Document omGpkgrVNv7K74oa2qEg8o9ZE

CLINICAL ANO METABOLIC ABNORMAL IT IES IN OCCUPATIONAL EXPOSURE TO POLYCHLORINATED BIPHENYLS (PCBs) Kenneth H. Chase, M.D.1, Otto Wong, Sc.O.2, Oavid Thomas, O.Stat.2 Bertram W. Berney, M.H.S.1 and Robert K. Simon, Ph.O.3 Presented at the American Occupational Health Conference, Atlanta, Georgia -- Hay 1, 1981 ^Washington Occupational Health Associates, Inc., Washington, O.C. 2GQ36 ^Biometric Research Institute, Inc., Washington, O.C, 20007 ^Professional Clinical Laboratories, Inc., Wilmington, Oelaware 198C6 HONS 209587 .ABSTRACT A cross-sectional study of 120 rale .vorkers was conducted to determine the prevalence of increased PCS absorption as well as the oresence of ootentially related clinical and metabolic abnormalities. Three exposure categories ("exposed", 'nominally exposed", "non-exposed") were defined. Complete work histories, clinical histories, physical exams and lab tests were obtained including PCS plasma lev el determinations. In addition, PCS fat levels were determined in a portion of each exposed aroup. Findings of dermatoxicity were observed and elevated PCS levels more frequently noted in the exposed group (p<0.0001) correlating well with age and duration of employment. These correlations were stronger for fat (o<0.00l) than for plasma (o<0.01) PCB levels. In the exposed group, a significant correlation was found between PCB plasma and serum triglyceride (p<0.0001) and SGOT levels (p<0.01). These correlations remained significant after controlling for either age or lenoth of employment. No significant correlations were found between PCB levels and cholesterol, HDL cholesterol or liver function tests other than SGOT. Further analyses relating frequency of reported direct contact with PCB levels suggested a dermal route of exposure. An analysis by union affiliation demonstrated that crafts with Greater direct exposure had correspondingly higher elevations of PCB levels. 1 HONS 209588 3ACKGRCUN0 3olyci*Iorinat5d biohenyls (PCBs) comprise a group of synthetic chlorinated hydrocarbon corpounds with considerable industrial utility based on their resistance to oxidation and dielectric properties. They have been used world-wide in capacitors and transformers and many other applications described elsewhere (1). The contamination of the general environment as well as the workolace by 3C3s was appreciated only recently leading to the ban of their continued manufacture under the Toxic Substances Control Act of 1975 [l). This action was promoted by greater awareness of ani~al toxicity studies (3), the stability and persistence of PCBs in the environment, the tendency of PCBs to accumulate in human tissues (A,5) and the documented chronic toxicity of PCBs in humans Jollowinq accidental oral ingestion in the "Yusho" incident (6). In soite of leoislatlve and regulatory actions to phase out the use of PC3s, solutions to oroblems of effective discosal have not been forthcoming. In 1977 Hecht (7) estimated that of the 1.C5 billion oounds produced in this country since the 193Cs, 7;J -lillicn pounds were still in use and 450 million sounds had r'3i-.ri their way into the general environment. A review 13) by the .`,'ational Institute for Occupational Safety and Health i'llOSH) of clinical and metabolic abnormalities in both environmentally and occupationally exposed groups revealed inconsistent relationshios between tissue ?C3 levels and abnormal findings. A more HONS 209589 recent publication (9) suggests alterations of lipid metabolism at PCB levels less than tkose usually reported in association with findings of clinical toxicity. An opportunity to re-examine these relationships was recently provided in a railroad passenger car and locomotive maintenance facility at which some employees have been exposed to PCB-containing transformer fluids {Pyranol, Inerteen, Askerel) for nearly 40 years. Inasmuch as the potential health hazards of PCBs were not apparent until the past few years, the handling of transformer fluids at this facility was on a casual basis with few or no special handling orocedures in affect for the greater part of this exposure period. A change in management followed by the documentation of elevated PCB levels in the work environment in early 1979 provided the impetus for developing more conservative handling procedures as well as a medical monitoring program which included the design and execution of an epidemiologic study herein described. It was the ouroose of this study to determine the prevalence of in creased PCB absorption among exposed workers as well as the pre valence of findings compatible with PCB toxicity. In addition two control grDuos were defined with either nominal or non-exposure against which comparisons could be made. Lastly, through detailed work histories, an attempt was made to determine those factors most predictive of elevated PCB levels as well as to shed light on the most probable route of exposure. HONS 209590 1 MATERIALS A'IQ ""ETHOOS One hundred and twenty male employees oarticioated in the study according to the following exposure grouos: Group I -* Exposed (36 subjects) Group II ** Nominally Exposed (15 subjects) Group III -- Non-Exposed (19 subjects) Subjects within each of these three groups were randomly selected by a computer program from the corresponding underlying populations. Group I subjects were defined as "exposed" on the basis of their frequent oooortunity for direct contact with PCB-containing trans former fluids in their assigned shop areas. Group II subjects were defined as "nominally exposed" on the basis of their employ ment at the same facility but without the opoortunity (work assignant) characterizing Group I subjects. Group III subjects were defined as "non-exoosed" on the basis of the absence of ?C3-containing fluids in their work environment and were employed elsewhere, ill three grcuos reside in the s~e geographic rsqion. All participants completed a detailed work history guestion-airs in which tasks entail fro cctental e-ioc sure to PCS-containing fluids were character!ted and quantified. I-fpr-ation related to ootantlal exoosure to other toxic substances '-'as also obtained. Ccrplet* medical histories and ohysical examinations were per formed on all participants. A comprehensive battery of clinical laboratory tests were performed in :he fasting state (12 hours) HONS 209591 as outlined in Table 1. Plasma =C3 levels were determined on all subjects. In addition, 50 rar :;rly selected subjects (36 in Group I, 5 in Group II and 9 in Group III) underwent subcutaneous fat biopsies upon which fat PCS levels were deter mined. Analytical determinations for PC3s in heparinized human plasma and fat specimens were performed as specified by EPA methods (10). In brief, hexane extraction was followed by speci fied silica gel and/or florisil cleanup to remove pesticides. The isolated PCS fraction was assayed by electron capture aas chrcmatcqraohy according to Uebb/VcCal1 (11). All of the laboratory testing was performed by Professional Clinical Labor atories, Inc. in WiImington, Delaware, This laboratory partici pated in the Canter for Disease Control (CDC) trial PCS testing program concocted during the course of this study. Student's t-tasts were used to compare average values of selected variables among the three rrouos. Correlation between selected variables in Group I was assessed usi-'g Pearson's sf-ole correction coefficients, and partial correlation coeffi cients to adjust for ccnfoundina vanities. The chi-square procedure was used to test for association i" the contingency tables. HONS 209592 RESULTS Table 2 shows the are distribution of the three groups. The two exposed oroups are older than the non-exposed group. The average aae for the exoosed nroup was 41.4 years; the nominally exposed group 37.2; and the non-exposed group 30.7. The difference in average aae between the exposed arouo and the non-exposed group is statistically significant (p <0.05). This age difference should be taken into consideration when comparisons between the groups are made, since many of the oarameters vary with age. Table 3 presents the distribution of lenath of emoloyment by exposure category. The average lenaths of employment were 17 years, 3 years and 9 months, and 4 years and 3 months for the exposed, nominally exposed, and non-exposed aroups, respec tively. The difference between the exposed aroup and the nominally exposed group or the non-exposed group is hiahly significant (p <0,0001). This was not surprising in view of the traditionally stable crafts represented by the exposed aroup. Review of medical histories and physical findings failed to reveal findings of oraan toxicity in the exposed group although several cases of chloracne were encountered in this group with none belna found in the other oroups. Other derma tological findings were more prevalent in the exposed group as well, but no significant association between these and PCS levels could be demonstrated within the group. Historically, HONS 209593 there was a corrmon reporting in Grouo I of transient skin and eye irritation attributed to accidental soi11s in which there was direct contact with PCB-containina fluids. The averaqes of some selected bfochemical parameters by exposure category are shown in Table 4. There is no statisti cally significant difference among the groups in trinlycerides, cholesterol, HDL, SGOT, SGPT. For v-glutamyl transoeptidase (yGTP), the highest value, 37.1 III/4, was reported among the exposed employees, the intermediate value, 34-1, among the nominally exoosed, and the lowest value, 29.B, amona the non-exposed: an indication of a trend by exposed category. However, because of the considerable variabilities in the measurement and the relatively small sample sizes of both the nominally exposed and the non-exposed arouos, these differences are not statistically significant. For PCB levels, the exposed group showed a clear elevation when compared to either the nominally exposed or the non-exoosed arouos (Table 5). The averaae plasma PCB level (33.4 opb) of the exposed group was more than twice that of either the nomin ally exposed group (14.2 ppb) or the non-exposed grouo (12.0 ppb). The differences in plasma PCB level between the exoosed qroup and the other two groups are highly significant (o <0.0001), but the difference between the nominally exposed and the non-exposed qroup is not statistically sionificant. The discrepancy in fat PCB level for the exposed group was 5.6 pom, more than three HONS 209594 times that of either the nominally exDOSed {1.4 pom) or the non-exposed group (1.3 opm). Again the difference between Group I (exposed) and either Group [[ (nominally exposed) or Group III (non-exposed) is hiahly significant (p <0.0001). but that between Group II and Group III is not significant. In an attempt to determine any association or correlation between PCS exoosure and other variables in Group I. Pearson's correlation coefficients were calculated. Similar analyses were not performed for Groups II or III as the PCB levels were well within the normal range (12,13) and the sample sizes for these groups were relatively snail. Table 6 shows the correlation coefficients and the corresponding p values for selected biochemical parameters, length of employment and age for the exposed group. Apart from the obvious expected association between length of employment and age, significant correlations were found between length of employment and plasma PCB level, as well as between length of employment and fat PCB level. However, the correlation between length of employment and fat PCB level (r 0.54, p <0.001) was stronger than that between length of employment and plasma PCB level (r 0.27, p <0.01). Similar significant correlations were observed between age and plasma PCB as well as fat PCB levels. A discussion on the possible confounding effect between age and length of employment is niven below. HONS 209595  Also in Group I, statistically significant correlation was demonstrated between olasma "C3 'avals and trinlycerides (r = 0.36, p <0.0001) but not between fat PCS levels and tri glycerides. Plasma C3 levels were also significantly corre lated with SCOT {r =* 0.28, p -0.01) but not with SGPT or ,GTP. Fat PCS levels were not significantly correlated with SCOT, SGPT, or yGTP. Since many of the variables examined in Table 6 are inter related, oartial correlation coefficients, which take into account confounding factors, were calculated and are presented in Table 7. The correlation between plasma PCS levels a.nd triglycerides remained highly significant even after adjusting for either age or length of employment {r 0.54 and 0.55 respectively, p <0.001). On the other hand, there was no significant correlation between fat PCB levels and triglycerides, adjusting for age or length of employment. Among the liver function parameters, SGOT was found to remain significantly correlated with plasma PCB levels even after adjusting for age (r * 0.26, p <0.05). To determine the possible relationship between the frequency of direct contact with PCB-containina fluids and PCB levels, all 120 subjects in the study were asked whether or not they ever had direct contact with such fluids and, if so, to retort the frequency with which this had occurred. Table 3 snows the HONS 209596 crosstabulation of plasma PCS levels and reoorted frequency of direct contact. The fioures clearly demonstrate a direct relationship between the two parameters. Arena those who reoorted"no direct contact with PCB-containinp fluids, 41% had PCB olasma levels less than or equal to 10 ppb, whereas only 4" of those with frequent contacts had plasma PCB levels in this ranae. On the other hand, only 3" of those who reoorted no direct contact had plasma PCB levels greater than 30 ppb, compared to 44% among those with frequent such contacts. This association is highly sionificant (chi-souare 28.33, p <0.005). Table 9 demonstrates a similar relationship between fat PCB levels and reported frequency of direct contact. As it was evident from the work histories that certain crafts inherently provided Greater opportunity than others for direct con tact with PCB-containinq fluids, an attempt was made to correlate plasma PCB levels with union affiliation for those subjects emoloyed at the maintenance facility {Groups I and II). The result was hiahly significant (chi-souare * 42.89, o<0.001) and is pre sented in Table 10. Similar calculations were not performed for fat PCB levels as the number of subjects with recorded levels was too small to permit same. HONS 209597 discussion In this study Groups I (exoosed) and II (nominally exposed) consisted of male subjects employed at a railroad passenger car and locomotive maintenance facility, whereas Group III (nonexposed) subjects were employed elsewhere. All subjects were of the same sex, shared the same employer, and lived in the same geographic area, the greater length of employment noted in Group I subjects was attributed to the traditional stability in the crafts represented by that group. These employees have remained at the same facility for a mean duration of 17 years with a ranoe of a few months to AO years. Comparable occupational PCB exposure studies have variously shown findings of dematotoxicity, hyperpigmentation and liver function abnormalities but without consistent relationshios to PCB levels when measured (8). The transient symptoms of skin and eye irritation associated with acute exposures, as well as the occasional findlna of chloracne reported in this study, are in keeoing with these other studies. A recent comprehensive study of capacitor manufacturing workers by Fischbein et al. (14) was similarly notable for Its relative absence of physical findings outside of a hloh prevalence of dermatologic abnormalities in spite of a mean plasma PCB level of 172 ppb (n2B9). Even rore recently Baker et al. (9) reported no symptoms or physical findings of oroan toxicity in a different grouo of PCB exposed capacitor workers whose mean plasma PCB level was 75.1 ppb (n18). HONS 209590 DISCUSSION In this study Groups I (exoosed) and II (nominally exposed) consisted of male subjects employed at a railroad passenger car and locomotive maintenance facility, whereas Group III (nonexposed) subjects were employed elsewhere. All subjects were of the same sex, shared the same employer, and lived in the same geographic area. The greater length of employment noted in Group I subjects was attributed to the traditional stability in the crafts represented by that group. These emplovees have remained at the same facility for a mean duration of 17 years with a range of a few months to 40 years. Comparable occupational PC8 exposure studies have variously shown findings of dermatotoxlclty, hyperpigmentation and liver function abnormalities but without consistent relationships to PCB levels when measured (8). The transient symptoms of skin and eye irritation associated with acute exposures, as well as the occasional finding of chloracne reported in this study, are In keeping with these other studies. A recent comprehensive study of capacitor manufacturing workers by Flschbeln et al. (14) was similarly notable for Its relative absence of-physical findings outside of a high prevalence of dermatologic abnormalities in spite of a mean plasma PCB level of 172 ppb (n289). Even more recently Baker et al. (9) reported no symptoms or physical findlnos of organ toxicity In a different group of PCB exposed capacitor workers whose mean plasma PCB level was 75.1 ppb (n18). HONS 209599 11 (IS) with structural similarities to PCBs. There have been less than a handful of reoorts outside the Yusho incident describing lipid abnormalities in human exposures to PC8s but these have sup ported the finding of elevated serum triglycerides (9). The addi tional finding of lowered HOL linonrotein levels in this latter study was not confirmed by the current investigation. Although no evidence of overt hepatotoxicity was evident in the exposed group, the association between plasma PCB levels and SGOT was statistically significant, a finding which has been re cently reported by others (14). [t is temoting to incriminate the liver as a target orcan for CB exposure responsible for both the trialyceride and SGOT abnormalities as has been suggested in animal studies (16), but this reguires further testing of alternative hy potheses. The demonstrated relationship between PCB levels and reported freguency of direct contrast with PCB-contalnlng fludls suggests a dermal route of exoosure. 4s PC3s generally exhibit low vapor pressures (19), an airborne route of exoosure is not likely unless the PCB-contalning fluids are heated to a significant degree. There is thus theoretical support for the use of suits impervious to PCBs In protecting the worker from further exposure to PCBs and the po tentlal, yet still unknown, risks for same. Those crafts with known direct exoosure to PCBs are at greater risks for PCB absorption as has been shown earlier (Table 10). The predominant crafts in this study are represented by the following HONS 209600 (18) with structural similarities to PC8s. There have been less than a handful of reports outside the Vusho incident describing lipid abnormalities in human exposures to PCBs but these have sup ported the finding of elevated serum triglycerides (9). The addi tional finding of lowered HDL lipoprotein levels in this latter study was not confirmed by the current Investigation. Although no evidence of overt hepatotoxlclty was evident in the exposed group, the association between plasma PCB levels and SGOT was statistically significant, a finding which has been re cently reoorted by others (14). It Is tempting to incriminate the liver as a target organ for PCB exoosure responsible for both the triglyceride and SGOT abnormalities as has been suggested in animal studies (16), but this requires further testing of alternative hy potheses. The demonstrated relationship between PCB levels and reported frequency of direct contrast with PCB-contalning fludis suggests a dermal route of exposure. As PCBs generally exhibit low vapor pressures (19), an airborne route of exposure Is not likely unless the PCB-contalning fluids are heated to a significant degree. There Is thus theoretical support for the use of suits impervious to PCBs In protecting the worker from further exposure to PCBs and the po tential, yet still unknown, risks for same. Those crafts with known direct exposure to PCBs are at greater risks for PCB absorption as has been shown earlier (Table 10). The predominant crafts In this study are represented by the following HONS 209601 13 unions: BMWE -- Brotherhood of Maintenace of Way Employees BRAC -- Brotherhood of Railway and Airline Clubs JCC -- Joint Council of Carmen IAM -- International Association of Machinists SMW1A - Sheet Metal Workers International Association I8EW -- International Brotherhood of Electrical Workers Analyses of this type provide a rational basis for the assignment of protective equipment and the development of appropriate handling procedures for PCB-containing fluids. 14 MONS 209602 (8) National Institute for Occupational Safety and Health: Criteria for a Recommended Standard...Occupational Exposure to PCBs, NIOSH Pub. No. 77-225, Washington, O.C. Government Printing Office, 1977. (9) Baker EL, landrlqan PJ, Glueck CJ, et al: Metabolic consequences of exposure to polychlorinated biphenyls (PCBs) in sewage sludge. Am J. Epidemiology 112: 553-563, 1980. (10) EPA 600/8-80-038 6/30 Manual of Analytical Methods for the Analysis of Pesticides In Humans and Environmental Samples. LI.S. EPA Health Effects Research Lab, Research Triangle Park, N.C. 27711. Ed. by R.R. Watts. (11) R.G. Webb and A.C. McCall, J. of Chromatographlc Sciences 11:366-373, 1973. (12) Price HA, Welch RL: Occurrence of polychlorinated biphenyls in humans. Environ Hlth Perspective 1:73-78, 1972. (13) Yobs, AR: Levels of polychlorinated biphenyls In Adipose tissue of the general copulation of the nation, Environ Hlth Perspective 1:79-31, 1972. (14) Plschbein A, Woolf MS, Libls R, et al: Clinical findings among PCB-exposed capacitor manufacturing workers. *nn NY Acad 5ci 323:703-715, 1979. MONS 209603 15 (15) Ouw Hk, Simpson GR, Siyall Os: Use and health effects of Arochlor 1242, a polychlorinated biphenyl, in an electrical industry. Arch Env Health 31:189-194, 1976. (16) Holub 8j, Plerkarskl J, Nlllson K: The effect of a PCB (2,4,2',4'-tetra-chloroblphenyl) on lipid synthesizing enzymes In rat liver mlcrosomes. Bull Environ Contam Toxicol 14:415, 1975. (17) Kuratsune, M, Masuda Y, Naoayama J: Some of the recent findings concerning Yusho.ln Proceedings of a National Conference on Polychlorinated Biphenyls. EPA-560/6-75-004, Washington, D.C. Environmental Protection Arency. March 1976, pp. 14-29. (18) Carlson, LA, Kolmadln-Hednan 3: Hyper-alpha-Hpoproteinemla in men exposed to chlorinated hydrocarbon pesticides. Acta Med Scand 192:29-32, 1972. (19) National Institute for Occupational Safety and Health, Health Hazard Evaluation Determination Report No. 76-52-386, April, 1977. HONS 209604 '.i ACKNOWLEDGMENTS ANO PUBLICATION INQUIRIES The authors wish to express their appreciation for the technical assistance of Roberta Kaplan, P.A.-C., Brian Olson, 8.A. and Cynthia M. 8urosh, M.Ed. in the preparation of this manuscript and for the assistance provided by the following persons in the performance of this study: Kenneth Harbert, P.A.-C., Vincent Kownacki, M.O., Judy Knotts, Tom Wampler, B.S., M.Ed., Wanda McMahon, R.N. and Kevin Rossbach. This study was supported by a contract from the National Railroad Passenger Corporation. Publication inquiries should be addressed to: Kenneth W. Chase, M.O. President Washington Occupational Health Associates, Inc. 1145 13th Street, Suite 312 Washington, O.C. 20036 The current address for Or. Simon is: Or. Robert K. Simon director of Toxicology Biomedical Reference Laboratories, Inc. P.0. 3ox 2230 Burlington, North Carolina 27215 HONS 209605 19 -Tablt 1. - Laboratory Tests Determined in Study Subjects Complete Blood Count Urinalysis Glucose BUN Uric Acid Creatinine Calcium Phosphorous Alkaline Phosphatase Total Protein Albumin Globulin A/G Ratio Total Bilirubin SGPT SGOT LOH y-Glutamyl Transpeptidase Triglycerldes Cholesterol HDl Cholesterol LDL Cholesterol VLDL Cholesterol Phenotype HONS 209606 Table 2. - Age Distribution of 120 Workers by PCB Exposure Category Age (years) 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60+ Total (N) Average Exposed No. % 1 1.2 14 16.3 17 19.8 5 5.8 7 8.1 6 7.0 2 2.3 11 12.8 17 19.8 6 7.0 B6 71.7 41.4 Nominally Exposed No. 1--------00 1 6.7 2 13.3 6 40.0 1 6.7 2 13.3 1 6.7 1 6.7 1 6.7 00 15 12.5 37.2 Non-exposed NoT 2 10.5 3 15.8 6 31.6 4 21.1 2 10.5 00 1 5.3 00 1 5.3 00 19 15.8 30.7 MONS 209607 PCS Exposure Category. Lenqth of Employment (years) <5 5- 9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45+ Total (N) Average Exposed Nominally Exposed Non-Exposed No. r No. 5-- 30 34.9 13 86.7 17 89.5 20 23.3 1 6.7 1 5.3 4 4.7 00 00 00 1 6-7 00 00 00 00 3 3.5 00 00 3 3.5 00 00 19 22.1 00 1 5.3 6 7.0 00 00 1 1.2 00 00 36 71.7 15 12.5 19 15.8 17.0 3.8 4.3 MONS 209608 Table 4. - Averages of Selected Biochemical Parameters in 120 Workers by PCS Exposure Category Parameter Triglycerides (mg*) Cholesterol (mg*) HDL Cholesterol (mg*) SGOT (IU/jt) SGPT (IU/i) yGTP (IU/l) Exposed Nominally Exposed Non-Exposed 135.8 128.7 145.5 231.7 238.3 238.2 58.8 57.2 61.7 19.6 19.1 19.4 21.7 21.6 21.3 37.1 34.1 29.8 HONS 209609 Table 5. - Averages and Ranges of Plasnia and Fat PCS Levels in 120 Workers by PCS Exposure Category. Plasma PCS Levels (ppb*) Average Range Fat PCS Levels (ppm**) Average Range Exposed Nominally Exposed Non-Exposed 33.4 10-312 14.2 10-30 12.0 10-27 5.6 1.0-21.6 1.4 1.0-1.a 1.3 1.0-2.0 *parts per billion it# parts per million HONS 209610 Table 5. - Averages and Ranges of Plasma and Fat PCS Levels In 120 Workers by PCS Exposure Category. Plasma PCS Levels (ppb*) Average Range Fat PCS Levels (ppm**) Average Range Exposed Nominally Exposed Non-Exposed 33.4 10-312 14.2 10-30 12.0 10-27 5.6 1.0-21.6 1.4 1.0-1.8 1.3 1.0-2.0 *parts per billion parts per million HONS 209611 Table 7. - Selected Partial Correlation Coefficients and Probabilities for Selected Variables in 86 Workers Exposed to PCBs. Variables Coefficient Plasma PCS level and triglycerides adjusting for age .54 Plasma PCS level and triglycerides adjusting for length of employment .55 Fat PCS level and triglycerides adjusting for age .20 Fat PCS level and triglycerides adjusting for length of employment .23 Plasma PCS level and SGOT adjusting for age .26 Probability <.001 <.001 ns* ns <.05 *ns not significant HONS 209612 Table 8. - Relationship 8etween Plasma PCS Levels and Frequency of Reported Direct Contact with PC8-Containinq Fluids Never Frequency of Reported Rarely Direct Contact Occasionally Frequently Plasma PC8 Levels (ppb) <10 16 * (4ri) 7 (30'S) 5 {15') 1 (47.) 29 11-20 17 (44%) 7 (305) 9 (27%) a (32%) 41 21-30 5 (13%) 6 (26%) 9 (27%) 5 (20%) 25 >31 1 (3%) 3 (13%) 10 (30%) 11 (44%) 39 100% 23 100% 33 100% 25 100% 25 120 chi-square * 28.83, p <.005 *The figures In parentheses are row percentages. MONS 209613 Table 9. - Relationship 8etween Fat PCS Levels and Frequency of Reported direct Contact with PCB-Containlng Fluids Never Frequency of Reported Direct Contact Rarely Occasionally Frequently Fat PCS Levels (ooml <1.5 3* (73?) 5 421?) 5 (31S) 2 (18t) 20 1.51-2.25 2 (18X) l (8?) 1 6?) 0 (0?) 4 >2.26 1 (9*) 6 (50X) 10 (63X) 9 (82X) 26 chi-square 13.06, p <.05 11 100X 12 100S 16 1001 11 loot The figures In parentheses are row percentages. HONS 209614 Table 10. - Relationship Between Plasma PCB Levels and Union Affiliation of Workers at Maintenance Facility BMWE BRAC MISC Union Affiliation ** JCC I AM SWMIA IBEW <10 5 (565)* 2 (50*!) 4 (575) 1 (75) 4 (125) 0 (05) 1 (45) 17 11-20 2 (225) 21-30 2 (22*) i31 0 (OX) 9 100? 1 (255) n (2P5) 1 (2SX) 0 (05) 0 (OX) 1 (145) 4 loot 7 loot 4 (275) a (535) 2 (135) 15 loot 15 (455) 7 (215) 7 (215) 33 loot (295) 2 (29X) 3 (435) 7 loot 9 (355) 4 (15S) 12 (465) 26 loot 35 24 25 101 chi-square 42.89* p<.001 * The figures In parentheses are row percentages. * See text. HONS 209615