Document Z4LgY12D29Onv2420b5gdQLdZ

CorporatHeealthPhysics 3M Center2,20-2B-02 CorporatoeccupationMaeldir-ine PO Box33220 CorporatPeroductResponsibility St.PaulM,N 55133-3220 CorporatTeoxicology 6517331110 3M MedicalDepartment FluorochemicaElxposure(Serum)AssessmentofDecaturChemical and FilmPlantEmployees Historicalelmyp,loyeesatthe3M Decaturchemicapllanthavevoluntariplayrticipaitned a biennialfluorochemicalmedical surveillancperogram. Totalorganicfluorinewas measured until1994 when perfluorooctanesulfona(tPeFOS) and perfluorooctanoate (PFOA) were assayed. However, thevoluntarynatureof themedicalsurveillance program does not necessarilpyrovide foran adequateunderstandingof the distributioonf fluorochemicalserum levelsintheDecatur workforce. Therefore,thepurpose of this assessmentwas torandomly sample employees attheDecatur chemical plantinorderto determinethe distributioonf employee serum fluorochemicallevelsaccordingtojobs, buildinglocationsand yearsworked. The overallgeometricmean of serum PFOS in chemicalplantemployees was 0.944 ppm. (95% CI 0.787-1.126).The highestserum PFOS valuemeasured was 10.600ppm. Among filmplantemployees,the overall geometricmean of serum PFOS was 0.136 ppm (95% Cl 0.114-0.161).Approximately 1/3ofthesefilmplantemployees had pastwork historyexperienceinthe chemical plant. The distributioonf serum PFOS levelsinthisrandorrisample assessmentwere similarto thoseobserved through the voluntarymedical surveillancperogram. Because of analyticaclhemistrymethod development,the studyalsoassayed fivefluorochemicals thathad not been previouslyassayed. Resultsfrom thisassessmentwillbe used to constructan exposure matrixintheupdated retrospectivceohortmortalitystudy ofthe 3M Decatur employee population. The datawillalsobe used torationalizseubcohort comparisonsintheongoing analysisof healthclaimsdatabetween chemicaland film plantemployeesfrom 1993-1998.Both thestudyprotocoland finalreportaresubmitted. 3M Company EPI-0006 Page Iof 29 PROTOCOL Epidemiology,220-3W-05 Medical Department 3M Company SL Paul,MN 55144 Date:September3, 1998 Title:FluorochemicalExposure AssessmentofDecaturChemical and Film Plant Employees Study StartDate: September 3, 1998 EstimatedDate of FinalReport:December 31, 1998 IRB ApprovalDate: September 3, 1998 ProtocolNumber IRB Approval Exempt Expedited x PrincipaIlnvestigator: Co-investigators: Geary W. Olsen,DVM, PhD' PerryW. Logan, MS2 Cathy A. Simpson, RN 2 JeanM. Bunis,RN, MPH' MicheleM. Buriew,M.S.' John C. Schumpert,MD, MPH' JeffreyH. Mandel,MD, MPH' Study Director: JeffreyH. Mandel,MD, MPH I- Medical Department,3M Company, 220-3W-05, St.Paul,NIN 55144 2..3M EHS&R P.O. Box 2206,Decatur,Alabama 35609-2206 ABSTRACT 3M Company Pa EPI-0006 ge 2 of 29 Employees atthe3M Decaturchemicalplanthave inthepastvoluntarily participateidna fluorochemicamledicalsurveillancperogram. Analysisof the surveillancdeatahas not shown significanatssociationbsetweentheemployees'clinical chemistryand hematology testsand eithertotalserum organicfluorineor serum PFOS levels.However, the voluntarynatureof themedicalsurveillancperogram does not allow fora completeunderstandingofthedistributiofnemployee fluorochemicalserum levels.In orderto addressthisissue,we proposetorandomly sample,forstatistical purposes,80 filmplantand 125 chemicalplantemployeestodeterminetheirserum levelsforperfluorooctanseulfonatep,erfluorooctanseulfonateamide, glycinederivative of perfluorooctanseulfonate,perfluorohexanseulfonatep,erfluorooctanoiaccid,N-ethyl perfluorooctanesulfonamideothanoland N-methyl perfluorooctanesulfonan-dedtohanol. A sub-sampleof employees (N = 30) willalsobe testedfortotalserum organicfluorine. A briefquestionnairweillalsobe administeredtoeach employee inquiringaboutcurrent and pastwork historyaswellaspossibleroutesoforalingestion.Inadditiontothose chosen fortherandom sample,allotherchemicaland filmplantemployees willbe offeredtheopportunitytohave theirbloodtestedforthesefluorochemicalsand respond to thequestionnaireU.pon completionofthisstudy,theresultwsillserveas a reference regardingfluorochemicaelxposurereductioreiffortisnthechemicalplantand allowfor adequatecharacterizationf thefilmplantpopulationwho willserveas a reference populationinhealth-basedstudiesofchemicalplantemployees. INTRODUCTION 3M Company EPI-0006 Page 3 of 29 Employees atthe3M Decaturchemicalplanthave inthepastvoluntarily participateidn a fluorochemicalmedicalsurveillancperogram.The surveillancperogram measur-edfortotalserum organicfluorinleevelsuntilthemid-I990'swhen serum perfluorooctanoatseulfonate(PFOS) determinationq,uantifiablbey high performance liquidchromatographmass spectrometryb,ecame incorporateidnthebiennialmedical surveillanceexaminations. Analysisof thesurveillancdeatahas not shown significant associationbsetween theemployees'clinicaclhemistryand hematology testsand either totalserum organicfluorinelevels[Roach,1982;Schuman, 1982]or serum PFOS levels [Olsenetal.,1998]. However, thevoluntarynatureof themedicalsurveillancperogram does not lenditseltfo an appropriatceharacterizatioofnthedistiibutioonf fluorochernical serum levels.For example,presentedinTable I isthedistributiofnfluorochemical serum levelsby thenumber of employees who voluntarilpyarticipateidn the lasttwo medicalsurveillanceexaminations(Fallof 1994 and Springof 1997).Ninetypercentof theserum PFOS measurements were below 6 ppm. Nevertheless,itispossiblethatthis distributiomnay be higheror loweras a resultofunknown selectiofnactorsthatmake employees decideto participatoer,notparticipatien,themedicalsurveillancperogram. Therefore,thepurpose of thisproposalisto randomly examine,based on appropriate statisticmaelthods, employees intheDecaturchemicalplantinorderto determinethe distributiofntheirserum fluorochemicalevels.Conductinga random sample analysiswillalsoallowresearchertsobetterunderstandserum fluorochemicallevelsin contextwith theemployees'currentand pastwork history.Inadditiona, random 3M Company EPI-0006 Page 4 of 29 sample of theneighboring3M DecaturFilmplantemployee population,locatedatthe same sitew,illalsobe testedfortheirfluorochemicaslerum levels.The filmplant employees have served as a comparisonpopulationinpriorhealthstudiesdue totheir nonoccupationalexposure tofluorochemicalsH.owever, theiractualserum fluorochemicalevelshave notbeen discerned.Healthstudiescan be mor-efully appreciatedifthedistributionosfemployee serum fluorochemicalevelsatboth the chen-Licalnd filmplantsareknown. Nffi'THODS OverallSummary-of Methods We proposeto randomly sample,forstatisticpaulrposes,80 filmplantand 125 chemicalplantemployees to detenrninteheirserum levelsforperfluorooctanseulfonate, perfluorooctanseulfonateaniidet,heglycinederivativoefperfluorooctanseulfonate(i.e., theacetateanion),perfluorohexanseulfonatep,erfluorooctanoaicid,N-ethyl perfluorooctanesulfonamideothanoland N-methyl perfluorooctanesulfonamiedtohanol. A sub-sampleof employees (N = 30) willalsobe testedfortotalserum organicfluorine. No otherserum testswillbe performed.A briefquestionnairweillalsobe administered toeach employee inquiringabout currentand pastwork historyas wellas possibleroutes of oralingestion.In additiontothosechosen,fortherandom sample,allotherchemical and filmplantemployees willbe offeredtheopportunittyohave theirblood testedfor thesefluorochemicalsand respondtothequestionnaire. SamRIe Size Issues 3M Company EPI-0006 Page 5 of 29 Four criticaflactorswere consideredtodecidethesample sizef,orthisstudy. Firsti,tisimportantthatsuch a samplebe randomly chosenfrom theemployee populationsof boththechemicaland filmplants.Second,thesample sizeisdrivenby theneed toprovideconfidencethattheexposureinthefilmplantisverysmallrelativteo thatof thechemicalplant.Third,thesample sizeshouldadequatelycharacteriztehe exposurelevelswithinthechemicalplantworkplace.Fourth,employees in thechemical and filmplantshouldbe offe*retdheopportunitytoknow what theirfluorochemicalserum levelsare. To addr-estshefirstwo pointsintheabove paragraph,we need to determinethe lower 95% confidencebound on themean of thedifferencbeetween theserum fluorochemicalevelsof thechemicalplantand thefilmplant.ProvidedinAppendix A arethedetaflsof themethods used by Dr.Timothy Church,associatperofessoratthe Universityof Minnesota,who servedas a consultansttatisticifaonrthisprotocol development. Brieflyw,e assumed,basedon pastmedicalsurveillancdeata,thatthe mean serum PFOS levelsof thechemicalplantemployee populationis1.85ppm. We do not know what themean serum PFOS levelisoffilmplantemployees. However, pooled serum datasuggeststhattheU.S.generalpopulationserum PFOS levelisin therange of 20 ppb to40 ppb. Therefore,we assumed thleDecaturfilmplantpopulationwillhave a mean higherthanthegeneralpopulationbecauseof itsphysicalproximityto thechemical plant.Ifthemean serum PFOS levelforthefilmplantpopulationwas assumed tobe 200 ppb,themean differencbeetween thechemicaland filmplantpopulationwould be 3M Company EPI-0006 Page 6 of 29 1.65ppm (1.85ppm minus 0.2ppm). Presentedbelow isthesample sizeneeded from boththechemical and filmplantpopulationstocalculatae lower95% confidencebound on thismean difference. Lower 95% confidencebound on difference1..2 1.3 1.4 1.5 1.55 1.6 Totalsample Size needed from bothplants:16 21 32 66 121 365 Thus, toobtaina lower 95% confidencebound of 1.50ppm, based on a mean difference of 1.65ppm, 33 employees inthefilmplantand 33 employees in thechemical plantwill have to have theirserum PFOS levelsdetermined.As can be seenfrom theabove chart, thesample sizebecomes increasingllyargeas thelower 95% confidencebound becomes closertotheactualmean differencoef1.65ppm. Data inAppendix A alsoprovide sample sizecalculationisfthefilmplantemployee populationisassumed to have a highermean serum PFOS level(300ppb). Inthiscase,toobtaina lower 95% confidence bound of 1.50ppm, based on a mean differencoef 1.55ppm (1.85ppm minus 300 ppb),a totalsample sizeof 760 employeesisrequiredfrom bothplants.Ifthisisreducedtoa lower95% confidencebound of 1.4fora mean differencoef 1.55ppm, a much smaller sample sizeisrequired(atotalof 125 employeesfrom bothplants).There isa degreeof uncertaintiynestimatingf,orstatisticsalmple sizeconsiderationw,hat themean serum PFOS levelof the filmplantpopulationisbecauseapproximately10 percentof thefilm plantemployees may have priorwork experienceinthechemicalplant. Inorderto addressthethirdissue(samplesizeshouldadequatelycharacteriztehe exposurelevelswithinthechemicalplantworkplace)w,e proposetostratiftyherandom 3M Company EPI-0006 Page 7 of 29 sample by buildinglocationwithinthechemicalplantT.o preventmisclassificatioofn potentiawlorkplace exposureexperiencew,e willanalyzesamplesfrom filmplant employees by thosewho have and have noteverworked inthechemicalplant.This is necessarybecauseof thesuspectedlonghalf-lifoef serum PFOS (estimatedto be at1000 days). An importantcaveatthatmust be consideredisthattheabove sample size calculationhsave used serum PFOS levelsfrom previousvoluntarysamples toestimate mean and standarddeviatiotvialuesforserum levelsofthischemical.However, inthis studywe willmeasure forfivespecififcluorochemical(sperfluorooctanseulfonateamide, theglycinederivativoef perfluorooctanseulfonateN,-ethylperfluorooctanesulfonamido ethanoland N-methyl perfluomoctanesulfonamideothanol)aboutwhich we have no informationa,tthistime,as to serum levelsI.tissuspectedthatthesechemicals arenot persistenats theyaremetabolizedtoPFOS; thusany differencbeetween serum levelsin employees in thefilmand chemicalplantsmay be quitesmall. Study esigli With.theabove issuesconsideredt,hestudyinvestigatorpsroposethefollowing researchdesign: 1. Randomly sample 80 filmplantemployees.We willassume 80% participatiownhich willprovidea studypopulationofapproximately65 employeesofwhich we estimate50 to55 willhave worked only inthefilmplant.The remainderwillhave worked inboth the 3M Company EPI-0006 Page 8 of 29 chemicaland filmplants.(Thisisbasedon theassumptionthat10 percentof Decatur employees have worked in bothplantsas mentioned by DecaturHuman Resources personnel.)The random samplewillbe chosenby thefollowingmeans: a)determineall full-timceurtentfilmplantemployees;b)usingarandom number generatoralgorithm, selecta sample sizeof 80 filmplantemployees. 2. Randomly swnple 125 chemicalplantemployeeswho areproportionatelsytratifibeyd thefollowingbuildinglocations1:,2/48,3,4 and 38/51. We willassume 80% participatiwohnich willprovidea sample sizeofapproximately100 employees. The random sample willbe chosen by thefollowingmeans:a)determineallfull-timceurrent chemicalplantemployees;b) usinga random number generatoralgorithm,selecta sample sizeof 125 employees proportionateltyothenumber ofemployees who work in thefivebuildinglocation(s1,2/48,3,4 and 38/51). Thisrandom sample of80 film plantand 125 chemicalplantemployees,if80% participatiiosnobtained,shouldprovide a 95% lower bound confidenceintervaolf,atleast1,.5ppm ifthemean PFOS difference is1.65ppm based on assuming themean employee serum levelsare1.85ppm forthe chemicalplantand 0.2ppm forthefilmplant. 3. In additiont,othe above random samples,w,e proposetosample allemployees (n 68) who areassignedtothewastewatertreatmentplant(building3s6 and 57). 3M Company EPI-0006. Page 9 of 29 4. Because ofconcernsthatmay be raisedby thisstudy,we willalsooffertotestallfull- timechemical or filmplantemployees who desiretoknow theirserum levelsof these seven fluorochemicals. 5. Studyparticipatiwoinllrequirethefollowing: a. Venipuncturewiththecollectioonfone via)ofblood(approximately10 cc)for thedeterminationof ihesevenfluorochemicalsI.tisanticipatetdhattotalserum organicfluorinecan alsobe determinedfrom thesame one via]of blood forthe30 individualwsho areaskedto participatienthisanalysis. b. Writtenresponsetoa briefquestionnairtehatinquiresaboutcurrentand past work historyalongwiththefrequencyof hand washing and use of gum, chew (tobacco)and cigarettheabitsoftheemployeewhileatwork (seeAppendix B). c. A signedconsentform by theemployee (seeAppendix C). 6. Each randomly chosenemployee(filmand chemical)willreceivea letteorf invitatiotnoparticipatbey plantmanagemeni. A copy ofthislettewrillbe senttothe employee'simmediate supervisorT.here willalsobe plant-widecommunication which willdescribethepurpose ofthisbloodcollectiopnrogram (i.e.t,odeterminethe distributionfemployee serum levelsof thesefluorochemicalasnd allowemployees to 3M Company EPI-0006 Page 10 of 29 know0 theirown individualvaluesshouldtheysodesire).Allstudyparticipantwsho are randomlychosen as wellasthosewho volunteerwillbeinformedoftheirown individual resultbsy a lettesrentto them from the3M Medical Department.Aggregateresultsof thestudywillbe communicated toplantmanagement and employees. FluorochemicalAnalyses Allblood willbe collecteadttheDecaturplantby MedAccess (an occupational healthcliniclocatedinDecaiur,Alabama) underthedirectioonfCathy Simpson, RN. Based on thisstudydesignand thefouremployee shiftasttheDecaturlocationi,tis estimatedthatblood collectiownilltakeapproximatelyfourdays.MedAccess will centrifugteheblood to obtaintheserum. MedAccess willprovidetheserum toCathy Simpson,RN who willshipthesamplestothe3M MedicalDepartment (St.Paul, Samples willbe cataloguedand thensenttothe3M EnvironmentalLaboratoryfor determinationof serum perfluorooctanseulfonatep,erfluorooctanseulfonateamide, glycinederivativoef perfluorooctanseulfonate,perfluomhexanesulfonate, perfluorooctanoiaccid,N-ethylperfluorooctanesulfonanteitdhoanoland N-methyl perfluorooctanesulfonamideothanol,viahighperformanceliquidchromatographymass spectrometryu,nderthedirectioonfDr.KrisHansen. Totalserum organicfluorinewill be determinedby the 3M SMD chemicalanalyticallaboratoruynderthedirectioonf Dr. VenkateswarluPothapragada. DataAnalysis 3M Company EPT-0006 Page II of29 Through the use of appropriatsetatistictaelchniques(e.g.s,tudentt-tests, multivariablreegressionanalysisu)singSAS softwarew,e willconduct thefollowing analyses: 1. Compare respondersand nonrespondersfrom thosechosen in therandom sample as to theirdemographic characteristi(ces.g.a,ge,gender,yearsworked). 2. Compare mean serum fluorochemicalevelsbetween thefilmand chemicalplant random samples. Determine thelower95% confidencebound of themean difference between thesepopulations. 3. Compare themean valuebetween thefilmplantand thechemicalplantpopulation who participate(drandom sample plusthosewho arenot partof therandom sample). Detenrninethelower 95% confidencebound of themean differencebetween these populations. 4. Characterizaend compare themean fluorochemicalserum measurements withinthe chemicalplantby buildinglocation. 5. Compare serum levelsby questionnairietems(e.g.e,xamine mean serum fluorochemicalevelsof thosewho may have frequenthand to mouth contactviathe 3M Company EPI-0006 Page 12 of 29 compilationof theiruse of gum, chew and cigarettetsothosewho do not have frequent use). Itshouldbe noted thatthisstudyispurelyan exposureassessmentexercise; medicalsurveillancienformation(e.g.s,erum chemistriesh,ematology,health questionnairesi)snot a component ofthisstudy.Thisinformationisvoluntarilcyollected on a biennialbasisand willbe nextofferedv,oluntarfltyo,Decaturchemicalplant employees in theSpringof 1999. StudyTime Frame ProRLDsedStudy Time Frame Task EstimatedTime Frame I- Obtaincurrentemployee lisftrom DecaturHR August, 1998 2. Detemiine random samples August, 1998 3. Send invitatiolnettetro participateorandom sample September, 1998 4. glood collection SepL/Oct.,1998 5. Serum analysesunder directiobny 3M Env. Lab Serum analysesunder directioonfDr. 'VI Oct.-Nov., 1998 Oct.-Nov., 1998 6. Data Analyses Nov. -Dec.,1998 7. Finalrel2ort Dec. 1998 DISCUSSION 3M Company EPI-0006 Page 13 of 29 The objectiveof thisproposedresear-cshtudyistocharacterizeviaappropriate statistictaelchniquesusingrandom sampling,thedistributiofnemployee serum levels of serum perfluorooctanseulfonatep,erfluorooctanseulfonataemide,glycinederivative of perfluorooctanseulfonate,perfluorohexanseulfonatep,erfluorooctanoiaccid,N-ethyl perfluorooctanesulfonanieitdhoanoland N-methylperfluorooctanesulfonamiedtohanol atthe3M Decaturchemicaland filmplantsI.nadditionf,or30 employees thepercentage of serum totalorganicfluorinethatisrepresentedby theseseven fluorochemicalwsillbe calculated. The dataobtainedfrom thisexposureassessmentinvestigatiownillbe important forseveralreasons.This informatiownillallowfora betteurnderstandingof the exposuredistributiofn serum fluorochemicalevelsinboththechemicaland filmplant populations.Second,thesedatamay serveas futurereferenceregardinghuman exposure assessmentforthefilmas wellas thechemicalplantintheareaof healthstudiesand exposurereduction.Third,thedatamay be used fortheconstructioonf an exposure matrixfortheanticipateudpdateof theretrospectivceohortmortalitystudyof the Decaturemployee population.Fourth,thisinformatiomnay providefurtherinsight regardingunderstandingcurrentfluorochemicalserum levelswith pastand present employee work historydataas wellas possibleroutesof oralingestion.Finallyt,his studywillallowfortheopportunityforchemicalplantemployeestoknow theirown serum levelsfortheseseven fluorochemicalsshould theyso desire. 3M Company EPI-0006 Page 14 of 29 The greatestlimitatiotnothesuccessof thisstudyisemployee participation. Because random sampling iscriticatlo theultimateinterpretatioofntheresultsfrom this study,itisimportantthattherebe high (i.e.8,0%) participatiloenvelsfrom thosechosen toparticipatIet.must be emphasized thatdataanalyseswillbe dependent upon high participationf thosechosen intherandom sample;nonparticipantwsilldecreasethe sample size.Although otheremployees may volunteertohave theirblood drawn, they cannotreplacea nonparticipanitntherandom sample withinthedataanalysesdevotedto therandom sample component of thisstudy.Inordertoobtainhigh participatiornatesof thoseemployees chosen in therandom sample,letterosfinvitatiownillbe senttothose chosentoparticipatbey 3M Decaturplantmanagement and the3M Medical Department. Supervisorswillbe asked to stronglyencourageparticipatiofnthoserandomly chosen employees.who reporttothem. Individuadlataareconsideredconfidentiailnformatioannd willnotbe disclosedto anyone outsidethe3M Medical Department withouttheemployee'swrittenconsent. Individuarlesultwsillbe communicated,by lettert,otheemployee from the3M Medical Department. Aggregate findingswillbe communicated toemployees and may be used in publ.icatioonrs publicpresentations. The proposed studymust obtain3M IRB approval.The protocol,any addendato theprotocol,dataanalyses,and a copy of the'finarleportwillundergo a Quality Assuranceauditaccordingto theStandardOperatingProcedureentitled",Procedurefor theCreation,Auditing,and Review and Approvalof3M EpidemiologyFinalReports. Permanent recordsof allotherdatageneratedduringthecourseof thisstudyaresubject 3M Company EPI-0006 Page 15 of 29 toprivacyand confidentialictoynsiderationsA.lldatagathcredorgeneratedincluding protocoladdenda and thefinalreportwillbe archivedby theMedical Department,3M Company, St.Paul,Minnesota,accordingto theStandardOperatingProcedureentitled, "Procedureforthe Archivingof a 3M Epiden-dologyStudy." 3M Company EPI-0006 Page 16 of 29 REFERENCES Roach DE (1982).FluorochemicalControlStudy.3M UnpublishedReport.3M Company:St.Paul,MN May 25,1982. Schuman LM (1982).LettertoDr.FrankLJbel(3M MedicalDirectorc)oncerning FluorochemicalControlStudy. April29, 1982. OlsenGW, BurrisJM, Mandel JH,ZobelLR (1998).An epidemiologicinvestigatiofn clinicaclhemistriesh,ematology and hor7nonesin relatiotno serum levelsof perfluomoctanesulfonateinmale fluorochernicparloductionemployees. 3M UnpublishedReport. 3M Company:St.Paul.,?AN.,April22,1998. TABLE I 3M Company EPI-0006 Page 17 of 29 Distributioonf Serum PFOS Levelsby Year, 3M DecaturMale Employees,MedicalSurveillancReesults* PFOS Level 1994 Data N % 0 -< I ppm 11 12 1 -<3ppm 59 66 3-<6ppm 16 18 6 Rpm Total 4 4 88 100 see referencOelsen etal(1998) 1997 Data N % 29 35 38 45 12 14 5 6 84 100 APPENDIX A 3M Company EPI-0006 Page 18 of 29 Sample size calculationfor the study of PFOS the blood of workers 3M Company EPI-0006 Page 19 of 29 levelsin Authored by Dr. Timothy Church Introduction The sample sizecalculationasrebased on theanalysisof datafrom a non-random voluntarysample from thechemicalplantinDecatur.There was a totalof 96 .observationosn PFOS levelsand 95 on PFOA levels.Table I givesthe summary .statistoincseach distributionT.able2 givesthesummary ofthelogdistribution. FiguresI and 2 are thesmoothed naturaldistributiofnosrPFOS and PFOA, respectively, and Figures3 and 4 areforthelogarithmicounterpartsN.ote thatthe logdistributions aremore symmetric thattheInaturaslcalevariables.The sample sizefocused on the PFOS levels. Method The sample sizeisdrivenby theneed toprovideconfidencethattheexposure inthefilm plantisverysmallrelativteothatof thechemicalplant.As shown in Table 1,themean PFOS levelinthevoluntarysamplesis1.855ppm and theacceptablelevelforthefilm plantisbelow thatand expectedtobe in therange200-300ppb.The statistiocf interest forthesample sizeisthedifferencbeetween thechemicalplant(assumed to have a mean of 1.855ppm) and thefilmplant(assumedtohave a mean of0.2-0.3ppm).The objectis tocompute thesample sizenecessarytoproduce 95% confidencethatthe differenceisat leastdelta,in ordertoplacethefilmplantmean exposurewellbelow thatof thechemical plant.Ifsuchconfidencedoes notresultt,henfurtherstudyof thefilmplantexposuresis warranted. The sample sizecalculationwsere carriedout by assumingthatthelog of the concentratiownas normallydistributebda,sed on Figure3,and thatthestandarddeviation on thelogscalewas 1,regardlessof themean, based on Table 2. Note thatboth PFOS and PFOA, whileshowing differenmtean values,bothhave a standarddeviationof about 1,even though themeans and medians arequitedifferent In orderto perform theanalysison thelogscale,themeans had tobe transformedtothe logscale,by solvingtheformula E(X)=exp(mu+sigma2l2), formu where E(X) isthemean on theoriginaslcale,and mu and signia= I arethemean and standarddeviationon thelogscale.For example,undertheassumption thatthe naturalmean is0.2ppm themean forthelogPFOS levelinthefilmplantwould be ln(O.2)- sigma2l2= -2.10944. Transformingthemeans to thelogscaleallowscomputationof theexpecteddifferencein the loeret-alz,n-ri-2Inwer ()i% confidencebound on thedifferenceT.he differenceon the 3M Company EPI-0006 Page 20 of 29 log scalecenterson i?zuajn.d thevarianceof thedifferenciessimply 2/11w,here niu, and iiiiairje.themeans forthechemicaland filmplants,respectivelayn,d iiisthenumber of observationpser plant.Under thenullhypothesist,hemean on thenaturalscaleforthe filmplantistheobservedmean ofthechemicalplant(1.855ppm)minus the minimum desireddifferenc(ee.g.d,elta= lppm). Under thealternatihvyepothesisitissimply either0.2ppm or0.3ppm. The mean forthechemicalplantisfixedat1.855ppm forboth thenulland alternativheypotheses. By thisformulationt,healternativheypothesisisthedifferencoef thetranformed altemative-hypothesimseans (e.g.1,.855ppm and 0.2ppm)and thenullhypothesisisthe differencoefthetransformednullhypothesismeans (i.e.1,.855ppm and (1.855delta)ppm).From thisformulationa,standardsingle-samplet,wo-sided,normal-theory samplesizecan be calculatedu,singalpha= 0.5and beta= 0.05.Iftheminimum differenciesrejectedinfavorofthealtemativet,heconfidencecoefficienotn the minimum differencdeeltawillbe atleast95%. Standardsample sizecalculationwsere performed usingS-Plus4.5 and thetransformed means, and sigma = 1. Results Table 3 givessample sizesbased on thelower 95% confidencebound forthe difference inmean PFOS levelbetween thechemicalplantand thefilmplantbeing atleastdelta. Sample sizeswere calculatedassuming means of 200ppb and 300ppb forthefilmplant. For 300ppb thesample sizesaregiven withdeltaequalto 1.0through 1.5ppm in O.Ippm steps,and for200ppb, with deltaequalto 1.2through 1.6and to 1.55.Sample sizesgiven arethesum from both plantsassumingequal samples from each,i.e.t,hetablegivesN where N=2n,=2n2, where niisthesamplefi-omplanti= filmorchemical.So,for example,ifthefilmplanthas a truelevelof 200ppb,thentheexpectedlowerconfidence bound on thedifferencewould be 1.55ifthetotalsample sizewere 122,61 from each plant. Recommendation A lowerconfidencebound of 1.55ppm,which would assurethatthefilmplanthas an upperbound of 300ppb (orabout0.5ppm) ifthetrueconcentratiownere 200ppb, is achievablewith 125 from bothplants.Ifthe,truceoncentratiownere 300ppb, a lower bound of 455ppb could be attainedwitha similarsample size.Given thatnearly100 volunteerswere obtainedpreviouslyt,hisalsoseems likea feasiblenumber of employees inwhom tofocussolicitatieofnforts. Table 1. summary statistics for data in: DecExp97 Min: lst Qu.: Mean: Median: 3rd Qu.: Max: Total N: NA's : Std Dev.: PFOS 0.10000 0.69125 1.85501 1.32SOO 2.48250 9.93000 96.00000 0.00000 1.73723 PFOA 0.05000 0.38900 1.32894 1.06000 1.87000 S.31000 96.00000 1.00000 1.0857S Table 2. *** summary Statistics for data in: DecExp97 Min: lst Qu.: Mean: Median: 3rd Qu.: Max: Total N: NA's : Std Dev.: Log.PFOS -2.302S9 -0.36930 0.22868 0.2813S 0.90926 2.29S56 96.00000 0.00000 0.93310 Log.PFOA -2.99573 -0-94418 -0.12126 O.OS827 0.62SS1 2.669S9 96.00000 1.00000 1.00902 3M Company EPI-OW6 Page 21 of 29 PFOS Density 4 to PF02 PPM Figure 1 PFOA Doesky 3MC o"inpany EPI-1)006 Page 22 4)f29 Log PFOS 0 ci L"PFO Figuri Log PPO A Donsn 0 4 PFOA pp. Figure4 -3 -1 Log(PFOA pp. Figure 3M Company EPI-0006 Page 23 of 29 Table 3. Sample sizetableforDecatur PlantStudy Assumingfilmplantconcentratio=n 300ppb Lower 95% ConfidenceBound on Differ-ence: 1.0 1.1 1.2 1.3 1.4 1.5 Totalsample sizefor both groups (N=2n 1=2n2): 20 26 36 58 126 760 Assumingfilm plant concentration= 200ppb Lower 95% Confidence Bound on Difference: 1.2 1.3 1.4 1.5 1.55 1.6 Totalsamplesizeforbothgroups(N=2n,=2n2): 16 22 32 66 122 366 APPENDIX B 3M Company EPI-0006 Page 24 of 29 I 3M Company EPI-0006 Page 25 of 29 DECATUR EMPLOYEE QUESTIONNAIRE Thank youforparticipatintghisresearchstudy.Pleaserespondtoeachquestionwitheithear shortansweror an 'x'intheappropriatbeox. NAME EMPLOYEE NUMBER 1. Have you everworked intheChemicalPlant? Ifno,pleasego toquestion2 Yes 0 If'yes' a. How many yearshave youworked inthechemicalplant?Years= b. What yeardidyou starwtorkinginthechemicalplant? Year No 0 2. Pleaseindicatfefyou have ever worked inthefollowinagreas.Mark an 'x'inallboxes thatapplyto you. 0 a. Buildin1g 0 b. Building2sand/or49 0 c.Buildin3g(OSCUOSF area) 0 d. Building3s(besidesOSCUOSF area) 0 e.Buildin4g North 0 f. Buildin4gmillroom/extruder 0 g. Buildin4g2 (PackagingFC inerts)oh. FilmPlant 0 i.Wastewater treatmentplant (Building3s6 and 57) 0 j. Other(Pleassepecify) 3. Thinkingaboutthejobthatyou workedforthelonciespteriodoftimewhileemployed at3M Decatur,please answerthefollowinqguestions. a. Job title: b. When didyouworkthereF:rom-(year) tQ_(year) c. Averagenumber ofhoursperweek on thisjob? Hours d. When you worked overtime,what was yourusualjobassignment? 3M Company EPI-0006 Page 26 of 29 4, Pleaseanswer thefollowinqguestionsregardingyourcurrenjtob. Currentplant: Chemical Film Currentjobtitle: What yeardidyou startworkinginthiscurrenjtob:Year Averagenumber ofhoursperweek on thisjob:Hours= When you work overtime,what isyourusualjobassignment? 5. Pleaseindicatienwhich area(s)you work inyourcuirentjob.Mark an @('inallboxes thatapplytoyou. [I a. Buildin1g 0 b. Building2sand/or49 El c. Buildin3g(OSCUOSF area) E:ld. Building3s(besidesOSCUOSF area) e. Buildin4g North 0 f. Building4smillroom/extruder g. Buildin4g2 (PackagingFC inerts)[h]. FilmPlant 01. Wastewater treatmentplant (Building3s6 and 57) j. Other(Pleasespec' 6. Whileatwork,do you chew gum? [:]always [:]frequently[-]sometimes rarely 0 never 7. Whileatwork,do you chew tobacco? always E] frequentlyEl sometimes rarely never 8. Whileatwork,do you smoke cigarettes? El always El frequently[:]sometimes rarely never 9. How frequentldyo you wash you hands beforeeatingwhileintheworkplace?Mark onlyone box. C] always 0 frequentl[y]sometimes [3 rarely never 10.What isyourheight? 11. What isyourweight? Feet= -inches Pounds= APPENDIX C 3M Company EPI-0006 Page 27 of 29 3M Company EPI-0006 Page 28 of 29 CONSENT FORM FOR COLLECTION OF BLOOD FOR SERUM FLUOROCHEMICAL LEVEL DETERMINATION INTRODUCTION You arebeing invitedtoparticipaitnea researchstudy.Your participatioinvolves donatingone testtube ofblood.Your blood willonly betestedfortheamount and kind offluorochemicalsitcontains.No othertestswillbe performedon your blood. Please review thisconsentform carefullaynd be sureyourquestionsareanswered beforeyou make a decisiontoparticipateT.he plantnurse,Cathy Simpson,RN, willbe availablteo answer any questionsthatyou may have about thisstudy. PURPOSE OF STUDY The purposeof thestudy istodeterminehow much and what typesof fluorochemicalsare foundinworkersattheDecaturplant.Thisinformatiownillassis3tM tobetter understandfluorochemicalsand how to reducesuch exposuresinthe workplace. STUDY PROCEDURES Your bloodwillbe drawn witha needlestickand requireone tube(10ml) ofblood.The bloodwillbe analyzedforsevenfluorochemicalsp:erfluorooctanseulfonate, perfluorooctanseulfonateamide,theglycinederivativoef perfluorooctanseulfonate, perfluorohexanesulfonatep,erfluorooctanoaicid,N-ethylperfluorooctanesulfonamido ethanoland N-methyl perfluorooctanesulfonamiedtohanol.A sub-sampleof employees willalsobe testedfortotalserum organicfluorineN.o othertestswillbe performedon yourblood. A lettewrillbe senttoyou informingyou ofyourbloodlevelsupon completionof theanalyses.A briefquestionnairweillalsobe administeredtoeach employee inquiringaboutcurrentand pastwork historyaswellas possibleroutesof oral ingestion. POTENTLKL RISKS/BENEFITS The onlydiscomfortyou may feelisfrom theneedlestick.You may alsohave some temporaryredness/bruising/swleilnginthisareaafterblood collection. BENEFITS ' There willbe no directbenefitfrom yourparticipatiionthisstudy.However, the informationgainedfrom thisstudywillfurthihrelpus understandhuman exposuresto fluorochemicalsY.our individuarlesultwsillbe communicated toyou only.The study's overallfindingsw,ithoutany specifiecmployeeidentifyinignformationw,illbe communicated to allDecaturemployees. 3M Company EPI-0006 Page 29 of 29 COMPENSATION Ifyou sufferinjuryor a medicalconditionthatappearstobe theresultof participatiinng thisstudy,you willbe referredtoanotherhealthcareprofessionaaltno costto you. In theeventof a researchrelatedinjuryc,ompensationwillbe determinedon a caseby case basisby 3M. The contactformedicalcompensationisJeffreyMandel, M.D., 612-7338670,of the3M Medical Department. CONFIDENTIALITY The overallresultscollectedinthisstudymay be usedinpublicationosr public presentationY.our name willnot be revealedinany publicatioonr otherdocuments intendedforpublicatioenxamination.Your individuarlesultwsillbe communicated to you only.Your individuarlesultwsillbe consideredconfidentiianlformationand willnot be disclosedtoanyone outsidethe3M MedicalDepartmentwithoutyour writtenconsent. SUBJECT RIGHTS/AVAU-ABMrrY OF E*4FORMAnON Ifyou have any questionsaboutthestudynow, or latero,r intheevent of a research relatedinjuryor emergency,contactCathy Simpson, RN (552-6341)or Dr. Jeffrey Mandel (612-733-8670).For answerstoquestionsaboutyourrightsinregardtothis researchy,ou may contactDr. LarryZobel,Chair,3M InstitutionRaelview Board at612733-5181. VOLUNTARY PARNCIPATION AND WM-EDRAWAL ParticipatiionthisstudyisvoluntaryR.efusaltoparticipawtiellinvolveno penaltyor lossof benefittsowhich you areotherwiseentitlddY.ou arefreetowithdraw from the studyatany time forany reason. SUBJECT CONSENT By signingtheconsentform,IcertiftyhatIam atleast18 yearsold.IconfirmthatIhave readthisconsentform, and thatI have been given adequateopportunitytoask any questionsImay have aboutthisconsentform or thestudy.IalsoconfirmthatI understandthescope of my participatiionthisstudy,and thatallof my questionshave been.answeredto my satisfactioIn.am signingthisconsentform voluntarilya,nd Idesire toparticipatien thestudy.IunderstandthatIam notwaivingor releasingany of my legal rightsby signingthisconsentform,orby participatiointhisstudy.IunderstandthatI willreceivea copy of thissignedconsentform. Signature Date PrintedName FINAL REPORT Epidemiology,220-3W-05 Medical Department 3M Company St.Paul,MN 55144 Title:FluorochemicalExposure Assessment ofDecatur Chemical and Film Plant Employees FinalReportDate: August 11,1999 Study StartDate: September 3, 1998 ProtocolNumber: EPI-0006 IRB Approval(#98082) Exempt: Expedited:X IRB Approval Date: September 3,1998 Study Archive Number: ARCH-021 PrincipalInvestigator: Co-investigators: Geary W. Olsen,DVM, PhD' PerryW. Logan, MS2 Cathy A. Simpson, RN 2 KristenJ.Hansen, PhD 3 Jean M. BurTis,RN, MPH' Michele M. Burlew, MS 1 John C. Schumpert, NM, NTH' JeffreyH. Mandel, MD, NTH' Study Director: JeffreyH. Mandel, NM, UTH 1. 3M Medical Department, 220-3W-05, St.Paul,N4N 55144-1000 2. 3M Decatur,EHS&R, P.O. Box 2206,Decatur,Alabama 35609-2206 3. 3M EnvironmentalLaboratory,2-3E-01,St.Paul,NIN 55144-1000 3M EPI-0006 Page 2 of 85 QUALITY ASSURANCE STATEMENT TITLE OF STUDY: FluorochemicalExposure AssessmentofDecaturChemical and Film PlantEmployees The above studywas examined forqualityassurancein keepingwith thespiriotfThe GuidelinesforGood EpidemiologyPracticesforOccupationaland Environmental EpidemiologicResearch aspublishedby theChemical ManufacturersAssociation EpidemiologyTask Group. The finalreportwas determinedto be an accuratereflection ofthedataobtained.The datesof QualityAssuranceactivitioens thisstudyarelisted below. Study InitiatioDnate:09/03/98 Study CompletionDate: 08/11/99 TYPE OF ALTDIT: ProtocolD,raftProtocol Addenda, Data File,Draft FinalReport FinalReport DATE OF AUDIT 06/28/99 08/09/99 DATE FINDINGS REPORTED TO PRINCIPAL INVESTIGATOR ANDSTUDY DIRECTOR 06/28/99 DATE FINDINGS REPORTED TO 3M MANAGEMENT 06/28/99 08/09/99 08/09/99 Archiving:All raw dataand thefinalreportwillbe filedintheOccupationalMedicine epidemiologyarchivesystem. Signatures(anddate)ofQA AuditTeam S/q lqq ABSTRACT 3M EPI-0006 Page 3 of 85 In the past,employees atthe 3M Decatur chemical planthave voluntarily participateidn a fluorochemicalmedical sur-veillanpcreogram. Analysis of the surveillancedata has not shown significantassociationsbetween theemployees' clinical chemistry and hematology testsand eithertotalserum organic fluorineor serum PFOS (perfluorooctanesulfonate)levels.However, thevoluntarynatureof the historical medical surveillanceprogram did not provide fora complete understanding of the distributioonf fluorochemicalserum levelsin theDecatur workforce. Therefore,the purpose of thisstudy was to collectdataby randomly sampling employees in theDecatur chemical plantin order to determine the distributioonf employee serum fluorochemical levelsaccordingto demographics, currentand longestheldjobs, years worked and buildinglocations.In addition,a random sample of theneighboring3M Decatur film plantemployee population,located atthe same site,was testedto determine fluorochemicalserum levelsin order to characterizethedifferencesbetween thetwo plantpopulations. A totalof 232 employees was randomly selectedforserum sampling: 186 (80%) participateidn the blood collectionwhich occurted in theFall,1998. An additional77 employees requestedblood testingforthe determinationof fluorochemicallevels.Of the random sample of employees who participated1,26 were from the chemical plantand 60 from the film plant. There were 61 volunteersfrom chemical and 16 volunteersfrom film;thus,allchemical participantnsumbered 'L87employees and allfilm participants numbered 76 employees. At thetime of blood collectione,mployees responded to a two- page questionnairethatinquired about theircurrentand longestheld jobs,the buildings 3M EPI-0006 Page 4 of 85 theyhad worked in(ifchemicalemployees),and possibleroutesoforalingestioonf fluorochemicaltshroughcigarettsemoking,chewinc,gum, chewing tobaccoand hand washing practices. Serasamples were extracteudsingan ion-pairinegxtractiopnrocedure.The extractwsere quantitativelaynalyzedforPFOS (perfluorooctanseulfonate)P,FHS (perfluorohexanseulfonate)P,OAA (perfluorooctanoaiccid),PFOSAA (N-ethyl perfluorooctanesulfonamidaocetateP)FOSA (perfluorooctanseulfonataemide),M570 (N-methylperfluorooctanesulfonamiadcoetatea)nd M556 (perfluorooctanesulfonamido acetateu)singhigh-pressurleiquidchromatography/electrosprtaayndem mass spectrometry(HPLC/ESMSMS) and evaluatedversusan extractecdurve.PFOS, PFHS, POAA, PFOSAA and PFOSA levelswere determinedby NorthwestBioanalytical Laboratory.M570 and M556 levelswere determinedby the3M Environmental Laboratory. The overallarithmetimceans (andrange)and thegeometricmeans and (95% confidenceintervalo)ftherandom sample ofchemicalemployees(n= 126)fortheseven fluorochen-iicaarlespresentedbelow (inppm): Chemical Plant ArithmetiMcean (andRange) GeometricMean (and95% CI) PFOS PFHS POAA PFOSAA M570 PFOSA M556 1.505 0.345 1.536 0.023 0.151 0.062 0.052 (0.091-10.600) (0.005-1.880) (0.021- 6.760) (0.001- 0.269) (0.008- 0.992) (0.0005- 0.612) (0.001- 0.406) PFOS PFHS POAA PFOSAA M570 PFOSA M556 0.941 0.180 0.899 0.008 0.081 0.013 0.022 (0.787- 1.126) (0.145-0.223) (0.722- 1.120) (0.006-0.011) (0.067- 0.098) (0.009-0.018) (0.018- 0.029) 3M EPI-0006 Page 5 of 85 The overallarithmetimceans (andrancrea)nd geometricmeans (95% confidence intervalo)f therandom sampleoffilmplantemployees (n= 60)fortheseven fluorochemicalasrepresentedbelow: Film Plant ArithmetiMcean (andRange) GeometricMean (and95% CI) PFOS 0.172 (0.015-0.946) PFOS 0.136 (0.114-0.161) PFHS 0.023 (0.001-0.210) PFHS 0.014 (0.011-0.018) POAA 0.071 (0.006-0.298) POAA 0.049 (0.039-0.062) PFOSAA 0.004 (0.001-0.038) PFOSAA 0.003 (0.002-0.003) M570 0.020 (0.001-0.454) M570 0.008 (0.006-0.011) PFOSA 85% ofsamples< LLOQ* PFOSA 85% ofsamples< LLOQ* M556 0.008 (0.0001-0.307) M556 0.003 (0.002-0.004) LLOQ = lowerlimitofquantitatiofnorPFOSA rangedfrom 0.001-0.010 ppm. The above valuesshowed highvariabilitayccordingto theemployees' demographics,work historyand buildinglocationsA.mong therandom sample (n 126)ofchemicalemployees,celloperatorhsad thehighestserum levelsof PFOS (geometricmean = 1.970ppm) and PFHS (geometricmean = 0.697ppm). However, sera from chemicaloperatorsand maintenanceworkershad thehighestlevelsof other fluorochemicaalnalytes(PFOSAA, M570, PFOSA and M556) acharacteristliickelydue totheirwork inBuildings3 and 4N withfluorochemicalalcoholsa,mides and acrylates. For example,chemicaloperatorshad a geometricmean levelof 0.131 ppm forM570 compared to0.033ppm forcelloperators0,.042formilloperatorsand 0.079ppm for wasteoperators.POAA levelswere abovethegeometricmean of 1.000ppm for employeeswith currentjobsofcelloperators(1.428ppm), chemicaloperators(1.887 ppm),maintenanceworkers(1.095ppm) ,milloperators(1.266ppm) and waste operators(1.542ppm). Employees withthejob categorieosf engineer/laabnd secretary 3M EPI-0006 Page 6 of 85 had the lowest serum fluorochemicallevels. PFHS, and to a lesserextentPFOS, were positivelyassociatedwith yearsworked in the chemical plant.The remaining fluorochemicalanalyteswere not routinelyassociatedwith years worked in thechemical plantby job categories.We did not observe an associationbetween hand-to-mouth usage or hand cleanliness(frequencyof washing hands) and serum fluorochemicallevels. Like theirmale counterparts,female chemical operatorsappeared to have increasedPFHS levelswith yearsworked. However, unliketheirmale counterparts, therewas no apparent modest linearassociationbetween PFOS and years worked among female chemical operators.Whether thisisdue to differentwork practicese,xposure patternsor pharmacokinetics once absorbed,remains to be determined. The sample size itsel(fn = 10 female chemical operatorsin random sample),isan important,limiting factorin theinterpretatioonf these data. The data also indicatesignificantllyower serum fluorochemicallevelsamong employees who have only worked in the filmplant (i.e.d,efinedas those employees in the random sample who have worked only in the film plantwith no priorwork on the D-1 maker locatedin the film plantor previouswork historyin chemical.The D- I maker uses FX-1801,amethylFOSEarnide). There were significantlyower serum fluorochemical levelsamong theseemployees who have only worked in the,film plantwhen compared to those who arecurrent'Chemicalplantemployees. Comparing the geometric means for each fluorochemical from the random sample of chemical operatorsand those employees who only have worked in thefilm plant,we observed the followingratios(inppm): PFOS (1.481/0.110)P;FHS (0.428/0.015)P;OAA (1.887/0.052)P;FOSAA (0.011/0.002); M570 (0.229/0.022);and M556 (0.044/0.003)E.xcept forPFOSAA, theseratiossuggest a 10-foldor greaterdifferencebetween chemical operatorsand film plantemployees who 3M EPI-0006 Page7of85 work severalhundredyardsaway from Building3. Thisonlyfilmplantemployee group had a geometric mean value forPFOS thatisapproximately 3-4 times higherthan the pooled geometric mean (0.029 ppm) from 64 samples obtainedfrom 18 U.S. blood banks. Thus, we suspect thatoccupationalexposure to PFOS does occur within the film plantalthough at much lower levelsthan among employees working atthe chemical plant.Additionallyemployees who worked on theD-1 maker have serum PFOS levels approximately 3 times higherthan thoseemployees who have never worked on theD-1 maker nor have worked in the chemical plant(i.e.t,he only filmplantemployees). We did not observe an associationbetween hand-to-mouth usage or hand cleanliness(frequencyof washing hands) and serum fluorochemicallevels.Itispossible an associationmight have been masked because industriahlygiene had instituteadn aggressiveeducationalcampaign severalmonths priorto thecollectionof blood samples in thisstudy;thus currentpracticesmay not be indicativeof pastpractices.Because the half-lifoef PFOS isestimatedto be 1000 days or more, such an associationmay not be discoverablewith thisstudy design. A limitationto thisstudy design which must be considered in theinterpretatioonf the datawas our inabilittyo more accuratelyquantifyan employee's work history experience.Decatur work historyrecordsprovide department numbers andjob titlebsut they do not provide informationregardingwhere someone worked (e.g.w,hat building(s) or with what specificfluorochemicals)S.elf-reportedwork historyinformationobtained by questionnairewas highlycorrelatedwith Decatur work historyrecordinformation; nevertheless,the specificitoyf where someone worked and with what chemicals was not known. Because many operationsarein batch mode, the likelihoodof determining specificitoyf historicawlorkload fluorochemicalexposure among chemical operators was notpossible. 3M EPI-0006 Page 8 of 85 The presentstudy'sserafluorochemicalevelso,bservedby job categoriesand buildinglocationss,tronglysupporttherecommendationsbome from recentlyconducted industriahlygieneassessments.These recommendations includespecifiecngineering controlstoreduceinhalatioenxposure,appropriatpeersonalprotectivequipment to preventoverexposureand appropriatepersonalhygienepracticeasmong employees to remove skinconcentrations. FinallyP,FOS and POAA serum levelsmeasured inthisstudyaresimilartothose thathave been previouslyreportedviapastbiennialmedicalsurveillancaectivities. Resultsof previousepidemiologicstudieshave notassociatedtheserum PFOS or POAA levelsobservedinthisstudypopulationwith hepaticl,ipidor hormone abnormalities. INTRODUCTION 3M EPI-0006 Page 9 of 85 In the past,employees atthe 3M Decatur chemical planthave voluntarily participateidn a fluorochemical medical surveillanceprogram. The surveillanceprogram analyzed fortotalserum organic fluorinelevelsuntilthe mid-1990's when serum perfluorooctanesulfonate(PFOS) and perfluorooctanoicacid(POAA) determination, quantifiableby high performance liquidchromatography mass spectrometry,became incorporatedin thebiennialmedical surveillancexaminations. Analysisof the surveillancedata has not shown significanatssociationsbetween the employees' clinical chemistry and hematology testsand eithertotalserum organicfluorinelevels[Roach, 1982; Schuman, 1982] or serum PFOS levels[Olsen et al.,1999]. However, the voluntarynatureof the medical surveillanceprogram may not lend itselfto an appropriatecharacterizatioonf the distributioonf fluorochemicalserum levelsas itisnot based on random sampling methods. Therefore,thepurpose of thisstudy was to collect data from thenecessary distributiobny randomly sampling employees in theDecatur chemical plantin order to determine thedistributioonf employee serum fluorochemical levelsaccording to demographics, currentand longestheldjobs, yearsworked and buildinglocations.In addition,a random sample of theneighboring3M Decatur film plantemployee population,locatedatthe same site,was testedto detem-iine fluorochemicalserum levelsin order to characterizethedifferencesbetween thetwo plantpopulations. The filmplantemployees have served as a comparison populationin a prior healthstudy(Mandel and Johnson, 1995) due to their(assumed) nonoccupational exposure to fluorochen-@cals.However, theiractualserum fluorochemical levelshad not been discerned.Epidemiologic studiesattheDecatur plantcan be more fullyappreciated 3M EPI-0006 Page 10 of 85 ifthed@istributioonfsemployee serum fluorochemicalevelsatboth thechemicaland filmplantsarebetterunderstood. MEETHODS Descriptioonf DecaturFacility The 3M DecatursiteislocatedinDecatur,Alabama which startedproductionin theearly1960's.The siteconsistosftwo plantsS,pecialtFyilm "filmplant"and SpecialtMyaterials"chemicalplant".Both plantsareintheSpecialtyMaterials ManufacturingDivision(SNIMD). The chemicalplantislocatedseveralhundred yards directleyastof theFilm Plant.The main buildingslocatedon thesiteareBuildings1,2, 3,5, 14,15,17,19,31,36,38,40,42,48,49,51,57,59 and 61 (seeAppendix A). Buildings14,15 and 19 areconsideredfilmplantbuildings.Buildings1,2,3,31,38,40, 42,48,49,51 and 61 areconsideredchemicalplantbuildings.Building5 istheboiler house thatcontrolssiteutilitiseusch as chilledwater,plantsteam,plantnitrogenand breathingair.Building5 islocatedsouthwestofthechemicalplant.Building17 serves as themaintenanceand stockroombuildinglocatedjustwest ofBuilding5 servicing. mainlythechemicalplant.Buildings36 and 57 aresitewastewatertreatmentbuildings locatedeastof thechemicalplant. The majorproductionbuildingsinDecaturfilmplantareBuildings14,15 and 19. Polyesterand non-polyesterfilmsareproduced inBuilding14. Maintenance,locker rooms,and diningfacilitiaersealllocatedin areasof Building14. Resin used infilm productionismanufacturedinBuildings15 and 19. The onlyprocessin thefilmplant usingfluorochemicalissrun on theD-1 filmline(calletdheD-1 maker).The process usesFX-1801 in the production of film used for a limited number 3M EPI-0006 Page II of 85 of products. Currently, no other processesin thefilm plantuse fluorochemicalsinproduction. The threemajor products produced in thechemical plantareprotectivechemicals, performance chemicals,and fluoroelastomers.The threeproduct groups are referredto as focus factories.Fluorochemicals identifiedin thisstudy are used in allfocus factory groups to some extent. Production forallfocus factoriestakesplace in Buildings 2, 3, 4, 3 8,40, 42, 49, 51 and 6 1. The chemical plant'smain officeareas,warehouse and quality controllabsare locatedin Building 1. The chemical plant'sdining facilitaynd locker rooms are locatedin Building 3 1. Raw materialsand intermediatesforeach product group may flow through many differentproductionbuildingsbeforethey arepackaged forshipping.The flow of protectivechemicals follow a path startingatBuilding 3 to Buildings2 or 49 to Buildings 3, 4, 38 or 5 1. The protectivechemicals group istheprimary producer of perfluorooctanesulfonylfluoride(POSF) and perfluorohexanesulfonylfluoride(PHSF) based chemistry. Octyl mercaptan or hexyl mercaptan isreactedwith chlorineand ammonium fluorideto produce octanesulfonylfluoride(OSF) or hexane sulfonyl fluoride(HSF) inBuilding 3 and isreferredtoas the 'celfleed'.The cellfeedissentto Buildings 2 and 49 where itisreactedinelectrochemicalcellsystems to produce POSF or PHSF. POSF isthe major sulfonatebased fluorochemicalproduced atDecatur. PHSF isproduced mainly for firesuppressionliquids.Most of thePOSF produced ispiped to Building 3 where amides, alcohols,acrylatesand otherfluorochen-iicaplolymers are produced. These fluorochemicalpolymers arethen used in allproduction buildingsto produce intermediatesand finishedgoods. 3M EPI-0006 Page 12 of 85 The performance chemicals aremostly made up of inertliquidsand fire suppressionliquids.The inertliquidsfollow a path startinagt Buildings2 or 49 to Buildings 40 or 42. Inertliquidsconsistof mostly perfluoronatedalkanes and do not contain sulfonateor carboxylicacid compounds. Firesuppressionliquidsare primarily based on sulfonatechemistriesstartingwith POSF and PHSF. Firesuppression products are made in Building 3 and packaged in Building 4. Fluorochemicals are used in theproduction of fluoroelastomerproducts. The first partof theflu6roelastomersiscalledlatex,which isproduced in Buildings4, 38 and 5 1. The latexisthen coagulated,washed and milled in Buildings4 and 61. POSFbased compounds aretheprimary fluorochemicalsof interestused in the majorityof fluoroelastomerproducts. POAA isalsoused in a limitednumber of fluoroelastomer product runs. POAA isused in the productionof latexthatiseventuallycoagulated, washed, and milled in Buildings4 and 61. This POAA containingproduct isrun infrequently,only severaltimes per year. POAA isalsoa by-product within the electrolyticcellsand is carriedthrough up to product. Itisbelievedto be a resulto f increasedoxidationwithin thecells.POAA was produced in Building2 and subsequently worked up in Building 3 more than 20 years ago and had not been produced in Decatur sincethe time of thisstudy. POAA production isexpected to resume in Buildings 2 and 49 in thenear future. Sam2le SizeDetermination 3M EPI-0006 Page 13 of 85 Three criticaflactorswere consideredtodecidethesample sizeforthisstudy. Firsti,twas importantthata samplebe randomly chosenfrom theemployee populations of boththechemicaland filmplants.Second,thesample sizewas drivenby theneed to provideconfidencethattheexposureinthefilmplantissmallrelativteothatof the chemicalplant.Third,thesample sizehad to adequatelycharacterizteheexposurelevels withinthechemicalplantworkplace.Inaddition,allemployeesinthechemicaland film planthad to be offeredtheopportunittyoknow theirfluorochemicalevelsviablood testinga,lthoughtheymay notbe partoftherandom sample.The random sample sizein thisstudyof more than200 subjectswas basedon: 1) thelower95% confidencebound ofthehypothesizedmean differencebetween theserum fluorochemicalevelsof the chemicalplant;and 2) toallowforadequatecharacterizatioofnserum fluorochemical differencebsyjob and buildingwithinthechemicalplant(seestudyprotocolfordetails). There was an added degreeof uncertaintyinestimatingsample sizebecause approximately10 percentof thefilmplantemployees may have had priorwork experienceinthechemicalplant.Also,an unknown number of filmplantworkershad worked on theD-1 maker where a PFOS-based fluorochemica(lFX 1801,a methylFOSE amide) has been used. The random sample was chosenby thefollowingmethods:a)allfull-timceurrent chemicaland filmplantemployees were identifievdiaacurrentplantrosterthatlisted departmentsand supervisorsb;) usinga random number generatoralgorithm,a sample of employees was chosen which was proportionatteo thenumber of employees who worked in thevariouschemicaldepartments,autoand chemicalmarketsgroup,DecaturEHS&R, Dyneon, and thefilmplant.We includedintherandom sampleallidentifieDdecatur 3M EPI-0006 Page 14 of 85 siteemployees who were assignedto thewastewatertreatmentplant(Buildings36 and 57). Altogethert,herewere 232 employeesrandomly chosento participatien thestudy (Table1).A totaolf 186 (80%) participataendd46 (20%) refused.The filmplant random sample had the lowestparticipatiroante(71%). Inadditiontothe 186 random sample participanttsh,erewere 77 employeesfrom thechemical(n= 61) and film(n 16) plantswho requestedtheirserum be testedforfluorochemicalevels.Hereafter, theseindividualwsillbe calledthe"volunteers." EmRloyee Study ParticiRation Studyparticipatiroenquiredthefollowing:1)a signedconsentform by the employee;2)a writtenresponsetoa briefquestionnair(eAppendixB) thatinquiredabout currentand pastwork historyalongwiththefrequencyof hand washing and use of gum, chew (tobacco)and cigarettheabitsoftheemployee whileatwork; and 3) a venipuncturewiththecollectioonftwo vialsofblood(approximatel2y0 cc)forthe determinatioonf theseven fluorochemicalsT.he studyprotocolwas approvedby the3M InstitutioRneavliew Board (IRB). Each randomly chosen employee (filmand chemical)receiveda letteorf invitatiotnoparticipatehatwas jointlsyignedby theplantmanager (Mr.Jim King) and the3M MedicalDepartmentdirecto(rDr.LarryZobel).There was,plant-wide communication which describedthepurposeofthisstudyand encouragedemployee participationA.llstudyparticipantwsh,o were eitherandomly chosenor who volunteeredw,ere informed of theirown individuarlesultbsy a lettesrenttothem from the3M MedicalDepartment inJuly,1999. Aggregateresultosf thestudywere also communicated atthattime totheemployees. 3M EPI-0006 Page 15 of 85 FluorochemicalAnalyses All blood was collectedinthemonths of October and November, 1998 atthe Decaturplantby MedAccess (anoccupationahlealthcliniclocatedinDecatur,Alabama) underthedirectioonf Cathy Simpson, RN who centrifugetdheblood toobtaintheserum and thenshippedthesamplestothe3M MedicalDepartment(St.Paul,NIN). Split sampleswere cataloguedby Diane Madsen and Jean Burrisand thensentto either Northwest Bioanalytica(lDr.David Vollmer)fordeterminationof perfluoro(>ctane sulfonat(ePFOS),perfluorooctanseulfonataemide (PFOSA), perfluorohexanseulfonate (PFHS) ,perfluorooctanoiacid(POAA) and N-ethylperfluorooctanesulfonamido acetate(PFOSAA) or to3M EnvironmentalLaboratory(Dr.KfisHansen) for determinatioonf N-methyl perfluorooctanesulfonamiadcoetate(M570) and perfluorooctanesulfonamiadcoetate(M556). Inboth laboratoriese,rasampleswere extractedusingan ion-pairinegxtraction procedure.The extractswere quantitativealnyalyzedforPFOS, PFHS, POAA, PFOSAA, PFOSA, M570 and M556 usinghigh-pressurleiquid chromatography/electrosprtaayndem mass spectrometry(HPLC/ESMSMS) and evaluatedversusan extractedcurve.There were minor differencebsetween the analyticamlethods used atNorthwestBioanalyticaalnd 3M EnvironmentalLaboratory. Most notably,NorthwestBioanalyticaelvaluatedanalytelevelsversusa curveextracted from human sera.Endogenous levelsofcertainfluorochemicalwsere determinedinthe standardmatrixand additionaflluroochemicawlas spikedintothematrix.The total amount of eachspecificfluorochemica(lendogenous+ spiked)was used toconstructan extractedstandardcurve. For theanalysisconductedatthe3M Environmental 3M EPI-0006 Page 16 of 85 Laboratory,the difficultiepsresentedby the endogenous levelsof fluorochemical in samples of "blank" testmatrix were circumvented by utilizinrgabbitseraas a surrogate matrix. Previous research had shown thatrabbitseracontainsthe lowest levelof endogenous fluorochemicalswhen compared to serafrom bovine,rat,monkey and human. As a qualitycontrolcheck,the 3M Environmental Laboratory screened PFOS levelsin approximately 10% of the seraanalyzed atNorthwest BioanalyticalW.hile most of the resultsagreed to within 25%, 14 of the40 samples checked showed lower (> 25%) values when analyzed at3M. Itisexpected thatthesediscrepanciesare due to differencesin curve slope and interceptsarisingfrom theanalyticaldifferencesdescribed above. Given thatNorthwest Bioanalyticalsatisfactoricloympleted a method validation forPFOS using human seraand given thatmost valueswere in close agreement with those obtained by the 3M Environmental Laboratory using a rabbitseracurve,data from both laboratorieswere considered accurateto withintheparameters defined by their methods. Detailsof both laboratoriesm'ethods and finalreportsarereported elsewhere [Vollmer, 1999; Hansen, 1999]. Data Analysis Each employee's questionnairedata and computerized work historyrecords were reviewed to detem-iinewhether the employee was: a) a currentchen-@calemployee (regardlesosf any work experiencein thefilmplant);b) a filmplantemployee with no historyin chemical;or c) a filmplantemployee with priorhistoryin chemical. Employees who were considered Decatur 'sitee'mployees (e.g.,safety,industrial 3M EPI-0006 Page 17 of 85 hycyienea)nd who statedtheycurrentlwyorked inone ormore chemicalbuildingswere consideredtobe chemicalemployees inthedataanalyses. Employees were askedtoprovidetheircurrentand Ion-est-heljdob. A reviewof thesejob titlebsy an industriahlygienist(PWL), epidemiologist(sGWO, JMB) and occupationahlealthnurse(CAS) categorizetdheentriesintoeightjob classificatiofnosr thechemicalplant:celloperatorsc,hemicaloperatorse,ngineers/laboratormya,intenance, milloperatorss,ecretariesu,pervisors/managementand waste operators.Film plant currentjobs (andlongestheldjobs)were categorizeidntofourjob classifications: engineers/laboratorfyi,lmprocessorsm,aintenanceand administrativeT.hese classificatiownesre done priortoany dataanalyses.The individualu'sualjob assignmentwhen he/sheworked overtimewas not analyzedas most personsreportedthis was thesame astheircurrent(or longestheld)job. Employees were askedon thestudy questionnairteoindicatethenumber of yearstheyhave worked inchemical.This informationcorrelatewditha reviewof recordsfrom theepidemiologyunit'Dsecatur work historydatabaseforthoseemployees with7000 leveldepartmentcodes;thusthese self-reporteddatawere used toassessyearsworked inchemical.On theotherhand, yearsworked infilmwere calculatefdrom theepidemiologyunit'Dsecaturwork history databasebecausethisinformationwas not requestedon thestudyquestionnaire. Chemical employees who had worked previouslyinthefilmplantwere identifieadnd classifieadstotheirtime ofserviceinthefilmplant(< 1980,1980-1989and 1990-1998). Age was calculatefdrom theemployee'sdateofbirthfrom theepidemiology unit'Dsecaturwork historydatabase.Body mass index(kg/m2)was calculatedbased on theinformationprovidedby theemployee on thequestionnaireA.n indexof hand-tomouth contactwas calculatedbased on whetherthepersonsmoked cigarettecsh,ewed 3M EPI-0006 Page 18 of 85 tobacco or chewed gum. An index of hand washing was based on whether or not the employee saidthey always washed theirhands beforeeatingwhile at work. Through the use of SAS and JMP and employing standardstatisticatlechniques (student'tstest,chi square, ANOVA, singleand multivariableregressionusing linear and nonlinearanalyses),data analysesconcentratedon thefollowing issues:1) compare responders and nonresponders in the random sample by theirdemographic characteristics (e.g.a,ge, gender, yearsworked); 2) compare mean serum nuorochemical levelswithin thechemical plantby a) employee demographics, b) self-reportewdork historydata based from the study questionnaireincludingcurrentjob,longest-heldjob, years worked in chemical and in which chemical buildings;c) work historyinformationsupplemented with data from the 3M epidemiology unit'scomputerized comprehensive work history recorddatabasefor theDecatur site,and d) personalhabits(alsoidentifiedon the study questionnaire)thatwere hypothesized to increasethe likelihoodof oralingestionof fluorochemicals(e.g.h,and washing, cigarettesmoking, chewing tobacco and chewing gum); and 3) likewise,compare mean serum fluorochemicallevelswithin the film plant by similarfactors.To preventmisclassificatioonf potentialworkplace exposure experienceto fluorochemicalswithinthe filmplant,we analyzed samples from film plant employees according to those who have and have never worked in thechemical plantas well as thosewho were identifiedas having worked on theD- I maker locatedin the film plant.Film plantemployees who had never worked on theD- I maker nor ever worked in chemical arehereafterreferredto as "only film plantemployees." Because the serum distributionfsorPFOS, PFHS, POAA, PFOSAA, M570, PFOSA and M556 appeared log normally distributed(a skewed distribution)n,aturallog transformationsof the fluorochemicalswere performed to calculategeometric means 3M EPI-0006 Page 19 of 85 (e(suImnxYn)and statisticcaallculatiornesgardincgentratlendencywere primarilbyased on thegeometric mean. The random variableX issaidto have a log normal distribution iflog X isnormally distributedt,hatis,ifX isof theform ey where Y isnormal (i.e.t,he normal bellshaped curve). The pertinentpropertiesof a log normal distributiocnan then be deiivedfrom propertiesof the normal distributionT.he mean and variance are of the normally distributeYd, thatis,of log X. The log normal distributiofnindsapplicationsin a wide varietyof fieldsincludingexposure assessmentsin nature(whether of humans, mammals, etc). Provided in Appendices C and D arethe histograms of the seven fluorochen-iicals as measured foremployees in the chemical and filmplants,respectively,using statistics derivedfrom the normal distributioanlong with thenaturallog transfonnationof the distribution.The Shapiro-Wilk W testsuggeststhenecessityof the log transformation. Measures of centraltendency routinelypresentedthroughoutthisreportwillinclude the arithmeticmean and range, and the geometric mean and associated95% confidence interval.Comparisons of geometric means were conducted using thestudent'sttestwith statisticaslignificanceconsidered atp < .05. All fluorochemicalmeasurements were reportedin partsper million(ppm) to the thirddecimal point. For statisticpaulrposes,serum fluorochemicalvalues thatwere less than the lower limitof quantitation(LLOQ) were assumed the midpoint between zero and the LLOQ. Of the totalnumber (n = 186) of employees consideredto be currently working in chemical who participateidn thestudy (126 from the random sample and 60 volunteers),thefollowingnumbers (inparentheseswith percentage)had reported LLOQ's by the measured fluorochemicalP:FOS (1,0.5%);PFHS (1, 0.5%); POAA (0, 0%); PFOSAA (49,26%); M570 (0,0%); PFOSA (36,19%); and M556 (8,4%). Of the 3M EPI-0006 Page 20 of 85 totalnumber (n = 76) of employees consideredto be currentfilm plant(60 in therandom sample and 16 volunteers),the followingnumbers (inparentheses)had reportedLLOQ's by themeasured fluorochemical:PFOS(1,1%);PFHS(2,2%);POAA(0,0%); PFOSAA (29,38%); M570 (0,0%); PFOSA (65,86%) and M556 (32,42%). Wechose not toanalyzePFOSA among thefilmplantemployees because 85% of them had serum PFOSA measured at lessthan LLOQ which resultedin minimum variabilitfyor statistical considerations.The LLOQ forPFOSA ranged,between analyses,from 0.001 to 0.010 PPM- Analyses focused on the random sample but aggregatedata analyseswere also conducted forallparticipants(random sample and volunteers)stratifiebdy the two plants. RESULTS ComRarison of random saMRIe resronders and nonresponders Responders (n = 186) and nonresponders (n = 46) from the random sample were compared by age, gender and yearsworked and found tobe alike.Among thechemical random sample, the average age was 42 years compared to 43 for nonresponders. Responders and nonresponders have worked, on average, 16 years. There was a similar5 to I ratioof male to female employees forthe respondersand nonresponders among chemical employees. Film plantemployees who responded were, on average,46 yearsof age,had worked 19 years and the ratioof male to female was 5 to 1. Nonresponders were 48 years of age,had worked 25 years and had a 7 to 1 male to female ratio.Thus, nonresponders in the film plantrandom sample were slightlyolder,worked longer and a greaterpercentage were males. 3M EPI-0006 Page 21 of 85 OverallFindings The arithmetimcean (andrange)of therandom sample as wellas thegeometric mean and (95% confidenceintervalo)f chemicalemployees (n= 126)fortheseven fluorochemicalasrepresentedbelow (inppm): Chemical Plant ArithmetiMcean (andRange) GeometricMean (and95% CI) PFOS PFHS POAA PFOSAA M570 PFOSA M556 1.505 0.345 1.536 0.023 0.151 0.062 0.052 (0.091-10.600) (0.005-1.880) (0.021-6.760) (0.001-0.269) (0.008- 0.992) (0.0005-0.612) (0.001- 0.406) PFOS PFHS POAA PFOSAA M570 PFOSA M556 0.941 0.180 0.899 0.008 0.081 0.013 0.022 (0.787- 1.126) (0.145-0.223) (0.722- 1.122) (0.006-0.011) (0.067-0.098) (0.009- 0.018) (0.018-0.029) The arithmetimcean (andrange)oftherandom sample as wellas thegeometric mean and (95% confidenceintervalo)f thefilmplantemployees(n= 60)forthesix fluorochemicalasrepresentedbelow: Film Plant ArithmeticMean (andRange) - GeometricMean (and95% CD - PFOS PFHS POAA PFOSAA M570 PFOSA M556 LLOQ = 0.172 (0.015-0.946) PFOS 0.136 (0.114-0.161) 0.023 (0.001-0.210) PFHS 0.014 (0.011-0.018) 0.071 (0.006-0.298) POAA 0.049 (0.039-0.062) 0.004 (0.001-0.038) PFOSAA 0.003 (0.002-0.003) 0.020 (0.001-0.454) M570 0.008 (0.006-0.011) 85% of samples< LLOQ* PFOSA 85% ofsamples< LLOQ* 0.008 (0.0001-0.307) M556 0.003 (0.002-0.004) lowerlimitofquantitatiofnorPFOSA rangedfrom 0.001-0.010ppm. Because theabove valuesmay be highlyvariableby employees'demographics,work historyand personalhabits,subsequentanalyseswillfocuson each plantseparately. 3M EPI-0006 Page 22 of 85 Tables I - 21 provide the resultsfrom the chemical plant.Tables 22 - 29 provide the resultsfrom thefilm plant. Chemical Plant Provided in tables2 and 3 arethedemographic characteristicbsy thenumber of chemicalemployees (and percent)from therandom sample (n = 126),volunteers(n = 60) and allchemical participants(bothrandom sample and volunteer,n = 186). The distributioonf demographic characteristicbsetween therandom sample and volunteers were comparable although the random sample had a higher percentage of chemical operators(37%) than did the volunteers(28%). The mean, median, range and geometric mean of the random sample, volunteers and allchemical participants,isprovided in Table 4 forthe seven fluorochemicals. The range of PFOS was from 0.091 - 10.600 ppm. Although the geometric means were consistentlyhigherin the random sample than volunteers,only with PFOSA did the geometric mean differsignificantlbyetween therandom sample (0.013 ppm) and the volunteers(0.006ppm). Itshould alsobe noted thatamong the random sample, five employees had serum PFOS levels > 5 ppm compared to none among the volunteers. Because thedemographic characteristicasnd geometric means did not substantialldyiffer between therandom sample and volunteers,subsequent tableswillreporton eitherthe random sample and/or allchemical participantsT.he volunteerswillnot be presented separately. . Presentedin Table 5 are the demographic characteristicosf the random sample of chemical employees by currentjob category(celloperator,chemical operator, engineer/lab,maintenance, mill operator,secretary,supervisor/management and waste 3M EPI-0006 Page 23 of 85 operator).Supervisors/management(mgmt) and waste operatorswere theoldestwith milloperatorstheyoungest.Milloperatorshave worked considerablylessyears,on average,thanallotherjobcategoriesT.hisisto be expectedsincemilloperatorisan entrylevelpositionfornew employees.The number (andproportiono)ffemale employees were similarbetween thechemicaloperatorsand theengineer/lagbroup. Providedintable6 isthemean, median and geometricmean foreach of theseven fluorochemicalevelsby gender,hand-to-mouthcontactw,ash hands and whetherthe individuahlad worked only in thechemicalplant.Geometricmean levelsformales were significanthliygherthanfemalesforPFOS, PFHS, POAA and M570. We didnot observe,as hypothesizedt,hathand-to-mouthcontact(viause ofcigar-ettecsh,ewing tobaccoor chewing gum) and lessfrequenthand washingresultedinhigher fluorochemicaslerum levels.Also,havingworked onlyinchemicaldidnotresultin higherserum fluorochemicalevelsW.e didobservethatthefurtherback intimethat chemicalemployees worked inthefilmplant,thelargertheirgeometricmean values were,as measured inthisstudy. For example,thegeometricmean valuesforchemical employees who lastworked inthefilmplantpriorto 1980,between 1980-1989,19901998 and neverworked in thefilmplantwer-e1.656ppm, 1.551ppm, 0.786ppm and 0.700ppm, respectivelyO.f course,thisisalsoa reflectioonfthenumber of years worked inthechemicalplant(tobe presentedlaterinthissection)T.hat is,the employees who worked in thefilmplantpriorto 1980 had subsequentlythelongest continuouswork historyinchemicalsince1980. Fluorochemicallevelsby currentjob categoryarepresentedinTable 7. Several observationwsere noteworthy.Firstt,hedistributiofnhigh-to-lowgeometricmean valuesvariesby currentjob categories.Celloperatorshave thehighestgeometricmean levelof PFOS. 3M EPI-0006 Page 24 of 85 The next group are thechemical operators,maintenance and waste operators.Supervisor/mgmt isnext,followedby the group consistingof milloperators, engineer/laband secretary.For PFHS, celloperatorshave the highestgeometric mean level.The next highestgroup appears to be chemical operators,waste operators, supervisor/mgmt and maintenance. For POAA, chemical operatorsappear to have the highestlevelsfollowed by thegroup consistingof celloperators,maintenance, mill operatorsand waste operators.Chemical operatorsand maintenance have significantly higherlevelsof M570 than allothercurrentjob categories.Chemical operators, maintenance and mill operatorshave the highestgeometric mean valuesforPFOSAA. PFOSA and M556 values were significantlhyigher forchemical operatorsthan formost otherjob categories. Fluorochemical ratios(PFOS/PFHS, PFOS/POAA, PFOS/(PFOSSA+M570+PFOSA+M556), M570/M556, PFOSAA/M556 and PFOSA/M556) are presentedby currentjob categoryin Table 8. The celloperatorshad the lowest PFOS/PFHS ratioand the milloperatorshad thelowest PFOS/POAA ratio. The largestPFOS/metabolite ratiowas forthe celloperators. Tables 9-11 are identicalto Tables 7-9,respectivelye,xcept thattheemployees' longestjob isanalyzed insteadof thecurrentjob category. Cell operatorsare not included as therewas only one celloperatorwho statedthiswas his longestjob held. The highestPFOS, PFHS and POAA levelswere observed among chemical operators. Maintenance and chemical operatorshad higherM570 and PFOSAA levels.Overall, resultsdid not vary substantiallbyetween currentjob and longestheld job. Table 12 isrestrictetdo only thosechemical employees who statedon the questionnairethatthey currentlywork injustone location(building).Because building locationissynonymous 3M EPI-0006 Page 25 of 85 with job category forcelloperators,Buildings 2/49 had the highestPFOS and PFHS levels.Building3 and Building4N representedthe areaswith the highestPOAA levelsalthough only one building,Building 1, had substantiallylower POAA levelswhen compared to the otherlocations.M570, PFOSAA and M556 levels were highestin Building 3. Buildings3 and 4MX (Na = n-@xer/extrudearrea)appeared to have comparable levelsof PFOSA. Among the 5 employees who only worked in Building 4N, therewas a wide range of PFOSA levels. Because employees may currentlywork in only one buildingbut have had a past historyof working in severalbuildings,we furtherrestrictetdhe analysesto only those employees who saidthey have only worked in one buildingthroughout their employment. This restrictetdhenumber of subjectstojust21 individuals(17% of the random sample) with representatioinn theseBuildings:1,3 and 4mx. Table 13 shows thatPFOS levelswere more than 5 times higher in the seraof Building 3 workers than in the seraof Building I or Building 4MX workers. PFHS levelswere almost 10 fold higher. POAA levelswere twice as high in seraof Building3 workers compared to Building4NlX workers and more than 15 times higherthan Building I workers. M570 and M556 levelswere 5 times higherin Building 3 workers than Buildings I or 4Na. PFOSAA and PFOSA levelswere comparable between Building 3 and Building4Nlx workers and lowest in Building 1. Tables 14 through 21 provide similardata analyses as the previous tablesbut now represent the 187 total(random sample and volunteers)chemical participantsT.here were no substantialdifferencesbetween theanalysesof the random sample and of all chemical participants.For example, among allchemical participantsm,illoperatorswere theyoungest employees (Tables 14, 17);most female employees were eitherin the 3M EPI-0006 Page 26 of 85 currentand longestjob categoryof chemical operatorsor engineer/lab(excluding secretary()Tables15, 18);celloperatorshad thehighestPFOS and PFHS serum levels and engineer/lab,secretaryand milloperatorshad the lowestPFOS and PFHS serum levels(Tables16, 19); and chemical operatorsand maintenance workers had the highest levelsof M570 and tended to alsohave the highestserum levelsof PFOSAA, PFOSA and M556. Fluorochemicallevelsstratifiebdy where employees only currentlywork (Table20),or have only ever worked (Table21),were alsocomparable with theresults from the random sample.All chemicalparticipantwsho have only worked inBuildingI had lower fluorochemicallevelsthanBuilding3 workers forallseven fluorochemicals (Table21). Building I workers had lower PFOS, POAA, PFOSAA and PFOSA levels thanBuilding4NIX employees. PFHS, M556 and M570 levelswere similarinBuilding I workersand Building4MX workers. A seriesof multivariablaenalyses(datanotshown) examining each fluorocheniicably severalindependentvariables(e.g.a,ge,body mass index,gender, currentjob,longest-heljdob,whether employed only inthechemical plant,yearsworked inthechemical plant)suggestedtheremay be up tothreeimportantexplanatory variablesT.hese were current(orlongest)job,yearsworked withinthechemical plant and gender. To bettervisualizethe influenceof yearsworked withinchemical on serum fluorochemicallevels,we stratifietdhe analysesby currentjob categories.In other words, the dependent variable(i.e.e,ach specificfluorochemical)was regressedon years worked inchemical foreach separatejob category.These linearregressionanalyses employed the untransfonnedas wellas transformed(naturalog)dependent variable. Analyses were conducted forthe random sample (n= 126) as well asforallchemical participant(sn 3M EPI-0006 Page 27 of 85 187).Presented in Appendix E are theanalyses foreach fluorochemical fortherandom sample (n = 126) and then separatelyforchemical operators,engineer/lab, maintenance,mill operatorsand supervisors/mgmt. Cell operatorsand secretariesarenot presentedbecause of theirinsufficienptopulation. From the scatterplotasnd models presentedin Appendix E, the following were observed. (Note:in Appendices fluorochemicalsare presentedin thefollowing order PFOS, PFHS, POAA, PFOSAA, M570, PFOSA and M556. For the scatterplotsu,pper and lower 95% confidence curves areprovided of the fittedline.First,forthe entire random sample, only the PFHS model fitthedata well with 22 percent of the variationof PFHS explainedby an increasein years worked in chemical. PFOS levelsincreased modestly with years worked in chemical although the varianceexplainedremained small (2r=.10). Although interceptsmay have been significanftorotherfluorochemical models fortheentirerandom sample, the varianceexplainedwas consistentlyquitesmall (i.e.l,essthan 3 percent);thus such models have minimum prediction.Among chemical operatorsthe most significantobservationwas thefindingof a linearincreaseof PFHS levelswith increasingyears worked in chemical. Thirty-fourpercentof the variationin PFHS.was explained. There were weaker positivelinearassociationsbetween POAA or PFOS and yearsworked in chemical. On the otherhand, thereappeared to be a suggestionthatthe highestlevelsof thefluorochemicalanalytes(PFOSAA, M570, PFOSA and M556) were most oftenobserved among chemical operatorswith justone or two years of experience. Among theengineer/labgroup, therewas a weak association between serum PFOS levelsand yearsworked inchemical. The strongestassociation observed among maintenance workers was thelinearincreaseof PFHS levelswith years worked in chemical. Like the chemical operators,a significanatmount of variationwas explained (26 percent)although the data were sparse.Among 3M EPI-0006 Page 28 of 85 the supervisor/mgmt group, PFOS, PFHS and POAA increasedwith years worked inchemical.. Approximately 15 percentof the variationwas explainedin each model. Model fitwas poor forthe milloperatorsbecause allbut two had worked for5 years or less;thus only scatterplotasre presented (notregressionmodels). The naturallog transformationsarepresentedin Appendix F forallchemical employees (n = 126) in the random sample as well as forthe two currentjob categories with the most numbers (chemical operatorsand engineer/lab).For the entirerandom sample, a weak association(r2= .08)isobserved forPFOS and years worked in chemical and a strongerassociation(r2= .23)forPFHS. For chemical operatorsthe strongest association(r2= .34)is with PFHS and yearsworked in chemical. Although the latter associationwas not observed among theengineer/labcategory with thenontransformed variable(seeAppendix E),the naturallog transformationof PFHS was significantly associated(r2= .19)with years worked in chemical (seeAppendix F). Presented in Appendix G are similarscatterplotasnd regressionmodels forall chemical participantsby currentjob category.There remained a positiveassociation between PFHS or PFOS serum levelsand yearsworked in chemical,with the strongerof thesetwo associationsfor PFHS. Because of more subjects,scatterplotasre alsonow shown forcelloperators.These plotssuggest,again,an increasein PFOS, PFHS and now alsoPOAA levelsamong currentcelloperatorswith years worked in chemical. Among chemical operatorsthe strongestassociationremained with PFHS, with weaker linearassociationsobserved forPFOS and POAA with years worked in chemical. Among theengineer/labgroup, thereremained a positivelinearassociationbetween eitherPFHS or PFOS with yearsworked in chemical. There were positive linear associationsforPFOS, PFHS and POAA 3M EPI-0006 Page 29 of 85 with years worked in chemical among both the maintenance and supervisor/mgmt groups. Too few milloperatorswith 5 or more work years in chemical were sampled to conduct a meaningful analysis.The scatterplotdata do show a wide range of serum POAA levelsamong milloperatorswithjustone year of work experiencein chemical. The scatterplotisn Appendix H representthelog transformationsforallchemical participantsand the two most numerous job categories:chemical operatorsand engineeraab. Again, the scatterplotssuggest a consistentlystrongpositiveassociation between serum PFHS levelsand years worked in chemical and a lesserassociationwith PFOS and years worked in chemical. Presented earlierin Table 6 was the observationthatserum fluorochemicallevels were lower among female workers. Whether thiswas due to a smaflerproportionof female workers injob categorieswhere exposure would be the highest,younger female workers and/or female employees with lesswork experience in chemical remained to be resolved. To address thisissuewe focused on those two job categoriesthathad the most female subjectswithin the random sample as well as allchemical participantsc:hemical operatorsand theengineer/labgroup. Presentedin Tables 22 and 23, by gender, arethe demographic characteristicasnd serum fluorochemicallevelsforthe random sample of chemical operatorsand the engineer/labgroup. Female employees had significantly lower geometric mean serum levelsof PFOS, PFHS and POAA. Multivariableanalyses of chemical operators of each fluorochemicallevelregressedon gender, years worked in chemical and with and without age are presentedin Appendix Iforthe random sample. For purposes of brevity,only thetransformed (naturallog)dependent models are presented. Gender appeared to be thebestpredictorof PFOS level(i.e.l,ower levels among female chemical operators)with years worked 3M EPI-0006 Page 30 of 85 in chemical not significantly associatedwith PFOS. Gender was alsosignificantlayssociatedwith POAA levels (lower POAA levelsamong female workers) adjustingforyearsworked in chemical and age. Both gender and yearsworked in chemical appeared to be important predictorsof PFHS levelsamong chemical operators.Among therandom sample of engineer/lab workers, gender was the most importantpredictorof PFOS, PFHS, POAA and PFOSAA levelsafteradjustingforyearsworked inchemical and age (Appendix J). Data for chemical operatorsand the engineer/labgroup from the allchemical participantshowed comparable results(Appendices K and L). To furtherclarifythisissue,regressionanalyseswere stratifiebdy gender as well as by job category.With male chemical operatorsas well as with themale engineer/lab group, therewas a consistentassociationof increasinglevelsof PFOS and PFHS (and POAA forchemical operatorsonly)with increasingyears worked, at leastfor thefirst severalyearsof work. Scatterplotasrefound in Appendix M. More questionableis whether such an associationremains linearor ispolynomial (quadratic)over time. Among female chemical operatorstheonly associationobserved was forPFHS and years worked. Scatterplotsarefound in Appendix N. Neither PFOS or POAA levels appeared to increasewith years worked in chemical among female chemical operators. The dataforthe female engineer/labgroup aredifficultto interprestince6 of the 9 individualshad lessthan 5 yearsof work in chemical. Use of an interactiotnerm (gender x yearsworked in chemical) in multivariablemodels was not an importantpredictorof fluorochemicallevels. Film Plant 3M EPI-0006 Page 31 of 85 Altogethertherewere 60 currentemployees who responded to the film plant random sampling. A totalof 36 employees had worked only in the film plant(i.e.,'only in thefilm plant'refertso film plantworkers with no known experienceon the D-1 maker or have had no previous work experiencein the chemical plant),6 film plantemployees were known to have worked on theD-1 maker and 18 employees had worked, atsome time previously,in the chemical plantbut were not on theD-1 maker (Table 24). For all filmparticipant(sn = 76, random sample and volunteers)a, totalof 49 had worked only in the film plant,7 were known to have worked on theD-1 maker and 20 had worked, at some time previously,in the chemical plant. Among the 60 employees of therandom sample, therewere no substantial demographic differences(Table 25) between the only film,theD-1 maker and prior chemical historygroups. However, therewere significandtifferencesin serum fluorochemicallevelsamong thesethreegroups of film plantworkers. Those employees who have only worked in the film plant(butnot on D-1 maker or previouschemical plant history)had significantllyower mean PFOS levels(Table 2-6).The geometric mean of PFOS foronly film plantworkers was 0.110 ppm (95% Cl 0.094-0.129)compared to 0.289 ppm (95% CI 0.159-0.527)foremployees known to have worked on the D-1 maker and the geometric mean was 0.178 ppm (0.137-0.233)forfilm plantemployees with priorhistoryin chemical. A similarsignificanatssociationa,lbeitat a lower ppm level,was observed for POAA. The only film plantemployees had significantllyower PFHS levelswhen compar-ed to film plantworkers with a previoushistoryin chemical; theirPFHS levelswere nonsignificantllyower than those who worked on the D-1 maker. There were no significandtifferencesin seralevelsof the remaining fluomchemical levelsamong the threegroups of film employees. 3M EPI-0006 Page 32 of 85 Interestinglya,llfilm plantworkers with a previoushistoryof having worked in thechemical planthad M556 values that were below the LLOQ. We do note thattheD-1 maker group had comparable levelsof M570 to the only film or film with previoushistoryin chemical groups (seeTable 26). We had hypothesized theD-1 maker group may have had higherlevelsbecause of their use of methyl FOSE amide which may metabolize to the analyteM570. Provided in Table 27 areratiosof fluorochemicals.The median ratioswere comparable forthese groups of filmplantworkers in therandom sample. Restrictingthe analysesto film employees with no D-1 maker or chemical plant experience,therewere no significandtifferencesby.age forthefour currentjob categoriesanalyzed:engineer/lab,filmprocessor,maintenance and administrative(Table 28). Although theirserum levelswere substantiallbyelow theircounterpartsin chemical,maintenance employees working in the film planthad significantlhyigher PFOS, POAA and M570 levelsthan theengineer/labgroup withinthefilm plant(Table 29). Engineer/lab,film processorsand administrativweorkers had comparable fluorochemicalserum levels. Median fluorochemicalratioswere comparable among thesejob categoriesof therandom sample of filmplantworkers (Table 30). Similar findingswere observed when allfilmplantparticipantwsere analyzed fordemographics and serum fluorochemicallevels(Tables31-33). Located in Appendix 0 are scatterplotosf the only film group foreach fluorochemicalregressedon years worked in film. Because maintenance workers had higherlevels,on average,than the otherthreejob groups among the only film employees, they arenumbered on the graphs. From theseanalysesthereissome suggestion that PFOS and POAA levelsmay increasewithin thefirstfew years of working atthe Decatur filmplantand then subsequentlyplateau.However, 3M EPI-0006 Page 33 of 85 unlikechemical workers, thereisno linear(orquadratic)associationobserved forPFHS. The remaining fluorochemicals showed no associationwith yearsworked in film. DISCUSSION The goal of thisresearcheffortwas to quantify,based on random sampling, the relationshipof employee serum levelsof seven fluorochemicalsat theDecatur chemical and film plants. In thatregard,thedata collectedand analyzed presenta convincing pictureof significantllyower serum fluorochernicalevelsamong employees who have only worked in the film plantwhen compared to those who arecurrentchemical plant employees. For example, comparing the geometric means foreach fluorochemical between chemical operatorsand those employees who only have worked in film,we observedthe followingratios:PFOS (1.481/0.110);PFHS (0.428/0.015)P;OAA (1.887/0.052)P;FOSAA (0.011/0.002)M;570 (0.229/0.022)a;nd M556 (0.044/0.003). These ratiose,xcept forPFOSAA, suggesta 10-foldor greaterdifferencebetween chemical operatorsand film plantemployees who work severalhundred yards away from Building 3. These only film plantworkers appear to have a geometric mean value for PFOS thatis approximately 3-4 times higherthan thepooled geometric mean (0.029 ppm) from 64 samples obtainedfrom 18 U.S. blood banks; thus,we suspect that occupationalexposure to PFOS occurs withinthe film plantalthough atmuch lower levelsthan among employees working atthe chemical plant. Among film plantemployees we alsoestablishedthefactthatworkers on the D-1 maker have serum PFOS levelsapproximately 3 times higherthan those who have never 3M EPI-0006 Page 34 of 85 worked on theD- Imaker nor have worked inthechemicalplant.Unexplainedisthe POAA levelsof theseworkers on theD- I maker as wellas thelevelsobservedamong otherfilmplantemployees. We confirmedseveralhypothesesforthechemicalplantemployees.Firstc,ell operatorshave thehighestserum levelsof PFOS and PFHS althoughtheirserum levels forotherfluorochemicalanalyteswere similartootherchemicalemployeeswho were involvedwiththechemicalreactors(i.e.c,hemicaloperatorsand maintenanceworkers). Second,chemicaloperatorsand maintenanceworkershad comparable serum fluorochemicalevels.Besidestheirhigherlevelsof PFOS and PFHS, theyboth had significanthliygherlevelsofM570 (themethylFOSE alcoholmetabolitea)nd toa lesser degreetoPFOSAA which istheethylFOSE alcoholmetabolite(aswellas an FC product itselFfC,-129). Chemical operatorsb,utnot maintenanceworkers,had higherlevelsof PFOSA. Both chemical operatorsand maintenanceworkers had moderatelyhigherlevels ofM556 thantheotherjobcategoriesT.hese data suggestthat,beyond generalplantbasedenvironmentalexposuretoPOSF and PHSF (which we assume isprimarily throughinhalatioannd conversiontoPFOS and PFHS, respectivelyt)h,echemical operatorsand maintenance workershave higherserum levelsas a resulotf their occupationaelxposureto thefluorochemicaplroducts.These occupationaelxposures may be from theFC alcoholsF,C amides,and FC acrylatesB.ecause these fluorochemicalhsave much lower vaporpressurethanPOSF and PHSF, thesedatamay indicatethattheexposureto thesechemicalproductswithinthechemicalplantis relativelyimitedtowithinBuilding3 and Building4N. Third,wasteoperatorswere comparable tochemicaloperatorsforserum levelsofPFOS and PFHS but,likethecell operatorsd,idnot have higherlevelsof thefluorochemicalanalytesF.ourth,mill operatorswere generallymuch younger employees 3M EPI-0006 Page 35 of 85 and theirhighestfluorochemical serum levelwas to POAA. Yet, the milloperators'POAA levelswere lower than those of celloperators,chemical operatorsand maintenance workers. This sugaests thereis plant-basedexposure of POAA well beyond the Building 4 area which may be due to the factthatPOAA isa by-product of theelectrolyticcellproduction.Finally,the data support thehypothesis thatthose individuals(e.g.,engineersand secr-etariewsh)o are much lesslikelyto have routineoccupationalexposure within thechemical plant,do, indeed have lower serum fluorochemicallevels.Employees who have only worked in Building 1 which isimmediately acrossthe walkway from Building 3, have serum fluorochemicallevelsthatrange between 7 (PFOS, PFHS) and 15 times (PFOSAA) lower than employees who have only worked in Building 3. Our analysesof fluorochemicallevelsin serum from randomly selected employees strengthenthe recommendations thatwere recentlymade in a Decatur industriahlygiene assessment analysis[Logan, 1998]. There is a strongcorrelation between the higheremployee serum levelsin the presentstudy and air,surfaceand personalmonitoring measurements which occurred duringthe industriahlygiene assessment.In the industrialhygiene assessment, Building 3 had the highestaverage airbornetotalfluorochemicallevelswith each value derivedfrom the totalmass of detectedtargetanalytesin each sample (POSF, PHSF, FC amides, FC alcohols,FC acrylates()seebelow): Bldg No. 1 3 4 Outside air 3 *mg/m Results of Fluorochemical Tube Air SamRIes No. Samples Average* 19 0.0145 Low* 0.000 66 1.6884 0.0070 10 0.1269 0.0047 3 0.0861 0.580 Mgh* 0.0601 38.0583 0.5216 0.1247 3M EPI-0006 Page 36 of 85 Surface wipe sampling was alsoconducted throughout the chemical plant(Buildings 1,2, 3,4, 17,38,49, 51 and 57). Sample resultsindicatedthatfluorochemicalswere found in nearlyallsamples with largevariationsin concentration.Building 3 had the highest surfacefluorochemical contamination with the average surfaceconcentrationgreaterthan 2. 100 ug/100cm Also, methyl FOSE alcoholwas the largestcontributorof fluorochemicalsfound throughoutsurfacewipes in Building3. Hand-wipe samplings indicatedthatemployees who had washed theirhands had very low levelsof fluorochemicalsdetected.Methyl FOSE alcoholand POAA were thecompounds found most oftenon employees' hands. Thus, the present study'sserafluorochen-ticallevels, observed by job categoriesand buildinglocations,stronglysupportthe recommendations bome from industrialhygiene assessments. These recommendations include specific engineeringcontrolsto reduce inhalationexposure, appropriatepersonalprotective equipment to prevent overexposure and appropriatepersonalhygiene practicesamong employees to remove skin concentrations. For thefirsttime we have shown a relationshipbetween serum PFHS levelsand thenumber of year-sworked in chemical. This findingwas observed acrossvarious currentjob categorieswithinchemical which suggestsPHSF, due to itshigh vapor pressure,islikelypresentthroughout the chemical plantpremises. The phannacokinetics of PFHS are unknown, although due to theshorterchain length,we suspect thebiologicalhalf-lifmeay be lessthan PFOS. We observed only a modest associationbetween yearsworked in the chemical plantand serum PFOS, and to a lesserextentPOAA, levels.These associationsappear to be more evidentamong employees within theirfirstfiveyears as demonstrated by significanqtuadraticassociationsfound with both male chemical operatorsand engineers/laboratopreyrsonnel. 3M EPI-OW6 Page 37 of 85 Like theirmale counterpartsf,emale chemicaloperatorsappearto have increased PFHS levelswith yearsworked. However, unliketheirmale counterpartst,herewas no apparentlinearassociatiobnetween PFOS and yearsworked. Whether thisisdue to differenwtork practicese,xposurepatternsorpharrnacokineticosnce absorbed,remains tobe determined.Gender-relatedifferenceisnthetoxicokineticosf POAA have been reportedforratsalthoughthemechanism of excretionmay be speciesdependentsince thesegenderdifferencewsere notobservedin nticer,abbitsor dogs [Griffitahnd Long, 1980;Hanhijarviand Ylinen,1988]. The half-lioffePOAA was estimatedtobe 7 times higher(7days)inmale ratsthanfemalerats. A limitatiotno thisstudydesignwhich must be consideredintheinterpretatioofn thedatawas our inabilittyomore accuratelqyuantifyan employee'swork history experience.Decaturwork historyrecordsprovidedepartmentnumbers and job titlebsut theydo notprovideinformationregardingwhere someone worked (e.g.w,hat building(s) orwithwhat specififcluorochemicalsS)e.lf-reportewdork historyinformationobtained by questionnairweas highlycorrelatewdithDecaturwork historyrecordinformation; neverthelesst,hespecificitoyf where someone worked and withwhat chemicalswas not known. Becausemany operationsareinbatchmode, thelikelihooodf determining specificitoyf workload fluorochemicaelxposureamong chemicaloperatorsisnot possible.Furthermore,such recordsdo notexistback intime.Neverihelessw,ith use of theemployees current(orlongestj)ob alongwithadditionaslurrogatevariablexposures (yearsworked inchemical,buildingnumber) we were abletocompare and contrast fluorochemicalevels.The leastpredictivoefthesethreevariableOsob type,building and yearsworked) was yearsworked withtheexceptionofPFHS where a stronglinear associatioenxistedacrossjob categoriesforPFHS with yearsworked. 3M EPI-0006 Page 38 of 85 We didnot observean associatiobnetween hand-to-mouthusage or hand cleanlines(sfrequencyofwashing hands)and serum fluorochemicallevels.Itispossible an associatiomnight have been masked becauseindustriahlygienehad instituteadn aggressiveducationalcampaign severalmonths priorto thecollectioonf blood samples inthisstudy;thus,currentpracticemsay notbe indicativoef pastpracticesB.ecause the half-lifoef PFOS isestimatedtobe@1000 days or more, such an associatiomnay not be discoverablweith thisstudydesign. The serum levelsobservedinthisstudyforPFOS and POAA arenotdifferent thanthosethathave been previouslyreportedforthisstudyand other3M occupational populations[Olsenetal.,1998a,1998b,1999]. Olsen etal.[1999]have not associated hepaticorlipidabnormalitiewsithPFOS levelsintheDecaturand Antwerp plant populationsthatunderwent voluntarymedicalsurveillancien 1995 and 1997. Hepatic lipidor hormone levelshave notbeen associatewdith serum POAA levelsamong 3M CottageGrove male workers who have experiencedhigherserum fluorochemicalevels thanthosedeterminedin thepresentstudyfortheseDecaturemployees [Gillilanadnd Mandel 1996;Olsen etal.1998a;1998b]. In summary, theobjectivoef thisproposedresearchstudywas to characteriz,e viarandom sampling,thedistributiofnemployee serum levelsofPFOS, PFHS, POAA, PFOSAA, M570, PFOSA and M556 atthe3M Decaturchemicaland filmplants.The dataobtainedfrom thisexposureassessmentinvestigatioanre importantforseveral reasons.Firstt,hesedataallowfora betteurnderstandingof theexposuredistributionf serum fluorochemicalevelsinboththechemicaland filmplantemployee populations. Second,thesedatamay serveasfuturereferenceregardinghuman exposureassessment 3M EPI-0006 Page 39 of 85 forthefilmaswellas thechemicalplantintheareaof healthstudiesand exposure reduction.Third,thedatamay be used fortheconstructioonf an exposurematrixforthe anticipateudpdateof theretrospectivceohortmortalitystudyof theDecaturemployee population.Finallyt,hisstudywillallowfortheopportunitfyoremployeestoknow their own serum levelsfortheseseven fluorochemicalsand encouragefurtherpracticelseading toa reductionintheirserum fluorochemicalevelsby thevarietyof exposure-reduction methods recommended intheDecaturindustriahlygieneexposureassessmentreport [Logan,1998]. 3M EPI-0006 Page 40 of 85 REFERENCES GriffitFh,Long J. Animal toxicitsytudieswithammonium perfluorooctanoatAem. Ind Hyg Assoc 1980;41:5760583. HanhijarviH, YlinenM. A proposedspeciesdifferencientherenalexcretioonf perfluorooctanoiacidinthebeagledog and rat.In:Beynen A, SolleveldH, eds.New Developments in Biosciencest:heirImplicationfsorLaboratoryAnimal Sciences. Dordrecht:MartinusNijhoff.1988:409-412. Hansen K. LaboratoryReport:AnalysisofFCs inSerum Samples CollectedFrom Employees at3M Decatur.St.Paul:3M EnvironmentalLaboratory.June,1999. Logan PW. 3M DecaturFluorochemicalIndustriaHlygiene Exposure Assessment. (Unpublishedreport.S)eptember 15,1998. GillilanFdD, Mandel JS. Serum perfluorooctanoiacidand hepaticenzymes, lipoproteinasnd cholesterola:studyof occupationalleyxposedmen. AJIM 1996;26:560568. OlsenGW, GillilanFdD, Burlew NW, BurrisJM, Mandel JS,Mandel JH. An epidemiologicinvestigationfreproductivheormones inmen with occupationaelxposure to perfluorooctanoiacid.J Occup Env Med 1998a;40:614-622. OlsenGW, BurrisJM, Burlew MM, Mandel JH. An epidemiologiicnvestigatiofn plasma cholecystokiniannd hepaticfunctioninperfluorooctanoiacidproduction workers.(3M unpublishedreport)1;998b. Olsen GW, BurrisJM, Mandel JH, Zobel LR. Serum perfluorooctanseulfonateand hepaticand lipidclinicaclhemistrytestsinfluorochemicaplroductionemployees. J Occup Env Med (1999,inpress). Roach DE. FluorochemicalControlStudy.3M UnpublishedReport.3M Company:St. Paul,N4N May 25, 1982. Schuman LM (1982).LettertoDr.FrankUbel (3M MedicalDirectorc)oncerning FluorochemicalControlStudy.April29, 1982. Vollmer D. QuantitativDeeterminationofPFOS, PFOSA, PFOSAA, POAA and PFHS inHuman Serum by LC/MS/MS. SaltLake City:NorthwesBtioanalyticalJ.uly1999. 3M EPI-0006 Page 41 of 85 ACKNOWLEDGEMEENT The investigatorgsreatlyappreciatethecontributionsof Kim Young to thisfinalreport. Table1. RandomsampleselectiboynDecatudrepartmenwtisthpercenptarticipation Dept Number Dept Name Employees with 090 locationcodes 7613 3M/Dyneon RelatedDecatur TotalN Sample Size Participated Yes 112 30 25(83) 7620 DecaturBldg 2 Operations 25 10 7(70) 7621 Bldg 49 Operations 1 7630 DecaturBldg 3 Operations 113 30 25(83) 7641 DecaturBldg 4N Operations 60 15 13(87) 7609 37 6825 8038 7604 7605 7616 7617 7622 DecaturSMD Maint-SA&C 54 15 Mfg ServcicesProcessEng 2 26 ProcessInstrumentatio&n CN 1 Supply Chain ResourceUnit 1 DecaturSMD Chem FactoryAdm 4 DecaturSMD Chem QualityAss 25 DecaturChem Ship Rcv Whse 24 DecaturSMD Logistics 10 DecaturPCPD FF Admin 24 14(93) 23(88) 5980 DecaturEHS&R 21 5 Employees regardlessof 090 or 190 location 6853 Auto & Chem Mkts Eng 8(090) 5 6853 Auto & Chem Mkts Eng 6(190) 4290 Auto & Chem IT NPUR&D 1(190) 4294 Auto & Chem IT Mfg - Quality/S 4(190) 4297 Auto & Chem IT CMG Mfg 6(190) Waste watertreatmentemployee 5984 DecaturWaste Treatment 6 6 5(100) 4(80) 6(100) Table 1 (continued) Dept Number Dept Name Heatingplantemployees 5982 DecaturHeatingPlant Film plantemployees Allremaining190 location Department codes Dyneo TOTAL Dyneon employees TotalN Sample Size 10 4 Participated Yes 3(75) 482 78 25 8 1025 232 55(71) 6(75) 186(80) Table2.Number(andpercenotf)randomsamplev,olunteaenrdalplarticipcahnetmicaelmployeebsydemographcic Sample (N 126) Volunteers(N 61) A Gender Female Male N -(0/0) N 24 (19) 9 102 (81) 52 (85) 1 Only Worked In Chemical Yes No 67 (53) 33 59 (47) 28 (54) 1 (46) CurrentJob CellOperator 5 (4) 4 Chemical Operator 47 (37) 17 Engineer/Lab 23 (18) 14 Maintenance 11 (9) 6 MillOperator 13 (10) 11 Secretary 4 (3) 1 Supervisor/Mgmt 18 (14) 8 Waste Operator 5 (4) 0 (7) (28) (23) (10) (18) (2) (13) (0) LongestJob CellOperator 1 (1) 2 Chemical Operator 57 (45) 20 Engineer/Lab 21 (17) 10 Film Processor 3 (2) 2 Maintenance 14 (11) 6 MillOperator 14 (11) 12 Secretary 6 (5) 1 Supervisor/Mgmt 7 (6) 8 Waste Operator 3 (2) 0 (3) (33) (16) (3) (10) (20) (2) (13) (0) Table2. (continued) Chew Gum Always/Frequently Sometimes Rarely/Never Chew Tobacco Yes No Smoke Cigarettes Yes No Hand toMouth Contact Yes No WashHands Yes No Sample (N 126) 22 (18) 32 (26) 70 (56) 19 (15) 105 (85) 41 (33) 82 (67) 84 (68) 40 (32) 101 (81) 23 (19) Volunteers(N 16) A N- - (%) - 14 (23) 20 (33) 26 (43) 6 (10) 54 (90) 1 14 (23) 1 46 (77) 42 (70) 18 (30) 5 42 (70) 14 18 (30) 4 Table 3. Percentageofemployeesfrom therandom sample,volunteerasnd allparticipantwsho respondedthat worked inDecaturbuildings/areas CurrentlyWork inBuildings... 1 2/49 30S 3 NOS 4N 4 Mill/Extruder 17 38/51 61 Film Wastewater Ever Worked inBuildings... 1 2/49 30S 3 NOS 4N 4 Mill/Extruder 17 38/51 42 61 Film Wastewater Sample (N 126) N (0/0) 42 (33) 22 (17) 39 (31) 41 (33) 37 (29) 45 (36) 18 (14) 24 (19) 15 (12) 23 (18) 12 (10) 14 (11) Sample N 52 (41) 33 (26) 68 (54) 72 (57) 64 (51) 80 (63) 22 (17) 36 (29) 29 (23) 25 (20) 41 (33) 19 (15) Volunteer(N = 61) N C/o) 22 (36) 16 (26) 21 (34) 23 (38) 22 (36) 29 (48) 10 (16) 21 (34) 14 (23) 14 (23) 9 (15) 10 (16) Volunteer N (%) 25 (41) 14 (23) 30 (49) 31 (51) 30 (49) 43 (70) 9 (15) 21 (34) 18 (30) 12 (20) 19 (31) 7 (11) AllPart N 64 38 60 64 59 74 28 45 29 37 21 24 Al N 77 47 98 103 94 123 17 57 47 37 60 26 Table 4. Serum fluorochemicallevels(ppm) of random sample,volunteerasnd allparticipancthemicalemployees Random Sample Geometric Mean Mean Median Range PFOS 1.505 0.941 1.140 0.091- 10.600 PFHS 0.345 0.180 0.170 0.005- 1.880 POAA 1.536 0.899 1.300 0.021- 6.760 PFOSAA 0.023 0.008 0.008 0.001- 0.269 M570 PFOSA 0.150 0.062 0.081 a 0.013 0.067 0.008- 0.992 0.012 0.0005- 0.612 M556 0.052 0.022 0.026 0.001- 0.406 Volunteers Geometric Mean Mean Median Range 1.259 0.758 0.877 0.052- 4.940 0.272 0.122 0.125 0.001- 1.580 1.206 0.649 0.908 0.015- 4.640 0.026 0.007 0.006 0.001- 0.234 0.173 0.068 0.054 0.004- 3.100 0.029 0.006 0.007 0.005- 0.443 0.041 0.020 0.018 0.003- 0.329 Geo Mean m 1.424 0. 0.321 0. 1.429 0. 0.024 0.( 0.158 0.( 0.051 0.0 0.048 0.0 a. significantdliyfferen(tp < .05)geometricmean thanvolunteerss,tudent'tstest Table 5. Demographic characteristiocfscurrentjob categorieosf random sample(N Age Mean SE (standard error) Median Range BMI Mean SE Median Range Years Worked In Chemical Mean SE Median Range Cell Operator' (N = 5) 45c 1.2 44 40-50 25.8 1.5 25.0 22.1-30.0 23 b.de 2.0 24 17-29 Chemical Operatorb (N = 47) 42 e,&h 1.2 43 25-62 28.3' 5.1 27.8 20.2-47.5 I Iall's 1.4 10 1-31 Engineer/ Lab' (N = 23) 419,' 1.7 42 23-58 27.6 5.0 27.6 18.5-38.4 Maintenanced (N = 11) 4 1g,h 2.5 42 27-52 26.9 2.8 26.6 22.8-32.5 15e 2.8 15 1-37 9&19 3.0 4 1-26 Gender* Female Male N-N 1(20) 4(80) @LN 10(21) 37(79) 6(26) 17(74) 0(0) 11(loo) Only Worked In* Chemical Yes No Hand to Mouth Contact Yes No Wash Hands Always Less frequently 3(60) 2(40) 3(60) 2(40) 5(100) 0(0) 23(49) 24(51 35(76 11(24) 40(87) 6(13) 14(61) 9(39) 11(50) 11(50) 16(73) 6(27) 6(55) 5(45) 8(73) 3(27) 10(91) 1(9) 126)of chemicalemployees Mill Operator' (N = 13) Secretaryf (N = 4) 35abtg.h 2.3 34 27-45 45 e 4.2 45 42-49 27.7 5.4 27.3 19.6-42.0 22.4bs 2.0 22.0 20.9-25.1 3a-bc@t&b 1.5 1 1-21 e 15 4.9 18 2-25 1(8) 12(92) 4(0) 0(0) 11(85) 2(15) 10(77) 3(23) 12(92) 1(8) 1(25) 3(75) 2(50) 2(50) 2(50) 2(50) *Current job types significantlydifferent,p < .05 chi square statistic (a-h) comparison for each currentjob category using student'st (p < .05) Table 6. Mean, geometricmean and median of serum fluorochemicalbsy demographic characteristiocfsrandom chemicalemployees Gender Female Male Mean 0.686 1.697 PFOS Geometric Mean Median 0.459 0.897* 0.412 1.310 Hand To mouth Yes No 1.362 1.810 1.008 0.674 1.140 0.954 Wash hands Yes No 1.581 1.179 0.897 0.459 1.190 0.735 Worked onlyin chemical Yes No 1.113 1.927 0.723 1.271 0.784 1.550 Mean 0.118 0.398 PFHS Geometric Mean Median 0.080 0.218* 0.082 0.223 0.278 0.474 0.173 0.185 0.168 0.181 0.360 0.259 0.188 0.136 0.176 0.126 0.247 0.456 0.126 0.270* 0.141 0.306 Mean 0.691 1.735 1.504 1.602 1.597 1.263 1.307 1.797 Table6. (continued) PFOSAA Gender Female Male Geometric Mean Mean Median 0.011 0.026 0.003 0.010 0.002 0.009 M570 Geometric Mean Mean Median 0.077 0.168 0.053 0.089* 0.052 0.073 PFOSA Geometric Mean Mean Median 0.03,7 0.012 0.068 0.013 0.014 0.011 Hand To Mouth Yes No 0.026 0.019 0.009 0.007 0.008 0.008 0.153 0.139 0.085 0.079 0.637 0.081 0.050 0.080 0.012 0.013 0.012 0.011 Wash hands Yes No Worked only inchemical Yes No 0.025 0.017 0.009 0.005 0.009 0.003 0.024 0.022 0.009 0.008 0.008 0.008 0.162 0.893 0.088 0.055 0.672 0.632 0.063 0.045 0.013 0.009 0.013 0.009 0.142 0.160 0.075 0.088 0.063 0.074 0.059 0.065 0.012 0.014 0.013 0.011 ttest,p < .05 Table 7. Mean, range,geometricmean and 95% confidenceintervaolfgeometricmean of serlunfluorochemice categorieasmong random sample (N = 126)of chemicalemployees PFOS Mean Range G. Mean 05% C.I. PFHS Mean Range G. Mean 95% C.I. Cell Operator' (N = 5) Chemical operatorb (N = 47) Engineer/ Labc (N = 23) 2.903 0.325- 6.840 1.970r-c" 0.732- 5.304 1.781 0.471- 7.260 1.481eegh 1.250- 1.755 0.634 0.095- 1.740 0.391al@dgh 0.256- 0.597 1.062 0.083- 1.880 0.428 0.071- 1.860 0.697c-"elg 0.308r-e,' 0.228- 2.130 0.246- 0.386 0.171 0.005- 0.905 0.078&.b.gh 0.046- 0.134 Maintenanced (N= 11) Mill Operatoe (N = 13) Secretar/ (N = 4) 1.672 0.291- 4.060 1.299ce,f 0.822- 2.054 0.718 0.230- 2.040 0.589al.,Ih 0.419- 0.828 0.497 0.220 - 1.140 0.397"',"' 0.195- 0.807 0.237 0.023- 0.790 0.1534 0.084-0.280 0.109 0.028- 0.374 0.074ab.&h 0.047-0.116 0.082 0.027- 0.172 0.066ab,g 0.031- 0.140 POAA Mean Range G. Mean 95% C.I. 2.213 0.126- 3.640 1.428c-' 0.422- 4.833 2.252 0.150- 6.760 1.887cfg 1.573- 2.265 0.376 0.035- 2,320 1.483 0.211-4.680 0.20gikbde.f.&h 0.134- 0.324 1.095c-f 0.670- 1.791 1.383 0.450- 2.340 1.266cf1 0.985- 1.629 0.183 0.095-2.611 0.172&-b,"c-g,' 0.113- 0.260 PFOSAA Mean Range G. Mean 95% C.I. 0.006 0.001- 0.016 0.003b-d-c 0.001- 0.009 0.036 0.001- 0.269 o.o I Iae-f 0.007- 0.018 0.014 0.001- 0.073 0.005bde 0.003- 0.010 0.034 0.001- 0.083 . 0.017"c'f 0.007- 0.043 0.020 0.004- 0.038 0.015&-clf 0.010-0.024 0.002 0.001- 0.004 0.002"dc 0.001- 0.003 M570 Mean Range G. Mean 95% C.I. 0.035 0.024-0.056 0.033bd 0.024 - 0.045 0.229 0.009- 0.992 0.074 0.008- 0.410 0.13la-c,r-fg 0.094 - 0.182 0.049bd 0.034 - 0.071 0.268 0.038- 0.701 0.045 0.025- 0.115 0.204a-c-e-fg 0.124 - 0.335 0.042 ld 0.034 - 0.051 0.039 0.010- 0.072 0.030bd 0.013 - 0.071 Table7. (continued) PFOSA Mean Range G. Mean 95% C.I. 0.003 0.001- 0.011 0.002'-e 0.001 - 0.005 0.099 0.001- 0.487 0.017 0.001- 0.063 0.028"c-'tg 0.005bg 0.017- 0.048 0.002- 0.010 0.065 0.001- 0.569 O.Oogb 0.003 - 0.025 0.040 0.012- 0.204 0.026a-c 0.016- 0.040 M556 Mean Range G. Mean 95% C.I. 0.014 0.003-0.028 0.0 11 bd 0.005- 0.218 0.074 0.001- 0.380 0.019 0.002- 0.127 0.044acctg O.Olobd 0.032- 0.061 0.006- 0.015 0.115 0.016- 0.406 0.013 0.002- 0.045 0.074a-c-e-fg O.Oogbd 0.041- 0.135 0.006- 0.015 (a-h)comparisonsforeach currenjtob categoryusingstudenfs t,p < 0.5 0.031 0.009- 0.080 0.021 0.009- 0.054 0.015 0.003- 0.030 O.Olobd 0.003- 0.031 Table8. Fluorochemicalratiobsy currenjtob categoriefsorrandom sample(N 126)of chemicalemployees PFOS/PFHS Mean Median Range PFOS/POAA Mean Median Range PFOS/Analytes Mean Median Range M570/M556 Mean Median Range PFOSAA/M556 Mean Median Range PFOSA/M556 Mean Median Range Cell Operator' (N = 5) 3.0 d@eh 3.3 1.5-3.9 1.5e 1.0 1.0-2.6 52.6b-h 56.1 4.9-93.3 3.5 2.8 2.0-7.5 0.4e 0.3 0.1-1.3 0.4 0.1 0.04-1.2 Chemical Operator' (N = 47) 5.5 d-c 5.2 1.1-14.8 0.9 c.fg 0.9 0.3-3.1 7.4k' 4.7 1.3-61.8 3.7c,e 3.0 0.5-8.9 o.g e 0.2 0.03-8.5 3.0 0.3 0.03-30.7 Engineer/ Lab' (N = 23) Maintenanced (N = 11) 5.9 dc 5.2 1.6-18.8 9.2 abcs 10.3 3.7-16.5 2.2b,d@e 1.7 0.7-4.4 1.2c-cf 1.3 0.7-1.5 7.4ah 3.9 0.8-36.8 5.8b 5.7 1.9-14.4 4.6k' 3.2 1.4-11.7 3.3 c,e 2.7 1.2-8.7 1.1 0.8 0.03-4.5 0.8e 0.1 0.01-3.1 2.5 0.3 0.03-18.3 0.40 0.1 0.02-2.7 Mill Operatoe (N = 13) Secretaryf (N= 4) 8.5 abg-h 8.3 4.3-14.9 6.1 5.8 4.8-8.3 0.5"cf9 0.5 0.2-0.9 2 .5 b,d,e,h 2.0 1.7-4.4 6.72 4.8 2.5-20.5 5.8Id 4.3 1.2-21.3 6.7a 6.9 3.2-9.7 3.2 3.2 2.2-4.3 2.4a-d,f-h 2.3 0.2-5.7 0.2 e 0.2 0.05-0.5 7.6d 2.1 0.8-52.3 2.5 2.7 0.9-3.7 Table 9. Demographic characteristiocfslongestjob categorieosfrandom sample(N = 126)of chemicalemployee Age Mean SE Median Range BMI Chemical Operatora (N = 57) 43 1.1 44 25-62 Engineer/ Lab' (N = 21) 40 2.3 41 23-58 Maintenance' (N = 14) 43 1.9 43 27-54 Mill operatord (N = 14) 35 1.8 32 27-45 Secretary' (N = 6) 46 d 1.9 45 42-54 Mean SE Median Range Years Worked InChemical Mean SE Median Range 27.8 0.7 27.5 18-47 14d 1.3 14 1-31 28.8 1.1 28.0 21-38 12d 2.7 8 1-37 27.2 0.7 26.7 23-33 11 3.0 8 1-36 27.3 1.4 26.9 20-42 3ab 1.4 1 1-21 24.2 1.4 23.9 21-30 20d 4.4 23 2-33 Gender Female Male Only Worked InChemical* Yes No Hand to Mouth Contact Yes No Wash Hands Yes No 8 (14) 49 (86) 27 (47) 30 (53) 39 (70) 17 (30) 47 (84) 9 (16) 5 (24) 16 (76) 15 (71) 6 (29) 11 (52) 10 (48) 15 (71) 6 (29) 0 (0) 14 (100) 6 (43) 8 (57) 9 (64) 5 (36) 86 (12) 2 (14) 2 (14) 12 (86) 12 (86) 2 (14) 11 (79) 3 (21) 93 (13) 1 (7) 6 (100) 0 (0) 3 (50) 3 (50) 3 (50) 3 (50) 3 (50) 3 (50) *Significantlydifferentpercentage among longestjob categories,chi square,p <.Os (a-h) comparisons for each longest job category using student's t,p <.05 Table 10. Mean, median,geometricmean and 95% confidenceintervalosf geometricmean of serum fluorochemic categorieasmong random sample (N = 126)ofchemicalemployees PFOS Mean Range G. Mean 95% C.I. PFHS Mean Range G. Mean 95% C.I. POAA Mean Range G. Mean 95% C.I. PFOSAA Mean Range G. Mean 95% C.I. M570 Mean Range G. Mean 95% C.I. Chemical Operator' (N = 57) Engineer/ Lab b (N = 21) 2.088 0.338- 7.880 1.697b,d,e,f 1.440- 1.998 0.520 0.095- 1.740 0.33oac.d.g 0.219 - 0.496 0.543 0.073 - 1.880 0.116 0.005 - 0.420 0.388b,c,d.e.f. 0.314 - 0.480 0.067kc 0.041 - 0.110 2.293 0.182 - 6.760 1.972b,C,C,f 1.694- 2.295 0.287 0.035 - 1.000 0.198 acd,g 0.134 - 0.295 0.032 0.001 - 0.269 O.Olob,e 0.007 - 0.016 0.014 0.001 - 0.073 0.005 a.c,d 0.027 - 0.010 0.213 0.009 -0.992 0. 120b.d,e,f 0.089 - 0.161 0.060 0.008 - 0.164 0.046ac 0.033 - 0.064 Maintenancec (N = 14) 2.250 0.291- 10.600 1.490b,le,l 0.933 - 2.379 Mill Operatord (N = 14) 0.735 0.230- 2.040 0.609a-c,' 0.441 - 0.842 0.297 0.023 - 1.250 0.176abd,e 0.101 - 0.307 0.107 0.028 - 3.744 0.074ac 0.048 - 0.112 1.667 0.211 - 4.680 1.229a.b,c,f 0.797 - 1.900 1.383 0.450 - 2.340 1.274',ef 1.009 - 1.609 0.029 0.001 - 0.083 0.014b,e 0.007 - 0.031 0.019 0.004 - 0.038 0.014',e 0.009 - 0.022 '0.269 0.038 - 0.701 0.200b,d,c,f 0.126 - 0.315 0.046 0.025 - 0.115 0.043 a.c 0.035 - 0.052 Secretary' (N = 6) 0.388 0.129- 1.140 0.295a-c,g 0.163 - 0.533 0.070 0.027 - 0.172 0.057ac 0.033 - 0.097 0.143 0.053 - 0.261 0.124 ,c,d.g 0.076 - 0.203 0.002 0.001 - 0.004 0.001 a,c,d,fg 0.001 - 0.002 0.032 0.010 - 0.072 0.025 a.c 0.014 - 0.046 SUF 0 0.09 0 0.20 0 0.01 0. 0.03 0. 0.02 0. 0.06 0. 0.00 0. 0.00 0. 0.01 0. 0.03( Table10. (continued) PFOSA Mean Range 0.088 0.001- 0.487 G. Mean 95% C.I. M556 Mean Range 0.021'-' 0.0126- 0.0340 0.070 0.001- 0.380 G. Mean 95% C.I. 0.039b.d.,,,l 0.029- 0.054 0.016 0.001- 0.063 0.004a,l 0.002- 0.009 0.014 0.002- 0.039 0.009a.C.C.Ig 0.006- 0.014 0.095 0.001- 0.612 0.010 0.003- 0.028 0.117 0.015- 0.406 0.060,d.c.1 0.038- 0.123 (a-h)comparisons for each currentjob categoryusing student'st,p < .05 0.042 0.012- 0.204 0.028b.1 0.018- 0.044 0.013 0.002- 0.045 0.0loa-C 0.006- 0.015 0.024 0.009- 0.080 0.017' 0.009- 0.033 0.011 0.003- 0.030 0.007a-c-g 0.003- 0.017 0 0.00 0 0.001 0 0.003 0 0.004 Table 11. Ratiooffluorochemicalevelsby longestjob categorieasmong random sample (N = 126)of chemicale PFOS/PFHS Mean Median Range PFOS/I?OAA Mean Median Range PFOS/Analytes, Mean Median Range M570/M556 Mean Median Range PFOSAA/M556 Mean Median Range PF6SA/M556 Mean Median Range Chemical Operator' (N = 57) 5.0c,d,g 4.9 1.1-14.8 i.ob,e,f 0.9 0.3-2.9 13.4 5.4 1.3-99.4 3.8b,f 3.0 0.5-9.5 O.gd 0.2 0.03-8.5 2.6f 0.3 0.03-30.7 Engineer/ Lab' (N = 21) 5.8c,d 4.8 1.9-18.8 Maintenancee (N = 14) 9. 1 ab,e,f 9.2 3.7-16.5 2 .0acd,tg 1.7 0.3-4.4 1.3b,d,e,f 1.3 0.6-2.3 Mill Operatord (N = 14) 8.8ab,e,f 8.3 4.3-14.9 0.5b,c,c,r 0.5 0.2-0.9 5.9 3.9 0.8-16.8 5.3 3.8 1.4-11.7 6.6 4.8 2.5-20.5 5.7kc 4.5 1.9-14.4 I.Id 0.6 0.04-4.5 3.5b,f 3.1 1.2-8.7 0.7d 0.1 0.01-3.1 5.6 4.2 1.2-21.3 2.3ab,c,e,g 2.2 0.2-5.7 2.2 0.3 0.03-18.3 0.5d,f 0.1 0.02-2.7 7.4c 2.1 0.8-52.3 (a-g)comparisons foreach longestjob categoryusingstudent'ts,p < .05 Secretary' (N = 6) 5.6cd 5.8 2.2-8.3 2.6ac,g 2.0 1.6-4.4 6.0 5.5 3.2-9.7 3.7 3.9 2.2-5.8 0.2d 0.2 0.05-0.5 2.6 2.7 0.9-3.7 Supe m (N 3.0 5: 0.7 5 3 1.3 6 3 2.5 I 1 0.02 9 0. 0.02 Table 12. Mean, range,geometricmean and 95% confidenceintervaolf geometricmean of serum fluorocherr, random sample(N = 126)ofchemicalemployeeswho currentloynlywork incertaibnuildings(asI PFOS Mean Range Bldg.I (N= 15) 0.768 0.109- 2.190 Bldgs.2/49 (N=7) 2.621 0.325- 6.840 Bldg.3 (N=22) 2.457 0.885- 7.260 Bldg.4MX (N=21) 0.607 0.23-1.620 Bldg.4N (N=5) 2.000 1.440-2670 G. Mean 95% C.I. PFHS Mean Range 0.481 0.270- 0.858 0.228 0.013- 0.713 1.945 0.824- 4.592 0.996 0.083- 1.880 2.135 1.693- 2.693' 0.568 0.151- 1.860 0.525 0.364- 0.757 0.101 0.028- 0.374 1.937 1.378- 2.723 0.450 0.161- 0.832 G. Mean 95% C.I. POAA Mean Range G. Mean 95% C.I. PFOSAA Mean Range 0.115 0.057- 0.234 0.554 0.051- 2.700 0.240 0.119- 0.487 0.013 0.001- 0.054 0.709 0.262- 1.922 1.879 0.126- 3.640 1.251 0.412- 3.802 0.005 0.001- 0.016 0.410 0.286- 0.586 2.777 0.261- 6.760 2.128 1.467- 3.087 0.055 0.003- 0.269 0.069 0.039- 0.121 1.303 0.450-2.110 1.188 0.865- 1.631 0.019 0.002- 0.038 0.397 0.192- 0.821 2.088 1.300- 2.860 2.024 1.420- 2.886 0.019 0.004- 0.043 G. Mean 95% C.I. M570 Mean Range 0.004 0.002- 0.010 0.101 0.015- 0.410 0.003 0.001- 0.008 0.028 0.010- 0.056 0.023 0.012- 0.043 0.382 0.063- 0.992 0.014 0.007- 0.026 0.037 0.025- 0.053 0.011 0.003- 0.048 0.099 0.054- 0.205 G. Mean 95% C.I. 0.071 0.044- 0.115 0.023 0.013- 0.043 0.308 0.224- 0.423 0.036 0.031- 0.043 0.088 0.047- 0.165 Table12. (continued) PFOSA Mean Range G. Mean 95% C.I. M556 Mean Range G. Mean 95% C.I. Yearsinchemical Mean 0.026 0.005- 0.161 0.009 0.004- 0.022 0.027 0.003- 0.127 0.017 0.010- 0.030 20 0.003 0.005- 0.106 0.002 0.001- 0.005 0.012 0.003- 0.028 0.009 0.004- 0.021 24 0.102 0.003- 0.569 0.036 0.018- 0.072 0.119 0.024- 0.380 0.092 0.066- 0.129 10 0.042 0.012- 0.204 0.027 0.015-0.047 0.008 0.016- 0.018 0.007 0.005- 0.011 5 0.251 0.003- 0.487 0.059 0.002- 1.595 0.076 0.026- 0.175 0.056 0.019- 0.163 13 3M EPI-0006 Page 60 of 85 Table 13. Mean, range,geometricmean and 95% confidenceintervalof geometric mean of senun fluorochemicalsforthoseemployees inrandom swnple (N = 126) who saidtheyhave only worked inone building/area PFOS Mean Range Bldg.I (N=6) 0.474 0.129- 1.700 Bldg. 3 (N=7) 2.561 1.450- 5.120 Bldg.4MX (N=8) 0.521 0.230- 0.838 G. Mean 95% C.I. PFHS Mean Range 0.302 0.114- 0.797 0.117 0.013- 0.420 2.293 1.453- 3.619 0.835 0.151- 1.860 0.554 0.340- 0.904 0.063 0.038- 0.152 G. Mean 95% C.I. POAA Mean Range 0.064 0.018- 0.223 0.164 0.053- 0.386 0.519 0.185- 1.450 3.021 0.366- 6.760 0.064 0.039- 0.103 1.082 0.450- 1.850 G. Mean 95% C.I. PFOSAA Mean Range 0.125 0.053- 0.294 0.001 0.001- 0.003 2.033 0.773- 5.351 0.030 0.005-0.118 1.030 0.719- 1.476 0.020 0.008- 0.037 G. Mean 95% C.I. M570 Mean Range 0.001 0.001- 0.002 0.082 0.015 - 0.201 0.016 0.005- 0.047 0.318 0.063- 0.480 0.015 0.008- 0.027 0.040 0.026- 0.053 G. Mean 95% C.I. PFOSA Mean Range 0.053 0.018- 0.159 0.023 0.009- 0.060 0.274 0.145- 0.520 0.158 0.003- 0.569 0.048 0.028 - 0.081 0.043 0.001- 0.204 G. Mean 95% C.I. M556 Mean Range 0.019 0.009- 0.037 0.022 0.003- 0.585 0.055 0.009- 0.324 0.097 0.033- 0.213 0.034 0.011- 0.108 0.010 0.004- 0.019 G. Mean 95% C.I. Years inchemical Mean 0.014 0.004- 0.045 23 0.079 0.042- 0.150 15 0.013 0.005- 0.034 1.6 Table 14. Age, BMI and yearsworked inchemicalamong currentjobcategorieosfallparticipan(tNs = 187)inc Age Mean SE Median Range BNE Mean SE Median Range YearsWorked In Chemical Mean SE Median Range Cell Operator' (N= 9) 45e 2.1 44 38-59 25.8 1.3 27.6 18-30 Chemical Operatorb (N= 64) 4 1e-gh 1.2 41 25-62 28.3e 0.7 27.8 19-47 Engineer/ Labc (N= 37) Maintenanced (N= 17) 42e-g-h 1.8 43 23-58 4le-g" 1.5 42 27-52 27.3 0.9 27.3 16-38 27.6 0.8 26.6 23-33 Mill Operator' (N= 24) Secretaryf (N= 5) 34ab.c,clfgji 1.3 32 25-45 44e 1.3 44 42-49 28.1* 1.1 26.9 20-43 22.7bc,g 0.8 23.0 21-25 Igb,de 3.2 21 2-30 I Oac,eg 1.1 10 1-31 15b,e,g 2.2 14 1-37 Ioa,c,g 2.6 3 1-30 28-bc.d@fgh 0.8 1 1-21 16e 3.9 20 2-25 (a-h) comparisonsforeachcurrentjob categoryusingstudenfst,p <.05 Table 15. Number (andpercentageo)f demographiccharacteristaimcosng currentjobcategorieosf allparticipants Cell Operator (N=9) Gender* Female Male 1(11) 8(89) Only Worked in Chemical* Yes No 5(55) 4(45) Hand toMouth Contact Yes No 7(78) 2(22) Wash Hands Always Less frequently 8(89) 1(11) Chemical Operator (N=64) 12(19) 52(81) 30(47) 34(53) 48(77) 14(23) 53(85) 9(15) Engineer/ Lab (N=37) Maintenance (N=17) Mill Operator (N=24) 9(24) 28(76) 0(0) 17(100) 3(13) 21(87) 22(60) 15(40) 10(59) 7(41) 22(92) 2(8) 16(44) 20(56) 23(64) 13(36) 13(76) 4(24) 15(88) 2(12) 21(88) 3(13) 20(83) 4(17) s Secretary (N=5) 5(100) 0(0) 1(20) 4(80) 3(60) 2(40) 3(60) 2(40) Significantdliyfferen(tp<.05)proportionbsetweenjob categoriecsh,isquaretest Table 16. Mean, range,geometricmean and 95% confidenceintervaolfgeometricmean of serum fluorochemicaleve among allparticipan(tNs = 187)ofchemical PFOS Mean Range G. Mean 95% C.I. PFHS Mean Range G. Mean 95% C.I. POAA Mean Range G. Mean 95% C.I. PFOSAA Mean Range G. Mean 95% C.I. M570 Mean Range G. Mean 95% C.I. Cell Operator' (N = 9). 2.266 0.325- 6.840 1.643c,d,' 0.914- 2.953 0.927 0.083- 1.880 0.664b,c,d@e,lg 0.346- 1.274 1.811 0.126- 3.640 1.263c,f9 0.619 - 2.576 0.009 0.001- 0.051 0.003b,e 0.001- 0.008 0.045 0.012- 0.169 0.032'-d 0.020- 0.054 Chemical Operatorb (N = 64) Engineer/ Labc (N = 37) Maintenance@ (N = 17) Mill Operatoe (N= 24) Secretaryf (N = 5) 1.839 0.189- 7.260 0.611 0.081- 1.740 1.480c,e,',' 0.390a-b-dg@h 1.253- 1.749 0.282- 0.540 1.772 0.291- 4.940 1.320c-c-f 0.904- 1.928 0.625 0.025- 2.040 0.50,ab.,,.o 0.374- 0.672 0.451 0.220- 1.140 0.366al.d,h 0.206- 0.649 0.411 0.034- 1.860 0.292acael 0.239- 0.357 0.148 0.005- 0.905 0.309 0.023- 1.240 0.075a-b-d-g-' 0. 1 781,d-c 0.0500 - 0.111 0.107- 0.297 0.082 0.001-0.374 0.0532,buh 0.035- 0.080 0.076 0.027- 0.172 0.063ab 0.035 - 0.113 2.196 0.150 - 6.760 0.343 0.024 - 2.320 1.512 0.211- 4.680 1.849c,'@e,'g 0.198a-b-d'e-g" 1.052c.fg 1.587- 2.154 0.141 - 0.277 0.691- 1.602 1.218 0.015- 2.340 0.987c.fg 0.668- 1.457 0.167 0.095 - 0.261 0.155abd,e,g,l 0.106- 0.226 0.042 0.001- 0.269 0.014ac.d,g 0.009- 0.020 0.010 0.001- 0.073 0.004b,e 0.002 - 0.005 0.037 0.001- 0.125 0.015b 0.007 - 0.032 0.022 0.003- 0.049 0.016ac 0.012- 0.023 0.002 0.001- 0.004 0.001 0.001- 0.002 0.284 0.009- 3.100 0.143ac,e,'g 0.107- 0.192 0.064 0.004- 0.410 0.041b.1 0.030- 0.056 0.237 0.038- 0.701 0.044 0.023- 0.115 0.186ac,e,lg 0.04lb,d 0.130- 0.267 0.036- 0.047 0.036 0.001 - 0.072 0.029b.d 0.015 - 0.057 Table16. (continued) PFOSA Mean Range G. Mean 95% C.I. 0.006 0.0005- 0.015 0.004 0.002-0.009 0.088 0.001- 0.487 0.02I,C,f 0.013- 0.033 0.011 0.066 0.001- 0.063 0.001- 0.569 0.003',d,e 0.012c 0.002- 0.005 0.005- 0.027 0.028 0.0005- 0.204 0.0 17ac 0.011- 0.025 M556 Mean Range G. Mean 95% C.I. 0.013 0.003- 0.035 0.0 1ol.d 0.006- 0.017 0.074 0.024 0.001- 0.380 0.002- 0.033 0.0458,c-c-19 0.0 1obd 0.034- 0.059 0.007- 0.014 0.088 0.008- 0.406 0.014 0.002- 0.045 0.054ac-e-19 0.0 11 b.1 0.034- 0.088 0.008- 0.014 (a-h)comparisonforeachcuffentjob categoryusingstudent'ts 0.025 0.003- 0.080 0.014 0.004- 0.042 0.015 0.003- 0.030 0.01 lb.d 0.004- 0.027 Tablel7. Age, BMI and yearsworked inchemicalby longesjtob categorieosfallparticipan(tNs 187)inchemic Age Mean SE Median Range BMI Allnegs Mean SE Median Range YearsWorked In Chemical Mean SE Median Range Cell Operator' (N = 3) Chemical Operator' (N = 77) Engineer/ Labc (N = 31) Maintenanced (N= 20) 43 2.5 45 38-46 42egJ' 1.0 44 25-62 41 e,g,h 2.0 41 23-58 43e,g 1.5 43 27-54 24.8 3.4 27.0 18-29 28.0 0.6 27.5 18-47 28.0 0.9 27.7 21-38 27.7 0.7 26.7 23-33 13 6.7 11 2-25 13e,g 1.2 12 1-31 13p'g 2.3 8 1-37 IItfg 2.6 6 1-36 (a-h)comparisonforeach longestjob categoryusingstudent'ts,p < .05 Mill Operator' (N = 26) 34b,cafgh 1.4 32 25-51 28.1 1.1 26.9 20-43 2bcdfg 0.8 1 1-21 Secretary (N= 7) 46" 1.7 44 42-54 24.1 1.2 23.4 21-30 20d,e 3.7 20 2-33 Table 18. Number (andpercent)ofdemographiccharacteristibcyslongestjobcategorieosf allparticipanetmploy Gender* Male Female Cell Operatoe (N=3) @L@ 3(100) 0(0) Chemical Operato? (N= 77) H@n 68(88) 9(12) Engineer/ Lab' (N= 31) Hin 23(74) 8(26) Maintenanced (N= 20) H@n 20(100.00) 0(0.00) Mill Operator' (N = 26) Hin 21(81) 5(19) Only Worked inChemical* Yes No 1(33) 2(67) 33(43) 44(57) 21(68) 10(32) 10(50) 10(50) 24(92) 2(8) Hand toMouth Contact* Yes No 3(100) 0(0) 56(75) 19(25) 14(45) 17(55) 14(70) 6(30) 22(85) 4(15) Wash Hands Always Less Frequently 2(67) 1(33) 63(84) 12(16) 20(65) 11(35) 17(85) 3(15) Significantdliyfferenptercentagesby longestjobcategoriesc,hisquare,p <.05 22(85) 4(15) Secretaryf (N= 7) nn 0(0.00) 7(100.00) 3(43) 4(57) 4(57) 3(43) 4(57) 3(43) Table 19. Mean, range,geometricmean and 95% confidenceintervaolfgeometricmean of serum fluorochemicals participan(tNs = 187)inchemical PFOS Mean Range G. Mean 95% C.I. PFHS Mean Range G. Mean 95% C.I. Cell Operatoe (N @ 3) Chemical operatorb (N= 77) Engineer/ Labc (N= 31) Maintenanced (N= 20) 1.298 0.700- 2.260 1.139c.' 0.571-2.274 2.008 0.093- 7.880 0.486 2.162 0.081- 1.740 0.291- 10.600 1.638r-e,'B- 0.313,b.4c.&h 1.385-1.936 0.224-0.436 1.450ce.Ig 0.988- 2.128 0.762 0.522 0.381- 1.200 0.009- 1.880 0,686'@',Le.fs 0.359c-(Le9@f 0.358 - 1.314 0.292 - 0.441 0.106 1.005- 0.420 0.061abd 0.041 - 0.091 0.340 0.023- 1.250 0.192&,bce.49 0.120 - 0.309 Mill Operator' (N= 26) 0.628 0.052-2.040 0.508b.,,.4h 0.386- 0.669 0.080 0.001- 0.374 0.052 ab.d 0.035 - 0.078 Secrew/ (N= 7) 0.370 0.129-1.140 0 0.291,b,,th 0.176- 0.480 0 0.067 0.027- 0.172 0 0.056ab.d 0.036 - 0.088 0 POAA Mean Range G. Mean 95% C.I. 0.578 0.470- 0.749 0.566b'c'4 0.428- 0.749 2.231 0.052-6.760 1.863&,cdtg 1.592-2.180 0.246 0.024 - 1.000 1.636 0.211-4.680 0. 170&,b,'%e,h 0.124 - 0.235 1.147bc,'g 0.780 - 1.686 1.203 0.015-2.340 0.985b-c'tg 0.686 - 1.413 0.138 0.053 - 0.261 0.12 ,,b.,Ie,h 0.080 - 0.184 PFOSAA Mean Range G. Mean 95% C.I. 0.019 0.001 - 0.051 0.006 0.001- 0.056 0.035 0.001 - 0.269 0.010 0.001 - 0.073 o.oiicf,9 0.008 - 0.016 0.004bde 0.002 - 0.006 0.033 0.001 - 0.125 0.014c'g 0.007 - 0.027 0.021 0.003 - 0.049 0.0 16%tg 0.012 -0.022 0.001 0.001-0.004 0.00 1bde 0.001 - 0.002 M570 Mean Range G. Mean 95% C.I. 0.064 0.001- 0.169 bd 0.027 0.004 - 0.164 0.259 0.009 - 3.100 0. 129a-r-e9-f 0.099 - 0.169 0.050 0.004 - 0.164 0.036bd 0.026 - 0.049 0.242 0.038 - 0.701 0.045 0.023 - 0.115 0. 186@,c,etg 0.042bd 0.131 - 0.263 0.037 - 0.048 0.031 0.010 - 0.072 0.026bd 0.015 - 0.042 Table19. (continued) PFOSA Mean Range G. Mean 95% C.I. 0.008 0.001 - 0.015 0.004 0.001- 0.031 0.079 0.001 - 0.487 0.017c4 0.011- 0.026 0.011 0.001- 0.063 0.003b.,Ie 0.002- 0.006 0.087 0.001- 0.612 0.012cs 0.005- 0.029 0.029 0.010- 0.204 0.0 17c,g 0.011-0.025 M556 Mean Range G. Mean 95% C.I. 0.013 0.003- 0.035 0.006b,dh 0.001- 0.034 0.074 0.001- 0.380 0.012 0.002- 0.040 0.042&-c-e-ts 0.008bdh 0.033- 0.055 0.006- 0.011 0.093 0.008- 0.406 0.014 0.002- 0.045 0.054a-c-c,'g 0.0 11 bd 0.034- 0.086 0.008- 0.014 (a-h)comparisonforeach longestjob categoryusingstudent'ts,p < .05 0.021 0.003- 0.080 0.0139 0.006- 0.028 0.012 0.003- 0.030 0.008bdh 0.004- 0.017 Table 20. Mean, range,geometricmean and 95% confidenceintervaolf geometricmean of serwn fluorochemica participan(tNs = 187)who currentlwyork inonlyone building/areianchemical PFOS Mean Range Bldg.I (N=23) 0.686 0.081-2.190 Bldgs.2/49 (N=12) 2.257 0.325- 6.840 Bldg.3 (N=30) 2.426 0.189- 7.260 Bldg.4MX (N=20) 0.556 0.052- 1.620 Bldg.4N (N=5) 2.000 1.440- 2.670 G. Mean 95% C.I. PFHS Mean Range 0.438 0.281- 0.681 0.194 0.013- 0.713 1.727 1.023- 2.916 0.870 0.076- 1.880 2.000 1.544- 2.590 0.567 0.139- 1.860 0.468 0.336- 0.653 0.079 0.001- 0.374 1.937 1.378- 2.723 0.450 0.161- 0.832 G. Mean 95% C.I. POAA Mean Range 0.101 0.059- 0.173 0.426 0.024- 2.700 0.585 0.292- 1.171 1.807 0.126- 3.640 0.417 0.369- 0.561 2.665 0.261- 6.760 0.050 0.030- 0.084 1.176 0.015- 2.110 0.397 0.192- 0.821 2.088 1.300- 2.860 G. Mean 95% C.I. PFOSAA Mean Range 0.200 0.119- 0.331 0.010 0.001- 0.054 1.247 0.631- 2.461 0.008 0.001- 0.051 2.111 1.587-.2.809 0.927 0.568- 1.515 0.049 0.002- 0.269 0.021 0.002- 0.049 2.024 1.420- 2.886 0.019 0.004- 0.043 G. Mean 95% C.I. M570 Mean Range 0.002 0.002- 0.006 0.077 0.006- 0.410 0.003 0.001- 0.007 0.043 0.010- 0.169 0.020 0.012- 0.034 0.481 0.063- 3.100 0.015 0.010- 0.024 0.040 0.024- 0.613 0.011 0.003- 0.048 0.099 0.054- 0.205 G. Mean 95% C.I. 0.050 0.033- 0.075 0.031 0.018- 0.052 0.348 0.260- 0.465 0.039 0.035- 0.044 0.088 0.047- 0.165 Table20. (continued) PFOSA Mean Range 0.018 0.0005- 0.161 G. Mean 95% C.I. M556 Mean Range 0.005 0.003- 0.010 0.021 0.003- 0.127 G. Mean 95% C.I. Years inchemical Mean 0.013 0.008- 0.020 20 0.006 0.0005- 0.021 0.003 0.002- 0.006 0.013 0.003- 0.347 0.010 0.006- 0.017 19 0.109 0.003- 0.569 0.037 0.020- 0.066 0.121 0.024- 0.380 0.095 0.073- 0.125 10 0.028 0.0005- 0.204 0.017 0.010-0.028 0.251 0.003- 0.487 0.059 0.002- 1.595 0.011 0.002- 0.030 0.010 0.007- 0.013 0.076 0.025- 0.175 0.057 0.019- 0.163 3 13 Table 21. Mean, range,geometricmean and 95% confidenceintervaolf geometricmean forallparticipant(sN 18 inone building/areianchemical PFOS Mean Range G. Mean 95% C.I. PFHS Mean Range G. Mean 95% C.I. POAA Mean Range G. Mean 95% C.I. PFOSAA Mean Range G. Mean 95% C.I. M570 Mean Range G. Mean Bldg.I (N = 11) 0.432 0.081- 1.700 0.282 0.150- 0.529 0.100 0.013-0.42 0.058 0.027-0.124 0.155 0.024- 0.386 0.115 0.064-0.201 0.003 0.001-0.011 0.002 0.001- 0.003 0.062 0.02-0.201 0.041 Bldg.3 (N = 8) 2.693 1.450- 5.120 2.427 1.619- 3.638 0.890 0.151- 1.860 0.578 0.236- 1.420 2.947 0.366- 6.760 2.079 0.923- 4.681 0.027 0.005- 0.118 0.014 0.005- 0.036 0.308 0.03-0.480 0.270 Bldg.4 MX (N=17) 0.501 0.052- 0.383 0.467 0.321- 0.681 0.053 0.01-.0.152 0.045 0.026-0.076 0.987 0.015- 1.850 0.817 0.467- 1.429 0.023 0.003- 0.049 0.017 0.011- 0.026 0.044 0.026- 0.061 0.045 95% C.I. Table 21. (continued) PFOSA Mean Range G. Mean 95% C.I. M556 Mean Range G. Mean 95% C.I. Years inchemical Mean 0.021- 0.080 0.158- 0.462 0.035- 0.057 0.013 0.0005 - 0.060 0.006 0.002- 0.016 0.017 0.04-0.059 0.012 0.006- 0.023 20 0.151 0.003- 0.569 0.059 0.013- 0.266 0.095 0.05-0.214 0.080 0.047- 0.136 17 0.031 0.0005- 0.204 0.018 0.009- 0.036 0.013 0.06-0.022 0.014 0.009-0.021 1.3 Table 22. Mean, median and range,by gender,ofchemicaloperatorasnd engineer/lafborrandom sample by ag chemical Age Mean Median Range BMI Mean Median Range Years Worked In Chemical Mean Median Range Chemical Operators Female Male 41.3 40 28-61 42.4 44 25-62 26.9 27.4 22.0-32.7 28.6 28.4 20.2-47.5 9.6 10.5 3-20 11.4 10.0 1-31 Female Engineer/L 36.0 35 23-54 24.0 22.2 20.7-30.2 9.6 1.3 1-33 Table23. Mean,rangeg,eometrmiecanand95% confidenicnetervoafgleometrmiecanofserumfluorochemic chemicalplantemployees whose currenjtob was chemicaloperatoror engineer/lab PFOS Mean Range Chemical Operators Female (N = 10) Male (N = 37) 1.183 0.471- 2.380 1.943 0.490- 7.260 Engineer/La Female(N = 6) 0.167 0.101-0.281 G. Mean 95% C.I. PFHS Mean Range 1.051 0.764- 1.447 0.190* 0.071- 0.404 1.625 1.345- 1.962 0.492 0.134- 1.860 0.157* 0.115- 0.214 0.032 0.005- 0.081 G. Mean 95% C.I. POAA Mean Range 0.167* 0.119- 0.235 1.342* 0.150-2.110 0.364 0.285- 0.465 2.497 0.745- 6.760 0.023* 0.011- 0.048 0.103 0.035- 0.300 G. Mean 95% C.I. PFOSAA Mean Range 0.129* 0.701- 1.818 0.012 0.001- 0.109 2.168 1.829- 2.570 0.041 0.001- 0.269 0.078* 0.043- 0.141 0.004 0.001- 0.018 G. Mean 95% C.I. M570 Mean Range 0.006 0.002- 0.016 0.114 0.029- 0.357 0.013 0.008- 0.022 0.260 0.009- 0.992 0.002 0.001- 0.005 0.046 0.020- 0.093 G. Mean 95% C.I. 0.087 0.055- 0.139 0.146 0.099- 0.217 0.039 0.024- 0.064 Table23. (continued) PFOSA Mean Range G. Mean 95% C.I. M556 Mean Range G. Mean 95% C.I. 0.063 0.003- 0.315 0.022 0.007- 0.062 0.041 0.007- 0.118 0.027 0.014- 0.050 p < 0.05 0.109 0.001- 0.487 0.030 0.016- 0.055 0.083 0.001- 0.380 0.050 0.035- 0.073 0.017 0.0005- 0.063 0.005 0.001- 0.024 0.010 0.003- 0.027 0.007 0.004- 0.014 3M EPI-0006 Page 76 of 85 Table 24. Distributioonffilmplantparticipantrsa:ndom sample,volunteerasnd allparticipants Have worked onlyinfilmplant (Have worked on D- Imaker) (Have notworked on D- I maker) Work infilmplant withpreviouswork in chemical Random Sample 42 (6) (36) 18 Film Plant Volunteers 14 (1) (13) All Participants 56 (7) (49) 2 20 Total 60 16 76 3M EPI-0006 Page 77 of 85 Table 25. Demographic characteristoifcrsandom sample(N 60)offilmplantemployees including subsets:employees with only film plantexperience;employees known to have worked on D- I Maker; and employees with priorchemical history Age Mean SE Median Range BNH Mean SE Median Range Years worked In film Mean SE Median Range Gender Female Male Currentjob Engineer/Lab Film processor Maintenance Administrative Longest job Engineer/Lab Film processor Maintenance Administrative Hand to mouth contact Yes No Wash hands Yes No All (N = 60) 46 1.1 47 23-59 28.0 0.6 27.8 18.0-41.8 13.7 10.0 14 0.1-36.0 11 (18) 49 (82) 16 (27) 23 (38) 10 (17) 11 (18) 13 (22) 26 (43) 11 (18) 10 (17) 37 (62) 23 (38) 50 (83) 10 (17) Only Film (N = 36) 44 1.5 46 23-59 28.2 0.8 27.8 18.0-41.8 13.7 1.7 14 0.1-29 6 (17) 30 (83) 10 (28) 12 (33) 7 (19) 7 (19) 7 (19) 15 (42) 8 (22) 6 (17) 26 (72) 10 (28) 28 (78) 8 (22) D- I Maker (N = 6) 46 3.6 48 30-55 26.9 1.9 27.5 21.7-31.7 Film w/ history of chemical (N = 18) 48 2.1 51 28-58 28.0 1.1 27.6 20.0 - 37.9' 9.2 4.1 6 1-21 1 (17) 5 (83) 0 (0) 5 (83) 1 (17) 0 (0) 0 (0) 5 (83) 1 (17) 0 (0) 4 (67) 2 (33) 6 (100) 0 (0) 15.4 2.4 16 1-36 4 (22) 14 (78) 6 (33) 6 (33) 2 (11) 4 (22) 6 (33) 6 (33) 2 (11) 4 (22) 7 (39) 11 (61) 16 (89) 2 (11) 3M EPT-0006 Page 78 of85 Table 26. Mean, range,geometricmean and 95% confidenceintervaolf geometricmean for random sample offilmplantemployeesby work historyo:nly film,D-1 Maker orfilmwithpriorchemicalwork history Only Film' (N = 35) D- I Makerb (N = 6) Film withprevious historyin chemical' (N = 18) PFOS Mean Range G. Mean 95% C.I. 0.122 0.032- 0.250 0. 11 obc 0.094- 0.129 0.367 0.122- 0.946 0.289a 0.159- 0.527 0.212 0.080- 0.692 0.178a 0.137- 0.233 PFHS Mean Range 0.015 0.001 - 0.075 0.023 0.005- 0.030 0.038 0.007- 0.210 G. Mean 95% C.I. 0.010C 0.008- 0.014 0.020 0.011-0.034 0.023a 0.015- 0.036 POAA Mean Range 0.052 0.006- 0.298 0.122 0.020- 0.197 0.090 0.012- 0.246 G. Mean 95% C.I. 0.03@b,c 0.028- 0.049 0.093a 0.044- 0.196 0.067a 0.044- 0.100 PFOSAA Mean Range G. Mean 95% C.I. 0.003 0.001- 0.009 0.002 0.002- 0.003 0.006 0.001- 0.022 . 0.004 0.022- 0.009 0.005 0.001- 0.038 0.003 0.002- 0.005 M570 Mean Range 0.022 0.0008- 0.454 0.018 0.0021- 0.053 0.018 0.0014- 0.069 G. Mean 95% C.I. 0.007 0.005- 0.010 0.010 0.006- 0.017 0.010 0.004- 0.026 M556 Mean Range 0.022 0.0001- 0.307 0.005 0.001- 0.014 All values< LOQ G. Mean 95% C.I. 0.003 0.001- 0.006 0.003 0.002- 0.004 (a-c)comparisonforeach currentjob categoryusingstudent'ts,p < .05 APPerldix D Page Ai)vendixD Distributionof Fluorocheniicalsand Their Natural Log Transformation Among Film PlantEmployees (N 60) in theRandom Sample 3M EPI-0006 Page 79 of 85 Table 27. Ratio of fluorochemicallevelsby random sample of film employees including subsets:employees onlywithfilmplantexperience;employees known to have worked on D- IMaker; and employees withpriorchemicalhistory PFOS/PFHS Mean Median Range PFOS/POAA Mean Median Range PFOS/Analytes Mean Median Range PFOSAA/M556 Mean Median Range M570/M556 Mean Median Range Only Film (N = 36) 14.9 10.4 1.8-107.6 3.3 2.8 0.7-9.2 10.0 7.8 0.2-37.6 1.9 1.0 0.003-14.0 5.0 2.3 0.3-45.0 D-1 Maker (N = 6) 18.8 12.7 5.0-46.6 5.7 2.4 0.9-21.0 25.6 11.5 2.1-91.8 2.8 1.3 0.3-10.9 6.9 3.4 0.8-28.2 Film With Previous HistoryIn Chemical (N = 18) 9.3 7.4 3.3-32.0 3.2 2.3 1.2-10.1 12.6 10.3 3.0-40.7 2.1 1.2 0.4-15.1 7.1 4.5 0.6-27.6 3M EPI-0006 Page 80 of 85 Table 28. Demographic characteristicosf random sample of film plant employees by currentjob categorieswho have worked only inthefilmplant(i.e.n,oton theD- I Maker or prior work inchemical) Age Mean SE Median Range BMI Mean SE Median Range Years worked In film Mean SE Median Range Gender Female Male Hand to mouth Contact Yes No Wash hands Yes No Engineer/Lab (N = 10) 46 2.8 48 23-58 26.8 1.5 27.3 21.6-31.7 14.8 2.9 15 0.1-29 2 (20) 8 (80) 8 (80) 2 (20) 8 (80) 2 (20) Film Processor (N = 12) 44 2.5 47 27-59 28.6 1.4 27.8 18.0-41.8 Maintenance (N = 7) 40 3.3 40 31-51 Administrative (N = 7) 48 3.3 50 40-55 28.7 1.8 29.5 24.1-32.9 29.2 1.8 27.9 24.4-41.8 14.1 2.6 17 0.5-29 2 (17) 10 (83) 4.6 3.4 3 0.5-12 0 (0) 7 (100) 20.4 3.4 25 5-28 2 (29) 5 (71) 10 (83) 2 (17) 10 (83) 2 (17) 4 (57) 3 (43) 6 (86) 1 (14) 4 (57) 3 (43) 4 (57) 3 (43) 3M EPI-0006 Page 81 of 85 Table 29. Mean, range,geometric mean and 95% confidenceintervalof geometric mean of serum fluorochemicalsforrandom sample of employees who have only worked inthefilmplant(i.e.n,oton theD- I Maker orpriorwork inchemical) PFOS Mean Range Engineer/Lab' (N = 10) 0.097 0.055 - 0.140 G. Mean 95% C.I. 0.0930 0.074.-0.116 PFHS Mean Range 0.016 0.001- 0.075 G. Mean 95% C.I. 0.009 0.005 - 0.018 POAA Mean Range 0.030 0.006 - 0.055 G. Mean 95% C.I. 0.022c 0.014 - 0.036 PFOSAA. Mean Range 0.002 0.001 - 0.005 G. Mean 95% C.I. 0.002 0.001 - 0.003 M570 Mean Range 0.006 0.002 - 0.017 G. Mean 95% C.I. 0.005C 0.004 - 0.007 M556 Mean Range 0.002 0.0001 - 0.003 G. Mean 95% C.I. o.ooib 0.001 - 0.003 Film Processorb (N = 12) 0.127 0.032- 0.250 0.106 0.074- 0.154 0.015 0.004- 0.047 0.011 0.007- 0.017 0.055 0.007 - 0.154 0.041 0.024 - 0.068 0.002 0.001 - 0.009 0.002 0.001- 0.003 0.048 0.003 - 0.454 0.010 0.004- 0.022 0.029 0.003 - 0.307 0.00a5 0.002 - 0.011 Maintenance' (N = 7) 0.159 0.137- 0.216 0.157a 0.139 - 0.177 0.016 0.001- 0.034 0.011 0.005- 0.026 0.098 0.021- 0.298 0.071a 0.038- 0.132 0.002 0.001- 0.006 0.002 0.001- 0.003 0.018 0.006- 0.046 0.014 ,d 0.009 - 0.024 0.005 0.001 - 0.016 0.004 0.002- 0.007 Administrativde (N = 7) 0.111 0.054 - 0.166 0.104 0.077 - 0.140 0.012 0.006- 0.033 0.010 0.006- 0.016 0.039 0.017- 0.063 0.035 0.024- 0.051 0.004 0.001- 0.006 0.004 0.002- 0.006 0.005 0.001- 0.009 0.004' 0.002 - 0.007 0.002 0.001 - 0.003 0.002 0.002 - 0.003 (a-d)comparisons foreach currentjob categoryusingstudent'st 3M EPI-0006 Page 82 of 85 Table30. Ratioof fluorochemicalevelsby currentjob among random sample of film employees who only have worked infilmand not on the D- I Maker PFOS/PFHS Mean Median Range PFOS/POAA Mean Median Range PFOS/Analytes Mean Median Range PFOSAA/M556 Mean Median Range M570/M556 Mean Median Range Engineer/Lab (N 10) Film Processor (N = 12) Maintenance (N = 7) 13.0 7.5 1.8-61.6 13.0 12.8 4.9-29.0 24.6 12.2 4.3-107.6 Administrative (N = 7) 11.1 10.4 5.1-16.5 4.0 3.0 1.3-9.2 3.2 3.2 1.2-6.3 2.7 2.1 0'.7-6.7 3.0 2.8 2.2-4.2 10.4 10.5 2.1-17.8 10.0 4.3 0.2-31.2 7.1 7.0 4.0-11.6 12.4 9.5 5.2-37.6 3.6 1.1 0.5-14.0 0.7 0.5 0.003-1.5 1.3 0.5 0.1-6.1 2.1 2.3 0.5-4.4 8.6 2.5 1.0-45.0 3.1 1.6 0.6-18.4 5.6 5.9 0.4-10.5 2.6 2.8 0.3-5.5 3M EPI-0006 Page 83 of 85 Table 31. Demographic characteristicsof allfilm plant participants(N = 76) by only film plant, D- I Maker or film plantwith previoushistoryin chemical Age Mean SE Median Range Bmi Mean SE Median Range Years worked In film Mean SE Median Range Gender Female Male Currentjob EngineerALab Film processor Maintenance Administrative Longestjob Engineer/Lab Film processor Maintenance Administrative Hand to mouth contact Yes No Wash hands Yes No All (N = 76) 45 1.0 47 '-)-359 28.3 0.5 27.9 18.0-41.8 Only Film (N = 49) 44 3.6 45 23-59 28.5 0.6 27.9 18.0-41.8 14.4 1.2 16.0 0.1-36.0 15.2 1.4 17.0 0.1-30 16 (21) 60 (79) 8 (16) 41 (84) 18 (24) 34 (45) 11 (14) 13 (17) 12 (25) 20 (41) 8 (16) 9 (18) 14 (18) 38 (50) 12 (16) 12 (16) 8 (16) 24 (49) 9 (18) 8 (16) 49 (64) 27 (36) 65 (86) 11 (14) 36 (73) 13 (27) 40 (82) 9 (18) D- I Maker (N = 7) 44 1.2 47 30-55 26.6 1.5 26.5 21.7-31.7 Film w/ history of chemical (N = 20) 47 2.1 51 28-58 28.5 1.0 28.0 20.0-37.9 8.1 3.5 2.0 1-21 2 (29) 5 (71) 0 (0) 6 (86) 1 (14) 0 (0) 0 (0) 6 (86) 1 (14) 0 (0) 14.6 2.5 15.0 1-36 6 (30) 14 (70) 6 (30) 8 (40) 2 (10) 4 (20) 6 (30) 8 (40) 2 (10) 4 (20) 5 (71) 2 (29) 7 (100) 0 (0) 8 (40) 12 (60) 18 (90) 2 (10) 3M EPI-0006 Page 84 of 85 Table 32. Mean, range,geometricmean and 95% confidenceintervaolf geometricmean of serum fluorochemicalfsorallfilmplantparticipaenmtployeesby work historyo:nlyfilmplant,D- I Maker orfilmplantwithprevioushistoriynchemical PFOS Mean Range G. Mean 95% C.I. PFHS Mean Range Only Film' (N = 49) 0.129 0.032- 0.264 0.1 16bc 0.101- 0.133 0.016 0.001- 0.075 D-I Maker b (N = 7) 0.347 0.122- 0.946 0.279a 0.168- 0.461 0.022 0.005- 0.030 Film withprevious historyin chemicalc (N = 20) 0.220 0.080- 0.692 0.185a 0.144- 0.238 0.038 0.007- 0.210 G. Mean 95% C.I. 0.0lic 0.009- 0.014 0.019 0.012- 0.030 0.024a 0.016- 0.036 POAA Mean Range 0.057 0.006- 0.298 0.146 0.020- 0.290 0.146 0.012- 1.220 G. Mean 95% C.I. PFOSAA Mean Range 0.040b,c 0.031- 0.051 0.003 0.001- 0.020 0.109a 0.054- 0.221 0.006 0.001- 0.022 0.078a 0.049- 0.124 0.006 0.001- 0.038 G. Mean 95% C.I. M570 Mean Range 0.004 0.002- 0.003 0.018 0.001- 0.454 0.004 0.002- 0.009 0.0390.002- 0.164 0.003 0.002- 0.005 0.017 0.001- 0.069 G. Mean 95% C.I. 0.007 0.005- 0.009 0.015 0.005- 0.046 0.010 0.006- 0.016 M556 Mean Range 0.009 0.0001- 0.307 0.006 0.001- 0.015 Allvalues< LOQ G. Mean 95% C.I. 0.003 0.002- 0.004 0.004 0.002- 0.008 (a-c)comparison foreach currentjob categoryusingstudent'ts,p < .05 3M EPI-0006 Page 85 of 85 Table 33. Mean, range,geometric mean and 95% confidence intervalof geometric mean of serum fluorochemicalsforallfilmplantparticipanetmployees who only worked infilmplant (i.e.n,ot on theD- I Maker or worked previouslyinchemical) PFOS Mean Range G. Mean 95% C.I. PFHS Mean Range Engineer/Lab' (N = 12) 0.108 0.055 - 0.170 0.102c 0.082- 0.127 0.018 0.001- 0.075 Film Processorb (N = 20) 0.133 0.032- 0.264. 0.114 0.088- 0.148 0.016 0.004- 0.052 Maintenancec (N = 8) 0.168 0.137 - 0.237 0.165ad 0.143 - 0.191 Administrativde (N = 9) 0.108 0.054 - 0.166 0.103c 0.081- 0.129 0.016 0.001 - 0.034 0.012 0.006 - 0.033 G. Mean 95% C.I. POAA Mean Range G. Mean 95% C.I. 0.011 0.006 -0.012 0.049 0.006 - 0.188 0.031" 0.017- 0.054 0.012 0.009- 0.017 0.055 0.007- 0.154 0.040 0.027- 0.060 0.011 0.006- 0.023 0.095 0.021- 0.298 0.072a 0.042- 0.124 0.010 0.007- 0.015 0.037 0.017 - 0.063 0.033 0.025- 0.046 PFOSAA Mean Range 0.002 0.001 - 0.005 0.005 0.001 - 0.020 0.004 0.001- 0.017 0.004 0.001- 0.006 G. Mean 95% C.I. 0.002 0.001 - 0.003 0.003 0.002- 0.005 0.002 0.001- 0.005 0.003 0.002 - 0.005 M570 Mean Range 0.006 0.002 - 0.017 0.031 0.002- 0.454 0.017 0.006- 0.046 0.005 0.001- 0.009 G. Mean 95% C.I. 0.005 0.003 - 0.007 0.008 0.005- 0.013 0.014 0.009- 0.022 0.004 0.002 - 0.006 M556 Mean Range 0.002 0.0001 - 0.007 0.019 0.001- 0.307 0.005 0.001- 0.016 0.003 0.001- 0.006 G. Mean 95% C.I. o.ooil 0.001 - 0.003 0.004a 0.002 - 0.006 0.014 0.002- 0.007 0.002 0.002 - 0.003 (a-c)comparisons foreach currentjob categoryusingstudent'st,p < .05 Appendix A Decatur Plant Maps i-- APpendix A Page I 3M Decatur,Alabama Plant Layout Revised 9/16/97 toti%,er Water pwnp hoLLse ......... ........... ........ Pic area ........... toFiriky IslandRoad bonr yard area ... 59 13 .20 oilEl:.Ell 12 stied contractor gate and puking to Finley IslandRoad 7@ @,7 dhi p@wking-ii 41611 41 yne61ft 3M 22 IZI ... corridor ri E]39 E] 080 o8 --- - - - - - - - - - - - - R 000 P,@ 37 i,pi paiking L. ............. idi........ [3101,*: ............... 0 o: OOL DyYUn 0 =3 56 Parking @(D p 4v Avenue B ............ 21 njam N Nebo Lane fjr(: traihinf, 43 E3 ...... scale I inch 4 70 feet .09 m Hes to State Docks Road El (D (2)6 57 ... 3M Decatur Chemical Plant ----------------------------------------------------------------------------- 41 3M Chemical :i.:Xi: Dyneon 56ED 0/42 @,x xvx x@ X, @x, I ...... ........ .. ............ ..................................................................... ............................................................ .... . puking ........... ................... ------------------------------------------------------------- . .. . .. . parking Dyneon rA rn ,a Appendix B Study Questionnaire APPendix B Page I DECATUR EMPLOYEE QUESTIONNAIRE AppendixB Page 2 Thank you forparticipatintghisresearcshtudy.Pleaserespondtoeachquestiownitheithear shortansweror an 'x'itnheappropriatbeox. NAME EMPLOYEE NUMBER 1. Have you everworked inthe ChemicalPlant? Yes E:l No [I Ifno,pleasego toquestion2 If'yes' a. How many yearshave-youworked inthe chemicalplant?Years= b. What yeardidyou startworkinginthechemicalplant? Year= 2. PleaseindicatIefyou have ever worked inthefollowinagreas.Mark an 'x'inallboxes thatapplytoyou. I--] Building1 [1 Building3s8 and/or51 El Building2s and/or49 0 Buildin4g2 (PackaginFgC inerts) El Buildin3g(OSCUOSF area) Building61 Buildin3g (besidesOSCUOSF area) FilmPlant(albluildings) El Buildin4g North El Buildin4g millroom/extruder Cl Building17 wastewatertreatmentplant (Building3s6 and 57) Other (Pleasespecify) 3. Thinkingaboutthejobthatyou worked forthelona6stperiodoftimewhileemployed at3M Decatur,please answer thefollowinqguestions. a. Jobtitle: b. When didyouworkthereF:rom-(year) to c. Averagenumber ofhoursperweek on thisjob? Hours= (year) d. When you worked overtimew,hatwas yourusualjobassignment? 4. Pleaseanswerthefollowinqguestionsregardinygourcurre job. Currentplant: Currentjobtitle: Chemical C] Film [3 What yeardidyou startworkinginthiscurrenjtob:Year Other APPendix B Page 3 Average number of hours perweek on thisjob: Hours When you work overtime,what isyour usualjob assignment? 5. Please indicateinwhich area(s)you work inyour currentjob.Mark an @x'inallboxes thatapply to you. El Building1 C] Buildings38 and/or51 El Buildings2 and/or 49. Building42 (Packaging FC inerts) El Building3 (OSCUOSF area) El Building3 (besidesOSCUOSF area) Building61 FilmPlant(albluildings) Building4 North Building4 millroom/extruder Building17 Wastewater treatmentplant (Building3s6 and 57) Other (Pleasspeecify)- 6. While atwork,do you chew gum? always frequently0 sometimes rarely never 7. While atwork,do you chew tobacco? El a. always F-IfrequentlyEl sometimes rarely never 8. While atwork,do you smoke cigarettes? [I always [I frequently[I sometimes El rarely F-I never 9. How frequentldyo you wash your hands beforeeatingwhileatwork? Markonlyonebox. El always E] frequently[]sometimes F_] rarely Cl never 10. What isyour height? 11. What isyour weight Feet = Pounds Inches Apperidix C Page Appendix C Distributionof Fluorochemicals and Their NaturalLog Transformation Among Chen-@dcalEmployees(N 126) in theRandom Sample Chemical Plant Random Sample PFOS ppm 2 --Jl maximum quartile median quartile n-tinimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 10.600 10.600 7.187 3.132 1.925 1.140 0.440 0.215 0.102 0.091 0.091 Moments Mean Std Dev Std ErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 1.5047 1.6122 0.1436 1.7890 1.2204 126.0000 126.0000 Test forNormality Shapiro-WilkW Test w Prob<W 0.734399 0.0000 Appendix c Page 2 Chemical Plant Random Sample InPFOS ppm Appendix c Page 3 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 2.3609 2.3609 1.9720 1.1415 0.6549 0.1310 -0.8215 -1.5388 -2.1793 -2.3936 -2.3936 Moments Mean Std Dev Std ErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights -0.0605 1.0263 0.0914 0.1204 -0.2415 126.0000 126.0000 Test forNormality Shapiro-WilkW Test w Prob<W 0.967746 0.0521 Chemical Plant Random Sample PFHS ppm 1.5 C.D maximum quartile median quartiie minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 1.8800 1.8800 1.7865 0.8777 -0.4200 0.1700 0.0794 0.0334 0.0137 0.0054 0.0054 Moments Mean Std Dev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 0.3450 0.4117 0.0367 0.4176 0.2724 126.0000 126.0000 Test forNormality Shapiro-WilkW Test w Prob<W 0.729906 0.0000 Appendix c Page 4 Chemical Plant Random Sample InPFHS ppm 'I! ooln quo maximum quartile median quartfle minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 0.6313 0.6313 0.5802 -0.1307 -0.8675 -1.7720 -2.5461 -3.4007 4.3022 -5.2269 -5.2269 Moments Mean StdDev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights -1.7152 1.2225 0.1089 -1.4996 -1.9307 126.0000 126.0000 Test forNormality Shapiro-WilkW Test w Prob<W 0.975283 0.2302 Appendix C Page 5 Chemical Plant Random Sample POAAppm 7 Apperidix c Page 6 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 6.7600 6.7600 5.6618 3.4300 2.0725 1.3000 0.3860 0.1281 0.0514 0.0209 0.0209 Moments Mean Std Dev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 1.5363 1.3359 0.1190 1.7718 1.3007 126.0000 126.0000 Test forNormality Shapiro-WilkW Test w Prob<W 0.875366 <.0001 Chemical Plant Random Sample InPOAA ppm APPendix C Page 7 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 1.9110 1.9110 1.7302 1.2318 0.7288 0.2624 -0.9519 -2.0550 -2.9685 -3.8680 -3.8680 Moments Mean StdDev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights -0.1061 1.2545 0.1118 0.1151 -0.3273 126.0000 126.0000 TestforNormality Shapiro-WilkW Test w Prob<W 0.903769 <.0001 Chemical Plant Random Sample PFOSAAppm 0.25 0.05 L D.DD maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 0.26900 0.26900 0.14915 0.06331 0.02812 0.00808 0.00276 0.00112 0.00112 0.00112 0.00112 Moments Mean Std Dev Std ErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 0.0233 0.0396 0.0035 0.0303 0.0163 126.0000 126.0000 Test forNonnality Shapiro-WilkW Test w Prob<W 0.600789 0.0000 Appendix c Page 8 Chemical Plant Random Sample InPFOSAA ppm Appendix c Page 9 -7 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% -1.3130 -1.3130 -1.9045 -2.7609 -3.5721 4.8184 -5.8916 -6.7944 -6.7944 -6.7944 -6.7944 Moments Mean StdDev Std ErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights -4-7813 1.4592 0.1300 -4.5240 -5.0386 126.0000 126.0000 Test forNormality Shapiro-WilkW Test w Prob<W 0.929527 <.0001 ChemicalPlant Random Sample M570 ppm i.D 0.6 0.4 O.D L maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% Moments Mean StdDev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 0.99200 0.99200 0.69103 0.41570 0.19425 0.06685 0.03773 0.02173 0.00965 0.00840 0.00840 0.1505 0.1862 0.0166 0.1833 0.1176 126.0000 126.0000 TestforNormality Shapiro-WilkW Test w Prob<W 0.712853 0.0000 Appendix C Page 10 Chemical Plant Random Sample InM570 ppm D- Appendix c Page 11 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% -0-0080 -0.0080 -0.3701 -0.8780 -1.6387 -2.7053 -3.2774 -3.8310 -4.6406 -4.7795 4.7795 Moments Mean Std Dev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights -2.5145 1.1167 0.0995 -2.3176 -2.7114 126.0000 126.0000 Test forNormality Shapiro-WilkW Test w Prob<W 0.957094 0.0035 Chemical Plant Random Sample PFOSAppm 0.6 D.1 % 0.1 .0 L maximum quartile median quartile minimum Quantiles 100.0% 99.5% 973% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 0.61200 0.61200 0.47948 0.23180 0.05625 0.01195 0.00269 0.00122 0.00050 0.00050 0.00050 Moments Mean Std Dev Std ErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 0.0618 0.1165 0.0104 0.0823 0.0412 126.0000 126.0000 Test forNormality Shapiro-WilkW Test w Prob<W 0.580929 0.0000 Appendix c Page 12 Chemical Plant Random Sample InPFOSA ppm 0 Appendix c Page 13 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% -0.4910 -0.4910 -0.7357 -1.4620 -2.8787 -4.4277 -5.9166 -6.7081 -7.6009 -7.6009 -7.6009 Moments Mean Std Dev Std ErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights -4-3545 1.9010 0.1694 -4.0193 -4.6896 126.0000 126.0000 Test forNormality Shapiro-WilkW Test w Prob<W 0.946788 0.0002 Chemical Plant Random Sample M556 ppm 0.20 0.10 0.05 0.00 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 0.40600 0.40600 0.32165 0.15000 0.05995 0.02615 0.00765 0.00300 0.00175 0.00140 0.00140 Moments Mean StdDev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 0.0519 0.0737 0.0066 0.0649 0.0389 126.0000 126.0000 TestforNormality Shapiro-WilWk Test w Prob.@W 0.671484 0.0000 APpendix C Page 14 ChemicalPlant Random Sample InM556 ppm Appendix c Page 15 -7 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5ct 0.5% 0.0% -0.9014 -0.9014 -1.1396 -1.8976 -2.8151 -3.6443 -4.8731 -5.8091 -6.3487 -6.5713 -6.5713 Moments Mean Std Dev Std ErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights -3.7960 1.3638 0.1215 -3.5556 -4.0365 126.0000 126.0000 Test forNormality Shapiro-WilkW Test w Prob<W 0.962731 0.0158 FilmPlant Random Sample PFOS ppm 0.7-11 0.6 0.2 0.1 0.0 maximum quartile median quartile minimum Quanbles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 0.94600 0.94600 0.81265 0.27350 0.20825 0.13750 0.08698 0.06720 0.02393 0.01500 0.01500 Moments Mean Std Dev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 0.17181 0.14780 0.01908 0.20999 0.13363 60.00000 60.00000 TestforNormality Shapiro-WilkW Test w 0.682603 P<ro.b0<0W01 ApperidiX D Page 2 Film Plant Random Sample Ln PFOS ppm -2.5 -3.0 -J-5 6L-j maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% -0.0555 -0.0555 -0.2197 -1.2965 -1.5712 -1.9841 -2.4421 -2.7002 -3.8019 -4.1997 -4-1997 Moments Mean Std Dev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights -1.99622 0.67992 0.08778 -1.82058 -2.17187 60.00000 60.00000 Test forNormality Shapiro-WilkW Test W Prob<W 0.975227 0.4827 AppendiX D Page 3 Film Plant Random Sample PFHS ppm Apperidix D Page 4 L maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 0.21000 0.21000 0.14670 0.04660 0.02660 0.01190 0.00718 0.00565 0.00131 0.00131 0.00131 Moments Mean StdDev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 0.02258 0.03053 0.00397 0.03053 0.01462 59.00000 59.00000 Test forNormality Shapiro-WilkW Test w Prob<W 0.578079 0.0000 Film Plant Random Sample Ln PFHS ppm AppendixD Page 5 -7 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% -1.5606 -1.5606 -2.0224 -3.0662 -3.6268 4.4312 -4.9365 -5.1761 -6.6377 -6.6377 -6.6377 Moments Mean Std Dev Std ErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 4.26780 0.95250 0.12401 4.01958 4.51602 59.00000 59.00000 Test forNormality Shapiro-WilkW Test w Prob<W 0.977452 0.5773 Film Plant Random Sample POAA ppm Appendix D Page 6 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 0.29800 0.29900 0.27200 0.15400 0.10800 0.05520 0.02400 0.01560 0.00651 0.00598 0.00598 Moments Mean Std Dev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 0.07084 0.06200 0.00807 0.08700 0.05469 59.00000 59.00000 Test forNoffnality Shapiro-WilkW Test w Prob<W 0.843094 <.0001 FilmPlant RandGm Sample Ln POAA ppm -2.5 - -3.5 -t.5 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% -1.2107 -1.2107 -1.3065 -1.8708 -2.2256 -2-8968 -3.7297 -4.1605 -5-0377 -5.1193 -5.1193 Moments Mean Std Dev Std Error Mean Upper 95% Mean Lower 95% Mean N Sum Weights -3.02097 0.91335 0.11891 --2.78295 -3-25899 59.00000 59.00000 Test for Normality Shapiro-Wilk W Test w Prob<W 0.975823 0.5122 Appendix D Page 7 FilmPlant Random Sample PFOSAAppm D.030 D.025 O.D20 O.Dl5 @.oio 0.005 0.000 maximum quartile median quartile minimum Quanflies 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 0.03780 0.03780 0.02975 0.00635 0.00487 0.00280 0.00112 0.00112 0.00112 0.00112 0.00112 Moments Mean Std Dev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 0.00397 0.00554 0.00072 0.00542 0.00253 59.00000 59.00000 Test forNormality Shapiro-WilkW Test w Prob<W 0.511689 0.0000 APPendix D Page 8 FilmPlant Random Sample Ln PFOSAA ppm -J.D -I,D -1.5 -6.0 ........... maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% -3.2754 -3.2754 -3.5529 -5-0593 -5.3247 -5.8781 -6.7944 -6.7944 -6.7944 -6.7944 -6.7944 Moments Mean Std Dev Std ErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights -5.95844 0.84775 0.11037 -5.73751 -6.17936 59.00000 59.00000 Test forNormality Shapiro-WilkW Test w Prob<W 0.843132 <.0001 Appendix D Page 9 FilmPlant Random Sample M570 ppm D.3 0.1 maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 0.45400 0.45400 0.25193 0.04805 0.01420 0.00690 0.00432 0.00251 0.00112 0.00080 0.00080 Moments Mean StdDev Std ErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 0.02024 0.05901 0.00762 0.03548 0.00499 60.00000 60.00000 Test forNormality Shapiro-WilkW Test w Prob<W 0.293209 0.0000 Appendix D Page 10 FilmPlant Random Sample Ln M570 ppm APPendix D Page IM I maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% -0.7897 -0.7897 -1.7780 -3.0356 -4.2546 -4.9779 -5.4434 -5-9875 -6.8371 -7.1309 -7.1309 Moments Mean StdDev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights -4.79892 1.10619 0.14281 -4.51316 -5.08467 60.00000 60.00000 Test forNormality Shapiro-WilkW Test w Prob<W 0.948604 0.0263 FilmPlant Random Sample M556 ppm 035 D.2D C.DS maximum quartile median quartile minimum Quantiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% 0.30700 0.30700 0.15407 0.00593 0.00250 0.00250 0.00250 0.00117 0.00021 0.00010 0.00010 Moments Mean Std Dev StdErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights 0.00816 0.03932 0.00508 0.01832 -0.00200 60.00000 60.00000 Test forNormality Shapiro-WilkW Test w Prob<W 0.162266 0.0000 Appendix D Page 12 FilmPlant Random Sample Ln M556 ppm -7 3 -8 -9 -10 maximum quartile median quartile minimum Quanfiles 100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% -1.1809 -1.1809 -2.7418 -5.1284 -5.9915 -5.9915 -5.9915 -6.7632 -8.6336 -9.2103 -9.2103 Moments Mean Std Dev Std ErrorMean Upper 95% Mean Lower 95% Mean N Sum Weights -5.93097 0.95792 0.12367 -5.68351 -6.17942 60.00000 60.00000 Test forNormality Shapiro-WilkW Test w Prob<W 0.682874 <.0001 Appendix D Page 13 AppendixE Page I Aor)endiEx Scatterplaontdsrec-,,reesqsuiaotniofnosrtluorochemicbaylysearsworkedInchemical (YRSCHEM) for random sample (n= 126)and forcurrenjtob cateogrie(schemical operatorse,ngineer/labm,aintenance,supervisor/mcmtand milloperators) 12.00 RandomSample PFOS ppm ByYRSCHEM IO.DO Appendix E Page 2 5.DD Term Intercept YRSCHEM 2.0 0.0: 0 5 10 15 20 25 30 35 10 IRSCUBI Source Model Error C Total = Lim ft Linear Fit PFOSdfppm 0.89178 + 0.0478 Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) YRSCliEM 0.10808 0.100887 -1.528756 1.504686 126 Analysis of Variance DF Sum of Squares 1 35.11712 124 289.79964 125 324.91676 Mean Square 35.1171 2.3371 F Ratio 15.0260 Prob>F 0.0002 Estimate 0.8917838 0.0478029 Parameter Std Error 0.208682 0.012332 Estimates t Ratio 4.27 3.88 Prob>ltl <.0001 0.0002 Lower 95% 0.4787397 0.0233943 Upper 95% 1.3048279 0.0722116 Random Sample PFHS ppm By YRSCHEM Appendix E Page 3 0 5 10 15 20 25 30 JS YISCHEW Term Intercept YRSCHEM Lim LinearFit PFHSdfppm 0.11968 + 0.01757 YRSCIREM Summary ofFit RSquare 0.223991 RSquare Adj 0.217733 Root Mean SquareError 0.364103 Mean of Response 0.344977 Observations(orSum Wgts) 126 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 4.744959 4.74496 124 16.438777 0.13257 125 21.183736 F Ratio 35.7919 Prob>F <.0001 Estimate 0.1196844 0.0175716 Parameter Estimates Std Ermr t Ratio 0.049702 2.41 0.002937 5.98 Prob>ltl 0.0175 <.0001 Lower 95% 0.02131 0.0117582 Upper 95% 0.2180589 0.023385 Random Sample POAA ppm By YRSCHEM 1.00 6.00 Appendix E Page 4 gas 4 o 5 io 15. .20. .25. .i.c .i.s .40. Lim Tenn Intercept YRSCHEM POAAppm Unear Fit 1.29399+ 0.0189 YRSCHEM Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.0246 0.016734 1.324636 1.536271 126 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 5.48740 5.48740 124 217.57785 1.75466 125 223.06524 F Ratio 3.1273 Prob>F 0.0794 Estimate 1.2939922 0.0188964 Parameter.Estimates Std Error t Ratio 0.180819 7.16 0.010685 1.77 Prob>ltl <.0001 0.0794 Lower 95% 0.9360979 -0.002253 Upper 95% 1.6518866 0.040046 Random Sample PFOSAA ppm By YRSCHEM 0.30 Appendix E Page 5 0.20 LC) 0.10 7. 93. 0.00 10 15 20 25 ID 35 10 Term Intercept YRSCBEM Lim R LinearFit PFOSAAdfppm = 0.03213- 0.00069YRSCHEM Summary ofFit RSquare RSquam Adj Root-Mean Square Error Mean of Response Observations(orSum Wgts) 0.0373 0.029536 0.03898 0.023293 126 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00729999 0.007300 124 0.18840938 0.001519 125 0.19570936 F Ratio 4.8044 Prob>F 0.0303 Estimate 0.0321302 -0.000689 Parameter Estimates Std Error t Ratio 0.005321 6.04 0.000314 -2.19 Prob>ltl <.Oool 0.6303 Lower 95% 0.0215985 -0.001312 Upper 950/o 0.0426619 -0.000067 Random Sample M570 ppm By YRSCHEM Appendix E Page 6 O.SO 0.70 0.60 0.50 0.40 0.30 0.20 0.10 O-OD D 5 10 15 20 25 30 35 40 INSCHIM Lim Term Intercept YRSCHEM LinearFit M570ppm 0.1791 - 0.00223YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareEffor Mean ofResponse Observations(orSum Wgts) 0.017688 0.009766 0.185242 0.150471 126 Source Model Effor C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.0766188 0.076619 124 4.2550321 0.034315 125 4.3316509 F Ratio 2.2328 Prob>F 0.1376 Estimate 0.1791 -0.002233 ParameterEstimates Std Error t Ratio 0.025286 7.08 0.001494 -1.49 Prob>ltl <.0001 0.1376 Lower 95% 0.1290506 -0.005191 Upper 95% 0.2291494 0.0007248 0.70 Random Sample PFOSA ppm By YFISCHEM 0.60 0,50 0.10 APPeridix E Page 7 0.20 Term Intercept YRSCHEM 0.00 0 5 10 15 20 25 30 is le )ISCHEM Source Model Error C Total Lim Linear Fit PFOSAdfppm 0.06731 - 0.00043 Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) YRSCHEM 0.001675 -0.00638 0.116893 0.061792 126 Analysis of Variance DF Sum of Squares 1 0.0028424 124 1.6943378 125 1.6971802 Mean Square 0.002842 0.013664 F Ratio 0.2080 Prob>F 0.6491 Esfimate 0.0673064 -0.00043 Parameter Std Error 0.015956 0.000943 Estimates t Ratio 4.22 -0.46 Prob>ltl <.0001 0.6491 Lower 95% 0.0357238 -0.002296 Upper 95% 0.0988889 0.0014363 0.50 0.10 0.30 r= 1 0.20 Random Sample M556 ppm By YRSCHEM Apperiaix E Page 8 0.00 0 5 10 is 20 25 30 is 40 )ISCHIV Lim Term Intercept YRSCHEM LinearFit M556dfppm 0.05953- 0.00059YRSCHEM Summary ofFit RSquare 0.007918 RSquare Adj -0.00008 Root Mean Square Error Mean of Response O.Or73716 0.051941 Observations(orSum Wgts) 126 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00537776 0.005378 124 0.67382941 0.005434 125 0.67920717 F Ratio 0.9896 Prob>F 0.3218 Estimate 0.0595259 -0.000592 Parameter Estimates Std Error t Ratio 0.010063 5.92 0.000595 -0.99 Prob>ltl <.0001 0.3218 Lower 95% 0.0396089 -0.001769 Upper 95% 0.0794428 0.0005854 Random Sample Chemical Operators PFOS ppm By YRSCHEM Apperidix E Page 9 6.00 1.00 i.uu 2.00 1.00 0.00 0 5 le 15 20 25 io is YRSCREM Term Intercept YRSCHEM um Fi Source Model Error C Total LinearFit PFOSdfppm = 1.41646+ 0.03312YRSCHEM Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.060486 0.039608 1.237904 1.781106 47 AnalysisofVariance DF Sum ofSquares Mean Square 1 4.439524 4.43952 45 68.958297 1.53241 46 73.397820 F Ratio 2.8971 Prob>F 0.0956 Estimate 1.4164581 '0.0331178 Parameter Estimates Std Error tRatio 0.280181 5.06 0.019457 1.70 Pmb>ltl <.0001 0.0956 Lower 95% 0.8521458 -0.006071 Upper 95% 1.9807704 0.0723065 Random Sample Chemical Operators PFHS ppm By YRSCHEM Apperidix E Page 10 0.00 0 man 5 10 15 20 25 30 35 YISCHBI Term Intercept YRSCHEM = Lim LinearFit PFHSdfppm 0.14813 + 0.0254 YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.342256 0.32764 0.333897 0.427751 47 Source Model Error C Total AnalysisofVariance DF Sum of Squares Mean Square 1 2.6105444 2.61054 45 5.0169158 0.11149 46 7.6274602 F Ratio 23.4157 Prob>F <.0001 Estimate 0.148129 0.0253956 Parameter Estimates Std Error t Ratio 0.075573 1.96 0.005248 4.84 Prob>ltl 0.0562 <.0001 Lower 95% -0.004082 0.0148254 Upper 95% 0.3003395 0.0359659 Random Sample Chemical Operators POAA ppm By YRSCHEM APPendix E Page 11 f.00 ol. to is 20 25 30 35 ISCREW Lim Term Intercept YRSCHEM LinearFit POAAppm 1.73387+ 0.04702YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.09937 0.079356 1.342508 2.251574 47 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 8.948633 8.94863 45 81.104760 I.SM3 46 90.053393 F Raflo 4.9650 Prob>F 0.0309 Estimate 1.7338672 0.0470188 Parameter Estimates Std Error tRatio 0.303857 5.71 0.021101 2.23 Prob>ltl <.0001 0.0309 Lower 95% 1.12187 0.0045187 Upper 95% 2.3458643 0.089519 Random Sample Chemical Operators PFOSAA ppm By YRSCHEM Djo APPerldix E Page 12 D.ID D.DD 0 5 10 15 20 25 30 35 Term Intercept YRSCFEEM Lim R LinearFit PFOSAAdfppm 0.0494 - 0.00118YRSCHEM Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observation(sorSum Wgts) 0.03628 0.014864. 0.057519 0.036"7 47 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00560458 0.005605 45 0.14887811 0.003308 46 0.15448269 F Ratio 1.6940 Prob>F 0.1997 Estimate 0.0494028 -0.001177 Parameter Estimates Std Error tRatio 0.013019 3.79 0.000904 -1.30 Prob>ltl 0.0004 0.1997 Lower 95% 0.0231822 -0.002998 Upper 95% 0.0756233 0.0006442 Random Sample Chemical Operators M570 ppm By YRSCHEM 0.80 0.60 Appendix E Page 13 0.2 men 0.00 05 10 15 20 25 IRSCREM 30 35 Lim Term Intercept YRSCHEM LinearFit M570ppm = 0.30244- 0.00666YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError 0.071071 0.050428 0.228431 Mean ofResponse 0.229083 Observation(sorSum Wgts) 47 Source Model FATor C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.1796529 0.179653 45 2.3481307 0.052181 46 2.5277836 F Ratio 3.4429 Prob>F 0.0701 Estimate 0.3024368 -0.006662 ParameterEstimates StdError tRaflo 0.051702 5.85 0.00359 -1.86 Prob>ltl <.0001 0.0701 Lower95% 0.1983041 -0.013894 Upper 95% 0.4065696 0.0005694 Random Sample Chemical Operators PFOSA ppm By YRSCHEM Appendix E Page 14 0.20 0 5 10 15 20 25 30 35 YRSCHgf = Lim Term Intercept YRSCHEM LinearFit PFOSAdfppm 0.12291 - 0.00214YRSCHEM Summary of Fit RSquare 0.023165 RSquare Adj 0.001457 Root Mean SquareError 0.131513 Mean of Response 0.099399 Observations(orSum Wgts) 47 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.01845661 0.018457 45 0.77830020 0.017296 46 0.79675681 F Ratio 1.0671 Prob>F 0.3071 Estimate 0.1229105 -0.002135 Parameter Estimates Std Error tRatio 0.029766 4.13 0.002067 -1.03 Prob>ltl 0.0002 0.3071 Lower 95% 0.0629591 -0.006299 Upper 95% 0.182862 0.002028 Random Sample Chemical Operators M556 ppm By YRSCHEM 7 '-I@ Appendix E Page 15 0.30 E 0.20 % D.ID 0 5 1D 15 20 25 30 35 Term Intercept YRSCHEM Source Model Error C Total Limfi j LinearFit M556dfppm = 0.09775- 0.00212YRSCHEM Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.069465 0.048786 0.073484 0.074438 47 Analysisof Variance DF Sum ofSquares Mean Square 1 0.01813944 0.018139 45 0.24299278 0.005400 46 0.26113221 F Ratio 3.3593 Prob>F 0.0734 Estimate 0.097747 -0.002117 Parameter Estimates StdError t Ratio 0.016632 5.88 0.001155 -1.83 Prob>ltl <.0001 0.0734 Lower 95% 0.0642487 -0.004443 Upper 95% 0.1312453 0.0002094 RandomSample Engineer/Lab PFOS ppm By YRSCHEM 2 00 1.50 APPendix E Page 16 0 5 10 15 20 25 30 35 40 IMSCREM Term Intercept YRSCI-IEM LinearFit PFOSdfppm 0.40446+ 0.01564YRSCHEM Summary ofFit RSquare 0.124933 RSquare Adj 0.083263 Root Mean SquareEffor 0.574244 Mean of Response 0.633961 Observations(orSum Wgts) 23 Source Model Error C Total AnalysisofVaOance DF Sum ofSquares Mean Square 1 0.9886669 0.988667 21 6.9248903 0.329757 22 7.9135572 F Ratio 2.9982 Prob>F 0.0980 Estimate 0.4044606 0.01564 ParameterEstimates Std Error t Ratio 0.178619 2.26 0.009033 1.73 Prob>ltl 0.0343 0.0980 Lower 95% 0.0330049 -0.003144 Upper 95% 0.7759163 0.034424 Random Sample Engineer/Lab PFHS ppm By YASCHEM Appendix E Page 17 0.60 0.2D 0.00 0 5 10 15 20 25 30 35 40 ISCHEW Term Intercept YRSCHEM PFHSdfppm UnearFft 0.10657 + 0.00439YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareEffor 0.065012 0.020489. 0.23077 Mean ofResponse 0.170933 Observations(orSum Wgts) 23 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.0777620 O.Cr77762 21 1.1183544 0.053255 22 1.1961164 F Ratio 1.4602 Prob>F 0.2403 Estimate 0.1065696 0.0043863 Parameter Estimates Std Error t Ratio 0.071781 1.48 0.00363 1.21 Prob>ltl 0.1525 0.2403 Lower 95% -0.042706 -0.003162 Upper 95% 0.2558458 0.011935 RandomSample EngineerA.ab POAA ppm ByYRSCHEM APPendix E Page 18 ?.DD 1.00 Term Intercept YRSCHEM 0.00 0 5 10 15 20 25 30 35 10 IRSCUBI Source Model Error C Total Litff Linear Fit POAAppm = 0.34907 + Summary RSquare RSquare Adj Root Mean Square Error Mean of Response Observation s (or Sum Wgts) 0.00185 of Fit YRSCHEM 0.002389 --0.04512 0.525717 0.376278 23 Analysis of Variance DF Sum of Squares 1 0.0138962 21 5.8039396 22 5.8178358 Mean Square 0.013896 0.276378 F Ratio 0.0503 Prob>F 0.8247 Estimate 0.3490696 0.0018542 Parameter Std Error 0.163524 0.008269 Estimates t Ratio 2.13 0.22 Prob>ltl 0.0447 0.8247 Lower 95% 0.0090046 -0.015342 Upper 95% 0.6891347 0.0190509 Random Sample Engineer/Lab PFOSAA ppm By YRSCHEM Appendix E Page 19 0.06 E i D.02 0.00 0 5 10 15 20 25 30 is to IISCUBI Lim Term Intercept YRSCHEM Source Model Error C Total LinearFit PFOSAAdfppm = 0.01789- 0.00027YRSCHEM Summary ofFit RSquare 0.034722 RSquare Adj -0.01124 Root Mean SquareError 0.019647 Mean ofResponse 0.013949 Observation(sorSum Wgts) 23 AnalysiosfVariance DF Sum ofSquares Mean Square 1 0.00029158 0.000292 21 0.00810598 0.000386 22 0.00839755 F Ratio 0.7554 Prob>F Estimate 0.0178899 -0.000269 ParameterEstimates StdError tRatio 0.006111 2.93 0.000309 -0.87 Prob>ltl 0.0080 0.3946 Lower 95% 0.0051812 -0.000911 Upper95% 0.0305987 0.0003741 Random Sample Engineer/Lab M570 ppm By YRSCHEM Appendix E Page 20 0 5 10 15 20 25 jo 35 4o YRSCHBI Lim Term Intercept YRSCHEM LinearFit M570ppm = 0.0747- 0.00004YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.000047 -0.04757 0.087863 0.074065 23 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00000754 0.000008 21 0.16211665 0.007720 22 0.16212419 F Ratio 0.0010 Prob>F 0.9754 Estimate 0.0746991 -0.000043 Parameter Estimates Std Error tRatio 0.02733 2.73 0.001382 -0.03 Prob>ltl 0.0125 0.9754 Lower 95% 0.0178643 -0.002917 Upper 95% 0.131534 0.0028309 Term Intercept YRSCHEM Random Sample Engineer/Lab PFOSA ppm By YRSCHEM O.DB Appendix E Page 21 0.05 0.04 E 0.02 O.DO 0 5 10 is 20 25 30 35 40 liscHEM Source Model Error C Total Lim Linear Fit PFOSAdfppm 0.01475 + 0.00015 Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) YRSCBEM 0.007656 -0.0396 0.023047 0.01689 23 Analysis of Variance DF Sum of Squares 1 0.00008605 21 0.01115404 22 0.01124009 Mean Square 0.000086 0.000531 F Ratio 0.1620 Prob>F 0.6914 Estimate 0.0147485 0.0001459 Parameter Std Error 0.007169 0.000363 Estimates I Ratio 2.06 0.40 Prob>ltl 0.0523 0.6914 Lower 95% -0.000159 -0.000608 Upper 95% 0.0296564 0.0008998 RandomSample Engineer/Lab M556ppm ByYRSCHEM APPendix E Page 22 E Term Intercept YRSCHEM 0.02 0.00 0 5 10 15 20 2; Jo' 35 to INSCREM Source Model Error C Total Lm ft M556dfppm RSquare .RSquare Adj Root Mean Square Mean of Response Observations (or Linear 0.0 188 Summary Effor Sum Wgts) Fit 0.00001 of Fit YRSCHEM .0.000046 -0.04757 0.027423 0.0186 23 Analysis of Variance DF Sum of Squares 1 0.00000073 21 0.01579191 22 0.01579264 Mean Square 0.000001 0.000752 F Ratio 0.0010 Prob>F 0.9754 Estimate 0.0187973 -0.000013 Parameter Std Error 0.00853 0.000431 Estimates t Ratio 2.20 -0.03 Prob>ltl 0.0388 0.9754 Lower 95% 0.0010588 -0.00091 Upper 95% 0.0365359 0.0008836 Random Sample Maintenance PFOS ppm By YRSCHEM APPendix E Page 23 0 2.00 1.00 0.00 0 5 10 15 20 25 30 Lim Term Intercept YRSCHEM Source Model Error C Total LinearFit PFOSdfppm = 1.36713+ 0.03289YRSCHEM Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.073544 -0.0294 1.245224 1.672091 II AnalysisofVariance DF Sum ofSquares Mean Square 1 1.107805 1.10780 9 13.955256 1.55058 10 15.063061 F Ratio 0.7144 Prob>F 0.4199 Estimate 1.3671255 0.0328884 Parameter Estimates Std Error tRaUo 0.52071 2.63 0.03891 0.85 Prob>ltl 0.0276 0.4199 Lower 96% 0.1891877 -0.055132 Upper 95% 2.5450633 0.1209093 AppendixE Page 24 Random Sample Maintenance PFHS ppm By YRSCHEM 1.00 0.80 0.60 0.40 0.20 0.00 0 5 10 is 20 25 30 YRSCH[M Term Intercept YRSCHEM Lim I Source Model Error C Total LinearFit PFHSdfppm = 0.1267 + 0.01194 YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareEffor Mean ofResponse Observations(orSum Wgts) 0.261552 0.179502 0.214098 0.237455 II AnalysisofVariance DF Sum ofSquares Mean Square 1 0.14611866 0.146119 9 0.41254111 0.045838 10 0.55865977 F Ratio 3.1877 Prob>F 0.1078 Estimate 0.1266974 0.0119444 Parameter Estimates Std Error tRatio 0.089528 1.42 0.00669 1.79 Prob>ltl 0.1907 0.1078 Lower 95% -0.075832 -0.003189 Upper 95% 0.3292263 0.0270783 Random Sample Maintenance POAA ppm By YRSCHEM APPendix E Page 25 2.00 10 15 20 25 JO ISCHR4 Term Intercept YRSCREM = Lieff LinearFit POAAppm RSquare RSquare Adj 1.24651+ 0.02555YRSCBEM Summary ofFit Root Mean SquareError Mean ofResponse Observation(sorSum Wgu) 0.039706 -0.06699 1.340539 1.483455 11 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.668731 0.66873 9 16.173404 1.79704 10 16.842135 F Ratio 0.3721 Prob>F .0.5569 Estimate 1.2465111 0.0255527 Parameter Estimates SidError t Ratio 0.560567 2.22 0.041888 0.61 Prob>ltl 0.0532 0.5569 Lower 95% -0.021591 -0.069206 Upper 95% 2.514613 0.1203111 Random Sample Maintenance PFOSAA ppm By YASCHEM O-OB APPendiX E Page 26 0.02 0.00 6 5 10 15 20 25 30 INSCHEW Term Intercept YRSCHEM Litff Unear Fit PFOSAAdfppm 0.0347- 0.00006YRSCHEM Summary ofFit RSquare 0.000472 RSquam Adj -0.11059 Root Mean SquareError 0.031301 Mean of Response 0.034106 Observation(sorSum Wgts) II Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00000416 0.000004 9 0.00881782 0.000980 10 0.00882198 F Ratio 0.0042 Prob>F 0.9495 Estimate 0.0346975 -0.000064 ParameterEstimates Std Error tRaUo 0.013089 2.65 0.000978 -0.07 Prob>ltl 0.0264 0.9495 Lower95% 0.0050878 -0.002276 Upper 95% 0.0643072 0.0021488 Random Sample Maintenance M570 ppm By YRSCHEM 0 BO 0.6D D.ID 0.20 APPendix E Page 27 0 5 10 is 20 25 30 )ISCHEW Term Intercept YRSCIREM Lim R LinearFit M570ppm = 0.26076+ 0.00079YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean ofResponse Observations(orSum Wgts) 0.001749 -0.10917 0.201468 0.268091 II Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00064018 0.000640 9 0.36530321 0.040589 10 0.36594339 F Ratio 0.0158 Pmb>F 0.9028 Estimate 0.2607598 '0.0007906 Parameter Estimates StdError tRatio 0.084247 3.10 0.006295 0.13 Pmb>ltl 0.0128 0.9028 Lower 95% 0.0701785 -0.01345 Upper 95% 0.4513411 0.0150317 Random Sample Maintenance PFOSA ppm By YRSCHEM AOPendix E Page 28 0.10 R 0.20 0.00 0 5 10 is 20 25 30 )ISCHEW Term Intercept YRSCHEM LinearFit PFOSAdf@pm = 0.09744- 0.00351YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.043743 -0.06251 0.174833 0.064939 II Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.01258420 0.012584 9 0.27509911 0.030567 10 0.28768331 F Ratio 0.4117 Prob>F 0.5371 Estimate 0.0974427 -0.003505 Parameter Estimates Std Error t Ratio 0.(Y73109 1.33 0.005463 -0.64 Prob>ltl 0.2153 0.5371 Lower 95% -0.067943 -0.015864 Upper 95% 0.2628285 0.0088531 Random Sample Maintenance M556 ppm By YRSCHEM 0 50 APPendix E Page 29 D.10 0 5 10 is 20 25 30 IMSCH(M Term Intercept YRSCHEM = um R LinearFit M556dfppm = 0.11026+ 0.00048YRSCHEM Summary ofFit RSquare 0.001793 RSquare Adj -0.10912 Root Mean SquareError 0.120441 Mean ofResponse 0.1147 Observation(sorSum Wgts) II Source Model Error C Total Analysisof Variance DF Sum ofSquares Mean Square 1 0.00023453 0.000235 9 0.13055477 0.014506 10 0.13(Y78930 F Raflo 0.0162 Prob>F .0.9016 Estimate 0.1102627 0.0004785 ParameterEstimates StdError tRatio 0.050364 2.19 0.003763 0.13 Prob>ltl 0.0563 0.9016 Lower 95% -0.00367 -0.008035 Upper 95% 0.2241958 0.0089921 Random Sample SupervisoUMgmt PFOS ppm By YRSCHEM APpendix E Page 30 10.00 8.00 6.oo 1.00 2.00 D.DD D 5 10 15 20 25 JD 35 lo IRSCIDI Term Intercept YRSCHEM Lim R j LinearFit PFOSdfppm -0-2688+ 0.10578YRSCHEM Summary ofFit RSquare 0.197186 RSquare Adj 0.14701 Root Mean Square Error 2.366822 Mean of Response 1.879072 Observations(orSum Wgts) 18 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 22.01465 22.0146 16 89-62951 5.6018 17 111-64416 F Ratio 3.9299 Prob>F 0.0649 Estimate -0.268787 0.1057769 Parameter Estimates Std Error t Ratio 1.218652 -0.22 0.053358 1.98 Prob>ltl 0.8282 0.0649 Lower 95% -2.8522 -0.007337 Upper 95% 2.3146273 0.2188905 Random Sample Supervisor/Mgmt PFHS ppm By YRSCHEM 2.00 s CL APPendix E Page 31 0.00 5 io 15 20 25 30 is 40 'ASCHIN Term Intercept YRSCHEM = Limri LinearFit PFHSdfppm 0.04613+ 0.01835YRSCHEM Summary of Fit RSquare RSquare Adj 0.194032 0.143659 Root Mean SquareError 0.414774 Mean of Response Observation(sorSum Wgts) 0.418777 18 Source Model Error C Total Analysisof Variance DF Sum of Squares Mean Square 1 0.6626723 0.662672 16 2.7525963 0.172037 17 3.4152686 F Ratio 3.8519 Prob>F 0.0673 Estimate 0.0461288 0.018352 ParameterEstimates StdError tRatio 0.213563 0.22 0.009351 1.96 Prob>ltl 0.8317 0.0673 Lower 95% -0.406602 -0.001471 Upper 95% 0.4988593 0.0381746 Random Sample Supervisor/Mgmt POAAppm By YASCHEM APPendix E Page 32 2.00 1.00 0.00 0 5 10 15 20 25 30 35 40 YRSCREW Term Intercept YRSCHEM um R Source Model Error C Total LinearFit POAAppm = 0.30941+ 0.05233YRSCHEM RSquare Summary of Fit RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.171838 0.120078 1.27387 1.370928 18 AnalysisofVariance DF Sum ofSquares Mean Square 1 5.387336 5.38734 16 25.963931 1.62275 17 31.351267 F Ratio 3.3199 Prob>F 0.0872 Estimate 0.3084084 0.0523265 Parameter Estimates Std Error t Ratio 0.655903 0.47 0.028718 1.82 Prob>lq 0.6446 0.0872 Lower 95% -1.082036 -0.008553 Upper 95% 1.698853 0.1132065 Random Sample Supervisor/Mgmt PFOSAA ppm By YRSCHEM 0,06 0.04 APPendix E Page 33 0 5 le 15 20 25 30 35 40 =-Lim Term Intercept YRSCHEM Unear Fit PFOSAAdppm 0.00595+ 0.00023YRSCHEM Summary of Fit RSquare 0.036993 RSquare Adj Root Mean Square Error -0.0232 0.012927 Mean of Response 0.010586 Observations(orSum Wgts) 18 Source Model Error C Total AnalysisofVariance DF Sum of Squares Mean Square 1 0.00010271 0.000103 16 0.00267372 0.000167 17 0.00277643 F Ratio 0.6146 Prob>F 0.4445 Estimate 0.0059463 0.0002285 Parameter Estimates Std Error tRatio 0.006656 0.89 0.000291 0.78 Prob>ltl 0.3849 0.4445 Lower 95% -0.008164 -0.000389 Upper 95% 0.0200562 0.0008463 Random Sample Supervisor/Mgmt M570ppm By YRSCHEM Appendix E Page 34 0.20 0.10 0.00 5 10 15 20 25 30 35 ID YRSCUBI Term Intercept YRSCBEM um R LinearFit M570ppm = 0.05229+ 0.00341YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.059466 0.000683 0.150515 0.121594 Is Source Model Error C Total AnalysisofVadance DF Sum ofSquares Mean Square 1 0.02291777 0.022918 16 0.36247508 0.022655 17 0.38539285 F Ratio 1.0116 Prob>F 0.3295 Estimate 0.052294 0.0034129 Parameter Estimates StdError tRatio 0.077498 0.67 0.003393 1.01 Prob>ltl 0.5095 0.3295 Lower 95% -0.111995 -0.00378 Upper 95% 0.2165826 0.0106062 Random Sample Supervisor/Mgmt PFOSA ppm By YRSCHEM APPerldix E Page 35 0.50 0.10 0.20 0.10 0.00 0 5 10 15 20 25 30 35 to ISCREM = Lim Term Intercept YRSCHEM LinearFit PFOSAdfppm -0-0334+ 0.00483YRSCHEM Summary of Fit RSquare RSquare Adj Root Mean SquareError 0.114214 0.058852 0.14906 Mean of Response Observation(sorSum Wgts) 0.064622 18 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.04583882 0.045839 16 0.35550303 0.022219 17 0.40134185 F Ratio 2.0631 Prob>F 0.1702 Estimate -0.033387 0.0048267 ParameterEstimates Std Error tRatio 0.07675 -0.44 0.00336 1.44 Prob>ltl 0.6694 0.1702 Lower 95% -0.196088 -0.002297 Upper 95% 0.129314 0.0119505 o.j5 0.30 0 25 D.20 0.15 Random Sample Supervisor/Mgmt M556 ppm By YRSCHEM Appendix E Page 36 Term Intercept YRSCEEM 0.05 5 10 15 20 25 30 35 to )ISCHEW Source Model Error C Total Lim R Linear M556dfppm -0.007 + RSquare Summary RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) Fit 0.00261 of Fit YRSCHEM 0.105382 0.049469 0.084362 0.046 18 Analysis of Variance DF Sum of Squares 1 0.01341369 16 0.11387207 17 0.12728576 Mean Square 0.013414 0.007117 F Ratio 1.8847 Prob>F 0.1887 Estimate -0.007018 0.002611 Parameter Std Error 0.043437 0.001902 Estimates t Ratio -0.16 1.37 Prob>ltl 0.8737 0.1887 Lower 96% -0.099101 -0.001421 Upper 95% 0.0850644 0.0066428 Random Sample MillOperators PFOS ppm By YRSCHEM J,OD 2.DD r= I.DD D ID is 2D 25 Appendix E Page 37 Random Sample MillOperators PFHS ppm By YRSCHEM D.JD 0.20 D.lD D.00 0 I life i---rTrr-rM 10 15 20 25 'ISCH[ii AppendiX E Page 38 Random Sample MillOperators POAA ppm By YRSCHEM 2.00 C) 1.00 0.00 0 10 is 20 25 )ISCHEM Appendix E Page 39 Random Sample MillOperators PFOSAA ppm By YRSCHEM O.Di APPP-ndix E Page 40 r-r-r10 YRSCHDA 20 25 Random Sample MillOperators M570 ppm By YRSCHEM 0.10 0.02 0.00 0 5 10 is 20 25 YRSCHEM AppendiX E Page 41 Random Sample MillOperators PFOSA ppm By YRSCHEM cE: r 0.00 Is 20 25 Appendix E Page 42 Random Sample MillOperators M556 ppm By YRSCHEM 0 06 D.DS 0.04 0.03 0.02 01 :,Oo ........io 15 20 25 Appendix E Page 43 Apperidix F Page I Appendix F Scatterplotasnd regressionequationsforfluorochen-@ca(lnsaturalog transformationb)y yearsworked inchemical(YRSCBEM) forallrandom sample(n= 126)and fortwo currentjob cateogrie(schemicaloperatorsand engineer/lab) Random Sample InPFOS ppm By YRSCHEM 2.00 I.DD r= APPendix F Page 2 .S-1.00 -2.DD D 5 10 is ?D 25 30 35 40 YRSCREM = Lim Term Intercept YRSCBEM LinearFit InPFOSdfppm -0.4008+ 0.02654YRSCHEM Summary ofFit RSquare 0.082224 RSquare Adj 0.074823 Root Mean Square Error Mean of Response 0.987128 -0.06052 Observations(orSum Wgts) 126 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 10.82508 10.8251 124 120.82819 0.9744 125 131.65326 F Ratio 11.1092 Prob>F 0.0011 Estimate -0.4008Or7 0.0265406 Parameter Estimates Std Error t Ratio 0.134748 -2.97 0.007963 3.33 Prob>ltl 0.0035 0.0011 Lower 95% -0.667512 0.0107798 Upper 95% -0.134101 0.0423014 RandomSample InPFHS ppm ByYRSCHEM I.OD O,DD -1.00 r_2.DD r_ -f.00 D5 r-r 10 15 20 25 30 35 4D INSCREM APPendix F Page 3 Term Intercept YRSCHEM Source Model Effor C Total Unear In PFHSdfppm = -2.4032 Summary RSquare RSquare Adj Root Mean Square Effor Mean of Response Observations (or Sum Wgts) Fit + 0.05366 of Fit Analysis of Variance DF Sum of Squares 1 44.25642 124 142.56346 125 186.81988 Estimate -2.403248 0.053664 Parameter Std Error 0.146366 0.008649 Estimates t Ratio -16.42 6.20 YRSCHEM 0.236894 0.230739 1.072243 -1.7152 126 Mean Square 44.2564 1.1497 F Ratio 38.4937 Prob>F <.0001 Prob>ltl <.0001 <.0001 Lower95% -2.69295 0.0365442 Upper 95% -2.113546 0.0707838 Random Sample InPOAA ppm By YRSCHEM an@ Apperidix F Page 4 Term Intercept YRSCBEM 0 5 10 15 20 25 30 35 4o INSCREW I = LitfRf LinearFit InPOAAppm -0.2007+ 0.00738YRSCHEM Summary of Fit RSquare RSquare Adj 0.004252 -0.00378 Root Mean Square Error 1.256877 Mean of Response Observations(orSum Wgts) -0.10609 126 Source Model Error C Total AnalysisofVariance DF Sum ofSquares K4ean Square 1 0.83656 0.83656 124 195-88769 1.57974 125 196.72425 F Ratio 0.5296 Prob>F 0.4682 Estimate -0.200686 0.0073781 Parameter Estimates Std Error t Ratio 0.17157 -1.17 0.010139 0.73 Prob>ltl 0.2444 0.4682 Lower 95% -0.540273 -0.01269 Upper 95% 0.1389006 0.0274458 Random Sample InPFOSAA ppm By YRSCHEM APPendix F Page 5 0 ms SEEN 10 15 20 25 30 JS to Term Intercept YRSCBEM Lim LinearFit InPFOSAAdfppm = -4.478- 0.02366YRSCHEM Summary ofFit RSquare 0.032321 RSquare Adj 0.024517 Root Mean Square Error 1.441-192 Mean of Response -4.7813 Observations(orSum Wgts) 126 Source Model Error C Total Analysisof Variance DF Sum of Squares Mean Square 1 8.60226 8.60226 124 257.55240 2.07704 125 266.15466 F Ratio 4.1416 Prob>F 0.0440 Estimate -4.477958 -0.023659 Parameter Esbmates Std Error t Ratio 0.19673 -22.76 0.011626 -2.04 Prob>ltl <.0001 0.0440 Lower 95% -4.867344 -0.04667 Upper 950/o -4.088572 -0.000649 Random Sample InM 570 ppm By YRSCHEM 0.00 APPeridix F Page 6 r=-200 3.00 -1.00 0 5 10 15 20 2S JO .35 fO YRSCRal um In570ppm Unear Fit -2.353- 0.0126YRSCHEM Summary ofFit RSquare RSquam Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.015641 0.007702 1.112421 -2.51453 126 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 2.43817 2.43817 124 153.44766 1.23748 125 155.88583 F Ratio 1.9703 Prob>F 0.1629 Term Intercept YRSCHEM Parameter Estimates Estimate Std Error -2.353036 0.151851 -0.012596 0.008974 t Ratio -15.50 -1.40 Prob>ltl <.0001 0.1629 000 -1.00 -2.00 -J.00 E 1.00 =-6.00 RandomSample InPFOSA ppm ByYRSCHEM own-- a so an an 8.00 am 0 5 10 15 20 25 30 35 to ISCREN APPendix F Page 7 Source Model Error C Total Term Intercept YRSCHEM Lim R Linear In PFOSAdfppm = -4.1363 Summary RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) Fit - 0.01701 of Fit YRSCHEM 0.009846 0.001861 1.899215 -4.35445 126 Analysis of Variance DF Sum of Squares 1 4.44768 124 447.26998 125 451.71766 Mean Square 4.44768 3.60702 F Ratio 1.2331 Prob>F 0.2690 Parameter Estimates Estimate Std Error t Ratio -4.13633 0.259252 -15.95 -0.017012 0.01532 -1.11 Lower 95% -4.649466 -0.047336 Upper 95% -3.623194 0.0133113 Random Sample InM556 ppm By YRSCHEM -1.00 -2.00 -J.OD APPendix F Page 8 -7-00 0 5 10 is 20 25 30 is 40 ISCUBI Term Intercept YRSCHEM Source Model Error C Total LinearFit In M556dfppm = -3.6365- 0.01244YRSCHEM Summary ofFit RSquare 0.010236 RSquare Adj 0.002254 Root Mean SquareError 1.362297 Mean ofResponse -3.79603 Observations(orSum Wgts) 126 AnalysisofVariance DF Sum ofSquares Mean Square 1 2.38002 2.38002 124 230.12586 1.85585 125 232.50588 F Ratio 1.2824 Prob>F 0.2596 Estimate -3.636469 -0.012445 Parameter Estimates Std Error t Ratio 0.18596 -19.56 0.010989 -1.13 Prob>ltl <.0001 0.2596 Lower 95% -4.004539 -0.034196 Upper 95% -3.268399 0.0093062 Random Sample ChemicalOperators InPFOS ppm By YRSCHEM ?.00 Appendix F Page 9 1.00 0.50 0.00 -0.50 -1.00 0 5 10 15 20 2S 30 35 YRSCUBI Lim Term Intercept YRSCHEM LinearFN InPFOSdfppm 0.25621+ 0.0124YRSCHEM Summary ofFit RSquare RSquare Adj 0.038311 0.01694 Root Mean SquareError 0.589163 Mean ofResponse 0.392725 Observation(sorSum Wgts) 47 Source Model Error C Total AnalysiosfVariance DF Sum ofSquares Mean Square 1 0.622261 0.622261 45 15.620109 0.347114 46 16-242370 F Ratio 1.7927 Prob>F 0.1873 Estimate 0.2562063 0.0123988 ParameterEstimates StdError tRatio 0.133348 1.92 0.00926 1.34 Prob>ltl 0.0610 0.1873 Lower 95% -0.01237 -0.006253 Upper 95% 0.524783 0.0310501 Random Sample Chemical Operators InPFHS ppm By YRSCHEM APperidix F Page 10 Is . 5 ID 15 20 25 30 35 YRSCREW Term Intercept YRSCHEM = Lim Source Model Error C Total LinearFit InPFHSdfppm -1.7176+ 0.0491YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError 0.345578 0.331035 0.640812 Mean of Response Observations(orSum Wgts) -1.17704 47 AnalysisofVariance DF Sum ofSquares Mean Square 1 9.758008 9.75801 45 18.478805 0.41064 46 28.236813 F Ratio 23.7629 Prob>F <.0001 Estimate -1.717649 0.0490992 Parameter Estimates Std Error t Ratio 0.145038 -11.84 0.010072 4.87 Prob>ltl <.0001 <.0001 Lower 95% -2.00977 0.0288128 Upper 950/o -1.425527 0.0693855 Random Sample ChemicalOperators InPOAA ppm By YRSCHEM i.oo o.oo -1.00 -?.DD 0 5 10 15 20 25 30 35 )ISCNEM Appendix F Page 11 Term Intercept YRSCHEM Unear Fit InPOAAppm 0.51048+ 0.01132YRSCHEM Summary of Fit RSquare 0.027722 RSquare Adj 0.006116 Root Mean SquareError 0.635686 Mean ofResponse 0.635094 Observation(sorSum Wgts) 47 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.518477 0.518477 45 18.184368 0.404097 46 18.702845 F Ratio 1.2831 Prob>F 0.2633 Estimate 0.5104788 0.0113177 ParameterEstimates SidError tRatio 0.143878 3.55 0.009992 1.13 Prob>ltl 0.0009 0.2633 Lower95% 0.2206941 -0.008806 Upper 95% 0.8002634 0.0314418 Random Sample Chemical Operators InPFOSAA ppm By YRSCHEM Appendix F Page 12 -5.00 *us. 10 15 20 2; 30' 35' Term Intercept YRSCHEM um R j Source Model Error C Total LinearFit InPFOSAAdfppm = -4.2679- 0.01959YRSCBEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response 0.013261 -0.00867 1.603004 -4.4836 Observations(orSum Wgts) 47 DF 1 45 46 Estimate -4.267867 -0.019594 AnalysisofVariance Sum ofSquares 1.55397 115.63304 117.18701 Parameter Estimates Std Error t Ratio 0.362816 -11.76 0.025196 -0.78 Mean Square 1.55397 2.56962 F Ratio 0.6047 Prob>F 0.4408 Prob>ltl <.0001 0.4408 Lower 95% -4.998614 -0.07034 Upper 95% -3.537119 0.0311532 Random Sample Chemical Operators InM570 ppm By YRSCHEM % APPeridix F Page 13 0 5 10 15 20 25 30 J5 YRSCHIM Lim Term Intercept YRSCHEM LinearFit In570ppm = -1.6206- 0.03729YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean of Response Observation(sorSum Wgu) 0.09218 0.072006 1.109871 -2.03122 47 Source Model Error C Total Analysisof Variance DF Sum ofSquares Mean Square 1 5.628506 5.62851 45 55.431655 1.23181 46 61.060161 F Ratio 4.5693 Prob>F 0.0380 Estimate -1.620635 -0.03729 Parameter Estimates Std Error 0.251203 tRatio -6.45 0.017445 -2.14 Prob>ltl <.0001 0.0380 Lower 95% -2.126582 -0.072425 Upper 95% -1.114688 -0.002154 Random Sample Chemical Operators InPFOSA ppm By YRSCHEM -100 -2.00 r Appendix F Page 14 -7.00 05 10 15 20 25 30 35 YRSCHDA Term Intercept YRSCHEM Lim R j LinearFit InPFOSAdfppm = -3.2174- 0.03217YRSCHEM Summary of Fit RSquare 0.026974 RSquare Adj Root Mean SquareError 0.0053511.832598 Mean of Response Observations(orSum Wgts) -3.57167 47 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 4.18951 4.18951 45 151.12876 3.35842 46 155.31826 F Ratio 1.2475 Prob>F 0.2700 Estimate -3.217438 -0.032172 Parameter Estimates Std Error t Ratio 0.414782 -7.76 0.028805 -1.12 Prob>ltl <.0001 0.2700 Lower 95% -4.052948 -0.090187 Upper 95% -2.382028 0.0258433 RandomSample ChemicaOlperators InM556ppmByYRSCHEM AppendixF Page 15 -2.00 -J.00 .s-5.00 -6.00 -7.DO 0 5 10 15 20 25 30 35 YRSCREM Source Model Error C Total Term Intercept YRSCHEM Lim Linear Fft In M556dfppm -2.7767 - 0.03141 Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) YRSCHEM 0.065942 0.045185 1.121235 -3.12253 47 Analysis of Variance DF Sum of Squares 1 3.993849 45 56.572602 46 60.566451 Mean Square 3.99385 1.25717 F Ratio 3.1769 Prob>F 0.0814 Parameter Estimates Estimate -2.77 6667 Std Error 0.253775 t Ratio -10.94 P<r,.%cMb)i>Iltl -0.031412 0.017623 -1.78 0.0814 Lower 95% Upper 95% -3.287794 -2.265539 -0.066907 0.0040837 Random Sample Engineer/Lab InPFOS ppm By YRSCHEM Appendix F Page 16 -0.50 .e -2.00 0 5 10 15 20 25 30 35 Lim Term Intercept YRSCHEM LinearFit InPFOSdfppm -1.4007+ 0.03146YRSCHEM Summary of Fit RSquare 0.170302 RSquam Adj 0.l3Cr793 Root Mean Square Error 0.963504 Mean ofResponse -0.93898 Observations(orSum Wgts) 23 Source DF Model 1 Error 21 C Total 22 Estimate -1.40069 0.0314649 AnalysisofVariance Sum ofSquares 4.001539 19.495151 23.496691 Parameter Estimates Std Error 0.299699 t Rago -4.67 0.015155 2.08 Mean Square 4.00154 0.92834 F Ratio 4.3104 Prob>F 0.0503 Prob>ltl 0.0001 0.0503 Lower 95% -2.023942 -0.000052 Upper 95% -0.777437 0.0629819 0.00 RandomSample Engineer/Lab InPFHS ppm ByYRSCHEM -1.00 -2.00 E APPerldixF Page 17 Term Intercept YRSCHEM .E-4.00 -5.00 05 10 15 20 25 30 35 INSCREM - = Lim Source Model Error C Total Linear In PFHSdfppm -3.1745 Summary RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) Fit + 0.04275 of Fit Analysis of Variance DF Sum of Squares 1 7.386077 21 30.914308 22 38.300386 YRSCREM 0.192846 0.15441 1.213305 -2.54721 23 Mean Square 7.38608 1.47211 F Ratio 5.0173 Prob>F 0.0360 Estimate -3.174495 0.0427483 Parameter Std Error 0.377399 0.019085 Estimates t Ratio -8.41 2.24 Prob>ltl <.0001 0.0360 Lower 95% -3.959334 0.0030601 Upper 95% -2.389657 0.0824366 RandomSample Engineer/Lab InPOAA ppmByYRSCHEM 100 050 -0.50 < 1.50 2.00 -2.50 --------- 0 5 10 15 20 25 30 35 APPendixF Page 18 Term Intercept YRSCHEM Source Model Error C Total Lim R Linear Fit In POAAppm -1.8235 + 0.01742 Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) YRSCHEM 0.047702 0.002355 1.079651 -1.56794 23 Analysis of Variance DF Sum of Squares 1 1.226171 21 24.478567 22 25.704738 Mean Square 1.22617 1.16565 F Ratio 1.0519 Prob>F .0.3167 Estimate -1.823526 0.0174176 Parameter Std Error 0.335826 0.016982 Estimates t Ratio -5.43 1.03 Prob>ltl <.0001 0.3167 Lower 95% -2.521909 -0.017899 Upper 95% -1.125143 0.0527339 Random Sample Engineer/Lab InPFOSAA ppm By YRSCHEM APPendix F Page 19 -5.00 -7.00 0 5 to is 20 25 30 35 IMSCHEM Lim Term Intercept YRSCHEM Unear Fft InPFOSAAdfppm = -5-15- 0.00367 YRSCEIEM Summary ofFit RSquare 0.001234 RSquare Adj -0.04633 Root Mean SquareError 1.44825 Mean ofResponse -5.2038 Observations(orSum Wgts) 23 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.054411 0.05441 21 44.045995 2.09743 22 44.100406 F Ratio 0.0259 Prob>F 0.8736 Estimate -5.149964 -0.003669 Parameter Estimates Std Error tRatio 0.450479 -11.43 0.02278 -0.16 Prob>ltl <.0001 0.8736 Lower 95% -6.086779 -0.051043 Upper 95% -4.213149 0.0437044 Random Sample Engineer/Lab Ln M570ppm By YRSCHEM AppendixF Page 20 -J.DD 30 35 10 15 20 25 YRSCH(M = Lim Term Intercept YRSCHEM LinearFit In570ppm -3.0598+ 0.00297YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareEffor Mean ofResponse 0.002026 -0.0455 0.915637 -3.01612 Observation(sorSum Wgts) 23 Source Model Error C Total AnalysiosfVariance DF Sum ofSquares Mean Square 1 0.035747 0.035747 21 17.606219 0.838391 22 17.641966 F Ratio 0.0426 Prob>F 0.8384 Estimate -3.059762 0.0029739 ParameterEstimates StdError tRatio 0.284809 -10.74 0.014402 0.21 Prob>ltl <.0001 0.8384 Lower95% -3.652051 -0.026977 Upper 950/o -2.467473 0.0329252 RandomSample Engineer/Lab InPFOSAppmByYRSCHEM -I.oo Appendix F Page 21 Term Intercept YRSCHEM -7.00 -8.00 0 5 10 15 2D 25 30 3S YRSCHEM Source Model Error C Total = Lim Linear In PFOSAdfppm = -5.5202 RSquare Summary RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) Fit + 0.00865 of Fit YRSCHEM 0.004124 -0.0433 1.864648 -5.39325 23 Analysis of Variance DF Sum of Squares 1 0.302390 21 73.015154 22 73.317544 Mean Square 0.30239 3.47691 F Ratio 0.0870 Prob>F 0.7710 Estimate -5.520173 O.OOW96 Parameter Estimates Std Error t Ratio 0.58 -9.52 0.02933 0.29 Prob>ltl <.0001 0.7710 Lower 95% -6.726339 -0.052345 Upper 95% -4.314007 0.0696438 -2 00 -2 so -J.00 Random Sample Engineer/Lab InM556 ppm By YRSCHEM -1.00 APPendix F Page 22 -6.00 -6.$D 0 5 10 15. .20. .25. .30. .JS. . Term Intercept YRSCHIEM Lim LinearFit InM556dfppm -4-7931 + 0.00973 YRSCHEM Summary ofFit RSquar6 RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.013025 -0.03397 1.174741 4.65037 23 Source DF Model 1 Error 21 C Total 22 Estimate -4.793115 0.0097276 Analysisof Variance Sum ofSquares 0.382462 28.980362 29.362824 Mean Square 0.38246 1.38002 F Ratio 0.2771 Prob>F 0.6041 Parameter Estimates Std Error t Ratio 0.365404 -13.12 0.018478 0.53 Prob>ltl <.0001 0.6041 Lower 95% -5.553008 -0.028699 Upper 95% -4.033222 0.0481544 APPendix G Page I Appendix G Scatterplotsand regressionequationsfor fluorochemicalsby years worked in chemical(YRSCHEM) forallchemical participant(sn = 187) forcurrentjob categories(cell operators,chemical operators,engineer/lab,maintenance, milloperatorsand supervisor/mgmt) AllParticipants PFOS ppm By YRSCHEM 1?.OD 10,ID APPendix G Page 2 D.DD -IT 0 5 1D 15 20 25 JD 35 iD iscmal Term lnmmept YRSCHEM um fi LinearFit PFOSdfppm 0.87788+ 0.04433YRSCHEM Summary ofFit RSquare RSquare Adj 0.109673 0.10486 Root Mean SquareError 1.424349 Mean of Response 1.424443 Observation(sorSum Wgts) 187 Source Model Error C Total AnalysisofVariance DF Sum ofSquares K4ean Square 1 46.23325 46.2333 185 375.32259 2.0288 186 421.55584 F Ratio 22.7888 Prob>F <.0001 Estimate 0.8778797 0.0443319 ParameterEstimates Std Error tRaflo 0.154783 5.67 0.009287 4.77 Prob>ltl <.0001 <.0001 Lower 95% 0.5725098 0.0260105 Upper 95% 1.1832495 0.0626534 2.00 AllParticipants PFHS ppm ByYRSCHEM APPeridixG Page 3 ki.00 Term Intercept YRSCIREM one am 0 5 10 15 20 25 JO JS to 'ISCREM Source Model Error C Total = Lim Linear Fit PFHSdfppm 0.12463 + 0.01594 Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) YRSCHEM 0.210847 0.206581 0.347846 0.321211 187 Analysis of Variance DF Sum of Squares 1 5.980711 185 22.384463 186 28.365174 Mean Square 5.98071 0.12100 F Ratio 49.4286 Prob>F <.0001 Estimate 0.1246314 0.0159447 Parameter Std Error 0.0378 0.002268 Estimates t Ratio 3.30 7.03 Prob>ltl 0.0012 <.0001 Lower 95% 0.0500559 0.0114703 Upper 95% 0.1992071 0.020419 AllParticipants POAA ppm By YFISCHEM 1.00 6.00 t.00 dim AppendixG Page4 5 10 15 20 25 JO 35 40 Lim Term Intercept YRSCHEM LinearFit POAAppm 1.20809+ 0.01788YRSCHEM Summary o.fFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.024711 0.019439 1.266529 1.42851 187 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 7.51900 7.51900 185 296.75766 1.60410 186 304.27666 F Ratio 4.6874 Prob>F 0.0317 Estimate 1.2080941 0.017878 Parameter Estimates Std Error t Ratio 0.137633 8.78 0.008258 2.17 Pmb>ltl <.0001 0.0317 Lower 95% 0.9365598 0.0015867 Upper 95% 1.4796284 0.0341694 AllParticipants PFOSAA ppm ByYRSCHEM 0.20 APPendix G Page 5 Term Intercept YRSCHEM one 0.00 7 --T- 7 0 5 ID 15 20 25 30 35 lo ISCHEM Source Model Error C Total Liw Linear PFOSAAdfppm = 0.03463 Summary RSquare RSquam Adj Root Mean Square Effor Mean of Response Observations (or Sum Wgts) Fit - 0.00084 of Fit YRSCHEM 0.052504 0.047383 0.04017 0.024293 187 Analysis of Variance DF Sum of Squares .1 0.01654217 185 0.29852064 186 0.31506281 Mean Square 0.016542 0.001614 F Ratio 10.2516 Prob>F 0.0016 Estimate 0.0346317 -0.000839 Parameter Std Error 0.004365 0.000262 Estimates t Ratio 7.93 -3.20 Prob>14 <.0001 0.0016 Lower 95% 0.0260196 -0.001355 Upper 95% 0.0432438 -0.000322 AllParbcipants M570 ppm By YRSCHEM Appendix G Page 6 l'OD IL Term In"t YRSCHEM 10 15 2D 25 3D 35 40 YRSCHN Lim Source Model Error C Total M570ppm LinearFit 0.1882 - 0.00247YRSCHEM RSquare Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.009497 0.004143 0.283957 0.157804 187 AnalysisofVariance DF Sum ofSquares Mean Square 1 0.143025 0.143025 185 14.916841 0.080632 186 15.059866 F Ratio 1.7738 Prob>F 0.1845 Estimate 0.188204 -0.002466 Parameter Estimates Std Error t Ratio 0.030857 6.10 0.001851 -1.33 Prob>ltl <.0001 0.1845 Lower 95% 0.1273257 -0.006118 Upper 95% 0.2490822 0.0011868 AlPlarticipants PFOSAppm ByYASCHEM 0,60 APPendixG Page 7 Term Intercept YRSCHEM 0.20 0.00 0 5 10 is 20 25 Jo is 10 Lw Source Model Error C Total Unear PFOSAdfppm 0.05391 RSquare Summary RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) Fit - 0.00022 of Fit DF 1 185 186 Estimate 0.0539099 -0.000216 Analysis of Variance Sum of Squares 0.0010985 2.1279346 2.1290331 Parameter Std Error 0.011655 0.000699 Estimates t Ratio 4.63 -0.31 YRSCBEM 0.000516 -0.00489 0.107249 0.051246 187 Mean Square 0.001098 0.011502 F Ratio 0.0955 Prob>F 0.7576 Prob>ltl <.0001 0.7576 Lower 95% 0.0309165 -0.001596 Upper 95% 0.0769033 0.0011635 AllParticipants M556 ppm By YRSCHEM 0.50 o.ts 0.30 0.25 0.20 0.15 o.io 0.05 0.00 0 .....I.L.... ..-Romeo -...go on -1 . F- * I- . T - I 5 lo is 20 25 30 35, ' 40 Appendix G Page 8 Term Intercept YRSCHEM Source Model Error C Total LinearFit M556dfppm 0.05481- 0.00053YRSCBEM Summary ofFit RSquare 0.007159 RSquare Adj 0.001763 Root Mean SquareError 0.07024 Mean ofResponse 0.048273 Observations(orSum Wgts) 186 AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00654604 0.006546 184 0.90780250 0.004934 185 0.91434853 F Ratio 1.3268 Prob>F 0.2509 Estimate 0.0548148 -0.000528 Parameter Estimates Std Error t Ratio 0.007667 7.15 0.000459 -1.15 Prob>ltl <.0001 0.2509 Lower 95% 0.039688 -0.001434 Upper 95% 0.0699416 0.0003767 AllParticipants CellOperators PFOS ppm By YRSCHEM APPendix G Page 9 5.00 t.00 2.00 1.00 0.00 0 5 10 15 20 25 30 JS YRSEREM = Lim Term Intercept YRSCHEM Source Model Error C Total LinearFit PFOSdfppm 0.41242+ 0.09869YRSCHEM Summary of Fit RSquare 0.23418 RSquare Adj Root Mean SquareError Mean ofResponse 0.124777 1.814425 2.265556 Observations(orSum Wgts) 9 AnalysisofVariance DF Sum ofSquares Mean Square 1 7.046913 7.04691 7 23.044960 3.29214 8 30.091872 F Ratio 2.1405 Prob>F 0.1869 Estimate 0.4124178 0.0986878 Parameter Estimates Std Error t Ratio 1.403612 0.29 0.067453 1.46 Prob>ltl 0.7774 0.1869 Lower 95% -2.906623 -0.060815 Upper 95% 3.7314586 0.2581907 AllParticipants CellOperators PFHS ppm By YRSCHEM E Appendix c Page 10 0 5 10 15 20 25 JO JS YRSCHEM = Liu Term Intercept YRSCHEM LinearFit PFHSdfppm -0-0673+ 0.05293YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observation(sorSum Wgts) 0.573083 0.512095 0.464481 0.926611 9 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 2.0272533 2.02725 7 1.5101985 0.21574 8 3.5374519 F Ratio 9.3966 Prob>F Estimate -0.067334 0.052932 ParameterEstimates StdError tRatio 0.359316 -0.19 0.017268 3.07 Prob>lq 0.8567 0.0182 Lower 95% -0.916987 0.0121003 Upper 95% 0.7823194 0.0937637 AllParticipants CollOperators POAA ppm By YRSCHEM J.00 2.00 1.00 0.00 0 5 10 15 20 25 30 35 INSCRaf Lim Unear Fit POAAppm = 0.25794+ 0.08268YRSCHEM Summary of FN RSquare 0.423489 RSquare Adj 0.34113 Root Mean SquareError 0.980819 Mean ofResponse 1.810556 Observation(sorSum Wgts) 9 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 4.946633 4.94663 7 6.734042 0.96201 8 11.680674 F Ratio 5.1420 Prob>F 0.0577 Term Intercept YRSCHEM ParameterEstimates Estimate Std Error 0.257943 0.758747 0.0826835 0.036463 t Ratio 0.34 2.27 Prob>ltl 0.7439 0.0577 APPendix G Page AllParticipants CellOperators PFOSAA ppm By YRSCHEM 0.06 APPerldix G Page 12 0.02 0 5 10 15 20 25 30 35 Urem Unear Fit PFOSAAdfppm 0.03031- 0.00112YRSCBEM Summary ofFit RSquare 0.418421 RSquare Adj 0.335339 Root Mean SquareError 0.013461 Mean of Response 0.009223 Observation(sorSum Wgts) 9 Source Model Error C Total Analysisof Variance DF Sum ofSquares Mean Square 1 0.00091259 0.000913 7 0.00126844 0.000181 8 0.00218103 F Ratio 5.0362 Prob>F 0.0597 Term Intercept YRSCHEM ParameterEstimates Estimate StdError 0.0303118 0.010413 -0.001123 0.0005 tRatio 2.91 -2.24 Prob>ltl 0.0226 0.0597 AllParticipants CollOperators -MS70 ppm By YRSCHEM Apper)dixG Page 13 O.ID F I r 0 5 10 15 20 25 30 35 INSCHEW Term Intercept YRSCHEM = Lim LinearFit M570ppm = 0.10376 - 0.00314YRSCHEM Summary ofFit RSquare RSquam Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.380736 0.292269 0.040686 0.044833 9 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00712436 0.007124 7 0.01158772 0.001655 8 0.01871208 F Ratio 4.3037 Prob>F 0.0767 Estimate 0.1037558 -0.003138 Parameter Estimates Std Error tRatio 0.031474 3.30 0.001513 -2.07 Prob>ltl 0.0132 0.0767 Lower 95% 0.02933 -0.006715 Upper 950/o 0.1781816 0.0004388 AJIParticipants CellOperators PFOSA ppm By YRSCHEM APPendix G Page 14 0.01 0 5 10 15 20 25 30 35 YISCRIN Term Intercept YRSCHEM Lim LinearFit PFOSAdfppm 0.01002- 0.0002YRSCHEM Summary ofFit RSquare 0.152809 RSquare Adj 0.031782 Root Mean SquareError 0.004794 Mean ofResponse 0.006259 Observation(sorSum Wgts) 9 Source Model Error C Total Analysisof Variance DF Sum ofSquares Mean Square 1 0.00002902 0.000029 7 0.00016088 0.000023 8 0.00018990 F Ratio 1.2626 Prob>F 0.2982 Estimate 0.0100194 -0.0002 ParameterEstimates StdError tRatio 0.003709 2.70 0.000178 -1.12 Prob>ltl 0.0306 0.2982 Lower 95% 0.0012498 -0.000622 Upper 95% 0.018789 0.0002212 AllParticipants CellOperators M556 ppm By YRSCHEM 0.01 Apperidix G Page 15 r= 0.02 kn 0.00 0 5 10 15 20 25 30 35 )ISCHEM Lim R Unear Fit M556dfppm = 0.01826- 0.00025YRSCHEM Summary of Fit RSquare 0.04781 RSquare Adj -0.08822 Root Mean SquareError 0.011547 Mean of Response 0.013478 Observations(orSum Wgts) 9 Source Model Error C Total Term Intercept YRSCHEM AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00004686 0.000047 7 0.00093337 0.000133 8 0.00098024 F Ratio 0.3515 Prob>F 0.5719 Parameter Estimates Estimate 0.0182567 Std Error 0.008933 -0.000254 0.000429 t Rado 2.04 -0.59 Prob>ltl 0.0803 0.5719 Term Intercept YRSCHEM Apperidix G Page 16 AllParticipants ChemicalOperators PFOS ppm By YRSCHEM 7.00 6.00 S.DD 3.OD 2.00 1.OD 0 5 10 15 20 25 30 35 ISCHEW Source Model Error C Total = Lin Linear Fit PFOSdfppm 1.45105 + 0.03765 Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) YRSCHEM 0.070586 0.055596 1.25103 1.839062 64 Analysis of Variance DF Sum of Squares 1 7.36950 62 97.03475 63 104.40425 Mean Square 7.36950 1.56508: F Ratio 4.7087 Prob>F 0.0339 Estimate 1.4510518 0.0376538 Parameter Std Error 0.237545 0.017352 Estimates t Ratio 6.11 2.17 Prob>ltl <.0001 0.0339 Lower 95% 0.9762066 0.002967 Upper 95% 1.925897 0.0723406 AJIParUcipant3 Chemical Operators PFHS ppm By YRSCHEM Apperidix G Page 17 % 5 le is 20 25 30 is WSCHEM Term Intercept YRSCHEM = LimR Source Model Error C Total Linear Fit PFHSdfppm 0.17914 + 0.02247 YRSCHEM Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgu) 0.284349 0.272806 0.326413 0.410705 64 Analysis of Variance DF Sum of Squares Mean Square 1 2.6246883 2.62469 62 6.6058219 0.10655 63 9.2305102 F Ratio 24.6344 Prob>F <.0001 Estimate 0.1791447 0.0224713 Parameter Std Error 0.061979 0.004527 Estimates t Ratio 2.89 4.96 Prob>ltl 0.0053 <.0001 Lower 95% 0.0552502 0.013421 Upper 95% 0.3030391 0.0315217 AllParticipants ChemicalOperators POAA ppm By YRSCHEM 7.00 5.00 r= 4.00 @no. 0.00 10 15 20 25 30 35 APPendix G Page 18 Term Intercept YRSCHEM LinearFit POAAppm 1.71456+ 0.04674YRSCHEM Summary ofFit RSquare 0.101987 RSquare Adj 0.087503 Root Mean SquareError 1.269983 Mean of Response 2.196234 Observation(sorSum Wgts) 64 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 11.35666 11.3567 62 99.99717 1.6129 63 11l.35384 F Rabo 7.0413 Prob>F .0.0101 Estimate 1.7145638 0.0467429 ParameterEstimates Std Error tRatio 0.241143 7.11 0.017615 2.65 Prob>ltl <.0001 0.0101 Lower 95% 1.2325247 0.0115305 Upper 95% 2.1966029 0.0819552 AllParticipants ChemicalOperators PFOSAA ppm By YRSCHEM 0.30 APPendix G Page 19 D.ID 0.00 0 5 10 15 20 25 30 JS )ISCHEM Term Intercept YRSCBIEM Livff LinearFit PFOSAAdfppm 0.05584- 0.00136YRSCEFM Summary ofFit RSquare RSquare Adj Root Mean SquareError 0.042288 0.026841 0.059297 Mean of Response Observation(sorSum Wgts) 0.041812 64 Source DF Model 1 Error 62 C Total 63 Estimate 0.0558352 -0.001361 Analysisof Variance Sum ofSquares 0.00962588 0.21799991 0.22762579 Mean Square 0.009626 0.003516 F Ratio 2.7376 Prob>F 0.1031 ParameterEstimates Std Error tRatio 0.011259 4.96 0.000822 -1.65 Pmb>ltl <.0001 0.1031 Lower 95% 0.0333282 -0.003005 Upper 95% 0.0783421 0.0002833 AllParticipants Chemical Operators M570 ppm By YRSCHEM Appendix r. Page 20 2.00 1.50 1.00 0.50 0.00 0 5 10 15 20 25 30 JS ,mscmai Term Intercept YRSCHEM = Lim R j LinearFit M570ppm = 0.37266- 0.00856YRSCHEM Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.031978 0.016364 0.431404 0.28442 64 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.381171 0.381171 62 11.538802 0.186110 63 11.919973 F Ratio 2.0481 Prob>F 0.1574 Estimate 0.3726642 -0.008563 Parameter Estimates Std Error tRatio 0.081915 4.55 0.005994 -1.43 Prob>ltl <.0001 0.1574 Lower 95% 0.208919 -0.020525 Upper 95% 0.5364095 0.0033979 AllParticipants Chemical Operators PFOSA ppm By YRSCHEM APPendix G Page 21 c- o.io 0.20 0.10 0.00 0 5 10 15 20 25 JO 35 IRSCROI Term Intercept YRSCHEM LM R Source Model Error C Total LinearFit PFOSAdfppm 0.10868 - 0.00198YRSCHEM Summary of Fit RSquare RSquara Adj Root Mean Square Error Mean ofResponse 0.018323 0.002489 0.132746 0.088272 Observations(orSum Wgts) 64 Analysisof Variance DF Sum of Squares Mean Square 1 0.0203920 0.020392 62 1.0925409 0.017622 63 1.1129329 F Ratio 1.1572 Prob>F Estimate 0.1086821 -0.001981 Parameter Estimates Std Error t Ratio 0.025206 4.31 0.001841 -1.08 Prob>ltl <.0001 0.2862 Lower 95% 0.0582964 -0.005661 Upper 95% 0.1590678 0.0016999 AllParticipants ChemicalOperators M556 ppm By YRSCHEM Appendix G Page 22 E now 10 15 20 25 30 is 'OSCHEW Tenn Intercept YRSCHEM Lim LinearFit M556df@pm 0.09703- 0.00222YRSCHEM Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgm) 0.072552 0.057593 0.072643 0.074167 64 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.02559413 0.025594 62 0.32717681 0.005277 63 0.35277094 F Ratio 4.8501 Prob>F .0.0314 Estimate 0.0970334 -0.002219 ParameterEstimates Std Error tRatio 0.013793 7.03 0.001008 -2.20 Prob>ltl <.0001 0.0314 Lower 95% 0.0694607 -0-004233 Upper 95% 0.1246062 -0.000205 AllParticipants Engineer/Lab PFOS ppm By YRSCHEM AppendiX G Page 23 E 0.50 0.00 0 5 ID 15 20 25 30 is to Lim Term Intercept YRSCHEM LinearFit PFOSdfppm 0.36243+ 0.01624YRSCHEM RSquare Summary ofFit RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.158643 0.134604 0.504597 0.611027 37 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 1.680348 1.68035 35 8.911651 0.25462 36 10.591999 F Ratio 6.5995 Prob>F 0.0146 Estimate 0.3624261 0.016237 Parameter Estimates Std Error tRatio 0.127461 2.84 0.00632 2.57 Prob>ltl 0.0074 0.0146 Lower 95% 0.1036677 0.0034058 Upper 95% 0.6211844 0.0290681 1.00 AllParticipants Engineer/Lab PFHS ppm ByYRSCHEM 0.80 0.50 0.10 APPerldixG Page 24 0.00 7 0 5 10 15 20 25 30 35 to YRSCREM ----------------- Lim Term Intercept YRSCBEM LinearFit PFHSdfppm 0.08056+ 0.00441YRSCBEM RSquare RSquare Adj Summary ofFit 0.091924 0.065979 Root Mean SquareError 0.186981 Mean ofResponse Observation(sorSum Wgts) 0.148053 37 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.1238708 0.123871 35 1.2236624 0.034962 36 1.3475332 F Ratio 3.5430 Prob>F 0.0681 Estimate 0.0805558 0.0044085 ParameterEstimates StdError tRatio 0.047231 1.71 0.002342 1.88 Prob>ltl 0.0970 0.0681 Lower 95% -0.015328 -0.000346 Upper 950/a 0.1764397 0.0091631 AllParticipants Engineer/Lab POAAppm By YRSCHEM J.00 AppendixG Page 25 2.00 1. R 1.00 0.50 0.00 ic is 20 25 30 35 10 Term Intercept YRSCHEM um ri j LinearFit POAAppm 0.30344+ 0.00257YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.005873 -0.02253 0.45092 0.342765 37 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.0420387 0.042039 35 7.1165096 0.203329 36 7.1585483 F Ratio 0.2068 Prob>F 0.6521 Estimate 0.3034436 0.0025682 Parameter Estimates Std Error t Ratio 0.113902 2.66 0.005648 0.45 Prob>ltl 0.0116 0.6521 Lower 95% 0.0722111 -0.008898 Upper 950/o 0.534676 0.0140344 AllParUcipants Engineer/Lab PFOSAA ppm By YRSCHEM D.Di E 0.04 O.Di Appendix G Page 26 O.Dv 0 5 10 15 20 25 jo 35 fo Term Intercept YRSCHEM Source Model Error C Total Lim Linear Fit PFOSAAdfppm = 0.01271 - 0.0002 Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) YRSCHEM 0.026565 -0.00125 0.016393 0.009642 37 DF 1 35 36 Estimate 0.0127146 -0.000201 Analysis of Variance Sum of Squares 0.00025666 0.00940502 0.00966167 Parameter Std Error 0.004141 0.000205 Estimates t Ratio 3.07 -0.98 Mean Square 0.000257 0.000269 F Ratio 0.9551 Prob>F 0.3351 Prob>ltl 0.0041 0.3351 Lower 95% 0.0043085 -0-000618 Upper 95% 0.0211207 0.0002162 AllParticipants Engineer/Lab M570 ppm By YRSCHEM E Appendix G Page 27 a 5 10 15 ZO 25 jo 35 40 YRSCIBI Term Int=ept YRSCBEM Lim R Source Model Error C Total LinearFit M570ppm = 0.06607- 0.00016YRSCHEM RSquare Summary of Fit RSquare Adj Root Mean Square ErTor Mean ofResponse Observations(orSum Wgts) 0.000864 -0-02768 0.075116 0.063565 37 AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00017081 0.000171 35 0.19748202 0.005642 36 0.19765282 F Ratio 0.0303 Prob>F Estimate 0.0660713 -0.000164 Parameter Estimates Std Error t Ratio 0.018974 3.48 0.000941 -0.17 Prob>ltl 0.0014 0.8629 Lower 95% 0.0275519 -0-002074 Upper 95% 0.10459(Y7 0.0017464 AllParticipants Engineer/Lab PFOSA ppm By YRSCHEM Appendix G Page 28 D.05 E D.D2 0 1; 15' 20 25' 30' 35' 40' Lim Term Intercept YRSCBEM LinearFit PFOSAdfppm 0.01025+ 0.00007YRSCHEM Summary of Fit RSquare RSquare Adj RootMean Sqila Error Mean of Response 0.002597 -0.0259 0.019338 0.011372 Observation(sorSum Wgts) 37 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00003408 0.000034 35 0.01308800 0.000374 36 0.01312208 F Ratio 0.0911 Prob>F 0.7645 Estimate 0.0102528 0.0000731 ParameterEstimates Std Error tRatio 0.004885 2.10 0.000242 0.30 Prob>ltl 0.0431 0.7645 Lower 95% 0.0003365 -0.000419 Upper 95% 0.0201692 0.0005649 AllParticipants Engineer/Lab M556 ppm By YRSCHEM APpendiX G Page 29 0.20 0.15 0.10 o.os 0.00 0 5 10 15 20 25 JD 35 10 )ISCHIV Term Intercept YRSCHEM Source Model Error C Total Unear Fit M556dfppm 0.03151 - 0.00049 Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations (or Sum Wgts) YRSCHEM 0.013353 -0.01494 0.056294 0.024078 37 Analysis of Variance DF Sum of Squares 1 0.00150111 35 0.11091707 36 0.11241818 Mean Square 0.001501 0.003169 F Ratio 0.4737 Prob>F 0.4958 Estimate 0.0315087 -0.000485 Parameter Std Error 0.01422 0.000705 Estimates t Ratio 2.22 -0.69 Pmb>ltl 0.0333 0.4958 Lower 95% 0.0026409 -0.001917 Upper 95% 0.0603766 0.0009462 AJIParticipants Maintenance PFOS ppm By YRSCHEM APpendiX G Page 30 I.DD 0.00 5 10 15 20 25 30 35 IISCREM Term Intercept YRSCHEM Lim R LinearFit PFOSdfppm = 1.03905+ 0.07695YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observation(sorSum Wgts) 0.332684 0.288196 1.203104 1.772294 17 Source Model Error C Total Analysisof Variance DF Sum ofSquares Mean Square 1 10.824249 10.8242 15 21.711881 1.4475 16 32.536130 F Ratio 7.4781 Prob>F 0.0154 Estimate 1.0390494 0.0769454 ParameterEstimates Std Error tRatio 0.396284 2.62 0.028138 2.73 Prob>ltl 0.0192 0.0154 Lower 95% 0.1943936 0.0169718 Upper 95% 1.8837052 0.1369191 4 AllParticipants Maintenance PFHS ppm By YRSCHEM Appendix G Page 31 I.OD 0 5 ID 15 20 25 30 is ISCHIM Term Intercept YRSCHEM Lim R LinearFit PFHSdfppm 0.07257+ 0.02482YRSCHEM Summary ofFit RSquare 0.551529 RSquare Adj Root Mean SquareError 0.521631 0.247055 Mean ofResponse 0.309053 Observation(sorSum Wgts) 17 Source Model Error C Total Ana"s ofVariance DF Sum ofSquares Mean Square 1 1.1259357 1.12594 15 0.9155453 0.06104 16 2.0414810 F Ratio 18.4470 Prob>F .0.0006 Estimate 0.0725662 0.0248165 ParameterEstimates StdError tRatio 0.081376 0.89 0.005778 4.29 Prob>ltl 0.3866 0.0006 Lower 95% -0.100882 0.012501 Upper 95% 0.2460149 0.037132 AllParticipants Maintenance POAA ppm By YRSCHEM Appendix G Page 32 ?.00 1.00 0.00 0 5 10 15 20 25 30 J5 Term Intercept YRSCHEM Lm ri Source Model Error C Total LinearFit POAAppm = 0.92588+ 0.06146YRSC]HEM Summary ofFit RSquare 0.221317 RSquare Adj 0.169404 Root Mean Square Error Mean of Response 1.272661 1.511529 Observations(orSum Wgts) 17 AnalysisofVariance DF Sum ofSquares Mean Square 1 6.905100 6.90510 15 24.295008 1.61967 16 31.200108 F Ratio 4.2633 Prob>F 0.0567 Estimate 0.9258836 0.0614567 Parameter Estimates Std Error t Ratio 0.419195 2.21 0.029764 2.06 Prob>ltl 0.0432 0.0567 Lower 95% 0.0323938 -0.001984 Upper 96% 1.8193733 0.1248977 AllParticipants Maintenance PFOSAA ppm By YRSCHEM 0.10 Apperidix G Page 33 0.02 0.00 0 5 10 is 20 25 30 35 )ISCHEN um Term Intercept YRSCHEM LinearFit PFOSAAdfppm 0.04575- 0.00095YRSCHEM Summary ofFit RSquare 0.064237 RSquare Adj 0.001853 Root Mean SquareError 0.040022 Mean of Response 0.036697 Observation(sorSum Wgts) 17 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00164936 0.001649 15 0.02402685 0.001602 16 0.02567621 F Ratio 1.0297 Prob>F 0.3263 Esfimate 0.0457483 -0.00095 ParameterEstimates Std Error tRatio 0.013183 3.47 0.000936 -1.01 Prob>ltl 0.0034 0.3263 Lower 95% 0.01765 -0.002945 Upper 95% 0.0738465 0.0010453 AllParticipants Maintenance M570 ppm By YASCHEM 0.10 Apperidix G Page 34 0.20 0.10 0.00 0 5 10 15 20 25 30 35 lescmoi I Term lnwrcept YRSCBEM LinearFft M570ppm = 0.21068+ 0.00273YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.03116 -0.03343 0.168111 0.236706 17 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.01363420 0.013634 15 0.42392181 0.028261 16 0.43755601 F Rabo 0.4824 Prob>F 0.4979 Estimate 0.2106824 0.0027309 Parameter Estimates Std Error t Ratio 0.055373 3.80 0.003932 0.69 Prob>ltl 0.0017 0.4979 Lower 95% O.M6575 -0.005649 Upper 95% 0.3287074 0.0111111 AllParticipants Maintenance PFOSA ppm By YRSCHEM 0.60 0.50 o,4D o.io 0.20 O.lD 0.00 0 s 10 15 20 25 30 35 YRSCIDI Lim R Unear Fit PFOSAdfppm 0.05937+ 0.00069YRSCHEM Summary of Fit RSquare 0.00258 RSquare Adj -0.06391 Root Mean SquareEffor 0.149807 Mean of Response 0.065944 Observation(sorSum Wgts) 17 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00087072 0.000871 15 0.33663010 0.022442 16 0.33750082 F Ratio 0.0388 Prob>F 0.8465 Term Intercept YRSCHEM ParameterEstimates Estimate Std Error 0.0593671 0.049344 0.0006901 0.003504 tRatio 1.20 0.20 Prob>ltl 0.2476 0.8465 Appendix G Page 35 AllParticipants Maintenance M556 ppm By YRSCHEM 0.10 D.30 0.20 APPendix G Page 36 0 5 10 Is 20 25 30 is Term Intercept YRSCHEM um LinearFit M556dfppm 0.07814 + 0.001YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.011725 -0.05416 0.101656 0.0877 17 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.00183898 0.001839 is 0.15500978 0.010334 16 0.15684876 F Ratio 0.1780 Prob>F .0.6791 Estimate 0.(Y781426 0.0010029 Parameter Estimates Std Error tRatio 0.033494 2.33 0.002377 0.42 Prob>ltl 0.0339 0.6791 Lower 95% 0.0067735 -0.004065 Upper 95% 0.1495118 0.006(Y704 12.00 AllParticipants Supervisors/Mgmt PFOS ppm By YRSCHEM 10.00 8.00 6.00 ;.oo 2.00 O.OD 0 5 10 15 20 25 30 35 10 )ISCHEM AppendixG Page 37 Term Intercept YRSCREM Source Model Error C Total Liw ri _j LinearFit PFOSdfppm = -0.2184 + 0.094 YRSCHEM Summafy of Fit RSquare RSquare Adj Root Mean Square Error Mean ofResponse Observations(orSum Wgts) 0.194588 0.161029 2.073295 1.732181 26 AnalysisofVariance DF Sum ofSquares Mean Square 1 24.92474 24.9247 24 103.16526 4.2986 25 128.09000 F Ratio 5.7984 Prob>F 0.0241 Estimate -0.218386 0.0940032 Parameter Estimates Std Error tRatio 0.906363 -0.24 0.039038 2.41 Prob>ltl 0.8116 0.0241 Lower 95% -2.089012 0.0134332 Upper 95% 1.6522407 0.1745732 AllParticipants Supervisors/Mgmt PFHS ppm By YRSCHEM Appendix G Page 38 0.00 0 1 7 --r T--l 5 10 15 20 25 JO 35 40 IRSCREN Lim Term Intercept YRSCHEM UnearFh PFHSdfppm 0.04486+ 0.01663YRSCHEM Summary ofFit RSquare O.l85Or7l RSquare Adj 0.151116 Root Mean SquareError 0.378289 Mean of Response 0.389914 Observations(orSum Wgts) 26 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.7799725 0.779972 24 3.4344674 0.143103 25 4.2144399 F Ratio 5.4504 Prob>F .0.0283 Estimate 0.0448618 0.016629 Parameter Estimates Std Error t Ratio 0.165373 0.27 0.007123 2.33 Prob>ltl 0.7885 0.0283 Lower 95% -0.296449 0.0019294 Upper 95% 0.3861726 0.0313297 AllParticipants SupeMsora/Mgmt POAA ppm ByYRSCHEM 500 t.00 3.00 2-OD 1.00 Appendix G Page 39 10 15 20 25 jo is ic YISCREM = Limri LinearFit POAAppm 0.17876+ 0.05352YRSCHEM Summary ofFft RSquare 0.193609 RSquareAdj RootMean SquareError 0.160009 1.184218 Mean ofResponse 1.2894 Observation(sorSum Wgts) 26 Source Model Error C Total AnalysiosfVariance DF Sum ofSquares Mean Square 1 8.080800 8.08080 24 33.656935 1.40237 25 41.737735 F Ratio 5.7622 Prob>F 0.0245 Term Intercept YRSCHEM Estimate 0.1787618 0.0535247 ParameterEsfimates StdError tRatio 0.517694 0.35 0.022298 2.40 Prob>ltl 0.7329 0.0245 Lower95% -0.889697 0.007505 Upper95% 1.2472201 0.0995444 AllParOcipant3 Supervisors/Mgmt PFOSAA ppm By YRSCHEM O.Di E 5f3 APPerldix G Page 40 0 5 10 15 20 25 30 35 40 liscHEW Term Intercept YRSCBEM Source Model Error C Total LinearFit PFOSAAdf@pm 0.00564+ 0.00016YRSCHEM Summary of Fit RSquare 0.023751 RSquare Adj -0.01693 Root Mean Square Error 0.011408 Mean ofIteWnse 0.009045 Observations (orSum Wgts) 26 Analysis of Variance DF Sum of Squares Mean Square 1 0.00007599 0.000076 24 0.00312335 0.000130 2.5 0.00319934 F Ratio 0.5839 Prob>F 0.4522 Estniate 0.0056392 0.0001641 Parameter Esfimates Std Error 0.004987 t Raflo 1.13 0.000215 0.76 Prob>ltl 0.2693 0.4522 Lower 95% -0.004654 -0.000279 Upper 95% 0.0159319 0.0006075 AllParticipants Supervisors/Mgmt M570 ppm By YRSCHEM Appendix G Page 41 Term Intercept YRSCHEM E 0.20 0.00 0 5 10 IS 20 25 30 35 40 YRSCREM Source Model Error C Total Lim R M570ppm RSquare RSquare Adj Root Mean Square Mean of Response Observations (or Unear 0.03728 + Summary Error Sum Wgts) Fit 0.00349 of Fit YRSCBEM 0.081466 0.043194 0.127205 0.109788 26 Analysis of Variance DF Sum of Squares 1 0.03444346 24 0.38834961 25 0.42279307 Mean Square 0.034443 0.016181 F Ratio 2.1286 Prob>F 0.1575 E3Umate 0.0372783 0.0034945 Parameter Std Error 0.055609 0.002395 Estimates t Raflo 0.67 1.46 Prob>ltl 0.5090 0.1575 Lower95% -0.077493 -0.001449 Upper 95% 0.1520492 0.0094378 AJIParticipants Supervisor4/Mgmt PFOSA ppm By YRSCHEM 0.80 0.60 o.to Appendix G Page 42 0.00 1 T I I 10 15 20 25 30 35 to Term Intercept YRSCHEM Lim R j LinearFit PFOSAdfppm -0.0169+ 0.00333YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.075283 0.036753 0.126709 0.052267 26 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.03137008 0.031370 24 0.38532631 0.016055 25 0.41669639 F Ratio 1.9539 Prob>F 0.1750 Estimate -0.016933 0.0033349 ParameterEstimates Std Error t Ratio 0.055392 -0.31 0.002386 1.40 Prob>ltl 0.7625 0.1750 Lower 95% -0.131256 -0.001589 Upper 95% 0.0973903 0.008259 AllParficipants SuperA.sors/Mgmt M556 ppm By YRSCHEM Appendix G Page 43 0.00 10 15 ZD 25 30 35 to )ISCHBI Term Intercept YRSCHEM = - LinearFit M556dfppm -0-0067+ 0.00235 YRSCHEM Summary of Fit RSquare 0.101017 RSquare Adj Root Mean SquareError 0.061931 0.072994 Mean of Response 0.04378 Observation(sorSum Wgts) 25 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.01377030 0.013770 23 0.12254662 0.005328 24 0.13631692 F Ratio 2.5845 Prob>F 0.1216 Estimate -0.006656 0.0023503 Parameter Estimates StdError tRatio 0.034603 -0.19 0.001462 1.61 Prob>ltl 0.8491 0.1216 Lower 95% -0.078239 -0.000674 Upper 95% 0.0649258 0.'0053745 AllParticipants MillOperators -PFOS ppm By YRSCHEM J.OD APPerldix G Page 44 D.DD 0 10 15 20 25 YRSCHEM AllParticipants MillOperators PFHS ppm By YRSCHEM 0.20 r 0.00 0 5 10 15 20 25 APPendix G Page 45 AllParticipants MillOperators POAA ppm By YRSCHEM 2.5 2.0 1.0 0 5 10 15 20 25 Appendix G Page 46 AllParticipants MillOperators PFOSAA ppm By YRSCHEM 0.02 0 D.01 0.00 D 5 ID 15 2D 25 APPerldix G Page 47 AllParticipants MillOperators M570ppm By YRSCHEM D.10 0.08 E 0.06 0.04 INSCHEM 20 25 Appendix r. Page 48 AllParticipants MillOperators PFOSA ppm By YRSCHEM 0.20 E C) 0.10 D.D5 lie. a T15 2D 25 YRSCHRI Appendix G Page 49 AllParticipants MillOperators M556 ppm By YRSCHEM Appendix G Page 50 10 15 20 25 IRSCHEM Appendix H Page 1 Appendix H Scatterplotasnd regressionequationsforfluorochemical(snaturalogtransformationb)y yearsworked inchemical(YRSCHEM) forallchemicalparticipan(tns= 187)and for two currentjob categorie(schemcialoperatorsand engineer/lab) AJIParticipants InPFOS ppm By YRSCHEM 2.00 i.oo ago o.oo 31 E2 1.00 a 0 5 10 15 20 25 io is to APPeridix H Page 2 Term Intercept YRSCHEM LinearFit InPFOSdfppm = -0.493+ 0.02935YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.098178 0.093304 1.002959 -0.13123 187 Source Model Error C Total Analysisof Variance DF Sum ofSquares Mean Square 1 20.25968 20.2597 185 186.09638 1.0059 186 206.35607 F Ratio 20.1403 Prob>F <.0001 Estimate -0.493042 0.0293465 Parameter Estimates Std Error t Ratio 0.108991 -4.52 0.006539 4.49 Prob>ltl <.0001 <.0001 Lower 95% -0.708069 0.0164454 Upper 95% -0.278015 0.0422476 AllParticipants InPFHS ppm By YRSCHEM 0.00 NJ Apperidix H Page 3 s io 15 20 25 30 35 1D YRSCHEM Term Intercept YRSCHEM Source Model Error C Total Lie%11 LinearFit InPFHSdfppm = -2-5211+ 0.05519YRSCHEM Summary ofFit RSquare 0.226387 RSquare Adj 0.222205 Root Mean SquareError 1.150356 Mean of Response -1.84074 Observation(sorSum Wgts) 187 AnalysisofVariance DF Sum ofSquares Mean Square 1 71.64131 71.6413 185 244-81410 1.3233 186 316.45541 F Ratio 54.1376 Prob>F <.0001 Estimate -2.52111 0.055185 Parameter Estimates Std Error t Ratio 0.125008 -20.17 0.0075 7.36 Prob>ltl <.0001 <.0001 Lower95% -2.767738 0.040388 Upper 950/c -2.274482 0.0699821 AllParficipants InPOAA ppm By YRSCHEM 2.00 1.00 APPendix H Page 4 2.00 son 0 5 10 15 20 25 lo is 4; lmscocw Term Intercept YRSCHEM Lieff Source Model Error C Total LinearFit InPOAAppm -0.3289+ 0.00943YRSCHEM Summary of Fit RSquare 0.006732 RSquare Adj Root Mean -SquareError Mean of Response 0.001363 1.291882 -0.21266 Observations(orSum Wgts) 187 AnalysisofVariance DF Sum of Squares Mean Square 1 2.09272 2.09272 185 308.75765 1.66896 186 310.85036 F Ratio 1.2539 Prob>F 0.2643 Estimate -0.328943 0.0094318 Parameter Estimates Std Error tRatio 0.140388 -2.34 0.008423 1.12 Prob>ll 0.0202 0.2643 Lower 95% @0-605913 -0-007186 Upper 95% -0-051973 0.0260493 AllParticipants InPFOSAA ppm By YFISCHEM -100 -2.00 -J.00 E APPendix H Page 5 -6.00 -7.oo ..... ..... 10 15 20 25 30 35 to ISCMEN Term Intercept YRSCHEM = - LinearFit InPFOSAAdfppm = -4.4107- 0.03347YRSCHEM Summary of Fit RSqlla 0.06056 RSquare Adj 0.055482 Root Mean SquareError 1.486622 Mean ofResponse -4-82337 Observation(sorSum Wgts) 187 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 26.35653 26.3565 185 408.85842 2.2100 186 435.21495 F Ratio 11.9258 Prob>F 0.0007 Estimate -4.410692 -0.033472 ParameterEstimates Std Error tRatio 0.16155 -27.30 0.009693 -3.45 Prob>ltl <.0001 0.0007 Lower 95% -4.729413 -0-052595 Upper 95% -4.091972 -0.01435 AllParticipants InM570ppm By YRSCHEM I.OD 0.00 Appendix H Page 6 0 5 10 15 20 25 30 35 40 YRSCHEM Term Intercept YRSCHEM Lim ft LinearFit In570ppm -2.4506- 0.00984YRSCHEM RSquare Summary of Fit RSquare Adj Root Mean Square Error Mean ofResponse Observations(orSum Wgts) 0.009202 0.003846 1.151312 -2.57193 187 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 2.27743 2.27743 185 245.22096 1.32552 186 247.49839 F Ratio 1.7181 Prob>F 0.1916 Estimate -2.450623 -0.009839 Parameter Estimates Std Error t Ratio 0.125112 -19.59 0.007506 -1.31 Prob>ltl <.0001 0.1916 Lower 95% -2.697456 -0.024649 Upper 95% -2.20379 0.0049701 AllParticipants InPFOSA ppm By YRSCHEM ApperidixH Page 7 E a -16 .S-6.00 -8.00 0 5 10 15 20 25 30 35 to ISCHEW Lim Term Intercept YRSCBEM LinearFit InPFOSAdfppm = -4.4035- 0.01431YRSCHEM RSquare RSquare Adj Summary ofFit 0.007678 0.002314 Root Mean Square Error 1.834221 Mean ofResponse Observations(orSum Wgts) -4.57987 187 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 4.81599 4.81599 185 622.40775 3.36437 186 627.22374 F Ratio 1.4315 Prob>F 0.2331 Estimate -4.403463 -0.014308 Parameter Estimates Std Error t Ratio 0.199323 -22.09 0.011959 -1.20 Prob>ltl <.0001 0.2331 Lower 95% -4.796706 -0.037902 Upper 95% -4.01022 0.0092855 AllParticipants InM556 ppm By YRSCHEM -I.DO Appendix H Page 8 .9-5.00 air 10 15 20 25 io is to liscREM Lim Term Intercept YRSCREM LinearFit InM556dfppm = -3.7337- 0.00771YRSCBEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.00447 -0-00094 1.298323 -3.82913 186 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 1.39270 1.39270 184 310.15820 1.68564 185 311.55090 F Ratio 0.8262 Prob>F 0.3646 Estimate -3.733702 -0.007709 Parameter Estimates Std Error t Ratio 0.141718 -26.35 0.008481 -0.91 Prob>ltl <.0001 0.3646 Lower 95% -4.013306 -0.024441 Upper 95% -3.454098 0.0090235 AllParticipants Chemical Operators InPFOS ppm By YRSCHEM AppendiX H Page 9 -1.00 -1.50 0 It 5 io 15 20 25 30 35 IMSCHBI Term Intercept YRSCHEM um r LinearFit InPFOSdfppm 0.19093 + 0.01954 YRSCHEM Summary ofFit RSquare 0.067842 RSquare Adj Root Mean SquareError Mean ofResponse 0.052807 0.663197 0.392284 Observations(orSum Wgts) 64 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 1.984648 1.98465 62 27.269497 0.43983 63 29.254145 F Ratio 4.5123 Pmb>F 0.0376 Estimate 0.1909269 0.0195403 Parameter Estimates Std Error t Ratio 0.125927 1.52 0.009199 2.12 Pmb>ltl 0.1346 0.0376 Lower 95% -0.060798 0.0011521 Upper 95% 0.4426522 0.0379285 AllParticipants Chemical Operators InPFHS ppm By YRSCHEM % EC- -1.5 Appendix H Page 10 0 5 10 15 20 25 30 is YRSCHDA Term Intercept YRSCBEM LinearFU InPFHSdfppm 1.7282+ 0.04829 YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean ofResponse Observations(orSum Wgts) 0.28459 0.273051 0.701066 -1.23054 64 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 12.122007 12.1220 62 30.472595 0.4915 63 42.594602 F Ratio 24.6636 Prob>F <.0001 Estimate -1.72818 0.0482922 Parameter Estimates Std Error tRatio 0.133118 -12.98 0.009724 4.97 Prob>ltl <.0001 <.0001 Lower 95% -1.994279 0.028854 Upper 95% -1.462081 0.0677304 AllParticipants Chemical Operators InPOAA ppm By YRSCHEM ? EC<C< Apperidix H Page .11 D 5 10 15 20 25 JD J5 YRSCHEM Lim Term Intercept YRSCHEM LinearFit InPOAAppm = 0.45333+ 0.01564YRSCHEM Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.051828 0.036535 0.612605 0.614523 64 Source Model Error C Total Analysisof Variance DF Sum ofSquares Mean Square 1 1.271823 1.27182 62 23.267630 0.37528 63 24.539453 F Ratio 3.3890 Prob>F 0.0704 Estimate 0.4533333 0.0156424 Parameter Esfimates Std Error t Ratio 0.116321 3.90 0.009497 1.84 Prob>ltl 0.0002 0.0704 Lower 95% 0.220811 -0.001343 Upper 95% 0.6858555 0.0326278 AllParticipants Chemical Operators InPFOSAA ppm By YRSCHEM APperidiX H Page 12 E < go's -7 0 5 to 15 20 25 30 J5 YISCREW Term Intercept YRSCBEM Lro ri Source Model Error C Total LinearFit InPFOSAAdfppm = -4.0532- 0.02268YRSCHEM RSquare Summary ofFit RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.016184 0.000316 1.619017 -4.28694 64 AnalysisofVariance DF Sum ofSquares Mean Square 1 2.67335 2.67335 62 162.51531 2.62121 63 165.18867 F Ratio 1.0199 Prob>F 0.3165 Estimate -4.053244 -0.022679 Parameter Estimates Std Error t Ratio 0.307418 -13-18 0.022456 -1.01 Prob>ltl <.0001 0.3165 Lower 95% -4.667764 -0.067569 Upper 95% -3.438725 0.0222112 AllParticipants Chemical Operators InM570ppm By YRSCHEM D- 16 E AppendiX H Page 13 0 5 10 is ?o 25 ic 3; YRSCHDA Term Intercept YRSCHEM Source Model Error C Total Linear Fit In570ppm 1.5009- 0.04316YRSCHEM RSquare Summary of Fit RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.10729 0.092891 1.139966 -1.94564 64 AnalysisofVariance DF Sum ofSquares Mean Square 1 9.683322 9.68332 62 80.570389 1.29952 63 90.253710 F Ratio 7.4514 Prob>F 0.0082 Estimate -1.500868 -0.043162 Parameter Estimates Std Error t Ratio 0.216456 -6.93 0.015812 -2.73 Prob>ltl <.0001 0.0082 Lower 95% -1.933557 -0.074769 Upper 95% -1.068179 -0.011555 AllParticipants Chemical Operators InPFOSA ppm By YASCHEM -2,00 J.00 E- APPendixH Page 14 0 5 10 15 20 25 30 35 YRSCHni Term Intercept YRSCHEM Source Model Error C Total LinearFit InPFOSAdfppm = -3.4933 - 0.03575 YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.030317 0.014677 1.851004 -3.8617 64 Analysisof Variance DF Sum ofSquares Mean Square 1 6.64144 6.64144 '62 212.42545 3.42622 63 219.06689 F Ratio 1.9384 Prob>F 0.1688 Estimate -3.49335 -0.035745 Parameter Estimates Std Error t Ratio 0.351467 -9.94 0.025674 -1.39 Prob>ltl <.0001 0.1688 Lower 95% -4.195923 -0.087068 Upper 950/o -2.790777 0.0155766 AllParticipants Chemical Operators InM556 ppm By YRSCHEM Appendix H Page 15 -5 0 5 10 15 20 25 JO J5 YRSCHEM Term Intercept YRSCHEM Lim ri Source Model Error C Total LinearFit InM556dfppm -2.7891- 0.03042YRSCREM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse 0.063375 0.048268 1.070608 -3.10248 Observations(orSum Wgts) 64 AnalysisofVadance DF Sum ofSquares Mean Square 1 4.808428 4.80843 62 71-064468 1.14620 63 75.872896 F Ratio 4.1951 Prob>F 6.0448 Estimate -2.789057 -0.030415 Parameter Estimates Std Error t Ratio 0.203286 -13.72 0.01485 -2.05 Prob>ltl <.0001 0.0448 Lower 95% -3.19542 -0.0601 Upper 95% -2.382693 -0.000731 AJIParticipants Engineer/Lab InPFOS ppm By YRSCHEM 00 APPendix H Page 16 -2.50 0 5 10 15 20 25 30 35 Term Intercept YRSCHEM Lm R LinearFit InPFOSdfppm 1.5212+ 0.03794YRSCHEM Summary of Fit RSquare 0.250379 RSquare Adj 0.228961. Root Mean SquareError 0.885804 Mean of Response -0.94033 Observations(orSum Wgts) 37 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 9.172752 9.17275 35 27.462686 0.78465 36 36.635438 F Ratio 11.6903 Prob>F 0.0016 Estimate -1.521161 0.0379363 Parameter Estimates Std Error tRatio 0.223754 -6.80 0.011095 3.42 Prob>ltl <.0001 0.0016 Lower95% -1.975403 0.0154116 Upper 96% -1.06692 0.060461 AllParticipants Engineer/Lab InPFHS ppm By YRSCHEM -100 -2.00 97 ii-i Do 1.00 Apperidix H Page 17 10 15 20 25 jo J5 ISCHBI Term Intercept YRSCREM Lm ri Linear Fit In PFHSdfppm -3.3667+ 0.05024YRSC]HEM Summary of Fit RSquare RSquare Adj Root Mean Square Error Mean ofResponse Observations(orSum Wgts) 0.291558 .0.271317 1.056855 -2.5975 37 Source Model Error C Total Analysisof Variance DF Sum of Squares Mean Square 1 16.088669 16.0887 35 39-092987 1.1169 36 55.181656 F Ratio 14.4042 Prob>F 0.0006 Estimate -3.366746 0.0502418 Parameter Estimates Std Error 0.266961 tRatio -12.61 0.013238 3.80 Prob>ltl <.0001 0.0006 Lower95% -3.908703 0.0233675 Upper 95% -2.824789 0.0771161 AllParticipants Engineer/Lab InPOAA ppm By YRSCHEM OD Appendix ii Page 18 2.00 -4.DD 0 5 10 15 ZO 25 30 is ISCREM Term Intercept YRSCHEM = Limri LinearFit InPOAAppm -2.0155+ 0.02576YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response 0.107494 0.081994 1.001589 -1.62112 Observation(sorSum Wgts) 37 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 4.228818 4.22882 35 35.111285 1.00318 36 39.340103 F Ratio 4.2154 Prob>F 0.0476 Estimate -2.015494 0.0257581 ParameterEstimates Std Error tRatio 0.253001 -7.97 0.012546 2.05 Prob>ltl <.0001 0.0476 Lower 95% -2.52911 0.0002892 Upper 95% -1.501878 0.0512271 AllParticipants Engineer/Lab InPFOSAA ppm By YRSCHEM APPendix H Page 19 E < -7.00 0 5 10 15 20 25 30 is YRSCREW Term Intercept YRSCHEM LinearFit InPFOSAAdfppm = -5.5422- 0.00745YRSCBEM Summary ofFit RSquare RSquare Adj Root Mean SquareError 0.005419 -0.023 1.362132 Mean of Response Observation(sorSum Wgts) -5.65628 37 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.353853 0.35385 35 64-939165 1.85540 36 65.293017 F Ratio 0.1907 Prob>F 0.6650 Estimate -5.542201 -0.007451 ParameterEstimates Std Error tRatio 0.344074 -16.11 0.017062 -0.44 Prob>ltl <.0001 0.6650 Lower 95% -6.240704 -0.042088 Upper 95% -4.843697 0.027186 AllParticipants Engineer/Lab InM570ppm By YRSCHEM Appendix H Page 20 -1.00 -5.00 0 10 15 20 25 30 35 YRSCH(M = Lin Term Intercept YRSCHEM Unear Fit In570ppm -3.3086+ 0.00738YRSCBEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.010486 -0.01779 0.967617 -3.19558 37 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.347256 0.347256 35 32.769885 0.936282 36 33.117141 F Ratio 0.3709 Prob>F '0.5465 Estimate -3.308597 0.0073812 Parameter Estimates Std Error t Ratio 0.24442 -13.54 0.01212 0.61 Prob>ltl <.0001 0.5465 Lower 95% -3.804792 -0.017224 Upper 95% -2.812402 0.0319863 AllParticipants Engineer/Lab InPFOSA ppm By YRSCHEM 3,)D ao APPendiX H Page 21 %--5.Do 7.00 10 15 20 25 30 JS Term Intercept YRSCHIEM Lim r Source Model Effor C Total LinearFit InPFOSAdfppm = -5.8688+ 0.0086YRSCBEM Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean ofResponse Observations(orSum Wgts) 0.005342 -0.02308 1.583297 -5.7372 37 AnalysisofVariance DF Sum ofSquares K4ean Square 1 0.471187 0.47119 35 87.739016 2.50683 36 88.210203 F Ratio 0.1880 Prob>F 0.6673 Estimate -5-868846 0.0085981 Parameter Estimates Std Error t Ratio 0.39994 -14.67 0.019832 0.43 Prob>ltl <.0001 0.6673 Lower 95% -6.680763 -0.031663 Upper 95% -5.056929 0.048859 AllParticipants Engineer/Lab InM556 pm By YRSCHEM -J.DD Appendix H Page 22 10 15 20 25 30 35 YRSCHIM Tenn Intercept YRSCHEM um Ft Source Model Error C Total Linear Fit InM556dfppm -4.7368+ 0.00765 YRSCHEM RSquare Summary ofFit RSquare Adj Root Mean SquareEffor Mean ofResponse Observations(orSum Wgts) 0.00737 -0.02099 1.198158 4.61966 37 AnalysisofVariance DF Sum ofSquares Mean Square 1 0.373036 0.37304 35 50.245432 1.43558 36 50.618468 F Ratio 0.2598 Prob>F 0.6134 Estimate 4.736793 0.0076503 Parameter Estimates Std Error 0.302654 tRatio -15.65 0.015008 0.51 Prob>ltl <.0001 0.6134 Lower 95% -5.35121 -0.022817 Upper 95% -4.122375 0.0381178 Appendix i Page I Appendix I Random sample currentjob chemical operators(n = 47): Regressionof fluorochemicalon gender,yearsworked inchemical and age; followed by regressionequation of fluorochemicalon gender and yearsworked in chemical: Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.9796 Random Sample ChemicalOperators InPFOS ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.123153 0.061978 0.575509 0.392725 47 AnalysisofVariance DF Sum ofSquares Mean Square 3 2.000299 0.666766 43 14.242071 0.331211 46 16.242370 Lack ofFit DF Sum of Squares Mean Square 41 13.910223 0.339274 2 0.331849 0.165924 43 14.242071 F Ratio 2.0131 Prob>F 0.1263 F Ratio 2.0448 Prob>F 0.3833 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENDER YRSCHEM AGE Nparm 1 1 1 ParameterEstimates Estimate Std Error 0.0464337 0.412448 -0.20812 0.102882 0.0092881 0.010914 0.0029634 0.010819 EffectTest DF Sum of Squares 1 1.3553627 1 0.2398607 1 0.0248502 tRatio 0.11 -2.02 0.85 0.27 F Ratio 4.0921 0.7242 0.0750 Prob>ltl 0.9109 0.0493 0.3995 0.7855 Prob>F 0.0493 0.3995 0.7855 Appendix I Page 2 Apperidix i Page 3 Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.8122 Random Sample Chemical Operators RSquare InPFOS ppm Summary of Fit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.121623 0.081697 0.569428 0.392725 47 AnalysisofVariance DF Sum ofSquares Mean Square 2 1.975448 0.987724 44 14.266921 0.324248 46 16.242370 Lack ofFt DF Sum ofSquares Mean Square 26 11.215936 0.431382 18 3.050985 0.169499 44 14.266921 F Ratio 3.0462 Prob>F 0.0577 F Ratio 2.5450 Prob>F 0.0221 Term Intercept GENDERFF-M YRSCHEM Source GENDER YRSCBIEM Estimate 0.152705 -0.207949 0.0109494 Parameter Estimates Std Error t Ratio 0.138482 1.10 0.101793 -2.04 0.008978 1.22 Prob>@l 0.2762 0.0471 0.2291 Lower 95% -0.126387 -0.413098 -0.007145 Nparm 1 EffectTest DF Sum ofSquares 1 1.3531877 0.4822506 F Ratio 4.1733 1.4873 Prob>F 0.0471 0.2291 Upper 95% 0.431797 -0.0028 0.0290438 RandomSample Chemical Operators Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.9884 InPFHS ppm RSquare Summary ofFit RSquare Adj Root Mean SquareEffor Mean ofResponse Observation(sorSum Wgts) 0.47973 0.443432 0.584505 -1.17704 47 AnalysisofVariance DF Sum ofSquares Mean Square 3 13.546050 4.51535 43 14.690763 0.34165 46 28.236813 Lack ofFit DF Sum ofSquares Mean Square 41 14.362568 0.350307 2 0.328195 0.164098 43 14.690763 F Ratio 13.2165 Prob>F <.0001 F Ratio 2.1347 Prob>F 0.3797 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENDER YRSCHEM AGE Nparm 1 1 1 ParameterEstimates Estimate Std Error -1 869781 0.418894 -0*-347847 0.10449 0.047003 0.011085 -0-000586 0.010988 t Ratio -4.46 -3.33 4.24 -0.05 EffectTest DF Sum ofSquares 1 3.7861950 1 6.1426340 1 0.0009719 F Ratio 11.0822 17.9795 0.0028 Prob>ltl <.0001 0.0018 0.0001 0.9577 Prob>F 0.0018 0.0001 0.9577 Appendix I Page 4 Appendix I Page 5 Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.8933 Random Sample ChemicalOperators InPFHS ppm Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.479696 0.456046 0.577843 -1.17704 47 AnalysisofVariance DF Sum ofSquares Mean Square 2 13.545078 6.77254 44 14.691735 0.33390 46 28.236813 Lack of Fit DF Sum ofSquares Mean Square 26 11.677832 0.449147 18 3.013903 0.167439 44 '14.691735 F Ratio 20.2829 Prob>F <.0001 F Ratio 2.6825 Prob>F 0.0170 Term Intercept GENDER[F-M] YRSCHEM Source GENDER YRSCHEM Estimate -1-890797 -0-34788 ParameterEsomates Std Error tRatio 0.140529 -13.45 0.103297 -3.37 0.009111 5.12 Prob>ltl <.0001 0.0016 <.0001 Lower 95% -2.174014 -0.556062 0.0283126 Nparm 1 1 EffedTest DF Sum ofSquares 1 3.7870700 1 8.7629557 F Rano 11-3418 26.2440 Prob>F 0.0016 <.0001 Upper 95% -1.60758 -0.139699 0.0650363 Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.9950 RandomSample Chemical Operators InPOAA ppm RSquare Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.197237 0.14123 0.590899 0.635094 47 AnalysisofVariance DF Sum ofSquares Mean Square 3 3.688886 1.22963 43 15.013959 0.34916 46 18.702845 Lack ofFit DF Sum ofSquares Mean Square 41 14.919738 0.363896 2 0.094221 0.047111 43 15.013959 F Ratio 3.5217 Prob>F 0.0228 F Ratio 7.7243 Prob>F 0.1211 Term Intercept GENDER[F-M] YRSCHIEM AGE Source GENDER YRSCHEM AGE Nparm 1 1 1 Parameter Estimates Estimate Std Error 0.3398931 0.423477 -0.318303 0.105633 0.008909 0.011206 0.0003393 0.011108 EffectTest DF Sum ofSquares 1 3.1703565 1 0.2206817 1 0.0003258 tRatio 0.80 -3.01 0.80 0.03 F Ratio 9.0799 0.6320 0.0009 Prob>ltl 0.4266 0.0043 6.4310 0.9758 Prob>F 0.0043 0.4310 0.9758 ApperidixI Page 6 Appendix I Page 7 Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.8758 Random Sample ChemicalOperators InPOAA ppm Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.197219 0.160729 0.584152 0.635094 47 AnalysisofVariance DF Sum ofSquares Mean Square 2 3.688560 1.84428 44 15.014285 0.34123 46 18.702845 Lack ofFit DF Sum ofSquares Mean Square 26 12.692319 0.488166 18 2.321965 0.128998 44 15.014285 F Ratio 5.4047 Prob>F 0.0080 F Ratio 3.7843 Prob>F 0.0025 Term Intercept GEND,ER[F-M] YRSCHEM Source GENDER YRSCBEM Estimate 0.3520616 -0.318283 0.0090993 ParameterEstimates Std Error tRatio 0.142063 2.48 0.104425 -3.05 0.00921 0.99 Prob>ltl 0.0171 0.0039 0.3286 Lower 95% -0.0657529 -0.528737 -0-000463 Nparm 1 1 EffectTest DF Sum ofSquares 1 3.1700833 1 0.3330469 F Ratio 9.2901 0.9760 Prob>F 0.0039 0.3286 Upper 95% 0.6383704 -0-107829 0.0276616 Random Sample Chemical Operators Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.9297 InPFOSAA ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.094899 0.031753 1.570558 4.4836 47 Analysisof Variance DF Sum ofSquares Mean Square 3 11.12095 3.70698 43 106.06607 2.46665 46 117.18701 Lack ofFit DF Sum ofSquares Mean Square 41 97.82374 2.38594 2 8.24233 4.12116 43 106.06607 F Ratio 1.5028 Prob>F 0.2274 F Ratio 0.5789 Prob>F 0.8095 Term Intercept GEND.tR[F-M] YRSCHEM AGE Source GENDER YRSCHEM AGE Nparm 1 1 1 ParameterEstimates Estimate Std Error -5.751677 1.125565 -0-441575 0.280763 -0.042433 0.029785 0.0352765 0.029525 t Ratio -5.11 -1.@7 -1.42 1.19 EffectTest DF Sum ofSquares 1 6.1015043 1 5.0062176 1 3.5213341 F Ratio 2.4736 2.0296 1.4276 Prob>ltl <.0001 0.1231 0.1615 0.2387 Prob>F 0.1231 0.1615 0.2387 Appendix I Page 8 ApperidixI Page 9 Random Sample ChemicalOperators Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.6366 InPFOSAA ppm RSquare Summary of Fit RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.06485 0.022343 1.578171 -4.4836 47 Analysisof Variance DF Sum ofSquares Mean Square 2 7.59961 3.79981 44 109.58740 2.49062 46 117.18701 Lack ofFit DF Sum ofSquares Mean Square 26 66.99736 2.57682 18 42.59004 2.36611 44 109.58740 F Ratio 1.5256 Prob>F 0.2288 F Ratio 1.0891 Prob>F 0.4335 Term Intercept GENDER[F-M) YRSCBEM Source GENI)ER YRSCHEM Estimate 4.486637 -0.439541 -0.022657 ParameterEstimates StdError tRatio 0.383804 -11.69 0.282119 -1.56 0.024883 -0.91 Prob>ltl <.0001 0.1264 0.3675 Lower 95% -5.260141 -1.008113 -0.072806 Nparm 1 1 EffectTest DF Sum ofSquares 1 6.0456409 1 2.0649335 F Ratio 2.4274 0.8291 Prob>F 0.1264 0.3675 Upper 95% -3.713133 0.1290304 0.0274916 Random Sample ChemicalOperators Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.9962 RSquare InM570ppm Summary ofFit RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.136842 0.076621 1.107108 -2.03122 47 Analysisof Variance DF Sum of Squares Mean Square 3 8.355583 2.78519 43 52.704577 1.22569 46 61-060161 Lack ofFit DF Sum ofSquares Mean Square 41 52.472382 1.27981 2 0.232195 0.11610 43 52.704577 F Ratio 2.2724 Prob>F 0.0937 F Ratio 11.0236 Pmb>F 0.0865 Term Intercept GENDER(F-M] YRSCHEM AGE Source GENDER YRSCHEM AGE Nparm 1 1 1 ParameterEstimates Estimate Std Error -1.704276 0.793426 -0.294632 0.197914 -0-038358 0.020996 -0.001758 0.020812 EffectTest DF Sum ofSquares 1 2.7163591 1 4.0909545 1 0.0087486 t Ratio -2.15 -1.49 -1.83 -0.08 F Ratio 2.2162 3.3377 0.0071 Prob>ltl 0.0374 0.1439 0.0747 0.9331 Prob>F 0.1439 0.0747 0.9331 Appendix I Page 10 Appendix I Page 11 Source Model Error C Total Source Lack of Fit Pure E TotalFxror Max RSq 0.7132 Random SAMPIG Chemical Operators RSquare InM570 ppm Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) Analysisof Variance OF Sum of Squares 2 8.346835 44 52.713326 46 61-060161 Lack of Fit DF Sum ofSquares 26 35.199258 18 17.514068 44 52.713326 0.136699 0.097458 1.094546 -'-).03122 47 Mean Square 4.17342 1.19803 F Ratio 3.4836 Prob>F 0.0394 Mean Square 1.35382 0.97300 F Ratio 1.3914 Prob>F 0.2365 Term Intercept GENDER[F-M] YRSCBEM Source GENDER YRSCHEM Estimate -1.767331 -0.294733 -0.039344 Parameter Estimates Std Error tRago 0.266189 -6.64 0.195665 -1.51 0.011258 -2.28 Prob>ltl <.0001 0.1391 0.0275 Lower 95% -2.303797 -0.689068 -0.074125 Nparm 1 1 EffectTest DF Sum ofSquares 1 2.7183290 1 6.2266074 F Ratio 2.2690 5.1974 Prob>F 0.1391 0.0275 Upper 95% -1.230864 0.0996013 -0.004563 Random Sample Chemical Operators Source Model Error C Total Source Lack of Fit Pure Error TotalError Max RSq 0.9763 RSquare InPFOSA ppm Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) Analysisof Variance DF Sum ofSquares 3 6.25657 43 149.06169 46 155.31826 Lack of Fit DF Sum ofSquares 41 145.38604 2 3.67565 43 149.06169 0.040282 -0.02667 1.861868 -3.57167 47 Mean Square 2.08552 3.46655 Mean Square 3.54600 1.83782 F Ratio 0.6016 Prob>F 0.6175 F Ratio 1.9295 Prob>F 0.4006 Term Intercept GENDER(F-K YRSCHEM AGE Source GENDER YRSCBEM AGE Nparm 1 1 1 Parameter Estimates Estimate Std Error -2.698905 1.334337 -0.197181 0.33284 -0.023905 0.63531 -0.01721 0.035001 EffectTest DF Sum ofSquares 1 1.2166245 1 1.5888669 1 0.8381204 tRatio -2.02 -0.59 -0.68 -0.49 F Ratio 0.3510 0.4583 0.2418 Prob>ltl 0.0494 0.5567 0.5020 0.6254 Prob>F 0.5567 0.5020 0.6254 Appendix I Page 12 Appendix i Page 13 Source Model Error C Total Source LackofFit Pure E TotalError Max RSq 0.5686 Random Sample ChemicalOperators InPFOSA ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observation(sorSum Wgts) 0.034886 -0-00898 1.845756 -3.57167 47 AnalysiosfVariance DF Sum ofSquares Mean Square 2 5.41845 2.70923 44 149.89981 3.40681 46 155.31826 Lack ofFit DF Sum ofSquares Mean Square 26 82.89148 3.18813 18 67.00833 3.72268 44 149.89981 F Ratio 0.7952 Prob>F 0.4579 F Ratio 0.8564 Prob>F 0.6485 Term Intercept GENDER[F-M] YRSCiiEM Source GENDER YRSCHEM Estimate -3.316074 -0.198173 -0.033553 ParameterEstimates StdError tRatio 0.44888 -7.39 0.329953 -0-60 0.029102 -1.15 Prob>ltl <.0001 .0.5512 0.2552 Lower 95% -4.220728 -0.863148 -0.092205 Nparm 1 1 EffectTest DF Sum ofSquares 1 1.2289470 1 4.5285163 F Ratio 0.3607 1.3293 Prob>F 0.5512 0.2552 Upper 95% -2.411419 0.4668018 0.0250987 RandomSample ChemicalOperators Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.9973 InM556 ppm Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.127024 0.066119 1.108876 -3.12253 47 AnalysisofVariance DF Sum ofSquares Mean Square 3 7.693385 2.56446 43 52.873065 1.22961 46 60.566451 Lack of Fit DF Sum ofSquares Mean Square 41 52.710027 1.28561 2 0.163038 0.08152 43 52.873065 F Ratio 2.0856 Prob>F 0.11@2 F Ratio 15.7707 Prob>F 0.0613 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENDER YRSCHEM AGE Nparm 1 1 1 ParameterEstimates Estimate Std Error -2.787872 0.794694 -0.340998 0.19823 -0.03131 0.02103 -0.004424 0.020846 EffectTest DF Sum ofSquares 1 3.6385764 1 2.7256630 1 0.0553796 t Ratio -3.51 -1.72 -1.49 -0.21 F Ratio 2.9591 2.2167 0.0450 Prob>ltl 0.0011 0.0926 0.1438 0.8329 Prob>F 0.0926 0.1438 0.8329 Apperidixi Page 14 Apperidixi Page 15 Random Sample Chemical Operators Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.7329 InM556 ppm Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) Analysisof VaOance DF Sum of Squares 2 7.638006 44 52.928445 46 60.566451 Lack ofFit DF Sum ofSquares 26 36.751620 18 16.176825 44 52.928445 0.12611 0.086387 1.096777 -3.12253 47 Mean Square 3.81900 1.20292 Mean Square 1.41352 0.89871 F Ratio 3.1748 Prob>F 0.0515 F Ratio 1.5728 Prob>F 0.1616 Term Intercept GENDER[F-Nfl YRSCHEM Source GENDER YRSCHEM Estimate -2.946517 -0.341253 -0.03379 ParameterEstimates StdError tRatio 0.266731 -11.05 .0-196063 -1.74 0.017293 -1-95 Prob>ltl <.0001 0.0888 0.0571 Lower 95% -3.484071 -0.736392 -0.068642 Nparm 1 1 EffectTest DF Sum of Squares 1 3.6441565 1 4.5927300 F Ratio 3.0294 3.8180 Prob>F 0.0888 0.0571 Upper 95% -2-408957 0.0538852 0.0010617 Appendix i Page I Appendix J Random sample currentjob engineer/labgroup (n = 23): Regression of fluorochemical on gender, years worked in chemical and age; followed by regressionequation of fluorochemicalon gender and years worked in chemical: APperidixi Page 2 RandomSample Engineer/Lab InPFOS ppm RSquare Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.391004 0.294847 0.867828 -0-93898 23 Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.9810 AnalysisofVariance DF Sum ofSquares Mean Square 3 9.187310 3.06244 19 14.309381 0.75313 22 23.496691 Lack ofFit DF Sum ofSquares Mean Square 18 13.862763 0.770154 1 0.446618 0.446618 19 14.309381 F Ratio 4.0663 Prob>F 0.0217 F Ratio 1.7244 Prob>F 0.5438 Term Intercept GENDER[F-K YRSCBIEM AGE Source GENDER YRSCBEM AGE Estimate -0.616826 -0.561666 0.0467532 -0.031175 Parameter Estimates Std Error tRatio 1.435087 -0.43 0.214754 -2.62 0.038427 1.22 0.047633 -0.65 Prob>ltl 0.6722 0.0170 0.2386 0.5206 Lower 95% -3.620476 -1.011148 -0.033675 -0.130872 Nparm 1 1 1 EffectTest DF Sum of Squares 1 5.1516007 1 1.1148580 1 0.3226016 F Ratio 6.8403 1.4803 0.4284 Prob>F 0.0170 0.2386 0.5206 Upper 95% 2.3868238 -0.112185 0.1271809 0.0685214 Appendix i Page 3 Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.8418 Random Sample Engineer/Lab RSquare InPFOS ppm Summary of Fit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) - 0.377275 0.315002 0.855336 -0.93898 23 AnalysisofVariance DF Sum ofSquares Mean Square 2 8.864708 4.43235 20 14.631983 0.73160 22 23.496691 Lack ofFit DF Sum ofSquares Mean Square 12 10.913717 0.909476 8 3.718265 0.464783 20 14.631983 F Ratio 6.0584 Prob>F 0.0088 F Ratio 1.9568 Prob>F 0.1735 Term Intercept GENDER[F-M] YRSCHEM Source GENDER YRSCHEM Estimate -1.538619 -0.537774 0.0233371 Parameter Estimates Std Error t Ratio 0.271378 -5.67 0.208582 0.013818 -2.58 1.69 Prob>ltl <.0001 0.0180 0.1068 Lower 95% -2.104699 -0.972865 -0.005487 Nparm 1 1 EffectTest DF Sum ofSquares 1 4.8631688 1 2.0866755 F Ratio 6.6473 2.8522 Prob>F 0.0180 0.1068 Upper 95% -0-972539 -0.102683 0.0521614 Source Model Error C Total Source Lack of Fit Pure Error TotalError Mu RSq 0.9849 Random Sample Engineer/Lab InPFHS ppm RSquare Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) Analysisof Variance DF Sum of Squares 3 16.373916 19 21.926470 22 38.300386 Lack ofFit DF Sum of Squares 18 21-348414 1 0.578056 19 21.926470 0,427513 0.33712 1.074255 -'-).54721 23 Mean Square 5.45797 1.15402 Mean Square 1.18602 0.57806 F Ratio 4.7295 Prob>F 0.0125 F Ratio 2.0517 Prob>F 0.5060 APperidixi Page 4 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENDER YRSCHEM AGE Estimate -2.462716 -0.741805 0.0546509 -0.030306 Parameter Estimates Std Error t Ratio 1.776447 -1.39 0.265837 -2.79 0.047567 1.15 0.058963 -0.51 Prob>ltl 0.1817 0.0117 0.2648 0.6132 Lower 95%. -6.180835 -1.298203 -0.044908 -0.153717 Nparm 1 1 1 EffectTest DF Sum of Squares 1 8.9859716 1 1.5233215 1 0.3048564 F Ratio 7.7866 1.3200 0.2642 Prob>F 0.0117 0.2648 0.6132 Upper 95% 1.2554025 -0-185407 0.1542097 0.0931054 Appendix i Page Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.8943 Random Sample Engineer/Lab InPFHS ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observation(sorSum Wgts) 0.419553 0.361509 1.054308 -2.54721 23 Analysisof Variance DF Sum ofSquares Mean Square 2 16.069059 8.03453 20 22.231326 1.11157 22 38.300386 Lack ofFft DF Sum ofSquares Mean Square 12 18.182976 1.51525 8 4.048350 0.5060-4 20 22.231326 F Ratio 7.2281 Prob>F 0.0043 F Ratio 2.9943 Prob>F 0.0638 Term Intercept GENDF-R[F-M] YRSCHEM Source GENDER YRSCHIEM Estimate -3.358798 -0.718579 0.0318879 ParameterEstimates Std Error tRatio 0.334507 -10.04 0.257103 -2.79 0.017033 1.87 Prob>ltl <.0001 0.0112 0.0759' Lower 95% -4.056563 -1.254884 -0.003642 Nparm 1 1 EffectTest DF Sum ofSquares 1 8.6829819 1 3.8959557 F Ratio 7.8115 3.5049 Prob>F 0.0112 0.0759 Upper 95% -2.661034 -0.182275 0.0674175 Appendix i Page 6 Source Model Error C Total Random Sample Engineer/Lab In POAAppm Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) Analysisof Variance DF Sum of Squares 3 8.441730 19 17.263008 22 25.704738 0.328411 0.222371 0.953194 -1.56794 23 Mean Square 2.81391 0.90858 Source Lack of Fit Pure Error TotalError Max RSq 0.9453 Lack ofFit DF Sum of Squares Mean Square 18 15.855759 0.88088 1 1.407248 1.40725 19 17.263008 F Ratio 3.0970 Prob>F 0.0514 F Ratio 0.6260 Prob>F 0.7776 Term Intercept GENDER[F-M] YRSCBEM AGE Source GENDER YRSCHEM AGE Estimate -1.020183 -0.663796 0.0323327 -0.03271 Parameter Estimates Std Error t Ratio 1.576253 -0.65 0.235879 -2.81 0.042207 6.77 0.052319 -0.63 Prob>ltl 0.5252 0.0111' 0.4531 0.5393 Lower 95% -4.319296 -1.157492 -0.056007 -0.142213 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 7.1953870 1 0.5331884 1 0.3551409 F Ratio 7.9194 0.5868 0.3909 Prob>F 0.0111 0.4531 0.5393 Upper 95% 2.2789291 -0.170099 0.1206719 0.0767938 Appendixi Page 7 RandomSample Engineer/Lab Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.8984 InPOAA ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.314595 0.246055 0.938567 -1.56794 23 AnalysisofVariance DF Sum ofSquares Mean Square 2 8.086590 4.04329 20 17.618149 0.88091 22 25.704738 Lack ofFit DF Sum ofSquares Mean Square 12 15.005915 1.25049 8 2.612233 0.32653 20 17.618149 F Ratio 4.5899 Prob>F 0.0229 F Ratio 3.8297 Prob>F 0.0326 Term Intercept qENDER(F-M) YRSCBEM Source GENDER YRSCHEM Estimate -1.987348 -0.638727 0.007764 ParameterEstimates Std Error tRatio 0.297785 -6.67 0.228879 -2.79 0.015163 0.51' Prob>lq <.0001 0.0113 0.6142 Lower95% -2-608513 -1.116156 -0-023865 Nparm 1 1 EffectTest DF Sum ofSquares 1 6.8604184 1 0.2309580 F Ratio 7.7879 0.2622 Prob>F 0.0113 0.6142 Upper 95% -1.366184 -0.161298 0.0393931 Apperidixi Page 8 RandomSample Engineer/Lab InPFOSAA ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.269018 0.1536 1.302561 -5.2038 23 Source Model Error C Total Analysisof Variance DF Sum ofSquares Mean Square 3 11.863793 3.95460 19 32.236613 1.69666 22 44.100406 F Ratio 2.3308 Prob>F 0.1067 Source Lack ofFit Pure Error TotalE r Max RSq 0.9997 Lack ofFit DF Sum ofSquares Mean Square F Ratio 18 32.222887 1.79016 130.4211 1 0.013726 0.01373 Prob>F 19 32.236613 0.0688 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENI)ER YRSCBIEM AGE Estimate -2.142278 -0.770545 0.0668187 -0.107688 Parameter Estimates Std Error tRatio 2.153984 -0.99 0.322333 -2.39 0.057617 1.16 0.071495 -1.51 Prob>ltl 0.3324 0.0273 0.2610 0.1484 Lower 95% -6.650586 -1.445192 -0.053899 -0.257327 Nparm EffectTest DF Sum ofSquares I- 9.6957614 1 2.2771575 1 3.8493437 F Ratio 5.7146 1.3421 2.2688 Prob>F 0.0273 0.2610 0.1484 Upper 95% 2.3660301 -0.095899 0.1875362 0.0419505 Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.5230 Random Sample Engineer/Lab InPFOSAA ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.181732 0.099905 1.343242 -5.2038 23 Analysisof Variance DF Sum of Squares Mean Square 2 8.014449 4.00722 20 36.085957 1.80430 22 44.100406 Lack ofFit DF Sum ofSquares Mean Square 12 15.048236 1.25402 8 21-037721 2.62972 20 36-085957 F Ratio 2.2209 Prob>F 0.1346 F Ratio 0.4769 Prob>F 0.8805 Appendix i Page 9 Term Intercept GENDER[F-M] YRSCREM Source GENDER YRSCHEM Estimate -5.326428 -0.688015 -0.014068 ParameterEstimates StdError tRatio 0.426179 -12.50 0.327563 -2.10 0.021701 -0.65 Prob>ltl <.0001 0.0486 0.5242 Lower 95% -6.215415 -1.371293 -0.059334 Nparm 1 1 EffectTest DF Sum ofSquares 1 7.9600380 1 0.7582318 F Ratio 4.4117 0.4202 Prob>F 0.0486 0.5242 Upper 95% -4.437441 -0-004736 0.0311988 Source Model Error C Total Source Lack of Fit Pure Error TotalError Max RSq 0.9946 Random Sample Engineer/Lab RSquare In M570 ppm Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) Analysisof Variance DF Sum ofSquares 3 0.754025 19 16.887940 22 17.641966 Lack of Fit DF Sum ofSquares 18 16.791797 1 0.096143 19 16.887940 0.04274 -0.10841 0.942783 -3.01612 23 Mean Square 0.251342 0.888839 Mean Square 0.932878 0.096143 F Ratio 0.2828 Prob>F 0.8372 F Ratio 9.7030 Prob>F 0.2481 Appendix i Page 10 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENI)ER YRSCFMM AGE Estimate -2.114124 -0.170074 0.0257524 -0.033236 Parameter Estimates Std Error t Ratio 1.559036 -1.36 0.233302 -0.73 0.041746 0.62 0.051747 -0.64 Prob>ltl 0.1910 0.4749 0.5446 0.5294 Lower 95% -5.3772 -0.658377 -0.061622 -0.141543 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 0.47234628 1 0.33824664 1 0.36666114 F Ratio 0.5314 0.3805 0.4125 Prob>F 0.4749 0.5446 0.5284 Upper 95% 1.1489527 0.3182297 O.il3l268 0.0750714 Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.4859 Random Sample Engineer/Lab InM570ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.021957 .-0.07585 0.928833 -3.01612 23 Analysisof Variance DF Sum ofSquares Mean Square 2 0.387364 0.193682 20 17.254602 0.862730 22 17.641966 Lack ofFit DF Sum ofSquares Mean Square 12 8.194523 0.68204 8 9.070079 1.13376 20 17.254602 F Ratio 0.2245 Prob>F 0.8009 F Ratio 0.6016 Prob>F 0.7939 Appendix i Page 11 Term Intercept GENDER[F-AC YRSCHEM Source GENI)ER YRSCHEM ParameterEstimates Estimate StdError tRatio --3.09685 0.294697 -10.51 0.144602 0.226505 -0.64 0.00(Y7884 * 0.015006 0.05 Prob>ltl <.0001 0.5305 0.9586 Lower 95% -3.711572 -0.61708 -0.030513 Nparm 1 1 EffectTest DF Sum ofSquares 1 0.35161749 1 0.00238175 F Ratio 0.4076 0.0028 Prob>F 0.5305 0.9586 Upper 95% -2.482128 0.3278751 0.0326895 Apperidixi Page 12 RandomSample Engineer/Lab Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.9950 InPFOSA ppm Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) AnalysisofVariance DF Sum ofSquares 3 0.846121 19 72.471423 22 73.317544 Lack ofFit DF Sum ofSquares 18 72.-104070 1 0.367353 19 72.471423 0.01154 -0.14453 1.95302 -5.39325 23 Mean Square 0.28204 3.81429 Mean Square 4.00578 0.36735 F Ratio 0.0739 Prob>F 0.9732 F Ratio 10.9044 Prob>F 0.2345 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENI)ER YRSCHEM AGE Estimate -5.985167 0.1760337 -0-001162 0.0171392 ParameterEstimates StdError tRatio 3.229618 -1.85 0.483297 0.36 0.086478 -0.02 0.107197 0.16 Prob>ltl 0.0794 0.7197 0.9840 0.8747 Lower95% -12.74479 -0.835511 -0-182762 -0.207225 Nparm 1 1 1 EffedTest DF Sum ofSquares 1 0.50603123 1 0.00158323 1 0.09750555 F Ratio 0.1327 0.0004 0.0256 Prob>F 0.7197 0.9840 0.8747 Upper 95% 0.7744532 1.1875781 6.1792383 0.2415031 Apperidixi Page 13 Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.8381 Random Sample Engineer/Lab InPFOSA ppm RSquare Summary of Fit RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) Analysisof Variance DF Sum ofSquares 2 0.748615 20 72.568929 22 73.317544 Lack ofFn DF Sum ofSquares 12 60.697300 8 11.871629 20 72.568929 0.010211 -0.08877 1.904848 -5.39325 23 Mean Square 0.37431 3.62845 Mean Square 5.05811 1.48395 F Ratio 0.1032 Prob>F 0.9025 F Ratio 3.4085 Prob>F 0.0452 Term Intercept GENDER[F-M] YRSCHEM Source GENDER YRSCHEM Esfimate -5.478392 0.1628986 0.0111116 ParameterEstimates StdError tRatio 0.604364 -9.06 0.464516 0.35 0.030774 0.36 Prob>ltl <.0001 0.7295 0.7218 Lower 95% -6.739063 -0.806057 -0-053081 Nparm 1 EffectTest DF Sum ofSquares 1 0.44622510 0.47306076 F Ratio 0.1230 0.1304 Prob>F 0.7295 0.7218 Upper95% -4.217722 1.1319545 0.075@039 Appendix i Page 14 Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.9991 Random Sample Engineer/Lab InM556 ppm RSquare Summary of Fit RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) AnalysisofVariance DF Sum ofSquares 3 1.164918 19 28.197906 22 29.362824 0.039673 -0-11196 1.218237 -4.65037 23 Mean Square 0.38831 1.48410 Lack ofFit DF Sum ofSquares Mean Square 18 28.172151 1.56512 1 0.025755 0.02575 19 28.197906 F Ratio 0.2616 Prob>F 0.8521 F Ratio 60.7697 Prob>F 0.1006 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENI)ER YRSCBEM AGE Estimate -5.844133 -0.1268Or7 -0-018287 0.0342181 ParameterEstimates Std Error tRatio 2.014541 -2.90 0.301466 -0.42 0.053943 -0.34 0.066866 0.51 Prob>ltl 0.0092 0.6787 0.7383 0.6147 Lower 95% -10.06059 -0.757779 -0.i3119 -0-105734 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 0.26258648 1 0.17056194 1 0.38865284 F Ratio 0.1769 0.1149 0.2619 Prob>F 0.6787 0.7383 0.6147 Upper 95% -1-627679 0.5041648 0.0946156 0.1741698 Source Model Error C Total Source Lack of Fit PuireError TotalError Max ASq 0.4851 Random Sample Engineer/Lab RSquare RSquare Adj InM556 ppm Summary of Fit Root Mean Square Error Mean of Response Observations(orSum Wgts) Analysisof Variance DF Sum ofSquares 2 0.776265 20 28.586559 22 29.362824 Lack ofFit DF Sum ofSquares 12 13.466606 8 15.119953 20 28.586559 0.026437 -0.07092 1.195545 -4.65037 23 Mean Square 0.38813 1.42933 Mean Square 1.12222 1.88999 F Ratio 0.2715 Prob>F 0.7650 F Ratio 0.5938 Prob>F 0.7996 Apperidixi Page 15 Term Intercept GENDER[F-M) YRSCHEM Source GENDER YRSCHEM Estimate -4.832364 -0.153031 0.0074147 Parameter Estimates Std Error 0.379318 tRatio -12.74 0.291545 -0.52 0.019315 0.38 Prob>ltl <.0001 0.6054 0.7051 Lower 95% -5.623603 -0.76118 -0.032814 Nparm 1 1 EffectTest DF Sum ofSquares 1 0.39380354 1 0.21064617 F Ratio 0.2755 0.1474 Prob>F 0.6054 0.7051 Upper 95% -4-041126 0.4551172 0.0477039 APPendix K Page I Appendix K A]Iparticipancturrenjtob chemicaloperator(sn= 64): Regressionof fluorochemicalon gender,yearsworked inchemicaland age;followedby regressioenquationof fluorochemicaoln genderand yearsworked inchemical Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.9661 AllParticipants Chemical Operators InPFOS ppm RSquare Summary ofFit RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.150439 0.107961 0.643599 0.392284 64 AnalysisofVariance DF Sum of Squares Mean Square 3 4.400964 1.46699 60 24.853181 0.41422 63 29.254145 Lack ofFit DF Sum of Squares Mean Square 573 2. 03..989611654375 00..431380652439 60 24.853181 F Ratio 3.5416 Prob>F 0.0198 F Ratio 1.2664 Prob>F 0.4953 APPerldixK Page 2 Term Intercept GIENDER[F-M] YRSCHEM AGE Source GENDER YRSCHEM AGE Estimate 0.062633 -0.250464 0.0171146 -0-000079 Parameter Estimates Std Error tRatio 0.401961 0.16 0.10427 -2.40 0.011052 1.55 0.010698 -0.01 Prob>ltl 0.8767 0.0194 0.1267 0.9941 Lower 95% -0.741408 -0.459035 -0-004992' -0.021478 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 2.3900299 1 0.9933570 1 0.0000227 F Ratio 5.7700 2.3981 0.0001 Prob>F 0.0194 0.1267 0.9941 Upper 95% 0.8666743 -0.04189'3 0.0392214 0.0213193 ...... -77-' Source Model Error C Total Source Lack ofFit Pure E TotalError Max RSq 0.6918 AllParticipants Chemical Operators In PFOSdfppm RSquare Summary of Fit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.150438 0.122584 0.638302 0.392284 64 Analysisof Variance DF Sum ofSquares Mean Square 2 4.400941 2.20047 61 24.853204 0.40743 63 29.254145 Lack ofFit DF Sum ofSquares Mean Square 31 15.838180 0.510909 30 9.015024 0.300501 61 24.853204 F Ratio 5.4009 Prob>F 0.0069 F Rago 1.7002 Prob>F 0.0748 APPendix K Page 3 Term Intercept GENDER[F-M] YRSCHEM Source GENDER YRSCHEM Estimate 0.0598248 -0.250543 0.017067 Parameter Estimates Std Error tRago 0.132618 0.45 0.10288 -2.44 0.008912 1.92 Prob>ltl 0.6535 0.0178 0.0602 Lower 95% -0.205362 -0.456265 -0.000753 Nparm 1 1 EffectTest DF Sum of Squares 1 2.4162931 1 1.4943613 F Ratio 5.9306 3.6678 Prob>F 0.0178 0.0602 Upper 95% 0.325012 -0.04482 0.0348868' APpendix K Page 4 AllParticipants Chemical Operators Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.9842 RSquare InPFHS ppm Summary of Fit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) Analysisof Variance DF Sum of Squares 3 17.697687 60 24.896914 63 42.594602 Lack ofFit DF Sum ofSquares 57 24.223557 3 0.673357 60 24.896914 0.415491 0.386266 0.644165 -1.23054 64 Mean Square 5.89923 0.41495 Mean Square 0.424975 0.224452 F Ratio 14.2168 Prob>F <.0001 F Ratio 1.8934 Prob>F 0.3353 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENDER YRSCHEM AGE Estimate -1.721892 -0.370838 0.0480263 -0.005737 Parameter Estimates StdError t Ratio 0.402314 -4.28 0.104362 -3.55 0.011061 4.34 0.010707 -0.54 Prob>ltl <.0001 0.0007 <.0001 0.5940 Lower 95% -2.52664 -0.579593 O.Oj59001 -0-027155 Nparm EffectTest DF Sum ofSquares 5.2393846 7.8221792 0.1191450 F Ratio 12-.6266 18.8510 0.2871 Prob>F 0.0007 <.0001 0.5940 Upper 95% -0.917144 -0-162083 0.(Y701525 0.01568 Appendix K Page 5 AllParticipants Chemical Operators Source Model Error C Total Soume Lack ofFit Pure Error TotalError Max RSq 0.8068 RSquare InPFHS ppm Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.412694 0.393438 0.64039 -1.23054 64 Analysisof Variance DF Sum ofSquares Mean Square 2 17.578542 8.78927 61 25.016059 0.41010 63 42.594602 Lack ofFit DF Sum of Squares Mean Square 31 16.785999 0.541484 30 8.230060 0.274335 61 25.016059 F Ratio 21.4321 Prob>F <.0001 F Raflo 1.9738 Prob>F 0.0329 Term Intercept GENDER(F-M] YRSCHEM Source GENDER YRSCHEM Esfimate -1.925193 -0.3765 0.0445154 Parameter Esfimates Std Error tRatio 0.133052 -14.47 0.103217 -3.65 0.008941 4.99 Prob>ltl <.0001 0.0005 <.0001 Lower 95% -2-191247 -0.582895 0.0266973 Nparm 1 1 EffectTest DF Sum ofSquares 1 5.456536 1 10.193737 F Ratio 13.3054 24.8568 Prob>F 0.0005 <.0001 Upper 95% -1-659138 -0.170105 6.0624535 APPendix K Page 6 Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.9807 AllParticipants Chemical Operators InPOAA ppm RSquare Summary ofFit RSquare Adj Root Mean SquareError Mean ofResponse Observation(sorSum Wgts) 0.21249 0.L73115 0.567525 0.614523 64 AnalysisofVariance DF Sum ofSquares Mean Square 3 5.214396 1.73813 60 19.325057 0.32208 63 24.539453 Lack of Fit DF Sum of Squares Mean Square 57 18.851100 0.330721 3 0.473957 0.157986 60 19.325057 F Ratio 5.3965 Prob>F 0.0024 F Ratio 2.0934 Prob>F 0.3009 Term Intercept GENDER[F-M] YRSCHEid AGE Source GENDER YRSCHEM AGE Estimate 0.4370637 -0.313225 0.0150521 -0-004228 ParameterEstimates Std Error tRatio 0.354448 1.23 0.091945 -3.41 0.009745 1.54 0.009433 -0.45 Prob>ltl 0.2224 0.0012 O.f277 0.6556 Lower 95% -0.271939 -0.497143 -0.004442 -0.023097 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 3.7378707 1 0.7683549 1 0.0647015 F Ratio 11-6053 2.3856 0.2009 Prob>F 0.0012 0.1277 0.6556 Upper 96% 1.1460661 -0.129307 0.0345458 0.0146413 APPendix K Page 7 Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.8150 AllParticipants Chemical Operators InPOAA ppm RSquare Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.209854 0.183947 0.563795 0.614523 64 AnalysisofVariance DF Sum ofSquares Mean Square 2 5.149695 2.57485 61 19.389758 0.31786 63 24.539453 Lack ofFit DF Sum ofSquares Mean Square 31 14.849379 0.479012 30 4.540380 0.151346 61 19.389758 F Ratio 8.1004 Prob>F 0.0008 F Ratio 3.1650 Prob>F 0.0011 Term Intercept GENDER[F-ml YRSCHEM Source GENDER YRSCHEM Estimate 0.2872479 -0.317397 0.0125091 Parameter Estimates Std-Error tRatio 0.117138 2.45 0.090872' -3.49 0.007871 1.59 Prob>ltl 0.0171 0.0009 0.1172 Lower 95% 0.0530152 -0.499106 -6-003231 Nparm 1 1 EffectTest DF Sum ofSquares 1 3.8778718 1 0.8027714 F Ratio 12.1997 2.5255 Prob>F 0.0009 0.1172 Upper 95% 0.5214806 -0.135688 0.0282488 Source Model Error C Total Source Lack ofFit Pure Error TotalError Max ASq 0.9025 AllParticipants Chemical Operators InPFOSAA ppm RSquare Summary of Fit RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.088735 0.043172 1.583933 -4.28694 64 Analysisof Variance DF Sum ofSquares Mean Square 3 14.65802 4.88601 60 150.53065 2.50884 63 165.18867 Lack ofFit DF Sum ofSquare$ Mean Square 57 134.42483 2.35833 3 16.10582 5.36861 60 150.53065 F Ratio 1.9475 Prob>F 0.1316 F Ratio 0.4393 Prob>F 0.9106 Appendix K Paige8 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENDER YRSCHEM AGE Estimate -5.252813 -0.52127 -0.04353 0.0265422 Parameter Estimates Sid Error tRatio 0.989248 -5.31 0.256614 -2.03 0.027199 -1.60 0.026328 1.01 Prob>ltl <.0001 0.0467 0.1148 0.3174 Lower 95% -7-231604 -1.034575 -0.097936 -0.026121 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 10.352296 1 6.426227 1 2.549914 F Ratio 4.1263 2.5614 1.0164 Prob>F 0.0467 0.1148 0.3174 Upper 95% -3.274023 -0-007964 0.0108755 0.0792053 APPendiX K Page 9 Source Model Error C Total Source Lack of Fit Pure Effor TotalError Max RSq 0.5858 AllParticipants Chemical Operators InPFOSAA ppm RSquare Summary of Fit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) Analysisof Variance DF Sum ofSquares 2 12.10810 61 153.08056 63 165.18867 0.073299 0.042915 1.584146 -4.28694 64 Mean Square 6.05405 2.50952 Lack of Fit DF Sum ofSquares Mean Square 31 84.66080 2.73099 30 68.41976 2.28066 61 153.08056 F Ratio 2.4124 Prob>F 0.0981 F Ratio 1.1975 Prob>F 0.3117 Term Intercept GENDER[F-M] YRSCHEM Source GENDER YRSCHEM Estimate -4.312304 -0.495076 -0.027566 Parameter Estimates Std Error tRatio 0.329134 -13.10 0.25533 -1.94 0.022117 -1.25 Prob>ltl <.0001 0.0571 0'.2174 Lower 95% -4.970449 -1-00564 -0-071792 Nparm 1 1 EffectTest DF Sum ofSquares 1 9.4347501 1 3.8984466 F Ratio 3.7596 1.5535 Prob>F 0.0571 0.2174 Upper 95% -3.654159 0.0154881 0.0166594 AlPlarticipants ChemicalOperators APPerldixK Page 10 Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.9962 InM570 ppm Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.164237 0.122449 1.12124 -1.94564 64 Analysisof Variance DF Sum ofSquares Mean Square 3 14.823015 4.94100 60 75.430695 1.25718 63 90.253710 Lack ofFit DF Sum ofSquares Mean Square 57 75.084356 1.31727 3 0.346339 0.11545 60 75.430695 F Ratio 3.9302 Prob>F 0.0126 F Ratio 11.4102 Prob>F 0.0335 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENDER YRSCBEM AGE Estimate -1.222446 -0.330479 -0.038776 -0.012926 ParameterEstimates Std Error tRatio 0.700272 -1.75 0.181653 -1.82 0.019254 -2.01 0.018637 -0.69 Prob>ltl 0.0860 0.0739 0.0485 0.4906 Lower 95% -2.623199 -0.693839 -0.077289 -0-050205 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 4.1610173 1 5.0990709 1 0.6047597 F Ratio 3.3098 4.0560 0.4810 Prob>F 0.0739 0.0485 0.4906 Upper 950/o 0.1783065 0.0328812 -O.OOOd63 0.0243533 -77- 7, Source Model Error C Total Source Lack of Fit Pure Error TotalError Max RSq 0.5860 AllParticipants Chemical Operators In M570 ppm RSquare Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) Analysisof Variance DF Sum ofSquares 2 14.218255 61 76.035455 63 90.253710 0.157537 0.129915 1.11646 -1.94564 64 Mean Square 7.10913 1.24648 Lack ofFit DF Sum ofSquares Mean Square 31 38.667300 1.24733 30 37.368155 1.24561 61 76.035455 F Ratio 5.7033 Prob>F 0.0054 F Ratio 1.0014 Prob>F 0.4993 Appendix K Page 11 Term Intercept GENI)ER[F-M] YRSCHEM Source GENI)ER YRSCHEM Estimate -1.680474 -0.343235 -0.04655 Parameter Estimates Std Error tRatio 0.231964 -7.24 0.179949 -1.91 0.015587 -2.99 Prob>ltl <.0001 0.0612 0.0041 Lower 95% -2.144315 -0.703066 -0.077719 Nparm 1 1 EffectTest DF Sum ofSquares 1 4.534934 1 11-117067 F Ratio 3.6382 8.9187 Prob>F 0.0612 0.0041 Upper 95% -1.216632 0.0165953 -0.015382 AllParticipants Chemical Operators APPendix K Page 12 Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.9823 RSquare InPFOSA ppm Summary ofFit RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.036452 -0.01173 1.87564 -3.8617 64 AnalysisofVariance DF Sum ofSquares Mean Square 3 7.98536 2.66179 60 211.08153 3.51803 63 219.06689 Lack of Fit DF Sum ofSquares Mean Square 57 207.20231 3.63513 3 3.87923 1.29308 60 211.08153 F Ratio 0.7566 Prob>F 0.5229 F Ratio 2.8112 Prob>F 0.2148 Term Intercept GENDER[F-hfl YRSCHEM AGE Source GENI)ER YRSCHIEM AGE Estimate -3-083552 -0.119692 -0.028921 -0.01353 Parameter Estimates Std Error tRatio 1.171434 -2.63 0.303874 -0.39 0.032208 -0.90 0.031176 -0.43 Prob>ltl 0.0108 0.6951 0.3728 0.6659 Lower 95% -5.426769 -0.727531 -0-093347 -0-075892 Nparm 1 1 1 EffectTest DF Sum of Squares 1 0.5458097 1 2.8366160 1 0.6625603 F Ratio 0.1551 0.8063 0.1883 Prob>F 0.6951 0.3728 0.6659 Upper 95% -0.740335 0.4881469 0.0355046 0.0488322 APPendix K Page 13 Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.9823 Term Intercept GENDER[F-M] YRSCBEM AGE Source GENI)ER YRSCBEM AGE AllParticipants Chemical Operators InPFOSA ppm RSquare Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.036452 -0.01173 1.87564 -3.8617 64 AnalysisofVariance DF Sum ofSquares Mean Square 3 7.98536 2.66179 60 211.08153 3.51803 63 219.06689 Lack ofFit DF Sum ofSquares Mean Square 57 207.20231 3.63513 3 3.87923 1.29308 60 211.08153 F Ratio 0.7566 Prob>F 0.5229 F Ratio 2.8112 Prob>F 0.2148 Esfimate -3.083552 -0.119692 -0.028921 -0.01353 Parameter Estimates Std Error t Ratio 1.171434 -2.63 0.303874 -0.39 *0.032208 -0.90 0.031176 -0.43 Prob>ltl 0.0108 0.6951 0.3728 0.6659 Lower 95% -5.426769 -0.727531 -'0.093347 -0-075892 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 0.5458097 1 2.8366160 1 0.6625603 F Ratio 0.1551 0.8063 0.1883 Prob>F 0.6951 0.3728 0.6659 Upper 95% -0.740335 0.4881469 0.0355046 0.0488322 APperidix K Page 14 AllParticipants Chemical Operators Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.5240 RSquare InPFOSA ppm Summary ofFit RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.033427 0.001736 1.86312 -3.8617 64 AnalysisofVariance DF Sum ofSquares Mean Square 2 7.32280 3.66140 61 211.74409 3.47121 63 219-06689 Lack ofFit DF Sum ofSquares Mean Square 31 107.46461 3.46660 30 104.27948 3.47598 61 211.74409 F Ratio 1.0548 Prob>F 0.3545 F Ratio 0.9973 Prob>F 0.5037 Term Intercept GENDER[F-M] YRSCHEM Source GENDER YRSCHEM Estimate -3.562968 -0.133044 -0.037059 Parameter Estimates Std Error tRatio 0.387096 -9.20 0.300294 -0.44 0.026012 -'1.42 Prob>ltl <.0001 0.6593 0.1593 Lower 95% 4.337015 -0.73352 -0.089073 Nparm 1 1 EffectTest DF Sum ofSquares 1 0.6813608 1 7.0457385 F Ratio 0.1963 2.0298 Prob>F 0.6593 0.1593 Upper 95% -2.788922 0.4674322 0.0149549 Appendix K Page 15 AllParticipants Chemical Operators RSquare InM556 ppm Summary of Fit RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.135522 0.092298 1.04555 -3.10248 64 Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.9974 AnalysisofVariance DF Sum ofSquares Mean Square 3 10.282420 3.42747 60 65.590476 1.09317 63 75.872896 Lack ofFit DF Sum of Squares Mean Square 57 65.396881 1.14731 3 0.193596 0.06453 60 65.590476 F Ratio 3.1353 Prob>F 0.0319 F Ratio 17.7790 Pmb>F 0.0177 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENI)ER YRSCHEM AGE Estimate -2.765873 -0.36628 -0.03039 -0-006153 Parameter Estimates Std Error t Ratio 0.653 4.24 0.16939 -2.16 0.017954 -1.69 0.017379 -0.35 Prob>ltl <.0001 0.0346 0.0957 0.7245 Lower 95% 4.072068 -0.705111 -0-066304 -0-040916 Npann 1 1 1 EffectTest DF Sum ofSquares 1 5.1113761 1 3.1321284 1 0.1370284 F Ratio 4.6757 2.8652 0.1253 Prob>F 0.0346 0.0957 0.7245 Upper 95% -1.459679 -0.027448 0.005523 0.0286099 APPendix K Page 16 Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.6199 AllParticipants Chemical Operators InM556 ppm RSquare Summary ofFit RSquare Adj Root Mean SquareEffor Mean of Response Observations(orSum Wgts) 0.133716 0.105313 1.038027 -3.10248 64 AnalysisofVariance DF Sum of Squares Mean Square 2 10.145391 5.07270 61 65.727505 1.07750 63 75.872896 Lack ofFit DF Sum of Squares Mean Square 31 36.884521 1.18982 30 28.842984 0.96143 61 65.727505 F Ratio 4.7078 Prob>F 0.0126 F Ratio 1.2376 Pmb>F 0.2807 Term Intercept GENDER[F-M] YRSCBEM Source GENDER YRSCBEM Estimate -2.983898 -0.372352 -0.034091 Parameter Estimates Std Error t Ratio 0.215668 -13.84 0.167307 -2.23 0.0144:92 -2.35 Prob>ltl <.0001 0.0298 0.0219 Lower 95% -3.415154 -0.706904 -0-06307 Nparm 1 1 EffectTest DF Sum ofSquares 1 5.3369633 1 5.9624437 F Ratio 4.9531 5.5336 Prob>F 0.0298 0.0219 Upper 95% -2.552642 -0-0378 -0-005112 APPendix L Page I Appendix L Allparticipancturrentjob engineer/lagbroup (n = 37): Regressionequationof fluorochemicalon gender,yearsworked in chemical and age; followed by regressionequationof fluorochemicalon gender and yearsworked in chemical APPerldix L Page 2 Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.9878 AllParticipants Engineer/Lab InPFOS ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.386611 0.330848 0.825205 -0.94033 37 AnalysisofVariance DF Sum ofSquares Mean Square 3 14.163658 4.72122 33 22.471780 0.68096 36 36.635438 Lack of Fit DF Sum ofSquares Mean Square 32 22.025162 0.688286 I 0.4466i8 0.446618 33 22.471780 F Ratio 6.9332 Prob>F 0.0010 F Ratio 1.5411 Prob>F 0.5735 Term Intercept GENDER[F-M] YRSCHIEM AGE Source GENDER YRSCHEM AGE Estimate -2.071342 -0.434286 0.0189436 0.0146474 ParameterEstimates Std Error tRatio 0.85018 -2.44 0.165902 -2.62 0.021692 0.87 0.026443 0.55 Prob>ltl 0.0204 0.0133 0.3888 0.5834 Lower 95% -3.801035 -0.771815 -0.025188 -0.039152 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 4.6662679 1 0.5193576 1 0.2089349 F Ratio 6.8525 0.7627 0.3068 Prob>F 0.0133 0.3888 0.5834 Upper 95% -0.341649 -0-096757 0.0630753 0.0684465 @4 APPendix L Page 3 Source Model Error C Total Source Lack of Fit Pure Error TotalE Max RSq 0.7264 AllParticipants Engineer/Lab InPFOS ppm Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) Analysisof Variance DF Sum ofSquares 2 13.954723 34 22.680715 36 36.635438 0.380908 0.344491 0.81675 -0.94033 37 Mean Square 6.97736 0.66708 Lack ofFit DF Sum ofSquares Mean Square 19 12.658602 0.666242 15 10.022112 0.668141 34 22.680715 F Ratio 10.4596 Prob>F 0.0003 F Ratio 0.9972 Prob>F 0.5100 Term Intercept GENDEREF-M] YRSC]HEM Source GENDER YRSCHEM Estimate -1.615211 -0.439047 0.0293537 Parameter Estimates Std Error tRatio 0.20928 -7.72 0.163982 -2.68 0.010721 2.74 Prob>ltl <.0001 0.0113 0.0098 Lower 95% -2.040516 -0.772296 0.0075663 Nparm 1 1 EffectTest - DF Sum ofSquares 1 4.7819718 1 5.0008180 F Ratio 7.1685 7.4966 Prob>F 0.0113 0.0098 Upper 95% -1.189906 -0.105797 0.051141 APPendix L Page 4 Source Model Error C Total Source Lack of Fit Pure Error TotalError Max RSq 0.9895 AllParticipants Engineer/Lab RSquare InPFHS ppm Summary ofFit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) Analysisof Variance DF Sum ofSquares 3 24.559883 33 30.621774 36 55.181656 Lack ofFit DF Sum ofSquares 32 3Q.043718 1 0.578056 33 30.621774 0.445073 0.394625 0.963293 -2.5975 37 Mean Square 8.18663 0.92793 Mean Square 0.938866 0.578056 F Ratio 8.8224 Prob>F 0.0002 F Ratio 1.6242 Prob>F 0.5616 Term Intercept GENDER[F-M] YP@SCHEM AGE Source GENI)ER YRSCHEM AGE Estimate -4.078592 -0.566055 0.0256075 0.0189228 Parameter Estimates Std Error t Ratio 0.992447 -4.11 0.193664 -2.92 0.025321 1.01 0.030868 0.61 Prob>ltl 0.0002 0.0062 0.3192 0.5441 Lower 95% -6.097727 -0.960065 -0.025909 -0.043879 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 7.9274937 1 0.9490174 1 0.3487090 F Ratio 8.5432 1.0227 0.3758 Prob>F 0.0062 0.3192 0.5441 Upper 95% -2-059457 -0.172045 0.077124 0.0817245 APPendix L Page 5 Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.8537 AllParticipants Engineer/Lab InPFHS ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.438754 0.40574 0.954409 -2.5975 37 AnalysisofVariance DF Sum ofSquares Mean Square 2 24.211174 12.1056 34 30.970483 0.9109 36 55.181656 Lack ofFit DF Sum ofSquares Mean Square 19 22.898600 1.20519 is 8.071882 0.53813 34 30.970483 F Ratio 13.2897 Prob>F <.0001 F Ratio 2.2396 Prob>F 0.0591 Term Intercept GENDER[F-M] YRSCBEM Source GENDER YRSCHEM Estimate -3.48932 -0.572206 0.0390561 ParameterEstimates StdError tRatio 0.244553 -14.27 0.19162 -2.99 0.012S28 3.12 Prob>ltl <.0001 0.0052 0.0037 Lower 95% -3.986308 -0.961623 0.0135966 Nparm 1 1 EffectTest DF Sum of Squares 1 8.1225045 1 8.8530880 F Ratio 8.9170 9.7191 Prob>F 0.0052 0.0037 Upper 95% -2.992332 -0.182789 0.0645156 AppendiX L Page 6 AllParticipants Engineer/Lab Source Model Error C Total Source Lack of Fit Pure Error TotalError Max RSq 0.9642 RSquare InPOAA ppm Summary of Fit RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.305199 0.242035 0.910104 -1.62112 37 AnalysisofVariance DF Sum ofSquares Mean Square 3 12.006567 4.00219 33 27.333536 0.82829 36 39.340103 Lack ofFit DF Sum ofSquares Mean Square 32 25.926287 0.81020 1 1.407248 1.40725 33 27.333536 F Ratio 4.8319 Prob>F 0.0068 F Ratio 0.5757 Prob>F 0.8031 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENI)ER YRSCBEM AGE Estimate -2.930493 -0.527939 -0.002986 0.0256936 Parameter Estimates Std Error tRatio 0.937648 -3.13 0.182971 -2.89 0.023923 -0-12 0.029164 0.88 Prob>ltl 0.0037 0.0068 0.9014 0.3947 Lower 95% -4.838141 -0-900194 -0-051658 -0.03364 Nparm 1 1 1 EffectTest DF Sum of Squares 1 6.8958131 1 0.0129048 1 0.6428967 F Ratio 8.3254 0.0156 0.7762 Prob>F 0.0068 0.9014 0.3847 Upper 95% -1.022846 -0.155684 0.0456859 0.0850277 Apperidix L Page 7 AllParticipants Engineer/Lab Source Model Error C Total Source Lack ofFit Pure Error TotalError K4ax RSq 0.7269 InPOAA ppm RSquare Summary ofFit RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.288857 0.247025 0.907103 -1.62112 37 AnalysisofVariance DF Sum ofSquares Mean Square 2 11.363671 5.68184 34 27.976432 0.82284 36 39.340103 Lack of Fit DF Sum ofSquares Mean Square 19 17.231908 0.906943 15 10.744524 0.716302 34 27-976432 F Ratio 6.9052 Prob>F 0.0030 F Ratio 1.2661 Prob>F 0.3249 Term Intercept GENDER[F-M] YRSCHEM Source GENI)ER YRSCHEM Esbmate -2.130374 -0.53629 0.0152146 Nparm 1 1 Parameter Estimates Std Error tRatio 0.232431 -9-17 0.182123 -2.94 0.011907 1.28 Prob>ltl <.0001 0.0058 0.2082 Lower 95% -2.602729 -0.906406 -0.008923 EffectTest DF Sum ofSquares 1 7.1348529 1 1.3541118 F Ratio 8.6710 1.6457 Prob>F 0.0058 0.2082 Upper 95% -1.65802 -0.166175 b.0394721 APPerldixL Page 8 AlPlarticipants Engineer/Lab InPFOSAA ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.181276 0.106847 1.272756 -5-65628 37 Source Model Error C Total Source Lack ofFit Pure E TotalE Max RSq 0.9998 AnalysisofVariance DF Sum ofSquares Mean Square 3 11.836056 3.94535 33 53.456962 1.61991 36 65.293017 Lack ofFit DF Sum ofSquares Mean Square 32 53.443235 1.67010 1 0.013726 0.01373 33 53.456962 F Ratio 2.4355 Prob>F 0.0822 F Ratio 121.6742 Prob>F 0.0717 Term Intercept GENDER[F-M] YRSCBEM AGE Source GENI)ER YRSCHEM AGE Estimate -3.677933 -0.571849 0.0270165 -0.063657 ParameterEstimates Std Error tRatio 1.311276 -2.80 0.25588 -2.23 0.033456 0.81 0.040785 -1.56 Prob>ltl 0.0084 0.0323 0.4251 0.1281 Lower 95% -6.345727 -1.092437 -0.04105 -0.146634 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 8.0905950 1 1.0563257 1 3.9462586 F Ratio 4.9945 0.6521 2.4361 Prob>F 0.0323 0.4251 0.1281 Upper 95% -1-010138 -0.051261 0.005083 0.0193199 ApperidixL Page 9 AllParticipants Engineer/Lab Source Model Error C Total Source Lack ofFit Pure Error TotalError Mm RSq 0.6097 RSquare InPFOSAA ppm Summary ofFit RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) AnalysisofVariance DF Sum ofSquares 2 7.889797 34 57.403220 36 65.293017 0.120837 0.069121 1.299358 -5.65628 37 Mean Square 3.94490 1.68833 Lack ofFft DF Sum ofSquares Mean Square 19 31-922124 1.68011 15 25.481096 1.69874 34 57-403220 F Ratio 2.3366 Prob>F 0.1120 F Ratio 0.9890 Prob>F 0.5166 Term Intercept GENDER[F-M] YRSCHIEM Source GENDER YRSCBIEM Estimate -5.660266 -0.551158 -0-018225 Parameter Estimates Std Error tRatio 0.332941 -17.00 0.260877 -2.11 0.017056 -1.07 Prob>ltl <.0001 0.0420 0.2928 Lower95% -6.336879 -1-081321 -6.052886 Nparm 1 1 EffectTest DF Sum ofSquares 1 7.5359446 1 1.9278040 F Ratio 4.4635 1.1418 Prob>F 0.0420 0.2928 Upper 95% -4-983654 -0.020995 0.016436 APPeridix L Page 10 AllParticipants Engineer/Lab Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.9971 RSquare InM570ppm Summary of Fit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.023179 -0-06562 0.990095 -3.19558 37 Analysisof Variance DF Sum of Squares Mean Square 3 0.767638 0.255879 33 32.349503 0.980288 36 33.117141 Lack ofFit DF Sum of Squares Mean Square 32 32.253360 1.00792 1 0.096143 0.09614 33 32.349503 F Ratio 0.2610 Prob>F 0.8529 F Ratio 10.4835 Prob>F 0.2406 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENDER YRSCBEM AGE Estimate -3.483257 -0-125208 6.001537 0.0047368 Parameter Estimates Std Error tRatio 1.02006 -3.41 0.199053 -0.63 0.026026 0.06 0.031727 0.15 Prob>ltl 0.0017 0.5337 0.9533 0.8822 Lower 95% -5.558572 -0.530181 -0.051413 -0.059812 Nparm 1 1 I EffectTest DF Sum of Squares 1 0.38786407 1 0.00341911 0.02185(Y78 F Raflo 0.3957 0.0035 0.0223 Prob>F 0.5337 0.9533 0.8822 Upper 95% -1.407941 0.2797653 0.0544869 0.0692859 APPendix L Page 11 AllParticipants Engineer/Lab InM570ppm Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observation(sorSum Wgts) 0.02252 -0.03498 0.975755 -3.19558 37 Source Model Effor C Total Source Lack ofFit PureE r TotalEffor Max RSq 0.3233 AnalysisofVariance DF Sum ofSquares Mean Square 2 0.745787 0.372894 34 32.371354 0.952099 36 33.117141 Lack ofFit DF Sum ofSquares Mean Square 19 9.959941 0.52421 15 22.411413 1.49409 34 32.371354 F Ratio 0.3917 Prob>F 0.6789 F Ratio 0.3509 Prob>F 0.9833 Term intercept GENDER[F-M] YRSCHEM Source GENDER YRSCHEM Estimate -3.335748 -0.126747 0.0049036 ParameterEstimates Std Error tRatio 0.250023 -13.34 0.195906 -0.65 0.012808 0.38 Prob>ltl <.0001 0.5220 0.7042 Lower 95% -3.843852 -0.524874 -0.021125 Nparm 1 1 EffectTest DF Sum ofSquares 1 0.39853117 1 0.13955278 F Ratio 0.4186 0.1466 Prob>F 0.5220 0.7042 Upper 95% -2.827644 0.2713797 0.0309325 Appendix L Page 12 Source Model Error C Total Source Lack ofFit Pure Error TotalError Max RSq 0.9958 AllParticipants Engineer/Lab RSquare InPFOSA ppm Summary of Fit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.022151 -0.06674 1.616733 -5.7372 37 Analysisof Variance DF Sum ofSquares Mean Square 3 1.953947 0.65132 33 86.256256 2.61383 36 88.210203 Lack ofFit DF Sum ofSquares Mean Square 32 85.888903 2.68403 1 0.367353 0.36735 33 86.256256 F Ratio 0.2492 Prob>F 0.8613 F Ratio 7.3064 Prob>F 0.2861 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENDER YRSCBEM AGE Estimate -5.523456 0.2341119 0.0199578 -0-00946 Parameter Estimates Std Error tRatio 1.665664 -3.32 0.325034 0.72 0.042498 0.47 0.051808 -0.18 Prob>ltl 0.0022 0.4764 0.6417 0.8562 Lower 95% -8.912253 -0.427171 -0-06804 -0.114862 Nparm 1 1 1 EffectTest DF Sum ofSquares 1 1.3560164 1 0.5764525 1 0.0871452 F Ratio 0.5188 0.2205 0.0333 Prob>F 0.4764 0.6417 0.8562 Upper 95% -2.134659 0.895395 0.10642 0.0959429 APPendixL Page 13 AlPlarticipants Engineer/Lab InPFOSA ppm Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.021163 -0.03642 1.593585 -5.7372 37 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 2 1.866802 0.93340 34 86.343402 2.53951 36 88.210203 F Ratio 0.3676 Prob>F 0.6951 Source Lack ofFit Pure E TotalError Max RSq 0.7369 Lack ofFit DF Sum ofSquares Mean Square F Ratio 19 63.139191 3.32312 2.1482 15 23.204211 1.54695 Prob>F 34 86.343402 0.0689 Term Intercept GENDER[F-M) YRSCBEM Source GENDER YRSCBEM Estimate -5.818037 0.2371866 0.0132347 ParameterEstimates StdError tRatio 0.408332 -14.25 0.31995 0.74 0.020918 0.63 Prob>ltl <.0001 0.4636 6.5312 Lower 95% -6.647863 -0.413027 -0.029275 Nparm 1 1 EffectTest DF Sum ofSquares 1 1.3956145 1 1.0165846 F Ratio 0.5496 0.4003 Prob>F 0.4636 0.5312 Upper 95% -4.988212 0.8873997 0.0557446 APPendix L Page 14 Source Model Error C Total Source Lack ofFit PureEffor TotalFxror Max RSq 0.9995 Term Intercept GENDER[F-M] YRSCHEM AGE Source GENDER YRSCHEM AGE AllParticipants Engineer/Lab InM556 ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observation(sorSum Wgts) 0.114795 0.034322 1.165251 -4.61966 37 AnalysisofVariance DF Sum ofSquares Mean Square 3 5.810766 1.93692 33 44.807703 1.35781 36 50.618468 Lack of Fit DF Sum ofSquares Mean Square 32 44.781948 1.39944 1 0.025755 0.02575 33 44.807703 F Ratio 1.4265 Prob>F 0.2526 F Ratio 54.3366 Prob>F 0.1071 Estimate -7.003379 -0.103922 -0.045945 0.0719096 ParameterEstimates StdError tRatio 1.200517 -5.83 0.234267 -0.44 0.03063 -1.50 0.03734 1.93 Prob>ltl <.0001 0.6602 0.1431 0.0628 Lower 95% -9-445834 -0.580538 -0.108262 -0.004059 Npaffn 1 1 1 EffectTest DF Sum of Squares 1 0.2671974 1 3.0550141 1 5.0357474 F Ratio 0.1968 2.2500 3.7087 Prob>F 0.6602 0.1431 0.0628 Upper 95% -4.560924 0.372694 0.0163723 0.1478779 APPendix L Page 15 AllParticipants Engineer/Lab Source Model Error C Total Source Lack ofFit PureError TotalError Max RSq 0.3732 InM556 ppm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.015311 -0.04261 1.210778 -4.61966 37 AnalysisofVariance DF Sum ofSquares Mean Square 2 0.775018 0.38751 34 49.843450 1.46598 36 50.618468 Lack ofFit DF Sum ofSquares Mean Square 19 18.113852 0.95336 15 31.729598 2.11531 34 49.843450 F Ratio 0.2643 Prob>F 0.7693 F Ratio 0.4507 Prob>F 0.9486 Tenn Intercept GENDER(F-M] YKSCBEM Source GENDER YRSCBEM Estimate -4.764061 -0.127295 0.0051619 ParameterEstimates StdError tRatio 0.310244 -15.36 0.243093 -0.52 0.015893 0.32' P.rob>ltl <.0001 0.6039 0.7473 Lower 95% -5.394548 -0.621316 -0.027136 Nparm 1 1 EffecTtest DF Sum ofSquares 1 0.40198223 1 0.15464711 F Ratio 0.2742 0.1055 Prob>F 0.6039 0.7473 Upper 950/a -4.133574 0.3667259 0.0374602 Appendix m Page I Api)endixM Scatterplot(sand regressionso)f fluorochemicallevelsof a chemicalparticipant male chemicaloperators(n = 52) and engineer/la(bn= 28) with yearsworked in chemical AllParticipants Male ChemicalOperators InPFOS ppm By YRSCHEM APPendix M Page 2 r= O.DD -D.5D -1.50 0 5 10 15 20 25 Jo JS YRSCREM Term Intercept YRSCHEM POMTD R dwn=2 Source Model Error C Total LinearFit InPFOSdfppm 0.28294+ 0.01961YRSCHEM Summary ofFit RSquare 0.077877 RSquare Adj 0.059435 Root Mean SquareError 0.65096 Mean of Response 0.494658 Observation(sorSum Wgts) 52 AnalysisofVariance DF Sum ofSquares Mean Square 1 1.789373 1.78937 50 21.187429 0.42375 51 22.976802 F Ratio 4.2227 Prob>F 0.0451 Estimate 0.2829416 0.0196069 ParameterEstimates Std Error tRatio 0.136981 2.07 0.009541 2.05 Prob>ltl 0.0441 0.0451 Lower 95% 0.0078068 0.0004424 Upper 95% 0.5580763 0.0387713 APpendix m Page 3 PolynomiFailtdegree=2 InPFOSdfppm = 0.07855+ 0.08713YRSCHEM - 0.00255YRSCHEMA2 RSquare Summary ofFit 0.152148 RSquare Adj Root Mean Square Error 0.117542 0.630532 Mean of Response 0.494658 Observations(orSum Wgts) 52 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 2 3.495868 1.74793 49 19.480934 0.39757 51 22.976802 F Ratio 4.3965 Prob>F 0.0175 Term Intercept YRSCHEM YRSCHEMA 2 Estimate 0.0785522 0.0871341 -0.002546 Parameter Estimates Std Error t Ratio 0.16534 0.48 0.033879 2.57 0.001229 -2.07 Prob>ltl 0.6368 0.0'132 0.0436 Lower 95% -0.25371 0.0190526 -0.005016 Upper 95% 0.4108141 0.1552157 -0-000076 AllParticipants MaleChemicalOperators InPFHS ppm By YRSCHEM 0.5D D.DD -0.50 r= I.DD -1.50 -2.00 Mae Apperidix m Page 4 0 5 10 15 20 25 30 35 YRSCHDA - Limft LinearFit InPFHSdfppm = -1.538+5 0.04363YRSCBEM Summary ofFit RSquare 0.293355 RSquareAdj 0.279222 RootMean SquareError 0.653321 Mean ofResponse -1-06736 Observatio(nosrSum Wgts) 52 Source Model Error C Total AnalysiosfVariance DF Sum ofSquares Mean Square 1 8.859630 8.85963 50 21.341385 0.42683 51 30.201015 F Ratio 20.7569 Prob>F <.0001 Term Intercept YRSCHEM Estimate -1.538462 0.043628 ParameterEstimates StdError tRatio Prob>ltl 0.137478 -11.19 <.0001 0.009576 4.56 <.0001 Lower95% -1.814595 0.0243941 Upper950/o -1.262329 0.0628619 Appendix m Page 5 PolynomiFaidltegree=2 InPFHSdfppm = -1.7114+ 0.10078YRSCHEM - 0.00215YRSCHEMA 2 Summary of Fit RSquare 0.33383 RSquare Adj 0.306639 Root Mean Square Error 0.640775 Mean of Response -1-06736 Observations(orSum Wgts) 52 Source Model Error C Total Analysisof Variance DF Sum ofSquares Mean Square 2 10.082003 5.04100 49 20.119012 0.41059 51 30.201015 F Ratio 12.2774 Prob>F <.0001 Term intercept YRSCHEM YRSCHEMA 2 Estimate -1.711447 0.1007796 -0.002155 Parameter Estimates Std Error tRatio 0.168026 -10.19 0.034429 2.93 0.001249 -1.73 Prob>ltl <.0001 0.0052 0.0908 Lower 95% -2.049106 0.0315921 -0.004665 Upper 95% -1.373787 0.1699671 0.0003549 AllParticipants Male Chemical Operators InPOAA ppm By YRSCHEM ?.OD 1.5D 1.00 Appendix m Page 6 0 51 0 15 20 25 30 35 Term Intercept YRSCHEM Poi@ R dmee:2 LinearFit InPOAAppm 0.55713 + 0.01691 YRSC]HEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.09096 0.072779 0.515758 0.739719 52 Source DF Model 1 Error 50 C Total 51 Estimate 0.5571329 0.0169091 AnalysisofVariance Sum ofSquares 1.330847 13.300322 14.631169 Parameter Estimates Std Error tRatio 0.108531 5.13 0.00756 2.24 Mean Square 1.33085 0.26601 F Ratio 5.0031 Prob>F 0.0298 Prob>ltl <.0001 0.0298 Lower 95% 0.3391425 0.0017251 Upper 95% 0.7751233 0.0320932 APPendix m Page 7 -PolynomiaFlitdegree=2 InPOAAppm 0.30559+ 0.10002YRSCHEM - 0.00313YRSCHEMA 2 Summary ofFit RSquare 0.26762 RSquare Adj 0.237727 Root Mean SquareError 0.467638 Mean ofResponse 0.739719 Observation(sorSum Wgts) 52 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 2 3.915590 1.95779 49 10.715579 0.21869 51 14.631169 F Ratio 8.9526 Prob>F 0.0005 Term Intercept YRSCBEM YRSCHEMA 2 Estimate 0.305@886 0.1000157 -0.003133 ParameterEstimates Std Error tRatio 0.122625 2.49 0.025126 3.98 0.000911 -3.44 Prob>ltl 0.0161 0.0002 0.0012 Lower 95% 0.0591644 0.0495226 -0.004965 Upper 95% 0.5520129 0.1505088 -0-001302 AllParticipants MaleChemicalOperators InPFOSAA ppm By YRSCHEM Appendix M Page 8 Term Intercept YRSCHEM -7.00 05 10 15 20 25 30 JS YRSCREM Source Model Error C Total Lim Linear In PFOSAAdfppm RSquare RSquare Adj = -3.8496 Summary Root Mean Square Error Mean of Response Observations (or Sum Wgts) Fit - 0.02457 of Fit YRSCHEM 0.021349 0.001776 1.604937 -4.11489 52 Analysis of Variance DF Sum of Squares 1 2.80958 50 128.79120 51 131.60078 Mean Square 2.80958 2.57582 F Ratio 1.0908 Prob>F 0.3013 Estimate -3.849596 -0.024568 Parameter Std Error 0.337727 0.023524 Estimates t Ratio -11.40 -1.04 Prob>ltl <.0001 0.3013 Lower 95% -4.527939 -0.071818 Upper 95% -3.171253 0.0226812 AllParticipants Male Chemical Operators In M570ppm By YRSCHEM 1,00 D.OD -1.00 Apperidix m Page 9 -S.DO 0 5 10 15 20 25 30 35 YRSCHEW Lim Term Intercept YRSCHEM Source Model Error C Total LinearFit In 570ppm = -1.3268- 0.04752 YRSCREM Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.129273 0.111859 1.189837 -1.83989 52 Analysisof Variance DF Sum of Squares Mean Square 1 10.509261 10.5093 50 70.785564 1.4157 51 81.294825 F Ratio 7.4233 Prob>F 0.0088 Estimate -1.326808 -0.047516 Parameter Estimates Std Error tRatio 0.250377 -5.30 0.01744 -2.72 Prob>ltl <.0001 0.0088 Lower 95% -1.829705 -0.082545 Upper 95% -0.823911 -0.012487 AllParticipants Male Chemical Operators InPFOSA ppm By YRSCHEM Appendix m Page 10 r 15 lo 20 25 30 35 YRSCHIM Term intercept YRSCHEM = Lim R Source Model Error C Total LinearFit InPFOSAdfppm -3.3047- 0.04865YRSCHEM Summary ofFit RSquare RSquare Adj 0.0581 0.039262 Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 1.890083 -3.83009 52 AnalysisofVariance DF Sum ofSquares Mean Square 1 11.01808 11.0181 50 178.62062 3.5724 51 189.63870 F Ratio 3.0842 Prob>F -0.0852 Estimate -3.304729 -0.048653 Parameter Estimates StdError tRabo 0.39773 -8.31 0.027704 -1.76 Prob>ltl <.0001 0.0852 Lower 95% -4.103591 -0.104298 Upper 95% -2.505866 0.0069914 AllParticipants Male Chemical Operators InM556 ppm By YRSCHEM -1.00 E Apperidix m Page 11 -6.00 15 0 5 10 20 25 30 35 YRSCHEM Term intercept YRSCHEM Lim r LinearFit In M556dfppm = -2.6395 - 0.0315 YRSCHEM Summary of Fit RSquare 0.075514 RSquare Adj 0.057024 Root Mean Square Error 1.063505 Mean of Response -2.97966 Observations (orSum Wgts) 52 Source Model Error C Total Analysisof Variance DF Sum of Squares Mean Square 1 4.619303 4.61930 50 56.552145 1.13104 51 61.171448 F Ratio 4.0841 Prob>F .0.0487 Estimate -2.639497 -0.031503 Parameter Esfimates Std Error tRatio 0.223793 -11.79 0.015588 -2.02 Prob>ltl <.0001 0.0487 Lower 95% -3.088999 -0.062812 Upper 95% -2.189996 -0.000193 AllParticipants MaleEngineer/Lab InPFOS ppm ByYRSCHEM Appendix m Page 12 C.DO E1.00 -2.DO -2.50 0 5 10 15 20 25 ID 35 40 YRSCHEM - Limft - P*m7if@tdevee=2 Source Model Error C Total UnearFit InPFOSdfppm = -1.251+5 0.03365YRSCHEM Summary ofFit RSquare 0.208121 RSquareAdj 0.177664 RootMean SquareError 0.886126 Mean ofResponse -0.66143 Observatio(nosrSum Wgts) 28 AnalysiosfVariance DF Sum ofSquares Mean Square 1 5.365639 5.36564 26 20.415689 0.78522 27 25.781328 F Ratio 6.8333 Prob>F 0.0147 Term Intercept YRSCHEM Estimate -1.251482 0.0336488 ParameteErstimates StdError tRatio Prob>ltl 0.28106 -4.45 0.0001 0.012872 2.61 0.0147 Lower95% -1.829204 0.0071897 Upper95% -0.673759 0.0601078 APPendix M Page 13 PolynomiaFlitdogree=2 InPFOSdfppm = -1.6361+ 0.13222YRSCHEM - 0.00282YRSCHEMA 2 Summary of Fit RSquare 0.307474 RSquare Adj 0.252072 Root Mean SquareError 0.845086 Mean ofResponse -0.66143 Observations(orSum Wgts) 28 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 2 7.927078 3.96354 25 17.854250 0.71417 27 25.781328 F Ratio 5.5499 Prob>F 0.0101 Term Intercept YRSCBEM YRSCHEMA 2 Estimate -1.636113 0.1322248 -0.002819 Parameter Estimates Std Error tRatio 0.336297 -4.87 0.053479 2.47 0.001489 -1.89 Prob>ltl <.0001 0.0206 0.0699 Lower 95% -2.328723 0.0220833 -0.005885 Upper 95% -0.943503 0.2423664 0.0002466 Appendix M Page 14 AlPlarticipants Male EngineedLab InPFHS ppm By YRSCHEM -0.50 -1.50 -2.50 - -I- -3.50 -4.50 0 5 10 15 20 25 30 35 10 YRSCHEW Term Intercept YRSCBEM - LimFi - POwTdR date:2 Unear Fit InPFHSdfppm = -2.9522+ 0.04106YRSCBEM Summary ofFit RSquare 0.243114 RSquare Adj 0.214003 Root Mean SquareError 0.978058 Mean of Response -2.23224 Observation(sorSum Wgts) 28 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 7.988791 7.98879 26 24.871516 0.95660 27 32.860307 F Ratio 8.3513 Prob>F 0.0077 Estimate -2.95222 0.0410581 ParameterEstimates Std Error tRatio 0.310219 -9.52 0.014208 2.89 Prob>ltl <.0001 0.0077 Lower95% -3.589879 0.0118541 Upper 95% -2.314561 0.0702621 . APPeridix m Page 15 Term Intercept YRSCHEM YRSCHEMA2 PolynomialFitdegree=2 InPFHSdfppm = -3.5713+ 0.19973YRSCHEM - 0.00454YRSCHEMA 2 Summary of Fit RSquare 0.445082 RSquare Adj 0.400688 Root Mean Square Error 0.854044 Mean of Response -2.23224 Observations(orSum Wgts) 28 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 2 14.625528 7.31276 25 18.234779 0.72939 27 32.860307 F Ratio 10.0258 Prob>F 0.0006 Estimate -3.571347 0.1997324 -0.004538 Parameter Estimates Std Error 0.339861 t Ratio -10.51 0.054046 3.70 0.001504 -3.02 Prob>ltl <.0001 0.0011 0.0058 Lower 95% 4.2713 0.0884234 -0.007636 Upper 950/a -2.871395 0.3110415 -0.00144 AllParticipants Male Engineer/Lab InPOAAppm By YRSCHEM Appendix M Page 16 5 20 25 30 35 40 YRSCHEM Term Intercept YRSCHEM Lim R P*ruW R devee=2 Source Model Error C Total LinearFit .InPOAAppm -1.6429+ 0.01806YRSCHEM Summary ofFit RSquare 0.063628 RSquare Adj Root Mean Square Error 0.027613 0.935202 Mean of Response Observations(orSum Wgts) -1.32623 28 AnalysisofVariance DF Sum ofSquares Mean Square 1 1.545191 1.54519 26 22.739675 0.87460 27 24.284865 F Ratio 1.7667 Prob>F 0.1953 Estimate -1.642879 0.0180572 Parameter Estimates Std Error t Ratio 0.296626 -5.54 0.013585 1.33 Prob>ltl <.0001 0.1953 Lower 95% -2.252597 -0.009867 Upper 95% -1.03316 0.0459816 low. Appendix M Page 17 PolynomialFitdegree=2 InPOAAppm -2.1643+ 0.15169YRSCHEM - 0.00382YRSCHEMA 2 Summary ofFit RSquare 0.257455 RSquare Adj 0.198052 Root Mean Square Error 0.849296 Mean of Response -1.32623 Observations(orSum Wgts) 28 Source Model Error C Total AnalysisofVariance DF Sum of Squares Mean Square 2 6.252264 3.12613 25 18.032601 0.72130 27 24.284865 F Ratio 4.3340 Prob>F 0.0242 Term Intercept YRSCHEM YRSCHEMA 2 Estimate -2.164287 0.1516874 -0.003821 Parameter Estimates Std Error t Ratio 0.337972 -6.40 0.053746 2.82 0.001496 -2.55 Prob>ltl <.0001 0.0092 0.0171 Lower 95% -2.860348 0.0409971 -0.006902 Upper 95% -1.468226 0.2623777 -0.000741 AllParticipants Male Engineer/Lab InPFOSAA ppm By YASCHEM -J.00 APPeridix M Page 18 -5.DD -6.00 -7-00 0 5 10 15 20 25 30 35 0 YRSCHEM Lim Term Intercept YRSCHEM Source Model Error C Total LinearFit InPFOSAAdfppm = -5.1411- 0.0164YRSCHEM Summary ofFit RSquare 0.024437 RSquare Adj -0.01308 Root Mean SquareError 1.399249 Mean of Response -5.4287 Observation(sorSum Wgts) 28 AnalysisofVariance DF Sum ofSquares Mean Square 1 1.275122 1.27512 26 50.905372 1.95790 27 52.180494 F Raflo 0.6513 Prob>F 0.4270 Estimate -5.141055 -0.016403 ParameterEstimates Std Error tRatio 0.443812 -11.58 0.020326 -0.81 Prob>ltl <.0001 0.4270 Lower 95% -6.053316 -0.058184 Upper 95% -4.228794 0.0253771 AllParticipants Male Engineer/Lab InM570ppm By YRSCHEM APPendix m Page 19 J.00 1.00 -5.00 0 5 10 15 20 25 30 35 40 YRSCHEM Term Intercept YRSCHEM Lim Source Model Error C Total LinearFit In570ppm = -3.1804+ 0.00328YRSCHFM Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.00186 -0.03653 1.024495 -3.12301 28 AnalysisofVariance DF Sum ofSquares Mean Square 1 0.050840 0.05084 26 27.289336 1.04959 27 27.340175 F Ratio 0.0484 Prob>F 0.8275 Estimate -3.180449 0.0032754 ParameterEstimates Std Error tRatio 0.324948 -9.79 0.014882 0.22 Prob>ltl <.0001 0.8275 Lower 95% -3.848383 -0.027315 Upper 95% -2.512515 0.033866 AllParticipants Male Engineer/Lab InPFOSA ppm By YRSCHEM APPendix M Page 20 6.DO --- - ----------------------------- -7.00 T 5 10 15 2D 25 30 35 to YRSCHEM Term Intercept YRSCBEM Un R LinearFit InPFOSAdf@pm = -6.0798+ 0.01464YRSCREM Summary ofFit RSquare 0.014659 RSquare Adj Root Mean SquareError -0.02324 1.620061 Mean of Response -5.82314 Observation(sorSum Wgts) 28 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 1.015219 1.01522 26 68.239528 2.62460 27 69.254747 F Ratio 0.3868 Prob>F 0.5394 Estimate -6.079806 0.0146365 ParameterEstimates Std Error tRatio 0.513848 -11.83 0.023534 0.62 Prob>ltl <.0001 0.5394 Lower 95% -7-136028 -0.033737 Upper 95% -5.023584 0.0630103 AllParticipants Male Engineer/Lab InM556 ppm By YRSCHEM -l.DO -2.00 Apperidix M Page 21 -6.00 0 5 10 15 20 25 30 35 10 YRSCREM Term Intercept YRSCHEM Lim r LinearFit InM556dfppm = -4.5528+ 0.00037YRSCHEM Summary ofFit RSquare 0.000015 RSquare Adj -0.03845 Root Mean Square Error 1.297208 Mean ofResponse -4.54625 Observations(orSum Wgts) 28 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.000661 0.00066 26 43.751480 1.68275 27 43.752141 F Ratio 0.0004 Prob>F 0.9843 Estimate -4.552795 0.0003734 Parameter Estimates Std Error t Ratio 0.411446 -11.07 0.018844 0.02 Prob>ltl <.0001 0.9843 Lower 95% -5.398529 -0.03836 Upper 95% -3.707061 0.039107 AppendiX N Page 1 Ai)i)endiNx Scatterplot(sand regressions)of fluorochemicallevelsof allchemical participant female chemical operators(n = 12) and engineer/lab(n 9) with years worked in chemical AllParticipants Female Chemcial Operators InPFOS ppm By YRSCHEM E Appendix N Page 2 10 Lim Term Intercept YRSCHEM LinearFit InPFOSdfppm 0.01226- 0.00779YRSCHEM Summary ofFit RSquare 0.008557 RSquare Adj -0.09059 Root Mean SquareError 0.578093 Mean ofResponse -0.05134 Observations(orSum Wgts) 12 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.0288427 0.028843 10 3.3419208 0.334192 11 3.3707635 F Ratio 0.0863 Prob>F 0.7749 Estimate 0.0122559 -0.007787 Parameter Estimates Std Error tRatio 0.273326 0.04 0.026506 -0.29 Prob>ltl 0.9651 0.7749 Lower 95% -0.596757 -0.066847 Upper 95% 0.6212687 0.0512728 AllParticipants Female Chemcial Operators InPFHS ppm By YASCHEM APPendix N Page 3 E -2.D -2.5 YRSCHEM -r.J 20 =-Lim Term Intercept YRSCHEM Source Model Error C Total LinearFit InPFHSdfppm -2.3774+ 0.05385YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.275351 0.202886 0.602463 -1.93766 12 AnalysisofVariance DF Sum ofSquares Mean Square 1 1.3791706 1.37917 10 3.6296113 0.36296 11 5.0087819 F Ratio 3.7998 Prob>F 0.0798 Estimate -2.377407 0.0538465 Parameter Estimates Std Error t Ratio 0.284848 -8.35 0.027623 1.95 Prob>ltl <.0001 0.0798 Lower 95% -3.012092 -0.007703 Upper 95% -1.742721 0.1153959 AllParticipants Female Chemcial Operators InPOAA ppm By YRSCHEM IID 0 so O.DO E C- 0.50 -1.00 -1.50 10 YRSCHEW Apperidix N Page 4 Term Intercept YRSCEEM ---L=R LinearFit InPOAAppm 0.32148- 0.03055 YRSCEEM Summary ofFit RSquaTe 0.079815 RSquare Adj -0.0122 Root Mean SquareEffor 0.715366 Mean of Response 0.072008 Observations(orSum Wgts) 12 Source Model Effor C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.4438781 0.443878 10 5.1174828 0.511748 11 5.5613609 F Ratio 0.8674 Prob>F 0.3736 Estimate 0.3214819 -0.030548 Parameter Estimates Std Error tRatio 0.33823 0.95 0.0328 -0.93 Prob>ltl 0.3643 0.3736 Lower 95% -0.432145 -0.103632 Upper 95% 1.075109 0.0425361 AllParticipants Female Chamcial Operators InPFOSAA ppm By YRSCHEM -iDo APPendix N Page 5 5.00 -a -6.00 le 20 ISCHEM Term Intercept YRSCHEM Source Model Error C Total LinearFit InPFOSAAdfppm 4.5678 - 0.0569YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.06068 -0.03325 1.543965 -5.0325 12 AnalysisofVariance DF Sum ofSquares Mean Square 1 1.539949 1.53995 10 23.838281 2.38383 11 25.378230 F Ratio 0.6460 Prob>F 0.4402 Estimate -4.56783 -0.056899 Parameter Estimates Std Error tRatio 0.729997 -6.26 0.070792 -0.80 Prob>ltl <.0001 0.4402 Lower 95% -6.194374 -0.214635 Upper 950/o -2.941286 0.1008375 AllParticipants Female Chemcial Operators InM570 ppm By YRSCHEM -IJ E- -2.3 APPendix N Page 6 2; YISCHEY Term Intercept YRSCHEM Lim R Source Model Error C Total Unear Fit In570ppm -2.1009- 0.0371YRSCHEM Summary of Fit RSquare RSquare Adj Root Mean SquareError Mean of Response Observation(sorSum Wgts) 0.111759 0.022934 0.721322 -2.40387 12 AnalysisofVariance DF Sum ofSquares K4ean Square 1 0.6546479 0.654648 10 5.2030495 0.520305 11 5.8576974 F Ratio 1.2582 Prob>F 0.2882 Estimate -2.100903 -0.037098 ParameterEstimates Std Error tRatio 0.341046 -6.16 0.033073 -1.12 Prob>ltl 0.0001 0.2882 Lower 95% -2.860804 -0-110791 Upper 95% -1.341001 0.0365942 AllParticipants Female Chemcial Operators InPFOSA ppm By YRSCHEM -2.00 -J.DD < 1.00 5c- Apperidix N Page 7 -6.00 0 10 20 Lim Term Intercept YRSCHEM Unear Fit InPFOSAdfppm -4.6226+ 0.0764YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.095234 0.004758 1.624037 -3.99866 12 Source Model Error C Total AnalysisofVariance DF Sum of Squares Mean Square 1 2.776184 2.77618 10 26.374947 2.63749 11 29.151132 F Ratio .1.0526 Prob>F 0.3291 Estimate 4.622563 0.0763964 Parameter Estimates Std Error t Ratio 0.767855 -6.02 0.074464 1.03 Prob>ltl 0.0001 0.3291 Lower 95% -6.333461 -0.08952 Upper 95% -2.911665 0.2423128 AllParticipants Female Chemcial Operators InM556 ppm By YRSCHEM -J.0E AppendiX N Page 8 0 10 20 YRSCOGI Term Intercept YRSCHEM Lim Source Model Error C Total UnearFd InM556dfppm -3.1494- 0.05942YRSCHEM Summary ofFit RSquare 0.159673 RSquareAdj 0.075641 Root Mean SquareError 0.940158 Mean ofResponse -3.63466 Observation(sorSum Wgts) 12 AnalysisofVariance DF Sum ofSquares k4eanSquare 1 1.679525 1.67953 10 8.838975 0.88390 11 10.518500 F Ratio .1.9001 Prob>F 0.1981 Estimate -3.149387 -0.059421 ParameterEstimates StdError tRatio 0.444513 -7.09 0.043107 -1.38 Prob>ltl <.0001 0.1981 Lower 950/o 4.13983 -0.155471 Upper 95% -2.158945 0.0366281 AllParticipants Female Engineer/Lab In PFOS ppm By YRSCHEM E Appendix N Page 9 -2.60 0 5 10 15 20 25 30 35 Term Intercept YRSCHEM Lim ri j Unear Fit InPFOSdfppm -1.8939+ 0.01024YRSCHEM Summary ofFit RSquare 0.058759 RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) -0.0757 0.505477 -1.80801 9 Source Model ErTor C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.1116544 0.111654 7 1.7885504 0.255507 8 1.9002048 F Ratio .0.4370 Prob>F 0.5297 Estimate -1.893913 0.0102397 Parameter Estimates Std Error t Ratio 0.212779 -8.90 0.01549 0.66 Prob>ltl <.0001 0.5297 Lower 95% -2.397059 -0.026388 Upper 95% -1.390766 0.0468678 AllParticipants Female Engineer/Lab InPFHSppmBYYRSCHEM -100 APPendix N Page 10 0 5 la 15 20 25 JO J5 YRscfl[M Term Intercept YRSCBIEM Source Model Error C Total LinearFit InPFHSdfppm -3-9868+ 0.03015 YRSCBEM RSquare Summary of Fit RSquare Adj Root Mean Square Error Mean of Response Observations(orSum Wgts) 0.138988 0.015987 0.925469 -3.73389 9 AnalysisofVariance DF Sum of Squares Mean Square 1 0.9678134 0.967813 7 5.9954503 0.856493 8 6.9632637 F Rato 1.1300 Prob>F 0.3231 Estimate -3.986788 0.030147 Parameter Estimates Std Error t Ratio 0.389573 -10.23 0.02836 1.06 Prob>ltl <.0001 0.3231 Lower 95% -4-90799 -0-036915 Upper 95% -3.065586 0.0972088 AllParticipants Female Engineer/Lab InPOAA ppm By YRSCHEM -1vo APpendix N Page -2.50 -J.00 0 5 10 15 20 25 30 35 YRSCHEM Term Intercept YRSCHEM LinearFit InPOAAppm -2.5628+ 0.00289YRSCHIEM Summary ofFit RSquare 0.001765 RSquare Adj -0.14084 Root Mean Square Error 0.848257 Mean of Response -2.53853 Observations(orSum Wgts) 9 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.0089035 0.008904 7 5.0367756 0.719539 8 5.0456791 F Ratio 0.0124 Prob>F 0.9145 Estimate -2.562785 0.0028915 Parameter Estimates Std Error tRatio 0.357071 -7.18 0.025994 0.11 Prob>ltl 0.0002 0.9145 Lower 95% -3.40713 -0.058575 Upper 95% -1.718439 0.0643584 AllParticipants Female Engineer/Lab InPFOSAA ppm By YRSCHEM APPerldixN Page 12 -6.00 -6.50 -1.00 5 io is 20 25 io J5 ISCHIM Term intercept YRSCBEM Source Model Error C Total LinearFit InPFOSAAdfppm = -6.1434- 0.02633YRSCHEM Summary ofFit RSquare 0.103266 RSquare Adj -0.02484 Root Mean SquareError 0.957088 Mean ofResponse Observations(orSum Wgts) -6.36431 9 AnalysisofVariance DF Sum ofSquares Mean Square 1 0.7384073 0.738407 7 6.4121225 0.916018 8 7.1505298 F Ratio .0.8061 Pmb>F 0.3991 Estimate -6.143411 -0.026333 Parameter Estimates Std Error t Ratio 0.402883 -15.25 0.029329 -0.90 Prob>ltl <.0001 0.3991 Lower 95% -7.096087 -0.095686 Upper 95% -5.190736 0.0430203 AllParticipants Female Engineer/Lab InM570 ppm By YRSCHEM AppendiX N Page 13 -I.oo io is 20 25 IQ is YRSCHIM = Lim Term Intercept YRSCHEM Source Model Error C Total LinearFit In570ppm -3.5233+ 0.01215 YRSCHEM Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean of Response Observations(orSum Wgts) 0.030399 -0.10812 0.846298 -3.42136 9 Analysisof Variance DF Sum ofSquares Mean Square 1 0.1571845 0.157185 7 5.0135469 0.716221 8 5.1707315 F Ratio .0.2195 Prob>F 0.6537 Estimate -3.523279 0.0121493 Parameter Estimates Std Error t Ratio 0.356247 -9.89 0.025934 0.47 Prob>ltl <.0001 0.6537 Lower 95% -4.365676 -0.049176 Upper 95% -2.680883 0.0734743 AllParticipants Female Engineer/Lab InPFOSA ppm By YRSCHEM -joo APperidix N Page 14 E S.DO 5.DD -7.00 -B.DO 0 5 10 15 ?0 25 io J5 YRSCREM = Lim Term Intercept YRSCHEM Source Model Error C Total LinearFit InPFOSAdfppm = -5.5285+ 0.007YRSCHEM Summary gf Fit RSquare 0.002879 RSquare Adj -0.13957 Root Mean SquareError 1.605932 Mean of Response -5.46983 Observation(sorSum Wp) 9 AnalysisofVariance DF Sum ofSquares Mean Square 1 0.052123 0.05212 7 18.053117 2.57902 8 18.105239 F Ratio 0.0202 Prob>F 0.8910 Estimate -5.528517 0.0069962 Parameter Estimates Std Error tRatio 0.676012 -8.18 0.049212 0.14 Prob>ltl <.0001 0.8910 Lower 950/o -7.127044 -0.109374 Upper 95% -3.929989 0.123366 AllParticipants Female Engineer/Lab InM556 ppm By YRSCHEM -j Apperidix N Page 1.5 E -5.5D -6.DD 0 5 10@ is 20 25 30 J5 YISCHIM = Lim Term Intercept YRSCHEM Source Model Error C Total LinearFit InM556dfppm -5.0701+ 0.02647YRSCHEM Summary ofFit RSquare . 0.119476 RSquare Adj -0.00631 Root Mean SquareError 0.886382 Mean of Response -4.84805 Observations(orSum Wgts) 9 AnalysisofVariance DF Sum of Squares Mean Square 1 0.7462412 0.746241 7 5.4997153 0.785674 8 6.2459565 F Ratio .0.9498 Prob>F 0.3622 Estimate -5.070124 0.026472 Parameter Estimates Std Error t Ratio 0.37312 -13.59 0.027162 0.97 Prob>ltl <.0001 0.3622 Lower 95% -5.952419 -0.037757 Upper 95% 4.187828 0.0907015 Appendix 0 Page 1 Ai)i)end0ix Scatterplots(and regressions)of fluorochemicallevelsof random sample who worked were only inthe filmplant(n = 36) with yearsworked in film RandomSample Only FilmEmployees (MaintenanceWorkers Numbered) INPFOS ppm By YrsFilm -2.DD r= 25-1.21 APpendix0 Page 2 0 5 10 is 20 25 30 IriFin - LimR - PokrarRdduee:2 LinearFit InPFOSdfppm = -2-3024+ 0.00313YrsFilm Summary ofFit RSquare 0.002948 RSquare Adj -0.02638 Root Mean SquareError 0.585965 Mean of Response -2.25946 Observation(sorSum Wgts) 36 Source Model Error C Total AnalysisofVaHance DF Sum ofSquares Mean Square 1 0.034516 0.034516 34 11.674079 0.343355 35 11.708595 F Ratio 0.1005 Prob>F 0.7531 Term Intercept YrsFilm Estimate -2.30237 0.0031336 ParameterEstimates StdError tRatio Prob>ltl 0.166902 -13.79 <.0001 0.009883 0.32 0.7531 Lower 95% -2.641553 -0.016952 Upper 950/o -1.963187 0.0232187 Appendix 0 Page 3 Term Intercept YrsFilm YrsFilmA2 PolynomiFailtdGgree=2 InPFOSdfppm = -2-5117+ 0.06209YrsFilm- 0.0021YrsFilm"2 Summary of Fit RSquare 0.083482 RSquare Adj 0.027935 Root Mean Square Error 0.570251 Mean of Response -2.25946 Observations(orSum Wgts) 36 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 2 0.977453 0.488726 33 10.731142 0.325186 35 11.708595 F Ratio 1.5029 Prob>F 0.2373 Estimate -2.511702 0.062089 -0.002097 Parameter Estimates Std Error tRatio 0.203701 -12.33 0.035933 1.73 0.001231 -1.70 Prob>ltl <.0001 0.0934 0.0980 Lower 95% -2.926132 -0.011017 -0.004602 Upper 950/o -2.097272 0.1351945 0.0004084 Random Sample Only FilmEmployees Ln PFHS ppm By YrsFilm APpendix 0 Page 4 .39 -4.DD E 1 7 05 10 is 20 yrsfin 25 30 - Limri - Pom-rdP ivge=2 Source Model Effor C Total Unear Fit InPFHSdfppm = -4.7215+ 0.00958YrsFilm Summary ofFit RSquare 0.011809 RSquare Adj -0.01814 Root Mean SquareEffor 0.882741 Mean ofResponse -4.58683 Observation(sorSum Wgts) 35 AnalysisofVariance DF Sum ofSquares Mean Square 1 0.307286 0.307286 33 25.714619 0.779231 34 26.021905 F Ratio 0.3943 Prob>F 0.5343 Term Intercept YrsFilm Estimate -4.721471 0.0095783 ParameterEstimates Std Error tRatio Prob>ltl 0.26122 -18.07 <.0001 0.015253 0.63 0.5343 Lower95% -5.252924 -0.021454 Upper 95% -4.190018 0.0406102 APPendix 0 Page 5 Term Intercept YrsFilm YrsFilMA2 InPFHSdfppm PolynomiFailtdegree=2 -5.3019+ 0.16523YrsFilm- 0.00548 YrsFilMA2 Summary of Fit RSquare 0.252796 RSquare Adj 0.206096 Root Mean Square Error 0.779496 Mean ofResponse -4.58683 Observations(orSum Wgts) 35 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 2 6.578235 3.28912 32 19.443670 0.60761 34 26.021905 F Ratio 5.4132 Prob>F 0.0094 Estimate -5.301864 0.1652333 -0.005481 Parameter Estimates Std Error t Ratio 0.292996 -18.10 0.050289 3.29 0.001706 -3.21 Prob>ltl <.0001 0.0025 0.0030 Lower95% -5.898674 0.0627984 -0.008957 Upper 95% -4.705053 0.2676682 -0.002006 Random Sample Only FilmEmployees (MaintenanceWorkers Numbered) InPOAA ppm By YrsFilm -39 -2.5- -23 -25 -3.5 Appendix 0 Page 6 5 10 15 20 25 30 irsrim - Limft - P@vWR iw-2 LineaFrit InPOAAppm = -3-5336+ 0.01719YrsFilm Summary of Fit RSquare 0.040923 RSquare Adj 0.01186 Root Mean SquareError 0.838584 Mean of Response -3.29191 Observations(orSum Wgts) 35 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.990187 0.990187 33 23.206369 0.703223 34 24.196556 F Ratio 1.4081 Prob>F 0.2438 Term Intercept YrsFilm Estimate -3.533607 0.017194 Parameter Estimates Std Error t Ratio 0.248153 -14.24 0.01449 1.19 Prob>ltl <.0001 0.2438 Lower 95% -4.038476 -0.012286 Upper 95% -3-028739 0.0466736 PolynomialFitdegree=2 InPOAAppm = -3.9585+ 0.13115YrsFilm- 0.00401YrsFilMA2 Summary ofFit RSquare 0.179823 RSquare Adj 0.128562 Root Mean Square Error 0.787509 Mean of Response -3.29191 Observations(orSum Wgts) 35 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 2 4.351095 2.17555 32 19.845461 0.62017 34 24.196556 F Ratio 3.5080 Prob>F 0.0419 Term Intercept YrsFilm YrsFilMA2 Parameter EsUmates Estimate Std Error -3.958504 0.296008 0.1311467 0.050806 -0.004013 0.001724 t Ratio -13.37 2.58 -2.33 Prob>ltl <.0001 0.0146 0.0264 Appendix o Page 7 Random Sample Only Film Employees (MaintenanceWorkers Numbered) InPFOSAA ppm By YrsFilm -5.00 m2l .07 Appendix 0 Page 8 -6.00 -6.50 .,. 23.1 -7.00- , i 05 EN42839 0.6 -i I 10 is as al I.I.I 2D 25 30 Lim r LinearFit InPFOSAAdfppm = -6.1143+ 0.00041YrsFilm Summary ofFit RSquare RSquare Adj 0.000031 -0.03027 Root Mean Square Error 0.739574 Mean of Response -6.10856 Observations(orSum Wgts) 35 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.000553 0.000553 33 18.050011 0.546970 34 18.050564 F Ratio 0.0010 Prob>F 0.9748 Term Intercept YrsFilm Estimate -6.114269 0.0004063 Parameter Estimates Std Error t Ratio 0.218854 -27.94 0.012779 0.03 Prob>ltl <.0001 0.9748 Lower 95% -6.559529 -0.025593 Upper 950/o -5.66901 0.0264053 Random Sample Only FilmEmployees (MaintenanceWorkers Numbered) InM570 By YrsFilm D.00 -1.00 -2.00 -3.DD 00 7 Appendix 0 Page 9 0 5 10 15 20 25 30 Unear Fit InM570 = -4.8046- 0.00844YrsFilxn Summary ofFit RSquare 0.006167 RSquare Adj Root Mean Square Error -0.02306 1.089533 Mean ofResponse -4.92021 Observations(orSum Wgts) 36 Source Model Error C Total AnalysisofVariance DF Sum ofSquares Mean Square 1 0.250458 0.25046 34 40.360826 1.18708 35 40.611285 F Ratio 0.2110 Prob>F 0.6489 Term Intercept YrsFilm Parameter Estimates Estimate Std Error -4.804612 0.310334 -0.008441 0.018377 t Ratio -15.48 -0.46 Prob>ltl <.0001 0.6489 Random Sample Only FilmEmployees (MaintenanceWorkers Numbered) InM556 ppm By YrsFilm -39 Appendix 0 Page 10 10 15 20 25 30 YI$FH Unear Fit InM556dfppm -6.0381+ 0.00926YrsFilm Summary ofFit RSquare RSquare Adj Root Mean SquareError Mean ofResponse Observations(orSum Wgts) 0.005982 -0.02325 1.213109 -5.91136 36 Source Model Error C Total Term Intercept YrsFilm AnalysisofVariance DF Sum ofSquares Mean Square 1 0.301128 0.30113 34 50.035524 1.47163 35 50.336652 F Ratio 0.2046 Prob>F 0.6539 Parameter Estimates Estimate Std Error -6.038109 0.345532 0.0092556 0.020461 tRatio -17.47 0.45 Prob>ltl <.0001 0.6539