Document OE1v9N73b0bRVMaEv2ZjLJY5Q

*> , .. I .. .. . .. 9 --.9.r. , ,* An Epidemiologic Study of'Mortality within a Cohort of Rubber Workers, A. 1. McMichaeI, M.D.; R. Spirtas, h4.S.; 1. 1. Kupper, Ph.D. 1.964-72 Interest i n the health status of workers in the U.S. rubber industry was brought to a new focus in 1970. In that year, the Occupational Safety and Health Act (OSH.4) was passed by Congress, and in ihe same year the United Rubber \Vorkers Union negotiated with the six malor U.S.rubber companies for a comprrclensive study of work-relaid health nrobkms. An earlier and longstanding interest i n the hral:h of rubber-workers stemmed irom knowledge of the physiological ef- -k c t j (of:en acute) of chemical com- oocnds used i n the industry.! 5 The in- dustry uses many chemicals. which "run the gamut of highly hazardous to innocuwc".l The rapid evolution of rubt w proc1;lction technology entails a cont i n u r u change in chemicals used. and in rt-action by-products produced.4 4nirJng the compounds n o w or previously used are some proven car- -cinogens and certain chemicals currently under investigation as carcinosens.6 11 Certain morbidity and mortality trends among rubber-workers have been sug- gested in earlier studies. In Britain, in the. 1950's. Parkes conducted an industrywide study of sickness absentee'ism. and found difierences in the morbidity experience (ior example. derma:itis) between production workers and office staff; -further differences were observed among the various production groups.12 In 1929-31. the British Registrar General reponed that rubber workers showed a 10% total mortality excess. compared with national mortality data, and that the excess derived from canccr d c d t h ~ . ~A] similar mortality cxccss had b w n record- ed in 1920-22.1' In the United States, a national study of occupational disahility, based on Social Security data from the period 1959-62. rhowed that skilled and semiskilled employees i n the rubber indu5tr). had experienced a higher morbidit). ratio for malignant neoplasms of the respiraton, system. chronic rheumatic &&t disease. crnphvscma. and arthritis. when comparcd with other industrial workers.15 The U.S. Department of Health, Education and IVcJiarce h d previously published figurer showing that, compared to male workers in all manufacturing industries. male rubber- workers age 20 to 64. i n 1950, had a 14% overall deficit of death5 (6 Howver. there was an 8%excess death rate from malignant neoplasms (especially of the bowel, 30%; respirator). system. lSe0; ..and the lymphatic and haernatopoirtic.., *- d. tissues, 54%). chronic rheumatic heart disease (10';). and arteriosclerotic heartd disease ( 1 5 % ) . These data were necessarily incomplete, being b s d on the "Usual Occupation" as recordcd on . the death ceniiicate Xlore recently, a group life insurance stud) by the U 5. Society of Actuaries, covering 1955-64. found an age-standardized mortalitr ' ratio of 89 (1.e. 11% dciicl:) for worker< in rubber manufacturiiip 1- hlancuso, in 1947. reported an ' eleiatcd proportion of deaths due to cancer for workers in the rubber and plastics industry in Ohio 18 In particular. he foiind a proportional mortalit) rxcess 1- from c a n c w r of thc- respiratory. genitourinar,, and ccwtral nrruouc sbstcmy. hlancuro also rcportrd a . rdativc e%cesrof ncop' :cms of the ccn- Ira1 W f b O U I \ y s t m arriisng Ohio m d l c - rcwclc.nts. and dcmomtratcd that :he CCI\ WJS locatt4 i n Summit Coun:y (tk industty Tl1l\ pt-\cnt qud\ i;wd an lii~:fmait p r q M < t i v < * tlt-ign IhLs I t h c l r l (abr pqwhiim at ri*k) CQmprrS(q1 all t h ' h u r l \ (I c. fat t t ~ y ) male wnoloyc.e.. agv 40 to 81 JI 1 I frl. within a nialor in Akron, Oho. Workins from cr>mj'dny :i3n io. d i c e +ut?aid tlwt deatn rates i9r rnjr,;. c a u s e (various mahgnant neolilasms. and direarcs oi the circulat on. respir.ttory. digestive; and gelnitourinao zystems) were high& in !hc Co-poundirrfi. milling. cdlcndering. and c w n g areas tbm in both the otkv pr4ut:ion areas and in Ohio State. i r i xcnCrai.f For r=xamr,lr..hc reported iive- crlculating inc!dencc. rates .4llllough a number of studies ha\e reponed txcessr\r respiratory cancer mortalit) in othw industrig (e.6. the asbwcn, chromate. arsenic, uraniuni mining. and steel ioundr).: industries), the c\i&nce regarding respiralow cancer deaths i n U.S rubber-workers ts meager. Tlie above-mentioned U.S. records i f pasrot!. twnsiotis. dvath claims. wcvance a\vJrds. t!amiwtrr* tn.- t w m hourly and 5d:rricd btatui. and employment cots. a u J h o r t of Mi5 male r u M m workcrs was identitid. dnd c x tensively crcns-chcc~x*d.(Female cm- ployces. comprising only lO% ni thc l l l i 6 1 population, \\we not :nclut!cd in the study.) ' told excess o! malignant neoplasnrs d -the stomach amo.rg rctired rubber :s.orks. comjsred :o Ohio State mor- tdli:?. data. Mancuro also observed a government report16 of an 18% excess respiratory cancer mortality. and htancuso's finding of an excess death rate among compounding, milling and calen- During ttw full nine-year period. I%+ 72. 1783 rife insurance death claims were paid on behdli of deaths occurring .within thc cohort The completeness ril . .. considerable b-ariatjon among the dering workers. suggest there may be a the deaths enumeration was establi5hed ..aricju: productioa departm*ents for respiratorv cancer excess within the rub- by cross-checking internally against dca!h rates from certain specific cause ber indusin. One outbreak of work- \ w i & s company and insurance files. r,i death, thereb? emphasizing the related respiratory illness has been and. externall); against the files oi the lwssible etiologic infuence of localized iactcrrs in the \ w r k environment. He r c r m r k d upon tht. "dilution effect" which can occur whcn such a ctudy documented in the US. rubber industry, in 1961. in which, at its peak, 312 (14%) of thc plant's 2.225 workers were affected, subsequently leaving 32 (1.4'0) US. lnterrul Revenue SCrsice. (This latter crosscheck entailed a random group of 250 members of. the cohon, a p prently alivc Dee. 31, 1972. The initial makes no distinction bawcen different pcrmanentlv disabled 14 tax-return search showed 221 (68%) qroups of workers experiwcing different From ~hesevarious studies, the death were alive, and the other 29 were con- ewpxures within an ifidustry. (Alancuso's o w n comparative data for different groups of korkers, h w mer. is hamered by the small numbers of percons in the various exposure group:.) rate from all causes within the U.S. rub- firmed as alive by subsequent personal ber industry appears not to be elevated follmv-up.) , -- _ _significanth. i f at all. However. the D-_ -e_-a.th ce..rtificates w_.e_re obtained from literature su-g-gests an excessive mor- the company's insurance carricr for all tality from certain malignant neopla.smF but'nine (O.S%) of these 1783 deaths. %Aa!icuso and Brennan subsequently (bladder. lymphatic and haematopoietic LVithin this cohort, the ijnl) b C d h extended the above cohort study through tissues, stomach, large bowel, and ineligible for life insurance ' brwcfits 1%3. and reported findings suggestive of possibly respiratory system and central would have been those among ex. a n cxcess mortality due to cancer of the nen'ous system). and from some types employees who quit from act#\p ern- gall bladder. hilc ducts and salivary of cardiovascular: disease. C1ands.l" The findings for thete relatively ploymenl during the 1964-71 perird. (Eighty-four such exits occurred. mtXly rare cancers in rubber-workers tuvc not Study Design of men in thc 40 to 49 age-rangc. A ye: been rep!ica:d elsewhere. Thc roults presented in this article are reccnt search of State Health DepJrtnicnt In 1954. Caw rnd Hosker had report- from an initial mortality study carried out records disclosid only ~ W J tlcathr c d an ewes\ of [>!adder tumor\ in British within the research prosram negotiated among this group of quitters. For prac- ru1;Ser workers e Subequtntly. despite between thc United Rubber M'orkers tical rcwarch purprnez, tIiitrviort*. thr rvmoval of the prcsumcd carcinogen (an and the six major US. rubber. com- 1783 conipany-recorrkd dcath flaifnc ant:oridant) irom the production paniet. When a series of epidemiologic represent the total niortalitv u ithin this process, eaccsc occurrence of bladdcr studies is initiated within a n industrial cohort for the pcriod 19f4-72.1 r tur:?ors i n Bri1ia.h rubt,wr-v,oikcrs was setting. an historical mortality study i s The iniormation ruIatinA to c a u w of [ rr-pnned by \'c)s,m and, mort? recently, commonly done iirst. for two reasons. death on the '&ath crrtifiralcs w r $ !Y! .\delstt:in." ';.cy5 identiftcd ik areas fir%tl\.. pccexisting data on death is c d c d by a Sation.11 Ccntcr k w Iioalth in tirc p1.w: of greatest r d . (mixing, usuall*. Jvailahlc both rratlilv and Stati\tic< no,oIirjkt. using thc EiRli!li !w;r.:cfiance, c l c m s and rli.iutch). In s\~tcmaIicJlfp.dlotlg rvith. d d l l On IhC Rcvi\ion of thc Internrtional i ti- U.5 ,firm (I idivrci* of a thIt!cv can- popula:ion at r i d . Secondly. cxccss mor- Clrssrftcatinn of D~WJ,PTSh.v ri*cul:ant t.1.r CXCVSI a m r q ritl)twr \vi,i;i*rs is still t a l ~ ~rr)prcsr*ntr a uiiiqut4y swiou- t y p e cauw rd dvarh coding (I! thc nii>rt,ilit\ h ~ ; .A piI,;:l.ition-fwtc~l I w v o n t r d o(. h r j l l h problrm such d slu!!v Can &la \vdS t t k r t h v dircctlv ciimlhirJMc * !I,.:? of l h : k b r a ,mct*r i n . \ b * a c liuwtt*, di*ic.!trtinc* tvIwtIwr t l w r c is ,111 c*xce>s to the* rirrwnt U 5 riattr,twl nitJr:,i!:!\ t < t : r . i t d a I I ~ I I Wriel. of 1.G kir r i ~ ~ I t . c tlr,uh rati: froin dny or ,ill c a u w . arid C!JIJ.,is (otn;rilid t j t . tt1p Sdt1lJll.d Cc*n!rr e,c:hrvi in t!w * * . t - h v iiiihpttw ?: Ctiirj wli(-:ty.i .in\. such i*xcc*- i s ~ w i ant d . ia,r II t u : t I i St.iti\tlc.. ---vp,:tr.d iiii J .c.ritr o f P t i l h d ~ l t ~witti wt.iin vxptnure .ili*d5 within t h ~ lhi- 1irb.t J ~ J W 01 rhi* n!ortdl:l? -1ud) ... ... .... ..... ... ... .i.............. ..... .. .. -4s w 852 101 SI0 a0 5s 64 2111 2166 ?&S 2310 ... C a l c u l a t i o n of Expected Mortality Figurer Tk:? p:tmary technique used for a j i ~ j i n gthe signiiicance of the ob5e-fd monalit). \\i!hin the cohort was to ccTpa:e it to the mortality expected \\ !thin ;hat cohort. had the death rates at each age been the same as in a selected "r:anda:d*' popula!ion. The ratio of ob- senrd :a expect4 deaths i s expressed as the 5:andardired .\tonality Ratio (SblR): S\!R = Observed deaths x 100 Expected deaths (14%) among a 20% random sam$r 0:' the cohort suggested that the racial composition of the cohort \\'a* similar to that of the 1968 U.S. nattr4nal male population within lite same agc.-rjngt. (blacks 100~). Results *e-specific aII:causes death rates . --were calcu_lat_ed-_.f.o.-r_the--c.ohort f 0 r . h . &rid - 19-64--7.2-.T..he._ra_-te_s a_re erlpre54 persoa-year, at iisk.'to avoid distortion due to the ageing of the cohort ___t-h.roug-h-o.ut the nine-year 0bservafio.n p e r i d . (lhe unit oi de3ominator . a t risl" observation, used in c.dculcting age-specific rates, is not the complete nine-year cxprricncc of an indni d c ~ l . buf is instead that pnrtion of tiis tordl "at risk" experience that falls within the specified age-mnp ( e g 45 54). Clturly. rnoit ind.~duaIsbb 111 ccmtribtife ... ... 110 :a i2 :0 2 0 33 6 2 1 L- 11 . I3 211 zs 4 13 14 20 1035 101 1073 103 5s I t ) * 1 3 I21 3.3 0 6.4 2 4 5 t 12 4.1 141 . 2.S 8C 3.5 i 9 L 24 -2N' I L5 315f 8.3 I57 rn. 103 10.6 , 15 1.l 213 13.6 % 11.1 78 YC I' si 9 1b I6 43 731 182 34 u 35 IO 9.5 105 ' 69 US 72 . 12.5 141.4 1765 22.1 st.4 286 among steelworkers aged 55 to 64. thc 45 to @ageI-range portion of the ruhberworker cohort experienced lower agespecific death rates than the comparison populations. This table indicates that there was no overall mortality excess within this population of rubber-workers during the period 1964 to 1972. In Table 2, the comparison with-the 1968 U.S. national male population morta!ity data is presentd in monr detail. These national data wcrc u w l as the "standard in :he cdlculattcin o f the number of expecld deaths within the rohofi. (This calculation makcas the. ... assumplion that, i f tte rubbcr industry is ficc of life-sh~teninghjzard. thcn OW might cxpctt that tht? Jgc:-slx*rific mortality ra!cr. w 4 l i i r i ttc naticm.!! m3lc pi~phtirm~ ~ i ~ i&i loi l tu*:**Jppliml within the rubber-worker cohort.) For each 5-year age-range, the ratio of the actual observed deaths to the exp c t e d deaths expresses the age-specific mortality ratio. The age-standardized mortality ratio (SUR) is calculated from the ratio of the total observed dea:hs for the full age-range to the sum of the agespecific expected deaths for the full agerange. The SVR for the age-range 40-84 was 99.1; for the-active working agerange 40-64the S M R was 93.2: and for ages 65-84 the S U R was 101.6. Using 1968 Ohio Stalc Mortality data (males) -as t h r standard. tlx! SS4Rs wcre very close to the above valucs 93.4 for the 40-84 age-range. 94.5 lor t t l e 40 to M aRc-ran!;i*, and 1018 for thr! age-rangc! 05.04. Usir:g tlw 5anx proc(&re prescntcd . .... ... .... ... rchiexe sta!.s:.ciz .'sig- 'KZXO" t. chance a;one ) Tw higks: S I R S M Y * for malignant neoplasms ofthe I:C.~L:T and o i the Ivmo'?atic a': haemaiopoietic $\stems :? each c a s ,i!-C -* the S v R was h i p : io! t k . act.ke" age- > . , /(arange !o 6 4 t tfran io- the kif a?- rmge (40 to 841.Fo: leuiema ded:% :? -the 40 to 6l age-range, the j V R n a s 3'5 i.e. approxima!c:y a threefold exces Monality from mdignr.-rt neop!asTi of the large intestine and the pros:r:a i was higher than expected. Haw?-.e'. malignant neoplasms of t h mpiraic? system and the bladder caused i e w c deaths than expected. For dea:hs iror? alf malignancies, the S.MR was SI@: elevated (103} for the iull a)Se-ranqt -Relatively, it wa5 mote elevate for ;be 40 io 6-1 age-range i.e. :he a'!- malignancies S i t R (100) was 85; hisher than the all-causes S.MR ''1' Tab!e 3 also present! thc S V R z *'Y . eight non-neoplasm causes oi de;:% There was a mode:ate etcess m o ~ sd from diabMes mellitus. a slihht dci.cct from chronic respinton di9eare It*oq- chitis. asthma. and empL3ses.a). atd some SUggC5tiOS of excess mozal.:? fro- liver cirrhosis among the older s q - ~ ? t of the cohort Deaths from ccciaec.:al and violent causcs \\e-e sig~.iic6-.::\ IxIo\\ the expected t r r m k w k ' e deaths from cu.cide \cc:c close tF.0 c i - . pcxted *;slue. With respect to t h ~cw?io\a$cL'm rys!m. thc number 0; d e a t h fro- :-- Icriosclcrosis- b\d5 coni!dcrzt!; t. ;-e? than orpcctcd ( H a s e e * . this d,ig?-.i' c crtq;n;ory reprc*seE:r a rela:i\C'. + ~ i ' : panion oi ca:<m\isrulr* I! 5 ~ 2 a' -~2 9 -_.- -- --- -_ - i- - e J ...*U'Wt as the catqytr. io!.# An+c1 micccllaneo~s" carrjirr; J.r e..J:i. tend tu be l u m m I i\,:h:q !c.. si 11, .............. age-range, mortaht~irqm 1.8 !*.inir bwt ........ disease \vas milJI, ai*<^, wtaiiw 110 the aII-crum WR t t 8 # ~ 4 - . ~ : . :hr*re \%a- a 54% excws oi death; trom th- caucc in the age-rangr u) t o ~I.C Simrlarl?. thee was a 3% CXCCII d clraths from Cerebrovascular disedsr in the age-range 40 to 54. On thc a6sumption that thc workinR en\ ircmment varies between different iub!ler pliints (because of diftmences in chemicals and p ~ e s s e s used. dif- ierenccr in environmental control procedures. and differerpnces in the local environment), it was decided to com- pare mmality data from five other smaller rubber manufacturing plants \\itinn this same complny. These five P!d?.!r are located in widdy differing ;art< of the U.S. For each plant, mortality d a t ~equivalent tp that analyzed above in the Akron cohort study WJS available. However. the populations at risk for each plant we:e not determined. Table 4 presents results of the proportional mor- talitr analysis, for white males, of the death certificate data from Akron and from these five other plants Lisin;: 1968 proportional mortality data ior white males as the standard. age- standardized proportional mortality rztioj (SP.\\R) were calculatcd for select- ed ma!ignanl ner~plasm deaths. (By rompwin;: with Table 3, it can be seen that the Akron SPhlR's afford a good ap' p:oximalion to the respective cause specifir Akron SMRs. The theoretic basis ior this approximation is discussed else\\ hwe.3 Note, however. that the S V R was calculated for the total male cohm, with an estimated 14% minority of blacks: whereas the SPMR was calculated for white males only.) Far deaths due to malignant neoplasm .I o i the s!omach, the prostate. and the v lymphatic and haematopoietic system. I * \ V thc proportional mortality excess at the five other plants w a s similiir to that at the Akron plant. However, for deaths due 10 malignant neopldcm of the rc$pi:atnry system, the bladder, and the central nwvous 5yStem. t h S~I%R for the five rJthrv plants combinerl was clcarly ht+r ow-rall than for the Akron plant. (The con~itl+rablevariation i n the figures ..... fur rhc. irdivirfua/ plant3 mav rcllcct the small niimhcrs u s t d in t h r~alrulatiom, r.i:h-r than a true rlifft-re~irc*in cause- specific mortalit); risks among those smaller plants.) Discussion The finding that the death rate from all causes within this cohort of rubberworkers was larver than expected. from general population mortality data. needs qualification on two counts. On the one hand, t he we1I-documented socioeconomic class gradient of okerall mortality might suggcst a higher mortrlitv within a blue-collar population than in the general population. But on the other hand. there is a strong selection process at play, wherein, to be employable in an industrial workforce. an individual must be relatively healthy and active. This selection factor acts to prducc! a "healthy workereffect". such that. in an industv free d significant life-shortenin8 hazards. death rates within the workforce in question will be lws than in the general population. lndividuiils whwe health does not mcct thc requirements for thc. specific industry do not enter that indu$tr),:hose whose Iwdth dc.cerioratt.c hclow t h a ~Irr.14do not rtnrrin in ~hti.n. dus!ry. Findings in other industries. free of serious life-threafening hazard. in. dicate an all-causn SVR of about Ki to 85.25 Clearly. this selxtiun prows> onl) directly affect< active workers. This 8s demonstrated in thi5 study b, the tinding o! an S U R of 101 b r t k p t . mirement 65 to 0t age-range. and an SVR d 93 for the 40 t o 64 age-range This latter S I R . by compJr!$on \\ith the atmementioned value. of Fd to 85. suggests a small ewes* mortnlitv among that portion of the rubber \vorhvr cohort in the active vmplovment agc-range Assuming tbrt tk " h d t h v worker" selection procc.55 a190 impinges in- directly upon the 65 to 8.1 agc-range, t'ne E\lR of 101 susficsts swne small excess mortalitv a; thi- age W p a ! t c m d cauzs-specific S\Wr (Table 3) is. to a mmltvatr ~ x t ~ cton. . sistcnt wi:h carlicr rcportrd mmtality datd for r r r t h r \\orher$. &atti$ wdhin this cohort wertb exrc4ve for malignant nculda%mhof t k 9 stoiiiarh, large ttmvcl. a i d thr* lymphatic m c l hacnirtolmietic cvstwn-. and for rvrtain t y p tit C ~ I c l i t n ~ w i i I . l r dl*CJ-t* in nlitlillv age f o r ---.-- - "t L.... ....-.. ... .. .... .... ......... ....... . -....-. .. .... ... .. i. ! .. ..-.... ............. .......-........... .../..r. :. , .:...... _... -t -- I cancer i s little d J t d to class This gradient could accoun1for a part d t k -vous rystem. This finding gives sonu! support to Sbnruso's conclusiom although the om? plant (Akron) that fell within his study d m i n showed a marked deficit in this sfudy. With respect to cardiovascular 11101tality, overall there was only a slight excess witlrin this cohon. Haw-. then appears to be a considerable increase in risk, for middle-aged rubber worken. of death from ischemic k a r t disease or c d r o - v a s c u l a r disease. Sqch deaths could result from hypenemion and/or vascular deterioration occurring in response to exposure to chemicals present in certain work areas. Carbon disulfide and carbon monoxide. for example, have been implicated elsewhere in cardiovascular mona1ity.u 3) However, until the second phase ad thirstuay is com$aed. a m c - my br&e-easrsumwdorkers is compaaretd&, igt -c-acn&no&t specific mortality excesses are at- tributable to work-environment exposure within the rubber industry. Such an ex- -cess could be a spurious (secondary) association perhaps the cohort d rub bcr workers was a "selected" population with respect to some important etiologic factor (e.g. dietary habits and stomach cancer), or perhaps all persons living in Akron have been exposed to an en- vironmental factor that increases the mortality from some specific cause above thal of the general population. However, the supporfive proportional mortality findings from the five other plants make this kind of explanation __ __unlikely. But. until it is-d-e.-m---o.n.stratedtlyt sp-e-.ci-f.i_c exposures or jobs within the'in. . -durt.r.y. are..assdciated with eXce55 deaihs. one can only suspect, rather'than con- clude with confidence, that working in from malignant neoplasms of pancreas and gall Madder.10 Some of Mancuso's evidence of excess brain neoplasm deaths in rubber workcrsn Jlrefers to certain jobs within the r u b industry. ie6tails an incrased risk of dying fro--m ~ spe. _cif.ic c.a_u..ie..s...sw U.U.'C t . "tumors" rather than exclusively to /I .malignant neoplasms. (Because of thc relative inaccessibility d the cranial cavity to tissue diagnosis. many death Summary The population at risk for this historical cohort study c o m p r i d hourly certificates for brain tumor deaths con- (i.e. factow) active and retired em- tain no information about the malignant ployees of a major tire manufacturing or benign nature of that tumor.) From plant in Akron, Ohio. Working from Table 4 it can hr, dduced that, in nm- pension, payroll, death claims and other . . ir.k:m plants within thc company under study. thew ha5 1nv.nan almmt two-fold p:almtional extc15 of deaths due to nirltgnant neoplasm cJ the central ncr- company fils, a cohort d 6678 male rublwr workers. agcd 40 to 84 at kn. 1. 196.1. was identified. With few exccptims, this fold cohort wds followed -_-- -- I- --- .s(Lf. JZ # < . X d b *I! 4Ct+:l*" ;*I*--. a. i a p(S..,:,w \\C and Conwar t\O Ckm.cat C a i c i n o t c n r w and Cancers *lnqitc.kl . ~lfinatrC. hrrl*iC Ihrmas. 1W. ~ n, m*9. 136 19. \tancKurrj TF. \(rcFar!ane E\I. hrtt'ndd )D: The &sinburion J canccc~m a l i b in O i ~ r /nPub H / h 45 39-?0. 1955 20. V ~ CSk Aetiologv d t u f n o ~d~Ih uinary bladder. U d hrr 24.276-280. 1969 21. Adclslein A\\ O c c u ~ l t o n rml rxtal*. Gncu. Ann Orcupat Hbg 10.53-5:. 1992 22 C d e 0. Hacwer R, Fncdcfl CH Oc- cupa~ionand cancer of the 'pvw urinaw tract. Cancer 29:1%SoH). 19% 23. Cuira A C B)ad&r carcinoma in rubkc ~0th/ U.r d 106543-51. 1971. 24. Outbreak of occupational illws k i i * $1 millionsenltmnt.OccuprrHaurds:.1749. November 1972. 25. Lloyd b4 and C i a c o h Long-mm mor- taLty study of steelworkers. / Occupar Mtd 11(6):299*310, lw. 26. State of Ohio, Depanmcnl d Health. Division of Vital katis~ics: Rcport of Vi1 Statistics for Ohio 1972. Ohio kale h p t . of Health. Columbus. Ohio. 27. U.S. De@. of Health, Education. and Welfare. Public HealthService. Heath Services and hkrual Health Administration. National -Cenfer for Health Statmics: Vital Statistics of the UnitedStates, 196& V d u m c 11 Mortality. .*1...;.:..'... ... ... .. ......... ._.. ........ . ...........~ .........._..................... .... . .. . .. .. . ,' . .-. . ... . . ,.....-.....