Document NEkpYqyoG2LMwn4wB31RBkzED

Mortality Patterns Among Worke in Three Texas Oil Refineries Terry L. Thomas, M.S.; Richard J .Waxweiler, Ph.0.; Rafael Moure-Eraso, M.S.; Sharon Itaya, M.D.; and Joseph F. Fraumeni, Jr., M.D. ' Thc cause-specific mortality experlence o f 2,509 active and retired members of the Oil, Chemical, and Alomic Workers Internullonat Union (OCA W) who worked at three oil refineries In the LiraumontfPort Arthur area o f Texas was examined to determine whether there niiyht be unusucrl patterns o f fatal disease related t o workpluce exposures. Deaths that occurred between 1943 and 1979 were identified from membership records of the OCAW heudyuurlers in Denver und from the records of union loculs in Texas which represent workers ut the three refineries. Deulh certilicutes were obtuincd from stole vitol records offices for the decedenls. Cuuse-specificProportionate hlortality Ratios (PMRr) were calculated using the mortality experience o f ;he U.S. gcnerd popdotion as a stondurd. Excerc rnortolity from stomach cancer occurred umony octive union incmbcrs who workcd ut Rcfinuy A and oniong active and rcrircd membcrs who worked at Rrfineries U onc C. PMRs for leukemia, muitiplc niyelomo, ond other lymphomas were elevated, 'especiully umong retired workers. Relorive frcqucncic; of broin tumor deaths were significuntly elevated umong aciive members at all three refineries and slightly eievoted among retirees at Refineries A and C. Findings suggest that oil refinery workers may have elevated risks of ,these cancers ond indicate that niorc drfinitivc studies urc nccemry. I n a study of mortality pattcrns among active members o f the Oil, Chemical, and Atomic Workcrs lntcrnalional Union (OCAW) in Tcxas, rclativc frcqucncics o f Icukcmia, of rnultiplc myelo%a,--and of cancers of thc stomach, kidnev and brain were found to bc incrcascd among s& in petrolcum rcfincrics and pctrochcmic.iI plants, pxliculJrly .~nioiig~lioscwho h a d bccil uilioii inciilbcrs i'ur a t lcait 10 years.' Proportionatc Mortdlity Ratios for canccrs o f thc livcr, pancrcas, lung, and skin wcrc also clcvatcd but wcre not associated with longcr union nicrnixrdiip. Wlicii ~ . I I Jfrom tliosc icfiiwrics wiili .it IC,I\I From the Environvcntal Epidemiology Branch, National Cancer Institute, Bethcsda. M D 20205 (Ms. Thomas, Occuprlional Studies Section; and Dr. Frauineni); the Division of Survcillaocc, H a t a r d Cvrlurlions, 2nd r i r l d Sludio, Nrrronal Inrtilulc for OccupJlional Safety and Health, Cincinnati (Dr. Waxweiler); and the Health and Safety Office, Oil, Chemical and Atomic Workers International Union, Dcnver (Mr. Mourc-Eraso and Dr. Ilaya). 100 deaths were analyzed separately, increased relative frequcncics o f cancers of thc brain (Refincries A and C),lung (Rcfineries B and C),and stomach (Refincries A and 8) werc found. Rclincries A, B, and C are located within 20 mrles of cach other in the Beaumont/Port Arthur area of Texas, a region with a heavy concentration of chemical plan&. All were opened 5hortly aftcr oil was discovered there during thc carly 1900s. Although technology has changed with timc, production proccsscs a t the thrcc plants arc very similar and c o n k t primdrily of standard rcfincry opcrations that involve separating crude oil into various components including fuels, petrolcum solvents, lubricating oils, pctrolcum wax, greases and other products. Each of the plants hds a small pctrochcmical opcralion. Bccausc o f thc limilcd ndlure of the original s c t of dau' (i.c., only dcdtlis among activc union mcmbcrs), cfforts wcrc m d c to asscrnblc a morc complctc sct of dcaths. With thc additions of rctirce dcaths, this study cxamincs Lhe causc-spccific mortality cxpcricncc of an cxpandcd group o f union mcmbcrs cmploycd a t Rcfincrics A, B, and C, Materials and Methods For rhc original study, data on the 3,105 active mcrnbcrs of OCAW locals in TCXJSwho died bctwccn 1947 dnd 1977 wcrc obtaincd from :hc Mcmbcrship Dcpartmcnt at OCAW hcadquartcrs in Denver, Colorado. Among these dcccndents wcrc 1,161 who had worked at Rcfineries A, B, or C. The prcsent study focuscs only on dcccased active and rctired OCAW mcmbcn who had bccn cmployed by Rcfincrics A, B, and C. Records o f thc union locals rcprcsenting workers at these refinerics were reviewed to dctermine whether data wcrc availablc to idcntify deaths among rctircd union mcmbori and d r l i t i o n d dc41hs .murig .iCtivc incinbcrs who 1i.d workcd .it the thrcc rcfincrics. Thc union l o c ~rlcprcscnting Rcfincrics A and B had maintained rccords of dcccascd active and rctircd rncrnbcrs since about 1943; however, most of 1l1c rclircc dc.ilhs wcrc rcportcd Jficr 1960. Kccords of rctircc deaths within the local represcnring Rcfinery C were available only from 1972. Data from the original study wcrc thus cxpanded to include 1,194 retiree dedtlls and 154 dditional active mcrnbcr dcaths that occurred as carly as 1943 and as late as 1978. This number includes all dcalhs which occurred during 1979 among mcmbcrs of the local rcprcscnting Rcfincry C . Journal of Occupational Medicine/Vol. 24, No. ZFebruary 1982 135 I -Table 1 Distribution of Male OCAW Decedents by Membership Status a t Death, by Refinery, and by Race ~~ Refinery A ' Refinery B Refinery C Total I Membership Whiu Nonwhitr Whiu Status No. X No. X No. X Nonwhite No. X White No. X Nonwhit8 NO. % White No. x N onwhitB No. 1 1 Active Retired Total 1 484 48.0 79 SOB 346 475 119 77.3 210 53.0 77 56.6 1,040 48.8 275 72.9 524 52.0 8 - 9 2 382 5 2 6 35 22.7 186 47.0 59 43.4 1,092 51.2 102 27.1 1,008 1OOD 87 10011 728 100.0 154 100.0 396 100.0 136 100.0 2,132 100.0 377 100.0 Information on ddtc and placc of death was abstracted f r o m uniori rccords. Cduw o f dcath, age, scx, and race were dctcrniincd from dcath ccrtificatcs obtaincd from state vital rccords officcs. Underlying C ~ U S Co f dcath was codcd t o thc Eighth Revision of the International Classification of a Dircascs* b y a traincd nosologist a l c o r d i n g to the rules in c f f c c t a t thc time of dcath. Obscrved numbcrs of dcaths wcre compared to expcctcd numbcrs that wcrc calculatcd using cause-specific proportions of all dcaths among U.S. rnalcs as a standard, w i t h Jppropriatc adjustmcnts for agc, calendar time, and racc.3 T h c ratio of obscrvcd to cxpcctcd dcaths i s the Proportiondtc Mortality Ratio (PMR). For cdch canccr sitc w i t h a significantly clcvatcd PMR in the total group, cxpcctcd deaths wcrc rc-calculatcd using as standards thc sitc-agc-timc-racc-spccific rclativc frcqucncics of all canccr dcaths among rnalcs in thc lot31 US. and among males in the two-county rcgion in which thc plants arc locatcd. Using this approach, thc ratios o f obscrvcd to cxpcctcd dcalhs arc thc Proportionatc Canccr Mortality Ratio (PCMK) .ind thc County Proportionatc cancer M o r l a l i l y K a t i o (CI'CMK). rcqm,livcly. Statistiial significance WJS dctcrniincd by a sunimdry chi-squarc tcst w i t h onc dcgrcc o of f r c ~ d o m . ~Ratios and chi-squarcs wcrc n o t computed whcn both obscrvcd and cxpcctcd numbcrs o f dcaths wcrc lcss than fivc. intcrprctation of thc rcsults i s subicct to scvcral limitations in mcthodology. I f the OCAW study group has a l o w c r m o r t a l i t y ratc for all c a w s than US. rnalcs, PMRs rcprcscnt inflated cstimJtcs of muse-spccific risks. Sincc the PCMR dnd CPCMR arc bascd on the proportion of total cancer dcaths, they might bc bcttcr cstimatcs o f causcspccific risks i f thc age-specific r a t a for total cancer in thc study group rcscmblc thosc of thc corresponding comparison population. Sincc thc ratios are bascd on thc rclativc frcqucncy of spccific causes of dcath, excesses for some causes will forcc the rclativc frequcncics of othcrs to bc l o w c r 111.iii cxpcitcd. I`inJlly, bccusc: of m u l i i p l c conipariSOIIS, one would cxpcct a few ratios to be siJtisticaIIy significanl bascd on chance alonc. Results About 92% of Ilic dcatli ccrtificatcs rcqucstcd f r o m statc vital rccords ullico wcrc locatcd. Bccdusc or vcry small numbcrs, fcmalcs were o m i t t c d from thc analyses. Tablc 1 shows thc distribution of dcccdcnts by r x c and by m c i i i h r s l i i p stdtus JL dcath for oach plant. About half of thc whitcs wcrc rctirccs. Less than 30%of thc nonwhitcs wcrc rctircd; howcver, 43% of the nonwhite deccdents w h o worhcd IA Pl.in\ C wcrc rctircd. Table 2 shows t h e distribution o f study subjects b y age at dcath. Scventy pcrccnt of thc white malc study subjects and about 60% of thc nonwhitcs were age 55 or older a t dcath. PMRs for selected causes o f death among white and nonwhite males are shown in Table 3 for the three refineries combined. PMRs for all cancers combined were significantly elevated among whites (PMR = 1.19) and nonwhitcs (PMR = 1.23). T h e PMR for circulatory disease was slightly in- creased for each racial group but was significantly elevated only among wliitcs. Thcre wcrc rclative deficits i n mortality from respiratory discasc and f r o m "all othcr causes" in.both racial groups. A significant cxccss of dcaths due t o accidcnts, suicide, and homicidc occurrcd among nonw h i b s but n o cxccss was sccn among whitcs. PMRs f o r cancers of the stomach, pancreas, and skin wcrc clevatcd among whites and nonwhites (Table 3) and reached statistical significance whcn data for both races wcrc combined. A significant excess of prostate canccr dcaths occurrcd among whites but n o t among nonwhites. The ob>civcd numbcr of kidncy cancer dcaihs was grcdtcr t l u i i c ~ p ~ t dcmdong whitcs; howcvcr, thcrc WAS a dcficit of deaths f r o m bladder canccr. A m o n g wllitcs, the obscrvcd n u m b c r of dcaths f r o m malignant brain tumors was more than twice that expected. Thcrc was also an increased frcqucncy of bcnign and unspccificd tumors of thc brain among whites. Although many of thcsc tumors actually m a y be malignant, the tcrm "brain t u m o r " i s often used clinically to describe brain cancer on death ccrtificatcs; thcrcforc, data for malignant, bcnign, and unspocificd brain tumors are shown grouped in subsequcnt analyses. A l l of thc dcaths in thc bcnign and unspccificd category wcrc o f an unspecified nature (i.e., thc underlying cause o f dcath was recordcd on thc death ccrtificate as "brain tumor"). T h c PMR for all hematopoietic and lymphatic I -Table 2 Distribution of Male OCAW Decedents by Age a t Death and b y Race -~~ ~~~ Age at Oeath White No. % Nonwhite No. % Total No. % <45 `45-54 55-64 65-74 75+ TOTAL 233 10.9 60 15.9 293 11.7 394 18.5 96 25.5 490 19.5 621 29.1 133 35.3 754 30.1 570 26.1 58 15.4 628 25.0 314 14.7 30 8.0 344 13.7 2,132 100.0 377 100.0 2,509 100.0 Mortality Patterns Among Workers in Three Texas O i l RefineriedThornas . .. _ _ _ -Table 3 Mortriity Experience of Male OCAW Members Employed by Refineries A, B, and C, by Race, 1943-1979 I Whits N onwhita Underlyinp Cause of Death (8thRevision, ICDA) Observed Expected 0 brewed Expected Observed Expected Deaths Deaths. PMRt Deaths Deaths. PhlRt Deaths Deaths' FPqlRt, All causes of death 2,132 2,132.0 1.00 377 All cancer (140-209) Stomach (151) Large intestine (153) Rectum (154) Liver (155,156) .) Pancreas (157) Larynx (161) Lung ( 162,163) Skin (172,173) Prostate (185) Bladder (188) Kidney (189) , Brain (191,192,225,23814 Malignant (191,192) . Benign and unspecified (225,238) Hematopoietic and lymphatic (200-209) Hodgkin's disease (201) Non-Hodgkin's lymphoma (200,202,208) IMultiple myeloma (203) .+Leukemia (204-207) Other malignant neoplasms 474 36 38 a 7 32 4 131 11 39 1 15 33 27 6 68 a 19 9 31 57 4008 25.5 36.6 14.5 9.o 22.5 6.3 119.5 6.8 26.8 12.9 10.0 15.6 11.8 3.8 39.5 5.4 12.9 4.4 16.4 59.0 1.190 1.41 B 1.04 0.55 0.78 1.425 0.64 1.10 1.61 1.466 0.088 1.51 2.1 1 B 2.28 g 1.58 1.72s 1.48 1.47 2.04i.agri j. 0.9 7 79 12 6 0 1 5 0 26 2 7 1 0 3 1 2 3 0 1 0 2 15 Circulatory disease (390-458) 1.177 1,124.1 1.059 176 Respiratory disease (460-519) 65 123.0 0.538 17 Accidents, suicide, homicide (800-998) 198 203.2 0 2 7 66 All other causes 218 281.0 0.786 39 ~ ~ ~~~~ *Based on proportionate mortality for U.S. males tObserved deathdewpected deaths $Observedand expected for ICDA 225 and 238 are not included in the total for cancer $Statistically significant a t the 0.05 level -Not calculated 377.0 64.0 6.1 4.0 1.6 1.9 3.5 1.2 18.1 0.3 6.6 1.5 -1.0 0-.8 4.5 0.5 2.3 0.2 1.6 12.8 171.4 21.7 46.3 73.5 1.00 1.235 196C 1.48 - - 1.-41 1.44 - 1.-06 -- -- - 1.17 1.03 0.78 1.42ij 0.538 2,509 553 48 44 8 a 37 4 157 13 46 2 15 36 28 8 71 a 20 9 33 72 1,353 a2 264 257 2,509.0 1.oo 464.8 37.6 40.6 16.1 10.9 26.0 7.4 137.6 7.2 33.4 14.4 11.0 - 1.19 b 1.52 1.08 0.50 0.73 1.428 0.54 1.14 1.81 t 1.38s 0.14 1-.37 1-2.7 2.-21$ 44.0 1.61 2 5.9 1.34 15.2 1.32 4.6 1.96 1,295.4 1.045 144.7 0.575 249.5 1.06 354.6 0.72: I -Table 4 Proportionate Cancer Mortality for Selected Cancer Sites among Male OCAW Members Employed by Refineries A, E, and C, 1943-1979 * Observed Deatlis Stomach (151) Pancreas (157) Skin (172,173) Prostatc (1 85) Brain (191,192) Leukemia (204-207) 36 32 11 39 27 31 White Expected Ex pected Dcatlis* PCMRt 0catlis.t CPClllRt 34.1 1.06 26.9 1.34 27.6 1.16 32.8 1.00 8.8 1.24 11.7 0.94 32.G 1.20 32.5 1.20 14.7 1.84s 18.4 1.47 20.9 1.488 29.1 1.06 Nonwhite O ~ S C N C J Expcctcd Expected Deatlis DeatliP P C M R t Deatiis I C P C M R t 12 a.9 1.34 13.0 0.92 5 2 7 1 2 4.6 0.5 8.9 1 .o 2.o 1.-09 0.-79 3.6 -- 1.-40 - -- - -- 'Based on proportionate cancer mortality lor U.S. males t Observed deathshxpected deaths $Based on proportionate canter mortality for males i n a twocounty area $Statistically significant at the 0.05 level -Nul t a l ~ ~ ~ l d l ~ d Journal of Occupational Medicine/Vol. 24, No. 2/February 1982 ' I Table 5 -Mortality Experience of White Male OCAW Mcrnberr Employed by Refineries A, B, end C by Menibenhip Status at Death, 1943-1878 ' Active Underlying Cause of Death (8th Revision, lC0A) -~ All causes o f death Observed Deaths 1040- Expected Deaths. 104U.O PMRt 1.oo All cancer (140-209) 231 186.5 1276 Stomach (151) Pancreas (157) Lung (162,163) S k i n (172. '173) Prostate (185) Kidney (189) Brain (191,192.225,238)$ -Hematopoietic and lymphatic (200-209) Hodgkin's dit5ase (201) Non-Hodgkin's lymphoma (200,202,238) Multiple myeloma 1203) Leukemia (204-Zpu AI1 other cancer 21 18 67 8 a 9 25 26 3 9 2 12 60 12.6 1.666 10.0 1.79 p 55.8 1.20 4.1 124 42 1.62 5.2 1.74 . 10.9 2.296 209 1.24 4.O - 6.6 1.37 2.o - 82 1.46 64.6 0 2 3 Circulatory disease (390-458) 518 482.6 1.07g: All other causes 285 . 3709 0.77 $ `Based on proportionate mortality for U.S. males *` tObserved deathdexpecteddeaths $Observedand expected lor ICOA 225 and 238 are not included in the total for cancer pStatirtically significant a; the 0.05 level - Not calculated 0 bserved Oeathr 1092 231 15 14 64 3 31 6 8 42 5 10 7 19 55 659 196 Retired Expected Deeths' 1092.0 2142 12.8 12.4 63.1 2.7 21.8 4.8 4.8 18.6 1.4 6.4 2.4 8.2 73.1 641.9 236.8 PMRt 1 .oo 1.1 1 1.17 1.13 1-.oo 3.424 1.25 1.65 2.265 3.658 1.57 2.959 2.328 0.75s 1.03 0.83 fj brain cancer. The CPCMR for pancreatic canccr was clcv ~ i c dmwng nonwhitcs. Thcrc wcrc x v c r d l diffcrcnces in morlali ty paltcrns among white .ic.tivc and rctircd union mcinbcr, (Tdblc 5). Although PMRs for all canccr and f o r circulatory disease wcrc clcvatcd in both membcrship groups, thcy wcrc >tatisticJi/y significant only among activc mcmbcrs. Rclativc cxccsscs of stomach, pancrcatic, skin, kidney and prostatc canccrs were sccn among active mcmbcrs and to a lcsscr cxtcnt among rctirees. Thc obscrvcd numbers of brain tumor dcaths wcrc prcalcr than cxpcctcd among active incmbcrs (PMR = 2.29) dnd rctirccs (PMR = 1.65), but thc PMR was significant only among active members. There wcrc significant exccsses of dcaths from Hodgkin's disease, multiplc mycloma and lcukcmia among rctirccs but n o t among activc mcmbcrs. PMRs for sclectcd CJUSCS of dcath among white male O C A W mcnibcrs who workcd a t Refinery A arc shown in T.iblc 6 . I3ccusc o f m a 1 1 numbcrs (N = 87), rdtios for n o n w h i t e ~ t i a l cw~ho workcd d t t h i s plmt were no1 cdlculatcd. Thcrc wcrc twofold rcldtivc exccssc~of stomach and b r i n canccr dcaths among activc nicmbers. Mortdlity due lo m u l t i p l c mycloma and to cancers of thc pancrcas and pro\t.ilc w.n \li~;litly grc.ifcr tIi.111 C ~ O C L I C ~r,cg:.irdlc\s o f mcmbcrship sldws. Among rctirecs thcrc wds d significant excess of dcaths f r o m non-Hodgkin's l y m p h o m a as well as increased frequcncics of Hodgkin's disease and Icukcmia. Tdbk 7 shows thc niortdlity cxpcrioncc of OCAW mcmbcrs who workcd at Rcfincry 8. An incrcased rclativc frcqucncy of stomach canccr dcaths occurrcd among whitcs and nonwhilcs. Obscrvcd numbcrs of dcaths for cmccrs of the stomach, pancrcas, and skin were greater than expected dmong whitcs rcgxdlcss of mcmbcrship status. Incrcascd rclativc Crcqucncics o f cmccrs of thc brain and kidncy anlong w h i k i were more p r o m i n c n t among aclivc union mcmbers. M o r t a l i t y f r o m brain cancer was more than twice that cxpcctcd among white activc mcmbcrs. Thcre was an cxccss of Icukemia dcdths among white active members ( P M R = 2.59) and rctirccs (PMR = 2.46). A m o n g white rctirecs, four deaths from multiple myeloma were obscrved and lcss than one was expected. There was a significant twofold cxcess of lung canccr deaths among nonwhite mdc5 Ai: i n g whites w h o worked a t Refinery C, there was a significant excess of brain tumors (Table 8), w i t h seven observed and only thrcc cxpccted. Brain and lung cancer deaths clustered among active union members. The increased relative frcqucncy of stomach cancer was not associatcd with mcmbcrship status. A significantly incrcascd rclJiivc frccjucncy or Icukcmia Jcaths was confincd to rclirccs (I'MR = 4.18). Anwng n o n w h i k s , PMRs wcrc not incrcdscd for any parttculdr cancer sitcs. Discussion A I ! l i t ~ ~ g hI i ~ n i ~ . ~ t i i musf tlic I'Mli m c ~ l i u d arc well known, its primary purposc i s to gcncratc hypothcscs that can be tcstcd b y more definitive analytic studies. Our study also has limitations rclatcd to the "hcdtliy workcr effcct," sccn cspccially among actively c m p l o y c d w o r k c r ~ .N~cvcrthcless, the unusual rnortdlity patterns, including rclativc excesses o f brain, stomach, and hematopoietic and lyrnp h A c maligndncics, arc coniistcnt with the findings rc- 138 Mortality Patterns A m o n g Workers in Three Texas Oil Refineriesflhomas .. 1 ITable 6 -Mortality Experience of Whita hlale OCAW Memben Employed by Refinery A by Membership Status a t Death. 1943-1978 (8th Revision, ICDA) Deaths Deaths. P M R t Oeaths Deaths. PMRt Deaths All causes of death 484 484.0 1.00 524 524R 1.00 1008 All cancer (140+209) 101 84.7 1.191 110 1013 1.08 211 Stomach 1151) Pancreas (157) lung (162, 163) Skin (172, 173) Prostate (185) Kidney (189) Brain (191, 192, 225. 238)$ Hodgkin's disease (2011 Non-Hodgkin's lymphoma (200, 202,208) Multipfe myeloma (2031 Leukemia (204-207) ALL other tanm 11 7 27 2 5 4 12 0 5 2 -- 3 25 5.7 1.928 4.5 1.55 -2 4 3 1.08 19 -2.2 2.22 2.3 -5.1 2.366 2.0 3.0 1.69 -0.9 -3.9 29.5 0.85 4 6.2 0.64 15 7 53 1.18 14 30 301 0.99 0 1.3 - 57 2 -15 10.1 1.49 3 2.3 -4 2.4 - 4 0.7 20 7 16 4 -7 3.0 2.33s 12 3 1.1 5 6 3.3 1s 9 27 35.3 0.17 52 Circulatory disease (390-458) 240 219.1 1 . 1 0 ~ 325 306.4 1.06 565 All other causes 143 180.2 0.798 89 115.6 0.77q 232 Deaths' P M R t I 1008.0 1.00 186.6 1.73$ 12.0 1.25 10.5 1.34 -55.1 1.03 3.2 12.3 1.628 4.6 1.51 -7.5 2.138 2.7 6.0 2.GO!j 2.1 2.43 4,- -IJ7.8 -1.15 I64.8 0.80 525.5 1.089: 295.8 0.78b 'Based on proportionate mortality for U.S. males tObserved deathdexpected deaths $Observed and expected for ICDA 225 and 238 are not included in the total for cancer gStatirttcally significant a t the 0.05 level -Not calculated portcd in other studies of pelroleum and pctrochcmical workers. Dcspitc biases that might cxist for data sets which arc l i m i k d IO dcaths among aclivc cmployces,5I6 the inclusion o f rctircc deaths did not change thc patterns n o t e d in our prcvious study limilcd to aciivc uniun mcmbcrs.1 Analyses utili~ingthc Proportiondtc C.tncer Morlality Ratio m c t h o d did not, subsidntially changc thc pattcrns seen for most canccr sites. CPCMRs for mosi cancer siies were vcry diffcrcnt from PCMRs. Howcver, i t i s not clear whcthcr a comparison utiliLing proportions from the twoc o u n t y arca i s appropriate, sincc about 2555 of thc iota1 workforcc in thcsc two counlics is cmploycd in t h c pctrolcum and chcmical i n d u ~ t r i e s . 7 . ~Mortality rates for brain canccr a n d Icukcmia in t h e two-county arca arc higher than cy; C J i-F. -Table 7 Mortality Experience of Male OCAW Memben Employed by Relinwy B by Membership Status at Ostth and by Race, 1943.1918 __-Whitt - Nonwhilr--I----_-II__ _ _ _ _ _ _ -----.--_--.-A._ct.i_vsI_ R~et.i_r.-a.-Id . Tat& .. - _ - _ _ 1I - Underlying CJUW a1 Octth Observed Expected Obtcnrd Expectrd Obvrvrd Expected Observed Expected (8th Revision, I C O N Osalhr Oerths. PMRt Oeilhr Oisthr. PMRt Oeathr oetthr' PMRt Oerths DctIhP P M R t All w u w of death All cancer (140.209) Sramrch (151) Pancreas (151) lung (162, 1631 Skin (172, 1711 Proscale (1851 K4d-y (1891 Brain (191.192.225.2381 t Hodgkin's direare (2011 Non44odgkin't lymphoma (200, 202.2081 Muhiole mvcloma 1203) rculatory disease (390-4581 . I othrf ciusrr 346 79 7 6 18 4 1 4 8 3 2 0 7 22 172 95 346.0 62.0 4.5 3.3 18.2 1.4 1.c I.? 3-1 1.3 2.1 0.1 2.1 21.) 162.1 121.9 1.00 1.57 1.80 -0.99 - - 2.---192 2.591 1.02 1.06 0.781 382 86 a 6 23 3 8 2 2 1 2 4 1 20 228 68 382.0 73.5 4.8 4.3 21.1 09 7.6 1.6 1.6 0.5 2.1 0.8 2.8 25.1 2269 81.6 1.00 1.17 1.67 1.60 1-.09 --1.06 --- 2.465 3.-78 1.00 0.82 728 165 12 41 7 9 6 10 4 4 4 14 42 400 163 728.0 1.00 154 135.5 1.22p 43 9.3 7.6 39.3 2.3 9.2 3.4 5.3 1.8 4.3 1.5 5~G 41.3 1.62 7 1.57 2 1.04 15 3 02a 0 0.98 3 1.78 0 -1.90 1 0 0 a*L710 2.575 1 1 0.89 14 389.0 1.03 73 203.5 0.80s 38 154.0 21.1 2.7 1.5 7.2 0.1 2.4 0-.4 -0-.2 0.6 9.5 71.6 56.3 1.00 1.654 -2.541 2.0--72 --- -- II . 1.47 1.02 0.676 ired on propoitionale morlalily lor US. malei bnrwd detihr/exprctcd deaths. 'brewedmd rapcctrd for ICOA 225 and 238 Jrs no1 included in tho total for cantti. 1rliClicJity i i g n i h n t at the 0.05 hwl Iu; calculrled irnal of Occupational Medicine/Vol. 24, No. Z/february 1982 I 1 -+\- 4,Y""a'.. c L I fm R t r * * AS k 5 7 6 f l e 4 j J f d q + .3 4 ., (,'*t . f * / r , * ..., . I . C) .? c' * 1. ' 1 " +..v.11 I ~ ' I1 I ' I - 139 .. I 1f i i i ;! i Table 8 -Mortality Experience of Male OCAW Members Employed by Refinery C b y Membership Status at Death, 1943-1979 Underlying Cause of D e a d (8th Revision, ICDA) Active Observed Expected ' Deaths DerthP PMRt Retired Observed Expected Deaths Deaths. PMRt Observed Deaths Total Expected Deaths' P M R t All causes of death 210 210.0 1.00 186 186.0 1.00 396 395.0 1.00 All cancer (140-2091 Stomach (151) Pancreas (157) 57 39.8 1.438 41 38.8 1.06 98 78.6 1 . 2 5 ~ 3 2.4 5 2.2 -3 1.8 -1 2.2 6 4.2 1.42 6 4.4 1.37 Lung ( 1 62,163) Skin (172,173) 22 12.6 2 0.8 -11 12.4 0.88 33 0 0.4 2 -25.1 1.32 1.3 Prostate (185) Kidney (189) Brain (191, 192,225,238): Hodgkin's disease (201) Non-Hodgkin's lymphoma (200,202,208) 8 Multiple myeloma (203) leukemia (204-207) 'Allather t a m e r 2 1.1 1 1.1 5 2.2 0 0.7 2 1.5 0 0.4 m2 1.6 -13 8 4 2 1.91 10 -1 0.8 2 -2 0.8 7 -0 0.2 0 -1 1.2 3 - -0 0.4 0 -6 1.4 4.185 8 5.3 2.0 1-.895 3.0 0.9 2.7 2.-35 8 - 0.7 3.1 2 . 6I_ Q 9 13.0 0.69 22 26.5 0.83 Circulatory disease (390-4581 160 101.5 1.04 106 108.1 0.98 212 209.5 1.01 AI1 other causes 47 68.7 0.688 39 39.1 1.00 86 107.9 0.805 'Based on proportionate mortality for 1I.S. males tObserved dearhslexpected deaths $Observed and expected for ICDA 225 and 238 are not included in the total f o r cancer EStatisticaIly significant at the 0.05 level -Not calculated thosc for thc total United StatCS." I f cxposurcs in thc pctrolcum and chcmical industries dccount for thc highcr ratcs in thc t w o countics, thc CPCMR would undcrestimatc tliow risks in o u r study group. Allhough Ihc incrcascd frcqucncy o f stomach canccr among mcmbc-s who worked at Rcfincry A was limited to activc union mcmbcrs, PMRs for stomach canccr wcre clcvatcd among both activc and rctircd mcmbcrs who workcd a t Rcfincry B, and to a lesser dcyrcc among Lhosc at Rcrtncry C. Thew findings arc consislcnt w i t h a report of cxccss n i o r t a l i t y from stomach cancer among pctrolcum mfincry worlccrs in Canada.ll Our c d i c r study indicalcd a slight cxcc5s of IcukcniiJ dcdtli\ .iinony x t i v c union r n c m k r s w h o w o r k c d d i Rcfincry '.lE Results of the present study indicated more Lhan a twofold inireased frequency of leukemia deaths among r c t i r c d 3s wcll Js activc union mcmbcrs at Rcfincry c.B and d fourfold cxccss among rctirccs a t Rcfincry In addi tion, PM Rs for mu1tiplc mycloma (Refineries A and B) and oilier lyniphomas (Rcfincry A) were clcvatcd. AltIiouKIi 11ic rc\poiisiiilc fdctors arc not ilc.ir, i t i, notc\VUI illy t h a t Icuhci>ii.i i s .isswhtcd will1 bciiLctic cxpo~ u r c ~ * a* n1d~that other pctroclwmicdl cxpoxtrcs m.iy bc rclatcd to lymphoprolifcrativc m a l i g n a n ~ i e s . ' ~ . ~ ~ T h c most consistcnt pattcrn in all thrcc rcfincrics was llw signifii.iti1 clc-v.ctioti oC I'Mlb for 1)iJiii u u r , p.trli~ularly among dclivc union mcmbcrs. A more dctdrlcd discussion of the brain tumor dcaths in this study group i s p r c w i i t c d clscwhcrc.*G Thc incrcascd risk Jmotig activc rncmbcrs may bc duc to a p c r i o d in midlirc of inircJscd susceptibility 10 brain tumors, for which d pcak in inortdli- ty i s c x h i b i t c d bctwccn agcs SO and 69 in thc g c n c r d popiil.itiim.10 Our lindiiigs .irc of 5pcii.d i i i l c r c \ l iii view ol thc cxccss risk of brain cancer rcportcd among pctrolcum r c f i n c r y workcrs in Canada" and among workers in a chcmical plant located in thc G u l f Coast arca of Tcxar.lB Brain luinors also appcdr to bc cxccssivc dmong workcrs cxposcd to v i n y l c h l o r i d ~ ,a~nd~ s~om~ c~ studies have wggcstcd high risks in thc pharmaccutica121 and rubber15*22industries. In sevcral of thcsc studics, brain tumors occurrcd p r i m a r i l y among cmployecs undcr the age of 65. I t i s n o t clear whcthcr this pattern reflects thc age distribution of the populations studicd or a biologic phcnomcnon rcldted l o the origins of brain canccr. In summary, our analysis of proportionatc m o r ~ a l i l y ddla cnconipassing activc .ind rctircd mcrnbcrs of thc OCAW crnpluycd in lhrcc Texas pctrolcuni rcfincrics sug gests that refinery workers have clcvatcd risks of brain, stomach and hematopoietic and lymphatic malignancies. Morc dcfinitivc studics arc nccdcd t o cstimatc cancer risks in thc pctrolcuin industry and to i d o n t i f y cxposurcs that arc carcinogenic. Acknowlcdgincnts Tlrc wrlrors ttlJl1h M L I)cborrlr W. \Yliiio l o r ccdinicil a,>i>tanrc rnd hls. tiilda Young lor rnrnuxript prcprralion. References 1. Tlitmi.ts 11.. I > i t i J i i n c P. .itid hlourc.rr.iso R : hlorl.ltfiy rtiwng worhcri ctirployrd iii (rcirolcuni rclining dnd pccrorltriiticJt planis. JON 22:97-103, 1980. 2. 'Eighth Revision In!crnational Classification of Diseases, Adapted lor use in the United Slates. Public Health Service Publica- lion No 1693. Washingion, D.C., US. Government Printing Office, 1967. 3. hlonron K R : Analysis of relative survivil and proportional mortality. c O t I J J l U /Biot,>CdA'rs 7:325-332, 1974. 1. hlrntcl N .itbd Il.tc.nvcl W: Siatiriic.tl dspccis of i l i c .Inalyris 140 Moriality Patiertis Among Workers in Three Texas Oil RefiiieriedThornas -.. ' . or d a i r from rctrobprclivc studies of d i r a r . j N ~ IC/uticw / t i h i 22: 719-748. 1959. 5 - MLMichrcl A I . tiayncs SG, dnd Tyrolcr HA: Obscrvalionr on .lhe &luatlon of occupational mortallty data. /OM 17:128-131, 1975. 6. Dccouflc P and Thomas TL: A methodological investigation of f a t a l disease risks In a large industrial cohort. /OM 21:107-110, 1979. 7. County Business Patterns, 1975-Texas. US. Department of Cornmcrcc, Bureau of the Census Publication No. CBP-75-45, Washington, D.C., US. Government Printing Officc, 1977. 8. County Business Patterns, 1969-Texas. US. Department of Commcrcc, Bureau of the Census Publication No. CBP-69-45, Wahington, D.C., US. Government Printing Olficc, 1970. 9. County Business Palterns, First Quarter 1962, Part 8B, West South Central S t a t c s (Tcxdr). U.S. Department of Commerce, bureau of the Census, Washington, D.C., U.S. Government Printing Office, 1963. _.1.0.._Mason TI. hlcKiv FW. Hoover R. e t al: Atlas of Cancer Mortality for US. Coun.tics: i950-1969. - 0 H E W Publication No. (NIH) 75-780, Washington, D.C., U.S. Government Printing Office, 1975. 11. Hanis NM, Stavraky KM. and Fowler JL:Cancer mortality In oil refinery workcrr./OM 21:167-174, 1979. 12. Infante P F , Wagoner j K , Rinsky RA, e t ai: Leukemia in benzene workers. Lancer 2 ~ 7 6 - 7 8 ,1977. 13. VipliJni I:C and S a i l 4 C: Ucn/cnc and Lcukcinid. N Crrql / Mrd 271~872.876,1964. 14. McMichael A I , Splrtas R, Kuppcr LL, e t al: Solvent ex- posure and leukemia among rubber workers: An epidemiologic study. / O M 17:234-239, 1975. I S . Monson R R and Fine L j : Cancer moriality and morbidity among rubber workers. / Nor/ Cancer in51 61; 1047-1053, 1978. 16. Thomas TL, W u w c i l c r Rj. Crandall MS. e t al. Brain cancer among OCAW members in three Texas oil refineries. A h N. Y. Acod Sc/. In press. 17. Theriaulr G and Gouiet L: A mortality study of oil refinery workers. /OM 21:367-370, 1979. 18. Alcxandcr V , Leffingwcll S S , Lloyd J W , e t al: Drain canccr in pctrochcmical workers - a case series report. A m / / n d Med. 1:115-123, 1980. 19. Waxweiler R j , Stringer W , Wagoner J K , et al: Ncop!astlc risk among workers exposed to vinyl chloride. Ann N Y A c a d Si/ 271 ~ 4 0 - 4 8 ,1976. 20. Monson R R and Pelers Jhl: Proportional mortality among vinyl chloride workers. Lancer 2:397-398, 1974. 21. Thomas TL and Decoufle P: Mortality among workers emploved in the pharmaceutical industry: A preliminary invertiga- tion./OAl 21:619-623, 1979. 22. hlancuso TF: Problems and perspective in epidemiological study of occupational h c d t h hazards i n the rubber industry. Environ Hrvllh Persprct 1 7 ~ 2 1 - 3 01, 976. I Journal of Occupational Medicine/Vol. 24, No. UFebruary 1982 141