Tobacco Documents Online

, i Toxicology and Industrial Health, Vol. 7, No. 3, 1991 141 BENZO-a-PYRENE: ENVIRONMENTAL PARTITIONING AND HUMAN EXPOSURE HOLLY A. HATTEMER-FREY' AND CURTIS C. TRAVIS2 'Lee Wan & Associates, Inc. 120 S. Jefferson Circle, Suite 100 Oak Ridge, TN 37830 ZOffice of Risk Analysis Health and Safety Research Division Oak Ridge National Laboratory* P.O. Box 2008, Building 4500S Oak Ridge, TN 37831-6109 A multimedia transport model was used to evaluate the environ- mental partitioning of benzo-a pyrene (BaP). Measured and pred dicted environmental concentrations were used to estimate the accumulation of BaP in the food chain and the subsequent ex- tent of human exposure from inhalation and ingestion. Results show that BaP partitions mainly into soil (82%) and sediment (17%) and that the food chain is the dominant pathway of hu- man exposure, accounting for about 97% of the total daily in- take of BaP. Inhalation and consumption of contaminated water are only minor pathways of human exposure. The long-term av- erage daily intake of BaP by the general population of the U.S. is estimated to be 2.2 micrograms (ug) per day. Cigarette smok- ink and indoor activities do not substantially increase human ex- posure to BaP, relative to exposures to background levels of BaP present in the environment. Since the increased lifetime risk as- sociated with human exposure to backkround levels of BaP is 3.5 x 10-°, we conclude that ingestion of food items contami- nated with BaP may pose a serious health threat to the U.S. population. 1. Correspondence should be addressed to: Holly A. Hattemer-Frey, Lee Wan & Associates, 120 South Jefferson Circle, Suite 100, Oak Ridge, TN 37830. 2. Key words: benzo-a-pyrene; food chain; risk; human exposure; environmental partitioning. "Research was sponsored by the U.S. Environmental Protection Agency under Interagency Agreements appucaoie unuer rvtanm rvmaneua nnergy aystems, rnc., ~.onuact t.v. uc-r.wra.vcc<<<+w wtm tne u.a. > Department of Energy. Toxicology and Industrial Health, 7:3, p. 141-157 Copyright 1991 Princeton Scientific Publishing Co., Inc. ISSN: 0748-2337

The combustion of fossil fuels is the primary anthropogenic source of background levels of polycyclic aromatic hydrocarbons (PAHs) present in the environment (Archer et al., 1979; Edwards, 1983). While there is no known intentional pro- duction or use of the most toxic PAH, benzo-a-pyrene (BaP), it has been detected in virtually all environmental media and food items consumed by animals and humans (Archer et al., 1979; Blummer, 1961; Faoro, 1975; Fazio and Howard, 1983; Lintas et al., 1979; Malanoski et al., 1968; Panalakas, 1976; Suess, 1976), indicating that environmental contamination of BaP is widespread. Because BaP is a potential human carcinogen, the environmental fate and accumulation of this compound in the food chain is of particular concern. This paper evaluates the environmental partitioning of BaP and identifies the major sources of human exposure. ENVIRONMENTAL PARTITIONING Various measurement and predictive techniques can be used to evaluate the movement and transfer of chemicals within and between environmental media. Multimedia partitioning models, for example, estimate the long-term, steady- state concentration of pollutant in various media. One such model, Mackay et al.'s (1985a,b) Level III Fugacity model, treats the environment as a "unit world" (a hypothetical region equal to 1 kmz) divided into six homogenous compart- ments: air, water, soil, bottom sediment, suspended sediment (in water), and aquatic biota (fish). Under conditions of continuous release, the model can be used to' estimate unknown (or nondetectable) concentrations in certain media from a chemical's physicochemical properties and from known (or detectable) concentrations in other media, thus providing a coherent account of concentra- tions in all media. Mackay et al.'s (1985a,b) Fugacity model was modified to account for uptake of chemicals through the food chain (Travis and Hattemer-Frey, 1987; Hattemer- Frey and Travis, 1989; Hattemer-Frey et al., 1990). The modified version, called the Fugacity Food Chain (FFC) model, estimates the concentration of chemical in all six media and then uses those concentrations to predict the amount of chemical that will accumulate in the food chain and the average daily intake by the general population. It should be emphasized that the FFC model is not an exact replica of the environment, since it contains many simplifying assumptions. Given the current state of knowledge concerning the environmental transport of chemicals, however, fugacity models are considered acceptable for exploring the steady-state partitioning and environmental behavior of organic chemicals chronically released into the environment (Cohen and Ryan, 1985; Mackay et al., 1985 a,b). In} inc liv frc ph toI DE Ba on ph hr- ph 2.4 10- Ah or wa* tha rad Th, (U, soi' (19 E>r Alt me. (Fa 4:' Phy Log cc Wat Vap Mol Her Bioc k Soil i Dail Dail Wat

Toxicology and Industrial Health, Vol. 7, No. 3, 1991 143 ~ I Input parameters required to predict the steady-state partitioning of a chemical include: 1) its physicochemical properties; 2) its bioaccumulation potential in living organisms; 3) degradation rates for processes that remove the compound from the system; and 4) an estimate of emissions into air, water, and soil. The physicochemical properties of BaP and its bioconcentration and biotransfer fac- tors (BCF and BTF) are presented in Table 1. Degradation Rate Estimates BaP released into the atmosphere may remain in the vapor phase or may sorb onto particulates. The fraction that remains in the vapor phase can undergo rapid photolysis with a half-life of six days (degradation rate coefficient of 4.81 x- 10-3 hr- ') (U.S. EPA, 1986). Since only 0.5% of BaP is assumed to exist in the vapor phase (Bidleman, 1988; Mackay et al., 1986), a degradation rate coefficient of 2.41 x 10'S hr-1 was used to represent the degradation of BaP in air (4.81 x 10-3 x 0.005). Although most of the BaP that partitions into water sorbs onto organic;matter or equilibrates with biota (U.S. EPA, 1984), microbial degradation of BaP in water can be an important degradative pathway. Herbes et a1. (1979) reported that the half-life of BaP in water is about 2.3 years (20,148 hr). Hence, a deg- radation rate coefficient of 3.4 x 10-5 hr-' was used in this analysis. The dominant mechanism of PAH removal from soil is microbial degradation (U.S. EPA, 1984). A degradation rate coefficient of 5.3 X 10`6 hr-' for BaP in soil was estimated from the half-life value of 15 years reported by Lu et al. (1977). Emission Rate Estimates Although the exact pattern of BaP input into the environment is not known, measured background concentrations of BaP in air [2.5 nanograms (ng)/m3; (Faoro, 1975)), water [8.7 ng per liter (L); (Santodonato et al., 1981)], and soil TABLE 1 Physicochemical properties of BaP Physicochemical properties Log octanol-water partition coefficient (log K,,) Value 5.98 Wate t b'i't rsou iiy 1 5 x 10-` mol/m' Vapor pressure 7.3 x 10-' Pa Molecular weight 252.3 Henry's Law'constant 1.6 x 10-" atm-m'/mol Bioconcentration (BCFs) and biotransfer factors (BTFs) Soil-to-root BCF 5.6 x 10-' Daily intake-to-beef BTF 4.0 x 10-'- d/kg Daily intake-to-milk BTF 1 . 2 x 10-= d/kg Water-to-fish BCF 1110 Reference Mackay and Paterson, 1990 Mackay and Paterson, 1990 Mackay and Paterson, 1990 Mackay and Paterson, 1990 U.S. EPA, 1986 Edwards, 1983 Travis and Arms, 1988 Travis and Arms, 1988 this paper

[3.0 micrograms (f.cg) per kg; (Shabad et al., 1971)] can be used to predict its cross-media partitioning using the methodology described in Travis and Hattemer-Frey (1990). The FFC model is calibrated by varying emission rate estimates until the concentration of chemical predicted by the model is consistent with measured background concentrations for the chemical in air, water, and soil. The following emission rates were found to best reproduce the available environmental data: Air = 1.32 x 103 mol/hr (93%) Water = 6.74 x 10° mol/hr (<1%) Soil = 9.55 x 10' mol/hr (7%). These data show that the majority (93%) of BaP entering the environment is emitted into the atmosphere, which is consistent with the findings of Suess (1976), who estimated that most BaP produced is released into the atmosphere. Thus, our best estimate of BaP emissions into the contiguous U.S. is 3.1 x 106 kg/year, which agrees well with estimates of 3.2 x 106 kg/year (Suess, 1976) and 1.2 x 101 kg/year (NAS, 1972). Environmental Fate of BaP The low solubility, low vapor pressure, and high octanol-water partition coef- ficent of BaP result in its partitioning mainly between soil (82%) and sediment (17%), with about 1% partitioning into water, and less than 1% into air, sus- pended sediment, and biota. The environmental profile agrees reasonably well with data reported by Ryan and Cohen (1986), who estimated that 99% of environmental BaP partitions into soil. Table 2 gives the predicted environmental concentrations for BaP. These values are approximations that are limited by accuracy and availability of reported data. Predicted concentrations show rea- sonably good agreement with the range of background values cited. ACCUMULATION OF BaP IN THE FOOD CHAIN The food chain is the primary pathway of human exposure to a large class of lipophilic compounds, including pentachlorophenol (PCP), dioxin (TCDD). DDT, and most pesticides (Beck et al., 1989; Hattemer-Frey and Travis, 1989; Lioy et al., 1988; Stevens and Gerbec, 1988; Travis and Arms, 1987; Travis and Hattemer-Frey, 1987). Although data on the occurrence of BaP in food are scarce, it has been reported to occur naturally in many food items including fruits, vegetables, and eggs (Archer et al., 1979). The FFC model was used to estimate the extent of human exposure to BaP from ingestion of agricultural produce (fruits, grains, and vegetables), beef, milk, fish, and eggs. [Note that for computational purposes, we use up to three significant figures to report model predictions. The reader should be aware that model estimates may not be ac- curate to three significant figures!] Com - Air" Water° soit Sedimen Suspend Fish/mo Fish Leafy ve ! Root crc Jrains a

Toxicology and Industrial Health, Vol. 7, No. 3, 1991 145 Z;. : TABLE 2 Comparison of predicted and measured environmental concentrations for BaP Phase Air' Water' ; Soil Sediment Suspended sediment Fish/mollusks Fish Leafy vegetables Root crops/fruits Grains and cereals ~. Forage Predicted concentration (p.g/kg)• Measured concentration (p.g/kg) Reference 2.5 2.5 Faoro, 1975 0.5-2.9 Santodonato et al., 1973 3.2 Suess, 1976 3.2 Fox and Staley, 1976 3.5 Grimmer, 1983 8.8 0.3-4.0 Basu and Saxena, 1978 0.6-114 1ARC, 1973 8.7 Santodonato et al., 1981 8.4 Andelman and Suess, 1970 5.6 Bometh and Kunte, 1983 20 Suess,1976 3.0 1.2-3.0 Shabad et al., 1971 . 1.5 Borneff and Kunte, 1964 0.4-4.0 Shcherbak and Logan, 1970 40 Blummer, 1961 137.3 260 Grimmer and Bohnke, 1975 160-190 Santodonato et al., 1973 137.3 NAd 9.7 4-60' Petrun and Rubenchik, 1966 2.0-6.0 Pancirov and Brown, 1977 <1.0-<11 Brown and Pancirov, 1979 1.1-7.9 Borneff and Kunte, 1964 11.5° Fazio and Howard, 1983 1.7-53° Fazio and Howard, 1983 1.3-6.9' __ Malanoski et a1., 1968 60 Mallett ct al., 1963 35.2 12.6-48 1ARC, 1973 7.0-24 U.S. EPA, 1985 23 Fritz, 1971 16.1 Archer et al., 1979 12.3-20.0 Kolar et al., 1975 7.4-24.5 Grimmer, 1968 0.2 0.1 Archer et al., 1979 0.1-0.5 1ARC, 1973 0.3 Wang and Meresz, 1981 0.3 Fazio and Howard, 1983 0.4 Shabad and Cohen, 1972 0.2 0.2-0.3 Kolar et al., 1975 0.2 Fritz, 1986 0.2 Lintas ct al., 1979 0.3 Grimmer and Pott, 1983 0.2-4.1 1ARC, 1973 35.9 NA

146 Hattemer-Frey and Travis TABLE 2 Comparison of predicted and measured environmental concentrations for BaP (continued) Predicted Measured concentration concentration Phase (µg/kg)• (pg/kg) Reference Dairy products Beef Eggs 'Nanograms per gram "Nanograms per cubic meter 'Nanograms per liter 'Measurement not available °Smoked fish 'Smoked cheese xSmoked meat 'Charbroiled meat 3.0 0.5' Panalaks, 1976 4.1-6.2' Fazio and Howard, 1983 2.3 0.2-2.OF Panalaks, 1976 3.0" Fazio and Howard, 1983 4.4 Ljinsky and Ross, 1967 18.8-24.1F Doremire et al., 1979 20 Fazio and Howard, 1983 0.2-0.6 IARC, 1973 <1.0-10.5 U.S. EPA, 1980 0.002 NA Accumulation on Vegetation Accurrtulation of organics in vegetation is a complex process that can potentially involve three pathways: deposition, root uptake, and air-to-leaf transfer. Root uptake of organics has been correlated with Kov, (Baes, 1982; Briggs et al., 1982) and can be estimated from B_ the soil-to-plant BCF. In the absence of measured data, B, can be estimated from the geometric mean regression equation devel- oped by Travis and Arms (1988). Because lipophilic compounds are most soluble in organic matter (and thus not very soluble in water), they tend to sorb strongly to soil, and root uptake is not expected to be a major pathway of vegetative contamination (Briggs et al., 1982). Since BaP is not very soluble in water, root uptake was not expected to be a major source of vegetative contamination for BaP. The concentration of BaP in vegetation due to root uptake (CVR) can be estimated by multiplying the con- centration of BaP in soil (CS) by a chemical-specific soil-to-root BCF (B,). Ed- wards (1983) reported that the concentration of PAHs in plants is generally lower than their corresponding concentration in soil. Soil-to-root BCFs for BaP range from 0.002 to 0.33 with a mean value of 0.056 (Edwards, 1983). This value is within a factor of four of the value obtained using the geometric mean regression equation developed by Travis and Arms (1988) (0.014). Using the measured x coi pla De SIat ata cen be t wh f Inte all ( day Usi r/Y veg (Tr et c (F, cro gro we~ tha: diff tha• ~ eqL pot ! & por pee Thi p t, hYF loc;

Toxicology and Industrial Health, Vol. 7, No. 3, 1991 147 itially Root 1982) sured ievel- >luble ongly tative be a aP in con- Ed- lower range lue is :ssion sured concentration of BaP in soil (3.0 pg/kg) and a B, value of 0.056, CVR for all plant groups is estimated to be 0.17 µg/kg. Deposition of BaP onto outer plant surfaces was expected to contribute sub- stantially to vegetative contamination, since most of the BaP released into the atmosphere sorbs to particulates (Edwards, 1983; U.S. EPA, 1984). The con- centration of pollutant on vegetation due to atmospheric deposition (CVD) can be estimated using the following equation derived from Travis et al. (1986): :. CVD (,ug/kg) = C. * FP * Vd * r/Y = T„Z (1) where: C, = the concentration of organic in air (,ug/m3); FP = fraction of chemical sorbed to particulates (unitless); Vd = atmospheric deposition velocity (m/sec); r/Y = vegetation-specific intercept fraction-to- productivity ratio (m2/mg) (Baes et al., 1984); T„Z = weathering half-life (in 2/T112 in sec) (Baes et al., 1984). , Hence, the model assumes that only the fraction of chemical sorbed to partic- ulates (FP) is available for deposition. Intercept fraction-to-productivity ratios (r/Y values) adjust for the fact that not all depositing particulates will settle onto edible plant parts. Vegetation-specific differences in the r/Y values for plants consumed by animals and humans require that different estimates of CVD be made for plants consumed by these two groups. Finally, plant concentrations are corrected for pollutant loss due to weathering. In the absence of measured data, it is generally assumed that organics that have accumulated on outer plant surfaces have a weathering half-life of 14 days (Baes`et al., 1984). Using the measured concentration of BaP in air (2.5 ng/m;; Faoro, 1975), an r/Y value of 0.4 m2/kg for vegetation consumed by animals and 0.32 m2/kg for vegetation consumed by humans (Baes et al., 1984), a Vd value of 0.002 m/sec (Travis and Hattemer-Frey, 1990), and a weathering half-life of 14 days (Baes et al., 1984), and assuming that 99.5% of BaP exists in the particulate phase (F, = 0.005) (Bidleman, 1988; Mackay et al., 1986), CVD for exposed food crops consumed by humans is estimated to be 2.81cg/kg, while CVD for forage equaled 3.5 /cg/kg. It was assumed that grains and protected produce, including potatoes and other root vegetables, legumes, and garden fruits, whose edible portions either grow underground or are protected by pods, shells, or nonedible peels, were not contaminated via the deposition or air-to-leaf transfer pathways. This assumption was made because there is no empirical data to support the hypothesis that chemical vapors absorbed by aerial plant parts are actually trans- located to other plant parts. If air-to-leaf transfer is a surface phenomenon only (i.e., no translocation occurs), then our assumption is valid. If, on the other hand, translocation does occur, then this assumption may not be valid. f

I While some organics are taken up by plants directly from contaminated soil and translocated to upper plant parts, other compounds volatilize from the soil, and their vapors are absorbed by aerial plant parts (Bacci and Gaggi, 1985, 1986, 1987; Bacci et al., 1990; Beall and Nash, 1971; Travis and Hattemer-Frey, 1988). Buckley (1982), Nash and Beall (1980), and Mosbaek et al. (1988) reported that PCB, DDT, and mercury residues present in plant foilage were mainly due to vapor transport from soil. The air-to-leaf BCF (B,,,) is defined as the equilibrium concentration of organic in upper plant parts (mg/kg) divided by the equilibrium concentration of organic in air as a vapor (mg/kg). Because an air-to-leaf BCF (B,,) has not been directly measured for BaP, it was estimated using'the following geometric mean equation developed by Travis and Hattemer-Frey (1990): Bv,=5.0 x 10-6*Ko,r~ H (2) where Ko,v is the octanol-water partition coefficient, and H is Henry's Law con- stant in atm-m3/mol. Using a log Ka,V of 5.98 (Mackay and Paterso'n, 1990) and an H value of 1.6 x 10-6 atm-m3/mol (U.S. EPA, 1986), a B,,, value of 3.0 x 10' was used in this analysis. The concentration in vegetation due to air-to-leaf transfer (CVA) can be esti- mated from the following equation (Travis and Hattemer-Frey, 1990): CVA (ug/kg) = C, * F~ * B,,, (3) where C, is the concentration of organic in air Ecg/kg, and F, represents the fraction of compound that exists in air as a vapor. Using the measured concen- tration of BaP in air (2.5 pglkg) and assuming that 0.5% of BaP exists in the vapor phase (F, = 0.005), CVA for exposed plants consumed by humans and animals is estimated to be 32 pg/kg. The total concentration of BaP on plants is determined by summing the contri- bution for each of the three pathways of vegetative contamination (CVR + CVD + CVA). Hence, the total concentration of BaP on exposed fruits and vegetables consumed by humans is estimated to be 35.0 µglkg, 91% of which is due to air-to-leaf transfer and 9% to deposition. The total concentration of BaP in forage is estimated to be 35.7 Eigl kg, with air-to-leaf transfer accounting for 90% and deposition for 10% of the total BaP concentration on forage. In both cases, root uptake was a negligible pathway of vegetative contamination. The = s total concentration of BaP in grains and roots crops is 0.2 Eig/kg. These results demonstrate that the concentration of BaP in exposed vegetation (both leaf}vegetables consumed by humans and forage crops consumed by animals) pri- marily results from air-to-leaf transfer, not direct deposition as speculated by Edwards (1983) ~ . ~ The FFC model predicts that air-to-leaf transfer accounts for about 90% of the total BaP concentration in exposed vegetation despite the fact that only 0.5% of sor dib dic exF con vall sm1~ witl via veg A& the roo and 198 of I Ho% BaF roo 'L exp dict me~ Acc Bec in li hal<. Altl posi shw dail The bv ; (iN-; fact hav regr mill dail of c inge catt

3i1 and il, and 1986; 1988). !d that due to •ontri- /R + :s and iich is f BaP -1s for both The esults leaf\ ) pri- ed by )f the 0.5% Toxicology and Industrial Health, Vol. 7, No. 3, 1991 149 of BaP is estimated to exist in the vapor phase (Bidleman, 1988). The result is somewhat surprising, since air-to-leaf transfer for TCDD (2,3,7,8-tetrachloro- dibenzo-p-dioxin), which is more lipophilic than BaP (log Kow = 6.85), is pre- dicted to account for only about 60% of total vegetative contamination for exposed crops (Travis and Hattemer-Frey, 1990). The reason for this apparent contradiction lies in the Henry's Law constant of the two compounds. The H value for BaP (1.6 x 10-6 atm-m3/mol) is about three orders of magnitude smaller than the H value for TCDD (3.6 x 10-3 atm-m3/mol). Since chemicals with a high Ko,,,, and a low H value are expected to accumulate to a large extent via this pathway, it is not surprising that more BaP than TCDD accumulates in vegetation via direct air-to-leaf transfer. Additionally, measured data support our finding concerning the importance of the air-to-leaf transfer pathway. The concentration of BaP in vegetation due to root uptake (0.2,ug/kg) is based on measured soil-to-root BCFs (Edwards, 1983) and the measured concentration of BaP in background soil (Santodonato et al., 1981). Furthermore, our CVR estimate is supported by measured concentrations of BaP in root crops and protected produce (Archer et al., 1979; Fazio and Howard, 1983; Wang and Meresz, 1981). The mean measured concentration of BaP in leafy vegetation is about 24 ug/kg. Since neither direct deposition nor root uptake appears to account for the relatively high concentration of BaP in exposed vegetation, air-to-leaf transfer is the only remaining pathway. Our pre- dicted concentration from air-to-leaf transfer is within a factor of two of actual measured concentrations (Table 2). Accumulation in Beef and Milk Because of its low water solubility and high lipophilicity, BaP readily accumulates in living organisms. Agricultural animals can be exposed to pollutants via in- halation and from the ingestion of contaminated forage, grains, soil, and water. Although ingestion of contaminated soil is often overlooked as a potential ex- posure pathway for terrestrial organisms, Travis and Hattemer-Frey (1990) showed that ingestion of contaminated soil accounts for 20% to 29% of the total daily intake of TCDD by beef and dairy cattle. The predicted concentration of pollutant in cow beef and milk is then calculated by multiplying the amount of pollutant taken in from all exposure pathways (inhalation and ingestion) by a chemical-specific, species-specific biotransfer factor (BTF). Again, since steady-state daily intake-to-cow beef and milk BTFs have not been measured for BaP, they were estimated using the geometric mean regression equation developed by Travis and Arms (1988). Using cow beef and milk BTFs of 2.3 x 10-'- d/kg and 7.6 x 10-3 d/kg, respectively, the predicted daily intake of BaP by beef and dairy cattle from inhalation and from ingestion of contaminated feed and water is presented in Table 3. These data show that ingestion of contaminated forage is the primary pathway of BaP exposure by cattle. Table 2 shows that the predicted concentration of BaP in cow beef (2.3

TABLE 3 Intake of BaP by beef and dairy cattle Intake by Percentage Intake by Percentage beef cattle of total dairy cattle of total (p.glday) daily intake (µglday) daily intake Inhalation 0.4 <1% 0.4 <1% Water 0.01 <1% 0.01 <1% Soil 0.2 <1% 0.5 <1% Forage 95.5 99% 395.0 ' 99% Grains 0.8 <1% 1.2 <1% TOTAL 96.9 100% 397.1 100% ,ug/kg) agrees well with the range of measured values (2.0 to 5.8Icg/kg), while the predicted concentration of BaP in whole milk and dairy products (3.0 ,ug/kg) is about six times higher than the single measured value (Table 2). Based on our experience modeling food chain exposures for other, lipophilic com- pounds, we expected the concentration of BaP in cow beef and milk to be similar, with the concentration in milk being slightly higher (Travis and Arms, 1988). Accumulation in Fish and Eggs The concentration of chemical in fish is generally estimated by multiplying the concentration of organic in water by a BCF for the chemical in fish. The U.S. EPA (1986) reported that the BCF for BaP in fish is approximately 11,100. Assuming that the concentration of BaP in water is 8.7 ng/L (Table 2), the corresponding predicted concentration of BaP in whole fish is estimated to be 92 µgl}cg. This prediction is higher than measured values for BaP in fish (Table 2). Hence, we suspect that the EPA BCF for BaP in fish is too high. Using the measured concentration of BaP in water and the mean measured concentration of BaP in whole fish (10,ug/kg), we estimate the BCF for BaP in whole fish to be 1100, or about one order of magnitude lower than the EPA-recommended value. The concentration of BaP in eggs can be estimated by multiplying the concen- tration of BaP in grains by a chemical-specific BCF. Using a BCF of 1.3 x 10-' for BaP (Belcher and Travis, 1990) and 0.11 ,ug/kg as the concentration of BaP in grains, the concentration in eggs is estimated to be 0.002 /cg/kg. QUANTIFYING THE EXTENT OF HUMAN EXPOSURE TO BaP An important objective in risk assessment is elucidating the pathways and extent of human exposure to pollutants chronically released into the environment. Daily intake of BaP from food and water was estimated by multiplying the concen- tration in beef, milk, vegetation, fish, eggs, and water times the average daily adult human consumption values reported by Yang and Nelson (1986) assuming Pr:_ that inhal back; sumi; sister Tabl< wate: beef, of cc The: U.S. of 2.: ` geom by L estim items by tt- predi data tities BaP t Assu; that t (U.S. to ba ~nce kf B~ Pu'

Toxicology and Industrial Health, Vol. 7, No. 3, 1991 151 ; the U.S. .100. . the o be .-able J the ation sh to nded icen- 10-' BaP aP Ktent Daily lcen- daily ming TABLE 4 Predicted average daily intake of BaP by the general population of the U.S. Daily intake Percent of total Source (µg/day) daily intake Food (total) 2.10 97% Produce 0.60 28% Dairy products 0.93 44% Beef 0.40 19% Fish 0.17 8% Eggs <0.01 <1% Inhalation 0.05 2% Water 0.01 1% ~: sistent with U.S. EPA assessment methodology (U.S. EPA, 1989). that 100% of BaP ingested is absorbed through the gut. The amount of BaP inhaled by humans was estimated by multiplying the concentration of BaP in background air (ng/m3) by the average adult inhalation rate of 20 m3/day as- suming that 100% of inhaled BaP is absorbed through the lung, which is con- Table 4 gives the predicted average daily intake of BaP by humans from air, water, and food. These data show that the food chain, especially dairy products, beef, and produce, accounts for 97% of human exposure to BaP. Consumption of contaminated eggs and water was not a major source of human exposure. The average, long-term daily intake of BaP by the general population of the U.S. is estimated to be 2.2,ug/day, which agrees well with the intake estimate of 2.2 pg/day reported by Suess (1976) but is about 70 times higher than the geometric mean value of 0.03 ug/day (range equals 0.01 to 1 ng/day) reported by Lioy et al. (1988). The relatively large difference between our exposure estimate and Lioy et al.'s stems from the fact that BaP concentrations in food items reported by Lioy et al. (1988) are generally lower than values predicted by the FFC model. The reason for this discrepancy is not understood. Our predicted concentration of BaP in food items is within the range of measured data (Table 2). One explanation might be that Lioy et al. sampled smaller quan- tities of the food items that we predicted would accumulate larger amounts of BaP (i.e., leafy vegetables and fish). Assuming that the general population takes in about 2.2 /ig of BaP per day and that the cancer potency factor for oral exposure to BaP is 11.5 mg/kg-day'' (U.S. EPA, 1986), the excess lifetime cancer risk associated with human exposure to background levels of BaP is 3.4 x 10-4, or a probability of 345 additional cancer cases per million persons. These results provide evidence that ingestion of BaP-contaminated food items may pose a serious health threat to the U.S. population.

Human Exposure to BaP from Smoking and Indoor Air Pollution Smoking represents another anthropogenic source of background human ex- posure to BaP. Butler and Crossley (1979) estimated that one cigarette delivers about 39 ng of BaP. Thus, average smokers (i.e., individuals who smoke 20 cigarettes a day) are taking in an additional 780 ng of BaP daily, which means that smokers get an additional 16% BaP from smoking. The total increased lifetime cancer risk from BaP associated with smoking 20 cigarettes per day is 4.7 x 10's. Heavy smokers (i.e., individuals who smoke 35 cigarettes a day) are likely to experience a total increased lifetime cancer risk of 5.7 x 10-1 from exposure to BaP. Similarly, Butler and Crossley (1979) reported that concentrations of BaP mea- sured indoors (2.2 nglm3) were comparable to outdoor air concentrations (2.5 ng/m;), demonstrating that indoor activities are not likely to substantially in- crease BaP intakes, since inhalation is not a major pathway of human exposure to BaP. CONCLUSIONS This paper presents a fugacity-based approach for evaluating the environmental partitioning of BaP. Although the modeling efforts used to obtain these results may not have been used extensively in the past, the consistent reproduction of measured environmental data provides evidence that this approach is valid. It should be emphasized, however, that environmental partitioning models are not exact replicas of the environment, since they contain many simplifying assump- tions. For example, the model assumes that all biochemical and chemical trans- forrriation reactions are first order. Furthermore, the models are difficult to validate for the following reasons: (1) reliable measured environmental data are often not available; (2) environmental concentrations vary spatially and tem- porally; and (3) data on the magnitude of emissions into air, water, and soil are rarely known precisely. Preliminary attempts to validate fugacity models, how- ever, show excellent agreement with measured environmental concentrations and estimates of human exposure (Hattemer-Frey et al., 1990; Hattemer-Frey and Travis, 1989; Mackay et al., 1985a; Travis and Hattemer-Frey, 1987). Fur- thermore, this approach allows researchers to quickly and accurately estimate the extent of human exposure relative to other accepted methods (e.g., mea- suring body burdens or conducting market basket studies). Multimedia transport models also represent useful tools for regulating human exposure to organic chemicals, since regulators can establish guidelines that take into account the significance of each exposure pathway. Organic chemicals tend to accumulate in the media in which they are most soluble. BaP, a very tipophilic compound that is virtually insoluble in water, partitions mainly into soil and sediment and accumulates to a large extent in the i - foc to : me exfn con to I Thr U.S hun not esti r valu risk we una, ANI w AR( 01 ta Pr BAC or E; BAC K BAC ca BAC lo 3: BAE ra BAE A n1 N BAS A PC BEA so

Toxicology and Industrial Health, Vol. 7, No. 3, 1991 153 food chain. Multimedia transport models, such as the FFC model, can be used to study the environmental fate of various pollutants released into the environ- ment, their accumulation in the food chain, the pathways and extent of human exposure, and the total environmental input. Results show that ingestion of contaminated food items remains the overwhelming source of human exposure to BaP. The long-term, average daily intake of BaP by the general population of the U.S. is estimated to be 2.2 Itg/day with the food chain accounting for 97% of human exposure to BaP. Inhalation and consumption of contaminated water are not major sources of human exposure. Our intake estimate agrees well with an estimate of 1.61tglday reported by Suess (1976) but is 70 times higher than the value of 0.03 kg/day reported by Lioy et al. (1988). Since the increased lifetime risk associated with human exposure to background levels of BaP is 3.4 x 10-4, we conclude that ingestion of food items contaminated with BaP may pose an unacceptable health risk to the U.S. population. REFERENCES ,ntal >ults n of J. It not .mp- ans- It to t are :em- ~ are low- ions Frey Fu r- nate nea- port :anic the nost 3ter, ) the I ANDELMAN, J.B., and M.J. SUESS, 1970. Polynuclear aromatic hydrocarbons in the water environment, Bull. WHO, 43:479-508. ARCHER, S.R., T.R. BLACKWOOD, and G.E. WILKINS, 1979. Status Assessment of Polynuclear Aromatic Hydrocarbons, EPA 600/2-79-210L, Industrial Environmen- tal Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Cincinnati, Ohio. BACCI, E., D. CALAMARI, C. GAGGI and M. VIGHI, 1990. Bioconcentration of organic chemical vapors in plant leaves: Experimental measurements and correlation, Environ. Sci. Technol., 24(6):885-889. BACCI, E., and C. GAGGI, 1987. Chlorinated hydrocarbon vapors and plant foliage: Kinetics and applications, Chemosphere, 16:2515-2522. BACCI, E., and C. GAGGI, 1986. Chlorinated pesticides and plant foilage: Translo- cation experiments, Bull. Environ. Contam. Toxicol., 37(6):850-857. BACCI, E., and C. GAGGI, 1985. Polychlorinated biphenyls in plant foliage: Trans- location or volatilization from contaminated soils? Bull. Environ. Contam. Toxicol., 35(5):673-681. BAES, C.F. III, 1982. Prediction of radionuclide Kd values from soil-plant concentration ratios. Trans. Amer. Nuc. Soc., 41:53-54. BAES, C.F. III, R.D. SHARP, A. SJOREEN and R. SHOR, 1984. A Review and Analysis of Parameters for Assessing Transport of Environmentally Released Radio- nuclides Through Agriculture, ORNL-5786. U.S. Department of Energy, Oak Ridge National Laboratory, Oak Ridge, TN. BASU, D.K., and J. SAXENA, 1978. Analysis of Raw and Drinking Water Samples for Polynuclear Aromatic Hydrocarbons, EPA CA/7-2999-A AND CA/8-2275-B, Ex- posure Evaluation Branch, Health Effects Research Laboratory, Cincinnati, Ohio. BEALL, M.L., Jr., and R.G. NASH, 1971. Organochlorine insecticide residues in soybean plant tops: Root versus vapor sorption, Agron. J., 63:460-4fi4.

BECK, H., K. ECKART, W. MATHAR, and R. WITTOWSKI, 1989. PCDD and PCDF body burden from food intake in the Federal Republic of Germany, Chemo- sphere,18(i-6):587-592. BELCHER, G.D., and C.C. TRAVIS, 1990. The food chain as a source of human exposure from municipal waste combustion, Proceedings of the International Confer- ence on Municipal Waste Combustion, Hollywood, Florida, April 11-14, U.S. Envi- ronmental Protection Agency, Research Triangle Park, NC. BIDLEMAN, T.F., 1988. Atmospheric processes. Environ. Sci. Technol., 22:361-367. BLUMMER, M., 1961. Benzpyrenes in soil, Science, L34:474-475. - BORNEFF, J., and H. KUNTE, 1983. Polycyclic aromatic hydrocarbons in river and iake water, biota, and sediments, Handbook of Polycyclic Aromatic Hydrocarbons, A. Bjorseth, Ed., Marcel Dekkar, Inc., New York. p, 629-652. BORNEFF, J., and H. KUNTE, 1964. Carciitogenic substances in water and soil, Arcg. Hyg. Bakteriol., 147:401-409. BRIGGS, G.G., R.H. BROMILOW and A.A. EVANS, 1982. Relationships between lipophilicity and root uptake and translocation of non-ionized chemicals by barley. Toxicol. Environ. Chem., 7:173-189. BROWN, R.A., and R.J. PANCIROV, 1979. Polynuclear aromatic hydrocarbons in Baltimore Canyon fish, Environ. Sci. Technol., 13(7):878-879. BUCKLEY, E.H., 1982. Accumulation of airborne polychlorinated biphenyls in foliage, Science, 216:520-522. BUTLER, J.D., and P. CROSSLEY, 1979. An appraisal of relative airborne sub-urban concentrations of polycyclic aromatic hydrocarbons monitored indoors and outdoors, Sci. Total Environ., 11:53-58. COHEN, Y., and P.A. RYAN, 1985. Multimedia modeling of environmental transport: Trichloroethylene test case, Environ. Sci. Technol., 19:412-417. DOREMIRE, M.E., G.E. HARMON, and D.E. PRATT, 1979. 3,4-Benzopyrene in chartoal grilled meats, J. Food Scf., 82:203-206. EDWARDS, N.T., 1983. Polycyclic aromatic hydrocarbons (PAH's) in the terrestrial environment-A review, J. Environ. Qual., 12(4):427-441. FAORO, R.B., 1975. Trends in concentrations of benzene soluble suspended particulate fraction and benzo-a-pyrene, J. Air Poll. Control Assoc., 25(6):638-640. FAZIO, T., and J.W. HOWARD, 1983. Polycyclic aromatic hydrocarbons in food, in Handbook of Polycyclic Aromatic Hydrocarbons. A. Bjorseth, Ed., Marcel Dekkar, Inc., New York, p. 461-505. FOX, M.A., and S.W. STANLEY, 1976. Determination of Polycyclic Aromatic Hydro- carbons in Atmospheric Particulate Matter by High Pressure Liquid Chromatography Coupled with Fluorescence Techniques, Anal. Chem., 48(7):992-998. FRITZ, W., 1986. On the formation of carcinogenic hydrocarbons during the thermal treatment of foods. Part III. The baking of bread and biscuits, Nahrung, 12(8):805. FRI1Z, W., 1971. Extent and sources of our contamination our food with carcinogenic hydrocarbons, Ernaehrungsforschung, 16(4):547-557. GRIMMER, G., 1983. Profile analysis of polycyclic aromatic hydrocarbons in air, Hand- book of Polycyclic Aromatic Hydrocarbons, A. B jorseth, Ed., Marcel Dekkar, Inc., New York, p. 149-181. GRIMMER, G. 1968. Carcinogenic hydrocarbons in the human environment, Deut. Apoth.-Ztg., 108(16):529-533. GR c GR i 1 HP r HA , HE IA KC L' LP L1 LL M M M M

Toxicology and Industrial Health, Vol, 7, No. 3, 1991 155 GRIMMER, G., and H. BOHNKE, 1975. Profile analysis of polycyclic aromatic hy- drocarbons and metals content in sediment layers of a lake, Cancer Lttr., 1:75-84. GRIMMER, G., and F. POTT, 1983. Occurrence of PAH. In Environmental Carcino- gens: Polycyclic Aromatic Hydrocarbons, G. Grimmer, Ed., CRC Press, Inc., Boca Raton, FL, p. 61-128. HATTEMER-FREY, H.A., C.C. TRAVIS, and M.L. LAND, 1990. Benzene: Envi- ronmental partitioning and human exposure, Environ. Res., 53:221-232. HATTEMER-FREY, H.A. and C.C. TRAVIS, 1989. Pentachlorophenol: Environmen- tal partitioning and human exposure. Arch. Environ. Conram. Toxicol., 18:482-489. HERBES, S.E., G.R. SOUTHWORTH, D.L. SHAEFFER, and W.H. CRIEST, 1979. Oak Ridge National Laboratory, Oak Ridge, TN. IARC, 1973. 1ARC Monographs on the Evaluation of Carcinogenic Risk of Chemicals to Man, Vol. 3, Certain Polycyclic Aromatic Hydrocarbons and Heterocyclic Com- pounds, Lyon, France, p. 91-93. KOLAR, L., LEDVINA, R. TICHA, J. and HANUS, F., 1975. Contamination of soil, agricultural crops, and vegetables by 3,4-benzpyrene in the vicinity of Cseke`Bude- jovice, Cesk. Hyg., 20(3):135-139. LINTAS, C., M.C. DE MATTHAEIS, and F. MERLI, 1979. Determination of benzo- a-pyrene in smoked, cooked and toasted food products, J. Cosmet. Toxicol., 17:325- 328. LIOY, P.L., R. HARKOV, J.M. WALDMAN, C. PIETARINEN and A. GREEN- BERG, 1988. The total human environmental exposure study (THEES) to benzo(a)- pyrene: Comparison of the inhalation and food pathways. Arch. Environ. Health, 43:304-312. LJINSKY, W., and A.E., ROSS, 1967. Production of polynuclear aromatic hydrocarbons in the cooking of food, Food Cosmet. Toxicol., 5:343-347. LU, P-Y., R.L. METCALF, N. PLUMMER, and D. MANDEL, 1977. The environ- mental fate of three carcinogens: Benzo-(a)-pyrene, benzidine, and vinyl chloride evaluated in laboratory model ecosystems. Arch. Environ. Contam. Toxicol., 6:129- 142. MACKAY, D. and S. PATERSON, 1990. Evaluation of the regional multimedia fate of organic chemicals: A level III Fugacity model, Environ. Sci. Technol., in press. MACKAY, D., S. PATERSON, and W.H. SCHROEDER, 1986. Model describing the rates of transfer processes of organic chemicals between atmosphere and water, En- viron. Sci. Technol.. 20(8):810-816. MACKAY, D., S. PATERSON, and B. CHEUNG, 1985a. Evaluating the environmental fate of chemicals: the fugacity level III approach as applied to 2,3,7,8-TCDD, Che- mosphere, 14:859-863. MACKAY, D., S. PATERSON, B. CHEUNG, and W.B. NEELY, 1985b. Evaluating the environmental behavior of chemicals with a level III fugacity model, Chemosphere, 14:335-374. ~ MALANOSKI, A.J., E.L. GREENFIELD, C.J. BARNES. J.M. WORTHINGTON, and F L. JOE, JR.. 1968. Survey of PAH in smoked foods, J. Assoc. Off. Anal. Chem., 51:114-121. MALLETT, L., A. PEDRIAU, and J. PEDRIAU, 1963. Pollution of benzopyrene type ~ polycyclic hydrocarbons of the western region of the Arctic Ocean, C.R. Acad. Sci. (Paris), 256:3487-3489.

MOSBAEK, H., J. TGELL, and T. SEVEL, 1988. Plant uptake of airborne mercury in background areas, Chemosphere, 17(6):1227. NASH. R.G., and M.L. BEALL, Jr., 1980. Distribution of silvex, 2,4-D, and TCDD applied to turf in chambers and field plots, J. Agric. Food Chem., 28(3):614-623. NATIONAL ACADEMY OF SCIENCES (NAS), 1972. Particulate Polycyclic Organic Matter, Washington, D.C. PANALAKS, T., 1976. Determination and identification of PAH in smoked and char- broiled food products by high pressure liquid chromatography and gas chromatog- raphy, J. Environ. Sci. Health Bull., 4:299-315. PANCIROV, R.J., and R.A. BROWN, 1977. Environ. Sci. Tchnol., 11:989-992. PETRUN, A.S., and B.L. RUBENCHIK, 1966. Possibility of carcinogenic 3,4-benzo- pyrene in electrostatically smoked fish, Vrach. Delo, 2:93-95. RYAN, P.A., and C. COHEN, 1986. Multimedia transport of particulate-bound organ- ics: Benzo(a)pyrene test case, Chemosphere, 15(l):21-47. SANTODONATO, J., P. HOWARD, and D. BASU, 1981. Health and ecological as- sessment of polynuclear aromatic hydrocarbons, J. Environ. Pathol. Toxicol., 5(1):98- 176. SANDOTONATO, J., D. BASU, and P.H. HOWARD, 1973. Multimedia human ex- posure and carcinogenic risk assessment for environmental PAH, in Polynuclear Ar- omatic Hydrocarbons: Chemistry and Biological Effects, A. Bjorseth and A.J. Dennis, Eds., Battelle Press, Columbus. OH. SHABAD, L.M., and Y.L. COHEN, 1972. Contents of benzo(a)pyrene in some crops, Arch. Geschwulstforsch., 40(3):237-243. SHABAD, L.M., Y.L. COHEN, A.P. ILNITSKY, A.Y. KHESINA, N.P. SHCHER- BAK, and G.A. SMIRNOV, 1971. The carcinogenic hydrocarbon benzo-a-pyrene in the soil, J. Natl. Cancer Inst., 47(6):1179-1191. SHCHERBAK, N.P., and Y.L. LOGAN. 1970. Benzo-a-pyrene content in the soils of regions differing in their degree of industrialization, Tr. Perm. Gos. Med. Inst., 99:85-. 89. STEVENS, J.B., and E.N. GERBEC, 1988. Dioxin in the agricultural food chain, Risk Anal., 8(3):329-335. SUESS, M.J., 1976. The environmental load and cycle of polycyclic aromatic hydro- carbons, Sci. Total Environ., 6:239-250. TRAVIS, C.C. and A.D. ARMS. 1988. Bioconcentration of organics in beef, milk, and vegetation, Environ. Sci. Technol., 22:271-274. TRAVIS, C.C., and A.D. ARMS. 1987. The food chain as a source of toxics exposure, In: Toxic Chemicals, Health, and the Environment, L.B. Lave and A.C. Upton, Eds., Plenum Press, pp. 95-113. TRAVIS, C.C., and H.A. HATTEMER-FREY. 1990. Human exposure to dioxin, Sci. Total Environ., in press. -- -- - TRAVIS, C.C. and H.A. HATTEMER-FREY. 1987. Human exposure to 2,3,7,8-' TCDD. Chemosphere, 16:2331-2342. TRAVIS, C.C., G.A. HOLTON, E.L. ETNIER. C. COOK, F.R. O'DONNELL, D.M. HETRICK, and E. DIXON, 1986. Assessment of inhalation and ingestion population exposures from incinerated hazardous wastes. Environ. Inter., 12:553-540. U.S. ENVIRONMENTAL PROTECTION AGENCY (U.S. EPA), 1986. Superfund

Toxicology and Industrial Health, Vol. 7, No. 3, 1991 157 ury Is Public Health Evaluation Manual, EPA-540/1-86-060, Office of Emergency and Re- medial Response, Washington, D.C. 9D (it U.S. ENVIRONMENTAL PROTECTION AGENCY (U.S. EPA), 1985. Environ- mental Profiles and Hazard Indices for Constituents of Municipal Sludge: nic 12 Benzo(a)pyrene, Office of Water Regulations and Standards, Washington, D.C. U.S. ENVIRONMENTAL PROTECTION AGENCY (U.S. EPA), 1984. Health Effects ar- /~ Assessment for Polycyclic Aromatic Hydrocarbons (PAHs), EPA 540/1-86-013, En- >g- f vironmental Criteria and Assessment Office, Cincinnati, OH and Office of Solid Waste and Emergency Response, Washington, D.C. U.S. ENVIRONMENTAL PROTECTION AGENCY (U.S. EPA), 1980. Ambient Water Quality Criteria for Polynuclear Aromatic Hydrocarbons, EPA 440/5-80-069, :o- (N Washington, D.C. n- Iff WANG, D.T., and O. MERESZ, 1981. Occurrence and potential uptake of polynuclear aromatic hydrocarbons of highway traffic origin by proximally grown food crops, is- fit presented at the Sixth Inter. Symp. of PAH, Battelle Columbus Laboratory, Colum- 3- ts bus, OH. YANG, Y-Y., and C.R. NELSON, 1986. An estimation of daily food usage factors for x- )F assessing radionuclide intake in the U.S. population, Health Phys., 50(2):245 257. . r- is, )s, 2- in )f h d