Method for deriving lead safety value based on AAL model of Monte Carlo simulation

Abstract

The invention discloses a method for deriving a lead safety value based on an AAL model of Monte Carlo simulation, and relates to the technical field of data derivation, and the method is technically characterized by comprising the following steps: S1, collecting data and carrying out statistical analysis; s2, adjusting an AAL model: establishing the AAL model through R software, inputting or modifying environmental medium lead concentration, exposure parameters and pharmacokinetic parameters in Excel, setting a lead exposure mode and operating a program, and simulating absorption, distribution, metabolism and excretion of lead by using a differential equation; s3, calculating an external lead intake value through a bisection method; and S4, selecting an uncertain coefficient, and calculating a health guidance value. According to the method, reference can be provided for dietary lead exposure risk assessment, in addition, the AAL model of Monte Carlo simulation can predict the blood lead level of each age according to external environment lead exposure of people, and therefore risk management personnel can utilize the model to carry out biological monitoring and take prevention and control measures according to monitoring data.

Description

Technical Field

[0001] The present invention relates to the technical field of data derivation, and in particular to a method for deriving a lead safety value based on an AAL model of Monte Carlo simulation. Background Art

[0002] As evidence of the toxicity of low lead exposure increases, re-updating the health-based guidance values ​​for lead has become an urgent issue.

[0003] After reviewing the epidemiological data on lead neurotoxicity published after 2010, the results showed that lead exposure is mainly related to attention deficit / hyperactivity disorder (ADHD), reduced IQ, and neurodevelopmental disorders in children. Reduced IQ is determined to be the most sensitive health hazard effect, and the corresponding blood lead threshold is 0.9μg / dL. The AAL model modified based on the monitoring data of lead exposure in the Chinese population combined with Monte Carlo simulation and dichotomy reverse calculation of the blood lead threshold of 0.9μg / dL corresponds to an external lead intake of 7.33μg / day, and finally the daily tolerable intake is calculated to be 0.12μg / day.

[0004] Based on the existing epidemiological research evidence, the blood lead threshold for lead is determined to be 0.9 μg / dL, and the tolerable daily intake of lead for the Chinese population is deduced to be 0.12 μg / day. This value can provide a reference for lead risk assessment and formulation of lead poisoning prevention and control measures, but in actual work, the dietary lead intake level of the Chinese population should be considered to make appropriate adjustments. Summary of the invention

[0005] The purpose of the present invention is to solve the above problems and provide a method for deriving lead safety value based on the AAL model of Monte Carlo simulation.

[0006] In order to achieve the above object, the technical solution of the present invention is as follows:

[0007] The present invention provides a method for deriving a lead safety value based on an AAL model of Monte Carlo simulation, comprising the following steps:

[0008] Step S1, collecting data and conducting statistical analysis;

[0009] Step S2, adjusting the AAL model: establishing the AAL model through R software, inputting or modifying the environmental medium lead concentration, exposure parameters and pharmacokinetic parameters in Excel, setting the lead exposure mode and running the program, and using differential equations to simulate the absorption, distribution, metabolism and excretion of lead;

[0010] Step S3, calculating the lead external intake value by dichotomy;

[0011] Step S4: Select the uncertainty coefficient and calculate the health guidance value.

[0012] The present invention is further configured as follows: the sub-steps of step S1 are:

[0013] Step S11, literature search;

[0014] Step S12, data extraction: extracting basic information of the literature, basic characteristics of the research subjects, analysis results, and health hazard effect values ​​or no health hazard effect values ​​in the literature;

[0015] Step S13, Bias risk assessment and meta-analysis of literature: After data extraction, the data were grouped according to the health hazard effects, and qualitative conclusions were drawn based on the number of studies with statistically significant BLL-related effects and the number of studies without statistical significance, and further bias assessment was conducted to evaluate its impact on the overall results and conclusions of the study;

[0016] Step S14, determine the most sensitive health hazard effect and the lowest blood lead level: select the study with the lowest blood lead level from the higher quality studies, determine the lowest health hazard effect value for each health hazard effect endpoint, then compare the lowest health hazard effect values ​​of different effect endpoints, and select the lowest value as the blood lead threshold and the corresponding effect endpoint as the most sensitive health hazard effect.

[0017] The present invention is further configured as follows: the sub-steps of step S2 are:

[0018] Step S21, sensitivity analysis: Use R language to perform sensitivity analysis on the kinetic parameters of the AAL model, and use the parameters with sensitivity higher than 0.1 as the parameters for subsequent Monte Carlo sampling; calculate the sensitivity coefficient according to the formula,

[0019]

[0020] Among them, x is the default reference value of the target parameter; △ is the range of change of the target parameter. Referring to previous studies, △ is set to 1% here; Φ is other parameters in the model, and its value is the default reference value; f is the function of the AAL model; parameters with a sensitivity coefficient greater than 0.1 are considered sensitive parameters, indicating that a 1% change in the parameter will have a 0.1% impact on the predicted blood lead;

[0021] Step S22, localization of parameters of the AAL model: In order to quantify the heterogeneity of food lead concentration and daily food intake within the population, the calculation formula of daily dietary lead intake is calculated with reference to the IEUBK model. The daily dietary lead intake is equal to the sum of the products of the daily intake of various types of food and the food lead concentration. The equation for daily dietary lead intake is as follows:

[0022]

[0023] Among them, INfoodtotal is the total dietary lead intake (μg / day) of each age group, food i represents the lead concentration of the ith dietary source (mg / kg), IR foodi is the intake of the ith food in each age group (g / day);

[0024] Step S23, constructing an AAL model based on Monte Carlo simulation: first determine the probability distribution of model parameters. The daily grain and vegetable intake of each age group obeys the lognormal distribution, and the daily water intake obeys the triangular distribution; the sensitive pharmacokinetic parameters are taken as default values, and their distribution is assumed to be normal distribution; the probability distribution of parameters is simulated by R software, and random numbers are generated. The above parameter values ​​are all within the 95% confidence interval. In the prediction of the AAL model based on Monte Carlo simulation, the environmental medium lead concentration, exposure parameters and sensitive parameters are randomly extracted from the defined distribution as input to simulate the blood lead values ​​of each age group. To ensure the stability of the results, the model based on Monte Carlo simulation is run 10,000 times each, and then the average blood lead value is calculated. The function used is as follows:

[0025] Parameters that obey the lognormal distribution: The function QLNORM (probability, meanlog, dlog) can generate lognormal distribution quantile values. First, we use the function QLNORM (probability, meanlog, sdlog) to calculate the upper and lower limits of the 95% confidence interval of the parameter to determine the range of random numbers. Then, we use RLNORM (n, meanlog, sdlog) to generate random numbers that conform to the lognormal distribution and determine whether the random number is within the 95% confidence interval. If it is outside the range, regenerate the random number, otherwise it will be used as the input value of AALM.

[0026] Parameters that obey the normal distribution: The return value of the function QNORM (probability, mean, sd) is the lower quantile of the given probability. Use the function QNORM (probability, mean, sd) to calculate the upper and lower limits of the 95% confidence interval of the parameter to determine the range of random numbers. Then use RNORM (n, mean, sd) to generate random numbers that conform to the normal distribution. Determine whether the random number is within the 95% confidence interval. If it is outside the range, regenerate the random number, otherwise it will be used as the input value of AALM. Explain the handling of negative numbers;

[0027] Parameters that obey triangular probability distribution: The return value of the function QTR (probability, min, max, mode) in the EnvStats package in the R language software is the quantile under a given probability. The function QTRI (probability, min, max, mode) is used to calculate the upper and lower limits of the 95% confidence interval of the parameter to determine the range of random numbers. The function RTR (In, min, max, mode) in the EnvStats package in the R language software can be called to generate random numbers that conform to the triangular distribution. It is determined whether the random number is within the 95% confidence interval. If it is outside the range, a new random number is generated, otherwise it will be used as the input value of AALM.

[0028] In order to improve the operation speed of the model, the foreach(itime, inorder, export, options.snow…) method in the foreach and doSNOW packages of the R language software is called to realize parallel computing.

[0029] The present invention is further configured as follows: the sub-steps of step S3 are:

[0030] Step S31: The model runs with the default external exposure setting, and it is observed that the maximum value of the 99% confidence interval of blood lead in a certain age group is higher than the target threshold; therefore, the total external exposure of this age group is taken as the target, and the food intake is reduced by the same amount to reduce the external exposure lead concentration to half;

[0031] Step S32: input the blood lead value calculated by the model;

[0032] Step S33: when the blood lead value obtained by external exposure is lower than the blood lead threshold, the external exposure lead concentration is set as the lower value, the original external exposure lead concentration is set as the upper value, and the middle value between the two is obtained;

[0033] Step S34: Adjust the food or water intake value so that the external exposure lead concentration is equal to the intermediate value, run the model to obtain the blood lead concentration under the external exposure setting, compare the blood lead concentration with the target threshold, if it is greater than the target threshold, use this value as the upper value, and the external exposure obtained in step S32 as the lower value; if it is less than the target threshold, use this value as the lower value, repeat this operation to make the result close to the target threshold.

[0034] The present invention is further configured as follows: in step S4, the default value of each uncertainty coefficient is set to 10, and after the uncertainty coefficient is selected, the lead external intake value corresponding to the blood lead threshold determined by the system review is substituted into the formula to calculate the health guidance value, and the calculation formula is as follows:

[0035] Health guidance value = lead intake value / uncertainty coefficient.

[0036] Compared with the existing technology, the beneficial effects of this solution are as follows: an AAL model based on Monte Carlo simulation is established, and the exposure parameters are adjusted to the data of the Chinese population, and the TDI of lead is derived to be 0.12μg / day or PTWI is 0.86μg / week, which can provide a reference for the risk assessment of dietary lead exposure. In addition, the AAL model based on Monte Carlo simulation can predict the blood lead level of people of all ages based on the external environmental lead exposure of the population, so risk managers can use the model for biological monitoring and take prevention and control measures based on the monitoring data. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a technical roadmap in the embodiment of the present invention;

[0038] Figure 2 This is a flow chart of literature screening in an embodiment of the present invention;

[0039] Figure 3 is a forest plot of lead and ADHD in an embodiment of the present invention;

[0040] Figure 4 is a funnel plot of lead and ADHD in an embodiment of the present invention;

[0041] Figure 5 is a forest plot of lead and reduced IQ in an embodiment of the present invention;

[0042] Figure 6 is a funnel plot of lead and IQ reduction in an embodiment of the present invention;

[0043] Figure 7 The exposure parameters in the embodiment of the present invention are compared with the blood lead distribution predicted by the national standard and the poisoning standard;

[0044] Figure 8 It is a comparison of the predicted blood lead of the AAL model in the embodiment of the present invention and the AAL model of Monte Carlo simulation;

[0045] Fig. 9 It is the blood lead distribution predicted by the AAL model of Monte Carlo simulation in the embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solution of the present invention will be further described in detail below in conjunction with the embodiments of the present invention and the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0047] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will be described in detail below in conjunction with the embodiments.

[0048] Example:

[0049] 1.1 Literature review results: Identification of the most sensitive health effects of lead and the associated minimum blood lead levels

[0050] A total of 374 publicly published articles were obtained through literature search, and 63 articles were initially included. The articles retrieved from the Chinese database were mainly cross-sectional studies, and the lead exposure level was mostly above 10 μg / dL, so Chinese articles were not included. After reading the full text, 27 articles were finally included. The inclusion and exclusion criteria are shown in Figure 2 The types of studies included cohort studies, case-control studies, and cross-sectional studies. A literature review determined that the main health hazards of lead exposure include attention deficit / hyperactivity disorder (ADHD), decreased IQ, and neurodevelopmental disorders.

[0051] Figure 3 The inclusion criteria of the literature in the study were as follows: similar research methods, cohort studies that the subjects were children without ADHD (<18 years old), case-control studies that the subjects were divided into two groups: the case group was children with ADHD and the control group was matched children without ADHD, the subjects of the cohort study did not have ADHD before follow-up, the exposure marker was blood lead, the exposure level was less than 10 μg / dL, the outcome determination method was consistent (ADHD scale or clinical diagnosis of ADHD), and the original data provided by the original literature were sufficient to calculate the combined effect value.

[0052] Figure 5 The inclusion criteria for the literature in the literature were as follows: similar research methods, research subjects were children (<18 years old), the outcome determination method was the Wechsler Intelligence Scale or the Raven Joint Test, the exposure marker was blood lead, the exposure level was less than 10μg / dL, and the original data provided by the original literature could be used to calculate the combined effect value.

[0053] In order to increase the credibility of the research evidence, a Meta-analysis was conducted on the literature that met the criteria. The Meta-analysis results of the correlation between lead and ADHD showed that the combined effect value was 1.352 and the confidence interval was (1.008, 1.815), indicating that an increase in children's blood lead levels can lead to an increased risk of ADHD. The funnel plot showed asymmetry, but the Egger's test result P>0.05, indicating that there was no publication bias. The results of the Meta-analysis showed that the research supported that lead exposure was associated with attention deficit / hyperactivity disorder (ADHD). The Meta-analysis results of the correlation between lead and reduced IQ showed that the combined effect value was -1.996 and the confidence interval was (-3.923, 0.068), indicating that increased blood lead levels were linearly correlated with reduced children's IQ. The funnel plot showed obvious asymmetry, and the Egger's test result P<0.05, indicating that there was a publication bias in the included studies, which may be caused by the small number of literature. The results of the Meta-analysis support that lead exposure is associated with ADHD and reduced IQ.

[0054] Some epidemiological studies have shown that lead exposure is statistically significantly associated with neurobehavioral problems, but the methods for measuring neurobehavioral problems are not uniform, so these studies are not included. Some other studies believe that lead exposure is related to other health hazards, including children's reduced academic performance, sensory impairment, autism, neurocognitive problems, and psychological problems, but the number of related literature is small and the evidence is still considered insufficient.

[0055] Several cohort studies have suggested that lead exposure is associated with reduced IQ. A large Swedish cohort study collected blood lead levels of 3,176 children aged 7–12 years from 1978 to 2007 and collected IQ test scores of these children in 2010. The results showed that when blood lead levels were in the range of 0.5–5 μg / dL, the average IQ loss was approximately 5 points. Nozomi Tatsuta et al. obtained data on umbilical cord blood, blood lead levels and IQ scores of 286 children from the TSCD birth cohort (a cohort study conducted in urban and coastal areas of Tohoku, Japan). Regression analysis was performed on 148 boys and 141 girls, and boys were divided into four groups according to blood lead quartiles: Q1 (0.33-0.57 μg / dL), Q2 (0.58-0.70 μg / dL), Q3 (0.71-0.89 μg / dL) and Q4 (0.90-1.71 μg / dL). The results showed that the blood lead level of boys was negatively correlated with IQ ((B = -16.362, p = 0.033). After adjusting for confounding factors between the groups, variance analysis showed that the WISC-IV scores of the groups were significantly different, and the multiple comparison test showed that there were significant differences between Q1-3 and Q4. Maitreyi et al. Mazumdar et al. reviewed a cohort study of a hospital in Boston and found that blood lead levels in children (6 months, 2 years, 4 years, and 10 years) were negatively correlated with children's IQ. However, the results of a study by Caroline M. Taylor et al. showed that prenatal B-Pb was positively correlated with IQ in 8-year-old girls. For every 1 μg / dl increase in prenatal blood lead, verbal IQ +0.71 points (p=0.021), performance IQ +0.57 points (p=0.099), and total IQ +0.73 points (p=0.017), but prenatal blood lead was affected by maternal factors, which was contrary to other research results. Although Homegnon AFBah et al. did not observe a statistical association between blood lead levels and children's IQ, the sensitivity of the Raven's Colored Progressive Matrices (RCPM) tool for measuring intelligence was not as high as the Wechsler Intelligence Scale, resulting in inconsistent research results with other studies, but several other cross-sectional studies support the association between lead and reduced IQ. Therefore, it can be considered that lead exposure is associated with reduced IQ.

[0056] By comparing the research quality and health hazard effect values ​​of all studies on the three effects (Table 1-1, Table 1-2, Table 1-3), the lowest blood lead level corresponding to reduced IQ in the study of Nozomi Tatsuta et al. was 0.9 μg / dL, and the research population belonged to the Asian population, with a smaller racial difference from the Chinese population. Therefore, the present invention selects reduced IQ as the most sensitive health hazard effect, and the corresponding blood lead level is 0.9 μg / dL (Table 1-2).

[0057] Table 1-1 Research on lead and ADHD

[0058]

[0059] Table 1-2 Research on the relationship between lead and reduced IQ

[0060]

[0061] Table 1-3 Studies on lead and neurodevelopmental disorders

[0063]

[0064] 1.2 AAL model parameter adjustment

[0065] According to the blood lead threshold corresponding to the sensitive health hazard effect re-determined by the literature review, combined with the blood lead distribution simulated by the Monte Carlo method and the obtained blood lead threshold, the external intake value of lead was calculated by the reverse dosimetry method through the AAL model. Finally, the appropriate uncertainty coefficient was selected to calculate the health guidance value of lead.

[0066] 1.2.1 Sensitivity analysis results

[0067] In order to detect whether the pharmacokinetic parameters in the model (transfer rate between tissues and organs, etc.) would affect the predicted blood lead, we conducted sensitivity analysis on the pharmacokinetic parameters. The parameters with sensitivity coefficients greater than 0.1 included rdiff (rate coefficient of transfer of exchangeable bone (cortical bone or trabecular bone) volume to bone surface or non-exchangeable bone volume), rcort (rate coefficient of transfer of lead from non-exchangeable cortical bone to diffusible plasma at time (t)), rcs2df (rate coefficient of transfer of lead from cortical bone surface to exchangeable cortical bone at time (t)), rrbc (rate coefficient of transfer of lead in red blood cells to diffusible plasma), tosof1 (the deposition fraction of lead from diffusible plasma to soft tissue compartment 1 at time (t), not scaled to bone surface deposition), and bldmot (lead concentration in maternal blood). These sensitive parameters were included in the model and set to default values.

[0068] Table 1-4 Sensitivity coefficients of ALLM pharmacokinetic parameters

[0069]

[0070] 1.2.2 Model exposure parameter adjustment

[0071] The exposure parameters of the AAL model are set for the American population, so we adjusted the model parameters based on the lead exposure data in China. The exposure parameters selected the intake of food and drinking water, and the intake was referenced to the "Handbook of Exposure Parameters for Chinese Populations" (see Table 2-1). Among them, food intake only includes grains and vegetables, because according to the third total diet, the main sources of dietary lead in each group of people are grains and vegetables, with a contribution rate of 72% to 80%. The lead concentration of food and drinking water refers to the national monitoring data, and the specific data is shown in Table 2-2.

[0072] Table 2-1 Input values ​​of exposure parameters of the AAL model based on Monte Carlo simulation

[0073]

[0074] Table 2-2 Input values ​​of external environmental lead concentration parameters of the AAL model based on Monte Carlo simulation

[0075]

[0076] 1.3 Model Validation

[0077] In order to verify the reliability of the Monte Carlo AAL model prediction, we input the national limit standards for food and drinking water as exposure concentration parameters into the model to predict the blood lead distribution of 0-90 years old. The results showed that the highest mean blood lead value predicted by the model was 8.80 μg / dL, 95% CI (6.26, 11.33) (see Figure 7 ), the standard of lead poisoning 10μg / dL is within the credible interval, indicating that although the model has prediction errors, the errors are within a reasonable range. Taking blood lead 10μg / dL as the standard, the lead concentrations of food and drinking water were reversely calculated. The results showed that the lead concentrations were: 0.00885mg / L for drinking water, 0.18mg / kg for cereals, and 0.09mg / kg for vegetables, which are close to the national lead limit standards, indicating that the Monte Carlo AAL model can be used for reverse calculation of lead intake by internal exposure (see Table 3).

[0078] Table 3 Comparison of lead concentrations predicted by the AAL model using Monte Carlo simulation and lead limit standards in China

[0079]

[0080] *The lead limit standards for cereals and vegetables refer to GB2762-2022 "National Food Safety Standard Limits of Contaminants in Food"; the lead limit standards for drinking water refer to GB5749-2022 "Sanitary Standards for Drinking Water"

[0081] Comparison of blood lead values ​​predicted by the AAL model and the AAL model using Monte Carlo simulation ( Figure 8) indicated that the mean blood lead values ​​of all age groups simulated by the AAL model were within the 99% confidence interval of blood lead predicted by the AAL model of Monte Carlo simulation, indicating that the AAL model of Monte Carlo simulation can quantify individual differences and is more suitable for group blood lead prediction.

[0082] 1.4 Calculation of health guidance values ​​for lead

[0083] 1.4.1 Calculation of lead external intake value

[0084] The 99% confidence interval of blood lead distribution for each age group predicted by the AAL model using Monte Carlo simulation is as follows Fig. 9 As shown, the external intake value corresponding to the bleeding lead threshold of 0.9 μg / dL was 7.34 μg / dL by reverse calculation using the dichotomy method.

[0085] Table 4 Changes of blood lead values ​​and intake values ​​by age

[0086]

[0087]

[0088] 1.4.2 Selection of uncertainty coefficient

[0089] In the process of deriving health guidance values, there are some uncertainties due to the toxicological evidence and the choice of derivation methods, so it is necessary to calculate the uncertainty factor (UF) based on the uncertainties. The uncertainties are as follows: due to the change in individual sensitivity (UFH) in the human population due to the extrapolation of health guidance values ​​(based on the lack of dose-response relationship data for susceptible populations, the default value of 10 times is used); uncertainty in model extrapolation (UFM) (UFM is 3 times); uncertainty factor (UFD) and correction factor (MF) of incomplete database: due to the lack of existing data, and taking into account UFD and susceptible populations, MF / UFD is set to 2. The uncertainty factor obtained according to the uncertainty factor calculation formula is 60.

[0090] Uncertainty coefficient calculation formula: UF = UFH * UFM * MF / UFD = 60

[0091] 1.4.3 Calculation of health guidance values ​​for lead

[0092] The external intake value of lead was calculated to be 7.34 μg / d by the AAL model based on Monte Carlo simulation. This value was used as the critical value of the external intake value. The tolerable daily intake (TDI) of lead was calculated to be 0.12 μg according to the health guidance value calculation formula, and the provisional tolerable weekly intake (PTWI) was 0.86 μg / week.

[0093] TDI = 7.34 / 60 = 0.12 μg / day

[0094] PTWI = 0.12 x 7 = 0.86 μg / week

[0095] Table 5 Lead limit standards in different countries

[0096]

[0097] 2. Discussion

[0098] The present invention reviews the human epidemiological studies related to lead and neurotoxicity published after 2010, determines that the neurotoxic effects of lead are mainly manifested in reduced IQ, ADHD, and neurodevelopmental disorders, compares the health hazard effect values ​​of these studies, determines that the most sensitive health hazard effect of lead is reduced IQ, and the corresponding blood lead threshold is 0.9 μg / dL, and uses the AAL model of Monte Carlo simulation to calculate the PTWI of lead to be 0.86 μg / week. Before using the AAL model of Monte Carlo simulation, the present invention uses the limit standard of lead in food and drinking water in China as the input value of the lead concentration parameter to simulate the blood lead distribution of various age groups, and compared with the blood lead poisoning standard of 10 μg / dL in my country, the mean blood lead values ​​of all age groups are lower than the standard, wherein the 95% confidence interval of the highest blood lead level includes 10 μg / dL, and the blood lead poisoning standard reversely verifies that the lead concentration parameter is close to the limit standard of food and drinking lead. In addition, the present invention also compares the AAL model of Monte Carlo simulation with the blood lead level predicted by the AAL model. The results show that the blood lead of all ages predicted by the AAL model is within the range of the blood lead of Monte Carlo simulation, reflecting that Monte Carlo simulation can quantify individual uncertainty and heterogeneity, so the Monte Carlo method is more suitable for the prediction of group blood lead.

[0099] Lead has been considered as one of the chemicals causing major public health problems internationally. Lead can enter the body through a variety of environmental media and accumulate in the body. Lead in the blood will eventually be deposited in the bones. Its half-life is as long as 20 to 30 years. Long-term low-lead exposure will have adverse effects on human health. Different authoritative organizations (such as JECFA, Health Canada, EPA, EFSA, etc.) have conducted scientific assessments on the toxicity of lead. They summarized the effects of lead on health, mainly including neurotoxicity and neurodegenerative effects, cardiovascular, renal, and reproductive toxicity. Among them, lower levels of lead exposure are most strongly associated with neurodevelopmental effects, especially reduced IQ scores and attention-related behaviors. IQ scores are the most widely used measurement endpoints for assessing developmental neurotoxicity of lead exposure. Infants and children are susceptible to lead exposure because they have higher levels of lead exposure and their gastrointestinal absorption and renal excretion efficiency are lower than those of adults. Therefore, choosing infants and children as susceptible populations and neurodevelopmental effects as key health effects is considered to be protective for the entire population. Therefore, based on these lead toxicity assessments, the present invention reviews the epidemiological research literature related to the neurotoxicity of lead published after 2010. The sensitive health hazard effects of lead include attention deficit / hyperactivity disorder (ADHD), reduced IQ, and neurodevelopmental disorders. Other studies have found that lead can cause violent behavior, psychological problems, decreased academic performance in children, sensory disorders, autism, and Parkinson's syndrome, but the evidence is insufficient.

[0100] All included literature was graded for evidence, and the weight of evidence and blood lead level were compared. It was determined that the blood lead threshold of 0.9 μg / dL was associated with decreased IQ, while the European Food Safety Authority calculated the lower limit of the benchmark dose of lead to be 1.2 μg / dL, and the latest reference limit of blood lead for children proposed by the US CDC was 3.5 μg / dL, indicating that the blood lead threshold of the present invention is of reference value for preventing lead poisoning. In addition, although the key research population corresponding to the most sensitive health hazard was determined to be the Japanese population, Japan and China are both in Asia, with small racial differences, so the determined blood lead threshold is also applicable to the Chinese population.

[0101] There are many exposure markers for lead, among which the concentration of lead in whole blood is the most widely used exposure biomarker in general population monitoring and is used as the main exposure indicator for investigating the association between lead exposure and health outcomes in epidemiological studies. BLL reflects recent comprehensive exposure and, to some extent, chronic exposure due to the transfer of lead between bone and blood. In contrast, bone lead is a biomarker of chronic exposure or total exposure due to its extremely long half-life. However, in practice, bone lead is not the preferred biomarker of lead exposure due to the limited availability of X-ray fluorimeter. Lead can also be measured in other biological matrices, such as breast milk, umbilical cord blood, urine, hair and teeth. However, most epidemiological and experimental studies tend to use blood measurements. In addition, the lead concentration in breast milk and umbilical cord blood is also a good biomarker of maternal and infant lead exposure. Therefore, the present invention uses blood lead and umbilical cord blood lead as markers of lead exposure.

[0102] Although the four Chinese Total Dietary Studies (TDS) conducted have shown that the dietary lead intake of the Chinese population has gradually decreased, there are still some people whose dietary intake has increased. According to the fourth total dietary study, the dietary lead risk assessment was conducted, and the results showed that the lead intake of 10 gender and age groups of Chinese residents was 48.7-116.7 μg / d. The lead exposure of children and adolescents aged 2-7 and 8-12 was the most pessimistic, and their mean and median exposure margin ratio (MOE) were less than 1 (0.1-0.3). Among the populations of various age groups, the population whose dietary lead intake was greater than the original PTWI was 9.7%-23.8%. Among them, the dietary lead intake of children and adolescents in the group exceeded the original PTWI by more than 20%, and the dietary lead intake of people over 13 years old also exceeded the original PTWI by more than 10%. The comparison of the TDI derived by the present invention with the dietary lead exposure of the Chinese population indicates that the government needs to take further management measures to limit lead exposure, and to reduce the lead content standards in industrial wastewater, waste gas, soil, drinking water and food by formulating policies and laws and regulations, and to eliminate lead exposure as much as possible. In addition, biological monitoring should be used to control the exposure of the population to lead, and the health guidance value can be updated based on existing modeling tools and possible new toxicological evidence. At the same time, all sectors of society need to pay attention to chronic lead exposure and blood lead levels in children, increase health education efforts, and cultivate good hygiene habits.

[0103] The blood lead threshold determined by the systematic review and the Monte Carlo AAL model, the obtained TDI is 0.12μg / day or PTWI is 0.86μg / week, which is lower than the lead intake limit value of the European Food Safety Authority and the US CDC. Assuming that the lead exposure comes from food and drinking water, this health guidance value is meaningful for explaining the risk of dietary exposure to lead. The present invention adjusts the input values ​​of the external environmental lead exposure parameters and concentration parameters of the AAL model based on the monitoring data of the Chinese population, so that the predicted health guidance values ​​are more suitable for the Chinese population. When formulating lead intake limits, each country refers to its own lead exposure data, so the standards formulated are different. The European Food Safety Authority (EFSA) calculated the reference point of lead content in the blood as 1.2μg / dL (0.5μg / kg body weight / day) based on the 95th percentile lower confidence limit of the benchmark dose (BMDL) for an additional risk of 1% of the full scale IQ of children. In 2017, the US FDA adopted the "temporary reference level" (IRL) to replace the "PTTDI" of lead level. Based on the no observed adverse neurodevelopmental effect level (NOAEL) as the basis of IRL, the reference value of 5μg / dL BLL (blood lead level) of the US CDC was used to obtain the IRL of 3μg / day for children and 12.5μg / day for women of childbearing age (WOCBA). Based on the qualitative evidence weight of published adult studies, it was proposed that the IRL of WOCBA is 12.5μg / d (corresponding to a BLL of 0.5μg / dL) for adults. my country has formulated lead limit standards for 22 kinds of food, including 0.2mg / kg for cereals and their products, 0.1mg / kg for fresh vegetables and fruits, and 1.0mg / kg for aquatic animals and their products. Therefore, the parameters of the model in the present invention are adjusted according to the lead exposure parameters of the domestic population to increase the credibility of the results.

[0104] Due to the heterogeneity and uncertainty of model parameters within the population, Monte Carlo simulation is introduced into the PBPK model to reduce the impact of individual differences on the prediction results. Barbara D.Beck et al. calibrated the Monte Carlo simulation of the O'Flaherty model using blood lead data of children in a certain area of ​​Utah, and accurately predicted the blood lead levels of children in two towns in Pennsylvania and Utah. Buqing Zhong et al. used Monte Carlo simulation combined with the IEUBK model to predict the children of the Chinese population. Compared with the biological monitoring data of preschool children, the results showed that the predicted blood lead 95% CI basically coincided with the observed blood lead distribution. The PBPK model using Monte Carlo simulation can obtain the blood lead level and error range of the population, which can improve the prediction effect of the blood lead level of the population.

[0105] The present invention assumes that lead exposure comes from food and drinking water. However, there are many ways of lead exposure, such as air, soil, dust, etc., which will affect the changes in blood lead, so the research results may have certain deviations. The Fourth Total Diet Study in China showed that the sources of dietary lead for all populations are mainly grains and vegetables. Therefore, the present invention mainly considers grains and vegetables as the main sources of dietary lead. But other foods also provide lead exposure, which may lead to deviations in the derived values. In addition, the toxicological evidence of lead is still being updated, and the health endpoints and blood lead thresholds may also change, so it is necessary to adjust the health guidance value of lead by continuously evaluating the toxicological evidence of lead to reduce the lead intake of the population as much as possible.

[0106] Secondly, the present invention directly quotes the default pharmacokinetic parameter values ​​recommended by the EPA. The default parameter values ​​are derived from studies based on European and American populations, which may also lead to certain differences between the predicted values ​​and the actual values. In the future, we need to further optimize the pharmacokinetic parameters, which will be beneficial for improving the accuracy of model prediction.

[0107] 3. Conclusion

[0108] This paper reviews the current epidemiological studies on lead neurotoxicity. The results show that lead exposure is mainly related to attention deficit / hyperactivity disorder (ADHD), reduced IQ, and neurodevelopmental disorders. The most sensitive health hazard effect is reduced IQ, and the corresponding blood lead threshold is 0.9 μg / dL.

[0109] The present invention establishes an AAL model based on Monte Carlo simulation, and adjusts the exposure parameters to the data of the Chinese population, and derives that the TDI of lead is 0.12 μg / day or the PTWI is 0.86 μg / week, which can provide a reference for the risk assessment of dietary lead exposure. In addition, the AAL model based on Monte Carlo simulation can predict the blood lead level of each age based on the external environmental lead exposure of the population, so risk managers can use the model for biological monitoring and take prevention and control measures based on the monitoring data.

[0110] The above specific embodiments are merely explanations of the present invention and are not limitations of the present invention. After reading this specification, those skilled in the art may make modifications to the embodiments without any creative contribution as needed. However, such modifications are protected by the patent law as long as they are within the scope of the claims of the present invention.

Claims

1. A method for deriving lead safety value based on the AAL model of Monte Carlo simulation, characterized in that: The following steps are involved: Step S1, collecting data and conducting statistical analysis; Step S2, adjusting the AAL model: establishing the AAL model through R software, inputting or modifying the environmental medium lead concentration, exposure parameters and pharmacokinetic parameters in Excel, setting the lead exposure mode and running the program, and using differential equations to simulate the absorption, distribution, metabolism and excretion of lead; Step S3, calculating the lead external intake value by dichotomy; Step S4: Select the uncertainty coefficient and calculate the health guidance value.

2. The method for deriving lead safety value based on the AAL model of Monte Carlo simulation as claimed in claim 1, characterized in that: The sub-steps of step S1 are: Step S11, literature search; Step S12, data extraction: extracting basic information of the literature, basic characteristics of the research subjects, analysis results, and health hazard effect values ​​or no health hazard effect values ​​in the literature; Step S13, Bias risk assessment and meta-analysis of literature: After data extraction, the data were grouped according to the health hazard effects, and qualitative conclusions were drawn based on the number of studies with statistically significant BLL-related effects and the number of studies without statistical significance, and further bias assessment was conducted to evaluate its impact on the overall results and conclusions of the study; Step S14, determine the most sensitive health hazard effect and the lowest blood lead level: select the study with the lowest blood lead level from the higher quality studies, determine the lowest health hazard effect value for each health hazard effect endpoint, then compare the lowest health hazard effect values ​​of different effect endpoints, and select the lowest value as the blood lead threshold and the corresponding effect endpoint as the most sensitive health hazard effect.

3. The method for deriving lead safety value based on the AAL model of Monte Carlo simulation as claimed in claim 1, characterized in that: The sub-steps of step S2 are: Step S21, sensitivity analysis: Use R language to perform sensitivity analysis on the kinetic parameters of the AAL model, and use the parameters with sensitivity higher than 0.1 as the parameters for subsequent Monte Carlo sampling; calculate the sensitivity coefficient according to the formula, Among them, x is the default reference value of the target parameter; △ is the range of change of the target parameter. Referring to previous studies, △ is set to 1% here; Φ is other parameters in the model, and its value is the default reference value; f is the function of the AAL model; parameters with a sensitivity coefficient greater than 0.1 are considered sensitive parameters, indicating that a 1% change in the parameter will have a 0.1% impact on the predicted blood lead; Step S22, localization of parameters of the AAL model: In order to quantify the heterogeneity of food lead concentration and daily food intake within the population, the calculation formula of daily dietary lead intake is calculated with reference to the IEUBK model. The daily dietary lead intake is equal to the sum of the products of the daily intake of various types of food and the food lead concentration. The equation for daily dietary lead intake is as follows: Among them, INfood total is the total dietary lead intake (μg / day) of each age group, food i represents the lead concentration of the ith dietary source (mg / kg), IR foodi is the intake of the ith food in each age group (g / day); Step S23, constructing an AAL model based on Monte Carlo simulation: first determine the probability distribution of model parameters, the daily grain and vegetable intake of each age group obeys the log-normal distribution, and the daily water intake obeys the triangular distribution; the sensitive pharmacokinetic parameters are taken as default values, and their distribution is assumed to be a normal distribution; the probability distribution of the parameters is simulated by R software, and random numbers are generated. The above parameter values ​​are all within the 95% confidence interval. In the prediction of the AAL model based on Monte Carlo simulation, the environmental medium lead concentration, exposure parameters and sensitive parameters are randomly extracted from the defined distribution as input to simulate the blood lead values ​​of each age group. To ensure the stability of the results, the model based on Monte Carlo simulation is run 10,000 times each, and then the average blood lead value is calculated. The function used for the parameters is as follows: Parameters that obey the lognormal distribution: The function QLNORM can generate lognormal distribution quantile values. First, we use the function QLNORM to calculate the upper and lower limits of the 95% confidence interval of the parameter to determine the range of random numbers, and then use RLNORM to generate random numbers that conform to the lognormal distribution. It is determined whether the random number is within the 95% confidence interval. If it is outside the range, a new random number is generated, otherwise it will be used as the input value of AALM; Parameters that obey the normal distribution: The return value of the function QNORM is the lower quantile of the given probability. The function QNORM is used to calculate the upper and lower limits of the 95% confidence interval of the parameter to determine the range of random numbers, and then RNORM is used to generate random numbers that conform to the normal distribution; determine whether the random number is within the 95% confidence interval. If it is outside the range, regenerate the random number, otherwise it will be used as the input value of AALM, explaining the handling of negative numbers; Parameters that obey triangular probability distribution: The return value of the function QTR in the EnvStats package in the R language software is the quantile under a given probability. The upper and lower limits of the 95% confidence interval of the parameter are calculated using the function QTRI to determine the range of random numbers. The function RTR in the EnvStats package in the R language software can be called to generate random numbers that conform to the triangular distribution, and determine whether the random number is within the 95% confidence interval. If it is outside the range, a new random number is generated, otherwise it will be used as the input value of AALM. In order to improve the operation speed of the model, the foreach method in the foreach and doSNOW packages of the R language software is called to implement parallel computing.

4. The method for deriving lead safety value based on the AAL model of Monte Carlo simulation as claimed in claim 1, characterized in that: The sub-steps of step S3 are: Step S31: The model runs with the default external exposure setting, and it is observed that the maximum value of the 99% confidence interval of blood lead in a certain age group is higher than the target threshold; therefore, the total external exposure of this age group is taken as the target, and the food intake is reduced by the same amount to reduce the external exposure lead concentration to half; Step S32: input the blood lead value calculated by the model; Step S33: when the blood lead value obtained by external exposure is lower than the blood lead threshold, the external exposure lead concentration is set as the lower value, the original external exposure lead concentration is set as the upper value, and the middle value between the two is obtained; Step S34: Adjust the food or water intake value so that the external exposure lead concentration is equal to the intermediate value, run the model to obtain the blood lead concentration under the external exposure setting, compare the blood lead concentration with the target threshold, if it is greater than the target threshold, use this value as the upper value, and the external exposure obtained in step S32 as the lower value; if it is less than the target threshold, use this value as the lower value, repeat this operation to make the result close to the target threshold.

5. The method for deriving lead safety value based on the AAL model of Monte Carlo simulation as claimed in claim 1, characterized in that: In step S4, the default value of each uncertainty coefficient is set to 10. After the uncertainty coefficient is selected, the lead external intake value corresponding to the blood lead threshold determined by the system review is substituted into the formula to calculate the health guidance value. The calculation formula is as follows: Health guidance value = lead intake value / uncertainty coefficient.

Images