Document bB9qBGq3MVN75Zj7MRR0XVEND

'y contribu. 1 statistical 5c research liaall, iat Pisapneort :a~ a whole lanuscripts lpkins Uni- Ireparat ion should be ete copies, te. Include iide to the tore than 8 Biological each year. itional Li- ieir meanI the text. text. The Medicus. are menfigures #fir nother's). r printed contribu,trictions, (requests ?d).A flat j map be ; in text. :d needs. :al. If the for rate . r ...-.....,~...i:...-.- ... . . NUMBER 3 --_ M A R C H 1976 0 RIG INAL CO-e2N-2TRIBUTIONS 25 1 26 1 27 0 284 297 304 31 2 $\ >* Hospital-Acquired . infections. I. Surveillance in Wenzel (U.of v i r i i n i a Medical Center. Charlottesville. Kathryn J. Hunteg and Jack M. Gwaltney, Jr. A n O u t b r e a i a f Nosocomial Proteus rerrgeri Urinary Wilfiam T..-ia&n. Alex C. Sonnenwirth and John V. ."Aflanra. GA-30333) b.&-2_ _-?.::.--;-:r-' The Epidemiology and Natural History of Streptococcal Pyoderma: A n Endemic Dis- ease o f -t h e Rural Southern United States. Kenrad E. Nelson ( U . of Illinois College of Medic!ne: Chicago. l L 60680).Alan L. Bisno. Paul Waytz. Joel Brunt. Vqai K. Moses and Ria,i-.Ul Haque M o r t a i i t y among Rubber Workers. I. White Male Union Employees in Akron, Ohio. Richard R. Monson (Harvard School of Public Health. Boston. M A 02175) and Kenneth K. Nakano Mortality among Rubber Workers. II. Other Employees. Richard R. Monson (Harvard School of Public Health. Boston. M A 021 151 and Kenneth K. Nakano Some Atherogenic Concomitants of Menopause: The Framingham Study. Marrhana C. Hjortland [National Heart and Lung Institure. Bethesda. M D 20014 ) . Parricia M. McNamara and William 8. Kannel Psychosocial Factors and the R i s k of Myocardial Infarctions in White Women. Moyses Szklo (Johns Hopkins U. School of Hygiene and Public Health. Baltimore, M D 212051, James Tonascia and Leon Gordis Childhood Lymphatic Leukemia: Prenatal Seasonality and Possible Association with Congenital Varicella. Nicholas J. Vianna (NY State Dept. of Health. Albany. NY 122371 and Adele K. Polan Life Dissatisfactions, J o b Dissatisfaction and Illness of Married M e n over Time. Joanne C. Gersten (Columbia U. School of Public Health. New York, NY 10032). Robert Friis and Thomas S. Langner LETTER TO THE EDITOR 342 Re-analysis of Seasonal Trend i n Incidence of Down's Syndrome in W e s t Jerusalem, 1964-1970. Richard P. Ament (Cornrnission on Professional and Hospital Activities. -* The School of Hv.a-tene and Public Health of The Johns HoDkins Universitv Sponsored by the Society for Epidemiologic Research - c11' American Journal of EPIDEMIOLOGY (Formerly the American Journal of Hygiene) Founded 1920,by W. H. WELCH a n d W. H. HOWELL BOARD OF OVERSEERS PHILIP E. SARTWECLhLa,irman \v.FREDERIK B. BANG GEORGE COhiSTOCK JOHN H. HASKS JOHN C. HUME ABRAHAhl hl. LILIEXFELD RAYhlOSD SELTSER BOARD OF EDITORS NEAL NATHANSOEXd,itor-in-Chief JOHN P. Fos GARDNER MIDDLEBROOK RALPH S. PAFFESBARCJERR. . TECHNICAL EDITOR FRAKCESSTARK ASSISTANT TECHNICAL EDITOR EDEN XLLES DELCHER A D M INISTRATlVE SECRETARY RUTH PRICE S UBSCR IPTI0 N M A NA G ER HARRIETTTELLJOHANK Published by The Johns Hopkins University School of Hygiene and Public Health Sponsored by the Sociery for Epidemiologic Research Supported in part by a bequest from the DeLamar Fund Subscription rates: US, S10.00; foreign, S45.00 per calendar year (12 issues). Advertising by medical publishers. manufacturers of drugs and laboratory equipment, and institutions seeking personnel in epidemiology is cordially solicited. Request further information from: American Journal of Epidemiology. 615 N. \\'olfe St.. Baltimore. MD 21205. Second class postage paid al Baltiniore, 3ld. 21202, and at addilional mailitig ofiices. ASSOCIATE EDITORS E. RUSSELL ALESANDERS, eattle, WA P n u ~C. BEAVER, N e w O i h n s , La. I'HILIP BR.\CH.\r.kN, AlhtZta, Ca. .\LFRED A. BUCKG, eneva, Srcitzerlaiid c.CHARLOTTE CAMPBELL,Sprin!jjieitl, Ill. C'HaRLES J. CAItPESTER, C / C d U d , Ohio JOHSCASSELC, l~apeHl ill, AT. C. ROBERT 31. CHANOCBKct,hesda, Afd. T. D.Y.CHIS, Katisas Cify, K a n s . D. I . C L E ~ I E NRe,w Orlea?is, L a . PHILIP COLE,Boston, illass. A. MICHAEDL AVIESJ,eritsaleiu, Israel LILA ELVEBACRKo,cliesfcr, -11inn. ~'REDERICK H. EPSTEISZ, uricli, Su~ilzerlatid ~LASSISG I:EISLEIB, Beflresda, X d . >[.iS\\.ELL F I S L . I S D , BOS1071, AfaSS. MICHAELL. FURCOLOM-, Lcxinirgfon, Ky. PAULH. H\RDYJR,.,Ralliiiiore, X d . DOSALD -1.HEXDERSOGS,enem, Swifzerland Lis T. HIGGISSA,nn Arbor, dlicli. TAKESHHIIRAY.AMTAok, yo, J a p a n KALTEURT.HOLLASDL,ondon, England ~ I I C H E L-1.IBRAHCIl~iaIp,el H i l l , N . C. I<IHL 11.JOHSSOCNa,?zal Zone DAVIDT. ILIRZOS,Naslcrillc, Tenti. IEYISGI . KESSLERB, alfi~ioi.ed, l d . LEIVISI ~ U L L,EPiRllsbur{q/i, Pa. LEOSARDT. IiURLASD, Rochesfer, N i m . .ILESASDD. ELARKGXIUIR, Boston, Mass. EDIVISH. LENNETTEB,erkeley, Calq. BRIAS~ ~ A C ~ I ABHostOonX, il,lass. JOSEPH L. ~ I E L N I CHKo,uston, Texas DAVIDL. MILLERL, ondon, England J.A.\lES W. MOSLEY,Los Angeles, CA STASLETA. PLOTKIPNii,iladelplria, Pa. K.EUGESESAKDERS, JR., Omaha, -1-E EVERETT L. SCHILLER, Baltimore, dld. NORTODS. SCHWEITZXEeRw, York, N . Y . IiEERTI Tr. SHAH, Ballintore, ilfd. 11.I.D. SHARMA, Delhi, I n d i a PAGLSHEEHE, S y a c u s e , N . Y . JAMSH. STEELE, Houston, Texas c.I<ESZABUHTOSUCHIPA, Tokyo, J a p a n HER~IAxA. TYROLER, Clrapcl H i l l , N . HERBERT -4.WENNERK, ansas City, dlo. I ~ S UWTESTLUNDT, m n s # ,N o t w a y \ ~ - A R R E K~YISKELSTEINJ,R., Berkele?., CA - VOL. 103 ~ RICHARD 7 Wenzel, fi C. A. Osterr tions. I. SL.V 1976. A n e w s, infections. nursing car2 65% of t h e : ogist requirc ganize line : cent accurn: tive reviev#s method wiij all patients (ternperatur i advantages bacteriolog fections w.: surveillanc.: the risk of outbreaks I hospital- :.. INT: -The efficient acquired infectic surveillance systi -developing a prog important problC-7 Received for public;. form Septemiler 10. i:! ' Department of Ir::, r RIC:I3ARI.~ R. MO>GC'l\l ASD I<E?:XETH I<. N.9KAh"::I M o n s o r , , ,R . F , . ( I h p t . of EFliderniology. H a r v a r d S c h o o l of Public H e a l t h , 677 H u n t i n g t o n Avc:., B o s t o n . M A 021 1 5 ) .and K. K. N a k a n o . Malrtality a m o n g rubber worker:;. I. W h i t e male union einplloyees in Akron. Ohio. Am J Epidemiol 103:284-.296. '1 976. Mortali::y a m o n g a c o h o r t of 13,571 w h i t e m a l e r u b b e r w o r k e r s in A k r o n , O h i o is d e s c r i b e d . M o r t a l i t y from all c a u s e s was 82% of t h a t e x p e c t e d o n t h e b a s i s of d e a t h r a t e s for US w h i t e m a l e s . E x c e s s d e a t h s f r o m c a n c e r (observed/expected numbers;) in specific sites among workers in specific w o r k areas i n c l u d e stomach: processing (18/9.9) ; l a r g e i n t e s t i n e : p r o c e s s i n g (14/10.51; l u n g : t i r e c u r i n g (20/12.4); b l a d d e r : all w o r k e r s (48/39.5); b r a i n : tire building (7/3.7); l y m p h a t i c : tire building (111'7.1 1; a n d l e u k e m i a : all ' workers (55/43.0). In general, excess c a n c e r d e a t h s occurred most often a m o n g m e n w h o s t a r t e d w o r k i n g b e t w e e n a g e s 35 a n d 44 a n d prior t o 1935. who w o r k e d 25-34 y e a r s a n d w e r e o v e r a g e 74 a t d e a t h , a n d who d i e d after 1954. Cancer, incidence of; occupational diseases; rubber workers In 1954, an excess rate of tumors of the urinary bladder among workers in the rub- ber industry was reported (1).Since then a number of investigators have studied total and cause-specific mortality among rubber workers (2-6). While no evidence has been Received for publication May 14. 1955. and in final form August 15. 1973. Abbreviation: SMR. standard mortality ratio. From the Department of Epidemiology. Harvard School of Public Health, Boston, MA 02115. Supported by a contract with the B. F. Coodrich Company and the United Rubber, Cork. Linoleum and Plastic Workers of America and by a center grant (5 P10 ES00002) from the National Institute of Environmental Health Sciences. The authors are indebted to I. Mikelsons, L. Pape, and L. Taylor for technical assistance; to W. H. Veigel. Ohio Department of Health, for assistance in obtaining death certificates: to Marian Heid. Sational Institute for Occupational Health, for coding death certificates; to officials of the I.R.S. and S.S.A. for assistance with the follow-up of terminated employees; to officers of URW Local #5 for maintaining R valuable source of occupational histories; and to the members of the Occupational Health Committee of the United Rubber. Cork, Linoleum and Plastic Workers of America and the B. F. Goodrich Comliany for support of this study and concern lor the heah h of the rubber ivorker. reported to suggest that overall mortality is increased among rubber workers, excesses of cancer of several other sites have been reported: salivary glands (41, stomach (3, 6 ) . large intestine (3, 6 , '71, gallbladder and bile ducts (4). lung (2, 5. 7), prostate (6), brain (2, 3, 51, lymphosarcoma (6. 7) and leukemia (5, 6). In 1970 the six major United States rubber companies, after negotiation with the United Rubber, Cork, Linoleum and Plastic Workers of America, agreed to support studies of the health effects of the work environment. In this paper the mortality experience between January 1, 1940, and June 30, 1974, of 13,571 white male employees of one of these companies is reported. Also, for workers who died between 1925 and 1939, a proportional mortality analysis is presented. METHODS . D e f i n i t i o n of p o p u l a t i o n . All members of the study population were employed by one company in Akron, Ohio, and were mem- 284 b e n of the :;i:iine wc1rkt.d in a plan. rubb :r prodcc. Smaller number plant, in whicl.1 rubbm plant are rubber reclaim j. Since 1900, ;I c company for e x name, social scci race, sex, date ployed, date ii: nated, fast deps; for termination ployment). This who were emplo: 1940. Since the receiving \vages I Since its date oi local union has each member. social security date of joining t on the departn: worked after 1E digit code has . department. X the same code department f: Whenever an ments, the r.ev entered seque;.date of the cl?: but each dues the year of the i During the : partments hn\ These depai t! Seven divisic : study, a sect::.. the divisions ments. As s c i : disparity in t '-I section or p r ! section, in \\k is a single dc; - 1In contrast to of managemeat. I MORTALITY AMONC'RUBBER WORKERS I. qorkedin a plant in which tires and other rubber products are manufactured. Smaller numbers worked at a chemical plant, in which chemicals for use in the rubber plant are manufactured, and at a rubber reclaim plant. ' Since 1900pa card has been kept by the .. company for each employee. Data include name, social security number (since 1937), race, sex, date of birth, date first- employed, date when employment termi. nated, last department worked, and reason for termination (death, pension, left em:. ployment). This file is complete for persons who were employees on or after January 1, :-.1940. Since the early 1940's, all employees ' receiving wages have been union members. Since its date of organization in 1933, the local union has maintained a dues file for each member. This file. includes name; social security number (since 1937) and date of joining the union. Also, information on the departments in which a n employee worked after 1933 was available. A fourdigit code has been used to identify each department. Almost without exception, the same code has been used for each department from 1933 to the present. Whenever an employee changed depart. ments, the new department number was 2426 employees. An employee's departmental work experience was categorized in two ways: ever and usual. For example, there were 319 employees whose longest employment was in the chemical division. There were an additional 465 employees who worked in the chemical division a t s o m e t i m e but whose longest employment was elsewhere in the company. T h e determination of a n employee's longest divisional employment,. . was based only on the work experienced after 1933 since no departmental information was available before that year. Prior to 1952, transfer from one division to another was uncommon because of work rules, so it is unlikely that much misclassification mas introduced. The average age started, year started, and years worked refer to total employment by the company rather than employment in the specific divisions or sections. Over 60 per cent of the employees were laid off for 3 months or more a t some time during their term of employment. This time spent out of employment was included in the years employed by the company and was considered to be an eighth "division." The 1432 persons whose usual s a t u s was in the "laid-off" division were but each dues card lasted 4 or 6 years, so the year of the change could be estimated. During the years of this study 271 departments have been in existence (table I). age of 1.5 years. The chemical plant produces chemicals used in making'rubber, mainly in closed systems. Products produced by the aero- section or per division. T h e tire curing section, in which 1092 employees worked, is a single department, while the process- 'In contrast to "salaried employees" who are part includes tire finishing and manufacturing of tubes and truck tires. Small rpbber products are made in the industrial prod- ucts division (e.g., rubber bands, hoses, belts, and molded rubber goods). The serv- p 286 MONSON AND N A K A S O ... TABLE 1 Nbmber of departments and characteristics of employees, according to plant division and section Division and section No. of dmepeanrtts- No. of employees' Ever Usual Average employment datat Age started Year started Years worked . ~..,.A :;+ tween rJ;inuar!r 1, 1s ..'.1:..__ the basis of Social ..-.... T h e other 77E e m p y-:.' to be live witidrn .T . .?. Approximatel:,? 70 "Laid off' Chemical Aerospace Tires Tire building Y 8482 11 784 18 3207 1432 3 19 1051 25 28 26 1937 .. 1941 1937 14 23 24 6 3491 1905 25 . 1933 2a known to be -.I' ive c or benefits p a d c u the remainin;; 30 1 but no death claim Social Securit] Acl Tire curing Miscellaneous Industrial products Plant 4 Hosebelt Miscellaneous 1 1092 538 30 1930 29 21 3141 1045 28 1930 28 27 1639 475 28 1933 24 29 1844 1248 27 1936 25 22 2889 624 27 1935 24 Follow-up of' act ployees was contiii Of the 2610 ernplo! July 1, 1971, 1561 later. Twenty had c Services Engineering Operating lVarehouse/shipping Elevatorkleaning Shops - Rubber reclaim Processing Total 26 548 165 35 24 1023 355 19 9 1625 736 28 5 974 418 38 27 2474 1834 32 10 1261 306 29 35 2426 111s 30 27 1 13.551 28 1933 1935 1931 1931 1933 1938 1931 1931 22 25 28 24 29 23 25 24 Ever is number ever employed in section; usual is number whose longest job was in section. f Age and year first employed by company; total years worked in company. most of the remai1 been laid off dur former employees 1 rolls on July 1, 15 have died during l-Employees who wt on July I, 1971, an( have died during considered to be li\ 1974. ices division includes a variety of depart- known to be deceased and 4164 who had ments, whose employees work throughout terminated employment. Among the group the p1ar.t or with a variety of products. The of terminated employees, 3694 worked less I For most of the I deceased, a copy certificate was o death was coded rubber reclaim division is a separate plant where used rubber is ground up and reprocessed. In the processing division, the raw materials of rubber are weighed and mixed than 13 years, 268 worked 15-24 years and 202 worked 25 years or more. The percentage of terminated employees and the distribution of years worked were similar among National Institute i and Health acco i effect at the tirassigned was the! together, and the resultant hot raw rubber is milled into a variety of forms. Follow-up. Approximately 32,000 white male union members were identified from the union and company files. Over 18,000 the employees of the various plant divisions. Determination of the status of the employees who had terminated employment prior to pension was conducted with the i digit Internation: I eases (ICD) Code, I certificates could 1 - cent) of the decc rection to the cau men who worked fewer than 5 years are not assistance of the Internal Revenue Service tality ratios (SM included in this report. For workers first and the Social Security Administration. the non-availabil employed in 1935 or later, follow-up started The last known year of filing an income tax Computation 1 on the fifth anniversary of employment. return was in 1972 for 695 of these men and deaths. The nun For workers first employed prior to 1935, in 1973 for 1954 men. These 2649 former assigned to speci follow-up started on January 1, 1940. The employees were considered to be live with- to the numbers e? study cohort consisted of 13,571 men. As of drawals on January 1st of the year in which expected numbei June 30, 1971, when this study was initi- the return was filed. viva1 of the stud: ated, there were 2610 active employees, Of the remaining 1515 terminated em- in 5-year age-tin. 2988 who were pensioned, 3809 who were ployees, 739 were found to have died be- of person-years i -='I ars -,ked .4 !3 !4 !a 29 !S 24 25 24 22 25 za 24 29 23 25 24 io had group ed less rs and xcentdistriamong t divi- ne em,yment ith the 3ervice .ration. m e tax :en and former e with1which ed emied be- MORTALITY AMONG RUBBER WORKERS I. 287 -- January I, 1940 and July 1, 1973, on multiplied by the age-time-cause-specific the basis of Social Security death claims. mortality rate for United States white .. T h e other 776 employees were considered males. T h e rates used were time-specific : to be live withdrawals on July 1, 1973. for each 5-year time group from 1940-1969. imately 70 per cent of these were The rates from 1965-1969 were used for nown t o be alive on the basis of earnings 1970-1974. T h e expected numbers in each or benefits paid during 1973. T h e status of' cell were summed over age and time to thepremaining 30 per cent was unkri*,wn, yield a total expected number for each 57 but no death claim had been filed with the categories of cau2e of death (8). Social Security Administration. T h e expected numbers of cause-specific Follow-up of active and pensioned em- deaths were also calculated using death was continued until July 1, 1974. rates for Ohio males. In general, the results Of the 2610 employees who were active on were the same as those obtained using July 1, 1971, 1561 were still active 3 years United States rates. Further, there are no later. Twenty had died, 508 had ret.ired and major differences between death rates for most of the remaining 521 employees had white males in Summit County, the loca- . been laid off during 1974. Of the 2988 tion of the plant, and for the United States former employees who were on the pension (9). Examples include (age-adjusted rates rolls on July 1, 19'71, 511 were known to per 100,000, Summit County/United . have died during the next three years. States): all cancer (185.5/174.0), stomach Employees who were active or pensioned cancer (16.2/15.2), large intestine cancer on July 1,1971, and who were not known to (19.6/16.5), lung cancer (39.0/38.0), brain ied during the next 3 years, were cancer (5.4/4.4), bladder cancer (8.0/6.8) ~ ..= considered to be live withdrawals on July 1, and leukemia (9.3/8.8). Any excess mortal- ity among rubber workers would, of course, most of the 5Oi9 persons known to be be included in the rates for Summit deceased, a copy or abstract of the death County. certificate was obtained. T h e cause of Proportional mortality, 1925-1939. Prior was coded by a nosologist a t the to 1940, it was not possible to compute National Institute for Occupational Safety expected numbers of deaths on the basis of and Health according to the system in death rates. However, the cause of death of effect at the time of death. T h e code 688 white male employees was known. assigned was then translated to the four- Therefore, expected numbers of cause- International Classification of Dis- specific deaths were calculated on the basis .. . eases (ICD) Code. Seventh Revision. Death of proportional mortality ratios. The ob- certificates could not be found for 97 (2 per served number from all causes was categ-ocent) of the deceased employees. No cor- rized into 5-year age-time-specific groups rection to the cause-specific standard mor- and age-time-specific proportional mortal- tality ratios (SMR) was made because of ity ratios for United States white males the non-availability of these certificates. were multiplied across each of these strata Computation of expected numbers of t o calculate the expected number of cause- : deaths. The numbers of deaths observed specific deaths (8). Only the last depart- assigned to specific causes were compared ment in which a man had worked was to the numbers expected. In computing the available for this group of employees. .:. expected numbers, the person-years of sur- Statistical considerations. No tests of viva1 of the study cohort were determined statistical significance are given. Since the in 5-year age-time intervals. The number overall SMR was 82, testing whether the of person-years in each age-time cell. was SMRs for specific causes of death differ , 288 MONSON AND NAKANO __ from 100 does not seem indicated. Also, sented. These are related using the SMR: inclusion of department-disease associa- (100 x observed/expected). The overall tions and of age-time categorizations de- SMR of 82 indicates that this group of pended to some extent on examination of rubber workers experienced mortality the data, and tests of statistical signifi- which was only 82 per cent of that ex- cance would not be appfopriate. pected. Causes in which the SMR was 95 or c RESULTS above include leukemia and cancers of the stomach, large intestine, bladder and lym- T h e causes of death of the 5079 de- phatic system. Causes in which the SMR ceased employees are presented in table 2. was 70 or below include all infective and 4 The numbers expected on the basis of parasitic diseases, cancer of the buccal United States mortality rates are also pre- cavity and pharynx, diseases of the respira- 4 TABLE 2 Observed and expected deaths according to cause* ICD N0.t and cause of death Observed Expected SMRS i All causes 5059 6186.9 82 I 001-139 Infective and parasitic diseases 1 1-10-205 Malignant neoplasms I 1-10-149 Buccal cavity and pharynx ! 130-159 Digestive organs and peritoneum 51 152.0 34 980 1046.4 94 19 36.1 53 364 371.6 98 150 Esophagus 21 26.4 91 i 151 Stomach I 153 Large intestine ; 154 Rectum 155. 156 Liver and biliary passages ! 157 Pancreas i 161 Laryns 98 93.9 104 104 103.1 101 12 17.2 89 30 33.8 89 53 60.3 88 12 17.0 71 162.163 Lung 234 233.1 92 li7 Prosta te 82 89.0 92 , 180 Kidney 19 24 .S cIt 181 Bladder 48 39.5 12:! 193 Brain 20 25.1 80 200-203 Lymphatic and myeloma3 31 50.4 101 204 Leukemia 55 13.0 138 260 Diabetes mellitus 81 90.i 93 330-334 Vascular lesions of central nervous 493 539.6 91 system 400-468 Diseases of circulatory system 2445 2913.2 83 420 Arteriosclerotic heart disease 1826 2155.9 85 ! 470-527 Diseases of respiratory system 530-587 Diseases of digestive system 1 590-637 Diseases of genito-urinary system j 800-998 Accidents, poisoning, violence 21s 321.4 68 215 284.0 76 74 140.0 53 278 446.8 62 600-936 Accidents 184 320.4 57 970-979 Suicide 72 105.8 68 I All other causes 142 222.8 64 Unknown causeofdeath 97 I ! Expected numbers based on 5-year age-time specific death rates for United States white males. t International Classification of Diseases. Seventh Revision. Standard mortality ratio: 100 x observed deaths/espected deaths. 5 1950 and later only for codes 200. 202. 203. -. tory and genitour dents, poisonings There were 25;: low-up accumul,: sioned employecs. accumulated by ! ing the date of occurring only wii status are considc all causes (4340 ( and 100 for all C E expected). Sevcn the terminated c. have died followir deaths were expc SMR was 81 if calculated using ployees on active In table 3, mo death and from according to five ment. In compi:t year when a n eiii person-years W C T ~ category. For t Standard rnortolify r all causes and f z selecti ~ Characteristic Aye started working Year started workin:: Yean worked A t e at death Year of death rI ; I ;the SMR: rhe overall is group of mortality of that exIR was 95 or ncers of the ,er and lym:h the SMR ifective and the buccal .the respira- SMRt 82 34 94 53 98 s1 104 101 89 89 88 71 92 92 cc If 122 80 101 128 93 91 83 85 68 76 53 62 57 68 64 males. MORTALITY AMOSG RUBBER WORKERS I. P tory a n d genitourinary systems, and acci- started working" categories, SMR's for w: dents, poisonings and violence. three independent cohorts were calculated. There were 253,026 person-years of fol- For the "years worked" grouping, the per- low-up accumulated by active and pen- son-years for a n individual were included sioned employees, and 88;414 person-years in more than one category. For example, accumulatkd by former employees follow- follow-up of a n individual in the "235" *in-g the date of termination. If deaths category did not begin until the 35th annioccurring only while on active or pensioned versary of employment. The 30 person- status are considered, the SMR was 85 for years accumulated prior to that time were all causes (4340 observed, 5082 expected) allocated equally to the other three cate- and 100 for all cancers (850 observed, 851 gories. .. expected). Seven hundred thirty-nine of In considering the data in table 3, it the terminated employees were known to must be emphasized that there are correla- have died following termination, while 1211 tions among t h e person-years of the five deaths were expected (SMR = 61). (The characteristics. A person who worked for 35 SMR was 81 if expected numbers were or more years must have started work calculated using the death rates of em- before age 30 and before 19.10. However. ployees on active or pensioned status.) the main point of emphasis is that the In table 3, mortality from all causes of S M R s within each characteristic do not death and from all cancers is considered differ greatly. For example, employees who according to five characteristics of employ- started working before 1925, when indus- ment. In computing the S M R s for age or trial hygiene was presumably not as highly year when an employee started working. all developed as in later years, have SMR's for person-years were grouped within a specific all causes and all cancers which are similar category. For example, for the "year to those for other employees. The highest SMR is that of 130 for deaths from cancer TABLE 3 in persons who died at age 73 and over. Standard mortality ratios (and observed deaths) ior The mortality experience was examined all causes and for all cancers, according to selected characteristics ?:$'Characteristic ( .Anlul ~ s i CanAClFlrZ for groups of workers categorized by plant division in two ways (table 1). Under the category "ever worked," an employee is counted in each division in which he Age started working <25 25-34 35-44 245 73 (9900) 81 ll9:;l 80 (2002) 91 13951 94 (14711 10912:Sl S3 (616) 98 (1141 worked (an average of 1.85 divisions per employee). Under the category "usually worked" each employee is counted only Year started working < 1925 86 (20401 91 13491 1925-1934 80 (1740) 100 (3781 ,.kz,: .. 2 1935 79 (1299) 68 (253) once. The S M R s under "ever worked'' tend to be less than those under "usually worked" because of an inverse relationship between the S M R and the number of divisions in which a person worked. Except for employees whose usual status was in the laid off division, the SMRs for all cancers were higher than the SMRs for all causes. T h e range of the SMR for all causes was 72 to 87, and for all cancers from 69 to 111. Calculations of observed and expected li- deaths, as shown in tables 2 and 4, were I 290 1'CBI.E 4 Standard olortolify;,,,fifg. (anc!obserwd d i ! a ~hs) /INa-ll causes and all cancers, according to plant divisions in ___-.__- which employees <wei-and usually worked -_--. ---.--.-- --.-- t Divi:;ion Averagt! survival ----_c._haracteristic:; i.Re at Year 01 ---vntry entv _-_._. Ever worked --. _- All causes -.__--- .AI1 cancers ~-~ Usually worked All -' causes All cancers "Laid off' 32 1945 81 (319) 69 (44) Chemical 34 1948 '14(161) 85 (32) 74 (68) 80(13) Aerospace 33 1945 14 (711) 92 (155) 81 (282) 89 (55) Tires 38 1943 79 (19683 92(397) 81 (1345) 93 (265) Industrial products 37 1945 76 (1600) 84 (309) 81 (796) 94 (159) Services 42 1944 83 (2238) 90 (405) 84 (1638) 94 (305) Rubber reclaim 37 1946 15 (313) 94 (68) 72 (93) Io0 (22) Processing 41 1943 79 (833) 98 (177) 87 (538) Ill (117) Total 38 1944 82 (5079) 94 (980) ,- Age and year of entry refer to follow-up. TABLE 5 Observed and expected deaths due to selected diseases. accordin3 to work area in which employee usual/>, worked Disease Work area No.of deaths Observed Expected Standard mortality ratio .ill gastrointestinal cancer Stomach cancer Large intestine cancer Pancreas cancer Lung cancer Prostate cancer Bladder cancer Brain cancer Lymphatic cancer and myeloma Leukemia .c-----s Asthma Diabetes Vascular diseases of central nervous system Pneumonia Processing Rubber reclaim Processing Processing Elevator/cleaning Tire curing Miscellaneous tire Warehouseh hipping Tire building Tire buildinp Tire building -Tires Processing Services W arehouse/shipping Operating services Tire curing 52 11 18 14 6 20 :2 8 9 7 11 18 10 11 11 19 12 38.2 7.1 9.9 10.5 2.9 12.4 8.8 2.6 4.8 3.7 7.1 11.7 4.2 7.0 5.8 14.1 7.7 140 140 190 130 210 160 1.10 310 190 190 160 lL 240 160 190 130 160 ~~ done for employees of each division and section. Work areas with excess deaths due to specific causes are listed in table 5. T h e following arbitrary criteria were used for inclusion in this table: at least 5 deaths from a specific disease, at least three excess deaths in the work area in which an employee usually worked, and a n SMR of 130 or greater. In general, excess observed deaths were not found among employees who worked in a section for short periods of time. In other words, any excess deaths among the "ew accounted for 1 grouping. Thcrc deaths from !u among emplcyc curing section, their usual H-Oi- bladder cancer ployees of the v whose usual VI section. The departn ployees who dit table 5 was ex: trointestinal ca division resultel of the esophag cesses of gastr leukemia in tl- spread througl ments. How:\. among men calendering, 2 bowel and 2 pharynx. A n - were only 3 a7.c Tire builCr and brain cEni lymphatic a?r the four brzii- type was sprii three were ;I trocytoma . -i\ I lymphoma, rb and one fro- i Excess d? i most often :n i processing 3i were also s e w cal divisioi '[. elevators i r r 1.' expected), 2i pected), and I served, 7.0 E:; cl The exccs diabetes and tral nervous I of departma: .', MCIFL'I'ALI'I Y' /LIViONG RUBBER WORKERS [. 291 among t'ke ' ' E , " ~ I worked" g:rourlin; were pneumonia in :he t i I e :,iring section was accountecl ;31. 'ki;r the "usu.ally w3rltec.l" consistent with the EX.[:E :is OF d.eaths from grouping. '['Ix::.e were two c!xce~lticm :,: 11 lung cancer. deaths froni 'iIJAg camce!: (8.5 ?xpe::ted) Cancers for which tl:e Sh!lI:tI. was greater among ernFlo!iee:1 who worked iii tht: tire than 90 (table 12) wer? ma.mined according curing, sect:on but for vrhcinn it vwi. not to the age-time ch ii*acti~ristiicsused in their usual ,work area., arid 4 deaths from table 3. In table-. 6 , the breakdown for bladder carice:. (:l.9 expected) amon:; em- bladder cancer, lympE,aI.ic:ca'ncer and leuployees of t . 1 ~warehouse/s,hipping sc.ction kemia are pres,ented f o r the ent.ire cohort. whose u s u d work area lidas in another T h e data for gastrointestinal cancer are section. given only for men who usually worked in T h e departmental experience of em- processing. (The patterns in the SMKs for ployees who (died of the diseases listed in stomach and large intestine cancer among table 5 was examined. The excess of gas- processing workers were similar.) The data trointestinal cancer in the rubber reclaim for lung cancer are given for all men who division resulted from 4 deaths from cancer ever worked in tire curing. of the esophagus (0.6 expected). T h e ex- T h e following observations are made cesses of gastrointestinal cancer and of based on the data in table 6: 1) Excess leukemia in the processing division were gastrointestinal cancer occurred in all cate- spread throughout a number of depart- gories, and was greatest in processing work-- ments. However, it is of interest that ers who died at age 75 and above. 2) Excess among men whose usual department was lung cancer was limited to tire curing calendering, 2 had cancer of the small workers who started work between ages 25 bowel and 2 had cancer of the naso- and 34 and who died in 1955 or later. 3) pharynx. Among the entire cohort there Excess bladder cancer was greatest in men were only 3 and 4 such deaths, respectively. who worked at least 35 years and who died Tire building had excesses of bladder a t ages 75 and above. No excess risk was and brain cancer as we11 as cancers of the present in men who started working after lymphatic and hematopoietic systems. Of 1934 or who died before 1955. (In men who the four brain tumors for which the cell started working between 1930 and 1934, type was specified on the death certificate. there were 7 deaths from bladder cancer three were gliomas and one was an as- and 5.1 expected). There was no associatrocytoma. There were seven deaths from tion of the SMR with the age men started lymphoma, three from Hodgkin's disease, working. 4) In men who started working and one from multiple myeloma. before age 25, a n excess was seen for Excess deaths from leukemia occurred leukemia but not for lymphatic cancer and most often among workers in the tire and myeloma. There was no evidence for an processing divisions. However, excesses excess of either category of cancer in men were also seen among workers in the chemi- who started working after 1934. cal division (2 observed, 0.7 expected), If excess cancer in fact had occurred elevators and cleaning (3 observed, 0.4 among rubber workers because of their expected), shops (7 observed, 6.6 ex- work experience, it might be expected that pected), and industrial products (9 ob- some minimum number of years of expo- served, 7.0 expected). sure would be necessary for induction of diabetes and vascular disease of the cen- ratios for each cause of death were calcu. tral nervous system occurred i n a number lated for men who worked a t least 25 years. of departments. The excess of' deaths i'roin Follow-up started on the 25th anniversary I .--:E .. . . * i i.I.. ! ! 292 hlONSO'N ASD N A K A S O of employment. T h e overall SMR was 86 ing after 1934, although only three such ~ and that for cancer was 98 (table 7). In men have died from 1eukt:mia. r general, the SMR's for each cause were no In the entire cohort there were only two more than 10 per cent greater than those in deaths from cancer of the salivary glands table 2, and some were !ess. The only large a n d eight deaths from cancer of the gall- difference was for leukemia, where the bladder and bile ducts. No common work i SMR was 160 rather than 128. T h e age- experience could be ascertained among time patterns for all causes, all cancer, these workers; only one person with cancer bladder cancer, and leukemia are pre- of the bile ducts worked ((for1year) in the sented in table 7. processing division. T h e patterns for all causes and all can- Table 8 gives the results of a propor- . cers are similar to those in table 3. Cancer tional mortality analysis for persons who was relatively uncommon before 1955 and died between 1925 and 1939. By definition, showed the greatest excess in men who died the expected number of deaths from all a t age 75 and over. Bladder cancer espe- causes is equal to the observed number. On cially had this pattern. Because a n excess a proportional basis, deaths from diseases of bladder cancer occurred only in men who of the circulatory system and from acci- died a t a relatively advanced age, it is dents. poisoning and violence were slightly .difficult to determine whether the men greater than expected. Deaths from cancer first employed after 1934 have a n excess were not in excess. The three employees risk. In contrast to the data in table 6. who died from brain cancer worked last in there is no evidence that the risk of leu- processing, tire building and a n .undeter- kemia is reduced in men who started work- mined department in the tire division. TABLE G Standard mortality ratios (and observed deaths) for selected cnncers. accordin? t o selected charncteristics Standard rnortc characteristic Age started working Year started workin: Ape at death Year of death Total Expected = 4.2. Obserccd nnd ex,: .-lkrori i ICD So.+and I Characteristic .\$e started working ~ Category (years) <25 25-34 2 35 Castrointestinal cancer. 120 (7) 120 (22) 160 (23) Year started working < 1925 1925-1934 2 1935 130 (21) 130 (22) 160 (9) Years worked Age at death 5-24 25-34 2 35 <65 65-74 2 75 150 (24) 140 (20) 100 (8) 110 (18) 110 (16) 250 (18) Sear of death < 1955 1955-1964 -> 1965 110 (13) 160 (21) 130 (18) Employees who usually worked in processing. iEmployees who ever worked in tire curing. $All employees. caLnucneer; 90 ( 4 ) 220 ( 2 4 ) 50 ( $ 1 130 (SI 200 ( 1-1) 120 (9) 130 (11) 140 (10) 200 (10) 140 (17) 140 (10) 220 (4) 70 ('1) 160 (12) 160 ( 17) Bladdef cancerr 110 (7) 130 (21) 0'1 ('0) 140 (24) 140 (19) GO (5) 110 (17) 100 (14) li0 (17) 100 (14) 90 (13) 220 (211 90 (8) 1.10( 1 7 130 (23) cancer and Leuliemiat m?-elomat 90(1.1) 110 (22) 110 (1.5) 180(19) 100 (18) 120 (18) 100 (14) 120 ( 2 2 ) SO (15) 1.10 (21) 1GO (25) 70 (9) 100 (24) 90 (13) 120 (14) 100 (29) 100 (15) 110 (7) 90 (1s) 170 (23) 150 (14) 120 (25) 150 (21) 110 (9) 130 (12) 90 (17) 100 (22) 120 (12) 150 (23) 110 (20) All causes 140-?Oj Slalipnanl : IjO-Ijg Digestkc peritcc 162. lG3 Lung 131 Bladder i 193 Brain 200-2iIj Lymphz: \ 00-334 Vascular ':r \'OUS sp:. 400-468 Diseases 10: I system I -800-993 .Accidenrs. violence .All ocher c. Expected nuill I cific proportional I males. I t Internatioiiel 1 i Revision. I A central among occupa I a comparison Ii ties esist as a I. MORTALITY AMOXG RUBBER WORKERS 1. 293 h TABLE 7 Standard mortality ratios (and obserccd deaths) in 7374 employees uho worked at feast 25 y e a k 0 IS Charrcterisic CateEory (years) All causes All cancers Bladder cancer Leukemia 1- .-k Age started working <E 25-31 i 6 (6.16) &I (1461) 86 (140) 97 (296) 130( 5 ) 140 (18) 260 (16) 110 (13) '8 135 100 (is3 117 (140) 100 (6) 170 (8) er r le Year started working < 1925 89 (li62) 95 (307) 1iO (21) 150 (20) 1925-1934 84 (915) 104 (215) 120 (9) 180 (14) )r- 30 Age at death n, 311 2 1935 <65 , 63-74 2i 5 52 (213) 72 (891) 89 (1140) loo (859) 91 (54) 81 (191) 99 (241) 131 (144) 60(1) 100 ( i ) 90(9) 210 (15) 140 (3) .180 (15) 180 (li) 100 ( 5 ) In ,es Year of death C1- cly <1955 1955-1964 2 1963 80 (4941 90 (1019) 85 (1377) 73 (75) 111 (1291 9i (272) 0(O)* 190 (15) 1.10 (16) 230 (8) 190 (15) I20 (14) :er Total :es 86 (2S90) 98 (576) 130 (31) 160 ( 3 i ) in * E x p e c t e d = 4.2. er- TABLE S rubber workers: 1) Another group of em- Obserced and expected deaths for ali idiite male ployed persons in Akron. Assuming that Akron employees. 1925-1939' such a group could be identified, the size c--s ICD So.+ and cause of death Ob- EX- and cost of this study would have been served pccted doubled. Also, there would be no guarantee 1: All causes 6s 6ss.o t h a t such a group would have no unusual 140-205 blalignant neoplasms 63 62.6 mortality patterns of its own. Finally, the 150-159 Dipotive oqans and peritoneum 31 33.7 death rates obtained from such a group 16'1. 163 Lunc 6 4.3 would be based on relatively small num- 1Sl Bladder 193 Brain 2 3 2.4 1.2 bers of deaths and would be quite unstable. 200-205 Lymphatic and heinrtopoieric 4 3.F 2) Salaried B.F. Goodrich employees in 330-334 Vascular lesions of central nervous system SS 36.2 Akron. The death rates for this group of 400-469 Diseases of circulatory 169 l5O.S employees are quite unstable. Also. any Mx)&8 system Accidents. poisoning, 141 19.3 confounding effects of socio-economic fac- violence All other causes .,--- 1 4 309.1 tors on mortality would likely be maximized in such a comparison. 31 Ohio death Expected numbers based on 5-year ape-time-spe- rates. T h e cause-specific mortality rates cific proportional mortalities for United States white males. t International Classification of Diseases. Seventh Revision. for Ohio white males are virtually identical to those for United States white males. United States rates were used because they are available in greater detail. 4) Summit DISCUSSION County rates. Such rates are not available. 5 ) Mortality rates .from an occupational A central issue in mortality studies cohort from Social Security records. The among occupational groups is the choice of development of such rates has been sug- a comparison group. A number of possibi1.i- gested (10). While these rates would be of ties exist as a comparison group for Akron great value. they currently are not avails- . .I 294 ' MONSON AND SAKANO ble. Until this resource has been developed, quency in some work area, and no obvious rates from the general population must be reason for selection with respect to that used. disease can be found, measures should be In studies of mortality where. expected taken to reduce possible harmful exposure. deaths are computed using death rates for Workers in the processing division had some general population, standard mortal- an excess of deaths from cancer of the ity ratios less' than 100 are common be- gaatrointestinal tract. Specific jobs held by cause of the relatively healthy status of men who worked in this division include most employed groups (10).In this study, rubber compounding, mixing, milling and death rates for all United States white calendering. In these areas there is consid- males were used because they are the only erable exposure to dusts and fumes. By rates readily available in sufficient cause- weight, the most common substance added sex-race-age-time-specific detail. Also, in- to rubber is carbon black, which strength- clusion of the terminated employees leads ens the rubber. Carbon black contains im- to a n underestimation of the SMR. Finally, purities, some of which have been shown to the non-availability of death certificates be carcinogenic in animals (12-14). In a for 2 per cent of those known to be deceased rubber worker who died from carbon pneu- would further reduce the cause-specific moconiosis, areas of the gastric mucosa SMRs: Thus, for methodologic reasons, were found to be infiltrated with carbon any SMR resulting from this study which black (15). However, no excess of cancer reflects a causal association is likely to be deaths was found in one study of carbon an underestimate of the true SMR in the black workers (two observed, five espected) population of rubber workers. The follow- (16, 17. ing S M R s for all causes of death among Excesses of cancer of the bronchus rubber workers have been reported: 56, among tire tvorkers have been reported U.S.Vital Statistics Special Study (7);59, (3.5). From the current data, workers who U.S.group life insurance records (11);97, cure tires appear to be a t increased risk of British rubber and cable workers (5); 99, bronchial cancer as well as being a t in- American rubber workers (6). creased risk of dying of pneumonia. In It is never possible to determine which producing tires. the "green" tire is cured in SMRs reflect causal association and which a press under heat and pressure. Repeated do not. A large number of S M R s have been exposure to the curing fumes was common calculated and random increases are likely. in curing workers since in the past the In selecting associations to be presented in cured tire had to be removed from the press this paper, arbitrary criteria were set up by hand. Most curing presses now are and all associations meeting those criteria automated. Workers in this area have also are presented. These are the associations been shown to have an excess of respiratory which are based on the largest number of signs and symptoms as compared to other observed deaths and thus are the most members of the plant population (18). The stable. The judgment as to whether an cigarette smoking habits of current curing association is causal must depend upon workers are similar to those of workers in external considerations: consistency with other areas of the plant (18). knowledge of the rubber-making process A greater incidence of cancer of the uri- and consistency with other studies of mor- nary bladder among rubber workers has tality. However, with respect to actions to been recognized for many years (1~9,201. be taken within the plant being studied, The probable cause is beta-naphthyla- caution must be exercised. If a disease is mine, a possible contaminant in antioxi- found to have occurred with excess fre- dants added in processing rubber. (There -. were 5 death: served and 4 workers who UI ing division). in shipping an cult to esplair. reported in shj rubber works, was seen amc Products in was a contam rubber manuf (5). If it is the cancer in rul employed as r' should not ha data reported cancer is seer work after 19; However, sinc cancer is gre: died at age 75 concluded t h . has been elini It has been cancer ;inion: to men tvho s (21). s o Stl aniong rhis gr The overid workers in s general espo used in I~M... particulnr 11: i n d u s t n an3 leukemogen!c' is no loiigir contaminar t used (24); A solvent e s p c in another i' been report t' : No hypcl?. cesses of t: aniong tire \ posed to org; well as 10 u m Based on MORTALITY AMONG RUBBER WORKERS I. . 295 were 5 deaths from bladder cancer ob- studies among rubber workers, attention served and 4.1 deaths expected among should be directed to identify specific sub- workers who usually worked in the process- stances which may be responsible for gas- ing division). The excess of bladder cancer trointestinal cancer in processing workers, in shipping and warehouse workers is diffi- lung cancer in tire curing workers, brain cult to explain. A similar finding has been cancer in tire builders and leukemia in all reported in shipping'workers in an English rubber workers. rubber works, although a greater excess was seen among processing workers (19). products in which beta-naphthylamine Was a contaminant have not been used in rubber manufacture since the late 1940's (5). Ifit is the agent responsible for bladder cancer in rubber workers, persons first REFEREXES 1. Case RAM, Hosker ME: Tumor of the urinary bladder as a n occupational disease in the rubber industry in England and Wales. Br J Prev SOC *- Med 839-50. 1954 Mancuso Tumors Of the nervous cancersystem: Industrial considerations. Acta Un Inter Cantre 19:448-489, 1963 employed as rubber workers in recent years not have an increased risk. In the data reported here, no excess of bladder Cancer is Seen in employees who started work after 1934, Or who died before 1955. However, since the excess risk of bladder cancer is greater primarily in those who died at age 75 and older, it may not yet be concluded that the cause of bladder cancer has been eliminated. It has been reported that excess bkidder 3. Mancuso TF. Ciocco A, El-Attar AA: An epidemiologic approach to the rubber industry. J OCCuP Med 10:213-232. 1968 4. Mancuso TF'.Brennan MI: Epidemiologic consid- erations of cancer of the gallbladder. bile ducts and salivary glands in the rubber industry. J Occup Med 12:333-3-11, 1970 5. F~~ AJ, Linden DC, Otven R A survey of occupational cancer in the rubber and cablemaking industries: results of five year analysis. 1967-71. Br J Ind bled 313140-151. 1974 6. McMichael AJ, Spirtas R. Kupper U:An epide- miologic study of mortality within a cohort of f ~ ~ ~ 4 $ , " $ ~19~64-'i2- OcCuP Med Cancer among was limited t(o21m).enswoho started working data before were age 25 found among this group of rubber workers. The overall e x e s from leukemia among workers in several divisions suggests a general exposure- Organic solvents are i.Curalnick L Mortality by industry and cause of death among men 20 to 6-1 years of age: United States. 1950. Vital Statistics-Special reports 53-2:397. 1963 8. Monson RR: Analysis of relative survival and proportional mortality. Computers and Biomedi- cal Research 73325-332. 9. Mason TJ. McKay FW: US Cancer Mortality by County, Washington. Department of Health, used in many processes, and benzene in particular has been used in the industr?- and has been suggested to be leukemogenic (22. 23). While pure benzene is no longer used as a solvent it is a contaminant in many of the d v e n t s still used (24). A general association between Education and Welfare Publication No. 74-615. 19i-L 10. Goldsmith J R \\'hat do ,ve espect from an occupational cohort? J &cup Xed 17:126-127. 19i5 11. Sibigtroth JC: 1965 Occupation Study. Society of Actuaries, Chicago. 1967 12. Von Haam E, Mallette FS: Carcinogenicity of carbon black extracts. Ind Hyg Occup Med solvent exposure and lymphatic leukemia in another g o u p Of rubber workers has been reported (23). No hypotheses are offered for the ex- of brain and lymphatic cancers among tire builders. These men are ex- 6237-242* l9j3 13. Neal J, Trieff NM: Isolation of a n unknown carcinogenic polycyclic hydrocarbon from carbon blacks. Health Lab Science 9:32-38. 1972 14. Gold A: Carbon black adsorbates: Separation and identification of a carcinogen and some oxygenated polyaromatics. Anal Chem 47:1469- 1472. 1975 posed to organic solvents and cements, as well as to uncured rubber. 15. Miller AA. Ramsden F: Carbon pneumoconiosis. Br J Ind Med 18103-113. 1961 16. Ingalls TH: Incidence of cancer in the carbon Based on the data in this and other black industry. Arch Ind Hyg 1:662-676, 1950 1: -':, F' I I,,;:(::. ,-I' '.j': :.! !I :I . I f . Ingalls TH,Risquez-Iribarren R Periodic search for cancer in the carbon black industry. Arch Env Health 2129-433, 1961 18. Fine L,Peters JM: Studies of respirntory morbid- ity in rubber workers I. Prevalence of respiratory symptoms and disease in curing workers. Arch Env Health. In press 19. Veys CA: Aetiology of tumors of the urinar?; bladder. Urol Int 24:276-288, 1969 20. Cole P, Hoover R, Friedell CH:Occupation and cancer of t h e lower urinary tract. Cancer 29~1250-1260,1972 21. Hoover R. Cole P: Temporal aspects of occupa- tional bladder carcinogenesis. N Engl J bled 288:1010-~1013.1973 12. Mailor\ T B Gall EA, Biicklii: \4'S. Chronic exposure to benzene (benzol). 111. The pathologic results. J Ind Hyg Tox 21:355-357. 1939 23. Vigliani EC, Saita G: Benzene and leukemia. N EnzI J Med 271:872-876, 1964 24. Pagnotto LD, Elkins HB, BruFsh HG, et al: Industrial benzene exposure from petroleum naphtha. I. Rubber coating industry. Am Ind Hyg ASSOCJ 12:417-421, 1961 25. McMichael AJ. Spirtas R. Kupper LL. et al: Solvenr exposure and leukemia among rubber workers: a n epidemiologic study. J Occup bled 17:?31-239, 1975 E.! 1 1 \ ( , C r s ,;, F > \ c::! :,i.il: 1: !I: l'lnlI , I -1 1.1 11t i I I'U hbr.. I'tl1,' tal.::on ' ' ey',i I : I ~:zli:c.? 3i#3CY< y 1nptnt: ?.A lyI''Ip t- ..~,a r In thti pr aniong whit Akron rut!: paper, ITICI~ a t the sZ.r.:e ail black 1'11 employers. ' among wLiiI rubber pl:~i' studying t 11 was to detc patterns an: ployees WI:! employees. Akron <:?I viously c'; . --assemble.:: Received C). form Augu.t ;= Abbrevio:;: From tho Z School of Pi:'.' Supported ' Company oxd and Plastic \S; (5 P10 ES OC Environment I. '"Salarici t agernexit in <.I: