Document KYY7yzMNd64BXKomG9NDYXLr

FILE NAME: WR Grace (WRG) DATE: 1986 DOC#: WRG342 DOCUMENT DESCRIPTION: 1986 Journal article - Cohort study of mortality of vermiculite miners exposed to tremolite. British Journal of Industrial Medicine 1986(43): 436-444 British Journal o f Industrial Medicine 1986;43:436-444 Cohort study o f n w rtu iity Table I M ate deaths hy act K>death (ICO m iei Cohort study o f mortality o f vermiculite miners exposed to tremolite JC MCDONALD, AD M CDONALD. B A R M STR O N G . AND P SEBASTIEN From the Schoolo f Occupational H ealth, M cG ill U niversity, M ontreal. Quebec. Canada AHaw es L*ryttx(16tj Trachea bmuchur ta i Iw f <1 H em m d rnnhwirmau (1611 Oewpbagw* Cota *ad reawn (152-1) Otherabdominal (154-9) O ther (140-14 9 .1 6 0 .164-20 Heart dbewe (J90-429) Respirator? lubemtM is (010-01 abstract ^Acohort of406 men'employed before 1963 for at least one year in a vermiculite mine i Montana was folUWed'upuntil July 1983. The vermiculite ore as fed to the mill contained 4-6% r . amphibole fibre in the tremolite series. Vital status was established in all but one of the 406 and death certificates were,obtained and coded for 163 of the .165 men who died. Compared with'white men in the United States^ Uie-'cdhort experienced excess mortality from aH causes`(SMR'.T 17), respiratory cancer (SMR2-45), non-malignant respiratory disease(SMR 2-55), and ac&f^&a(MR .... ^2-.14). Foitirdeaths.were.fjfflmftitjtiignant mesotheliocra.ip^ptnii9.nal,mortality 2-4%). Compared with Montana deathiatesrtheSMR for respiratory cancer.was somewhat higher : ^ analyseso & e sp re to iy ^ i^ a ^ 'es^ m ate d cumulativeexposuregavea relationthatm ffnpttfii^ri " " signfficarfiS''fiom lineaxitycfFbe results o f this and case-refereht'analyses indicate an'Jncrt6il;risk . ' ..o f mortaliQ'.'from rcspiratoiyaricerin -this cohort of about.1% 'for each fibre y e a r'o ^ p l^ ..'In - relationto estimated exposure thtraortalityexperienced by the cohort from both l& g tracer and Other respirt? (400-519) Ccrebm astnlar (430-45S) Es-.-u causee Acu6eats(E 000-949) OrbwlE 950-990) - Other aw es . -- Causesot Irnuan- '; ' -IncludesRwrdeathsftomwahj a i*. . '(* t . ; i ` .. df99:cases'of,WMiBgoant d b*: ieoBtiolsTodreases with m '.t il.'icirfaitolitefg-dried4ung we ;:..v v o.cnooot.aromblir^ofLo!ri;?oi.tThbecr.caostbneps .3 9 J 3 .a m o s itc .I 2 - l, crocit mcsothdial tumours was higher'than inchrysotile mining. ` M'. in ting tissue from cl . . : . .v.V.'v '' . .. ' -' thequantity o f tremolite owners o f a vermiculite m m e m .y b b y , .Montana,, to. investigate th e heaith'of th w a n p k s e s . Morbidityand mortality studies in .the sa.mopopsdatton had been initiated by N iO SH m February 1982 but, because o f the urgent decisions facing them, the com pany wished to have an independent inquiry as quickly as possible. Having assured ourselves that NIOSH and the employees' union had no objection to parallel and complementary studies, and indeed wel comed them, p u r research proposals were finalised and accepted in August 1983 and we began work in September. Our studies had four components: a cohort mortality study (described in this paper), a cross sectional radiographic survey, an investigation o f asbestos bodies in sputum, and an electron micro scopic analysis o f lung tissue samples a t necropsy from ex-employees. Findings from the last three stud ies are reported separately.1'-3 as a re those o f '' . * . "In ...1 n r*1 *Ti-- . ........ chiysdtile, tthhough only id and the reliability of both sets of results asbestos fibre mined." 11 ..increased..-.,.:.. . . . .... u r . . - : "T- -mesothAoma-'Cnooi the ^ n c e r n T b r the_health o f u w k e js jo.tbB.miDe ii.r trcaSofilew asibc predom'r stem m ed!^noPIfrom ' exposure to .vermiculite, a' ' tissue at necropsy.19 Expe micaceous mineral for which'ho adverse effect ts sus lave shown that the card pected, but because the ore body is contaminated with ofite is similar to that o f c fibrous amphibole deposits in the tremolite series. Evidence o f pulmonary Fibrous tremolite and related minerals fail within the an x ray surrey o f Libby et definition o f asbestos* but have limited value and are Grace acquired the compa produced only in some small mines in northern Ital-. bee. examined radiograp On the other hand, they are common contaminants i 19"us pulmonary function various mining operations, notably the chrysotiL- a respiratory symptom mines o f Quebec and certain talc mines in New York efforts were made to mini State. O ther members o f the amphibole family tous antismoking campaij indude crocidofite and amosile which cause pul the risk further. Efforts m onary fibrosis, lung cancer, and malignant meso* investigate pulmonary th d ial tum ors. Existing evidence suggests indeed that included the search for most occupationally related mesotheliomas are known deceased employee caused by amphibole fibres and few by chrysolite.9' 7 terms o f duration and du. NIOSH. Although the NIOSH inquiries and our own were designed, conducted, and analysed quite sepa The experience o f men in the Libby mine exposed only to the tremolite group o f fibres was therefore of estimated. An analysis o f cancers and two maligna* rately. we discussed and compared o u r data and practical concern and scientific importance. o f !ii9 deaths. This inform methods at every stage. This was a valuable exercise: unnecessary differences and discrepancies were elbni. Few observations on the effects o n health o f trenr olitc have been made: Cohort studies o f employees < talc mines in New York State, some o f which ma. contain fibrous tremolite. showed an excess o f pul cn:.: -any to the US 1 Agency on 24 Match I' control analysis o f these d that the lung cancer risk Accepted JO September IMS m onary disease and lung cancer and one o r two cases o f malignant mesothelioma.* 4 In a postmortem study cumulative exposure. 436 m i C ohort stu d y o f m o rta lity o f verm ku lite m in o rs exp o sed to trem o lite t ` 437 j T ab let M ale deaths h r age and certified coast l iiw/ iifden'll fICD cadet Age at death tyra n t Total liners <45 43-04 A S T IE N ia n a vcrmiculitc mine in nill contained 4-6% of nit one of the 406 and . Compared with white ill causes (SMR 1*17), 5), and accidents (SMR ility 2*4%). Compared igher (3*03). Man-year :ion that did not depart licate an increased risk re year of exposure. In vboth Jung.canr,aiid-.j{ 11 n 75 IftV ......... nenph mn: L-OnI (1*0 0 0 0 0 Trachea b n a d w end lung (162) t 16 4 2t Pleura and meduulimim (MO) 0 2 0 2 O tw chapu and tomach <ISO--t) 0 I I 2 Colon and n d n (IS2-3) 0 2 0 2 Olber abdominal (154-) 0 2 I 3 Other (140-14. 160.164^-208) 0 8 5 13 Heart A eaae (390-4291 1 21 32 54 Rcspiraion tabcrcuhnB (0KMUS) 0 1 0 1 Other fopm toty (480-51) 0 * 12 21 Caritbrovarctllar (430-458) 0 4 7 II Extemal cauxc AcadatU (E 800-949) 4 t 6 U O ther(E 950-998) 2 2 1 5 O t^ ' CUBES 2 3 5 10 ' u tt known 1 0 1 2 ..a four deaths from malignant meaothcSoma coded ICD 143 (13.199 (2). and 515 (I). of 99 cases o f malignant mesothelioma an d matched controls four cases with more than 10 million-films o f tremolite/g dried lung were observed compared with one control.10 The comparable figures, in cases and controls, for other asbestos fibres were: chrysotile ! '3933,am osite42:l, croddolite 9 4 ,'ta d antbophylKte M aterial and methods STUDY GROUP O ur cohort comprised all male employees with a com pany record o f employment starting before 1 January .'.1963 and a net period of-at least-one-year.-These.ente-. . 2:0.Jn-lung tissue from chrysotile miners an d miners ria were m et by 406 menJTwdve were first employed ' the quantity o f tremolite appears sim ilar to that o f ' before 1940 and the remainder thereafter. T he vital p f .both .te^of.jfcs'ujts chrysotile, although .only a minute fraction o f the status o f each man on I July .1983 was sought. sis fibre'mined.11,5 In a study ofsix,.cases o f Inquiries were first.made o f relatives and friends by a pf. workers in ..this.jjqinc . *? ' " Glioma' from the Quebec chrysotile mines' recently retired long term company employee. F or all r uCr.;.`iiie WSs'the predom inant fibre found in the lung -those reported dead .th e date and plaocofdcath .were, ^risSUca t necropsy.11 Experimental studies in-animals investigated; for those-reportedahve. the address and -T-rohave^hoWn'thathtbe cattanogenic.potential-'pf-trem- --telephone number were-rcquested.:ln cases in which;,, S?<dhd!is:sitnikar,td th a to fo tb e r fibrous am p h ib o ietM - these inquiries failed to.produce-a.satisfactory result,:'-- E-wdeiu^'dPjpUlmdpary-fibrosis' was- firsKnDted-in'' information -was ' l ^ t t e d v t o . -.-driversy licence y rave limited.,value antLare. II mines in northern J u l y . . comment contam inants in. , notably, the chrysotile. a Ulc mines in New Yoric ` the amphibole family nosite which cause pulcr, and malignant mesolencc suggests indeed that ted mesotheliomas ate and few by chrysolite.*-1 the Libby mine exposed o f fibres was therefore o f ific importance, effects on health o f tremrt studies o f employees o f ate. some o f which may bowed an excess o f puliccr and one o r two cases ** In a postmortem study an-x ray'& u i s ^ f Libby employeesin. 1959 an d , since Grace quirdd th e com pany in-.1963,employces have been examined radiographically every;year. In the 1970s-pUlmOnary function tests were added a n d .b le r, a -respiratory- -symptom questionnaire Intensive efforts were made to minimise exposure and a vigo- ntismoking campaign was instituted to reduce - further. Efforts made by the company to r.-- -..gate pulm onary disease am ong workers included the search for death certificates for all known deceased employees. Individual exposures in terms o f duration and dust concentration were also estimated. An analysis o f these data showed 16 lung cancers and two malignant mesotheliomas in a to u t of 109 deaths. T his information was reported by the company to the US Environmental Protection Agency on 24 M arch 1983. A preliminary casecontrol analysis o f these data by ourselves indicated that the lung cancer risk was probably related to rive exposure. bureaus. Finally, the remits o f searches o f sot^i$ cu ?{ i rily files fo r NIOSH were consulted fo r.tb o te m ta -t who could n o t otherwise be traced. *"' ' v V - - The num ber alive o n 'l July 1983 was 226 and I6S were reported dead. A further 14 men could be traced ' only through sociat security files: these men were known to be living o n 31 December 1981, but their subsequent status was not established. They were included in the man-years analysis until the date a t which they were lost to view. Death certificates were sought in the state in which each o f the 165deaths was thought to have occurred and copies were obtained for 163. T h e registered cause o f death was coded by a single qualified nosologist according to the eighth revision o f the International Classification o f Dis eases. Tabic 1 shows the causes o f death. PRODU CTION PROCESS Vcrmiculitc is a mineral derived from geological transformation o f triociahedra! micas.1* On heating. '438 McDonald. McDonald. Armstrong, and Schasticn Cghort m J y ofnmrutHl, the material expands about 13 times in volume and panying minerals arc augitc, biotite. calcitc, diopside. m isek a lo g v acquires useful properties. Expanded vermiculite is hornblende, magnetite, quartz, sphcnc. and trcmolitc- |. ;s well established tha mainly used as an insulator against heat and noise in actinofite.'6 Milting takes advantage o f differences in 1.,bbj. is contaminated b; the construction industry and also as an absorbent particle shape to separate the thin flakes o f ver presence in airborne dust low density filler and chemical carrier. O f the 325000 miculite from the btocky waste particles. The finished m the report o f an indu tons o f vermiculite concentrate (before expansion) concentrate in 1984 contained 80-95% vermiculite. Montana State Board ol used in th e United States in 1984,34% came from the Before 1934. the milling was basically a dry screen has further informalior Libby mine. According to legend, the vemuculite m ountain m Libby was discovered in 1913 by Edward process. A second mill, operating a wet process, was added in 1955 to accomplish the first stages o f sepa nature o f Libby fibres. A systematic study by N Alley: walking in a short tunnel in search o f metal ration. W ork m the dfy mill was very dusty but tron microscopy o f airbe deposit, the flame o f his mining candle came in con* became much cleaner after 1965 when ventilation bnute fillers in various lc tact with vermiculite an d the immediate trans formation by heal attracted his attention. Commer equipm ent was installed. In 1974. for hygienic and technical reasons, both dry and wet m ilk were shut is now in progress. P rd ` from the wet mill indica cial exploitation started in 1923. Libby production, dow n and ore was processed in a new mill built morphologies. They ma initially low, increased from 20000 tons in 1940 to .150000 in 1950.an d 200000 in 1970. nearby which operated o n a n entirely wet process in which separation was made by vibrating screens, roll tl .meter, have a lath or they may be found mien A tp resen train in g isc o n d u c te d o n b e n c h e so rlc v c ls.r scr.eens,.- Hump...h..r.e..y....s..e..p..a..r.a..t.o...r.s.^ and .(.f.olation. .T. he. ,. ticks (fig IX S o la r, the ' ithat extend laterally 'from the surface to metre than -"n ew wet min operates continuously, producing 1100 and aspect ratio are 300 -vertical-:feet.'below the original, surface o r the ,-to n s o f concentrate each day in -I9M uhiVis'thcn <|e-;. . . 3-100gm respectively. - . : ' : m ountain-A lthough equipment has changed over the .,.,,w atw w M ried. and hauled.to plant. All - , (62%) o f fibres have be i.-. V)years,-ore;has.ajhrays.Mjeen;extracted by thq d a s a c ., _ % T O ^ ! ; y t o B . n e e h i ^ ^ . a l ^ ^ a I ^ t o n |n ^ . ; . . , 7 . -..Comparison o f d ectrt ^ ^ t S ^ . t e t * w w ^ * P ^ J ^ M i n g , loading (with shovels.::;J n g s ^ b n d .t..^ .,,>i:!. . kVicviipLuc .jicpmisshovR^bat abou' ft. w ot fro n u a d kwder^^and hailing to the mill o r trans-. ^ .. .The. scraau p g ..q p cratM a.4 g ^^ , y fibres are detect! byt!tl ; i4{*fcr-.potet::D^nwas..ion^dered the most dusty,.": riiiailite'.concentrate, diffenn^' in jgttipk.sm.aiKr^. ,. p)easuTOgexposjuc;to: . :...=J.operation-and,>in.mO^^ney'drills:,with baghouses.,; industrial.application..-TTie t ^ i|m u j ^ MSiJmg; " ' were introduced. .x; a n d shipping concentrate'at Libby have,been danger! " ' energy dispersive specto .A fter removal o f coarse rock, the ore is blended . several times with the object o f conuofling exposure :::d complicated result' . - and fed a t controlled rates to the to p o f the mill. A t. to d u s t. A t present most o f the concentrate is shipped. , .nrony fibres, waij. comp ~" this point the granules arc ab o u t I cm in diameter and - by rad, the loading station being equipped to reduce - tremofite-actinofite sen contain approximately 20% vermiculite. Accom- dust exposure; bagging is still used to a small extent. some was too high to r according to the dassi Mincralogical Associati 7-either displayed no-'cal sites in the theoretic occupied by sodium). : neskwiebeckite, o r displ response (sodium distr 'iiet and could be calk .itysis o f Libby samp what they called " sodi suited whether it was observed by optical mi> massive am phibok cry ments resembling fibre the complexity o f the fibres are probably a asbestiform and mass there may also be sorw mineral was idcniifk ahbough not reported. - .-tu or company surv e \ i>i i s i : r e e s t im a t e Regardless o f their n. caeh member o f the exposure as measured irm strong, and Sehastten Cohort study o f m ortality o f vtrm indbe miners exposed to treoudite 439 . biotitc. calcitc. diopsidc. I -tz, sphcnc. and trcmoluc- . Jvantage o f differences in , the thin flake* o f ver- J mc particles. The finished ! icd 80-95% vcrmiculilc. [ is basically a dry screen t a ting a wet process, was j i the first stages o f sepa- ; till was very dusty but 1965 when ventilation 1974, for hygienic and and wet mills were shut ed in a new milt built i entirely wet process in iy vibrating screens, roll >ns. and flotation. The uously. producing i 100 in 1984; this is then dethc screening plant. AH er clarification in a tail* qparates grades o f verg in particle size and . chniqucs fqcJutndimg . hby have been changcd { . jf controlling exposure " -concentrate is shipped ing equipped to .reduce used tb a small extent. m in e r a l o g y technique1* (the so-called " regulatory fibres"). Work Ii is well established that the vcrmkuiitc deppsil in histones based on 42 defined jo b categories were ! l'h \ is contaminated by ashestiform fibres.'^ Their obtained from the company together with detailed ec in airborne dust was first mentioned in 1956 ' information on th e process, its changes with time, and report o f an industrial hygiene survey by the the introduction o f control measures. Analysis o f the j...... tana State Board o f Health" but only recently process and observations made during two field visits has further information become available on the (by PS) showed that the 42 jo b categories described nature o f Libby fibres. the working operations with good specificity. Our A systematic study by analytical transmission elec general strategy for estimating cumulative exposures tron microscopy o f airborne fibres collected on mem was ta t to take maximum advantage o f existing air brane filters in various locations o f the mine and mill measurements and (b ) to concentrate efforts on the is now in progress. Preliminary findings on samples post dry min period (before 1975) when substantia! from the wet mill indicate that film s exhibit several exposures occurred. Altogether, results o f 1363 air morphologies. They may be straight with uniform measurements were available for the period before diameter, have a lath o r needle shape, o r be curved; 1975. They were extracted from reports o f surveys thev may be found intermingied with other platy par carried out by the company, the Montana State ' (fig I). So far, the ranges for diameter, length, Board o f H ealth, the National Center for Urban and aspect ratio are 0-1-2 pm , 1-70 pm , and Industrial Health, and the Bureau o f M in . Efforts 3- 00pm respectively. A fairfy high proportion to monitor the work environment that before 1967 (62%) o f fibres have been found longer than_5pm. were virtually restricted to measurements o f dust con Comparison o f electron and optical microscope centration in the dty mil) using the midget impinger counts shows that about a third o f the total airborne method increased substantiaUy in the 1970s with the fibres are detected by the optical technique used for introduction o f the membrane filter method.11 In the measuring exposure to asbestos at the work place.1* entire period before 1965 only 48.environmental m - . -Elemental chemical analysis o f individual fibres by surements were made; this increased to 385 for foe . energy dispersive spectrometry*'yielded unsuspected period 1965-9 and to 930 fo r the period 1970-4. Both " and complicated results. Although the chemistry o f static and personal sampling had been used. As most many film s was compatible with minerals o f the ' o f the personal samples were short term, they proba- . trcmolite-acunoiitc series, the sodium content o f bly reflected ambient concentrations rather than time - was too high to meet the criteria for tremolite weighted average exposures. .... . .. Jing to the classification o f the International ' ; A modified version o f the exjxmire zone concept" _ MincralogicalAssodation.*0 Some sodium-rich fibres was' applied .in using air mcasinmnents-data tb esti--. - .either-:disp)yed:-no calcium response'(with .-both B- Ornate .exposure. Behring m mind th a t at'any-tim e a ...' .-sit - in .tbe.'itheoretka] formula' of- ahiphiboles *- w orker (s somewhere, doing something, we defined 28 -* occupied, by.sodium ), an d so could be cattedim ag-'1' "operation locations, (table 2). These.-were defined to - .nesioriebeckite,o r displayed an intermediate, calcium ensure that every available,airmeasuremcnl.could be - - .response .(sodium-distributed between A site and B . unequivocally allocated. Information on location w as: -- -ale) and could :be called richterite. .Tray diffraction used to ebssify static samples and that on operation analysts o f Libby samples by Atkinson e ta l detected to classify short term personal samples. Operation - what they ailed "sodium treraoiite" but it was n o t ' locations were further divided in periods bounded by stated whether it was in fibrous form .16 They also dates o f significant change in process o r in control oh.-rved by optical microscopy that comminution of procedures. In the calculation o f cumulative exposure ve amphibole crystals produced cleavage frag grouped air measurements data were averaged using . resembling fibres. These observations indicate the technique o f Oldham,*3 which provides "a min tne complexity o f the lib b y situation; the airborne imum variance unbiased tim atc" of the arithmetic fibres arc probably a mixture o f different types o f mean in skewed log normal distributions. asbcstiform and massive amphiboles. In addition, In the drv mill midget impinger measurements were there may also he some exposure to q u a m since this made in 1944. 1956. and 1958 and much more regu mineral was identified in some ore samples," larly between 1962 and 1969. High dust concen although not reported in air samples taken in govern trations (38-9 m pef geometric mean) significantly ment or company surveys. dropped by a factor o f about 4-6 in 1965 after the installation o f a m ajor piece o f ventilation equipment, exposure estimates Regardless o f their nature, we had to estimate for known as tlic scirocco fan. There was no d ear indi cation o f further consistent decrease in concen `- member o f the cohort the cumulative fibre trations in the dry mill after 1965 and concentrations re as measured by standard optical microscope were taken a s constant. Measurements by the mem- -mstrong. and Sbastien ow so it is unlikely that [ iceuraey o f our cumu- J Cohort study o f mortality o f eermiadite maters'exposed to /rentable Ml Tabic 3 DeathsandSMRs bycauseand timesineefirst employment (reference: UnitedStates tehltemen). '95% Confidmce mterreh hi parentheses) vailable in terms o f jo b . ion. Wc therefore used ; reach jo b the duration ! by workers in different ' nulative exposure was . p o f company represen- j ort was compared with j nited States using the to compute expected standardised mortality :tions were made using 14 from which United ath rates, available to onal analyses o f mor*erc made using Monsi figures lo r numbers opriatdy interpolated censuses. Death rates ntana were computed nd the-United States dated to,this.year fo r- sween mortality and ' irformed in two ways... d .to itie-end.of.each US-82 - ' a Mr * " :> 08-I 7;-. J lti Iv'idt ..-0 6 tsos 0-5 0-2 06 0-7 0-2 0-S 01-*22 04 06 0-J melunchroom(09). mine l), nver stitioa conveyor Canv <-ttiratk tCDtthArerl Thorsuorpru rmptorntm t.trorsl < to /0-/V u SM R 0 SUP *20 0 SM tt CnmMclr cohort 6 SUK Alt ..iuv.'i Mahpunt ntopbun: RopirMory Abdominal Other Circulatory 4rMme NMMnalifiuua: . Respiratory disease External cases: Aaddems Other 160-161 ISO-159 390-458 460-SI9 800-949 950-998 17 0-79 45 I I I 103 1-31 165 1-17 4 5*02 - 4 16 13 2-42 23 2-45 ll-J J -tl* 5) 10-45-4-54) (1)5-3-99) i 0-86 i 044 5 Ml 7 0-88 0-- 2 069 II 1-83 13 126 3 0-31 20 093 42 097 65 0-87 0-- 7 3-36 14 3-30 21 2-S5 5 1*80 1 1-02 4 1-46 3 2-22 9 3-11 1 0-72 . 18 M 4 5 1-28 month o f each man's fife, o r until the end o f 1982 for excess o f that expected for white men in the United " who survived until then, was calculated, Slates by 24 deaths (SM R H 7 ). The SMR for all a-ycars a t risk were divided into four groups causes rose from 0-79 for the first nine years from first . .tding to cumulative exposure to fibres, and exposure, when " healthy worker" selection effects expected deaths and SM Rs computed for each group. would be expected to be strongest, to 1*31 after 20 A linear relation between SM R and cumulative years from first exposure. There was a substantial exposure was then fitted using the C U M program in excess number o f deaths from respiratory cancer the m anner described by Hanley an d Liddell.14 Sec (SMR 2*43), from non-malignant respiratory disease ondly, (dative rides were estimated from case-control (SM R 2-55). and from accidental cau$cs (SMR 214). sets chosen-from within the cohort..Controls for each T h a t was no evidence o f ahexcess number o f deaths . ' casd were djosen-as men surviving beyond the age o f from cancers o f non-respiratory rites. Eight o f the ' - death o f the case, who had been te r n atid had started 'deaths from non-malignant respiratory disease were work a t Libby mine within three years o f the case. For certified as due to.pneumoconiosis (ICD 314 or SIS). - ` bach control/ exposure was cumulated to the age o f O f the 18- deaths due tef 'accidental causes, five "-.'4 ->-o f the matching case. Estimates o f linear occurred whOe the men were employed at the mine . . .sein relative risk (odds ratio) per unit exposure and the other 13 after they had fcft'.'Four deaths were ; :wbiev'm ade'tiring the method.-.oC conditional like- from 7fte$othelioihh;'on^c(rteirby'liK'hbsoiogist to \ li.hqtod;1* ^ - .-'-"-I. f . ffis.jj ' :* respiratory,can<w(lC'165)rone!topheunKKomosis; .; ' (ICD 513), an d tw o to'neoplasm wUhoutypefiification `-R esults---..... : 'V;--,* >ofsite (K ^ T 9 9 )-T h ree o fth e e iiie s v tw pleural and *" 'ooe peritoneal. In'addttion. one tVath'w,M<tified as - ' ' 'T h e m ortality experience o fth e jjhbrt'aS a whole is due to mediastinal'cahccf (ICD 163) in a man nine ' summarised -in table 3. M ortality fo r all causes was in years after first employment A t necropsy the pathol- `..k Table 4 Deathsfrom respiratory cancer (ICD 60-163). SAfRs. andemulative exposure (reference: Montanamenj M ur. m l duraiim o f se n k e (yew ) Mean concentration ((/ml) Mom cumulative exposure" (font) Time from first employment (yearn): <10 0 SMR 10-19 0 SMR 220 0 SMR All 0 SMR C um deiiw tx p e tu rt (frnyt 0~<25 25-C20 3-0 5-8 12-5 2 667 1 1-35 4 1-68 7 204 13-8 100 77*3 0 0-00 2-94 3 1-8} 5 1-97 Z& -< 500 11-0 49*7 1 9-09 1 3-27 5 980 7-53 500 16-0 64*2 836-1 1 20-60 0 000 3 6-77 4 3-38 Alt 8-7 176 144-4 4 J-SO 4 206 1} 3-8J 2X 303 'Cum ulated to end o f service. 442 McDonald. McDonald. Armstrong, and Sebastien ` " l/m t years, ` .. Fig 2 - Respiratory cncerSMRs 20 armoreyearsfrom first " employment e` xposure (flndyears).: ' ~' ogist did hot consider th isa mesothelioma. ' ' Using M ontana death rates instead o f United States national rates reduced the number o f expected 'deaths from respiratory cancer'from 9-41'to 7-60, increasing`the SM R from 2-45 to 3-03. Table 4 gives th eS M R s for respiratory cancer (using Montana 'death-rates) by cumulative exposure. It-is reasonable . to focus attention o h the SM R s20orm ons years from first employment, when risk due to exposure to fibres would be expected to be most evident. A linear relation between these SMRs and cumu lative exposure (see fig 2) is given by the equation SM R = 1-52 + 0-01 i f/y or, expressed as a relative risk, RR = 1 + 0-007f/y. Alternatively, by forcing the line through SM R - 1a t zero exposure, the equa tion SM R => 1-00 + 0-013 f/y was obtained. The observed deviation from linearity could be explained by chance (y1 (3df) = 3-47) and the trend has high statistical significance (y1 (Idf) -- 18-9, p < 0-001). The constrained line may be interpreted to imply a r increase in SMR for respiratory cancer o f about I >3% for each fibre year o r cumulative exposure. There is also evidence from the data m table 4 o f an excess mortality from lung cancer less than 10 years from first employment The two cases in the lowest exposure category could not have been due to work at Libby. One was a man who was first employed a t age 66 and died at 69; the other joined a t 25 and died at 28. The two other deaths in the short latency category were 8-9 and 9-1 years from first employment. Both had substantial levels o f cumulative exposure (470 and 522 f/y) so a causal hypothesis is less easily dis missed. Case-control analyses o f exposure-response wen carried out for the 23 deaths coded to respiratory can cer (IC D 160-163), the eight coded to pneu moconiosis (ICO 514-515), and (with some-overtap) the four mesotheliomas. Results are presented in table 5. A statistically significant linear relation was found between relative risk for respiratory cancer and cumulative exposuie which is dearest in cases occur ring 20 o r m ore y e a n from first em p lo y m en t^ 1 (Idf) = 7*04, p 001). In this group the linearjnercase m relative ride is estimated a t 1-0% per fibre year (95% confidence ipterval 0-1-9-7). This relation.is described by the equation RR = 1-00 + 0-010 f/y. Although positiverelations with cumulative exposure .were also, observed with pneumoconiosis and with meso thelioma, 'numbers were small, confidence intervals unde, and conventional levels o f statistical significance Were not reached. . DfansfcM - / - ...... 4 - . - The cohort studied was not large but sufficient to show that the workers in this mine experienced a seri ous hazard from lung cancer, pneumoconiosis, and mesothelioma. This was presumably attributable to the fibrous amphibolc contaminant o f the vermtculik ore and mainly to dust conditions before 1974, panic ularly in the dry mill. N o precise comparison with the New York State studies can be made but the crude SM Rs for lung cancer were similar, 2-4 in this study, 2-8 in the study by Kleinfeid etui.* and 3-2 in that by TableS Estimateso fexposure: responsefrom case-referent analysis Caute o fdeath Rcsplnaoty cancer (ICO 160*163) PMunHictxuosis (ICD 515) Mesothcfatima (ICD various) No cases A A IS 23 $ A Tim*from first employment (years) 0-9 10-19 *20 All Alt A IIO 20) fi on % increase in retati** risk (O k) perj/m l year t9$% CO 2*1 (04). 66) 0 4 < -0 -l.x - 1-0 to-1.9-7 0-2 401. >0) 0 -3 (0 0 .4 1 ) 0-2 ( -0-2. >100) r ` ttd fiM Hefi " 0 3-08 1-23 7-04 7-40 242 o-to Cohort study o f n. Brown era/.* Noi the parallel study o in. In all these reported. The findings in ourown. were not Their cohort wa employed before gated 18 months on work history r> party work histor who had started b probably more c according to the fc o:.r cohort. Convc Mi-Gill cohort, al indicated that the: we used the same treat it or tnterpresame death certifk logist but with soi various -minor di ..higher than those '. Our analyses c ' relation to secure v. .selves establish x . findings from oil seere.; reasonable ... response pattern i - bon, our findings ..... (about 1% increaexposure) is the sr Vij -.Peto27T o r chrysoi substantially steep chrysolite miners tional mortality ft lower.1* Having re dian m itten were I't-eit in tow cor ; sent at mecroffc s..:`iiiar amounts amounts in cases t question the retail tile even in the mi The rarity o f meso olina chrysolite tc> great excess appe nation based sold; whereas the textile well separate the carcinogenic fibril -phibole fibres { > -iVptible to tong pe-A-tratc more d thrlioma. penctracancer, which is ii A rm stro n g , a n d S eh a sticn ; i * interpreted to imply an i tory cancer o f about 1-3% j lativc exposure. ; m the data in table 4 o f an > cancer less than lt> soars : two cases tn the lowest - liavc been due to work at was first employed at age -joined at 25 and died at .lie short latency category . i first employment. Both i jmulalivc exposure (470 ; xnhesis is less easily dis- - exposure-response were coded to respiratory can- eight coded to pneuand (with some overlap) jits are presented in table inear relation was found respiratory cancer and s clearest in cases occur- ; rst employment (y2 (ldf) lup the linear increase in -0% per fibre year (95% rhis relation is described . * + 0;0l0f/y. Although j, ative exposure were also ^ . iosis and with -roeso- i til. confidence intervals levels o f . . statistical large but sufficient to..- .. mine`experienced.a.seri-|:| r. pneumoconiosis, and ; -umably- attributable to' !- tinant o f the vcrmiculite " ions before 1974,'particisc comparison with the be made but the crude milar, 2-4 in this study. n a l* and 3-2 in that bj disk / ' lld flfo r C ll H .:f - 0 3-08 1-23 7-04 749 242 (M0 C ohort stu d y o f m o rta lity o f rerm icu lite m in ers exp o sed to tre m o iite 443 Brown e la t* N ot surprisingly, the SM R obtained in the parallel study by NIO SH ,10 2-2. was d o se to o ur own. In ail these studies cases o f mesothelioma were reported. flic findings in the NIOSH study, although do se to .\n. were not identical for four main reasons. (1) cohort was larger and induded men first on.-,-toyed before 1969, but observation was termi nated 18 months earlier. (2) Their cohort was based on work history records, ours on computerised com pany work history files. Sixteen men (alive in 1981) who had started before 1963 according to the former probably more correct records had not done so according to tire latter, and were thus not induded in our cohort. Conversely, two men were induded in the McGill cohort, although their work history records indicated that they had started in 1963. (3) Although wr used the same basic exposure data, we did not - >i o r interpret it in exactly the same way. (4) The Jeath certificates were coded by the same noso- logt but with some differences. As a result o f these various m inor differences o u r risk estimates were higher than those o f NIOSH. _ O ur analyses o f respiratory cancer mortality in relation to accumulated exposure d o not in them selves establish a linear relation. But, given the .^findings from other asbestos exposed cohorts,3 it seems reasonable to expect a similar exposure 'response pattern in this group also. O n that assump tion, o u r findings suggest that the observed slope (abont -1% increase in Jung cancer per fibre year of ire) is the sam e as that calculated by Doll and . - for chrysotile textile workers. This slope was substantially steeper than dial estimated fo r .Quebec' chrysotile miners an d millers in whom the propor tional mortality, from mesothelioma was also much .loweri*?:Having regard fo rth e fact th at ( a ) the Cana-, dian miners were also exposed to tremoiite fibres,, .'albeit m low concentrations, and ( b ) tremoiite is' - present a t necropsy in lung tissue o f these miners in similar am ounts to chrysotile12 and in higher amounts in cases o f mesothelioma.'9 one is forced to question the relative carcmogetuc potency o f chryso tile even in the mines and mills where it is produced. ' . rity of mesothelioma (I case) in the South Car- hrysotile textile plant where lung cancer was m great excess appears to exclude any simple expla nation based solely on fibre type per sc. Nevertheless, whereas the textile process, carding in particular may well separate the chrysotile bundles into finer more carcinogenic fibrils, few o f which reach the pleura, amphibolc fibres (including tremoiite) are much less susceptible to longitudinal division but remain able to penetrate more deeply than chrysotile. F or meso thelioma, penetration is the prerequisite. For lung cancer, which is initiated in the airways, as indeed is asbestosts. length, thinness, and durability o f the fibrils together with tobacco smoke and perhaps other carcinogens are more important parameters than ability to penetrate. If a linear relation between the risk o f lung cancer and accumulated exposure be accepted a I% increase in mortality per fibre year has practical implications for control in this industry- At present airborne fibre concentrations, reported by the company to average about 0-1 f,ml the estimated risk o f lung cancer over a working lifetime might be increased by about 5%. This estimate, however, is associated with consid erable statistical uncertainty and could vary in either direction by a factor o f peritaps two o r three. From the viewpoint o f epidemiological meth odology. U may be worth noting that the exposure response conclusions from this study confirmed the preliminary case-control analysis o f incomplete data' made in 1983 mentioned in the introduction. As exposure levels had been underestimated, however, the slope o f the line proved less steep than it had then appeared. Rapid analyses o f this type2* deserve to be used more often for immediate assessment of occupational hazards. This research was supported by a grant to McGill University by W R Grace & C o with whose help the. cohort was identified and past exposures estimated. We received help from several agencies in the follow up period and in particular we acknowledge the invaluable contribution o f M r Earl Lovick in all aspects o f the study. References . , ..''M cD onald JC .Stbosticn P. Armstrong B. Radiological survey of p u t to d present vcraucaHtc miners exposed to tremoiite. Sr J . 'Sebnstien P, Armstrong B. C ue B. McDonald JC Estimation of amphibolc exposure from asbestos body ind macrophage counts in sputum: a survey in vcrmkuitlc miners. Aim Orrup far (in press). 'S ebssuen P, G reen K - Case 8. Harley R . McDonald AD. M cDonald J C Trem olite-asd chrysotile fibres in human lung. Am O ctyi Hyg (in press). - ` Ross M. K untze R A . C hiton RA. A definition for asbestos. In: levadie B, ed. Dcfimtiom fo r osfireMs and inker krahh-rrlatrd stScare*. Philadelphia: ASTM. 1984:139- 47. 'M cDonald JC. Mineral fibres and cancer. Am Arad kM S m tya re 1984.13:345-52. ` Royal Commission on matters o f health and safety arising from the useof asbestos in Ontario. Report. Toronto: Queen's Pnoier tor Ontario. 19(4. 'Achesoa ED. Gardner MJ- Asheiux: tkeam troitm a far ashatax. London: Health and Safely Commission. HMSO. 1983. " Kleinfekl M. Messile J. Kooyama K. Mortality experience in a group ofasbestos withers. Arch Environ Hretlk 1967:19:177-80. ` Brown DP. Dement JM . Wagoner JK. Mortality patterns among mine and millers occupationally exposed to asbestiform talc. In: Lcmcn R, Dement JM , eds. O u t end duetue. P ari Forest South. Illinois; P alhotoi. 1979:317-24. 444 McDonald, McDonald, Armstrong, and Sebastien 10McDonald AD. McDonald 1C. PtxXey FD. Mineral conical of hrog in tncsoihcKal tumour* in North America. Arm O ttup Hyp 1982;26:417-22. " Pootey FD. An examination o f the fibrous mineral conient of asbestos in lung tissue from the Canadian chrysotde mining industry, fim m Rmeaeeh 1974;12:2111-0. 11Rowlands N. C M * GW, McDonald AD. Asbestos fibres in the lungsofchrysolitem eets and millet*. A preliminary feport. Ann Occup Mix I982S **! 1- 5. " Chtirg A. Wiggi B. Depaoti L. Kampe B.Stevens B. Lung asbestos conient in chrysoliteworkers with mesothelioma. Am Her Respir Ob 19*4:130:1042-5. l*WagnerJC.Chamberlain M. Brown RC.tin t. Biologicaleffectsof uemotilc. Sr J Canter I9S2H&352-40. ' RousseauxJM.Rooshci PG. Vidvoya LA. Herbillon AJ. The ver- miculiaatioa o f triaetahedml micas. The K. level and its cor relation with chemical composition. Clay M inerals 1973;10:1-16. '* Atkinson GR. Rose D. Thomas K. Jones D. Chatfidd E l. Going IE . Collection, analysis and characterization o f rerm icoilte tarn ' ' pies fo r fibre conient and asbestos contam ination. Washington D C US Environmental Protection Agency. 1981. (.Contract Ho . . . . 4*4)1.5915.) ' - US Department o f Labor. Asbestform onJ/orfibrous m inerab m . ' 'm ine,m ilk andquarries. Pittsburgh: Mine and Safety and Health ' ! Adminstratioo. 1980.. - .. . : *.*Walton WH. The nature, hazards and assessmento f occupational' -- exposure to airborne asbestos dust:'a review. A m Occup H y f 19*235:117- 2*7. ` ''Sbastien P. Gaudicbet A. BrfionGrdUnd MA. Jansoa X. Spe.. uomciric X purdispenton denetgieen microscopieileettonique i transmission: application la caractrisation des fibtes minr ale. Journal dr M imscnpre el Je Spectrascople O rrtrnnifnr 1910*83-97. H tnlem alioasl Mineralogical Assoctaltoa. Nomenclature of amphiboles. Canadian Mineralmcha 1978:14:501-30. " Edwards GH. Lynch JR. The method used by the US PubSc Radiologic; miners exp< Health Strives for enumeration o f asbestos dust on membrane th em Ann Otcup iiv% 1448:11:1-4. " Com M. Csmcn NA. Work placeexposure rones for classification JC MCDONALD. I o femployee exposure* to physical and chemical agents. Am lu i H fg A*** J 1979:40:47-57. From the School o f Oct **Oldham PD. On estimating the arithmetic means of tognomalty. distributed populations. Binmetnes 1945:213,235-9. " Monsoa R. Analysis o f relative survival and proportional mo tahty. Computers amt Biomrettcol Research 1974;7:325-32. 11 Hanley J. Liddell FOR. F init relationships between exposure abstract Chest rat employees (164 men. andstandardized mocratily ratios.JO etup Med 198537:555-40. . . . rsident within 200 n ''Thom as DC. General telaiive risk models for survival time and matched case-control analysis. Biometries I W 'J I ilM i. " " Doit R. Peto J. Effects an health o fexposure to asbestos. London: ^1s 43 o f th e 80 an d 24 o " 367 films were asses: Health and Safety Commission. HMSO. 1983. . ,.r ..classification. M edia " LiedxdpeollsuFrDe aKn.dTmhoomrtaalsityDf.roGmibpbnseuGmWoc.oMniocsDiso, nreasldpirJaCtorFyi'abnred, i " la tiv c e x p o s u re to th . `abdominal malignancies in chrysolite production tn Quebec, .c ,.;':;yses show ed indepen ' 1926-75. Ann Acad Med Singapore 1984;13:340-4. ` ' ' 'o isltc i D . McDoniM JC. Restricted cohort study designs. ScantJ Work Enrirott Health I9S2A30-3. -.N.-'and. o f ge a n d p ro " retirem ent age th e ir. 100f/ml years. This ; not m udi so. Destruction o f m anuscripts From 1 July 1985 articles submitted for publication will not be returned. A uthois whose papers are rejected will be advised o f tbe decision and the manuscripts will be kept under security for three months to deal with any inquiries and then des troyed. The survey described ii series o f studies underu o f work in a vermicul Two .complementary p initiated by NTSH an of contamination of amphibole deposits in and description or the were given in our repot in this group o f work siar.iiol excess mortality while men from respir non-maiignant respira and malignant mesothe = 2-4%). The NIOSH broadly defined cohort independently assemble Vermiculite. a micao heat treatment has mar tries, agriculture, and p vermiculite deposit, at commercially mined sir 5' ill scale but. after n ' :cd, production i> 'upped in I960. 150 ahout 21)0000 annually The Libby deposit Accepted 30 September 1485