Document byNr4ZdRm4ovrNXqdj8e08E76
FILE NAME: ZK/L
DATE:
DOC#: BO/L
DOCUMENT DESCRIPTION::ZKZ^Z &Z>
(,---_.,._--_. - .--~--. ~
......,:.~,.'
....
o~cupations and Cigarette Smoking as Factors in Lung Cancer
LSTER BRESLOW, M.D F.A.P.H.A., teMAR HOAGLIN. GLADYS RASMUSSEN. and HERBERT K. ABRAMS, M.D", F.A.P.H.A.
sludit.-s have implicated cigarette smok
1M, 1-",'- Itudy off!" _.II.
'0 ,i-.l niJcnee ,...tl/n, connec'
"'-r, rigarf>f'. ,molei.....i,Ia railY. ,,, a.dido... i, Ii,,, lftI.ral
'0 ,.",K".,;"n. ,hal -... laft... tom.
,Ia. r,;ulc>f:;uU rellllloa,hlp to
di..
ru'" TII,,-"-.I lur,"a, " ..Jr, ..
",,,,.,d. it i
ing 3-1 and certain occupalions' - es. peci~lly those involving exposure to ratliolldiva 1 and claromate ores I and asbestos I-in the cau~ation of lung
e:lneer. However, the number of per
sons known to hllvc such occupational
expo!oures in the United States is rela til-ely small
:,11\.' !ooh:trp incre.uc in lung cancer
rlurl,llily constitutes one of the mo!t
.j...I:trul:lr di!U1C phenomena of the
j'.1'1 1\\0 dCf':1des. In the United States
I:,, :I;:~3dju5Ied mortality ratc for lung
=t(~llu6llI".'ir n"n'imibned19f3r0o
m to
2.7 p 11.0
er in
100,000 1948,
-' ..ro' Ihan fl)urlold jnc~ase. In at least
"m.' ~'alr. C:llilomia, pulmonary c.lnccr
I."" :tl't'tlllllts for more deaths than docs
"ullllun:try tuberculosis, aot{ among
I;~L... :t~ed '15-64 it ca~ 4 per cent of
.111 .1.':111""
AII""{lls to explain this rising mor ulit~ '" ~purious, e.g" as due to jm
,,,,;\\.,1 tli"gnosis of the discl'lse, do not
"OUIJ up. '1bc lung cancer death rate-
.. hid. unfortull:llely is 8tiU our best
"j"',..,im:ttion of incidence rate - has ir nc'.rI~I ~-d two and half times more
:li'l\ ,. among men tb:an among women. ~ t:.t' \'.&11 l~rJly presume that physiciana
:..m.' OIll'l'licd diagnostic improvements '., ':I\~' I more completely to men thnn
:"-'U":I"Ink.'nr. ILikewise,
the f
increasing
d
pro-
i..f .1.I..I11110 a ung C41ncer 'oun as a caul!C
. . Ulons autoptlle5 supports the
''''a I.".'lin lIl'lt We :lre d('~1illg with.
:".11
Il1('fC~!-e
'.t
In ule
l I'15C8~1
}.n\ ir"llmt:ntal changes appear to
'.,....r I~ Lest explanation for the
.. "':'lIk'tIt\n. PrC\'iously publi~hed
TIle prC!ent jnvestiglltion WDS under t:lken lour years ago: (1) to determine whether' a!ly additional oceup~lions might be involved in lung C:lncer llnd thlls oeservc intensive inYcstigation, and (2) to obtllin dill. on the role of tobacco U$lge in lung C:1ncer.
Methodological Considerations
Definitive study of whether an occu palion or cigarette smoking is. caus. tive factor in lung cancer requires that Ihe inci,lcnce rate lor the exposed group be dclennincd. ntis rate js then com pucd with that of a control population, llae moa acceplable control being all the rest of the population Irom which the
Dr. Breslow is chief, Bureaa of ChroDic Di~...- . California Slate Dcpanment of Publie Heallh. Berkeley, ulif.; aDd Mr. lIoaglia and 1\ln. naamulScn are aleo auoeialed with Ihi, Jepanmcnt. Dr. Abrams. who Wit formrrl,. chief, Bureau of Adult Health. California 51alo Oep.1rtment of Public He~llh. ifsunndo,wCmhieccalill::o310dffiirCefc'. toTrh. eaIfteear llabDdanAdmuWseemlfaernot DuiLJin. ].1niton' Union. Local 25, ChicaKo, ilL TIlit paper is ba.ed oa a .tud,. supponed by a reaearr.h j;T&nl (CS 9M7) 'rona Ihe Na tional Cancer lnalitule. PlIS. aDd wu prer.cnlcd lIelore Ihe Indu.trial Hyc~ne S..ctioll of Ihc AmericOlD Public Health Allodalion It Ihe Eil:hly6nt Annu:.1 Mectine In New York. N, Y.. Novcnlber 10, J953,
171
(-
~--...."...- .
172 FED.. 19:>4 A~lERICAN JOURNAL Ot' punuc HEALTH
exposed group came, AMuming the
aame diagnostic and reporting standards for both groups, one may thus deter-
mine how the chllnce Ulilt ii member 01
an expo!~ group will de,oelop lung
canter compares with the chance that such a person would develop the dis. eA!e if he were not a member of the exposed group. If the likclihood of developing lung cancer is lubstantially in creased by being in the exposed group (eog. a ~ertain occupation) one may conclude that the exposure is a causative flletor in the disea~. The key to this analysis lies in determining comparative incidence rates for a suspect group and a control populntion.
But nrst it is necessary to identify .uspect exposed groups. Thnt was the
aim 01 the present study. The im-
portance of pin pointing these groups arises from the fact that incidence studies require I!~mbling sizablc popu. lations Ilnd observing the occurrence of the disell!C in them-& large and ex peNlive tuk. If. for example, data were obtained implicating occuPlltions beyond those already known to be cauMtive {/lctors in lung cancer. one would be justificd in I!lsembling populations of the occupational groups and observing them for lung cancer' inci dcnce rntC5-1l !ltudy which would de velop more definite proof of the pre!ence or absence of an etiological relationship.
Several methods of investigation have advanced our kno",-ledge in this sphere. First, is clinical observation. of a!!locia' tion betwccn a case or a Cc\\' case! of lung cancer and some exposure. e.g., ashe!to!l. One delect of this method is that such clinic:l1 obsen'fttil'ln depends brgely on chance and often involves such .mall numbers of cases that the validity is questioned. Where made, of course, such observations are helpful
Another common method of idl!nti. fying suspect Occup:ltiom in a highly fatlll discase, such as lung cancer, is to
noIe tlle frequency of the nriou occu.
pations on death certific:llcs where the
disCll!C "ppears I! the cnuse of deeth. Thc frcquencies of the occupations on such death certificates may thcn be eom. par~ with those on all dcath certificates {or the general popuilltion. If certain
occupations appear with much gruter frequency on death certificate1 lor lung
cancer than on certificatc! where death is due to other caUSC!l. one ma,. conclude that the occupation may be in,olved and should be intensiTely inve.atigated. Important achievements mar be credited to the use of this method. However, major difficulties with this use of death certificate data arise Irom the following
circurrutances: only one OCCtJp8tion may be recorded; while this is supposed to be the "usual occupation," it is often not correctly recorded and duration in the
occup:ltion is not recorded. All these circumstances tend to blur the identifi. cation of .uspect occupations from death
certificates.
The method selected in the present study was to determine the lifelong occuplltional and tobaceOoUMge history of lung cancer patients (ond a control group) by actual interview with the patients. The work history included tho number of years engaged. name of compnny, industry,.a description of dutic:t, and listing of products and hazardous materials known to the patient.
'" Data Colle.ction
," .
." .
Patients ohserved during 1949-1952-<'0
in 11 CaliCornia hospitals (three county gener:ll hospitals, five vetera~ or mHi.
tory. two university, and one. private
hospital) comprised the study group.
An i!'terviewer especially qualified in
occupatiollnl anlllysis obtained the d:lta
fro In, e:lch patient in the cooperating
hO!lpltals who!C condition permitted the
.interview. The session with each pa.
tient oIten luted two hours or more,
Slnce many details about smoking wue
.,
I:nSb -oed and the inlonnation about
~o Ztaal looccf u5p1a8tiopnatiwenuts
highly detailed. who finally had
h- topathologicaUy proved diagnoses of
cancer were ultimately included in
the analysi&.
Fot e\'ery patient in the case aeries anoilier patient adlllitted to the hospital aLoul the lime time-of the same aso (within five yean), !e..'C, and race-for a condition other than cancer or a ch~t diM-aN: was chosen in a random manner as a matched "contro!'" Two inter-
1'ie9(t'1"!, one working in Northern Cali. lornia and one in Southern uJifornia ~pitllis. obtained the data from tho lung cancer patient and control patients lI!'ing the !amc achedule in identical m:lIlner. That interviewer bias did not tnler into the i:clermination of tobaccoUl':lI;e hi~lory acems likely, among other ~:lSons. becaulC the results closely _ """'" p:!r~lld thOle obtain~ in other investi. 1- - 'sallons where IUch bias apparently wu
excluded.... and because the results for the two interviewers were very similar. It is difficult to conceive how inter
viewer biu might have affected tho occup3tional data. since both lung rant'Cl' p4tienb and control patients tne1ltioned hundreds of occupations and none of lhnc was suspected by the intfTTiewers. Preliminary findings wero kept from tho interviewers unlil the end of the datA collection.
LUNC CANCER VOl. '" 173
Method of Analysis
Each record "'as rmewed by a
physician, particularly with re!pect to the pathologic diagnosis_ An indu_st~ial hygiene chenli!t completed the hsttng of materials usuaUy involved in tha occupations mClltioned by the patients but not neces!arily known to them. TIu:rea{ter. the data Were coded and tabulations prt'pared..
Attempts to analyze the data concern ing type of illtiU!lry and exposure to nloterillls already identified as toxic or carcinogenic did not prove fruitful Hcnce the major analytical effort with rC!'pcct to work hi!tory went into the data on oupation itself. This had ~n coded according to the Dictionary oE Occupational Titles (2nd ed., 1949). prepared by the Division of Occupa. tional Analysis. U. S. Employment ~rvice, a code with 367 three-digit titles. Particular attention "as devoted to the qUe!lion of years of exposure because pre!Ct\t knowledge of carcinogenesis due to environmental agents indicates that duration of exposure and a long latent period are critical factors.
Data on tobacco usage were analyzed so as to reveal the type and degree of smoking, as well u the fact of smoking itself.
Finally. an attempt was made to !Cpa rate the effects of tobacco u!age aDd occupations which appeared implicated.
Table l--Ac- Distribullon 01 Lila. Call1cet' rAllenl. _d Controll
174 FID.. 19:>4 AMERICAN JOURNAL OF runuc HEALTH
T.M. 2--Sex nnd RII~. D1..ributlnn or Lun, Caneer
P.tlentnd Control.
Sex and Race
Total
TOlal male White NelP'O Other
Tocal Female While Nl'To Other
Adenocarcinoma
46
42 41
1 4 l 1
Calel Other Types of Carciaom.
472
451
420 20
n
21 IS 3
3
Total
518
493 461 20 12
2S 18 4 3
Coalrol.
TotAl
518
493 461 20
U
25 18
4 3
Findings on Smoking
Tables 1 and 2 reveal the distribution oillie lung CIIIll.'er patients according to their biological characteristics - age. sex, rae!!, and the essential similarity of
the matched control lUi~ in fe5pect to
these characteristics. Tobacco usage by lung cancer pa-
tients and the control group is shown in Table 3. Here one may note thnt 481 (93 per cent) of the cases gave his-
tories of having smoked cigarettes;
whereas only 394 (76 per cent) of the '
controls gave a limilar'history. No luch greater frcquency of tobacco usage in other forms (pipe, cigar, chew, and Inuff) occurs among the lung canur p4tients when compared with the control group. In Cact, 68 of the control group u:ted tobacco in a form other than eig3TcUes exclusively; whereaa only 15 of the lung cancer patient. did 10relative frequencies which are not much
TaLle 3-Touaeeo U.a.e Durin. 20 Yean Prior If) StadT Reported by Lan. Cancer 1'.lieat. and Co"trol.
Manner of U.inc Tobacco
Total Tocal not smoking
ci,an:ue. No use of tobacco Tobacco other lhan cillrell~ only
Totalllllokini ciprelles Ciprclles, only Cipreltes and olher
Adenocarcinoma
46
_ 6 4
2 .0 31
9
Caoes OtherTypea of Carcinoma
472
28 15
13
441
298
116
Total
518
S4 19
-IS 4M l29
ISS
Coutrol. Total 518
124
56
68
m'
2.w
154
.'
I
LUNC CANCER VOL. 44 175
Table 4--Amouat.r CJc.~II. Sraokh" Durin, 20 y~." Prlol' to Stady Reported by Lun, Cancer Patlenll and ControU
Packa;eeol
Cipl'dl" Smuked pet Da,.
Ca.es
Number
Percm&
Controll
Number
Per cen&
Total
518
100
518
100
" None
M
7
124
24
<~
19
4
SO
10
~-
71
14
lOS
20
1-
3M
59
199
38
2 011 Dlont
80
IS
.22
riot recorded 10 2 18 3 ",
difl'l'lCllt (or the nonusers of tobacco in
the t\\'O aeries. nlis .ugg~tion of an eHect by cig.
:&rt'ltcs only is consi~tent with other in \~ligatioDs in the field. Some support (Llr lhe hypothesis that cigarette smoke iug .ffecla the de...e1opment of epithelial cOireinonla of the lung more than ade
1I''''ucinon1& may be noted. Six out II! 46 (13 per cent) of the C:l.5eS of ( _ , illll"tloc:l.rcinoma did not .moke cig. - ':areUcs; whereu only 28 out of 472
1.6 per cent) of the patients with other t~ pes of carcinoma, described variously
:l squamous, epithelial, undifferen ti:att'd, did not .moke cigarettes.
One commonly used measure of the amuunt of Dposure to cigarette smoke
ing, the a\'erllge number of packs mloked per day, appeaR in Table 4Se,"enty.(our per cent of all lung c:l.ncer putients reported smoking one or more packs of cigarettes on the average per day over the preceding 20 years, com pared with a frequency of only 42 per ccnt among the controls. "Excessive cigarclte smoking" (two or more packs per day) occurred almost four times as commonly anl0ng the cue series as among the control group.
Other items which may indicate Ihe umoullt of cigarette smoking include the age at which the individual began liIRoking, whether or not he smokes be fore breakfast, and whether or not he "inhales." Table 5 reveals that, by all
l.LIe 5-Ace or Bt';rinnin, Ci,nrelle Snlokinc. Slnokln, Hoblt.. and "Ci,IIreu. Couch" Repor1ed Ly LUDI' Caneer Pallen'. and Conlroia
usn
Control.
P~rcen&
Per cent
Orareue Smokin,
~raclerisUc:a
Number
Baaed Oil Ciprc:ue
Pereent
81~d
Sued Oil Nunlber Cilarello
Percmt Sawd
"
Snwken
0111
Smoken
011
Only
Toul
Onl,.
TOlal
sa TOlal
518
100
518
100
TOl,l c:I:;AretIe a&Ilokera 484
100
93
394
100
76
Ik;;,~n J5 yean of a.:e 166
3'
12
116
29
22
n~'&Il 15-2'
286
S9
55
2-13
62
41
Jk.;&n 2S fe.... 01 'Ie II
7
6
3S
9
7
Mllihale'"
449
93
87
327
83
6.l
-!'rnoLe brrore weald..t" 373
77
72
229
.. \
""ciC:U~1te cou"h'"
204
42
39
112
28
22
" ....~ ~ ~ i~( ..
.! ~~. I~
~
;:..~ ~1"
..... ~
I
~...-.
I :.
.;
~
,za ~
J 't,: ~; '~
~!::
.
I
:,. ~:r. .!.
-
li6 FEB.. 1954 A:'IERICAN JOURNAL OF punuc HEALTH
or Table 6--ullmaled Pre..lencf! natel Lun, CaacM' pel' 100,000 ~'aJca
.1 ~Iec,ed A,es _d Varl.-m. Cll:llrclle SmoklD, nahit
Number
Lanc Cancer
Control
FAtimated PrenJenee R.at~ Quantit)" Smoked Daily t
All Qutn. H~avy titiea
None
Re1a!iyc Pr.,,..lcnc:e :
Heary
All
QIWI' titica
P.tient ,;ed 50-59 ye.n
1. Doll lind Ilill
276
275
94
42
2. 'i\'ync!er .nd Cr.h.m 258
210
117
47
, 7
!3
n
6
11
'" 3. S.dow.ky, Cilli.m,
.nd Cornfield
198
208
87
]2
7
4
.. Present
213
213
102
46
6
17
7
p.tien...ged 60-69 yean
L Doll .nd Hill
197
199
137
63
2. "WyndeT anc! Cr.h.m 187
160
138
70
3. S.do...~ky. Cilliam,
.. .. 0
5
29
15
.nd Cornfield
76
71
44
71
5
9
4
.. Present
168
169
433
78
17
2S
S
t.u.a.04 aceo.41., ..... _ ...... 01 C.... 40ld.u Dol& ,. . . .Jioto ., DoU .... R1II, ... , . . . u4 c.a.Ma, aa4
104......,. CUlla.....eI c....,oItI .....~. Ia.........r.U
J,. t 0.11, _011.4 ...U,....... ,.. ,......... la .11 1 _ 01 _.~I., n~ ci............do rI.- u4
= etrua ....."e4 eb.h ~.,t.... aa C11Ah"" f.U.",,, I. DoU 4 HlU-l OL ef l.lAces ......k 4 .-ipnttH
.1,.. = .1,........ = ...." 2. W,.eI....... C,.tIA_1
$
1 plpolul %~ cl........ : _ J. Sad....'. Clw.... ...
= = .d,. _ ..., ....... c.....1.lol--1 ci",r 10 .1......... I __ pI...._
_1..... oal, elcanne _ ........
20 .1....11... 6. fM ,...at ..
<i&ll,
~ 1I.1a1l p....Ie Ie .h nil., Hn.,....._. ~~ ....Ura/nll _. _ _........ aD (USud_
".ft' ............... aM-C Mlokenlnle .. H..., _oll., I.......... 110. f.IIIb.. of el..............1..1"1. el .1Iantt........, ..... DoU
ad HIlI-.o Of .,.... , 2. . ., ......... Cr......-U .r .... 1 S. S.d......'. ClUw.. M4 Cera'.~1 .. _ , M4
......,-.. 6. ,~I
ow _ _
of the5e criteria, the lung cancer p:1. tient, Imoked cigarettes to a gr...nter ex tent than did the control group. AI!lo, approximately two-fifth. of the lung cancer patients litated that they had had a "cigarette cough" u~ually antcdating the onM!t of their illness by many yearll; 120 said that they had had a cough for five years or more. The frequency of IUch "cigarette cough" was only about one-hoalf as great among the control group and only 39 of the control patients indicoated that their cough had ll1llted five years or more.
Significllnce of Findings on Cigarette Smoking
The data clearly show cigarette smoking to be more frequent and inten~e among lung cancer patienh than among
the control group. Our data parallel that of other recent, well controlled !ltudies; oall indicate a politivc correloa tion between cigarette smolcing and lung cancer, a correlation which increases !tteadily with the amount of cigarette smoking. One can hardly escape the
conc1u~ion of Doll and Hill thoat
"smoking is'a factor and an imporUtlt
factor in the production of carcinoma of the lung," and the more recent con. clusion of 0011,10 "The results amount, 1 bdie\'c, to proof that amoking is cau~!" of bronchial c:ucinoma."
The cigarette smokcr or prospective cigarette ~m(Jker may wim to know how much his chances of developing lung
callrer are incrcased Ly luch .moking.
Acknowledging that a correlation exists, one may still inquire how msny times dot's cigarette smoking increase the
-
LUNC CANCER YOX- '" 177
liktlihooc1 of lu~ cancer? Further,
,...hat is that likdihood in absolute
terms? To anawer these questions
definitely would require that records of
cigarette smoking {or large numbers of
penous in tho lung cancer ago be
:llfCnlbled. Tht'!'e persons would then
be pi.1.ced in eat.c:gories with respect to
dgaretle smuking, e.g., nonsmokers.
moderate smok~rs. and heavy smokers.
Finally thcy would be observed {or a
wfIicient length of time to determine tho
r:lte of lung c3ncer among the variolU
.:attgories of cigllrclte smokers and the
nonsmokers. (Such studies Ilre now
under way.)
.
However, statilltical devic~ have been
deyelo~ lor data luch as tho!e in the
present investigatiun which, i{ one is
willing to make certain as.sumptions,
yield D.n answer to the question, how
m:lny times does cigarette smoking in
crc:ue the likelihood of lung cancer? Application of Cornfield', formula 11.12 to the data {rolll (our Etudies. including the pr~nt one (Tahle 6), indicates thllt for men aged 50-59 the prevalence of IUllg cancer is trom four to 11 times as great among amokers as among non smokers. and from beven to 27 times as great among hcavy smokers as among noruomokers. Similar ratios were found (ur the 6(}..69 age' group. This formula involves the assumption that in 10 far as amoking history is concerned. the pa tients are a representative ..mple of all lung cancer cases and the controls are a rtpresentati\'e sllnlple of all other ~r !ons in the general population.
Findings on Occupations
As noted earlier, the records for lung c:mcer patient and controls showed aU
(- )
TaLlo 1-Twenl,. I'er cent ~.ml.le Crona Cumplete U.t of
Oeeupational Group,' In Which There Were at u ..t 5
Ca_ Of' 5 Controla Employed COl" 5 Yean 01" ~Iore
Occupalioa
Pnsons Employed ill
Ot-ruralion .t LUI'
SYurw
c.~es
Conlroll
006 Al:tbon. editors. and reporten 088 Slaip c.&ptaint, lII.t~ pilOlS. and cn&inet'n
7
9
8
8
, lOS Oerlca, c~a1 offiee
ISS Can..._ n aad solicilors
2
8
3
11
180 S.ICWllcn to consumen
7
'ZZ1 Wail,"" aacl w.itresses. except pri..te family
14
10
266 PUIk:cD_ and ddcetiwe, public tenice
6
3
307 Animal uIII Jivatock fumen 317 Fann I.-nda. :!nimal and livestock
6
5
10
10
433 Ceneral w-.J..-arkia, occupations. 110& t:lHwhcre el...~ed
8
1
.s 516 Painl."... <:seept coDilraclioll aad mainlenaace
3
6
524 Brick aad atoaemuona IIId Iile setten
5
1
SJ2 Collalnaetlcm occupations. not el~..here d...med
2S
542 l.ucomoCiye fi~lZIen
S
6
S60 Attendant.. fillin, atalaon. and parkin, lOCI
10
9
S83 ~f~'(:hllliCl and rcpainn~ not elsewhere cl.auified
15
16
1'(.... !-,&e naJ-.I, .....,.4 r.- ,lot II _.~'I. . . . . . . . . .101, - llot .....~ , " - , .
......,.......u _ h ....... S ....... J c_tnJo wit" S ,..... _ .... _,1.,-,). o. ~
--hot ,. u... cla..aM
s.4~ u, _ . Ie ,lot DI.:,_., .. O"....leooal Tlta.. (2.4 N.)
Ea,Ie,.... .we..... D. Co, U. So
Serrlc I..'.
~ ~
~
:~~
:;.
~
7~
~
.;'; """
...;:
,~
~~
.-:: :;!
. -' "~;~r~..1.~''\..'~,
7#~~~
.~,,.;
.....~~
~
~"':! '.:<;~'..
.!~ '
. ~
.,
FF.II., 19.ii A'It:nICAf'i JOUItNAL Of rUDI.1C HEALTH
or ;; Tubl. S--orru[NIllon Crotlr' - In Wl,leh the Fn"<(Urnc:y
Emf-lormenl AmonI&' Ctl~. Wn t un.' Twice tI9
Yur
CrTal Amnnl( Control.
Occultation
Penon. Ernr10yeJ in Oecur~linn II Wtt
5 Yeun
Cues
Control,
016 [nllnren<, ci,U
5
138 Stoc:k dprkt
7
266 Policemen and detectives, public lC"ice
6
4.'lO Lumbrnnen. raftamen. anti wood choplX"fS
11
4.3 ~lachille Ihop and relalN occupAtionl, nol
elsewhere c:lauified
8
485 Welder. and flame culten
10
520 OccupatiOlll In prOtluc:lion of petroleum
5
523 Conllruclion machinc:ry operalors, nol
ebewherc cl.uilied
7
52-1 Brick and Iionemasoni snd tile setlen
5
5J8 Brakemen. railroad
10
570 Fircmm. other than procen firemen
11
573 Cr.nemen. dcrrickmcn. h.,i,lmen, and ahovelmeD
15
575 Drillers, ntraelion of miner.ls and construction
7
1 3 3 5
3 1 2
3
1
5 2 6 2
'e .0...,.,1... ,ta..ill.. ~..di.r
.Ita
1"'. 'truh;n,'c.a. D. C U. S. F._,le,_o' S.nic.. pI.,...... . AI k ..e S .._ .Ilk S ,.... ...... eI
DIc....uy eI o...,..lJ..u Tic... (:.4 "-).
J.c1.~" . .I, _ ...~ 18 .~j,.""'_
occupation! followed and their duration! in years. A listing was prepared of the occupations puraued by at least Ii"e patients or controls for live YCllrs or more. This listing re\'ealed 81 occupations in tbi!! arbitrary frequency-andduration category
Table 7 presents a random lample of thc!-e occupation!! and the numher of patients and controls who at some time during their work history had been engaged in them for five yean or more. It will be noted that among the 518 lung cancer patients aeven had worked u authors. editors, or reporters (occupation) for fh'e years or more, and. among the 518 controls, nine had been similarly engaged. Relative frequencies of occupation would appear to 1I:\\'c the aame logical significlUlce as ~uch frequencics of cigarette smoking. The lat ter is. of course, much marc common among both case and control groups than is any aingle occupation. Hence.
the number!! pertaining to anyone oc cupation would be less imprCMivc than in the case of cigarette smoking. However. the frequency with which the lung cancer patients had pursued ~rtain occupations wu lubstanti.Jlly grcater than the frequency with "'hich control patients Iud pUr!!ued thcse S!lmc occupation:! .
The occupations for which the fre quency among case!! Wa! at least twice all great as among controls. using the !ollme frequency.duration category as before, appear in Table 8. Such a tabulation. of c;purse. show! the ~cupatjons classified according to an arbitrary code which was not devised to reflect common occupational e:tpl\SureS that might Le respon~i}'le for lung cnneer. Further analysis therefore included the deLnilcd examination of records of individual cues and control! and regrouping of occupntions according to possible common exposures. For ex.
LUNG CANCER VOL. '" 179
:lIupl~ one occup3tion31 gruup. number
530 (plumbcf!l, gas filters. and steam
fitters) showed 16 cases and 10 controls
with a five year etn}Jloyment hi!tory.
EuDlination of thC3C 26 individual records mc3lcd that suh~lanti3lly all
the: difference bct"'een C3..ce and control ~cries was due to the steam fitters who
hnd be:avy exposure te; asbestos. It
,,'as also noted that other types of work
in\'oh'ing asbestos occurred more fre-
quently in the employment history of
ca..ccs than among the employment hi.
tury of controls. Thus, Il ne\ category,
including !Iteam filters. asbestos workers,
lind boilennllkers-all involving expo-
sure to ashe;toa--was found to occur 10
times among the cases and only one
time among the conuols.
Table 9 presents the data for occupa-
tions grouped according to posaiUly
!isnificant exposures. Although num-
r _ ) bers in each o( tbe categories are small,
,
tlae differences between ca!lCS and con-
trols appear to implicate the OCCUpAtions
listed. It is of interest to note that
tcVeral of those with the molt striking
differences involve exposure to metallic
particles and fumes and products of
metallic combustion. This observation
tends to corroborate the sugg~tion of
Wynder and Graham 13 that "hot metal"
occupations may be invoh'ed in lung cancer.
In reeent data of Dollio the only
occupation \vith an important excC!'lS in the cancer group was thllt direetly concl'rncd with the production of gas. Little or no support was found indicating that
other occupations suffer special risks. Howc\,er, Doll's !tudy was bllsed upon experience in' England where the em-
ployment ll:!ttcrn diffcrs from thllt in the United Stllt~, especially California. Also, Doll lI~ed an occupation code of
only 76 cnh'snrit"s \"hich may ha\'c heen
too coarse to rel'eal the differences found in the pre~ent investigation where a code of 367 categories was U!ed and anlllysis carried to subgroupings of
tht.~.
It mlly be noted that 77 person! with lung cancer had an exposure of more than five years to suspeet occupations (unduplicatcd couut in !C\'en occupation groups) according to the data in
Table 9. TIlis number reprC!enls a not
insigniric.mt proportion of the total (518) lung cancer patients in the entire
study. A further quC!tion up"n ,,,bieb the
present data might he expected to throW'
some light is the e:ttent to which ciga. rctte smoking and occupations operate
Tilille 9--Ca_ and Controla Employed Co.. lit Lea~' 5 Yean La Selected Oecoration Croura
Occupatio..
1. Wduen and aheet metal worken doing ",eldinl 1. S:ram CUIen. boilerolaltrn. an<i 3~bnlo. worken
J. Electric bridge r.nnc operalort-metlll indllJlrl
(. OCCUl"llioll1 in the exlraction of 1~:lJ. ~inc:.
and cOpf'd' ore
1:C5, 'fir;
"incen firemen. oil lind wipen
ConSIn.Jdion lind m'lntenanc ainlrr1l
Coou. commercia e:u: u iOi cannery cooks)
Penons Employed ill Occupation a' Leu.
5 Yean
Cases
Controls
14
:z
10
1
5
1
9
3
12
6
%2
1%
J5
21
nc....,..tJ..., ......... ........ .. ""-'11i". MCer4i.. ,. c...- - ecn,.lt..al ~... .h..
4.t.Lt-A .............. f'ft' .....
100 n:D~ 1951 .\~lt:nrCAN JOURNAL QIo' I'UDLIC H,\tTH
TliLle 100Example oC AdJII~lmrnt Cor Smokln ,\pplletl 10 W~I,le? and Sheet l\I~lnl Workera Dnin, Weltlinc
NOlie < Y.. 'h -
1 -
201' more
Not reo eonled
Total
All c:a~ anti conlrols,
male -
l.Ca~
2. COlllrol. 1. Tolal 4. Prorartion of
C:lH"5 anIon, total c.~,
and controll
Weldera anti ,heet md.1
"'orken doin, welding, male
5. Casr-t 6. Conlrol. 7. TOlal
8. Expected Dumber of welder C85ct t
22 110 132
0.167
0 1 1 0.167
16
69
296
80
4S lOS
193
22
61 174
489
102
0.263 0.396 0.60S 0.783
0
2
11
1
0
0
1
0
0
2
12
1
0.0 0.792 7.260 0.783
10 493 18 493
28 966
0.357
0
14
0
2
0
16
0.0 9.002
Ex,...o4 .._ I.. ., .... _ _ _!.I.... =-'.00-1 =. ~
= '-I = ~................i.. ., ............Id...
16 '.11$
11
" = NI_~.' ,.... ..........,__ U.i.. S I_ .~....... _ - (0.611. '.tS4I_.holo.lI, .ipli__ .. " . .
.... 1...1.
11
eI."... ,,_ ,... "IiH. r....,.. ..
1 ............ NI, 1..... 1.c1od... I...... p. . . .Ioiclo . . . . . . .
.iIa. _lee ft' ... t Ot.,.i.o4 h ."I,iac 'M
.f .._ . .- r ....1 ......
'nl. Ia ...11 _Mia. cau.., .. tIoe
u.. ._k ., IJo.. Ia
_M'" ea.....' (Ii If.. I X
No.1).
"u. I M.IoI. A. 5tI.d..1 T.~I" ... r_.IoL " ... Y.rk, J..-' Wil.,.
cr.o,. pO... : T~ .... oJ......,,11." n ~......, ..,..
110.... Ia ToWo t.
".., .,... M ..... "'-'
tu ..............,... ........ _0""''' . . . . . Ita. ~",4"11.. 0' ...hacic.l .ip,,~ ~. I' IS . . . . ., LtnL ICOM" tM
... 1iI.., .....up. . . . . . . . . .5 ,.... ~..............tit'ir.' "PI~ca.C
as separate factors in lung cancer. For example', both cigarett4 smoking and
welding appear to be related to the
d~l1se. but do welders get lung cancer more frequently merdy because as a
group they are heavy cigarette smokers?
Although data from the prC!ent study
involve small numbers for many of the occupations, analysis does indicate that welding as an occupation oper(lte3 as a lactor distinct from cigarette smoking in relation to lung cancer (Table 10), The group of ,team fitten, boilermakers, and as~to. workers lies on tlle border.
line of Ittltistical lignificance when the
.effect of cigarette mloking is controlled.
In the other occupational groups the
Ilumhers nre such that the differences might arise Cram chance variation more than five times in a hundred.
However. all the occuptllionl listed in Table 9 de$Crve intel15ive etudy to
cletermine their precise relationship to lung cancer, One important step is to as~ernLle populations of individuals em-
ployed in these occupations, and with varied smoking histories, in order to observe their lung cancer incidence
rah~5. These rates may then be com-
pared with the rate for the corre3pond. ing a~e~x group in the general population. One may thus determine
LUNG C.\NCER VOL 44 181
how being in a particular occupational
frl'uP :lfTccts the chancf'S of dcvt'loping lUll" c.ncer and likc\Vi~ how cigarette tlll:1dng mOllifies the!-e chance!.
and mech:H1i~m r~~pollsible should be identified in order that prolc!<:ti\e mes'!-
ures may be illtroduced.
Conclusion
The d3ta in this 6tudy con~titule still
,"ulllt"r liuk in the ch;1in of evidence c(ltlnecting lung cancer with cigarette
,",,,killS. C\'idence which the New Eng-
I.llld Journal of Medicine not~ to be ~w ftrnng as to bl! considered proof
,,ilhill the cnryctay meaning of the
",ord." ,. It is time lor those concerned
lit ",ith health education and with the
tlhics of aJvt"rtifing. leaft in he3lth :and mcdic:1l journnl~, to lake note. rurrhcr invC!tigntion should leek the c~:lct component of cig;1rcUe smoke-t:IT, arsenic, or other-which is respon.iLIe.
TIle data also suggest that sever:1l
"("'up:ati,,ns (Tablc 9), io addition to lh/J~ prcvioll~Iy identified 4S having an
diological relationship. may be involved in the development of lung cancer.
TIII.'!C occupations should noW be in-
tcn~ively 6ludied to determine to what
extent persons engaged io them suffer
a speci:!l ri5k of lung c:!ncer. TIle"," :ttcr. if an etiological relationship is
"~'3rly rl'lahli5hed, the prcc:ise c:tpofure
1. D M. ['" DnU. R., 4 Kf''''I,. E. L. c..en
at tbe ~"I !. R~I.Ii T4Mcc" 8rat. J. Callcl'
S:I. ItSl. 2. Schr.k, I\~ D,k.,. L. A. I.Uard. C. 1'.d
DoJ,04. S. Tobae,. SWlekin. . . I. ['{ule,ita' FIt'-
'''' I. Di...N. C....r 1I...."'h 10:~'~. I9W. 3. 'IV,.du. E. L." C.. ~.... ,_ A. To~.<co !.,ok.
la. . . Po..,bl. [dol_etc riciN 1ft Drcbi....f.
c."'........ c..... A IIt.d, 01 '" r.......
,-'.U.A.
I'" :J:t3J6. I~.
,II.
Dell. R." 01 C,,,,,..._
.I.Ii111.'"A". "I...,.
A,r5il..1M, .01I.
A.,'olo.., ::1271-12"
19SL
s. Lrw , W. L. C.IJ,t.i II...... c:..,h.,dt. P. R. CalM" 1114 r~eC'c. !lIlo~ia" A r U,.'nar, . . .
r .... J.A.~I.A. 1"':336-lll, 19W.
Hper. W. Co <H...... I.....I Ta...... """ AW"
DI..,._ Srrito.Ao' III no..... 1"2.
7. 11_... 1. 0.. 5o.P4'. E.. .... !.h",..I. C. DI.
U,,,I....,1.h..it II", ["b"llev.. f. SC'h".....YI la
s.c~... t"SCh"Hbto'It" La......h ....'. ZlIcil,. I.
It....ol_k 2l:3~'. If:L
..01 M..klc. W..
Cr...._.. F. c..cor...N
1M 5...... ftn"in'ory S,...m ia
C.il'"
PuL
It..hk R~. 6):111"'1127, I.....
~. C~ S. L ..........._ . , BI....eclc&1 !O"C,
"'.'erial c..... t.a... .. N"....J'
Ia u.s
1 :110,
111.51.
10. 0.11. I. " ... ~"I C."I. ._ . wid &OJ Ad\.
r... .1-.7.
I .... P... II. 'rtl. M. J. S:1-S2f.
""Ht...,. D.... I. c.r.ta. ~$90. I~.
n. Con,.I I. A U.'~'" of t ..l.lllo. Co_ranll..
I.... CII.I.&1
A"IIe.h.. c. cr of
lhe
8,.........
J. Nat. c:....c., t....
II :1:6'-127$. I~I.
.".1 e - 1%. 5od...kJ'. D Cim A~ C.r.hl J. Th SI I.. Ic.1 A_,",.. ..._ S.~kift.
,'n_ u.n,. of lJoe
11>1 u:lu:-n:;l. I?U.
Il. W,..d.,. L L.. .d C....... E. A. Iiolo.ic F......
r. 1'" III .rMMhi....... u'iflao !lpdi.. R"~,
c..... w l.tI....,ieJ [.,..ru: It,....,... lli t"r~rt
"",10. lod II,.. lOre.,. M... 4:!ll. 1951.
S.. E.,I... II. Edlrlol. C.N.. 01 .h ""...
I.
...... 2" :4" -1104, I'..,.
Journal 25 Years Ago
_\nD~Dl1)1 to the January, 1954 "25 Years AJ;o." In addition to the articles "II the ('OIllUlun cold mentioned 1:1~t month, the 1929 Juurnal (p. 419) d"~t.:riLcs
311 ~ishtp:!oe booklct "That Mean Cold," by prc."Cnt.day a~sociate editor of the Journll). RAymonJ S. PallerlSOn. It W3!! reportedly his hr!'t production :15 director
of h(,:lllh eJucation for the 10hn HUllcock Mutual Life In!urance CO'"P:lII~', the January i!sue (p. 9.1.) having rcported his nppointmcnt as director of health
,.,luIalion, Life: Comcrvation Scrvice of thnt cornpnny. Dr. Patterson hnd alrc:tdy
Lc"t'll til.: Ilut!tor of A.J.P.H.'s "Annolated Public He:thh Bibliography" for !e\'cral
j.:n~. it having fir~t appeared unJcr his by.line in March, 1926.