Document 0XRm8wVzx96MzR4MyDLp6yDd

6. TEMPORAL TRENDS 6.1. INTRODUCTION Small amounts of dioxin-like compounds may be formed during natural fires, suggesting that these compounds may have always been present in the environment. However, it is generally believed that greater amounts of these compounds have been produced and released into the environment in association w ith human industrial and combustion practices. As a result, environmental levels are likely to be higher in modern times than in earlier times. The trend of increasing levels may now be reversing (i.e., releases and environmental levels may be gradually decreasing), however, w ith changes in industrial practices. As discussed in Volume 1, the potential for environmental releases of dioxin-like compounds has been reduced as a result of the switch to unleaded automobile fuels and associated use of catalytic converters and reduction in halogenated scavenger fuel additives (remaining uses of leaded fuel include chain saws, logging machinery, and mowers), process changes at pulp and paper mills, improved emission controls for incinerators, and reductions in the manufacture and use of chlorinated phenolic intermediates and products. This chapter describes trends in the levels of dioxin-like compounds that have been observed in various environmental media and foods, as well as evidence of downward trends in exposure to dioxin-like compounds in humans. The downward trend in human exposure is supported by a modeling exercise that reconstructs the most likely past doses of dioxin-like compounds contributing to observed body burdens. Reviews of several studies and the modeling exercise are followed by several observations with regard to temporal trends of dioxin-like compounds. 6.2. SEDIMENT CORE STUDIES OF TEMPORAL TRENDS Questions regarding the contribution made by natural sources to the overall environmental burden of CDDs and CDFs can be partially addressed using the results from analyses of the temporal distribution of CDD/CDFs in sediment core samples. Sediment cores provide a historical record of contaminant inputs into the environment and have been used by several researchers to study temporal trends in CDD/CDF deposition. Studies at various sites in the United States and Europe suggest that environmental DRAFT--DO NOT QUOTE OR CITE 6-1 December 2003 concentrations of CDD/CDFs began to increase rapidly in the 1930s and peaked around the 1970s. Czuczwa and Hites (1984) analyzed sediment core samples taken in Lake Huron by the University of Michigan's Great Lakes Research Station. Sedimentation rates within the core samples were determined using Cs-137 and Pb-210 techniques. Those rates were used as a basis for relating depth of core sample to time of deposition. CDD/CDFs were detected in the core samples with no appreciable degradation over time. The most abundant CDD/CDFs were OCDDs and HpCDD/CDFs. Analysis of sample depth showed that the concentration of CDD/CDFs increased steadily beginning in approximately 1940 and leveled off around 1960. Correlations were observed over time between the levels of CDDs and CDFs in the sediment cores and the total volume of synthetic chloro-aromatics produced by the petrochemical industry in the United States. However, coal consumption did not show a good correlation with CDD/CDF concentrations over time. Czuczwa and Hites (1984) concluded that the history of sedimentation rates of CDD/CDFs in core samples from Lake Huron reflected of atmospheric deposition from the combustion of synthetic chloro-aromatics. In a similar study, Czuczwa et al. (1985a) reported on the temporal variability of CDD/CDFs in sediment core samples taken from a wilderness lake, located in an uninhabited and undeveloped island (Siskiwit Lake, Isle Royale) in Lake Superior. The only mechanism of contaminant input into the lake was believed to be atmospheric transport and deposition. The historical record of CDD/CDF concentration in the core samples showed that CDD/CDFs were virtually absent from the sediments until around 1940. All CDD/CDF homologue groups were detected in sediment samples near the surface, with HpCDDs and OCDDs accounting for the highest percentage of total CDD/CDFs. Comparisons were made between the congener profiles found in the lake sediments and congener profiles found in urban air particles. A correlation coefficient of 0.997 was observed, leading to the conclusion that CDD/CDFs entered the lake system from aerial transport and deposition. Smith et al. (1992, 1993) analyzed sediment core layers from Green Lake, located near Syracuse, New York, to determine temporal trends in the deposition of CDDs and CDFs since the beginning of the industrial era (circa 1860). This deep lake (200-foot depth) is thought to be affected only by atmospheric deposition because no industrial DRAFT--DO NOT QUOTE OR CITE 6-2 December 2003 inputs are present and motorboats are not allowed. Relatively low concentrations of CDDs and CDFs (10 pg/g or less) were observed in sediments deposited from 1860 to 1930. However, concentrations increased rapidly thereafter, reaching a peak in the mid1960s when total CDD concentrations exceeded 1,300 pg/g and total CDF concentrations exceeded 250 pg/kg. The concentrations of CDDs and CDFs have declined rapidly since the mid-1960s and, in 1986-1990, were measured at 750 pg/g as total CDD/CDF. In the most recent samples, HpCDD and OCDD dominated the mixture of CDD/CDFs. This observation is consistent with that of Czuczwa et al. (1985a). The authors speculated that the decline in CDD/CDF concentrations over time may be due to the switch to unleaded fuels for vehicles. Sediment cores from the Hudson River were analyzed by Smith et al. (1995). The results indicated that the subsurface sediment layers, dated between 1950 and 1980, had the highest concentrations of CDD/CDFs. OCDD, HpCDF, and OCDF accounted for the highest percentage of total CDD/CDFs in these cores. Pearson et al. (1995) studied sediment cores collected from the Great Lakes (i.e., Lakes Superior, Michigan, and Ontario) and remote inland lakes. The researchers calculated CDD/CDF accumulation rates as the product of concentration (pg/g) and sedimentation rate (g/cm2-yr). The results of this study indicated that CDD/CDF accumulation began in the 1930s-1940s and peaked in the early to mid-1970s. Lake Ontario (350 to 575 pg/cm2-yr) and Lake Michigan (25 to 100 pg/cm2-yr) were found to have higher CDD/CDF accumulation rates than Lake Superior and the remote inland lakes (5 to 10 pg/cm2-yr). CDD/CDF profiles for these lakes indicated that OCDD dominated. However, the homologue profile for Lake Ontario differed from the other lakes, leading the authors to speculate that different nonatmospheric sources were responsible for the CDD/CDFs found in this lake. MacDonald et al. (1992) observed similar temporal results in Canada. MacDonald et al. (1992) observed that OCDD and PCB-77 concentrations in sediments collected from the Strait of Georgia, British Columbia, began to increase in about 1940, reaching a maximum in about 1970. However, 2,3,7,8-TCDF concentrations did not begin to increase until about 1965 as a result of discharges of chlorine bleach effluent from local pulp mills. Lebeuf et al. (1995) observed decreasing dioxin-like PCB (i.e., PCBs 77, 126, and 169) trends in sediments from the Lower Estuary and Gulf of St. Lawrence, Canada. Two sediment cores were collected from sites approximately 50 km apart and sliced into 25 DRAFT--DO NOT QUOTE OR CITE 6-3 December 2003 samples. PCB concentrations were found to increase with increasing depth. These results indicate that recent inputs to this water body have decreased substantially (Lebeuf et al., 1995). Rappe et al. (1997) analyzed sediment core samples from five lakes in southern Mississippi. The sediment cores were collected from five man-made recreational lakes with no known industrial point source of CDD/CDFs and low atmospheric deposition rates. Cores were subdivided into sections to evaluate temporal trends in deposition of CDD/CDFs. No observable trend for levels of CDDs, homologues, or I-TEQs correlating to the age of the strata could be identified. Recently, EPA/DOE conducted a time-trend study of dioxin-like compounds in sediment cores (Cleverly et al., 1996; Versar, 1996). Cores from 11 lakes/reservoirs were collected, sectioned, and dated using 137Cs and 210Pb dating techniques, and analyzed for CDD/CDFs and PCBs. The lakes were located in various geographic locations throughout the United States (10 within the continental United States and 1 in Arctic/Alaska) and were selected to represent background conditions (i.e., no known CDD/CDF sources). For several of the lakes, dated samples were available for time periods ranging from the 1700s to the present. The results of the study indicated that CDD/CDF and PCB inputs to U.S. lakes have increased over time, with significant increases occurring after the 1930s. This is consistent with the findings of other researchers. In general, a minimum of one order of magnitude increase in concentration occurred from the pre- to post-1930s timeframe. With few exceptions, this observation was consistent for all 2,3,7,8-substituted CDD/CDF congeners, CDD/CDF homologue groups, and PCBs. The observed temporal trends were consistent across lakes (except the Arctic/Alaska lake), especially for lakes within the same geographic region. For some lakes, a downward trend appeared to exist for the most recent periods. The point of inflection for this downward trend varied across lakes, but appeared to occur between the 1950s and 1970s. Figure 6-1 depicts the changes in concentration over time for Beaver Lake, Washington. The data also indicate that the CDD/CDF and PCB profiles in these lakes were similar across all periods and across all lakes. This may suggest that the relative congener-specific inputs from the atmosphere have remained consistent over time and are similar across all geographic regions. Relationships also were observed between CDD/CDF and PCB trends and indicators of anthropogenic activities (i.e., PCP production, leaded gasoline sales, carbon monoxide DRAFT--DO NOT QUOTE OR CITE 6-4 December 2003 emissions, and PCB releases) that may be associated with production of CDD/CDFs or PCBs. However, it should be noted that correlations between these variables do not necessarily reflect causal relationships. Several studies have also evaluated sediment cores from European lakes. Czuczwa et al. (1985b) studied temporal trends in three Swiss lakes (Lakes Zurich, Lugano, and Baldegg). No CDD/CDFs were detected in the sediments prior to 1945, but increasing levels were observed in subsequent time periods. The most abundant congeners found in the uppermost sediment samples were OCDD, HpCDD/CDFs, and OCDF. Using sedimentation rates and CDD/CDF surface sediment concentrations, the estimated accumulation rates were 300 pg/cm2-yr for Lake Zurich, 270 pg/cm2-yr for Lake Lugano, and 190 pg/cm2-yr for Lake Baldegg. The authors noted that the similarities in congener profiles and fluxes of CDD/CDF in Switzerland and the United States may be indicative of similar sources of CDD/CDFs. Similar trends were noted by Beurskens et al. (1993, 1994). These researchers evaluated sediment cores from Lake Ketelmeer, a sedimentation area of the Rhine River in The Netherlands. CDD/CDF concentrations were shown to increase after the 1940s and peak between 1960 and 1980. Similar results were obtained for coplanar PCBs. Sediment cores from two lakes in the Black Forest region of Germany were analyzed for temporal trends in CDD/CDF deposition by Schramm et al. (1994). CDD/CDFs were found to have increased by a factor of 13 since the 1930s. Recently, Hagenmaier and Walczok (1996) evaluated CDD/CDF levels in dated sediment core samples collected from Lake Constance, Germany. Cores were collected in December 1995 and April 1996. Based on a preliminary dating scheme, the results indicated that the I-TEQdf concentrations for these samples began increasing around 1940, reached their peak around 1970 to 1975, and then began decreasing. The homologue profiles for the 1940s sediments were similar to those in recent deposition samples. Using dated sediment core analyses, Alcock et al. (1997a) reported that CDD/CDF concentrations in a remote lake in Scotland began increasing in the 1860s and 1870s and peaked in the 1950s and 1960s. According to Alcock et al. (1997a), concentrations appear to have decreased in recent years. Brzuzy and Hites (1995) reported on changes over time in the CDD/CDF homologue profile for Lake Windermere in the United Kingdom. CDFs accounted for a significant fraction of the total CDD/CDF concentration in sediment DRAFT--DO NOT QUOTE OR CITE 6-5 December 2003 core sections dated 1946-1950. In contrast, sections dated 1988-1992 had a significantly lower fraction of CDFs, and the profile was dominated by OCDD and HpCDD. These results provide evidence that the sources of CDD/CDF deposition have changed over time, with CDFs accounting for a much higher percentage of inputs during the earlier period. Sediment core samples from the Baltic Proper, near Sweden, showed detectable levels of CDD/CDFs dating back as early as 1882 (Kjeller and Rappe, 1994). CDD/CDFs increased slowly in the sediment strata dated between 1882 (92 pg/g) and 1962 (233 pg/g) and then increased rapidly in the 1970s. Total CDD/CDFs were estimated to be 520 pg/g in 1970 and 1,803 pg/g in 1978. CDD/CDFs decreased to a concentration of 1,454 pg/g in the most recent layer, dated 1985. DeWit et al. (1990) also observed that CDD/CDF concentrations were higher in deeper sediment layers from the Baltic Sea. I-TEQdf concentrations in surface sediments (i.e., most recently deposited) were approximately 20 times higher than in sediments collected at depths of 22 to 28 cm. Sediment cores were collected to a depth of 5 -1 8 m at three locations near Osaka Bay, Japan, and analyzed for CDD/CDFs (Sakai et al., 1998). CDD/CDFs found at Yodo River, which is influenced by urban activities, increased slowly from 1980, reached a peak in 1993, and then dropped to a lower concentration before stabilizing. Comparison between the southern and northern parts of Lake Biwa, representing areas more and less affected by human activities, respectively, showed that CDD/CDF concentrations at the southern part of Lake Biwa began increasing around 1955 and increased dramatically in 1964, reaching a peak in 1973. At the northern part of Lake Biwa, CDD/CDF concentrations increased dramatically in the late 1960s and continued to increase until the 1980s, when they leveled off. Homologue profiles for the sediment core sampled from the northern part of Lake Biwa showed that OCDD was the dominant congener since 1842, followed by TCDD after 1935. The concentrations of OCDD and TCDD continued to increase after 1935. The sources of these two congeners were suspected to be herbicides, preservatives, and municipal solid waste incinerators (Sakai et al., 1998). CDD/CDFs have also been detected in remote Arctic sediment core samples, but at very low concentrations (Tan et al., 1993; Vartiainen et al., 1995). Tan et al. (1993) analyzed a sediment core collected from Wonder Lake, Alaska. With the exception of 1,2,3,4,6,7,8-HpCDF and OCDF, which had concentrations of 13 and 15 pg/g, DRAFT--DO NOT QUOTE OR CITE 6-6 December 2003 respectively, in all sediment sections dated between 1590 and 1790, CDD/CDF concentrations were less than 8 pg/g. The authors suggest that these results support de novo synthesis of CDD/CDFs. Vartiainen et al. (1995) found total CDD/CDF levels of 2.29 pg/g in sediments dated 1890 and 55 pg/g in sediments dated 1994. OCDD dominated in these samples, followed by the hepta- and hexa-chlorinated congeners. The results of these sediment core studies provide evidence that deposition of dioxin-like compounds began increasing dramatically after the 1930s and continued throughout the 1960s. Decreases appear to have occurred only during the most recent periods. In all of these studies, the higher chlorinated compounds dominated the homologue profiles. These observations are consistent among cores collected in various locations throughout the United States and Europe. CDD/CDFs have been observed both in relatively remote lakes, as well as in lakes close to industrialized areas. This suggests that atmospheric transport and deposition may be an important mechanism of entry into these lakes. 6.3. TEMPORAL TRENDS IN SOIL, VEGETATION, AND AIR Temporal trends in CDD/CDF and PCB deposition have also been studied in other types of environmental media including soil, vegetation, and air samples. Kjeller et al. (1991) analyzed archived soil samples dated from 1846 to 1986 from semirural plots in the United Kingdom. Herbage samples dated from 1891 to 1988 that originated from a grassland area were also tested for CDD/CDFs (Kjeller et al., 1991, 1996). All CDD/CDF homologue groups were detected in soil and herbage samples from all time periods, and the CDD/CDF concentrations increased over time beginning at about 1900 (Kjeller et al., 1991). The concentrations of total CDDs in soil increased from 31 to 92 pg/g between 1893 and 1986. CDD/CDFs in the vegetation samples remained essentially constant between 1861 and 1945, increased to peak levels in the early 1960s, declined, and then reached a second peak in the late 1970s (Kjeller et al., 1996). CDD/CDFs were about 7 to 8 times higher in the 1960s to 1980s than in the sample from 1891 to 1900. In the most recent sample (1991 to 1993), total CDD/CDFs declined to levels similar to those observed in pre-1946 samples (Kjeller et al., 1996). In a similar study, Alcock et al. (1997b) presented evidence that CDD/CDFs were present in UK soils before the widespread development of chloroaromatics (around DRAFT--DO NOT QUOTE OR CITE 6-7 December 2003 1930s). A previously unopened bottle of soil, collected in 1881 from Rothamsted Experimental Station as part of an agricultural experiment, was analyzed for CDD/CDFs. The soil sample contained 0.7 ppt I-TEQDF, with OCDD, 1,2,3,4,6,7,8-HpCDD, and 1,2,3,4,6,7,8-HpCDF as the dominant congeners. Great care was taken to avoid contamination with modern air and dust for the initial analysis, and the sample was subsequently exposed to laboratory air for 32 days to determine whether current air concentrations would alter the CDD/CDF concentrations detected in the archived soil. The results indicated that such exposure did not alter the soil concentration of CDD/CDF. Modern soil samples collected from the same field plot were found to contain 1.4 ppt ITEQdf. The authors speculated that the increase was presumably a result of cumulative atmospheric deposition of CDD/CDFs. The results also indicated that although the soils from 1881 had lower concentrations of CDD/CDFs, they contained similar congener profiles of CDD/CDF as modern soil. This may indicate long-term persistence of CDD/CDFs in soil, and similarities in source inputs over time (Alcock et al., 1997b). Alcock et al. (1997a) also summarized temporal trends in sediments, archived vegetation, soil, food groups, and direct air measurements from the United Kingdom. They suggested that concentrations in UK media were the highest between the 1950s and the 1970s and were directly related to human activities. Since the 1970s, CDD/CDF concentrations have shown a consistent decline. According to Alcock et al. (1997b), current herbage concentrations are similar to pre-1946 levels, and ambient air monitoring data from London and Manchester indicate that CDD/CDF air concentration have steadily declined since the 1970s. Archived soil, herbage, and air samples collected in the United Kingdom between 1942 and 1992 were used to evaluate changes in PCB emissions over time (Harner et al., 1995). The concentrations of PCB congeners 28, 52, 138, and 153 in these samples rose from near zero in 1935 to a maximum in the late 1960s and then fell steadily to their present levels. Hiester et al. (1995) observed a decrease of CDD/CDF concentrations in Germany's ambient air over a 6-year period. Ambient air samples were collected over 12 sampling intervals from four sites in the heavily industrialized Rhine-Ruhr region of Germany during 1987-1988 and 1993-1994, and analyzed for CDD/CDFs. Total I-TEQDFs for these sites ranged from 0.13 pg/m3 to 0.33 pg/m3 during 1987-1988, and from 0.04 pg/m3 to 0.12 pg/m3 during 1993-1994. Reductions in CDD/CDF I-TEQDFs at these sites ranged from 46 DRAFT--DO NOT QUOTE OR CITE 6-8 December 2003 to 69 percent over the 6-year period (i.e, from 0.22 pg/m3 to 0.13 pg/m3 at Dortmund, and from 0.13 pg/m3 to 0.04 pg/m3 at Kln). These reductions were attributed to abatement actions taken since 1989 (Hiester et al., 1995). 6.4. TEMPORAL TRENDS IN WILDLIFE Temporal trends in CDD/CDFs and PCBs have also been studied in wildlife, including fish and bird eggs. Hebert et al. (1994) analyzed pooled herring gull eggs for CDDs annually between 1981 and 1991. The eggs were collected from colonies in the Great Lakes and the Gulf of St. Lawrence River. Analyses results indicate that CDD levels declined between 1981 and 1984, but that CDD levels have remained relatively constant since 1984. DeWit et al. (1994) evaluated temporal trends in the levels of CDD/CDFs and coplanar PCBs in the biota of Sweden. Guillemot eggs were collected from the Island of St. Karlso in the Baltic Proper (Sweden) between 1969 and 1992, and pike samples were collected from Lake Storvindeln in Lapland, Sweden, between 1968 and 1992. During these time periods, the concentrations of CDD/CDFs and PCBs decreased in both species. Roos et al. (1998) confirmed that PCB concentrations decreased significantly in the Baltic Sea between 1989 and 1997 at an annual rate of 2 -4 percent by analyzing samples from 54 juvenile grey seals caught off the Swedish coast. Roos et al. (1998) also found that the PCB concentrations in herring, cod, and guillemot eggs caught from the Baltic Sea during the period 1969-1996 decreased over time by 9 -1 0 percent annually. Decreases in PCBs since the 1970s were also observed in fish from Finland (Korhonen et al., 1995). Pike samples from both inland lakes and coastal areas of Finland had significantly lower total PCB concentrations in 1994 (1.7 to 2.1 yvg/g) than in 1971 (> 7 to 10 yvg/g). U.S. EPA (1994a) reported a decline in PCB concentrations in lake trout from Lake Michigan since the late 1970s; however, PCB concentrations currently appear to be approaching equilibrium in the Great Lakes system. U.S. EPA (1994a) attributed the decline in tissue concentrations to reductions in pollutant loadings to the water column. The levels of PCBs in coho salmon also declined during the early 1980s, but have remained relatively constant since that time. According to U.S. EPA (1994a), the leveling off of PCB concentrations in the Great Lakes is related to "(1) historically contaminated sediments; (2) tributaries inputs resulting from point sources, spills, and runoff from both DRAFT--DO NOT QUOTE OR CITE 6-9 December 2003 urban and rural areas, and resuspension from contaminated sediments; and (3) atmospheric deposition of pollutants." Hilbert et al. (1997) analyzed cod livers collected from Danish waters between 1973 and 1996 for total PCBs (Aroclor 1260). The results indicated that PCB levels in cod livers decreased during the past three decades. Total PCBs ranged from approximately 4 to 8 mg/kg (fresh weight) in 1973 at five locations in Danish waters to less than 1 mg/kg (fresh weight) in 1996 in those same waters. Huestis et al. (1997) examined the temporal and age-related trends of CDD/CDFs and coplanar PCBs in Lake Ontario lake trout. Archived samples of 4-year-old lake trout, collected between 1977 and 1993 from the eastern basin of Lake Ontario at Main Duck Island (MDI), were analyzed for CDD/CDF/PCBs. Three- to 9-year-old trout were collected from the western end of the basin at Port Credit. The results of the temporal trends analysis indicated that CDD/CDF/PCB concentrations were at their highest levels in 1977 and lowest in 1987. The total TEQDFP-WHO94 was 583 ppt in 1977, compared to 124 ppt in 1993. The most important contributor to the total TEQDFP-WHO94 was PCB 126. This PCB congener contributed between 40 to 50 percent of the total TEQDFP-WHO94, depending on the year examined. The contribution of 2,3,7,8-TCDD accounted for 15 to 20 percent of the total TEQDFP-WHO94. CDD/CDF contaminant profiles for 1977, 1982, and 1991 suggest an increase of the proportion of 2,3,7,8-TCDF as the proportion of 2,3,7,8-TCDD decreases. The study also evaluated age-related effects of CDD/CDF/PCB concentrations. The results of the analyses of 3- to 9-year-old fish indicate that as the age of the fish increases, the level of contamination also increases. Boumphrey et al. (1998) examined 10 gannet eggs taken from Ailsa Craig, a colony in the northern part of the Irish Sea, every 2 years from 1977 to 1992 to analyze temporal trends in total PCB and PCB congener concentrations. The results indicated that the average total PCB concentration was 6.08 yvg/g in 1977 and 2.5 yvg/g in 1992 with an annual rate of decline of approximately 0.25 yvg/g. Individual PCB congeners declined at various rates, which resulted in different congener profiles between 1977 and 1992. 6.5. TEMPORAL TRENDS IN FOOD PRODUCTS Recent studies have evaluated trends in concentrations of dioxin-like compounds in food by analyzing levels of PCBs and CDD/CDFs in food products from different time DRAFT--DO NOT QUOTE OR CITE 6-10 December 2003 periods. The United Kingdom's Ministry of Agriculture, Fisheries, and Food (MAFF) compared the levels of CDD/CDFs found in commercially available cows' milk in 1990 to the levels in samples collected in 1995 (MAFF, 1997a). The 1995 samples were collected from 12 locations in England and corresponded, where possible, to the 1990 sampling locations. Analysis of the 1995 samples was performed on a pool of 105 pints of full fat milk from each location. As shown in Table 6-1, the lipid-based I-TEQDF concentrations in 1990 ranged from 1.1 to 3.3 ppt, while the 1995 I-TEQDF levels decreased to between 0.67 and 1.4 ppt. Table 6-1 also reports PCB levels for the 1995 samples. Lipid-based TEQP-WHO94 levels ranged from 0.75 to 2.2 ppt. No PCB analysis for the 1990 cows' milk samples was presented. Frst and Wilmers (1995) compared the levels of CDD/CDFs found in German dairy products in 1990 to the levels in 120 dairy samples collected in 1994. Over the 4-year period, mean I-TEQDF concentration in milk fat decreased by almost 25 percent from 1.35 ppt to 1.02 ppt. Similar reductions were noted in human milk fat (Frst and Wilmers, 1995). To examine trends in CDD/CDF and PCB concentration in American food products, Winters et al. (1998) analyzed 14 preserved food samples from various decades of the 20th century for 7 dioxin-like coplanar PCBs and the 17 2,3,7,8-substituted dioxin and furan congeners. The authors compared the concentrations found in historical samples to the current dioxin concentrations observed in the national food surveys for beef (Winters et al., 1996a; 1996b), pork (Lorber et al., 1997), poultry (Ferrario et al., 1997), and milk (Lorber et al., 1998). As shown in Table 6-2, all 10 samples, dated from 1957 to 1982, had I-TEQDFconcentrations higher than the current mean concentrations (when nondetects were set to one-half the limit of detection). Similarly, mean TEQP-WHO94 concentrations were higher than current mean concentrations for 12 of the 13 samples taken between 1945 and 1983. If these samples are indicative of past CDD/CDF concentrations, normalized I-TEQDF results suggest CDD/CDF levels 2 to 3 times higher, and PCB levels over 10 times higher during the 1950s, 1960s, and 1970s than current concentrations. As shown in Figures 6-2 and 6-3, I-TEQDF and TEQP-WHO94 concentrations in food products began to increase with the 1957 sample and continued to increase throughout the 1960s and 1970s. The peak concentration was observed in the 1968 poultry sample. This trend in CDD/CDF and PCB concentrations in food products is consistent with the DRAFT--DO NOT QUOTE OR CITE 6-11 December 2003 pattern observed in sediment cores, as discussed in Section 6.2 (Cleverly et al., 1996; Versar, 1996; Smith et al., 1992, 1993, 1995; Czuczwa et al., 1984, 1985a). 6.6. TEMPORAL TRENDS IN HUMAN EXPOSURE Several studies have examined trends in the dietary intake of CDD/CDFs and PCBs. In 1995-1996, MAFF reported on its analysis of Total Diet Study samples collected in 1982 and 1992 from 24 locations in the United Kingdom (UK) (MAFF, 1995, 1996, 1997b). Of the 11 food groups examined, 10 groups represented the major dietary contribution to intake of dioxins and PCBs, and 1 group represented bread because it is a dietary staple. The results indicated that total dietary intake of PCBs and CDD/CDFs by consumers in the United Kingdom decreased dramatically between 1982 and 1992. Dietary intake of CDD/CDFs and PCBs for specific food items was estimated by multiplying the TEQ concentration of CDD/CDFs and PCBs in the food item (calculated by setting nondetects to the limit of detection) by the average daily intake for that food item, as estimated in the UK's National Food Consumption Survey. Total dietary intake was calculated by summing the dietary intakes for all food groups. The estimated upper bound dietary intakes of CDD/CDFs and PCBs by the average adult UK consumer in 1982 and 1992 are presented in Table 6-3. The total dietary intake of CDD/CDFs was estimated to be 240-pg/day I-TEQDF in 1982 and 69-pg/day I-TEQDF in 1992. The total dietary intake of PCBs decreased from 156-pg/day TEQP-WHO94 in 1982 to 46-pg/day in 1992 (MAFF, 1997b). Harrison et al. (1998) also reported on composite human milk samples collected as part of the MAFF study. Lipid-based CDD/CDF I-TEQ levels in Birmingham, England, were 37 ppt in 1987-1988 and 21 ppt in 1993-1994. I-TEQ levels were 29 ppt in Glasglow in 1987-1988 and 21 ppt in 1993-1994. Liem et al. (1997) analyzed duplicate portions of 24-hour diet samples collected in The Netherlands in 1978, 1984 to 1985, and 1994. This study was conducted to estimate the dietary intake of CDD/CDFs and PCBs in the Dutch population 18 years of age and older to evaluate trends in dietary exposures. Dietary intake was estimated by combining the results of the chemical analyses of foods with data on consumption rates from the Dutch National Food Consumption Survey. Liem et al. (1997) reported a significant reduction of CDDs and CDFs in the diet over the three time periods. The mean daily dietary intake of CDD/CDF decreased from 4.2 pg/kg I-TEQDF in 1978 to 1.8 pg/kg I- DRAFT--DO NOT QUOTE OR CITE 6-12 December 2003 TEQdf in 1984-1985 and 0.5 pg/kg I-TEQDF in 1994. When PCBs were included in the TEQ calculation, the daily dietary intake was 11 pg/kg TEQdfp-WH0 94 in 1978, 4.2 pg/kg TEQdfp-WH0 94 in 1984-1985, and 1.4 pg/kg TEQdfp-WH0 94 in 1994. The percentage decrease in 1994 samples was consistent for all the measured CDD/CDF/PCBs. The results of this study suggest that a reduction in dietary ingestion of CDD/CDFs and PCBs occurred in The Netherlands beginning in the late 1970s. In a study similar to the United Kingdom's MAFF (1995, 1996, 1997b) study, Frst and Wilmers (1997) found that CDD/CDF dietary levels also dropped in Germany in recent years. Several hundred food samples were randomly collected and analyzed for CDD/CDFs during 1989 and 1995. Fish products showed the greatest decline in CDD/CDF food concentrations over this period. Significant decreases also were noted for meats. Samples of more than 300 dairy products were collected in 1990 and 1994 from several dairies in North Rhine and Westphalia and analyzed for CDD/CDFs. The results indicate that from 1990 to 1994, CDD/CDF levels in cows' milk decreased approximately 25 percent. To test whether the reduction of CDD/CDF concentrations in food had a positive effect on human exposure, dietary intake of CDD/CDFs was estimated using the results of the food sample analysis described above and standard food consumption data. This analysis indicated that in the past few years, CDD/CDF intake by humans decreased by approximately 50 percent. The current average daily intake is estimated to be 69.6 pg I-TEQdf compared to a daily average intake of 127.3 pg I-TEQdf in 1990. This decrease in daily intake was also reflected in a decrease in breastmilk concentrations (Frst and Wilmers, 1997). A study of more than 1,000 individual breastmilk samples from the North Rhine-Westphalia region showed a decrease from 34 ppt I-TEQdf in milk fat in 1989 to 14.2 ppt I-TEQdf in milk fat in 1996. This represents a 60 percent reduction since 1989. 6.7. TEMPORAL TRENDS IN HUMAN BODY BURDENS OF DIOXIN-LIKE COMPOUNDS IN THE UNITED STATES Long-term, nationally representative environmental monitoring for dioxin-like compounds has not been conducted. However, this section reviews various dioxin body burden monitoring studies which have been conducted in the United States in past decades. Studies which sampled either blood or adipose tissue were compiled; studies on DRAFT--DO NOT QUOTE OR CITE 6-13 December 2003 other matrices such as breast milk were not included. Also, this compilation only included dioxin and furan congeners; dioxin-like PCBs were not included. Other characteristics of studies included in this compilation are: 1. All or a toxicologically significant subset of the 17 dioxin-like dioxins and furans were measured and individual-specific or overall study average concentrations of these congeners were available (this was needed in order to calculate TEQDF WHO98 concentrations). 2. All results were reported on a lipid-basis. 3. Although not possible for all studies, data were obtained which had reported detection limits so that average congener concentrations could be derived assuming non-detects (NDs) equal one-half detection limit ( / DL). When detection limits were not supplied, NDs were assumed to equal zero. In only one study from the 1970s (Kang et al., 1991; U.S. EPA, 1990), NDs were reported as zero and detection limits were not supplied. In this case, it made no difference in the calculation of TEQDF -WHO98 concentrations. 4. All measurements were on background populations. The first study described below included Vietnam Veterans, non-Vietnam Veterans, and civilians. All three groups were included in this analysis as the concentrations in the three were indistinguishable - the civilian population had, in fact, the highest concentrations. Another reason to include all three populations was that this was the only study available with the full suite of congener concentrations in the 1970s. Two other studies (Schecter et al., 1989; Schecter et al., 1993) were comprised of blood samples taken from Vietnam Veterans only. In both of these, small subsets of individuals appeared to have elevated concentrations of 2,3,7,8-TCDD and TEQ. Congener-specific averages, not including these individuals, could be developed with the information in the articles, so these amended profiles were included in this compilation. 5. Only studies which sampled adult populations were included; data on children or infants were not included. Studies on "average adult populations" which included demographic information were comprised of approximately the same number of males and females and the average age was usually in the 40s. Studies on DRAFT--DO NOT QUOTE OR CITE 6-14 December 2003 Vietnam Veterans were all male populations and the maximum age was in the 40s, and one study that sampled breast adipose tissue was on females alone. Twelve studies were found for this compilation. A summary of the results from these studies is shown in Table 6-4. Table 6-5 provides the congener-specific average concentrations from each of the years in this compilation. Figure 6-4 shows the TEQ concentrations collected from these studies as a function of year. Figure 6-5 shows the age trend for three studies. This trend refers to the common finding for surveys in the 1980s and 1990s that older individuals had higher dioxin concentrations as compared to younger individuals. Following now are study-by-study summaries. Study 1: Dioxins and Dibenzofurans in adipose tissue o f US Vietnam veterans and controls This study was undertaken by the U.S. Department of Veterans Affairs and the U.S. EPA (Kang et al. 1991; U.S. EPA, 1990). They used EPA's National Human Adipose Tissue Survey (NHATS) Repository to collect samples. In the repository at the time of the study were approximately 8,000 tissue samples, with up to a 1000 a year collected annually starting in 1970. The target population for NHATS was all non-institutionalized persons in the US. Due to the invasive nature of adipose tissue sampling, the population was actually derived from individuals who died from external causes (90%) and surgical patients (10%). From that repository, they were able to identify samples from 494 males who potentially served in Vietnam, those born between 1936 and 1954. These 494 samples were taken between 1972 and 1981. Searching through military records, they were able to identify 134 veterans from this potential population of 494, 40 of whom served in Vietnam. They selected all of these 40, as well as 80 of the remaining 94 veterans who had not served in Vietnam (these 80 selected randomly from the population of 94). They matched each of the 40 Veterans with 2 civilians in terms of age and sex from the potential population of 494. The final population of this study was, therefore, 200 individuals: 40 Veterans serving in Vietnam, 80 non-Vietnam Veterans and 80 civilians. Five samples were discarded as inappropriate (not enough lipid, records of individuals not certain with regard to military status, etc.), for a final sample count of 195. DRAFT- DO NOT QUOTE OR CITE 6-15 December 2003 While a comprehensive set of data, this was not a good background representation because the oldest individual was no older than 45 years old (the last year sampled, 1981, minus the earliest birth date, 1936), and they were all males. Still, it was the only data set from these years which measured all the 17 CDD/CDF congeners, and results suggest that these individuals were not influenced by unique high exposures. The results from these three sample sets were indistinguishable - TEQDF -WHO98 concentrations were 71.9, 65.4, and 72.0 pg/g lipid for the Vietnam Veterans, the non Vietnam Veterans, and the civilians, respectively. The 2,3,7,8-TCDD concentrations followed the same trend, with concentrations of 13.4, 12.5, and 15.8 pg/d lipid for the same three groups. For this reason, these study populations were merged for further analysis here. These full study summaries can be considered to represent the years 1972 to 1981, for this limited male population. It would be desirable to compile concentrations by sampling year, to see if there was a temporal trend in the data. However, the full sample-specific study results in U.S. EPA (1990) did not indicate the year in which each of the samples were taken. This information was retrieved from EPA's files by personal communication (personal communication, J. Remmers, Office of Prevention, Pesticides, and Toxic Substances, U.S. EPA, to M. Lorber, Office of Research and Development, U.S. EPA, 2000), which allowed for the year-by-year compilations shown in Table 6-4. Only 178 of the 195 samples could be identified by year of sample collection. As seen in Table 6-4, there appears to be a clear trend, with concentrations (in ppt) declining from the 80s to the 50s between 1972 and 1981. The 5 samples averaging 129 ppt TEQ in 1976 appear to be anomalous. There were other samples with concentrations above 100 ppt, such as a sample at 151 ppt TEQDF -WHO98 from 1981, at 166 ppt TEQdf -WHO98 from 1978, and at 179 ppt TEQDF -WHO98 from 1973, but 4 of 5 samples in 1976 ranged from 99 to 215 ppt TEQDF -WHO98. There is no reason to believe that the high average in 1976 is indicative of any exposure for that year. Data from 1976 notwithstanding, these data are the earliest that could be found and seem to clearly suggest a declining trend of TEQs throughout the 1970s, starting as high as in the 80s ppt TEQ lipid during the early 1970s, and declining to the 50s ppt TEQ lipid by the late 1970s. DRAFT--DO NOT QUOTE OR CITE 6-16 December 2003 Study 2. Control samples taken to compare to individuals exposed to a PCB transformer fire in a building in upstate New York Schecter et al. (1986) sampled intra-abdominal and subcutaneous adipose tissue in five individuals in Binghamton, during 1983 and 1984 (these samples were assumed to be taken in 1983 for tabular and figure display). One individual had been exposed to fumes from a PCB transformer fire in an office building, but that individual is not included here. Two of the remaining 4 individuals had died from unknown causes - two types of adipose tissue were sampled and averaged for purposes here. Subcutaneous adipose tissue samples were taken from an additional two controls. No information was available as to the age and sex of these four individuals. Schecter et al. (1986) reported the concentrations of 14 of 17 of the CDD/CDF congeners; 1,2,3,7,8-PCDF; 2,3,4,6,7,8HxCDF; and 1,2,3,7,8,9-HxCDF were not reported. The average of the four control samples was 34 ppt TEQDF-WHO98 lipid. Study 3. Background population o f St. Louis, Missouri Graham et al. (1986) reported on the analysis of adipose tissues from autopsy patients who had died suddenly or violently out of the hospital during 1985. It was unclear as to whether to include the results from this study in this compilation, since only these 6 CDD/CDF congeners were measured: 2,3,7,8-TCDD; 1,2,3,7,8-PCDD; 1,2,3,6,7,8-HxCDD; 1,2,3,4,6,7,8-HpCDD; OCDD; and 2,3,4,7,8-PCDF. However, a comparison between these results and the results from the previous two studies suggests that these six congeners comprise over 90% of total TEQDF-WHO98. For this reason, TEQDF-WHO98 concentrations were calculated and included in this compilation. Graham et al. (1986) also listed age and sex in the tabular summary of results. Of the 35 samples, 16 were male and 19 were female, and the age range was from 21 to 88 years, with an average age of 43. The average (congener limited) TEQDF -WHO98 concentration from this population was 47 ppt lipid. Figure 6-5a shows the results from this study, with concentrations displayed as a function of age. There is a clear age trend, with concentrations ranging from less than 20 ppt TEQDF -WHO98 lipid for the youngest sampled individual to over 100 ppt TEQDF-WHO98 lipid for the oldest sampled individual. The highest sample was about 140 ppt TEQDF -WHO98 lipid for a 60 year-old individual. DRAFT- DO NOT QUOTE OR CITE 6-17 December 2003 Study 4: Four background individuals from Atlanta, Georgia Patterson et al. (1994) reported on the analysis of mono- and di-ortho substituted polychlorinated biphenyls, dioxins, and furans in serum and adipose tissue samples collected in Atlanta. The results included from this study for the present purposes were adipose tissue samples from four individuals who had died suddenly. These four included 2 men and 2 women, ages 19, 25, 35, and 55 years. Patterson et al. (1994) reported on 13 of 17 congeners, not evaluating 2 hexa- and 1 hepta-furan congeners. As above, the reported results comprised the bulk of the TEQ. Therefore, these data were included in the current compilation. The average TEQ concentration from this small set was 31 ppt TEQdf -WHO98 lipid. Study 5: NHATS 1987 The National Human Adipose Tissue Survey of 1987, FY87 NHATS, was undertaken, in part, for comparison with a very similarly designed survey in 1982, FY82 NHATS. For both surveys, composite samples of adipose tissue from individuals characteristic of average background conditions were analyzed for the full suite of dioxin and furan congeners. FY82 NHATS was not included in this temporal trend compilation because there appears to have been unexplained high findings of 1,2,3,7,8-PCDD. Table 6-6 compares average congener concentrations of selected dioxin congeners from FY82 NHATS and FY87 NHATS for the 15-44 year age group. As seen in Table 6-6, all dioxin congeners had reasonably comparable concentrations, except 1,2,3,7,8-PCDD. The finding of 125.0 ppt for 1,2,3,7,8-PCDD is much higher than results for that congener for any of the studies found here, even the highest ones in the 1970s. The overall average 1,2,3,7,8-PCDD concentration for FY82 NHATS at 73.6 is itself significantly higher than ever found anywhere. For that reason, FY82 NHATS was not included in this compilation. Results from 38 of the 48 composite samples from NHATS '87 were used. These were results for the adult population, described in the age groups: 15-44 and > 4 5 years. The 10 composites representing the < 1 5 year age group were not included. The 38 samples represented 666 individuals (the 48 samples represent 865 individuals). According to the 1980 census, these two age groups comprised 46 and 31% of the population, respectively. Average congener concentrations for these age groups presented in U.S. EPA (1991) were weighted to calculate the adult average concentrations DRAFT- DO NOT QUOTE OR CITE 6-18 December 2003 shown in Table 6-5. Again, not all congeners were reported, but those that were reported comprised over 90% of TEQ concentrations, as suggested by other studies which had results for all congeners. The overall adult average concentration was 37 ppt TEQDFWHO98 lipid. Like other studies, there appears to be an age trend, with the older set having an average of 53 ppt TEQDF-WHO98, while the younger age range had an average that was almost half that at 28 ppt TEQDF-WHO98. Study 6. Vietnam Veterans in Massachusetts in 1988 Schecter et al. (1989) reported on a study in which blood was taken from 28 Vietnam Veterans from the state of Massachusetts. The year in which the sampling occurred was not provided. Therefore, it is assumed to have occurred one year prior to the year of study publication. Of the 28 individuals sampled, it appeared as though 2 individuals had been exposed to Agent Orange in Vietnam or to another source of 2,3,7,8TCDD. The concentration of this congener in their blood was 34 and 29 ppt lipid, and the TEQdf -WHO98 concentrations were 54 and 62 ppt lipid. The other congeners in the blood of these two individuals did not appear elevated. For the other 26 individuals, the average 2,3,7,8-TCDD concentration was 4.8 ppt and the average TEQDF -WHO98 concentration was 29.6 ppt lipid. Study 7: San Francisco and Los Angeles residents in the late 1980s This study was undertaken by the California Air Resources Board for the purpose of determining a preliminary estimate of the body burden levels of CDD/CDFs in the California population, so that future efforts evaluating the impact of specific air sources could use the knowledge gained, both in terms of survey design, implementation, and analysis, as well as the background levels found (Kramer et al., 1989). A total of 57 adipose tissue samples, from an initial target population of 60, were taken from surgical patients who were in surgery for reasons other than cancer. Samples were selected based on the following stratification variables: 1) age - three age groups of 12-34, 35-49, and > 5 0 years were sampled, 2) location - San Francisco and Los Angeles were targeted, 3) sex half of the samples were sought from each sex. The average age of the respondents was about 50 years. The average concentration found was 31 ppt TEQDF -WHO98 lipid. Figure 6-5b shows the results as a function of age. Again, a trend is suggested showing higher DRAFT--DO NOT QUOTE OR CITE 6-19 December 2003 concentrations with age. However, the trend is not nearly as well defined as in the Missouri study. Concentrations ranged from < 1 0 ppt TEQDF -WHO98 lipid for the youngest individuals sampled (about 21 years old) to approximately 80 ppt TEQDF -WHO98 lipid for an individual over 60 years old. Study 8: Pooled blood sample from blood bank donors in upstate New York In a survey article describing various sampling efforts in the United States and around the world, Schecter (1991) presented the results of a pooled blood sample comprised of the blood from 100 individuals. This same profile was identified in Schecter et al. (1991) as the control population for an AIDS study, and having originated from a Syracuse Red Cross Blood Bank. Schecter et al. (1991) claims that this control population contained samples collected from approximately an equal number of males and females. No additional demographic information was available for these individuals. The average concentration reported was somewhat higher at 50 ppt TEQDF -WHO98 lipid, compared to other values presented for the 1980s in this compilation. Study 9: Fifty Michigan Vietnam Veterans Blood samples from 50 Michigan Vietnam Veterans were collected by Schecter et al. (1993) in 1991. The purpose of this collection was to compare impacts of exposure to Agent Orange in these veterans to levels of dioxin in Vietnamese exposed to Agent Orange. The average age of the 50 veterans at the time of sampling was 48 years, with a range of 41 to 66 years. Levels of 2,3,7,8-TCDD and other dioxins were elevated in 6 of these 50 Vietnam veterans. The average TEQDF -WHO98 concentration in these 6 veterans was 85 ppt lipid. The average TEQDF -WHO98 concentration in the other 44 veterans was much lower at 29.1 ppt lipid. The difference in 2,3,7,8-TCDD concentration was more striking: 46 ppt in the 6 veterans versus 4.1 ppt in the other 44 veterans. The average of the 44 veterans was used for this compilation. Study 10: Pooled blood sample from Binghamton Schecter et al. (1997) reported on the result of a pooled blood sample representing 100 adult men and women. The samples were collected at a Binghamton hospital laboratory from specimens ready to be discarded after having been used for routine DRAFT- DO NOT QUOTE OR CITE 6-20 December 2003 medical purposes. The average TEQ concentration from this pooled sample was 32 ppt TEQdf -WHO98 lipid. Study 11: Blood samples from 6 study sites around the country This is the extensive compilation prepared by the Centers for Disease Control (CDC, 2000) that forms the basis of this Reassessment's representation of current background body burdens in the United States. Details on this compilation are provided in Chapter 4, Section 4.3.2 and are not included here. Briefly, these were background populations from site-specific studies. Sites and populations were: Manchester, Missouri (n = 61); Times Beach, Missouri (n = 67); Jacksonville, Arkansas (n = 57); Oregon (n = 9); Wisconsin (n = 93); and North Carolina (n = 29). The average TEQDF -WHO98 concentration is 20 ppt lipid. The age of the individuals sampled were available for 214 individuals located in Missouri, Arkansas, and North Carolina. The average age was 45 years, with a range of 20-70 years. The concentrations of these individuals are graphed in Figure 6-5c. While an age trend is evident, it does appear as though that the trend is much less marked, as compared to earlier years. Study 12. Breast adipose tissue in San Francisco Petreas et al. (2000) presented a summary of results of analyses of breast tissue samples taken from women undergoing breast surgery during 1998. These 45 individuals comprised the control group in a breast cancer case-control study in the San Francisco Bay area. Specific results from each individual were not available in Petreas et al. (2000), but were supplied by personal communication (personal communication from M. Petreas, Hazardous Material Laboratory, California EPA, to M. Lorber, Office of Research and Development, U.S. EPA, 2000). The average age of these women was 45 years, with a range of 28-67 years. The average TEQDF -WHO98 concentration was 25 ppt lipid. Two important trends that can be ascertained from these data are: 1) While certainly not a statistically collected set of data, these studies suggest a downward trend in average adult body burdens of CDD/CDFs from the 1970s through the 1990s. The adult concentrations appear to be in the range of 70 - 90 ppt TEQdf -WHO98 in the early 1970s; in the range of 30 - 50 ppt TEQDF-WHO98 in DRAFT- DO NOT QUOTE OR CITE 6-21 December 2003 the 1980s; below 30 ppt in the 1990s; and perhaps approximately 20 ppt by the year 2000. 2) An age trend that has been identified by others in the literature is displayed is discussed here and displayed in Figure 6-5. Specifically, older individuals appear to have higher body burdens as compared to younger individuals. Based on an examination of three studies in the 1980s (NHATS FY87 and the two studies graphed in Figure 6-5), and one from the 1990s (Figure 6-5c), one might speculate that the age trend is less pronounced through the 1990s. This is probably due to a dose trend discussed earlier in Section 6.7 (i.e., that doses to CDD/CDFs, as exemplified by 2,3,7,8-TCDD, were possibly higher in the 1960s and 70s, declining to current background levels). With lower exposures through the 1990s, older individuals could be in a period of depuration - body burdens are actually declining as they age. For example, individuals who were in their 30s and 40s in the 1970s and 80s may have had concentrations above 40 ppt TEQDF -WHO98, but may have had concentrations less than 30 ppt in the 1990s, as lower background doses could not sustain the higher body burdens. 6.8. ADDITIONAL EVIDENCE OF TEMPORAL TRENDS IN BODY BURDENS Schecter (1991) analyzed liver tissues estimated to be 100 to 400 years old recovered from the frozen bodies of two Native American (Eskimo) women. The women died in their igloo in Point Barrow, Alaska, when they were trapped and frozen by an ice overflow. Oil was used for cooking and heating, and ventilation was poor. One woman had soot-laden lungs. The results indicated that dioxin levels were much lower in these ancient tissues than in livers of people currently living in industrial areas. Tong et al. (1990), as cited in Schecter (1991), found a lipid-based total I-TEQDF level of 0.24 ppt in one of the ancient liver samples. I-TEQ levels (i.e., CDD, CDF, and total) were nondetectable in the other ancient sample. Analysis of two liver samples from modern times showed lipid-based total I-TEQDF levels of 13.3 ppt (Ryan et al., 1986, as cited in Schecter, 1991). Recently, Papke et al. (1997) analyzed 180 whole blood samples collected in Germany in 1996 for CDD/CDFs. The samples were taken only from individuals who had no known exposure to CDD/CDFs (i.e., their only exposure would be through food DRAFT--DO NOT QUOTE OR CITE 6-22 December 2003 ingestion). The results of this study were compared to a similar study conducted in 1994. The results indicated that concentrations declined from 19.1 ppt I-TEQDF in blood lipids in 1994 to 16.5 ppt I-TEQdf in blood lipids in 1996. These results suggest human exposure to CDD/CDFs have decreased over the past several years. Wittsiepe et al. (1998) measured CDD/CDFs in 507 blood samples collected in Germany between 1991 and 1996. The samples were intended to represent the general population (i.e., the individuals were not exposed to dioxin-like compounds as a result of occupational or accidental contacts). The results indicated that blood levels of CDD/CDFs declined significantly between 1991 and 1996, with the more recent concentrations representing approximately one-half the earlier concentrations. The mean lipid-based total CDD/CDF concentrations were 71.8 pg/g in 1991 and 334.4 pg/g in 1996 (Table 6-7). Liem et al. (1996) detected downward trends in the levels of CDD/CDFs in human breastmilk between 1987-1988 and 1992-1993. Breastmilk samples were collected from women in 11 countries as part of a World Health Organization (WHO)-coordinated exposure study. Protocols were developed to ensure that the samples collected during the two periods were comparable. The protocols included criteria for selection of donors, sampling areas, etc. The samples were analyzed for the 17 CDD/CDF congeners, as well as 6 marker PCBs (IUPAC numbers 28, 51, 101, 138, 153, and 180). The results indicated that CDD/CDF levels are decreasing in some countries (Table 6-8). Liem et al. (1996) estimated an overall annual CDD/CDF decrease of 7.2 percent, based on the data from those countries. In a similar study, Schecter et al. (1997) assessed CDD/CDF concentrations in blood and breastmilk samples collected in Germany and the United States during two time periods. More than 100 blood samples were collected in Germany during the years 1989 and 1994 from "persons for whom there was concern about dioxin exposure but abnormal dioxin blood levels were not found." American blood samples were collected in 1984-1989 (male veterans with no dioxin abnormalities) and 1996. The German milk samples were collected in Westphalia in 1991 and 1995, and American samples were two pooled samples from 1988 (Binghamton, NY, and Los Angeles, CA), five individual analyses from the period 1995-1996 (Binghamton, NY), and one pooled sample from 1997 (Binghamton, NY). The results of the study indicated that from 1989 to 1994, the CDD/CDF concentrations in German blood declined from 43 ppt I-TEQDF to 19 ppt I-TEQDF. DRAFT--DO NOT QUOTE OR CITE 6-23 December 2003 CDD/CDF concentrations in United States blood samples declined from 28 ppt I-TEQDF in 1984-1989 to 25 ppt I-TEQDF in 1996. The results of the German breastmilk analyses indicated a decline from 23 ppt I-TEQDF in 1991 to 16 ppt I-TEQDF in 1995. American milk samples showed a reduction from 17 ppt I-TEQDF in 1988 to 9 ppt I-TEQDF in 1995-1997. Kiviranta et al. (1998) measured the concentrations of the 17 toxic CDD/CDF congeners and 6 PCB congeners (IUPAC 28, 52, 101, 138, 153, and 180) in human milk samples from primiparae mothers in Finland. Samples were collected from women in both rural and urban areas between 1992 and 1994 and compared to data from 1987 (Table 6-9). Total lipid-based CDD/CDF concentrations declined from 339 ppt (n = 37) to 217 ppt (n = 28) in rural areas between 1987 and 1992-1994; urban concentrations were similar: 375 ppt (n = 47) and 381 ppt (n = 14) for the two sample periods, respectively. ITEQdf values decreased from 20.1 ppt to 13.6 ppt in rural areas, and 26.3 ppt to 19.9 ppt for urban areas between 1987 and 1992-1994. Total lipid-based PCB concentrations declined from 396 ng/g (n = 37) to 198 ng/g (n = 28) in rural areas, and from 496 ng/g (n=47) to 296 ng/g (n = 14) in urban areas for the same two sample periods. 6.9. A MODELING EFFORT TO RECONSTRUCT PAST DOSES OF 2,3,7,8-TCDD Previous sections in this chapter describe evidence supporting temporal trends in environmental concentrations and human exposure to CDD/CDF/PCBs. Levels of dioxin like compounds appeared to increase in the environment starting from the 1930s through the 1960s, and loadings began to decline perhaps starting in the 1970s to the present. Recent evidence collected on animal food products in the United States (Winters et al., 1998), combined with body burden data, are the best evidence that human exposures to dioxins may have followed the same trends. (See Sections 6.5 and 6.6.) This section describes a third way of evaluating past exposures to dioxins. Pinsky and Lorber (1998) described an effort to statistically reconstruct the pattern of past human exposure to the most toxic dioxin congener, 2,3,7,8-TCDD (abbreviated TCDD), through use of a simple pharmacokinetic (PK) model that included a time-varying TCDD exposure dose. This section summarizes the procedure and presents some key results from this modeling exercise. The original reference (Pinsky and Lorber, 1998) should be obtained for further detail. DRAFT--DO NOT QUOTE OR CITE 6-24 December 2003 A first-order, one-compartment PK model was used to compute an individual's body lipids' TCDD concentration over time. Key inputs for that model include: (1) a timevarying dose of TCDD (expressed in units of pg/kg-day), (2) a fraction of dose absorbed into the body lipid compartment (assumed to be constant), (3) the volume of the body lipid compartment (assumed to be time varying), and (4) a rate of TCDD loss from the lipid compartment (modeled as a function of the percent of body fat). To calculate the rate of TCDD loss, a model was needed to predict how body lipid volumes vary over time, in addition to a model of how overall body weight varied over time. In this modeling exercise, all inputs were fixed except the time-varying dose of TCDD. Using Bayesian statistical approaches, the dose was "calibrated" to best fit a set of data on TCDD concentration in body lipids. These data, shown in Table 6-10, were obtained from studies that focused on persons with no known direct exposure to dioxins and, as such, measured background exposure levels. In terms of this modeling exercise, the most important data from this set, were from the 1970s, suggesting that body lipid concentrations of TCDD were above 10 ppt during those years (VA/U.S. EPA, 1988). Current data from the 1980s into the 1990s show TCDD concentrations below 10 ppt (U.S. EPA, 1991; Michalek et al., 1998; Andrews et al., 1989). The feature of the Bayesian approach that is most relevant to this calibration modeling exercise was the use of constraints on the input functions. In other words, much of the evidence described earlier in this chapter suggests an expected trend on the dose function that was being calibrated in this modeling exercise (i.e., that the TCDD dose may have increased from the 1930s to the 1970s and declined thereafter). An examination of the existing trend data suggests, specifically, the following for the current dose modeling purposes: (1) a peak in environmental levels appears to have occurred in the 1960s or 1970s, (2) early century levels are from 2 to > 3 3 times lower than the peak, (3) late 1980s levels are from 1 to 20 times lower than the peak, and (4) late 1980s levels are higher than early century levels; in all cases, the ratio of peak to 1980s levels is lower than the ratio of peak to early century levels. Also, and importantly for this modeling exercise, the estimate of TCDD exposure dose based on the 1994 release of this dioxin reassessment document (U.S. EPA, 1994b) was 0.17 pg TCDD/kg-day. Using these trends, the following Bayesian "plausibility criteria" were established for calibration modeling purposes: DRAFT--DO NOT QUOTE OR CITE 6-25 December 2003 1. A range of 0.0 to 0.50 pg TCDD/kg-day for the exposure dose in 1990. The same plausible range was used for the 1900 dose. 2. Ranges of 2 to 200 for the ratio of peak to 1900 dose, and 1 to 100 for peak to 1990 dose. 3. Peak year set between 1945 and 1980. Finally, to ensure a smooth exposure curve, a limit of 20 percent was set on the rate of decrease from the peak exposure level going forward or backward 1 year. Pinsky and Lorber (1998) detailed how well the calibrated doses duplicated the measured body burdens shown in Table 6-10. In general, several slightly different calibrated dose functions fit the data equally well. An example of a family of similar dose curves is shown in Figure 6-6. In that figure, the dose appears to increase from the 1940s through the 1960s, then begins to drop through the 1970s, with a baseline level being reached by the 1980s. This qualitatively fits some of the trend data described previously. However, while the calibrated model fits the data reasonably well, it does not fit the data perfectly. Obviously, the lack of a perfect model fit beginning in the 1970s could be partially attributed to the lack of observed body concentrations of 2,3,7,8-TCDD. As seen in Figure 6-6, the dose curves appear to converge after the 1970s, when there were observed data, while the dose curves seem to diverge prior to 1970, when there were no data on which to base the calibration. Also, some data were inconsistent, particularly the National Human Adipose Tissue Study (NHATS) data from 1982 and 1987 (U.S. EPA, 1991). Specifically, in a comparison of the NHATS 1982 and 1987 data, the mean TCDD concentration increased considerably from 1982 to 1987 in the oldest age group (4 5 + ), but decreased considerably in the two younger age groups. Further, the NHATS 1982 data do not display the trend of increasing TCDD concentrations by age that is seen in most other studies done in the 1980s (mean was 6.9 pg/g in the 15-44 age group and 5.5 pg/g in the 4 5 + age group), while the age trend in NHATS 1987 data (mean of 4.4 pg/g in the 15-44 age group versus 9.4 pg/g in the 4 5 + age group) seems exaggerated. These trends are difficult to explain with the current modeling structure; subsequently, all models with a good data to model fit overpredicted the 1982 mean and underpredicted the 1987 mean in the highest age group. More data from prior to and during the 1970s would have provided a much more useful database from which to DRAFT--DO NOT QUOTE OR CITE 6-26 December 2003 calibrate the model; however, no other TCDD body concentration data could be found from those time periods. Despite the lack of a perfect fit of the model to the data, several informative findings resulted from this exercise. 1. The model calibration exercise was regenerated using two changes to the initial Bayesian constraints on the shape of the exposure/dose curve. One change essentially dropped all constraints in order to test whether the imposition of the constraints restricted how well a calibrated dose curve could fit the data. It was found that a best-fit solution for exposure with no constraints provided only an insignificant improvement. The second change was to constrain the exposure dose, making it constant over time. Results showed that the temporally varying dose provided a significantly improved fit to the data, as compared to the constant dose. Further, the best-fit constant dose in this exercise was 0.35 pg TCDD/kgday, compared with the current average adult dose of 0.17 pg TCDD/kg-day, as determined by the previous version of this dioxin reassessment (U.S. EPA, 1994b), and with the revised 0.09 pg TCDD/kg-day of this current reassessment. The result provides strong evidence that past doses were, in fact, higher than current doses. 2. The exposure/dose curves in Figure 6-6 suggest that dioxin exposure followed a sharp bell curve, with a precipitous drop to a flat baseline in dose before 1980. This drop is counterintuitive and probably more the result of the simplicity of the pharmacokinetic model than a real-world trend. However, it may be reasonable to treat some generalizations from these calibrated dose curves as reasonable hypotheses. For example, the late 1960s were estimated as years of peak TCDD exposure, an observation that coincides with peaks found in sediment core studies. The estimates derived suggest that TCDD exposures may have been 20 times higher during the 1960s than the 1980s. Over a 10-year peak period in the 1960s and early 1970s, daily exposures could have been as high as 1.5 to 2.0 pg/kg-day, possibly dropping to as low as 0.10 pg/kg-day and below into the 1980s. Without body burden data, it may be difficult to go much further with the model results. DRAFT--DO NOT QUOTE OR CITE 6-27 December 2003 3. In addition to an exposure dose, the results of this exercise also include temporal body burden levels, as described by body lipid concentrations of TCDD. An example of these results is shown in Figure 6-7. The "specimen year" on the xaxis refers simply to the year in which a cross-section of the population can be examined. For example, in 1986, young individuals have a body burden of about 2 pg/g lipid; whereas older individuals have a body burden exceeding 5 pg/g lipid. The modeled build-up of TCDD in an individual's body can be ascertained by following a curve corresponding to the individual's birth year, shown on the z-axis, and progressing to the left. This figure displays two important trends: (a) body burdens in general tend to be dropping (Schecter, 1991; U.S. EPA, 1991; MAFF, 1995), and (b) body burdens are higher in older individuals than younger individuals (Orban et al., 1994; Andrews et al., 1989; Van der Molen et al., 1996). As seen from specimen years after approximately 1970, body concentrations in individuals of all ages appear to be dropping. At the same time, the cross section of all birthyear populations after about 1970 suggests that concentrations are higher in older individuals. Interestingly, this trend may not have been present in the U.S. population in the mid-1970s and earlier. In the mid-1970s, peak body concentrations appear to have peaked for individuals in their 20s (roughly), with a constant body burden for older individuals. In the 1960s and earlier, differences in body burden do not appear to be a function of age. 4. This model also had a breastmilk feeding component. To search for the best-fit exposure/dose curve, it was arbitrarily assumed that half the population was breast fed and the other half was bottle fed. Therefore, the average concentration for all individuals was calculated as the midpoint of body lipid concentrations modeled with and without breast-feeding. Breast-fed infants were exposed to milk concentrations modeled to occur in 25-year-old females (where breastmilk lipid concentrations were assumed to equal body lipid concentrations). Infants were assumed to be breast fed for 4 months, and their consumption of breastmilk lipids was 26 g/day. Bottle-fed infants were assumed to be exposed to the general exposure dose, which as can be expected, turned out to be much lower than the breast-feeding dose. Model predictions were compared to a limited subset of the available body burden data, specifically to the under-15 age groups for the two DRAFT--DO NOT QUOTE OR CITE 6-28 December 2003 NHATS data sets, where the mean concentrations were 4.2 pg/g lipid (NHATS 1987) and 2.0 pg/g lipid (NHATS 1982). For the under-15 age group in 1982, the predicted mean concentrations in one possible solution were 3.8 pg/g in breast-fed children versus 0.3 pg/g in bottle-fed children; in 1987, the expected means in this age group were 1.8 pg/g for breast-fed and 0.2 pg/g for bottle-fed individuals. Assuming a 50 percent breast-fed rate, predictions yielded averages of 2.0 pg/g and 1.0 pg/g for the 1982 and 1987 NHATS, respectively. Two observations could be made. First, the modeling exercise shows the impact of breast-feeding since modeled predictions of bottle-fed body concentrations were much lower than breast-fed body concentrations (i.e., 0.3 and 0.2 pg/g lipid for bottle fed versus 3.8 and 1.8 pg/g lipid for breast fed). Therefore, if the low body burden found for bottle-fed infants reflects reality, then the NHATS data showing 4.2 and 2.0 pg/mL also show the influence of breast-feeding on the body burden of children. Second, both the model and the data show a drop in the body concentrations of the under15 age group between 1987 and 1982, suggesting a trend toward declining exposures through the 1980s. DRAFT--DO NOT QUOTE OR CITE 6-29 December 2003 REFERENCES Alcock, R.E.; Coleman, P.; McLachlan, M., Johnston, A.E.; Jones, K.C. (1997a) Reconstructing air concentrations and deposition fluxes of PCDD/Fs in the UK. Organohalogen Compounds. 33:88-92. Alcock, R.E.; McLachlan, M.; Johnston, A.E.; Jones, K.C. (1997b) Confirmation of the presence of PCDD/F in UK soils prior to 1900. Organohalogen Compounds. 32:187-191. Andrews, J.S.; Garrett, W.A.; Patterson, D.G.; Needham, L.L.; Roberts, D.W.; Bagby, J.R.; Anderson, J.E.; Hoffman, R.E.; Schramm, W. (1989) 2378tetrachlorodibenzo-p-dioxin levels in adipose tissue of persons with no known exposure and in exposed persons. Chemosphere. 18: 499-506. Beurskens, J.E.M.; Mol, G.A.J.; Barreveld, H.L.; van Munster, B.; Winkels, H.J. (1993) Geochronology of priority pollutants in a sedimentation area of the Rhine River. Environmental Toxicology and Chemistry. 12:1549-1566. Beurskens, J.E.M.; Winkels, H.J.; de Wolf, J.; Dekker, C.G.C. (1994) Trends of priority pollutants in the Rhine during the last fifty years. Water Science Technology. 29:77-85. Boumphrey, R.; Jones, K.; Osborn, D. (1998) PCB congeners in gannet (Su/sa bassana) eggs from the Scottish coast: temporal trends (1977-1992) and spatial comparisons. Organohalogen Compounds. 39:125-127. Brzuzy, L.P.; Hites, R.A. (1995) Estimating the atmospheric deposition of polychlorinated dibenzo-p-dioxins and dibenzofurans from soils. Environmental Science and Technology. 29:2090-2098. CDC (2000). Personal communication from D. Patterson, CDC, Atlanta, GA, to M. Lorber, U.S. EPA, Washington, DC. April, 2000. Cleverly, D.; Monetti, M.; Phillips, L.; Cramer, P.; Heit, M.; McCarthy, S.; O'Rourke, K.; Stanley, J.; Winters, D. (1996) A time-trends study of the occurrences and levels of CDDs, CDFs and dioxin-like PCBs in sediment cores from 11 geographically distributed lakes in the United States. Organohalogen Compounds. 28:77-82. Czuczwa, J.M.; Hites, R.A. (1984) Environmental fate of combustion-generated polychlorinated dioxins and furans. Environmental Science and Technology. 18:444-450. Czuczwa, J.M.; McVeety, B.D.; Hites, R.A. (1985a) Polychlorinated dibenzo-p-dioxins and dibenzofurans in sediments from Siskiwit Lake, Isle Royale. Chemosphere. 14:623-626. DRAFT--DO NOT QUOTE OR CITE 6-30 December 2003 Czuczwa, J.M.; Niessen, F.; Hites, R.A. (1985b) Historical record of polychlorinated dibenzo-p-dioxins in Swiss lake sediments. Chemosphere. 14:1175-1 179. DeWit, C.A.; Jansson, B.; Strandell, M.; Jonsson, P.; Bergqvist, P.A.; Bergek, S.; Kjeller, L.O.; Rappe, C.; Olsson, M.; Slorach, S. (1990) Results of the first year of the Swedish dioxin survey. Chemosphere. 20:1473-1480. DeWit, C.A.; Jamberg, U.G.; Asplund, L.T.; Jansson, B.; Olsson, M.; Odsjo, I.L.; Lindstedt, L.L.; Anderrson, O.; Bergek, S.; Hjelt, M.; Rappe, C.; Jansson, A.; Nygren, M. (1994) The Swedish dioxin survey: summary of results from PCDD/F and coplanar PCB analyses in biota. Organohalogen Compounds. 20:47-50. Ferrario, J.; Byrne, C.; Lorber, M.; Saunders, P.; Leese, W.; Dupuy, A.; Winters, D.; Cleverly, D.; Schaum, J.; Pinsky, P.; Deyrup, C.; Ellis, R.; Walcott, J. (1997) A statistical survey of dioxin-like compounds in the United States poultry fat. Organohalogen Compounds. 32:245-251. Frst, P. (1993) Contribution of different pathways to human exposure to PCDDs/PCDFs. Organohalogen Compounds. 13:1-8. Frst, P.; Wilmers, K. (1995) PCDD/F levels in dairy products 1994 versus 1990. Organohalogen Compounds. 26:101-104. Frst, P.; Wilmers, K. (1997) Decline of human PCCDD/F intake via food between 1989 and 1996. Organohalogen Compounds. 33:116-121. Graham, M., F.D. Hileman, R.G. Orth, J.M. Wendling, J.D. Wilson (1986) Chlorocarbons in adipose tissue from a Missouri Population. Chemosphere. 15: 1595-1600. Hagenmaier, H.; Walczok, M. (1996) Time trends in levels, patterns, and profiles for PCDD/PCDFs in sediment cores of Lake Constance. Organohalogen Compounds. 28:101-104. Harner, T.; Mackay, D.; Jones, K. (1995) Model of long-term exchange of PCBs between soil and the atmosphere in the southern UK. Environmental Science and Technology. 29:1200-1209. Harrison, N.; Wearne, S.; de M. Gem, M.G.; Gleadle, A.; Startin, J.; Thorpe, S.; Wright, C.; Kelly, M.; Robinson, C.; White, S.; Hardy, D.; Edinburg, V. (1998) Time trends in human dietary exposure to PCDDs, PCDFs and PCBs in the UK. Chemosphere. 37(9-12):1657-1670. Hebert, C.E.; Norstrom, R.J.; Simon, M.; Braune, B.M.; Weseloh, D.V.; MacDonald, C.R. (1994) Temporal trends and sources of PCDDs and PCDFs in the Great Lakes: herring gull egg monitoring, 1981-1991. Environmental Science and Technology. 28:1268-1277. DRAFT--DO NOT QUOTE OR CITE 6-31 December 2003 Hiester, E.; Brickman, P.; Bhm, R.; Eynck, P.; Gerlack, A.; Mlder, W.; Ristow, H. (1995) Pronounced decrease of PCDD/PCDF burden in ambient air. Organohalogen Compounds. 24:147-152. Hilbert, G.; Lillemark, L.; Nilsson, P. (1997) PCB in cod liver-time trend study and correlation between total PCB (Aroclor 1260) and CB congeners. Organohalogen Compounds. 32:340-343. Huestis, S.J.; Servos, M.R.; Whittle, D.M.; Van Der Hevvel, M.; Dixon, D.G. (1997) Evaluation of temporal and age-related trends of chemically and biologically generated 2,3,7,8-tetrachlorodibenzo-p-dioxin equivalents in Lake Ontario lake trout, 1977 to 1993. Environmental Toxicology and Chemistry. 16:2154-2164. Kang, H.K., K.K. Watanabe, J.Breen, J. Remmers, M.G. Conomos, J. Stanley, M. Flicker (1991) Dioxins and dibenzofurans in adipose tissue of US Vietnam Veterans and controls. American Journal of Public Health. 81:344-348. Kiviranta, H.; Purkunen, R.; Vartiainen, T. (1998) Levels of PCDD/Fs in PCBs in human milk in 1994 in Finland: decrease in concentrations from 1987 to 1994. Organohalogen Compounds. 38:121-124. Kjeller, L.O.; Rappe, C. (1994) PCDD and PCDF in a sediment core from the Baltic Proper. Organohalogen Compounds. 20:115-120. Kjeller, L.O.; Jones, K.C.; Johnston, A.E.; Rappe, C. (1991) Increases in the polychlorinated dibenzo-p-dioxin and -furan content of soils and vegetation since the 1840s. Environmental Science and Technology. 25:1619-1627. Kjeller, L.O.; Jones, K.C.; Johnston, A.E.; Rappe, C. (1996) Evidence for a decline in atmospheric emissions of PCDD/CDFs in the U.K. Environmental Science and Technology. 30:1398-1403. Korhonen, M.; Mannio, J.; Vartiainen, T. (1995) The levels of PCBs in aquatic biota in Finland. Organohalogen Compounds. 24:453-456. Lebeuf, M.; Gobell, C.; Clermont, Y. (1995) Non-ortho chlorophenyls in fish and sediments of the Estuary and Gulf of St. Lawrence. Organohalogen Compounds. 26:421-426. Liem, A.K.D.; Ahlborg, U.G.; Beck, H.; Haschke, F.; Nygren, M.; Younes, M.; Yrjnkeikki (1996) Levels of PCBs, PCDDs, and PCDFs in human milk. Results from the second round of a WHO-coordinated exposure study. Organohalogen Compounds. 30:268-273. Liem, A.K.D; Hoogerbrugge, R.; Cuijpers, E.J.C.; den Hartog, R.S.; Hijman, W.C.; Linders, S.H.M.A.; Marsman, J.A.; van der Velde, E.G.; Zomer, B. (1997) Trends in dietary exposure to dioxins and PCBs in The Netherlands. Results from a duplicate diet study. Organohalogen Compounds. 33:112-115. DRAFT--DO NOT QUOTE OR CITE 6-32 December 2003 Lorber, M.; Saunders, P.; Ferrario, J.; Leese, W.; Winters, D.; Cleverly, D.; Schaum, J.; Deyrup, C.; Ellis, R.; Walcott, J.; Dupuy, A.; Byrne, C.; McDaniel, D. (1997) A statistical survey of dioxin-like compounds in United States pork. Organohalogen Compounds. 32:238-244. Lorber, M.N.; Winters, D.W.; Griggs, J.; Cook, R.; Baker, S.; Ferrario, J.; Byrne, C.; Dupuy, A.; Schaum, J. (1998) A national survey of dioxin-like compounds in the United States milk supply. Organohalogen Compounds. 38:125-129. MacDonald, R.W.; Cretney, W .J.; Crewe, N.; Paton, D. (1992) A history of octachlorodibenzo-p-dioxin, 2,3,7,8-tetrachlorodibenzofuran, and 3,3'3,3'tetrachlorobiphenyl contamination in Howe Sound, British Columbia. Environmental Science and Technology. 26:1544-1550. Michalek, J.; Rahe, A.; Kulkampi, P.; Tripathi, R. (1998) Levels of 2,3,7,8tetrachlordibenzo-p-dioxin in 1,302 unexposed Air Force Vietnam-era veterans. Journal of Exposure Analysis Environmental Epidemiology. 8(1):59-64. Ministry of Agriculture, Fisheries, and Food (MAFF). (1995) Dioxins in food - UK dietary intakes. MAFF Food Surveillance Information Sheet No. 71. United Kingdom, Food Safety Directorate. Ministry of Agriculture, Fisheries, and Food (MAFF); Joint Food Safety Directorate. (1996) Polychlorinated biphenyls in food - UK dietary intakes. MAFF Food Surveillance Information Sheet No. 89, May 1996. (http://www.maff.gov.uk/food/infsheet/1996/no89/89pcb.htm, September 9, 1998) Ministry of Agriculture, Fisheries, and Food (MAFF); Joint Food Safety Directorate. (1997a) Dioxins and PCBs in retail cows' milk in England. MAFF Food Surveillance Information Sheet No. 136, December 1997. (http://www.maff.gov.uk/food/infsheet/1997/no136/136dioxi.htm, July 16, 1998) Ministry of Agriculture, Fisheries, and Food (MAFF); Joint Food Safety Directorate (1997b) Dioxins and polychlorinated biphenyls in foods and human milk. MAFF Food Surveillance Information Sheet No. 105, June 1997. (http://www.maff.gov.uk/food/infsheet/1997/no105/105dioxi.htm, July 15, 1998) Orban, J.E.; Stanley, J.S.; Schwemberger, M.S.; Remmers, J.C. (1994) Dioxins and dibenzofurans in adipose tissue of the general U.S. population and selected subpopulations. American Journal of Public Health. 84: 439-445. Papke, O.; Herrmann, T.H.; Ball, M. (1997) PCDD/PCDFs in humans, follow up of background for Germany, 1996. Organohalogen Compounds. 33:530-534. DRAFT- DO NOT QUOTE OR CITE 6-33 December 2003 Patterson, D.G., G.D. Todd, W.E. Turner, V. Maggio, L.R. Alexander, and L.L. Needham (1994) Levels of non-ortho-substituted (coplanar), mono- and di-ortho-substituted polychlorinated biphenyls, dibnezo-p-dioxins, and dibenzofurans in human serum and adipose tissue. Environmental Health Perspectives Supplements. 102(1):195204. Pearson, R.F.; Swackhamer, D.L.; Eisenreich, S.J. (1995) Concentrations , inventories, and accumulations of polychlorinated dibenzo-p-dioxins and dibenzofurans in the sediments of the Great Lakes. Organohalogen Compounds. 24:267-271. Petreas, M., J. She, J. Winkler, P. Visita, M. McKinney, P. Reynolds, D. Smith, D. Gilliss, S. Hurley, S. Jeffrey, E. Mahoney (2000) Body burdens of organohalogens in California populations. Organohalogen Compounds. 48: 17-20. Pinsky, P.; Lorber, M.N. (1998) A model to evaluate past exposure to 2,3,7,8-TCDD. Journal of Exposure Analysis and Environmental Epidemiology. 8(2):187-206. Rappe, C.; Andersson, R.; Bonner, M.; Cooper, K.; Fiedler, H.; Lau, C.; Howell, F. (1997) PCDDs and PCDFs in lake sediment cores from southern Mississippi, USA. Organohalogen Compounds. 32:18-23. Roos, A.; Bergman, A.; Greyerz, E.; Olsson, M. (1998) Time trend studies on 3DDT and PCB in juvenile grey seals (Halichoerus grypus), fish and guillemot eggs from the Baltic Sea. Organohalogen Compounds. 39:109-112. Ryan, J.J.; Schecter, A.; Sun, W.F.; Lizotte, R. (1986) Distribution of chlorinated dibenzo-p-dioxins and chlorinated dibenzofurans in human tissues from the general population. In Chlorinated Dioxins and Dibenzofurans and Perspectives. (ed. C. Rappe et al.) Chelsea, Michigan: Lewis Publishing, p. 3. Sakai, S.; Deguchi, S.; Urano, S.; Takatsuki, H.; Megumi, K.; Sato, T.; Weber, R. (1998) Time trends of PCDDs/DFs in sediments from Osaka Bay and Lake Biwa. Organohalogen Compounds. 39:359-362. Schecter, A., J.J. Ryan, G. Gitlitz (1986) Chlorinated dioxin and dibenzofuran levels in human adipose tissues from exposed and control populations. Chapter 4, p. 51-65 in, Chlorinated Dioxins and Dibenzofurans in Perspective, C. Rappe, et al., eds. Lewis Publishing, Chelsea, Michigan. Schecter, A., J. Constable, J. V. Bangert, K.Wiberg, M. Hansson, M. Nygren, C. Rappe (1989) Isomer specific measurement of polychlorinated dibenzodioxin and dibenzofuran isomers in human blood from American veterans two decades after exposure to Agent Orange. Chemosphere. 18:531-538. Schecter, A. (1991) Dioxins and related compounds in humans and in the environment. In: Biological Basis for Risk Assessment of Dioxins and Related Compounds. Banbury Report #35. Edited by M. Gallo, R. Scheuplein, and K. Van der Heijden. Plainview, NY: Cold Spring Harbor Laboratory Press. DRAFT- DO NOT QUOTE OR CITE 6-34 December 2003 Schecter, A.J., B.J. Poiesz, P.W. Brandt-Rauf, O. Papke, M. Ball (1991) Dioxin levels in blood of AIDS patients and controls. Medical Science Research. 19:273-275. Schecter, A., H. McGee, J. Stanley, K. Boggess (1993) Chlorinated dioxin, dibenzofuran, coplanar, mono-ortho, and di-ortho substituted PCB congener levels in blood and semen of Michigan Vietnam Veters compared with levels in Vietnamese exposed to Agent Orange. Chemosphere. 27: 241-252. Schecter, A., O. Papke, P. Furst, J.J. Ryan (1997) Temporal changes in dioxin and dibenzofuran levels in general population human blood and milk from Germany and the United States. Organohalogen Compounds 33: 473-478 (with supplemental data supplied by personal communication, A. Schecter to M. Lorber, February, 2001). Schramm, K.W.; Juttner, I.; Winkler, R.; Steinberg, C.; Kettrup, A. (1994) PCDD/CDF in recent and historical sediment layers of two German lakes. Organohalogen Compounds. 20:179-182. Smith, R.M.; O'Keefe, P.; Aldous, K.; Briggs, R.; Hilker, D.; Connor, S. (1992) Measurement of PCDFs and PCDDs in air samples and lake sediments at several locations in upstate New York. Chemosphere. 25:95-98. Smith, R.M.; O'Keefe, P.W.; Hilker, D.R.; Bush, B.; Connor, S.; Donnelly, R.; Storm, R.; Liddle, M. (1993) The historical record of PCDDs, PCDFs, PAHs, PCBs, and lead in Green Lake, New York - 1860 to 1990. Organohalogen Compounds. 20:215 218. Smith, R.M.; O'Keefe, P.W.; Hilker, D.; Connor, S.; Posner, E. (1995) Direct and indirect contributions of atmospheric PCDDs and PCDFs to Hudson River national estuarine research reserve sediment cores. Organohalogen Compounds. 24:141-145. Stanley, J.S., K.M. Bauer, K. Turman, K. Boggess, P. Cramer. 1989. Determination of body burdens for polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) in California residents. Final Report for the State of California Air Resources Board, Research Division, 1800 15th St, Sacramento, CA 95812. ARB Contract No: A6-195-33. Tan, Y.L.; Kong, A.; Monetti, M.A.; Lau, B.P.Y.; Gubala, C.; Landers, D. (1993) Polycyclic aromatic hydrocarbons and polychlorinated dibenzo-p-dioxins and dibenzofurans in sediment from a sub-Arctic lake in Alaska. Organohalogen Compounds. 309-312. Tong, H.Y.; Gross, M.L.; Schecter, A.; Monson, S.J.; Dekin, A. (1990) Sources of dioxins in the environment: second stage study of PCDD/Fs in ancient human tissue and environmental samples. Chemosphere. 20:987. DRAFT--DO NOT QUOTE OR CITE 6-35 December 2003 U.S. EPA (Environmental Protection Agency). (1990) Dioxins and Dibenzofurans in Adipose Tissue of U.S. Vietnam Veterans and Controls. Department of Veterans Affairs and the U.S. Environmental Protection Agency. EPA-560/5-89-002. U.S. EPA (Environmental Protection Agency). (1991) Chlorinated dioxins and furans in the general U.S. population: NHATS FY87 results. Washington, DC: Office of Toxic Substances. EPA-560/5-91-003. U.S. EPA (Environmental Protection Agency). (1994a) Deposition of air pollutants to the great waters. Research Triangle Park, NC: Office of Air Quality Planning and Standards. EPA-453/R93-055. U.S. EPA (Environmental Protection Agency). (1994b) Estimating exposure to dioxin-like compounds. Washington, DC: Office of Research and Development. EPA/600/688/005cb. Van der Molen, G.; Kooliman, A.; Slob, W. (1996) A generic toxicokinetic model for persistent lipophilic compounds in humans: an application to TCDD. Fundamental Applied Toxicology. 31: 83-94. Vartiainen, T.; Mannio, J.; Korhonen, M.; Kinnunen, K.; Strandman, T. (1995) Levels of PCDDs and PCDFs in dated sediment lake sediments in subarctic Finland. Organohalogen Compounds. 24:369-372. Versar, Inc. (1996) Time-trend analyses of dioxin-like compounds in sediment cores. Draft report prepared by Versar, Inc., for U.S. Environmental Protection Agency, National Center for Environmental Assessment. EPA Contract No. 68-D3-0013. Veterans Administration/U.S. Environmental Protection Agency (VA/U.S. EPA). (1988) Dioxins and dibenzofurans in adipose tissue of U.S. Vietnam veterans and controls. EPA/560/5-89-002. Winters, D.; Cleverly, D.; Meier, K.; Dupuy, A.; Byrne, C.; Deyrup, C.; Ellis, R.; Ferrario, J.; Harless, R.; Leese, W.; Lorber, M.; McDaniel, D.; Schaum, J.; Walcott, J. (1996a) A statistical survey of dioxin-like compounds in United States beef: a progress report. Chemosphere. 32(3):469-478. Winters, D.; Cleverly, D.; Lorber, M.; Meier, K.; Dupuy, A.; Byrne, C.; Deyrup, C.; Ellis, R.; Ferrario, J.; Leese, W.; Schaum, J.; Walcott, J. (1996b) Coplanar polychlorinated biphenyls (PCBs) in a national sample of beef in the United States: preliminary results. Organohalogen Compounds. 27:386-390. Winters, D.L.; Anderson, S.; Lorber, M.; Ferrario, J.; Byrne, C. (1998) Trends in dioxin and PCB concentrations in meat samples from several decades of the 20th century. Organohalogen Compounds. 38:75-77. DRAFT--DO NOT QUOTE OR CITE 6-36 December 2003 Wittsiepe, J.; Schrey, P.; Ewers, U.; Wilhelm, M.; Selenka, F. (1998) Decrease of PCDD/F levels in human blood from Germany (1991-1996). Organohalogen Compounds. 38:211-214. DRAFT--DO NOT QUOTE OR CITE 6-37 December 2003 Table 6-1. Lipid Based Concentrations of CDD/CDFs and PCBs in Samples of Pooled Retail Milk Purchased in the United Kingdom During 1990 and 1995 Location Bristol Cambridge Carlisle Central London Commuter Londonc Crewe Exeter Northallerton Norwich Nottingham Slough Worchester CDD/CDFs (1990)a 2.0 1.4 1.1-1.4 1.4-3.3 1.7-2.1 -- 1.1--3.3 -- -- -- Concentrations (ppt) TEQ (lipid) CDD/CDFs (1995) 0.92 PCBs (1995) 2.2 CDD/CDF/PCBsb (1995/1996) 3.1 0.84 0.67 1.5 0.75 2.3 1.4 0.99 1.7 2.7 0.88 1.4 1.3 1.7 2.1 1.7 2.6 3.5 2.9 1.1 1.7 1.1 2.1 2.8 3.2 1.1 0.80 2.3 1.6 3.5 2.4 1.0 2.3 3.4 Note: I-TEFDFs were used in calculating I-TEQDFs; TEFP-WHO94s were used in calculating TEQP-WHO94s. ppt = parts per trillion a Some locations were sampled twice in 1990. Samples were also purchased from Beverley, Leeds, and Preston in 1990. Fat contents were not measured for the 1990 samples, and the whole milk concentrations were converted assuming 4 percent fat content, which is typical for UK whole milk. b The combined concentrations of CDD/CDFs and PCBs were calculated before rounding. c The individual pints forming this pool were purchased in outer London and the home counties. Source: MAFF (1997a). DRAFT--DO NOT QUOTE OR CITE 6-38 December 2003 Table 6-2. CDD/F and PCB TEQ Concentrations and Percent Differences from Current TEQ Levels Description 1908 Beef ration 1945 Beef and pork 1957 Dried cream 1968 Bacon bar 1968 Deviled ham 1971 Beef 1971 Bacon wafer 1977 Raw chicken 1977 Cooked chicken 1979 Pork slices 1980 Beef steak 1982 Ham slice 1983 Beef in bbq 1983 Turkey w ith gravy i-t e Qdf, pg/g Lipid 0.34 (0.15) 0.98 (0.75) 2.05 (0.81) 3.01 (2.94) 3.73 (3.71) 1.36 (0.02) 1.75 (1.62) 1.29 (1.18) 1.33 (1.20) 1.46 (1.20) 0.94 (0.73) 1.36 (1.04) 0.50 (0.03) 0.55 (0.23) t e q p-w h o 94, pg/g Lipid Percent of Current ITEQDF Levelsa Percent of Current t e q p-w h o 94 Levelsa 0.07 (0.07) 38 (42) 15 (15) 0.36 (0.36) 89 (197) 140 (146) 3.56 (3.54) 244 (96) 827 (824) 1.05 (1.05) 231 (638) 1747 (2620) 0.61 (0.61) 287 (805) 1019 (1529) 2.48 (1.98) 153 (7) 540 (540) 1.98 (1.98) 135 (352) 3301 (4952) 2.72 (2.72) 202 (287) 970 (970) 2.83 (2.83) 209 (292) 1009 (1009) 0.04 (0.04) 1 12 (262) 72 (105) 0.93 (0.93) 106 (207) 203 (203) 0.07 (0.07) 105 (227) 119 (178) 0.79 (0.79) 56 (8) 171 (171) 0.32 (0.31) 85 (57) 113 (113) Note: (results assume ND = % LOD; results calculated at ND = 0 shown in parenthesis) pg/g = picograms per gram a Current CDD/CDF/PCB levels are the levels observed in national food surveys for beef (Winters et al., 1996a, 1996b), pork (Lorber et al., 1997), poultry (Ferrario et al., 1997), and milk (Lorber et al., 1998). Source: Winters et al. (1998). DRAFT--DO NOT QUOTE OR CITE 6-39 December 2003 Table 6-3. Estimated Upper Bound Dietary Intakes of CDD/CDFs and PCBs by the Average UK Consumer in 1982 and 1992 Food Group Bread Other cereal products Carcass meat Offals (internal organs) Meat products Poultry Fish Oils and fats Eggs Milk Milk products TOTAL CDD/CDF Intake (mean) (pg I-TEQDF/person/day)a 1982 1992 34 14 17 16 4 31 15 3 82 73 38 6 22 3 48 17 66 9 240 69 PCB Intake (mean) (pg TEQP-WHO94/ person/day)b 1982 1992 22 13 3 10 3 0.3 0.2 83 41 13 6 40 8 52 28 11 34 7 156 46 Note: Estimated total dietary intakes were calculated before rounding. a MAFF (1995). b Adapted from MAFF (1997b). DRAFT--DO NOT QUOTE OR CITE 6-40 December 2003 Table 6-4. Summary of Studies with Body Burden Data of Dioxins and Furans Year 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1983 1985 Mean TEQ, pg/g lipid 87 89 70 68 129 69 69 67 55 54 34 47 N 7 14 14 14 5 17 17 22 42 26 4 35 1986 31 4 1987 1988 1988 overall: 37 15-44: 27.5 > 45: 53.4 30 37 666 26 57 1989 1991 50 100 29 44 1996 1996 32 100 20 316 1998 25 45 CO II IX 00 00 II II IX IX Ages (years) 18-45 Study #; Reference; Location 1; Kang et al., 1991; U.S. EPA, 1990. Vietnam Veterans, non-Vietnam Veterans, and Civilians. NA (21-88) (19-55) NA 2; Schecter et al., 1986; Binghamton 3; Graham et al., 1986; St. Louis 4; Patterson et al., 1994; Atlanta 5; Orban et al., 1987; U.S. EPA,1991; NHATS '87 XI XI XI II II l NA (12-88) NA (41-66) NA (20-70) (28-67) -P> -p> CJ1 o U l U1 6; Schecter et al., 1989; Massachusetts 7; Stanley et al., 1989; San Francisco and Los Angeles 8; Schecter et al., 1991; Syracuse 9; Schecter et al., 1993; Michigan Vietnam veterans 10; Schecter et al., 1997; Binghamton 11; CDC, 2000; background populations from site-specific studies in MO, OR, WS, AK, and NC. 12; Petreas et al., 2000; San Francisco DRAFT--DO NOT QUOTE OR CITE 6-41 December 2003 Table 6-5. Average Congener Concentrations for Body Burden Studies of Dioxins and Furans (colum ns are study number and sampling year; all results in pg/g lipid) Congener 2378-D 12378-D 123478-D 123678-D 123789-D 1234678-D OCDD 2378-F 12378-F 23478-F 123478-F 123678-F 234678-F 123789-F 1234678-F 1234789-F OCDF WHO- TEQ 1 ;1 9 7 2 22.4 22.1 NA1 170.6 18.5 277.0 1273.4 3.8 1.2 30.9 27.6 11.0 3.7 0 (ND) 52.7 1.9 3.7 87 1 ;1 9 7 3 20.2 24.4 NA1 189.4 24.7 448.1 1768.0 3.2 0.6 25.9 29.3 13.9 4.8 0.2 57.4 1.5 5.9 89 1;1974 14.5 20.1 NA1 168.4 18.1 250.5 1291.4 2.3 0.2 19.5 24.5 11.1 3.1 0 (ND) 44.2 0.8 4.2 70 1; 1975 15.3 18.8 NA1 148.9 17.1 221.5 939.4 1.6 0.3 23.8 17.1 8.7 2.6 <0.1 31.3 0.6 3.7 68 1 ;1 9 7 6 26.8 38.4 NA1 304.6 36.2 588.2 2790.0 4.2 1.3 32.8 34.1 17.0 6.6 0 (ND) 49.7 1.7 2.9 129 1 ;1 9 7 7 11.8 18.1 NA1 152.9 16.5 246.1 1154.7 1.1 0.1 24.4 20.7 11.1 3.4 0.1 35.8 0.8 2.8 69 1;1978 11.6 18.1 NA1 175.0 18.6 308.8 1325.1 1.3 0.2 25.3 24.3 11.5 3.5 0 (ND) 37.2 1.5 1.9 69 1;1979 12.5 18.2 NA1 154.9 17.1 265.7 1294.1 1.2 0.1 25.0 20.8 10.9 3.1 <0.1 32.3 1.1 1.8 67 1 ;1 9 8 0 11.0 14.1 NA1 133.4 14.6 214.8 1004.2 1.9 0.3 18.0 16.6 8.6 2.6 <0.1 30.6 0.7 1.6 55 1;1981 11.9 16.0 NA1 124.5 12.8 178.4 810.4 1.3 0.2 15.6 14.6 7.5 1.7 0.1 23.7 0.4 1.8 54 DRAFT--DO NOT QUOTE OR CITE 6-42 December 2003 Table 6-5. Average Congener C oncentrations for Body Burden Studies of Dioxins and Furans (columns are study number and sampling year; all results in pg/g lipid) (continued) Congener 2378-D 12378-D 123478-D 123678-D 123789-D 1234678-D OCDD 2378-F 12378-F 23478-F 123478-F 123678-F 234678-F 123789-F 1234678-F 1234789-F OCDF WHO- TEQ 2; 1983 3; 1985 4; 1986 5; 1987 6; 1988 7; 1988 8; 1989 9; 1991 10; 1996 1 1 ;1 9 9 6 12; 1998 6.7 8.4 4.4 6.4 4.8 6.2 5.2 4.1 4.2 2.1 3.9 10.3 19.2 11.6 12.9 7.8 12.7 21.0 8.3 9.8 5.2 6.3 NA NA 5.1 NA 9.1 14.0 13.0 NA 10.6 6.2 NA 55.9 107.5 94.2 90.7 64.1 70.1 84.0 76.2 67.9 73.1 57.6 8.0 NA 16.9 13.3 13.3 12.8 15.0 10.9 10.7 7.1 NA 88.2 253.3 55.6 129.7 126.3 124.8 187.0 108.8 116.5 79.2 68.6 579.5 1273.3 446 876.4 1054.6 700.7 1174.0 731.0 879.8 664.0 528.3 1.3 NA 1.1 1.9 NA 2.7 3.1 2.1 ND (2.00) 0.7 NA NA NA NA NA NA 1.0 2.8 1.2 ND (1.9) 0.8 NA 13.9 12.0 3.7 12.0 9.1 8.0 13.0 8.5 9.3 6.2 10.5 13.4 NA 3.7 NA 13.1 8.8 15.0 10.5 14.0 6.5 5.9 8.5 NA 5.8 7.0 7.3 5.4 14.0 5.5 7.9 5.3 4.1 NA NA NA NA 1.9 2.2 3.6 2.9 ND (4.1) 6.2 NA NA NA NA NA NA 0.6 ND (1.2) 2.7 4 0.7 NA 16.0 NA 12.0 NA 26.0 11.0 36.0 19.0 13.9 13.2 NA 9.1 NA NA NA NA 60.1 ND (1.8) 3.2 4.9 1.3 NA 5.8 NA NA NA NA 1.2 4.2 9.9 ND (5.00) 2.1 NA 34 47 31 37 30 37 50 29 32 20 25 1 For these data, 123478-HxCDD and 123678-HxCDD were measured together and reported as, 123478/678-HsCDD. DRAFT--DO NOT QUOTE OR CITE 6-43 December 2003 Table 6-6. Comparison of the 15-44 Age Group Average Concentration of Selected Congeners from NHATS FY82 and NHATS FY87 Congener 2,3,7,8-TCDD 1,2,3,7,8-PCDD 1,2,3,4,6,7,8-HpCDD OCDD NHATS FY82 6.87 125.0 114.0 760.0 Source: U.S. EPA (1991). NHATS FY87 4.33 9.48 99.8 726.0 DRAFT--DO NOT QUOTE OR CITE 6-44 December 2003 Table 6-7. Trends in Blood CDD/CDF Levels in a German Population, 1991-1996 Year Number of samples Mean age (yrs) Fat content (mg/g) Mean total CDD/CDF concentration (pg/g) 1991 95 44.7 5.7 718.4 Source: Wittsiepe et al. (1998). 1992 157 42.4 5.7 703.2 1993 17 40.5 6.0 534.5 1994 74 46.5 6.0 376.7 1995 69 45.2 5.7 431.6 1996 95 37.7 5.1 373.1 DRAFT--DO NOT QUOTE OR CITE 6-45 December 2003 Table 6-8. Comparison o f Results from the First and Second Round of WHO-Coordinated Human Milk Study C o u n try A u s tria Belgium Canada C roa tia Denmark Finland Germany Hungary N ethe rla nd s N orw a yd U n ite d Kingdom Area Vienna (urban) Tulln (rural) Brabant W allou Liege Brussels All Provinces 1981 All Provinces 1982 M aritim es Qubec O n ta rio b Prairies British Columbia Kirk Zagreb Several Regions/Cities Helsinki Kuopio Berlin North Rhine-W estphalia Budapest Scentes Rural Area Urban Area All Regions Trom so (coastal) Hamar (rural) Skien/Porsgrum m (ind) B irm in g h a m G la sg o w CDDs and CDFs (pg I-TEQDF/g) 1 9 8 7 /8 8 a n 1 9 9 2 /9 3 n 17.1 18.6 33.7 4 0 .2 38.8 15.6 18.1 17.6 19.4 23.0 12.0 11.8 17.8 18.0 15.5 32.0 31.6 9.1 11.3 37.4 39.6 34.2 18.9 15.0 19.4 37.0 29.1 54 10.7 13 51 10.9 21 20.8 27.1 26.6 8 20 6 28.6 200 14.5 100 19 10.8 20 34 13.4 20 76 18.1 20 31 14.6 20 23 15.7 20 14 8.4 10 41 13.5 13 42 15.2 48 38 21.5 10 31 12.0 24 40 16.5 10 79 20.7e 100 8.5 20 50 7.8 10 13 13 10 22.4 17 11 10.1 10 10 9.3 10 10 12.5 10 17.9 15.2 20 23 3 [M arker PCBs] (ng/g) 1 9 8 7 /8 8 n 1 9 9 2 /9 3 558 609 500c 450c 830c 150 203 762 416 392 272 562c 507c 533c 381 303 12 275 21 3 0 6 260 212 112 86 137 128 58 70 14 218 41 2 19 10 209 38 189 31 133 375 143 61 45 10 10 96 253 10 2 7 3 (5 3 6 c) 10 2 6 5 (4 c 8 3 c) 8 3 0 2 (4 6 8 c) 129 131 n 13 21 8 20 6 200 100 20 20 20 20 20 10 13 48 10 24 10 20 10 17 10 10 10 20 23 NOTE: Results are expressed on a fa t basis. 3 (m arker PCBs) and I-TEQDFs are calcu la te d assum ing n o n -d e te c t values are equal to zero. a Calculated using Nordic TEF-model. b O n ta rio -1 9 8 8 d enotes prop o rtio na l mean o f tw o pooled sam ples analyzed in th e firs t round. c Analyzed using packed colum n technique. d To com pare results betw een firs t and second round, samples fro m 1 9 9 2 /9 3 have been reanalyzed using (old) packed colum n technique (Becher and Skare, personal com m unication). e Dioxin levels in hum an m ilk sam ples fro m N orth R hine-W estphalia colle cte d in 1 99 2 as reported by Frst (1 99 3 ). Source: Liem et al. (1 99 6 ). DRAFT--DO NOT QUOTE OR CITE 6-46 December 2003 Table 6-9. Comparison of CDD/CDF Concentrations in Human Milk from Finland in 1987 and 1992-1994 Selected Congeners 2,3,7,8-TCDF 2,3,7,8-TCDD 2,3,4,7,8-PeCDF 1,2,3,7,8-PeCDD 1,2,3,6,7,8-HxCDD OCDD 3CDD/CDF I-TEQdf Conc. pg/g fat in 1992-1994 Urban Area n = 14 Rural Area n = 28 1.93 0 .7 4 *** 0.49 0.44 2.66 1.46 1.71 0.68 16.3 7.0* 10.4 4.65 6.22 2.16* 4.36 1.56 33.2 8.94 26.9 8.16 230 80.9 *** 126 55.7 381 1 2 0 * * * 217 76.6 19.9 7.42* 13.6 4.57 Conc. pg/g fat in 1987 Urban Area n = 47 Rural Area n = 37 2.98 2.89 6.75 4.29xxx 3.37 1.85 2.50 1.25x 20.1 12.5 13.1 5.41 9.78 4.87xxx 7.53 3.23xxx 48.2 15.8xxx 41.5 15.3xxx 187 83.6 171 7 1 .5xxx 375 132 339 108xxx 26.3 11.9 20.1 6.54xxx Note: Asterisks indicate a statistically significant difference between urban and rural areas in 199 2 -1 99 4 . * p<0.01 ** p<0.005 * * * p<0.001 x indicates a statistically significant difference between 1987 and 1 9 9 2 -1 9 9 4 results x p<0.01 xx p < 0 .0 0 5 xxx p < 0.0 01 Source: Kiviranta et al. (1998). DRAFT--DO NOT QUOTE OR CITE 6-47 December 2003 Table 6-10. Mean Human Lipid TCDD Concentrations Reported in Various U.S. Studies Study (Reference) Andrews et al. (1989) Air Force (Michalek et al., 1997) NHATS 82 (U.S. EPA, 1991) NHATS 87 (U.S. EPA, 1991) VA/EPA (VA/U.S. EPA,1988) Age/Gender Group 18-29, both 30-39, both 40-49, both 50-59, both 60-79, both 35-39, male 40-44, male 45-49, male 50-54, male 55-59, male 60-64, male 65-69, male 0-14, both 15-44, both 45 + , both 0-14, both 15-44, both 45 + , both 20-36, male 23-39, male 26-42, male 29-45, male Sample Size 14 30 25 22 37 168 280 165 232 142 33 35 178 312 273 146 318 401 27 29 57 82 Year 1986 1986 1986 1986 1986 1987 1987 1987 1987 1987 1987 1987 1982 1982 1982 1987 1987 1987 1971-1973 1974-1976 1977-1979 1980-1982 TCDD Mean, pg/g 4.0 5.9 5.5 8.0 9.5 3.8 4.0 4.6 4.7 4.8 5.0 6.2 4.2 6.9 5.5 2.0 4.4 9.4 19.8 17.3 11.6 12.6 Standard Error of the Mean 0.95 0.65 0.71 0.76 0.59 0.23 0.18 0.23 0.20 0.25 0.52 0.51 0.69 0.87 0.84 0.82 0.52 0.41 1.2 1.2 1.2 1.2 DRAFT--DO NOT QUOTE OR CITE 6-48 December 2003 Sediment Levels, Beaver Lake, Olympic Peninsula, WA Nondetects = 1/2 LOD Residue Levels (pg/g, dw) 1974.4 1964 1955 1946 1932 Year 1921 1909 Total CDDs Total CDFs Total CDD/Fs 1897 Figure 6-1. CDD/CDF Levels in Sediment, Beaver Lake, Washington 1884 DRAFT--DO NOT QUOTE OR CITE 6-49 December 2003 4 A en en --ttCcor----oL'' 2 c 0oc 1 A A A A A A A 0 1900 A _I_ 1920 __ I_________ 1940 Year _I_ 1960 _I_ 1980 Current CDD/CDF C oncentrations (pg I-T EQ/g lip id ): beef=0.89; chicken=0.64; pork=1.30; m ilk=0.84 2000 Figure 6-2. I-TEQDF Concentrations of Historical Food Samples from the U.S. (results calculated at ND = % LOD) Source: Adapted from Winters et al. (1998). Year Current CDD/CDF C oncentrations (pg TEQ-W HO94/g lip id ): beef=0.46; chicken=0.28; pork=0.06; m ilk=0.43 Figure 6-3. TEQP-WHO94 Concentrations of Historical Food Samples from the U.S. (results calculated at ND = % LOD) Source: Adapted from Winters et al. (1998). DRAFT--DO NOT QUOTE OR CITE 6-50 December 2003 140 120 Q. 100 o 80 'S 60 O 40 20 0 1960 1970 1980 Year 1990 2000 Figure 6-4. Average Adult Population TEQDF -WHO98 Concentrations as a Function of Year (all results in ppt lipid). DRAFT--DO NOT QUOTE OR CITE 6-51 December 2003 (a) Missouri, 1985 (n = 35), average = 47.0 ppt lipid (Graham et al., 1986) (b) California, 1988 (n = 57), average = 31.0 ppt lipid (Orban et al., 1989) 160 140 I . 120 100 Q. 80 o 60 40 20 0 0 20 40 60 Age 80 100 (c) CDC data, 1995-1997 (n - 216), average = 18.3 pg/g lipid (CDC, 2000) 160 140 1.120 100 80 60 40 20 0 0 20 40 60 Age Figure 6-5. Age Trend Relationships for Three Studies. 80 100 DRAFT--DO NOT QUOTE OR CITE 6-52 December 2003 e pg/kg-day -2logProb(YI0) ------------147.9 -----------------148.5 TM ............ 149.1 ----------- 149.4 148.9 Figure 6-6. Examples of Temporal Exposure Curves for 2,3,7,8-TCDD, e(t) in Units of pg/kg-day. Source: Pinsky and Lorber (1998). DRAFT--DO NOT QUOTE OR CITE 6-53 December 2003 TCDD pg/g lipid 1956 Specimen Year 1946 1963 Birth Year Figure 6-7. Predicted Mean TCDD Lipid Concentrations (pg/g) in Males by Birth Year and Specimen Year Derived Using e(t) Curve Labeled A in Figure 6-6 Source: Pinsky and Lorber (1998). DRAFT--DO NOT QUOTE OR CITE 6-54 December 2003