Cadmium Exposure Via Diet and Its Implication on the Derivation of Health-Based Soil Screening Values in China

ORIGINAL ARTICLE Cadmium exposure via diet and its implication on the derivation of health-based soil screening values in China

Mao-Sheng Zhong1,2,3, Lin Jiang1,2,3, Dan Han1,2,3, Tian-Xiang Xia1,2,3, Jue-Jun Yao1,2,3, Xiao-Yang Jia1,2,3 and Chao Peng1,2,3

The cadmium (Cd) intake rates via diet of adults from different regions in China were between 0.160 and 0.557 μg/(kg BW·day), which were less than the provisional tolerable monthly intake (0.833 μg/(kg BW·day)) issued by Food and Agriculture Organization/ World Health Organization in 2010, but higher than the one (0.365 μg/(kg BW·day)) issued by the European Food Safety Authority in 2011, to protect children, vegetarians and people living in heavily contaminated regions, and the intake rate of children (1.007 μg/(kg BW·day)) at the national scale was higher than the values recommended by the above institutes and those of adults. Vegetables were the critical contributors, followed by rice, flour, meats and aquatic products. Cd concentration in vegetable was the most sensitive factor in calculating the intake rate, followed by its contents in rice and aquatic products, and the intake rate of flour, indicating that more attention should be given to these parameters in future total diet surveys. When dietary exposure was incorporated, the derived national screening value of Cd under commercial scenario was reduced from 825 to 458 mg/kg, while the values of the north, south, Beijing and Shanghai were reduced to 627, 365, 693 and 489 mg/kg, respectively, indicating that the hazard would be underestimated if dietary exposure was not taken into account, especially for the south. The great variance between the screening values was due to the varied Cd intake rates, which indicated that deriving a screening value for each specific area based on its corresponding exposure characteristics was more appropriate. The national screening level for the residential scenario derived theoretically based on the dietary exposure characteristics of children was a negative value, meaning that the dietary intake rate was above the tolerable value. The method used in the United Kingdom to derive soil guideline values when non-soil exposure accounted for more than half of the maximum tolerable daily intake dose may be an appropriate estimate, but the exact ratio assigned to soil exposure should be assessed comprehensively based on a more sophisticated dietary exposure survey and the corresponding economic implications.

Journal of Exposure Science and Environmental Epidemiology (2015) 25, 433–442; doi:10.1038/jes.2015.5; published online 4 March 2015 Keywords: Cd; contaminated sites; dietary exposure; soil screening value

INTRODUCTION while maintaining the accuracy of results.10,11 The assessment Land contamination is a serious problem in China. In recent years, begins with the simple process of taking soil samples, analyzing many old and polluting industries are being shut down or them for relevant contaminants and comparing the results with relocated away from urban centers due to rapid urban growth in their corresponding soil screening values (SSVs), which are defined China. As a result, a large number of contaminated sites (often as the allowable chemical concentrations in environmental media referred to as “brownfields”) are emerging. These brownfields below which no additional regulatory attention is warranted. If the pose a dual problem: an environmental and health hazard in results are below the SSVs, the risk to human health is deemed China’s most densely populated areas, and as brownfields can not acceptable and no further investigation is needed. If the results be developed safely, they pose an obstacle to urban and econo- are high, a potential risk to human health is indicated and a more mic development. Therefore, a health-risk assessment must detailed site-specific assessment (often called tier two assessment) be carried out and remediation must be implemented before is needed, and some form of action or management may be the redevelopment of brownfield sites, in accordance with necessary. Therefore, SSVs are very important tools in identifying the provisional rules for the Environmental Management of contaminants of concern to which more resources must be Contaminated Sites and notes on strengthening pollution control allocated to investigate, and in some countries it is used as a tool throughout the decommissioning, relocation and redevelopment to screen out sites with high health risks for which more active of industrial enterprises issued by Ministry of Environmental management measures should be taken. Protection of China and the related administrative rules issued by Although they may be modified according to the analysis local environmental protection bureaus.1–6 detection limits of contaminants or background concentrations According to the technical guidelines of China, a tiered assess- found naturally, most SSVs are health-based and are derived using ment approach is recommended.7–9 This approach is also used a risk-assessment approach, combining toxic potency estimates, worldwide and is advantageous in saving investigation costs, acceptable target risks and hazards, and default conservative

1Beijing Municipal Research Institute of Environmental Protection, Beijing, China; 2National Engineering Research Centre of Urban Environmental Pollution Control, Beijing, China and 3Beijing Key Laboratory for Risk Modeling and Remediation of Contaminated Sites, Beijing, China. Correspondence: Professor Lin Jiang, Beijing Municipal Research Institute of Environmental Protection, No.59 Beiyingfangzhongjie, Xicheng District, Beijing, China. Tel.: +86 010 88362281. Fax: +86 010 88362281. E-mail: [email protected] Received 16 September 2014; revised 19 November 2014; accepted 23 November 2014; published online 4 March 2015 Cadmium exposure via diet Zhong et al 434 exposure values. For carcinogenic contaminants screening values also analyzed. The food consumption rate of the whole nation was directly are derived assuming the acceptable individual excess cancer risks obtained from the CNNHS survey report,28 while those of the sub-regions from exposure to contaminated soils to be 1 × 10 − 6, whereas for (north, south, Beijing and Shanghai) were calculated by weighting the non-carcinogenic pollutants the acceptable hazard indexes are population proportions of the involved provinces (Supplementary Tables S1–S3). For example, the food consumption of the north region was always set to 1, meaning that the exposure dose of the individual Σ contaminant from soil should not exceed its daily tolerable intake. obtained using the following formula: food consumption rate = R × P, where R is the food consumption rate of the provinces in the north region However, the daily tolerable intakes are the maximal allowable and P is the population proportion of the provinces in the north region.30 doses that receptors can take from all potential exposure Owing to the lack of original diet consumption data for children aged 2–7 pathways. Therefore, hazards from non-carcinogens may be years at the provincial level, only the dietary pattern at the national level underestimated by applying SSVs derived from the assumption was compiled based on the age-weighted average values of boys and girls that the acceptable hazard index is 1 from soil exposure if there (Supplementary Table S4). are some other important exposure pathways. Cadmium (Cd) is a toxic heavy metal that is ubiquitous in Cd in Different Foods environmental media, especially in soils. Concentrations of Cd in The database of Cd concentrations in different food types was compiled soils from smelting plants have been reported to be between 6.35 12–19 from the published literature since 2000 (Supplementary Tables S5). A total and 117.9 mg/kg, which are much higher than the national of 22,421 data points for Cd concentrations were collected, but a portion 20 5 2 background values of 0.02–0.33 mg/kg. More than 1.3 × 10 km of the data (3,726), which exceeded the standards for contaminants in of soils are contaminated by Cd in China and 1.46 × 108 kg of food,31 was removed with the aim of preventing the negative effect of agricultural products are polluted by Cd every year.21 Currently, substandard foods in quantifying the impact of Cd dietary exposure on the SSV of Cd is determined by its non-carcinogenic toxicity SSVs. Owing to the lack of original data, concentrations reported in the and derived by assuming the acceptable hazard index to be 1. literatures for different food types were averaged and then used for Cd However, apart from soil exposure, dietary intake is also a critical exposure. As a portion of the data featuring Cd concentrations in rice was pathway.22 It was reported that Cd intake from diet was 2.5 μg/kg obtained from the literature and based on dry weight, and because rice consumption rate is based on fresh weight, the concentrations were body weight per week in Shanghai,23 which contributed to 35.7% μ converted from dry weight into equivalent fresh-weight concentrations, of the provisional tolerable monthly intake (PTMI) of Cd (7 g/kg with the assumption that the water content of rice grain is 10%.32 body weight per week) issued by the World Health Organization (WHO) in 2010 (ref. 24). According to Huang et al.,25 Cd exposures through diet in three cities of the Zhejiang Province during 2009 Calculation of Cd Daily Intake Rate and 2010 were between 6.4 μg/day and 11.4 μg/day, contributing The estimated daily intake of Cd was calculated by multiplying daily food 15.6–42.6% to the PTMI. The result from a dietary survey carried consumption rates with corresponding Cd concentrations according to equation (1) (ref. 30): out in 2000 revealed that the national average dietary Cd X μ 26 exposure was 22.2 g/day, which was equivalent to 44.4% of EDIDCd ¼ ci ´ IRDi=BW ð1Þ the PTMI. Therefore, if the screening values derived under the μ current assumptions are applied in a tiered one risk assessment, EDIDCd is the estimated daily intake of Cd via diet, g/(kg BW day); ci is the many Cd-contaminated sites may be screened out as “clean sites” average Cd concentration in food i, mg/kg; IRDi is the daily intake rate of food i, g/day; and BW is body weight, kg. The reported date for each with negligible hazards and the health of future users will be food type in all literatures were averaged and used to calculate the threatened. corresponding EDIDCd at the national level, while for EDIDCd of the south, At present, there are still no national SSVs in China, although north, Beijing and Shanghai the statistics reported for the corresponding the national risk-assessment guidelines have come into effect. sub-region were used. Only exposure to the food studied in the present The screening values for non-carcinogens issued by the Beijing assessment was considered, which account for ~ 85% of the total daily diet Environmental Protection Bureau (BEPB) were derived assuming intake,28 while the contribution of some food types such as oil, salt, sugar an acceptable hazard index of 0.2 (ref. 27), which was in and pastry was excluded, as their contribution to dietary Cd exposure was 22,30 accordance with some states in the United States. The reason considered negligible. for the conservative assumptions is that receptors may be Owing to limited data on intake rate and concentration of vegetable in exposed to five contaminants doing harm to the same organ each sub-category (e.g. root and tuber vegetables, leafy vegetable, bulb vegetable etc), Cd exposure from vegetable consumption was calculated simultaneously on a contaminated site instead of dietary based on the averaged concentration of all reported data in literatures exposure. Therefore, the aim of this study was to quantify, at multiplied by the total vegetable intake rate recorded in CNNHS, and the the national and regional levels, the average dietary Cd intake by variance in concentration and intake rate of the individual sub-category compiling a database of Cd in Chinese foods and the dietary was neglected. For the meat category, the intake rate includes pork, beef, habits of people in different regions. Furthermore, the impact of mutton, chicken and duck, and the corresponding concentration was the dietary Cd exposure on SSVs under residential and commercial average value of all reported data on each sub-category. Edible offal scenarios was quantified. (e.g., kidney, liver, etc), always considered to be of high Cd concentration, was not considered in the present assessment, in order to avoid the uncertainty due to limited available concentration data and less intake MATERIALS AND METHODS frequency (4.7 g/day for adults while 3.2 g/day for children according to Jin28). Food Consumption Data Khan et al.33 and Gemma et al.34 reported that cooking practices did Trends in regional and national dietary habits were obtained from the 2002 not reduce the concentration of Cd in food. Therefore, the reported 28 China National Nutrition and Health Survey (CNNHS). This database is concentrations on raw food in the literatures were assumed to be China’s principal nutrition reference, containing the dietary patterns of equivalent to the ingested concentrations in the present study. 68,962 individuals from 31 provinces (excluding Hong Kong, Taiwan and Macau). In the present study, the food types are categorized into one of nine groups: rice and its products (hereinafter referred as rice), wheat flour Derivation of SSV and other grain products (hereinafter referred as flour), pulses, vegetables, SSVs for residential and commercial exposure scenarios were derived in fruits, meats, milk, eggs and aquatic products. Given the diversified dietary accordance with the national risk-assessment guidelines of China. Children habits of different regions, the food consumption patterns were analyzed younger than 7 years of age were the critical receptor for the residential at national, northern and southern levels. Classification of “north” and scenario, while adults were the critical receptor for the commercial “south” was achieved using a geographical divide running from the Huai scenario. Owing to the low volatility of Cd and the negligible hazard from River to the Qinling Mountains.29 Furthermore, the dietary patterns in the particulate inhalation pathway, only incidental oral ingestion and Beijing and Shanghai, the most developed and largest cities in China, were dermal adsorption were used as predominant exposure pathways in

Table 1. Deﬁnition and values of parameters

Parameters Deﬁnition Values Sources

Residential Commercial

RfD Reference dose, μg/(kg BW·day) 1 37 IRs Soil ingestion rate, mg/day 200 100 2 EF Exposure frequency, day/a 350 250 2 AR Skin exposure area, cm2 2448 2856 2 AH Soil adherence factor, mg/cm2 0.2 0.2 2 AF Dermal adsorption factor, unitless 0.001 2 BW Body weight, kg 15.9 56.8 2

deriving the SSV.11,35 Based on the above assumptions, equation (2) concentration, and the differences between pulses and meats (ref. 36) was used to derive the SSV of Cd. were not obvious. Cd concentrations in each food group in Beijing 1 and Shanghai were lower than the corresponding average values SSV ¼ ´ 103 ð2Þ 1 þ 1 of the whole nation, the south and the north. - ´ ´ - ´ ´ ðÞRfD EDIDCd BW 365 ðÞRfD EDIDCd BW 365 IRs ´ EF AR ´ AH ´ AF ´ EF The definition and values of the parameters in equation (2) are sum- Estimated Daily Cd Intake marized in Table 1. Based on the national nutrition and health survey in China, the average daily intake for Cd was calculated and the data are listed Sensitivity Analysis in Table 2. There was a large variation in Cd intake between regions. Adults in the south had the highest Cd intake rate, up to The sensitivity of the variables (ci and IRDi) used to calculate dietary Cd μ intake was assessed by calculating the parameters between each input and 0.557 g/(kg BW·day), whereas adults from the north had output during simulations and then evaluating each input’s contribution to 0.240 μg/(kg BW·day). For the adults of the whole nation, the Cd the output variance by normalizing to 100%. The Monte Carlo simulation intake rate was 0.444 μg/(kg BW·day), which was higher than that and sensitivity analysis were conducted using @Risk5.5 (Palisade, Palisade in the north but lower than that in the south. Furthermore, the Cd Corporation, New York, USA. 2010) software. intake rate of people in Beijing was much lower than that of people in Shanghai, but they were both lower than the national average value and the values of people in the north and south. RESULTS Compared with adults, the intake rate of children of the whole Chinese Dietary Patterns nation was more than two times higher due to lower body As shown in Figure 1, the diet intake rates and patterns of the weights. whole population and each sub-region are different. For adults, The breakdown of Cd contributions from different food groups the total diet intake rates of Beijing and Shanghai are 1066.1 and also varied between regions. In general, vegetables were the main 1150.2 g/day, respectively, which are higher than the average contributor of Cd intake, while the contributions from pulses, values of the whole population (874.3 g/day), the north (779 g/ eggs, milk and fruits (except Beijing) were negligible. For the total day) and the south (956.2 g/day). Vegetables, rice and flour are the population, vegetables contributed 27.8% to dietary Cd exposure, three major food groups, contributing 74.9% and 75.3% of the followed by rice and flour (Figure 2). For the population in the total diet for the whole population and people of the north, north, 34.6% was attributed to vegetables, which was comparable respectively. However, in the south, the dominant components are to that of the whole population. Flour and aquatic products vegetables, rice and meat, contributing 79.2%. The diet patterns in were also of great importance, accounting for 27.4% and 13.4%, Beijing and Shanghai, the most developed cities in China located respectively. In contrast, vegetables contributed 34.4% of the in the north and the south, respectively, are slightly different from total Cd for the south, followed by rice and aquatic products. those of the whole population, the north and the south. In Beijing, The main contributors were different between Beijing and in addition to vegetables, rice and flour, which are 60.5% of the Shanghai. For people in Beijing, the most important contributor total diet, milk and meat also account for a large percentage, up to was vegetables, followed by flour, rice and fruit, but vegetables, 21%. In Shanghai, 23.5% of the total diet is composed of meat and rice, meat and aquatic products were the most important aquatic products, and half is composed of vegetables and rice. contributors for people in Shanghai. The main contributors to The diet intake rate of children at the national level is 589.2 g/day the Cd intake rate of children were similar to those of adults due and the pattern is similar to that of adults. to similar dietary patterns.

Cd in Different Food Groups Screening Value A total of 18,695 data points of Cd in Chinese foods that were Screening values for commercial and residential scenarios below the standards of contaminants in food were compiled incorporating dietary exposure of Cd were derived and the results based on the published literature, and the average concentrations are listed in Table 3. are listed in Table 2. The highest mean concentration was found in When dietary Cd exposure was considered, its corresponding aquatic products throughout the country, even up to 0.093 mg/kg screening values would be reduced. The impact under the in the south, followed by the concentrations in rice, ﬂour, pulse, commercial scenario is greatest to the south, while it is negligible meat and vegetables, and the lowest mean concentration was in the north and Beijing. For the residential scenario, the screening found in fruits. Furthermore, the concentrations in rice, ﬂour, milk level derived theoretically with dietary exposure incorporated was and vegetables from the south were higher than those in the a negative value, meaning that the dietary Cd intake rate of north and the whole country, while Cd in eggs and fruits were children, the sensitive receptors under the scenario, exceeded the lower. Aquatic products from the south had a slightly lower maximum tolerable dose.

© 2015 Nature America, Inc. Journal of Exposure Science and Environmental Epidemiology (2015), 433 – 442 Cadmium exposure via diet Zhong et al 436 Sensitivity Analysis whole population, the Cd concentration in vegetables contributed As shown in Figure 3, Cd concentration in vegetables was the ~ 20% to the variance of Cd intake rate, followed by the ﬂour most sensitive variable in estimating dietary Cd intake. For the intake rate and Cd concentration in rice and ﬂour. For the people

Figure 1. Dietary patterns of people in different regions.

Table 2. Average concentrations of Cd and intake rates

Food type Cd concentration (mg/kg) Cd intake rate (μg/(kg BW·day))

NL NH SO BJ SH NL NH SH BJ SH NL-C

Rice 0.025 0.011 0.031 0.010 0.021 0.105 0.022 0.180 0.023 0.099 0.246 Flour 0.043 0.015 0.053 0.009 0.014 0.106 0.066 0.051 0.031 0.014 0.226 Pulses 0.024 0.022 0.025 0.002 0.013 0.007 0.005 0.009 0.001 0.005 0.015 Meat 0.026 0.023 0.027 0.003 0.018 0.036 0.019 0.051 0.006 0.044 0.089 Egg 0.008 0.011 0.006 0.003 0.005 0.003 0.006 0.002 0.002 0.004 0.010 Aquatic products 0.084 0.093 0.080 0.019 0.043 0.044 0.032 0.064 0.009 0.101 0.095 Milk 0.006 0.002 0.009 0.001 0.001 0.003 0.001 0.004 0.002 0.002 0.012 Vegetables 0.028 0.021 0.035 0.011 0.025 0.135 0.083 0.192 0.062 0.138 0.296 Fruit 0.006 0.007 0.005 0.013 0.001 0.005 0.006 0.004 0.024 0.001 0.018 Total / 0.444 (25.2) 0.240 (13.6) 0.557 (31.6) 0.160 (9.1) 0.408 (23.2) 1.007 (16.0) Abbreviations: BJ, Beijing; NH, north; NL, national level; NL-C, national level of children; SO, south; SH, Shanghai. Values in parentheses are total Cd in take rates in μg/day.

in the north, except for the Cd concentration in vegetables, flour concentrations in the other regions and the concentration at intake rate, Cd concentration in aquatic products, vegetable intake the national level are higher than the values reported in Finland rate and Cd concentration in rice were also very sensitive (0.006 mg/kg), France (0.007 mg/kg), Germany (0.011 ~ 0.116 mg/ variables. In contrast, in addition to the Cd concentration in kg), Ireland (0.03 mg/kg), Italy (0.0033 mg/kg), the Netherlands vegetables, the sensitivity of the Cd concentration in rice and (0.01 mg/kg), Norway (0.001 ~ 0.05 mg/kg), Portugal (0.025 mg/kg), aquatic products, and the vegetable intake rate were sensitive Switzerland (0.004 ~ 0.034 mg/kg), the United Kingdom (0.013 mg/ variables for the people in the south. For the people in Beijing and kg), Canada (0.0005 ~ 0.008 mg/kg), Japan (0.01 ~ 0.02 mg/kg) and Shanghai, in addition to the Cd concentration in vegetables, the Korea (0.015 ~ 0.021 mg/kg).38–43 For vegetables, the statistic Cd concentrations in aquatic products and meat were important. of Cd concentration of the south and the national are higher Given the sensitivity of the above variables to the estimated Cd than the one reported for Japan (0.02 mg/kg), Korea (0.02 mg/kg), intake via diet, the more accurately their statistic distributions are Norway (0.0069 ~ 0.012 mg/kg) and the United Kingdom (0.002 ~ characterized in future total dietary surveys, the less prone to 0.012 mg/kg),38–43 but lower than the one reported in Greece errors the derived screening values will be. (0.05 mg/kg) and Italy (0.06 mg/kg).38 Owing to the different dietary patterns and Cd concentrations in food groups, the dietary Cd intake rate of adults varies between DISCUSSION 0.160 and 0.557 μg/(kg BW·day) among different regions in China. Through the collection of literature data and based on the They are all below 0.833 μg/(kg BW·day), which is the PTMI rate national nutrition survey, the present study provides a framework issued by the Food and Agriculture Organization (FAO) and the for assessing dietary Cd intake and its impacts on the SSVs in WHO in 2010 (ref. 44). However, compared with 0.365 μg/(kg China for the first time. The published data revealed that the BW·day), the tolerable value issued to protect children, vegetarians Cd concentration of rice from the south, Shanghai and at the and people living in heavily contaminated regions by the Euro- national level are higher than the values reported by Canada pean Food Safety Authority (EFSA) in 2009 and 2011 (refs 45, 46), (0.002 ~ 0.009 mg/kg), Norway (0.013 mg/kg), Portugal (0.015 mg/ the average Cd intake rate at the national level, as well as – kg) and Switzerland (0.011 mg/kg),38 40 lower than the value the value of the people in the south and Shanghai are higher. reported in Japan (0.05 mg/kg) and Iran (0.4 mg/kg),41,42 and Meanwhile, the average value for the whole population in the similar in comparison with those values reported in Korea present study is somewhat higher than the values reported by (0.02 mg/kg), France (0.024 mg/kg), Finland (0.022 mg/kg), Ger- Gao et al.26 based on the total dietary surveys carried on in 1990 many (0.029 mg/kg), the Netherlands (0.029 mg/kg) and the (13.8 μg/day), 1992 (19.4 μg/day) and 2000 (22.2 μg/day), indicat- – United Kingdom.38 40,43 The differences in Cd concentrations in ing that the deterioration of food quality due to the conta- rice between the north and Beijing are negligible, and they are mination occurred during the industrial development process. both similar to the values reported in the previously mentioned Compared with the intake rates of people in Korea (14.41 μg/day), foreign countries. Cd concentration in flour from the south and its the United States (11.5 ~ 14.2 μg/day), Belgium (16.3 μg/day), concentration at the national level are higher than the values Denmark (16 μg/day), Finland (16 μg/ day), France (10.6 μg/day), reported in Japan, while for the north and Beijing, the concentra- Norway (15.8 μg/day), Germany (19.2 μg/day), the Netherlands tion in flour is similar to that of Japan.41 The Cd concentrations (19.3 μg/day), the UK (12.1 μg/day), Portugal (16.5 μg/day) and in meat from the north, south and Shanghai, and its concen- Canada (16.06 μg/day), the average intake rate of the adults in tration at the national level are similar to the values reported in China is higher. However, in contrast with Japan (25 ~ 30 μg/day), Japan (0.02 mg/kg) and Belgium (0.024 mg/kg); higher than Italy (20.2 μg/day), Ireland (25.1 μg/day) and New Zealand values reported in Canada (0.00022 ~ 0.004 mg/kg), Denmark (16.3 μg/day), the difference is negligible.38,43,47–51 (0.0022 mg/kg), Finland (0.001 ~ 0.004 mg/kg), France (0.004 mg/ The breakdown of Cd contributions from different food kg), Germany (0.007 ~ 0.016 mg/kg), Greece (0.0074 mg/kg), the groups for adults shows that vegetables are the most important United Kingdom (0.0046 mg/kg and Korea (0.012 mg/kg); and contributor, while the contributions from pulses, eggs, milk and lower than values reported in the Netherlands (0.05 mg/kg) and fruits (except Beijing) are negligible. In comparison, rice, vege- Norway (0.046 mg/kg).38–40,43 The Cd concentration in meat in tables, shellfish, mollusk, seaweed and meat account for 88.6% of Beijing is relatively low and is similar to the values reported in the Cd intake for people in Korea.43 Among them, the contribution Canada, Denmark, Finland, France, Germany, Greece and the from rice is the greatest (25.2%) (rice intake rate is also the United Kingdom. Except for aquatic products in Beijing, the Cd highest), followed by shellfish (14.5%) and mollusk (11.7%), which

© 2015 Nature America, Inc. Journal of Exposure Science and Environmental Epidemiology (2015), 433 – 442 Cadmium exposure via diet Zhong et al 438 also have high Cd concentrations, up to 0.5 and 0.2 mg/kg, However, in New Zealand it was reported that for the male adults respectively.43 For the whole population in Ireland, the dominant older than 25 years, the main contributor was oyster (26%), potato contributor is vegetables, up to 70%, followed by ﬂour (23.6%).48 (26%) and bread (10%), contributing 62% to the total intake.48

Figure 2. Cd intake from different food types.

Journal of Exposure Science and Environmental Epidemiology (2015), 433 – 442 © 2015 Nature America, Inc. Cadmium exposure via diet Zhong et al 439 Compared with the industrial screening value (1300 mg/kg) Table 3. Soil screening values of Cd (mg/kg) issued by New Zealand,11 the results in this study are much lower. Land use Dietary Dietary exposure considered The difference is because the dietary intake rate for people in exposure not New Zealand (18.2 μg/day) is lower than the values in this study, considered and because the exposure parameters used by New Zealand are different from the ones used in this study.11 However, in National North South Beijing Shanghai comparison with the soil guideline value of the United Kingdom Commercial 825 458 627 365 693 489 (230 mg/kg) for the commercial scenario, the values in the study Residential 83 − 1// / / are much higher because the tolerable daily intake (TDI) rate of Cd in the United Kingdom was set to 0.36 μg/(kg BW·day) and the allowed exposure dose from soil was set to half that of the TDI.58 Compared with the regional screening level (100 mg/kg) issued by the US EPA,59 the results in this study are much higher, because its acceptable hazard index is assumed to be 0.1, assuming that 10 In Italy, the key contributors for adults were flour (29%), toxic elements will do harm to the same organ simultaneously. vegetables (26%) and aquatic products (24%),52 while for adults For the residential scenario in which children were the sensitive in France, flour contributed 26%, followed by vegetables (19%), risk receptors, the screening value derived theoretically was a fish and aquatic products (13%) and potato (8%).52 negative value, meaning that the dietary Cd exposure of children The calculated dietary Cd intake rate of children at the national exceeded the maximum tolerable dose, and that the Cd screening level was 1.007 μg/(kg BW·day) in this study, which is much higher value of Cd should be set to 0 or its concentration in soil should be than the allowable doses issued by the WHO and EFSA. The critical remediated to 0 if no other risk control measures are taken. contributor was vegetables, followed by rice, flour, meat and However, this may not be rational as the background concentra- aquatic products, which was similar to that of the adults at the tion of Cd in China is 0.02–0.33 mg/kg20 and lowering the Cd national level. Compared with the daily intake rates reported for μ μ concentration in soil to 0 is not cost effective and is not practical children in Germany (0.737 g/kg BW·day), Australia (0.833 g/kg from a technical perspective. In this situation, except for reducing BW·day), Denmark (0.31 μg/kg BW·day), New Zealand (0.383 μg/kg μ 51,53–56 the dietary Cd exposure of children through improving food BW·day) and the Netherlands (0.32 g/kg BW·day), the quality or changing dietary patterns, the method used to derive value for children in China is much higher. The main contributors the Cd soil guideline value in the United Kingdom may be a good were different among these countries, which may due to the benchmark on which to set the allowable dose through soil different dietary patterns and Cd concentrations in each food exposure to be 50% of the total maximum allowable intake rates if group. In Germany, the main contributors were fruits and the non-soil exposure dose was more than a half,58 and the vegetables, accounting for 37.1% of the total intake, followed by 53 screening value for residential scenario will be 41 mg/kg if this cereals and bakery goods and beverages. In Australia, vegeta- method is applied. However, in China the exact value assigned to bles accounted for > 50%, followed by cereals and other grain- 54 the allowed exposure dose through soil should be assessed based foods, which was similar to that of China. In contrast with comprehensively based on our current food quality so that the Australia, cereal and cereal products were the predominant 55 exact Cd exposure status of children and the economic implica- contributor to the dietary exposure of children in Denmark, tions can be assessed thoroughly. accounting for nearly 50%, followed by vegetables. In New Zealand, potatoes were the main source of dietary Cd exposure, accounting for nearly 30%,51 followed by bread. In the Nether- CONCLUSIONS lands, wheat was the most important contributor to Cd exposure, 56 The average Cd intake rates for adults in different regions of China accounting for nearly 50%, followed by wheat. are all below the PTMI issued by the FAO/WHO, and range from The present results show that when the dietary exposure to Cd 0.160 μg/(kg BW·day) and 0.557 μg/(kg BW·day). However, the is considered, the SSVs will be reduced for the commercial mean values of the nation, the south and Shanghai exceed the scenario in which adults were considered as the sensitive risk values issued by the EFSA to protect children, vegetarians and receptors, especially for the south, where the intake rate was the people living in heavily contaminated regions. The average intake highest. Therefore, it can be concluded that the hazard will be rate of the whole nation is higher than the values of the north, underestimated and some potential contaminated sites would Beijing and Shanghai, but it is lower than the value of the south. be screened out as clean sites using screening values with- Vegetables are the critical contributor to Cd intake, followed by out considering dietary exposure. However, compared with the rice, flour, meats and aquatic products. The screening values for 57 screening value (150 mg/kg) issued by the BEPB, assuming the the commercial land-use scenario will be reduced by 16–55.6%, acceptable hazard index to be 0.2, the derived values in this study meaning that the hazard will be underestimated if dietary are much higher. Furthermore, due to the different dietary intake exposure is not considered, and deriving a screening value for rates among different regions, the difference among the derived each specific area based on its corresponding exposure char- values is very large. Under the same exposure scenario, the acteristics of Cd is more appropriate due to the large variation screening value for the south is the lowest, followed by the value among different regions. The Cd concentration in vegetables is for the whole population, Shanghai and the north, and the value the most sensitive factor in calculating the intake rate, followed by for Beijing is the highest. The derived value for the north is 1.4 Cd concentrations in rice and aquatic products, and the intake rate times higher than the national average value, while the one for of flour, indicating that more attention should be given to these the south is 1.3 times lower. The difference between the values of parameters in future total diet surveys. The screening value for the Beijing and the north (where Beijing is located) is negligible, while residential land-use scenario derived theoretically based on the the value of Shanghai is 1.3 times higher than the value of the whole dietary exposure status of children was a negative value, south where it is located. The above differences indicate that if the meaning that the dietary intake rate exceeded the maximum average Cd intake rate for the whole population is to be used to tolerable dose. The method used in the United Kingdom can be derive the national SSV, the hazard for the people in the north, applied to derive screening values for the residential scenario Beijing and Shanghai will be overestimated, but for people in based on comprehensive assessment of the exact dietary Cd other southern cities other than Shanghai, the hazard will be exposure status of children in China as well as the economic underestimated. impact of each decision.

Figure 3. Sensitivity analysis contribution to variance (%). AP, aquatic product; C, Cd concentration; EG, egg; FL, ﬂour; FR, fruit; IR, intake rate; MI, milk; ME, meat; PL, pulses; RI, rice; VE, vegetable.

Journal of Exposure Science and Environmental Epidemiology (2015), 433 – 442 © 2015 Nature America, Inc. Cadmium exposure via diet Zhong et al 441 CONFLICT OF INTEREST 23 Liu H, Wu CF, Lu Y, Fang YM, Gao WW, Xin ZH. Assessment on the dietary The authors declare no conflict of interest. exposure of lead and cadmium in Shanghai residents. Chinese J Food Hygiene 2011; 23:218–223 (in Chinese). 24 Satarug S, Garrett SH, Sens MA, Sens DA. Cadmium, environmental exposure, and ACKNOWLEDGEMENTS health outcomes. Environ Health Perspect 2010; 118:182–190. 25 Huang LC, Tang J, Zhang RH, Gu F, Fang YQ, Zhang HX et al. Exposure This work was financially supported by the Beijing Municipal Science and Technology assessment on dietary lead and cadmium of residents in three areas of Commission (D08040000360000) and an international cooperative project between Zhejiang Province from 2009 to 2010. Chinese J Prevent Med 2012; 46: the Beijing Municipal Environmental Protection Bureau and the Italian Ministry for 42–46 (in Chinese). Environment, Land and Sea. 26 Gao JQ, Li XW, Zhao JL. 2000 Chinese total diet study—the dietary lead and cadmium intakes. J Hygiene Res 2006; 35: 750–754 (in Chinese). REFERENCES 27 CRAES (Chinese Research Academy of Environmental Science). Technical Docu- ment on Deriving Soil Screening Values of Beijing. Beijing Environmental Protection 1 MEPPRC. (Ministry of Environmental Protection of the People’s Republic of China). Bureau: Beijing, 2011. Provisional Rules for the Environmental Management of Contaminated Sites (Draft) 28 Jin SG. The Tenth Nutrition and Health Survey Report about Chinese Residents: Beijing: MEPPRC 2009 (in Chinese). Dataset about Nutrition and Health in 2002. People’s Medical Publishing House: 2 MEPPRC. (Ministry of Environmental Protection of the People’s Republic of China) Beijing, 2008 (in Chinese). Technical Guideline for Risk Assessment of Contaminated Sites 2014, HJ25.3-2014 29 Han YF, Zhang JG, Zhang ZS. Regional Geography of China. Guangdong Higher Beijing: MEPPRC (in Chinese). Education Press: Guangzhou, China, 2009 (in Chinese). 3 JSEPD (Environmental Protection Department of Jiangsu Province) Notes on Pol- 30 Li G, Sun GX, Williams PN, Nunes L, Zhu YG. Inorganic arsenic in Chinese food and lution Prevention and Control on Industrial Contaminated Sites 2014 Nanjing: JSEPD its cancer risk. Environ Internat 2011; 37: 1219–1225. (in Chinese). 31 MoH (Ministry of Health of the People’s Republic of China). Contaminants Limit in 4 SYEPB (Shenyang Environmental Protection Bureau) Notes on Contaminated Sites f#ood 2012 GB 2762-2012 (in Chinese). Beijing: MoH Pollution Control and Remediation Shengyang: SYEPB 2007 (in Chinese). 32 Williams PN, Price AH, Raab A, Hossain SA, Feldmann J, Meharg AA. Variation in 5 CQMPG (Chongqing Municipal People’s Government) Notes on Contaminated arsenic speciation and concentration in paddy rice related to dietary exposure. Sites Remediation within Chongqing Municipality 2008 Chongqing: CQMPG (in Environ Sci Technol 2005; 39: 5531–5540. Chinese). 33 Khan SI, Ahmed AK, Mottashir Y, Mohammad R, Mahfuzar H, Samar K et al. Arsenic 6 PGZJ (The People’s Government of Zhejiang Province) Notes on Supervision on the and Cadmium in Food-chain in Bangladesh--An Exploratory Study. J Health Redevelopment of Industrial Sites 2013 Hangzhou: PGZJ (in Chinese). Population Nutr 2010; 28: 578–584. 7 Jiang L, Li LX, Huang HL, Wang JL, Wang Y, Zhang LN et al. Environmental site 34 Gemma P, Roser MC, Juan ML, Josel D. Effects of Various cooking processes on the assessment guideline 2009 DB11/T 656-2009 Beijing: Beijing Municipal Quality concentrations of arsenic, cadmium, mercury, and lead in foods. J Agric Food and Technology Supervision Bureau (in Chinese). Chem 2008; 56: 11262–11269. 8 Jiang L, Yu LF, Ding Q, Li FS Technical Research on Environmental Risk Management 35 Zhang HL, Li XP, Qi JY, Chen YT, Li BJ. Primary research on health risk assessment on Contaminated Sites of POPs Pesticides. China Environmental Science Press: of heavy metals in the surrounding soil and air of a municipal solid waste 32 – Beijing, China, 2012 (in Chinese). incinerator (MSWI), south China. J Agro Environ Sci 2013; : 1670 1676 9 Jiang L, Zhong MS, Liang J, Yao JJ, Xia TX, Fan YL. Application and benefit (in Chinese). assessment of tiered health risk assessment approach on site contaminated by 36 MfE (Ministry for Environment). Methodology for Deriving Soil Guideline Values benzene. Environ Sci 2013; 34:1034–1043 (in Chinese). Protective of Human Health 2010 Wellington: Ministry for Environment. 10 ASTM. Standard Guild for Risk-Based Reaction Action 2004 (E2081-00). 37 John AC, Richard LB, Thomas EM Software Guidance Manual-RBCA Tool Kit for Washington DC: American Society for Testing and Materials. Chemical Releases. GSI Environmental Inc: Houston, 2012. 11 MfE (Ministry for Environment). Guidelines for Assessing and Managing Petroleum 38 EC (European Commission). Assessment of the Dietary Exposure to Arsenic, Cadmium, Lead and Mercury of the Population of the EU Member States Contaminated Sites in New Zealand 2011. Wellington: Ministry for Environment. 12 Zhu PF, Ning P, Zeng XD, Wang HJ, Zhao R, He B. Soil pollution and heavy metals 2004. Brussels: European Commission. 39 Yuriko K, Testuo N, Nami K, Takamoto U, Kazaguki O. Cadmium concen hyper-accumulators around two sites contaminated by smelting. Safety Environ tration in current Japanese foods and beverages. J Occup Health 2002; 44: Engineer 2006; 13:48–51 (in Chinese). 240–247. 13 Zhang Y, Liu JJ, Liang CH, Chen XZ. Study on pollution status of heavy metal in 40 Health Canada. Trace Element Concentrations in 2007 Total Dietary Study Compo- soil from factory district of a former smelter in Shenyang City. J Anhui Agric Sci sites. Available at http://www.hc-sc.gc.ca/fn-an/surveill/total-diet/concentration/ 2008; 36: 6481–6483 (in Chinese). conc-food-alim_2007-eng.php. 2007 (accessed 7 March 2014). 14 Zeng YM, Ye ZR. Characteristics and assessment of soil contamination by heavy 41 Tsukahara T, Ezaki T, Moriguchi J, Furuki K, Shimbo S, Matsuda-Inoguchi N et al. metals around zinc smelting plant of Huixian. Sci Technol Inform 2010; 7: Rice as the most influential source of cadmium intake among general Japanese 346–347 (in Chinese). population. Sci Total Environ 2003; 305:41–51. 15 Du P. Distribution of Heavy Metals and Their Chemical Speciation in Soil Around a 42 Zazouli MA, Shokrazdeh M, Izanloo SF. Cadmium content in rice and its daily Pb-Zn Smelter. Chinese Research Academy of Environmental Science: Beijing, intake in Ghaemshahr region of Iran. Afr J Biotechnol 2008; 7: 3686–3689. China, 2007 (in Chinese). 43 Kim M, Wolt JD. Cadmium exposure in the South Korean population: implications 16 Jiang Y, Zhan XW, Sun LD, Xiao RM, Huang ZL. The formation distribution of Zn, of input assumptions for deterministic dietary assessment. Hum Ecol Risk Assess 38 Cu and Pb and Cd in the soil nearby the smelter. Guangdong Chem Ind 2011; : 2008; 14:835–850. – 110 112 (in Chinese). 44 FAO/WHO (Food and Agriculture Organization/World Health Organization). 17 Li XH, Tang ZL. Concentrations of cadmium, copper, lead, zinc in soil around Working Document for Information and Use in Discussions Related to Contaminants 5 smelting plant of Baiyin Region, Gansu Province. Geotech Invest Survey 2007; : and Toxins in the GSCTFF (General Standard for Contaminants and Toxins in Food – 23 27 (in Chinese). and Feed) 2011. Geneva: FAO/WHO. 18 Niu YL Study on Heavy Metal Dynamic Distribution Characteristic and its Pollution 45 EFSA (European Food Safety Authority). Scientific Opinion of the Panel on Con- Warnimg. Hunan University: Changsha, China, 2007 (in Chinese). taminants in the Food Chain on a Request from the European Commission on 19 Xu ZJ, Wu CH, Liu F, Qiu XY. Study on combined pollution character of heavy Cadmium in Food 2009. metals in soils around a typical Lead-Zinc smeltery. J Hum Univ Sci Technol 2007; 46 EFSA (European Food Safety Authority). Statement on Tolerable Weekly Intake for 22: 111–114 (in Chinese). Cadmium 2011. 20 Zhao XJ, Lu SJ, Xu RJ, Li BL, Wu GP, Wui FS. Soil heavy metal Cadmium standard 47 Egan SK, Bolger PM, Carringto CD. Update of US FDA’s Total Diet Study food list limit and range of background value research. Environ Sci 2014; 35: and diets. J Expos Sci Environ Epidemiol 2007; 17:573–582. 1491–1496 (in Chinese). 48 FSAI (Food safety authority of Ireland). Report on a Total Diet Study Carried Out by 21 Yu H, Wang JL, Fang W, Yuan JG, Yang ZY. Cadmium accumulation in different the Food Safety Authority of Ireland in the Period 2001–2005, Dublin:Food safety rice cultivars and screening for pollution-safe cultivars of rice. Sci Total Environ authority of Ireland 2011. 2006; 370: 302–309. 49 Health Canada. Dietary Intake of Contaminants and Other Chemicals for Different 22 Yuan XP, Wang J, Shang YE, Sun BG. Health risk assessment of cadmium via Age-Sex Groups of Canadians. Available at http://www.hc-sc.gc.ca/fn-an/surveil/ dietary intake by adults in China. J Sci Food Agric 2014; 94:373–380. total-diet/intake-apport/index-eng.php (accessed 7 March 2014).

© 2015 Nature America, Inc. Journal of Exposure Science and Environmental Epidemiology (2015), 433 – 442 Cadmium exposure via diet Zhong et al 442 50 Ikeda M, Ezaki T, Tsukahara T, Moriguchi J. Dietary cadmium intake in polluted 55 NFI (National Food Institute). Chemical Contaminants 2004–2011, 2013. and non-polluted areas in Japan in the past and in the present. Int Arch Occup 56 RIVM (The Netherlands National Institute of Public Health and the Environment). Environ Health 2004; 77: 227–234. Dietary Intake of Heavy Metals (Cadmium, Lead and Mercury) by the Dutch Popu- 51 MAF (Ministry of Agriculture and Forestry). 2009 New Zealand Total Diet Study- lation 2003. Agricultural Compound Residual, Selected Contaminant and Nutrient Elements 2011. 57 Zhou YY, Li FS, Li LX, Huang HL, Cao YZ, Yan ZG et al. Screening Levels for Soil Wellington: Ministry of agriculture and Forestry. Environmental Risk Assessment of Sites. Beijing Municipal Quality and Technology 52 EFSA (European Food Safety Authority). Cadmium Dietary Exposure in the Euro- Supervision Bureau: Beijing, 2011 DB11/T 811-2011. (in Chinese). pean Population 2012. 58 EA (Environmental Agency). Soil Guideline Values for Cadmium in Soil 2009. 53 DGHCP (Directorate-General Health and Consumer Protection). Assessment of the London: EA. Dietary Exposure to Arsenic, Cadmium, Lead and Mercury of the Population of the EU 59 US EPA. Regional Screening Level (RSL) Industrial Soil Table (TR = 1E-6, HQ = 0.1). Member States 2004. Available at http://www.epa.gov/reg3hwmd/risk/human/rb-concentration_table/ 54 FSANZ (Food Standards Australia New Zealand). The 23rd Australian Total Diet Generic_Tables/docs/indsoil_sl_table_01run_NOV2013.pdf. accessed 7 March Study 2011. 2014.

Supplementary Information accompanies the paper on the Journal of Exposure Science and Environmental Epidemiology website (http:// www.nature.com/jes)