Document N2ZoyybvXnvav5mw0xej850Yb

DownloadRandom document
3 M Company Psge 1 of 121 FINAL REPORT Epidemiology, 220-3W-05 Medical Department 3M Company St. Paul, MN 55144 Date: October 11,2001* Title: A Cross-sectional Analysis of Serum Perfluorooctanesuifonate(PFOS) and Perfluorooctanoate (PFOA) in Relation to Clinical Chemistry, Thyroid Hormone, Hematology and Urinalysis Results from Male and Female Employee Participants of the 3,000Antwerp and Decatur Fluorochemical Medical SurveillanceProgram Study Start Date: March I, 2000 Protocol Number (not applicable) IRB Approval Exempt Expedjted -x IRBApproval Date: (not applicable as these data are from a medical surveillance --* pro,oram? Principal Investigator: Co-investigators: Study Director: Geary W. Olsen, D.V.M.P,h.D.' Michele M. Burlew, M.S.' Jean M. Bums, R.N., M.P.H.' Jeffrey H. Mandel, MD.,M.P.H.' Jeffrey H. Mandel, M.D.,M.P.H.' 1. 3M Medical Department. 230-3W-05,St. Paul, MN 55144-1000 *(Correctionsmade from previous version) 3M Company : Page 2 of 121 ABSTRACT ( The 3M fluorochemical medical surveillance program is conducted on a routine periodic basis at the company's Antwerp (Belgium) and Decatur (Alabama) fluorochemical manufacturing plants. In the most recent occurrence in 2000, there were 255 Antwerp employees (206 male and 49 female) and 263 Decatur employees (215 male, 48 female) who participated in the program. This represents approximately 75 percent and 50 percent of the eligible employees at these two locations, respectively. Seventy three percent of the participating Antwerp male employees and 75 percent of the Decatur employees were engaged in production activities. Only 12percent of the participating Antwerp female employees were engaged in production activities compared to 63 percent of the Decatur female employees. Employees' sera were quantitatively analyzed forPFOS (pcrfluorooctanesulfonat), PFOA (perfluorooctanoate!, PFHS (perfluorohexanesulfonate), PFOSAA (N-ethyl perfluorooctanesulfonamidoacetate), M570 (N-methyl perfluorooctanesulfonamidoacetate), PFOSA (perfluorooctanesulfonatearnidc) and M5S6 (perfluorooctanesulfonamidoacetate) using high=pressure liquid chromatography/lectrospray tandem mass spectrometry (HPLC/ESMSMS) and evaluated versus an extracted curve from a human serum matrix. A total organic fluorine index (TOF) was also determined by calculating the percent of each specific fluorochemical's molecular weight that was attributed to organic fluorine and multiplied by the ppm measured for each ftuorochemical and then summed across all seven fluorochemicals. 3M Company Page 3 of 121 Mean serum PFOS levels for Antwerp production and non-production male workers were 1.16 and 0.42 ppm, respectively. Among Decatur production and non- . production male workers, their mean serum PFOS levels were 1.63 and 0.73 ppm, respectively. Mean serum PFOA levels for Antwerp male production and non-production workers were 1.28 and 0.34 ppm, respectively. Among Decatur male production and non-production workers, their mean serum PFOA levels were 2.34 and 0.59 ppm, respectively. The mean PFOS and PFOA levels for the Antwerp female employees (primarily nonproduction) were 0.13 ppm and 0.07 ppm, respectively. The mean PFOS and PFOA levels for Decatur female production and nonproduction employees were 0.93 and 1.23 ppm, respectively. Separate reports have been written which analyzed the employees' serurn levels in relation to their job and building location work assignments as obtained from a self-reported work history questionnaire. A standard set of hematological and clinical chemistry tests were analyzed. These included the following hematological tests: hematocrit (percent), hemoglobin (gm/dl), red blood cells (RBC, 1000/ram3), white blood cells (WBC, I000/mm 3) and platelet count ( I000/mm3); and the following clinical chemistry tests: alkaline phosphatase (IU/L), gamma glutarnyl transferase (GGT, I'U/L), aspartate aminotransferase (AST, I-U/L), alanine aminotransferase (ALT, IU/L), total and direct bilirubin (mg/dl), blood urea nitrogen (BUN, mg/dl), serum creatinine (mg/dl), blood glucose (m_dl), cholesterol (m_dl). high density Cholesterol (HDL, mg/dl) and triglycerides (mg/dl). Urinalyses were only assessed for Decatur employees via the standard urine microstick analysis, which tested for urine glucose, albumin and red blood cells. Six thyroid hormones were also assayed: thyroid stimulating hormone (TSH;/_IU/ml); serum thyroxine (T4;/z_dl.,); 3M Company Page 4 of 121 free thyroxine (free T4; n_dL); serum triiodothyronine (T3; p_mL); thyroid hormone binding ratio (THBR, %, previously referred to as T3 Uptake) and free thyroxine index (FTI). Statistical analyses wer_ conducted on the entire surveillance population as well as subgroups by gender, production worker (yes/no) and location. Univariate analyses categorized mean levels by serum PFOS quartile distributions. Multivadable regression was used to analyze the clinical chenaistry and thyroid hormones as dependent variables in relation to the independent effects of PFOS, PFOA or TOF adjusted for several demographic variables (age, body mass index, number of alcoholic drinks per day, cigarettes smoked per day and years worked). There was a modest positive association between PFOS or PFO_ with cholesterol as well as a stronger positive association between PFOA and triglyceddes. These associations are inconsistent wi_ the known toxicological evidence that has shown the hypolipidemic (not hyperlipidernic) effect of PFOS (in rats and primates) and PFOA (in rats but no effect in primates) at dosages that produced serum PFOS or PFOA levels higher than those measured in this population. Therefore, it is unlikely the observed positive associations between PFOS or PFOA and lipids arc causal. Because of the potential confounding positive association with serum triglycerides, this variable was added to the hepatic clinical chemistry models as an independent variable. In these models, no significant.associations were observed with PFOS, PFOA or TOF in relation to alkaline phosphatase, GGT, AST, ALT or total bilirubin. Although T3 was positively associated with PFOA, no other thyroid hormones were associated with PFOS, PFOA or TOF: thus there is unlikely a causal explanation (e.g., hypothyroidism or 3M Company Page 5 of 121 hyperthyroidism) for this positive T3 association with PFOA. Hematological and urinalysriessultwsereunremarkable. Insummary,thefindingfsromthe2000fluorochemimceadlicaslurveillance programcontinuteosuggestthatAntwerpandDecatufrluorochemicparloductioand non-productieomnployeedso nothavesignificcahnatngeisnserumcholesterol, lipoprotcoirnhsepatiecnzymesthatareconsistewnitthtoxicologifcianldingisn laboratoarnyimalsL.imitatioonfsthestudyincludietcsross-sectidoensailgnt,he voluntaprayrticipatriaotneasndthelowerlevelosfserumPFOS andPFOA measured amongthesemployeecsomparedwiththosesuspectetdocauseeffectisnlaboratory animalsA.lon_tudinaanlalysisreportesdeparatefloyrthefluorochemicmaeldical surveillaAnnctewerpandDecatuprrogramdatafrom1994throug2h000. INTRODUCTION 3M Company Page 6 of 121 - The 3M fluorochemical medical surveillance pro=o'ramis conducted on a routine periodic basis at the company's Antwerp (Belgium) and Decatur (Alabama) fluorochemical manufacturing plants. Prior to 1994, total organic fluorine was measured rather than any specific fluorochemical analyte. Serum perfluorooctanesulfonate (I'FOS) and perfluorooctanoate (PFOA) have been routinely assayed since 1994/95 rather than total organic fluorine. An analysis of the 1994/95 and 1997 medical surveillance pi'ogram data in relation to Antwerp and Decatur employees' serum PFO$ levels have been reported elsewhere (Olsen et al, 1998a, 1999a). In the 1994/1995 medical surveillance program, a total of 178 employees participated (Antwerp = 88; Decatur = o 90) and 149 employees participated in 1997 (Antwerp -- 65; Decatur = 84). A total of 61 Antwerp and Decatur employees participated in both years. The Antwerp male employee population was si_ificantly younger than that at Decatur, had lower Body Mass Indices (BMI) and had hi_er self-reported daily consumption of alcohol. In addition, the employees' clinical chemistry profiles were different for several tests. The Antwerp employee population had lower mean alkaline phosphatase and triglyceride values and higher total bilirubin and HDL values than the Decatur employee population. The findings from this prior epidemiologic analysis suggested that si_ificant clinical chemistry and hematological abnormalities were not associated with serum perfluorooctanesulfonate (PFO$) levels up to 6 parts per million (Olsen et al 1998a; 1999a). Nor were there consistent associations reported between serum PFO$ and several hormone tests including testosterone, estradiol and thyroid stimulating hormone (T$I'D. It was not possible to derive inferences from the few employees who had serum i : . .. 3M Company Page 7 of 121 PFOS levels > 6 ppm. An important limitation of this prior analysis was the low voluntary participation of male employees (less than 50%) and insufficient sample size of female employees which prevented a separate analysis. Also, although serum perfluorooctanoate (PFOA) was measured, it was not included in the analyses. Because the voluntary nature of the medical surveillance program may not provide for a complete understanding of the distribution of serum fluorochemical levels in the Decatur workforce, a random sample of 232 employees was selected for fluorochernical testing in the Fall, 1998. The distributions of employee serum PFOS and PFOA levels were comparable to the results reported in the voluntary Decatur medical surveillance program (Olsen et al 1999b). This finding suggested that the distribution of serum fluorochemical levels observed in the prior voluntary medical surveillance program likely reflected the distribution of serum PFOS and PFOA levels of all employees in the chemical plant. . Detailed discussions of the toxicology and epidemiology of PFOS and PFOA have been reported elsewhere (3M Company 2000; Alexander 2001a; 2001b; Butenhoff et al 2001: Gilliland and Mandel 1993:1996; Haughom and Spydevold 1992; Olsen et al 1998a; 1998b; 1999a; 2000; Pastoor et al 1987; Seacat et al 2001a; 2001b; Sohlenius et al 1993). For the purpose of brevity, this information will not be summarized in this Introduction. Suffice it to mention that for the purpose of employee medical surveillance, PFOS has been reported to be an inducer of peroxisome proliferation and hypolipidemia in rodents (Pastoor et al 1987; Ikeda et al 1987; Haughom and Spydevold 1992; Seacat et al 2001a: Sohlenius et al 1993 ) and primates (Seacat et al 2001b). PFOA has been inconsistently reported to produce hypolipidemia in rodents (Pastoor et al 1987; 3M Company Page8 of t21 ( Haughom and Spydevold 1992;) and not in primates (Butenhoffet al 2001). The mechanism of action pertaining to this hypolipidemia remains to be fully elucidated. The purpose of this report was to conduct a cross-sectional analysis of the 2000 fluorochemicaI medical surveillance program for Antwerp and Decatur male and female employees. Unlike the earlier report for Antwerp and Decatur employees which only analyzed for PFOS (Olsen et al 1998a; 1999a), the present study examined associations for both PFOS and/or PFOA as well as a calculated measure for total organic fluorine (TOF). Longitudinal analy_es of employees who participated from 1994/95 through 2000 were not analyzed as this was a focus of a separate analytical report (Olsen et al 2001a). o. METHODS The fluorochemical medical surveillance program is available, on a voluntary basis, to all Antwerp and Decatur chemical plant employees and those site employees who may work in the chemical plant area. In 2000, approximately 340 Antwerp and 500 Decatur chemical plant and site employees were eligible to participate. In addition to the fluorochemical testing program, a standard battery of clinical chemistry, pulmonary -. function and urinalysis (Decatur only) tests were performed on employees. In addition, several thyroid hormones were measured, A site=specific work history was also administered to all employee participants, Analyses of these self=reported workplace questionnaire data in conjunction with the employees' serum fluorochemical levels have been reported elsewhere for Antwerp (Olsen et al 2001b) and Decatur (Olsen et al, 2001c). 3MCompany Page9 of 12i - Hematology. Clinical Chemistryand Urinalysis Allina Laboratory Services (St. Paul, Minnesota) performed the standard hematological and clinical chemistry tests. These included the following hematolo_cal tests: hematocrit (percent), hemoglobin (grrgdl), red blood cells (RBC, 1000/ram3), whim blood cells (WBC, 1000/mm 3) and platelet count (1000/ram3); and the following clinical chemistry tests: alkaline phosphatase (IU/L), gamma glutarnyl transferase (GGT, 1U/L), aspartate aminotransferasc (AST, IU/L), alanine arninotransferase (ALT, IU/L), total and direct bilirubin (mg/dl), blood urea nitrogen (BUN, mg/dl), serum creatininc (m_dl), blood glucose (mg/dl), cholesterol (mg/dl), high density cholesterol (HDL, m_dl) and triglyce'rides (rag/all). Urinalyses were only assessed for Decatur employees via the standard urine rrdcrostick analysis which tested for urine glucose, albumin and red blood cells. =oO. ThyroidHormones Six thyroid tests were conducted by LabCorp(Kansas City, MO): thyroid stimulating hormone (TSH; #IU/ml); serum thyroxine (T4; #g/dL); free thyroxine (free T4: ngdL): serum triiodothyronine (T3; pg/mL); thyroid hormone binding ratio (THBR, %, previously referred to as T3 Uptake) and free thyroxine index (FrI). TSH, free T4 and T3 were determined bYan immunochemiluminometric assay (ICMA). '1"4and THBR were determined by a cloned enzyme donorimmunoassay (CEDIA). FTI was calculated by multiplying T4 and THBR. 3MCompany Page10of 121 Fluorochemical Analyses Sera samples were extracted using an ion-pairing extraction procedure (Hansen et al, 2001). The extracts were quantitatively analyzed for PFOS (perfluorooctanesulfonate), PFOA (perfluorooctanoate), PFHS (perfluoroheanesulfonate), PFOSAA (N-ethyl perfluorooctanesulfonamidoacetate), M570 (N-methyl perfluorooctanesulfonamidoacetate), PFOSA (perfluorooctanesulfonateamide) and M556 (perfluorooctariesulfonamidoacetate) using high-pressure liquid chromatography/electrospray tandem mass spectrometry (HPLC/ESMSMS) and evaluated versus an extracted curve from a human serum matrix. Endogenous levels of certain fluorochemical were determined in the standard serum matrix and additional fluorochemical was spiked into the matrix. The total amount of each specific fluorochemical (endogenous + spiked) was used to construct an.extracted standard curve. All serum fluorochemical analyses were determined by Northwest Bioanaltyical Laboratory Inc. (Salt Lake City, LIT). A description of the distribution of the serum fluorochemical levels is reported elsewhere for Antwerp(Olsen et al, 2001b) and Decazur (Olsen et al, 2001c). For Antwerp, all employee serum values for PFOS and PFOA values were above the lower limit of quantitation (LLOQ). There was one employee (0.3 percent) with a PFI-IS value below the LLOQ (0.0027 ppm) and one employee (0.3 percent) with a M570 below the LLOQ (0.0057 ppm). There were 111 employees (44 percent) with PFOSAA vaiues below the LLOQ (0.006 ppm): 88 employees (35 percent) were below the LLOQ (0.001 ppm) for PFOSA; and 13 employees (5 percent) were below the LLOQ (0.0043 3M Company Page 1! of 121 ppm) for M556. For Decatur, all employee serum values for PFOS, PFHS, PFOA and M570 were above the respective lower limit of quantitation (LLOQ). There were 8 (3 percent) employees with PFOSAA values below the LLOQ (0.006 ppm); 111 employees (42 percent) were below the LLOQ for PFOSA (0.001 ppm); and 13 employees (5 percent) were below the LLOQ for M556 (0.0043 ppm). For statistical analysis purposes, serum fluorochemical values that were less than the LLOQ were assumed to be the midpoint between zero and the LLOQ. A total organic fluorine index (TOF) was determined by calculating the percent of each specific fluorochemicars molecular weight that was attributed to organic fluorine (PFOS (64.7%); PFHS (61.9%): PFOA (69.0%); PFOSAA (55.3%); PFOSA (64.7%); M570 (56.6%)_d M556 (58.1%)) multiplied by the ppm measured for each fluorochemical and then summed across all seven fluorochemicals. .-. Data Analyses Serum PFOS and PFOA levels were the predominant fluorochemicals as the other five analytes were measured at considerably lower levels (Olsen et al 2001b; 2001c); therefore. PFOS and PFOA were the only two specific fluorochemicals analyzed as explanatory, variables in regression models. TOF was also considered in the analyses which took into account these other analytes in an aggregate index (see above definition). Descriptive simple and stratified analyses, Pearson correlation coefficients, ANOVA and multivariable regression were used to evaluate associations between PFOS, PFOA and TOF and each hematolo_cal and clinical chemistry test and thyroid hormone assay. For stratified analyses, employees were divided into quartiles of their serum PFOS 3MCompany Page12of 121 / distribution. Age, body mass index, Current alcohol consumption (drinks per day) and cigarette use (cigarettes smoked per day), years worked at Antwerp or Decatur, and type of job (production versus non-production) were potential confounding factors that were considered in the analyses. Production jobs included cell operators, chemical operators, mill operators and crew supervisors. Non-production jobs included engineers, QAJAC laboratory and research workers, secretaries and managers. Multivariable regression models were fitted with PFOS and/or PFOA analyzed as a continuous variable(s). Significance of coefficients was considered at t6< .05. Natural log transformations of the dependent variables were performed, when necessary, to normalize variables and to enhance model fit. Study results were analyzed using the SAS System (1990). RESULTS . oo. Altogether. there were 255 Antwerp employees (206 male and 49 female) and 263 Decatur employees (215 male. 48 female) who participated in the 2000 fluorochemical medical surveillance program (Table 1). Seventy three percent of the Antwerp male employees and 75 percent of the Decatur employees worked in production activities. Only 12 percent of the Antwerp female employees worked in production activities compared to 63 percent of the Decatur female employees. Provided in Table 2 are the mean PFOS, PFOA and TOF values, demographic values and clinical chemistry, and thyroid values for male employees stratified by location and production or non-production work activities. Regardless of the production categorization, Antwerp male employees compared to Decatur employees had lower 3MCompany Page13of 121 ..... serum PFOS and PFOA levels; were si_ificantly younger; had lower mean BMIs; worked fewer years; drank, on average, more alcoholic beverages per day; had higher mean HDL and total bilirubin values; and had lower mean triglyceride, alkaline phosphatase, GGT, AST and ALT values. Mean thyroid hormone values tended to be higher among Antwerp employees. All mean values were within reference ranges. Comparable findings were observed for Antwerp female employees compared to Decatur female employees (Table 3). Given the differences between Antwerp and Decatur employees, univariate analyses were initially stratified by location. Antwerp data, stratified by gender and production, are provided in Tables 4 through 12. In a similar fashion Decatur employee data are provided in Tables 13-24. The Decatur data also include employee urinalysis results. Antwerp production male employee data (n " 150!,.stratified by quartile of serum PFOS distribution, is presented in three sequential tables for clinical chemistry (Table 4) and thyroid hormones (Table 5) and hematology (Table 6) results. The highest quartile (4in)mean serum PFOS level was 2.61 ppm (range 1.76 - 6.24 ppm) compared to the lowest quartile (I st) mean serum PFOS level of 0.29 pprn (range 0.04 - 0.41 ppm). Production workers in the highest quartile of serum PFOS levels were older and worked more years at Antwerp. There were no significant mean differences between the quartiles for BMI, cigarettes smoked or drinks per day. There was only one si=maificandtifference between the four quartile levels for any clinical chemistry, thyroid hormone or hematology comparisons. This significant difference was the comparison of the mean BUN value between the Ist and 3requartiles. 3M Company Page 14 of 121 In a similar fashion for the 56 non-production Antwerp male employees, their clinical chemistry, thyroid hormone and hematology results are presented in Tables 7, 8 . and 9, respectively, for their quartile distribution of serum PFOS. In this analysis, the highest quartile had a mean serum PFOS level of 0.90 ppm (range 0.49 - 1.76) compared to a mean of 0.13 ppm (range 0.05-0.20 ppm) in the lowest quartile. No significant mean differences were observed for demographic (Table 7), clinical chemistry (Table 7), thyroid hormone (Table 8) or hematology (Table 9) comparisons between the serum PFOS quartile distributions. Among the 49 Antwerp production and non-production female employees analyzed as a _oup (Table 10), the highest quartile mean serum PFOS level was 0.26 ppm (range 0.15- 0.55) compared to the lowest quartile mean serum PFOS level of 0.06 o ppm (range 0.04 - 0.08 ppm). The highest serum PFOS quartile did not significantly diffedremographicatlhlayntheotherthreqeuartil(eTsableI0).The lowerthree quartilheasdsomesignificdainftferencbestweenthemselvefsorthemeancomparisons ofyearsworkedanddrinkpserday.OnlyonecliniccahlemistrByU,N, wassignificantly differebnettweenthequartilaessthe3'aand4" quartilheasdhighemrean BUN values than the Ia quartileA.llmean valueswere withinreferenceranges. No significanmtean thyroihdormone(TableII)orhematolog(yTable12)differenwcaesobservebdetween thequartiles. A totaolf161l_aturproductimoanleemployeewserestratifbiaesdedontheir quartidliestributoifosnerumPFOS (Table13).The highesqtuartihlaeda 3.22ppm mean serumPFOS leve(lrang2e.31- 10.06c)omparedto0.55ppm mean serumPFOS leveilnthelowesqtuartilTeh.erewereno significmaenatn demographidcifferences / 3M Company Page 15 of 121 between the four quartilesand the only clinical chemistry test thatwas significantly different was ALT (Table 13). The highest quartile had a significantly higher meanALT level (44 IU/ml) comparedto the Ist(33 IU/ml), 2_a(32 IU/ml) or 3_ (33 lU/ml) quartiles. There were no significant mean differences for the Decatur male production employee quartiledistributions for thyroid hormones (Table 14), hematology (Table 15) or urinalysis (Table 16) results. Among the $4 Decatur non-productionmale employees (Table 17), their hi_est quartile mean serum PFOS level was 1.66 ppm (range 1.00 - 2.95 15pro)comparedto the lowest quartile mean of 0.19 ppm (range 0.06 - 0.29 ppm). The highest quartileworked almost twice as long as the lowest quartile(Table 17). There were no significant differences in other demographics, clinical chemistries (Table 17), thyroid hormones (Table 18), hematology (Table 19) or urinalysis (Table 20) results among the quartile distributions. o.o. Among the 48 Decatur productionandnon-production female employees (Table 21), the highest quartilehad a mean serum PFOS level of 2.04 ppm (range 1.38 - 3.62 ppm) compared to the lowest quartile mean serum PFOS level of 0.20 ppm (range0.06 0.31ppm).Them werenosignificdainftferencbeestweenthequartiliensrelatitoon demographic(sTable2I),cliniccahlemistri(eTsabl2eI)orthyroihdormones(Tabl2e2). The thirqduartihlaeda si2nificanltolwyermean platelceotuntthanthe1*tquartile (Table23):howevert,hefourtqhuartiwlaes notsignificanldoywerthantheIstquartile. Urinalysfiisndingdsidnotdiffebry quarti(lTeabl2e4). PresenteidnTable25 arcthenumber(andpercentagoef)AntwerporDecatur employeewshichhadabovereferenrcaengevaluefsorhepaticcliniccahlemistrtyests. 3M Company Page 16of 121 These findin_ in Table 25 are stratified by serum PFOS quartile distribution within each of the gender and production/non-production categories. Because each sub-population has a different serum PFOS quartile distribution, comparisons should only be done within each location-, production- and gench,-r-specific category. Also presented is the number and percentage of employees who had one or more liver enzyme and bilirubin tests above the reference ranges (see aggregate total liver panel). The percentage of Antwerp employees whose liver enzyme tests were above reference range values was comparable for production and non-production male employees. Among Decatur employees, there was a higher percentage of production male employees in the 4m quartile for ALT, GGT and the total liver panel than the other quartiles. For non.-production male employees, the highest percentages occurred in the second or third quartiles. Neither Antwerp or Decatur female employees had percentages consistent with any trend in the quartile distributions. Provided in Tables 26 and 27 are the serum PFOS quartile distributions for the combined 421 Antwerp and Decatur production and non-production male employees. The highest quartile (4m) had a mean serum distribution of 2.69 ppm (range 1.69 - I0.06 ppm) compared to 0.27 ppm mean (range 0.04 - 0.42 ppm) compared to the lowest (I st) quartile distribution. It is important to note that the number (and percentages) of Antwerp versus Decatur employees in each of these four quartiles differ (see footnote to Table 26). In the Iowost (I _t)quartile, there is a greater percentage of Antwerp than Decatur male employees and more non-production than production employees. In the subsequent higher serum PFOS quartiles, the percentage of Decatur production male employees increased and the percentage of non=production male employees, whether 3M Company Page 17of 12! from Antwerp or Decatur, decreased. These differences were also reflected in the demographics between quartiles. For example, demographically the trend from the lowest to highest quartile increased with age, BMI and years worked and decreased with the mean number of alcohol drinks per day. Likewise, the means of the clinical chemistry and thyroid hormone tests were reflective of the higher percentage of Antwerp employees in the lower quartiles and higher percentage of Decatur employees in the higher quartiles. Mean triglyceride and alkaline phosphatase levels were lower and total bilirubin levels were higher in the lowest quartile compared to the highest quartile. For thyroid hormones, T3 was lower in the I stquartile compared to the 4_ quartile and THBR was significantly higher. Combinc_d analyses of Antwerp and Decatur production and non-production female employees (Tables 28 and 29) presented a similar distribution of employees by location and production pattern as was observed with the production and non-production male employees (Tables 26 and 27). Antwerp female employees predominated in the lowest quartile and Decatur female employees predominated in the highest quartile. This distribution difference is then seen with the lower mean age, BMI and alkaline phospham,sc findings and the _-eater number of drinks per day and higher total bilirubin levels in the lowest quartile compared to the highest quartile. Also observed was a lower mean GOT and blood glucose level in the lowest quartile when compared to the highest quartile. There were no thyroid hormone differences between the quartile distributions (Table 29). Summarized in Table 30 are the combined number of Antwerp and Decatur employees (and percentages) who had hepatic clinical chemistry, tests above reference 3M Company Page 18 of 121 ...... range values stratified by quartile of the serum PFOS distribution. Among male employees, twelve percent of the employees had above reference range values for ALT and GGT in the 4_ quartile compared to 4 to 8 percent in the Is_through3_ quartiles. For the total liver panel, 23 percent of the male employees had one or more liver clinical chemistry tests above the reference range value compared to 14 to 16 percent in the lower three quartiles. No differences were observed within the female employee population. These percentages were not adjusted forpotential confounding factors (e.g., BMI). Because the hi_er liver enzyme function test resultsin the 4'hquartile might be confounded by demographics (higher BMI, older age) and/or clinical chemistry tests (triglycerides) reflective of dietary differences, multivariableregression analyses were conducted on ti'iecombined Antwerp and Decatur male employee participants. Each regression model had the following variables: productionjob (yes = 1; no = 0); .... Antwerp/Decatur (l = Antwerp; 0 = Decatur); age, BMI, cigarettes per day, drinks l_r day and yearsworked. For the analyses that invoh'ed hepatic clinical chemis_y tests, triglycerides was also considered a potentialexplanatory variable. Re=_'essionmodels analyzed serum PFOS, serum PFOA, serumPFOS and PFOA, and total organic fluorine (TOt=).. Provided in tables 31 through 34 are the analyses for these fluorochemical comparisons in relation to their effect on cholesterol, adjusted for the other explanatory variables. Serum PFOS was positively associated with cholesterol although its explanation of the variability of cholesterol in the model was less than I percent (see partial R:). (Note: This positive association is opposite that of the well.established negative association between serum cholesterol and PFOS that have been shown to occur 3M Company Page 19 of 121 (< in toxicological studies at threshold serum doses that were approximately 2 ordersof magnitude hi_er than those serum PFOS levels observed in these employees.) Like PFOS (Table 31), there were positive significant associations each for PFOA (Table 32) and TOF (Table 34) with cholesterol but the model that jointly examined the effects of PFOS and PFOA found neither to be significant (Table 33). Again, this is contrary to the toxicological research that has shown PFOA lowers serum cholesterol. Age and drinks per day were significant variables in the model with cholesterol. PFOS and TOF were not significantly associated with HDL, but PFOA was significantly negatively associated (Tables 35 through 38). As to be expected, BMI and drinks per day were stron_y associated with HDL. Analysis of triglycerides showed PFOS, PFOA and TOF were positively associated (Tables 39 through 42). PFOA appeared to be the more significant predictor than PFOS. (Note: PFOS and PFOA have decreased serum triglyceride levels ( at toxicological doses, not increased serum triglyceride levels.) Age, BMI and cigarettes smoked per day were significant variables in the triglycedde models found in Tables 39 through 42. Provided in Figures 1 through 3 are scatter plots of the simple linear regressions between the natural log of serum trigiycerides and PFOA for Antwerp male, Decatur male and Antwerp and Decatur female employees; Multivariable regression model results for the hepatic clinical chemistry analyses axe found in Tables 43 through 62. Because of the potential confounding positive association with serur0 triglycerides, this variable is added to these models. No significant associations were observed with PFOS, PFOA and TOF in relation to alkaline phosphatase (Tables 43 through 46), GGT (Tables 47 throu_ 50) or AST (Tables 51 through 54). Although PFOS or PFOA were not significantly associated with ALT 3MCompany Page20of 121 /- (Tables 55 - 57), TOF was positively associated with ALT (Table 58). PFOS, PFOA or ! TOF were not si_ificant predictors of total bilirubin (Tables 59-62). Multivariable regression analyses of the thyroid hormones resulted in no significant associations of PFOS, PFOA or TOF with TSH (Tables 63 - 66), T4 (Tables 67 - 70), Free T4 (Tables 7 t - 74), TI-IBR(Tables 75 - 78) or FTI (Tables 79 - 82). PFOS, PFOA and TOF were positively associated with T3 although contributed minimally to the variation explained in the model (see partial R2). DISCUSSION Although voluntary participation rates ranged from 53 percent (Decatur) to 75 percent (Antwerp), the 2000 fluorochemical medical surveillance program had the most (in absolute numbers) employee male and female participants ever for both locations. This is !!.kely due to a combination of factors including 1) greater knowledge of the collective (individual and research) value of the fluorochemical medical surveillance proem'am:2) employee awareness about the persistence and prevalence of PFOS in human tissue and the environment; and 3) the company's May 16, 2000 phase out announcement that it would cease production of perfluorooctanyl chemistry in certain repellents and surfactants by the end of 2000. Serum PFOS and PFOA levels were comparable to those previously reported for employees at these manufacturing operations. Serum levels appeared to be log normally distributed with the highest values for PFOS at 10 ppm. This upper tail of the serum PFOS distribution was also reported in a random sample analysis of Decatur employees conducted in 1998 (Olsen et al 1999b). Separate reports examine the employees' serum 3MCompany Page21of 121 PFOS, PFOA, PFHS, PFOSAA, M570, PFOSA and M556 levels measured in the 2000 fluorochemical medical surveillance program with their workplace operations in Antwerp (Olsen et al, 2001b) and Decatur (Olsen et al, 2001c). We continued to observe consistent differences between Antwerp and Decatur employees regarding their demographics and lifestyle differences. In particular, Antwerp male employees, on average, were younger (and thus worked less), had much lower BMIs and drank more alcoholic beverages than their Decatur counterparts. All three differences can be important confounding variables when analyzing lipid and hepatic clinical chemistry tests. We have also consistently seen higher total bilirubin values among Antwerp employees since 1995 which may be partially attributable to a greater prevalence of Gilbert's syndrome (Olsen et al 1998a; 1999a). An inconsistent finding from these aggregate analyses was the positive associations in the multivariable models reported between PFOS and serum cholesterol and PFOA and serum cholesterol and triglycerides. There is a substantial body of toxicological literature to suggest these associations are spurious because PFOS (in rats and primates) has been reported to decrease serum cholesterol and triglyceride levels (3M Company 2000; I-Iaughom and Spydcvold 1992; Ikeda et al 1987; Pastoor et al 1987; Seacat et al 2001a; 2001b; Sohlenius et al 1993). On the other hand, there is inconsistent evidence for hypolipidemia with PFOA in rodents (Pastoor et al 1987; Haughom and Spydevold 1992) and no effect observed in primates (Butenhoff et al 2001). In primates, there was no association observed between PFOA and cholesterol or triglyceddes (Butenhoff et al 2001). There is no toxicolo_cal evidence that at the serum PFOA levds observed in our medical surveillance program that PFOA would have resulted in 3MCompany Page22of 12! (" hypcrlipidemic associations. In addition, the PFOA levels observed among Antwerp and Decatur employees in 2000 was lower than those measured in 3M's Cottage Grove manufacturing employees whose serum PFOA levels have been assayed as high as 100 ppm. Hypolipidemic or hyperlipidemic effects have not been associated with serum PFOA levels among these Cottage Grove employees (Gilliland and Mandel 1996; Olsen et al, 2000). Most recently, the 2000 Cottage Grove fluorochemical medical surveillance program analysis again showed no association between serum PFOA levels and serum cholesterol or h"iglyceddcs (as seen in Figure 4). (Note: The serum PFOA levels graphed in Figure 4 are substantially higher than those cited in Figures I through 3 for the Antwerp and Decatur male and female employees.) We therefore believe that it is highly unlikely that these arc causal associations observed in the 2000 fluorochemical medical surveillance data between PFOA and serum cholesterol and triglycerides. : Previous toxicological and.e.pidemiological research has also not suggested positive associations between elevated serum liver enzymes results and serum PFOS or PFOA that were at the levels observed in the Antwerp and Decatur employee population (3M Company, 2000; Butenhoff et al 2001; Gilliland and Mandel 1996: Olsen et al 1998a; 1999a; 2000; Seacat et al 2001a; 2001b). In this 2000 fluorochemical medical surveillance program we observed, among Decatur production employees, a significantly greater mean ALT among those workers in the highest serum PFOS quartile distribution compared to the other,three quartiles. This highest quartile of Decatur employees also had the greatest percentage of employees with ALT (28%) and GGT (15%) values above the reference range as well as the total liver panel (35%). A comparable percentage (36%) was observed among Decatur non-production employees in the second lowest 3M Company Page23 of 121 quartile with one or more hepatic clinical chemistry tests above the reference range. When male employees were combined by production status and location (as seen in Table 30), we reported an upward trend in the percentage of employees in the highest quartile with values above the reference range for ALT (12%), GGT (12%) and total liver panel (23%). However, after adjusting the employees' individual liver function values by potential confounding factors including age, BMI, number of alcoholic drinks per day, cigarettes per day and serum trigiyceride values, we found no association between liver function values and PFOS or PFOA. We therefore suspect that the univariate associations were influenced by known confounders of liver function analyses A battery of thyroid hormone tests were included in the 2000 fluorochemical medical surveillance program due to preliminary, albeit biologically inconsistent, findings in toxicolo_cal studies that have yet to be completed. Our surveillance data do not s.ug. gest any biologically sig'nificant associations between thyroid hormones and employees' measured serum PFOS, PFOA or calculated TOF levels. A retrospective cohort mortality study of Decatur employees from 1961-1997 reported 3 deaths from bladder cancer compared to 0.2 expected in the subgroup of workers with the highest potential exposure to perfluorooctancsulfonyl fluoride ('POSF)based chemistry, and materials (Alexander 2001b). It was not determined whether this association was fluorochemical-related or possibly due to other non-fluorochemical occupational or non-o0cupational exposures. An analysis of episode of cares (Olsen et al 2001d) reported a higher reoccurrence of cystitis among female Decatur chemical plant workers than their counterparts in the film plant although the actual prevalence of unique individuals with episodes of care reg .arding cystitis was similar. No differences were 3MCompany Page24of 121 reported among male chemical and film plant employees. The analysis of these 2000 fluorochemical medical surveillance data showed no association between the prevalence of abnormal urinalyses and employee serum PFOS levels among the Decatur employees. Limitations of this study design include its cross-sectional nature which does not adequately allow for the assessment of temporal changes. However, the large participation of employees in 2000 who may have participated in the 1994/95 and/or 1997 fluorochemical medical surveillance programsat these two manufacturing sites has enabled a longitudinal analysis to be performed. This longitudinal analysis is the focus of a separate 3M investigation (Olsen et al, 2001a). Althou_ still very limited in numbers, we were able to provide separate cross-sectional analyses for female employees, for the first time, which showed no biologically relevant associations between se_'m PFOS and/or PFOA levels with clinical chemistries, thyroid hormones or hematology results. Because 3M has announced a phase-out of the production of pcrfluorooctany[ chemistry- related materials, we anticipate that the Antwerp and Decatur employee population mean PFOS and PFOA serum levels should be lower when measured during the next fluorochemical medical surveillance program. These future analyses may be hindered by the fewer employees in the workfor_e as a consequence of the phase-out announced by the company. Another study limitation was the lower serum PFOS and PFOA levels measured among these employees compared with those suspected to cause effects in laboratory animals. " In summary, the findin_ from the 2000 fluorochemical medical surveillance program continue to suggest that Antwerp and Decatur fluorocheraical production and non-production employees do not show substantial changes in serum hepatic enzymes, 3MCompany Page25 of 121 cholesterol, or lipoproteins associated with the serum PFOS and PFOA levels measure& A separate lon_tudinal analysis is reported for the fluorochemical medical surveillance Antwerp and Decatur program data from 1994 through 2000. ACKNOWLEDGEM_I'S The investigators acknowledge the contributions of Kimbcrly Young in the prcpartion of this report. f REFERENCES 3MCompany Page26 of 121 3M Company (2000). SIDS Initial assessment repot: Peffluorooctane sulfonic acid and its salts. St. Paul (_E_):3M Company, (unpublished report). Alexander BH (2001a). Mortality study of workers employed at the 3M Cottage Grove facility. Minneapolis (Ml_:University of Minnesota, (unpublished report). Alexander BH (2001b). Mortality study of workers employed at the 3M Decatur facility. _finneapolis (MN):University of Minnesota, (unpublished report). Butenhoff TL, Costa G, Elcombe C, Farrar D, Hansen K, Iwal H, Jung R, Kennedy G, Lieder P, Olsen GW, Thomford P. Toxicity of ammonium perfluorooctanoate (APFO) in cynomolgus monkeys after 26 weeks of oral dosing. St. Paul (MN):3M Company, (unpublished report) Gilliland FD, Mandel JS (1993). Mortality among employees of a perfluorooctanoic acid production plant. J Occup Med 35:950-954. Gilliland F'D, Mandel. JS (1996). Serum perfluorooctanoic acid and hepatic enzymes, lipoproteins and cholesterol: a study of occupationally exposed men. Am J Ind Med 129:560-568. Hansen KJ, Clemen LA, Ellefson ME, Johnson JHO (2001). Compound-specific, quantitative characterization of organic fluorochemicals in biolo_cal matrices. Environ Sci Technol 35:766-770. "'" Haughom B, Spydevold O (1992). The mechanism underlying the hypolipemic effect of perfluooctanoic acid (PFOA). perfluoroctanesulphonic acid (PFOSA) and clofibric acid. Biochemica et Biophysica Acta 1128:65-72. Ikeda T, Fukuda K, Mori I, Enomoto M, Komai T, Suga T (1987). Induction of c?xochrome P-450 and pcroixome proliferation in rat liver by perfluorinated octanesulfonic acid. In: Peri.rosmes in Biology and Medicine. (HD Fahrni and H Sies, eds) New York:Springer Verlag, pp 304-308. Olsen GW. Burr'is J'M, Mandel JH, Zobel LR (1998a). An epidemiologic investigation of clinical chemistries, hematology and hormones in relation to serum levels of perfluorooctane sulfoeate in male fluorochemical production employees. St. Paul:3M Company (unpublished report). Olsen GW, Gilliland FD, Burlew MM, Bun'is JM, Mandel JS, Mandel JH (1998b). An epidemiolo_c investigation of reproductive hormones in men with occupational exposure to perfluorooctanoic acid. JOEJv! (40(7):614-621. 3M Company Page27 of 121 Olsen GW, Bun'is JM, Mandel JH, Zobel LR (1999a). Serum perfluorooctane sulfonate and hepatic and lipid clinical chemistry tests in fluorochemical production employees. JOEM 41 (9):799-806. Olsen GW, Logan PW, Simpson CA, Hansen KJ, Burris JM, Burlew lk,_VI,Schumpert JC, Mandel JH (1999b). Fluorochemical exposure assessment of Decatur chemical and film plant employees. St. Paul:3M Company (unpublished report). Olsen GW, Burris J'M, Burlew MM, Mandel JH (2000). Plasma cholecystokinn and hepatic enzymes, cholesterol and lipoproteins in ammonium penCluorooctanoate production workers. Drug Chem Toxicol 23(4):603-620. Olsen GW, Burlew MM, Burris JB, Mandel J'M (2001a). A longitudinal analysis of serum perlfuorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) levels in relation to lipid and clinical chemistry test results from male employee participants of the 1994/95, 1997 and 2000 fluorochemical medical surveillance program. St. Paul, MN:3M Company (unpublished report). Olsen GW, Schmickler MN, Tierens J'M, Logan PW, Bun'is JM, Burlew MM, Lundberg JK. Mandel JH !2001b). Descriptive summary of serum fluorochemical levels among employee participants of the year 2000 Antwerp fluorochemical medical surveillance program. St. Paul:3M Company (unpublished report). Olsen GW, Logan PW, Simpson CA, Bun'is JM, Burlew MM, Lundberg .IK, Mandel JI-I (2001c). Descriptive summary of serum fluorochemical levels among employee participants of the year 2000 Dec_t'ur fluorochemical medical surveillance program. St Paul:3M Company (unpublished report). Olsen GW, Burlew _'VLMH, ocking BB, Skratt JC, Burris J1VI,Mandel J'H (2001d). An epidemiolo_c analysis of episodes of care of 3M Decatur chemical and film plant employees, 1993-1998. St. Paul 1_1:3M Company (unpublished report). P_toor "i'P, Lee KP, Perri lvIA, Gillies PJ (1987). Biochemical studies of ammonium perfluorooctnoate-induced hepatomegaly proliferation. Exp Mol Pathol 47:98-109. and morphological and peroxisome SAS Institute, Inc. (1990). SAS Users Guide:Statistics Version 6. Cary, NC:SAS Institute Inc. Seacat AM, Thomford PJ, Hansen KI, Clemen LA, Case biT, Butenhoff J'L (2001a). Sub.chronic dietary, toxicity, of potassium perfluorooctanesulfonic acid in rats. Toxicol Sci (submitted 200 la). Seacat AM, Thoford PJ, Hansen KJ, Olsen GW, Case M'T, Butenhoff JL. Subchronic toxicity studies on pea'fluorooctanesulfonate potassium salt in cynomolgus monkeys. Toxicol Sci (submitted, 2001b). 3M Company Page 28 of 121 Sohlenius AK, Eriksson AM, Hogstrom C, Kimland M, DePierre JW (1993). Perfluorooctansulfonic acid is a potent inducer of peroxisomal fatty acid B-oxidation and other activities known to be affected by peroxisome proliferators in mouse liver. Pharmacol Toxicol 72:90-93. r. oo.. Table 1 Number of Employee Participants in the 2000 Antwerp a,ld Decatur Medical Surveillance Programs '_... Comp:'l_y Page29 of l_l Antwe,p (N = 255) Male (N = 206) Production Non-Production Female (N = 49) Production Non-Production 150 (73)* 56 (27)* " 6 (12)* 43 (88)* Decatur (N = 263) . Male (N = 215) Production Non-Production Female (N = 48) Production Non-Production 161 (75)* 54 (25)* 30 (63)* 18 (37)* *Percent in parenthesis PFOS PFOA TOF Age BMI Years Worked Cigaretles/day Drinks/day Cholesterol , tlDL Triglycerides Alk Phos GGT AST ALT Table 2 Mean ValueforMale Eml)loyeeParticipantSse'rum FluomchemicalLevels, Demographics,ClinicaClhemistrieasndThyroidResults ... Company Page 30 of 121 All Antwerp(N = 206) Decatu(rN = 215) 0.96d 1.41) 1.1)3'* 1.90 . 1.6(: 2.65 37d 43 24.8d 28,8 13" !6 4 I. I '+ 2 i8 55d 6 0. I 215 44 1244 191 60d 74 23" 3i 23c 26 23 d 35 Production Antwerp(N = 150) Decatur(N = 161) I.IC 1.63 1.28" ,2.34 !.97 d 3.18 36d 42 24.64 28.9 12' 15 5 I. Id 215 55d 6 O.l 217 43 1244 198 " 60d 76 23d 31 23d 26 22d 36 Non-Production Antwerp(N = 56) De.calu(rN= 54) 0.42b 0.73 0.34b {).59 0.61 b 1.07 40b 45 25.2d 28.4 15= 22 2 I. !a 225' 55c 5 0.2 209 45 122b 169 60" 67 26b 29 24 25 25 31 Total Biliruhi,1 Direct Bilirubi. BUN Creatinine Glucose TSH T4 Free T4 T3 THBR F['I 1.0 d ().I_ 19d 1.2 85d 2.0' K.2 I. I_ 13ib 34d 2.7" p < .05 compared to Dcca;ur (t It=st) Up < .01 comparedto Decatur (t tesl) o p < .OOl comparedto Decatur (t lest) d p < .O00l compared to Decatur (t to.st) 0.7 (|. I 15 I. 1 95 2.9 8.4 I.I 125 31 2.5 Table 2 (continued) 1.0d O.Ih 19d I. I d 84 d 2.0" 8.3 I. Ic 132' 34 d 2.7" 0.7 0.1 15 1.2 95 3.I 8.4 I.I 127 30 2.5 I. Id 0. ! 19d 1.2 87' 1.9 X.I 1.1 126 35 d 2.7' _.l Company Page31 of 121 0.8 0.1 15 I. 1 94 2.2 1_.5 I.I 120 31 2.5 Table 3 3M Compan Page 32 of 12 ( Mean ValuesforFemaleEmployeeParticipanStesr'umFluorochemicLaelvels, DemographicCsl,inicCahlemistriaensdThyroiRdesults PFOS PFOA TOF Age BMI YearsWorked Antwerp (N = 49) 0.13d 0.07d 0.17n 36 22.8d 12a Decatur (N = 48) 0.93 1.23 1.76 42 27.7 13 Cigarettes/day Drinks/day Cholesterol HDL 2d 0.5d 208 68a 5 0.I 200 59 Trigiycerides 94a 133 Alk Phos 46a 65 .... GGT 12 d 18 .... AST 18 " 20 ALT 13d 19 Total Bilirubin 0.8b 0.6 Direct Bilirubin 0.I 0. I BUN 16 12 Creatinine 0.9 0.8 Glucose 85 87 TSH 2.3 2.3 T4 I0.2 9.3 Free T4 " 1A b 1.0 T3 148 b 128 THBR 3 28 FTI 2.9a 2.5 p < .05 compared to Decatur (student t test) bp < .01 compared to Decatur (student t test) c p < .001 compared to Decatur (student t test) dp < .0001 compared to Decatur (student t test) PFOS PFOA TOF Table 4 _ .Company Page 33 of 121 Fluorochemical, Antwelp Male Production Demographic and Clinical Chemistry Employee (N = 150) Results by Quartile of Serum PFOS Distlibution Mean QuartilIe(H = 37) Median SD R_u,_ 0.29"1'4 0.33 0.11 0.04-0.41 Mean 0.58_'4 Qu_util2e(H = 38) Median Sl) RanBe 0.57 0.12 0.41-0.78 Mean Quan!le3 (N = 38) Median SD RanBe 1.18'_4 1.16 0.22 0.79- 1.66 Mean 2.61t'z3 Quartil4e(N : 37) Media,, SD RanEe 2.27 1.06 1.67-6.24 0.944 0.42 1.06 0.02 - 4.03 1.51 0.72 1.70 0.07 - 7.04 !.02 1.00 0.60 0.21 - 3.2"/ 1.66' i.64 0.81 0.25 - 3.59 0.922.3.4 0.60 0.78 0.05 - 3.03 1.63j'4 1.08 1.26 0.42 - 5.69 !1.82''4 1.80 0.55 1.03 - 3.14 3.5 i "_ 3.30 I. 18 1.92 - 7.36 Age 334 BMI 24.3 Years Worked 82']'4 Cigarettes/day 5 Drinks/day 1.2 Cholcllearol 207 HDL 57 Triglyceddel 102 AIk Phm 60 OGT 20 AST 24 ALT 23 'loud Billmbin 1.0 Direct BJlimbJn 0. I BUN I83 (_rcalJnil_ 1.1 Glucose 85 34 23.8 5 0 1.0 202 57 102 61 16 24 22 1.0 0. I 17 J.1 87 7 23- 48 2."/. 19.2- 33.2 6 2-25 7 0 - 20 I 0- 4 39 145-308 13 32 - 85 49 34-221 15 .34-96 II 8-53 8 13 - 58 10 I I -71 0.3 0.6 - i .6 0.04 0.0 - 0.2 4 I I - 25 0.2 0.9- 1.5 19 31 - 131 37 24.9 121 5 I.I 216 52 125 60 24 24 22 !.0 0. I lit I.I 86 36 9 24.3 3.I II 9 0 8 0.9 1.0 217 41 49 10 113 87 60 15 20 16 23 5 21 8 0.9 0.4 0. I , 0. I 18 4 I.I 0.2 88 16 21 - 56 19.0-34.7 2-20 0- 25 0- 4 148-295 38 -72 35-546 30- 113 8-89 16- 41 10-43 0.5 - 2.0 0.0 - 0.3 12 - 25 0.8- 1.7 49- !13 37 25.0 12' 4 0.9 212 54 140 59 21 22 22 1.0 0. I 20' 1.1 84 36 25.3 13 0 0.7 196 53 113 59 19 21 20 1.0 0. I 19 I.I 85 9 2.8 '7 6 i 0.9 41 12 124 15 I! 5 9 0.3 0.03 5 0.2 21 22- 55 17.5- 32.3 1-29 0- 20 0- 4 105-297 29 - 80 41-731 30-94 10-64 13 - 33 9-46 0.5 - 2.3 0.0 - 0.2 14 - 31 0.8-2.0 45- 168 39' 24.3 15i 7 I. I 226 57 130 61 26 23 20 1.0 0. i 19 !.1 80 39 8 28 - 55 24.7 3.0 17.8- 30.9 15 6 5-29 0 8 0- 25 0.7 1.2 0- 5 232 46 122-316 51 19 26- 119 105 75 42- 346 62 14 21-89 19 19 7-85 22 6 15 - 39 20 9 8-45 0.9 0.3 0.4 - 2.2 O.! O.I 0.0 - 0.4 19 4 I I - 30 1.0 0.2 0.8 - 1.5 83 " 20 40- 120 IMaim hi Sillnificanlly difl_:renl(P < .05, I]onfer, mi (i)tUill) | te_) film! I_g IllCallI)f dlc I" quartile z Meau il IdlPllificantlydifferent (P < .05, Itoltfenoni (Dunn) ! lest) from die memt of Ihe 2" quartile 3Melm i| lilnificamtly different {P < .05, Bonferroni (Dunn) ! lest) from the nw.anof the 3'_ qnarlile 4Mean hi lieniflcanilv different (P < .05, Boiifffoni (I)tlnn) I lesl) frmn file mean of the 4kquartile :'Table 5 , Antwerp Male Production Employee (N = 150) Thyroid Results* by Quartileof SerumPFOS Distribution** _.., Company P.',gc34 of 121 TSH ' T4 ' Free T4 T3 THBR I_TI Quartile I (N = 37), Mean Median SD RMge i .8 1.6 |. I 05 - 5.7 8.2 8.3 1.5 5.4 - I 1.5 1.2 1.1 0.2 0.9- 1.5 127 127 15 95 - 155 34 34 3 28 - 40 2.8 2.6 0.5 2. I - 4.2 Quartile2 (N = 38) Mean Median SI) Itan_e 2.0 2.0 0.9 0.7 - 5.5 8.5 8.6 1.4 6.6 - 12.0 1.1 I.I 0.1 0.9- 1.4 134 132 17 34 34 2 102 - 169 29 - 39 2.8 2.8 0.4 2. I - 4.0 Quartile3 l_l = 38) Mean Median SD Range ,2.0 1.7 - 1.1 0.8 - 6.1 8.2 8. I 1.4 5.0 - 11.5 I.I 1.1 0.2 0.6- 1.6 132 132 19 97 - IS0 34 34 3 29 - 43 2.7 2.7 0.4 1.9 - 3.9 Quarlilc 4 (N = 3"/) Mean Median SD RanBc 2.2 1.6 3.0 0.5 - 19.4 8. I 8.2 1.4 4.7 - 11.0 1.1 1.1 0.2 0.8- 1.6 136 137 22 98 - 185 34 34 2 29 - 41 2.7 2.7 0.4 1.6 - 3.6 *No slsnificandy different (P < .05, Bonfermni (Dunn) t test) mean valu_ **See Table 4'for r_.xumPFOS quartiledistribution Table 6 Antweq_Male ProductionEmployee (N = 150) Hematology Results* by Quallile of Serum PFOS Distribution** _ _. Company Page 35 of J21 IIC"r IIGB RBC W]BC Platelets Quartile I (N = 37) Mean Median SD RanBe , 46 46 3 41-53 15.5 15.4 0.8 14.0- 1"/.4 5.2 5.2 0.3 4.6 - 5.9 7;0 6.4 1.8 4.2 - I 1.4 ,Quartile2 (N = 38) Mean Median SD Range 46 46 3 39-51 15.5 15.5 0.9 13.2- 18.1 5. I 5.2 0.3 4.4 - 5.9 7.3 7. I 1.8 4.4- 11.5 244 242 57 138 - 380 254 250 5 ! 167 - 373 Quartile3 (N = 37) Mean Median SD Range 46 46 3 40-51 15.4 15.5 0.8 13.6- 17.3 5. I 5.1 0.3 4.7 - 5.9 7.6 7.4 1.6 5.2- I1.1 253 242 .73 106 - 427 Mean 46 Quartile4 (N = 37) Median SD Range 46 " 3 41-51 15.3 15.3 0.9 13.4- 17.1 5.0 5.0 0.3 4.0 - 5.8 7.2 6.9 2.1 4.5 - 15.6 249 247 55 !37 - 369 *No IdlPaificantlydifferent (P < .05. Bonferroni (Dunn) t teat) mean values **See Table 4 for aeaumPPOS quartile distribution PlZOS PFOA TOP Table 7 _,.. Company Page 36 of 121 Antwerp Male Non-Production Employee (N = 56) Fluorochemical, Demographic and Clinical Chemistry Results by Quartile of Serum PFOS Distribution Mean Quartile I (N = 14) Median Sb Range 0.133.4 0.13 0.05 0.05-0.20 Mean 0.274 Qnarlilc2 (N = 13) Median SD Range 0.28 0.03 0.21 -0.31 Mean Quarlil 3 (N = 15) Median SD Range 0.401"4 0.41 Q.0$ 0.32-0.48 Mean Quartile 4 (N = 14) Median SD Range 0.90 t'z'! 0.64 0.4'7 0.49- i.'/6 0.184 0.06 0.46 0.01- I.?8 0.194 0.15 0.13 0.05-0.51 0.3-/ 0.32 0.23 0.06-0.85 0.621"2 0.42 0.49 0.12- I.,/8 0.243.4 0.17 0..'14 0.06- 1.38 0.374 0.35 0.11 0.22-0.64 '0.604 0.54 0.19 0.3-/-0.96 1.22I'L_ 1.09 0.69 0.56-3.01 Age 40 36 13 23 - 58 BMI 24.6 2S.I 3.3" 19.9- 31.3 Ye,m| Worked 15 15 10 1-29 Ciprelles/day I 0 5 0 - 20 Drinks/day I.I 0.7 0.9 0.1 -2.9 Choleslerol 215 218 3-/ 140-293 HDL 55 54 II 40--/8 "rri$1yc_idrs, 94 -/8 39 45 - 177 Alk Plms 59 59 I,/ 30 - 91 20 18 14 8-65 Ab'r 24 22 ? 15- 3-/ ALT 24 21 10 12-41 Tozal Bilimbin 1.2 Direct Bilirubin 0.1 BUN 18 1.2 0.3 0..'i- 1.9 0.1 0.04 ILl -0.2 18 4 15-2"/ Creallnlue 1.2 1.2 0.2 1,0- I.'/ Gtucos 84 86 14 60- 104 41 42 6 31- 53 2S.4, 17 24.3 16 3.3 2 I.'-734.2 6 6-29 0 0 0 0- 0 t.0 0.'/ 1.3 0.0-5.0 244 231 43 191-331 61 58 27 31 - 121 159 118 129 36 - 463 62 63 II 43 - 80 32 21 26 25 23 8 13- !11 15- 44 2-/ 25 14 10-61 1.2 1.2 0.4 0.8-2.0 0.1 0.1 0.04 0.1 -0.2 22 19 15 14-,/! I.$ I.I 1.3 0.9- 5.8 88 92 15 50- 107 38 26.1 13 4 I.I 219 53 120 61 26 25 24 0.9 0.1 18 1.2 8"/ 40 25.I 13 0 I.! 223 49 99 61 16 22 21 0.9 0.1 18 1.2 95 9 25 - 56 40 3.5 2 I.I- 33.9 8 3-26 24.4 15 ,/ 0 - 20 2 0.9 0.1-3.4 !.3 39 15,/-2,/,/ 225 10 40--/7 54 6t 49 - 254 I I,/ 18 30- 94 5-/ 21 "/-80 25 ,/ 14- 38 24 ti Ii -46 24 0.3 0.5- !.'7 I.I 0.1 0.0-0.3 0.1 4 13-25 19 0.2 0.8- 1.5 I.I 20 48- 114 91 41 24.2 15 0 0.9 231 5-/ 95 56 16 23 21 1.0 0.1 19 I.I 90 9 26 - 55 3.0 20.4- 30.I 9 2-27 4 0- I0 1.6 0.0-6.4 33 I,//8-2,/,/ 18 31 - 100 -/2 3-/- 262 12 39 - 77 26 6- 107 8 16- 49 I! 12-44 0.3 0.7- 1.9 0.03 0.1 -0.2 3 14-24 0.2 0.8- 1.6 14 68 - 115 1Mean is adllnificanllydifferent(P < .05. Bmd'er,mi (Dunn) ! test)from;luen_an of Ihe I'_quaflile sMean is si|niflcanlly different(P < .05. Bonferroni(Dunn) 1lell) fromthe meanof Ih 2" quartile 3Me,anis significantly differenl(P < .05, Bonferroni(Dunn) 11_1)from Ihe meanof Ihc 3_ quartile 4 ,f.,., ;, ,;,,,,ltir.nn*lu ,llrferenl (P < .05. Bonfemmi (Dunn) I lest) from the mean of II_ 4d'quartile Table 8 Antwerp Male Non-ProductionEmployees (N = 56) ThyroidResults* by Quartileof Serum PFOS Distribution** ..._ ComPany Page 37 of 121 TSH T4 Free T4 '1"3 THBR I_TI Quartile I (N = 14_ Mean Median SD Ranp' 2. I 2.0 I. I 0.4 - 4.2 8.9 8.9 I. I 6.8 - 10.4 1.2 1.2 0.2 1.0- !.5 131 128 16 106- 164 33 34 2 30 - 36 2.9 3.0 0.3 2.3 - 3.4 Quartile 2 (N = 13) Mean Median SD RanRe 2.0 1.9 1.2 0.7 - 5.4 7.8 8.6 !.5 5.0 - 9.4 1.1 1.2 0.2 0.9- 1.5 120 118 12 35 33 4 103- 145 30 - 42 2.6 2.7 0.4 2.0 - 3.4 Quartile 3 (N = 15) Mean Median SD Range ;1.6 1.7 1.0 0.03 - 4.3 7.9 7.7 1.3 5.7 - 9.8 1.1 1.1 0.2 0.9 - 1.5 128 126 25 9[ - 161 35 34 3 28 - 41 2.7 2.7 0.4 2.1 - 3.5 Quartile 4 (N = 147 Mean Median SD Ranse 2.0 1.6 1.4 1.0 - 6. I 7.9 7.7 1.9 4.2 - 11.4 i.1 1.2 0.2 0.9- 1.4 125 129 18 87- 147 35 34 2 32 - 41 2.7 2.7 0.6 1.7 - 3.7 NOsJpiflcamtly different (P < .05. Bonfexronj (Dean) t test) nwamvalues *See Table "/'forserum PFOS quartile distribution "Fable 9 Antwerp Male Production Employee (N = 49) Hematology Results* by Quartile of Serum PFOS Distribution** _.., Company Page 38 of 1:1 ItCT HGB RBC WBC PlateleAs Quartile 1 {N = 14) Mean Median SD Ranl| 44 44 2 40 - 48 15.1 15.0 i.I 12.2- 17.0 5.1 5.1 0.2 4.8-5.6 6.6 6.4 1.3 5.1 - 10.1 228 226 23 183 - 258 Quartile2 {N :13) Mean Median $1) Range 47 47 2 42 - 49 15.6 15.5 0.7 14.3- 16.5 5.1 5.1 0.3 4.6-5.6 6.9 6.5 1.7 4.7- 10.7 267 270 49 206 - 353 Quartile3 (lq : 15) Mean Median SD Range 46 46 3 42 - 51 15.4 15.3 0.8 14.0- 17.0 5.1 ,5.1 0.3 4.5-5.7 6.5 6.0 2.1 3.8- 11.0 225 221 i 53 129 - 335 Mean 45 Quartile4 (1'4= 14) Median SD Range 45 2 41 - 50 15.3 15.3 0.8 14.1- 17.1 5.0 5.0 0.3 4.7-5.5 6.5 6.4 1.3 4.9-9.6 234 234 39 172 - 306 *No significantly different (P < .05, Bonferroni (Dunn) t test) mean values **See Table 7 for serum PFOS quartile distribution PIzOS PI_A TOP ABe TuL 0 Antwerp Female Prodttclion* and Non-Production Employee (N = 49) Fiuorochemical, Demographic and Clinical Chclnistry Rgsults by Quartile of Serum PFOS Mean Quartile i (N = 12) Median SD Ranl|C Quartile 2 (H = 12) Mean Median SD Ranae Mean Quartile 3 (N = ! 3) Median SD Range 0.116 0.06 0.01 0.114-0.08 0.09 0.09 0.01 0.08-0.10 0.11 0.11 0..01 0.10-0.14 ...... J..... .' _, .,., 39 of 121 Mean Quartile'4 (N = 12) Median SD Range 0,26 i'2J 0.21 0.13 0.15-0.55 0.03 0.02 0.02 0.01 - 0.08 0.03 0.02 0.01 0.01 - 0.06 0.04 0.03 0.01 0.02- 0.07 0.202 0.09 0.31 0.02 - I. I I 0.08 0.07 0.02 0.05-0.12 0.09 O.l_J 0.01 0.07-0.11 0.12 0.11 0.02 0.09-0.15 0.40j_.3 0.26 0.34 0.13- 1.25 32 31 8 24 - .TAJ 36 34 9 24 - 52 38 36 5 3 ! - 48 3"/ 36 6 29 - 52 BM! Yc_lWorked 23.8 23.5 7_ 5 2.6 18.8-28.3 7" 0.8-22 21.9 21.8 13 12 2.5 :18.4-26.3 3 8 4-29 23.'/ 21.3 151 14 4.6 17.3-32.3 6 9-28 21.8 22.0 1.7 18.3 -25.0 13 13 7 5-29 C|gKeUeadchy I 0 3 O- I0 2 0 6 0-20 2 0 5 O- 15 2 0 4 O- 13 DdnllraFday 0.2 _ 0.1 0.3 O- 1,0 0,6 0.5 0.4 0.1 - !.3 0.6 a 0.4 0..4 0.0- 1.4 0,6 0.5 0.5 0.1 - 1.6 Cholesterol 205 195 30 I,i5 - 253 214 224 45 132- 274 208 19./ 39 160- 302 207 197 32 !64 - 2./I HDL 62 60 II 46- 85 71 "/2 19 46 - 121 68 64 18 43 - 104 "/2 6"/ 16 53 - 104 'rriglyceddes III 99 57 46-248 73 80 27 26- 112 98 93 43 46- 172 94 8"/ AIk lPhos 53 54 14 22- 70 44 48 14 .25 - 61 44 45 10 26- 61 42 42 AST 21 20 5 14 - 31 I./ 17 5 I ! -2"/ 18 17 6 9 -26 I./ 15 ALl' 14 12 ./ 8-35 12 12 2 8- 17 IS 13 ./ 6-34 II II GGT 12 I0 8 3 - 32 Ii II 5 2 - 19 14 10 10 7- 41 10 10 Total Blllmbin 0.8 0.7 0.2 0.5 - 1.2 1.0 1.0 0.3 0.5- 1.7 0.8 0.8 0.3 0.3 - 1.3 0.7 0.7 Dbr,cl Billmbin O.I O.I 0.07 0.0 - 0.2 0. I 0. I 0.09 1)I. - 0.4 0. I O.I 0.06 0.0 - 0.2 O.I 0. I BUN 123"4 12 2 9-16 15 15 3 11-23 18_ 19 4 9-22 16t 16 CreaUnlna 0.9 0.9 0.2 0.6- I. I 0.9 0.9 0.2 0.7 - !.3 1.0 1.0 0.2 0.7- 1.4 1.0 1.0 Glucose 75 ./6 16 38-98 86 *Number of Female employees by produclioncalegmy by quartile QI Q2 Q3 Q4 Produ_m 3 I 0 2 Non-Produ_ion 9 II 13 I0 90 12 65 - 101 89 91 II 65- 105 88 90 *Mean is significlmlJy different (P < .05. Bonfermnl (Dunn) t tat) from the mean of Ihe I" quarlile 2 Mean is significantly different (P < .05. Bonferroni (Duim) t test) from the mean of the 2"*quimile 3 Mean is significantly different (P < .05, Bonfenoni (Dunn) t teit) from the mean of the 3atquarlile * Mean is significantly different (P < .05, BonfeiToni(Dunn) I lest) from the mean oflhe 4" quarlile 44 Ii 6 3 4 0.2 0.0 3 O.I 17 32- !./I 20- 59 12 - 33 7- 18 5 - 23 0.3 - 1.2 0. I - 0. I 13-23 0.8 - !.2 49- 117 r' TSH ; 2"4 Fn_T4 2'3 THBR ]PTI i ;Table I I AntwerpFemaleProductionandNon-ProductionEmployee(N =49) ThyroidResulls*by Quartileof ScramPPOSDistribution** ".. Company Page 40 of 1.!.1 Qum,c1(N.. 12) Mm M_xiilm -qD Rmap 2.0 1.9 1.4 0.03 - 4.9 10.7 11.3 2.2 6.6 - 13.8 I.I I.I 0.1 0.8- 1.3- 157 164 29 106- 191 27 27 5 19 - 34 2.8 2.9 0.5 2.I - 3.6 Q,mtit2e(N= 12) Mean Median Sb Rang 2A 2.4 1.0 0.6 - 4. ! 9.7 9.7 1.8 6.9 - 12.3 - 1.1 1.2 0.1 0.9- 1.3 128 134 22 98- 163 31 31 4 25 - 36 2.9 3.0 0.4 2.3 - 3.6 Q,mi_ 3(N= l_) Mean Median SD RanBc ,2.2 1.8 - 1.8 0.03 - 6.7 10.5 10.7 3.6 4.6- 18.3 1.4 1.1 1.0 0.7-4.6 159 145 66 81-345 31 32 7 22 - 46 3.2 2.9 1.6 1.8- 8.4 Quart,4e(N= 12) . Mean Median SD Range 2.6 2.0 !.6 !.0 - 6.5 10.0 9.8 2.3 63 - 13.3- 1.0 1.0 0.1 0.9- 1.2 - 144 135 34 109-228, 29 29 4 24 - 35. 2.8 2.7 0.4 2.2 - 3.4, *No siplflr.a,ntly different (P < .05, Bonfeamni (Duma)t lem) racan values **So8 Table l0 for mum PPOS quartiledistribution Table 12 Antwerp Female Production and Non-Production Employee (N = 49) Hematology Results* by Quartile of Serum PFOS Distribution** ._M Company Page 41 of 121 HCT HGB RBC WBC Plateleta Quartile I (hi = 12) Mean Median SD Range 40 42 4 29-43 ., Quartile 2 (N = 12) Mean Median SI) Range 41 41 3 37 -45 13.3 13.6 4.6 4.7 1.5 9.4- 14.9 0.4 3.7 - 5. I 13.5 13.5 0.7 12.5- 15.0 4.5 4.5 0.3 4. I - 5. i 7.'/ 7.4 1.9 4.8 - 10.1 6.3 6.0 1.7 3.9 - 9.3 261 246 50 189 - 379 275 282 49 211 - 374 Quartile 3 (N = 13) Mean Median SD Range 40 41 2 35 -44 13.3 13.3 0.8 11.7- 14.8 4.5 4.5 0.2 4.2 - 5.0 7.3 7.2 !.4 5.4 - 9.5 277 251 68 202 -426 Quartile 4 (i'4= 12) Mean Median SD RanBe 41 41 2 38 -46 13.5 13.6 0.8 12.5- 15.2 4.5 4.5 0.3 4.2 - 5. l 6.5 6.3 1.3 4.4 - 9.4 269 260 64 181 - 406 *No significantly different (P < .05, Bonferroni (Dunn) t test) mean values **See Table 10 for serum PF:OSquartiledislribution /J -. : i PFOS PFOA TOI _ Age BMi Table 13 Comnp-',J_y Page 42 of 121 DecaturMale ProductionEmployee(N = 161) lquoruchcnlicalD, emographicandClinicalChemistryResultsby Quartileof ScramPFOSDistribution Mean 0.55TM Qu.a_ile ! (N ,, 40) Median Sb itanBe 0.55 0.16 0.11 -0.'/5 Mean 1.01TM Quarlile 2 (N : 40)" Median SD Range 0.99 0.18 0.76- 1.30 Mean Qumllc 3 (N'_ 41) Median SD Range 1.74TM 1.74 0.28 !.32-2.29 Mean 3.22t'2's Qumile 4 (N : 40) Median SD RanBc 3.03 1.22 2.31 - 10.06 1.24)' 1.24 0.6/ 0.06-2.72 1.824 1.53 1.05 0.35-4.61 2.42_'4 2.3/ 1.16 0.76-'7.48 3.88''z_ 3.68 1.86 1.52- 12.70 1.34TM 1.34 0.52 0.14-2.52 2.20TM 2.04 0.79 0.89-4.22 3.43la'4 3.32 1.06 !.75-6.61 5.75j'_') 5.31 I.'/7 3.00-12.23 43 44 9 26-63 42 41 9 26-61 42 43 8 28-57 41 41 10 27-60 29.0 28. I 3.7" 24.5 - 37.6 28A 27.3 5.2 1"/.2- 50. I 29.9 ' 29.2 5.0 22.6 - 45.5 28.3 28.3 4.0 19.9- 39.2 YeauraWodced 12 4 13 2-38 Cigarettes/day 8 0 13 0- 40 Drinks/day 0.2 0 11.3 0- I Choleiteml 214 220 43 121-296 HDL 42 41 8 29 - 59 13 5 12 2-34 5 0 I0 0 - 30 0.1 0 0.2 0- I 224 219 42 155-308 45 44 8 33 - 75 17 22 12 2-38 9 0 13 O- 40 0.1 0 0.2 0- I 213 208 44 147'384 43 42 11 29 - 70 16 14 II 3-38 4 0 9 0- 30 0.1 0 0.2 0.0- 1.0 216 210 39 160-319 43 43 8 28 - 64 Triglyceridel 232 19B 139 32 - 633 165 137 I01 32 - 550 202 16/ 138 44 - 792 195 175 128 39 - 796 AIk Phos 76 73 20 44 - 142 74 69 22 39 - 160 78 75 22 39- 139 75 71 20 44- 126 GGT 33 AST 26 ALl" 33 * Total Bilirubin 0.7 28 22 7 - 144 29 23" 17 I0 - 87 25 "/ 16 - 42 26 25 7 15 - 51 31 12 12-_3 324 28 17 6- 103 0.7 0.2 0.3 - i.5 0.8 0.8 0.2 0.4- 1.2 29 27 25 24 334 31 0.7 0.7 13 11 - 80 7 7 - 39 12 10-58 0.2 0.4- I. I 34 29 441._) 0.7 30 16 26 I! 37 23 0.7 0.2 10 - 71 15 - 69 12-99 0.4- 1.3 Direct Bilimbin 0. I BUN 15 Ctltlinlne I.I Glucose 97 0. I 0. I 0.0 - 0.2 0. I 0. I 0. I 0.0- 0.7 0. I 0. I .05 0.0- 0.2 0. I 15 5 9-33 15 15 4 8 -30 15 14 3 8-23 15 ,I.I 0.2 0.7- 1.6 I.! 1.1 0.2 0.8- i.7 1.4 1.0 2.2 0.8- 15.0 1.0 91 19 75- 184 93 93 l0 75- 113 99 93 39 74-381 92 0. I 0. I 0.0- 0.6 15 5 6-26 I.I 0.2 0.8- !.4 90 12 72 - 129 t Mean is silgnlficantly different (P < .05. Bonferroni (Dunn) t teat) from the mean of the I" quartile a Mean Is significantly different (P < .05, Bonfemmi (Dum0 t tea0 from the mean of the 2" quartile ) Mean 1| silgnlficantly different (P < .05. Bonfenoni (Donn) I lelt) flora Ihe ilgan of the 3ntquartile 4 Mean Is significantly different (P < .05. Bonfem)ni (Dunn) t test) from the mean of the 4 d*quartile Table 14 Decatur Male Production Employee (N = 161) Thyroid Results* by Quartileof Serum PFOS Distribution** ,ompany Page 43 of 121 TSH '1'4 l_r_ T4 '!"3 TIIBR FTI QmuliS:I {N ,,40) Mean Median SD Range 4.5 2.4 10.4 0.5 - 65.3 7.9 8.2 1.3, 4.6 - 10.7 1.0 1.0 0.1 0.6- 1.3 122 1IS 19 96 - 186 31 31 2,4 2.4 3 24-38 0.3 1.2- 3.0 Quartil2e_N = 40) ...... Qmmil 3 Q_I=41) Mean Median SD Range Mnn Median SD Range 2A 1.8 2.9 0,2 - 18.8 :2.4 2.1 1.5 0.8 - 8,6 8.5 8.5 1.5 3,3 - 1i.4 8.5 8.2 1.7 4.7 - 12.9 i.1 !.1 0.2 0.4- 1.4 I.I 1.1 0.2 03- 1.5 124 122 19 93 - 196 127 119 24 87 - 172 30 30 2 26-37 2.5 2.5 0.5 1.0 - 3.4 31 31 ,3 2.6 2.5 0.5 26-37 1.5 - 4.1 Quartil4e(N = 40) Mean Median $D Ranlce 3.0 2.4 3A 0,8 - 21.5 8.5 8.4 1.2 5.1 - 11.4 l.l 1.0 0.1 0.8- 1.3 135 136 23 97 - 190 30 30 3 25-38 2.5 2.4 ,, 0.4 1.9 - 3.4 *No slliniflcq_y diffee=nt (P < .05, Bonfemmi (Dunn) t m0 mean values **San Table 13/'ix mum _ quartiledlmibution "Fable 15 Decatur Male Production Employee (N = 161) Hematology Results* by Quartile of Serum PFOS Distribution** _ . Company Page 44 of 121 HCT HOB RBC WBC Phlteletl Quartile i (N.= 40) Me,an Median SD Ransc 45 45 2.7 39-51 15.2 15.3 4.9 5.0 0.9 13.3- 17.2 0.3 4.0 - 5.5 6.1 6.0 1.3 4.3 - 10.2 206 200 45 126 - 332 Quarlile 2 (N = 40) Mrain Median SD Ranjsc 45 45 2.2 40-50 15.1 15.2 5.0 5.0 0.8 13.4- 17.3 0.3 4. I - 5.6 6.2 5.9 1.6 3.3 - 10.2 224 222 42 146 --353 Quartile3 (N = 41) Mean Median SD Range 45 45 2.9 38-52 15.2 15.1 1.0 12.1 - 17.5 5.2 5.0 1.8 4.1 - 16.0 6.4 5.9 1.8 4.1 - 11.6 223 220 47 122 - 328 Mean 45 Quartile 4 (N = 40) Median SD Ranl_e 45 2.4 39-50 15.2 15.1 0.8 12.9- 16.6 5.0 5.0 , 0.4 3.8 - 5.9 6.5 6.3 1.8 3.8 - 13.5 2!6 213 46 132 - 307 *No significanlly different (P < .05, Bonfermni (Dunn) t lest) mean values **See Table 13 for serum PPOS quartile dislfibulion 3M Company Page 45 of 121 Table 16 / ( Decatur Male Production Employee (N = 161) Urinalysis Results by Quartile of Serum PFOS Distribution* Albumin Blood Sugar Quartile I N (%) 1 (3) 3 (8) 3 (8) Quartile 2 N (%) 2 (6) 4 (10) I (3) Quartile 3 N (%) 1 (3) 4 (10) 2 (1.2) Quartile N (%) 1 (3) 1 (3) 0 (0) Number of Employees: Q1 = 40; Q2 = 40; Q3 = 41; Q4 = 40 *See Table 13 for serum PFOS quartile distribution / - _, / / r ' / Table 17 PP(_ Decatur Male Non-Production Employee (N = 54) Clinical Chemistry Results by Quartile of Serum PFOS Distribution MeM Qumrllle I {N - 13) ..... Mediu SD R;mllc 0.19_'4 0.20 0.08 0.06-0.29 {_arlile 2 i_N= 14) Mean Mdiml Sb RamB 0.394 0.39 0.06 0.32-0.49 Mean (_ianile 3 (N -,14} Media. SD RJmee 0.711_I0".470 0.16 0.50-0.98 PPOA 0.344 0.21 0.84 0.04 -2.10 0.344 0.30 0.15 0.16-0.61 0.544 0.48 0[28 0.19- !.25 TOP 0.42)'4 0.37 0.46 0.08- 1.90 0.644 0.60 , 0.21 0.40- 1.12 !0.981'4 0.96 " 0.24 0.64- 1.39 Age BM! 42 44 10 27-59 42 36 13 28-60 48 51 8 30-36 29.$ 27.4 6:0 22.7 - 40.8 26.3 2S.5 4.0 21.7 - 35.4 29. I 28.7 3.4 25.5 - 37.3 Yeats Wodmd 15.1 18.7 11.9 0.8- 37.9 19.6 17.3 13.8 '2.2 - 38.5 24.3 28.6 11.6 2.3 - 34.2 CipleUeMday 5 0 Ddnkl/day 0.2 0 207 199 ItDL, 46 40 13 0-40 5 0 11 0-40 0.3 0 - 0.8 0.3 0 0.7 0 - 2.0 45 158- 305 204 192 42 153 - 278 14 32 - 82 49 46 14 35- 80 2 0 8 t 0.2 0 (J.s 0-30 0 - 1.6 203 197 48 144- 281 43 40 8 34 - 59 TriBiyt::eddel 157 AIk Phos 39 185 63 38 - 2.q4 129 96 64 59-241 61 15 26-79 62 62 17 : 39-98 161 154 65 62-284 72 72 15 48-105 GOT 22 AST 25 ALl" 28 Tolal IllUmldn 0.8 Bllimbin 0. I BUN 14 +_teallalM 1.0 21 22 23 0.8 0. I 13 " 1.0 8 7 16 0.2 0.07 3 0.1 II -35 16 - 42 I $ - 74 0.S - 1,0 0.0 - 0.2 8-22 0.9- 1.2 36 25 29 13- !19 29 29 28 27 I0 16 - 48 27 25 33 32 20 14 - 91 35 33 0.8 0.8 0.2 0.5- 1.0 0.8 0.'/ 0. I 0. I 0.03 0. I - 0.2 0. I O.I 14 14 3 10- 18 15 15 1.0 I,I 0.1 0.8- I.I I.I I.i 8 15-41 6 16 - 39 12 24 - 66 0.3 0.4-1.4 0.06 0.0- 0.2 5 8-24 0.2 0.7- 1.4 Ohtool 88 119 8 76 - 99 89 89 6 _'/9- 100 94 92 10 70- 112 i Moan is allpdflcanlly different (P .05, Uonfe,oai (Dunn) I test) from the ruefulof the Iu q,;artile 2 Mum il IdpIficanily different (P < .0S, Bonfe,oni (Dunn) 1lOSt)from Ihe mean of the 2 "t quartile SMall JlIJlniflr, anlJydjffetenl (P < .0S, Bonfenoni (Dana) t test) from the meatl of Ihe 3" quartile " ......... ,,n . n rt,,,,,,,..,,,_nIF;hmnn!_its0 from IhCmean of Ihc 4 d'quartile _o_.Compaqy Page 46 of 121 Mean Quartile 4 (N = 13) Median SD R_c 1.66i'_] 1.19 0.73 1.00-2.95 1.17t'2") 1.07 0.60 0.35 -2.05 2.29ia'3 1.88 0.91 1.13-3.74 49 51 6 35-56 28.7 29.4 2.3 24 - 32 27.81 31.8 8.6 4,5 - 35.4 8 0 15 0-40 0. I 0 0.2 0 - 0.8 222 233 42 ! 39 - 297 43 42 12 24 -'/3 232 122 I'/3 69 - 512 75 73 15 52-104 29 23 22 9-89 22 23 5 14 - 29 28 27 6 20- 4 I 0.8 0.8 0.2 0.4 - I.! 0. I 0. I 0.07 0. I - 0.3 16 14 4 12-22 I.I 1.1 0.1 1.0- 1.3 103 91 26 83 - 166 t Table 18 _ompany Page 47 of 121 Decatur Male Non-Production Employee (N = 54) Thyroid Results by Quartile of Serum PFOS Distribution** TSH T4 l_reeT4 1'3 THIBR Fl'I Qumilet(N=13) M_n Median $D 2.1 1.8 1.0 Kan_ 0.8-4.2 9.1 9.1 1.2 6.9-103 I.I 1.1 0.1 0.1-1.3 124 129 17 _J- 150 30 30 2 29-34 2.7 2.8 0.4 2.0-3.4 Q_rtilc2(N=14) M_n M_ian SD Ran_ 2.1 1.9 1.3 0.03-5.2 8.2 7.9 1.6 6.2-10.7 1.1 l.I 0.1 0.9-1.3 I_ 112 23 _- 180 31 31 3 25-35 2.4 2.6 0.4 1.8-3.2 Quanil3(N=14) Mean M_i_ SD Ran_ 3.0 2.0 .* 2_ 1.6-11.8 8.0 8.2 1.0 1.1 !.3 6.3-10.2 0.1 0.8-1.2 117 !14 23 86- 164 31 31 3 26-35 2.4 2.5 0.5 l.?-3.1 Qua_ile'4(N=13) Mean Median SD Ra._ 1.6 1.5 0.9 0.4-3.6 8.7 a.7 1.2 7.0-10.9 1.2 i.2 0.1 _9-1.4 118 12l 13 91 - 136 30 31 3 25 -36 2.6 2.5 0.4 2.0-3.2 *NO 818nifieanllydiffere_ (P < .05, Boafcn'oni(Dunn) t test)meanvalues *eS_ Table_7 tarserumPI_S quartidlieslribulio. / Table 19 Decatur Male Non-Production Employee (N = 54) Hematology Results* by Quartile of Serum PFOS Distribution** .... Company Page 48 of !21 Hc"r HOB RBC WBC Ph_lUl Quarlile I (N -- 13) Mean Median SD Range 45 45 3 38-51 15.2 15.4 5.0 5.0 1.2 12.0- 16.9 0.3 4.7 - 5.5 6.1 5.7 2.2 4.1 - 13.1 245 223 62 134-337 Quartile 2 (N = 14) M_n Median SD Range 45 45 .1 41-53 15.4 15.0 I.I I,i.2- 17.8 5.6 4.9 2.8 4.5 - 15.1 6.0 6.1 1.6 3.0-8.2 219 217 30 172-266 Quartile 3 (H =, 14) Mean Median SD RanBe 44 45 3 41 -49 14.9 15.0 0.8 13.6- 16.7 4.8 4.8 0.3 4.3 - 5.3 6.0 5.7 l.l 4A-8.6 230 206 58 149-361 Quartile4 (14= 13) Mean Median SD Range 45 45 ! 43 -49 15.0 15.1 0.5 14.0- 16.0 5. I 5. ! 0.3 4.6 - 5.5 6.4 6.0 1.6 4.4-9.8 212 203 34 167- 258 *No II|niflcantly different (P < .05, Bonfenoni (Dunn) t test) mean values **._e Table 17 for z-,rumPFOSquartile diaeribudon t Table 20 Decatur Male Non-Production Employee CN= 54) Urinalysis Results by Quartile of Serum PFOS Distribution* 3M Compan: Page 49 of 121 Albumin Blood Sugar Quartile I N (%) 0 (0) 2 (15) o (o) Quartile 2 N (%) 0 (0) 0 (0) o (o) Quartile 3 N (%) 0 (0) 2 (15) o (o) Quartile 4 N (k) 1 (8) 0 (0) o (o) Number of Employees: QI = 13; Q2 =14; Q3 = 14; (24 = 13 *See Table 17 for serum PFOS quartile distribution )' -o .., 7' PFOS PFOA "Fable21 Decatur Female Productionand Non-Production Employee (N = 48) Clinical Chemistry Resulls by Quartile of Senam PFOS Distribution Mean 0.203,4 Qtiarlile I (N = 121 Median SI) I,tu.f.e 0.20 0.08 0.06-O.31 Quu,lil2 (N = 12) Mean Median .',I;) RanEe, 0.493.4 O.SO O.13 0.32 -0.70 , Mean Quartile3 {N = 121 Median SD Rani]e 0.99 ='2"40.92 0.16 0.77 - 1.30 0.40 _'4 0.28 0.47 0.08 - 1.81 0.783'= 0.60 0.92 0.10 - 3.50 1.77 la 1.28 I. 17 0.25 - 4.00 TOF Age BMI 0.48 _.4 36 27.5 0.34 36 27.4 0.38 0.21 - 1.60 8 2.'i- 47 6.8 21.5 - 45.3 1.024 43 25,9 0.94 43 25.5 0.72 II 5.4 0.33 - 3.02 26 - 58 20.0 - 39.3 2-14m'4 1.83 44 43 27.5 28.2 0.88 0.86 - 3.54 6 32 - 50 4.5 20.3 - 33.6 Yean=WoAed II 7 10 2-27 14 13 II 2-27 12 6 10 4-32 Cigmettes/day 2 0 5 0 - 15 3 0 9 0 - 30 4 0 9 0 - 30 Drinks/day 0.0 Cholesterol 184 0 0.1 0.0-0.3 170 40 138-266 0.1 0.0 0.2 202 210 31 0.0- 1.0 139-262 0.0 0 0.1 0.0- 0.1 206 200 43 161-313 HDL 55 56 II 33-69 60 58 12 40-81 66 66 12 50-91 Triglycerides 96 AJk Phos 59 100 49 24 - 198 109 89 58 61 16 27-81 63 5g Ig 41 - 233 34-91 186 94 68 69 281 42 - 1049 22 41 - 100 GGT AST ALl" 14 15 6 6-26 14 13 6 7-30 28 17 27 10-97 22 21 8 13 -43 18 18 5 I 1-26 21 19 9 7 -39 20 16 13 9 - 58 18 17 7 I 1 - 36 22 17 12 6 - 47 Total Bilirubin 0.6 0.6 0.2 0.2 - 0.9 0.6 0.6 0.2 0.3 - 1.0 0.6 0.5 0. ! 0.4 - 0.8 Direct Bilimbin 0.1 BUN 12 0.1 0.1 0.0-0.2 13 4 5-20 03 0.1 0.05 0.0-0.1 12 13 3 8- 17 0.1 0.1 13 13 0.04 0.0-0.1 4 6-23 Creatinine 0.8 0.8 0.2 0.6- I.I 0.8 0.8 0.1 0.7- 1.2 0.9 0.9 0.2 0.6- 1.2 Glucose 89 90 IS 73- 125 86 86 9 72- 110 82 85 8 67-90 .,,! Company PageSOofi2t Mean Qoartile 4 (N = 12) Median SD Range 2.04 _'z'3 1.80 0.78 1.38 - 3.62 1.98la I..cut 1.27 0.85 - 5.4 I 3.391.2J 2.76 44 46 29.9 27.8 1.65 1.99 - 7.8 I 7 30 - $2 6.8 21.0 - 41.5 15 17 10 3-32 13 5 15 0 - 40 0.1 0.0 0.1 0.0-0.3 209 206 42 129- 287 55 55 12 36-78 142 113 94 46 - 3')4 70 70 15 44-95 16 13 9 6-30 18 17 5 I I - 30 17 14 6 10 - 29 0.5 0.5 0. ! 0.3 - 0.7 0.1 0.1 0.05 0.0-0.1 12 13 5 I - 18 0.9 0.8 0.1 0.7 - I.I 92 89 13 78 - 123 *Number of Female emldoyees by production category by quartile Production _QI Q2 3 I Q3 _ Q4 0 2 t Mean is significantly different (P < .05, Bonferroni (Dunn) t tesl) from Ihe mean of the I" quartile 2Mean is significantly difl_renI (P < .05, Bonfcrroni (l')ul|n) t lesl) fronl lEe mean of the 2*'=qu,trlile _tMean is significantly difi_rent (P < .05, Bmd_:rmni(Dunn) t test) from tile ineali of the 3'aquartile 4Mean is significantly different (P < .05, Bonferroni (Dunn) t lest) from Ihe mean of the 4't'quartile L Table 22 Decatur Female Produclion and Non-Production Employee (N = 48) Thyroid Resulls* by Quartile of Serum PFOS Distribution** TSH "!"4 FreeT4 3"3 THBR FTI Quanilel(N=12) Mean M_ian SD R_p _;' 2.0 2.0 1.3 0.03-4.8 10n 9.8 2.7 6.5-15.1 1.0 I.I 0.1" 0.9-1.3 132 126 26 i(_- 188 28 29 3 24 -34 2.7 2.7 0.6 1.7-3.8 Quarlilc2(N =12) M_n M_ian Si) Ran_ 2.6 2.2 1.3 0.7-4.6 9.0 8.9 2.0 5,8-12.2 1.1 1.0 _1 0.9-1.3 127 120 29 86 - 176 28 28 3 23- 36 2.5 2.5 0.5 1.7-3.1 *No significantly different (P < .05, Bonferroni (Dunn) t test) nw.anvalues **See Table 21 for serum PFOS quartiledistribution ,.Quarlile3)N=12 ) , Mean M_ian SD Ran_ ,2.4 2.1 1.2 1.0-5.2 9.2 8.9 1.7 6.7-11.9 1.0 1.0 0.1 0.7-1.1 126 119 26 91 - I_ 26 26 4 22 - 32 2.3 2.4 0.4 1.6-2.7 ..,f Company Page 51 of 12 [ Quaffilc4(N =!2) Mean M_ian SD RanBe 2.2 2.2 0.6 1.4-3.6 9.1 8.4 2.5 5.8-14.2 1.0 1.0 0.1 0.8-1.2 127 122 32 86 - 196 28 28 4 18- 32 2.4 2.4 0.4 1.8-3.0 /_ Table 23 Decatur Female Production and Non-Production Employe_ (I'4= 48) Hematology Results by Quartile of Serum PFOS Distribution* ,l Company Page 52 of 12! HCT HGB RBC WBC Platelets Quartile! (N = 12) Mean Median SD RanBe 38 38 12.6 12.7 3 .31 - 43 1.2 9.9- 14.4 4.3 4.3 0.3 3.8 - 4.8 6.7 6.3 2.0 4.2 - ! 1.7 2803 260 68 212- 450 Quartile 2 (N = 12) Mean Median SD Ranse 40 40 13.3 13.3 2 36 - 43 0.6 12.3- 14.5 4.3 4.3 0.3 3.7 - 4.8 6.6 6.5 1.9 4.3 - 10.4 228 216 42 185 -302 Quartile3 (N = 12) Mean Median SD Ranse 3.9 39 l 36 - 41 12.9 12.8 0.5 12.0- 13.8 4.4 4.3 0.3 3.9 - 5.0 5.9 6.2 1.7 2.8 - 8.4 209' 206 34 147 -272 Mean Quartile 4 (N = 12) Median SD Ransc 40 40 4 34 - 49 13.4 13.4 1.4 11.3- 16.2 4.3 4.4 0.4 3.9 - 5.0 7.6 7.5 1.9 4.2 - 10.4 258 254 55 159 - 339 *See Table 21 fi)r serum PFOS quartile distribution t Mean is significantly different (P < .05, Bonferroni (Dunn) t lest) from the reean of the i" quartile a Mean is significantly different (P < .05, Breffcrroni (Dunn) t test) from the ntean of the 2-' quartile 3 Mean is significantly different (P < .05, Bonferroni (Dunn) t test) from the mean of the 3_dquartile 4 Mean is significantly different (P < .05, Bonferroni (Dram) t teat)from the mean of ate 4" quartile Table 24 Decatur Female Production and Non-Production Employee (N = 48) Urinalysis Results by Quartile of Serum PFOS Distribution* :--_ Company P"___53 of 121 Albumin Blood Sugar - Quartile 1 N(_) 0 (0) 2 (17) 0 (0) Quartile 2 N(_) 0 (0) 0 (0) 0 (0) Quartile 3 N(_) 2 (17) 3 (25) 0 (0) Quartile 4 N(_) 0 (0) 0 (0) 0 (0) . Number of _mployees: Q1 = 12; Q2 = 12; Q3 - 12; Q4 = 12 *See Table 21 for serum PFOS quartile distribution k / Table z3"" Number of Participants (Percent in Parenthesis) Stratified by Antwerp or Decatur Employee Populations Who Had Above Reference Range Values for Hepatic Clinical Chemistry Tests by Quartile of Serum PFOS Distribution l.iiilJ,*i*)' o4 54 of 121 Antwer 0 Alkaline I_Jsphalase QI Q2 Q3 (,14 Mik 15rodm:liona O(O) 0(0) O(O) U(O) Muk Non-Productiona 0 (0) _nulc Productim_ . and Nowl_nod_mitm 0 (01 0 (O) 0 (0) 0 (0) 0 (0) 0 (01 O (01 Deealur Male Pmdaclion* I(3) 2(6) MaleNon-PnxhJ_iJ_ s O(0} O(0) PcmaleProduclim* and Nea-Pmduc_tim 0 (0) 0 (0) 2(6) O(O) i) (0) I(31 O(O) 0 (0) A_'I" QI 02 03 Q4 I(31 0(01 0(0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0(01 I (7) 0 (0) 0(0) I(3) 0(01 4(101 0(0) 2(14) 0(0) 0(0) ! (8) 0 (01 0 (01 0 (01 ALT QI Q2 Q3 Q4 I(31 0 (01 0 (0) 0(01 I (8) 0 (0) 0(01 0 (01 0 (0) 0(01 0 (0) 0 (0) GOT Total Lives"I"-,nel* Qi Q2 Q3 Q4 . QI Q2 Q3 Q4 I(3) I (7) 0 (0) I(3) I (8) 0 (0) 2(5) 4(II) 2 (13) I (7) 0 (0) 0 (0) 3(j)t 2 (14) 0 (0) $(131 2 ( 151 J (8) 4(II) 3 (201 0 (0) .5(14 I (81 0 (11) 3(R) I (8) I (8) 5(131 3(I_) 11(281 I (7) 2(14) 0(01 0 (01 0 (0) 0 (01 4(101 3(8) 2(6) 6(151 0(0) 3(21) 0(0) I (8) 0 (01 0 (0) 2 (1"7) 0 (0) "7(|8) 8(20) I (8) 5 (36) I (8) 0 (Q) 7(18) 14(35 2(14) I (8) 2 07) 0 (0) *Include Alkaline Phoqlmte, AST, ALT, GOT, Total and Din:or Bilimbin I See Table 4 for serum PFOS quartile distribution 2 Sec Table 7 for serum PVOS quartile distribution ) See Table !0 for serum PFOS quartile distribution 4 See Table 13 for serum PFOS qua,tile distribution ". s See Table 17 for serum PFOS quartile distribution 6 See Table 21 for serum PFOS quartile distribution J .,M Company Table 26 Page55of 121 Antwerp and Decatur Male Production and Non-Production* (N = 421) Fluorochemical, Demographic and Clinical Chemistry Results by Quartile of ScrumPFOS Distribution PFOS PFOA TOF Age BMI Mean Qumll I (N = 105) Median SI) Range 0.27TM 0.29 0.11 0.04-0.42 0.54TM 0.25 0.'/7 0.01-4.03 0.62TM 0.43 0.58 0.05-3.03 38 ) 36 10 23- 60 25.8 25.1 4.0 19.2- 40.8 Year|Worked 12_ II 10 1-38 C'issrelles/day 4 0 9 0-40 I_rinkgdly 0.9 ).+ 0.7 1.0 0- 5 Cholesleml 214. HDL 54 209 41 140-331 53 15 31 - 121 Triglycerides 1314 104 95 32-527 AlE Phos 61''4 62 16 26-98 (lOT 24 20 16 7- III AST 25 24 8 13-58 ALl[" 264 23 13 10-91 TotalBilJmbin 1.0).4. 0.9 0.3 0.5- 2.0 Direcl Bilimbin 0.1 0,1 0.1 0.0-0.3 ]BUN 18 I'/ 7 8-71 Crem, inl_ 1.2 I.! 0.$ 0.8- 5.8 Glucose 87 89 16 31 - 131 Mean Quarlile2 (N = 10.5) Median SI) Range 0.60t').+ 0.59 0.12 0.43-0.81 1.21_'+ 0.86 1,19 0.06-7.04 1.40TM 1.14 0.89 0.38-5.69 41 40 10 21 - 63 26.9 26.3 4.0 19.0- 37.3 15 11 12 5 0 10 0.6 0.3 0.9 214 217 43 47 45 !I 155 130 102 67 66 18 29 22 22 25 24 6 28 26 II 0.9 0.8 0.3 0.1 0.1 0.1 17 . 17 4 I.I I.I 0.2 2-38 0-40 0-4 121 -308 29 - 80 35-633 30- 142 7 - 144 16-49 10-63 0.3- 2.0 0.0-0.7 9-30 0.7- 1.7 91 90 17 _9- 184 Mean Quartile3 (N = 106) Median SD Ranlle 1.191"2"4t.17 0.24 0.82- 1.68 '1.45I'+ !.20 1.10 0.12-7.48 2.12''1'4 1.88 0.8"/ 0.98-6.61 421 43 9 22 - 61 27.3 26.7 4.5 17_ - 50. I 16t 16 ,II 1-38 6 0 10 0.5 a 0. I 0.9 0-40 0- 6 215 216 39 105 -303 48 46 13 24 - 100 169 134 123 32-731 69t 67 21 30- 160 26 23 15 6 - 89 24 24 7 7-51 28 26 14 6- 103 0.8_ 0.8 0.1 O.1 17 16 1.3 I.I 0.3" 0.4- 2.0 0.1 0.0-0.3 $ 8-31 1.4 0.8- 15.0 91 91 17 45- 168 Mean qu_ilc 4 <N: 105) Median SD Rangc 2.691'2") 2.46 1.09 1.69- 10.06 2.'/0_'2") 2.43 1.63 0.25-12.70 4.41L2'3 4.06 1.72 1.92-12.23 40 40 9 27 - 60 27.2 26.8 4.5 17.8- 45.5 15 6 0.5 j 222 48 1"7"/_ "7_0 30 25 331 15 0 0.0 214 45 155 67 25 24 29 10 10 0.9 44 15 123 19 17 .9 19 2-38 0-40 0- 5 122-384 26 - 119 39-796 21- 126 7 - 85 13-69 8 -9Q 0.8j 0.1 17 I.I 0.7 0.1 16 1.0 0.3 0.4- 2.2 0.1 0.0-0.6 5 6-30 0.2 0.8 - 1.5 91 89 30 40- 331 Table 26 (continued) *Nmnbcr of ,hale emph)yccs by Iocatio,i, producli_mcategoryand quartile (i)crccnl !.nparentlmsis) Quartilc I Pr_luclion Ntm-PrLxlucti.n Quartile 2 Production Hi)n-PllxJuL'ii)n Quartile 3 PrtKluction Non-Production Anlwcq) 38 38 Decatur 7 22 38 12 40 15 38 4 51 13 Total 45 (43) GO(57) 78 (74) 27 (2(,) 89 (84) 27 (16). I Mean i! dpilictnlly diffacnt (P < .0.%Bonfc.oai (l_um) I lUl) from liD"metal ol'lh I" quartile zbleml i| Idpificlmlly diffc:rcnt (P < .05, Uemfenoni(Dram)I test)fromIhe mcmlof Ihe 2-' qumlilc zMe.ami| idlnlflcanll), different(P < .05oBcmfenoni(Dunn)t teal)hornIhe meanof Ih 3_ quartile 4Mean is sillnificandydifferent(P < .05, Bonfc.oni (Du,u0 1test)fromIhc meanof Ihc 4'hquartile .,M Compa,ly Page 56 of 121 Quarlile 4 Production Non-Production 36 2 63 4 99 (94) 6 (6) Table 27 Antwerp and Decatur Male Produclion and Non-Production Employee Thyroid Results by Quartile of Serum PFOS Distribution* (N = 421) ,4Co,11p-',lty I'age 57 of 12 I TSH T4 Free T4 '!3 THBR Frl Quartile ! (N = 105) Mean Median SI) Ranlle 2.0 1.9 1.2 0.03-$3 8.3 8.5 1.4 5.0 - 11.5 1.1 !.1 0.2 0.9- 1.5 1244 123 17 94- 164 33 _'4 33 3 26 - 42 2.7 2.7 0.5 1.7 - 4.2 Quarlile 2 (N : 105,): Mean Median SD Range 3.1 2.0 6.6 0.5-65.3 8.2 8.4 1.4 4.2 - 12.0 l,l !.1 0.1 0.6- 1.4 128 127 20 86- 186 32 33 3 24 - 41 2.6 2.5 0.4 i.2 - 4.0 . Quartile 3 (N : 106) Mean Median SD Range : 2.1 1.7 - 2.0 0.2-18.8 8.3 8.2 1.5 3.3 - 12.9 !.1 I.l 0.2 0.4- 1.6 127 126 2l 9J - 196 32' 32 3 25 - 43 2.6 2.6 0.5 1.0 - 4. i Quartile4 (N : 105) Mean Median SD Range 2.5 1.9 2.8 0.5-21.5 8.4 8.2 1.4 4.7 - I1.4 1.1 1.2 0.2 0,8- 1.6 1321 131 22 87 - 190 32' 32 3 25 - 4 I 2.6 2.6 0.4 1.6 - 3.6 *See Table 26 for serum PFOS distribution Meanis sillnifR:anildyifferent(P < .05, BonfetTmt(iDunn)t lest) fromthe meanof the I" qumtile M_m is sl|nificantlydifferent(P.< .05, Bonfenoni(Dunn)t lest) fromIhe meanof the 2mdquartile Meanis siBnificamlldyifferent(P < .05, Boaferroni(Dunn)t teat)fromthe meaaof the3"_quartile 4Mc_ is al|nifieantlydifferent(P< .05,Bmferroni(Dunn)I IrJt)fromthe meanof the4_. quartile / Table 28 Anlwerp and Decatu," Female Production and Non-Production (H = 97) FI.orochemical, Demographic and Clinical Chemistry Results by Quartile of Serum PFOS Distribution _,,_MCoinpa,_y Page 58 of 121 PFO$ PFOA QumtiJe I _H,_24) Me'-,,t Median SD Raalle 0.073.4 0.08 0.02 0.04-0.10 0,043.* 0.02 0.04 0.01 -0.23 Quartile2 (N = 24) Mean Median SD RanBe 0.134 0.13 0.03 0.10-0.19 0.0"/4 0.05 0.07 0.02-0.34 TOF* Ale BMI 0.09 i'+ 344 22.84 0.09 34 23.4 0.04 0.05 -0.26 9 24 - 52 2.7 18.4 - 28.3 0.17 s'4 0.14 3"/4 36 23.94 22.2 0.07 ,0.09-0.35 7 25 - 52 4.3 17.3 - 32.3 Years Worked I I C'igaretteJ/day 14 Drinks/day 0.4' Chol_l 207. HDL 66 'rrillyceddes 93 AIk Phol 504 GGT II 4 AST 19 ALT 13 Total BJUmbln 0.8 j'4 Direct BHImbtn 0.1 BUN 14 Creatlnine 0.9 Glucose 80: g 0 0.3 203 61 90 52 10 19 12 0.8 0.1 13 0.9 82 8 I - 29 4 0-20 0.4 0- I 39 132 - 274 16 46- 121 48 26- 248 16 22-81 ? 2-32 4i II-31 5 8-35 0.2 0.5- 1.2 0.1 0.0-0.4 3 9-23 0.2 0.6- 1.3 14 38- 98 15 14 7 24 0 5 0.44 0.3 0.4 203 198 39 65 64 16 91 88 41 441.4 44 il 13 10 8 I it 16 7 16 13 11 0.83.4 0.8 0.3 0.1 0.1 0.1 16 16 4 1.0 1,0 0.2 93 t 92 13 3 - 29 0- 15 0- I 138 - 302 33- 104 24 - 172 20-65 5-41 9-43 6-58 0.2- 1.7 0.0-0.2 7-22 0.7- 1.4 65- 125 Mean 0.391 Quartile3 (N: 25) Median SD Range 0.37 0.15 0.20-0.70 0.611 0.36 0.74 0.04-3.50 0.80 t'_'+ 0.59 39 38 25.5 23.6 0.61 0.21-3.02 9 25 - 58 6. I 18.3 - 45.3 12 10 9 2 0 7 0.3 0 0.4 200 200 32 63 61 15 107 91 53 59" 56 16 14 12 6 19 19 5 16 15 6 0.6 t'2 0.6 0.2 0.1 0.1 0,1 14 14 4 0.9 0.8 0.1 85 87 12 2 - 27 0-30 0-2 !39 - 27 i 38- 104 32 - 233 32-91 7-30 11-33 7-36 0.3- 1.0 0.0-0.2 5-23 0.7- !.2 49- II0 Men. Quartile4 (H ,=24) Media. SD Ra.B 1.51''2'_ !.34 0.76 0.77-3.62 !.88 c2'] 1.39 1.20 0.25-5.41 2.77 t'2'3 2.66 44 ''2 45 2S.7 ''2 27.8 1.44 0.86-7.81 6 30- 52 5.7 20.3 - 41.S 14 12 8''2 0 0 t'2 0 208 202 60 58 164 104 69 s'2 70 22' 14 19 18 19 16 0.5 j'2 0.5 0.1 0.1 13 13 0.9 0.8 87 87 I0 13 0.1 42 13 206 18 21 7 IO 0, ! 0.04 5 0.2 12 3 - 32 0-40 0- 1 129 - 313 36-91 42 - 1049 41 - 100 6-97 7-39 6-47 0.3 -0.8 0.0- 0.1 I -23 0.6- 1.2 67- 123 "Fable 28 (continued) *Number of fmale cmployccs by location, production category and quartile (percent in parenthesis) Quartile I Ptt,Nlttctl(m Non-Production Quartile 2 Pmdm-Iion Non-Pr,_uclicm ", Quartile 3 Production Non-Production Antwerp 3 20 De_e_ur 0 I 2 17 I 4 ! 6 7 II Total 3 (12) 21 (88) 3 (12) 21 (88) 8 (32), 17 (68) t Mean is significantly diffexen! (P < .05, ,Boafenoni (Dram) ! test) from the mcmt of ihc 1"qmmil :Mean is siilnificanlly different (P < .05, Bonfcrroni (Dum0 ! leatt)from tlte mean of Ilte 2" quartile 3Mean is slllniflcanlly different (P < .05. Bonferroni (Dunn) t teat)from the nwamof the 3" quartile 4Mean is sillnificantly different (P < .05, Iltmfenolli (Dunn) I lest) from the mean of the 4a'qumlilc ) "- t M Company Page 59 of 121 Quanile 4 Produclion Non-Pr_luclion 0 0 22 2 22 (92) 2 (8) ' l Compal_y Table 29 l'agc60ot 12! AntwerpandDecaturPemaleProductionandNon-ProductioEnmployee(N = 9"7) ThyroidResults*by Quartileof SerumPFOSDistribution** TSH 1"4 lZreeT4 T3 TitBR PT! QumrlileI Mean Median SD R,mg 2.2 2.2 1.2 ,0.03 - 4.9 10.2 I.I 145 10.2 I.I 147 2.0 6.6- 13.8 0.1 0.8 - 1.3 28 98- 191 29 29 2.9 2.9 4 19 - 36 0.5 2. I - 3.6 Qua,tilc 2 Mean Median SD 2.2 2.0 1.5 Ranse 0.03 - 6.7 9.8 9.8 3.1 1.2 I.I 0.7 147 139 .'i3 4.6- 18.3 0.7 -4.6 81 -345 31 32 6 22 - 46 3.0 2.8 1.3 1.7 - 8.4 Quartile3 Mean Median SD ; 2.5 2. t 1.4 Range 0.7 - 6.5 9.9 9.5 2.3 5.8- 15.1 I.I I.I 0.l 0.9- 1.3 | 133 129 3"1 86-228 28 27 ,3 23 - 36 2.7 2.7 0.5 t.7 - 3.8 Quartile4 Mean Median SD 2.3 2.2 1.0 Ranl_e 1.0 - 5.2 9.1 1.0 127 8.7 . 2,1 1.0 0.1 120 28 5.8- 14.2 0.7 - 1.2 86- 196 27 27 4 18 - 32 2.4 2.4 0.4 1.6 - 3.0 *No significantly different(P < .05, Bonferroni{Dunn) t test) mean;'values **See Table 28 for serum PFOS quartiledistribution Table 30 ._ Company Page 61 of t2l Number of Participants (Percent in Parenthesis) by Employee Population Which Had AboveReferenceRangeValuesforHepaticClinicalChemistryTestsby Quartileof SerumPFOSDistribution _amJUIma AlkaU_Phmpluttate OI 02 O] Q4 AST QI Q2 Q) Q4 ALT QI Q2 Q3 Q4 t "" GOT QI Q2 03 04 TotalLiverPani* QI Q2 Q3 Q,I IM)duciloalut_ O(O) I (I) 3(3) 2(2) Hm-IMMmioa 3(3) I (I) I (I) 4(4) Vumbeaq_,_ PmtuaJomnds 0 (o) 0 (0) u (o) 0 (113 Hm-Pmdaakat 0 (03 I (4) o(03 0 (o1 4 (4) 4(4) 7(7) 13(12) ? 0 (o) I (4) 0 (It) 0 (03 6(6) 11(8) 6(6) 12(12) 15(14) 17(16) 17(16) 24(23) 0 (03 0 (0) 0 (0) 2(S) 0 (03 2 (u) 0 (03 2 (u) *lnchs_ Alkaline_ AST.ALl; GOT.TotalandDirectBllimbin _ Treble26 formum PlC3Squmile dimibutkm a See Table211fogt,emm Plq3Squadlledi_ibutimt p Intercept PFOS ProductionJob (yes/no) Anlwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked R2= .08 ^djR =.O6 *Natural log Table 3 I MultivariablcRegression Model of Cholesterol* by PFOS and OtherPotential ExplanatoryVariables for Antwelp and Decatur Male Employee Participants. 2000 Medical Surveillance Program Parameter 5.072 0.020 - 0.010 - 0.025 0.006 0.001 0.0007 0.035 - 0.002 SE ; 0.081 0.009 0.023 0.025 0.002 0.002 0.001 0.012 ?i 0.001 I p value .0001 .04 .66 .31 .0002 ".62 -.49 .004 .15 A Company Page 62 of 121 Partial R 2 - < .01 < .01 < .01 .04 < .01 < .01 .02 < .01 Intercept PFOA Production Job (yes/no) Antwerp/Decatur Age. BMI Cigarettes/day Drinks/day Years Worked R2= .08 Adj R2= .06 *Natural log Table 32 Multivariable Regression Model of Cholesterol* by PFOA and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants. 2000 Medical Surveillance Program Parameter 5.069 0.015 - 0.012 - 0.032 0.007 0.001 0.0006 0.03 - 0.002 SE 0.080 0.008 0.024 0.025 ' 0.002 0.002 0.001 0.01 0.001 p value .0001 .05 .63 .22 .0001 .64 .52 .005 .21 .Q_lnpally Page63of 12! Partial R2 < .01 < .01 < .01 .05 < .01 < .01 .02 < .0l Intercept PI_)S ' PFOA ProductionJob (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked R2 = .08 Adj R2 = ,06 *Nalurai log Table 33 MultivariableRegression Model of Cholesterol* by PFOS and PFOA and Other Poten.tialExplanatoryVariables for Antwerpand De.,ca_rMale Employee Participants, 2000 Medical Surveillance Program Parameter 5.066 0.015 0.009 - 0.018 - 0.033 0.007 t 0.001 0.0007 0.035 - 0.002 SE 0.08! 0.010 0.008 0.024 0.025 0.002 0.002 0.IX}1 0.0! ._ 0.001 p value < .0001 .16 . .26 .46 .20 .0001 .62 .50 .004 .15 ..M Compa.y Page 64 of 121 PartialR2 - < .01 < .01 < .01 < .01 .05 < .01 < .01 .02' .004 ,L Intercept TOF ProductionJob (yes/no) Antwerp/Decatur Age. BMI Cigarettes/day Drinks/day Years Worked R2=.08 Adj R2 = .07 *Natural log .t Table 34 MultivadablcRegressionModelof Cholesterol*by TOF andOtherPotentialExplanatoryVariables for AntwerpandDecaturMale EmployeeParticipants, 2000 MedicalSurvcillanccProgram Parameter 5.065 0.015 - 0.018 - 0.034 0.007 0.001 0.0006 0.034 - 0.002 SE 0.081 0.006 0.024 0.025 0.002 0.002 " 0.001 0.012 0.001 o p value t < .0001 .02 .45 .18 .0001 .63 .51 .005 .16 / .,4 Company Page 65 of 12 i PartialR2 < .01 < .01 < .01 .05. < .01 <.01 .02 < .01 Intercept PFOS Production Job (yes/no) Antwerp/Decatur Age. BMi Cigarettes/day Drinks/day Years Worked R2= .33 Adj R2---.32 *Natural log Table35 MultivariaRbelgeressioMnodelofHDL* byPFOS andOtherPotentiEaxlplanatoVrayriables forAntwerpandDecatuMraleEmployeeParticipants, 2000MedicalSurveillanPcreogram Parameter 4.313 -0.005 0.009 - 0.059 0.002 - 0.019 - 0.004 0.083 - 0.00l SE 0.090 0.011 0.026 0.027 0.002 0.003 0.00l 0.014 0.002 p value < .000 l .64 .73 .03 .37 < .0001 .0004 < .0001 .51 .'. t _.Clompany Page66 of 121 PartialR2 .01 < .01 .17 < .01 .07 .01 . .06 < .01 Intercept PFOA Production Job (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked R2 = .34 Adj R2= .32 *Natural log Table 36 Multivariable Regression Model of HDL* by PFOA and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 4.326 - 0.018 0.028 - 0.043 0.001 - 0.019 - 0.004 0.084 - 0.001 SE 0.090 0.009 0.027 0.028 0.002 0.003 0.001 0.014 0.002 p value <.0001 .04 .30 .13 .50 < .0001 .0004 < .0001 .54 Company Page67of 121 PartialR2 .04 < .01 .14 < .01 .07 .01 .06 < .01 Intercept " PFOS PFOA Production Job (yes/no) Antwerp/Decatur Age BMI Ci garettes/day Drinks/day Years Worked R2 = .34 Adj R2 = .32 *Natural log Table 37 Mullivariablc Regression Model of HDL* by PFOS and PFOA and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 4.324 0.006 - 0.020 0.025 - 0.043 0.001 - 0.019 - 0.004 0.084 - 0.001 SE 0.090 0.012 0.010 0.271 0.028 0.002 0.003 0.001 0.014 0.002 p value < .0001 .60 , .04 .36 .13 .50 < .0001 .0004 < .0001 .49 ...: Company Page68 of 121 Partial R z .01 .03 < .01 .14 < .01 .07 .01 .06 < .01 Intercept , TOP Production Job (yes/no) Antwerp/Decatur Age, BMI Cigarettes/day Drinks/day Years Worked 14_- ,31 AdjII]_--...I1 *Nllltiralloll Table 38 Multivadahle Regression Model of HDL* by TOF and Olher Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2{)00 Medical Surveillance Program Parameter 4.322 - 0.010 0.022 - 0.050 0.001 - 0.019 - 0.004 J 0.084 - 0.0009 SE _ 0.090 0.007 0.027 0.028 :. 0.002 0.003 0.001 0.014 0.002 p value < .0001 . .14 .41 .08 .45 < .0001 .0004 < .0001 .58 i .o _..o.ol.o.o.j _69 of 121 Partial R2 .03 < .01 .15 < .01 .07 .01 .06 < .01 Intercept PFOS ProductionJob (yes/no) Antwerp/Decatur Age. BM! Cigarettes/day Drinks/day Years Worked R2= .28 Adj R== .27 *Natural log Table 39 Multivariable Regression Model of Triglycerides* by PFOS andOlhcr Potential ExplanatoryVariables for Antwerpand Decatur Male Employee Participants, 2000 Medical Surveillance Program Paramcter 2.768 0.066 0.023 0.151 0.013 0.055 0.008 0.033 - 0.007 SE 0.224 0.026 0.065 0.068 0.005 0.007 0.003 0.034 0.004 p value t < .0001 .01 , .72 .03 .009 < .0001 .002 .33 .07 ..i Company Page 70 of 121 Partial R2 .03 < .01 .10 .02 .10 .02 < .01 < .01 Intercept PFOA Production Job {yes/no) Antwerp/Ik, catur , Age BMI Cigarettes/day Drinks/day Years Worked R2 = .29 Adj Rz = .27 *Natural log Table 40 Multivariable Regression Model of Tdglyccddcs* by PFOA and Olher Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2(X_0Medical Surveillance Program Parameter 2,742 0,066 - 0.004 0.111 0,014 0,055 0.008 0.029 - 0.007 _ SE 0,224 0.021 0.066 0.070 0.005 0.007 0.003 0.034 0.004 p value < .0001 .002 .95 .12 .005 < .0001 .003 .15 .11 .._ Co,npaay Page 71 of 12! Partial R 2 .05 < .01 .08 .02 .10 .02 < .01 .005 /. 7 Intercept " PFOS PFOA Production Job (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked R_ = .29 Adj R2 = .27 *Natural log t " Table 4 ! Multivariable Regression Model of Tdglycerides* by PFOS and PFOA and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical. Surveillance Program Parameter 2.734 0.037 0.053 -0.021 0.109 0.014 0055 0.008 0.030 - 0.007 SE 0.224 0.029 0.023 0.067 0.070 0.005 0.007 0.003 0.034 0.004 p value < .0001 .20 . .02 .76 .12 .004 < .0001 .002 .15 .07 Company Page72of 121 Partial R 2 .03 .02 < .01 .08 .02 .10 .02 < .0 l < .01 /" r Intercept TOF ProduclionJob (yes/no) Antwerp/Decatur Age. BMI Cigarettes/day Drinks/day Years Worked R2= .29 Adj R2 - .27 *Natural log Table 42 MultivariablcRegression Model of Triglycerides* by TOF, and Other Potential ExplanatoryVariables for Antwerpand Decatur MaZeEmployee Participants, 2000 Medical Surveillance Program Parameter 2.736 0.056 - 0.017 0.113 0.014 0.055 0.008 0.030 - 0.007 SE 0.223 0.017 0.067 0.070 0.005 0.007 : 0.003 0.034 0.004 p value < .0001 .0009 .81 . 10 .005 < .0001 ,002 .37 .07 Company Page 73 of 121 Partial R2 .06 < .01 .08 .02 .10 .02 < .01 < .01 : Intercept PFOS Production Job (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R 2 =. 18 Adj R 2 = .16 *Natural log Table 43 Multivariable Regression Model of Alkaline Phosphatas* by PFOS and Other Potential Explanatory Variables for Antweq) and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 3.718 0.013 0.036 0.149 0.0008 - 0.004 0.002 - 0.024 - 0.002 0.083 SE 0.362 0.013 0.032 0.034 0.002 0.004 0.001 0.016 0.002 0.024 p value < .0001 .32 , .26 < .0001 .73 .26 .17 .14 .41 .0006 r , . Company "' Page 74 of 121 Partial R2 .02 < .0l .11 < .01 < .01 < .01 < .01 < .01 .02 / , . Intercept PFOA Production Job (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R2 = .18 Adj11'.tO *Natural log !.Table 44 Muitivariab|e Regression Model of Alkaline Phosphatasc* by PFOA and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 3.746 0.0001 0.047 0.154 0.0006 - 0.004 0.002 - 0.024 - 0.001 0.086 SE 0.128 0.010 0.032 0.035 0.002 0.004 0.001 0.017 0.002 0.024 p value t < .0001 .99 , .15 < .0001 .80 .24 .18 .14 .52 .0004 .,M Company Page75 of 121 Partial R 2 .03 < .01 .10 < .01 < .01 < .01 < .01 < .01 .03 "Fable45 Intercept PFOS PFOA Multivariable Regression Model of Alkaline Phosphatase* by PFOS and PFOA and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program : Parameter i SE p value 3.747 0.128 < .0001 0.015 0.014 .28 - 0.005 0.012 .65 ProductionJob (yes/no) 0.040 0.033 .23 Antwerp/Decatur Age BMI 0.153 0.0007 - 0.004 0.034 0.002 0.004 < .0001 .78 .25 Cigarettes/day 0.002 0.001 .17 Drinks/day - 0.024 0.017 .15 Years Worked - 0.002 0.002 .42 Triglycerides* 0.085 0.024 .0005 R2=.18 Adj R2 =. 16 *Natural log _,4iCompany Page76of 121 Partial R 2 .. .02 < .01 < .01 .10 < .01 < .01 < .01 < .01 < .01 .02 : ..... -4 Intercept TOP Production Job (yes/no) Antwerp/Decat ur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R2 = .18 AdJR2=.16 *Natural log "Fable 46 Muitivariable Regression Model of Alkaline Phosphatase*by TOF and Olher Potential ExplanatoryVariables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 3.747 0.006 0.037 0.147 0.0008 - 0.004 0.002 - 0.024 - 0.002 0.084 AIk Phos SE 0.128 0.008 0.033 0.034 0.002 0.004 0.001 0.016 0.002 0.024 p value < .0001 , .47 .27 < .0001 .72 .25 .18 .14 .45 .0006 .l Compa,ly Page77 of 121 Partial Rz .04 < .01 . I0 < .01 " < .01 < .01 < .01 < .01 .02 Intercept PFOS ProductionJob (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R 2 = .25 Adj R2 = .23 *Natural log Table 47 MultiwtriableRegression Model of GGT* by PFOS and Other Potential ExplanatoryVariables forAntwerp and Decatur Male Employee Participants, 2(}00 Medical Surveillance Program Parameter 1.246 0.028 - 0.003 0.255 0.0003 0.006 0.003 0.117 0.002 0.294 SE 0.239 0,024 0.059 0.063 0.004 0.007 0.003 0.031 0.004 y 0.045 p value < .0001 .24 .96 < .0001 .95 .36 .28 .0002 .45 < .0001 Company Page 78 of 121 Parlial R2 .03 < .01 .07 < .01 . .02 .01 .03 < .01 .08 : Intercept PFq3A ProductionJob (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R2 = .25 ^dlr _'-.': _. _illUi.uhl ,g ' Table 48 Multivariable Regression Model of GGT* by PFOA and OtherPotential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 1.245 0.032 - 0.020 0.235 0.0009 0.006 0.003 0.116 0.003 0.289 SE 0.238 0.019 0.060 0.065 0.004 , 0.007 0.003 0.031 0.004 . 0.045 p value < .0001 0.10 . .74 .0003 .84 .35 .28 .0002 .40 < .0001 A Cumpany Page 79 of 121 Partial R2 , .04 < .01 .06 .01 .02 .01 .03 < .01 .08 rs Intercept " PFOS PFOA Production Job (yes/no) Antwerp/Decatur Age BMi Cigarettes/day Drinks/day Years Worked Trigl ycerides* R2 -- .25 Adj Rz = .23 *Natural log Table 49 Multivariable Regression Model of GGT* by PFOS and PFOA and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 1.246 0,014 0.028 - 0.026 0.23 0.0009 0.006 0.003 0.116 0.003 0.288 SE {).239 0.027 0.021 0.062 0.065 0.004 0.007 0.003 0.031 0.004 0.045 p value t < .0001 .60 .20 .67 .0003 .82 .34 .27 .0002 .46 < .0001 i Company Page 80 of 121 . Partial R2 .03 .02 < .001 .06 .01 .02 .01 .03 < .01 .08 Intercept TOP Production Job (yes/no) Antwerp/Decatur Age. BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R2 = .25 atural log Table 50 Multivariablc Regression Model of GGT* by TOF and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2(X)0 Medical Surveillance Program Parameter 1.246 0.029 - 0.028 0.235 0.001 0.006 0.003 0.116 0.003 0.288 SE ' 0.238 0.016 0.061 0.064 0.004 0.007 0.003 0.031 0.003 0.045 t p value < .0001 .06 , .64 .0003 .82 .33 .28 .0002 .48 < .0001 I Company Page81 o1'121 Partial R2 .04 < .01 .06 .01 .02 .01 " .03 < .01 .07 Intercept PFOS Production Job (yes/no) Antwerp/Decatur Age. BMI Cigarettes/day Drinks/day Years Worked Triglycerides* S 2 -- .09 Adj R2= .07 *Naturallog "Fable51 Multivariable Regression Model of AST* by PFOS and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 2.725 0.013 - 0.022 0.114 0.003 0.002 - 0.003 0.052 - 0.004 0.055 SE 0.133 0.013 0.033 0.035 0.002 0.004 0.001 0.017 0.002 0.025 p value < .0001 .33 , .50 .001 t .28 .53 .04 .002 .05 .03 a "3MCompany Page82 of 121 Partial R2 - < .01 < .01 .03 < .01 < .01 < .01 .02 .01 .01 Intercept PFOA Production Job (yes/no) Anl werp/l)ecatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R2 = .09 Adj R2 = .07 *Natural log Table 52 Multivariable RegresSion Model of AST* by PFOA and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2(X)0 Medical Surveillance Program Parameter 2.725 0,015 - 0.030 0.105 0.003 0.002 - 0.003 0.051 - 0.004 0.053 SE 0. i 33 0.011 0.034 0.036 0.002 0.004 0.001 0.017 0.002 0.025 p value < .0001 .17 , .37 .004 .23 .51 .04 .003 .05 .04 ,_Company Page 83 of 121 Partial R 2 .01 < .01 .02 < .01 < .01 < .01 .02 .01 < .01 / Intercept , PFOS PFOA Production Job (yes/no) Antwerp/Decatur Age BM! Cigarettes/day Drinks/day Years Worked Triglycerides* R2 = .09 Adj R2 = .07 *Natural log Table 53 Multivariable Regression Model of AST* by PFOS and PFOA and Olhelr Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 2.725 0.006 0.013 SE ; 0.132 0.015 0.012 p value < .0001 .68 .28 - 0.033 0.104 0.003 0.002 0.034 = '0.036 0.002 0.004 .34 .004 " .23 .50 - 0.003 0.001 .04 0.052 : 0.017 .003 - 0.004 0.002 .04 0.052 0.025 .04 ,. Company Page 84 of 121 Partial R2 < .01 < .01 < .01 .02 < .01 < .01 < .01 .02 .01 < .01 Intercept TOP ProductionJob (yes/no) Antwerp/Decatur Age BM! Cigarettes/day Drinks/day Years Worked Triglycerides* R2 - .09 Adj R2= .07 *Naturallog Table54 MultivariablcRegrcssionModel of AST* by TOF andOlherPotentiaEl xplanatoryVariabl_ for AntwerpandDecaturMale EmployeeParticipants, 2000 Medical Surveillance Program Parmmter 2.725 0.011 - 0.031 0.106 0.003 0.002 - 0.003 0.052 - 0.004 0.053 SE 0.133 0.009 !. 0.034 0.036 0.002 0.004 0.001 0.017 0.002 0.025 p.value < .0001 .17 , .36 .003 .24 .51 .04 .003 .04 . .04 i .t Company Page85of!21 .. PartialR2 .01 < .01 .02 < .01 < .01 < .01 .02 .01 < .01 Inlcrccpt " PFOS Production Job (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycendes* R2 = .27 Adj R2= .25 *Natural log Table 55 Muliivariable Regression Model of ALT* by PFOS and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Paramelcr 1.744 0.021 0.017 0.172 - 0.002 0.025 - 0.007 - 0.006 - 0.004 0.189 SE O.16.5 0.0 !8 0.043 0.042 0.003 0.004 0.002 0.024 0.002 0.032 p value < .0001 .25 .69 < .0001 .50 < .0001 .0003 .79 .10 < .0001 ,, Company l_agc86 of J2I Partial R2 .0 I < .01 .06 < .01 .13 .01 < .01 < .01 .05 /' In tercept PFOA Production Job (yes/no) Antwerp/Decatur Age BM 1 Cigarcltcs/day Drinks/day Years Worked Triglycerides* R2 = .27 Adj R2= .25 Natural log Table 56 Mullivariablc Regression Model of ALl'* by PFOA and Other PolenUalExplanatory Variables for Antwerp and Decatur Male Employee Participants, 2(XX)Medical SurveillanceProgram * i Parameter SE p value 1.761 0.165 < .0001 0.005 0.003 .13 0.027 0.041 .51 0.186 - 0.002 0.024 0.04 1 0.003 0.004 < .0001 .44 < .0001 - 0.007 - 0.005 - 0.004 !, 0.002 0.024 0,002 .0002 .83 .15 0.190 0.032 < .0001 _,.,,dCompany Page87 of 121 Partial R2 - < .01 < .01 .07 < .01 .12 .01 < .01 < .01 .05 Intercept PFOS PFOA ProductionJob (yes/no) Antwerp/Decatur Age BMI Cigarettes/day : Drinks/day Years Worked Triglycerides* R2 = .27 Adj R2 = .26 *Naturallog Table 57 MultivariableRegression Modal of ALT* by PFOS and PFOA and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 1.757 0.019 0.004 0.013 0.176 - 0.002 0.025 - 0.007 - 0.006 - 0.004 0.187 SE 0.165 0.018 0.003 :, 0.043 0.042 0.003 0.004 0.002 0.024 0.002 0.032 p value , < .0001 .29 , .15 .76 < .0001 .50 < .0001 .0003 .80 .11 < .0001 . , Company Page 88 of 121 Partial R2 .01 < .01 < .01 .06 < .01 .12 < .01 < .01 < .01 .05 ,, Intet_cept " TOF Production Job (yes/no) Antwerp/Decatur Age BM! Cigarettes/day Diinks/day Years Worked Triglycerides* R2 = .32 Adj R2= .31 Natural log Table 58 Multivariable Regression Model of ALl'* by TOF and Other l'otenlial Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 1,965 0.029 - 0.054 0.296 - 0.003 0.012 - 0,007 0.01 - 0.002 0.199 SE 0.193 0.013 0.050 0.051 0.004 0.005 0.002 0.025 0.003 0.04 p value < .0001 .02 , .27 < .0001 .38 .02 .0003 .62 .44 < ,0001 ._!Company Page89 of 121 Partial R2 .06 < .01 .15 < .01 .04 .01 < .01 . < .01 .05 i _ Lk_ " Intercept PFOS Production Job (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R2= .29 Adj Rz = .27 *Natural log Table 59 Multivadabl Regression Model of Total Bilirubin* by PFOS and Other Potential Explanatory Variables for Antweq_ and Decatur Male Employee Participants, 2{){)0 Medical Surveillance Program Parameter 0.209 - 0.017 - 0.068 - 0.262 0.001 - 0.005 - 0.008 0.005 0.0002 - 0.015 SE 0.145 0.015 0.036 0.038 0.003 0.004 0.002 0.02 0.002 0.027 p value < .0001 .25 .06 < .0001 .58 .19 < .0001 .80 .94 .57 .,,t Cl_mpany Page 90 of 12I Partial R2 .03 .01 .18 < .01 < .01 .06 < .01 < .01 < .01 / Intercept PFOA Pro/.luction Job (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R2 = .29 Adj R2 -.2 *Natural log Table 60 Multivariable Regression Model of Total Bilirubin* by PFOA and Other Potential Explanatory Variables for Antwerp and DecatUrMale Employee Participants, 2000 Medical Surveillance Program Parameter 0.210 - 0.004 - 0.078 - 0.265 0.002 - 0.005 - 0.008 0.005 - 0.0002 - 0.018 SE 0.145 0.011 0.037 0.039 0.003 0.004 0.002 0.019 0.002 0.027 p value < .0001 .74 , .04 < .0001 .54 .21 < .0001 .80 .91 .52 , Company Page 91 of 121 Partial Rz .05 .01 .17 < .01 < .01 .06 < .01 < .01 < .01 /, Intercept PFOS PFOA Produclion Job (yes/no) Antwerp/Decatur Age BMI Cigareues/day Drinks/day Years Worked Triglycerides* R2 = .29 Adj R2= .27 *Natural log Table 61 Multivariable Regression Model of Total Bilirubin* by PFOS and PFOA and Other Potential Explanatory Variables for Antwerp and Decalur Male Employee Participants, 2000 Medical Surveillance Program Parameter 0.209 - 0.018 0.002 - 0.070 - 0,264 0,002 - 0.005 - 0.008 0.005 0.0002 - 0,016 SE 0.144 0.016 0.013 0,037 0.039 : 0.003 0.004 0.002 0.002 , 0.002 0.027 p value < .0001 .27 .86 .063 < .0001 .57 .20 < .0001 .81 .95 .56 ,t Company Page92 of 121 Partial R 2 - 0 .03 .02 .01 .16 <.01 < .01 .06 < .01 < .01 < .01 ., Intercept TOF Production Job (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Tdglyceddes* R2 = .29 Adj R2 = .27 *Natural log Table 62 Muitivariablc Regression Model of Total Bilimbin* by TOff and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants. 2000 Medical Surveillance Program Parameter 0.210 - 0.011 - 0.064 - 0.257 0.001 - 0.005 - 0.008 0.005 0.00007 - 0.014 SE 0.144 0.009 0.037 0.039 0.003 0.004 0.002 0.019 0.002 0.027 p value , < .0001 .25 .09 < .0001 .62 .19 < .0001 .78 .98 .60 : Company Page 93 of 12i Partial R2 .06 .01 .16 < .0 ! < .01 .06 < .01 < .01 < .01 Intercept PFOS Produclion Job (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycefides* R2 = .07 Adj R2 = .05 Naturallog Table 63 MultivariableRegression Model of TSH* by PffOS and Other Potential ff_planatoryVariables forAntwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter - 0.539 0.015 0.109 0.184 0.005 - 0.005 - 0.005 0.057 - 0.008 0.204 SE 0.327 0.033 " 0.081 0.086 0.006 0.009 0.003 0.042 0.005 0.061 p value .10 .65 .18 .03 .36 .56 .17 .17 .13 .001 Company Page 94 of 121 ' PartialR2 < .0! < .01 .02 < .01 < .01 < .01 < .01 < .0 l .02 Intercept PFOA ProductionJob (yes/no) Antwerp/Decatur Age BM! Cigarettes/day Drinks/day Years Worked Triglycerides* R2 - .07 Adj R2 = .05 *Naturallog Table 64 ! m Mullivariahlc Regres'sionModel of TSH* by PFOA and Other Potential Explanalory Variables for Antwerp and Decatur Male Employee Participants 20(X}Medical Surveillance Program Parameter - 0.539 0.018 0.100 0.173 0.006 - 0.005 - 0.005 0.056 - 0.008 0.201 SE 0.327 0.027 0.083 0.088 0.006 0.009 0.003 0.042 0.005 0.062 p value .10 .51 , .23 .051 .33 .56 .17 .18 .13 .001 , Company Page 95 of 121 . Partial R 2 .02 < .01 < .01 < .01 < .01 < .01 < .01 < .01 .02 Intercept PFOS PPOA ProductionJob (yes/no) Antwe. rp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R2--.07 Adj Rz -.05 *Naturallog Table 65 MultivariableRegression Model of TSH* by PFOS and PFOA and Other Polential Explanatory Variables for Antwerpand Decatur Male Employee Participants, 2{X)0Medical Surveillance Program Parameter - 0.539 0.007 0.015 0.096 0.173 0.006 - 0.005 - 0.005 0.056 - 0.008 0.200 SE 0.327 0.036 0.029 0.084 0.089 0.006 0.009 ! 0.003 ,D 0.042 0.005 0.062 p value .10 .85 .61 .25 .05 .33 .57 .17 .18 .13 .001 !. .,/I (',roll,ally Page 96 of 12 I PartialR 2 - < .01 <.01 < .01 < .01 < .01 < .01 < .01 ." < .01 < .01 .02 Intercept TOF ProductionJob (yes/no) Antwerp/Decatur Age. BM! Cigarettes/day Drinks/day Years Worked Triglycerides* R2 = .07 Adj R2 = 05" ' *Naturallog Table 66 MultivariahlcRegression Model of TSH* by TOF and Olher Polenlial ExplanatoryVariables for Antwerp and Decatur Male Employee Participants, 20(}0 Medical Surveillance Program Parameter - 0.539 0.015 0.097 .174 0.006 - 0.005 -0.005 0.056 - 0.008 0.201 SE 0.327 t 0.021 0.084 0.088 0.006 0.009 _ 0.003 0.042 0.00_ o 0.062 p value .10 .49 .25 .05 .33 .57 .17 .18 .12 .001 ._!Company Pag_97of 121 . PartialRz .02 < .01 .01 < .01 < .01 <.01 < .01 < .01 .02 ,/ Intercept PFOS Produclion Job (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R2= .02 Adj R2 = .01 *Natural log Table 67 Multivariable Regression Model of T4* by PFOS and Olbcr Polenlial Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2(}00 Medical Surveillance Progrdm Parameter 2.263 - 0.003 - 0.003 0.011 - 0.003 0.001 0.0009 - 0.024 0.001 - 0.018 SE 0.090 0.009 0.022 0.024 0.002 0.003 0.0009 0.012 0.001 0.017 p value , < .001 .78 .91 .64 .08 .66 .32 .04 .35 .29 ! Company Page98 of 121 Partial R2 - < .01 < .01 < .01 < .01 < .01 < .0 l < .01 < .01 < .01 Intercept PFOA Production Job (yes/no) Antwerp/Decatur Age BMI. Cigareltes/day Drinks/day Years Worked Triglycerides* R2 = .03 Adj R2 = .01 *Natural log Table 68 Multivariable Regression Model of T4* by PFOA and Other Potential Explanatory Variables for Antwerp :md Decatur Male Employee Participants, 200(} Medical Surveillance Program Parameter 2.263 0.0003 - 0.005 0.010 - 0.003 0.001 0.001 -0.02 0.001 - 0.019 SE ; 0.090 t ' 0.007 0.023 0.024 0.002 0.003 0.0009 0.01 0.001 0.017 p value < .0001 .97 , .82 .69 .09 .64 .31 .04 37 .28 L Uiiq)dlty .._c99 of 121 Partial R2 <.01 < .01 < .01 < .01 < .01 < .01 < .01 < .01 < .01 Intercept PFOS PFOA ProductionJob (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R2 = .03 Adj R2 = < .01 *Natural log Table 69 Mullivariable Regression Model of T4* by PFOS and PFOA and OtherPotentialExplanatory Variables for Antwerpand Decatur Male Employee Participants, 2(100Medical Surveillance Program Parameter 2.263 - 0.003 0.001 - 0.004 0.010 - 0.003 0.001 0.0009 - 0.024 0.001 - 0.018 SE 0.090 0.010 0.008 0.023 0.024 0.002 0.003 0.0009 0.012 0.001 t ' 0.017 p value < .0001 .74 , .86 .87 .68 .09 .65 ".32 .04 .35 .29 . Company Page 100of 121 PartialRz < .01 < .01 < .01 < .01 < .01 < .01 < .01 < .0 I < .0I < .01 r Table 70 Multivariable Regression Model ofT4* by TOF and OtherPotentialExplanatoryVariables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program intercept " TOF ProductionJob (yes/no) Antwerp/Decatur Age BMI Cigarclte_@day Drinks/day Years Worked Triglycerides* Parameter 2.263 0.0002 - 0.005 0.010 - 0.003 0.001 0.001 - 0.024 0.001 - 0.019 R2 -- .03 Adj R2 = < _01 *Natural log t SE 0.090 0.006 0.023 0.024 0.002 0.003 0.0009 t " 0.012 0.001 0.017 p value < .0001 .97 , .82 .68 .09 .64 .31 .04 .37 .28 C(_mp;my Page 101 of 121 ParlialR2 < .01 < .01 < .01 < .01 < .0[ < .01 < .01 < .01 < .01 Intercep[ " PFOS Production Job (yes/no) Anlwerp/Decatur Age BM! Cigarettes/day Drinks/day Years Worked Triglycerides* R 2 = .06 Adj R2 = .04 *Naturallog s _l'able71 Multivariable Regression Model of Free T4* by PFOS and OIher Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 0.299 - 0.004 - 0.030 - 0.021 - 0.003 - 0.003 - 0.0009 0.006 0.002 0.003 SE : 0.076 0.008 0.019 0.020 0.001 0.002 0.0008 0.010 0.00 !. 0.014 p value < .0001 .63 . .11 .28 .03 .13 .27 .56 .21 .85 ._MCompany Page 102of 121 PartialR 2 < .01 < .01 .03 .01 < .01 < .01 < .01 < .0 l < .01 Intercept PFOA ProductionJob (yes/no) Antwerp/Decatur Age. BM! Cigarettes/day Drinks/day Years Worked Triglycerides* R2= .07 Adj R2= .04 *Natural log lt Table 72 MultivadablcRegressionModcl of FreeT4* by PFOA a,ldOtherl)otcntialExplanatoryVariables for Antwerpa,ldDecaturMale EmployeeParticipants, 2000MedicalSurveillanceProgram Parameter 0.299 - 0.006 - 0.025 - 0.017 - 0.003 - 0.003 - 0.0009 0.006 0.002 0.004 SE ; 0.076 0.006 0.019 ", 0.021 0.001 0.002 0.0008 0.010 0.001 0.014 p value .0001 .31 , .19 .41 .02 .13 .27 .53 .20 .78 . !' M CUml);my _ _gc 103 of 121 PartialRj .01 < .01 .02 .01 < .01 < .01 < .01 < .01 < .01 .t .. Table 73 MultivariableRegression Model of Free T4* by PFOS and PFOA and Other Potential ExplanatoryVariables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Intercept PFOS PFOA ProductionJob (yes/no) Antwerp/Decatur Age BMi Cigarettes,/day Drinks/day Years Worked Tdglycerides* R2 = .07 Adj g a --.04 *'" "ruralog Pman_ter 0.299 - 0.0005 - 0.006 - 0.025 - 0.020 - 0.003 - 0.003 - 0.0009 0.006 0.002 0;004 SE 0.076 0.008 0.007 0.020 0.210 0.001 0.002 0.0008 0.0!0 0.001 0.014 ". p value .0001 .96 . .37 .21 .41 .02 .13 .27 .54 ".20 .77 3M Company Page 104of 121 Partial R 2 - < .01 < .01 < .01 .02 .01 < .01 .002 < .01 < .01 < .01 " Intercept , TOP ProductionJob {yes/no) Antweqd[k, catur Age: BMI Cigaretles/day Drinks/day Years Worked Triglycerides* Rz = .06 Adj R2 = .04 *Natural log Table 74 MultivariableRegressioMn odel of FreeT4* byTOF andOtherPotentiaEl xplanatoryVariables for AntwerpandDecaturMale EmployeeParticipants, 2000MedicalSurveillanceProgram Parameter 0.299 - 0.004 - 0.025 - 0.018 - 0.003 - 0.003 - 0.0009 0.006 0.002 0.004 SE 0.076 0.005 0.020 0.020 0.001 0.002 " 0.0008 0.010 0.001 0.014 s p value .0001 .37 . .19 .38 .02 .13 .27 .01 .19 .79 3MCompany Page105of 121 Partial Rz .01 < .01 .02 .01 < .01 < .01 < .01 < .01 < .01 Intercept PFOS Production Job (yes/no) Antwerp/Decatur Age. BM ! Cigarettes/day Drinks/day Years Worked Triglycerides* R2= .35 Adj R2 = .34 *HI,!,l,qI,,l,!I Table 75 Multivariable RegressionModel of THBR* by PFOS and Olher Potential Explanatory Valiables for Antwerp and Decatur Male Employee Participants, 2000 Medicai Surveillance Program Parameter 3.589 - 0.003 - 0.006 - 0.090 - 0.0005 - 0.001 - 0.0007 - 0.015 - 0.0002 - 0.003 SE 0.041 0.004 0.010 0.011 0.0008 0.001 t 0.0004 0.005 0.0007 0.008 p value < .0001 .40 , .55 < .0001 .50 .29 .13 .005 .77 .75 , -ge 1o6o{ 12] Partial R2 .03 < .01 . .30 < .01 < .01 < .01 .01 < .01 < .01 Intercept PFOA ProductioJnob (yes/no) Antwerp/Decatur Age BMI CigareUes/day Drinks/day Years Worked Triglycerides* R2 = .35 Adj R2 = .34 *NillUFill 101l "['able76 Multivariable Regression Model of THBR* by PFOA and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 3.589 -0.003 -0.006 - 0.089 - 0.0005 - 0.001 - 0.0007 0.015 - 0.0002 - 0.002 SE 0.041 0.003 0.010 0.011 0.0008 0.001 : 0.0004 0.005 0.0006 0.0O_ p value < .0001 .43 .58 < .0001 .48 .29 .13 .004 .71 .76 L 3M Coml)any PageI07 of' |21 PartiaRlz .04 < .01 " .28 < .01 < .01 < .01 .01 < .01 < .01 " Intercept PFOS PFOA Produclion Job (yes/no) Antwerp/Decatur Age BMI Cigaretles/day Drinks/day Years Worked Triglycerides* R_= .35 Adj R2= .34 *_.tural log "Fable 77 Multivariablc Regression Model of THBR* by PFOS and PFOA and Olher Potcntial Explanatory Variables for Antwclp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 3.589 - 0.003 - 0.002 - 0.005 - 0.088 - 0.0006 - 0.001 - 0.0007 0.015 - 0.0002 - 0.002 SE 0.041 0.005 0.004 0.011 0.011 0.0008 0.001 0.0004 0.005 0.0007 0.008 p value < .0001 ' .58 , .64 .66 < .0001 .47 .28 .13 .004 .78 .79 3M Company Page 108 of 121 Partial Rz .03 .02 < .Ol .28 < .01 < .01 <.01 .01 < .01 < .01 ,. Intercept TOF Production Job (yes/no) Anlwelp/Decalur Age BM! Cigarettes/day Drinks/day Years Worked Triglycerides* R2= .35 Adj R2= .34 *Natural log Table 78 Mullivariable Regression Model of THBR* by TOF and Olher Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 3.589 - 0.003 - 0.004 - 0.088 - 0.0006 - 0.001 - 0.0007 0.015 - 0.0002 - 0.002 SE 0.041 0.003 0.011 0.011 0.0008 0.001 0.0004 0.005 0.0007 0.008 p value < .0001 .29 , .69 < .0001 .45 .28 .13 .004 .77 .80 3MCompany Page 109of 121 Partial R2 .05 < .01 .28 < .01 < .01 < .01 .0l < .01 < .01 Intercept PPOS Production Job (yes/no) Antwerp/Decatur Age. BMI Cigarettes/day DJinks/day Yeli0'tlWorked '1_llJl)o',.IIol_,_* III -' Iit *Natural og Table 79 Multiva0"i:d_lcl_cgn'ession Model of 171"i*by PFOS and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program 0 o Parameter 1.239 - 0.006 - 0.009 - 0.078 - 0.003 - 0.0001 0.0002 - 0.008 O.0flI SE 0.085 0.009 0.021 0.022 0.002 0.002 0.0009 0.011 0.001 p value < .0001 .45 , .65 .0004 .03 .96 .82 .44 .41 II I1! .I II IIItl Ill Partial R z .01 < .01 .07 .02 < .01 < .01 < .01 < .o I III r*- Intercept PFOA Production Job (yes/no) "Antwerp/Decatur Age BMI Cigarettes/day Drinks/day YearsWorked Triglyeerid_* R2 - .10 Adj R2 -- .08 *Natural log t _. Table 80 Mullivariable Regression Model of FTI* by PFOA and Other Potential Explanatory Variables for Antwerp and DecaturMale Employee Participants, 2000 Medical Surveillance Program Parameter 1.239 - 0.002 - 0.012 - 0.079 - 0.003 - 0.00007 0.0002 - 0.008 0.001 - 0.023 SE 0.085 0.007 0.021 0.023 0.002 0.002 0.0009 ' 0.011 0.00 ! 0.016 p value < .0001 .77 .56 .0006 .03 .98 .80 .44 A7 .15 3M Company Page 111 of 121 PartialR2 .01 < .01 .07 .02 < .01 < .01 < .01 < .01 < .01 Intercept PFOS PI'OA ProductionJob (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R_=.10 Adj R2 = .08 *Natural log Table 8 I MultivariableRegression Model of FTI* by PFOS and PFOA and Other PotentialExplanatoryVariables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 1.239 - 0.006 0.0002 - 0.010 - 0.079 - 0.003 - 0.0001 0.0002 - 0.008 0.001 - 0.022 SE : 0.085 0.009 0.008 0.022 0.023 0.002 0.002 0.0009 0.011 0.00 w 0.016 p value < .0001 .49 , .98 .66 .0007 .03 .96 .82 .44 .41 .17 JM COml_mty P+'gc 112 of 121 Partial R 2 .01 < .01 < .01 .06 .02 .02 < .01 < .01 .002 .004 Intercept TOF Produclion Job (yes/no) Anlweq_)ecatur Age BM ! Cigarettes/day Drinks/day Years Worked Triglycerides* .i R2 = .10 ^djR'=.0S *Natural log " Table 82 Multivariable Regression Model of FI'I* by TOF and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants. 2(X}0 Medical Surveillance Program Parameter 1.239 -0.003 - 0.010 - 0.078 - 0.003 - 0.0001 0.0002 - 0.008 0.001 - 0.022 SE 0.085 0.006 0.022 0.023 i 0.002 0.002 0.0009 0.011 0.001 0.016 p value < .0001 ' .62 , .63 .0007 .03 .97 .81 .44 .44 .16 3M Company Page I13of121 Partial R2 .01 < .01 .06 .02 < .01 < .01 < .01 < .01 < .01 f ...... / / Inlerccpl PFOS Produclion Job (yes/no) Anlwerp/Decatur Age BMI ," CigareUcs/day Drinks/day Years Worked Triglycerides* R2 = .12 ^,!j =-l.0 *Nalu,al og Table 83 Mullivariable Regression Model of T3* by PFOS and Olhcr Potential Explanatory Variables for Antwerp and Decatur Male Employe_ Participants, 2(_)0 Medical Surveillance Program Parameler 4.702 0.015 0.022 - 0.099 - 0.002 0.005 0.003 - 0.027 - 0.0006 0.020 " SE : 0.074 0.007 0.018 0.019 0.001 0.002 0.0008 0.009 0.001 0.014 p value < .0001 .04 .23 < .0001 .24 .02 .001 .004 .60 .15 A Company Page 114 of 121 Partial R2 .01 .01 .03 < .01 .02 .02 .02 < .01 < .01 Intercept PFOA ProductionJob (yes/no) Antwerp/Decatur Age BMI : Cigarettes/day Drinks/day Years Worked Triglyceddes* R2 = .13 Adj R2 = .11 *Natural log Table 84 Multivadable Regression Model of T3* by PFOA andOtherPotentialExplanatoryVariables forAntwerpandDecaturMale EmployeeParticipants, 2000MedicalSurveillanceProgram Parameter 4.702 0.016 0.015 - 0.109 - 0.001 0.005 0,003 - 0.028 - 0.0004 0.018 SE 0.073 ? 0.006 0.019 0.020 0.001 0.002 ". 0.0008 0.009 0.001 : 0.014 p value < .0001 .01 , .41 < .0001 .33 .02 .001 .003 .70 .20 Company Page llSof 121 Partial R2 .02 < .01 .03 < .01 .02 ' .02 .02 < .01 < .01 Intcrcpl " PFOS PFOA Production Job (yes/no) Antwerp/Decatur Age BM! Cigarettes/day Drinks/day Years Worked Triglycerides* Itj = .13 Adj R2 =. 11 *Naturallog "Fable85 Multivariable Regression Model ofT3* by PFOS and PFOA and Other PotefltialExplanatory Variables for Antwerp and Decatur Male Employ_ Pmicipants, 2000 Medical Survcillanco Program Parsmetcr 4.703 0.009 0.013 0.012 - 0.109 - 0.001 0.005 0.003 - 0.028 - 0.0006 0.017 SE 0.073 0.008 0.007 0.019 0.020 0.001 0.002 0.0008 0.009 0.001 0.014 p value < .0001 .29 .05 .54 < .0001 .35 .01 .001 .003 .59 .23 i Company Pagel16of 121 Partial R2 .0I < .01 <.01 .03 < .01 .02 .02 .02 < .01 < .01 / Intercept TOF Production Job (yes/no) Antwerp/Decatur Age BMI Cigarettes/day Drinks/day Years Worked Triglycerides* R2 = .13 Adj R2=.1 r *Natural log : Table 86 Multivariable Regression Model of'l"3* by TOF and Other Potential Explanatory Variables for Antwerp and Decatur Male Employee Participants, 2000 Medical Surveillance Program Parameter 4.702 0.014 i SE 0.073 0.005 p value < .0001 .004 0.011 =0.109 a - 0.001 0.005 0.019 0.020 0.001 0.002 , .54 < .0001 .35 .01 0.003 " 0.0007 .001 - 0.028 0.009 .003 - 0.0007 0.001 . .56 0.017 0.014 .22 .._ Company Page l 17of !21 Partial R2 .02 < .01 .03 < .01 .02 .02 .02 < .01 < .01 3M Company , ?:2_e If8 of 121 / Figure 1. Linear Regression Model of Triglgycerides* by PFOA* for Antwen_ Male Emi_ioyees, 2000 Medical Surveillance Program J |3 _ / iI _ _11 Q =" " ::1 --- .. I " I ' I " I " I " I ' i ' I I" ' ... 'l_ Summary of Fit /...... RSquare _o .' o 0.029 .... Analysis of Variance Source Model Error C Total DF Sum of Squares Mean Square I 1.863 1..863 204 61.193 0299 205 63.056 F Ratio 6.211 Prob>F 0.014 Parameter Estimates Term Intercept In PFOA Estimate 4.695 0.073 Std Error 0.042 0.029 t Ratio Pmb>lt! 111.43 <.0001 2.49 0.014 *naaa_ log -o 3M Company Page 119 of 121 / ' Figure 2. Linear Regression of Tdglyceddes* by PFOA" for Decatur Male Employees, 2000 Medical Surveillance Progri_m /) 6._- 1.3", s.s- o I, -= " "-",l ,, ,@_'Ib_ _. 1.I" :' II ].*_ ' * = = = , = I ' , ' J -_.) 4.0 -J.O -Z.O -_) .0 1.O Z.O ,T._ ,.0 S.O Summary of Rt / RSquare 0.028 .... "'" Analysis of Variance. Source Model Error C Total DF Sum of Squares Mean Square I 2.164 2.164 213 73.969 0.347 214 76.133 F:Ratio 6.232 Prob>F 0.0t3 Parameter Estimates Term Intercept In PFOA Estimate Std Error t Ratio Prob>Jtl 5.052 0.041 122.27 <0001 0.098 0.039 2.50 0.013 naturallog 3M Company "_zg_e 120of t21 . Figure 3. Linear Regression of Tdglyceddes* by PFOA" for Antwerp an_" _- ..... Female Employees, 2000 Medical Surveillance Program 1) i ........ ( N 5.) t. % t.I -_'d .o. 41 I) I'l'.',.i.,.i,l.i'l. -_i -! -.l -2 -1 I 1 2 ] t. Summary of Fit P-,Squa_ --AnalysofiVsariance 0.078 Source Model Error C Total DF Sum of Squares Mean Square I 2.519 2.519 95 29.877 0.314 96 32396 Parameter Estimates F Ratio: 8.0I Prob>F O.(X_ Term Intercept In PFOA Estimate Std Error t Ratio Prob>ltl 4.690 0.081 58.14 <.0001 0.091 0.032 2.83 0.006 *mmral Iog i -o . 3M Company , Page 121 of 121. (" , Figure 4. Linear Regression of Tdglycerides* by P..FOA*for Cottage Grove Male Employees, 2000 Medical Surveillance Program .. 7) S.i " t.) " _.i- . = ". " " gO :" . I,..".,'_ -....,,--- ,.t " "" " " = I_ " I " I "I'I'I" I'I'I'I' -S -! -I -t -I I I t ] ! S _KU Summaw of Fit RSquare 0.008 _,. L Analysis of Variance Source Model E.n'or C Total DF Sum of Squares Mean Square 1 0.452 0.452 129 54.251 0.421 130 54.704 F Ratio 1.076 Prob>F 0_302 Term Intercept In PFOA Parameter Estimates Estimate 5.022 0.032 Std Error 0.057 0.031 t Ratio Prob>ltl 88.04 <.000I 1.04 0.302 log L *.