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.

-------