- Email: [email protected]
706 Assessment and modifying factors in risk assessment: The case of triallate, a neurotoxic pesticide
Poster Session P37. Risk assessment toxic agents. A considerable number of the occupational diseases is due to carcinogenic agents present in work environment. Therefore, it has been attempted to assess the share or contribution of the occupational factors in the process of carcinogenesis. In addition to the necessity of assessing exposure type and determining high-risk population, occupational disease prophylaxis requires a system of occupational disease prevention based on precise legal regulations. Questions of prophylactic actions are discussed in the Ordinance of the Polish Minister of Health and Social Welfare on the protection of worker health from the risks attrributable to workplace carcinogens. One provision of the Ordinance requires that the workers exposed to carcinogenic agents should receive relevant information on health risks associated with exposure to those agents. It is the employer’s duty to provide the information, and he is also obliged to determine whether or not carcinogenic agents are present in the work environment. When carcinogens are found to be present there, the employer is obliged to perform a qualitative assessment of the health risk resulting from exposure to those carcinogens. To make the task easier to the employers and relevant labour safety services, a team of experts has been order to prepare guidelines useful in assessing health risk associated with exposure to individual carcinogenic agents. Up to now, guidelines for 71 carcinogenic agents have been prepared. The reports are designed for employers, employees, and work safety monitoring services. For Diesel engine emissions the value of 8.5 × 10−3 represents the risk of cancer development during 40 year period employment under the exposure. The indicates that 8 out at 10.000 people exposed during 3 years to employment to Diesel engine emissions as particulate at 4 mg/m3 will develop lung cancer. 704
RISK ASSESSMENT OF NEW CARBOFURAN PESTICIDES FOR OPERATOR AND ENVIRONMENT
Y. Chaika. L.I.Medved’s Institute of Ecohygiene and Toxicology, Kyiv, Ukraine Toxicological evaluation of several new carbofuran pesticides was done. Carbofuran pesticides are classified to 1 hazard class (extremely hazardous) by acute peroral and inhalation toxicity according to modern pesticides’ classifications by hazard. Carbofuran is stable in soil, water and vegetating plants. Work conditions formed during use of carbofuran pesticides were examined at seed dressing factory and during seeding on field. Complex of factors forming during work was studied, including except chemical factor - dust pollution, noise and vibration controls. Subjectively workers had a number of complains. Seeds after dressing by formulation didn’t change their color. Recommendations to improve carbofuran formulation were given, application rate was decreased. pesticide’s usage was examined again after improvement. Subjectively workers had no complains. Seeds after dressing by formulation were with characteristic red color. Operator risk assessment of improved formulation was done using different existing models, including so-called German model, EURO POEM and model based on our own methodical approaches. The values of carbofuran risk were lower than acceptable level – 1, if pesticide’s application is limited by 4 hours per day and is done with appropriate operator protective bodywear (mask, gloves, suit, boots, hat). Pesticide behavior in soil was examined in laboratory, modeling natural conditions with different application rates. Investigation show that carbofuran may be classified as highly migrative from soil to water formulation. Researches of quantities of active ingredient in water, soil and air environment around seed dressing factory were provided. The results of chemical analysis allows to make conclusion, that risk of carbofuran for environment is minimal.
705
s189
IS IT POSSIBLE TO ESTABLISH A CAUSAL RELATIONSHIP BETWEEN CFS (CHRONIC FATIGUE SYNDROME) AND CHEMICAL EXPOSURES? PARACELSUS PARADIGM IMPLICATIONS.
A. Ferrer-Dufol 1 , S. Nogué-Xarau 2 , E. Vilanova Gisbert 3 . 1 Toxicology Unit, University Clinic Hospital, Zaragoza, Spain, 2 Toxicology Unit, Clinic Hospital, Barcelona, Spain, 3 Toxicology Department, Elche University, Spain Paracelsus paradigm stated that there are not “non-toxic substances” and that any substance may become toxic in a dose-dependent way. His “third defense” has been considered the beginning of the scientific approach to toxicology and is the base of the strategies of prevention that assume for each substance a non-toxic level of exposure. We will discuss if some new admitted clinical pictures such as CFS implies a modification of this paradigm. CFS has been accepted as a clinical entity by CDC and defined as severe fatigue lasting more than 6 months and at least 4 of 8 signs or symptoms related to neurological, muscle-skeletal and immunologic impairment. Related for some experts to the old neurasthenic pathology, its origin has been attributed to a broad spectrum of causes: infections, trauma, stress and chemical exposure. Its relationship to chemicals has been founded on various grounds: clinical reports of CFS after acute or chronic exposure to solvents and pesticides and comorbidity with other syndromes associated to chemicals (Multiple Chemical Sensitivity, Gulf syndrome, Chemical intolerance, Sick building syndrome). On the other way there are sound arguments against: absence of any demonstrated dose-response effect, absence of relationship with any biomarker of exposure or effect, scarce documentation of exposure in most cases. A pathogenetic hypothesis has not been established. Some authors have proposed an altered sensitivity of GABAa receptor, cholinergic ways impairment and a toxicant-induced loss of tolerance to chemicals and other substances as some kind of food. We can be able to maintain the usefulness of the “third defense” admitting three different ways or response to chemical exposure: 1) toxic, implying a specific, with defined targets, and doses-dependent response, 2) allergic, less specific and mediated by the immune system; and 3) unspecific effects produced by chemicals acting as stressors in a individual-dependent way. 706
ASSESSMENT AND MODIFYING FACTORS IN RISK ASSESSMENT: THE CASE OF TRIALLATE, A NEUROTOXIC PESTICIDE
F. Broeckaert 1 , A. Li 2 , D. Goldstein 2 , J. Acquavella 2 , M. Martens 1 . 1 Monsanto Europe, Brussels, Belgium, 2 Monsanto, St-Louis, Missouri, USA Triallate is a pre-emergent herbicide used on barley, lentils, peas, triticale, wheat, and canary grass (seed only) in the fall or in the spring before targeted weed species germinate. Triallate is a thiocarbamate producing a peripheral neuropathy (degeneration of nerve fibers) in experimental animals. For the risk assessment of pesticide applicators, the systemic Acceptable Operator Exposure Level (systemic AOEL) obtained from toxicology studies is compared to the absorbed dose taking in to account inhalation and dermal routes of exposure. For triallate, the systemic AOEL has been derived from a 90-day neurotoxicity study in the rat which is appropriate for an application window of a couple of weeks per year. The No Observed Effect Level (NOEL) of this study (6.4 mg/kg bw/day) has been divided by an assessment factor (AF) of 100 to turn the experimental safe dose into a human equivalent dose. This AOEL has been converted into a systemic AOEL using the fraction of the dose absorbed from the gastro-intestinal tract of the rat. Despite the fact that very sensitive techniques were used to derive the NOEL in rats, the regulatory authorities in Europe imposed a additional modifying factor (MF) of 5 on top of the AF of 100 to derive the systemic AOEL. This was imposed in 1999 unless evidence was delivered that triallate is not neurotoxic in triallate manufacturing/packaging workers. To this end, a cross-sectional neurophysiological study was conducted to evaluate the relationship
s190
Poster Session P37. Risk assessment
between occupational exposure to triallate and several indicators of neurological function. The results of these studies were consistent with the absence of an association between triallate exposure and neurological abnormalities. The overall analysis of the complete data set on the neurotoxicity of triallate clearly demonstrates that an assessment factor of 100 is sufficient to assess the risk to health of professional applicators who constitute one of the most exposed individuals to triallate. 707
A COMPARATIVE HEALTH RISK ASSESSMENT FOR THE GENERAL POPULATION EXPOSED TO ALACHLOR AND BENZENE IN EUROPE
F. Broeckaert, M.-A. Reding1and, M.A. Martens. Monsanto Europe S.A., Brussels, Belgium Recently, alachlor (2-chloro-2’,6’-diethyl-N-methoxy-methylacetanilide), a herbicide used for weed control on corn, sunflowers and soybeans has been detected in the atmosphere of several rural and urban areas in Europe during and several weeks after its normal application time period (1–3). Levels of alachlor ranged between no detect (1/2 LOD = 0.08 ng m−3 ) and a maximum of 17.83 ng m−3 (3). The objective of this study was to determine the risks associated with short-term exposure to atmospheric alachlor based on a 28-day inhalation toxicity study in rats. Risk was expressed in term of margins of exposure (MOE), the ratio between the highest systemic level of exposure without any adverse effect in the experimental species and the absorbed dose in man derived from atmospheric monitoring data. Exposure was calculated using the maximum level of alachlor (17.83 ng/m3 ) (3), 100% pulmonary absorption and physiological breathing-activity patterns for European populations. Health risks of alachlor were compared with the risks associated with exposure to ambient levels of benzene using the most appropriate short-term toxicological study (4) and 95th percentiles of average benzene levels reported by the European Environmental Protection Agency. For the maximum level of alachlor in air, the short-term MOE by inhalation ranges between 186,951 (outdoor workers) and 434,239 (adult females). These figures are compared against the MOEs for benzene exposure in several EU countries. The risks associated with short-term respiratory exposures to alachlor are extremely low, including when food and drinking water sources are considered. 708
EPIDEMIOLOGICAL VALIDATION OF ENVIRONMENTAL CANCER RISK ASSESSMENTS: A CASE STUDY IN POPULATIONS EXPOSED TO POLYCYCLIC AROMATIC HYDROCARBONS.
M. Camus 1 , A. Vyskocil 2 , C. Viau 2 . 1 Environmental Health Sciences Bureau, Health Canada, Montreal, Qc, Canada; 2 Department of Environmental and Occupational Health, University of Montreal, Montreal, Qc, Canada Can epidemiology validate local environmental cancer risk assessments (RA)? We assessed this approach in populations exposed to polycyclic aromatic hydrocarbon mixtures (PAH) in 5 aluminium smelting and 5 other communities in Quebec. Lifetime lung cancer risks predicted for these communities on the basis of PAH and benzo(a)pyrene (BaP) levels measured in the 1990s were converted to annual incidence rates among women. We estimated the time required for these predicted rates to become statistically significant in these population, using a Poisson sample size calculation. This calculation was redone assuming 10 times higher historical exposures. Using population-weighted linear regression, we estimated the exposureresponse gradient between female lung cancer rates (1989–1993) and “dispersion-adjusted” BaP exposure estimates. Risk predictions based on recent low exposure levels could not be detected statistically before 300 years. However, had pre-1970 exposures been merely 10 times higher than today, the predicted risk might be detected statistically with about 3 years of data. The linear regression model estimated an exposure-response gradient (R2 ≈ 0.8) across the 5
aluminium-smelting communities. The rates in the 5 other areas where too heterogeneous for the trend to be significant. The gradient across aluminium-smelting communities would match the BaP-based RA if pre-1970 exposures had been some 56 (95% CI 12–101) times higher than today. Although RAs seem the only way to assess health impacts of recent low levels of carcinogens in small populations, epidemiological studies may have the power to assess local impacts of past environmental exposures at least one order of magnitude higher than today. 709
EVALUATION OF FOOD RISK EXPOSURE USING EXTREME VALUE THEORY-APPLICATION TO HEAVY METALS FOR SEA PRODUCTS CONSUMERS
J. Tressou 1 , P. Bertail 2 , A. Crepet 1 , M. Feinberg 3 , J.-Ch. Leblanc 4 . 1 Lab. de recherche sur la consommation,INRA, Ivry/seine, France; 2 Lab. de statistique, CREST/ENSAE, 92245 Malakoff, France; 3 Lab. de chimie analytique, INAP-G/INRA, Paris, France, 4 Direction scientifique NHSA, INRA, Paris, France This paper presents new statistical methods for evaluating food risk exposure related to some contaminants. We focused on the estimation of the probability to be exposed over the so-called provisional tolerable weekly intakes (PTWIs), when both consumption data and contamination data are independently available. For some toxic contaminants, PTWIs belongs to the exposure tail distribution, which suggests the use of Extreme Value theory to evaluate the risk. Our approach consists in modelling the exposure tail by a Pareto type distribution characterized by a Pareto index which may be seen as a measure of risk. Using propositions by Hall and Feuverger, we correct the bias of the Hill estimator to precisely estimate the risk index. We compare the results with empirical plugin methods. To illustrate our approach, we present some evaluations of risk exposure to heavy metals via sea product consumption. We also focus on the assumptions about exposure calculation such as aggregation of data or use of probabilistic calculus mode. Because of the strong impact of these assumptions, conclusions about exposure to heavy metals for consumers of sea products can not be easily summarised. As far as food risk is concerned, according to the data used in risk characterisation process, methylmercury intake via the consumption of sea products seems high for a significant part of the French population and especially for children aged 3 to 8 years old. Concerning the feasibility of our method based on tail estimation, the use of Pareto tail adjustment is nonsense if PTWI does not belong to the distribution tail, but at the opposite case it allows to quantify very low risk. Use of empirical plug-in methods is clearly relevant for very risky food contaminants. Developments are needed concerning confidence interval for such probabilities to exceed a given toxicological level. 710
RISK ASSESSMENT FOR CHROMIUM EXPOSURE STUDY OF A POPULATION LOCATED IN THE PROVINCE OF SANTA FE ARGENTINE REPUBLIC
D. Gotelli 1 , M. Gotelli 1 , L. Signorini 1 , A. Lo Balbo 1 , R. Castro 2 , S. Britos 3 , C. Gotelli 1 . 1 Center for Toxicological Research, Buenos Aires; 2 Univ. of Flores, Buenos Aires; 3 Univ. of Buenos Aires Aiming to determine if the industrial activity confined within the limits of the city of Esperanza, with a population of about 30.000 inhabitants, could generate a chromium contamination capable of producing alterations to the human health and the environment, a Risk Assessment Study was implemented applying the methodologies proposed by the World Health Organization and the Agency for Toxic Substances and Disease Registry of the U.S.A. Applying the bioaccessibility criterion, chromium was quantified in the air, soil, water, food and drinks; microfauna was studied; ecotoxic tests with bioindicators were performed; the vegetal covering was studied; the evaluation of cancer and teratogenesis incidence as well as other pathologies registered in the population was analysed. Using the whole information generated, the evaluation equations and criteria specified in this methodology were applied establishing that the presence of chromium in the area does not represent a risk either for human health or the ecosystem.
RISK ASSESSMENT OF NEW CARBOFURAN PESTICIDES FOR OPERATOR AND ENVIRONMENT
Y. Chaika. L.I.Medved’s Institute of Ecohygiene and Toxicology, Kyiv, Ukraine Toxicological evaluation of several new carbofuran pesticides was done. Carbofuran pesticides are classified to 1 hazard class (extremely hazardous) by acute peroral and inhalation toxicity according to modern pesticides’ classifications by hazard. Carbofuran is stable in soil, water and vegetating plants. Work conditions formed during use of carbofuran pesticides were examined at seed dressing factory and during seeding on field. Complex of factors forming during work was studied, including except chemical factor - dust pollution, noise and vibration controls. Subjectively workers had a number of complains. Seeds after dressing by formulation didn’t change their color. Recommendations to improve carbofuran formulation were given, application rate was decreased. pesticide’s usage was examined again after improvement. Subjectively workers had no complains. Seeds after dressing by formulation were with characteristic red color. Operator risk assessment of improved formulation was done using different existing models, including so-called German model, EURO POEM and model based on our own methodical approaches. The values of carbofuran risk were lower than acceptable level – 1, if pesticide’s application is limited by 4 hours per day and is done with appropriate operator protective bodywear (mask, gloves, suit, boots, hat). Pesticide behavior in soil was examined in laboratory, modeling natural conditions with different application rates. Investigation show that carbofuran may be classified as highly migrative from soil to water formulation. Researches of quantities of active ingredient in water, soil and air environment around seed dressing factory were provided. The results of chemical analysis allows to make conclusion, that risk of carbofuran for environment is minimal.
705
s189
IS IT POSSIBLE TO ESTABLISH A CAUSAL RELATIONSHIP BETWEEN CFS (CHRONIC FATIGUE SYNDROME) AND CHEMICAL EXPOSURES? PARACELSUS PARADIGM IMPLICATIONS.
A. Ferrer-Dufol 1 , S. Nogué-Xarau 2 , E. Vilanova Gisbert 3 . 1 Toxicology Unit, University Clinic Hospital, Zaragoza, Spain, 2 Toxicology Unit, Clinic Hospital, Barcelona, Spain, 3 Toxicology Department, Elche University, Spain Paracelsus paradigm stated that there are not “non-toxic substances” and that any substance may become toxic in a dose-dependent way. His “third defense” has been considered the beginning of the scientific approach to toxicology and is the base of the strategies of prevention that assume for each substance a non-toxic level of exposure. We will discuss if some new admitted clinical pictures such as CFS implies a modification of this paradigm. CFS has been accepted as a clinical entity by CDC and defined as severe fatigue lasting more than 6 months and at least 4 of 8 signs or symptoms related to neurological, muscle-skeletal and immunologic impairment. Related for some experts to the old neurasthenic pathology, its origin has been attributed to a broad spectrum of causes: infections, trauma, stress and chemical exposure. Its relationship to chemicals has been founded on various grounds: clinical reports of CFS after acute or chronic exposure to solvents and pesticides and comorbidity with other syndromes associated to chemicals (Multiple Chemical Sensitivity, Gulf syndrome, Chemical intolerance, Sick building syndrome). On the other way there are sound arguments against: absence of any demonstrated dose-response effect, absence of relationship with any biomarker of exposure or effect, scarce documentation of exposure in most cases. A pathogenetic hypothesis has not been established. Some authors have proposed an altered sensitivity of GABAa receptor, cholinergic ways impairment and a toxicant-induced loss of tolerance to chemicals and other substances as some kind of food. We can be able to maintain the usefulness of the “third defense” admitting three different ways or response to chemical exposure: 1) toxic, implying a specific, with defined targets, and doses-dependent response, 2) allergic, less specific and mediated by the immune system; and 3) unspecific effects produced by chemicals acting as stressors in a individual-dependent way. 706
ASSESSMENT AND MODIFYING FACTORS IN RISK ASSESSMENT: THE CASE OF TRIALLATE, A NEUROTOXIC PESTICIDE
F. Broeckaert 1 , A. Li 2 , D. Goldstein 2 , J. Acquavella 2 , M. Martens 1 . 1 Monsanto Europe, Brussels, Belgium, 2 Monsanto, St-Louis, Missouri, USA Triallate is a pre-emergent herbicide used on barley, lentils, peas, triticale, wheat, and canary grass (seed only) in the fall or in the spring before targeted weed species germinate. Triallate is a thiocarbamate producing a peripheral neuropathy (degeneration of nerve fibers) in experimental animals. For the risk assessment of pesticide applicators, the systemic Acceptable Operator Exposure Level (systemic AOEL) obtained from toxicology studies is compared to the absorbed dose taking in to account inhalation and dermal routes of exposure. For triallate, the systemic AOEL has been derived from a 90-day neurotoxicity study in the rat which is appropriate for an application window of a couple of weeks per year. The No Observed Effect Level (NOEL) of this study (6.4 mg/kg bw/day) has been divided by an assessment factor (AF) of 100 to turn the experimental safe dose into a human equivalent dose. This AOEL has been converted into a systemic AOEL using the fraction of the dose absorbed from the gastro-intestinal tract of the rat. Despite the fact that very sensitive techniques were used to derive the NOEL in rats, the regulatory authorities in Europe imposed a additional modifying factor (MF) of 5 on top of the AF of 100 to derive the systemic AOEL. This was imposed in 1999 unless evidence was delivered that triallate is not neurotoxic in triallate manufacturing/packaging workers. To this end, a cross-sectional neurophysiological study was conducted to evaluate the relationship
s190
Poster Session P37. Risk assessment
between occupational exposure to triallate and several indicators of neurological function. The results of these studies were consistent with the absence of an association between triallate exposure and neurological abnormalities. The overall analysis of the complete data set on the neurotoxicity of triallate clearly demonstrates that an assessment factor of 100 is sufficient to assess the risk to health of professional applicators who constitute one of the most exposed individuals to triallate. 707
A COMPARATIVE HEALTH RISK ASSESSMENT FOR THE GENERAL POPULATION EXPOSED TO ALACHLOR AND BENZENE IN EUROPE
F. Broeckaert, M.-A. Reding1and, M.A. Martens. Monsanto Europe S.A., Brussels, Belgium Recently, alachlor (2-chloro-2’,6’-diethyl-N-methoxy-methylacetanilide), a herbicide used for weed control on corn, sunflowers and soybeans has been detected in the atmosphere of several rural and urban areas in Europe during and several weeks after its normal application time period (1–3). Levels of alachlor ranged between no detect (1/2 LOD = 0.08 ng m−3 ) and a maximum of 17.83 ng m−3 (3). The objective of this study was to determine the risks associated with short-term exposure to atmospheric alachlor based on a 28-day inhalation toxicity study in rats. Risk was expressed in term of margins of exposure (MOE), the ratio between the highest systemic level of exposure without any adverse effect in the experimental species and the absorbed dose in man derived from atmospheric monitoring data. Exposure was calculated using the maximum level of alachlor (17.83 ng/m3 ) (3), 100% pulmonary absorption and physiological breathing-activity patterns for European populations. Health risks of alachlor were compared with the risks associated with exposure to ambient levels of benzene using the most appropriate short-term toxicological study (4) and 95th percentiles of average benzene levels reported by the European Environmental Protection Agency. For the maximum level of alachlor in air, the short-term MOE by inhalation ranges between 186,951 (outdoor workers) and 434,239 (adult females). These figures are compared against the MOEs for benzene exposure in several EU countries. The risks associated with short-term respiratory exposures to alachlor are extremely low, including when food and drinking water sources are considered. 708
EPIDEMIOLOGICAL VALIDATION OF ENVIRONMENTAL CANCER RISK ASSESSMENTS: A CASE STUDY IN POPULATIONS EXPOSED TO POLYCYCLIC AROMATIC HYDROCARBONS.
M. Camus 1 , A. Vyskocil 2 , C. Viau 2 . 1 Environmental Health Sciences Bureau, Health Canada, Montreal, Qc, Canada; 2 Department of Environmental and Occupational Health, University of Montreal, Montreal, Qc, Canada Can epidemiology validate local environmental cancer risk assessments (RA)? We assessed this approach in populations exposed to polycyclic aromatic hydrocarbon mixtures (PAH) in 5 aluminium smelting and 5 other communities in Quebec. Lifetime lung cancer risks predicted for these communities on the basis of PAH and benzo(a)pyrene (BaP) levels measured in the 1990s were converted to annual incidence rates among women. We estimated the time required for these predicted rates to become statistically significant in these population, using a Poisson sample size calculation. This calculation was redone assuming 10 times higher historical exposures. Using population-weighted linear regression, we estimated the exposureresponse gradient between female lung cancer rates (1989–1993) and “dispersion-adjusted” BaP exposure estimates. Risk predictions based on recent low exposure levels could not be detected statistically before 300 years. However, had pre-1970 exposures been merely 10 times higher than today, the predicted risk might be detected statistically with about 3 years of data. The linear regression model estimated an exposure-response gradient (R2 ≈ 0.8) across the 5
aluminium-smelting communities. The rates in the 5 other areas where too heterogeneous for the trend to be significant. The gradient across aluminium-smelting communities would match the BaP-based RA if pre-1970 exposures had been some 56 (95% CI 12–101) times higher than today. Although RAs seem the only way to assess health impacts of recent low levels of carcinogens in small populations, epidemiological studies may have the power to assess local impacts of past environmental exposures at least one order of magnitude higher than today. 709
EVALUATION OF FOOD RISK EXPOSURE USING EXTREME VALUE THEORY-APPLICATION TO HEAVY METALS FOR SEA PRODUCTS CONSUMERS
J. Tressou 1 , P. Bertail 2 , A. Crepet 1 , M. Feinberg 3 , J.-Ch. Leblanc 4 . 1 Lab. de recherche sur la consommation,INRA, Ivry/seine, France; 2 Lab. de statistique, CREST/ENSAE, 92245 Malakoff, France; 3 Lab. de chimie analytique, INAP-G/INRA, Paris, France, 4 Direction scientifique NHSA, INRA, Paris, France This paper presents new statistical methods for evaluating food risk exposure related to some contaminants. We focused on the estimation of the probability to be exposed over the so-called provisional tolerable weekly intakes (PTWIs), when both consumption data and contamination data are independently available. For some toxic contaminants, PTWIs belongs to the exposure tail distribution, which suggests the use of Extreme Value theory to evaluate the risk. Our approach consists in modelling the exposure tail by a Pareto type distribution characterized by a Pareto index which may be seen as a measure of risk. Using propositions by Hall and Feuverger, we correct the bias of the Hill estimator to precisely estimate the risk index. We compare the results with empirical plugin methods. To illustrate our approach, we present some evaluations of risk exposure to heavy metals via sea product consumption. We also focus on the assumptions about exposure calculation such as aggregation of data or use of probabilistic calculus mode. Because of the strong impact of these assumptions, conclusions about exposure to heavy metals for consumers of sea products can not be easily summarised. As far as food risk is concerned, according to the data used in risk characterisation process, methylmercury intake via the consumption of sea products seems high for a significant part of the French population and especially for children aged 3 to 8 years old. Concerning the feasibility of our method based on tail estimation, the use of Pareto tail adjustment is nonsense if PTWI does not belong to the distribution tail, but at the opposite case it allows to quantify very low risk. Use of empirical plug-in methods is clearly relevant for very risky food contaminants. Developments are needed concerning confidence interval for such probabilities to exceed a given toxicological level. 710
RISK ASSESSMENT FOR CHROMIUM EXPOSURE STUDY OF A POPULATION LOCATED IN THE PROVINCE OF SANTA FE ARGENTINE REPUBLIC
D. Gotelli 1 , M. Gotelli 1 , L. Signorini 1 , A. Lo Balbo 1 , R. Castro 2 , S. Britos 3 , C. Gotelli 1 . 1 Center for Toxicological Research, Buenos Aires; 2 Univ. of Flores, Buenos Aires; 3 Univ. of Buenos Aires Aiming to determine if the industrial activity confined within the limits of the city of Esperanza, with a population of about 30.000 inhabitants, could generate a chromium contamination capable of producing alterations to the human health and the environment, a Risk Assessment Study was implemented applying the methodologies proposed by the World Health Organization and the Agency for Toxic Substances and Disease Registry of the U.S.A. Applying the bioaccessibility criterion, chromium was quantified in the air, soil, water, food and drinks; microfauna was studied; ecotoxic tests with bioindicators were performed; the vegetal covering was studied; the evaluation of cancer and teratogenesis incidence as well as other pathologies registered in the population was analysed. Using the whole information generated, the evaluation equations and criteria specified in this methodology were applied establishing that the presence of chromium in the area does not represent a risk either for human health or the ecosystem.