1.7
4.0
2.4
Charlotte
fluted)
N=17g)
i.q
•t
M -'•
l.()
r.| :
i.
1.0
i
l.P
1.0
Charlotte
Prevalence
Rate (%}
(4.1)
(5.8)
(2.3)
(8.7)
(2.9)
(7.6)
(2.9)
(1.2)
(2.9)
Adult Femala iijn-Smokers
Birmingham
(N=388)
1.1
1.5
2.5
1.9
4.2 -
1.0
0.2
1.1
0,5
GrnsLoro \
(Less P
(N=3(7)
I.I)
1.4
2. ;>,
1.7
1..-J
1.1
1.1
2.1
1.1
Charlotte
olluted
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0 '
1
1.0
Charlotte
Prevalence
Rate ( %)
(3.2)'
(6.5) •
(2.4)
(5.7)
(0.8) .
(7.7)
(3.3)
(1.6)
(2.8)
-------
TABLE 3. Irritation Symptom °rcvalcnce Ratios Ai'iorn Mains in Three
Southeastern Cities Distributed by r.mokin-i and Lpisode Status
** ' e
\
\ '
'e Discom'fprt
iroat Disqppfort
V
lest Di-sco|nf
-------
TABLE 3. Irritation Symptom "rcvnlcncc Itah'ns Asnom Moles in Three
Southeastern Cities Distributed by r.mokin
-------
TABLE 4. Irritation Symptom Prevalence Ratios Among Children in Three
Southeastern Cities Distributed By Episode Status
Symptom
Eye Discomfort
Throat Discomfort
Chest Discomfort
Cough
Shortness of Breath
• Headache
oint Pains
Medical Visits
restricted Activity
Biminqham
Episode
(11=1837)
1.3
1.0
1.4
' 1.5
. v ,., . ,.
1.3
~» A , . _
' 'T.H*
0.2
1.4
0.3
Greensboro
(Less
(N=1819)
1.5
1.5
3.0
. 1.7
1.2
i •*
-- ,vo - - '
1.4
5.0
1.6
Charlotte
diluted)
(H=1305)
«
1.0
1.0
1.0 ~
1.0
1,0
1.0
1.0
1.0
1.0
1
Charlotte
Prevalence
Rate (%)
(2.9)
(6.4)
(1.9)
(6.6)
(1.6)
(7.2)
(2.6)
(0.5)
(3.Q)
-------
TA3LE 5. Irritation Symnton Prevalence Ratios* In
Charlotte Distributed'by Age and Smoking
1
f
i
t
Eye:. Discomfort .
Throat Disconfort
Chest Disconfort
' Cough
„ Shortness of Breath
..H^ache/
Joint Pains
i'edlcnl Visits
Restricted /-ctivity
Adult Females
• Smokers
1.9
1.5
2.4
•1*
.1,6 .
2.1
5.1
0.3
" 0.8
Non-Smokers
2.9
1.3
1.4
- ~rar
.. J5-2
2.1
3.9
0.8
1.1
Adult Kales
Snokers
1.1
1.0
2.7
r-TTb™
3.3
1.1
2.9
0.1
0.4
N on -Smokers
0.6
0.3
0.4
-B.^
0.6
1.0
4.3
1.1
0.2
Children
1.3
1.0
1.0
1.5
0.7
0.7
1.1
i
8.0
0.7
*Chr~lrtte orcvalence r?te in ^pril 1971 divided by Charlotte prevalence rate in
November 1971.
------- |