Lead is a well known toxic metal, and current exposures in children constitute a reason for concern 1 . In France, lead has multiple anthropogenic sources and is now mainly present in its inorganic form in the environment 2 3 . The relative importance of different sources depends on the blood lead range. For the general European population 1 and for children 4 , food is usually the major source of exposure, with cereals and vegetables products contributing mostly to dietary lead exposure. Tap water can also, in some cases, be an important contributor because of the presence of lead pipes in old homes and public plumbing systems. Degradation of old lead-based paint results in the contamination of indoor dust that can be inhaled or ingested, thus adding to the sources already mentioned. Other incidental sources of lead exposure include consumer products, notably toys, and hobbies or occupations involving lead 3 . After the ban of leaded petrol, air concentrations have decreased substantially and are now due almost entirely to industrial emissions 5 6 . In France, the targeted regulations to decrease elevated B-Pb concentrations, control measures and screening strategies have progressively reduced risks from lead pipes, lead-based paint in houses built before 1949 and contamination at specific industrial sites 3 . Children under six years of age have the highest exposure to lead because of several factors such as greater hand dust contamination, frequent hand-to-mouth transfer and higher absorption rates than adults. Also, lead can pass through the placenta so that the child is born with lead from the mother's cumulated body burden 7 . Overall, lead poisoning is still a serious hazard for children and causes significant neurologic damage linked to cognitive and behavioral impairment 1 8 . Although frequently overlooked, the timing of the dose in regard to windows of highest vulnerability in children is also important 9 10 .

The first national study carried out in France in 1999 by the National Institute of Health and Medical Research (INSERM) showed that 2% of French children aged one to six years of age had B-Pb concentrations > 100 μg/L (i.e. approximately 85,000 children); the geometric mean blood-lead concentration was 37 μg/L 2 . This exposure level was similar to other Western European countries 11 . In a new survey, 2008-2009, the National Institute for Health Surveillance (InVS) 12 found that the geometric mean B-Pb had decreased to 15 μg/L (standard deviation [SD], 1.6) among children aged 1-6 years, and the prevalence of B-Pb concentrations > 100 μg/L had dwindled to 0.11% (i.e. 5,333 children) 12 . Nonetheless, many children are still at risk because there is no evidence for a lead toxicity threshold. The B-Pb concentration intervention value in the US and France is 100 μg/L; above this limit the subject is considered as lead poisoning by public health authorities and is supposed to be reported in the French National system of surveillance of children's B-Pb concentrations. At lower values lead toxicity may still cause damage to nervous system functions, including decreased nerve conduction velocity and cognitive deficits 1 , and significant neurologic damage may occur as a result of both intrauterine and postnatal exposures 13 14 . The intellectual decrement may be expressed in terms of a loss of IQ points for every μg/L unit increase of the B-Pb, but this loss slope is steeper at B-Pb concentrations lower than 100 μg/L than at higher levels 14 . At the individual level, this drop may seem small and inconsequential, but at the population level, small effects in many individuals are likely to have an impact on the overall societal benefits 11 . The effects include lower school performance and educational attainment, which may influence societal adaptation and economic success, with some affected children showing juvenile delinquency 11 15 . Therefore, improvements in cognitive ability will benefit society by raising both economic wealth and overall wellbeing. Several economic studies, mainly in the US, have estimated the costs and risks associated with infantile lead poisoning and lead toxicity, in some cases weighing them against the costs associated with lead-based paint control and other efforts. These studies have also calculated the potential increased financial earnings that would result if the level of lead in children's blood were to be reduced [ 8 16 , and 17 ]. In France, studies are mostly epidemiological, focusing on targeted screening and lead exposure. There have been few economic assessments of lead's impact on the children's health, with the exception of the studies by Chanel 18 19 20 , while Fassin and colleagues highlighted the social aspects of lead exposure 21 . The present paper aims to fill the gap and contribute at least in part to a cost benefit analysis (CBA), while taking into account that there is "no single estimate that accurately reflects the costs and the benefits of lead hazard control" 8 . We first summarize the childhood lead exposure situation in France and related information on the main exposure media and risk factors. We then estimate the monetary benefits that can be expected from pollutant abatement, with estimates of investment costs to achieve this reduction, as based on available information. Lastly, we compare the main findings of this study and discuss the role of CBA in a societal perspective of public policy development.

The aim of this paper was to provide an economic evaluation of the health impacts of children with lead exposure in France. Based on the assumption of the EFSA report 1 , that there is no threshold of lead exposure, our study provides a range of annual benefits and partial costs estimated in order to highlight the economic impact for society of lead exposure reduction policies below the conventional B-Pb screening value of 100 μg/L. Several hypothetical threshold values for intoxication (15, 24, 100 μg/L, respectively) were chosen following a "what if" approach. We have no strong data to choose levels lower than 15 μg/L but also do not assume it to be a safe exposure level. The partial cost benefit analysis documents a clear cost effectiveness of lead hazard control, which should result in benefits greatly superior to the costs, as suggested by the comparison of the sum of benefits to that of congruent costs for one year. This study showed that by reducing childhood lead exposure, large social benefits might be produced for the birth cohort of 2008 (and subsequent years): € 22.72 billion, € 10.72 billion and € 0.44 billion, respectively. The benefits were mainly due to the social avoided costs, specifically the lost life time earnings, at exposures corresponding to B-Pb <100 μg/L. There are some limitations to our analysis, due in particular to access to figures related to avoided costs and to costs of exposure reduction as we will see below. This is the reason why we could not perform a complete CBA. Direct health costs were also estimated but they were probably underestimated. Lead exposure provokes other health impacts besides cognitive disorders which were not assessed in this paper, such as cardiovascular diseases and cancer that lead to premature mortality. This would yield higher social costs than IQ decrement alone 56 . We disregarded for instance, drug costs and medical intervention costs such as intravenous chelation. Among other costs, the pretium doloris calculated on the basis of €8,000 per child in the Metal Blanc judgment was certainly underestimated, because only a small part of the children have been compensated, while also neglecting the psychological and economic suffering of the family or household of the children affected. We also estimated the need for special education to be 10% for children with B-Pb ≥ 100 μg/L. The somewhat uncertain data on crime costs suggest that the economic impact is comparatively low, but the costs of crime and rape were probably underestimated, because they did not include the value of statistical life, which may be greater than that of accidents (between €₁₉₉₉ 0.5 to 1.5 million in Europe and French estimations were the lowest bracket estimate) 57 58 .

They highlight the additional social consequences of lead pollution. In regard to annual costs to invest in pollution abatement, our preliminary estimates are affected by the paucity of available data. We could not make a complete CBA because of lack of available data on the abatement costs, we had a very small part of the industrial costs. Official data from the ministry of Environment show that the major industrial sources of lead in France are the metals and non metallic minerals sectors 59 . Three quarters of the 2007 emissions took place through water, and two waste treatment facilities alone amounted to 60% of total emissions of the ten most emitting facilities 60 . We had also quite imprecise cost estimates for substitution of lead pipes, whose yearly estimates are certainly exaggerated. So far, clean-up costs of industrial lead-contaminated sites cannot be evaluated in France. Partial data stem from the experience of the highly polluted MetalEurop site remediated by SITA-Suez Environment, which amounted to €28 million 61 . Unfortunately, these findings cannot be extrapolated to the national situation. As to contaminated sites, we point out the need for a specific evaluation. However, costs to decontaminate French houses with lead-based paint were available. And we calculated these costs once-for-all in one year, even if we overestimated the annual expenses, they appeared to be the most important efforts to be made in order to control the hazard. We could express an equivalent annual cost by using the capital recovery factor of standard interest calculations for loans which is the appropriate conversion factor. However, uncertainties remain regarding the time horizon and the social discount rate to use. A 0.05 conversion factor between one-time cost and annual cost is a compromise.

Some of the costs were paid within one year or paid over no more than five years, costs would be substantially less subsequent to that, while benefits would continue to accrue for each new birth cohort being born during the following years.

Our first estimates of total net benefit induced by reducing exposure to soils and dust in respect of the costs incurred by the decontamination of French houses with lead-based paint highlight that policies aimed at reducing lead exposures had an overall positive societal and economic impact. Additional estimates of total net benefit were performed, that considered the costs associated with dust and soils and drinking water lead pipes substitution. The expected health gains, according to the different B-Pb hypothetical threshold values, were calculated to be € 3.9 to 4 billion, € 1.86-2 billion and €0.12-0.25 billion respectively. The corresponding figures per child range from €1,661 to €1,721, €2,666 to €2,861, and €21,939 to €47,815, respectively. These estimates should be considered with caution, because of the uncertainty in the quality of data on costs of lead water pipes removal; a specific evaluation is also needed here.

Various uncertainties exist in our calculations: benefits linked to the dose-response function, and monetary valuation of the abatement costs linked to houses remediation, which yield uncertainties in the partial cost benefit estimates. According to Rabl and colleagues, there is a factor two uncertainty, both in the dose-response function and in the monetary valuation 62 63 . Should the scientific literature show some day evidence of lower toxicity level values than the one we used in this sensitivity analysis, the health cost figures would be substantially increased.

The overall return of investments is important and must be taken into account by the policy makers. They are in line with several US findings that illustrate how reduction of childhood lead exposure has a high social benefit, in particular the studies from Schwartz 16 , Salkever 17 and Grosse and colleagues 64 . Between 1976 and 1999, Grosse et al. 64 estimated the economic impact of the trend of reduced lead exposure over 25 years in a cohort of children starting at 2 years of age in 2000. The estimate cost was valued from $110 to $319 billion (US) for the cohort each year, comparing it as if the blood lead concentration were that same as in 1975. Landrigan et al. 34 estimated the total annual costs of childhood lead poisoning to be $₁₉₉₇43 billion in each birth cohort exposed to lead in the US. Their methodological approach was based on the contribution of environmental pollutants by using an Environmentally Attributable Fraction (EAF) model, which was estimated at 100% for lead poisoning. Recent studies calculated the economic impact of childhood poisoning below100 μg/L. The most recent major U.S. study was that of Gould 8 . It was more comprehensive than those previously published, and produced a CBA by comparing the estimated costs in 1996 related to cleanup of lead-containing paint in the U.S. ($ 1 - $11 billion (US)) and secondly, by calculating the monetary benefits and social benefits by reducing lead exposure for a cohort of children <6 years ($192 - $270 billion) with earning losses amounted to 87% of total avoided costs. Total net benefits amounted to $ 181 - $ 269 billion. Therefore, a specific calculation induced by lead-based paint was not performed in this study. Muennig et al. 65 , whereas, provided information on the benefits that might be realized if all children in the United States had a blood lead level of less than 10 μg/L. The net societal benefits showed improvements in high school graduation rates and reductions in crime would amount to $50,000 (SD, $14,000) per child annually at a discount rate of 3%. This would result in overall savings of approximately $1.2 trillion (SD, $341 billion) and produce an additional 4.8 million QALYs (SD, 2 million QALYs) for the US society as a whole.

Researchers in other European countries with prevalence of lead exposures similar to French figures may use this as a guide as to undertake similar economic assessments. Additionally, these data may motivate the revision of the current French policies as to whether or not to intervene in regard to lead pollution, and, in a more general sense, revamping France's overall policy on reducing pollution that may be affecting children's development. The introduction of unleaded petrol has greatly decreased emissions of lead in the atmosphere in France and globally. (Paris ambient air concentrations decreased by 97% between 1991 and 2005) 66 . The relative benefits of this action were substantial 3 and likely much greater than the benefits from further reduction of B-Pb levels today. Nonetheless, much abatement remains to be done, as other sources are only slowly being removed, if at all. The screening of houses for sale or rent with lead-based paint was implemented through the 2004 Public Health Act and its stringent policies on industrial emissions were triggered by EU regulations. The French 2004 national environmental health action plan has also contributed to the steady decrease in exposure of the general population and of its most vulnerable young segment over the last years in France.

EFSA recommends that "work should continue to reduce exposure to lead, from both dietary and non-dietary sources" 1 . The major prevention campaigns aim to reduce lead exposure to the lowest possible level in order to protect children and childbearing age women. The obtained benefits for exposure levels <100 μg/L in this study are in line with the EFSA recommendations. They are a first step evaluation which should be expanded and refined. Our results emphasize the substantial monetary advantages obtained from preventing losses of a few IQ points because of lead exposures among children. While 1-point change in Full Scale IQ score is within the standard error of an individual's single measurement, it may be highly significant on a population basis 25 .

The primary economic benefits of policies focused on lead exposure abatement are the further reduction of low blood lead levels. In contrast, prevention of cases with B-Pb >100 μg/L accounts for much lower benefits. This is because children with milder exposures are much more common and they still benefit from decreased exposure, as there is no known safe level of lead exposure. Lead toxicity is still a serious public health issue, despite the present low prevalence of unacceptably high B-Pb concentrations. Public policies to prevent lead exposure will reduce future medical expenses and the reduce the burden on special education classes. More importantly, they will also increase the productivity of children during their adult lives. Our CBA results suggest that overall reduction of costs due to toxicity can be achieved by further control of major contact media, including food, through diffusion of lead in the environment from industrial releases and also by further control of residential sources (leaded paint, deteriorated housing, old water pipes). In addition to abating the burden of developmental impairment in general, these policies will also help to reduce health disparities. This objective calls for prioritized policies focused on the most highly exposed communities and individuals. This combined strategy is a policy issue that our data aim to inspire. Yet, additional documentation of the B-Pb values for further evaluation is needed. A more thorough evaluation of the marginal costs of the measures to be taken is also needed in order to balance lead exposure abatement options.

AFSSA: French Food Safety Agency; ANAH: National Agency of the Housing Environment; B-Pb: blood-lead; CBA: cost benefit analysis; CEPA: California Environmental Protection Agency; COI: Cost of illness; CSHPF: High Council of Hygiene; EAF: Environmentally Attributable Fraction; EFSA: European Food Authority Safety; GDP: Gross Domestic Product; IEGRE: European Institute Reasoned Management for the Environment; IME: Medical Educational Institutes; INERIS: National Institute for Industrial Environment and Risks; INSERM: National Institute of Health and Medical Research; INSEE: National Institute for Statistics and Studies; InVS: French Institute for Public Health Surveillance; IQ: Intellectual Quotient; ITEP: Educational and Therapeutic Institutes; LERES: Laboratory study and research in environment and health; PPP: Purchasing Power Parity; QALY: Quality-Adjusted Life Year; SNSPE: French National system of surveillance of children's B-Pb concentrations.

The authors declare that they have no competing interests. PGr is an editor-in-chief of Environmental Health, but was not involved in the editorial handling of this manuscript.

CP performed the literature review, drafted the manuscript and carried out the analysis. MB, DZN, PGr, PGl and PH contributed substantially to defining the methods of the analysis, interpreting the results of the study and editing the manuscript. All authors read and approved the final version.