Document DDnKp7Q6700dZkj330j1w6aYB

FILE NAME Brakes BRK DATE 1983 DOC BRK083 DOCUMENT DESCRIPTION Journal Article - Lung Cancer and Occupation in Alameda County - American Journal of Industrial Medicine Ww (% vrin cals American Journal of Industrial Medicine 565-575 1983 Lung Cancer and Occupation in Alameda County A Death Certificate Control Study K.L. Milne BS D.P. Sandler PhD R.B. Everson MD and Steven M. Brown MD A control comparison of lung cancer deaths versus deaths from all other cancers occurring in Alameda County California between 1958 and 1962 was conducted to investigate possible associations between lung cancer and occupation adjusted specific analyses indicated that a history of work in certain industries and occupations was associated with an increased risk of lung cancer Among males a significant positive association was found for glass metal furniture professional and photographic equipment rubber leather and plastic manufacturing water and air transportation auto repair services and construction Specific occupations associated with a significant positive risk for males were electricians aircraft mechanics painters plasterers machine operators construction workers bus and truck drivers and guards and nightwatchmen Among females beauticians had a significantly high risk associ- ated with lung cancer These results were consistent with other studies of lung cancer related to occupation Key words lung cancer epidemiology occupation INTRODUCTION Much attention has been focused on the occupational factors associated with lung cancer Doll and Peto 1981 have attributed 15 of all male and % of all female lung cancer deaths in the US to occupational exposures Several studies have investigated the occurrence of lung cancer among people working in a broad range of industries and occupations DeCoufle and Stanislawcyzk 1977 Dunn and Weir 1968 Enterline and McKiever 1963 Kennaway and Kennaway 1936 Menck and Henderson 1976 Menck et al 1977 Milham 1976 Williams et al 1977 Although they have used different methodologies these studies have observed many of the same associations with certain occupational settings However each has also identi- Epidemiology Branch Biometry and Risk Assessment Program National Institute of Environmental Health Sciences Research Triangle Park NC Address reprint requests to Karen L. Milne Epidemiology Branch N.I.E.H.S Building 101 Room A359 P.O. Box 12233 Research Triangle Park NC 27709 Accepted for publication November 30 1982 '1983 Alan R. Liss Inc. ce - gemmee3 re PO ku ku wy a ae ae pire % 566 Milen t e al fied elevated risks not reported by other investigators Some of these may have escaped notice because there is no apparent hazardous chemical or industrial exposure In addition occupational exposures may vary from one geographic locale to another studies undertaken in a particular area would miss risks associated with industries not located in the study region Most of these studies of occupational factors have relied on proportionate mortality ratios PMRs as indicators of risk associated with employment This method may introduce bias because the population under study may not share the same distribution of mortality as the standard population used to compute the rates and because the proportions for various causes of death or occupations are not independent To broaden the geographic and methodologic range of such studies we under- took a death certificate control study of lung cancer and occupation in Alameda County California We then compared our results with those of other investigators who have reported on occupational associations with lung cancer looking for consistency in the observed patterns of risk METHODS This control study analyzes death certificates from all cancer deaths recorded among residents of Alameda County California between 1958 and 1962. We compared lung cancer deaths cases to deaths from all other malignancies controls to determine whether persons dying from lung cancer were more likely to have been employed in certain occupations than persons dying from other cancers Death certificate causes of death were coded according to the seventh revision of the ICDA World Health Organization 1965 A total of 7,490 cancer deaths including deaths from benign brain tumors were collected One hundred forty deaths occurring under the age of 18 were eliminated because they would not contribute to an analysis of occupational factors Usual occupation and industry of employment were coded using the Bureau of Census Industrial and Occupational Classification System US Bureau of Census 1971 Detailed employment histories were not available The 925 deaths from lung cancer were compared to all other cancer deaths for risk associated with exposure to specific industries and occupations In addition a smaller comparison group of cancers not known to be strongly associated with occupational risk factors Doll and Peto 1981 Higginson 1980 was used to mini- mize potential exposure bias To form this reduced control group we eliminated cancers of pancreas ICDA 157 nasal sinus ICDA 160 kidney ICDA 180 bladder ICDA 181 bone ICDA 196-197 and all hematopoietic cancers ICDA 200-204 This left 4,880 cancers in the second comparison group All analyses were stratified on age at death < 50 years 50-59 years 60-69 years 70 7y9ears and > 80 years and analyzed separately for males and females Combined risk estimates were provided by a computer program that computes the maximum likelihood estimate MLE of the odds ratio using Gart's modification of the extended Cornfield method Gart 1970 Individual odds ratios and 95 confidence intervals are computed for each stratum specific table A correction factor of 0.5 was added to each cell of the 2 ^ 2 contingency tables The result of the test of significance for the summary MLE is indicated by a p value If the odds ratio was 1.0 then persons dying from lung cancer experienced the same risk as persons dying from other cancers for the occupation of interest A risk u | I. ee ee ee rr ST eT oseitat sicearin ne crsoe reread eT vt tag georvap-cs rove THAT AR _aen Lung Cancer and Occupation 567 than 1.0 indicated an excess risk of lung cancer death associated with the greater occupation es RESULTS The study group consisted of 925 lung cancer deaths and 6,420 deaths with other cancers Among the cases 826 death certificates had lung cancer coded as a of death and 99 had lung cancer listed as a significant condition also primary cause of cancer sites for the control group present at death Table I gives the distribution of cancer on the death certificate Numbers reflect the fact that 895 for any mention sites recorded on their death certificate Other and controls had multiple cancer unspecified cancers are not shown Race was categorized as white and white Blacks Indians Chinese Japaand those with Spanish surname composed the white category Because the nneusmeber of white deaths was small 10.8 of all deaths we did not conduct specific analyses Table II shows the distribution of cases and controls by age and sex Eighty percent of the lung cancer cases were male Analyses were conducted separately for males and females because of differing occupational pat- terns In Table III we have compared usual industry of employment for male cases with that for the control groups all cancer deaths and nonoccupationally related TABLE I. Distribution of Cancers Among Controls Seventh revision Site ICDA code No. of cancers Buccal cavity and pharynx Esophagus Stomach Colon Rectum Liver Pancreas Nasal sinus Breast Cervix Uterus Ovary Prostate Testis Kidney Bladder Melanoma Brain Thyroid Bone Lymphoma Hodgkin's disease Multiple myeloma Leukemia 140-148 150 151 153 154 155 157 160 170 171 172 174 177 178 180 181 190 193 194 196-197 200 201 203 204 183 148 504 798 - 331 225 359 23 714 246 54 232 472 17 141 267 41 193 29 32 _ 181 100 108 290 Number of cancers do not equal number of controls since multiple sites included on a death certificate are listed and other and unspecified sites are not shown in this table coi dg RD LT } 568 Milne et al TABLE II Distribution of Lung Cancer Cases and Control Cancers by Sex and Age Age yrs < 50 50-59 60-69 70-79 > 80 Total Cases 73 195 258 175 46 747 Males Controls 312 475 868 990 485 3,130 Females Cases Controls 32 488 37 574 45 815 48 860 16 553 178 3.290 cancer deaths Overall adjusted odds ratios were computed for broad industrial and specific industries when numbers were sufficiently large or an a priori groups association Because of the large number of hypothesis suggested an occupational potential categories we did not pursue speceinfviicraonnamleynsteaslfohrazianrdudstarnideswohrerocecunpuamtbieornss where there was no reason to suspect an employed in these categories were too small five exposed to assess a potential effect Male lung cancer cases had significant risk associated with employment in ten industries Stone clay and glass manufacturing all metal manufacturing profes- sional and photographic equipment manufacturing water transportation and auto at the % level Construction furniture manufactur- repair services were significant and services to ing rubber leather and plastics manufacturing air transportation dwellings were all significant at the % level The associations with certain industries namely agriculture forestry and fish- ing transportation manufacturing specifically shipbuilding and retail trade were significantly lower than expected odds ratio = 0.6 0.7 0.8 respectively We also computed odds ratios for broad job categories and specific occupations the criteria established for analysis by industry Table IV We among males using codes 401-575 particularly elec- observed significant excess risk among craftsmen associations for tricians aircraft mechanics painters and plasterers Significant bus drivers truck drivers construction laborers and guards and machine operators associated with a significant deficit nightwatchmen were also observed Occupations of risk were professional technical and kindred persononceclupaantdiofnaarlmseertstings with the Among employed females there were no specific The risk of of beauty shops which had a significant risk for lung cancer times that of other exception cancer associated with employment as a beautician was seven lcuanngcers Table V gives the odds ratios for women within broad occupation categories Numbers of working women were too sparse to pursue more detailed occupational analysis DISCUSSION prospective Results from our study are compared in Table VI with those of several earlier the studies with which we drew comparisons were ) a studies Among for an average of seven years to assess the study that followed 68,000 employed men and Weir 1968 2 a PMR study risk of developing lung cancer by occupation Dunn TABLE III Risk A Industry and JCS co Agriculture forestr Mining 047-058 Construction 067- Lumber furniture r Sawmills 108 Furniture mfg 1 Stone clay glass m Glass mfg only 11 All metal mfg 130 Blast furnaces an iron steel 139 Machinery mfg 17 Transportation mfg Motor vehicle m Shipbuilding 22 Professional phot mfg 239-267 Food and kindred Textile mfg 309Paper mfg 328-3 Printing publishin Chemical mfg 34 Paint mfg 359 Petroleum coal n Rubber plastic Rubber mfg 3 Leather mfg 3 Transportation 4 Railroads 407 Trucking 417 Water 419 Air 427 Communications Utilities 467-47 Wholesale trade Retail trade 60 Business repair Auto repair ( Serv to dwel Personal serv ( Laundries 779 Entertainment a Professional an Banking real e Public adminis No. = number OR = odds rat US Census Bu Potentially oc significant at significant at a yes - eo ad cere gpreggy, MEER cee Te yt TOE mang T beg ERIPERRO EE a a _ .eerboot aayte istrial priori er of Lung Cancer and Occupation 569 TABLE III Risk Associated With Industry of Employment for Males Industry and ICS codesa Agriculture forestry fishing 017-028 Mining 047-058 Construction 067-077 Lumber furniture mfg 107-118 Sawmills 108 Furniture mfg 118 Stone clay glass mfg 119-138 Glass mfg only 119 All metal mfg 130-169 ; Blast furnaces and other primary iron steel 139,147 Machinery mfg 177-209 Transportation mfg 219-229 Motor vehicle mfg 219 Shipbuilding 228 Professional photographic equipment mfg 239-267 Food and kindred products mfg 268-298 Textile mfg 309-327 Paper mfg 328-337 Printing publishing 338-339 Chemical mfg 347-369 Paint mfg 359 Petroleum coal mfg 377-378 Rubber plastic leather mfg 379-397 Rubber mfg 379 Leather mfg 388,397 Transportation 407 429 Railroads 407 Trucking 417 Water 419 Air 427 Communications 447-449 Utilities 467-479 Wholesale trade 507-588 Retail trade 607-698 Business repair services 727-759 Auto repair 757 Serv to dwellings 728 Personal serv 769-799 Laundries 779 Entertainment and recreation 807-809 Professional and related services 828-897 Banking real estate etc 707-719 Public administration 907-947 Total group No. control 16/144 3/16 100/290 11/34 3/17 7/63 11/23 7/12 29/82 OR 0.6 0.9 1.6d 1.4 0.8 4.2d 1.9 2.2 1.6 16/38 22/102 24/144 5/28 17/110 1.8 0.8 0.7 0.7 0.7 10/23 25/106 5/13 3/18 8/60 9/48 3/14 3/27 7/10 4/6 3/4 86/327 34/181 11/37 24/61 5/4 5/31 11/62 24/125 70/373 40/87 10/16 13/24 28/93 7/21 4/25 47/207 27/130 64/278 1.9 1.0 1.9 0.6 0.6 0.8 1.0 0.5 3.0d 2.5 3.2 1.1 0.8 1.1 1.6 4.2d 0.7 0.8 0.8 0.8 1.9d 2.3 2.4d 1.4 1.5 0.6 0.9 0.9 0.9 Reduced control groupb OR 0.5 0.7 1.6d 1.3 0.8 3.6d 1.8 2.0 1.6 22226 22226 22226 22226 22226 12 12 2.5 0.7 0.6 0.8 0.7 0.5 3.3 2.6 4.5 1.1 0.8 1.0 1.9d 5.3d 0.7 0.8 0.8 0.8 1.80 2.2 1.8 1.2 1.2 0.5 1.0 1.0 0.8 No. = number of exposed controls OR = odds ratio Maximum Likelihood Estimate US Census Bureau Industrial Classification System U.S. Bureau of Census Potentially occupationally related cancers as defined in text removed significant at < .05 dsignificant at < .01 1971 ** Mita, cacy yeaa : i Boy , aa | 1 ! | | i ' ' | | : on | : ; i 1 ; : . t i i 570 Milne et al TABLE IV Risk Associated With Occupation for Male Lung Cancer Cases Occupation and OCS code" All professional technical and kindred 001-195 Engineers accountants lawyers 001-056 Physicians and nurses 061-085 Managers administrators 201-245 Sales workers 260-285 Clerical workers 301-395 All craftsmen 401-475 Bakers 402 Brickmasons 410 Carpenters 415,750 Cabinetmakers furniture finishers 413,433 Typesetters 422 Electricians 430 All foremen 441 Machinists 461 Aircraft mechanics 471 Auto mechanics 473 Heavy equip mech 481 Painters 510 Plasterers 520 Plumbers 522,523 All operatives except transport 601-695 Assemblers 602 Laundry cleaning operatives 630 Machine operatives 645-660 690-692 Welders 680 All transport operatives 701-715 Bus drivers 703 Taxi drivers 714 Truck drivers 715 Other transport 704-713 All laborers except farm 740-785 Constr laborers 751 Longshoremen 760 Farmers & farm laborers 801-824 All service workers service workers 901-26 901-26 Janitors 901-903 Food service workers 910-916 Guards 962 Total group No. control OR 51/283 139 8/52 109/488 40.205 44/199 215/771 2/21 317 25/207 4/19 2/15 18/40 1337 1337 842 842 5 11/33 15:49 24.59 11.10 8:34 64 215 59 4 15.28 5/16 36.109 4.5 27 23.53 7.44 . 328 10/23 7:27 14.95 0.8 0.8 0.6 0.9 0.8 0.9 1.2 0.4 2.1 1.2 1.0 0.6 1.8 1.5 0.7 3.6 1.2 1.2 1.7 4.3d 1.1 1.2 2.3 2.0 2.1d 1.2 1.3 3.5 1.2 1.6 0.7 0.9 1.9 1.0 0.8 7/73 216 1.1 36/125 1.2 23:91 12.32 1.0 1.9 Reduced control groupb OR 0.8 0.8 0.9 1.0 0.8 0.9 1.2 0.4 1.8 1.1 1.1 0.7 1.8 1.7 0.7 4.4d 1.1 1.1 1.8d 4.4d 1.1 | 1.2 1.6 1.4 2.5 1.1 2.8 1.0 1.3 0.6 0.9 1.7 0.9 - 0.7 0.51.2 1.2 1.0 2.0 No. === number of exposed controls OR = odds ratio Maximum Likelihood Estimate US Census Bureau Industrial Classification System see U.S. Bureau of Census 1971 Potentially occupationally related cancers as defined in text removed significant at < .05 significant at p < .01 Mortality of workers in Washington state Milham 1976 3 a PMR Standard Mortality Ratio study of deaths from lung cancer among US men by occupation using 1950 Vital Statistics data Enterline and McKiever 1963 4 a retrospective interview study comparing lung cancer patients with persons with nonneoplastic diseases using persons w which oth study com and 6 st iindnustdryustrya a Menck ar tions 1970 Cens risk amon Thes compariso There wer lung cance occupation mechanics mechanics Find our study painters and Hend significant reported b plasterers demonstra _. was repor Som hazardous ers and cu 1979 Dol TORRE cree gS mag act SHR Eno Sanam TB gee RT ae Lung Cancer and Occupation TABLE V. Risk Associated With Occupation for Female Lung Cancers Occupation and OCS code" All professional technical and kindred 001-195 Physicians and nurses 061-085 Managers administrators 201-245 Sales workers 260-285 Clerical workers 301-395 All service workers 901-965 Cleaners 901-903 Food service workers 910-916 Health service workers 921-926 Personal service workers 931-954 Beauticians 944 All private household workers 980-986 No. control OR 10/232 0.8 5/81 1.3 7/89 1.4 4/74 1.0 21/318 = 1.1 15/184 1.4 1/55 0.3 3/61 0.9 3/31 1.7 8/35 4.1 7/17 7.0 4/49 1.5 No. = number of exposed OR = odds ratio Maximum Likelihood Estimate US Census Bureau Occupational Classification System see U.S. Bureau of Census 1971 significant atp < .01 571 persons whose primary lifetime employment was clerical work as the standard to which other occupations were compared DeCoufle and Stanislawcyzk 1977 5 a study comparing cases of lung cancer to other cancer controls Williams et al 1977 and 6 a study calculating expected incident cases and deaths for each occupation and industry assuming equal specific rates across occupation categories Menck and Henderson 1976 In this latter study the occupation popula- tions were derived from two Public Use Samples of the Los Angeles County 1970 Census The authors used the ratio population as a measure of risk among the individual occupations Menck and Henderson 1976 These studies provide results obtained by at least four different methods for comparison Not all occupational groups however were included in each study There were 48 occupational groups that appear to be associated with an increased lung cancer risk in one or more of these studies Among these 48 there are several occupations observed in our study that are hitherto not reported namely aircraft mechanics guards and night watchmen and machine operators Findings were consistent across studies for a number of occupations Including our study there are six reports of significantly high lung cancer risk associated with painters DeCoufle and Stanis awcyzk1977 Enterline and McKiever 1963 Menck and Henderson 1976 Milham 1976 Williams et al 1977 We have also observed significantly elevated odds ratios for several other occupational groups that have been reported by others as having an increased lung cancer risk These occupations include electricians construction workers and bus and truck drivers We did not plasterers each of which demonstrate significant risk for asbestos workers cooks or plumbers was reported by three or more studies Some of our observations may be due to occupational exposures known to be hazardous such as arsenic and benzene for painters asbestos for construction workers and cutting oils for mechanics Cole and Goldman 1975 Kusnetz and Hutchison 1979 Doll 1959 Selikoff et al 1967 Potential hazards such as dusts plasterers 572 Milne et al TABLE VI Occupations Reported to be Associated With Lung Cancer Risk Occupation Aircraft mechanics Auto mechanics Artists Asbestos workers Bartenders Blacksmiths forgemen hammerers Bus drivers Clothing ironers Construction workers Cooks Dental lab technicians Decorators Electricians Elevator operators Entertainers Excavating grading machine operators Guards nightwatchmen Heat treaters Inspectors Janitors cleaners Longshoremen Machine operators Machine shop workers Mechanics except auto Metal molders Mine operatives Officers enlisted men Armed Services Oilers greasers except auto Orchardists Painters Paperhangers Photoengravers Plasterers Plumbers Pressmen Railroad conductors Restaurant bar managers Roofers Sheetmetal workers Shoe repairers Stationary engineers Tailors and furriers Taxi drivers Telephone linemen Truck drivers Welders Writers Author Abbreviation DeC 2 D E Me Mil Mn W 9 ** x x X X X : x x** x x *** x X x X X ** X x x x ** Xx ^ x ** x X X x ** X ^ . ^ ^ ^ x X *** X X X ^^^ ^^^ ral ^^^ ** X Numbers in brackets indicate reference source Mn is current Milne study x denotes a reported association significant at p < .05 The absence of association was found or that occupation was not evaluated * adjusted for smoking ** Significant SMR but not significant PMR an x indicates either no are PMT eg 4 oe > Va ~ ose eee Ly 4 diesel fum be respon exposures of study Man to evaluate tions for \ one or twC The drawing c implied as exposures compared be occupat ing potenti particularly In additior guards was control ser controls g ing servic reflects a p category Two categories chance alor versus non- sons associ for testing risk DeCo to compare workers in using this g We al as the case Since color and Peto 15 with particu cantly assoc may reflect that the sign not due to c published re Menck and 1977 enhar Occup The industr ment but m Lung Cancer and Occupation 573 diesel fumes bus and truck drivers and hairsprays and dyes beauticians may also be responsible for increased lung cancer risks Other observations may result from exposures not yet identified or may be due to biases introduced by the chosen methods of study Many death certificate studies have used proportionate mortality ratios PMRs to evaluate risks associated with occupational factors By definition all the proportions for various categories must sum to unity so a very high or very low PMR for one or two occupations affects the PMRS for all other job classifications The odds ratio analysis avoids this pitfall However one must be cautious in drawing conclusions about causality using other cancer deaths as controls The implied assumption is that persons with other cancers are not biased with respect to exposures of interest Axelson 1979 In an effort to minimize this bias we also compared the lung cancers with a group of controls whose cancers are not known to be occupationally related In several instances an association was increased by deleting potentially occupationally related cancers from the comparison group This was particularly true for the water transportaion industry and for leather manufacturing In addition the risk among aircraft mechanics painters machine operatives and guards was increased when the occupation cancers were eliminated from the control series In other cases significant associations were reduced by removing these controls glass manufacturing professional and photographic equipment manufacturing services to dwellings bus and truck drivers and construction laborers This reflects a proportionally greater number of controls dropping out of the nonexposed category Two problems arise when multiple comparisons are made between exposure categories First when there are many comparisons significant results can occur by chance alone Second our exposure groups were defined as for example carpenters versus noncarpenters or shipbuilding versus nonshipbuilding Thus when those persons associated with high risk occupations become part of the nonexposed standard for testing other occupations the result is a more conservative estimate of relative risk DeCoufle and Stanislawcyzk 1977 used clerical workers as a group with which to compare each occupation Since the lung cancer risk associated with clerical workers in our study was less than 1.0 when compared with all other occupations using this group as a standard would have produced less conservative odds ratios We also performed a separate analysis using colorectal cancers ICD 153 154 as the case group to evaluate whether our approach produced spurious associations Since colorectal cancer is not generally considered to be occupationally related Doll and Peto 1981 Higginson 1980 we would not expect to see significant associations with particular occupational groups In our data set colorectal cancer was significantly associated only with collar occupations typesetters and farmers which may reflect socioeconomic or demographic factors Therefore we feel fairly certain that the significant lung cancer risks among multiple industries and occupations were not due to chance alone In addition the overall consistency of our results with other published reports DeCoufle and Stanislawcyzk 1977 Enterline and McKiever 1963 Menck and Henderson 1976 Menck et al 1977 Milham et al 1976 Williams et al 1977 enhances the validity of our observations Occupational histories obtained from death certificates may not be accurate The industry and occupations coded on death certificates should reflect usual employment but may in fact reflect last known employment In addition occupations are 4 ; - 8 : oe |! oe E 4 = FREI ARES Age te Ee aeat be ae Dare erR St ne eae oF , j SEO TT Er POURRA NS REE RRPET EE R ro ase ~ - 574 Milne et al often reported by next or someone else less familiar with the decedent Enterline and McKiever 1963 noted that craftsmen foremen and kindred workers were overstated on death certificates compared with census schedules and that clerical and kindred workers and operatives were underreported Although adjust- for this bias did not reduce the significance of their results for these major icnagtegories the associations for some subgroups were appreciably weakened Enter- line and McKiever 1963 Since our study was based on death certificate analysis we did not have any information on tobacco use or socioeconomic status which are potential confounders Doll 1959 The association of smoking and occupational status has been evaluated in a recent report Covey and Wynder 1981 According to the authors the following were found to have a greater proportion of current cigarette smokers than occupations taxi truck and bus the national average 38 painters machinists carpenters drivers and farmers These groups also tended to be heavier smokers 21+ per day and smokers of the high tar 21+ mg cigarettes Williams et al 1977 reported that construction workers also are more likely than others to be smokers We did not find machinists or farmers to be associated with a higher risk for lung cancer ctaharnpeonttheerrs cancers Our results for construction workers and bus and truck drivers are consistent with those for studies that adjusted for smoking in the analysis Dunn and Weir 1968 Williams et al 1977 Our analyses of female deaths was limited by small numbers The only signifi- cant association we observed was among beauticians This finding has been reported previously Menck et al 1977 Garfinkel et al 1977 one of these reports utilized the same Alameda County lung cancer cases but used a different comparison group et al 1977 As larger numbers of women enter the work force the aGnaarlfyisnikseolf occupationally related deaths in women should be more statistically sound The analysis of our data with regard to industrial classifications also produced results consistent with those of other investigators Significant associations with auto services construction transportation primary metal manufacturing and leather rmeapnauifracturing have been reported previously Menck and Henderson 1976 Williams et al O1n9e77association usually reported for lung cancer that of employment in the shipbuilding industry ie asbestos was not found in our study The lack of a demonstrated effect may be explained by lack of a long enough latent period Lung cancer originating from asbestos exposure usually requires a minimum year 1975 Selikoff et al 1980 If the height of exposure was during latency Higginson increased risk due to the 1940s at a naval ship repair base in Alameda County an asbestos would not have been apparent in the period covered by our study 1958- 1962 The lack of association may also be due to factors not accounted for by this study design corroborate the findings of other investigators who have used dOiuffrersetnutdymestehrovdess ttoo identify occupations that may contribute to the etiology of We have also identified some additional occupations that may have an liunncgrecasaendcerrisk It is beyond the scope of a death certificate study to assess the interactions among occupation and socioeconomic or lifestyle factors Future studies as well as related factors may identify etiologic which focus on specific exposures components of lung cancer risk among workers in the occupational settings that we and others have identified as potentially hazardous ACKN T Health for pro REFER Axelson E Cole P iy Covey I DeCouf ( Doll R Doll R. Dunn J Enterli Garfink Gart JJ Higgin Higgin Kenna Kusne Mench Mencl Milha Seliko Seliko US E Will Work Lung Cancer and Occupation 575 ACKNOWLEDGMENTS The authors wish to thank the Department of Biomedical and Environmental Health Sciences of the School of Public Health University of California Berkeley J for providing access to the Alameda County registry data REFERENCES Axelson O 1979 The referent control study in occupational health epidemiology Scand J Environ Health 91-99 Cole P Goldman MB 1975 Occupation In Fraumeni JF Jr ed Persons at High Risk of Cancer New York Academic Press pp 167-183 Covey LS Wynder EL 1981 Smoking habits and occupational status J Occup Med 537-542 DeCoufle P Stanislawcyzk K 1977 A retrospective survey of cancer in relation to occupation DHEW NIOSH Publ 77-178 Washington DC US Government Printing Office Doll R 1959 Occupational lung cancer A review Br J Ind Med 181-190 Doll R Peto R 1981 The causes of cancer Quantitative estimates of avoidable risks cancer in the US today JNCI 66 1193-1308 ~ Dunn JE Weir JM 1968 A prospective study of mortality of several occupational groups Arch Environ Health 71-76 Enterline PE McKiever MF 1963 Differential mortality from lung cancer by occupation J Occup Med 283-290 Garfinkel JS Selvin S Brown SM 1977 Brief communication Possible increased risk of lung cancer among beauticians JNCI 141-143 Gart JJ 1970 Point and interval estimation of the common odds ratio in the combination of 2 ^ 2 tables with fixed marginals Biometrika 471-475 =e Higginson J 1975 Cancer etiology and prevention In Fraumeni JF Jr ed Persons at High Risk of Cancer New York Academic Press pp 385-398 Higginson J 1980 Proportion of cancers due to occupation Prev Med 180-188 Kennaway NM Kennaway EL 1936 A study of the incidence of cancer of the lung and larynx J Hygiene 236 267 Kusnetz S Hutchison MK 1979 A guide to the relatedness of disease DHEW NIOSH Publ 79-116 Washington DC US Government Printing Office Menck HR Henderson BE 1976 Occupational differences in rates of lung cancer J Occup Med 797-801 : Menck HR Pike MC Henderson BE Jing JS 1977 Lung cancer risk among beauticians and other female workers JNCI 59 1423-1425 Milham S Jr 1976 Occupational mortality in Washington State 1950-1971 volume I. DHEW Publ 76-175 Washington DC US Government Printing Office : Selikoff IJ Bader RA Bader ME Churg J Hammond EC 1967 and Neoplasia Am J Med 487- 495 Selikoff IJ Bader RA Bader ME Churg J Hammond EC ( 1967 Editorial Asbestosis and Neoplasia Am J Med 487 495 US Bureau of the Census 1971 1970 Census of Population Classified Index of Industries and Occupations Washington DC Williams RR Stegens NL Goldsmith JR 1977 Associations of cancer site and type with occupation and industry from the Third National Cancer Survey Interview JNCI 59 1147-1185 World Health Organization International Classification of Diseases seventh rev Geneva WHO 1965