Document dn8nQJYd4wQo0vaGy3R7LpMRq

Taft/ Taft Stettinius & Hollister LLP 425 Walnut Street, Suite 1800 / Cincinnati, OH 45202-3957/ Tel : 513 .381 .2838/ Fax : 513 .381 .0_205 /www.taftlaw.com Cincinnati / Cleveland / Columbus / Dayton / Indianapolis / Northern Kentucky / Phoenix / Beijing ROBERT A. BILOTT 513.357 .9638 bilott@taftfaw.com IIIIIIIIIII!I~ I ; I ?. i! tE iii 1,11 i : --..1 March 6, 2009 ~f~ .rc73oj_o-Ot37V4=4 - 3M TSCA Confidential Business Information Center (7407M) c.,.: EPA East - Room 6428, Attn : Section 8(e) & FYI U .S. Environmental Protection Agency 1200 Pennsylvania Avenue, NW Washington, DC 20460-0001 Re: Submission To TSCA 8(e)/FYI Database Re: PFOA/PFOS To TSCA 8(e)/FYI Database: We are hereby providing the following information for inclusion in the TSCA 8(e)/ FYI databases with respect to PFOA/PFOS : 1 . Costa, G., et al ., "Thirty Years of Medical Surveillance in Perfluorooctanoic Acid Production Workers," 51 J .O.E .M. 1-9 (March 2009); and 2. Joensen, U .N ., et al., "Do Perfluoroalkyl Compounds Impair Human Semen Quality?," Environ. Health Persp. (doi: 10 .1289/ehp.0800517) (online March 2, 2009). RAB :mdm Enclosure 11361469 .1 . . I) i i) 1 i , . J _ -I w 3 r ~gao 0 JOEM " Volume 51, Number 3, March 2009 p. 2 1 Thirty Years of Medical Surveillance in Perfluooctanoic Acid Production Workers Giovanni Costa, MD erfluorooctanoic acid (PFOA, Samantha Sartori, BS Dario Consonni, MD C7F15COOH, CAS No . 335-67-1) is a fully fluorinated carboxylic acid, P widely used as emulsifier in flu- Objective: To report health outcomes of 30 years (1978-2007) of medical surveillance ofworkers engaged in a per,fluooctanoic acid (PFOA) production plant. Methods: Fifty-three males workers (20 to 63 years) were Submitted every year to medical examination and blood chemical chemistry tests, and serum POA dosage. Results: In the latest survey PFOA serum levels ranged from 0.20 to 47.04 t.t,g/mL in currently exposed workers, and from 0.53 to 18.66 l.eg/mL in those formerly exposed. No clinical evidence (fany vPecific trouble or disease has been recorded over the 30years, and all the biochemical parameters, including live; kidney and hormonal fiunctions, turned out to be within the reference ranges, but a significant association of total cholesterol and uric acid with and PFOA serum level was evidenced. Conelusions: A probable interference ofPFOA on intermediate metabolism deseruesfurther investigations. (J Occup Environ Med. 2009,51 :000-000) oropolymer polymerization . It is soluble in water, where it forms a mixture with its dissociated perfluorooctanoate anion (C7F,SCOO-); it is absorbed by oral, dermal and inhalation routes, and mainly bound to serum albumin. It is mostly found in liver, serum and kidney, and is excreted; not metabolized, mainly in urines, with large differences in elimination speeds between species and sexes. It is a mild irritant for skin and mucoses, it is not a sensitizer and has neither teratogenic nor fetotoxic effects, but it delays sexual maturation in animals.--5 It is not a genotoxic, but in two chronic studies in rats it has been reported to increase the incidence of tumors (adenomas) of liver, pancre- atic acinar cells, and testicular Ley- dig ce11S,6~' like other peroxisome proliferators, as it acts mainly as a peroxisome proliferator-activated receptor alpha (PPARa) agonist.s~9 In recent years, many studies have been carried out with the aim of better understanding the toxico- kinetics and mode of actions, al- though with not clear and definite conclusions . 10-17 From the Department of Occupational and Environmental Health (Mr Costa, Mr Sartori), Universitv of Milano, Italy; Unit of Epidemiology (Mr Consonni). 1RCCS Maggiore Hospital, Mangiagalli and Regina Elena Foundation, Milano, Italy ; and Institute of Medical Statistics (Mr Sartori) and Biometrics "G .A . Maccacaro," University of Milano, Italy. Address correspondence to: Prof. Giovanni Costa, Department of Occupational and Environmental Health, University of Milano . Clinica del Lavoro "L . Devoto," Via S. Barnaba 8, 20122 Milano: E-mail : giovanni-costa@unimi .it . Copyright 2009 by American College of Occupational and Environmental Medicine D:011 10.1097/aOM.0b013e3181965d80 Its half-life of elimination in serum markedly differs between species and sexes, being 1 to 9 days in rats, 18 days in mice, 20 to 40 days in dogs, 20 to 32 days in monkeys, and 3.8 years in humans .' g-2 Hence, its biopersistence in the human blood, its detection in the general population21-24 and its presence in several biota and animals all over the p. 3 2 Medical Surveillance of Workers Exposed to PFOA " Costa et al worldZS'' are a concern for possible chronic effects.`b~2y Epidemiological studies on large cohorts of occupationally exposed workers, mainly of US plants (3M and DuPont) in the period 19472002, could not find any association between PFOA exposure and significant adverse health effects in terms of morbidity and mortality.3-3'' For mortality in particular, neither Gilliland and Mande13' nor Alexander'12 and Leonard et al'3 found excess of death for cancers of any type in quite large 3M and DuPont cohorts. Also the slight excess of prostate cancer signaled by Gilliland and Mandel,3' and of bladder cancer found in the first study by Alexander,32 were not confirmed by subsequent more specific surveys both in 3M3 and DuPont33 workers. However, some contrasting and somewhat inconsistent findings, mainly related to lipid metabolism, have been reported as concerns perturbation of biochemical variables . The first study, made by Ubel et al ; in 1980 in about 300 workers of 3M Cottage Grove plant, having total serum organic fluorine (90% PFOA) up to 71 Rg/mL, did not report any significant deviations from the normal values as concerns hematology, liver enzymes, total cholesterol, glucose, and uric acid . In a subsequent study on 115 workers of the same plant,?5 after adjusting for age, smoking, alcohol intake and BMI in a multivariate regression model, a total serum fluorine level above 10 ppm was associated with slight increases in hepatic enzymes through an interaction with obesity. A slight interference with HDL cholesterol was also observed, when in association with alcohol consumption. In other two cross-sectional studies on 191 3M production workers, ;6 serum PFOA was not significantly associated with any of the 1 I measured hormones, except for a 10% increase in mean estradiol level among employees having the highest levels of serum PFOA (?30 ppm), although this association was con- founded by body mass index . Extending the survey in the same production plant to 265 workers for a longer period,3' a negative associa- tion was recorded between serum PFOA and plasma Cholecystokinin (although inside the normal range), and no effect on hepatic enzymes, bilirubin, triclycerides, HDL, and LDL cholesterol . However, in a subsequent survey concerning 518 workers of both sexes from two US and European plants,38 the same authors found a modest positive association between PFOA and cholesterol and triglycerides in both cross sectional and longitudinal (over a 6 year time period) analysis, after adjusting for potential confounders, such as age, BMI, smoking, alcohol, and location . More recently, Olsen and Zobel39 have re-examined the same parameters in 506 3M workers from three US and European 3M plants, by multiple and logistic regressions and analysis of covariance: serum PFOA concentrations were negatively associated with HDL and bilirubin, and positively with triglycerides, after adjusting for possible confounders. Very recently two reports, one longitudinal and another cross-sectional, on DuPont production workers have been published. In the retrospective study of 454 production workers, with repeated measurements concerning liver function and lipids over a period of 25 years, 40 a positive association between serum PFOA and total cholesterol and AST was found, whereas a negative association with bilirubin, after adjustment for age, BMI, gender, and decade of hire, but without control for lipid lowering medications. In the other cross-sectional study on 1025 active workers,4' including hematology, liver and renal function, lipid and purine metabolism, and hormones, a modest but statistically significant positive relationship between PFOA and total cholesterol, LDL, VDLD, GGT and uric acid was recorded, after excluding people taking lipid lowering medications. The present study is aimed at providing a further epidemiological contribution to the understanding of possible effects on humans . The data refer to the workers of a chemical plant (Miteni, Trissino, Italy), where PFOA has been produced by electrochemical fluorination since 1968 . Since 1978 all the workers engaged in PFOA production have been submitted every year to the medical surveillance program by an expert Occupational Health Physician, according to the Italian legislation concerning health and safety at work . Materials and Methods Subjects As a whole, 53 workers, all males, engaged in the PFOA production department have been examined every year from 1978 to 2007, for a total of 919 persons-years. Subjects' age ranged from 20 to 63 years and length of work exposure varied from 0.5 to 32 .5 years in the period . At the latest PFOA biomonitoring, in 2007, 37 were active workers in the PFOA department, whereas 16 were no longer exposed being retired or transferred to other departments in the meantime . All the other male workers, 12 executive clerks and 95 blue collars from the other departments of the company, who have never been exposed to PFOA but submitted to the same periodical medical surveillance, were considered as "controls." Table 1 summarizes the main characteristics of the examined workers. Methods All the workers (exposed and not exposed) have been submitted every year to physical examination, including recording of blood pressure, height and weight, and to blood chemical chemistry tests, including hematology (Ht, Hb, WBC, RBC, PLTS), liver (albumin, a,-a2-(3-'Y p. 4 JOEM " Volume 51, Number 3, March 2009 globulins, bilirubin, AST, ALT, ALP GGT) and renal (urea nitrogen, cre- atinine) functions, glucose and lipid metabolism (total and HDL Choles- terol, triglycerides), and uric acid . All the variables have been measured over the years by autoanalysers (SMA) in the same laboratory of the nearby regional public general hos- pital, undergoing a periodical quality control program. In addition, in the workers of the PFOA department, also Apo-A and Apo-B lipoproteins, protein C reactive, immunoglobulins IgA-IgG- IgM, Testosterone, Estradiol, FSH, TSH, FT3, FT4, and Prostatic Serum Antigen (PSA) were tested in blood in 2002 and 2006. The biological monitoring of se- rum PFOA started in April 2000 and was repeated in May 2001, Decem- ber 2001, June 2002, September 2003, September 2004, May 2006, and May 2007 . It included not only the workers of the PFOA department currently exposed in those years, but also those formerly exposed who had been transferred to other departments or had retired in the meantime, as well as some non-exposed workers (clerks), taken as controls. PFOA serum levels were determined by HPLC-Electrospray-Tandem Mass Spectrometry42 in two internationally accredited laboratories : the first one carried out the analysis in 2000, 2001, and 2002, the second one in the following years. A double blind comparison between the two laboratories carried out in 2003 gave a correlation index equal to 0.90. In 2000 and 2001, the method had an upper scale detection limit fixed at 45 .5 Rg/mL, thereafter no detection limit was present. Statistical Analysis ANOVA, t test and multiple linear regression models were used to evaluate the relationships between PFOA exposure and hematochemical effects while adjusting for relevant covariates . Natural log transformation of the dependent variables was performed where appropriate . General- TABLE 1 Demographic Characteristics of the Workers Examined PFOA Currently Workers PFOA Previously Exposed Workers No Sex Age (mean and SD)' Years of work (mean and SD)' Years elapsed since PFOA exposure (mean and SD)' Person-years of observation BMI (mean and SD) Smoking (%) Alcohol consumption (%) 37 All males 41 .6 (8 .31) 14 .3 (9 .1) 487 .2 26 .2 (2 .9) 38 .2 73 .5 16 All males 52 .0 (8.7) 16 .3 (9 .9) 9.5 (8 .3) 431 .8 25.7 (3 .1) 47.6 71 .4 'At the latest medical check. Control Group 107 All males 41 .9 (9.2) 14 .0 (9.3) 1499 .3 25 .1 (2 .8) 36.5 62.6 ized Estimating Equations (GEE) models with an exchangeable correlation matrix were fitted when analyzing repeated measurements and relationships between PFOA serum level and the different (continuous or dichotomous) health outcomes, while accounting for within-subject correlations . ; In the regression models the following covariates were included : age, years of exposure, year of PFOA sampling, BMI, smoking, and alcohol consumption. Stata 1e software was used to perform all the statistical analyses . TABLE 2 PFOA Serum Level (i.Lg/mL), Recorded in the Year 2007, in Currently and Formerly Exposed Workers Currently Formerly Exposed Exposed No. Min 25th percentile Median Mean arithmetic Mean geometric 75th percentile Max Standard deviation 39 0 .20 2_25 5.71 12.93 4 .02 23.55 47 .04 14 .43 11 0 .53 2.61 4.43 6.81 3.76 9.24 18.66 6.06 Results From the clinical point of view, all the workers were in good health and no excess of morbidity for any disease has been recorded over the years. In the latest survey, carried out in May 2007, the PFOA serum levels ranged from 0.20 to 47 .04 p,g/mL (ppm) in the currently exposed workers, and from 0.53 to 18 .66 Wg/mL in the formerly exposed workers (Table 2) : in the seven non-exposed subjects (clerks/staff) examined in 2006, the PFOA serum levels ranged from 0.05 to 0.181 Wg/mL. The highest level recorded was 91 .9 Wg/mL in 2002 . Table 3 shows the PFOA serum levels recorded over the years in the production workers. As above mentioned, in 2000 and 2001 the method had an upper scale detection limit fixed at 45 .5 p,g/mL ; consequently, in order to compare the data over the years the maximum level in 2000 and 2001 was set by default as the highest level recorded in 2002, when the method had no upper scale detection limit. A significant decrease of both peak and mean levels was recorded after 2002, following a renovation of the plant carried out in that year, including a partial automation of the process, and the adoption of more strict working procedures and use of suitable protective devices. In the 27 currently production workers who had their serum PFOA measured in all the four last occasions, median value decreased from 13 .6 p,g/mL in 2002 to 7.1 pg/mL in 2007 (-47 .8%), mean value decreased from 22 .2 p,g/mL to 14 .0 La.g/mL (-37 .0%), and peak level decreased from 86.3 wg/mL to 47 .0 wg/mL (-45 .5l0), respectively . All the biochemical parameters, turned out to be on average within p. 5 4 Medical Surveillance of Workers Exposed to PFOA " Costa et al the laboratory reference ranges, while some subjects had some parameters above the upper reference values, as shown in Table 4, which refers to the results of the latest check carried out in 2007 . As compared with the control group some differences, both in terms of mean TABLE 3 Serum PFOA Levels No. Min 25th percentile Median Mean arithmetic Mean geometric 75th percentile Max Standard deviation (Wg/mL) 2000 25 1 .54 5.53 11 .92 18 .8 11 .7 32 .0 86 .3' 20 .0 Recorded 2001 42 0.73 4.35 11 .07 19.7 10.2 19.72 91 .9' 23.6 in Production Workers Over the 2002 2003 2004 2006 46 0.34 4.57 10 .15 19 .3 9 .3 20.80 91 .9 23.0 41 0.38 4 .11 6.25 13 .7 6 .9 14 .20 74 .7 16 .6 34 0.54 2.84 6.82 11 .4 6 .5 18 .97 46 .3 12 .0 49 0.54 2.36 5 .27 10.8 5 .8 16 .31 41 .9 11 .8 Year 2007 50 0 .20 2.18 3.89 11 .6 5 .4 18 .66 47.0 13.3 'Upper scale detection limit at 45 .5 wg/mL (discussed in text) . values and number of persons above the upper references limits, were noted for some parameters such as uric acid, cholesterol, triglycerides, and liver enzymes. As the crude correlations between PFOA and the biochemical variables, recorded in the last 7 years, gave some indication on a possible association between PFOA serum levels and some lipids and uric acid (data not shown), three more in-depth analyses in this respect were performed . The aim was clarifying this possible association by controlling the most important confounding factors, and after preliminary exclusion TABLE 4 Haematology: Results of the Last Check in the Currently Exposed Workers and Control Group Exposed Control Group Glucose (mg/dL) Urea nitrogen (mg/dL) Creatinine (mg/dL) Uric acid (mg/dL) Cholesterol total (mg/dL) Cholesterol HDL (mg/dL) Triglycerides (mg/dL) Bilirubin total (mg/dL) Bilirubin conjugated (mg/dL) AST (U/L) ALT (U/L) yGT (U/Lj ALP (U/L) Proteins total (g/dQ Albumin (%) a1 globulins (%) a2 globulins (%) (3 globulins (%) -y globulins (%) WBC (X 109/L) RBC (x10'2/L) Hb (g/dL) Ht (%) Platelets (x 109/L) Protein C reactive (mg/L) Apo A (g/L) Apo B (g/L) IgG (g/L) IgA (g/L) IgM ON PSA (ng/mL) Testosterone (ng/mL) Estradiol (pg/mL) TSH (UI/mL) FT3 (pg/mL) Ff4 (ng/dL) Reference Values Min-Max 70-110 7-22 0.80-1 .30 3.5-7.2 191-240 >39 30-180 <1 .0 <30 <50 <50 8-61 40-129 6.4-8 .2 52-67 3-7 5-9 5-12 10-22 4 .0-10.9 4 .Fr5 .9 13 .5-17.5 41-53 140-440 0.0-5 .0 1 .10-2.05 0.55-1 .40 7 .0-16 .0 0.70-4 .0 0.40-2.30 <4 .0 0.1-8 .0 10-50 0.27-4 .20 2.0-4 .4 0.80-1 .70 Mean (SD) 92 .7 (8.7) 15 .2 (3 .1) 0.92 (0 .12) 6.2(l .1) 239.7 (50.9)' 55 .9 (14.7) 178.9 (118.6) 0.58 (0.22) 0.15 (0.05) 32 .4 (8.8) 47 .8 (22.3) 53 .8 (46.7) 71 .8 (22.8) 7.71 (0.50) 61 .6 (3.0) 4 .1 (0.9) 8 .9 (1 .1) 5 .9 (0.9) 14 .7 (2 .6) 7 .02 (2 .13) 5 .06 (0 .37) 15 .6 (0.88)1 45 .9 (2.2) 241 .3 (55.0) 2.59 (4 .21) 1 .43 (0 .24) 1 .25 (0 .31) 10.6 (2 .4) 3.1 (1 .2) 1 .2 (0.76) 0.80 (0 .94) 5.88 (0 .57) 30.8(l .3) 1 .99 (5 .48) 3.62 (0.67) 1 .31 (13.5) % Outside Reference Range 2.6 0 0 17.9 35.9 0 7 .7 0 2.6 2.6 17 .9 28 .2 0 1 0 0 12 .8 0 0 2 .6 0 2 .6 0 0 5.1 5 .1 5 .1 2.6 12 .8 5.1 2.6 5.1 2.6 2.6 2.6 2.6 Mean (SD) 89 .3 (16.5) 15 .2 (3 .0) 0.93 (0 .11) 5.7(l .2) 213.5 (39.7) 56.7 (12.3) 146.3 (97.9) 0.67 (0 .27) 0.18 (0 .11) 29 .7 (10.7) 40.6 (21 .8) 44 .5 (41 .4) 67 .7 (19.6) 7.81 (0.56) 62 .3 (3 .9) 5.5(g .4) 8.8 (1 .4) 5.8 (0 .5) 14 .7 (2.3) 6 .96(l .88) 5.11 (0 .39) 15 .7 (0 .86) 46 .0 (2 .4) 237.1 (54.4) % Outside Reference Range 6 .5 0 .9 0 4.7 26.2 6.6 27.1 10 .5 7 .6 4.8 26 .2 17.9 0 0 0 0.9 14 0 0 5 .6 1 .8 0 .9 0 0 p. 6 JOEM " Volume 51, Number 3, March 2009 of people under treatment of primary hyperlipidemias and with a history of chronic hepatitis. In the first analysis, 34 currently exposed workers were matched with 34 workers from the other depart- ments with the same age (-~ 1 year) and job seniority ( 1 year) in the company, never exposed to PFOA, having the same working hours (shift work or daywork), and being resident in the same area with similar housing and living conditions. Table 5 shows that the two groups signifi- cantly differ only for total choles- terol and uric acid mean levels, being slightly higher in workers exposed to PFOA . In the second analysis, the same 34 currently exposed workers were compared with all the other 107 male workers of the company. The multi- ple regression analysis, adjusted by age, job seniority, body mass index, smoking and alcohol consumption, confirmed a significant effect of PFOA exposure as concerns an increase in mean total cholesterol and uric acid (Table 6) . The third analysis considered the 56 subjects (currently, formerly and never exposed) who had serum PFOA assessed concurrently with the biochemical parameters in the last 7 years. Multivariate GEE models (including age, job seniority, body mass index, alcohol consump- tion, and years of observation as potential confounders) show that total cholesterol and uric acid levels were weakly but significantly corre- lated to serum PFOA (Table 7) as well as some liver enzymes and alpha2 globulins; however, total bilirubin appears to be inversely related to PFOA . As expected, also body mass index was found to be significantly correlated to total (Coef. = 0.028, Y < 0.01) and HDL (Coef. = - .029, P < 0.001) cholesterol, triglycerides (Coef. = 0.065, P < 0.001) and uric acid (Coef. = 0.022 . P < 0.001), as well as with sotne liver enzymes (ALT, ALP), suggesting a possible interaction, like in previous surveys." TABLE 5 Comparison Between 34 Exposed and Non-Exposed Workers, Matched Work Seniority, Day/Shiftwork, and Living Conditions Non-Exposed Exposed t Test No . BMI Glucose (mg/dL) Urea nitrogen (mg/dL) Creatinine (mg/dL) Uric acid (mg/dL) Total cholesterol (mg/dL) HDL cholesterol (mg/dL) Triglycerides (mg/dL) Total bilirubin (mg/dL) AST (U/L) ALT (U/L) GGT (U/L) ALP (U/L) Total proteins (g/dL) Albumins (%) 1 globulins (%) a2 globulins (~) P globulins (%) y globulins (%) WBC (X 109/L) RBC (x10'2/L) Haemoglobin (g/dL) Haematocrit (%) Platelets (x109/L) 34 25 .94 89 .41 14 .82 0.93 5.73 206.4 56.68 150.03 0.58 31 .62 42 .74 48 .52 69 .41 7.72 62.87 3.82 8.76 5.68 14 .44 6 .64 5 .08 15.69 45 .86 233.03 34 26 .16 93 .56 19.47 0.94 6.29 237 .0 57 .82 155 .35 0_55 31 .44 35 .29 43.88 67.65 7 .67 61 .51 3.94 8.90 5.76 14 .47 7 .53 5_00 15 .37 45 .36 238.97 0.34 1 .30 1 .14 0 .19 2 .10 3 .06 0 .21 0.77 0.18 0.41 1 .35 0.05 0.24 0.22 0.99 0.59 0.20 0.49 0.10 1 .86 0.60 0.40 0.74 0.84 By Age, P NS NS NS NS 0 .039 0 .003 NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS TABLE 6 Multiple Regression Analysis Comparing the 34 Exposed Workers (E) With All the Other 107 Workers (ALL) of the Plant Parameter Coef. (E vs ALL) 95% C.1. P Glucose (mg/dL) Urea nitrogen (mg/dL) Creatinine (mg/dL) Uric acid (mg/dL) Total cholesterol (mg/dL) HDL cholesterol (mg/dL) Triglycerides (mg/dL) Total bilirubin (mg/dL) AST (U/L) ALT (U/L) GGT (U/L) ALP (U/L) Total proteins (g/dL) Albumin (%) a1 globulins (r6) a2 globulins (r6) R globulins (%) y globulins (%) WBC (x109/L) RBC (x10'?/L) Haemoglobin (g/dL) Haematocrit (9/0) Platelets (x109/L) 3.53 3.92 0.004 0.50 21 .7 2.42 -.15 -0.09 1 .35 -5 .18 0.32 -0.78 -0.20 -0.73 -1 .82 0.27 -0 .003 -0.53 0.58 -0.08 -0 .27 -0 .51 1 .31 -2.21/9.28 -0.95/8.79 -0 .04/0 .05 0.06/0 .94 6.83/36 .6 -2.30/7.13 -34 .6/34 .3 -0.19/0 .01 -2 .72/5.41 -13.7/3.32 -17 .5/18 .1 -8.51/6.95 -0.57/0 .17 -3 .44/1 .97 -8 .18/4.54 -0 .75/1.28 -0.37/0.36 -2.29/1 .24 -0.19/1 .35 -0 .23/0.07 -0 .60/0.07 -1 .42/0.40 -18 .8/21 .4 NS NS NS 0.027 0.005 NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS p. 7 6 Medical Surveillance of Workers Exposed to PFOA " Costa et al TABLE 7 Results of the Multivariate Analysis (GEE Model) in the 56 Subjects (Currently, Formerly, and Never Exposed) Who Had Serum PFOA Assessed Concurrently With the Biochemical Parameters in the Last Six Yr (Controlled for Age, Job Seniority, Body Mass Index, Alcohol Consumption, and Yr of Observation) Coef. Pfoa 95% C.1 . P Glucose Urea nitrogen Creatinine Uric acid Total cholesterol HDL cholesterol Triglycerides Total bilirubin Conj . bilirubin AST ALT GGT ALP Apo A Apo B IgG IgM Estradiol Prot C reactive PSA Testosterone Total proteins Albumins 1 globulins 2 globulins 0 globulins y globulins WBC RBC Haemoglobin Haematocrit Platelets 0.005 0.017 -0 .011 0.026 0.028 -0 .018 0.055 -0 .080 -0.034 0.038 0.116 0.177 0.057 -0 .001 0.023 -0 .017 0.048 -0 .012 -0 .020 -0 .032 -0.005 -0.001 -0.009 0 .026 0.026 0.011 0.013 0.029 -0.002 0.008 0.122 1 .987 -0.011/0 .020 -0.015/0 .049 -0.025/0.004 0 .001/0.053 0.002/0.055 -0 .047/0.012 -0 .036/0.147 -0 .137/-0.024 -0 .099/0 .031 -0 .003/0 .080 0.054/0 .177 0.076/0278 0.007/0 .107 -0.038/0 .035 -0.041/0_087 -0 .115/-0.080 -0 .093/0 .190 -0 .080/0.057 -0 .268/0 .228 -0 .286/0 .222 -0 .021/0 .011 -0.010/0 .008 -0.017/0 .001 -0.001/0 .053 0.007/0 .045 -0.008/0 .030 -0.005/0.031 -0 .011/0.071 -0 .009/0 .006 -0 .113/0 .129 -0 .132/0 .376 -7 .36/11 .34 NS NS NS <0.05 <0.05 NS NS <0.01 NS NS <0 .01 <0 .01 <0 .05 NS NS NS NS NS NS NS NS NS NS NS <0 .01 NS NS NS NS NS NS NS 60 o -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 2627,2829 30 years after exposure Fig. 1 . PFOA serum level (Il.g/mL) in the 16 formerly exposed workers, who retired or were transferred to other departments . Discussion In our cohort neither clinical evidence of specific disturbances nor health disorders have been recorded over 30 years of medical observations of workers exposed to PFOA, having serum levels ranging from 0.20 to 91 .9 p,g/mL. A significant decrease in PFOA blood levels (-37% in mean level and -45 .5% in peak level) was recorded in the 4 last years after plant renovation and improvement of working conditions . However, in evaluating this trend it is necessary to take into account the long biological half-life of the substance, so that the present blood levels largely reflect the exposure conditions of the previous years. In fact, contrary to the rather rapid rates of elimination reported in laboratory animals, PFOA appears to be slowly eliminated in humans .3'3 Very recently Olsen et al` have determined the serum elimination half-life of PFOA in 26 retired fluorochemicals production workers over a 5 year span ; the arithmetic mean half-life was 3.8 years (95% CI = 3.1 to 4.4) and the geometric mean half-life of 3.5 years (95% CI = 3.0 to 4.1), with a range from 1 .5 to 9.1 year . A similar trend can be found also in our data (Fig . 1) . In the 16 formerly exposed workers, who retired or were transferred to other departments, the decrement of PFOA blood concentration over the last 7 years appears to be more or less steeper depending on both extent . and duration of the past exposure, and years elapsed since leaving the job. Although the data are not complete and homogeneous, in terms of PFOA level, number and years of measurements, time elapsed since exposure, the observed trend in the decrement suggests an estimated half-life of 5.1 (sd 1 .7) years, varying from 2.6 to 9.7 years (geometric mean 4.8). It is noteworthy that even after 20 years from exposure some people still have I to 3 p.g/mL of PFOA in their blood. p. 8 JOEM " Volume 51, Number 3, March 2009 7 As regards biological effects, no significant perturbation of liver and renal functions, immunology, and hormonal secretion (including estradiol, testosterone, thyroid, and prostate hormones) were recorded in our workers, except for a significant interference with lipids (cholesterol) and purines (uric acid) metabolism, beside the influence of BMI. As mentioned in Introduction, some previous cross sectional studies in 3M workers did not find any consistent association between serum PFOA and liver, renal and lipid functions, and hormonal secretion. As concerns liver function and lipids metabolism in particular, Gilliland and Mande135 did not record any significant difference between exposed and non-exposed workers in any biological parameter of liver function, and no significant correlation between total serum fluorine (above 10 wg/mL) and cholesterol, LDL or HDL, after adjusting for age, BMI, alcohol, and smoking. The same negative results were obtained in three successive cross-sectional investigations in workers with serum PFOA up to 114.1 Vtg/mL.3' More recently, in two other 3M groups, including both European and US occupationally exposed workers, Olsen et a1311 found no significant effect on liver function but a positive, although modest, association between serum PFOA (range : 0.01 to 12 .7 wg/mL) and cholesterol as well as triglycerides, both in a crosssectional and in a 6-year longitudinal analysis . Moreover, in a further study in workers with serum PFOA ranging from 0.007 to 92 .03 Rg/mL, Olsen and Zobel ;' recorded a weakly negative association with HDL and total bilirubin, and a positive one with triglycerides . after excluding people taking cholesterol lowering medications; however, these findings were explained by demographic differences across the three locations. However, Sakr et al in a cross sectional study on DuPont workers,4t having serum PFOA from 0.005 to 9.55 ltg/mL, reported a modest, but significant positive correlation between serum PFOA and ,total cholesterol, LDL, and VLDL, whereas in another longitudinal studY40 they found a positive association of serum PFOA (range : 0 to 22 .66 p.g/mL) with total cholesterol and AST, and a negative association with total bilirubin, after adjusting for potential confounders (age, BMI, gender, and decade of hiring). However, it is worth mentioning that serum cholesterol appears to move in the opposite direction than observed in experimental studies in rodents, in which PFOA has a significant hypolipidemic effect in serum, being a strong PPARa agonist.45 No study reported any hypolipidemic effect in exposed workers, whereas three studies including the present one recorded a significant increase in cholesterol levels, besides other nonconsistent increases of triglycerides and liver enzymes. Also in cynomolgus monkeys, after 6 months of oral dosing of PFOA from 3 to 30/20 mg/ kg/die (with serum PFOA ranging from 20 to 467 wg/mL) no decrease of cholesterol, but a significant increase of triglycerides was observed .'a This might be due to different mechanisms of action and/or receptor agonisms, ie, PPARa/y and/or CAR29 in rodents rather than in primates, which deserves fiuther investigation. For example it is well known that there are different expressions and properties of PPAR isoforms (a, (3, and y) in animals and humans"3-15,46-48 as well as there might be different bindings with serum proteins (ie, albumin in animal, CETP) and in renal read.sorption (Oatpl and Oat3),49,5 or interference with hepatic metabolism of fatty acids, ie, via CYP450 . As to the latter, it is worth mentioning that three studies (Olsen and Zabel,39 Sakr et a1,4' and the present one), also recorded a significant negative association between serum PFOA and total bilirubin, that was not documented in rodents and primates. It is worth noting in this regard that perfluorolauric acid proved to cause a marked bilirubin destruction in vitro when added to a pure bacterial enzyme (CYP102 or BM3) specialized in fatty acid metabolism . By analogy, it is possible that a similar effect might take place in humans on interaction of PFOA with a fatty acid metabolizing CYP enzyme.s' Moreover, we recorded a positive significant association between serum PFOA and uric acid, that was also reported by Sakr et al in DuPont workers4' This may be suggestive of some interference in liver purines metabolism, which deserves specific investigations . In our study, we tried to control as many confounding variables as possible, in particular age, BMI, smoking, alcohol consumption, and shift work . The first four factors were taken into consideration in the GEE analysis, whereas the possible interference of shift and night work on metabolic parameters, as evidenced in literature, 52,53 was controlled by comparing exposed workers and controls matched for (shift) work seniority (Table 5) . So, beside some inconsistent findings among the studies, it is probable that PFOA exerts some interference on intermediate metabolism also in humans. Owing to the small number of exposed subjects it is not possible to draw any firm conclusion, though the long time span of the careful medical surveillance may allow to exclude a severe impact of PFOA un human health . However, the possible interference on the intermediate metabolism should not be underestimated and deserves further investigations on mechanisms for action, and as a possible risk factor for metabolic disorders and/or CVD diseases . References 1 . Kennedy GL, Butenhoff JL Jr, Olsen GW, et al. The toxicology of perfluorooctanoate. Crit Rev Toxicol. 2004 ;34: 351-384 . 2. Lau C, Anitole K. Hodes C, Lai D, Pfahles-Hutchens A, Seed J . Perfluoroalkyl acids: a review of monitoring and p. 9 8 Medical Surveillance of Workers Exposed to PFOA " Costa et al toxicological findings . Toxicol Sci. 2007 ; 99 :366-394 . 3. Butenhoff JL, Kennedy GL, Frame SR, O'Connor JC York RG . The reproductive toxicology of ammonium perfluorooctanoate (APFO) in the rat . Toxicology. 2004 ;196:95-I 16. 4. Lau C, Thibodeaux JR, Hanson RG, et al . Effects of perfluorooctanoic acid exposure during pregnancy in the mouse . Toxicol Sci. 2006;90:510-518 . 5 . Abbott BD, Wolf CJ, Schmid JE, et al . Perfluorooctanoic acid (PFOA)-induced developmental toxicity in the mouse is dependent on expression of peroxisome proliferator activated receptor-alpha (PPARa). Toxicol Sci . 2007 ;98 :571-581 . 6. Sibinski L1, Allen JL, Erickson EE. Two Year Oral (diet) Toxicity/Carcinogenicity Study of Fluorochemical FC-143 in Rats . St . Paul, MN : Riker Laboratories, Inc., Expt . No_ 0281CR0012 . U.S . Environmental Protection Agency Administrative Record 226-0437-0440; 1987. 7. Cook JC, Murray SM, Frame SR, et al . Induction of Leydig cell adenomas by ammonium pertluorooctanate: a possible endocrine related mechanism. Toxicol Appl Pharmacol. 1992 ;192 :209-217 . 3 . Biegel LB, Hurtt ME, Frame SR, O'Connor JC, Cook JC. Mechanisms of extrahepatic tumor induction by peroxisome proliferators in male CD rats . Toxicol Sci. 2001 ;60:44-55 . 9. Maloney EK, Waxman DJ . Transactivation of PPARa and PPARy by structur- ally diverse environmental chemicals. Toxicol Appl Pharmacol. 1999 ;161 :209218. 10 . Butenhoff JL, Gaylor DW, Moore 1A, et al . Characterization of risk for general population exposure to perfluorooctanoate. Regul Toxicol Pharrnacol. 2004 ;39: 363-380. 11 . Andersen ME, Clewell H1, Tan YM, Butenhoff1L, Olsen GW . Pharmacokinetic modeling of saturable, renal resotption of perfluoroalkylacids in monkeys-probing the determinants of long plasma halflives. Toxicology. 2006 ;227 :156-164. 12. Klaunig JE, Babich MA, Baetcke KP, et al . PPARalpha agonist-induced rodent tumors : modes of action and human relevance. Crit Rev Toxicol. 2003 ; 33 :655-780. 13 . Palmer CN, Hsu M-H. Griffin Kl, Raucy JL . Johnson EF. Peroxisomc proliferator activated receptor-Of expression in human liver. Mol Pharmacol. 1998-.53:14-22- 14, Cheung C, Akiyama TE, Ward JM, et al . Diminished hepatocellular proliferation in mice humanized for the nuclear receptor peroxisome proliferator-activated re- eeptor alpha. Cancer Res . 2004 ;64 : 3849-3854 . 15 . Morimura K, Cheung C, Ward JM, Redd JK, Gonzalez FJ . Differential susceptibil- ity of mice humanized for peroxisome proliferator-activated receptor alpha to Wy-14,643-induced liver tuntorigenesis. Carcinogenesis . 2006 ;27:1074-1080 . 16 . Rosen MB, Lee JS, Ren H, et al . Toxi- cogenomic dissection of the perfluorooctanoic acid transcript profile in mouse liver: evidence for the involvement of nuclear receptors PPARalpha and CAR . Toxicol Sci . 2008 ;103 :46-56. 17. Kudo N, Kawashima Y. Toxicity and toxicokinetics of perfluorooctanoic acid in humans and animals. J Toxicol Sci . 2003 ;28 :49-57 . 18 . Butenhoff J, Costa G, Elcombe C, et al. Toxicity of ammonium perfluorooctanoate in male cynomolgus monkeys after oral dosing for 6 months . Toxicol Sci. 2002 ;69 :244-257 . 19 . Harada K, Inoue K, Morikawa A, Yoshi- naga T, Saito N, Koizumiet A . Renal clearance of perfluorooctane sulfonate and perfluorooctanoate in humans and their species-specific excretion . Enriron Res . 2005 ;99 :253-261 . 20. Olsen GW, Burris JM, Ehresman DJ, et al . Half-life of serum elimination of per- fluorooctanesulfonate, perfluorohexane- sulfonate, and perfluorooctanoate in retired fluorochemical production work- ers. Environ Health Pernpect . 2007 ;115 : 1298-1305 . 21 . Calafat AM, Kuklenyik Z, Reidy JA, Caudill SP, Tully JS, Needham LL . Se- rum concentrations of 1 I polyfluoroalkyl compounds in the U.S. population : data from the National Health and Nutrition Examination Survey (NHANES) 1999- 2000 . Environ Sci Technol. 2007 ;41 : 2237-2242 . 22 . Emmett EA, Shofer FS, Zhang H, Free- man D, Desai C Shaw LM . Community exposure to perfluorooctanoate : relationships between serum concentrations and exposure sources. J Occup Environ Med. 2006 ;48 :759-770 . 23 . Kannan K, Corsotini S, Falandysz J, et al . Perfluorooctanesulfonate and related fluorochemicals in human blood from several countries. Environ Sci Technol. 2004 ;38 :4489-4495 . 24 . Olsen GW, Huang H-Y, Helzlsouer, KJ, Hansen KJ, Butenhoff JL, Mandel JH . Historical comparison of perfluorooctanesulfonate, perfluorooctanoate, and other tluorochemicals in human blood. Environ Health Perspect. 2005 : 113:539-545 . 25 . Giesy JP, Kannan K. Global distribution of perfluorooctane sulfonate and related perfluorinated compounds in wildlife . Environ Sci Technol . 2001 ;35 :13391342 . 26 . Nakata H, Kannan K, Nasu T, Cho H-S, Sinclair E, Takemura A. Perfluorinated contaminants in sediment and aquatic organisms collected from shallow water and tidal flat areas of the Ariake Sea, Japan: environmental fate of perfJuo- rooctane sulfonate in aquatic ecosystems. Environ Sci Technol. 2006;40:49164921 . 27 . Tao L, Kannan K, Kajiwara N, et al . Perfluorooctanesulfonate and related fluorochemicals in albatrosses, elephant seals, penguins, and polar skuas from the Southern Ocean. Envirort Sci Technol . 2006 ;40 :7642-7648. 28 . US Environmental Protection Agency . PerJluorooctanoic Acid. 2008 . Available at : http ://www .epa.gov/oppt/pfoa . Ac- ces.eed October 3, 2008 . 29. UK Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment. Second Draft Working Paper on the Tolerable Daily Intake for PerJluorooctanoic Acid. 2006 . Available at : http ://cot .gov .uk/pdfsltox-2006- 18.pdf. Accessed October 3, 2008 . 30. Ubel FA, Sorenson SD, Roach DE . Health status of plant workers exposed to fluorochemicals-a preliminary report . Am hid Hyg Assoc J. 1980;41 :584-589 . 31 . Gilliland FD, Mandel JS . Mortality among employees of a perfluorooctanoic acid production plant. J Occup Med. 1993 ;35:950-954 . 32. Alexander BH . Mortality Study of Work- ers Employed at the 3M Cottage Grove Facility . St . Paul, MN : 3M Company; Washington DC : U.S . EPA Public Docket AR-226-1030018 ;2001. 33 . Leonard RC, Kreckmann KH, Sakr CJ, Symons JM . Retrospective cohort mortality study of workers in a polymer production plant including a referencepopulation of regional workers . Ann Epidemiol. 2008 ;18 :15-22 . 34 . Alexander BH, Olsen GW. Bladder cancer in perfluorooctanesulfonyl fluoride manufacturing workers . Ann Epidemiol. 2007 ;17 :471-478 . 35. Gilliland FD, Mandel JS . Serum perfluorooctanoic acid and hepatic enzymes, lipoproteins and cholesterol : a study of occupationally exposed men. Am J Ind Med. 1996 ;29 :560-568 . 36 . Olsen GW, Gilliland FD, Burlew MM, Burris JM, Mandel JS, Mandel JHL. An epidemiologic investigation of reproductive hormones in men with occupational exposure to pefluorooctanoic acid . J Occup Environ Med. 1998 ;40:614-621 . 37 . Olsen GW, Burris JM, Burlew MM, p. 10 JOEM " Volume 51, Number 3, March 2009 Mandel JH . Plasma cholecystokinin and hepatic enzymes, cholesterol and lipoproteins in ammonium pertluorooctansulfonate production workers. Drug Chem Toxicol. 2000;23:603-620. 38 . Olsen GW, Burris JM, Burlew MM, Mandel JH . Epidemiologic assessment of workers serum perfluorooctansulfonate (PFOS) and pefluorooctanoate (PFOA) concentrations and medical surveillance examinations . J Occup Environ Med. 2003 ;45:260-270 . 39. Olsen GW, Zobel LR. Assessment of lipid, hepatic, and thyroid parameters with serum perl7uooctanoate (PFOA) concentrations in fluochemical production workers. ]tit Arch Occup Environ Health . 2007 ;81 :231-246 . 40. Sakr CJ, Leonard RC, Kreckmann KH, Slade MD, Cullen MR . Longitudinal study of serum lipids and liver enzymes in worker with occupational exposure to ammonium perfluooctanoate. J Occup Environ Med. 2007 ;49:872-879 . 41 . Sakr CJ, Kreckmann KH, Green JW, Gilles PJ, Reynolds JL, Leonard RC . Cross-sectional study of lipids and liver enzymes related to a serum biomarker of exposure (ammonium perfluooctanoate or APFO) as part of general health survey in a cohort of occupationally exposed workers . J Occup Environ Med. 2007 ;49: 1086-1096 . 42 . Hansen KJ, Clemen LA, Ellefson ME, Johnson JHO. Compound-specific, quantitative characterization of organic f7uorochemicals in biological matrices. Environ Sci Technol. 2001 ;35:766-770. 43 . Twisk JWR. Applied Longitudinal Data Analysis for Epidemiology. Cambridge: Cambridge University Press; 2003 . 44 . StataCorp. Stata Statistical Software : Release 10. College Station, TX : StataCorp LP ; 2007 . 45 . Xie Y, Yang Q, Nelson BD, DePierre JW . The relationship between liver peroxisome proliferation and adipose tissue atrophy induced by peroxisome proliferator exposure and withdrawal in mice . Biochem Pharnuicol . 2003 ;66:749-756 . 46 . Michalik L, Desvergne B, Dreyer C, DeGavillet M, Laurini RN, Wahl W. PPAR expression and function during vertebrate development. Int J Dev Biol . 2002 ;46:105-114 . 47 . Rosen ED, Spiegelman BM . PPARy: a nuclear regulator of metabolism, differentiation, and growth . J Biol Chem . 2001 ; 276:37731-37734 . 48 . Toth B, Hornung D, Schotz C, Djalali S, 9 Friese K, Jeschke U . Peroxisome proliferatoractivated receptors: new players in the field of reproduction . Am J Reprod Immunol . 2007 ;58 :289-310 . 49 . Katakura M, Kudo N, Tsuda T, Hibino Y, Mitsumoto A, Kawashima Y. Rat organic anion transporter 3 and organic anion transporting polypeptide I mediate perfluorooctanoic acid transport . J Health Sci. 2007 ;53:77-83 . 50 . Kudo N, Katakura M, Sato Y, Kawashima Y . Sex hormone-regulated renal transport of perfluorooctanoic acid . Chem Biol Interact. 2002 ;139 :301-316 . 51 . Pons N, Pipino S, De Matteis F. Interaction of polyhalogenated compounds of appropriate configuration with mammalian or bacterial CYP enzymes . Increased bilirubin and uropofobylinogen oxidation. Biochem Pharmacol . 2003;66 :405-414 . 52 . Karlsson B, Knutsson A, Lindahl B, Alfredsson L. Metabolic disturbances in male workers with rotating threeshift work . Results of the WOLF study. /tit Arch Occup Environ Health. 2003 ; 76 :424-430 . 53 . Biggi N, Consonni D, Galluzzo V, Sogliani M. Costa G. Metabolic syndrome in permanent nightworkers . Chronobiol Int. 2008 ;25 :1-12. p. 11 IZT ehponline.org ENVIRONMENTAL HEALTH PERSPECTIVES Do Perfluoroalicyl Compounds Impair Human Semen Quality? Ulla Nordstrom Joensen, Rossana Bossi, Henrik Leffers, Allan Astrup Jensen, Niels E. Skakkebaek, and Niels Jrargensen doi: 10 .1289/ehp .0800517 (available at http://dx.doi .org/) Online 2 March 2009 md% NIEHS ~ National Institute of ~ Environmental Health Sciences National Institutes of Health U.S. Department of Health and Human Services Do Perfluoroalkyl Compounds Impair Human Semen Quality? Ulla Nordstrom Joensenl, Rossana Bossi2, Henrik Leffers" Allan Astrup Jensen3, Niels E. Skakkebaek', Niels Jorgensen' . 1 University Department of Growth and Reproduction, Rigshospitalet, Copenhagen, Denmark Z National Environmental Research Institute, University of Aarhus, Roskilde, Denmark 3 FORCE Technology, Brondby, Denmark Address of institution where work was done: University Department of Growth and Reproduction Rigshospitalet, section 5064 Blegdamsvej 9 2100 Copenhagen Denmark Corresponding author: Ulla Nordstr6m Joensen, MD University Department of Growth and Reproduction, Rigshospitalet Blegdamsvej 9 2100 Copenhagen Denmark Telephone : +45 3545 5064 Fax : +45 3545 6054 E-Mail: ulla.nordstroem .joensen@rh .regionh .dk p. 13 Page 2 of 26 Acknowledgements/grant support : The authors greatly appreciate the support from the European Union (contract no. QLK4CT- 2002-00603), the Danish Agency for Science, Technology and Innovation (grant nos . 9700833 and 271070678), the Danish Ministry of Health (ISMF) (Grant no. 7-302-02-9/3), the Danish Environmental Protection Agency, and the University of Copenhagen . The authors declare they have no competing financial or non-financial interests . Running title: PFAAs and Human Semen Quality . EHP article descriptor : Reproductive Health Key words: Endocrine disruptors, male reproductive health, perfluoroalkyl compounds, PFAA, PFC, semen quality, sperm morphology, testosterone . Abbreviations : BMI : CI: ESI : FAI: FSH : HPLC: LC-MS-MS : LH : LOD: PFAA : PFC: PFDA : PFDoA PFHpA: PFHxS : PFNA: PFOA : PFOS : PFOSA: PFTrA : PFUnA: SHBG : TDS : body mass index confidence interval electrospray ionization free androgen index follicle stimulating hormone high performance liquid chromatography liquid chromatography-tandem mass spectrometry luteinizing hormone limit of detection perfluoroalkyl acids polyfluorinated compounds perfluorodecanoic acid perfluorododecanoic acid perfluoroheptanoic acid perfluorohexane sulfonic acid perfluorononanoic acid perfluorooctanoic acid perfluorooctane sulfonic acid perfluorooctane sulfonamide perfluorotridecanoic acid perfluoroundecanoic acid sex hormone binding globulin testicular dysgenesis syndrome 2 Outline of section headers Abstract Introduction Materials and methods Study population PFAA analysis Reproductive hormone analysis Semen analysis Statistical analysis Results Levels of PFAAs in serum PFAAs and semen variables PFAAs and reproductive hormones Smoking and BMI Discussion and conclusions References Tables Figure legends Figures p. 15 Page 4 of 26 Abstract Background : Perfluoroalkyl acids (PFAAs) are found globally in wildlife and humans, and are suspected to act as endocrine disruptors. There are no reports of PFAA levels in adult men from Denmark, and no reports of a possible association between semen quality and PFAA exposure . Objectives : To investigate possible associations between PFAAs and testicular function. The hypothesis was that higher PFAA levels would be associated with lower semen quality and lower testosterone levels . Methods : We included 105 Danish men (median age 19 years) from the general population, and analyzed serum samples for levels of 10 different PFAAs and reproductive hormones, and assessed semen quality . Results : Considerable levels of PFOS, PFOA and PFHxS were found in all young men (median 24.5, 4.9 and 6.6 ng/mL, respectively) . Men with high combined levels of PFOS and PFOA had a median of 6.2 million normal spermatozoa in their ejaculate in contrast 15.5 million among men with low PFOS-PFOA (p=0.030). In addition we found non-significant trends with regard to lower sperm concentration, lower total sperm counts and altered pituitarygonadal hormones among men with high PFOS-PFOA levels . Conclusion : High PFAA levels were associated with fewer normal sperms . Thus, high levels of PFAAs may contribute to the otherwise unexplained low semen quality often seen in young men . However, our findings need to be corroborated in larger studies . ' Page 5 of 26 p. 16 ' Introduction Perfluoroalkyl acids (PFAAs) are degradation products of many man-made polyfluorinated compounds (PFCs) used in consumer and industrial products, for example for impregnation of carpets, textiles and paper (Jensen et al . 2008; Jensen and Leffers 2008; Kissa 2001). Studies of environment, wildlife and humans suggest widespread presence and exposure, as well as persistence in the environment and bioaccumulation (Giesy and Kannan 2001 ; Kannan et al. 2004) . For perfluorooctanoic acid (PFOA), pertluorooctane sulfonic acid (PFOS) and perfluorohexane sulfonic acid (PFHxS), three of the most abundant PFAAs, half-lives for humans have been estimated as 3 .8, 5 .4 and 8.5 years, respectively (Olsen et al. 2007). Some studies suggest that men may have higher serum concentrations of PFAAs than women, and younger individuals may have higher levels than older (Calafat et al. 2006) . Thus, young men may have particularly high levels of exposure and may therefore be a group at risk for potential adverse effects of PFAAs. PFAAs can cross the placental barrier and therefore have the potential to affect the fetus. .In humans, levels of PFOS and PFOA in umbilical cord blood have been inversely related to birth weight (Apelberg et al . 2007) . In addition, PFOS, PFOA and PFHxS have been detected in human seminal plasma samples (Guruge et al. 2005). However, data on effects in humans are sparse, and most come from studies of occupationally exposed individuals . These studies have not given conclusive evidence of adverse effects . A recent study, however, measured PFAA levels in early pregnancy found that higher levels of PFOS and PFOA was associated with significantly longer waiting time to pregnancy (Fei et a1 . 2009). Animal studies provide some evidence for adverse reproductive effects on animals exposed as adults or in utero. Exposure of adult male rats to PFOA reduced their testosterone levels and increased their estradiol levels, which may partly explain earlier 5 p. 17 Page 6 of 26 findings of induction of Leydig cell hyperplasia and/or adenomas in the testes of exposed animals (Biegel et al. 1995; Cook et al. 1992). Our objective was to investigate the associations between PFOS, PFOA, PFHxS and other PFAAs and testicular function . Our primary hypothesis was that high concentrations of PFAAs would be associated with low testosterone levels and secondly that high PFAA levels are negatively associated to semen quality variables . Materials and methods Study population . Since 1996, semen quality of young men in Denmark has been surveyed in a cross-sectional study (Andersen et al . 2000; Jorgensen et al. 2002) . All young Danish men must report for military draft, and annually new cohorts of approximately 300 men from the Copenhagen area in Denmark have been included . They each provided one semen sample and had a venous blood sample drawn. Of the 546 men examined in 2003, we selected 105 for the investigation of associations between PFAAs and testicular function. The 105 men included the 53 men (group 1) with the highest testosterone levels (median 31 .8 nmol/l, range 30.1 - 34.8), and the 52 men (group 2) with the lowest testosterone levels (median 14.0 nmol/l, range 10.5 - 15.5). We chose the men examined in 2003, as this was the latest year from which we had completed analyses of reproductive hormones . The median age of men in group 1 was 18,9 years (range 18,2 - 24,6), median age in group 2 was 19,0 years (range 18,2 - 25, 1), and median age for all 105 young men was 19.0 years (range 18.2 - 25 .2) years . Information on ejaculation abstinence period and hour of blood sampling was recorded. All samples of semen and blood were collected between 8.30 a.m. and 1 .15 p.m. Serum was stored at -20C until chemical analysis . The Danish National Committee on Biomedical Research Ethics, Copenhagen Region, approved the research, and all young men gave written informed consent. 6 Page 7 of 26 , p. 18 PFAA analysis. Thawed serum samples were analyzed in January 2008 for ten different perfluorinated chemicals, with carbon chain length from C6 to C13 : PFHxS (perfluorohexane sulfonic acid), PFHpA (perfluoroheptanoic acid), PFOA (perfluorooctanoic acid), PFOS (perfluorooctane sulfonic acid), PFOSA (perfluorooctane sulfonamide), PFNA (perfluorononanoic acid), PFDA (perfluorodecanoic acid), PFUnA (perfluoroundecanoic acid), PFDoA (perfluorododecanoic acid), and PFTrA (perfluorotridecanoic acid) . One mL of serum was spiked with the surrogate standards 13C8PFOA, 13C2-PFDA and 13C4-PFOS and extracted according to the ion pairing method described previously (Hansen et al. 2001) . Matrix-matched standards were prepared by spiking rabbit serum (Sigma Aldrich, Schnelldorf, Germany) with the analytes and the surrogate standards . Blank samples consisted of rabbit serum spiked with only surrogate standards . Standards and blanks were extracted together with each batch of samples . Instrumental analysis was performed by liquid chromatography-tandem mass spectrometry (LC-MS-MS) with electrospray ionization (ESI). The extracts (20 pL, injection volume) were chromatographed on a C 18 Betasil column (2.1 x 50 mm, Thermo Hypesil-Keystone, Bellafonte, PA) using an Agilent 1100 Series HPLC (Agilent Technologies, Palo Alto, CA) . The high performance liquid chromatography (HPLC) was interfaced to a triple quadrupole API 2000 (Sciex, Concorde, Ontario, Canada) equipped with a TurboIon Spray source operating in negative ion mode. Chromatographic conditions and transition MS-MS ions have been described in details previously (Bossi et al . 2005) . The limits of detection (LODs) ranged from 0.1 to 0.5 ng/mL. Reproductive hormone analysis. Thawed serum samples were analyzed for the levels of testosterone, estradiol, SHBG (sex hormone binding globulin), LH (luteinizing hormone), P. 19 Page 8 of 26 FSH (follicle stimulating hormone) and inhibin B as described previously (Paasch et al. 2008). Free androgen index (FAI) was calculated as [testosterone x 100 / SHBG] . Ratios between hormones were calculated by simple division . Semen analysis. Semen volume was assessed by weight, and the sperm concentration by use of a Burker-Turk haemocytometer . Total sperm count was calculated as semen volume x sperm concentration . The percentage of motile spermatozoa (WHO class A+B+C) was assessed on fresh samples . Sperm morphology slides were fixed, Papanicolaou stained and all assessed according to strict criteria (Menkveld et al . 1990) by one trained technician over a period of one week. Further details of the semen analysis can be seen in previously published work (7t6rgensen et al. 2002) . Statistical analysis. Medians and 5-95`h percentiles were used to describe the levels of PFAAs in group 1 and 2. Mann-Whitney U-test was used to compare the groups 1 and 2 with respect to PFAA levels, BMI (body mass index) and smoking status . Samples with values below LOD were set to 0 ng/mL. We used Pearson's correlation coefficients and related p-values to describe correlations between levels of different PFCs . Univariate regression analysis was done for comparison of hormone levels between group I and 2, and to describe associations between PFAAs and hormones or semen variables . Sperm concentration, semen volume and total sperm count were adjusted for the effect of ejaculation abstinence period. Sperm motility was adjusted for time between ejaculation and assessment of motility. Sex hormone concentrations were adjusted for hour of blood sampling. Semen variables and hormone levels and ratios, except sperm morphology and total testosterone, were In transformed to obtain normality of the residuals . Smoking and ' Page 9 of 26 . BMI were tested for confounding effects but were found to be non-significant, and therefore not included in the final analyses. p. 20 We proceeded to analyze for associations between PFAAs and testicular function (hormone levels and semen quality) for the whole group of 105 men . We singled out PFOS and PFOA, and results were calculated as estimated changes in endpoint (reproductive hormones and semen variables) with a change in serum concentration of 1 ng/mL of PFOS, PFOA concentrations, and the summed concentration of PFOS and PFOA. We divided the men into three groups from the combined concentrations of PFOS and PFOA. Each sample was given a quartile score of 1 to 4 for PFOS and PFOA levels separately. Score 1 was given to samples with levels within the lowest quartile, and score 4 within the highest quartile. We then summed the quartile scores for PFOS and PFOA, giving each sample a possible score from 2 to 8. Samples were then divided into 3 quartile groups for PFOS and PFOA combined: "Low PFAA" group (N=29) with summed quartile score from 2 to 3, "intermediate PFAA" (N=48) group with score 4 to 6, and "high PFAA" (N=28) group with score 7 to 8. Analysis for association between quartile group and hormone levels or semen variables was done using univariate regression analysis, adjusted for the abovementioned confounders . Statistical analysis was performed using SPSS statistical software version 16 .0 (SPSS Inc., Chicago, IL, USA) . Results Levels ofPFAAs in serum. The serum levels and number of samples above LOD for all PFAAs are shown for the low- and high testosterone groups and the entire group of 105 men (Table 1). Except for PFOSA, which was detected in only 56 men, there was no significant difference in levels of any PFAA between group 1 and 2. The median PFOS, 9 p. 21 Page 10 of 26 PFOA and PFHxS concentrations for the whole group of men were 24.5, 4.9 and 6.6 nglmL, respectively, and only these were included in the final regression analyses . The remaining PFAAs were found in much lower concentrations, and therefore these results are not discussed further. PFOS levels were positively correlated with PFOA (r = 0.594, p < 0.0005) and PFHxS (r = 0.304, p = 0.002) levels . PFOA and PFHxS levels were positively correlated, but not statistically significantly (r = 0.136, p = 0.2) . PFAAs and semen variables . In the whole group of 105 men, there was a tendency toward reduced levels of all semen variables in the "high" PFAA quartile group compared to the "low" group (groups constructed to include PFOS and PFOA levels, see statistical analysis), Table 2. The difference in percentage of morphologically normal spermatozoa as well as in the total number of normal spermatozoa (total sperm count x % morphologically normal sperms) was statistically significant (p = 0.037 and 0.030, respectively) . In the high PFOS-PFOA group the median number of normal spermatozoa in the ejaculate was 6.2 million vs. 15 .5 million in the low group (Figures 1 and 2). When analyzing associations of semen variables to PFOS and PFOA separately, as well as the simple summed concentration of PFOS and PFOA, estimated changes in semen variables with a change in serum PFAA concentration of I ng/mL indicated negative but non-significant associations between the PFAAs and semen variables (Table 3). PFAAs and reproductive hormones. There was no significant association between testosterone levels and PFAA levels, and no significant difference in PFAA levels between the high- and low-testosterone groups . For the whole group of 105 men, adjusted medians for the hormones point to a negative association to PFAA levels - however, none of these tendencies were statistically significant (Table 2). 10 Estimated associations between reproductive hormones and PFOS and PFOA separately and the summed concentration of PFOS and PFOA, showed no significant associations (Table 4). Smoking and BMI. 35% of the 105 men were smokers, and there were more smokers in the high testosterone group compared to the low testosterone group (49% and 21% smokers, respectively, p = 0.003) . Smoking status was significantly associated to higher testosterone, lower estradiol and higher SHBG when entered as a confounder in univariate analyses for these three variables only (p = 0.004 to 0.009). Smoking was not associated with any semen quality variables (p = 0.1 to 0.4), or levels of any PFAA (p = 0.1 to 1 .0) . Including smoking as a confounder did not considerably change the presented results or significance levels. Therefore, smoking was not included in the final analyses. BMI was not associated to levels of any PFAA (p = 0.1 to 1 .0), nor was there any confounding effect on any semen variables (e .g. p = 0.5 for morphologically normal sperms, and p = 0.9 for total number of morphologically normal sperms in analyses for difference between high and low PFAA groups). Discussion and conclusions This study examined PFAA levels in young adults . High serum concentrations of PFAAs were significantly associated with reduced numbers of normal spermatozoa . In addition, sperm concentration, total sperm count and sperm motility showed some tendency towards lower levels in men with high PFAA levels, although not at a statistically significant level . A tendency toward lower inhibin B/FSH ratio with high PFAA levels was in agreement with these findings, as these hormones reflect the spermatogenetic activity. ll p. 23 Page 12 of 26 Testosterone, FAI, testosterone/LH ratio and testosterone/estradiol ratio could suggest a poorer Leydig cell function in the "high" compared to the "low" PFAA quartile group. However, the associations between reproductive hormones and PFAAs were not completely consistent, and all were non-significant. This study therefore cannot demonstrate an adverse effect of PFAAs on Leydig cell function. We singled out PFOS and PFOA, as there are no data that specifically support PFHxS as an endocrine disruptor, and divided the men into three groups from the combined concentrations of PFOS and PFOA to account for a potential different effect at same concentration levels . As we could find no significant association between testosterone levels and PFAA levels, and no significant difference in PFAA levels between high- and low-testosterone groups, we could analyze associations between PFOS-PFOA levels and reproductive hormones or semen variables for the whole group . Controlling for confounding effect of smoking or BMI did not change estimates or significance levels . To our knowledge, there have been no consistent reports of associations between PFAA levels and smoking or BMI. Our material included only 105 men, and in addition, we had selected the men based on their serum testosterone values . The selection of two groups with high and low testosterone was done to test our primary hypothesis and therefore affects the homogeneity of the group when coirelations are analyzed for the group as a whole. This could potentially influence the subsequent analysis of semen quality by bias or confounding, and may affect the general applicability of the results . A larger follow-up study should preferably include randomly selected men from the general population . Our study is, to our knowledge, the first report of a correlation between semen quality and PFAAs . Very few studies of other endocrine disruptors (e.g. phthalates and pesticides) have previously demonstrated such an association (Duty et al . 2003 ; Hauser et 12 al. 2003; Hauser et al. 2007 ; Meeker et al. 2008 ; Swan et al . 2003) . If the results from our preliminary study of an association between high levels of PFAAs and decreased number of normal sperms are confinned, then high levels of PFAAs may be regarded as another endocrine disrupting factor contributing to the low semen quality seen among many young men . However, the importance of mixture effects of low-dose exposure to multiple compounds is becoming evident from animal studies but remains to be studied in humans (Hass et al. 2007 ; Kortenkamp et al. 2007) . The mode of action by PFAAs is not clear and only a few animal studies have explored mechanistic issues. These show decreased testosterone levels and reduced expression of steroidogenesis genes associated with Leydig cell hyperplasia in adult animals (Biegel et al. 1995 ; Shi et al. 2007), suggesting a direct testicular effect. Recent studies have indicated that the fetal gonad is particularly sensitive to exogenous factors (Skakkebaek et al . 2001) . However, our results could indicate that exposures later in life may contribute to impairment of semen quality, in line with other recent studies (Hauser et al . 2007) . We speculate that morphology is perhaps more susceptible to this than sperm concentration or total sperm count . Sperm morphology has proven an important indicator of semen quality and fertility in a clinical setting, even in men with normal sperm concentration (Guzick et al. 2001) . Interestingly, a recent study showed higher levels of maternal PFOS and PFOA levels in early pregnancy was associated with significantly longer waiting time to pregnancy (Fei et al. 2009) . We speculate that men and women living together may have similar exposure to PFAAs and that decreased semen quality caused by high PFAA levels may contribute to the longer waiting time to pregnancy found in that study. The use and emission polyfluorinated compounds continue to increase, and they are not readily cleared from the environment (Jensen et al. 2008; Prevedouros et al. 200b) . 13 p. 25 Page 14 of 26 Therefore, humans and wildlife worldwide will be exposed for years to come. We found positive correiations between levels of different PFAAs, as has been found previously (Apelberg et al. 2007; Calafat et al. 2007; Fei et al. 2007), suggesting common sources of exposure . The PFAAs levels we have detected are comparable to those found in other countries like Sweden and the Faroe Islands (Karrman et al. 2007 ; Weihe et al. 2008), but lower than earlier results from Denmark from 1996-2002 (Fei et al. 2007) . Thus, the effects we have indicated may also be true for other than the Danish population . In conclusion, our results indicate that higher PFAA levels were associated with lower numbers of normal sperms . In addition, we found non-significant negative associations between PFAA levels and other semen variables and reproductive hormones . Thus, high levels of PFAAs may contribute to the otherwise unexplained low semen quality seen in many young men . However, results from this first and preliminary study should be corroborated in larger studies . References Andersen AG, Jensen TK, Carlsen E, Jorgensen N, Andersson AM, Krarup T et al. 2000 . High frequency of sub-optimal semen quality in an unselected population of young men . Hum Reprod 15:366-372 . Apelberg BJ, Witter FR, Herbstman JB, Calafat AM, Halden RU, Needham LL et al. 2007 . Cord serum concentrations of perfluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA) in relation to weight and size at birth. Environ Health Perspect 115 :1670-1676. Biegel LB, Liu RC, Hurtt ME, Cook JC. 1995 . Effects of ammonium perfluorooctanoate on Leydig cell function: in vitro, in vivo, and ex vivo studies . Toxicol Appl Pharmacol 134 :18-25 . Bossi R, Riget FF, Dietz R. 2005 . Temporal and spatial trends of perfluorinated compounds in ringed seal (Phoca hispida) from Greenland . Environ Sci Technol 39:7416-7422 . Calafat AM, Kuklenyik Z, Caudill SP, Reidy JA, Needham LL. 2006. Pertluorochemicals in pooled serum samples from United States residents in 2001 and 2002. Environ Sci Technol 40:2128-2134. Calafat AM, Wong LY, Kuklenyik Z, Reidy JA, Needham LL. 2007. Polyfluoroalkyl chemicals in the U.S . population : data from the National Health and Nutrition Examination Survey (NHANES) 2003-2004 and comparisons with NHANES 19992000. Environ Health Perspect 115:1596-1602 . Cook JC, Murray SM, Frame SR, Hurtt ME. 1992. Induction of Leydig cell adenomas by ammonium perfluorooctanoate : a possible endocrine-related mechanism . Toxicol Appl Pharmacol 113 :209-217 . Duty SM, Silva MJ, Barr DB, Brock JW, Ryan L, Chen Z et al . 2003. Phthalate exposure and human semen parameters. Epidemiology 14:269-277 . 15 p. 27 Page 16 of 26 Fei C, McLaughlin JK, Lipworth L, Olsen J. 2009 . In press . Maternal levels of perfluorinated chemicals and subfecundity . Hum Reprod. Fei C, McLaughlin JK, Tarone RE, Olsen J. 2007. Perfluorinated chemicals and fetal growth: a study within the Danish National Birth Cohort . Environ Health Perspect 115:1677-1682 . Giesy JP, Kannan K. 2001 . Global distribution ofperfluorooctane sulfonate in wildlife . Environ Sci Technol 35 :1339-1342 . Guruge KS, Taniyasu S, Yamashita N, Wijeratna S, Mohotti KM, Seneviratne HR et al. 2005 . Perfluorinated organic compounds in human blood serum and seminal plasma : a study of urban and rural tea worker populations in Sri Lanka. J Environ Monit 7 :371-377 . Guzick DS, Overstreet JW, Factor-Litvak P, Brazil CK, Nakajima ST, Coutifaris C et al. 2001 . Sperm morphology, motility, and concentration in fertile and infertile men . N Engl J Med 345 :1388-1393 . Hansen KJ, Clemen LA, Ellefson ME, Johnson HO. 2001 . Compound-specific, quantitative characterization of organic fluorochemicals in biological matrices. Environ Sci Technol 35:766-770. Hass U, Scholze M, Christiansen S, Dalgaard M, Vinggaard AM, Axelstad M et al. 2007. Combined exposure to anti-androgens exacerbates disruption of sexual differentiation in the rat. Environ Health Perspect 115 Suppl 1 :122-128. Hauser R, Chen Z, Pothier L, Ryan L, Altshul L. 2003. The relationship between human semen parameters and environmental exposure to polychlorinated biphenyls and p,p'-DDE. Environ Health Perspect l 11 :1505-1511 . Hauser R, Meeker JD, Singh NP, Silva MJ, Ryan L, Duty S et al. 2007 . DNA damage in human sperm is related to urinary levels of phthalate monoester and oxidative metabolites . Hum Reprod 22:688-695. 16 Jensen A.A., Poulsen P.B ., Bossi R. 2008. Survey and environmental/health assessment of fluorinated substances in impregnated consumer products and impregnating agents . Survey of chemical substances in consumer products no. 99. Danish Agency for Environmental Protection . Http ://www.mst.dk/Udgivelser/Publications/2008/10/978-87-7052-845-O .htm [accessed 10 February 2009]. Jensen AA, Leffers H. 2008 . Emerging endocrine disrupters : perfluoroalkylated substances. Int J Andro131 :161-169 . Jrgensen N, Carlsen E, Nermoen I, Punab M, Suominen J, Andersen AG et al. 2002. EastWest gradient in semen quality in the Nordic-Baltic area: a study of men from the general population in Denmark, Norway, Estonia and Finland. Hum Reprod 17 :2199-2208 . Kannan K, Corsolini S, Falandysz J, Fillmann G, Kumar KS, Loganathan BG et al. 2004. Perfluorooctanesulfonate and related fluorochemicals in human blood from several countries . Environ Sci Technol 38 :4489-4495 . Karrman A, Ericson I, van Bavel B, Darnerud PO, Aune M, Glynn A et al. 2007. Exposure of perfluorinated chemicals through lactation : levels of matched human milk and serum and a temporal trend, 1996-2004, in Sweden. Environ Health Perspect 115 :226-230. Kissa E. 2001 . Fluorinated surfactants and repellents, 2nd edition . Surfactant science series, Vol . 97. New York: Marcel Dekker . Kortenkamp A, Faust M, Scholze M, Backhaus T. 2007 . Low-level exposure to multiple chemicals : reason for human health concerns? Environ Health Perspect 115 Suppl 1 :106-114 . p. 29 Page 18 of 26 Meeker JD, Barr DB, Hauser R. 2008. Human semen quality and sperm DNA damage in . relation to urinary metabolites of pyrethroid insecticides . Hum Reprod 23:19321940. Menkveld R, Stander FS, Kotze TJ, Kruger TF, van Zyl JA. 1990. The evaluation of morphological characteristics of human spermatozoa according to stricter criteria . Hum Reprod 5 :586-592 . Olsen GW, Burris JM, Ehresman DJ, Froehlich JW, Seacat AM, Butenhoff JL et al. 2007 . Half-life of serum elimination of perfluorooctanesulfonate,perfluorohexanesulfonate, and perfluorooctanoate in retired fluorochemical production workers . Environ Health Perspect 115 :12981305. Paasch U, Salzbrunn A, Glander HJ, Plambeck K, Salzbrunn H, Grunewald S et al . 2008. Semen quality in sub-fertile range for a significant proportion of young men from the general German population: a co-ordinated, controlled study of 791 men from Hamburg and Leipzig. Int J Androl 31 :93-102 . Prevedouros K, Cousins IT, Buck RC, Korzeniowski SH. 2006. Sources, fate and transport of perfluorocarboxylates . Environ Sci Technol 40:32-44 . Shi Z, Zhang H, Liu Y, Xu M, Dai J. 2007. Alterations in gene expression and testosterone synthesis in the testes of male rats exposed to perfluorododecanoic acid. Toxicol Sci 98:206-215 . Skakkebwk NE, Rajpert-De Meyts E, Main KM. 2001 . Testicular dysgenesis syndrome: an increasingly common developmental disorder with environmental aspects . Hum Reprod 16:972-978 . Swan SH, Kruse RL, Liu F, Barr DB, Drobnis EZ, Redmon JB et al. 2003. Semen quality in relation to biomarkers of pesticide exposure. Environ Health Perspect 111 :14781484. 18 Weihe P, Kato K, Calafat AM, Nielsen F, Wanigatunga AA, Needham LL et al. 2008 . Serum concentrations of polyfluoroalkyl compounds in Faroese whale meat consumers. Environ Sci Techno142:6291-6295 . Table l p. 31 Page 20 of 26 PFAA median concentrations (5-95`h percentiles) for high and low testosterone groups and all study subjects (ng/mL), and p-values for difference between the two groups . PFHxS PFHpA PFOA PFOS PFOSA PFNA PFDA PFUnA PFDoA PFTrA Samples High testosterone > LOD N=53 105 6.6 (4.0-13.0) 98 0.2 (0.01-0.9) 105 4.4 (2.6-7 .0) 105 25 .5 (14.2 - 39.6) 56 0.1 (0.00-3.7) 105 0.8 (0.4-1 .8) 104 0.9 (0.2-1 .1) 101 0.1 (0.04-0.3) 102 0.08 (0.04-0.8) 7 0.00 (0.000 - 0.4) Low testosterone N=52 6.6 (3.5-12.1) 0.3 (0.00-1 .3) 5.0 (2.7-7 .5) 23.9 (12.8 - 45.2) 0.00 (0.00-3.5) 0.8 (0.4-2 .0) 0.8 (0.4-1 .2) 0.2 (0.00-0.4) 0.08 (0.02-0.8) 0.00 (0.00 - 0.06) Whole group N=105 6.6 (4.0-12.1) 0.2 (0.00-1 .1) 4.9 (2.7-7 .2) 24.5 (14.2 - 42.1) 0.06 (0.00-3.5) 0.8 (0.4-1 .8) 0.9 (0.3-1 .1) 0.1 (0.02-0.4) 0.08 (0.04-0.8) 0.00 (0.00-0.2) pvalue 0.8 0.8 0.1 0.4 0.008* 0.6 0.9 1 .0 0.8 0.2 Page 21 -of 26 p. 32 Table 2 Adjusted means (95% CI) for PFAA quartile groups, and p-value for difference between the low and high PFAA quartile groups . Sex hormones' Low PFAA Intermediate PFAA High PFAA N Testosterone (nmol/L) Estradiol (pmol/L) SHBG (nmol/L) LH (IU/L) FSH (IU/L) Inhibin-B (pg/mL) FAI Testosterone/LH FAI/LH Estradiol/Testosterone Inhibin/FSH N=29 25.2 (21 .7 - 28.7) 77.6 (70.3 - 85.8) 27.8 (24.0 - 32.2) 3.4 (2.8-4.1) 2.7 (2.1-3.5) 181 (142-232) 84.1 (74.1 - 95 .5) 6.9 (5 .6-8.4) 24 .7 (19 .9 - 30.7) 3 .3 (3 .0-3 .7) 66.8 (42.1 - 106 .0) N=48 22.3 (19 .6 - 25.0) 72.4 (67 .0 - 84.7) 25.4 (22 .6 - 28.5) 2.9 (2.5-3 .4) 2.7 (2.2-3-3) 175 (144-212) 80.2 (72.6 - 88.7) 7.0 (5.9-8.2) 27.4 (23.1 - 32.5) 3.6 (3.2-3.9) 66.0 (45 .9 - 94.8) N=28 22.3 (18.8 - 25.8) 76.6 (69.3 - 84.7) 26.1 (22.5 - 30.3) 3.7 (3.0-4 .4) 3.0 (2.3-3 .8) 152 (119-195) 77.8 (68.5 - 88.3) 5.5 (4.5-6 .8) 21 .2 (17 .1 - 26.4) 3 .8 (3 .4-4 .2) 51 .3 (32.3 - 81 .4) Semen qualityb Volume (mL) 4.0 Concentration (mio ./mL) 59 Total count (mio .) 228 Motile sperms (%) 73 Morphologically normal (%) 8 .8 Total morphologically normal (mio.) 15 .5 (3.2 - 5.0) (36-96) (l34 - 389) (69-77) (7 .2 - 10.4) (7.3 - 33.0) 3 .4 (2.9 - 4.1) 51 (35-74) 172 (114-261) 70 (66-73) 7.7 (6.4 - 9.0) 10 .0 (5 .6 - 17 .9) 3.5 (2.9 - 4.4) 40 (25-64) 143 (86 - 237) 71 (66-75) 6.3 (4.6 - 8 .0) 6.23 (3.0 - 12 .8) p-value 0.2 0.9 0.5 0.6 0.6 0.3 0.4 0.1 0.3 0.1 0.4 0.3 0.2 0.1 0.4 0.037' 0.030~ a Hormone levels are adjusted for time of blood sampling. b Volume, concentration and total count are adjusted for duration of abstinence. Motility is adjusted for time between ejaculation and semen analysis. Morphology is not adjusted for confounders . p. 33 Page 22 of 26 Table 3 Estimated change in semen variables' with a change in PFAA of 1 ng/mL (95% CI). All 105 men included . PFOS PFOA PFAA sumb In Volume In Concentration In Total count In Motility Morphology 0.000 -0 .020 -0 .018 -0 .006 -0 .085 (-0.012 - 0.011) (-0.044 - 0.005) (-0.045 - 0.010) (-0.019 - 0.007) (-0.200 - 0.026) -0 .002 -0 .080 -0.074 -0.027 -0 .540 (-0.070 - 0.066) (-0.230 - 0.066) (-0.230 - 0.086) (-0.110 - 0.053) (-1 .200 - 0.110) 0.000 -0 .018 -0 .016 -0 .006 -0 .082 (-0.010 - 0.010) (-0.040 - 0.004) (-0.041 - 0.008) (-0.018 - 0.007) (-0.181 - 0.018) a Volume, concentration and total count are adjusted for duration of abstinence. Motility is adjusted for time between ejaculation and semen analysis . Morphology is not adjusted for confounders . b "PFAA sum" is PFOS and PFOA mass concentrations summed (ng/mL) . Table 4 Estimated change in reproductive hormones' with a change in PFAA concentration of 1 ng/mL (95% CI). All 105 men included . PFOS PFOA PFAA sumb Testosterone In Estradiol In SHBG In LH In FSH In Inhibin-B In FAI In Testosterone/LH In FAI/LH In EstradioU Testosterone In Inhibin/FSH -0 .087 -0 .001 0.002 0.000 0.004 -0.004 -0.006 -0.003 -0.006 0.003 -0 .009 (-0.32 - 0.15) (-0.008 - 0.005) (-0.007 - 0.012) (-0.014 - 0.012) (-0.13 - 0.22) (-0.21 - 0.12) (-0.015 - 0.002) (-0.017 - 0.011) (-0.020 - 0.009) (-0.005 - 0.010) (-0.039 - 0.022) -0 .98 -0 .012 -0 .009 -0 .010 -0.037 0.012 -0.038 -0.037 -0.028 (-2.33 - 0.37) (-0.051 - 0.027) (-0.067 - 0.048) (-0.084 - 0.064) (-0.14 - 0.064) (-0.084 - 0.11) (-0.087 - 0.011) (-0.12 - 0.045) (-0.114 - 0.058) 0.035 (-0.010 - 0.081) 0.049 (-0.13 - 0.23) -0 .093 -0 .001 0.002 0.000 0.003 -0 .003 -0 .006 -0 .003 -0.005 (-0.303 (-0.007 (-0.007 (-0.012 (-0.013 (-0.018 (-0.014 (-0.016 (-0.018 - 0.116) - 0.005) - 0.011) - 0.010) - 0.018) - 0.012) - 0.001) - 0.009) - 0.008) 0.003 (-0.004 - 0.010) -0.006 (-0.034 - 0.022) a Hormone levels are adjusted for time of blood sampling. b "PFAA sum" is PFOS and PFOA mass concentrations summed (ng/mL) . Figure legends p. 35 Page 24 of 26 Figure 1. Morphologically normal spermatozoa (%) and PFAA quartile groups (adjusted means and 95% CI). All 105 men included . * p = 0.037 for difference compared to Low PFAA quartile group. Figure 2. Total morphologically normal spermatozoa (mio.)a and PFAA quartile groups (adjusted means and 95% CI). All 105 men included . a Total morphologically normal spermatozoa is adjusted for duration of abstinence. * p = 0.030 for difference compared to Low PFAA quartile group . Figure 1 . Morphologically normal spervmatozor M. 4 z 0 Low PFAA tntennediate PFAA High PFAA Morphologically normal spermatozoa (%) and PFAA quartile groups (adjusted means and 95% CI). All 105 men included . * p = 0 .037 for difference compared to Low PFAA quartile group . 201x285mm (150 x 150 DPI) Figure 2. Total norpbologteallp nornWi spermatozoa (mio.) . Low PFAA Intermediate PFAA High PFAA p. 37 Page 26 of 26 Total morphologically normal spermatozoa (mio .)a and PFAA quartile groups (adjusted means and 95% CI) . All 105 men included . a Total morphologically normal spermatozoa is adjusted for duration of abstinence . * p = 0 .030 for difference compared to Low PFAA quartile group. 201x285mm (150 x 150 DPI)