Document nkZ6EK2qZZEK7oJJmMmkJo6E8

Soft Tissue Sarzuina and Xon-Hodgkin's Lymphoma in Workers Exposed to Phenoxy Herbicides, Chlorophenols, and Dioxins: T w o Nested Case-Control Studies .. *! b8 396 Epidemiolqn. Juls 14%. \ d u m c 6 Numkr 4 HERBICIDES. DlfiXINS. CANCER 397 I ,' X..' 398 KOGEVINAS ET AL Epidemiology Julv 1995. Volume 6 Number 4 ndntj ut e x p u r r . \K'iiiiin c x h iii.ir'hcJ x , t s < ' x p w t r c s :AClJ l l ; L ? \ \,<>E = ;.;. ;5 ... mcurring less than 5 years betore the dare ot selection ure to any polychlorinated diben- i!i!,v ,he .id\ '.kc'ic i,ii..r,.j. F..r c,i<b. : R = q.6:9jrY, CI = 1.1-23\, a n d ined, rub)ecrs were grouped into rhLw nn e.xp,)sure to TCDD (L7R = 5.2; 9j'.',o Cl = 2.9-3:i. inp a cumulative exptwre x o r e ICE) otless than G.L\L\j, Exposure to chlorophennls and to all other chemicals , ~ . r r q ~ n . i i ntgL) .11i t s p " u r c ~t le<: Thin ! h v ] , i n d e x m i n e d ( n u .marerials and pr&xesschemicak) was nor those ever exposed (CE 2 0.1)05). Ever-exposed workers associated with rlsk of soft tissue sarcoma (Table I ) . were furrher classified into those with low exposure Odds ratios for exposure to different chemicals were only (C.005 9 CE < I . corresponding. for most jublects. to minimally modified when adjusted for period of first an exposure of less than 1 year); medium exposure ( I 5 employment and time since tirst employment. CE < IO): and high exposure (CE 2 IO). The odds ratios s h w n in Table 2 for selected major In addition. two alternative clwcificntinn~of exposure chemicals are based o n cumulative exposure score in \\ere wed. 111 rhe tircr. iurnu1,itive cxyosurc >c.Irc tLwrc.ite<<xic-In#inesroseJ.!ow. medium. hirh cxp.- esamined i\ ithLutrime Legin:. In the -emnJ. - u b j c c r ~ wrc\ \\'e t+ierved 'in incrzaiine riik with exposure ror in e x h m.icc-hd ,er u c r c :iiihcJ ILL, rdin; tL' L c i ~ . ~ t l ~ -2 , + P \CF;i. r hi;he.t ;.irc.'?:rv = l 1 . T : Qj",, CI = rive exywire icore (rank 1 corresponding to the lowest 3.9-309; P-value for trend = 2.21), 2.45-T (OR tor CE, rank 6 IC) the highesr)." Rank within each mnrched iiighesr c3re:gm = 7.7; 3 5 ' ' ~SI = 2.5-47;; P-vaiue for set was entered in the regression equations as a contin- trend = O.O7), any polychlorinated dihenzodioxin or uous variable. Resulrs for these two exposure classifica- furan (OR for highest caregorv = 19.0: 95% CI = ? tions were similar to those obtained tor the main analysis 1.3-1236, P-value for trend = G.W8), and TCDD (OR i (based on cumulative exposure using a 5-year lag). and for highesr category = 10.6;95% CI = 0.6-671; P-value are not reporred here. Results are available from the for trend = 0.04). authors upon request. A model including exposure both to 2.4-D . i d Exposure to six other pesticides (or classes of pecti- ?.4,5-T (as dichotomous variables) resulted in Iiwer i d e s ) was also examined (arsenicals. dichhrdiphenyl- ORs than those obtained in the univariare analvsis rriihluroeth.ine (DPT).dintyseb, Iindinc. ~~r:an~>rhos- (2.4-L): LX = i.,;9.5,'" CI = ?.S-lS: ;.-l,j~TC. R = yhates. rriarines). These were prduceJ in the >ame or 1.i.950/, CI = C 2-13.:), Similarly decreased ORs were djacent Jep.irtments r,) rho-c prtdiiiiiiq the rhrn<)x\- tLiund tor J rnldcl including l.4.S-T \C'R = 1.4: 35'-'i herbicides or were sprayed by the same uorkers spraying CI = 3.2-1 1.3) dnd bICP.4 (OR = 9.9: CI = I.2-lX')8 phenoxv herbicides. .+parr from the phenoxv herbicides a model including ?,4-D (OR= 1.5; 95% CI = 0.1- .and related chemicals, [ h e x six pesricides constituted 16.1) and MCP.4 (OR = 8.0; 95% CI = J.4-182). and rhe m w r ccvnmonlv tv'ciirrint! chemicals amonp the a m d e l including the three herbicides (2.4-D: OR = >L L h L ~ r tI>\ l i t : i.1-c.i .'t - . i r c L v n$~,r !trr.yhtm.i. .A Jich(v- I - I . ~ i " o i ' i = ~ . l - l j . ~ . ~ . 4 . 5 - T : L 7 Kl=. 4 . ~ j ' < ) C I = .viiLu Li'iwiiL.iiiLm <>I c i p ) , u e i\ irrilz.! J :-I! ,: J!c:?i,. c7R = ; JS'j, (1: r i-1;:). however. m a v encompass c(msidcrhle mi~i.issiiication. decrea.*e in d R > uhen multiple exposures are titred m c e intormation on dates and quantities oi prduction ccmcurrentlv is largely the result ot correlation between and jprnvlng d the i i x pesticides was not available in all ditferent exposures. cohorts. Of I I c m s with sarcoma. i were production workers, and 4 were spravers. The OR for exposure to any phe- ST+TICTIC +L As +LYSIS noxv herbicide in production workers was 3.5 (CI = ! We uaed cmditional loqisric regression .inalvsis.-' L'unJititnil es.icr :ntrrencr \vnc ow1 in The d r ri-iie Garcoma studv whenever cumuiarivr c x p w r e w.is cnrrrci ? . Q - V ) , t'or r\-r('sure ro .inv chlorophenol was 2.3 (CI = C.4-19.3), and tor expsure to anv polvchlorin irrJ J i b e n : ~ ~ ! \iiGn i i r ;ur.in i\rl> 4.5 (CI = :.S-l5'. in rhe mcdeli .icxewrical v,iri.ible nith t'cwr levels. ii-inc rhr .i.rri.riinl p i ~ h , i ~Lc. cS.ict,Turh\ Ycnr.; Repression models limited to jprqers were based on v e n .mdl number\. .All t8wr .rra-.cr iaaes were exposed to 3t trom tircr c u p , s u r r until Jiwase <,r wlecri\m control (as .I dichoromous vmable. less. Lir more. than I2 wars) lc.i>t {me yhcnG,xi herbiiidc. Z t rhe I: p , i i e r i<>nrrtIL in the i w r mnrched 3et5, eight were s x p ~ ~ eadn.d tour dnd vear of tirst e x p u r e (as 3 Jichotomous variable. betore o r atrer 1970) were ex.imined as potential confounding varihles. were nunexpobed to any phenoxy herbicide. Exposure ro dinoseb (OR = 5.4:95% CI = J.7-45) and DDT (OR = 1.3; 9506 C1 = 3.+-)C) was wociared with increased risk. whereas maller or no excess risks were o b w n e d tor marines (OR = 0 . i ; 95% CI = 02.)-I 1.3). menicals (OR = 2.6; 9 5 8 CI = 2 6 - 4 . 6 ) . orennqhixyhares (OR = 1.5: 95% CI = 3.3-6.j), and lindane (OR = 2 ) . .+djusrmenr tor exposure ro any phenoxv herbicide reduced the OR for dinoseb iOR = z.4; 95% CI = 2.3-11.1) m d tor DDT (OR = 2.3: PO CI = 3.1-0. i). ORs k r phenoxy herbicides. chlorophe- nols, and their cuntaminanrs were little modified when adjustment was made tor exposure to dinoseb (OR for Epidemiology J u l y 1995. L'olume h Numhrr 4 HERRICIDES. DIOXINS. .4N@ C.4SCER 399 TABLE I . Odds Ratios and 95% Confidence Intenals (CI) for Soft Tissue Sarcoma and Non-Hodgkin's Lymphoma by Cumulative Exposure to Specific Chemicals. Cumulative Exposure Is Lagged by 5 Years. Risks Are Relative to That among Sonexposed Chcrn~ai' 5 m T ~ m rSdrCLirnaT EXl*'"LJ C,w\/Ct~niroI> 17R 05% Ci Ncm-Hdgkin'r Lvrnyhornat - EXpd CaserlCmrrols OR 05% CI 19/85 12/56 I0/3 5 15/76 1.25 I I1 185 C 36 -$ , _" i3 .:.--c- 1 !l-;vv4 ?3/i8 11/39 I s4 I93 t! f 111 . r: 400 KCGEVINAS ET AL Epidemiolw July 1995. \'ohme 6 Number 4 T-LE 2. OdJ5 Ratios for Soit TirJue Sarcoma by Level of Exposure ro Selected Chemicals: Cumulative Exposure Ha*Been LdggeJ hv 5 Fears Chemical' JnJ t\p<'-urc An) phenoxy herbicide Nonexpovd Low itedium H:gh Numkr .dC.i,ei/ i,r8,r,:, 1!5 :!4 h13 5/13 i'R 1.0 16.99 i.64 11.96 95'C: Referent 3.65-1327.5 0.5M34.9 1.03-701.9 I.4.i.T.TP Smcxpxed Law MdlW High MCPAIPIB Noncxpovd Low \tdium High 6 0 1 1.3 Rclereni l;5 1.91 3.33-13.:9 113 i.52 0.09-63i.l 316 i.65 3.46-476.J 1/16 1.0 Referent 1/6 13.17 0.55-9613 4/12 10.28 3.90-141 i 4/11 10.87 3.84444.; .\nv Jtoxin in iurm Nonerpxd Lou 4ledium !txh -< !'" 1/31 I 2 Referenr 4/12 4 40 C 59-53.41 I/6 3.18 3.N-35 31 416 IJ.Y5 l.26-1:36.3 . .I > SC',.. .., TABLE 3. OJdz Raiioi f,xr S o n - H d g k i n ' s L\niphun;d $, Level of Exposure to Selected Chemicals: Cumulative Expo. ."re Hac &en Lagged bv 5 Years - - ..Chemical' and E-.. ,.il:< Zumivr Ill C 7 w r l r-7T,-l. (3R J+.. t;! Anvt phcnoxv herbicide Nonexposed Low kledium High IJ/il 6\22 6/33 il3C I?.. 1 54 1 26 1 36 :.J-DiDPIDB+ Sanexpused LOW Medium H!m 2.4.5-TKP Sonexpad Low Mdium High MCP'4/P/Bt Nonexposed Low Medium High 17/82 6/29 411C 511 i Referent C.3c-3.15 :18-2.18 2.424.16 .4nvt chlorophenol Y Lmlrc.*F.bcJ LOW Medium Hich Pcnrachlorophend Noncxpomf LOW Medium High I : Referent 4.19 :19-29.59 in, JI, <.r, 'r !ur.>aX;.mciroxd hw Medium Hi& TCDD Nonexpusnl LOW \tedium H:ch 11/119 4/lS 118 4/13 I.? 142 J.63 3 56 Referent 2.j4-4.6) (1-13.67 ?66-19.20 . than those ohrewed in the cohort .in.il\>i>.but iiverdll, they x c o r J well w i t h rhe r e u l r s ,)t rh.ir .inilv>is. -(. : Discussion .4 5-year time-laeged malvsis was dpphed. ignoring 3 ; The two case-control studies were conducted within an any chemical exposures in a marched set wcuning less r ; international cohort ot utxkers expused to phenoxv than 5 vean betore disease occurrence (cases) or selecherbicides. chloruphenols. and Jioxins. T h e study pop- tion into the ttudv (controls). The underhing .issump- ulatiun incluJes workers wirh rubsr~ntinlexposure IO tion is that exposures in J 1.m .iraee ut J carcinogenic these chemic3ls. .is documented bv measurements of process do not affect :he timing di occurrence of the serum levels ot dioxins in workers':'"' and bv intoma- disease. although the .election \reCitlcdllV o! J 3-year cion on the job hisrories or' the workers m d on the limit is xhirrary. A loncer time limit would prubablv be prduction historv ut each plant. In comparison with the less Appropriate, since ,me oi the mdpr chemicals ex- previous cohorr morralirv Jnalvcis,:.!y the rwo case-con- amined. TCDD, hJa been shewn to be J cGmplete car- rrol studies included two to three times more cases and cinoqen with a potent rumor-promoting action. '* The used modeling to estimate individual exposure. Risk semiquantitative tvpe ai analvris (due to the use of estimates in the nested case-control studies are higher relative scures of expowre) d ~ dnor J U ~ CaI m~ ore exten- 402 ICCGEVINAS ET AL Epidemiology July 1995, Volume 6 Number 4 t