SYMMARY
A lead (Pb) smelter continues to operate in the City of Shymkent, South Kazakhstan. Investigations have been conducted to evaluate the exposure pathways of Pb poisoning in children living nearby the smelter including contaminated air, soil and dust. Pb in soil is a long-term environmental pollutant from the smelter with an area of approximately 50 sq. km. having surface soil concentrations over 400 mg Pb/kg. Soil Pb was determined to be the main reason for elevated lead concentrations in blood collected from children in the area, i.e., 95% of children living close to the smelter have blood Pb concentrations >10 9g pb/dl. Children attending one kindergarten had a geometric mean blood Pb concentration of 27.7 9g/dL. Various strategies for environmental remediation were identified and the calibrated IEUBK model applied to evaluate the effectiveness of the possible remediation scenarios. A significant correlation between IEUBK predicted and measured blood Pb concentrations was observed (r=0.99, P<0.01). The IEUBK model combined with Monte Carlo simulation predicts that if soil is remediated to a cleanup level of 250 mg Pb/kg and emissions are controlled from the smelter, the expected blood Pb concentrations would be <10 9g pb/dl for 90-92% of the children living close to the smelter.
Keywords: lead in blood, bio-kinetic model, soil remediation
Introduction. Lead (Pb) in the environment, especially high soil Pb, will remain a serious environmental health hazard for young children until adequately addressed to prevent future Pb exposures. The Blacksmith Institute lists approximately 20 sites in the world with critical levels of lead poisoning of children.[1] At the same time, many areas in developing countries with high lead concentrations in the environment have not been well assessed, and the countries have difficulties with addressing such problems. A comprehensive investigation of the lead poisoning of children in Kazakhstan by a group of USA and Kazakh scientists started in 1997 with the sponsorship of the Civilian Research & Development Foundation (CRDF). In Shymkent, the scientists defined a zone of 50 sq. km. with very high environmental contamination in the soil around a lead smelter that had been operating for approximately 60 years without adequate pollution control. In 2008, a feasibility study was conducted to further characterize the extent of environmental contamination and identify and test potential remediation approaches. The potential remediation strategies were then evaluated using the IEUBK model to predict blood Pb levels in children in the vicinity of the smelter after remediation was completed.
The Integrated Exposure Uptake Biokinetic (IEUBK) Model for Lead in Children was used to predict the risk of elevated blood lead (PbB) levels in children (under the age of seven) that are exposed to environmental lead (Pb) from many sources. The model also predicts the risk (e.g., probability) that a typical child, exposed to specified media Pb concentrations, will have pbb levels greater or equal to the level associated with adverse health effects (10 9g/dL). [2] 2. The IEUBK model includes more than 120 equations to describe the uptake and bio-kinetic processes of children’s lead exposure.[4] Some of the equations are linear and some have complex mathematical structure, including parabolic, hyperbolic and exponential functions. Various studies have shown the advantages of the model compared to the linear equation used prior to IEUBK.[5-8] However many times the model can be difficult to accurately apply to certain Pb exposures without additional investigations.[9] The United States Environmental protection Agency (usepa) states that the ieubk is not calibrated for children with lead in blood geometric means above 30 9g/dL.[10].
USEPA also proposed a validation strategy for the IEUBK model.[11] The authors of the strategy state that model validation should be considered an iterative process that aims to test applications of a model 3 under a variety of conditions (combinations of input variables) that ideally span those conditions where applications of the model will be made. While the collection of data relevant to evaluation of the IEUBK model's performance is quite extensive, it is typically observational, not experimental in nature. When observational data (such as those obtained in epidemiologic studies) are used in model evaluation, inconsistencies between the modeled and observed systems can be particularly problematic. An observational study does not allow the investigator to fix the values of some variables or isolate processes to be measured quantitatively as can be done in an experimental approach. In fact, values for some variables may not be measurable under field conditions. On the other hand, even if experimental studies on children were acceptable by rigorously measuring lead exposure and blood lead levels over a full range of the lead concentrations and exposures of interest, such data would be unlikely to reflect real-world exposure conditions, due to the practical difficulties in measuring actual exposure.
Nevertheless, the IEUBK has been validated using various regional studies with unique sources of Pb exposure. For example in northeastern Pennsylvania (USA), blood Pb data were collected during emergency remediation activities at a former lead-acid battery recycling plant.[12] Estimates of household dust Pb concentrations based upon surface soil concentrations resulted in IEUBK blood lead predictions which were consistent with data collected during the blood lead monitoring program. The validity of the IEUBK predictions were confirmed using the given assumptions including estimates of household dust Pb levels[12]. In addition, simulations using the IEUBK model were conducted to estimate the blood lead concentrations in children living in the vicinity of a non-ferrous plant located in Hoboken, Belgium.[13] Pb concentrations in soil ranged from 59 - 2,425 mg Pb/kg and concentrations in house dust ranged from 234 - 73,394 mg Pb/kg. Measured blood lead concentrations in children aged 2-7 years were between 3 and 35 9g/dL. Exposure sources included indoor dust in the houses and the schools, outdoor dust and soil near the homes surrounding the plant and in the school's playground, and suspended dust in the air. The IEUBK model was able to predict the measured blood lead concentrations adequately except for the lowest exposure group which had higher predicted values compared to measured values. Analysis of the data revealed that the neighborhood’s school is an important source of exposure to lead. In addition, indoor house dust dominated exposure for children going to school outside the area because of the high concentrations of lead in indoor dust (3 to 5 times higher than in outdoor soil)[13].
Some authors showed that the IEUBK does not always fit the theoretical concepts of lead uptake, as well as the observed data resulting from biological monitoring. For example, E. J. O'Flaherty suggested that the IEUBK model “...is not physiologically based”.[14]. Since 1989 in Trail, British Columbia (Canada), the community has been monitoring blood lead levels in children, studying exposure pathways and conducting comprehensive education and management programs.[15] From 1989 through 1996, mean blood lead levels of pre-school children declined at an average rate of 0.6 9g/dL per year. From 1996 to 1999, mean blood lead levels fell at an average rate of 1.8 9g/dL per year, from 11.5 9g/dL in 1996 to 5.9 9g/dL in 1999. In 1998, the annual arithmetic mean air lead level in Trail was 0.28 9g/m3, compared to 1.1 9g/m3 in 1996. Reductions of approximately 50% were observed in Pb concentrations in outdoor dust, street dust and indoor dust after smelter emissions were reduced. Slight reductions (statistically insignificant) were observed in household carpet dust and soil lead concentrations.
During the summer of 2001, the smelting and refining operations at Trail were shut down completely for 3 months. During this period, average air lead levels in Trail dropped to 0.03 9g/m3. The average blood lead level in Trail pre-school children at the end of the shutdown was 4.7 9g/dL. These results can be compared to the prevailing theories concerning the 4 relative importance of various environmental lead sources. For example because of the importance of Pb soil concentrations, simulations with the IEUBK model would not have predicted the dramatic decline in children's blood lead levels observed in Trail following the reductions in air lead levels. The Trail experience suggests that more attention should be paid to active sources of highly bioavailable lead bearing dusts from air emissions.
In another study [16] Monte Carlo simulations were used to make IEUBK predictions more consistent with the observation data. Based on the available deterministic information concerning exposure, absorption and biokinetic processes of lead, the IEUBK model was used to produce a geometric mean blood lead concentration for a lognormal probability distribution. This distribution was used to describe the blood lead level’s inter-individual variability for children exposed to similar environmental concentrations. This modeling approach allowed differences in behavior, population heterogeneity and individual patterns of lead uptake and biokinetics to be evaluated. The lognormal distribution was used because it is generally accepted as a plausible model of a number of independent components and is the typically observed distribution of environmental data. However without actual data, the lognormal distribution is arbitrary and its use can lead to unrealistic decision-making in risk management. In the study, the distribution was used in conjunction with a deterministic geometric standard deviation derived from empirical studies, independent from input parameters.
One purpose of the evaluations in Shymkent, Kazakhstan was to calibrate the IEUBK model when applied to measured data for the area in the vicinity of the smelter. High levels of environmental contamination, as well as the variability of the contamination in the whole area, enabled evaluation of various assumptions and evaluation of the validity of the IEUBK model for use in selection of remediation plans for the City.
Materials and methods. Background. Shymkent is situated in the southern part of Kazakhstan and is one of the most populated regions of this Central Asian country. The lead smelter has been operating in Shymkent since the 1930s. The smelter provided the majority of lead for bullets used by the Soviet army during World War II. During 1998-2008, many state-of-the-art methods were used to determine environmental exposures of concern and the potential for lead poisoning in the children of Shymkent. For measurements of lead in soils, dust and paint XRF instruments were used. Also the samples from Shymkent were analyzed by electronmicroprobe techniques at the University of Colorado. Lead in blood was determined using portable devices. In addition, zincprotoporphryn (ZnPP) in blood was measured using a portable fluorometric instrument. Site specific studies were conducted in Shymkent during 2002 and 2003 (soil and blood samples) and 2008 (soil samples). Specifically, the following methods were used:
Metals in soil using X-Ray Fluorescence (XRF): After collection (usually 4-5 composites at each location) and drying, the soil was analyzed using a portable XRF spectrum analyzer (NITON XL-700 in 2002-2003 and InnovXsystem Alpha-4000 in 2008). Simultaneous K-shell and L-shell analyses were performed for the following metals Pb, As, Mo, Zr, Sr, Rb, Hg, Zn, Cu, Ni, Co, Fe, Mn and Cr by x-ray detection measured in mg/kg. The accuracy of the analyses was checked by re-analyzing selected samples in the USA with a laboratory XRF (Spectrace 7000 according to EPA approved method [17]). Calibration of the portable XRF 5 instrument was checked before and after each series of analyses. Confirmation of the analytical results was performed using split sample analyses in USA and Kazakh laboratories (USEPA Method for ICP/MS [17]).
Air Samples: Air samples were collected on filters using personal hygiene sampling pumps. The pumps recorded the volume of air pulled through the filters. The pumps and filters were typically mounted at a height similar to a person’s breathing zone. The pump ran for 24 hours before the filters were removed. The concentration of lead on the filters was determined by analyzing the filters using the XRF techniques described above (the instruments had special attachments to hold the air filters). The filters were weighted to determine the mass of lead and the air concentration was determined by dividing the mass by the total volume of air through the filter in 24 hours. Lead in Blood by Anodic Stripping Voltametry (ASV): Blood samples were collected and analyzed on site using the ESA LeadCare Blood Lead Testing System (model 3010B analyzer). In particular, 50 microliters of blood were quantitatively transferred and mixed with a treatment reagent. In the reagent, the lead was removed from the red blood cells and made available for detection.
The sample was transferred to the asv instrument where the lead was detected and results displayed as 9g/dL lead in the blood sample. Analysis took approximately ten minutes per sample. Calibration was checked using two blood lead standards (low and high) twice per day on each instrument. The procedure has been approved by the United States Environmental Protection Agency (USEPA). During this time over 1000 measurements of Pb in soil were made.[18] Based on these data, an iso-concentration map of the surface soil (0 to 2 cm depth) Pb concentrations was prepared using Kriging techniques. The Kriged iso-concentration map is provided in Figure 2. As shown, an area of approximately 50 sq. km. has concentration of Pb in surface soil greater than 400 mg/kg. Many areas had concentrations above 10,000 mg/kg. Outside of the area with values above 1,000 mg/kg, lead concentrations decreased, posing very little risk to residents in remote districts of the City.
The contaminated zone is located close to in the center of the City, with about 60,000 children less than 7 years old living in the area. Approximately 6 schools and 8 day-care centers operate within in the contaminated zone. In 1998, high levels of contamination and Pb blood levels were measured at day-care center (kindergarten) “Sholpan” (“Morning star” in Kazakh) which is close to the lead smelter (less than 1 km). The maximum soil lead concentration measured at Sholpan was 24,000 mg/kg. In 2008, several different methods for soil treatment and remediation in Shymkent were identified and evaluated. These included excavation, cover, mixing, and stabilization/treatment using phosphate or sulfide chemicals. [17]
Bioavailability Determination. Historically the relative bioavailability (RBA) of metals in soils has been determined using in vivo, or animal based studies. Typically the in vivo studies for lead have utilized young swine and have determined RBA coefficients through the comparison of the amount of lead ingested versus amount of lead excreted. The in vivo studies are generally accepted as the best means of assessing bioavailability; however, they are extremely expensive and can take many months to complete, therefore researchers have developed an in vitro, or laboratory based method to determine RBA coefficients. 19-20. Dr. John Drexler at the University of Colorado Laboratory of Environmental and Geological Studies (LEGS) has worked in conjunction with EPA Region 8 to develop an in vitro method that can be used to obtain RBA data for lead in soils. The in vitro method simulates the human gut by reproducing stomach residence time, pH, temperature, agitation, and solid to liquid ratio. Prior to analysis, the sample is sieved with a 250 micro sieve and the <250 micron sieved fraction of the sample is used in the analysis. The sub <250 micron fraction is analyzed as those particles are most likely to adhere to human skin and thus be ingested via hand to mouth processes. In general, the in vitro method consists of placing one gram of sample test material into a plastic bottle with 100 mL of an extraction fluid (0.4 M glycine, pH 1.5).
Each plastic bottle is placed in a water bath maintained at 37.C and rotated end-over-end for 1 hour. After 1 hour, the bottles are removed and placed upright to allow the soil to settle to the bottom. After settling, a 15-mL sample of supernatant fluid is removed directly from the extraction bottle into a disposable 20-cc syringe, and then filtered through a 0.45-um cellulose acetate filter to remove any suspended particulate matter. This filtered sample of extraction fluid is then analyzed by ICP/MS 7 techniques (method 6200) to quantify the fraction of lead or arsenic in the sample which had dissolved. The in vivo procedure outline above has been validated and accepted by USEPA (method 9200). Based on the analytical results (XRF and ICP/MS), four representative soil samples ranging from 3,030 to 18,924 mg/kg of Pb (<250 micron fraction) were evaluated using the in vitro procedure described above. The RBA values ranged from 93 to 113%. Therefore the lead in the soil from Shymkent has a greater bioavailability than the standard IEUBK assumption (60%). Based on an estimation of 100% lead bioavailability in Shymkent, 50% uptake was used in the IEUBK model for lead and indoor dust. The IEUBK model was used to evaluate the possible effectiveness of the remediation methods. STATISTIKA software was used to produce descriptive statistics, correlation and multiple regression analyses. @RISK software was utilized to provide Monte Carlo simulations.
RESULTS. The IEUBK model uses lead concentrations in outdoor soil, indoor dust, air, food, and water to predict a distribution of lead concentrations in children’s blood. International consensus has generally agreed on criteria of 10 9g/dL in blood of infants and young children as an upper limit before lead poisoning causes health problems. The IEUBK model estimates the probability distribution of blood lead levels in exposed children up to age 7 and can beused to estimate blood lead levels greater than 10 9g/dL for specific exposure scenarios. Although the focus of this study is on soil and dust, consideration and evaluation of all of the exposure pathways and exposure routes allowed in the IEUBK model are important. Therefore, the IEUBK model must be calibrated using site specific information in order to apply it to complex environmental conditions similar to those found in Shymkent.
Table 1 - A summary of results of biological and environmental testing for four locations in Shymkent are provided in the
Tablel.Sum mary results of biological and environmental testing in Shymkent # |
Location |
Number of blood samples |
Lead in Blood, Geometric Mean,Geometr ic Standard Deviation, range, 9g/dL |
Zincprotoporp hryn in blood (ZnPP), Mean ± Standard Deviation, 9g/dL |
Average air lead concentr ation, 9g/m3 |
Average soil lead concentrati on, mg/kg |
Average indoor dust lead concentrati on, mg/kg |
1 |
Sholpan |
73 |
27.7, 1.8, 2.1-103 |
69.54±46.85 |
7.9 |
1313 |
4625 |
2 |
House of Invalids |
10 |
12.8, 1.6, 4.7-22.8 |
51.75±23.6 |
5 |
794 |
433 |
3 |
Pink Flower |
59 |
7.72, 1.6, 3.0-20.5 |
50.19±19.38 |
1 |
104 |
1400 |
4 |
Orphanag e 3 |
14 |
18.4, 1.4, 9.7-29.9 |
53.64±19.98 |
5 |
1530 |
795 |
The IEUBK model requires additional calibration to accurately predict the measured blood lead concentrations. The data from the four Shymkent locations provided in Table 1 were used in the IEUBK model and then predictions of blood lead levels using different ratios of ingested soil and indoor dust were made. The default ratio in the IEUBK model is 45%/55% ingested soil/indoor dust. The average of squared residuals of modeled and measured blood lead data as a function of the soil/dust ingestion weighting factor is shown on figure 4 (the x-axis provides the percent of soil ingestion). As shown, the weighting factor of 70%/30% for Shymkent’s conditions proved to be the best estimate for the calibration of IEUBK model. The comparison of measured and modeled data for each of the four locations is illustrated in figure 5. The default weighting factor of IEUBK model, 45%/55%, was replaced with 70%/30% to account for local conditions. n figure 6, the distributions of the observed and predicted levels of blood lead for children in Sholpan kindergarten are shown. Blue bars demonstrates the measured values, red line reflects the log-normal distribution with 9=3.28, : =0.47 (calculated by the ieubk model). The distributions match each other with high statistical significance (p<0.01). The IEUBK model predicts the blood lead effectively, including values greater than 30 9g/dL. The calibrated ieubk model parameters were utilized to determine the effectiveness of remediating soil under differing exposure assumptions. The results of IEUBK predictions are provided in the table 2.
Table 2 - Results of IEUBK predictions of lead in blood distribution for different remediation scenarios
# |
Scenario |
Predicted distribution of lead in blood |
1 . |
Soil is remediated to 2500 mg/kg of lead, lead in the soil is treated with chemicals, smelter continues to operate producing 5 mg/m3 average lead air concentration |
95% of children with lead in blood less than 21.5 9g/dL |
2 . |
Soil is remediated to 2500 mg/kg of lead, lead in the soil is treated with chemicals, indoor dust is cleaned up to 290 mg/kg, no smelter emission are occurring |
95% of the children with blood-lead concentrations less than 15 9g/dL, 79% of the children with blood-lead concentrations less than 10 9g/dL. |
3 . |
Soil is remediated to 400 mg/kg of lead, lead in the soil is treated with chemicals, indoor dust is cleaned up to 290 mg/kg, no smelter emission are occurring |
95% of the children with blood-lead concentrations less than 16 9g/dL, 72% of the children with blood-lead concentrations less than 10 9g/dL |
4 . |
Soil is remediated to 250 mg/kg of lead, lead in the soil is treated with the chemicals, indoor dust is cleaned up to 290 mg/kg, no smelter emission are occurring |
90% of the children with blood-lead concentrations less than 10 9g/dL |
5 . |
Soil is remediated to 200 mg/kg of lead, lead in the soil is treated with chemicals, indoor dust is cleaned up to 290 mg/kg, no smelter emission are occurring |
95.1% of the children with blood-lead concentrations less than 10 9g/dL |
The first remediation scenario assumes that soil is remediated to 2500 mg/kg lead and that the lead in the soil is stabilized, treated with chemicals, resulting in 14% bioavailability (7% effective uptake). This value is based on actual tests of the stabilization agent with soils from Shymkent having high Pb concentrations. A further assumption is that that the smelter continues to operate in such a manner that produces a 5 mg/m3 average lead air concentration. Under such circumstances, the IEUBK model predicts a 2250 mg/kg indoor lead dust concentration. Fourteen percent bioavailability (7% bodily uptake) was assumed for tracked-in dust in indoor areas from outdoor soil, and 100% bioavailability (50% uptake) was assumed for smelter emissions that settle in indoor spaces. The assumptions result in an average of 16.6% uptake of ingested indoor dust. Overall, this scenario results in a median blood-lead concentration of 9.9 9g/dL compared with a measured median blood-lead concentration of 29.1 9g/dL for children, an approximate 2/3 reduction in blood-lead concentration, which is a significant improvement. The predicted blood-level distribution for this scenario would be somewhat similar to or less than that associated with USA children when leaded gasoline was in use. Ninety five percent of children are predicted to have blood-lead levels less than 21.5 9g/dL.
The second remediation scenario is the same as the first (remediation to 2500 mg/kg using soil stabilization/treatment methods) but assumes that no on-going smelter emissions are occurring and that existing indoor dust has been cleaned up. The predicted median blood-lead level of 6.8 9g/dL represents a reduction of more than 75% compared with levels measured in 2002 at Sholpan kindergarten. In addition, approximately 95% of the children would have blood-lead concentrations less than 15 9g/dL, and 79% of the children would have blood-lead concentrations less than 10 9g/dL. In the third remediation scenario, the soil is cleaned up to 400 mg/kg by various remediation techniques including mixing, addition of cover materials and stabilization/treatment. The assumption of no smelter emissions, with essentially no contribution to indoor dust levels from airborne lead, was used. As a result, approximately 95% of the children would have lead blood levels less than 16 9g/dL. Seventy two percent of the children would have lead blood- lead concentrations less than 10 9g/dL.
The calibrated ieubk model suggests a higher percentage of children above 10 9g/dL than at this same soil concentration for most USA scenarios because of the assumptions of higher uptake (50% instead of 30%) and a higher contribution from soil (70% instead of 45%). In the fourth remediation scenario, the soil was cleaned up to 250 mg/kg using various techniques and no smelter emissions were assumed to be present. The IEUBK model predicts that 90% of the children would have blood lead levels of less than 10 9g/dL. If soil was remediated to a level of 200 to 250 mg/kg and such remediation was technologically possible and affordable, this alternative soil cleanup goal should be considered, especially in combination with controlling smelter emissions.The calibrated IEUBK model predicts that a 200 mg/kg soil remediation goal (scenario five) would result in 95.1% of the children below 10 9g/dL. Of course, at these levels, the influences of other environmental sources of lead such as that from lead paint, lead glaze on eating utensils, food sources, homeopathic remedies, etc. become more important to control. Figure 8 compares two of the possible scenarios.
Discussion. The IEUBK model used to assess different soil remediation scenarios for Shymkent is an effective evaluation tool if applied together with reliable statistical procedures including calibration and distribution simulations. The high environmental lead contamination found in Shymkent provided a unique opportunity to calibrate the IEUBK model at high concentrations resulting in accurate evaluations at both normal and high concentrations. The environmental studies revealed that Shymkent is currently among the top 10-15 most lead contaminated locations in the world (along with such a places as Tianying, China, La Oroya, Peru, Haina, Dominican Republic). The 1000 mg/kg isoconcentration encloses a large area including the high-density populated zone.
The study showed that the IEUBK model resulted in accurate predictions of the levels and distribution of lead in blood for different locations in Shymkent. However, the reliability of the IEUBK model for predicting lead concentrations depends on specific calibration of the model. The calibration includes the assumptions that the dietary intake and the lead contamination of food and water are equal across the affected area and are less important than soil and dust contamination. The measured concentrations of lead in the air, soil and indoor dust enabled evaluation and resultant determination that the soil/dust ingestion weighting factor is the main calibration parameter of the model for Shymkent. For such an evaluation, consider the function f: lbpredicted = f(R, Lsoil, Lindoor dust, Lair, ;), where LBpredicted = IEUBK prediction of blood lead concentration, Lsoil = measured lead in soil concentration,
Lindoor dust = measured lead concentration in indoor dust,
Lair = measured lead concentration in air,
R = soil/dust factor,
; = parameters assumed to be constant.
Analysis using the Shymkent data shows that LBpredicted is a linear function of SD for every fixed set of parameters. For example, for the case of Sholpan kindergarten:
lbpredicted 1 < -0.206 R +41.15 (r=-0.999, P<0.05), where lbpredicted 1 is the ieubk-predicted blood lead concentration geometric mean for this location. The negative slope indicates that measured indoor dust concentrations are higher than soil concentrations. Considering the linearity of the model in this simulation, readers can understand why the average of the squared residuals of the predicted and measured blood lead levels are a quadratic function of SD, with the single minimum (Figure 5). If LBi is a measured geometric mean blood lead concentration for the location i, and =iR+>i is an approximation of the ieubk predicted blood-lead concentration as a function of soil/dust factor, we find R to minimize the sum of squared residuals:
-(=iR+>i ))2 ? min, this means
=i+2=i2R+2 =i >i =0, or R= = > = which is the only minimum of the function .
Generally, multiple linear regressions can be found to fit IEUBK estimations of blood lead in Shymkent conditions. For example, the model
LBpredicted=4.12+1.3*Lair+0.00485*((100-R)/100* Lindoor dust +R/100* Lsoil ), (1) reproduces the predictions of IEUBK with good convergence (r=0.95, F=785, p<0.01).
It allows us to combine IEUBK model with Monte-Carlo simulation, to take into account uncertainties in prognostic evaluation of remediation results. For example, in remediation scenario 4 we can assume that lead in air will be eliminated entirely, lead in indoor dust has a log-normal distribution with geometric mean=290 mg/kg, and lead in soil has log-normal distribution with geometric mean= 250 mg/kg, geometric standard deviation 1.9 for both cases. Then we can run the Monte Carlo simulation for the function [1], superimposing on this function a log- normal distribution with GSD=1.6. The simulation results with the above assumptions (100000 iterations) is provided in figure 9. In this case, we will be able to predict that 8.4-10.8% of the children will have lead in blood at levels more than 10 9g/dL.
Using linear approximation of IEUBK model, we can also determine requiring remediation goals with more accuracy. If x = (100-R)/100 *Lindoor dust +R/100 *Lsoil (total lead “burden” from soil and indoor dust), so in accordance with the linear model, to reach the goal of a geometric mean of blood lead at the level of, for example, 5.3 9g/dL, we must reduce x to the level of x < 261 - 268*Lair. This level could be achieved by the remediation scenario with elimination of air emissions, remediation of soil to 300 mg/kg and cleaning up the indoor dust to 170 mg/kg or less (assuming soil/dust factor R=70%). For the situation in Shymkent, remediation of the soil combined with an effective on-going community educational program and/or control of smelter emissions will dramatically improve the neurobehavioral status of Shymkent’s young children. Additional measures should also be implemented such as environmental and biological monitoring and creation of “clean playgrounds” that could allow children to spend time in safe environmental conditions. Kazakhstan also needs state-of-the-art procedures for the treatment of children diagnosed with the lead poisoning.
Other lead pathways should also be considered for subsequent improvement of the IEUBK model in Shymkent’s conditions. Of particular concern is the effect of wind blowing contaminated soil from the contaminated industrial zone to residential areas and the effect of on-going smelter emissions on surface soil lead concentrations after remediation is complete. To the extent possible, these sources of 17 potential complication to the predicted extent of blood lead improvement should be modeled, monitored, and controlled.
CONCLUSIONS. 1. The environmental situation in Shymkent, Kazakhstan, seriously impacts public health because of the critical level of lead contamination in air, soil and indoor dust. In the areas close to the functioning lead smelter, more than 94% of children in the age of 0-84 months have blood lead values exceeding the safe level. These conditions require urgent measures to reduce the contamination, control the emissions and execute a comprehensive health protection program for the children. 2. The IEUBK model, developed by the USEPA, proved its validity in the conditions of Shymkent, including blood lead levels above 30 9g/dL, assuming the uptake level for lead in soil and dust is 50%, and the soil/dust weighting ingestion factor is 70/30. 3. Based on the IEUBK prediction, the most effective remediation scenario for Shymkent in the 50 square km. contaminated area would be full reduction of smelter emissions, clean up of the lead in soil to the level of 250 mg/kg and indoor dust to 290 mg/kg. 4. For the investigated range of lead concentrations in the environment and consequent biological effects, linear regressions are found between the IEUBK model predictions and basic input parameters, such as lead in air, soil and indoor dust. The results could be used by the local professionals along with the IEUBK software to analyze the remediation dynamics and progress. This approach also allows use of the IEUBK model together with Monte-Carlo simulations.
REFERENCES
- Succop, P. A., O'Flaherty, Bornschein, R. L. [et al.]. A kinetic model for estimating changes in the concentration of lead in the blood of young children/ P.A.Succop, O'Flaherty, R.L.Bornschein [et al.] Heavy Metals in the Environment: International Conference, New Orleans. September 1987. - ep Consultants Ltd., Edinburgh. - .289-291.
- Cornelis C., Berghmans .,Sprundel . [et al.]. Use of the ieubk Model for Determination of Exposure Routes in View of Site Remediation. Human and Ecological Risk Assessment/ c.cornelis, .Berghmans, .Sprundel [et al.]//An International Journal, - 2006.-Vol.12.- Issue 5. .963 - 982
- Drexler, J.W., Brattin, W.J. An in vitro procedure for estimation of lead relative bioavailability with validation. Hum./J.W. Drexler, W.J. BrattinTEcol. Risk. Assess.-2007.- №13.- P. 383-401.
- Ellen J. O'Flaherty. A Physiologically Based Kinetic Model for Lead in Children and Adults. - Environmental Health Perspectives Supplements. December 1998, Volume 106, № S6.
- Canfield R. L., Henderson C. R., Jr., Cory-Slechta D.A., Cox C., Jusko T. A., Lanphear B. P. (a) Intellectual impairment in children with blood lead concentrations below 10 pg per deciliter. //N. Engl. J. Med.- 2003.-V.348.- P.1517-1526.
- Chaparro, C.M., Fornes, R., Neufeld, L.M., Tena Alavez, G., Egui'a-Liz Cedillo, R., Dewey, K.G. Early umbilical cord clamping contributes to elevated blood lead levels among infants with higher lead exposure/ C.M. Chaparro, R.Fornes, L.M.Neufeld, G.Tena Alavez, R.Egui'a-Liz Cedillo, K.G.Dewey //J Pediatr.- 2007. -V.151.- №56.-P.506-512. 7. Bellinger, D., Leviton, A., Waternaux, C., Needleman, H.L., Rabinowitz, M. Longitudinal analyses of prenatal and postnatal lead exposure and early cognitive development./ D. Bellinger, A.Leviton, C.Waternaux, H.L.Needleman, M.Rabinowitz, //NEngl J Med., 1987, V.316, P.1037-1043.
- Dietrich, K.N., Krafft, K.M., Bornschein, R.L., Hammond, P.B., Berger, O., Succop, P.A., Bier, M. Low-level fetal lead exposure effect on neurobehavioral development in ealy infancy/ K.N.Dietrich, K.M.Krafft, R.L.Bornschein, P.B.Hammond, O.Berger, P.A.Succop, M.Bier// Pediatrics.-1987.-V.80.- P.721-730.
- [Referat] Mickle, M.H. Structure, Use, and Validation of the IEUBK Model. Environmental Health Perspectives Supplements.- December 1998- Vol.106.№S6.
- Bellinger, D., Leviton, A., Rabinowitz, M., Allerd, E., Needleman, H., Schoenbaym, S. Weight gain and maturity in fetuses exposed to low levels of lead/ D.Bellinger, A.Leviton, M.Rabinowitz, E.Allerd, H. Needleman, S.Schoenbaym // Environ Res, 1991.- V.52, P.151-158.
- Dussias, V., Stefos, T., Stefanidis, K., Paraskevaidis, E., Karabini, F., Lolis, D. Lead concentrations in maternal and umbilical cord blood in areas with high and low air pollution/ V.Dussias, T.Stefos, K.Stefanidis, E.Paraskevaidis, F.Karabini, D.Lolis//Clin Exp Obstet Gynecol.-1997.- V.24.- №4.- P.187-189.
- Baghurst, P.A, McMichael, A.J., Wigg, N.R., Vimpani, G.V., Robertson, E.F., Roberts, R.J., Tong, S.L. Environmental exposure to lead and children-s intelligence at the age of seven years/ P.A.Baghurst, A.J.McMichael, N.R.Wigg, G.V.Vimpani, E.F.Robertson, R.J.Roberts, S.L.Tong //N Engl J Med.-1992.- V.327.- P.1279-1284.
- Schroeder, H.A., Mitchener, M. Toxic Effects of Trace Elements on the Reproduction of Mice and Rats/ H.A.Schroeder, M.Mitchener //Arch Env Hlth. -1971- V.23 - P. 102-106.
- Ernhart, C.B., Wolf, W.W., Kennard, M.J., Erhard, P., Filipovich, H.F., Socol, R.J. Intrauterine exposure To low Levels of lead: the status of the neonate/ C.B.Ernhart, W.W.Wolf, M.J.Kennard, P.Erhard, H.F.Filipovich, R.J.Socol//Arch Environ Health.-1986.-V.41.-P.287-291.
- Sandhir, R., Gill, K. D. Effect of lead on lipid peroxidation in liver of rats/ R.Sandhir, K. D.Gill // Biol. Trace Elem. Res.-1995.- V.48.- P. 91-97.
- [Referat] Landri, R., Bonnard, R. Ducharme, G. Assessment of the Output Distribution in the IEUBK Model with Monte Carlo Simulation/R.Landri, R.Bonnard, G.Ducharme.- Epidemiology: November 2006. - Vol.17 - Issue 6 - P.S472
- Foster, W.G. Reproductive Toxicity of chronic lead-exposure in the female cynomolgus monkeys / W.G. Foster // J Reprod Toxicol. -1992.-V.16.- P.123-131.
- Routh, D.K., Mushak,P., Boone, L. A new syndrome of elevated blood lead and microcephaly/ D.K.Routh, P. Mushak, L.Boone // J Pediatr Psychol.-1979.-V.4.-P.67-76.
- Gulson, B.L., Mizon, K.J., Palmer, J.M., Patison, N., Law, A.J., Korsch, M.J., Mahaffey K.R., Donnelly J.B. Longitudinal study of daily intake and excretion of lead in newly born infants/B.L.Gulson, K.J.Mizon, J.M.Palmer, N.Patison, AJ.Law, M.J.Korsch, K.R.Mahaffey, J.B.Donnelly // Environ Res.-2001.- V.85.- №3.- P.232-245.
- O'Flaherty E.J. Physiologically based models for bone-seeking elements. V: Lead absorption and disposition in childhood/ E.J. O'Flaherty // Toxicol Appl Pharm.-1995.- V.131.- P.297-308.