Tobacco Documents Online

(- J~) TC 900 r' .. Time Trends in, Lu'ng C~ancer~ M'Ort~~aiity, Amo~ng~, N~o~nsmok~~ers~~ and a lN~o#;e~~ on P~assiv~e~ Smo~kaing r~~ ~ Lawr,ence, Garf0nke11 M:A: a~'.a ABSTFiACt--Zunq cancer mortality rates were computed for nonsmokers in the American Cancar Society's prospective study for three a--year periods from 1960 to 119T2' and in the Dorn study of veterans for%ttiree 5-year periods 1rom,1954 10 11969: There was no evidence of any Drenp in these rates by 5-year age groups or for tne total groups. No timetren0,weat observed in nonsmokerm for cancers of' other selected sites e:crot fer a decrease in cancer of tne uterus: Compared to nonsmoking women married' to ncnsmotuing, rousbands. nonsmokers marriedito smoking hus- tsands showed very'littley if any,: increased risk ot'tung cancer:-- JNCt 1981, 6bt;t@'GL-1ofib: Afortaliiy rates (irom' llnng cancer in, men in the UnutrdStaites have bocm risingsteadiily since1193b ('thc first yrar thcse cancers were classified separatel'+y) and in women since the mid-1960's. It has grncrallj- brcni arc,-ptrd that the major reason fior the increase has brcn the ciyarc+ttr smoking paucrns which 1x-6-nn in y,ottng men :trmtrnd 1'1'or)d' l1'ar 1'.:tin'di in young vvomen in the l9'30's and 1940's. A larga body of cs•idrnce from epitirmiolvgic and panhologir studies on smokrrs con- firms this conctusimn (J,). A n•ccnt es'timatc ofl the pcrcrnugc of cartcrrs attriibtut:lbla to snto{:in>; in men was 34.5A for total ctncrrs,and 82.8A' for lurng,ctncer. in ..•orncn the rornparatilc perccntugcs uvrre, 5.4 and' 43'-1% ' (2). This analysis was tyasrd on dita Grorn the lhrbc cpidcmiologic study of the ACS and c:on•rred the pcriodl 1967-71. It was based, on a nuntbcr of' assump- lions that would give slightly different figures if the smoiiing, distnibtutions in the study population differed from nhosr of the general pop'u0atiion or if smoking dl~scritsutiirnnsch:tui,ged in thelate1c970'scomqsa,rrd to the latc I9Wrs (as they irndecd'': ha.•e in wontcn)i Tlsc•rr ht,s been a suggrstion. liowc.•rr, that tlic Itnng ancer trend in nnnsmnkc•rs has also iurcrcau-d inI the United Staars over the ycars. E'nstaom (31 sr:ttrd thnn "a more cornplt,tr tundr.stand,inr of liung cnncrr ctiiolo5y is ncrdrd." This analy,sis indicated a l:tr4c rclati.•c nnrTrasc' inr lung, r.tnccr noortalut'y' in nrntosnac,kcrs in bc>tih white mcn amd wliiirc- wontrn lit.•lwncn 1,914 and )9T5 on the basis of ani isatY-rlrrc•t:,tion of dat.i in santpitn of nntic;na1 inort:nliny statistics am& sr•.•cral, rTixJc•rniolugim suutlicsin diJft7rrnt Nrriodsof nimr(3):. Einstlcnrm rncogtiiz,rd th:tu ntbst~ off thcinrnrstsv occurrcdlir•tl.ccn a 1914 sur%c•y of eft,:+uli rryistr, tion artas in 24 st.itrs and n:,tnonal mortality stauis'tics rr)xrrtr•d in 1I955 an<]i t'hat most of tlnat incrraso t.as lrrubaluli- attriluuu:tlulic to incompirtvrnrss of rrportistg lung, Mtncrr attd' to cltangcs iin diagnoseic e7itKriaL tlW poss;ihilitY cxists th:in lujtl;,cancer is increasing in norrsmoi.cr,s who have lnad ificrra.ing cXp<rsurc to otlic•r fac,tors-occtrj>aticotr:ll cx'l>osures, general air, po'Il~uition, and F?crhapscvcn uopyssiti•csTrloking, (inhaling the smoke (tom slnokers): Esrn, if these (actors were related' to the alleged increase in lung, cancer, they could have had' only, minimal' cif'rct, on the tup.+•ard' trend for, lting cancer in men. since the mortality ratesamongsmol.ers and nonsmokersdiiffcr so' g=atlj•. Moreover, in the last 50 years, and until recently; most men had a history of cirarrrte smoking. )I.Imong women lung cancer ratrs remained low up to . abous. 1960. SSince nhien, there has brcn a threrlold increase in r-aues attributable in', large part to the changes in smoking patterns among women during nhee preceding two or three decades. In', this paper, informatuon is pro,.•idrd' on trrndli for lung eenrrr (and cancars of' srvcrctl othen satKsy in nonsmokers over a 12-%.cari period (1)fn0-72) fnom data in the prospxrtirr study of the ACS. In additicm, dat:t for no'nsntokrrs brom the Dorn studn• of vetcrans fut the years 1955-69 are givcn. Whili, such data do not provide evidence os•cr a.•rn• lom;, tiirtc spxnt they are based on the tM'•o~ )•argest pros).x•ctin•r studies in thc United States and covcr a l7'ryv:rc Ixriiod (ront 1953 to 1972. MATERIALS AND METHODS Proct•diures in the col)rction of d:rta in the prc»prr- tive sau'dy of the A'CS have. Ix-cn p:rrsontrdlin a nutnluc•,r' of ptoblicrrt,ions (4-6). Thcre were 94,000 malr andi :;75100rJ' fttm:ilc nonsmokers aa the start of'. the studv. In the ACS saudy. a'"nonsrrtol:er'" is one who rclu)ruc•d firv or slae h:nd: nrvcr, snnnJond or sntokrd only cuc.t.sintoallh- t but had never srnol:rd rex;trJar)p. Cainssifhc.tniinn was rnadr as of' tdtc stsra' of the s:ardy,. and vc•ry frn• nonsmohcrareportnd that thoyst:tnu-& toctnolcotn any of Lour later yursuionnaires. FnrolPntrrtt of, su}ijc-res in thc AiC_5 snidy Ix-.t,:rn in Oraolx-rCL959' and' rarrracii•d th'roueh Marrh 1960. Fezt- low-un wn!s complrac fcor 98'.4 0 of Ai subjcxus throuc;ii Jitnr 19T1'. and 92.8b, complete for the 1'tit.h %•car vf tlbc study. Deaths wrrc rcportrd by the AC:.S voJutotrr•rs. and death ct:'r'tilicattes ..°erc obtuined front stutr, lec•al'th dc- A~~~xan~~'uilt0~ t•t't.D. AICS-.1+ncrican. C'nrrr S..ti,tl': t Rnn-inrd. t7rrober23'. 1980; xor,p+rd'. J:+nu.+rl' 2b. 196J. +' rleT+artnurnt't of' Yi,tdr'mnlog) - andl Sr.in.a+rn. Amcri<.nC+u+rr$ix,itap,. 777' T1urd tA.u,.. 1:c,.- 1'ort., \.1';, 1U017. t I' ths+nk F.u.g.rne Ro6nt (ot +u.ptii.-inR, tVir d.rnaa for, rlir IJnrn ae++dy o(' %rtiRhn.and Hcnrl 1'.~qrrc+ .nd t..ind-1lrtdin+r for. 2»ro+in/; in tiK 11 r«vs•iorgof tter daeas in; thiti %rucl., 1061 J,\cJ. s.nt. tz: so 6 )trxf Iwnu

1362 . Garfinkiet partment.s. Mmrtality data for this anal'wsis bcgin with observation starting on Judy 1. 1960. Data arc presented fmr, three 4'•Vea'r periods: pcriod 1. July 1, 1960, . through June 30, 19'64; period,2., July 1, 1964',, throu;gh Junc'30, 196L;, and': period 3; J~uJy 1, 11963, through June 30; 1972. Pcrson-vcs!rs of observation in nonsmok€rs and deaths at singfe years of attained ages 35-89 years were computed and'combine'd by 5•year aruined age groups. In the Dorn stndy of', veten!ns; questionnaires were mailed starting in January 1954 to 293,000 veterans holding U.S. Government' life insurance.. About ~ 63' : of the qpestiionrtaires were' retri.-ed over a period' of several months. In, January 11957a second'questionnaire w-as mailed'' tb thosr, not responding to the firsti mau:ling and the replies raised the uotal to 85, (7)1 About 34.0001 of those who replied were nonsmokers. The same cla'ssific2iti'on' of nonsmokers was usrd'. in uh'is study as was used for the ACS study. Person-years of observa- tion andlmortality by singie y;ean of'antained.age were computeA starting with: January 1., 1955. for the res spondbrs to the first'mailin'g,and starting with January 1, 1958, for the responders to the secondmauling.. Death certificates were supplied to the Veterans' Ad- ministration in supporti of insurance claims through, 1962. For the period 1962'-69; deauh certificates wcre' obtained through field work at hralth dcparu:rocnts bg ACS personnel'(8)'. Drath rates by , 5-year age groups were ad'.ittstrd to titrc distribuuion, of the stationary population ;,.L,,):ci whitc men and white women of ages 35 years and' over inthK abridgrd lnfe tables for the U.S. population in 1965 (9)', Differences in death rates for periods I and 2' atoc periodk 1' and 3' were tested for signiifionr at the P<0.05 level by the Man1tei-Haenszel1 procrdktrr (10), RESULTS Time Trends fn Lung; Cancer Mortality Among Nonsm'okers Table 1: shows the 5.vear atrainrdi age dc;tnh rates for lung cancer arnon'g, nonsmokers in rhreo periods ofl Lime. The table includes men'. andi women iin the ACSI study and men in the Dorn study of veterains. There were 1!95 deaths from lung cancer among male non- smokers and 564 deaths from 1ung, c+nccr among . ftmale nonsmokers in; the ACS stntdy during the• 1!2- veair perir.d. Thert• were 168 deaths f'romitung caincrr among nonsmol:crs in the 15-)•ear pariocd in tiie Dorn study of vrterans. Some of the r:utcs comvuucd for 5- TABLE 1': rDeafh rares Jrom ltrnq eancer per 100:000: persor-yrara aunonq nonarnokrri. aqrs J.5-89 vrora, by :imc prriori: ACS proapeciiw, study and thn L."arn aLudy o-f :-rtrranu ACS prospective studya Dorn'i study of veterans' A'ttained age smup, yr' Period 1: Period 2 July 1960- July 19@9r June 11964 June,196'8 Period'3: Period 1: Period 2: Period 3: Jiuly 196&rJan: 19'55- Jan: 1960-- Jar.. 1'965- June 1972 Dec. 1959 Dec. 196AI Dec. 1'969' 44airs 35-29 40-44 d'5-a9 - (4.0) (8.7) (5_1) (1a.3) (103.5)(8 6,) Sfu-64 (5.3) 8.8 (8:8) 5:L59 10.51 11.6 8:3' (120), - - 60-fiii' 17.0 173 1'7:5' 1T2 (10.7) (a8'0) 65-69 18_6 29.4 3d.3' 25_1 16'.9 ;3'.S 70'7y 32.3 26.4 1!9.2' 39.9 a0~5 382 75-79' 32.7 41.5- 58.6 (37-8)1 (15:4> 47.2 80-81 (47:9) 1I06:8' 51'.9 - (2(10,61 (20!6i) 85-89 61 _8 152.7 (69,.9)' (595 2) - - Na. of dkaths 52' 74 69 38 52 78 Asne-standardizrd dearth,rate, 12:5' 18.5 15.8 18.9 13 4 Fernales 35-39 10-11 - (3.5) (3:5f 15-l9: 5.9 (3!3) (1.6) 5U-5r 5.2 7:7 (3.0) 55-69 7.4 8!0 60-64 1 A'.0 1'2:3'. 11.5 65-69 15.6 1151. 17.7 70-7/ 19.4 21.1 22.0 71 37.3 30S 36.3 80-34 51.5 . l'S'.1 40.8 85-89 53.4 ' 44.5 5915 ' No. of deaths 1'75' 184' 205. Age-standarditrd death ratt 13.8I 12.9 13;1 ' Some 5'-yr age gsoups were, combined': in the standardiuition of ra'tts, to a~•oid'. 0 ctses in t'he<r groups. ~ X'i+oulrrs ii,, parrwUrses indicate <'5' deaiths' ini group. 11y7. t-;DL. 6G: \'O~. 6:, JPN~t; t9tt1:

Lunpi Cancer Trenda In Nionsrno'ken year age groups were small and subject to eonsidt-:a!blr sampfing variation. There was no appearance of' any consistent inceasr in the lung cancer death rate among nonsmokers with time by 5-year age groups. The age- st,andardized rates fbr, males shown in table' 1 and ini text-tigurr I showed' no trend. T}se rates for women were based on many more eses. and the age-standard- i¢ed rate was vinually the' same in all three periods. The differences in rates bev..'een' periods I andi 2 and periods I andl 3' were not: statistiolly signifiontu in both the ACS' study, and the Di©rn study of veterans. The analvsis w•as baset3' on the' tnndt:rlying cause of death on death certificates. The deathl rates for the three periods were also standardia*d to the distribution of the stationary popula~tion of white men and women comibined; of ages 35 years and over, in the abridged life table for the U.S. population in 11965. This standardiratnon raised thr raues for males slightly and dKCreased the rates for females siightJy. but it changetji the pattern of' the uends verry little. An attempt was made in the first 6,veara of' (oldow'~up in the A'(S studil• to obtain confirmatiion ofi diagnosis (or all cases +.•ith, cancer ftom physicians who signed the death certificates or from hospitals in which death occurred. Information was ret•eived, confirming the priimary site of cancer' in 78% of' the cases, and microscopic confirmaaion was obtained in b9A of the cases in the first 6.•ears (6). Table 2 rsho..•s a comparison of tfoa dcarch crrrificatt; diagnosts and the final diiagnosis from the medical report. Among nonsmo'king men., 74 were rrporiod to have died of lung cancer according to the death crrtifieatcs. Six of these (8,1iq) were reportt-d': to havee d'ied' of cancer of another siie' on the final report_ However„ 9(0.8:A): of the deaths reported as being du€ to cainccr of a siitr other than luing, on: death certificates proved to be due to lning cancer on the fsnal report. Thus among, nonsmoking men there were 74 deaths from liung cancer rel>ortrdl on dt'ath certi(icatcs and, 77 deaths from lung cancer according to the final mcdical report: 0 ~ 0 30 o :o 0 0 1t to. r-y 0 DORN rUID'r;, ..,c.s: sTUOX 4EN wEN ------ ---- iE AC:S: STl>tiTt; w Or.tEN nso nso YEAR TiXT,/n(a'lut 1.-LLnc ti..n mort.diip nsrs in iJu c 4•.rlrrni..di lai nnnsmmarr.a iniIr.4,(:S, trnr1>Tti.r siud) .and.t.nd nnn.n,rplqt in i1biR S•yr 1rTiodi in 11ir Birrn uudr oi %rictini, TABt.L' 2.-L.cotp cancrr dsaLlu arnonp +o+vrno.cm in' fs+=1'6' yrars q/ rtudy tne dcatJt urrltijlieatrs aved owt Jiwat' reaort, Fina!' Death certificate diaRrwsic reimK Lung oancer' Other cancer disgnosia 1;Ib: Percent No. Perren:' Mslb Lune,rsncer 6& 91.9 9 0.8 Other eanrzr 6 8.1 1;1'53 99.2 Total 74 100.0 1.162 100.0 Females Lung cancer 169 8v!3 10 02. Other tancrs 34 16.7 5.160 99.8 Total 208' 100!0 5.170 100.0 In ..~omem the picture ..as, somet+•hat different. Two hundred and three cases of'. lung cancer among non- smokers were' repo'ned to be lung cancers on drath crrtifitares; and 34' (16:iq), were reported to be cancrrs of otht•r sites on the (inaJ' medii-tll report. A smaller number, 10 (0.8%), of those cancers rhati were reported as being of a sioe' othen than uhe Iiung on death certificates were reported: to be iung cancers on the final report. Thus on death certificate reports. 203' nonsmoking women %.•rrr reported to have dird of luns;, cancrr in the first 6 years. Ontlte final re}?ort, 179' (a decrease of 11.8%) were reported to have died of liung etncer. About one-third of the 34 frmales whose c,uses of death were antribtned to lung cancer on the death ccrtificaues and changed on medical confiranaaion dicd fi-om breast cancrrs. Hoxvever, brcast, cancer ..as,under- diagnoscd on d'cathi crrtiifintcs in nonsrzool:ing,ti+•omrn.. There were 1.31!0' breast cena'ss reported oni dbac't, certibicdues in the first 6 years of, the study and 11,371 on the final reporr. 'I'ablie 3 slrows thc ai~r•sw'nt:aroiizr•d rates fc~r total morraliitw for all cancers and for cultt'crs of sclcctt-d! sites among nonsntol.crs in the tltrtY' titrnc pcriods. Oitrr:nll morta,lity in, men drent-rsrd 3'. frrnno Pcrio& I to 3. T'nis slight difference was staitistiraJ'ly signilir:tnt at' the P<D.05 ' )rve3l bccausc ofl the largp nitmhcr of dt•:tuhs involv€d.. Ntone of thc diffk•rrnc•rs in total r=tnrt'r or in: c-Anccrs of othcr s'Jtes in mc'n in table 3' bt'tn.•ean Prrior{s I a'nd 2 an& pcriads I and 3 were s'tairistically s;iGrnifir cznt: l1°omen had an 8% dt'crt:ax' in total dr.ath rrtrs bctwecn periods I and 3. Thr di'fft'rrnrc in ratt-s was statisticaJly significant. The dccrcases in riotal c:nncc•r and u'tcrinK canccr, bco+•rcn periods I and'2'arsd lxr,iot?s I and 3 M•ere stauistic;idlry•, signif6nnt. Nonc ot thc d6fferenres for r.tnicers of other sites were stanistir:rli.• sig~niifiicainit escelpc fortlbr 29% deccasc in crrnrctsof thobuccad c:in•iay, pttairynx. lar.•,nx: aitd exu}phigus tit'twrcn rxriods I and 3.. Passllve' Smoking, A, nunabc'r of stuoiiics have cst:ibliishcd that non- smokers exposed to s,ntol.c f-rom t',igarr'ttrs in a poorly J1U. t•lnl. 61 . N~o.. 6. Jt•\L' f"r

1064 _Garilinkei TASiL 3:-7}rt.di ir nmrtnluty rof.um fmm eavurrs of aeLtdmd niW, in tAree taine periods for wo,unnokrrs A CS prroapec•lnue r:irdl 19'6D-:! Parameter No. of deaths Period 1: Period 2: Period 3: July 1960- Jiuiy 19641- Tuly up6c- 3une 1964 June I196e :une 1972' I;Lales Total'l deatths i 9i05 L.608.7, L588. 6 1.5 59.9 Totzl i cancera 3.15'11 247.8 2'52.i 251_6 Cancers of buceal avity„pharynz, llirynx. and'eaophagw 62 6.86' 6.79 5.46' Cancer of mion-rtctaam 636 51.9 4S'.0 50.4 Canaer of panernss 199 15.0 17.6 14.0 Cancer of proatate siand 573 69.5 6311 59.6 Ferrtaies Total deaths 52.965 1.49K'.S' 1.495,8 ' i.374.2 Total cancers 13275 317.9 30a.6', 295:1. Caneera of bucnl'nvity; pharynx• larletz, and esophagus 159, 4188 4.21 3;48' Cancer of colon-rectum 2,429 58J0 59.8~ 56:7 Cancer of pancreas 688 17:4 1'6.2' 14.6 Cancer of brmLst. 3.186 69.3: 6E-0 75.0 Cancer of uterus 833 22.1 18.4 15.0 ventilated room will show increased levels of carbon monoxide in, their blood. These higtier, levels of carbon monoxide mn result in de•tcrioration of psl•ditzmotor performance: Many nonsmokers have acute rye and throat irritation responses in the environment of ciga- rette smokcrs (J1'):, One paper reported changes in lung function, tests in people classifirdi as passive smil com!parcd to nonsmokcrs, and these changes: were intrrpretrd as demonstrating a, greater reduction in the function of small airw^ag•s (12). M,iray,ama (13)' reported' lung mnccr mortaliiy rauios in Japan ranging up to 2:1 in nonstttol:iing, women with husbands who smoleed 20'or, more ciearcttrs a day compared' to nonsmoking women with nonsmoking husbands: Trichopolous et a1. ('11') reported similar findings in a, easr-control study in Greect. A, similar analysis was made of nonsmokers in thr. AC5 stiudy, even though classifying nonsmoi:ing,t.+omrn on, the basis of', the; smoking habits of their htusbands is not an arcurate mrssurc of their drgrae of passive smoking. N4oreover: rxposmrrs in Japan and C;nrce may be very different than they are in the 1Jniitcd Staucs., Lung cancrr mort5lity, among porsons who were rnarriedl to cigarette smokers was romnarrd A.•ith the mort2liiy among, those married to nonsmokers. A totail'of.1D,6:739 nonsmoking Armnion wcre idrnnificd who were marri!cd'. a) to men who never smoked, b)i to, men who cturrenuly smoked cigstrctacs rcgtularJj• but less than 20, cigarrttes a day, and c) to men who cwrn•ntlR•y smoked 20'or more c,igar~cnies a day. A1ost husbands had smokrd for 20 or naore years before tbe study tac-gan, and presttmalily their wives %cc•rc noorr, 16{;cJc to have been }dassive smokras thain .roro the womrn maruird to nonsmok€rs.T;.•c-nt.=t•.ig.ht wercc•nt of the hnrKlD,rtds of rnonsmokirtg, Womrn were nonsmokers cornp;rtc7d to 21 -_ of men in the total', st:udj• popvdation.T;rble 4 shows the resulls of this anudi-sis. Exprctrd nnnilx•.ra of dcaiths %+•crc based ont the 11,ng cancer rates for the 12- ycar Ix•riod (1196D-72): by 5-ycar age groups of thc womeni with nonsmoking htrsbands. No ataontipt ,+•as made in this first analti•sis to aditust for otlocr possitul'r confounding factors. The obsemed! versus exlPrttKc lung cancer mortalii}• naio for wonten whose ltustrtncf! smoked less tlian 20 cigarettes ai das• was 1.27: for thosc whose husbands smoked 20 or mort•, cig.trcttcs a: d:ty•, i, was 1.10. Nriither of these diiffcairncrs tvas stauistinll~ sigpiiicaint au,1'<0.0a bs• thc AfantK•J'-1{•tc•nszcl ,zrcaccdure. A separate m•rtchrd~Grottps :rnal\sis was rn:rtdc nf' thc 6ung, cancrr deaths antong thr winor 3 grota}zs of women -to eliminate the possible rffrcts of potvnti:rl conifoundins, factors. The wontrn iin~ the 3 Groulns %.•cre matched by age (5-}•r age groups). race ('w'titet non- white)• hrgfnest, edvcational stanas of hiusbiund or wiife (not a high school Sraduatc, h:igli school gradiuime, or higher), residence (rural, not ruraJ), aind' }3tusirrnd occuoauionadly rapost•d, to dusr•, fwmos: o,r rapors (yes , or no). The aill nas rrstrictrdl to rtorrsnoof,ing women w•ho~ wrrr, non sick and who h;ud+ no srrious disease at the s.tairt of the stuld}•: "A,ripwstt-dr' nu rrttic•rs of deaths for, each m;tuahed diiad were conuNuued., as describcd itn onht•r otublicttions (1S~, 16). In this pro- TARL 41•-Qb.rnxd urrrut erprcttd' fvnq'conrrr dicthai amonq nmesrnoZ-inp uomcn uit/i eiqarene +rmoJn•ny km.cbcndw; ACiS arudy!: 1960-72° 'iusLand' ivusbana l$ usba nd smoKed <20 srrn.-:.ed ?20i Paraimettr did',not't smoke cigarettrs/ day ci;Rrcc:es/ . day Obsemiedlde=ths 65 39 49 Expected'deaths 65:00 30 67 44 67' 1+lortalit.• ntio 1.00 1.27 1'. 10 ' Expected deaths are based on Lhe lung cancer ratrs by S-yr age Zrc.ps in wo-en w;tfi nensmok:r,g iausbsnrs appiied to .+-ie per~on-yez.~rs of --omen k•ith, smoking husbands. ' The 95% eonfidence hmit for women witnihusbal Fmoking. <20 cig'airTttes/day -•as 0.85, 1'.89. for women with huRl+ands smoking ?20 cigarethcs/dag-, it was 0 77, 1.61. 31C., [•oL, GG. KO;. 6: )ll u9Rt (i ~ 7S 3Y

Lung Cancer Trends In Nonsmokers 'D,.ats' b.->i'fatched proup .tudy.• .tdjts+teC hms{r caAcc- dessAu amo+,p .oo+ncn soitJa' wnrrno+carp ho+.+banda marciuad" wi:n ro+nrn soihh nno.CinQ husonracu Group )r" o. af sdjusted lung cancer desth(r Ra' tio Nonsmoking husband 25:6 1.00 Husband smoked <20: 35.0 1•37 NS eigzretatt/day Nonsmoking husband 34.5 1.00 Htuband I smohed ZL20 °,;5.8 1•0Y NS cigarrettes/day ' See text for deftnitson of ad';usted deaths. ~ Iilatr}ied on the basis of a) wi:e'r 5-yr age grouD: b) husband'% oaupational exposure, c) highest educational level of husband or wife, d) raoe, e) urban*r.:r.l tasidenee, and Jl) absencrof:serious d"oxsse at' the stsrt of the study. ` NS'-not signiticanL ccdure women whose husbands never smoked were comfpared to women from each of the 2 groups in which the husband smolcrd ciza!rrtnes. The number of lung cancer deaths in cach matchrdi diad' was adjiustrd to the proportion, of persons flor each grioup andi surnmrd over all groups to give an '"adjustrd„' nwmbrr of' It¢ng cancer draths: Vairiances were computed for each ofl the matched groups arnd snimuttcd over all matched groups, and probabilities .+•rrc coanptued under the nntJ1 hy,prxhesis of obscn•ing, no diffkrcncrs: The results of this analwsi's are shown in table 5. The ratio of adjusted lung cancer dcaths ini wotnrn whose husbands smokrd'.liess than 20 cigaretucs a day to those in women whose husbands never smoked Was 1.37. The comparable ratio for women whose husbands smokedi 20' or, morr cigarcttrs a day was 1.04. \onr of'; these diffcricncrs were statistiical0y signiiflicaint (P>0.05): Dt5'CUSSiON Dirta from the two proslnrtivc stiutJirs rcp(trtcd in this pctlx•r indintc that thc asc-i(djiusrcd mortnlity: rate for, Itunfi ran(Zr in nonsmoking mcn 35,89 ycars old'. was bctwrcn 12 and 1'9/!00!(600 in the 1',9501s a!nd'. 1!9G0's: Thc obsrn•cd ra¢o for womcni was ibout 13/ 100,000! The rate may actuallj• bc, abouit 1'0'. less because lung oncer in nonsmohi!nit;, .,omen, may be ovcr-rcporucd oni dcalth ccrtifirates. The lung ctncer ratrs shown in table I may be sliis)anlj• diflfrrcrnt from those shown in other p,ub1ica1ions because diiflfcrcnt ycars, aigc groups, or, mcthods of standardiz.ition were cm ployed. T}no rates for mnile and it•malr na+nsnrokcrs by ' agr gryoup in this anal}•sis were in alibun the same rattg; as t)iat of nhc 1958 rates for nonsmokers in H:rcnszrl's rcport, of a 110: sample ofl deauh ccraifit-:ucs in the Uniurd Staitcs (17, 18)! The P9GG-68 rstiimatcs dctivcd', by Enstrom from sr.•rraf sonrccs: are not dirc•..ctily cotn- pariibdc1wcauseof a~ difJiIrcnt clZssifieatinn oftron• smokers ("never: smoked algarettes") (3). The male rates in ttnc pcriodl 1968-72 are about one•half those rrportrd by Ens"rom for active Mormons in 1958-75 (19), Enstrom ocfined active Morrnons as a conort that cautl be considered "'aimost entirely as white males who rnevcr smoked," and he used this cDhon to serve as the nonsmoker lung cancer rates in the 1968-75 penod "in lieu of recent national mortality da:ta on nonsmoktn."' The moruiaty rates for lung cancer in both malr and (hmalt nonsmokers by 5-year age groups showed no consistent trends over the period' in this study. Long-term effects of passive smoking are difficult to establish because of the problems in ciassificarcion. It may be misleading to classifj• a women as a passik•r smoker or not on, the basis of her husband's smoking habit. Wives of nonsmokers may be more eupbsed': to cigarette smokle of others than wives of, cigarcue• srnol:ing men; wives ofl smokers may be very )ittl'r exposed to the agareatc smoke from their husbands or others. In addition. 13% of the women nonsmokers ..•ho dird of lung cancer in the ACS study rrporaod nhar they were previously marricd', and the classification, of their rxposure to thcir husbanid5' smoking may not be pert i nrn t. In au(lops}• studies ofl citzarctte sntokers, there was a dosr-rrlatcd slxcnrum of histolbgic findiniGS. including basal cell h}•prrplasiat mctapi~rsia, and cells ..•ith atN Ypi- mi nuclei, in the mucosa of the tracheobnoncl¢ial grrr that may lead to in%asivc carcinoma. In cuntrast, advanced histoliogic chanGrs in specimens from thr• trachrobronchial trce, such as lesions with six or morce cell ro..•s: icsions having 50A or more cc•,lls witJn aty,piioil' nuclci, and, carcinoma in siitu, sverr fround ini lrss than 0.1' of the slides of nonsmolcrrs (20). Since there is such littlr variation in the avlxarancc of these his,tolbgic changes in, nonsmokers ofl diiffrrcnt agc, scx, and residence, it seems doubtful that those nonsmokers who had be-cm hc•aviily €xposcd to ciiy,trctar smoke from others in their livrs could havc had many morc precursor Icsions for the drsclbpmrnt of larng cancer than nonsmokers not so cxho.cd. Thrrefore, there is r.•idi•ncr (¢orn thrsc sttoduos that jnaRsivn smnkitog rrnnot play more nhan a vrr)- smallt role in the d'cvcJoptttrnt of 1l¢rng cancrr. Mortality ratios for male srnokcrs of lrss tltan 10 cigTIret¢cs a day compared trJi those of npncmnkers rangr frurn 2 to l, in Japan to ncarl}• 5 to: I in the United Sta'trs. Mortality ratios in t.•ornrn arr r.•rn loWer. Ju appcars unlikely on a biologic basis• thorcfiore,, that' svi.rs, trnth hiusbands who smoke 20 or more ciga¢ectrs a day can ha.•c morrtalitn• ratios uhut :thlrrtnach those of regneJar cisarictnc smokers• To ofsraiin data on passive smukinS, in nonsmoking women, an t•pidcmiiolcrgic study shotulidl be sprcifi-ll~ drsignc•.d to mc•asurc tliK-ir rxuosure as accur:trclp- as possible. This is vcry difficui¢ to do. Naiahc•r nhr. Japanrsc stntdy nor uhe ACS saudy s.•as df sigrocd to obtain drfiutiti.-c iotformntivn on passis•r strtct'king. Da¢a for luns, rrutrrr risks in occupation:i)ly capox-d' nonsm@I:Prs's conlparrdl too notlraf?oscd nonsnlCl:krrsJre ,L\'(Z. 1',()l:. K,r 10. 6:. )L'yE. IYnI

"7(D66 'Gartlnkat not very eztensive: One uudH ,<hm«red an inar-cd' risk in he=viJv exposed asbestos workrrs on the hasis of a small nurnbcr of ases (W). it would be iratrrest,ing to corrtaimre srudies of Isng nna:r trends in nonsrnokcrs over a lorag' period: of tirne, but the magor public hcallh probiern in kang aneer is. with cigarette smokers. Cigarette smokers' who are octupa!tionally exposeKl to asbestos have' a grcaciy e3evaued nak compared to the risi: among, ciganettr smokers not so exposed (2)'): Lumg' onrrr rates are rising at an alarzning rate' in women who smoke cigarrtnes. Edwotional efforts should focus on. smoking.- crssation programs fbr these groups and: partieularlyy on persuading young, people not to start. Even if the estimates from this analvsis are in error and there was a slight increase in lung cancer trends in nonsrnokers: it didl not aippcar to be an important problem in the overal'J, pict~ure for tt9e' time period of tiois uudy. RIEFERIENICES (1) Public Health Service. S,moking and healih: A repon of the Surgron Genrral. SS'ashrngron; D.C.: U.S. Go.-c Prtnt OIf. 1979Jn'k9EW publiratron No. (irH/s)79.50066). (21 Hwr+wono EC. Ssm.rr,w HI Smoking and rancer in rhr lllnired Surrs. Prr.• Med 1980: 9r1!69-173.. (3) L%*•Mor+ JE. RSsing lung'ranrer mornafiir among nnnsmokrrs. J)+tCS' 1979: 62755-760. (I) Hi,ws+oKo E'C S,mokin6 in relation to the dFath rates of' one millbn men and womrn: Nail Cancrr Inst Stonogr 1906: 19: , 1'27-2DM. (3) HAarsrowD EC. GAUtrrws;u L Coronarv hr.rn disrae, suokr andl aonic anrnrnsm. Faeaon in rhr errologv. Areh Environ Idralth 1969; 19:1167-182. (lS). Ga'RFaue[u L. CGnrmrmonaliiy, inn nonsmokers_ Pros/lncti~ uudy br the Amrricon Canca Sarir•ty: JI:OI 1980; 61 :'1169- 1'173, (7) 1:.omw HA_ Thr Dan uudy of smoking and morulktv , amont; L'5.: ve+cnns:. Reaorr on erx:hta~nd onr•half years.of otlrrn,,. uon. Nasl' CGncrr Inu 1•'fonoprr 1900: 1!9i1'-125 {?) RbGr.i,'r' )J4U.. )ducrur JL, Smoling and ousas of druh amonE (Y) U.S. vrtrrans: 16' 7rars of obsrn.a tion. Pvbluc Hcalrh Rcv 1980; 95.215-2:'D. Pubiie Health Si~ica. ):.ilo'ubles: Im National Center fkA Hralrh Suusuc. l'iul suusuics of the Unired Surrv-1965. S'oI )1 Mornality;., part: A. 1t'.shrngron: D.C: U.& Govt Print OQ. 1967:1-8. (JO) MA'A7El. A:, HAEaaE' 1- tC. Searrstu:al aspnctr ot' the anad.,us ol data from rettrmslsreniut uudrn of'disr.sr. J! tiatl CGne" lnsri 1959: 22c7 ) 9-748. (11) Public Health Scnei:r. lnsoLrntan• smoling. !n: The hralth consrqtannaes oi smoking. 1975,. Atlanu. C:a:: C'enrrr (nr D'rseasr Conrrol. 1975:83-I I12: (12) SltskrCl JR. Fmou HF. Small-air+.a,•sd.sllrssrrinn irvnnnrnn'.ns chronically r:posedl to tobacco smok'e: N Engli J Mrdl 1'9R0. !ir-.~720.7M, (73) Hitusw.ux T. )i:on•smolinK ..•inrs of hra.-v smokrrs ha+< a higtirr risk ol' lung onccr. A studs•' ftom J.(un. Sr \7rd J 19811: 2112:1183-185: (1if) Tiut-HO.otruos D. kALA.ntntiA. SrrRtuos L. MArW.,'ran. Bl Lung nncrr and passiur srnol'rng. Int J Gnnrr 19Mi: 27:1-;I (JS) Hw.raroau EC. GARrt.%atiL L Scanst.. H., 12sr EA. Tar and nitotine aorrrent of nlrarrrrr smole in rrl.unun in drirh tatrs.. E'm•rron Rcs 1976; C2:263-274. (16), HArusosn EC. GarlinNt:ZL, L Avprrun .url r rnunun hr.irt drv-asn Findingr of a~ prnspnru.,r stud.. Hr: .t'lbd; J' 1975'. ?''b9-274 (J,7)', Hwt wEL N'. Lcat'E +Nn 1)B. SrKr tx SUC.. Lurll(-<lanfsi mntt: Gn as rrlarrd to rrsid'nrr'and s'mulinr; hisunrlrs. I. ll'luir rnalrs J h.ltl Cancer Inst 1962. :8:9i -ItNJ1. !l8)' HALXUS.L W: Twbvltrt ):'E. LnX-aannrt wnrraGr, • a. ru'I:rar11 uu rrsidr+scr and'' snuskrrnv hutnnrs. 10. 11'hito fk-mairs. J\.rrl C.+ru rr lnsr 1964. 32:ri03'.838. (19) E.Nctttov JE. Cirrcrr .ud rowl'rnurraiirl'.nuunK,wr,nWr \frorinuns: CGnt+nr 1978: 12:'1943.-1951t. (217) Ar2rt.Arat O: GAas'tNxEL L,)HAsr.to.%;n EC. CDswrnr;n irr Lnunr hr.rl epithrliunrt in rcLrron to rrg:mrnr snurLin%-. 1950-1977. N. Ens;f J Mrd 1!A79[ J110:SR1-31t6'. (27) Hy.rruo!cn EC. Se.r.tcorr (J• Si.utte.r: HL A.la•.'t<n r>q> r„rre r t> - arnrr' svnolinR .. andl dr:rtlr. e»rs. AnnK1'Arad Sil~n. lw7b. !au 173-490: