Tobacco Documents Online

.. 1''. ~.+ ? . L " TRIs tnctori5l tr;iy be protscted by, coDYri0 Int. J Cancer: 40. 604-GQ9 (1987) /gW (~ttle 17 U:S. COd4 = 1987 Alan:R. Liss. Inc. LUNG CANCER AMONG CHINESE WOMEN --~ PuDM'Cit Wn 01'}tN 1n1rn1'tK1n11 Un,O-iQt,^tt CMCf. P,IO/.c!{i0n of I'UA.On IOif~nftiOnfK COntrt If C1nct, Yu-Tang GAOi. VWilljam I. BLOT'. Wei'ZHENG1, Abby G. ERSHOwz„Cheng Wen Hsu3'. LytmI. Ll:vtN=, Rong ZHAt.G' and Joseph F. Ftz~t.Hevt. 1R.'- tShanghai Cancer Instirure. Shanghai. People's Republic ojChina: =Nattonal Cancer Institure, Berhesda, MD 20892, US.1; and 3Sluinghai Chest'Hospiial. Shanghai. People's 1?epublic ojCluna. A case-controt study involving intor+riYws with &72 female k+ns cancer patients and 735 poputatior*bas.d contsols was conductad to investigate the hish rates of hft cancer, norablr adenocarcinoma, among women In Shanthai ~ Cigarette smok• ing was a strong risk factor, but accmunted for only about one- fourth of, all newly diagnosed cases of lung cancer. Most pa- tients, particularly with adenocarclnoma, were IHabng non- smokers. The risks of lung cancer were higher among wanen reporting tuberculosis and other preixistin= lung diseases. Hormonal factors were suggested by an Mcr.wd risk associ- at.d with late menopause and by a gradiint In Me r(sk of adenocarcinwma with decreasing menstrual'''cYcle ksngth. with a 3-fold'exceu among woman who had shorter qrNes: Perhaps most intriguing were associations found between lung cancer and maasures of sxposure to cooking oil vaport. Rltks In- creased with the numbers of m.als cooked by either stir frying, deep frying or tioitinj;,widi the frequency of smokiness during cooking; and with the frequency of eye i'ritstion dur- in= cooking. Use of rapess.d oil, whose volatiles following hi:h-cemperature cooking may be rnuta=enic, was also re- ported more often by the cancer patients. The fkndin=s thus confirm that factors other than smoking are r.sponsible for the high risk of lung cancer among CJwnese women and pro vide clues for further research, {nchidn'g the assessment of cooking practice& The age-adjusted annual lung cancer incidence rate among females in Shan¢hai during the 1970's and 1980's has been about 20 per 100.000 population, one of'the highest rates im China and'in the world (National CancerControl!Office, 1980; Gao. 1982: Waterhouse et al., 1982: Shanghai Cancer Regis- try. 1983)1 Elevated lung cancer death tates have also been observed among Chinese women in other parts of the world, including Hong Kong (Kung er al:- 1984: Koo a' al,, 1985),. Singapore (MacLennan ec al.. 1977) and the United States (Fraumeni' and' Mason. 1974: Hinds er al,, 1981; Green and Brophy. 1982). The high rates of this cancer are surprising since few Chinese women smoke cigarettes (Deng and! Gao. 1985). Funhermore. in Shanghai (Zheng and Gao. 1986) and elsewhere, hospital records have indicated that most of the lung tumors are adenocarcinomas, a type of lung cancer less strongly relbted to smoking (Lubin and 81ot„ 1984). To explore reasons for the high rates of lung car,cer among women in Shanghai, the Shanghai Cancer Institute, in collaboration with the US National Cancer Institute, carried out a population- based case-control investigation. Herein we report the results of this study: quantifying the role of smoking and evaluating a variety of suspected risk factors. MATERIAL AND METHODS All newly diagnosed cases of primary lung cancer (9th Revision ICD 162) during the 2-year period February I984- Febtuary 1986 among female residents of urban Shanghai aged 35-69 years were identified by a rapid' repotting system for lung cancer estublished~ for this study. The system was built upon the existing Shanghai' Cancer Registry, the oidest in China (Gao, 1982). Trained staff contacted medical facilities in Shanghai to ascertain new cases, so that interviews could be rapidly scheduled (typically within 2 weeks of dvagrasi's). The staff reviewed relevant medical records, abstracting data on the basis of diagnosis. histologic type. and the site of'the tumor within the lung. Two senior pathologists and 4 senior clinicians were appointed to review the diagnostic information from,all cases collected in the study. including X-ray films. cytologic and histologic slides. Female controls were randomly selected within 5-year age strata from the general population of the Shanghai urban area. The number and age distribution of the controls were deter- mined in advance from the number and age distribution of lung cancer cases reported to the Shanghai Cancer Registry during the period 1980-81. The selection procedure invoived randomly choosing' a neighborhood committee from among the approximately 1.300 committees in urbamShanghai, then randomly choosing a household group within the comminee and ascertaining from existing rosters the names of all females in the appropriate age range. Among these persons. 2 were randomly selected. If the first was absent during the period of study or could not be interviewed, the second was accepted as a control. Tables of random numbers were used in the random sampling. The cases and controls were interviewed by trained inter- viewers. A structured questionnaire was used to obtain infor- mation on demographic charaeteristics. exposure to tobacco- dietary and cooking practices. medical conditions. family his- tory of lung cancer, menstrual artd'reproductive factors. job history and other variables. All completed questionnaires and medical abstracts were checked by a field supervisor. and'the information was then abstracted on coding sheets for key- punching and cotnputeri.zation in the United States. Statisticalianalyses of the collected data were based on mul- tivariate techniques for case-control data (Breslow and Day;. 1980). Logistic regression analyses were used to estimate summary relative risks (RR) of lung cancer associated with various facton, after adjusting for age (,<55; 55-59, 60-64. 65-69), smoking (non-smoker; smoked less than 20 years or less than 10 cigarettes/day; smoked 20 or more years and 10- 19 cigarettes/day; smoked 20 or more years and 20 or more cigarettes/day). education (no formal education, primary school, secondary school and higher) and other variables. and to evaluate statistical significance. Poptdation' attributable risk (PAR) estimates for smoking. adjttsted for age, were also derived (Whittemore, 1983). RESULTs A total of 765 lung cancer patients were identified during the 2-year period and interviews conducted with 672 (88%). We excluded the 93 patients who died, including 38 ascer- tained by death certif~ate only: There were no patients who refused interview. Forty-three percent of the cases were diag- nosed by tissue biopsy. 3896 by cytology, and 1'996 by repeated Addnessre~n ot nequests from China to Dr, Geo and frorn oMer coun- aies to Dr. Blor. Received: March 30~ 1987 and in revised form May 29. 1987.

l LUNG CANCER IN CHINESE WOMEN 605 TABt.E1- RELATIVE R1SKS OF LU1:G.CANCER ASSOC'IATEDwTTH.KUMBt7t OF CIGARETTES SMOKED PER'.DMAND DLRAT7O+. OF: SMOKING " tArnwn otaeotunr Vumflcrof'cipnrncs smNtd er da <.30 Yurs > 30.Tur, ~ . p Cue~ Camroe RR' 9!Sf CI Casc• Conrrol. RR 95:Q; < 10 36 45 1.4 0.9-2.2 34 29 2.4 1.4-1 1' 10-i9 19 17' 2.6 1.2-5.7 56 33 3.2 2.0-51 20+ 13 2 8.9 2.0-40:2 78 10 14 1 7 1~.28 0 'Ad)usred fa aEc rd ehbcawn Nansmoicen ur wed :asdr. rekrencs rroap.. TiAaI.E 11 -RELATIVE RISKS OF LUNG CANCF'll AMOwG IqN-SMOKING M'OME.tiASSOCIATED WTTHYEARS LIVED wITH'A.SMOKINGHUSBAND YeanIrvd .nrA.mot,rq l.se.W Cua Cavds RR' "7F C7 <20 57' 99 1.0 - 20-29 63 93 1'.I 0;7-1'.8 30-39 78 107 1.3 0;8-2:1 ;t 40 48 76 1.7 1.0-2.9 ' hcymsted 4a "c and edWCauoe X-ray films, Among the 542 interviewed cases pathologically or eytological}v diagnosed. adenocarcinoma was the predomi, nant cell, type accounting for 61 % of'all cases. 22 % were squamous carcinomas. 6% were oat-cell (or small-cell undif- ferentiated) cancers. and' 114E were miztures and other cell types. A total of 735 controls were interviewed. Among these 71 (9.7 %) were "second" controls, chosen, mainly because the first selected'control1 had moved from the Shanghai urban area or was found to be outside the eligible age range. The distri- butions by age, education and' maritali status were generally similar between,cases and controls. More controls than cases (32 S( vs., 204E ). however. were in the oldest age group of 65- 69 years. but all subsequent analyses were age-adjusted. Cigarette smoking Cigarette srnoking was associated with a signnificantly in- creased risk of lung cancer. even though only 35% of the lung ,ancer patients (compared to 18 % of the controls) had ever smoked. There was a 33-fold excess risk (95% CI = 2.5- 4.2) of lung cancer among smokers, but risks were higher for squamous-cell carcinoma (RR = 7:2, 95 % CI' = 4.6-11.1) and oat-cell cancer (RR = 7.2, 95% Cx - 3.6-17.0) than for adenocarcinorna (RR = 1.5, 95 % CI = l.0-2.1). The RR for all lung cancers combined tended to rise with increasing num- bers of cigarettes smoked per day'and with increasing duration of smoking (Table I). The excess reached 14-fold for females who srrtoked 20 or more cigarettes per day for more than 30 Yws. Si;milar trendt existed for adenocareiaonu and' for squa- motu/oat-cell cancers, but the magnitude of the increase was coauidenbly greater for the latter (not shown). We calculated population-atttibvubk rislt (PAR) estimates for smoking in each age group. The PAR rose with age, from 8% to 27% to 28% to 40% at'ages <55, 55-59, 60-64, and 65-69. respectively, primarily because the prc valence of sawking rose with age. In total, we estimated that 2496 of all female lung cancers in Shanghai were due to smoking.. Passivt smok'ing No significant increase in risk was observed for overall exposure to environmental tobacco smoke during childhood (RR = 1.1. 95% Cl - 0.7-1.7) or adult life (RR - 0.9. 95% CI = 0.6-1.4). For these calculations, exposure was said to Occur if the subject had ever lived with a smoker. When exposure was defined in tertns of husband's turtokittg, bow= ever, ltmg cancer risks among non-smoking women tended to ibcrease with the number of years a woman lived with a husband who smoked: the RR reaching 1.7 among those with 40 or more compared to less than 20 years' exposure (Table 11). The risk in this heavily exposed group was even higher (RR = 2.9„ 95 %' CI = 1.0-8.9) for squamous- and oat-cell carcinoma.. Previous lung diseases Since lung cancer in its early stages may be confused with other lung diseases, we excluded non.rrtalignant lung diseases occurring withimthe 3 years preceding interview in evaluating the effect of prior lung disease upon lung cancer risk. Table III shows that previous tuberculosis. pneumonia and emphy- sema were significantly' associated with lung cancer risk even after adjusting for smoking. Although, some individuals re- ported having 2 or more of these diseases. the excesses for each persisted when those with, multiple conditions were ex- cluded. Further analysis (not shown) indicated that' the effectt of tuberculosis was not related to the use of isoniazid or streptomycin. While tuberculosis and pneumonia were related~ to, both squamousioat-cell carcinoma and adenocarcinoma of the lung. emphysema and chronic bronchitis were associated~ only with the squamous- and oat-cell types. Cooking pranices Soybean and rapeseed oils were the oils used mosr often for cooking in Shanghai, with over 95% of women reporting the use of both products. Rapeseed oil, however, was reported as the most often used cooking oil by 52 % of the cases compared to 45 % of the controls. The overall increase in risk associated' with rapeseed compared to soybean as the most often used oil was 1.4 (95% CI = 1.1-1.8). Table IV'shows that the excess lung cancer risk associated with use of rapeseed oil existed at each level of reported frequency of eye irritation when eook- ing, a subjective variable representing sever'ity of exposure too cooking vapors. The calculations for this Table excluded~ the few women who never cooked, and employed as the reference group women who most often used soybean' oil'1 but never or rarely reported eye itriiation. Table IV also shows that risks of lung cancer were independently related to eye imtation, with the highest risks (RR - 2.8, 95% CI - 1.g4.3) among those using rapeseed oil and' fr+squendy reporting irritation. The patterns were similar for squamous/oat-cell cancer and ~ adenocarcinoma. We also observed, after adjusting for eye irritation, a 60% higher risk for lung eancer among womeno who reponed considerable or somewhat smoky cottditions ln their homes when cooking. another rough measure of exposure to cooking vapors and to house ventilation (Table V). in, addition, the risk ratios increased with the number of differentt dishes per week prepated by stir frying. deep frying, or boiling (Table VI). In contrast. no significant case/control differences were associated with the type of fuel used for cooking- The RR and 95% CI associated with coal. gas and wood as the usual, fuels were 0.9 (0.7-1~.3). 1.1 (0.7-1.5). and'1.0 (0.6- 1',.8) respectively. There was no trend in risk with increasing years of use of coa1L the most cominon cooking fuel in Shang-

606 GAO ET AL. TAiLE nI.- RELATIVE RISKS OF LUNG CANCER.ASSOCIATED WITH PREVIOUS LUNG DISEASES Putlunl d/seasn . Carrols All pus RR~ 9S'R Ct .. ~v aa<eH RR' %T Ct '"° orclroma RR' 4S4 None 554 418 1.0 80 1.0 - 229 1.0 - Tuberculosis 61 80: 1.7 1.1-2.4 20 2.0 1.1-3.7 42 1.6 1.0-2.5 Pneumonh 35 65 1.9 1.2-3.0 16 1.8 0.9-3.8 26 1.5 0.9-2.7 Emphysema 18 37 2.0 1.0-3.7. 19 4.5 2.0-10.3 6 0.7 0.3-2.0 Chronic bronchitis 86 112 1.2 0.8-1.77 35 1.4 0.8-2.5 33 0.8 0.5-1.3 wit110W emr)hvSema Others 30 30 1.3 0.7-2.2 8 1.7 0.7-4.2 13 1.0 0.5-2.0 1 All nslu ad)uutd foe .ge. Rduanon atd >n"AS and Maove so wamrn wph w reponed pux lunR dueau. hrwes wuA ewse IAan oar Np 14 durase am included w acii lung dusase cregory. TAiLE rV - RELATIVE RiSKS OF LUNG CANCER ASSOCIATED WITH COOKJNG OIL USED M06T OFTEN AND FREOUENCY OF EYE nIRJTATION WHEN COOKING TAiLE V - RELATIVE RlSKS OF LUNG CANCER ASSOCIATED wtTH~ FREQUENCY OF EYE IRRITATION AND HOUSE SMOKINESS WHEV COOKING Eye mw.oon when cnotie Oil Wd Miss otw- Cars Cooo4 RRI 9S* CI Eye lrrrrso.. Mwse .nroklKu Ca.n Canols~ RR' 954 Cl I g Never or rarely Soybean 140 214 1.0 - Neverharely None/slight 244 380 1.0 - Rapeseed 145 193 1.2 0.9-1.7. Somewhat/ 55 55 1.6 1.0-2.5 Sometimes Soybean 70 72 1.5 1.0-2.3 considerable Rapeseed 87 63 2.0 1.3-3.0 Sometitnest' hitme/ilight 212 200 1.6 1.2-2.1 Frequently Soybean 59 56 1.4 0.9-2.3' frequently SomewhaU ' 109 60 2.6 1.8-3 7 Rapeseed 90 50 2.9 1.8-4.3 considerable . Total Soybean 269 342 1.0 - ' Rapeseed 322 306 1.4 1.1-1.8 Adpuated for afe..d"we a.d unolunR. IAdjyaaed (ar.R .Oucluom aed makwa. TAaLE % 7- RELATIVE RISKS OF LUNGCAYCER ASSOCIATED WITH NUMBER OF DIFFERENT DISHES PER WEEK.rRE1ARED BY DIFFERE!tT METNODSQF COOKING NYIIIisR Jf QlSilef per .xeh Caca Cartrol, RR' 95% CI Sdir fning 42 0 336 408 1.0 - 20-24 1,98 211 1.2 0.9-1.5 25-29 48 47 1.2 0.8-1.9 ~t 30 34 15 2.6 1.3-5.0 Dtep f-ing 0 502 594 1.0 - I 85 68 1.5 1.0-2.1 2 2li 15 1.6 0.8-3.2 > 3 8 4 1.9 0,5-6.8 Boiling <3 96 124 1.0 - 4-7 390 483 1.0 0.7-1.3 8-11 63 40 1.8 1.t-3.0 ;a 12 67 33 2.2 1.3-3.7 'Adrrwed ioe ape..duano. +.d e.mkft. hai. Exposures to coW: an ©ther fuel fitntes were generally associated only with cooking. since nearly aU homes inShang- hai were unheated. Diet The women were asked about their usual frequency of con- sumption during adulthood of 32 commonly eaten foods, in- cliiding the major contributors of vitamin A. Using Chinese food composition tables to estimate the rttinol and carotene content of each food and applying these escfatates to its fre- quency of intake. an index of vttamtn A consumption in ret- inol-equivalent units was eonstructed. The risks for lung cancer R:nded to be lower among Blose with low values of this index TAiLE V71- RELATIVE RISKS OF LUNG CANCER ASSOCIATED WTTH DIETARY'INTAKE OF VITAMIN A V..mm Qu.ruk k.el d consumpan~ eiPOlYIL vanaple I IH11A/ 0 nt 1%, Vitamin A 1.0 0.6 0.8 0.5 index (0.5-0M (0j6-1.1) (0.4-0.7) RetinDl-rich 1.0 0.9 l'>0 0.9 foods (0.7-1.3) (0.:7-1.3) (0.7-1.2) Caraene-rich 1.0 0.6 0.5 0.5 foods (0.5-0:8) (0:4-0:7) (0.3-0.6) IRak edativa to lufheu quamk of eonwmpqwn aud adjusted for ap. .ducuan aed smoWng. 95S C(m R.r<ndscses. (Table VII). This association was accounted for mainly by a lower risk among those with a neduced' consumption of earo- tene-rich foods (the dominant source being dark green vegeta- bles). No effect on risk was found for consumptton of the nttinol-rich foods (mainly fish. eggs and liver). The paaerlt5 were generally similar for squamous/oattelt cancer and ade- tsocarcinoma. and for smokers and non-srtakers. Mensnrool and rrproductive fiteTors The risks of lung cancer were higher among women with shorter metsstrud' cyck kngths (Table WI). The association was primarily seen for adenocarcinoma. which showed a strong dose-response relationship. Among women aged 55 years and over with a natural menopatse. tfie risk of adenocarcinoma tended to increase with the total number of inenstnnl cycles aver their lifetime. Some increased risk of adenocarcinoma was seett when natural menopause occurred at age 30 or later (RR - 1.3. 95Ae CI = 0.9-1.7, after adjusting for mettstrual cycle length). No associations were seen with age at menarehe. age at first pregnsn<y or parity.

LUNG CANCER IN CHINESE wOMEN TARt.E Y711-RELATR'E RlSKSOF LUNG CANCER ASSOCIATEDWIT11 LENGTH OFMEySTRUAL CYCLE 607 ceartA ul mnsrrual ccck Idn,, AII ~. csxs Cnrrols RR' 4S'4 CI ~f RR' 951 CI~ ~~I RR' 9~V CI'~ > 33 43' 60 1.0 - 12 1.0 - 18 1.0 - 30-33 272 327 1.6 1.0-2.6 62 0.9 0.4-2.0 124 1.9 1.0-3.5 26-29 241 268 1.6 1.0-2.7 51 0.8 0.4-1.9 127 2.1 1.1-3.9 < 26 98 78 2.2 1'.3-3.7 23 1.6 0.7-3.9 54 2.9 IL5-5.7 'AdPuned r« aae. .euc>Mroe. smak" .nd ntipsi" a aiasrnuswn. TABU IX - RELATIVE RtSKS OF LUNG CANCER FOR WOMEN EVER EMPLOYED IN M AlOIt OCCUPATIONAL GROUPS ottypn= pnpn . , Ca+es Ca.rols RR= 9~..L CI 1-Il Professiosvls and tnchnicians: leaders of 113 116 1.0 0.7- L4 QI-IV state organizxions. party and mass orp- niiations and enterprise units Office and rslated personnel: sales workers 75 96 0.7 0.5-1.0 v' Service workers 1159 160 1.0 0.8-1.4. VI Agricvltural.,forestry. arutnal!husbandry 24 21 1.1 0 6-?.:1 YII-lX and fishery workers Production. tratuportation and other 436 471 1. C 0.9-1.4 0 related workers Never worked 61 75 I_ I 0.7-1.6 'Women emplarod in more tAan one occvp.noiul!caleEon in` included in each pit(oM1 1111vtIKl1 ltle)' wo/ke4 The LVi1inE KIKTe wL bKCO On Me symem rsed in the . N6: Popula~wn Census of Me People's Reput,irc of Chm (Populatan Ceews Office. I48i I: =Adjuwed (o6 age. educauwn and +molung Oecuparior; bined (81ot and Fraumeni. 1986). an overall 30% excess of Most women reported working outside the hotne, but ease/ lung cancer (RR= 1.3. 9596 C1 = 1'.1-1.5) was found among to major occupational categories non-smoking women tnarried to smokers. with the RR reach- cont s accordin l diff e c g ro er n e ations were associated ing 1•7 among those most heavily exposed: No rnajar occu were sma11 (Table IX) p . with increased risk of lung cancer. A decreased risk. however, was observed for women ever employed in the cotton textile indtutry, the largest employer of women in Shanghai. There was a slight increase in the relative risk of lung cancer among women ever employed as cooks (RR = 1.2. 95% CI = 0.6- 2.1), but few worked longer than 20 years. Fmnils hisron• Although the causal significance of the relation of'prior lung disease to lung cancer remains to be clarified. the high preva- lenee of'previous pulmonary infections may have contributed in part to the high incidence of lung cancer among Shanghai' women. Earlier in this century; non-malignant lung disease was one of the leading causes of death in China (Kan: 1981). With the advent of antibiotics and improved living conditions, the incidence and mortality of chronic lung diseases. panicu• The cancer patients reported about the same frcquency of larly tubercuiosis, declined. Nevertheless, a substantial por- lung cancer in their mothers (1.09fc ) and fathers (1.7 Yi ) as dte tion (38 9E ) of the women with lung cancer in this study reported controls (1.0% and 1.596, respectively). The RR, adjusted for prior lung disease. including 12% who were long-term survi- age. education and smoking, associated with having a parent vors of tuberculosis. whereas significantly lower percentages with lung cancer was 1.1 (95% CI = 0.6-2.3). More stbs of controls reported these diseases. To some extent it is posst- were reported to have lung cancer, but the numbers affected ble that recall or ascertainmenr bias may contribute to the were small (6 cases. 3 eontrols: RR - 3.0. 95% CI - 0.7- associations observed with prior lung diseases. The elevated 12.5): Only one child, of a cantrol, had lung eancer, risk of lung cancer following tuberculosis, however, is consis- tent with recent studies in other countries. and is not explained ~~ by cigarette smoking or treatment with isottiazid. apulrttorury carcinogen in laboratory animals (Howe er al.. 1979: Hinds sr The high incidence of lung cancer among women in Shang- al.. 1982: Bakris er al'.. 1983). hai, together with the low prevaknce of srnoking in the general F,mphysenu was also' signifMcantly related to lung cancer. Population. led us to consider a number of possible etiologic aflter adjustment for smoking habits, with the excess limited to factors. .N+hik eiprette smoking was aa inVortant cattse of~- squamous- and oat-cell carcinomas. This finding adds to the ~bing cancer. showinE a cleardote-retponse tread, the majority evidence that chronic obstructive pulmonary disease enhances of lung ttunors, particularlr adenacircinonws, occurred among the risk of lung cancer (Skil)rud er a/.. 1q86). even when "on-smokers. controlling for smoking practices. Also noteworthy is the ele- Lnvironmental tobacrn smoke nt:y accottnt for sotne. but vated risk associated with prior pneumonia. especially since Probably few, of the cancers among non-smokers. sina tiyere an usociation with lung adenocarcmoma has previously been was little or no association with ever having lived with a>eported among, women in Los Angeles (Wu er al:. 1985). smoker. Among non-smokin women rrurried to smokers. While pneumonia typically occurred during adulthood in our liowever. there was an upward trend in risk associated with smdy. the finding in Los Angeles primarily'concerned child- ~neuing years of exposure. This latter finding is eonsistetn bood tnfection: n IvIth early~zen studies euatin~ tNvtksmokinn i~ n~un~~ conicate sisteet w~findings from B Pass smoking were corn- for cookin8 gcancer risk

608 GAO ET A11.. Hong Kong (Koo er al., 1983). The risks of lung cancer among Shanghai women increased, however, with vanous measures of exposure to cooking oil vapors. These included the number of different dishes prepared per week by either stir frying, deep frying. or boiling;, the frequency of eye irritation when cooking: and the smokiness of the house when cooking. ln Chinese wok cooking. regardless of the method used„ oil is usually poured into a wok and heated to high temperatures before meat or vegetables are added. Even boiling may entail some exposure to cooking oil vapors, since oil is often added to the water before heating. ConseQuentty, the living quarters may become smoky during cooking, wRh oppotmnity for exposure to inhalabie cooking oil vapors. The plausibility of the hypothesis that lung cancer may be related to cooking oil vapors. partucularly from rapeseed oils used in Shanghai. is enhanced by recent experimental ~ investi - gations. Ln one study the mutagenicity of products from cook- ing oil was assayed by the Ames test (Qu er al,, 1986). The extracts of condensed volatiles of rapeseed oil, refined rape- seed oil, and soybean oil heated at about 270°C were all positive in tester strain TA98 activated with S9. The mutage- nicity of the extract from rapeseed oil volatiles was stronger than that from soybean oil volatiles. There was no evidence of mutagenicity in the oils themselves, either heated or unheated. In another siudy: the extracts of condensed volatiles of rape- seed oil enhanced the yield of micronuclei in polychromatic erythrocytes of the bone marrow of mice, with a clear dose- response relationship. reflecting datnage of'chrvmosomes and cell genotoxicity by rapeseed oil volatiles (Chen. 1987). Al- though these tests often correlate with carcinogenic potential, no bioassay studies have yet been carried out, to our knowl- edge. If the effect of rapeseed oil smoke on lung cancer incidence is real. the problem is of great importance to popu- lations of eastern central China and other areas of the world where the oil is often used for cooking. Chinese rapeseed oil, which, is pressed from seeds of Brassica campesrris„ contains about 50% erucic acid (Chinese Academy of Medical Sci, ences. 1981). in conLrast to rapeseed'oil with <2% erucic acid (Canbra oil). which was rectntiy approved for sale in the United States (Federal Register. 1985): Several studies have shown that the risk of lung cancer is elevated by a low intake of foods containing vitamin A, partic- ularly as its precursor beta-carotene (Colditz et a(., 1987). Although reported mainly in Western countries, this associa- tion has also been noted among Chinese women in Singapore (MacLennan eral:, 1977): However, we found no evidence of' a protective effect among women in Shanghai, where intake of fresh. carotene-rich, dark green vegetables is high by worid standards. In fact. a positive association was observed between carotene intake and lung cattcer, risk in females (in contrast to no association in males): We have no ready, explanation for this unusual finding in females, which was observed also in a case-control study of lung cancen in, Hawaii (Hinds et al., 1984). However. the protective effect of carotene-rich, foods was mainly confined to current smokers in one large-scale study (Ziegler tr a/., 1986), and thus the effect may be less evident in Shanghai where few women smoke. A clue to hormonal factors was suggested by an association between menstrual cycle kngth and1ung adenocarcinoma. A 3-fold difference in adenocarcinotna risk was found among Shanghai women reporting short (<26 days) compared to long (>33 days) mensrrual cycles, with only a weak trend for squamous- and oat-cell cancers. A relation of short menstrual cycles to breast cancer risk has been suggested in data from Sweden (Ulsson et al., 1983) and, to a lesser extent, the United' States (Sherman et al.. 1982). Some increases in lung adeno- carcinoma risk were also associated with late menopause and with a high estimated total number of rnenstrual! cycies among women aged 55 and over having a natural menopause. A role of hormonal factors is also suggested by the observation that among non-smokers adenocarcinotna affecu proportionately more females than males (Lubin and Blot. 1984), and by the findings of' estrogen and progesterone receptors in some lung adenocarcinomas of women (Chaudhuri et a1., 1982). We discovered no relation to oral contraceptives or replacement estrogen therapy: but use of these compounds among worrten in the study group was rare in Shanghai. It seems unlikely that the menstruallpatterns of Chinese women contribute greatly to their high lung cancer risk, but the internal consistency of the trends suggests that future studies of lung cancer in China and elsewhere should examine endocrine hypotheses in more detail. This large population-based case-control study of lung can- cer in urban Shanghai has confirrned'tltat cigarette smoking is a strong risk factor among Chinese women, but only accountss for about one-fourth of all, newly diagnosed cases. Causes of the remainder are utxlear, but occupational factors did not appear to be important. nor did familial tendency to lung cancer. Our data suggest, however, that prior lung diseascs,. hormonali factors, and cooking practices may be involved.. Most provocative are the associations with cooking oil vola- tiles, and further investigations are needed to evaluate their contribution to the high lung cancer rates among Chinese women in various parts of the world. ACKNOWLEDGEMENTS This work was supported in pan by the NCI. We thank Dr. BJ. Stone. Dr. Slm Xiao-ou and Ms. R. Parsons for comput- ing support. Ms. C. Chen for computational and~ translation assistance, and Dr. B. Henderson for advice and helpful suggestions. tt;tcFFAENCFs BAiuus. G.. MuLOnnns, G:P., KnRCxac. It.. Emnru. E., Ro. J'., and YboN. B.. Pulmonary scar carcinoma. Cawer. 32. 493-497 (1983). Bwr. W.J.. and FRAUwem. J.F.. JR.. Passive stnotinj and lung aancer.. J. nar. Cancer Gur.. Tf, 993-1000 (1986): BRECLaw:,N.E.. and DAY. N,E.. SuvisrioaJ xeelsodr in cancer research.. The anahsis of case-aonrrol sr.idies. L4RC Seiewrific P>iNicwion, 31. pp. 192-246. IARC. Lyon (1980), CNkutHnnu; P.K.. TnowAs. P.A.. WkuRFa. M.J.. BatELE. H.A.. GurrA. T:D.. and BeArnE. C.W.. Steroid >ecepors in human lung c.rcer cyto- sols. Cawer Jtrt.. 16, 327-332 (19B2). CNEN: T.D.. Micrawtclau test of condensed votatiks of rapeseed oil. Tirwwr, 1987 (in pras). CHtNESE ACADEMY OfMEDICAL SCIENCFS . Food cowpositlOq A7Mes. POo- p,ie's Health Publiihin` Co.. Beijing (1981). Cot.txrz, G.A.. StAwrsea„ MJ., and WnLES, W.C., Diet and lung C2 tWrtWls. Arch. uUrrn. caattr[ a rtWeY+of tht evidta i~ Med, 147, 157-160 (1991). DENG. J., aed Gn.o, Y.T.. Prevakm of smoking arrnn~ 110.0f)0 adulh residents in ShanBhai urban uer. C7lrtsex J: pnetir: Med:. tf, 271-274 (1985). FeDERAL REatsrt>:. Vol. !1. No. I8. US Governmeen Prinrina Office. Washin8ton, DC (1983). FRAUStfnt, J.F.. and MASo++. T.JS. Crmeer mona)iry, uffwna Chinese Amoticans. 1930-19'69. J. not: Cawcer Inst:. 52, 659-b63 (1974). GAo;, Y.T.., Cancer incidence in Shanahai' during 1973-77. Naa: Cancer hsu. Mtusalr.. 62, 43-46 (1982): GREEn, J.P.. and BRO.wY. P.. Carcinortu of the lung in non-smoa'in` Chinese women. West. J. Med.„t><, 291,294 (1982).

LUNG CANCER IN'CHINESE WOMEN 609 p1.r~ r/ W. CoNES. H,I..~nK A,~ .r ~~~Tubercnlis and l778 ancer ru1: in tnn-snwkini sp. . Ia8_ I W.m. M W:., Kot.o~rtt.. L.N.. HANtctx. LH.. 1ad Lt6. I.. Dietary .aantn A. caroaene. vwmin C and risk of lunj cower in Hawaii: Amer. ! Lprdr'"'td.'. 119. 227-237, (1964). Ht.m. M' W. STeMMe1tMA.. G.N.. YANG. H.Y:. KaoNn. L-N., aad Lt i. 1. DdfernKes m hmg cower risk from sawkiat atno.~ Japanese. (~,nc,e. and Hati+.aiian women in Hawaii. 6u., J. Ca~nrr, 37, 297-30Q 1191t1 t Wwt. G. 1.Ut.vsAV. 1.. CoPPoCK. E.. asd MnLE7t. A.B., hooisid' e.p..urc m relation ro cower iacidence and mortal' ~ty m a odson of wtenvknss p.uems. Mu. J. Ep(deedd.. f, 3QS-312 (1979): K.•. G. Trbercylosis asd ils conuol ia Beijing. OIiR.,wed. J., !4, 6!S- 6a11t1981i K, ... L C. ikt. J.H. aad L¢e. N.. Anana1ysu of aome risk factors for h,y ca.m at Hoej Kon;. hu. J. Cawcn, 3S, 149-1SS (1985): K..,: L C. Lu.. N.. and Ho: )'.. Do cookSnB fuels poae a riak for >MV1 an<n' A cau-caNtol sad} of women iu Hong Kong. EcoJ. Dis.. 2, :S~-2eSt1983t Ki K,. 1. Sn K.. and LASt. T.. Lung cower in HM Koe~ Chitkse: ..Wit,, and hts,olopctypes 1973-1982. Brii: J. Cancer, 6Y, 381-388 tw/N, Lt u•. I. rtdi BLoT. W.1.. Assessment of lung cower risk facton by Ynx*.la cateFor.. J. ear, Cancer Jdst..,73, 383-398 (1984). Mv 1.U...%. R. D, CorrA.1.. DAY. N.E.. LAw; C.H.. NG: Y:K., ard S--f-e.TAM, K.. Rtsk factors for lun; cancer in Singapore Chinese. a prywlsuan .nh hiEh female incidence rates. Inr. J: Cmtcer. 20, 834- IEUr14771 K.n-u C+.csN Co.-rnoL OFFlCE. Arlas of Cancer Afonnlirn, in dte "r , RrprAliC of Chtna. China Map Press. Beijing (1980); N+t t:MIt. Ctsus OFFxE. 10 Percent Sampltn4 Tabularion 08 the 1982 Population Cettsus of the People's Republic of China. China Sutisttca), PublisM»t8 House. Beijtng (1985),, OLSsoN: H.. LA!+Dn:-0Lssor:. M.. and Gn.i.aFxa. B.. Retrospective as- aessment of inensuval cycle length in patients with bre:sr cancer. in paientc with bsnigo bnrast disease. and in women witlsout:breast dutase. !: twt. Concer /nu.. 70, 17-21 (1983 ): Qu. Y.H., Xu, G:X.. HUANG. F., FATw- LC.. and GAO, Y:L. AeAmes sest on the Ey-produces of the heating of oookinp oils. T.onor. 4, 58-60 (1986): SNAN6lIA] CANCv . R6GiSfRY. AMIYOI Reporu 011 CatlCer J71(7deNfe and WOrfdlh': DxR fn. S10/1gIIN (1rb0/( Area, SlYnjhii ~ CGOttr Jnfiitu4e. sao8haf (19d3). SNptMAN, B.M., WAS.uce. R.B.. and BeAN. l.A.. CyCIK ovarian func- tion and brtat cancer. Cancer Res. (Sarppl.l; 42, 3286-3288 (1982). SaXSLLatro. D.M.. OFPoRD:,K.P.. and MtLLER. R.D.. Higher risk of lung cancer in chronic obstrssetivepu lmonary disease: a prospecuve matched controlled smdy. Ahn. inrern: Aled.,. 105, 503-507'(198d). WArFx+ovss, l.. Mtmt. C:, S+uNMUCAnA'n+ASa. K.. and Powui. 1.. Q'ascer iwcidena fr Jfve caiuiaenrs, Vol: 1V, IAAC. Lyon (1982) , WtttrrEwonE, A.S.. Eairnaun~ amibunbk risks from case-control sad- ies. Mter. J.,Epidenuol.. 117, 76-85 (1983). Wv. A.. HENDEtsoN. B.E.. PncF. M.C.. and Yc. M.C.. Smoking and odier risk faeron for lunt cancer in women, J. ear. Cancen Inn.. 74„ 747-731(1983). Zxewo. W.. and GAO. Y.T.. A hapiul-b.sed casc-comrol study on asso- ttiasion ofsquantous cell carcinoma end adeewcarcisartu with stnokin8. Tantor.l.l -7 2O(198b). 7mcttx. R.G., MASON. T.I.. SraHAGeN: A.. HoavEa. R.. ScNoeNmG. I.. GuDt.an•. G.. Vmeo. P.W.. and Fa,.uMaNt. I.F:. la.. Caraenoid iraake, re=etables. and the risk of lung cancer amon8 white men in New lertey. Amer. J. Epidtntiol:, 123, 1080-1093 (1986).