Document Z4m3Ydq5LE9Qd5qaQrX4wL2vO

ASBESTOS, DENTAL X-RAYS, TOBACCO, AND ALCOHOL IN THE EPIDEMIOLOGY OF LARYNGEAL CANCER M. Ward Hinds, MD, MPH,*-f David B. Thomas, MD, Dr PH,? and H. P. O'Reilly, MPH* A case-control study of 47 laryngeal cancers in males of three counties of Washington State was conducted. Personal interview was used to obtain in formation on smoking, alcohol use, exposure to asbestos, and other substances, and x-rays of the head and neck area. Smoking and alcohol consumption were found to increase risk of laryngeal cancer independently, with a clear doseresponse relationship. Neither asbestos exposure nor exposure to other sub stances was found to significantly increase the risk of laryngeal cancer, although the relative risk with asbestos exposure was 1.75. Lifetime history of exposure to dental x-rays on five or more occasions was associated with sig nificantly increased risk of laryngeal cancer among heavy smokers but not among light smokers. The importance of tobacco and alcohol in the epidemiology of laryngeal cancer was re-affirmed, the importance of asbestos exposure was brought into question, and a possible relationship of laryngeal cancer with exposure to dental x-rays among heavy smokers was demonstrated. Cancer 44:1114-1120, 1979. ertain risk factors for cancer of the operatives in the leather industry.19 Smoking C larynx have been well established by epi history but not alcohol intake was controlled demiologic studies. These include smokinfgo,r in the analysis. A mortality study of mall- alcohol, and male sex.9-11-16'20'21 In addition, workers in a nickel refinery disclosed 5 various occupational exposures have been in laryngeal cancer deaths when only 1.4 deaths vestigated as possible causes of cancer of the were expected based on national mortality larynx. Chemical exposures have been im rates in Norway.13 An association oflaryngeal plicated by two studies. Cancer mortality cancer with wood dust exposure among non- was examined among 139 U. S. counties with smokers but not among smokers has also been the highest proportion of employment in the noted.11 chemical industry.7 It was found that these Considerable interest has been aroused chemical industry counties showed a statisti concerning the possible association of asbestos cally significant 10% elevation in mortality exposure and risk oflaryngeal cancer. Asbestos from cancer of the larynx when compared is known to be associated with bronchogenic to total Li. S. mortality. A case-control study carcinoma and mesothelioma and possibly carried out at the Roswell Park Memorial in other cancers.1 In 1973, Sic!! and McGill re stitute lound significantly elevated relative ported a case-control study of 100 males with risk for laryngeal cancer among barbers, laryngeal cancer.17 They found that 31 cases operatives in the chemical industry, and reported exposure to asbestos as compared to only 3 controls for a relative risk (RR) of 14.5. The same year a brief report on a From the *,()ccup;iiimi;il Health Section. Washington cohort of workers at an asbestos textile and Slate Depart men i nl'Social and Health Services, Olympia, WaOiiitgurti. tC-aiuer Center of Hawaii, Epidemiology IVogiam. University l Hawaii, Honolulu. Hnunii. Hud IloltltiiiM>n dinner Reseatcli Ck'titei, StMillc, Washington cement factory indicated that 2 deaths from cancer of the larynx had occurred, whereas only 0.37 were expected.11 A second tasccuntrol study in 1975 found that 10 of -13 Stippoitcd ut p.tit hv NCI (uant Nit. NCl/CA 157*4 males with laryngeal cancer gave histories ul and Cnimatl MO!-Cl**:Vl!i 1 1. Address lot leptinls: Ur. \1. W. Hinds, Canter Center of I I.iwaii, University of Hawaii, 1236 l^iuhala. Suite *407, Honolulu. HI 9681:*. Accepted for publication September 18, 1978. exposure to asbestos compared with none of the age-matched neighborhood controls.13 The present study was initiated following an investigation of mesothelioma incidence in OOU8-543X/79/090WI 114 $0.85 American Cancer Society 1114 HWBUI0006302 N... 3 Kimdkmioi.ocy ok Lakync.cai. Canckk lli>t'L\ ft al. 11 IT) I lie area of l*uj^ei Sound. Washington State, where there is considerable shipbuilding activity and. thus, asbestos use.0 because mesothelioma incidence was round to be sig nificantly and greatly elevated in the one county that contains a very large Naval ship yard, incidence rates for cancer of the larynx were also examined for that county. Sur prisingly. that county was found to have the lowest, rather than the highest, age-adjusted incidence of laryngeal cancer in men of the six counties examined. There was no correlation between incidence of mesothelioma and in cidence of laryngeal cancer by county (Table 1). The question was thus raised of whether asbestos exposure is related to risk of laryngeal cancer in the Puget Sound area. This study was designed to explore that question and to examine other exposures which might have a relationship to risk of laryngeal cancer. Matkhiai-s and Mkthous Ail males diagnosed as having laryngeal cancer while residents of Pierce, King, or Kitsap counties between January 1. 1076 and June 30, 1077 were identified through the Cancer Surveillance System (CSS) of the Fred Hutchinson Cancer Research Center in Seattle. The CSS is a population-based cancer registry and a member of the Surveillance, Kpidemiology and End Results (STER) Group of the National Cancer Institute. Sixty-nine such cases were identified and 47 of these were included in this study. The remaining 22 cases were excluded because their private physicians were unwilling for their patients to he interviewed (10), the patient was deceased with no close relatives (8), the patient could not be located (2) or the patient refused to be interviewed (2). A neighborhood control was chosen for each case with matching for sex, race, and 10-year age group. Controls were chosen by a 1'am.k 1. Imulcnte of MrsitliK'liom.i .HitI l.,n\n^eal ((.Hirer in Six Wavliinginn ('.mimics. I!l74-7li Average annual agc-.iiljiiMrfl itKHleme per 100.000 males County I'ictce C'.reys Miirlwr Thurston Kin* Snohomish Kitsap Larvngcafranter (N = 220) ion H.6 8.0 6.8 5.5 4.0 Mcvothclionia (N* O.'l 0.0 0.8 1.0 0.8 1.3 door-to-door survey starting at the third house to the right of the case's address and follow ing a written set of directions until a control was found. A potential control address vvas not to be abandoned until at least two visits at different times of the day and two phone calls, one at night, were unsuccessful. No addresses had to be abandoned with this pro cedure, although 8 potential controls refused to be interviewed and were replaced by others in the same neighborhood. All cases and controls read and signed an informed consent form and received a copy prior to being interviewed. The interviews were carried out by two of the authors (MWH and HPO) with matched cases and controls always interviewed by the same person. Inter views were conducted cither in the subject's home or at his place of work. In three in stances where the case was deceased, a pri mary relative (wife, brother) who lived with the deceased for many years was interviewed. The same relative of the matched control for these cases was interviewed rather than the control himself. The interviewer employed a standard questionnaire and over a period of 20 to 30 minutes elicited a detailed history of places and nature of'cmploy meni through out the subject's life and specific histories of Tahi.k 2. Comparison of Usual Cigarette Consumption levels for Case* awl Controls Cigarettes per l.iv Cases I Vi tent (No) l.rss ill.ill 10 10-20 31-10 Mine ih.m *10 1 tllAI. 1.3 27.7 -Hi ft 21 100 (1 (31 (lb t----4 (Ml) (17) Test lor trentl, p < .001, Controls Percent 3-1.0 20.8 27.7 H.!i 10(1.0 (No.) (Mil (Ml (13) III H7) HR- 1.00 7 13 15.-VI 20.00 p ^.025 < 01 <.ni * Risks arc expicsscil irlalivc m a risk of 1.00 lor consumption of less than 10 cigarettes per clay. 1116 Cancer September 1979 Vnl. 44 T.uui, 3. Companion of Tnoil Pack-Years of Cigarette Exposure lor Cases amt Comrols Pack-) e;i rs exposu ft* Less than 10 10-33 40-69 70 or more TiriAi. Cases Percent (No.) 4.3 35.5 40.4 29.8 100.0 (2) (12) (19) (14) (47) Controls Percent (No.) 34.1 31.9 17.0 17.0 100.0 (16) (15) (8) (8) (47) RR* 1.00 IS. 40 19.00 14.00 P <05 <.ooi <01 Test lor [rend, p < .005. * Risks are expressed relative to a risk of 1.00 for exposure to less than 10 pack-years. exposure to asbestos, arsenic, nickel, uranium, chromium, cobalt, and other chemicals or materials felt to be of importance to the sub ject. In addition, a detailed smoking history, alcohol consumption history, history of x-ray exposure in the head and neck area, and his tory of hobbies were obtained. Ail cases were classified according to site of origin of the tumor (glottic, supraglottic, other) and all data were coded and key punched for analysis. For 2 x 2 matched-pair tables with small tell frequencies, the method suggested by Bennett and Underwood was used for testing significance.1 Clti-squarc analysis was used where matching was not maintained. Results All 47 laryngeal cancers were in Caucasians. There was no significant difference between the mean age at diagnosis of the cases (63.0 8.3*) and the controls (63.1 i 9.2), in dicating that matching on age was successful. In addition there was no significant difference between the mean age of the cases and the mean age of all male laryngeal cancer cases identified by CSS in die three counties under study for the period January 1, 1975 through Meal! standard ties union. June 30, 1977 (63.2 9.2, N = 153). A comparison'of the 47 study cases to these same 153 total cases likewise revealed no significant difference for site of tumor, county of residence, or marital status. Thus, the study population of 47 cases would appear to lie representative of all laryngeal cancer cases occurring in males during this period and in this geographic area. The site of the cancer was glottic for 35 of the cases (74.5%), supraglottic for 9 (19.1 %), and subglottic or other lor the remaining 3 cases. This is a higher proportion of glottic cases than has been reported in a study by Wyndcr el at.11 but may reflect the substantial population of Swedish ancestry in the l'uget Sound area, since glottic cancers in Sweden may constitute up to 89% of all laryngeal can cer,10 In the 1970 census, the three counties surveyed in this study had a proportionate Swedish population more than four times greater than Wynder's population."1 A comparison of the 47 cases interviewed with the 22 cases not interviewed revealed that the cases not interviewed were slightly younger (60.7 7.5 years), less likely to be married (77.3% versus 80.9%), and less likely to have a glottic cancer (54.5% versus 74.5%), but none of the differences were statistically significant. Tabu: 4. Comparison of Usual Level of Alcohol Consumption lor Cases .111(1 Comrols Cases Controls Altnhtff per flay Pen eni (No.) Percent |N<>.) RR* V f .css than 1 unit 1 -`J units .4-ti units Muir than <i tnuts Total l':i * 19 2 31.9 j5 r 100.0 (ID (9) (15) tl-'l 1-17) 53.2 21.2 19.2 IV. 4 100.0 (25) (10) 19) (3) (-17) l.tm 2.05 \\Y 3.79 <.115 '.Mil <.01 Trsl lot nend. |> < .001. * Risks arc expressed it-laiive to a risk of 1.00 (in consumption of less than 1 mm pci day. .\ Cl at I rt ex Sll co in cn wi ol sn Cl I (h ha ol on Sll Sll it an do Ah Inalh I (3.- CO I 11 ex | tlu- sis Jar. stai CX| am the yea no si ra (slii alsi of x pox boo' c;in< snu Tin cat i V of cJi r nifii HWBUI0006304 No. :i Kimdkmioi.ugv of Laryngeal Cancer I limit el al. 1117 The relationship of smoking and alcohol consumption to risk of laryngeal cancer were analyzed without maintaining matched pairs. Taiu> Relative Risks of I-irvngoal (lamer Ai u tiding to Picscncc or Absence ot 1 wo Kxposutes -- Heavy Drinking and Draw .Smoking* The sinoking association, with a dose-response relationship, was apparent whether one I IlMVV smoking examined the usual level of cigarette con No Yes sumption (Table 2) or the total pack-years Htxivv consumption calculated from the year of start drinking ing sinoking to the year of diagnosis for the No Yes 1 (Hit 2.25 2.1 1 x.lo ease (Table 3). Exsmokers were combined with current smokers for all analyses. Only 1 of the eases hut 8 of the controls were nonsmokers. Usual alcohol consumption was cal culated on the basis of units per day. with * Heavy drinking is defined as one or more units of alcohol per day: heavy smoking as more than 20 cig arcues per day. t Risks arc expressed relative lo a risk of 1.00 for persons who arc neither heavy drinkers nor heavy smokers. the approximation that one unit = 1 ounce of hard liquor = 4 ounces of wine = 12 ounces of beer. The results in Table 4 clearly dem onstrate an association between alcohol con sumption and risk of laryngeal cancer and strongly suggest a dose-response relationship. Although the number of eases was small, it was possible to demonstrate that smoking and alcohol consumption appear to act as in dependent risk factors ('fable 5). This would agree with the findings of previous studies.11 Although synergism was suggested,3 the broad, open-ended categories do not really allow such analysis. for a history of exposure to asbestos, 23 (53.2'r) of the eases and 19 (40.49r) of the controls gave a positive response. There were 14 matched pairs for which only the ease was exposed to asbestos and 8 pairs for which only the control was exposed. Matched pair analy sis of asbestos exposure thus gave a RR for laryngeal cancer of 1.73. which was not statistically significant. When the duration of exposure to asbestos was examined for cases and controls with an exposure history, again to any one of these and risk of laryngeal cancer. There was some suggestion of an associa tion between a history of exposure to dental x-rays and risk of laryngeal cancer. Cases and controls were asked to estimate the number of times in their lives that a visit to a dentist had involved dental x-rays. Cases reported a mean of 14.3 dental x-ray visits and controls a mean of 12.3, a nonsignificant dil Terence. The frequency distribution of the number oi dental x-ray visits was highly skewed however, with 15 (31.9%) eases and 22 (46.8'/7) controls giv ing a history ol' only 0-4 dental x-ray visits. Because of a concern about the possibility of confounding of any association of dental xrays and laryngeal cancer by either smoking or alcohol use, the relationships between smoking and dental x-rays and between alcohol use and dental x-rays were examined for all 94 subjects. No statistically significant association could be found in any analysis of number of dental x-ray visits versus number there was no significant difference (8.4 8.2 years and 8.3 7.5 years, respectively), and nr> dose-response effect could be demon Table 6. Matched-Pair Analysis of Asbestos Kxpoxurc in Relation to Risk of Uii)iigc;l (lamer strated. 'file sources of asbestos exposure (shipyards, brake repair, construction) were also similar for both rases and controls. A lack ol significant association between asbestos ex Pairs analysed Pairs Paii* w/otilv w/onlv cases controls N cxjxiscd exposed KK l> posure and htrynge.il cancer was round for both glottic and sttpraglotiic cases, and for lase-ionliol pairs fortuitously matt lied for smoking and al< nliol t nnsuinptiou ( fable ('>). 1 he RR was consistently greater than 1.00 for each category of pairs analyzed, however. Verv lew cases or controls gave a history All Glottic cancer sSupraglottic earner Smoking- matched* Alcohol mauhedt 47 35 y 12 12 14 9 3 A H 1.75 .21 7 1.29 1 l.ou .22 1 3.011 .37 n 1.33 .73 of exposure to arsenic, nickel, uranium, chromium, or cohalt and there was no sig- * Matched on smoking categories in 'I able 2. t Matched on alcohol consumption categories in nili<am association found between exposure Table 4. 1118 Cancer September 1979 Vol. 44 Taiii.k 7. Cutnpatbou ol Numlier of Denial X-ray Visit* fv Oaves aiul liy Controls Ainoiij' Smokers of 21 or More Cigarettes a Day No. of denial x-ray sitiis 0-4 5-9 10 or more Till Al. Cases Percent (No.) 21.9 18.8 59.3 100.0 (') l> (19) (32) Controls Percent (No.) 01.7 11.8 23.5 100.0 (ID (2) (4) (H) RR* 1.0 4.7 7.5 P N..Y <.02 Test for trend, p < .005. * Risks are expressed relative to a risk of 1.0 for exposure to less than 5 denial x-ray visits. of cigarettes a dav smoked or number of units the number of persons giving such an ex- of alcohol a week consumed. Specifically, posurc history was too small lor meaningful there was no evidence that more heavy analysis, smokers were present among the groups with more dental x-ray visits. For the categories of Discussion 0-4, 5-9, and 10 or more dental x-ray visit^^---^" the percentage of heavy (2l4-/dav) smokers! This study found, as have others, that the was 48.6, 44.4. and 56.4, respectively. The! most readily identifiable risk factors for dental x-rav data were then analyzed in twol larvngeal cancer are tobacco and alcohol con- ways to determine association with laryngeal sumption.9-20,21 The findings in this study, cancer risk. First, the matching of pairs was however, arc not supportive of the hypothesis disregarded anti evidence of a dose-response that asbestos exposure is an important risk was sought. Among light smokers (20 or fewer factor for cancer of the larynx, although the cigarettes per day), there was no increased RR of 1.75 found is consistent with a small risk of laryngeal cancer found in association risk associated with asbestos exposure, both with dental x-ray visits. However, among the studies ofSieli and McGill17 and Shcuignra heavy smokers, a dose-response relationship and Morgan15 found very high relative risks was suggested and a significant relative risk for laryngeal cancer associated with asbestos ol 7.5 was found when a history of 10 or more exposure (RR = 14.5 anti , respectively). It denial x-ray visits was reported (Table 7). A is difficult to understand why such a strong matched pair analysis revealed 12 case-control relationship would have been missed in the' pairs in which only the ease was exposed to current study if it is a real one. The investiga foe or more dental x-ray visits versus 5 pairs tion of Stcll and McGill used hospital con in which only the control was so exposed. This trols with no apparent matching other than difference was not significant, but when the sex and age. No alcohol consumption history matched pair analysis was stratified by smok was taken. We do not know if both eases and ing level, a significant association was found controls were interviewed by the same in among heavy smokers (21 or more cigarettes dividual using a standardized formal. Some per day) between exposure to 5 or more dental possibilities for bias thus exist, but it is dif x-ray visits and laryngeal cancer (Table 8). No ficult to see bow such biases would produce association was found between laryngeal can a RR of 14.5. One question which needs to cer and oilier x-rays of die head and neck, but be answered is whether the laryngeal cancer [ afill 8. Comparison of Maidicd Case-Control Pairs for Exposure lo Denial X-ravs Pairs i\/only Pairs w/only N rases exposed* ion!itils exposed RR I> All null hut jr.iil X i**ins flunked Ini snmkin^t 1 iKlt Heaw -17 III 12 12 * 5 5 2.10 .10 2 1.00 1.00 0 * .0-1 * Exposed = a nr more denial visits uiili x-rays taken. t L.igln smoking = 20 or Inver dgaieiics per day; licavy smoking = 21 or moic cigarettes per day. N< ta co i iny cm na: pat int- 'i did wel gro Fen. vie\ trol inte v/cn asbi mm in ti the sum of I. pare A six smol (100 war. tenti. tO S( high Th the | above were asbcsi Stcll ; Shctti posur the p in th aitswe posurc or in sible ti two p and tl ferent ex post coinini ingtnn While Tororu 1-iverp. theliou commo buildin HWBUI0006306 No. 3 Eimokmiologv of Laryngeal Cancer Hinds el al. 1119 cases were likely to gain any workmen's source of occupational exposure to asbestos, iompensation benefits by linking a past work with 12 of the 44 asbestos-exposed subjects ing exposure to their illness. Such a situation identifying a shipyard job as the site of asbestos could indeed introduce a large bias if the contact. However, occupational exposures nature of the investigation was at all ap also occurred frequently in construction of parent to the cases prior to or during the buildings and homes, brake lining repair, and interview. the plumbing industry. The investigation by Shcttigara and Morgan At least three previous studies present evi did match on neighborhood of residence as dence which is not in favor of a relationship well as sex and age, thus using a control between asbestos exposure and laryngeal can group very similar to that of the current study. cer. Sclikoff et al. have been following three Four of the cases were apparently inter large cohorts of asbestos workers for several viewed by clinic personnel while all the con years and as yet no significant excess of trols and the remainder of the cases were laryngeal cancer has been observed.14 Ad interviewed in their homes by a single inter mittedly, laryngeal cancer is an uncommon viewer. It seems possible that a search for neoplasm, but not so rare as mesothelioma asbestos exposure might have been somewhat for which excess mortality has been easily- more emphasized for the 4 cases interviewed demonstrated in these cohorts. The case- in the clinic, thus introducing some bias into control study of 3 14 cases of laryngeal cancer the study. It is surprising that alcohol con by Wynder el al. was unable to detect any oc sumption was not found to be related to risk cupational exposure, other than possibly of laryngeal cancer even though this is ap wood dust in nonsmokers, which conferred parently a well-established risk factor.11,10,21 excess risk of laryngeal cancer.21 Asbestos Asbestos-exposed cases were more often exposure was specifically questioned for and smokers than non-asbestos exposed cases none of the 13 nonsmoking cases were aware < 100% versus 90.9%) although the difference of such exposure. Finally, a retrospective was not significant. Again, although some po survey of over 14,000 cancer cases admitted tential for bias can be shown, it is difficult to Roswell Park Memorial Institute in Buffalo to see how bias alone might produce the was carried out with comparison to nonneo- high RR found for asbestos exposure. plastic controls." Each person was classified The primary difference in the findings of into as many different occupational groups the present study as compared to the two as were present in the history. Alter adjust above was in the proportion of controls who ment for smoking history, there were no oc were found to give a positive history for cupations usually associated with asbestos ex asbestos exposure. Only 3% of the controls of posure which showed an increased risk for Stell anti McGill and none of the controls of laryngeal cancer. Shcttigara and Morgan reported asbestos ex In summary, the relationship of asbestos posure as compared to 40.4% so doing in exposure to laryngeal cancer appears to be the present study. We included any person open to question. A recommendation for in the asbestos-exposure category if he further studies may not be warranted, how answered positively to an inquiry about ex ever, since asbestos exposure is already com posure as a result of working directly with ing under strict control because of its clear re or in the vicinity of asbestos dust. It is pos lationship to mesothelioma and lung cancer.1 sible that definitions of exposure used by the The finding in the present study of a pos two previous studies were more restrictive sible relationship between exposure to dental and this is in part responsible for the dif x-rays and laryngeal cancer among heavy ferent findings, or it may Ire that asbestos smokers is one which does call for further exposure in the general population is more investigation because of the common use of common in the Puget Sound area of Wash dental x-rays. Of course this relationship was ington State than in Liverpool or Toronto. based on a small number of cases and may While this might not be unexpected for not be real, or if real may not indicate that Toronto, it would he somewhat surprising for dental x-rays themselves tire responsible for Livetpool, whith is in an area where meso excess laryngeal cancer among heavy smokers. thelioma <lue to asbestos exposure is not un The frequency of visits to the dentist's office common.5 In the Puget Sound area, the ship and the use of dental x-rays arc undoubted), building industry was found to be a common correlated with many variables, particularly 1120 Cancer September 1979 Vol. 44 djclary habits, which were not controlled for in this study. Also arguing against a causal association between dental x-rays and laryn geal cancer is the fact that although thyroid cancer has been found in excess among per sons who have received large doses of external irradiation to the head and neck, there is as yet no evidence that these persons have an increased risk of laryngeal cancer.* It will re quire much larger studies and more detailed interviews to determine if dental x-ray ex posure poses a cancer risk for heavy smokers. ADDENDUM After acceptance of the manuscript for publication, an additional case-control study of laryngeal cancer and asbestos exposure appeared in the literature.** This study was con ducted in the province of Trieste, Italy, where 60 laryngeal cancer patients and 60 ageand sex-matched hospital controls were interviewed concerning smoking habits and oc cupational exposure to asbestos. Fifty percent of cases and 35% of controls gave a history ofoccupational exposure to asbestos. The relative risk of 1.66 was not statistically significant. REFERENCES 1. Becklake, M. R.: Asbestos-related diseases of the lung and other organs: Their epidemiology and im plications for clinical practice. Am. Rev. Respir. Dis. ! 14: 187-227. 1976. 2. Bennett, B. M., and Underwood, R. E.: On McNcmar's test for the 2x2 table and us power func tion. Biometrics 26:339-343. 1970. 3. Cole. P.. and MacMahon. B.: Attributable risk per cent in case-control studies. Dr. J. Prev. 5oc. Med. 25: 242-244, 1971. 4. Favus, M. J., Schneider, A. B.. et al: Thyroid can cer occurring as a late consequence nt head and neck irradiation. ,V Engl.J. MM. 294:1019-1925, 1976. 5. Greenberg, M., and Lloyd Davies. T. A.: Meso thelioma register 1967-68.Br.J. IndusL MM. 31:91 -104, 1974. 6. Hinds. M. W.: Mesothelioma in the United States: Incidence in the 197Q's./. Occupation. MM. 20:469-471, 1978. 7. Hoover, R., and Fraumcni, J. F.: Cancer mortality in U. S. counties with chemical industries. Environ. Hrt. 9:196-207. 1975. 8. Hmiicn, L., Bross. 1. D. J.. and Viadana, E.: A Retrospective Survey nl Cancer in Relation to Occupa tion. DHKW(NIOSH) Publication 77-178, Washington, D C., U. S. Govt. Printing Office, 1977. 9. Kahn. H. A.: The Dorn study of smoking and mortality among U. S. veterans: Report on eight and one-hall scars of observation. Sad. Cancer lust. Motiogr. 19:1-41), 1966. 10. Martcnsson, B.: Epidemiological aspects of laryn geal caiiuiotna in Scandanavia. Laryngoicope 85:11851189. 1975. 11. Momon. R. K.. and Lyon. J. L.r Proportional mortality among alcoholics. Cancer 36:1077- 1079. 1975. 12. Newhouse, M. L., and Berry, G.: Asbestos and laryngeal carcinoma. Ixincet 2:615. 1973. 13. Pederson. E-. Hogetveit. A. C.. and Andersen. A.: Cancer of the respiratory organs among workers at a nickel refinery in Norway. Int.J. Cancer 12:32-41, 1973. 14. Sclikoff, I. J., Hammond, F.. C.. and Scrdman. II.: Cancer risk in insulation workers in the United States, m Biological Effects of Asbestos. P. Bogovski, ct al.. Etls. Lyon, International Agency for Research on Cancer, 1973; pp. 209-216. 15. Sheltigara, P. T., and Morgan, R. W.: Asbestos, smoking and laryngeal carcinoma. Arch. Environ. Health 30:517-519. 1975. 16. Stcll, P. M.; Smoking and laryngeal carcinoma. Lancet 1:617-618, 1972. 17. Stell, P. M., and McGill, T.: Asbestos and laryngeal carcinoma. Lancet 2:416-417, 1973. 18. U. S. Bureau of the Census, Census of the Popula tion--1970. Washington, D. C., U. S. Govt. Priming Office, 1973. 19. Viadana. E.. Bross, I. J., and lloutcn, L.: Cancer experience of men exposed to inhalation of chemicals or to combustion products. J. Occupation. Med. 18: 787-792. 1976. 20. Vincent, R. G., and Marchetta, F.: The relation ship of the use of tobacco and alcohol to cancer of the oral cavity, pharynx or larynx. Am. J. Surg. 106:501-505, 1963. 21. Wvndcr, E. L., Covey. L. S , Mjbtirhi, K., and Mushinski, M.: Environmental factors in cancer of the larynx. Cancer 38:1591 -1601, 1976. 22. Bianrhi, C., Di Benito, L., Castdli, M., and Brnilo. A.: Exposition a lamiante dans le cancer du larynx. Paihologtca 70:403-408, 1978. Vc. ;in<i built (fed paiic A ,, (him j in M., IhJ.ii.-. idcnn nKl>l I 1 Niipi.u v Icm. 'lilcnit (> (Its niimifi h\d rnc ()< lulurcMiliin Clu-M r. lion <il m;iss. In Jn ('ini'll, rif'lll lot I-! olii il UiUnii 11 C.lnrl llini.i, j. Si llfiiii. 1 Adillrv HWBUI0006308