Document 4a257BGDMMvKR90JvJLpLzZEj

Arf. /. Cancer: |7,310-317 (1976) CoCvi O*. CV /QLcio- C* nfrywM; d* *# <t*a~ t^uTH K. PETERS FtMM- Htuksst J0u-fri A COHORT STUDY ON MORTALITY FROM CANCER AND OTHER CAUSES AMONG WORKERS AT A METAL REFINERY VITAL AN Of STUDY POPU Shmkan Tokudome and Masanori Kuratsune Department ofPublic Health, Faculty ofMcdlcine, Kyushu University, Fukuoka, Japan Summary. A non-concurrent prospective study ms made on deaths from cancer end other causes occurring among 2,675 male workers at a metal refinery from 1949 to 1971. The expected number of deaths computed by applying age- and cause-specific death rates of Japanese mates to these workers was compared with the observed number ofdeaths. Among 839 copper smelters, significantly Increased mortalities were notedfor lungcancer (SMR - 1.189) aiptcoM) cancer, but not_for cancer of the stomacnfTlver (primary) and biliary passages, pancreas and skin or for leukemia, tuberculosis, cerebrovascular diseases, heart diseases and liver cirrhosis. A dose-response relationship was demonstrated between the mortality from fung cancer and the degree of exposure. A very high excess mortality from lung cancer (SMR = 2,500) was Seen among copper smelters who were ' considered to have been most heavily exposed to , arsenic or workers who had engaged in sintering and blast furnace operations for 15 years of more before 1949. The latent period oflung cancer was 37.6 years on average, and not related to level of exposure. Twenty-six of 29 deaths from lung cancer atnong copper smelters occurred after they had left the refinery. Other production workers and clerical workers showed no significant excess mortality from any kind ofcancer. As reported, a case-control study made by using mortality cards demonstrated hidden cases of occupational lung cancer among copper smelters at a metal refinery in Ooita prefecture (Kuratsune er at., 1974). This necessitated a more extensive inves tigation of the current and former employees of the retineiy for their mortality experience of cancer and other diseases. A so-called non-concurrent pro spective (Lilienfeld et al., 1967) or retrospective cohort (MacMahon and Pugh, 1970) study was undertaken, the results of which are reported in this paper. MATERIAL and methods" Nominal rolls of the following four groups of workers were provided by the refinery. 0) copper smelters who retired before December 31. 1948; (2) workers employed up to December 31,1948 and who retired after January t. 1949; (3) workers who were employed and who retired after January 1.1949; and (4) workers stilt employed on August I, 1971. Name, sex, date of birth, date of employment, date of retirement, history of work (work section and job), Honseki (place of permanent family residence), present residence, cause of retirement, vital status (alive or dead) on December 31, 1971 and date of death were listed for each worker on these rolls. Deaths which occurred among these workers during the 23-year period from January ), 1949 to December 31,1971 were analyzed in this study. The survey for vital'status, was made In the following way. The refinery questioned the village, town or city office at the Honseki of each of the workers about his vital status since vital records of the Japanese are kept there, irrespective of place of residence. When the death of a worker was noted, the Ooita Labor Standards Office collected a copy of his death certificate from the District Legal Affairs Bureau at his Honseki, since the certificates arc legally preserved there for 27 years alter death. Thus, practically all the deaths during the above study period could be traced. From death certificates thus collected, we assigned an underlying cause of death to each death according to the 8th Revision ofthe International Classification of Diseases, Injuries and Causes of Death (WHO, 1967). In contrast to our previous study (Kuratsune et at., 1974), the validity of diagnosis was not examined by making any particular investigation. Although the nominal rolls listed 4,797 persons in all, female workers, mate workers whose Honseki was outside Ooita prefecture, those who bad died before or on December 31, 1948, those who had been employed less than 1 year before December 3). 1971 and those listed in duplicate, were exchidedfrom the analysis. The above exclusion of workers having their Honseki outside Ooita prefecture was simply due to the fact that their mortality information was not provided by the Labor Standards Office by the time of this analysis. Furthermore, persons whose Honseki, date of birth, date of employment or vital status was unknown were excluded. Thus, the remaining 2.675 workers including 839 copper smelters were used for the current analysis and their vital and employment status on December 31.1971 is shown in Tabic 1. Received: October 14. 1975. and in revised form December 17. 1975. Known to be alive Still employed Ex-workers Known to be dece Total study popufc 1 Ffgura in piftruhtst As shown in Ta into five cohorts ac both the work sect the work they wep the workers (coppe the copper smeltini of these workers h were included in worked in the secti transferred to othei the following cohor the workers belon. section, while coho: the maintenance cohort 4 those btlo lead electrolysis or of these workers 1 section. They werei the section to wh longest period of ti engaged exclusively the work sections ti For more detaile subdivided into t exposure (Table \ COHORTS BY JOB, AND TOTAI CetiOA 1 Coppet smt'. 2 Fetro-nickel smelting 3 Maintenantx transports a Copper or 1< elccirolysi product ioi sulfuric tic 5 Clerical wor Total f iR AND OTHER SRY , >F employment, date of ; (work section and job), nent family residence), ' retirement, vital status er 31, 1971 and date of worker on these rolls, among these workers rom January 1, 1949 to ilyzed in this study, uus was made in the questioned the village, Homtkl of each of the 45 since vital records of irrespective of place of if a worker was noted, ifTk-e collected a copy of : District Legal Affairs ice the certificates are years after death. Thus, uring the above study t collected, we assigned o each death according rnational Classification uses of Death (WHO, jidy (Kuratsune gnosis was not icular investigation, listed 4,797 persons in ere whose Honsektwas i who had died before those who had been 're December 31, 1971 ere excluded front the \ of workers having uefecture was simply ality information was indards Office by the nore, persons whose employment or vital preluded. Thus, the ;!uding 839 copper tnt analysis and their Deaember31,197lis and in revised form , t.l St,, Nl KC ; N\tr ,,; J KS Ji TABLE I VIT\L AND EMPLOYMENT 3TATW* OF STUDY POPULATION ON DECEMBER H. Iv7 Known to be alive Still employed Ex-workers Known to be deceased Total study population 12261 (324) 1,089 (338) 2,330 (682) 325(157) 2,675 (839) 1 Fijpires in pwtthaas on number ofcoppr As shown in Table 11, the workers were divided into five cohorts according to their jobs, considering both the work sections to which they belonged and the work they were engaged in. Cohort I included the workers (copper smelters) who had belonged to the copper smelting section for at least 1 year. Some of these workers had engaged in lead smelting but were included in this cohort. Those who had worked in the section for less than I year and were transferred to other sections were included in one of the followingcohorts; 2,3 or4. Cohort 2consi$ted of the workers belonging to the ferro-nickel smelting section, while cohort 3 comprised those belonging to the maintenance or transportation section, and cohort 4 those belonging to the section of copper or lead electrolysis or ofsulfuric add production. Some of these workers had belonged to more than one section. They were assigned to cohorts according to the section to which they had belonged for the longest period of time. Cohort 5 consisted of those engaged- exclusively in clerical work, irrespective of the work sections to which they belonged. For more detailed analysis, cohort 1 was further subdivided into several cohorts by degree- of exposure (Table V), First, it was divided into 3 TABLE It COHORTS av JOB, NUMBER OF WORKERS. DEATHS ANO TOTAL PERSON-YEARS AT RISK Cohan Number Number Total ei of pereonyctra wotfcm dcaftbft stffak 1 Copper smelting 839 157 15.521.3 2 Fetro-nlckel smelting 268 6 3.599.3 3 Maintenance or transportation 821 108 15,396.0 4 Copper or lead electrolysis or production of sulfuric acid 389 22 5.838.0 3 Clerical work 358 32 6.068.0 Total 2,673 325 46,443.0 cohorts, namely cohorts 11.12 and 13. according to length of employment. It was also divided into 3 cohorts designated as cohorts 21,22 ami 23 by level ofexposure. This subdivision, however, was approxi mate due to the lack of quantitative information in regard to exposure to arsenic and other related compounds released by the smelting processes. Therefore, after consulting the technical managers and physicians of the refinery, the subdivision was made according to work engaged on for the longest period of time, os described at the bottom of Table V. Furthermore, cohort I was divided into 3 cohorts designated as cohorts 31. 32, and 33 according to time of exposure (time when exposure occurred) and to length of employment as copper smelters. For each cohort, populations at risk were calcu lated bv the ordinary person-years method (Doll and Hill, 1936), computing age-specific mid-year popu lation In each of the calendar years from 1949 to 1971. In order to examine possible excess risks of mortality from cancer and other causes of death in the cohorts, the observed number of deaths from a specific cause was tabulated for each cohort over the calendar years and compared with the corresponding expected number of deaths which was computed by applying age- and cause-specific death rates by calendar year for male Japanese to the populations at risk of each of the cohorts and by accumulating overall age groups and the calendar years. In this calculation, the sex- and ago-spedfic death rates for malignant neoplasms and other diseases for 'all Japanese males in 1930. 1935, I960-, 1965 and I97(), appearing in the Vital Statistics, 1971 (1973) and in the Death Statistics of Malignant Neoplasms (1973) were used. For convenience, death rates *.g. for 1955 were applied to the populations at risk not only for 1955 but also for the 2 preceding and 2 following years, namely 1953,1954, 1956 and 1957. This lineal interpolation of death rates was based upon the fact that it yielded a good approximation even for lung cancer death which was increasing most rapidly. Although the ICD was revised three times, giving the 6th, 7th and 8th Revisions during the period from 1950 to 1970, the categorization of the specific diseases studied was practically the same except for " malignant neoplasm of respiratory system" and "malignant neoplasm of trachea, bronchus, and lung." Thus, the expected number of deaths calcu lated from the above mortality rates was considered to be comparable with the observed number of deaths tabulated according to the 8th Revision of ICD. For respiratory cancer, however, " malignant neoplasm of respiratory system specified as second ary" was included in " malignant neoplasm of respiratory system " in the 6th and 7th Revisions of ICD but not in the 8th Revision, and " malignant 312 TOKUOOMB AND KURATSUNE neoplasm of pleura * was included in. " malignant neoplasm oftrachea, bronchus, and lung " m the 6th and 7th Revisions of ICD but not in the 8th. Therefore, the expected number of deaths from "malignant neoplasm of respiratory system" or from " malignant neoplasm oftrachea, bronchus and lung" was not completely comparable to the corresponding observed number of deaths Classified by the 8th RevisionofICD. However,thediscrepancy must be very small because deaths from " malignant neoplasm of respiratory system, specified as second ary " orfrom " malignant neoplasm of pleura " were rare m Japan. The standardised mortality ratio (SMR) was calculated by observed and expected number of deaths and their differences were examined by the Chi-square test when the expected number of deaths was over 10 or by Poisson distribution when the number was 10 or less. RESULTS Mortality experience ofcohorts doing differentjobs Among copper smelters (cohort 1), significantly increased SMRs were observed for total malignant neoplasms, malignant neoplasm of large intestine, except rectum (abbreviated as colon cancer), malignant neoplasm of respiratory system, and malignant neoplasm of trachea, bronchus, and lung (hereafter abbreviated as lung cancer), as shown in Table in and IV. The increase was most marked for lung cancer, the ratio of observed to expected deaths being as high as 12:1, followed by the ratio of approximately 3:1 for colon cancer. No such increased mortality was observed for malignant neoplasms, of stomach,, liver (primary) and biliary passages and pancreas. Besides one primary malig nant neoplasm of liver, 10 deaths from unspecified malignant neoplasm of liver were seen. These 11 I I table nr OBSERVED and EXPECTED NUMBER OP DEATHS FROM SELECTED CAUSES, WITH STANDARDIZED MORTALITY RATIOS <SMR) * AMONG COHORTS BY JOB, 1MMV7I CkiMofdMtb Drilled Hit No, * Cohort 1 (ooptw Cohort 2 (faro-nfekfil (Ctroteheoprto3rt amahiAt) mental) or wntotawm) p Tuberculosis 010019 Obt. Exp. SMR Malignant neoplasm 140-209 Ota. Exp. SMR Cerebrovascular diseases 430438 Ota. Exp. SMR Heart diseases 393-398,410414, Obs. 420429 Exp. SMR Ulcer of stomach and duodenum 331-333 Ota. Exp. SMR Cirrhosis ofUver 571 Ota. Exp. SMR Accidents E800-E949 Ota. Exp. SMR All other causes Residual Obs. Exp. SMR Total causes 000-E999 Ota. Exp. SMR 13 13.95 82 55 28.82 191 * 26 34.18 76 7 14:90 47 8 4.15 193 2 3.42 38 911.30 SO 37 39.20 94 157 131.92 103 i 1.81 55 1 1.86 54 0 1.36 -- 1 1.01 99 0 0.27 -- I 0.33 303 1 2.32 43 1 3.89 26 6 13.05 46 ii 13.98 79 24 20.80 115 12 23.91 SO* 3 11.08 27* 1 3.03 33 7 2.64 263* 18 10.70 168* 32 31.60 101 108 117.74 92 Cohort < oscoQiyiPi duUVrifi idd projection) 3 4.01 75 3 3.99 75 0 _ 3_ .A89 3 2.18 138 0 0.62 -- 3 0.62 484* 2 3.81 52 8 7.70 104 22 26.82 82 Cohort 3 (started) 6 4.39 131 4 715 56 5 9.26 54 4 4.33 92 0 1.12 -- 1 090 in 0 4.1*4 12 13.38 90 32 44.87 71 'SMR - fotawvoIMpaciM)v 100. -- :8!h Rrtuton of inienutlmia] CUniAution or Dnm. Injuria and Gnnct of Daub. -- 1 affniAcuotat 3% MS; "* WiMnt at 171 level, ' QBSERVl WITH MaUfARAt mfipttjfi Stomach Large intestine (exc< rectum) Liver (primary, seco and unspecified) s biliary passages Liver (primary) and passages Pancreas Respiratory system Trachea, bronchus a lung Alt other malignant neoplasms Total malignant neo * SMR (observed; Vtofftcent el S% leva deaths were sign.il number of deaths,! from malignant no ary and unspecifiet Nodeaths from ski1 No excess mortal vascular diseases, and duodenum, tin causes of death wat hypertensive disea nephritis and nephi were not elevated sponding expected In cohort 2 cot mortality from all than the expected r; mortality for any C; In cohort 3 cons maintenance or trar mortalities were no for accidents, while rrent jobs ignificantly 1 malignant ie intestine, n cancer), r^tem, and is, and lung is shown in marked for Kited deaths he ratio of . No such - malignant i and biliary mary malig- unspecified n. These 11 9 8 12 12 31 70 82 HM PHIL -- LUNG CANCER AMONG COPPER SMELTERS 313 table tv OBSERVED AND EXPECTED NUMBER OP DEATHS PROM SELECTED MALICNANT NEOPLASM. WITH STANDARDIZED MORTALITY RATIOS (SMR) < AMONG COHORTS BY JOB, IMP-1971 Millgftftnt ncootwra - Dauiltf lift No. ' Cohort t (copper UnftHirii) Cohort 2 (tavorttfakit HTWhtftfJ Cohort 1 (transport or molmaitftc*) Cohort 4 (copper or lead ctoctrofrgfr. firltWk odd production) Cohort 5 (chrtdal) Stomach 131 Large intestine (except rectum) 133 Liver (primary, secondary and unspecified) and biliary passages Uver (primary) and biliary passages 133, 136, 197.7,197.8 155, 156 Pancreas ^ . t JT ' Respiratory system 160-163 Trachea, bronchus and lung 162 AU other malignant neoplasms Residual Tout malignant neoplasms 140-209 Obs. Exp. SMR Obs. Exp. SMR Obs. Exp. SMR Obs. Exp.. SMR Obs. Exp. SMR Obs. Exp. SMR Obs. Exp. SMR Obs. Exp. SMR Obs. Exp. SMR 10 14.71 68 3 0.S9 308** 11 3.26 337*" 1 0.72 139 2 0.91 220 29 3.18 912** 29 2.44 1,189** 0 _6..1*7* 53 28.82 191** 0 0.87 -- 0 0.05 -- 1 0.20 500 0 0.04 -- 0 0.06 -- 0 0.18 -- 0 0.13 -- 0 0.50 1 1.86 54 12 10.46 115 1 0.45 222 5 2.34 214 0 0.52 -- o' 0.67 -- 2 2.27 88 2 1.74 113 4 4.61 87 24 20.80 115 i 1.94 52 0 0.10 -- 2 0.44 453 1 0.10 1,000 0 0.13 -- 0 0.39 -- 0 0.29 -- 0 0.99 3 3.99 7$ 0 3.61 . * 0 0.16 -- I 0.79 127 0 0.17 -- 0 0.21 -- 2 0.73 267 1 0.57 173 1 1.63 61 4 7.15 56 SMR - (obswvsd/expued)x 100. -- 'Sth Rerittoa or InunuilliinH CUuillctllon of Otaum, IRIiuln Hid Cum or Daft. ~ niticHn t ** HsbIBuM at IX lunL deaths were significantly more than the expected number of deaths, 3.26, calculated from the-mortality from malignant neoplasm of Uver (primary, second ary and unspecified) and biliary passages (p<0.01). No deaths from skin cancer or leukemia were noted. No excess mortality from tuberculosis, cerebro vascular diseases, heart diseases, ulcer of stomach and duodenum, cirrhosis ofliver, accidents and total causes of death was observed. Similarly, deaths from hypertensive diseases, pneumonia and bronchitis, nephritis and nephrosis, diabetes mellitus and suicide were not elevated as compared with the corre sponding expected number of deaths. In cohort 2 consisting of ferro-nickel smelters, mortality from all causes was considerably lower than the expected rate with no significantly increased mortality for any causes of death examined. In cohort 3 consisting of the workers engaged in maintenance or transportation,significantlyincreased mortalities were noted only for cirrhosis ofliver and for accidents, while significantly lower mortalities for cerebrovascular diseases and heart diseases were observed. An increased mortality from cirrhosis of liver and a decreased mortality from cerebrovascular diseases were noted in cohort 4 consistingof workers engaged' in copper or lead electrolysis or in production of sulfuric add. Cohort 3 consisting ofclerical workersexpericnced a similar mortality to the general population except for malignantneoplasmofthestomach andaccidents, both of which were significantly lower. Mortalityfrom lung cancer among copper smelters by length ofemployment As described, a distinct excess SMR was observed for lung cancer among copper smelters, A possible association of such excess mortality with length of employment was then sought. Significant excess SMRs were observed in all of the cohorts and a definite positive gradient was seen between SMRs and length of employment, as shown in Table V. 314 TOKUDOME AND KURATSUNB . ; table v . OBSERVED AND EXPECTED NUMBER OP DEATHS FROM MALIGNANT NEOPLASM OF TRACHEA, BRONCHUS AND LUNG, WITH STANDARDIZED MORTALITY RATIOS (SMR) AMONO COHORTS OP COPPER SMELTERS BY LENGTH OP EMPLOYMENT, LEVEL OF EXPOSURE, LENGTH OP EMPLOYMENT AND TIME OP EXPOSURE, 1*9,1971 Length of JLovI ofcdtjMvsrv Lflrfth gfnpleiaw u4 thru ofoxpoturt Cohort 11 (ow 20 mu) Cohort 12 (10-19 ymn} Cohort is (1-9 jmn) Cohon 21 <be*vy l) Cohort 22 (medium *) Cohort 21 OHM*} Cohort 11 (IS reus or non, with lSth ]mr completed fcefofv Duunybof It 1949) Cohort 32 (15 jwet* or more* with 13th yew completed during 1949*1971) Cohen as (1-14 WHS, Uweotthmof tune of nponre) Obs. 20 Exp. 1.05 SMR 1,903 *** s 0.68 733** 4 0.71 363** 15 10 1.01 0.80 1,485 ** 1,250** 4 0.63 635 ** 17 0.83 2,048** 5 047 1,064** 3 1.14 614 * 1 Stoteriat by " Pm * Omoawslt" method. cnaMut of afateti, Uut AnM oporatlon. -- * Opsntioa ofconvener, production of by-proAucts fmdnly ansnk trioxhl). operation of Iwl Uut l\sm operation of lend olodrlc Atman aad wmtwwy Aimna fin lead lane*. -- Ttsmiwrt. aampllap and (mam] mejateaann at the copper etaUUna eaetlaa, IIuh Atman aparatlan htdadlap on blendiaaend oottrel operation. -- 4 *i*niftc*nt at IXIatoL Mortalityfrom lung ameer among copper smelters by level ofexposure A significantly increased mortality was noted in ail of the cohorts examined and again a distinct positive gradient could be seen between mortality and level of exposure, ratios of observed to expected mortality as * high as approximately 15, 13 and 6 being noted in ' the heavy, medium and light exposure groups, * respectively (Table V). Mortalityfrom lung cancer among copper smelters by ' length ofemployment and time ofexposure Significantly increased mortality ratios were seen in a31 of .the cohorts 31, 32 and 33 as shown in Table V. SMR of cohort 31 was considerably higher than that of cohort 32, indicating that those exposed for 15 years or more before 1949 experienced a much higher mortality risk than those exposed for a similar duration but with the 15th year completed between 1949 and 1971, Cohort 33 consisting of copper smelters exposed for less than 15 years irrespective of time of exposure also showed a significantly increased mortality as mentioned but the increase was much less marked as compared with cohorts 31 and 32. Furthermore, an interesting fact was observed, namely that 28 of 29 copper smelters who died from lung cancer had been engaged in the smelting work before 1949. byjob, respectively. However, the smelters who were employed for 15 years or more with the 15th year completed during 1949-1971 (cohort 32) did not show any definite mortality gradient associated with their jobs. Table VI also shows that the smelters most heavily exposed to arsenic and other related compounds experienced a strikingly diflferent mortality from lung cancer depending on time ofexposure in spite of their approximately equal duration of exposure. Similar differences in mortality experience were also noted among those moderately exposed and those lightly exposed, but they were less marked. Mortalityfrom colon cancer among copper smelters by degree ofexposure As mentioned, a significant excess mortality was also observed for colon cancer among copper smelters. A possible association of such excess mortality with length of employment, level of exposure and time of exposure was examined but no table vj- OBSERVED AND EXPECTED NUMBER OF DEATHS FROM MALIGNANT NEOPLASM OF TRACHEA. BRONCHUS AND LUNO. WITH STANDARDIZED MORTALITY RATIOS (SMR) AMONG COTTER SMELTERS BY LENGTH OF EMPLOYMENT, TIME OF EXPOSURE AND LEVEL OF EXPOSURE. 1949.1971 Mortalityfrom lung cancer among copper smelters by length of employment, time of exposure and level of exposure As shown in Table VI. the markedly increased mortality from lung cancer among smelters employed for 15 years or more before 1949 (cohort 31) was closely associated wiih their jobs, the mortality risk ratios being .25, 18 and 14 in those heavily exposed, moderately exposed and lightly exposed, as classified Hwvy Medium Light exposure 09tpOUfl exposure Cohort 31 (over t5 years. before 1949) Cohort 32 (over 15 years. 1949-1971) Obs. Exp. SMR Obs. Exp. SMR 10 0.40 2.500** 1 0.18 356 4 0.22 1,818 ** 3 017 1.763 * 3 0.21 1,429** 1 0.13 769 * significant at 1 ! level. significant rdatior very small number Latent period of smelters As shown in Tat had occurred am study period. The 9, 13 and 5 cases 60-fi9 and 70-79 ; age was 42 and tl 61.7 years. It was occurred in 26 wor. The latent period ( of first employme death) ranged fro average. The distri for 10-t9, 20-29, latent period, resp 50 years with an t smelters withheav sponding figures v of 34.1 years in 1 exposure (cohort average of 45.3 ye exposure(cohort 2 be seen in the lenj with levels of expc Cohorts 21,22 ment, being 24.7, average,' respect! similar analysis w employment. Nine heavy exposure smelters in their length of latent ) deaths occurring The latent periods 34 to 40 years wr the corresponding to 47 years, with again, no signifies exposure could be Our previous i records from 19 mortality from lu in the refinery. Oi by extending obs forwards up to 1 definitely increas (SMR . 1,189) i clear dosc-responwell supports the 1 ,CHa, bronchus per smelters by F EXPOSURE. I**S-I7I and ilim ofexposure rt 32 ears ore* K year 1 during 970 Cohort (1*14 years, irrespective ef lime of exposure) s 7.47 4** 3 1.14 614** of coqwtv. oroSictlon of vtwaHor foniK* for tend inclti'llne ore Meodinf end he smelters who were a with the i 5th year cohort 32) did not client associated with ->melters most heavily related compounds xent mortality from sure in spite of . of exposure, w^erience were also / exposed and those ess marked. mg copper smelters by excess mortality was icer among copper ion of such excess nployment, level of was-examined but no UMBER OF DEATHS SM OF TRACHEA. H STANDARDIZED CTO COPPER SMELTERS TIME OF EXPOSURE IRE. IMS-1971 Median) UsM e ostpaaco expowre 43 10 0.22 0.21 1318** 1,429** 3I 8 0.17 0.13 1,765** 769 LUNO CANCER AMONO COPPER SMELTERS 313 significant relationship was seen at all due to the very smalt number of deaths occurring. Latent period of lung cancer seen among copper smellers As shown in Table IV, 29 deaths from lung cancer had occurred, among copper smelters during the study period. The distribution of age at death was 2, 9, 13 and 5 cases Tor the age-classes 40-49, 50-59, 60-69 and 70-79 years, respectively. The youngest age was 42 and the oldest 76. with an average of 61.7 years. It was also noted that these deaths had occurred in 26 workers after they had left the refinery. The latent period (designated as period between time of first employment as copper smelter and time of death) ranged from 13 to 50 years, 37.6 years on average. The distribution was 1,2,15.8 and 3 cases foe-10-19. 20-29, 30-39, 40-49 and 50-59 years of latent period, respectively. It also ranged from 31 to 50 years with an average of 37.9 years in 15 copper smelters with heavy exposure (cohort 21). The corre sponding figures were 13 to 47 years with an average of 34.1 years in 10 coppers smelters with medium exposure (cohort 22) and 35 to SO years with an average of 45.3 years in 4 copper smelters with light exposure(cohort 23). Thus, no definite gradient could be seen in the length of latent period in association with levels of exposure. Cohorts-21, 22 and 23 differed in age on employ ment, being 24.7, 25.3 and 19.0 years of age on average, respectively. Therefore, an additional similar analysis was made by adjusting for age on employment. Nine deaths in the workers undergoing heavy exposure who were employed as- copper smelters in their twenties were compared for the length of latent period with the corresponding 7 deaths occurring in those with moderate exposure. The tatent periods of the former group ranged from 34 to 40 years with an average of 36.8 years, while the corresponding figures of the latter group were 25 to 47 years, with an average of 37.6 years. Thus, again, no significant difference attributable to level of exposure could be seen in the length of latent period. DISCUSSION Our previous case-control study with mortality records from 196T to 1969 revealed an excess mortality from lung cancer among copper smelters in the refinery. Our present study confirmed this fact by extending observations backwards to 1949 and forwards up to 1971. It demonstrated not only a definitely increased mortality from lung cancer (SMR -- 1,189) among copper smelters but also a clear dose-response relationship. Our study therefore well supports the observations by Lee and Fraumeni (1969) and Ott et at,, (1974) who reported an excess mortality from respiratory cancer among copper smelters and workers exposed to arsenical* by similar cohort .analysis to ours. The refinery is known to have started its copper smelting operations in 1916. The operations con sisted of sintering by the * Pot " method, followed by processings with blast furnaces and " Mabuki '* hearths which are oldJapanese traditional converters. As shown in Figure I. the " Pot" method was discarded in 1945, replaced by the " Ctecnawalt " method which had beenintroduced in 1936. This new sintering method was also discontinued in 1953. The blast furnaces were used until quite recently and replaced by a modern type offlash smelting furnace, while the " Mabuki" hearths were replaced by converters in 1930. Figure t also shows that a drastic decline in copper production as well' os in the amount of arsenic contained .in the processed ores occurred after the end of World War II. It is alsoknown that a fairly large amount of " Kmkasekl * ore from Taiwan,which is particularly rich in arsenic was used before World War ff. Therefore, it is reasonable to consider that the pre-war exposure was far worse than the recent one. These historical facts seem to be well associated with our findings. Although our study docs not specify the causal* agent or agents, we consider that arsenic compounds and sulfur dioxide to which copper smelters must have been heavily exposed are probably responsible for the excess mortality observed. Other carcinogens such as polycyclic aromatic hydrocarbons might also be involved because the sintering and smelting processes formerly used must have released a fairly large amount of these compounds. Since smoking is generally accepted as one of the most dominant causes of lung cancer and our previous case-control study showed that lOoutofll copper smelters who died of lung cancer had been smokers, a causative role of smoking cannot be overlooked for. the increased mortality from lung, cancer among copper smelters. It does not seem probable, however, that such an increasedrisk can be explained exclusively by this factor. This is because their smoking habits cannot be considered to be unusual compared with those of other workers. According to Segi et al. (1965) andSegi and Kuriham (1972), the mortality from lung cancer among males in Ooita prefecture where the refinery is located was very dose to the national average. Therefore, the excess mortality from lung cancer seen among the copper smelters cannot be attributed to any local peculiarity in cancer mortality either. Besides lung cancer, a significant excess mortality was noted for malignant neoplasm of liver (primary, secondary and unspecified) and biliary passages 4 TOKUDOME and KURAT8UNS AflffMffK n! enwornm Amowit *4 pnotMnd IMiic TcnAArth TWMoinh 50,000 ' 500 40.000 400 50.000 300 could explain the obs the specific types of workers and the ot ships. ACKt This work was su] the Ministry of Li Yssuhiko Baba, fa Industrial Health; t UNE ETL 20.000 200 10.000 I( 0* "Po"irtfinng 0 Bta&t hmuu FWi viwNiq tamm "Mitokr im(4) CUwtfftm ' Let auteurs ant marts causes de die thioriquc de diets, a iti campari au narnb tation significative d ct da cfiton, mats pa. peau, nlpar leucemh ividettce mt rapport, nette surmortaliti pi indent considiris Ct frlttage ou aux hauls est en moyenne de 37 tatis parml les fond travalllant d laprodm Figure 1 Long-term trends of the annual amount of copper ores processed C*------*), the amounts of arsenic in these (o----6) and in " KJnkaseki " ore (#--), and of the smelting operations in tire refinery, during 1927-1972. This figure urea provided by the Ministry of Labor, Diath Statistics or Report or Vital 3tai Health and Welfare, Tt Doll, R., and Hill, A among the copper smelters. Since the majority of these deaths were unspecified malignant neoplasms of liver without adequate diagnostic validity, any further reference can hardly be made with confidence. to the considerably long latent period. Another important aspect of latent period is its possible association with level of exposure. It was expected that the latent period of lung cancer seen among death in relation to mortality of British d. 095S). DruCKRZY, H., Quant gourds. VICC Mtmogr It seems highly desirable that similar epidemiologic copper smelters would be in inverse proportion to studies with particular attention to thissite of cancer level of exposure, in view of the observations by Lee trntc hatards from drtt Springer-Veriag, New 1 should be made at other copper refineries. and Fraumeni {1969) and of the results from the International Statu Injuries and Causes oi r It seems noteworthy that no deaths from skin extensive animal experiments conducted by Druckrey World Health Organist cancercould beseen among all theworkersexamined. Since arsenic is known to cause skin cancer in man and some of the workers in the refinery suffered from (1967), but this could not be demonstrated in our study. Our results were practically the same as those obtained by Ott et at., (1974). Kurativns, M.,Tokui Tokumtsu, Y., Hayai lung cancer among o 552-358 (1974). , dermatitis and pigmentation due to a heavy exposure Finally, the fact that, as many as 971 workers were Lei, A, M., and Fraum to arsenic compounds particularly arsenic trioxide, excluded from the present analysis due to the fact some excess mortality from skin cancer was expected. that their Honsekl was outside Ooita prefecture, Our negative result might be due to the well-known must be considered. If they had had any unusual, favorable prognosis of this type of cancer. either high or low, experience in incidence or Our previous study indicated the necessity of mortality from cancer, their elimination must have establishing a thorough life-long health care system for copper smelters. The present study reconfirmed it because as already described, 26 of 29 lung cancer deaths seen among copper smelters bad occurred after they left the refinery. This may be partially due seriously distorted the results. However, it is hard to imagine that the Honseki is firmly connected with cancer risk, because it is decided primarily for reghnative purposes. Even if It is in some way related to cancer, it is still difficult to see how it r 123** 1965 19701972 CMento Mints of arsenic in these r. during 1927-1972. This uent period Another period is its possible osure. It was expected ng cancer seen among . inverse proportion to he observations by Lee if the results from the conducted by Druckrey c demonstrated in our ically the same as those ny as 971 workers were nalysis due to the fact side Ooita prefecture, had had any unusual, ience in incidence or elimination must have . However, it is hard to firmly connected with decided primarily for if , in some way LUNO CANCER AMONG COPPER SMELTERS 317 couldexplain theobserved increase in mortality from the specific types of cancer among the very specific workers and the observed dose-response relation ships. ACKNOWLEDGEMENTS This work was supported in part, by a grant from the Ministry of Labor. We are grateful to Dr. Yasuhiko Baba, former Senior Expert Officer of Industrial Health; the Industrial Health Division, Ministry of Labor staffofthe Ooita LaborStandards Office; and to the-staff of the refinery and Dr. Seiji (noue. hospital director of the refinery, for their co operation. We are also grateful to Dr. Fusa Ueda of the Ministry of Health and Welfare for her technical advice in regard to classification of underlying causes ofdeath, to Mr. Masato Ikeda for hi* advice on statistics and staff, Mrs. Kayo Tsuno and Miss Motoko Taniwaki or this department for their general assistance. une Etude de la mortality par cancer et des autres causes DE Dficfes CHEZ LES EMPLOYES D'UNE FONDERIE Lea auteurs out efftttui uat etude prospective non chronologico-concordantedes diets par cancer et des autres causes de diets chez 2673 hammes ayant travatUi dans une fondcrie entre 1949 et 1971. Le nombre tMorique de diets, ditermini d'aprts les taux de mortaliti par dge et par cause Calcutta pour its Japonah, a etc comport tut nombre de diets observe. Dans une cohorts de 839fondeurs de culvre, an a constati tore augmen tation significative de la mortaliti par cancers du povmon (taux comparattf de mortaliti (SMR) -> 1189) et du cdlon, malt pas par cancers de I'estamac, du foie (prtniitlf) et des vales btlialres, du panertas et de la peau, nt par leucimie, tuberetdose, maladies ciribrovascuiatres, cardiopathies ou ctrrhase dufoie. On a mis en tvldenee un rapport dosc-riponse entre la mortaliti par cancers du poumon et le degri d'exposition. Une'tris nette surmortallti par cancers du poumon (SMR -- 2500) a Hi observte parml les fondeurs de culvre qut italent considiris comme les plus fortement exposts A {'arsenic ou Us ouvriers qut avaient iti emptayis au frittage ou aux haulsfourneauxpendant au mains 13 ansavant 1949. La piriode de latenee du cancer putmonalre est en moyenne de 37,6 ans et n'est pas Hie au degri tfexposition. Sur les29 diets par cancer dupoumon camtatis parml Us fondeurs de culvre, 26 St son! produits aprts lew depart de i'uslne. Pour d'autres employis travadlant A laproduction ou dans Us bureaux, la surmortalltipar Us divers typesde cancer n'estpassignificative. REFERENCES Death Statistics or Mauonant Neoplasms, A Special Report op Vital Statistics, Japan, pp. 6-67, Ministry of Health and Welitoe, Tokyo (1973), Doll, R., and Hill, A. B., Lung cancer and other causes of death la relation to smoking. A second report on the mortality of British docton. Brit. mod. If, 1071-1081 (1936). Druckriv, H., Quantitative asserts In chemical carcino genesis. Vice Monograph Series, VoL 7. Potential carcina- Srnfc haiards from Sets (evaluation of risks), pp. d0-78, pringer-Verlatg, New York (1967). Inurnational Statistical CiasoticattOn op Disiasis, Iniumis and Causes or Dsatil 1965 Revision, Vol. 1 and 2, World Health Organization, Geneva (1967). Kuratsunc, M.,Toxubom, S., Skikarusa.T., Yojhida. MTokummsv. Y,, Hayano, T., and Siita, M., Occupational lung cancer among copper imeitei*. lot. J. Cancer, 13, 532-55* (1974). Let, A. M., and Fradmuo, 3. F.. Ja., Arsenic and respiratory cancer In man; an occupational study. S. not. Cancer hist., 42,1045-1052 (1969). UuTNreLD, A. M7 Pedersen, R, andDowo, J. E., Cancer epidemiology: methods of study, pp. SO-83, Johns Hopkins Press, Baltimore (1967). MacMahon, B.. and Puoh, T. F.. Epidemiology, principle* and methods, pp. 43-44, Uttie. Brown and Co., Boston (1970). Ott, M. G., Howsr, B. B.,and Gordon, H. L-, Respiratory cancer and occupational exposure to axsenkals. Arch, enrtran. Hith, 29,250-255 (1974). Ssoi, M,, Kurihaka, M, and MATSUYAMA. T., Cancer mortality In Japan (1999-1962), p. 122, Dept, of Public Health, Tohoku Univ. School or Med-, Sendai (1965). Seoi, M., and Kurihara, M-, Age-adjusted death rotes and specific death rates from cancer by prefecture h Japan, 1939-1967, p. 15, Segi Institute of Cancer Epidemiology. Nagoya (1972). Vital Statistics, 1971, Japan, Vol. 1, pp. 110-143, Ministry of Health and Welfare, Tokyo (1973). '