Document jm8v86on5bkDR63OGojqE2gEN

DownloadSearchable PDFRandom document
-19-70 WED 15:20 FALK LIBRARY (PIT) I JOEM Volume 37,Number 5, May 1995 FAX NO, 6489020 P, 03 597 Occupationally Related Cancer Risk Among Coke Oven Workers: 30 Years of Follow-Up joseph P. Costantino, DrPH Carol K. Redmond, ScD Audrey Bearden, MS I3-r he cause-specific mortality panems of two la& steelworker -cohorts have been monitored shce the early 1960s by the D m e n t of Bios&ustics of tbe Graduate School of Public Health, University of Rasburgh. `-I4 Information obtained from this research was used by fed- eral agencies to estabBsh occupetionrl exposure Iimits and to esamate carcinogenic risk in communities exposure to emissions from coke plants. The cohorts include appmxjmateiy 5 9 . W men from steel company plants in Alleghcny County, Pennsylvania (Allegheny County Cohort) and 9,800 men from coke oven plants outside of Allegheny Cwnty, Peansylvanta (non-Alleghm y County cobM). Among other findings, the results OC previous investigators have shown a highly significant excess mortality among steelworkers em loyed as coke oven workers/-'*' `-'`This paper focuses on mortality risk among coke oven workers and presents the first report of updated patterns of caw-specific mortality experienced by coke oven workers through 1982. This updated information provides 30 years of WO& h i s ~ l yand vital status follow-up for 15,818 individuals representing all the workers included in the non-Nlegheny County cohort and a subset of workers in the Allegheny County cohort. This represmts a substanml enhancement of follow-up i m that previously reponed. On average, the updated data contained 14.5 additional years of work expenewe and 7 years of additional viral s t ~ m sinformation for each worker- The number of ob- IOV-18-70 WED 15:21 FALK LIBRARY (PIT) FAX NO. 6489020 Pa 04 998 Coke Oven Workers: X-Yeer Mortality Update Costadno et a, served total deaths was increased by 7 8 , and the observed number of deaths due to respirdtoq cancer and cancer UT the prostate were each increased by mom than 25%. In ad- dition to providingdata for assessing the loog-term carcinogenicrisk from occupational exposure to coke oven emissions, an import;mt ;rspect of obtaining long-term fdlow-up data is that it can be used to assess patterns of carcinogenic risk across periods oE time before d after the implementation of environmental emission controls and occupational exposure limits. Materfals and Methods The non-Altegheny County cohort comprises coke oven workers and frequency-matched non-oven steelworkers employed from 1951 to 1955 at 10 coke oven plants outside of Allegheny County. These plants 32e lwated in areas thmughout the United StaLes and Canada. The original Allegheny County oohon consists of workers employed in 1953 at seven steel plants within the county. ~ W Oof which are coke oven facilities. The Allegheny County cohort subset reported in this study includes all coke oven workers from these two facilities.and a matched smple of mn-cike oven workers. For both cohorts. the sampling vf non-oven workers was performed at a 2:1 ratio (when possibte) where non-oven workers were matched by age, mce, and date of first employment. The specific rnehdofogy for constructing the cohorts. classifying work a r e i ~ ~an, d obtaining updated work hisrory and vital status information has k e n &$ailed in previous I.3-6. I 2 The popuhtiorr of individuals classified as coke oven ;u#i non-coke oven control workers has changed slightly from that describecl in prior analysis. Coke oven work history experienced by individuals subsequent to previous assessments required the reclassification from non-oven to oven worker of 23 indi- viduais. Also. in the process of up- ciilting information it was possible to verify work histories included in .prior analysis. On occasion. when verifying the classification of work histories previously obtained, de- scriptions of some workers' jobs were discovered t h t wccc muLe detailed than chose available previously. .ha result of ibis more detided information, the classification of 32 *additidworktrs was rnodikd. Previously, work history infomation on file was complete only through I966 for the non-Allegheny County cohort and through I970 for the Allegheny County cohort. Vital status was complete only through t975. In this report, all work history and vital status informationwere cx- tended though 1982. Original plant work records were reviewed and ab- stractedby staffof thc Department of Biostatistics. 'Iheidentificationof vital status for those no longer employed in the sael industry wps obtained from employerbenefir ceconfs iind tiom the Social Security Administration. Verification of all deaths was made by obtaining P copy ofthe deah certificate from the appropriate state bureau of vital statistics. To be consistent with the clavsification scheme wed in prior updiues, the underlying cause of death far each newly identified decedent was c d e d according to the sevenrh nvisim of the international classification of dis- eascri.Is To maintain continuity of job categorization. 311jobs were also coded accordiag IO the previously defined c1;lssification strategy. These classitications were defined m the basis of the physical location in which il worker was stationed during the performance of his assigned job. Coke ovenjobs were subsequently grouped i n b thvx main caiegories. The categories inulude jobs in which: ( I ) all of the work effort was spent on the top of the ovens (topside fiit!-time); (2)part of the work effort was on top and some of the side of the ovens (topsidepart-time):,and ( 3 ) all ut'thc work effort was on the side of the ovens (side). Workers ut jobs on the top area of rhc ovens recejved the highest levels of exposure to coke oven emissions. Exposure coke oven emissions for each groq was measured io terms of coal ta pitch volatiles (CTPV),the class o polynuclear aromatic hydrocarbon d13t is tht: benzene-soluble frroctiol of total particulate The av m g e daily exposure for the job cot. egcwies w8$ determined in previousl! published work as 3.15 m@m3 fa topside full-time jabs, 1.99 mgim for topside part-time jobs, and O.S{ mdm3 forside jobs.I4 The m a a m ments used to develop these average w e n obtained by an extensive indus trial hygjenr survey conducted in t h ~ Iate 1960s. Unfortunately. exposun measuEments for subsequent ped. ods are not available. 'rhus, the dc veiopment of a definitive curnulotivc exposure measure that would incop prate changes in e x p u r e pat- over time was not possible. How- ever, a wei&ed exposure i&x %- counting for the relative JifTerenm in exposure intensity of the thrse classifications of coke oven jabs was determined for each worker and used as an indirect measure of e x p o w response. This index was det'inedas W sum. over all jobs. of the pmducl of the number of months ut ajob ami the weight for that job. The weights chosen forLopside futl-time. topa part-time:, a d side oven jobs wett thc average exposure IO CTPV of 3. i 5. I .99,and 0.88,mpectively. lnternnlly standardi7ed esdmadcs of relative risks for cluw-specific mortality were cdculated using an extension of the Mantel-Haend stutistic, using non-oven workers the baseline. The cstimotes vm% tested for sWistical signiticance us- ing a suhrnory x' with one d e p e of In addition, tests Of trend were pertomed to evaiuate thc expnsure-response re~ationshipfor years of employment and for ledE of weighted exposure index." w- tality among coke oven workers u*F asses.& by race for 33 categoria Of cause of death. Patterns of mOrmliv from cancer of the lung. trachea, imnchus ond OM pramre Wmcr were further assessed through FAX NO. 6489020 P,05 JOEM Volume 37,Number 5, May 1995 599 TABLE 1 EJurnberofWorkers by Cohort and Race Allegheny County Non-Akgheny County Totrd White Nonwhite -.- Tdal Whrte Nommrte Total Wh#o - -bhwttcte TOM 985 la60 2025 1d 3 6 --1860 3296 2401 - -2920- 5321 1967 2065 a2 2886 3579 I___ 64s 4863 5644 10,407 2932 3125 6057 4322 -_5439 9761 7254 8564 15.818 @sons of re[ative risk by yean of coke Oven employment, CTPV. cumulative exposure levels and period of follow-up. All relative risks were adjusted for age, race, coke plant and period of follow-up, as appropriate. These annlyses were accomplished using the software package OCM A P . ~ O - * ~ In the circumstances when tbe expccted number of deaths fop a given analysis was less than 5, the relative risk was not reponed. Results The distributionof workers by cohort, wwk area, and face is shown in Table 1. A total of I5,818 me0 are hcIuded in tbs analysis. Of these, 5,321 are coke oven workers and 10,497 are non-oven workers. The Propartion of nonwhites from each cohort is similar: 56% for nonAllegheny County and 52% for Allegheny County. There is substantial variation in the racisJ.distribution by work area, which js evident in both cohorts.Table 2 provides the distribution of coke oven workers by work areas rlr time of entry into the study. Overall, 69% of coke oven workers who worked topside full-time are RonwhitE. Among those who worked topside part-time, only 9%m nonwhite. Among non-cwea workers, 54% are nonwhite. The relative risk of cause-specific mortaIity among coke oven wcvkers is shown in Table 3. Statidcdly significant excess mortality is present for seve~alcategories of deaths. The relative risk for ail causes of death is 1.11 (P< .001). This increase in total mortality is primarily the result of excesw in deaths due to cancer, particularly deaths due to respiratory and genitourinary cancer. The rdarjve risk for ail cancers is 1.38 (P c .mi)f,or respiratory cancer 2.07 (P < -001). and for genitwtrinary cancer 1.49 CP < .01). The excess death from respiratory cancer is atujbutable to elevated risks from cancer of the lungs, bronchus, and trachea, where the dative risk is 2.13 (P< .001). The excess death from genitourinary cancer is due to elevated risks from prostate cancer where rhe relative risk is 1.60 ( P < .Ol). These findings rn consistent acms racial categories (Tabk 4). However, the excess risks are greater among the nonwhite workers than orthe whitt worker;. To further assess the nature of rcspinrory cancer risk. determinations of relative risks by three measures of exposure to coke o v a emis- sion were undertaken. This included evaluations by years of empIoyment at coke wen jobs, by years of em- ployment at jobs on top of the coke ovens, and by C P V cumulative exposure. Table 5 provides the relative risk of death from cancer ofthe lung, bronrhus, and trachea among coke oven workers by 5-year increments of employment at coke oven jobs. The risks for individuals who worked lessthan 5 years at ovenjobs TABU 2 Distribution ofW e Oven Workers by Cohort, Race, and Work Area at Time of Entry into the Study C- 'L----, Megheny County Race WIt9 Nonwhite Tatal Topside FulCflme Tapside PW-Tlme Number Pmtxmt Wwnbcr & r a m i 77 ~- 226 303 25.4 -74 6 tco.0 304 a __I 312 97.4 2.6 c__ 1W.o -Side ovnr W m b -a -P-~rccnt 584 41.4 -_ 828 --- 58.6 1410 100.0 Nan-~i-cg~nty 1 Total 4- white NonMie 286 584 32.5 67.5 ____l__- Totel 880 100.0 w\M? 363 30.7 Nonwhie a20 68.3 ------------ Total 1183 700.0 324 85.9 826 53 14.1 1213 -PI-- 377 1W.O 2039 628 gT.1 1410 61 8.9 2039 669 100.0 3449 40.5 59 5 100.0 40.9 SB.. 100.0 Totd _I_-__I Number Percent 965 47.7 -io60 __52.3_ 2025 7oao 14% - 1-8-8-0 3296 43.6 56.4 100.0 2401 2920 45.1 54.9 5321 100.0 NOV-19-70 WED 15:24 -I-- FALK LIBRARY (PIT) FAX NO. 6489020 P, 06 TABLE 3 Relative Risk of Cause-Specifi Mortality Among All Coke Oven Workers CWSOdDaelh -- All causg DN.oe.thofs 2291 R1.k 1-08"" 964L Con(ldrclCs Lhlb I 02-1.14 5B6 12 199 25 25 69 264 255 9 87 sa 1 12 16 40 4 40 1079 163 701 83 132 I25 81 36 45 15 23 130 66 66 140 1.341.25 0.88 0.6s 1.11 0.m 1.89- 1.95- 0.m 1.497.57 t 2.03 1.14 t .26 t 1.01 1.02 0.90 1.03 1.03 1.12 1.18 0.82 0.69 0.97 1.13 0.60 1.04 1.04 1.17 1-04 ?.1%1.50 0.60-2.66 a.70-1.ti 0.41-1.12 0.66-1 .a6 .z0.67-1 1.55-225 1.5s-2.33 0.41-2.34 I . 11-2.01 1.09-2.30 t O.Eb1.53 0.131-2.12 am-1.90 t 0.137-1.B 1 0.9c1.10 0.74-1.09 0.94-1.74 0.78-7.38 0.84.1.4`1 a.a-i.4a 0.82-1.08 0.96-1 -03 0.661.41 0.57-2.28 0.37-1.02 0.63-1 30 0.78-1.41 0.84-1 -63 0.84-1.30 t Not determinsd when capsctsd values were less than 5. CNS,central nervoussystem: HO, Asart disease. is 1.33. The rites peak for those wha worked 15 to 19 years at suchjobs is 2.91. T h i s trend of increasing risk W N I i~n~cIreasing years of employment is highly significant (P < ,001).Tn- ble 6 provides this same type of analysis, but is based on years of employment on topside coke oven jobs, rhc jobs that have the highest levels of exposure to coke oven ermssions. An even stronger exposure-response relationship is evident using this d4nitioa uf cuke oven exposure. The relative risk for an individual wodcing less than five years at top-side jobs is L6?. The rclative risk for an individual who worked 15 yean or more at topside jobs is about 2.6 times as high at about 4.4 (trend &st: P < .001). Relative risk by levels of weighted exposure index also revealed a strong exposure-respnnse relarionship (Table 7). The nsks across cat- egories of exposure exhibited more -=than a 3-fold increase @end test: P .OOI). Efforts to assess exposure-re- spoase relationship for deaths due to cancerof the prostate were restricted due to thereiative mfrequmcy of this event. Assessment was thus limited to the analysis acms weighted exposure groupings. which had to be mi- Lapsed to only four categories. The results of this analysis is S ~ O W Oin Table 8. From this data the existencl of an exposure-responserelationshil is not dear. The relative risks fa those in the highest and lowest expo sure categories are similar at 1.4 whereas the relative risk for those t the middle exposure category is 1.9 To determine if the risk of lunl cancer mortality among the coho1 was constant over time, relative ri& by categories of weighed exposun index were determined for thm p ri& of fonow-up: r b u g h 1965 1966 to 1975, and 1976 to 198; (Table 9). When controlling fm ex @.sure index, there appears to be 3 amnuation of risk over time for rll categories except that of 1 to 199; Tbe attenuation is apparent in 1966-to-1975 interval and is sw. tained in the post-1975 interval. De pending OR the category of exporn index, the relative risks before 1966 are about 1.5 to 4 times greatertha those after 1975. Discussion The resulrs presented here an consistent with the findings rrported from previous updates of both thc Allegheny m d nun-Allegheny County cohons. Occupational e x p sure lo coke oven emissions is assw ciatcd with excess mortality fconr cancer of the respiratory system and cancer of the prostate. The naturc af the association for cancer of the respintory system is very strong. Gem enlly speaking, the relative risks by subgroups are large and stndsticajly highly significant. Exposure-respowe rehtionships- are evident when considering both years of em ployment at coke oven jobs rmd weighted exposure index. In addition. the magnitude of the exposures response relationship is enhanced with a condition on years of emplay- rnent at those jobs with the intense exposure, spccificdty jobsd top of the coke oven, These fin- provide strong evidence SUP*: the contention that the noted @: ition between respiratory c m occupational e x p u r e to emission is in fact B casud dOEM Volume 37,Number 5. May Y 895 FAX NO, 8489020 P, 07 6M Ail cau6es 891 1.04 0.96-1.13 All I;ancers Buccal and pttarynaael Digestive aod pernsrneal Largeintesfne PC#XEitS Other digestive Respiratcr], lungs. bronchus,trachea Other respfmtory Ganito-&nary PlUstatB lastis and othsr genital Kidney Bladder and other winery LWWBnbMmmFJOieuc skln All other mcer Cafdiww-rend disease Vascular ;edom of CNS Arteriosclerotic HD Hypertensive HD other csrbiwaacutar-mal brnalignant nsptratory Nonmalignant digestive c i s i s of iiw othw nonmahgnant q e s - tlve 7ubercutosis ~#betesmennus Acclelenm Hdmicidrs and suicides K-defind causes mAll CaUBes 213 7 38 8 11 27 86 92 3 37 19 0 8 10 17 2 19 450 54 33 14 49 $0 33 12 21 3 6 57 19 20 40 1.24 1.52 0.87 0.441.05 1-09 1 59" 1.62" t 1.57. 1.77 t t am 1.14 t 7.73 asa 0.89 1.oo D.92 1SI9 1.12 0.71 OSO' 0.92 I 0.47 1.73 1.14 1.74 0.92 1.0&7.48 0.543-4.31 0.a-1.23 020-0.94 0.52-2.1 6 0.88-1.74 l.l*ZlZ 1.2&2-?7 t 1-03-2.44 0.M6 t t 0.461.97 0.64-205 t 0.65-13 7 am-1-11 0.64-1.22 0.88-1.14 0.4&1.71 0.76-1 56 0.79-7 59 0.47-1.05 0.25n.32 0.55-1.54 t 0.2c-1.11 B.82-1.56 0.65-1 -99 0.943.27 0.633-1.34 'P < .05, "P .01,"'P c .w1. T Nol detennined if the obaemd md expected v a l w were less than 5. CNS, centraJ newus system; HD. heart dlseasa 1400 353 5 73 71 14 42 179 173 6 50 39 I 4 6 23 2 21 629 109 368 # 83 75 40 24 24 123 17 73 47 48 f 00 1.11- 7 .e f 0.89 1.O? 1.17 0.79 2.12'2.220.89 1.42 1.4f t t 7.93 1.39 t 0.90 1.05 0.91 7 .a7 1.06 1.14 1.24 0.96 0.# 1.03 0.98 0.71 0.96 1.oo 1.oo 1.71 1.04-1 2 0 127-1.66 t 0.66-1.22 asi-2.02 0.5G2.43 0.54-1 .18 7.64-2.70 1.?0-2.84 0.3Q-2.69 o.gc212 093-2.37 t t 0.61-5.84 0.78-2.48 t 0.50-1.59 0.94-3.16 0.71-1.16 0.9.3-1 2 4 0.78-1.45 0.84-1.53 0.88-1 .'tD 0.85-1.41 0.53-1.53 0.54-1.79 o.46-2.ia 0.3&1.37 0.71-1.31 0.69-1 .d4 0.m-1 .a 0.8s: -4s hi.Similar fiidings have &en reparted from studies conducted throughout the world.*" Smoking hisrories of the workers hluded in this study are not available. Although we were not ab!e to Wntrol for smoking in OUT analysis, is not likdy that differtnCCS be- been coke oven workers and nonwen controj workers m regard to Smoking habits could be SO substantial as to change inferences regarding thr? association of c&c oven e x p and iespiraiory cancer. Thb conclusion is based on h e find- ings. FirsL atre is a strong exposureresponse relationship. If the association between coke oven employment and respiratory cancer was actifactlial, caused by a differ- ence in sindung panem, one would mt expect to see such signifint exposure-response findings. Second, the non-oven workers were pIant- specific internal c m ~ n o l smatched by age. race, and date of hire. They represeat blue-collar cdleagues of the coke oven workers who for the m m past reside in the same communities as the coke oven warkcrs. Based on socioemnomic and demop p h i c factors. it is un&cly that there is any substantial differences in smoking habits between the coke oven worlrers and non-oven conmIs. Third, the respiratory cancer risk for some top-si& categary coke oven workers with at least 10 years of employment show a 4.5-fold increase compared with non-oven workers. Differences in smoking habits between these worken and controls would have to be very ex- mme to negate a relative risk of this m@tude." Estimates of the effect i -19-70 WED 15:26 FALK LIBRARY (PIT) FAX NO, 6489020 P, 08 - 602 Cake Oven Workers: -Year Mottality U W e 9 Castantino a J -TABLE 5 Relative Risk of Dmth from Cancer ofme Lung. Bronchus, and TracheaAmang All Coke Oven Workers Yeam oi Cake No. at No. of Relathe i%%conRd.nco WhnpioVmont Risk R i a urnits 0 10.487 203 1.oo 1-5 1724 39 1.33 0.92-1 .a9 543 524 ia 1.37 0.a-2.51 10-14 503 26 I-&? T 26-2.99 15-19 533 47 2.91" 2.27-4.52 120 P-value 2037 125 2.71 1 .%2.85 Trend tact c.001 'P e .M.-f < .001. ~~ ~ TMLE 6 RelaWe R id Oeath from Cancer of the Lung, Bronchmand Trachsa among All Coke Oven Workers Who Worked Top-Side Jobs 0 1-5 5-9 10-1 4 15-13 220 Pvalue Trend lest 'P < .os. " P .Dol. 10,497 856 362 23t 152 248 203 31 24 24 20 27 1.a0 1.w 2.584.254.45" 4.34- .W1 1.41-2.51 1.75-4.23 291-7.14 279-7.56 ZH94B7 -~ -~ ~ TABLE 7 Relative Risk ofDeath from Cancer of the Lung, Bronchus, and Trachea Among Ail Coke Oven Worlcers by Weighted Exposure Index -Weigbd Erposure No. at Index (mghn3 mo) Wsk No. of Deaths Retathe Rlrk 95% Confidence Limits 0 1-49 50-184 200-349 360-499 500- 649 2650 P-value Trend test 10.497 1576 1062 976 882 348 677 203 34 43 56 39 27 S6 la0 7 24 1.58 1.96' 2.02' 2.er 3.13' c,001 o.aw.81 1.13-2.27 1.482.79 1.55-3.19 1.98-4.57 2.4S4.56 'P .= ,001. of confounding indicate that a differ- primarily began in the early 1970s to ence as large as 90% smokers in implement emissions controls and to coke oven workers and 30%smokers reduce levels of occupational expoin contrnis would be sufficient to sure through the use of protective negate conctusions of rssociation for devices (respirators, work-am isolarelative risks of on~y2.5 OI ~ower.~" tion cabins, etc) have been benefi- The reduction of respiratory can- cial. Mowcvw, an alternative expta- cer mom~icyrisk Over intervals of nation couId be that intensity of Cotlow-up suggests that &om that exposun is a critical fxtm and thar those who obtain a ceruin level of exposure in rhe first interval of fa- low-up have a greater intensity of exposun: than those who obtain the same level of exposux after 30yem a follow-up. Additional work should be undertaken to resolve this issue. A worthwhile assessment would be to compam the reSUlts from this study with those from a simi~m investigation conduced among a new cohort of workers using a point in time. around 1975 as the time window of employment far c- entry. However. the most efficient time 10 undertake such m invesliga- tiocl would be after 1995, when available follow-up would exceed 20 years, thereby providing a reasonable l i k e i i b d of detecting occupation. ally d a t e d lung cancers. Sam+ size considerations may necessitate an even Ionger period of follow-up. Employment in the steel indusmy has been declining for many years, and the rnnnber of individuals entering the industry as new employees hap been low. Thus, a LSn5-based cohm would be much smaller than those available From eariier time periods and would contain only P small proportion nf ernptoyees whose work history was concenttared in periods after enactment of occupationai ex- posure standards. ln contrast to the findings regard- ing respiratory cancer, excess mortality tiom cancer of the prostab:has not been a consistent finding in Othef studies among steelworkers. Inter- preting the nature of the noted ciniiori iwtwwn exposure to C& oven crnivsions and prostate in this study is not as sraigtttfomd *as it is for respiratory cancer. This dificulty is primarily due to sparseness of data for stratifiede d ucltion. There were only 58 canCfls of the prostate observed in the 4 worker population. Thus, cWs' classificationof risks by type o f c d oven job and duration of mP!Oy' men1 is limited. As such, it is &*; wir w make sound regarding expsure-responM tionship with this disease. n nn 1 ai JOEM Vdume 37,Numbor 5, May 1995 803 of 31- of he 31s Irk his 31 m b a lint me olt at 9La-- 20 >k m!e- 3ct IP Las nd ng `Bs lzt 5c ds *0- ds x- dm- I er P 0. a ci rc i Y J r r S Y I mu 8 Rdative Risk of Death from Cancer of the Prostate m n g NICake oven Workera by weighted Exposure Index WdgMadEXVSW3 *.at mderr(rn@n+`-mo) - R i Y 0 10,497 Naof beam 61 Rsrativc Rick 1.oo 95KConfldence --urnits 7-199 200-389 2400 P-uakrs 2838 1276 1407 21 20 17 1.43 1.93` 1-44 0.a1-237 1.I1-3.21 0.80-2.41 Trend test 0.51 - -~- ~~ TABLE 9 Math Risk of Des# from Cancer of the Lung, Bronchus, and Trachea Among AI1 cake Oven Workers by WeQhted Expasure tndex zvrd pniodof Fdlcvzr-Up No. of deaths P t-lW 2Cw399 2clM99 2800 Rprtative risk P-value Tmnd test 0 1-199 200-399 400-589 -a0 `P < .05. `* P C ,061. 10,497 3072 1304 534 411 44 17 22 9 27 IM) 1.09 3.00" 3.356.7F <.001 9426 2314 1181 550 465 72 33 m 17 22 1.00 1.66 1.71 2.40- 2w <.Dol 8010 1840 855 453 439 a7 27 23 17 15 1.oo 120 1.w 2.29" 1.67 ,021 Rostate cancer is essentially a dis cse of oider men. Death from this ascast among workers under 60 of age is relatively rare, with thc majority of deaths occurriog afkr %e 74. Based on age of last follow- UP, the proportion of workers over Ihe age of 74 was oniy 16.7% and 10.0% for the AUcgheny County and Nan-Allegheny County cohorts, rewtively. To adequaWy cvalnatc association between prostate cm- Ccr and occupational exposure to coke oven emissions, it may be nee- to follow the cohosts for sev- aal more years to obtain the in%sed numbers of observed cilsw required for an adequate evaluation of this cancer risk. Approximately 59% of all themen kingfollowedin these cohorts were still alive at the end of 1982. About haIf are under age 60,and 80% are under age 70. Thus. additionai follow-ilp could provide information usefut for this t y p of evaluation. Th: association betwceo CTPV exposure and respiratory cancer could be expfored further by tnathematicat m~dehgwith muItistage" and twostagc" models of carcinogenesis. Work to incorporate time- dependent exposure patterns into tbese modtls and to develop com- puter soft war^ ta handle large-cohort datasets has been recentlycompleted by our g r ~ u p . ' ~N~o'nupdaced information from sub.sets of tbe cohurt was used for this work. Findings from the prellirnjnary assessments made in Coajunction whh this work suggat a possible late-stage carcinogcoic effect of cxymure lo cokeoven emissions in addition TO a0 earlystage effect. We uliN attempt in future to assess the validq of these tindings by nndertaking a comprehenshre quantitative risk assessment applying the extended models to the full complement of data now available for the cohort. References 1. Uoyd IW, C i o A. Long-wm r n d - ~ t ystudy of S t e C l ~ ~ kI. aM~~ I o ~ o I - Ogy I OCCV Mcb. t96k 1I:ZW-3 IO. 2. Robinson H. b l p e n n manr\ity study ut stea~warkcts1:1. Morralicy by level of ineomc in whites and norrwhitu.1 &cup Med i~tl:41L416. 3. Redmond CX, Smiih EM, Uoyd ru/, Rush HW. tang-- m o d t r y ~hrdyof srctlworkcrs: UI. Pollaw-up. J 0cctt.p Md, 1%Y;11:513-521. 4. Lloyd JW,Ludn Fe. R m CK. Ckisner PB.Lonpmn mcotaIiry study of stcclwcnk~:N.Mortality by wark m.J occup M e d 1970;12:151-157. 5. Uoyd IW.hag-cetm i n d i t y shrdy of steelworkus. V. RespintMy cancer in cake plant worbxs. J Ocrrcp Mrd. 1971; 13:53-M. 6 Rsdmond CK, (3aceo A, Lloyd JW, Rush HW.Long-wm martalky sWry & rutlwarkers: VI. M d i t y from malig- oant oc4plums mmng d e wea work- USJ. OLXWMed. 1932;14:@1-629. 7 . LEnr TJ. Redmoad CK. Bmlin PP, Salvin L,KushHW.Long-term mmaiity w-study of steclworkcrr. MI.Mcmaljty term tmong m e opcrataa. J &up Med. i 9?41h&8-6 14. 8. Redtnond CK.Gustin J. KammE.law- kmlImmlity sol* d stctlwakar:Wl Mortality panems pmmg open hanh workcrr. 3 Ofcvp Med 1975;17:40-43. '1. Manrmdar S. Liar T. RutmLIod CX. ipng-mm a l i r y study of sfeelwork- ar: IX. M d i ~ py a W s among rhea and an mrll w&rS I Orcup M d l975;17751-755. 10. HE,Rcdmcmd CK. h n p m mamIity srudy ofsteelvmkers X Mor- tality putem among masons. .IOmup M e d 19?6;18;543-545. 1t. MatumdarS.Redmond CK.SalleeitoW, S m i n a n N. An epidemiological study of 9-70 WED 15:ZQ FALK LIBRARY P I T) FAX NO, 6489020 P, 10 Coke Oven Workers 30-Year Mortality Update Costantino et ab wpnrum lo coal IRr pitch valati'lcsamong w h oven wo*em. J Air Pdluciun CUR- rrd Aswc. 1975;25:3-389, 12. National Institute for Occupational Safety and Health. Long-rtnn morfaliiy silldy of seehvwkas. CinciMlli: Unitecr S&i& Depanment of Health and H u m Scrvicca (NIOSH PuMic3tia No. $1- 120): 1991. 13. Radmnd CK Cancer rnanaliry among coke oven workmi. Environ Health Per- s p m . 1983;52:67-13. 14. Mazumdar S, Rcdmad CK,Cormtino JP. Patwnrdhm RN, zhw SYJ. Recent developments in mulurpec modeling of ~ I W Tdata for Carci-ic rkl: B S ~ S - mart. Envimn Healrh Perspecr. 1491:W. 271-277. 15. Manwl of the hernational Srotisricui ClilFsfiiafim of Discass, injwies and Cansr of Drutk Geneva Switzerland: Warld Health Organidon: 1957. 16. Fannick N,Canshon LT.andSbockley3. Exposum (0 coal tv pitch volatjks at cokc avens. Am I d Hyg Assoc J. L972 33:461-468. 17. Muriel N. Hotzwzd W. Smisticai eyncts of the analysis of data from mtrnrpmive studies of dirrasc. J Narf Cancer lmi. 1954;22:719-748. 18. U;moel N. Gvnluation of i w i v r l data and two new rank onkr smisrics arising in its consideration. Cuncer CRemorhcr REP. 196630:163-17O. 19. Hakulinen T.A Manrel-Hnenszcl ststfistic for testing h e asroeiaiion betwecn a p~yEh0tomou4expasue and a rare mt- anne. A m J &&mid lYS1;I 13:192-197. 20. Marsh GM,Preininger ME.OCMAP A uscr-anemed c c c u ~ a a adl o n m a l ity d y i s p r 0 5 m . The Amcriwn Stai- M. IbDp CE. HenltJi Cow- ,# 1980,34:245-246. Sncdring, C a m cmd Ctvonie h n g Dis. 21. Caplan RJ. Maoh GM. Erhiins PE A eaw in du Workpkm: A Repon o f t & gcnaaliztd e M v e expasure modelhg Surgean GmrraL Wadtiqtm 0.C.;U.S. pmgram for arscssing dose-response in Government Printing Office: 1985:114- epidemiologic investigations. Compwt 116. Biomcd A~J. 1983;16:587-S96. 31. Anfkrml u, et al. Quainnalive ap. Z. Manh GM,EMandJ, Paik M.Preiningu poachcsin use tomesscvlcurisl~Risk L1. Rhc cqdin DcMAp/pc.A urn-prientai aeeup.rional~wrtalicy~pm grim for the LEM pc AnI sa 1986,m. 508-309. Bioi WJ,Brown LM,Portan Lu,Stone BJ, Fraumeai IF. Lung cancer among - 32. 33. hd 1983;3:277-295. Mml~uvkarSH.Knudron AG. MUand cinocc: a model for hnmn au&. gcnsis. I Nad Cnnccr fny. 1489.81: 4 t 1-420. Dong M Y Redmond CK. M w u W s, Cosmntim JP. A multistage npprorcb to 10ng-tC~1~ t l c d w o t iA~#~I f~E. p a m O l . h e CDhan ;uralysis of lifetim lung m- 1983;f 17705-716. Fer risk among soclworkecen exposed 24. &rtmd JP, chw N. Patris & et al. coke own emissions. Am J &pub&, Mortality due to respintory catcurs in 1988;1283860-873. the coke o m planw of the haaim coal 34. EAazUmdrrr S. Rslrond CK Costn* mining industcy. Br I fnd Mud 1987;44: JP,wrwardhan RN. Zhou SYJ. Rreert 559-565. developments in the miciaage mc&ng 25. Snkak H,Tsuchijn K,Tok8kun N. et al. of cohort d u a for cxchogmic risk u- Lung wnm among coke oven warkers. sesunent. Environ Health Penpen Ind Heal&. 1975:1357-68. 26. Dnvics GM.A m a l i t y study of coke oven mrkm in twn Sarth Wales integrated steelwaters. Er J I d Me& 19n. 34:291-297. 27. llurley JF,Archibald RM. Collins PL, Fanning DM,Iacobscn M, St& RC. The morrallty ofcdre w o r h in Britain. A n f lnd Md. 1983$591-704. 28. Swscn GMH, Slangen JJM. Valonics A. Hayes RB. Scheffar T, Stumans F. hknvdiky of coke plant workers in the Naherfud~&. JlnslMrJ. 1931;4%:130- 199r;w-271-2n. a.35. &ti SYI, Mantmdar S, Rcdmood Dong MH,Coumtho JP. Compulatioo of adjusted mte and likime risks fmn mupadonal cohon dara: a p k a g e Using Fortnn Wd G U M . Cc~n- pu# Biunwd Res. 1991;24:29-46. 36. MPZUIIWIW S, W a g S. Tw-MUtcllla Modeiing of Coke Own Workers M o d - ify Dara Prrocotsd at Spring Maxhgof The Biometrics Socicry (WAR), Haw ton. TCXPMS.amh 24-27. 1991. 37. Hedmond CK Mazumdm S. Den'@, Anaiysi.s a d Inlcrprrtcuiun sf &qp 135. Term Mwrdiity Studies of Coke o*n 29. Axelsm 0.Aspects on confuunding in Workurz Pracienced nt the Jlhh A d oocupyiona~health epi&mia(wy. S o d Meeting of tk lruemaionol swistid I Work Envrrchn Hsolrh 1978;498-102. Instirute. Cilm. Qypr. Sepmnbu 1991. The Radar That May Save Lives, If Allowed To As thunderstormsrumble acms tht summer sky. the Federal Aviuion Administmiion is two ycm bchiad scheduk on a pfanned 47-airpon radar system that could warn pilots of violent wind shits. "his d s , ..c r d of a US Air DC-9 might have been pnvenred if charlotte's jetport httd what f known as Terminal Doppler Weather Mar. Charlotte was schrdulrd to get a 120-foorhigh h p g e r tower a year ago, but thc governmwr and the property owner could not agree on a price for the land. Charlotte is not alone. Tht towers must sit 8 to 12 miles from the a i r p m and io llne w i l runways. New Yorkes. objjecting to ugly towers near their homes and womcd about cadiation (a fear h e FAA terms groundless)lmveblocked systemsiu Kennedy and L3Cmrdiu airports. Uthrr sites am snagged in cfispures over wetlands. hazardous dumps. and hdm ZdfraCts. h d,only 10 towers have ken delivered. The fist one to help p h c t I d - d t Houston-is .&.duled 10 beyin service this month. 1From dOUTLOOK." in US.News & World Report, July 18, 1994, p 8.