Health risk assessment after decontamination of the beaches polluted by the wrecked ERIKA tanker.

Following the wreck of the oil tanker ERIKA off the north-west coast of France in December 1999, cleaning up of the beaches involved considerable work, which in any case could not be perfect. This raised the question of the short- and long-term health risks for the future bathers related to the toxicity of the remaining oil polycyclic aromatic hydrocarbons (PAHs). This risk assessment study was conducted to help health authorities plan risk management policies and inform the public. Thirty-six beaches were selected, representing a sample of the most frequently encountered topographic and beach usage situations; seven "control" beaches, unspoiled by ERIKA, were also investigated. Samples of water and sand were taken from each site, as well as from the surface of rocks. The 16 PAHs selected by the U.S. EPA were quantified in each environment. Several scenarios of exposure were contemplated: (1) a child between 2 and 4 years accidentally ingesting a small ball of fuel, (2) a child daily exposed throughout his holiday-time stay, (3) an adult (including a pregnant woman) spending his/her holidays on the coast, (4) an adult working on the beach, (5) and an adult practicing water sports. Among the available and significant toxicological values, the most conservative ones were selected for computing risks. The sand and water, after decontamination, were slightly polluted (respectively, 7.8 microg/kg and 23.3 ng/l of total 16 PAHs), with values similar to those found in the control beaches. By contrast, the rocky areas in some places were still highly polluted (up to 23 mg/kg on the surface layer). No lethal risk was found for a young child who had accidentally ingested a small ball of fuel. The life-long excess risks for skin cancer and for all other cancers were about 10-5 in scenarios including contact with the polluted rocks. In all other cases, excess risks were considerably lower. The hazard quotient for teratogenic effects was very small, except in scenarios where pregnant women would walk among rocks containing high pollution levels. In conclusion, exposure was mainly associated with polluted water among children, and with spoiled rocks for adults. Despite uncertainties, mainly dealing with the prediction of long-term risks following a short-term exposure, this study showed that beaches where pollution was no longer visible after decontamination did not entail any significant health risks and could be opened to the public.

Slow-stirring method for determining the n-octanol/water partition coefficient (Pow) for highly hydrophobic chemicals: performance evaluation in a ring test.

The n-octanol/water partition coefficient (Pow) is one of the most important parameters employed for estimating a chemical's environmental fate and toxicity. The currently adopted test guidelines for its determination do not allow for reliable determination of log Pow greater than 5. The slow-stirring experiment, in contrast, has been demonstrated to provide reliable log Pow data up to log Pow of 8.3. To validate this method and to obtain its approval as an official Organization for Economic Cooperation and Development (Paris, France) test guideline, a ring test was performed to evaluate the accuracy and precision of the slow-stirring experiment for determination of log Pow, particularly for highly hydrophobic compounds. Up to 15 volunteer laboratories tested 1,2,3,4-tetrachlorobenzene, hexachlorobenzene. 2,2',3,3',5,5',6,6'-octachlorobiphenyl, and decachlorobiphenyl (4.5 < log Pow < 8.2). The ring-test results for the respective chemicals were 4.62, 5.50, 7.39, and 8.18. The results deviated by less than 0.1 log Pow units from the reference log Pow. In addition, the relative standard deviations of log Pow were less than 2%. Using the protocol of the ring test, log Pow of p,p'-DDT was determined to be 6.24 +/- 0.05 (mean +/- standard deviation). In combination, these results indicate that the slow-stirring method is precise and accurate and, thus, allows for reliable determination of log Pow of highly hydrophobic chemicals.

Environmental benzene exposure assessment for parent-child pairs in Rouen, France.

There is a lack of data on environmental benzene exposure in children. In this study, we compared personal benzene exposure and inhalation uptake in a group of children to those of their parents. We also compared levels of urinary benzene metabolites, trans,trans-muconic acid (MA) and hydroquinone (HQ), for those two groups, and assessed the correlation between personal benzene exposure and urinary MA and HQ concentrations. The study was performed on 21, 2-3-year-old children and their parents recruited on a voluntary basis among non-smokers from the three largest day-care centers of the town of Rouen in France. Average benzene concentrations were measured over 5 consecutive days with diffusive samplers. The following simultaneous measurements were carried out: personal exposure of the parents, concentrations inside and outside the day care centers, and inside the volunteer's bedrooms. Morning and evening urine samples were collected during the same period. Benzene personal exposure levels were 14.4+/-7.7 microg/m(3) and 11.09+/-6.15 microg/m(3) in parents and children, respectively. Benzene inhalation uptake estimates were 2.51+/-1.23 microg/kg/day in the group of parents and 5.68+/-3.17 microg/kg/day in the group of children. Detectable levels of MA and HQ were found in 85% and 100% of the samples, respectively. Intra-individual variation of urinary MA and HQ concentrations expressed as a coefficient of variation (CV) ranged from 63 to 232% and from 13 to 144%, respectively. Mean values of MA and HQ (in mg/g creatinine) were 1.6- and 1.8-fold higher in the group of children than in the group of parents (P=0.008 and P<0.0001, respectively). Significant correlations between metabolites levels and benzene were not found.