Document 4vJ9x5Mn7k57rGpam3b28QNoV

FINAL p. 1 /!~72f0\-,2 Zb ~ ~7 STUDY TITLE Ammonium Perfluorooctanoate : Phase II . Retrospective Cohort Mortality Analyse s Related to a Serum Biomarker of Exposure in a Polymer Production Plant AUTHOR: Robin C . Leonard, Ph .D. STUDY COMPLETED ON : September 15, 2006 PERFORMING LABORATORY : DuPont Epidemiology Program DuPont Haskell Laboratory for Health and Environmental Science s 1090 Elkton Road, Newark, DE 19714 LABORATORY PROJECT ID : DuPont-14809 (Phase II) WORK REQUEST NUMBER : 15259 SERVICE CODE NUMBER : 1242 SPONSOR: E.I. du Pont de Nemours and Company Wilmington, Delaware 1989 8 U .S .A . 1 of 73 p. 2 FINAL Good Epidemiology Practice Compliance Statement ..... .. .. . . . . . . . . . . . .. . . .. ..... ... . . . . . . . . . . .......... .. . . . . . . . . 5 DuPont Approval ... ..... .. . . .. . . . . .. . . . . . . . . . . . . . . . . ........ . .. . . . . .. . . .. ...... .. .. .. .. .. . . . . . . . . . . . . ... ..... .. . . . . . . . . . .. .. ........ .. .. . . 6 Epidemiology Review Board .. .. . . . . . . . . . . . . . . . .. ........ . . . . . . . . .. ............ .. .. . . . . . . . . . . . . . . . .. ... .. . . . . . . . . . . . .. ............ .. 7 Abbreviations and Symbols . .. .. . . .. . . . . . . . . . . . . . .. ...... .. . . . . . . . . .............. .... . . . . . . . . . . . . . . . ...... . .. . . . . . . . . . ............ .. 8 Executive Summary . . . .. ................. . . . . . . . . . . . . . .. ......... . .. . . . . .. . . .. ............ .... . . . . . . . . . . . . . .. ..... . . . . . . . . . . . .. .. ........ 9 Introduction ..... . . . . . . . . . . . . . .. .... .. .... ..... .. . . .. . . . . . . . . . . . .......... . . . . . . . . .. .. .. .............. . . . . . . . . . . . . . .. ..... . . . . . . . . . . . .. ...... 11 Methods . . . . .. ...... .. . . . . . . . . . . . . . . . . . .. .... ......... .. . . . . . . . . . . . .. ........ .. . . . . . . . . . . . . .. .... ........ .. . . . . . . . . . . . .. ...... . . . . . . . . . . .. . . .. .. 1 2 Cohort Ascertainment . . . .. . . . . .. .... ..... .. . . .. . . . . . . . .. ........ . . . . . . . . . . . . .. .... ........ ... . .. . . . . . . . . . . . .. ..... . . . . . . . . . . . .. .. 13 DuPont Mortality Registry . . .. .. .. ....... .. .. . . . . . . . . . ........ . . .. . . . . . . . . . . . . . . .. .......... .. . . . . . . . . . . . .. ..... . . . . . . . . . . . .. .. 13 Exposure Categorization . . . . . . . . . . . .. ......... . . . . . . . . . . . ........ .. . . . . . . . . . . . . . . . . . . .. ........ .. . . . . . . . . . . . .. ... .. . . . . . . . . . . . . . 1 3 Establish Job Exposure Categories ..... .. . . . . . . . . . .. ........ . . . . .. . . . . . . . . . . .. ........ .. . . .. . . . . . . . . . . . ..... . . . . . . . . . . . . . 14 Exhibit 1 . Group changes based on similar job titles . . .. . . . . . . . . . . .. . . .. .......... . . . . . . . . . . . ..... .... . . . . . . . . . 15 Application of Job Exposure Categories to Historical Job Titles .. ........... . . . . . . . . . . . . .. ..... .. . . . . . . . 16 Calculation of Individual Exposure Metrics . . . . . . .. ............ . . . . . . . . . . . . . . . . .. ........ . . . . . . . . . .. . ...... . . . . . . . . . 16 Exhibit 2 . Example of calculation of average intensity of exposure* . ......... . . . . . . . . . ..... .. . . .. . . . 16 Validate Exposure Classification . . . . . . . .. ..... . . . . . . . . . . . .. ........... . .. .. . . . . . . . . . . . . .. ........ .. . . . . . . . . . .. ..... . . . . . . . 16 Exhibit 3 Etiology of Validation Data Set .. . . . . . . . . . .. .............. .. . . .. . . . . . . . . . . .......... .. . . . . . . . . . .. .. ..... . . . 17 Exhibit 4 . Results of mixed model used to validate cumulative exposure to PFOA : serum PFOA as a function of estimates of PFOA exposure based on job . . . . . . . . . . ..... .. . . . . . . . . . . . . . ....... . 18 Exhibit 5 . Validation of exposure using repeated measurements (mixed model) to predict serum PFOA . . . . . . . . .. .... .................. .. . . . . . . . . . . . .. ..... .. . . . . . . . . . . . . . .. .... ............ .. . . . . . . . . .. .. ...... .. . . . . . . . . . ...... 1 9 Mortality Analyses and Development of Occupational Reference Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Cox Proportional Hazards Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Description of Methods Specific to Ischemic Hea rt Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Categories of exposure for average intensity : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Exhibit 6 . Average intensity of exposure categories for propo rt ional hazards analyses for ischemic heart disease mo rtality, stratified by exposure lag period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Categories of exposure for cumulative exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Exhibit 7A. Cumulative exposure catego ri es for proport ional hazards analyses for ischemic heart disease mo rt ality, stratified by exposure lag period ; quartiles determined by cumulative exposure distribution of cases among white males . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Exhibit 7B . Cumulative exposure categories for propo rtional hazards analyses for ischemic heart disease mo rtality, stratified by exposure lag period ; quartiles determined by cumulative exposure distribution of entire cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 4 Results .. ... .. . . . . . . . . . . . .. . . .. ...... .. .. . . . . . . . . . . . . . . . . . . . . .. ...... ... . . . . . . . . . . . . .. ..... ... . . . .. . . . . . . . . . . . . . . . . . ...... .. . . . . . . . . . . .. .... .. .. . 25 Cohort Description ...... ...... .. .. . . . . . . . . . . . . . . .. ...... .. . . . . . . . . . . . .. .. .. ... .... .. .. . . . . . . . . . . . . . . . . . .. ...... .. . . . . . . . . .. ...... ... 25 Mortality Analyses on Entire Cohort . . . . .. .. ... . . . . . . . . . . . . . . .. .. ... . ... .. .. . . . . . . . . . . . . . . . . . . . ...... . . . . . . . . .. .... ..... 25 All Causes of Death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 All Malignant Neoplasms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Cancer of Biliary Passages and Liver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Cancer of Pancreas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Urinary Tract Cancers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Cancer of Bronchus, Trachea, Lung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Cancer of Prostate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Cerebrovascular Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 All Heart Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Ischemic Heart Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 2 of 73 ~ p. 3 FINA L Diabetes Mellitus ...... ...... .... .. .. . . . . . . . . . . ...... ...... ...... .................... .. .. . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . .. .. ... 28 Cox Proportional Hazards Modeling ...... ...... ...... .................... .. . . . . . . . . . . . . . . . .. . . .. . . . . . . . . . . .. .. . . .. .. ... 28 Ischemic Heart Disease ........... .. . . . . . . . . . . .. . . .. . .. . .. .... .. .... ...... ............. . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . .. . . .. . 2 8 Discussion. .. .............................. ... . .. . . . . . . . . . . .. .... .. .... ...... .................... .. .. . . . . . . . . . . . . . .. . . .. . . . . . . . . . . . . .. . . .... ... 29 Conclusions .............................. .... .. .. . . . . . . . . .. .... .. .... ...... .................... .. .. . . . . . . . . . . . .. .. . . .. .. . . . . . . . . .. .. .. .. .. ... 32 References .................................... .. .. . . . . . . . . .. .... .. .... ...... ........................ . . . . . . . . . . . . . .. .. .. .. . . . . . . . . .. .. ......... 33 TABLES................................... .... .. . . . . . . . . . . .. .... .. .... ...... ...................... .. . . . . . . . . . . . . . .. .. .. . . . . . . . . . . .. .. .... ..... 36 Table 1 Washington Works mortality study cohort .............. ... . . . . . . . . . . . . . ... . . .. .. . . . . . . . . .. .. .... ..... 37 Table 2 SMRs for selected causes of death in Washington Works males, females compared to DuPont Region 1 (West Virginia (less Washington Works), Ohio, Virginia, Kentucky, Indiana, Pennsylvania, Tennessee, and North Carolina), U.S.A. national population, West Virginia state population .. .. .... ...... .................. . . . . . . . . . . . . . . . . . .. .. . . . . . . . . .. .. .. ....... 38 Table 3 White male workers included in the risk-sets of the proportional hazard analysis for IHD stratified by case/non-case status .... .. .... ...... .............. .... .. . . . . . . . . . . . .. .. .. . . . . . . . . . . .. .. ......... 39 Table 4 White male workers included in the risk-sets of the proportional hazard analysis stratified by never-APFO-use/ever-APFO-use status .... ........ .... .. . . . . . . . . . . . . . . . .. . . . . . . . . . . .. .. ......... 40 Table 5 Mortality rate ratios for IHD by exposure category for no-lag analyses using case calendar-year and year of hire (pre-1954) as potential confounders . . . . . . .. ... . . . .. . . . . .. .. ......... 41 Table 6 Mortality rate ratios for IHD by average intensity exposure category, including increasing 5-year lags of exposure, using case calendar-year and year of hire (pre-1954) as potential confounders ............... .. . . . . . . . . . . . . . . . . . . ... . .. .... ...... ............ .. . . . . . . . . . . . . . . . .. . . . . . . . . . . .. .. ......... 42 Table 7 Mortality rate ratios for IHD by cumulative exposure category, including increasing 5-year lags of exposure, using case calendar-year and year of hire (pre-1954) as potential confounders ; A) exposure categories based on case distribution, B) exposure categories based on cohort distribution. . .. . . . . . . . . . . .. .... ...... ........ .... . . . . . . . . . . . . . . . . . .. . . . . . . . . . . .. .. ......... 43 FIGURES ............. .... .. .. .. .. ................. . . . . . . . . . . . . . .. . . . . .. .. .. . . .. .. .... .............. .. . . . . . . . . . . . .. .. .. .. . . . . . . . . .. .. ......... 44 Figure 1 . Time in Job vs Serum PFOA-Cross-Sec tional Study-Job Exposure Category 1 . . . .. . . .. .................. ...... ...... .. ............... . .. . . . . . . . . . . .. . . . . . . .. . . .. .. .. .... ............ .. . . . . . . . . . . . . . .. .. .. .. .. . . .. .. .. .. ......... 45 Figure 2. Time in Job vs Serum PFOA-Cross-Sectional Study-Job Exposure Category 2 . . . .. . . ..--............... ...... ...... ................ .. . . . . . . . . . . . . .. .. . . . . .. .. . . .. .. ................ .. . . . . . . . . . . . .. .. .. .. .. .. .. . . .. .. .. ....... .. 46 Figure 3. Time in Job vs Serum PFOA-Cross-Sectional Study-Job Exposure Category 3 . . . . . . . .. .................. ...... ...... .............. .. . . .. . . . . . . . . . . .. . . . . . . .. . . . . .. .. ................ .. . . . . . . . . . . . . . .. .. .. .. .. . . . . . . .. .. ....... .. 47 Figure 4. Serum PFOA vs Cumulative Exposure - FLAIR Data . . . . .. .... .. .... .. .. . . . . .. .. ....... .. 48 Figure 5. Serum PFOA vs Average Intensity of Exposure - FLAIR Data .. .. . . . . . . .. .. ......... 49 Figure 6. Serum PFOA vs Concurrent Job Intensity Factor - FLAIR Data .. .. . . .... ....... .. 50 Figure 7. Decreasing IHD mortality rates in the U .S.A..... .. . . . . . . . . . . . . . .. .... .... .. . . .. .. . . .. .. ....... .. 5 1 APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Appendix A Washington Works vs Region 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 All-Cause Mortality Surveillance Report : Males . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 All-Cause Mortality Surveillance Report : Females . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 All-Cause Mortality Surveillance Report : Totals (Males and Females) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 7 Appendix B Washington Works vs USA .... ............ .. .... .. .. . . . . . . . . . . . .. .. .. ............ .. .. .. ......... .. . . . . 59 All-Cause Mortality Surveillance Report : Males ...... ...... .. . . . . . . . . . . . . . .. .. .. ..... . .... .. ... . .. ..... .... .. . . 59 All-Cause Mortality Surveillance Report : Females . ...... .. . . . . . . . . . . . . . . . .. .. .. ..... ....... .. .... ......... . . . . 61 All-Cause Mortality Surveillance Report : Totals (Males and Females) .... .. ..... ...... ..... .. .. .. . . 63 Appendix C Washington Works vs West Virginia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 All-Cause Mortality Surveillance Report : Males . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 All-Cause Mortality Surveillance Report : Females . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 3 of 73 p. 4 FINA L All-Cause Mort ality Surveillance Repo rt: Totals (Males and Females) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Appendix D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Job Exposure Catego ry Development based on Division and Job . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4 of 73 15 p. 5 FINAL Good Epidemiolo u Practice Compliance Statement This study was conducted according to guidance provided by the American College of Epidemiology Ethics and Standards Practice Committee which can be accessed at http ://www .acepidemiology2 .org/policystmts/EthicsGuide .asp. Applicant / Sponsor : E .I. du Pont de Nemours and Company Wilmington, Delaware 1989 8 U .S .A . ~ nipel R" C. L edda* Ph~- ~~J [?ft Pka m~* lkacvcb Epidaaibl f* 5 of 73 p. 6 FINAL DuPont Approval : We, the undersigned, declare that this report provides an accurate evaluation of data obtained from this study. A4pp row+F+ bxz ObMw. ~L+a ~ r[ Z~~ La ~~-^ A~ew:i n u, twf AA~" 4w im Ruft wd fs- ~~~,~~4~~ za~ * .~ b'`t Ph~. ~~ Durua U&*ZU Ubmawy rM a am md Apprwnd br ~ w U*,, M a , Clidumi"OffiRw PAMM A ,, bMMUd, ~ . ,~~,~~ . ~.+~n I~AMBO" ft 6 of 73 ~ p. 7 FINAL Epidemiology Review Board 7~e ~ E~pid~i3~r ~~1di~ ~a ~ ~dvdnr~r H+n+~!,, 1M ~~ou~~rw 8+~d ~S~s~, wl~ir ct~riwai~ a~rts in ttwt a~o~ ~la~pt ~bwme aw okr4 Md 00"r* aad mred cra a M e n* cti,rw ct'the B H w cam"w do qaatigr, a baa&4, arA ~rl~~ aft~ n.reu~ ~et7v~ rpo~r~r+ed by ~fara I7aP~a [be~r . T.1ra =a msabrn l r ra nvWw d i n *aft. Pd xqm& ed cn~ d# d do raw* is mkvm% *04 o. k ie dak+ily awJ. Md &M dw aaw*wkn rirer aQpnepW A* ftm 60 d .ta ~e 1m~~lia . KC law Tim L . B6UCh&1* P1LIl. #t . Guawb"M tli~v ~ ~;L~C14 a MA ~ ~~~~'J eoM4~' BNOR, U13 MUN&M Um". Mix. ( ~', i~ O P~, Iobog ft" Bbmbn ~f ~'ublk ~t~I Xe A S. Sa bcio, Pb.I?, C .LJf . ~ hyc+~ epi~rtCzt~st, ~E~r, o~ n } Each ERB member was asked to sign off on this report using their own separate signature page, and then to fax their signature to the DuPont Epidemiology Department . Once received by the DuPont Epidemiology department, each fax was then scanned into the computer system as jpeg files . The jpeg files were then cropped to display only the signature . Each cropped image was then inserted into this page, using Microsoft Word . The original faxes will be retained in the DuPont archives . 7 of 73 8 p. 8 FINA L Abbreviations and Symbol s AIC Akaike Information Criterion, a partial likelihood test statistic APFO Ammonium perfluorooctanoate BMI Body mass index CI Confidence interval CPHM Cox proportional hazards model CSHS Cross-sectional health survey FLAIR Fluoropolymers Laboratory Analysis Information Retrieval GI Gastrointestinal HDL High-density lipoprotein HWE Healthy worker effec t HWSE Healthy worker survivor effect IHD Ischemic heart diseas e LDL Low-density lipoprotein MRR Mortality rate rati o NDI National Death Index (USA) OCMAP Occupational Mortality Analysis Program (University of Pittsburgh) PFOA Perfluorooctanoic aci d PPE Personal protective equipment PPM Parts per millio n PPM-Years Parts per million per year, accumulated SMR Standardized mortality rati o VLDL Very low-density lipoprotein 8 of 73 ~ p. 9 FINA L Executive Summary Ammonium perfluorooctaonate (APFO) is the ammonium salt of the fully fluorinated, 8-carbon carboxylic acid . APFO is used to aid in the emulsion polymerization of fluoropolymers . APFO and its salts are soluble in water and readily dissociate to the perfluorooctanoate anion (PFOA) . APFO is a surfactant that enables the fluoropolymer components to remain emulsified during polymerization and is not incorporated into the polymer itself. It is the ammonium salt (APFO) to which workers may be exposed ; the biomarker measured in serum is the dissociated anion (PFOA) . As a result of the presence and biopersistence of PFOA in the blood of hum ans, the potential health effects of the chemical have been examined, primarily in occupational cohorts. The plant site where this study was conducted comprises several businesses with a diverse range of polymer manufacturing processes, most of which do not involve the use of APFO . Approximately one-half of the employees at the site have been assigned to APFO areas at some time in their careers . As part of a large project on occupational exposure to PFOA, this study's objective is to determine whether workplace exposure to PFOA is related to increased mortality risk for any cause (Phase II) . The overall project comprised two studies : Phase I, a cross-sectional surveillance that analyzed several types of clinical data (blood chemistries such as lipids, enzymes, and blood counts, among others) and a biomarker of exposure (serum PFOA) for potential relationships (to be issued in a separate repo rt); and Phase II, a retrospective cohort study that examined site-wide standardized mo rtality analyses, and also utilized job histo ry information as well as serum PFOA data to classify each member of the historical cohort by level of potential occupational exposure for a more detailed analysis of ischemic hea rt disease . Based on the results of the cross-sectional study, we concluded that workers in all areas across the entire plant site show some measurable level of serum PFOA (range : 0.005 ppm to 9 .55 ppm) . The cohort for Phase II was defined as all individuals who have ever worked at the Washington Works plant at any time between Janua ry 1, 1948 (plant start -up) and December 31, 2002 . The cohort (n = 6,027) was asce rt ained primarily through the DuPont Epidemiology Registries ; additional members were identified from plant-based work history records . Standardized mort ality ratios were calculated for the study cohort for all causes of death, death from all cancers combined, and disease-specific causes of death by comparing the cohort to rates for three reference populations : the general population of the U .S .A ., the West Virginia general population, and the population of DuPont workers residing in West Virginia and seve n neighboring states in the region (DuPont Region 1) . In addition, Cox propo rtional hazards models provided an internal comparison based on exposure categories of serum PFOA for mortality due to ischemic hea rt disease . Mortality rates at this site are generally well within expected values and support the presence of a healthy worker effect . Analyses specific to PFOA categories were conducted for ischemic hea rt disease mortality . The analyses based on average intensity of exposure showed no relationship to PFOA exposure levels . The analyses based on cumulative exposure indicated an increasing trend for the mortality rate ratio with increasing exposure category if those categories were based 9 of 73 !C~ FINA L P . 10 on the distribution of case exposures, but not if those categories were based on the distribution of exposures in the entire cohort. In no case were the mo rtality rate ratio estimates statistically significant . Prostate cancer and cerebrovascular disease, both repo rted as increased in previous 3M Company occupational epidemiology repo rt s [1,2], were reduced in this cohort against all reference populations, cerebrovascular disease significantly so for the U .S .A . and West Virginia populations . The few cases of each of these causes of death did not allow meaningful inte rn al comparisons . Comparisons using DuPont Region 1 reference rates, but not U.S .A. or West Virginia rates, do indicate statistically non-significant elevations in SMRs for kidney cancer mo rtality (SMR=185 ; 95% CI=95-323, p>0 .05), and a statistically significant increase in diabetes mortality (SMR=197 ; 95% CI=123-298, p<0.05) in males and females combined at this plant site . While few kidney cancer cases had been employed in APFO areas, the data from this study are inadequate for examining in appropriate detail rare outcomes such as kidney c ancer. Similarly, the difficulties in examining mortality for diabetes prevent drawing conclusions based on these data . The results reported here show no convincing evidence of increased mo rtality risk associated with APFO exposure for workers at this plant . These results do show statistically non-significant elevations in relative risk for kidney cancer and a statistically significant increase in diabetes mortality for workers at this site . However, given the size and length of follow-up of the study population, the evidence to thoroughly examine mo rtality events like kidney cancer or even diabetes, may not be adequate. Proportional hazards analyses for ischemic hea rt disease mo rtality showed an increase in the model based on equal distribution of cases across cumulative exposure categories in one lagged analysis (the 10-year lag period) . Other exposure lags showed no effect, and results for a second set of models using a different set of exposure cutpoints were attenuated toward the null. None of the hazard estimates themselves were statistically significant. Thus the positive finding in the proportional hazard analysis, as well as the increased diabetes mortality, might be due to chance . Because of the complexity of the exposure assessment and limited power for some analyses, additional investigations are needed . 10of73 C~ P . 11 FINAL Ammonium Perfluorooctanoate : Phase II . Retrospec tive Cohort Mortality Analyses Related to a Serum Biomarker of Exposure in a Polymer Production Plant Introductio n Ammonium perfluorooctanoate (APFO) is the ammonium salt of a fully fluorinated carboxylic acid, perfluorooctanoic acid (PFOA) . APFO is used to aid in the emulsion polymerization of fluoropolymers . The salts of PFOA are soluble in water and readily dissociate to the carboxylate anion (PFOA) . APFO is a process additive, i .e., it is the surfactant that enables the fluoropolymer components to remain emulsified in order for polymerization to occur. Neither APFO nor PFOA is incorporated into the fluoropolymer . As a result of industrial use of APFO and biopersistence of PFOA in the blood of humans, the potential health effects of APFO have been examined in multiple studies, primarily in occupational cohorts [1-7] . A study of community exposure to PFOA was conducted to determine the relationships between serum concentrations and exposure sources, and also to examine the relationships between PFOA and hematologic and biochemical clinical markers . This community study indicated that water, not air, was the likely source of exposure ; no associations were seen with adverse health effects [8-9] . Some general population samples have been used to examine biomonitoring data that indicated that age, gender, and possible duration of exposure had little to do with the background levels in the population . [10-12] . A retrospective cohort mortality study was conducted at a 3M plant in Cottage Grove, Minnesota, that produced APFO [ 1] . The cohort consisted of 3,537 workers employed for at least six months between January 1947 and December 1983 . Follow-up was nearly complete (99 .5%) for the study participants, and 398 deaths were recorded . Since APFO production was limited to the Chemical Division, the two exposure categories were "exposed" (worked at least one month in the Chemical Division) and "not exposed" (worked one month or less in the Chemical Division) . Standardized mortality ratios (SMRs) were calculated comparing the Cottage Grove cohort with mortality rates for the populations of the U .S .A . and the state of Minnesota, using stratification for duration of employment and 3 exposure latency periods . When exposure status was taken into account, most SMRs were significantly lower than the expected rate, a not surprising finding, considering the potential for healthy worker bias . The SMR for prostate cancer was elevated in the Chemical Division (area of APFO production), but this increase was not statistically significant as it was based on only four cases . Internal comparisons were performed using proportional hazards modeling for various categories of causes of death, including prostate cancer. The only mortality rate ratio that was statistically significantly greater than 1 .0 was for the association between prostate cancer and duration of work in the Chemical Division . However, given that there were only four cases in the Chemical Division and an additional two cases in the rest of the cohort, it is difficult to emphasize this finding . This study was updated through 1997; the updated cohort consisted of 3,992 workers who had worked for at least one year at the Cottage Grove Plant . [2] The exposure categories were changed to comprise three groups : "definite" (based on tasks performed in the Chemical Division); "probable" (tasks involving transient, lower exposures) ; and "non-exposed" (primaril y non-Chemical Division jobs) . A few SMRs were elevated : cancer of the prostate (N=1 ; II of73 17 FINA L p . 12 SMR=1 .30, 95% CI: 0.03-7 .20) ; pancreatic cancer (N=1 ; SMR=1 .34, 95% Cl : 0.03-7 .42) ; and cerebrovascular disease (N=5 ; SMR=2 .58, 95% Cl : 0 .84-6 .03) ; but none were statistically significant, and all elevated SMRs were based on very few cases . The DuPont Epidemiology Program conducted a cross-sectional health survey (Phase I) of 1,025 employees at the Washington Works, West Virginia polymer production facility . That investigation used epidemiologic and statistical analyses of several types of clinical data and a biomarker of exposure (serum PFOA) to determine the presence of any association between occupational exposure to APFO and measurable changes in clinical laboratory measurements or physical examination endpoints . The results of this study indicated a positive association between serum level of PFOA in workers at the polymer production plant and serum cholesterol, triglycerides, and LDL cholesterol . No association was seen with HDL cholesterol [3] . These results were similar to those published by Olsen et al . [4] from a cross-sectional study on 3M workers at two plants . After adjustment for potential confounders including body mass index (BMI), current alcohol use, smoking, and age, a statistically significant association between increased serum levels of PFOA and increased levels of both cholesterol and triglycerides were observed in multivariable linear regression analyses . Longitudinal analyses for the 3M workers also showed that PFOA was positively associated with serum cholesterol (log-linear regression coefficient = 1 .03 ; 95% Cl : 1 .01-1 .05) and serum triglycerides (log-linear regression coefficient = 1 .10 ; 95% Cl : 1 .05-1 .16) . However, as in our own study, the percent of variation explained by the model, as well as by serum PFOA, was small, and there was no association of serum PFOA with HDL cholesterol . The 3M investigators had concluded that since their results were opposite to those expected based on animal studies, their findings were probably spurious . DuPont conducted a cross-sectional medical surveillance for altered liver function on the workers at the Washington Work plant site in 1979 . These results indicated no changes in levels of liver enzymes associated with work area assignment [5] . Gilliland and Mandel examined clinical chemistries in 3M workers and reported no abnormalities or associations with total organic fluorine levels [6] . They did suggest that serum total organic fluorine levels might modulate hepatic responses to obesity and alcohol, but this suggestion was not supported by results of subsequent surveillance examinations among those workers [7] . The current study examines all-causes of death combined and cause-specific mortality rates for the DuPont employees at the Washington Works, West Virginia, polymer manufacturing facility . This facility produces several types of polymers, most of which are made by processes not involving APFO . Approximately one-third of the employees at the plant works in APFO-using areas . SMRs were calculated by using three different reference populations : the general U .S .A . population, the state population of West Virginia, and an eight-state regional DuPont employee population (DuPont Region 1) . Because increased lipids levels are a risk factor for cardiovascular diseases, we also used Cox proportional hazards models to estimate mortality rate ratios (MRR) for an internal comparison of mortality due to ischemic heart disease associated with categories of exposure to APFO . The exposure assessment for PFOA was based on a combination of work history information for each subject and serum PFOA levels obtained from the Phase I cross-sectional survey of the active workers in 2004 . Methods 12 of 73 13 FINAL p . 13 Cohort Ascertainment The cohort was defined as all individuals who have ever worked at the Washington Works plant at any time between January 1, 1948 (plant start-up) and December 31, 2002 . The cohort was ascertained primarily through the DuPont Epidemiology Registries ; additional members were identified from plant-based work history records . DuPont Mortality Registry The DuPont Company has maintained a Mortality Registry for all active and pensioned U .S .A. employees since 1957 . This Registry provides the expected numbers of deaths used in the DuPont Epidemiology Surveillance Program SMR calculations to compare each plant site in the U.S.A. to the rest of the U .S .A. DuPont population . Deaths are reported to the Registry by the corporate Benefits division through death certificates that accompany life insurance claims filed by beneficiaries of deceased employees and pensioners . Until recently, employment duration of at least 15 years was required for pensioning . However, additional changes in vesting strategies and insurance policies created fiduciary responsibility on the part of the Company that requires notification of death of additional former employees . Deaths are ascribed to the observed numbers for the plant site at which the employee worked at the time of death, or the site at which the pensioner worked at the time of retirement . For those who left the Company between 1950 and 1979, 91 .7% were either pensioned or covered by some other vested benefit . Of the employees leaving the company between 1980 and 2005, only 60% were pensioned or covered by other vested benefits . However, deceased non-pensioned employees terminating after 1979 have been added to the Registry through the use of the National Death Index database, NDI Plus . In addition, the social security numbers for each cohort member were submitted to the Social Security Administration for confirmation of vital status . The Employee Registry, which provides the demographic information on all individuals ever employed by DuPont in the United States, is updated from a monthly upload from Corporate Human Resources . The Epidemiology Employee Registry currently includes approximately 265,000 individuals, 6,027 who ever worked at Washington Works, approximately 2000 of whom are presently located at that site . Exposure Categorization While the half-life of PFOA in humans is estimated to be about 4 years [13], the kinetics of PFOA in humans are not well characterized . APFO can be absorbed via inhalation, as well as orally and dermally . Dermal absorption is very slow, and is an issue only for occupational exposure [14,15] . Serum PFOA levels were considered the best measure of exposure, because serum levels integrate all routes of exposure and provide an estimate of the amount of the compound delivered internally to the organ tissues . Information regarding the relationship between an individual's job and the measured serum PFOA level was based on the Phase I crosssectional health survey conducted in 2004 and incorporated into the exposure assessment for the retrospective cohort mortality study . There were four major steps in determining exposure categories in Phase I . 13 of 73 FINA L p . 14 1) Establish relative exposure categories for current job titles using serum PFOA . a. Link individually measured serum PFOA levels measured with the job title held by the individual at the time of sampling . b. Examine the median, range, and distribution of serum levels for each job to determine the typical exposure for that job title . c. Based on the "typical" exposures, assign each job title to one of three relative exposure categories (Job Exposure Category : low, medium, high) . 2) Apply Job Exposure Categories to historical job titles . a. Link unique job titles from work history files with job titles and assign the historic job titles to the corresponding Job Exposure Category . b. Apply appropriate Job Exposure Category to each record in the cohort work history . 3) Calculate individual exposure metric s a. Multiply the time each individual spent in each Job Exposure Category by the intensity factor associated with that category and sum across all categories to calculate individual Cumulative Exposure . b. Calculate Average Intensity by dividing cumulative exposure by duration of hire . 4) Validate exposure classification by plotting exposure variables (average exposure intensity, cumulative exposure, and concurrent job intensity factor) versus the measured serum PFOA levels collected as part of a plant-site voluntary biomonitoring program Establish Job Exposure Categorie s Exposure and employment data were collected for 1,025 Washington Works employees as part of the Phase I cross-sectional health survey conducted in 2004 . These data were combined to establish exposure categories for job titles . In the cross-sectional study, work divisions at the plant site were designated as "APFO-use" or "no APFO use" based on the potential for exposure to APFO, with the understanding that some individuals within APFO-use divisions may not have had exposure to APFO and some individuals in no APFO-use divisions may have had undocumented exposure to APFO . Median, minimum, and maximum serum PFOA levels were calculated for each no APFO-use division, using only individuals who had never worked in an APFO-use division, to establish the criteria for the low-intensity job exposure category . Next, the median, minimum, and maximum serum PFOA levels were calculated for each job title in the APFO-use divisions . The divisions designated as "APFO-use" Divisions included : TEFLON Maintenance, TEFLON Polymers Production, TEFLON Copolymers Production, Research, and Technical . The distribution of serum PFOA levels within each job title was examined . See Appendix D . APFO-use jobs were then grouped into three job-exposure categories based on job-specific PFOA serum level information : 14 of 73 ~~ FINA L p . 15 Job Exposure Catego ry I consisted of all no APFO-use division jobs, and APFO-use jobs within the same serum level range as those employees who had never had a job assignment in any APFO-use division--median<0 .25 ppm . Job Exposure Category 2 consisted of APFO-use jobs with a median serum level >0 .25 ppm and <0 .75 ppm . Job Exposure Catego ry 3 comprised all APFO-use jobs with a median serum level >0 .75 ppm . Some latitude was allowed in the use of median serum levels as the only criterion in the assignment of a given job title to a job exposure catego ry. In instances where a job title was categorized differently than similar job titles as a result of a median serum value from a very small sample, the job title was grouped with the similar job titles . (See Exhibit 1 .) Exhibit 1 . Grou changes based on similar j ob titles Job Exposure Serum PFOA Cate o Division Job Median Min Max n From To Reaso n Moved to be in same category APFO-USE@ as other "Specialists" (n=11) COPOLYMERS AREA with similar exposure (Range : PROD . SPECIALIST 0 .255 0 .255 0 .255 1 2 1 0 .025 -0 .272 ) Moved to be with other "Sr APFO-USE@ Engineers" in Job Exposure POLYMERS SR Category 2 (n= 8 ; Range : 0 .097 PROD . ENGINEER 0 .765 0 .412 1 .59 4 3 2 -0 .576) TECHNICAL TECH SPEC 0 .783 0 .783 0 .783 1 3 2 Moved to be with other APFO-USE@ "Specialists" in Job Exposure COPOLYMERS Category 2 (n= 7 ; Range : 0 .134 PROD . TECH SPEC 1 .46 1 .46 1 .46 1 3 2 - 1 .28 ) The resulting numbers of cross-sectional study pa rticipants with jobs in Job Exposure Categories 1, 2, and 3 were 784, 107, and 134, respectively . The mean serum levels within the respective groups were 0 .21, 0.43, and 1 .69 ppm. Those mean serum levels served as the intensity factors for the three Job Exposure Categories . Because PFOA is believed to have a half-life of about 4 years in humans [13], the length of time spent in the assignments used to define the Job Exposure Categories was examined to ensure that the job titles would not be misclassified as a result of individuals with either very short or very long stays in the job . Each participant's serum PFOA level was plotted against the time in the concurrent job by Job Exposure Category . The resulting correlation indicated that time in job was not strongly associated with serum PFOA level and should not substantially contribute to misclassification of job titles . 15 of 73 16 FINA L p . 16 Application of Job Exposure Categories to Historical Job Title s Complete job histories for all Washington Works employees were obtained via electronic files from the Human Resources department at the plant site . In addition, historical work divisions were designated as "APFO-use" or "no APFO-use" based on the potential for occupational exposure to APFO . All historical job titles in the "no APFO-use" work divisions were assigned to Job Exposure Category 1 . Approximately 1600 unique "APFO-use" job titles were identified in the work history files . Most of those unique job titles resulted from variations in spelling or abbreviations, or division name changes for common job titles . Historic job titles were matched with similar titles identified in the cross-sectional survey and assigned to the corresponding Job Exposure Category . Calculation of Individual Exposure Metric s Cumulative exposure was calculated for each individual in the full cohort by multiplying time in the various job exposure categories by the intensity factor associated with job exposure categories 1 through 3, either 0 .21, 0.43, or 1 .69 ppm, respectively . Average intensity was also calculated for each cohort member by dividing the individual's cumulative exposure by their duration of hire, as shown in Exhibit 2 below . Exhibit 2 . Example of calculation of average intensity of exposure* . Time in Intensity Exposure Catego ry Factor JobExpCatl 2 .50 0.21 0.52 JobExpCat2 12 .00 0 .43 5 .16 JobExpCat3 7 .25 1 .69 12 .25 17 .94 / 21 .75 *Values are at end of follow-up Validate Exposure Classification To validate the assignment of Job Exposure Categories to historical job titles, relationships were examined between calculated exposure values and measured PFOA serum levels in the Phase III Longitudinal Study dataset [16] (see Figures 4, 5, and 6) . The Longitudinal Study dataset was comprised of sampling data for individuals with more than one sample from the Fluoropolymers Laboratory Analysis Information Retrieval (FLAIR) biomonitoring database together with the Phase I study . The FLAIR biomonitoring database archived serum PFOA data collected on a voluntary basis to ensure the effectiveness of workplace controls . Samples had been collected between 1979 and 2002 . 16 of 73 17 FINAL p . 17 Since the assignment of Job Exposure Categories to historical job titles was based on blood PFOA measurements that were taken in Phase I, those values were removed for the validation analysis . There were also 40 observations from the FLAIR database (corresponding to 23 employees) where the PFOA blood sample was taken after the employee had stopped working at the plant . These 40 observations were also removed from the validation analysis . Our validation was performed on dataset 6 as described in Exhibit 3 . Exhibit 3 Etiology of Validation Data Se t Data Date of Number of Number of set Collection Participants Observations I FLAIR database 1979-2002 891 194 7 2 Cross-Sectional Study 2004 1025 1025 3 All FLAIR data and Cross- 1979-2004 891 2148 sectional data of FLAIR participants 4 FLAIR and Cross-sectional 1979-2004 461 1718 participants with more than 1 measurement each 5 FLAIR and Cross-sectional 1979-2002 461 1517 participants with more than 1 measurement each minus the Cross-sectional sample s 6 FLAIR and Cross-sectional 1979-2002 451 1477 participants with more than 1 measurement each minus the Cross-sectional samples and retiree sample s Cumulative exposure, average intensity, and concurrent job intensity factor were calculated for each individual up to the time of the sampling (from hire date to sample date) . The relationships between each serum PFOA value and the corresponding average intensity, cumulative exposure, and concurrent job intensity factor were analyzed . There were 21 missing values for jo b intensity factor and 6 missing observations for average intensity . First, the observations were treated as being independent (although many employees had more than one observation) and examined in a general linear model . Then the associations between serum PFOA and the exposures of interest (average intensity, cumulative exposure, and concurrent job intensity factor) were examined in mixed models . 17 of 73 ~~ FINA L p . 18 Exhibit 4 . Results of mixed model used to validate cumulative exposure to PFOA : serum PFOA as a function of estimates of PFOA exposure based on 'ob . Correlation P value Coefficient Average Intensity 0 .40 <0 .0001 Cumulative Exposure 0.36 <0 .0001 Intensity Factor 0 .39 <0 .000 1 Prior to fitting the mixed models, sample variograms were created for the outcome variable (blood PFOA) to evaluate the serial correlation, the measurement error and the random effect for each mixed model . The spatial power covariance structure was found to have the best fit for the data and was therefore used . This structure allows the correlations between errors to be modeled in such a way that two points that are close in time are more correlated than two points that are further apart in time . This covariance structure is appropriate for unequally space d measurements, which was the case in our dataset since employees were having their blood PFOA checked voluntarily and at different time intervals . Each model included the exposure of interest (average intensity, cumulative exposure, or we calculated the numbers of years since January 1, 1979 that would correspond to each calendar date (example : (blood sample date-January 1, 1979)/365 .25). An interaction term between the exposure variable and the date variable was also added to the model . All exposure variables were positively and significantly associated with the outcome (serum PFOA), suggesting that historical job titles were properly categorized (see Exhibit 5) . Time was negatively associated with serum PFOA, supporting the observation that PFOA is reduced over time . The model that gave the best fit was the mixed model that included the exposure variable and the time variable, without the interaction term . The concurrent job intensity factor model had the best fit for the data as suggested by the lowest fit statistic test . Therefore intensity factor explains serum PFOA better than cumulative exposure and average intensity . This suggests that concurrent exposure has greater influence on serum PFOA levels than past exposure and, therefore, the use of concurrent measured serum levels to characterize job titles into relative exposure categories is a valid approach for this compound despite the concerns around half-life . 18 of 73 ~9 FINAL P . 19 Exhibit 5 . Validation of exposure using repeated measurements (mixed model) to predict serum PFOA A . Predictor variable : average intensity Coefficient P Average intensity 0 .899 <0.0001 Time * -0.036 0.000 6 Fit statistic : AIC = 5445 .7* * B . Predictor variable : cumulative exposure Coefficient p Cumulative exposure 0 .0002 <0 .0001 Time * -0 .034 0.0014 Fit statistic : AIC = 5477.6 * * C . Predictor variable : intensity factor Coefficient P Intensity factor 0 .796 <0 .0001 Time * -0 .036 0.0006 Fit statistic : AIC = 5407 .8 ** * Time is calculated as number of years since 197 9 ** For AIC (Akaike Information Criterion), the smaller value indicates better fit . Mortality Analyses and Development of Occupational Reference File s All SMRs were calculated using OCMAP (Occupational Mortality Analysis Program) developed by the University of Pittsburgh [ 17] . This software compares observed numbers of deaths in the study population to expected numbers of deaths based on rates for chosen reference populations for specific gender, 5-year time, and 5-year age categories by cause of death . State and U .S .A. reference rate files were acquired directly from the University of Pittsburgh . Estimating relative risks by SMRs is a standard epidemiological approach to adjust for confounding by age and other characteristics that differ between populations . Typically, the general U .S .A. population is used as the reference group ; however, it is not an appropriate comparison group for a worker cohort . Because healthier people are selectively hired to work, these populations may not be comparable in terms of health status . This may introduce a downward bias in estimates of the SMR due to confounding by the healthy worker effect [ 1820] . While the effect is generally stronger for chronic diseases, the downward bias in comparative estimates has also been demonstrated for cancer [21 ] . One approach for reducing healthy worker bias is to choose a reference population composed of workers unexposed to the particular hazardous agent of interest . Restricting the comparison group to the same geographical region as the exposed cohort also improves comparability b y 19 of 73 O C>2 FINAL p . 20 reducing likelihood of unmeasured confounding by commonly shared regional characteristics such as diet and lifestyle . Preliminary results from our work on another large occupational coho rt indicated that the most appropriate compa rison for occupational cohorts is the working population of the same company drawn from the same region as the study plant site (manuscript in preparation) . Potentially biased estimates of reduced SMRs usually seen when comparisons are made to general non-occupational populations are not observed using this comparison, presumably because the healthy worker effect is reduced . Additionally, comparing mortality rates for workers from the same general region adjusts for local socio-cultural factors, although not all local effects are likely to be removed. A second component of healthy worker bias arises from the healthy worker survivor effect (HWSE) . This bias is introduced when less healthy workers leave the workforce earlier than healthy workers, thereby having no opportunity to accrue cumulative exposures as large as more healthy workers . One approach that has been proposed to reduce this healthy worker survivor effect, is to assign zero weight to exposures in the 5 to 10 years proximate to the date of death (or diagnosis) in order to discount the effect of exposures during periods of time closer to the event of interest [22-23] . For the DuPont employee comparisons, we created a DuPont regional reference file (DuPont Region 1) that included all DuPont employees in West Virginia and seven neighboring states : Ohio, Virginia, Kentucky, Indiana, Pennsylvania, Tennessee, and No rth Carolina (excluding those employees at the Washington Works site) . For the DuPont worker mo rtality rates, race was not an adjustment variable . No follow-up methods or effo rts additional to those used in Registry ascertainment were applied to the coho rt files. The first level of cohort analysis was the calculation of SMRs for the entire Washington Works coho rt, with follow-up from 1948 to 2002, the last year for which the Registry has been updated through NDI Plus . SMRs were calculated based on compa risons to the U .S.A. general population, the state of West Virginia, and the DuPont Region 1 reference file . Cox Proportional Hazards Modelin g SMRs are useful for comparing mo rtality between an exposed to an unexposed reference group . However, in order to take full advantage of the exposure assessment for PFOA and examine exposure-response relationships we tu rned to Cox models for one outcome-ischemic heart disease . Cox propo rtional hazards models (CPHM), with age as the time metameter, were used to estimate adjusted mort ality rate ratios ( MRRs) for ordinal PFOA exposure categories . These categories were based on the cumulative exposure calculated for each member of the historical coho rt based on the categorization of jobs . The cumulative exposure thus calculated was then used to derive the average exposure intensity for each coho rt member, based on the job histo ry data . Ischemic hea rt disease mort ality was chosen based on the fact that increased lipids are a risk factor for ischemic hea rt disease, and there were sufficient cases to enable division into exposure groups . For ischemic heart disease, we estimated mo rtality rate ratios using lagged exposure ( 5, 10, 15, and 20 years) to reduce any bias introduced by the healthy worker survivor effect . 20 of 73 ~~ p . 21 FINAL CPHM is a statistical model used to investigate the relationship between survival time (or time to event) and one or more independent variables [24] . An assumption that must be met for these analyses to be meaningful is that the hazard rates are proportional to one another at all ages . This means that at any given age (t), the hazard rate for those exposed to a risk factor [hl(t)] is a constant multiple of the underlying hazard [ho(t)] for that age . A significant advantage to the approach is that the baseline hazard function does not have to be explicitly described, since the different risks are relative . The model can be stated as follows : hl(t) = h0(t) x B Estimating the constant multiplication factor for changes in risk (B) is conveniently done using an exponential function, B = eb . Reformulating equation 1 yield s hI (t) = ho(t) x eb . Therefore, if ho(t) represents the hazard in the unexposed group at any given time, the hazard ratio (HR) comparing the exposed and unexposed i s HR = [hl(t)] ho(t)] = eb, or taking logarithms, log(HR) = b . All CPHM was conducted using SAS Proc PHREG, version 9 .01 . Description of Methods Specific to Ischemic Heart Diseas e There were only three cases of IHD among women and only one non-white male case . Therefore, women and non-whites were excluded from all proportional hazards analysis . Person-time for the risk set of each index case was comprised of people who had started working by the age of the case that defined the risk set (case age at death), and were still alive at that age . In addition to exposure to PFOA, the regression models also included calendar year of death in order to control for secular trends over the follow-up period . There were 235 cases of ischemic heart disease available for analysis . Since cases were hired on average 20 years earlier than the non-cases, year of hire was also included to adjust for confounding . Because half of the cases were hired before 1954, we created a binary variable for year of hire (before or after 1954) that was included in the model . The correlation between calendar year of death and the binary variable, hired pre- or post-1954, was - 0 .44 (p<0 .0001) . We chose to use average intensity and cumulative exposure as the exposure metrics for CPHM analysis. Because heart disease mortality is known to be strongly affected by the healthy worker survivor effect, we chose to lag both exposure metrics 5, 10, 15, and 20 years . These lags eliminated the more recent exposures to reduce bias engendered by healthier workers staying in the workplace longer periods of time thus accumulating more exposure . Categories of exposure for average intensity : In each Cox model, workers whose jobs were categorized as having the lowest exposure to PFOA (lowest average intensity) and who also never worked in any APFO-using division were considered the reference group, thus enabling an internal analysis . Both the APFO areas of the plant site and the non-APFO areas comprise a wide diversity of jobs (mechanics, engineers, supervisors, administrative, etc .) This diversity of job types in all non-APFO areas of the plant 21 of 73 22 FINA L p . 22 should make this group of workers comparable to those workers who ever worked with APFO in all characteristics except for exposure to APFO . Five categories of average intensity of exposure to APFO were specified to ensure that an adequate number of cases would be in each category, thus increasing the stability of the MRR estimates. The exposure categories are ordinal, with 0 comprising the reference group, 1 being the lowest exposure, and 4 being the highest exposure category, for those analyses that comprised four categories . These exposure category definitions are presented in Exhibit 6 . Exhibit 6 . Average intensity of exposure categories for proportional hazards analyses for ischemic heart disease mortality, stratified by exposure lag period . No lag of exposure N = 4,460 o Reference : x = 0.21 ppm and Never APFO-use 167 cases o Category 1 : x = 0 .21 ppm and Ever APFO-use 28 cases o Category 2 : 0.21<x<= 0.250 ppm 12 cases o Category 3 : 0.250<x<=0 .371 ppm 14 cases o Category 4 : > 0 .371 ppm 14 cases Exposure lagging 5 year s N = 4,44 0 o Reference : x = 0 .21 ppm and Never APFO-use 162 cases o Category 1 : x= 0 .21 ppm and Ever APFO-use 30 cases o Category 2 : 0 .211 <= x <= 0 .254 ppm 12 cases o Category 3 : 0.261 <= x <= 0 .551 ppm 13 cases o Category 4 : 0.592 <= x <= 1 .524 ppm 12 cases Exposure lagging 10 year s N = 3,989 o Reference : x = 0 .21 ppm and Never APFO-use 152 cases o Category 1 : x= 0 .21 ppm and Ever APFO-use 30 cases o Category 2 : 0 .211 <= x <= 0 .256 ppm 11 cases o Category 3 : 0 .261 <= x <= 0 .555 ppm 12 cases o Category 4 : 0 .565 <= x<= 1 .524 ppm 12 cases Exposure lagging 15 years N = 3,98 6 o Reference : x = 0 .21 ppm and Never APFO-use 142 cases o Category 1 : x= 0 .21 ppm and Ever APFO-use 30 cases o Category 2 : 0 .21 < x <= 0 .269 ppm 11 cases o Category 3 : 0 .269< x <= 0 .591 ppm 12 cases o Category 4 : x > 0 .591 ppm 12 cases 22 of 73 e>?3 FINA L p . 23 Exposure lagging 20 years N = 3,44 0 o Reference : x = 0 .21 ppm and Never APFO-use 130 cases o Category 1 : x = 0 .21 ppm and Ever APFO-use 28 cases o Category 2 : 0 .21< x <= 0 .330 ppm 16 cases o Category3 : x > 0 .330 ppm 15 cases Categories of exposure for cumulative exposur e Cumulative exposure was estimated as the total attained exposure for each case event at the time of death . Analytic strata were created for each case event by matching all eligible non-case subjects (workers who had not died by the date of the case event) and assigning cumulative exposure based on total attained exposure for each worker at the same age as the case event . Given that there is debate on the implications of the statistical aspects of the categorization of cumulative exposure, two different approaches were used to determine four cumulative exposure categories . The two approaches thus provided a form of sensitivity analysis . In the first analysis (exhibit 7A), quartiles of the cumulative exposure were determined by the distribution of exposures for case subjects . In a second analysis designed to test the model sensitivity to cumulative exposure categorization, quartiles were determined by the distribution of exposures for all workers in the cohort (exhibit 7B) . For each lagged analysis, the exposure values for the four categories are shown for both strategies, and the number of cases of IHD mortality are listed . Cox models analyzing the proportional hazards for cumulative exposure categories with the lowest exposure group serving as the referent also included variables adjusting for calendar year of the case event and pre-1954 hire period . Exhibit 7A. Cumulative exposure categories for proportional hazards analyses for ischemic heart disease mortality, stratified by exposure lag period ; quartiles determined by cumulative exposure distribution of cases among white males . No lag of exposure N = 4,460 o Reference : x <= 3 .81 ppm years 58 cases o Category 1 : 3 .81<x< = 5 .45 ppm years 59 cases o Category 2 : 5 .45<x<= 6 .78 ppm years 59 cases o Category 3 : x>6 .78 ppm years 59 cases Exposure lagging 5 years N = 4,44 0 o Reference : x<= 3 .42 ppm years 57 cases o Category 1 : 3 .42<x< = 5 .28 ppm years 57 cases o Category 2 : 5 .28<x<= 6 .51 ppm years 57 cases o Category 3 : x>6 .51 ppm years 58 cases 23 of 73 FINA L p . 24 Exposure lagging 10 years N = 3,98 9 o Reference : x <= 3 .12 ppm years 54 cases o Category 1 : 3 .12<x< = 4 .90 ppm years 54 cases o Category 2 : 4.90<x<= 6.40 ppm years 54 cases o Category 3 : x>6 .40 ppm years 55 cases Exposure lagging 15 year s N = 3,986 o Reference : x <= 2.43 ppm years 51 cases o Category 1 : 2.43<x< = 4 .19 ppm years 52 cases o Category 2 : 4 .19<x<= 5 .66 ppm years 52 cases o Category 3 : x>5.66 ppm years 52 cases Exposure lagging 20 year s N = 3,44 0 o Reference : x<= 1 .66 ppm years 47 cases o Category 1 : 1 .66<x< = 3 .48 ppm years 47 cases o Category 2 : 3 .48<x<= 5.07 ppm years 47 cases o Category 3 : x>5 .07 ppm years 48 case s Exhibit 7B . Cumulative exposure categories for proportional hazards analyses for ischemic heart disease mortality, stratified by exposure lag period ; quartiles determined by cumulative exposure distribution of entire cohort. No lag of exposure N = 4,46 0 o Reference : x <= 0 .99 ppm years 10 cases o Category 1 : 0 .99<x< = 4 .29 ppm years 61 cases o Category 2 : 4 .29<x<= 6 .98 ppm years 114 cases o Category 3 : x>6 .98 ppm years 50 cases Exposure lagging 5 years N = 4,44 0 o Reference : x <= 1 .97 ppm years 26 cases o Category 1 : 1 .97<x< = 4 .42 ppm years 61 cases o Category 2 : 4 .42<x<= 6 .24 ppm years 76 cases o Category 3 : x>6 .24 ppm years 66 cases Exposure lagging 10 year s N = 3,98 9 o Reference : x <= 2 .28 ppm years 37 cases o Category 1 : 2 .28<x< = 3 .85 ppm years 46 cases o Category 2 : 3 .85<x<= 5 .23 ppm years 37 cases o Category 3 : x>5 .23 ppm years 97 cases Exposure lagging 15 years 24 of 73 .25 FINAL p . 25 N = 3,986 o Reference : x<= 1 .86 ppm years 38 cases o Category 1 : 1 .86<x< = 3 .16 ppm years 34 cases o Category 2 : 3.16<x<= 4.27 ppm years 33 cases o Category 3 : x>4 .27 ppm years 102 cases Exposure lagging 20 years N = 3,44 0 o Reference : x <= 1 .19 ppm years 27 cases o Category 1 : 1 .19<x< = 2 .38 ppm years 34 cases o Category 2 : 2.38<x<= 3 .24 ppm years 28 cases o Category 3 : x>3 .24 ppm years 100 cases Results Cohort Descriptio n The Washington Works cohort consists of individuals who worked at the plant at any time between 1948 and 2002 . First, 5,476 individuals were originally identified from the Epidemiology Employee Registry . Of these persons, 22 individuals were excluded for the following reasons ; 1 had no verifiable birth date, and 21 had transferred to the Washington Works location after December 31, 2002, the end of the mortality surveillance period . This resulted in 5,454 individuals who were included from the Epidemiology Employee Registry with an additional 573 individuals included based on work history records obtained from the plant site Human Resources Department . The resulting cohort for all analyses included 6,027 individuals . Table 1 shows the descriptive statistics for the historical Washington Works cohort . Total person-years were 127,513 .2 for males, and 18,224 .5 for females . Person-years ascribed to the three cumulative exposure categories were 74,603 .6 for Group 1(lowest potential exposure to APFO) ; 52,461 .8 for Group 2; and 18,672 .3 for Group 3 (highest potential exposure to APFO) . Mortality Analyses on Entire Cohor t While the DuPont Regional population appears to be the most appropriate reference group for mortality rate comparisons, we also report SMRs based on both the U .S .A . and West Virginia rates . The U .S .A . comparisons provide some context for other studies in the published literature, and the comparisons to the West Virginia state population are presented in response to a request from study participants . As would be expected, almost all SMRs comparing Washington Works mortality rates for defined causes to the U .S .A . and West Virginia population mortality rates were below 100, the standard metric of the SMR indicating no observed differences in the mortality rates between the compared populations . Further, many SMR estimates were statistically significantly below this estimate of no effect indicating that Washington Works employees had lower mortality rates for many causes of death compared to the general population . Due to concerns about statistical precision, only those causes of death for which there were at least five deaths observed were considered relevant for consideration of increased or decreased risk . 25 of 73 12 C, FINA L p . 26 Table 2 shows selected SMRs for Washington Works males and females when compared to the three reference populations . These causes of death were selected based on results from animal studies and other occupational studies, and are detailed below . Complete SMR analysis results are presented in Appendices A-C . There were only 33 deaths among females workers ; limiting the statistical power to detect significant differences in disease-specific mo rtality rates among female workers and restricting interpretations of SMRs to all causes of death combined and all cancers combined. All Causes of Deat h For males, the SMR for all causes of death was 94 (95% Cl: 87-100) based on the Region 1 DuPont population . The SMRs for all causes based on comparisons to West Virginia and total U.S .A. were 58 (95% Cl : 54-62) and 66 (95% Cl : 62-71), respectively. For females, the SMR for all causes of death was 147 (95% Cl : 101-207) based on the Region 1 DuPont population . The SMRs for all causes based on compari sons to West Virginia and total U .S .A . were 73 (95% Cl : 51-103) and 81 (95% Cl : 56-113), respectively . All Malignant Neoplasm s For males, the SMR for all malignant neoplasms was 100 (95% Cl : 88-114) based on the Region 1 DuPont population. The SMRs for all malignant neoplasms based on comparisons to West Virginia and total U .S .A. were 68 (95% Cl : 60-78) (WV) and 74 (95% Cl : 64-84) (U .S .A . ) For females, the SMRs for all malignant neoplasms were 149 (95% Cl : 77-260) ; 79 (95% Cl : 41-139), and 87 (95% Cl : 45-15 1) when comparing against the Region 1 DuPont Population, West Virginia, and total U .S .A ., respectively . Cancer of Biliary Passages and Liver Based on seven deaths, the SMRs in males for cancer of biliary passages and liver were 133 (95% Cl : 53-274), 104 (95% Cl : 42-215), and 90 (95% Cl : 36-185) when comparing against the Region 1 DuPont Population, West Virginia, and total U .S .A ., respectively . There was only one death due to cancer of biliary passages and liver among females . Cancer of Pancrea s For males, the SMRs for cancer of the pancreas were 100 (95% CI : 50-180), 83 (95% Cl : 41-148), and 71 (95% Cl : 36-128) when comparing against the Region 1 DuPont Population, West Virginia, and total U .S .A ., respectively . There were no reported cases of cancer of the pancreas in females . 26 of 73 FINA L p . 27 Urinary Tract Cancer s There were 12 deaths from kidney cancer in males ; the SMRs were 185 (95% Cl : 95-323), 155 (95% Cl : 80-272), and 156 (95% Cl : 80-272) when comparing against the Region 1 DuPont Population, West Virginia, and total U.S.A., respectively . No deaths from kidney cancer were seen in females . There were 7 deaths from cancer of the bladder and other urina ry organs in males ; the SMRs were 131 (95% Cl : 53-269), 105 (95% Cl : 42-216), and 101 (95% Cl : 41-209) when comparing against the Region 1 DuPont Population, West Virginia, and total U .S.A., respectively. One death from bladder cancer was seen in females . Because there were few deaths from kidney cancer, there was not sufficient statistical power to fit Cox proportional hazard models for assessing the association of this outcome with exposure categories . Examination of job histories showed that only half the cases had ever worked in the APFO-use divisions . Cancer of Bronchus, Trachea, Lung For males, the SMR for cancer of the bronchus, trachea, and lung was 81 (95% Cl : 63-104) based on the Region 1 DuPont population The SMRs for cancer of the bronchus, trachea, and lung based on comparisons to West Virginia and total U.S.A. were 49 (95% Cl : 38-163) (WV) and 61 (95% Cl : 47-77) (U .S.A.) Cancer of Prostat e The SMR for cancer of the prostate was 65 (95% Cl : 34-114), and 58 (95% Cl : 30-100)), based on the Region 1 DuPont population and West Virginia, respectively . The SMR for cancer of the prostate based on comparisons to the total U .S .A . was 52 (95% Cl : 27-91) . Cerebrovascular Disease For males, the SMR for cerebrovascular disease was 86 (95% Cl : 60-120) based on the Region I DuPont population . The SMRs for cerebrovascular disease based on comparisons to West Virginia and total U.S.A . were 60 (95% Cl : 42-84) (WV) and 61 ((95% Cl : 42-85) (U .S .A.). For females, only one death was due to cerebrovascular disease . 27 of 73 C~18 FINA L p . 28 All Heart Diseas e For males, the SMR for all heart disease was 110 (95% Cl : 98-123) based on the Region I DuPont population . The SMRs for all heart disease based on comparisons to West Virginia and total U .S.A. were 66 (95% CI : 59-74) (WV) and 80 ((95% Cl : 71-89) (U .S .A.). For females, the SMRs for all heart disease were 143 (95% Cl : 46-333), 51 (95% Cl : 17119), and 64 (95% Cl : 21-150) when comparing against the Region 1 DuPont Population, West Virginia, and total U .S .A., respectively . Ischemic Heart Diseas e In males, the SMRs for ischemic heart disease were 109 (95% Cl : 96-124), and 69 (95% Cl : 61-78), based on the Region 1 DuPont population and West Virginia, respectively . The SMR for ischemic heart disease based on comparisons to total U .S.A. was 81 (95% Cl : 71-93) . There were only three deaths due to ischemic heart disease in females . Diabetes Mellitu s Mortality from diabetes mellitus among males was significantly elevated based on comparison to the DuPont Region 1 population (SMR= 183 ; 95% CI=112-283), but was below 100 .0 in comparisons to both the West Virginia (SMR= 67 ; 95% CI=41-104) and U .S .A. population (SMR= 81 ; 95% CI=50-125) . There were only two deaths attributed to diabetes among females . Cox Proportional Hazards Modeling Ischemic Heart Diseas e Table 3 presents the descriptive statistics for the white males used in the proportional hazards models for IHD, stratified by case/non-case status, and Table 4 shows descriptive statistics for this subset, stratified by never-APFO-use/ever-APFO-use . The first CHPM for ischemic heart disease was fit to data for 4,460 white males using the average intensity as the exposure of interest, with zero lag for exposure . Two models are presented in Table 5 : each includes PFOA exposure variables, and one adjusts for calendar year of the event and the other adjusts for hire date (pre or post-1954) in the model (1954 was the median date of hire) . Table 5 also shows the mortality rate ratios by exposure category for the no-lag models using case calendar-year or the binary variable for hired before 1954 as a potential confounder . When we looked separately at calendar year of death and date of hire, these two variables were both statistically significant . For calendar year of death, the MRR is less than one, which means that the background death rate for ischemic heart disease is going down over the study period . This is also true for the national mortality rates from ischemic heart disease in the U .S .A . (Figure 7) . As for the date of hire, those hired prior to 1954 had higher risk of death from IHD than those hired 28 of 73 FINA L p . 29 after 1954. However, when both of these variables (date of hire and calendar year of death) were accounted for in the model, only calendar year of death remained significant which means that the effect of the date of hire was confounded by the calendar year of death . Table 6 presents a summary of the CPHM analyses on IHD conducted for all white males in the cohort without lagging average intensity exposures, and those with lags of different time intervals (5, 10, 15, 20 years) to adjust for potential effects of HWSE with the inclusion of both potential time confounders (case calendar-year and hired before 1954) . These results show no significantly increased MRRs for IHD mortality between exposure categories for all analyses, with and without lags . Additionally, no trends of increasing MRRs are seen across exposure categories with the exception of category 4, the highest exposure group . The increase in MRR with increasing lag provides evidence that analysis using lagged exposures compensates for the healthy worker survivor effect for this cause of death . Details of these analyses are presented in Table 6 . Table 7 displays the results from both analyses of cumulative exposure categories for Cox proportional hazard models of the association between cumulative exposure and IHD mortality . In the first part of the table (section A), results are reported for categories determined by quartiles of cumulative exposure among case subjects only . This method ensures that each exposure category contains one-fourth of the cases (see exhibit 7A) for the corresponding lag model . Though no estimates of the hazard ratio are statistically significant in any model, the results from the 10-year lagged exposure model suggest an increasing trend in the mortality rate ratio for the highest two exposure categories . Due to the lack of consensus on a universally "best" approach to cumulative exposure categorization, we performed a second set of analyses using quartiles determined by the entire WW cohort as a form of sensitivity analysis . Table 7B lists the results of these models corresponding to cutpoints described in exhibit 7B . Although all estimates were still not statistically significant, mortality rate ratio estimates for this analysis attenuated towards the null value of 1 .0 for all lagged exposure models including those for the 10-year lag . Discussion In this retrospective cohort mortality study, we assessed whether workers at a polymer production plant exhibited increased mortality from any specific cause of death, as well as the more general categories of all causes and all malignant diseases . SMRs were generated using three different reference populations-a regional population from the same company, which reduced the bias from both the healthy worker effect and regional socio-cultural attributes ; the state population in which the plant was located ; and the general U .S.A. population . Although the site of the study is a manufacturing plant that produces a wide variety of products from many different chemicals, the only occupational exposure examined was APFO . The areas where this chemical is used employs about one-third of the site's workers . Exposure to APFO has been shown to cause benign neoplasms in rodent toxicology studies . Liver, Leydig-cell, and pancreatic acinar-cell tumors were observed in rats, but all of those findings are hypothesized to be mediated via PPARa . Humans have low PPARa receptor expression, and are not as responsive to PPARa agonists [25] . This study had low power to 29 of 73 so FINA L p . 30 detect excess risk for these rare tumors, and the increased SMR of 135 for cancer of the biliary passages and liver, based on seven cases in males, was not statistically significant . Prostate cancer and cerebrovascular disease, both reported as increased in previous 3M Company occupational epidemiology reports [1,2], were reduced in this cohort against all reference populations, cerebrovascular disease significantly so for the U .S.A and West Virginia populations . The few cases of each of these causes of death did not allow meaningful internal comparisons . Despite limited statistical power to evaluate mortality rates for specific cancers, some elevated relative risks did emerge that bear further scrutiny by worker surveillance and exposure monitoring . Although animal toxicology data and published occupational studies on workers exposed to PFOA do not provide any a priori reason to suggest a potential effect on risk for kidney cancer, comparisons against DuPont Region 1 returned increased, but non-significant SMRs . However, examination of work histories showed that few cases had spent appreciable time in the APFO areas . We did identify an increased mortality risk for diabetes mellitus in this cohort of workersdriven largely by 20 cases in males and two in females-when comparisons were made to the regional worker population from the same company (SMR= 183 ; 95% CI=112-283) . However, comparisons to West Virginia (SMR= 67 ; 95% CI=41-104) and to the general U .S .A. population (SMR= 81 ; 95% CI=50-125) did not indicate an increased risk of mortality due to diabetes . Although Cox proportional hazard modeling could be done, the small number of cases would severely limit the value of the estimates . There is a substantial literature supporting the under-reporting of diabetes, especially of lateonset or type II diabetes, on death certificates . Differences between countries' mortality reporting for diabetes has been shown to depend, among other factors, on physician differences in reporting this disease in Part I of the death certificate or as the underlying cause [26] . In the U .S .A ., a study of the frequency of reporting diabetes on death certificates of 540 known diabetics showed that diabetes was recorded on just 39 percent of the death certificates and as the underlying cause of death for only 10 percent of decedents with diabetes . In addition, diabetes was significantly less likely to be reported on the death certificates of decedents dying of cancer [27] . Our SMR analyses are based on the underlying cause of death, and it is reasonable to assume that the prevalence of diabetes in this cohort has been under-ascertained . We know of no reason why the same under-ascertainment should not apply to the reference populations as well ; thus, any bias in ascertainment would be non-differential . Cardiovascular death rates are higher in diabetics than in non-diabetics [28-29] . Although cardiovascular disease and late-onset diabetes share several risk factors (diets high in refined carbohydrates, sedentary lifestyle, age, and body mass index, for example), a study conducted in Iceland identified an independent effect of diabetes on coronary heart disease after adjustment for blood pressure, serum lipids, uric acid, smoking, and height and weight [30] . As discussed below, no increase in cardiovascular disease was noted for this cohort . Given the number of endpoints examined in the SMR analyses, it is not surprising to find an isolated increase in one of the causes of death . The lack of agreement with other studies of PFOA workers, and the lack of any animal toxicology findings to support this association suggest that the finding is due to chance . However, we will follow up on these results in future surveillance . 30 of 73 ~! p . 31 FINAL Our initial proportional hazards models of ischemic heart disease, an outcome potentially influenced by increased serum lipids, utilized cumulative exposure . This metric seemed to be the most biologically appropriate . The results of these models, conducted without lagging of exposure, showed significantly reduced rate ratios, which were assumed to be due to confounding by the healthy worker survivor effect (age was controlled for by using it as the time metric in the Cox model). We then turned to average intensity of exposure . With these analyses, we introduced the five-year lag of exposure, which appeared to mitigate the effects of the healthy worker survivor effect, prompting the re-analysis of cumulative exposure with lags . In order to stabilize the estimates of the mortality rate ratio, the exposure cutpoints were determined first by dividing the cases into quartiles, as is commonly done in occupational studies . The multi-dimensionality of occupational exposure metrics contributes to discussion about which metrics are the more robust, least biased, and most biologically meaningful [31] . In light of this debate, we have presented analyses using both average intensity and cumulative exposure, with lagging and without . In addition, we conducted analyses on cumulative exposure with th e cutpoints driven both by case and total cohort distributions of exposure, which provides a sensitivity analysis for the impact of cut-point selection on the observed exposure-response relationships . For ischemic heart disease mortality, no increases in mortality rate ratios were seen in the results of any analyses conducted using average exposure intensity. Many of the MRRs were below 1 .0 . It is well known, however, that for a chronic disease such as IHD, mortality underestimate s morbidity and is therefore not an ideal endpoint for epidemiologic analysis . Furthermore, the rate ratio for IHD may also be biased by the HWSE, even in internal analyses . The use of lagged exposures by 5 year periods to adjust for the healthy worker survivor effect did not have much effect, as estimates of the mortality rate ratio were still less than 1 .0. There was, however, an increased trend in MRRs as the lag time increased from 0 to 20 years by 5-year intervals for Exposure Category 4 (highest exposure category) . While these findings remain statistically nonsignificant, the increase in the MRR as the lag increased for the highest exposure group demonstrated that this approach did indeed compensate for the healthy worker survivor effect among those who worked long enough to achieve exposures in the highest category . For proportional hazards models of cumulative exposure, no significant increase in the MRR was observed with the exception of the 10-year lag model based on the set of exposure categories with an equal distribution of cases assigned to each exposure group . An elevated relative risk of 1 .6 was found in the highest exposure category, and there was an increasing trend in MRR with increasing exposure . While neither MRR estimate for the upper two exposure categories in the 10-year lag model was statistically significant, the apparent trend cannot be ignored . In order to investigate the sensitivity of the model to the categorization of cumulative exposure, we redefined cumulative exposure categories based on quartiles of the entire WW cohort . Effect estimates from this model were attenuated towards a null value . The highest relative risk was again seen in the 10-year lagged model where the MRR was 1 .3 for the highest exposure category, but there was no apparent trend . The overall absence of positive effect estimates using either the 5-, 15-, or 20-year exposure lags suggest that the positive exposure-response trend for cumulative exposure lagged by 10 years requires further investigation before firm conclusions can be reached . 31 of 73 3Z FINA L p . 32 When we looked separately at calendar year of death and date of hire (before or after 1954 for IHD), both time variables were statistically significant . For calendar year of death, the MRR was less than one, which indicates that the background death rate for ischemic heart disease was going down over the study period . This is also true for the national mortality rates from ischemic heart disease in the U .S.A. As for the date of hire, those hired prior to 1954 had higher risk of death from IHD than those hired after 1954 . However, when both of these variables (date of hire and calendar year of death) were accounted for in the model, only calendar year of death remained significant, evidence that the effect of the date of hire was confounded by the calendar year of death. This finding underlines the importance of exploring in detail all potential timevarying confounders in this type of analysis . Strengths of this study include the availability of biomonitoring data to support retrospective exposure classification and a large cohort with over fifty years of follow-up . In addition, there were sufficient mortality data to enable several types of analysis using both external and internal comparisons, including Cox proportional hazards analysis for ischemic heart disease . Limitations of this study include the potential loss to follow-up of decedents prior to 1979, a period where exposures may have been less well-controlled, causing potential bias towards the null due to the healthy worker survivor effect . Although this loss could have been as high as 10 percent, due to the ascertainment procedures for the DuPont Mortality Registry, it was likely much smaller due to the establishment of vital status of all cohort members through the Social Security Administration . Moreover, the lagged exposure approach applied to the ischemic heart disease analysis was designed to reduce such bias . A major limitation is likely to be the lack of accounting for confounding by other occupational and non-occupational risk factors . Most importantly, information was not available for members of the cohort about the major risk factors for cardiovascular disease (smoking, diet, and other life-style factors) . Additionally, no information was available for cohort members who were being treated with medications such as statins or anti-hypertensive medications . Conclusions The results reported here show no convincing evidence of increased mortality risk associated with APFO exposure for workers at this plant . These results do show statistically non-significant elevations in relative risk for kidney cancer and a statistically significant increase in diabetes mortality for workers at this site . However, given the size and length of follow-up of the study population, the evidence to thoroughly examine mortality events like kidney cancer or even diabetes, may not be adequate . Proportional hazards analyses for ischemic heart disease mortality showed an increase in the model based on equal distribution of cases across cumulative exposure categories in one lagged analysis (the 10-year lag period) . Other exposure lags showed no effect, and results for a second set of models using a different set of exposure cutpoints were attenuated toward the null . None of the hazard estimates themselves were statistically significant . Thus the positive finding in the proportional hazard analysis, as well as the increased diabetes mortality, might be due to chance . Because of the complexity of the exposure assessment and limited power for some analyses, additional investigations are needed . 32 of 73 ~,53 FINAL p . 33 References [1] Gilliland, F .D. and Mandel, J .S . (1993) . Mortality among employees of a perfluorooctanoic acid production plant . J. Occup . Med. 35 :950-954 . [2] Alexander, B .H . (2001) . Mortality study of workers employed at the 3M Cottage Grove Facility . Final Report . April 26, 2001 . Division of Environmental and Occupational Health, School of Public Health, University of Minnesota . US EPA AR226-1030a018 . [3] Haskell Report . (2006) . Ammonium Perfluorooctanoate : Cross-Sectional Surveillance of Clinical Measures of General. Health Status Related to a Serum Biomarker o f Exposure and Retrospective Cohort Mortality Analyses in a Polymer Production Plant . In review . [4] Olsen, G .W ., Burris, J .M., Burlew, M .M., Mandel, J .H. (2003) . Epidemiological assessment of worker serum perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) concentrations and medical surveillance examinations . J. Occup . Environ. Med. 45 :260-270. [5] DuPont Haskell Laboratory (1981) . Liver study of Washington Works employees exposed to C-8 : Results of blood biochemistry testing . Unpublished report, HL-673-96 . [6] Gilliland, F .D. and Mandel, J .S . (1996) . Serum perfluorooctanoic acid and hepatic enzymes, lipoproteins, and cholesterol : a study of occupationally exposed men . Am . J. Ind. Med. 29 : 560-568 . [7] Olsen, G .W ., Burris, J .M., Burlew, M .M., and Mandel, J .H. (2000) . Plasma cholecystokinin and hepatic enzymes, cholesterol and lipoproteins in ammonium perfluooctanoate production workers . Drug Chem. Toxicol . 23 : 603-620 . [8] Emmett, A .E ., Shofer, F .S ., Zhang, H ., Freeman, D ., Desai, C ., and Shaw, L .M . (2006a) . Community exposure to perfluorooctanoate : Relationships between serum concentrations and exposure sources . JOEM 48 :759-770 . [9] Emmett, A .E ., Zhang, H ., Shofer, F .S ., Freeman, D ., Rodway, N .V., Desai, C ., and Shaw, L.M. (2006b) . Community exposure to perfluorooctanoate : Relationships between serum levels and certain health parameters . JOEM 48 : 771-779 , [10] Olsen, G .W ., Burris, J.M., Lundberg, J .K., Hansen, K .J., Mandel, J .H., and Zobel, L .R. (2002a) . Identification of fluorochemicals in human sera . I . American Red Cross adult blood donors . US EPA Public Docket AR-226 . [11] Olsen, G .W ., Burris, J .M ., Lundberg, J .K ., Hansen, K.J ., Mandel, J .H ., and Zobel, L .R . (2002b) . Identification of fluorochemicals in human sera . II . Elderly participants in the adult changes in thought study, Seattle, Washington . US EPA Public Docket AR-226 . 33 of 73 ,-.)/ FINA L p . 34 [12] Olsen, G .W ., Burris, J.M ., Lundberg, J .K., Hansen, K.J., Mandel, J .H ., and Zobel, L .R. (2002c) . Identification of fluorochemicals in human sera. III . Pediatric participants in a group A Streptococci clinical trial investigation. US EPA Public Docket AR-226 . [13] Burris J .M. ., Lundberg J .K., Olsen G .W ., Simpson D ., and Mandel G . (2002) . Determination of serum half-lives of several fluorochemicals . 3M Company . Interim Report #2. January 11, 2002 . US EPA AR 226-1086 . [14] Washburn, S .T ., Bingman, T .S ., Braithwaite, S .K ., Buck R .C ., Buxton, W ., Clewell, H .J ., Horoun, L .A ., Kester, J .E ., Rickard, R.W ., and Shipp, A .M . (2005) . Exposure assessment and risk characterization for perfluorooctanoate in selected consumer articles . Environ . Sci . Technol . 39 : 3904-3910 . [15] Kennedy, G .L ., Jr ., Butenhoff, J .L ., Olsen, G .W ., O'Connor, J .C ., Seacat, A .M ., Perkins, R.G ., Biegel, L .B ., Murphy, S .R ., Farrar, D .G . (2004) . The toxicology of perfluorooctanoate . Crit . Rev . Toxicol . 34 :351-384. [16] Sakr, C .J ., Leonard, R.C ., Kreckmann, K .H ., Slade, M .D ., Eisen, E .A ., and Cullen, M .R. (2006) . Longitudinal study of serum lipids and liver enzymes in workers with occupational exposure to ammonium perfluorooctanoate (APFO) . Manuscript in review . [17] Marsh, G .M., Youk, A.O ., Stone, R .A ., Sefcik, S .S., Alcorn, C .A . (1998) . OCMAPPLUS : A new occupational cohort mortality analysis program for multifactor work history and exposure- based analysis . J. Occup . Environ . Med. 40: 351-362 . [18] Gilbert, E . S . (1982) . Some confounding factors in the study of mortality and occupational exposures . Am . J. Epidemiol . 116 : 177-188. [19] Morgan, R . W ., Kelsh, M . A ., Zhao, K., Heringer, S . (1998) . Mortality of aerospace workers exposed to trichloroethylene . Epidemiology 9 : 424-431 . [20] Checkoway, H . and Pearce, N .E . Research Methods in Occupational Epidemiology, Oxford University Press, New York, 1989 . [21] Gun, R .T ., Pratt, N .L ., Griffith, E .C ., Adams, G .G ., Bisby, J .A ., and Robinson, K.L . (2004) . Update of a prospective study of mortality and cancer incidence in the Australian petroleum industry . Occup . And Environ . Med . 61 : 150-156 . [22] Arrighi, H .M . and Hertz-Picciotto, I . (1994) . The evolving concept of the healthy worker survivor effect . Epidem . 5 : 189-196 . [23] Hertz-Picciotto, I ., Arrighi, H .M ., Hu, S .-W . (2000) . Does arsenic exposure increase the risk for circulatory disease? Am . J . Epidemiol . 151 : 174-181 . [24] Szklo, Moyses, Szklo, M ., Nieto, F .J . Epidemiology : Beyond the Basics . Aspen Publications, 1999 . 34 of 73 ..35 FINAL p . 35 [25] Klaunig, J .E., Babich, M .A., Baetcke, K.P ., Cook, J .C., et al . (2003) PPARa agonistinduced rodent tumors : Modes of action and human relevance . Crit. Rev in Toxicol . 33 :655-780. [26] Lu, T.H., Walker, S ., Johansson, L .A., and Huang, C .N. (2005) . An international comparison study indicated physicians' habits in reporting diabetes in part I of death certificates affected reported national diabetes mortality . J. Clin. Epidem . 58 :1150-1157. [27] McEwen, L .N., Kim, C ., Haan, M ., Ghosh, D., Lantz, P.M., Mangione, C .M., Saford, M .M., Marrero, D ., Thompson, J.J., and Herman, W.H . (2006) . Diabetes reporting as a cause of death-Results from the Translating Research Into Action for Diabetes (TRIAD) study . Diabetes Care 29: 247-253 . [28] Roper, N .A., Bilous, R .W ., Kelly, W .F., Unwin, N .C., and Connolly, V .M. (2002) . Causespecific mortality in a population with diabetes-South Tees Diabetes mortality study . Diabetes Care 25 :43-48 . [29] Fox, C .S., Coady, S ., Sorlie, P .D., Levy, D ., Meigs, J .B., D'Agostino, R.B., Wilson, P.W.F., Savage, P.J. (2004) . Trends in cardiovascular complications of diabetes . J .A.M.A. 292 :2495-2499. [30] Vilbergsson, S ., Sigurdsson, G ., Sigvaldason, H ., and Sigfusson, N . (1998) . Coronary heart disease mortality among non-insulin-dependent diabetic subjects in Iceland : the independent effect of diabetes . The Reykjavik Study 17-year follow-up . J. Internal Med . 244 :309-316. [31 ] Loomis, D . and Kromhout, H . (2004) . Exposure variability : Concepts and applications in occupational epidemiology. Am . J . Indust . Med. 45 :113-122. 35 of 73 FINA L TABLES p . 36 36 of 73 37 FINAL p . 37 Table 1 Washington Works mortality study cohort Washington Works (as of 12/31/2002 ) Mortality Study Cohort Males Females n Mean Min Max n Mean Min Max Cohort 4872 115 5 Age at Hire 29 11 70 27 17 56 Year of Hire 1974 1948 2002 1986 1948 2002 Age at Termination 50 19 74 32 18 71 Year of Termination 1989 1955 2004 1993 1967 2004 Duration of Hire (Yrs) 19 0 49 10 0 44 Yrs of Follow Up 26 0 55 16 0 55 Age at End of Follow 55 20 96 43 Up % White 95 .40% 92 .29% 21 85 #Active 1650 429 Duration of Hire (Yrs) 17 0 41 14 0 4 0 Age at End of Follow Up 46 23 68 44 22 67 #Dead 773 33 #Cancer Deaths 222 1 2 AgeAtDeath 66 22 96 58 %White 99 .09% 96 .97 % 22 85 37 of 73 FINAL p . 38 Table 2 SMRs for selected causes of death in Washington Works males, females compared to DuPont Region 1(West Virginia (less Washington Works), Ohio, Virginia, Kentucky, Indiana, Pennsylvania, Tennessee, and North Carolina), U .S .A . national popula ti on, West Virginia state populatio n Cause of Death MALES FEMALES N=4872 N=1155 Total Mortality : 773 Total Mortality : 3 3 Person Years : 127,513 .2 Person Years : 18,224 . 5 West West DuPont U .S .A. Virginia DuPont U .S .A. Virginia Region 1 National State Region 1 National Stat e N SMR SMR SMR N SMR SMR SMR All Causes of Death (773) 93 .6 66 .2** 58 .1** (33) 147 .2* 80.7 73 . 4 All Malignant Neoplasms (222) 100 .4 73 .7** 68 .3** (12) 149 .0 86.6 79 .4 Cancer of Biliary Passage s & Liver (7) 133 .1 89 .7 104 .2 (1) 384 .8# 394.5# 551 .8# Cancer of Pancreas (11) 100 .5 74 .0 82 .9 0 0 0 0 Cancer of Bronchus , Trachea, Lung (64) 81 .3 60 .6** 49 .0** (2) 132 .9^ 69 .5^ 56 .6^ Cancer of Prostate (Male s only) (12) 65 .3 51 .8** 57 .5 N/A 0 0 0 Cancer of Breast (0) 0 0 0 (2) 77 .4^ 61 .1^ 63 .5^ Cancer of Kidney (12) 184 .7 155 .7 155 .2 0 N/A N/A N/A Diabetes (20) 183 .1 * 81 .2 67 .0 (2) 796 .1 ^ 160 .8^ 121 .7^ CerebrovascularDisease (34) 86 .1 60 .9** 60 .1** (1) 90 .5# 48 .7# 49 .7# All Heart Disease (309) 109 .9 80 .0** 66 .3** (5) 142 .7 64 .4 51 .1 Ischemic Hea rt Disease (236) 109 .3 81 .4** 69 .0 (3) 135 .0 64 .0 49 . 7 (*) SIGNIFICANT AT 5% LEVEL ; (**) SIGNIFICANT AT 1% LEVEL ; (#) BASED ON 1 CASE ; (^) BASED ON 2 CASES N/A Not Applicable 38 of 73 ~ 4~ FINAL p . 39 Table 3 White male workers included in the risk-sets of the proportional hazard analysis for IHD stratified by case/non-case statu s Variable CASES NON-CASES N(%) or Mean (SD) N(%) or Mean (SD) [Minimum-Maximum] [Minimum-Maximum] Total Number 235(5 .27%) 4225 (94 .73 %) Males 235(100%) 4225(100%) White Race (v/s non white) 235(100%) 4225(100%) Age at hire 33 .72 (9 .82) 29 .06 (8 .77) [18.21-65.85] [11 .23-70 .17] Year of birth 1921 (11 .20) 1945 (15 .97) [1892-1958] [1890-1973 ] Year of hire 1955 (8 .92) 1974 (15 .71) [1948-1995] [1948-2002] Year of death 1987 (11 .30) N A [1958-2002] Ever-APFO-use 68(28 .94%) 2185 (51 .72%) 39 of 73 ~ FINAL p . 40 Table 4 White male workers included in the risk-sets of the proportional hazard analysis stratified by never-APFO-use/ever-APFO-use statu s Variable NEVER-APFO-USE EVER-APFO-USE N(%) or Mean (SD) N(%) or Mean (SD) [Minimum-Maximum] [Minimum-Maximum] Total Number 2207 (49 .48 %) 2253 (50 .52 %) Males 2207(100%) 2253 (100%) White Race 2207(100%) 2253 (100%) (v/s non white) Age at hire 30.89 (9.42) 27 .74 (8 .04) [14.40-70.17] [11 .23-64.15] Year of birth 1941 (18 .37) 1947 (14.10 ) [1890-1972] [1903-1973] Year of hire 1972 (17 .10) 1974 (14 .72) [1948-2002] [1948-2002 ] Average intensity 0.21 (0) 0 .42 (0 .35) (at end of follow-up) [0 .21-0 .21] [0 .21-1 .69] Cumulative exposure 4 .06 (2 .51) 9 .10 (10 .00) (at end of follow-up) [0 .00-9 .02] [0 .01-71 .85] Time since hire 18 .83 ( 12 .20) 21 .00 (12 .63) (at end of follow-up) [0 .00-42 .27] [0 .05-48 .54] Cases 167(7 .57%) 68(3 .02%) 40 of 73 7/ p . 41 FINAL Table 5 Mortality rate ratios for IHD by exposure category for no-lag analyses using case calendar- year and year of hire (pre-1954) as potential confounders Hired pre-1954 Case-calendar NO LAG yea r N = 4,460 NO LAG N = 4,460 MRR MR R (95 % CI) (95 % CI) Reference 1 1 Category 1 0 .858 0 .996 (0.569-1 .293) (0 .657-1 .509) Category 2 0 .575 0.715 (0.319-1 .035) (0 .394-1 .298 ) Category 3 0 .767 0.944 (0.444-1 .327) (0 .542-1 .645) Category 4 0 .558 0.64 6 (0.323-0.966) (0 .372-1 .123) Case Calendar- --- 0.965 Year (0.952-0 .978) Hired before 1 .42 2 1954 (1 .075-1 .881) 41 of 73 FINAL p . 42 Table 6 Mortality rate ratios for IHD by average intensity exposure category, including increasing 5-year lags of exposure, using case calendar-year and year of hire (pre-1954) as potential confounders NO LAG 5-YEAR LAG 10-YEAR LAG 15-YEAR LAG 20-YEAR LAG N = 4,460 N = 4,440 N = 3,989 N = 3,986 N = 3,440 MRR MRR MRR MRR MRR (95 % CI) (95 % CI) (95 % CI) (95 % CI) (95 % CI) Reference* 1 1 1 1 1 Category 1 0 .996 1 .035 1.048 0 .976 0.884 (0.657-1 .510) (0 .689-1 .557) (0 .693-1 .582) (0 .643-1 .482) (0 .574-1 .362) Category 2 0 .715 0 .657 0.688 0 .763 0.976 (0.394-1 .298) (0 .361-1 .195) (0 .369-1 .284) (0 .409-1 .427) (0 .573-1 .663 ) Category 3 0 .944 0 .738 0.802 0 .943 0.828 (0.541-1 .646) (0 .416-1 .310) (0 .442-1 .457) (0 .519-1 .715) (0 .481-1 .424) Category 4 0 .646 0 .842 0.890 0 .975 (0.372-1 .123) (0 .466-1 .521) (0 .491-1 .611) (0 .538-1 .769) Case Calendar- 0 .965 0 .965 0.963 0 .963 0.964 Year (0.951-0.979) (0 .950-0 .980) (0 .947-0.979) (0 .946-0 .981) (0 .944-0.985) Hired before 1 .001 1 .042 1.089 1 .087 1.053 1954 (0.738-1 .360) (0 .764-1 .421) (0 .791-1 .501) (0 .780-1 .513) (0 .744-1 .492) * Exposure distributions for each category by lag period are listed in exhibit 6 . 42 of 73 ~~ FINAL p . 43 Table 7 Mortality rate ratios for IHD by cumulative exposure category, including increasing 5-year lags of exposure, using case calendar-year and year of hire (pre-1954) as potential confounders ; A) exposure categories based on case distribution, B) exposure categorie s based on cohort distribution . NO LAG 5-YEAR LAG 10-YEAR LAG 15-YEAR LAG 20-YEAR LAG A) N = 4,460 N = 4,440 N = 3,989 N = 3,986 N = 3,440 MRR MRR MRR MRR MR R (95 % CI) (95 % CI) (95 % CI) (95 % CI) (95 % CI) Reference* 1 1 1 1 1 Category 1 1 .046 0 .864 0 .996 0 .935 0.647 (0.711-1 .539) (0.579-1 .291) (0 .647-1 .531) (0.598-1 .461) (0.407-1 .029) Category 2 1 .156 1 .204 1 .377 1 .102 0.692 (0.745-1 .793) (0.757-1 .914) (0 .829-2 .287) (0.644-1 .887) (0 .388-1 .233 ) Category 3 1 .110 1 .077 1 .610 1 .089 0.764 (0 .698-1 .767) (0 .657-1 .764) (0 .942-2 .753) (0 .597-1 .984) (0 .393-1 .489) Case Calendar- 0 .959 0 .958 0 .948 0 .958 0 .969 Year (0 .942-0 .977) (0 .940-0 .976) (0 .928-0 .968) (0 .935-0 .982) (0 .943-0 .997) Hired before 0 .946 0 .952 0 .903 1 .032 1 .096 1954 (0.672-1 .333) (0 .671-1 .352) (0 .633-1 .289) (0 .710-1 .501) (0 .733-1 .639) B) Reference t 1 1 1 1 1 Category 1 1 .018 0 .980 1.062 0 .914 0.967 (0.515-2.010) (0 .607-1 .582) (0.662-1 .680) (0 .557-1 .502) (0 .568-1 .647 ) Category 2 1 .132 1 .014 0.825 0 .860 0.931 (0.573-2.235) (0 .603-1 .705) (0 .483-1 .409) (0 .494-1 .497) (0 .505-1 .717) Category 3 1 .027 1 .019 1.256 1 .069 0.88 2 (0.496-2.127) (0 .571-1 .817) (0 .721-2.191) (0 .598-1 .910) (0 .461-1 .688) Case Calendar- 0 .961 0.961 0 .954 0 .958 0 .967 Year (0.945-0.978) (0 .943-0.980) (0 .934-0.975) (0 .936-0.981) (0 .940-0.994) Hired before 0 .969 1 .008 0 .974 1 .027 1 .095 1954 (0.692-1 .357) (0 .711-1 .429) (0 .679-1 .398) (0 .710-1 .484) (0 .742-1 .614) * Exposure distributions for each category by lag period are listed in exhibit 7A . t Exposure distributions for each category by lag period are listed in exhibit 7B . 43 of 73 44 FINAL p . 44 FIGURES 44of73 FINAL p . 45 Figure 1 . Time in Job vs Serum PFOA-Cross-Sectional Study-Job Exposure Category 1 Job Exposure Category 1 APFO-Use Jobs 10 CL JobExpCatl 0LL 0.1 G. E 2 -3 Years 8Years 0 0 .25 ppm .75 ppm 0 .0 1 0.001 0 2 4 6 8 Years In Job 10 12 45 of 73 <~ ~ FINAL p . 46 Figure 2 . Time in Job vs Serum PFOA-Cross-Sectional Study-Job Exposure Category 2 Job Exposure Catego ry 2 10 ,~ _ . .. . e . _ 0 JobExpCat2 CL Q -3 Years OU. -8 Years a 0 .25 ppm E U) 0 .75 ppm 0 .1 0 .01 0 2 4 6 8 10 12 14 16 18 2 0 Years In Job 46 of 73 FINAL p . 47 Figure 3 . Time in Job vs Serum PFOA-Cross-Sectional Study-Job Exposure Category 3 Job Exposure Catego ry 3 10 LL 1 Q. E _ . __ . . 0 , ., _, _ . . . P N 0 JobExpCat3 - 3 Year s 8 Years .25 ppm 0 .75 ppm 0.1 0 2 4 6 8 10 12 Years In Job 14 16 18 47 of 73 FINAL 10 0 10 Figure 4 . Serum PFOA vs Cumulative Exposure - FLAIR Dat a Serum PFOA vs Cumulative ExposurePotential p . 48 oa. 1 OU. IL E ~a`+ 0 .1 0 .01 0 .001 0.01 0 .1 1 10 Cumulative Exposure (ppm years) 10 0 48 of 73 7 / FINAL p . 49 Figure 5 . Serum PFOA vs Average Intensity of Exposure - FLAIR Data Serum PFOA vs Average Intensity 100 10 E 1 CL a 0LL CL E m 0 .1 y 0 .01 0 .001 0 .1 1 1 0 Average Intensity (ppm) 49 of 73 '50 FINAL p . 50 Figure 6 . Serum PFOA vs Concurrent Job Intensity Factor - FLAIR Data Serum PFOA vs Concurrent Job Intensity facto r 100 10 E 1 CL 0LL a E 0 .1 U) 0 .01 0 .001 0 .1 1 1 0 Concurrent Job Intensity Factor (ppm) 50 of 73 p . 51 FINAL Figure 7 . Decreasing IHD mortality rates in the U .S .A . US rates of death from IH D 18 16 14 12 -~- 20-24 -t- 25-29 30-34 10 35-3 9 40-44 --0-45-49 8 +50-54 -55-59 -60-64 6 65-69 4 2 0 - - - 1960-64 1965-69 1970-74 1975-79 1980-84 1985-89 1990-94 1995-99 2000-01 51 of 73 FINAL p . 52 APPENDICES 52 of 73 ~~~ FINAL p . 53 Appendix A Washington Works vs Region 1 All-Cause Mortality Surveillance Report : Males Cause of Death Observed Expected 95% 95% 99% 99% SMR Lower Upper Lower Upper All Causes of Death 773 826 .1 93 .6 87 .1 100 .4 85 .1 102 . 6 Tuberculosis 0 0 .4 N/A 0 .0 866 .9 0 .0 1245 .0 All Malignant Neoplasms 222 221 .2 100 .4 87 .6 114 .5 83 .8 119 .1 Can cer of Buccal Cavity & Pharynx 4 3 .2 123 .5 33 .6 316 .1 20 .7 388 .7 Can cer of Digestive Organs & Peritoneum 49 52 .6 93 .2 69 .0 123 .2 62.5 133 . 3 Can cer of Esophagus 4 4 .8 83 .5 22 .8 213 .8 14.0 262 .9 Can cer of Stomach 2 5 .7 35 .0 4 .2 126 .3 1 .8 162 .1 Can cer of Large Intestine 17 20.9 81 .5 47 .5 130 .4 39.5 147 .5 Can cer of Rectum 5 3 .7 135 .3 43 .9 315 .6 29.2 382 .8 Can cer of Biliary Passages & Liver 7 5 .3 133 .1 53 .5 274 .2 38.7 325 .7 Can cer of Pan creas 11 10.9 100 .5 50 .2 179 .8 39.5 208 .1 Cancer of All Other Digestive Organs 3 1 .3 232 .8 48 .0 680 .4 26.2 852 . 0 Can cer of Respiratory System Cancer of 70 82 .3 85 .1 Larynx 3 1 .5 195 .6 66 .3 107 .5 61 .2 115 .0 40 .4 571 .7 22 .0 715 .8 Cancer of Bronchus, Trachea, Lung 64 78 .7 81 .3 62 .6 103 .8 57.5 111 .3 Cancer of All Other Respiratory 3 2 .0 151 .0 31 .2 441 .2 17 .0 552 . 5 Can cer of Breast C an cer of Prostate males only) Cancer of Testes and Other male genital Organs 0 0 .3 N/A 0 .0 1425 .3 0.0 2046 .9 12 18 .4 65 .3 33 .8 114 .1 26.9 131 .5 1 0 .6 169 .7 4 .2 945 .7 0.8 1261 .1 C ancer of Kidney C ancer of Bladder and O th er Urinary Organs 12 6 .5 184 .7 95 .4 322 .6 76 .1 371 .6 7 5 .4 130 .7 52 .6 269 .4 38 .0 320 .0 Malignant Melanoma C ancer of Eye C ancer of Central Nervous System of Skin 2 4 .4 45 .5 5 .5 164 .4 2 .3 211 .0 0 0 .4 N/A 0.0 886 .7 0 .0 1273 .5 9 6 .9 130 .1 59 .5 246 .9 45 .3 289.0 C ancer of Thyroid & Other Endocri ne Glands 3 0 .5 C ancer of Bone Cancer of All Lymphatic, Haematopoietic Tissue 29 2 0 .3 24 .0 Non-Hodgkins Lymphoma 9 8 .2 Hodgkins Disease 2 1 .1 Leukemia & Aleukemia 12 10 .3 Cancer of All Other Lymphopoietic Tissue 6 4 .4 All Other Malignant Neoplasms 22 15 .5 633 .2 * 130 .7 1850 .4 71 .3 2316 .9 648 .3 78 .4 2342 .1 33 .4 3006 .3 1 20.7 109 .9 179 .8 116 .1 136 .7 141 .9 80 .8 173 .4 70 .8 191 .4 50 .3 208 .6 38 .2 244 .2 21 .8 649 .5 9 .3 833 .7 60 .0 202 .8 47 .8 233 .6 50 .2 297 .6 35 .0 356 . 9 88 .9 214 .8 76 .0 240 .0 53 of 73 64 FINAL p . 54 Appendix A Washington Works vs Region 1 All-Cause Mortality Surveillance Report : Males (Continued) Cause of Death Observed Beni Neoplasms 1 Diabetes Mellitus 20 Expected 95% 95% 99% 99% SMR Lower Upper Lower Upper 2 .4 42 .0 1 .0 233 .8 0 .2 311 .7 10 .9 183 .1 * 111 .8 282.8 94.8 317 .4 Cerebrovascular Disease 34 39 .5 86 .1 59 .6 120 .3 52 .8 132 .0 All Heart Disease Rheumatic Heart Disease 309 281 .0 109 .9 5 1 .7 302 .5 98 .0 122 .9 94 .5 127 . 1 98 .2 706 .0 65 .2 856 .2 Ischemic Hea rt Disease 236 215 .9 109 .3 Chronic Endocard. Dis. ; Other Myocard . Insuff. 11 10 .3 106 .4 95 .8 124 .2 91 .8 129 .0 53 .1 190 .3 41 .8 220 .2 Hypertension with Heart Disease 1 6 .2 16 .2 * 0.4 90 .1 0 .1 120 .1 All Other Heart Disease 56 Hypertension w/o Hea rt Disease 5 46 .9 119 .4 2 .3 214 .4 90 .2 155 .0 82 .3 166 .9 69.6 500 .4 46 .2 606 .8 Non-malignant Respiratory Disease 46 50 .5 91 .1 Influenza & Pneumonia 14 15 .4 90.7 66 .7 121 .6 60.2 131 .8 49 .6 152 .1 40.4 173 .8 Bronchitis, Emphysema, Asthma Bronchitis Emphysema Asthma 11 11 .2 98 .6 49 .2 176 .5 38.7 204 .2 5 3 .7 133 .8 43 .4 312 .1 28.8 378 .5 6 6 .7 88 .9 32 .6 193 .6 22.8 232 .1 0 0.7 N/A 0 .0 551 .5 0.0 792 .0 Other Non-malignant Respiratory Disease 21 23 .9 88 .0 54 .4 134.4 46 .4 150 .6 Ulcer of Stomach & Duodenum 0 1 .5 N/A 0 .0 242 .9 0 .0 348 .8 Cirrhosis of Liver 8 9 .2 86 .9 Nephritis & Ne hrosis 8 6 .0 132 .5 37 .5 171 .2 27 .9 201 .8 57 .2 261 .1 42 .6 307 .8 All External Causes of Death 41 65 .2 62 .9 45 .1 85 .3 40 .5 92 . 9 Accidents 31 40 .4 76 .7 52 .1 108 .9 45 .9 119 .9 Motor Vehicle Accidents 20 22 .8 87 .6 53 .5 135 .3 45 .3 151 .9 All Other Accidents 11 17 .6 62 .6 31 .3 112 .0 24 .6 129 .7 Suicides 8 19 .1 41 .8 ** 18 .1 82 .4 13 .4 97 .2 Homicides & Other External Causes 2 5 .7 35 .3 4 .3 127 .4 1 .8 163 .5 All Other Causes of Death 78 72 .1 108 .1 85 .5 134 .9 79 .2 143 .8 CERTAIN INFECTIOUS AND PARASITIC DISEASES I 1 .0 100 .1 2 .5 557 .9 0 .5 743 .9 Unknown Causes (In All Causes Category Only) 0 (*) SIGNIFICANT AT 5% LEVEL ; (**) SIGNIFICANT AT I% LEVE L 54 of 73 FINAL p . 55 Appendix A Washington Works vs Region 1 All-Cause Mortality Surveillance Report : Female s Cause of Death Observed Expected 95% 95% SMR Lower Upper 99% 99% Lower Upper All Causes of Death 33 22 .4 147 .2 * 101 .3 206 .7 89 .5 226 .9 Tuberculosis 0 0 .0 N/A 0.0 491866 .7 0.0 706400 .0 All Malignant Neoplasms 12 8 .1 149 .0 77 .0 260 .3 61 .4 299 .8 Can cer of Buccal Cavity & Pharynx 0 0 .2 N/A 0.0 2064 .0 0 .0 2964 .2 Can cer of Digestive Organs & Pe ritoneum 2 1 .5 130 .2 15 .8 470 .4 6 .7 603 . 8 Cancer of Esophagus 0 0 .0 N/A 0.0 14899 .0 0 .0 21397 .4 Can cer of Stomach 1 0 .0 2586 .7 64.7 14412.8 12 .9 19218.8 Cancer of Large Intestine 0 0 .8 N/A 0.0 440 .6 0 .0 632 .7 Cancer of Rectum 0 0 .1 N/A 0.0 4151 .5 0 .0 5962 .2 Cancer of Bilia Passages & Liver 1 0 .3 384 .8 9 .6 2144 .2 1 .9 2859 .2 Cancer of P an creas 0 0 .3 N/A 0.0 1467 .2 0 .0 2107 .1 Can cer of All Other Digestive Organs 0 0 .0 N/A 0.0 10537.0 0 .0 15132 . 8 Can cer of Respiratory System 2 1 .5 132 .2 16 .0 477.5 6 .8 613 .0 Cancer of Larynx 0 0 .0 N/A 0 .0 43969 .0 0 .0 63146 .6 Cancer of Bronchus, Trachea, Lung 2 1 .5 132 .9 16 .1 480.2 6 .8 616.4 Cancer of All Other Respiratory 0 0.0 N/A N/A C an cer of Breast 2 2 .6 77 .4 9 .4 279 .8 4 .0 359 .1 All Uteri ne Cancers (Females only) 0 0.0 N/A 0 .0 17174.1 0 .0 24664 .8 C an cer of Ce rvix Uteri Females only) 0 0.0 N/A 0 .0 20449 .0 0 .0 29368 .1 C an cer of Other Female Genital Organs 0 0.7 N/A 0 .0 516 .9 0 .0 742 .3 C an cer of Kidney 0 0 .1 N/A 0 .0 2793 .4 0 .0 4011 .8 C ancer of Bladder and Other Urinary Organs 0 0 .0 N/A 0 .0 20841 .8 0 .0 29932 .2 Malignant Melanoma of Skin 1 0 .0 2138 .6 53 .5 11916 .2 10.7 15889 .6 C ancer of Eye 0 0 .0 N/A 0 .0 40988 .9 0 .0 58866 .7 C ancer of Central Nerv ous System 0 0 .2 N/A 0 .0 1903 .6 0 .0 2733 .9 C ancer of Thyroid & Other Endoc ri ne Glands 0 0 .0 N/A 0 .0 105702 .0 0 .0 151805 .2 C ancer of Bone 0 0 .0 N/A N/ A Cancer of All Lymphatic, Haematopoieti c Tissue 3 0 .8 395 .0 81 .5 1154 .4 44.5 1445 .4 Non-Hodgkins Lymphoma 0 0 .1 N/A 0 .0 2890 .2 0 .0 4150 .7 Hodgkins Disease 0 0 .2 N/A 0 .0 2072 .1 0 .0 2975 .9 Leukemia & Aleukemia 1 0 .3 292 .7 7 .3 1630 .9 1 .5 2174 .7 Cancer of All Other Lym ho oietic Tissue 2 0 .1 1783 .8 * 215 .8 6444 .0 91 .9 8271 .5 All Other Malignant Neoplasms 2 0 .3 611 .0 73 .9 2207 .3 31 .5 2833 .3 55 of 73 FINAL p . 56 Appendix A Washington Works vs Region 1 All-Cause Mortality Surveillance Report : Females (continued ) Cause of Death Observed Expected 95% 95% 99% 99% SMR Lower Upper Lower Upper Benign Neoplasms 0 0.0 N/A 0.0 18528 .4 0 .0 26609.7 Diabetes Mellitus 2 0.3 796 .1 96 .3 2876 .1 41 .0 3691 .7 Cerebrovascular Disease 1 1 .1 90 .5 2 .3 504 .4 0 .5 672 .6 All Heart Disease 5 3 .5 142 .7 46 .3 333 .0 30 .8 403 .8 Rheumatic Heart Disease 0 0.1 N/A 0 .0 4456 .4 0 .0 6400 .1 Ischemic Hea rt Disease Chronic Endocard. Dis. ; Other Myocard . Insuff. 0 3 2.2 135 .0 27 .8 394 .4 15 .2 493 .8 0.1 N/A 0.0 3005 .8 0 .0 4316 .8 Hypertension with Heart Disease 0 0 .2 N/A 0 .0 1724 .2 0 .0 2476 .3 All Other Heart Disease 2 0 .9 232 .2 28 .1 838 .7 12 .0 1076 . 5 Hypertension w/o Heart Disease 0 0 .1 N/A 0 .0 3461 .6 0 .0 4971 .4 Non-malignant Respiratory Disease 3 1 .0 294.6 60.8 860 .8 33 .2 1077 .8 Influenza & Pneumonia Bronchitis, Emphysema, Asthma 0 2 0 .3 N/A 0 .0 1292 .5 0 .0 1856 .3 0 .3 679 .9 82 .3 2456 .1 35 .0 3152 . 6 Bronchitis 1 0 .1 756 .4 18 .9 4214 .8 3 .8 5620 .3 Emphysema 1 0 .1 1411 .6 35 .3 7865 .6 7 .1 10488 .4 Asthma 0 0 .1 N/A 0 .0 4048 .1 0 .0 5813 . 7 Other Non-malignant Respiratory Disease 1 0 .4 227 .9 5 .7 1269 .6 1 .1 1692 .9 Ulcer of Stomach & Duodenum 0 0 .0 N/A 0 .0 8221 .5 0.0 11807 .4 Cirrhosis of Liver 1 0 .1 804 .8 20 .1 4484 .5 4 .0 5979 .9 Nephritis & Ne hrosis 0 0 .1 N/A 0 .0 2671 .3 0 .0 3836 .4 All External Causes of Death 4 3 .1 130 .8 35 .7 335 .0 22 .0 412 . 0 Accidents 4 1 .6 247 .3 67 .4 633 .3 41 .6 778 .7 Motor Vehicle Accidents 3 1 .4 215 .7 44 .5 630.4 24.3 789 .3 All Other Accidents 1 0 .2 441 .4 11 .0 2459.5 2 .2 3279 . 6 Suicides 0 0 .6 N/A 0 .0 660.3 0 .0 948 .3 Homicides & Other External Causes 0 0 .9 N/A 0 .0 418.6 0 .0 601 .2 All Other Causes of Death 5 2 .2 223 .2 72 .4 520.7 48.1 631 .5 CERTAIN INFECTIOUS AND PARASITIC DISEASES 0 0 .1 N/A 0 .0 6277 .0 0 .0 9014 .8 Unknown Causes (In All Causes Category Only) 0 (*) SIGNIFICANT AT 5% LEVEL; (**) SIGNIFICANT AT 1% LEVE L 56 of 73 ~..J ~ FINAL p . 57 Appendix A Washington Works vs Region 1 All-Cause Mortality Surveillance Report : Totals (Males and Females ) Cause of Death Observed Expected 95% 95% 99% 99% SMR Lower Upper Lower Upper All Causes of Death 806 848 .5 95 .0 88 .5 101 .8 86 .6 104 . 0 Tuberculosis 0 0 .4 N/A 0 .0 865 .4 0.0 1242 .8 All Malignant Neoplasms 234 229 .2 102 .1 C ancer of Buccal Cavity & Pharynx 4 3 .4 117 .0 C ancer of Digestive Organs & Peri toneum 51 54 .1 94 .3 C ancer of Esophagus 4 4 .8 83 .1 C ancer of Stomach 3 5 .8 52 .1 C ancer of Large Intestine 17 21 .7 78 .3 Cancer of Rectum 5 3 .8 132 .1 Cancer of Bilia Passages & Liver 8 5 .5 144 .9 Cancer of Pancreas 11 11 .2 98 .2 Cancer of All Other Digestive Organs 3 1 .3 226 .7 89 .4 116 .0 85 .7 120 .6 31 .9 299 .6 19 .7 368 .4 70 .2 123 .9 63 .7 133 . 9 22 .6 212 .7 14 .0 261 .5 10 .7 152 .2 5 .9 190 .6 45 .6 125 .4 38 .0 141 .9 42 .9 308 .2 28 .5 373 .8 62 .6 285 .6 46 .6 336 .6 49 .0 175 .8 38 .6 203 .4 46 .8 662 .4 25 .5 829 .4 Cancer of Respiratory System Cancer of Larynx 72 83 .8 85 .9 3 1 .5 194 .6 67 .2 108 .2 62 .1 115 .6 40 .1 568 .6 21 .9 711 .9 Cancer of Bronchus, Trachea, Lung 66 80.3 82 .2 63 .6 104 .6 58 .5 112 .1 Cancer of All Other Respiratory 3 2 .0 151 .0 31 .2 441 .2 17 .0 552 . 5 Can cer of Breast 2 2 .8 70.4 8 .5 254 .3 3 .6 326 .4 All Ute rine Cancers (females only) 0 0.0 N/A 0 .0 17174 .1 0 .0 24664 .8 Can cer of Cervix Uteri (females only) 0 0.0 N/A 0 .0 20449.0 0 .0 29368 .1 Can cer of Other Female genital Organs 0 0.7 N/A 0 .0 516 .9 0 .0 742 .3 Can cer of Prostate (males only) Cancer of Testes and Other male genital 12 18 .4 65 .3 33 .8 114 .1 26 .9 131 .5 Organs 1 0 .6 169.7 4 .2 945 .7 0 .8 1261 .1 Cancer of Kidney 12 6 .6 181 .0 93 .5 316 .2 74.6 364 .2 C an cer of Bladder an d Other Urinary Organs 7 5 .4 130.3 52 .4 268 .5 37.9 318 .9 Malignant Melanoma of Skin 3 4 .4 67 .5 13 .9 197 .4 7 .6 247 .2 C ancer of Eye 0 0 .4 N/A 0 .0 867 .9 0.0 1246 .5 Cancer of Central Ne rvous System 9 7 .1 126 .5 57 .8 240 .2 44.0 281 .1 Cancer of Thyroid & Other Endocri ne Glands 3 0 .5 628 .6 * 129 .7 1836 .9 70.8 2300 .0 Cancer of Bone 2 0 .3 648 .3 Cancer of All Lymphatic, Haematopoieti c Tissue 32 24 .8 129 .1 78 .4 2342 .1 33 .4 3006 .3 88 .3 182 .3 77 .9 200 .4 Non-Hodgkins Lymphoma 9 8 .3 108 .2 49 .5 205 .4 37 .7 240 .5 Hodgkins Disease 2 1 .3 155 .0 18 .8 559 .9 8 .0 718 .7 Leukemia & Aleukemia 13 10 .7 121 .8 64 .8 208 .2 52 .3 238 .8 Cancer of AlI Other Lym ho oietic Tissue 8 4 .5 177 .8 All Other Malignant Neoplasms 24 15 .8 151 .6 76 .8 350 .3 57 .1 412 .8 97 .1 225 .5 83 .7 251 .0 57 of 73 FINAL p . 58 Appendix A Washington Works vs Region 1 All-Cause Mo rtality Surveillance Report : Totals (Males and Females) (continued) Cause of Death Observed Expected 95% 95% SMR Lower Upper 99% 99% Lower U er Beni Neoplasms 1 2 .4 41 .6 1 .0 231 .8 0.2 309.2 Diabetes Mellitus 22 11 .2 196 .9 123 .4 298 .1 105 .5 333 .1 Cerebrovascular Disease 35 40.6 86.2 60.1 119 .9 53 .3 131 .4 All Heart Disease 314 284.5 110.4 Rheumatic Hea rt Disease 5 1 .7 288.1 98 .5 123 .3 94 .9 127 .5 93 .5 672 .3 62 .1 815 .3 Ischemic Heart Disease Chronic Endocard . Dis . ; Other Myocard. Insuff. 11 239 218.2 109 .5 10.5 105 .1 96 .1 124 .4 92.1 129 .2 52 .5 188 .1 41 .3 217 .7 Hypertension with Heart Disease 1 6 .4 15 .6 * 0 .4 87 .1 0 .1 116 .1 All Other Heart Disease 58 47 .8 121 .4 Hypertension w/o Heart Disease 5 2 .4 205 .1 92 .2 156 .9 84 .3 168 . 8 66 .6 478 .5 44 .2 580 .3 Non-malignant Respiratory Disease 49 51 .5 95 .2 70 .4 125 .8 63 .8 136 .1 Influenza & Pneumonia 14 15 .7 89 .0 48 .7 149 .4 39 .6 170 .6 Bronchitis, Emphysema, Asthma 13 11 .4 113 .6 60 .5 194 .2 48 .7 222 . 7 Bronchitis 6 3 .9 155 .0 56 .9 337 .4 39 .7 404 .6 Emphysema 7 6 .8 102 .7 41 .3 211 .5 29 .9 251 .3 Asthma 0 0 .8 N/A 0 .0 485 .4 0 .0 697 . 1 Other Non-malignant Respiratory Disease 22 24 .3 90 .5 56 .7 137 .0 48 .5 153 .1 Ulcer of Stomach & Duodenum 0 1 .6 N/A 0 .0 235 .9 0.0 338 .8 Cirrhosis of Liver 9 9 .3 96 .5 Nephritis & Nephrosis 8 6 .2 129 .6 44 .1 183 .1 33 .6 214 .3 55 .9 255 .3 41 .6 300 .9 All External Causes of Death 45 68 .2 65 .9 48 .1 88 .2 43 .4 95 . 7 Accidents 35 42.0 83 .3 58 .0 115 .9 51 .5 126 .9 Motor Vehicle Accidents 23 24 .2 95 .0 60 .2 142 .5 51 .7 158 .9 All Other Accidents 12 17 .8 67 .4 34 .9 117 .8 27 .8 135 . 7 Suicides 8 19 .7 40 .7 17 .6 80 .1 13 .1 94 .4 Homicides & Other External Causes 2 6 .6 30 .5 3 .7 110 .3 1 .6 141 .5 All Other Causes of Death 83 CERTAIN INFECTIOUS AND PARASITI C DISEASES 1 Unknown Causes (In All Causes Catego ry Only) 0 74 .4 1 .1 111 .6 94 .6 (*) SIGNIFICANT AT 5% LEVEL; (** ) SIGNIFICANT AT I% LEVEL 88 .9 138 .3 82 .6 147 .2 2 .4 526 .9 0 .5 702 .6 58 of 73 FINAL p . 59 Appendix B Washington Works vs USA All-Cause Mortality Surveillance Report : Males Cause of Death Observed Expected 95% 95% SMR Lower Upper 99% Lower 99% Upper All Causes of Death 773 1167.0 66 .2 61 .6 71 .1 60.3 72 . 6 Tuberculosis 0 2 .0 N/A 0 .0 183 .7 0 .0 263 .9 All Malignant Neoplasms 222 301 .2 73 .7 64 .3 84 .1 61 .6 87 .5 C ancer of Buccal Cavity & Pharynx 4 7 .6 52 .9 14 .4 135 .5 8 .9 166 .6 C ancer of Digestive Organs & Peri toneum 49 74.1 66 .2 49 .0 87 .5 44 .4 94 .6 Cancer of Esophagus 4 9 .6 41 .5 11 .3 106 .4 7 .0 130 .8 Cancer of Stomach 2 9.8 20 .5 2 .5 74 .0 1 .1 95.0 Cancer of Large Intestine 17 24.5 69 .4 40.4 111 .1 33 .7 125 .7 Cancer of Rectum 5 5 .3 94 .7 30.7 220 .9 20 .4 267 .9 C ancer of Bilia Passages & Liver 7 7 .8 89 .7 36.1 184 .9 26 .1 219.6 C ancer of Panc reas 11 14 .9 74 .0 36 .9 132 .3 29 .1 153 .2 Cancer of All Other Digestive Organs 3 2 .2 135 .9 28 .0 397 .2 15 .3 497. 3 C ancer of Respiratory System Cancer of Larynx 70 110 .6 63 .3 3 3 .9 76 .7 49.3 80 .0 45 .5 85 .5 15 .8 224 .0 8 .6 280.5 Cancer of Bronchus, Trachea, Lung 64 105 .6 60.6 46.7 77 .4 42 .9 83 .0 Cancer of All Other Respiratory 3 1 .0 293 .1 60 .5 856 .5 33 .0 1072 . 4 Cancer of Breast 0 0 .4 N/A 0.0 952 .7 0 .0 1368 .3 Cancer of Prostate (males only) Cancer of Testes and other male Genital 12 23 .2 51 .8 * 26 .8 90 .5 21 .3 104.2 Organs 1 1 .2 86.9 2 .2 484 .0 0 .4 645 .4 Can cer of Kidney 12 7 .7 155 .7 80 .4 271 .9 64.1 313 .2 Can cer of Bladder an d Other Urinary Organs 7 6 .9 101 .4 40 .8 208 .9 29.5 248 .1 Malignant Melanoma of Skin 2 5 .1 39 .0 4 .7 140.7 2 .0 180 .7 C an cer of Eye 0 0 .2 N/A 0 .0 2299 .0 0 .0 3301 .8 C an cer of Central Nervous System 9 8 .6 105 .0 48 .0 199 .4 36 .5 233 .3 C an cer of Thyroid & Other Endocri ne Glands 3 0 .9 332 .2 68 .5 970.7 37 .4 1215 .3 C an cer of Bone Cancer of All Lymphatic, Haematopoietic Tissue 29 2 0 .8 251 .6 29.6 98 .0 30 .4 909 .0 13 .0 1166 .8 65 .6 140 .7 57 .4 155 .3 Hodgkins Disease 2 1 .9 103 .3 12 .5 373 .0 5 .3 478 .8 Non-Hodgkins Lymphoma Leukemia & Aleukemia 9 11 .3 80 .0 12 11 .1 107 .8 36 .6 151 .9 27 .8 177 .8 55 .7 188 .3 44 .4 216 .9 Cancer of All Other L ho oietic Tissue 6 5 .3 113 .8 41 .8 247 .8 29 .2 297 . 1 All Other Malignant Neoplasms 22 24 .5 89 .9 56.3 136 .1 48 .2 152 .1 59 of 73 ,~ _ ~"~ FINAL p . 60 Appendix B Washington Works vs USA All-Cause Mortality Surveillance Report : Males (continued) 95% 95% 99% 99% Cause of Death Observed Expected SMR Lower Upper Lower Upper Beni Neoplasms 1 2 .9 35 .1 0.9 195 .3 0 .2 260 .5 Diabetes Mellitus 20 24.6 81 .2 49 .6 125 .3 42 .0 140 .7 Cerebrovascular Disease 34 55 .8 60 .9 ** 42 .2 85 .1 37 .3 93 .3 All Heart Disease 309 386 .5 80.0 71 .3 89 .4 68 .7 92 . 5 Rheumatic Heart Disease 5 4 .0 125 .7 40.8 293 .4 27.1 355 .8 Ischemic Heart Disease 236 289 .8 81 .4 Chronic Endocard. Dis. ; Other Myocard. Insuff. 11 13 .1 83 .7 71 .4 92 .5 68 .4 96 .1 41 .8 149 .7 32 .9 173 .2 Hypertension with Hea rt Disease 1 11 .8 8 .5 ** 0 .2 47 .3 0 .0 63 .1 All Other Heart Disease 56 67 .8 82 .6 Hypertension w/o Hea rt Disease 5 4 .6 108 .3 62 .4 107 .3 57 .0 115 . 6 35 .2 252 .8 23 .4 306.6 Non-malignant Respiratory Disease 46 85 .4 53 .9 39 .4 71 .9 35 .6 77 .9 Influenza & Pneumonia 14 25 .8 54 .3 * 29 .7 91 .1 24 .2 104 .1 Bronchitis, Emphysema, Asthma 11 23 .3 47 .2 23 .6 84.5 18 .5 97 .7 Bronchitis 5 11 .2 44.7 14 .5 104 .4 9 .6 126 .6 Emphysema 6 10 .3 58.5 21 .5 127 .3 15 .0 152 .7 Asthma 0 1 .9 N/A 0.0 197 .5 0 .0 283 .7 Other Non-malignant Respiratory Disease 21 36 .3 57 .9 35 .8 88 .5 30 .5 99 .0 Ulcer of Stomach & Duodenum 0 3 .5 N/A 0 .0 105 .7 0 .0 151 .7 Cirrhosis of Liver 8 29 .0 27 .6 ** 11 .9 54 .4 8 .9 64 . l Nephritis & Nephrosis 8 10 .7 74 .7 32.3 147 .2 24 .0 173 .5 All External Causes of Death 41 124 .4 33 .0 ** 23 .7 44 .7 21 .2 48 . 7 Accidents 31 73 .9 42 .0 28 .5 59 .6 25 .1 65 .6 Motor Vehicle Accidents 20 36 .2 55 .2 33 .7 85 .2 28 .6 95 .7 All Other Accidents 11 37 .7 29.2 14 .6 52.3 11 .5 60 .5 Suicides 8 28 .5 28 .0 12 .1 55 .2 9 .0 65 .1 Homicides & Other External Causes 2 22 .0 9 .1 1 .1 32 .9 0 .5 42 .2 All Other Causes of Death 78 127 .4 61 .2 48 .4 Ce rtain Infectious and Parasitic Diseases 1 14 .3 7 .0 0 .2 Unknown Causes (In All Causes Category Only) 0 76 .4 44 .9 38 .9 0 .0 81 .5 51 .9 (*) SIGNIFICANT AT 5% LEVEL ; (**) SIGNIFICANT AT I/u LEVEL 60 of 73 ~~ p . 61 FINAL Appendix B Washington Works vs US A All-Cause Mortality Surveillance Report : Female s Cause of Death Observed Expected 95% 95% 99% 99% SMR Lower Upper Lower Upper All Causes of Death 33 40 .9 80 .7 55 .5 113 .3 49 .1 124 .4 Tuberculosis 0 0 .0 N/A 0 .0 8275 .0 0.0 11884 .3 All Malignant Neoplasms 12 13 .8 86 .6 44 .8 151 .4 35 .7 174 .3 C ancer of Buccal Cavity & Pharynx 0 0 .1 N/A 0 .0 2486 .0 0.0 3570 .3 C ancer of Digestive Organs & Peri toneum 2 2 .4 84 .0 10 .2 303 .4 4.3 389 .5 Cancer of Esophagus 0 0 .1 N/A 0 .0 3070 .1 0.0 4409 .1 Cancer of Stomach 1 0 .3 397 .1 9 .9 2212 .9 2 .0 2950 .8 Cancer of Large Intestine 0 0 .9 N/A 0 .0 388 .6 0 .0 558 .0 Cancer of Rectum 0 0 .2 N/A 0 .0 2150 .8 0 .0 3088 .9 Cancer of Bilia Passages & Liver 1 0 .3 394 .5 9 .9 2197 .9 2 .0 2930 .9 Cancer of Pancreas 0 0.5 N/A 0 .0 698 .4 0 .0 1003 .0 Cancer of All Other Digestive Organs 0 0.1 N/A 0 .0 3466 .8 0 .0 4978 .9 Can cer of Respiratory System 2 3 .0 67 .7 8 .2 244 .5 3 .5 313 .9 Can cer of Larynx 0 0.0 N/A 0 .0 8045 .8 0 .0 11555 .1 Can cer of Bronchus, Trachea, Lung 2 2.9 69 .5 8 .4 251 .0 3 .6 322 .2 Can cer of All Other Respiratory 0 0.0 N/A 0 .0 11919 .2 0 .0 17117. 9 Can cer of Breast 2 3 .3 61 .1 7 .4 220.6 3 .1 283 .1 All Uterine C ancers (Females only) 0 0.9 N/A 0 .0 419 .5 0 .0 602 .5 Can cer of Cerv ix Uteri (Females only) 0 0.6 N/A 0 .0 636 .4 0 .0 914 .0 C an cer of Other Female Genital Organs 0 0.9 N/A 0.0 416 .4 0 .0 598 .0 C an cer of Kidney Cancer of Bladder and Other Urinary Organs 0 0 .2 N/A 0.0 1793 .7 0 .0 2576 .0 0 0 .1 N/A 0.0 3722 .5 0 .0 5346 .1 Malignant Melanoma of Skin 1 0 .2 422 .2 10.6 2352 .3 2 .1 3136 .7 C ancer of Eye 0 0 .0 N/A 0.0 52400 .6 0 .0 75255 .7 Cancer of Central Ne rvous System Cancer of Thyroid & Other Endocrin e Glands 0 0 .4 N/A 0.0 834 .8 0 .0 1199 .0 0 0 .1 N/A 0.0 6334 .1 0 .0 9096 .8 C ancer of Bone 0 0 .0 N/A 0 .0 8833 .8 0.0 12686 .8 Cancer of All Lymphatic, Haematopoieti c Tissue 3 1 .2 245 .5 50 .6 717 .4 27.7 898 .2 Hodgkins Disease 0 0 .1 N/A 0 .0 3891 .4 0.0 5588 .6 Non-Hodgkins Lymphoma 0 0 .5 N/A 0 .0 811 .3 0 .0 1165 .2 Leukemia & Aleukemia 1 0 .5 207 .0 5 .2 1153 .4 1 .0 1538 .0 Cancer of All Other L ho oietic Tissue 2 0 .2 1057 .3 * 127 .9 3819 .5 54 .5 4902 .7 All Other Malignant Neoplasms 2 1 .0 197 .7 23 .9 714 .1 10 .2 916 .6 61 of 73 ~r~ FINAL p . 62 Appendix B Washington Works vs US A All-Cause Mortality Surveillance Report : Females (continued) Cause of Death Observed Expected 95% 95% 99% 99% SMR Lower Upper Lower Upper Benign Neoplasms 0 0 .1 N/A 0 .0 2467 .4 Diabetes Mellitus 2 1 .2 160 .8 19.5 580 .7 Cerebrovascular Disease 1 2 .1 48 .7 1 .2 271 .6 0 .0 3543 .6 8 .3 745 .4 0 .2 362 .2 All Heart Disease 5 7 .8 64 .4 20 .9 150 .2 13 .9 182 . 2 Rheumatic Heart Disease 0 0 .2 N/A 0 .0 1694 .6 0.0 2433 .7 Ischemic Heart Disease Chronic Endocard . Dis. ; Other Myocard . Insuff. 3 0 Hypertension with Heart Disease 0 4 .7 64 .0 13 .2 187 .1 7 .2 234 .3 0 .4 N/A 0 .0 1017 .5 0 .0 1461 .3 0 .4 N/A 0 .0 852 .1 0.0 1223 .7 All Other Heart Disease 2 2 .1 96 .7 11 .7 349 .4 5 .0 448 . 5 Hypertension w/o Heart Disease 0 0 .2 N/A 0 .0 2038 .1 0 .0 2927 .1 Non-malignant Respiratory Disease 3 2 .5 119 .6 24 .7 349.5 13 .5 437 .6 Influenza & Pneumonia 0 0 .7 N/A 0 .0 535 .7 0 .0 769.4 Bronchitis, Emphysema, Asthma 2 0 .9 228 .8 27 .7 826.4 11 .8 1060. 8 Bronchitis 1 0 .4 236 .2 5 .9 1315 .9 1 .2 1754 .7 Emphysema 1 0 .2 447 .2 11 .2 2491 .9 2 .2 3322 .9 Asthma 0 0 .2 N/A 0 .0 1623 .5 0 .0 2331 . 6 Other Non-malignant Respiratory Disease 1 0 .9 105 .7 2 .6 589 .2 0 .5 785 .6 Ulcer of Stomach & Duodenum 0 0 .1 N/A 0 .0 4846 .9 0 .0 6961 .0 Cirrhosis of Liver 1 0 .9 105 .5 2 .6 588 .0 0 .5 784 .1 Nephritis & Nephrosis 0 0 .4 N/A 0 .0 915 .7 0 .0 1315 .1 All External Causes of Death 4 5 .3 75 .5 20 .6 193 .2 12 .7 237 . 6 Accidents 4 3 .1 129 .6 35 .3 331 .7 21 .8 407 .9 Motor Vehicle Accidents 3 1 .9 156 .4 32 .3 457 .0 17 .6 572 .2 All Other Accidents 1 1 .2 85 .6 2 .1 476 .7 0 .4 635 . 7 Suicides 0 1 .1 N/A 0 .0 328 .2 0 .0 471 .3 Homicides & Other External Causes 0 1 .1 N/A 0 .0 338 .6 0.0 486 .3 All Other Causes of Death 5 5 .8 86 .1 27 .9 200 .8 18 .6 243 . 6 Cert ain Infectious and Parasitic Diseases 0 0 .7 N/A 0 .0 543 .5 0 .0 780 .6 Unknown Causes (In All Causes Category Only) 0 (*) SIGNIFICANT AT 5% LEVEL ; (**) SIGNIFICANT AT 1% LEVE L 62 of 73 ~i'~ FINAL p . 63 Appendix B Washington Works vs US A All-Cause Mortality Surveillance Report : Totals (Males and Females ) Cause of Death Observed Expected 95% 95% 99% 99% SMR Lower Upper Lower Upper All Causes of Death 806 1207 .9 66.7 ** 62.2 71 .5 60.8 73 .0 Tuberculosis 0 2 .1 N/A 0.0 179 .7 0.0 258 .1 All Malignant Neoplasms 234 315 .0 74 .3 65 .1 84 .4 62.4 87 .8 C an cer of Buccal Cavity & Pharynx 4 7 .7 51 .9 14.1 132 .9 8 .7 163 .4 C ancer of Digestive Organs & Peritoneum 51 76 .4 66.7 49.7 87 .7 45 .1 94 . 8 Cancer of Esophagus 4 9 .7 41 .0 11 .2 105 .1 6 .9 129 .2 Cancer of Stomach 3 10 .0 30.0 * 6.2 87 .6 3 .4 109 .6 Cancer of Large Intestine 17 25 .4 66.8 38 .9 107 .0 32 .4 121 .0 Cancer of Rectum 5 5 .5 91 .7 29.8 213 .9 19 .8 259 .5 Cancer of Bilia Passages & Liver 8 8 .1 99.3 42.9 195 .7 31 .9 230 .6 Cancer of Pancreas 11 15 .4 71 .4 35 .7 127 .8 28 .1 147 .9 Cancer of All Other Digestive Organs 3 2 .3 129.7 26.8 378 .9 14 .6 474 .5 C ancer of Respiratory System Cancer of Larynx 72 113 .5 63 .4 49.6 79 .9 45 .8 85 .3 3 4 .0 75 .8 15 .6 221 .4 8 .5 277 .2 C ancer of Bronchus, Trachea, Lung 66 108 .5 60.8 47 .0 77 .4 43 .3 82 .9 Cancer of All Other Respiratory 3 1 .1 284 .5 58 .7 831 .4 32 .1 1041 . 0 C ancer of Breast 2 3 .7 54 .6 6 .6 197 .2 2 .8 253 .2 All Uteri ne Cancers Females only) 0 0 .9 N/A 0 .0 419 .5 0 .0 602 .5 C ancer of Cervix Uteri (Females only) 0 0 .6 N/A 0 .0 636 .4 0 .0 914.0 C ancer of Other Female Genital Organs 0 0 .9 N/A 0 .0 416 .4 0 .0 598 .0 C ancer of Prostate ales only) Cancer of Testes and Other Male Genital Organs 12 23 .2 51 .8 * 26 .8 90.5 21 .3 104 .2 1 1 .2 86 .9 2 .2 484 .0 0 .4 645 .4 C ancer of Kidney Cancer of Bladder and Other Urinary Organs 12 7 .9 151 .6 78 .4 264.9 62 .5 305 .1 7 7 .0 99 .9 40 .2 205 .9 29 .1 244.6 Malignant Melanoma of Skin 3 5 .4 55 .9 11 .5 163 .2 6 .3 204 .4 Cancer of Eye 0 0.2 N/A 0 .0 2202 .4 0 .0 3163 .0 Can cer of Central Nerv ous System Cancer of Thyroid & Other Endocrin e Glands 9 9.0 99 .9 45 .7 189.6 34 .8 221 .9 3 1 .0 312 .0 64 .4 911 .9 35 .2 1141 .7 Cancer of Bone 2 0.8 239 .1 Cancer of All Lymphatic, Haematopoieti c Tissue 32 30 .8 103 .8 28 .9 863 .6 12 .3 1108 .5 71 .0 146 .6 62 .6 161 .1 Hodgkins Disease Non-Hodgkins Lymphoma 2 2 .0 98 .4 9 11 .7 76 .9 11 .9 355 .6 5 .1 456 .5 35 .2 146 .0 26 .8 170 .8 Leukemia & Aleukemia 13 11 .6 111 .9 59 .6 191 .4 48 .0 219 .5 Cancer of All Other L ho oietic Tissue 8 5 .5 146 .5 63 .3 288 .7 47 .1 340 . 3 All Other Malignant Neoplasms 24 25 .5 94 .2 60 .3 140 .1 52 .0 155 .9 63 of 73 FINAL p . 64 Appendix B Washington Works vs USA All-Cause Mo rtality Surveillance Report: Totals (Males and Females) (continued ) Cause of Death Observed Expected 95% 95% 99% 99% SMR Lower Upper Lower Upper Benign Neoplasms 1 3 .0 33 .3 0 .8 185 .6 0 .2 247.5 Diabetes Mellitus 22 25 .9 85 .0 53 .3 128 .7 45 .5 143 .8 Cerebrovascular Disease 35 57 .9 60.4 42 .1 84 .1 37 .4 92.1 All Heart Disease 314 Rheumatic Heart Disease 394 .2 79.6 5 4 .2 119.2 71 .1 89 .0 68 .5 92. 0 38 .7 278 .2 25 .7 337 .4 Ischemic Heart Disease Chronic Endocard. Dis. ; Other Myocard. Insuff. 239 294 .5 11 13 .5 81 .2 81 .4 71 .2 92 .1 68 .3 95 .7 40 .6 145 .7 32 .0 168 .6 Hypertension with Hea rt Disease 1 12 .2 8 .2 0 .2 45 .6 0 .0 60 .8 All Other Heart Disease 58 69 .8 83 .1 Hypertension w/o Hea rt Disease 5 4 .8 104 .2 63 .1 107 .4 57 .7 115 . 5 33 .8 243 .3 22 .5 295 .0 Non-malignant Respiratory Disease 49 87.9 55 .7 41 .2 73 .7 37 .4 79 .7 Influenza & Pneumonia 14 26.5 52 .9 * 28 .9 88 .7 23 .5 101 .4 Bronchitis, Emphysema, Asthma 13 24.2 53 .8 * 28 .6 91 .9 23 .1 105 . 4 Bronchitis 6 11 .6 51 .7 19 .0 112 .5 13 .2 135 .0 Emphysema 7 10.5 66 .8 26 .8 137.6 19 .4 163 .5 Asthma 0 2 .1 N/A 0 .0 176 .1 0 .0 252 . 9 Other Non-malignant Respiratory Disease 22 37 .2 59 .1 37 .0 89 .4 31 .7 99 .9 Ulcer of Stomach & Duodenum 0 3 .6 N/A 0 .0 103 .4 0.0 148 .5 Cirrhosis of Liver 9 29 .9 30 .1 Nephritis & Ne hrosis 8 11 .1 72 .0 13 .8 57 .1 10.5 66 .8 31 .1 141 .9 23 .1 167 .2 All Exte rnal Causes of Death 45 129 .7 34.7 25 .3 46 .4 22.8 50 .4 Accidents 35 77 .0 45 .5 31 .7 63 .2 28 .1 69 .3 Motor Vehicle Accidents 23 38 .2 60 .3 * 38 .2 90 .5 32.8 100 .9 All Other Accidents 12 38 .8 30 .9 16 .0 54 .0 12 .7 62 . 2 Suicides 8 29 .7 27 .0 11 .6 53 .2 8 .7 62 .6 Homicides & Other External Causes 2 23 .1 8 .7 1 .0 31 .3 0 .4 40 .2 All Other Causes of Death 83 133 .2 62 .3 ** 49 .6 Ce rt ain Infectious and Parasitic Diseases 1 15 .0 6 .7 ** 0 .2 Unknown Causes (In All Causes Catego ry Only) 0 77 .3 46 .1 82_ 2 37 .2 0 .0 49 .6 (*) SIGNIFICANT AT 5% LEVEL ; (** ) SIGNIFICANT AT 1% LEVEL 64 of 73 ~~'" _...~ FINAL Appendix C Washington Works vs West Virginia All-Cause Mortality Surveillance Report : Male s p . 65 Cause of Death Observed Expected 95% SMR Lower 95% Upper 99% 99% Lower Upper All Causes of Death 773 1331 .3 58 .1 54 .0 62.3 52 .8 63 . 7 Tuberculosis 0 2 .1 N/A 0 .0 176.8 0 .0 253 .9 All Malignant Neoplasms 222 325 .2 68 .3 59 .6 77.9 57 .0 81 .0 C an cer of Buccal Cavity & Pharynx 4 6 .4 62 .1 16 .9 159 .0 10 .4 195 .5 Cancer of Digestive Organs & Peritoneum 49 68.6 71 .5 * 52 .9 94.5 47 .9 102 . 2 Cancer of Esophagus 4 8 .4 47 .4 12 .9 121 .4 8 .0 149 .2 Cancer of Stomach 2 8.1 24.5 * 3 .0 88.7 1 .3 113 .8 Cancer of Large Intestine 17 23.9 71 .1 41 .4 113 .8 34 .5 128 .7 Cancer of Rectum 5 5.8 86.3 28 .0 201 .4 18 .6 244 .2 Cancer of Bilia Passages & Liver 7 6.7 104 .2 41 .9 214.6 30 .3 254 .9 Cancer of Pancreas 11 13.3 82.9 41 .4 148.3 32 .6 171 .6 Cancer of All Other Digestive Organs 3 2.3 131 .8 27 .2 385.2 14 .9 482 . 3 Can cer of Respiratory System Cancer of Larynx 70 136 .1 51 .4 3 4.5 67.2 40 .1 65.0 37 .0 69 .5 13 .9 196.3 7 .6 245 .8 Can cer of Bronchus, Trachea, Lung 64 130.7 49.0 37 .7 62.5 34.6 67 .0 Can cer of All Other Respiratory 3 0.9 319.1 65 .8 932 .6 36.0 1167 . 7 Can cer of Breast 0 0.4 N/A 0 .0 987 .4 0.0 1418 .0 Can cer of Prostate (Males only) 12 20.9 57 .5 29.7 100.4 23 .7 115 .6 Cancer of Testes and Other Male Genita l Organs I 1 .3 75 .7 1 .9 421 .9 0.4 562 .6 C an cer of Kidney Cancer of Bladder and Other Urinary Organs Malignant Melanoma of 12 7 .7 155 .2 80.2 271 .2 63 .9 312 .4 7 6 .7 104 .7 42 .1 215 .6 30.5 256 .2 Skin 2 5 .5 36 .4 4.4 131 .4 1 .9 168 .7 C an cer of Eye 0 0 .2 N/A 0.0 2230 .9 0 .0 3203 .9 C an cer of Central Ne rv ous System Cancer of Thyroid & Other Endocrin e Glands 9 8 .0 112 .0 51 .2 212 .5 39 .0 248 .8 3 1 .0 301 .1 62 .1 880 .0 33 .9 1101 .9 C ancer of Bone 2 0 .9 230 .9 Cancer of All Lymphatic, Haematopoietic Tissue 29 30 .3 95 .8 Hodgkins Disease 2 1 .9 107 .6 Non-Hodgkins Lymphoma 9 11 .2 80 .7 27 .9 833 .9 11 .9 1070.5 64 .2 137 .6 56 .2 151 .9 13 .0 388 .6 5 .5 498 .8 36 .9 153 .2 28 .1 179.3 Leukemia & Aleukemia 12 12 .0 100 .2 Cancer of All Other L ho oietic Tissue 6 5 .3 114 .1 All Other Malignant Neoplasms 22 31 .2 70 .5 51 .8 175 .1 41 .3 201 .6 41 .9 248 .3 29 .2 297 .8 44 .2 106 .7 37 .8 119 .2 65 of 73 FINAL p . 66 Appendix C Washington Works vs West Virginia All-Cause Mortality Surveillance Report : Males (continued) Cause of Death Observed Expected 95% 95% 99% 99% SMR Lower Upper Lower Upper Benign Neoplasms 1 3 .5 28 .3 0 .7 157 .5 0 .1 210 .1 Diabetes Mellitus 20 29 .8 67.0 41 .0 103 .5 34 .7 116 .2 Cerebrovascular Disease 34 56.6 60 .1 All Hea rt Disease 309 465 .8 66 .3 Rheumatic Heart Disease 5 4 .3 115 .7 41 .6 83 .9 36 .9 92 .1 59 .1 74.2 57 .0 76 . 7 37 .5 269 .9 24 .9 327 .4 Ischemic Heart Disease Chronic Endocard . Dis . ; Other Myocard. Insuff. 236 342 .1 69 .0 11 15 .7 70 .1 60 .5 78.4 58 .0 81 .4 35 .0 125 .5 27 .6 145 .2 Hypertension with Heart Disease 1 9 .7 10 .3 0 .3 57 .5 0 .1 76 .7 All Other Heart Disease 56 94 .0 59 .6 Hypertension w/o Heart Disease 5 4 .2 117 .8 45 .0 77.3 41 .1 83 . 3 38 .2 274 .8 25 .4 333 .3 Non-malignant Respiratory Disease Influenza & Pneumonia 46 116 .3 39 .6 ** 29 .0 52.8 26 .2 57 .2 14 27 .5 51 .0 27 .9 85 .5 22 .7 97 .7 Bronchitis, Emphysema, Asthma 11 29 .5 37 .3 18 .6 66 .7 14 .6 77 . 1 Bronchitis 5 16 .8 29 .8 9 .7 69 .5 6 .4 84 .3 Emphysema Asthma 6 11 .1 54 .1 19 .9 117 .8 13 .9 141 .2 0 1 .6 N/A 0 .0 224 .3 0 .0 322 . 1 Other Non-malignant Respiratory Disease 21 59 .2 35 .4 21 .9 54 .2 18 .7 60.7 Ulcer of Stomach & Duodenum 0 3 .5 N/A 0 .0 106 .0 0 .0 152 .3 Cirrhosis of Liver 8 27 .5 29 .1 12 .6 57 .4 9 .4 67 .6 Nephritis & Nephrosis 8 13 .5 59 .2 25 .6 116 .7 19 .0 137 .5 All Exte rn al Causes of Death 41 146 .0 28 .1 20.2 38 .1 18 .1 41 . 5 Accidents 31 94 .6 32 .8 22 .3 46 .5 19 .6 51 .2 Motor Vehicle Accidents 20 45 .4 44 .0 ** 26.9 68 .0 22 .8 76 .3 All Other Accidents 11 49 .1 22 .4 ** 11 .2 40 .1 8 .8 46 . 4 Suicides 8 33 .1 24 .2 10 .4 47 .6 7 .8 56 .1 Homicides & Other External Causes 2 18 .3 All Other Causes of Death 78 138 .1 Certain Infectious and Parasitic Diseases 1 3 .2 10 .9 1 .3 39 .4 0 .6 50.6 56 .5 ** 44 .6 70 .5 41 .4 75 .1 30 .9 0 .8 172 .1 0.2 229 .5 Unknown Causes (In All Causes Category Only) 0 (*) SIGNIFICANT AT 5% LEVEL; (** ) SIGNIFICANT AT 1% LEVEL 66 of 73 ~ ' . FINAL p . 67 Appendix C Washington Works vs West Virginia All-Cause Mo rtality Surveillance Report: Females Cause of Death Observed Expected 95% SMR Lower 95% Upper 99% Lower 99% Upper All Causes of Death 33 45 .0 73 .4 50 .5 103 .1 44 .7 113 .2 Tuberculosis 0 0 .0 N/A 0 .0 9008 .5 0 .0 12937 .7 All Malignant Neoplasms 1 2 15 .1 79 .4 41 .0 138 .7 32 .7 159 .8 C ancer of Buccal Cavity & Pharynx 0 0 .1 N/A 0 .0 3027 .2 0 .0 4347 .6 C ancer of Digestive Organs & Pe ritoneum 2 2 .3 87 .0 10 .5 314 .1 4 .5 403 .2 Cancer of Esophagus 0 0 .1 N/A 0 .0 4248 .5 0 .0 6101 .6 Cancer of Stomach 1 0 .2 551 .8 13 .8 3074 .9 2 .8 4100 .2 Cancer of Large Intestine 0 1 .0 N/A 0 .0 356 .1 0 .0 511 .4 Cancer of Rectum 0 0 .2 N/A 0 .0 1992 .8 0 .0 2861 .9 Cancer of Bilia Passages & Liver 1 0 .2 441 .4 11 .0 2459 .7 2 .2 3279 .9 Cancer of P ancreas 0 0 .5 N/A 0 .0 778 .4 0 .0 1117 .9 Cancer of All Other Digestive Organs 0 0 .1 N/A 0 .0 3346 .9 0 .0 4806 .8 C ancer of Respiratory System 2 3 .6 55 .2 6 .7 199 .3 2 .8 255 .8 Cancer of Larynx 0 0 .1 N/A 0 .0 5906 .2 0 .0 8482 .2 Cancer of Bronchus, Trachea, Lung 2 3 .5 56 .6 6 .8 204 .3 2 .9 262 .3 Can cer of All Other Respiratory 0 0 .0 N/A 0 .0 13298 .5 0 .0 19098 .8 C an cer of Breast 2 3 .1 63 .5 7 .7 229 .5 3 .3 294 .6 All Uterine Can cers (Females only) 0 1 .2 N/A 0 .0 319 .3 0 .0 458 .6 Can cer of Cerv ix Uteri (Females only) 0 0 .8 N/A 0 .0 442 .8 0 .0 635 .9 Can cer of Other Female Genital Organs 0 0 .9 N/A 0 .0 397 .2 0 .0 570 .4 Can cer of Kidney Can cer of Bladder and Other Urina ry Organs 0 0 .2 N/A 0.0 1690 .2 0 .0 2427 .4 0 0 .1 N/A 0 .0 2837 .0 0 .0 4074 .4 Malignant Melanoma of Can cer of Eye Cancer of Central Nervous System Cancer of Thyroid & Other Endoc ri n e Glands Skin 1 0 .3 338 .9 8.5 1888 .5 0 0 .0 N/A 0 .0 48475 .7 0 0 .5 N/A 0.0 764 .4 0 0 .1 N/A 0.0 6825 .2 1 .7 2518 .2 0 .0 69618 .9 0 .0 1097 .9 0 .0 9802 .0 Cancer of Bone C an cer of All Lymphatic, Haematopoietic Tissue 3 0 0 .0 N/A 0.0 7639 .3 0.0 10971 .2 1 .3 235 .1 48 .5 687 .2 26.5 860 .4 Hodgkins Disease 0 0 .1 N/A 0.0 3287 .9 0.0 4721 .9 Non-Hodgkins Lymphoma 0 0 .5 N/A 0 .0 804 .2 0 .0 1155 .0 Leukemia & Aleukemia 1 0 .5 187 .6 4 .7 1045 .3 0 .9 1393 .8 Cancer of All Other L ho oietic Tissue 2 0.2 1165 .6 " 141 .0 4210.9 60 .0 5405 . 1 All Other Malignant Neoplasms 2 1 .3 151 .5 18 .3 547 .4 7 .8 702 .6 67 of 73 FINAL p . 68 Appendix C Washington Works vs West Virgini a All-Cause Mortality Surveillance Report : Females (continued ) Cause of Death Observed Expected SMR 95% 95% 99% 99% Lower Upper Lower Upper Benign Neoplasms 0 0 .2 N/A 0.0 2200 .4 0 .0 3160 .2 Diabetes Mellitus 2 1 .6 121 .7 14 .7 439 .6 6 .3 564 .2 Cerebrovascular Disease 1 2 .0 49 .7 1 .2 277 .2 0 .2 369 .6 All Heart Disease 5 9 .8 51 .1 16.6 119 .3 11 .0 144 . 6 Rheumatic Heart Disease 0 0 .3 N/A 0 .0 1272 .7 0.0 1827 .8 Ischemic Heart Disease 3 6 .0 49 .7 10 .3 145 .3 Chronic Endocard. Dis. ; Other Myocard. Insuff. 0 0 .5 N/A 0 .0 792 .8 5 .6 182 .0 0 .0 1138 .5 Hypertension with Hea rt Disease 0 0 .3 N/A 0 .0 1239 .2 0.0 1779 .7 All Other Heart Disease 2 2 .7 74 .1 9 .0 267 .7 3 .8 343 . 6 Hypertension w/o Hea rt Disease 0 0 .2 N/A 0 .0 1940 .3 0.0 2786 .5 Non-malignant Respiratory Disease 3 3 .1 95 .7 19 .7 279 .7 10.8 350 .2 Influenza & Pneumonia 0 0 .7 N/A 0 .0 526 .5 0 .0 756 .2 Bronchitis, Emphysema, Asthma 2 1 .2 173 .9 21 .0 628 .2 9 .0 806 . 3 Bronchitis 1 0 .7 146 .8 3 .7 817 .7 0.7 1090 .4 Emphysema 1 0 .3 383 .1 9 .6 2134 .5 1 .9 2846 .2 Asthma 0 0 .2 N/A 0 .0 1776 .0 0 .0 2550 . 7 Other Non-malignant Respiratory Disease 1 1 .3 77 .9 1 .9 434.0 0 .4 578 .7 Ulcer of Stomach & Duodenum 0 0 .1 N/A 0 .0 5737 .2 0 .0 8239 .5 Cirrhosis of Liver 1 0 .7 150 .1 3 .8 836 .6 0 .8 1115 .5 Nephritis & Nephrosis 0 0 .5 N/A 0 .0 806 .0 0 .0 1157 .5 All Exte rnal Causes of Death 4 5 .8 69 .4 18 .9 177 .8 11 .7 218 . 6 Accidents 4 3 .6 109 .9 30 .0 281 .5 18 .5 346 .1 Motor Vehicle Accidents 3 2 .5 118 .0 24 .3 344.7 13 .3 431 .7 All Other Accidents 1 1 .1 91 .3 2 .3 508 .6 0 .5 678 . 1 Suicides 0 1 .1 N/A 0 .0 340.4 0 .0 488 .9 Homicides & Other Exte rnal Causes 0 1 .0 N/A 0 .0 355 .2 0 .0 510 .1 All Other Causes of Death 5 5 .9 85 .0 27 .6 198 .5 18 .3 240 .7 CERTAIN INFECTIOUS AND PARASITIC DISEASES 0 0 .1 N/A 0 .0 2506.3 0 .0 3599 .4 Unknown Causes (In All Causes Category Only) 0 (*) SIGNIFICANT AT 5% LEVEL; (**) SIGNIFICANT AT 1/u LEVE L 68 of 73 =~ ~ ., . FINAL p . 69 Appendix C Washington Works vs West Virginia All-Cause Mortality Surveillance Report : Totals (Males and Females ) Cause of Death Observed Expected 95% 95% 99% 99% SMR Lower Upper Lower Upper All Causes of Death 806 1376.3 58 .6 54 .6 62 .8 53 .4 64 . 1 Tuberculosis 0 2 .1 N/A 0.0 173 .4 0 .0 249 .0 All Malignant Neoplasms 234 340 .3 68 .8 60 .2 78 .2 57 .7 81 .2 C ancer of Buccal Cavity & Pharynx Cancer of Digestive Organs & Peritoneum 4 6 .6 60.9 16 .6 156 .0 10 .2 191 .8 51 70.9 72 .0 * 53 .6 94 .6 48 .7 102 .2 Cancer of Esophagus 4 8 .5 46.9 12 .8 120 .1 7 .9 147 .7 Cancer of Stomach 3 8 .3 36 .0 7 .4 105 .3 4 .1 131 .8 Cancer of Large Intestine 17 25 .0 68 .1 39 .7 109 .1 33 .1 123 .4 Can cer of Rectum 5 6.0 83 .6 27 .1 195 .1 18 .0 236 .7 Cancer of Bilia Passages & Liver 8 6.9 115 .2 49 .7 226 .9 37 .0 267 .4 Cancer of P ancreas 11 13 .7 80.0 39 .9 143 .2 31 .4 165 .7 Cancer of All Other Digestive Organs 3 2 .4 125 .7 25 .9 367 .4 14 .2 460 .0 Cancer of Respiratory System Cancer of Larynx 72 139 .7 51 .5 3 4 .5 66 .2 40 .3 64.9 37 .2 69.3 13 .7 193 .6 7 .5 242 .4 Cancer of Bronchus, Trachea, Lung 66 134 .2 49 .2 38 .0 62.5 35 .0 67 .0 Cancer of All Other Respiratory 3 1 .0 310 .0 64 .0 905 .9 34 .9 1134. 2 Cancer of Breast 2 3 .5 56 .8 6 .9 205 .1 2 .9 263 .3 All Ute rine Cancers ( Females only) 0 1 .2 N/A 0 .0 319.3 0 .0 458 .6 Can cer of Cerv ix Uteri (Females only) 0 0 .8 N/A 0 .0 442 .8 0 .0 635 .9 Can cer of Other Female Genital Organs 0 0 .9 N/A 0 .0 397.2 0 .0 570.4 Can cer of Prostate (Males only) Cancer of Testes and Other Male Genita l Organs 12 20 .9 57 .5 29 .7 100.4 23 .7 115 .6 1 1 .3 75 .7 1 .9 421 .9 0 .4 562 .6 Can cer of Kidney Cancer of Bladder and Other Urinary Organs 12 7 .9 151 .0 78 .0 263 .7 62 .2 303 .8 7 6 .8 102 .7 41 .3 211 .5 29 .9 251 .3 Malignant Melanoma of Skin 3 5 .8 51 .8 10 .7 151 .4 5 .8 189 .5 C an cer of Eye 0 0 .2 N/A 0 .0 2132 .7 0 .0 3063 .0 Cancer of Central Ne rvous System 9 8 .5 Cancer of Thyroid & Other Endocrin e Glands 3 1 .1 C an cer of Bone 2 0 .9 Cancer of All Lymphatic , Haemato oietic Tissue 32 31 .5 105 .6 285 .6 218 .7 101 .5 48 .3 200 .5 36 .8 234 .7 58 .9 834 .7 32 .2 1045 .2 26 .5 789 .9 11 .3 1013 .9 69 .4 143 .3 61 .2 157 .5 Hodgkins Disease Non-Hodgkins Lymphoma 2 2 .0 101 .4 9 11 .6 77 .5 12 .3 366 .5 5 .2 470 .4 35 .4 147 .1 27 .0 172 .2 Leukemia & Aleukemia 13 12 .5 103 .9 Cancer of All Other Lymphopoieti c Tissue 8 5 .4 147 .3 All Other Malignant Neoplasms 24 32 .5 73 .8 55 .3 177 .7 44 .6 203 .9 63 .6 290 .3 47 .3 342 .1 47.3 109 .8 40 .7 122 .2 69 of 73 FINAL p . 70 Appendix C Washington Works vs West Virgini a All-Cause Mortality Surveillance Report : Totals (Males and Females) (continued) Cause of Death Observed Expected 95% 95% 99% 99% SMR Lower Upper Lower Upper Benign Neoplasms 1 3 .7 27 .0 0.7 150 .4 0.1 200 .6 Diabetes Mellitus 22 31 .5 69 .9 43 .8 105 .8 37 .5 118 .2 Cerebrovascular Disease 35 58 .6 59 .7 41 .6 83 .1 36 .9 91 .0 All Heart Disease 314 475 .6 66.0 Rheumatic Heart Disease 5 4 .6 108 .4 58 .9 73 .7 56.8 76 .3 35 .2 253 .0 23 .4 306 .8 Ischemic Hea rt Disease Chronic Endocard. Dis. ; Other Myocard. Insuff. 11 239 348 .1 16 .1 68 .7 68 .1 60.2 77 .9 57 .7 81 .0 34.0 121 .9 26.8 141 .1 Hypertension with Hea rt Disease 1 10 .0 10.0 ** 0 .3 55 .8 0.1 74 .4 All Other Heart Disease 58 96 .7 60.0 45 .5 77 .5 41 .6 83 .4 Hypertension w/o Heart Disease 5 4 .4 112 .7 36.6 263 .0 24.3 319 .0 Non-malignant Respiratory Disease 49 119 .4 41 .0 30 .4 54 .3 27.5 58 .7 Influenza & Pneumonia 14 28 .2 49 .7 27 .2 83 .4 22.1 95 .2 Bronchitis, Emphysema, Asthma 13 30 .7 42 .4 22 .6 72 .5 18.2 83 . 1 Bronchitis 6 17 .5 34 .3 12 .6 74 .7 8.8 89 .6 Emphysema 7 11 .4 61 .7 24 .8 127 .1 17.9 150 .9 Asthma 0 1 .9 N/A 0 .0 199 .1 0.0 286 .0 Other Non-malignant Respiratory Disease 22 60 .5 36 .3 ** 22 .8 55 .0 19 .5 61 .5 Ulcer of Stomach & Duodenum 0 3 .5 N/A 0 .0 104 .1 0.0 149 .5 Cirrhosis of Liver 9 28 .1 32 .0 Nephritis & Ne hrosis 8 14 .0 57 .3 14 .6 60 .7 11 .1 71 .1 24 .7 112 .8 18.4 133 .0 All External Causes of Death 45 151 .8 29 .7 ** 21 .6 39 .7 19 .5 43 . 1 Accidents 35 98 .2 35 .6 24 .8 49 .6 22 .0 54 .3 Motor Vehicle Accidents 23 48 .0 47 .9 30 .4 71 .9 26 .1 80.2 All Other Accidents 12 50.2 23 .9 ** 12 .3 41 .7 9 .8 48 . 1 Suicides 8 34.2 23 .4 10 .1 46 .1 7 .5 54 .3 Homicides & Other External Causes 2 19.4 10 .3 1 .2 37 .3 0 .5 47 .9 All Other Causes of Death 83 CERTAIN INFECTIOUS AND PARASITI C DISEASES 1 Unknown Causes (In All Causes Category Only) 0 144 .0 3 .4 57 .6 ** 45 .9 71 .5 42 .6 76 .0 29 .5 0 .7 164 .6 0 .1 219 .5 (*) SIGNIFICANT AT 5% LEVEL; (** ) SIGNIFICANT AT 1% LEVEL 70 of 73 ri i p . 71 FINAL Appendix D Job Exposure Category Development based on Division and Job TECHNICAL PROJECT COORD 0.008 0 .008 0 .008 I TEFLON@ POLYMERS PROD. SYSTEMS ANALYST 0.024 0 .024 0.024 ---- - RESEARCH HR SPEC 0.025 0 .025 0.025 1 S 1 t 1 1 T EFLON POLYMERS PROD. BUSINESS ANALYST 0.034 0 .034 0.034 1 1 1 MANUFACTURING 0.037 0 .037 0.037 1 RESEARCH DIVISION CHEMIST 0.041 0.04 0.043 2 2 1 TEFLON POLYMERS PROD. RES SUPERVISOR 0.056 0 .056 0.056 1 1 _.P_O_L. Y ENG DESIGN 0.067 0 .012 0.139 2 2 CONT ADM/BUS SVC/SAFETY .~ '-- 0.072 0 .026 0.149 8 2 t0 :~` RESEARCH 4423 NL ANALYST 0.072 0 .036 0.605 6 4 1 it 1 RESEARCH 4421 LAB ANALYST 0.073 0 .073 0.073 1 1 1 TEFLON@ POLYMERS PROD. ENGINEER 0.084 0 .037 0.131 1 I 2 .084 0 .023 0.377 14 4 1SPECIALTYOMUNDPR0 TEFLON@ POLYMERS PROD. ADMIN ASSISTANT 0.089 0.051 0.127 1 1 BUSINESS SERVICES 0.092 0.007 0.183 11 8 19 TEFLON POLYMERS PROD. PROJECT ENGINEER 0.093 0 .049 0.136 1 I 2 1 RESEARCH STAFF ENGINEER 0 .095 0.026 0.462 2 1 TEFLON@ COPOLYMER S PROD. PROD'N COORDINATOR 0.1 0 . 1 0.1 1 t 3 1 L HUMAN RESOURCES 0.103 0.033 0.173 I I _2 TEFLON POLYMERS PROD. TEFLON@ COPOLYMERS PROD. -~_ MASTER SCHEDULER 0.104 0 .104 0.104 I ENGINEER 0.105 0 .083 0.118 l 2 1 1 3 1 TECHNICAL AREA SPECIALIST 0.106 0 .106 0.106 1 I 1 B[JTACITE PRODUCTION 0 .107 0 .019 0.23 23 2 1 RESEARCH RESEARCH .. .. . - .-.. -. 0.116___ 0,1_16 0.116 l . -~- --- --- ~-- _TECHNOLOGYLYST MANAGER 0.118 0 .031 0.204 ~-_ I 1 _1- 2 TECHNI CAL 't---- AREA SUPT TECH 0.119 1 0 .119 0.119 I RESEARCH ADMIN ASSISTANT 3 0.11 9 0 .076 0.171 ! I-~ 3 E -T_F_L_O_N@ . _ _ _ ---- POLYMERS PROD-. ~.---- SPECIAL _ -- - ASSIGNMENT -- 0-~---_ .12 0.12 0_.1 -12 -- -I~ E-----~- E-.-P-. C`O-M-P-O-U-N-D-IN-G-P-R-O-D..._--'-t--------. ._... 0 .12 0 .026 0.652 21 11 I .-------------~ - --- --^------~'-------~----- {----_- RESEARCH 4422 LE ANALYST 0.127 0 .127 0.127 I 1 1 -1-- 4 1 13 T-E - C . . H . N --IC - A . ._--L -_ J PR.O_ ..D'N COORDINATOR 0... .128 0 .128 0.128 . --.---- ._--- - - _ ~ . - - - . ..... --B ....... ....-U --- -T ----~A ----C ....-----I--T -----E - . . ._ ..~ - ---a ------ -M ------A ---`--I ---N ._ ..--T ----E ----N -- A .._. NCE !+ ~ 1 , - TECHNICAL DIVISION ENGINEER 0.131 0 .107 0 .154. ..- 2 jTE -F~ LO~ N@ ~ C~ OP- OL- Y- ME+ RS -~ -----#~ - --- --- ~-- ------ ^-`--~ - -!- -~ - --- --~--- 2 - 1 PROD . SPECIAL ASSIUNMENT 0..I., 31 0 .032 0 .297 2 I 3 .. . + .- - --t ---"-- --- --'- - --- - ' - ------+ -- -__ 1 R . ._ .E ... SEARCH 3 SR ENGINEER 0.134 0 .081 2.07 ~ 1 2 4 .------ --- --. . TEFLON@ COPOLYMER . l SENIOSPROD R SPECIALIST -- - --- . _ .-- -- 0.136 --r 0 .0...3.-2~0-.2-39 2 t - r-----~---~-~--- --- - _ F_ ILEANMT PRODUCT_ ION _ _ _ _ 0.136 0 .0 05 # 0 .657 1 - 8 25 I f _. ~ 44 '4 8 ~-- - - - ~----- SHE&EA .. - . - _ - ....-'-- --~0.13-7'+- --0 .-01-8-~-0 .- -27-9-----3 --~'------ '-_----~.. .__- .-~'~ -1 . TE. .F. L.....O. ...N . ..(. ~ POLYME. R.. S PROD STAFF BUS ANALYST 0.138 . - 0 .138 { 0 .13 8 RESEARCH ~ SPECI.A _-_.LIST 0.138 0.038 I 0 .261 i 2 -- -~ - 3 1 SPECIALTY COMPOUN D MAINT 013 ' 9 0067 0 .2 11 1 1 2 r ZYT..EL. a PRODUC~ TION - ~ _--- 0.14- -0.~ 006 0 .746'- 2' 2- 2 - 2- 3 -4 - 7 - 1 ACRYLICS 0 .145 0.052 0.481 6 4 l0 .1 71 of 73 FINAL p . 72 7ECHNICAL SR TECH ASSOC 0 1 47 i 0. 146 ~ 0.148 -'- -j-- -- -t--'-'--._'.----- 4 2 2 TEFLONa POLYMERS PROD . MFG SERVICE REP. . . 0. .15 . . . -._.__ . .. . . - . _ TE_ CH_ NIC _AL SRCHEMIS._T ._ .. .._ . .._0.153 1 0 .15 0.15 . a _._ f f 0..105235 0 .153__I__ - - 1 _ } _ _1 .. - 1_ ---.._._. . TEFLON , POLYMERS PROD . SPECIALIST 0.153 0 .272 1 4 5 1 1 1 TEFLON(a) POLYMERS PROD . PLA- NT S- UPT TEFLON@ COPOLYMERS 0.157 0 .157 - ~ ~ PROD.-S- TAF -F ENGINEER E [ 0 .159 ....0.. .0...1.7 0 .15 7 -- 1 E ! 1 r -'i -0.3-5 I -- ' 2'_ . - _ ~ ttt 3 1 1 TEFLO- N~a POLYME- RS PROD_ . DI V. ISION E Y NGINEER 0 .161 0 .078 0 .242 _ '. 1 . 3 1 RESEARCH TECH ASSOC 0 .164 0 .028 0 .426 3 6_ 9 1 [ FILAMENT MAINTENANCE RESEARCH _____j ARIAS_.____. .-----..- 167 0 .104 0 .213 4 4 - - K44- 0.17 0 .104 0.191 3 I_ ! ( 3 1 TE CHNICAL SENIOR TECHNICIAN 0 .171 0 .053 0.377 --' '---_. .__-_--_ -- ---1----- ----- - 1 4 5 1 POLY ENG CONSTRUCTION 0 .172 0 .172 0.I72 1 --~ I _-~ 7`~ .275 1 3 t ttt~--- 4 1 RESEARCH TECHNICIAN 0 .181 0.042 0 TEFLON@ COPOLYMER S PROD, ARE A SUPT PROD 0 .184 ------_-__._--- ------'- - '-- -- 0.184 ._0_ . ._--'_1 t. -----...8 - ~- ~4 --_ - -- .. .. . ......1 .... ... ... .:. . ...----I -. 1 RESEARCH STORES COORDINATOR 0 .184 0.184 0.184 I- l 1 TEFLON@ COPOLYMER S PROD . PROCESS --~ ----- ., t -._..- D- ESIG -N -E- R_ 0..1 -. 9 _ 0 . .181 0.198 2 2 1 ----- . ----- -- --T- - --- DELRIN@ MAINTENANCE 0 .194 0.078 0 .43 2 7 ...~ .`~ _., . -- '---'-'- - ---- - '--- --- -"- - ' ---I-,--- ZYTEL@ MAINTENANCE jA 0.194 0.069 0 .43 1 3 TECHNICAL ADM IN ASSISTANT 0.197 0.054 0 .34 1 t TEFLON@ COPOLYMER S PROD . SENIOR TECHNICIAN 0.198 0. 1 98 0.198 1 ; 1_1 ~1 RES EARCH__ 4420 LABORATORIAN 0 .201 _ 0.055 W 1 .38 3 t- 37 5 -~-{-----------' - DELRIN@ PRODUCTION 0.203 0.044 0 .457 5 ! 2_ i - 4 7 . ---- 31 36_ B&ESMAINTENANCE ~~- 0.209 0.063 0 .464 3--L 1 0 TECH NICAL ---- __ C ...-E--R --T - -C.O _ORD.I....N . A . _T ... .O-R --0 .2-19._ ._OWER& SERVICES 0.221 0 ..I..8.. 5. 0 .~ 252 .--- - - -~ 2 -- -' -~ -- ... . ~ 0 .2 21 0 .22 1 . - - - -{( - - --- -- - - - - . TECHNICAL 4420 LABORATORIAN 1 0 .225 0.07 0 .352 3 1 2 E . P. COMPOUND `: - MAINTENANCE 0 .234 0 .107 0 .476 3 - _,~------- - TEFLON POLYMERS PROD . SR - -- ~--- . -............---.__ P.R. ..O --I_E-C-T--S-U--P-V--R-_. . ;-~--- 0.....--2-4.........._0t._ 24,1 ... _0. ..2.4 ---~----- -.. - EM_.P .__ ..L ...----O - ._ ..Y _-E _-E -. . ....R ..... E.. L ._'A ---T ---I--O . ._ .N ... . S. a.. 0.24. .` 0 ...._ .106.-..0._ ...3-8-- - ._ ! 3 TEFLON@ MAINTENANCE AREA SUPT MAINT 0 .242 - 0 .242 ~ 0 .242 ~r--1 --- ( -~---__ ` RESEARCH TECH FELLOW 0 .244 0 .174 0.31 4 2 3 1~ .. . .!I( . .. ....._~ . - .. ._. .- ..-----. ..f ; t z 1 1 2 POWER & SERVICES RESEARCH SR CHEMIST 0 .245 0 .045 0.963 f .6 __1 _- %4 _~-2 ---' --- 2_ 2.---- 1 0 .246 0 . . . y- . .246( 0.246, .~.~ ~ - ~ ., .. . ~ I l. ._... 1 . RESEA.R ..- CH DIV.ISI, O. N ENG - INEER .... 0247 j 0_12 5 0,581 2 1 __ 3 1 T.EFL.O._N..PO-L_Y_ MERS PROD. DIVISION CHEMIST 0.248 __.__ 0.248 0.248 -1 I 4 _ . _ ~_ L 1 --E-- _ TECHN_I_C- .A._L_ - --- . ..._ .__.-__t-_S..P._E_C-I_A.L_ AS-SIG_ NM-ENT ( TEFLON@ COPOL Y ME R S . 0__.-2....4_.9~0-.2.4_9- 3i _0_.. -2 . _4 _ -9 -, , _I . _ _) _ ...... ....+ .. ._... ., . _,1 ..~.._._... 4 PROD . AR _ E. fA SPECIALIST .. . 0.255 0.255 0 .255 - . f . f RESEARCH ? SUPERV ISOR 0.258 0.159 0,357 TEFLON@ COPOLYMER S .. - _ 2 PR_ OD . TEFLO~N@ COPOLYMER S SRCHEMIST PROD . ~ SPECIALIST -- ----- .. _--- - TEFLON@ COPOLYMERS ~ 0.265 0 .171 0 .359 0.273 0 .134 128 4_ _ .... . I _ LL . . . 2 5 , P. ROD . AREA SUPT . _ 0...2_79._ __ ,:_.0_-.._1..3_~7._0.. .369 3 . 3 1 2 2 2 2 TEFLON@ POLYMERS PROD . TECH FELLOW 0.282 0 .282 0 .282 I 1 ..~ : _~. ....._.. ..-.1.._. . 2 TE.C...H.. NICAL .. ........ . . . . TEFLON@ COPOLYMER S , _ PR_OD...._. ._ _ _ .. _.. . .__ a SR. ..EN . GINEER 0.282 0 .097 0 ... 316 1~2 _ . .L. RES.EN,_ G.IN .EER, -__ 0~ _286 - , ,,,w0 ~ .133 0. ..4-,3.9. ~ ~ 3 2 2 2 TEFL-O-.N_ a PO_ LYM._E..R..S_PR-O.D.- .. . . . TECH ASSOC .1- - -----~ ~ 0.288 - - - ;-0 .1`31 . .k 0 .471 .__ ._- ---'r -- - __ 6 ~-- -~-` 2 TECHNICAL TE.. C . . . . HNIC . . IAN . . 0 .289 .0. ..099 0 .5. 62 I 1 . ~ - - , 1---- 3 ....... 2 RESEARCH MFG SUPT 0 .292 0?92 ~~ _0 , .~. ^92 ~ I ~ 1 1 2 72 of 73 ~3 FINAL p . 73 TEFLON@ COPOLYMER S PROD SR TECH ASSOC 0 .292 --_-_._----------- ----'------------ 1 0.197 0 .387 2 2 2 RESEARCH AREA SUPT TECH 0 .296 0.296 0 .296 TEFLON@ COPOLYMERS PROD . SUPERVISOR 0 .307 0.123 2 .39 I_ 6 2 3 1 2 11 2 RESEARCH SENIOR TECHNICIAN 0 .31 0 .31 0.31 TEFLON@ COPOLYMERS { ! PROD . - -- DIVISION ENGINEER 0 .33 0.33 0.33 -- - - TEFLON@ POLYMERS PROD. ENVIR CONT CONSULT 0 .34 0 .329 0.35 I 1 2 I 1 2 2 2 SEARCH SR TECH ASSOC 0.344 0 .058 0 .559 2 TEFLON@ COPOLYMERS PROD . 6810 OPERATOR 1 0.349 0 .349 0 .349 ----- ---- ---'- 3 3 8 1 1 TEFLON@ MAINTENANCE 6720 STC/SPS MECH 0 .35 0 .159 0.54 1 1 2 22RE 2 2 TEFLON POLYMERS PROD. SR CHEMIST 0.356 0 .308 0.405 2 2 2 TEFLON@ POLYMERS PROD. MAINT SUPT 0.363 0 .363 0.363 1 TEFLON@ COPOLYMERS ~ PROD. ._ ._. ---------- SR ENGINEER .- ---- 0.363 0 .101 0.576 4 I ~-t' - - -'--}--- TECHNICAL TECH 0.369 0 .148 0.589 1 I TEFLON@ COPOLYMER S PROD. 4420 LABORATORIAN 0.382 0 .166 0.708 7 2 I 5 2 9 2 2 2 2 TEFLON@ POLYMERS PROD. AREA SUP T 0 .112 0.837 TEFLON@ COPOLYMER S PROD. PROCESS ENGINEER 0.427 0.28 0.574 3 1 2 1 2 5 2 2 TEFLON POLYMERS PROD. SENIOR TECHNICIAN 0.427 0 .244 0 .61 l l 2 2 TEFLON@ POLYMERS PROD.TEFLON@ COPOLYMER S PROD. TEFLON@ COPOLYMER S PROD. QUALITY COORD OA44_ 0 .444_ TECH ASSOC 0.459 0 .196 ADMIN ASSISTANT 0 .488 ! 0 .488 0.444 l 1 0.715 2 I 0.488 1 2 3 2 1 2 TEFLON e POLYMERS PROD._ PROD'N COORDINATOR 0.55 _~0 .528 0572 2 I .._... ---- -- -' -------- - --- ------- 2 2 TEFLON POL..YMERS PROD S R TECH ASSOC 0.581 0.17 1 .57 3 1 4 ~ --- .--------- ----- ----- -- ----I _.-- 2 TECHNICAL SPECIALIST 0.635 0.61 0.659 TEFLON@ COPOLYMERS ~ PROD. -- ~ SPECIALIST 0.763 0 .134 2 .39 2 4 2 2 2 6 2 -T-E-F-L--O-N-(-~a-P-O-L-Y..M.E-R- S P- RO- D.-S -R --E -N -- G-I-N -E -E liR --0-.7- 65 --0-.4- 12 --1- .59-- 2i~ 4 ---rr------------ 2 TECHNICAL TECH SPEC 0.783 0 .783 0.783 ~ l f f 1 --~Y----'--- ----- -------~-- ----~-- -- -}- --f ----- -~ TEFLON POLYMERS PROD. 6810 ADVM OPR u 0.805 0 .299 1 .53 2 17 4 ! 23 -- ~ TEFLONa POLYMERS PROD_ . 6810 TRNC OPR 11 1 .07 1 .07 1 .07 4 TEFLON@ COPOLYMER S PROD. 6810 OPERATOR 11 1 .136 0 .188 5 .015 18 12 --------------------- --------- ----------- ------ TE_L_O_ N /~ POLYMERS PROD ._ SUPERVISOR 1 .21 0 .233 3 .18 1 19 3 I 4 50 8 2 r~._T..E_.F_L~O~N_a_)_P. OL- YM- ERS- PRO- D 6810- STC- /SPS OP 11 1 .3 1 .3 1 .3 1 ---~'---'--- TEFLON@ COPOLYMER S 1 ~~;.. . . . PROD. TECH . - _ . . . .. . . .. . ... ................. . . .. ... . . . SPEC --- - _ - - - 1 - - -.4.6 . . .. ;. 1 .. -.-4- 6 ----1-_. - _ r4 -_. ------~... 6 - -- -- r---- ---- - -I ---- a I Z ------ -TE-F-LO-N--M -- A-IN -T -E -- NAtN-C-E-_. . 67.2- 0 M -E- CH -A- NIC- 1 - .72- 6 - O .- 15- 5 - 6 :81 ~ - 4 6 7 2 19 -- - ----- 3 +"~ '~i' 4TEFLONa POLYM ERS PROD. 6810 OPERATORII 3 .31 1 O .1 99 955 I 4 2 18 LL 8 32 ~43-.,:' 73 of 73 ~~