| Deoxynivalenol(DON)is one of the most commonly found mycotoxins,produced by Fusarium species growing in cereal crops including wheat,barley,oats,rye,maize,etc.The frequent occurrence and high level of DON across the world make it an important substance to control for governments.Existing data shows that DON presents various toxicities on humans and animals,which can cause both acute poisoning and chronic damage to growth,immunity and reproductive functions.The extent of these toxic effects depends on the exposure.Therefore,accurately estimating the exposures in target population is the key to scientific risk assessment and rational decision-making.There are two methods for estimating human DON exposure: external exposure assessment and internal exposure assessment.The former is based on the analysis of human external environment(air,food,water,etc.),and is the mainstream method of chemical exposure assessment at present.For the mycotoxin DON,external exposure assessment is based on its content in food and human consumption of the contaminated foods.The reliability of external exposure assessment depends on the representativeness of food contamination data and the accuracy of food consumption data.It is usually more difficult to conduct external exposure assessment when there are diverse sources of exposure.The high heterogeneity of mycotoxin distribution in food further increases the difficulty of external exposure assessment.In view of this,in recent years,internal exposure assessments based on human biomarkers have developed rapidly,through analyzing the biological samples(such as urine or blood)for the target chemical and its metabolites to obtain the actual exposure burden.The method of internal exposure assessment avoids the uncertainties in food sampling and consumption survey,eliminates the influence of food processing and individual variation in metabolism,and provides a foundation for toxicity research by directly measuring the exposure on individual level.So far,some studies have conducted DON exposure assessments from wheat and/or maize in Chinese population,but only focused on limited areas.Few studies were conducted using internal exposure assessment method in China.The internal exposure assessments conducted abroad haven’t been compared with the external exposure.Therefore,in view of the unclear overall risk and geographical distribution of DON exposure in Chinese population,and lack of studies on the correlation between internal and external DON exposures,this study carried out a systemic risk assessment for DON and its derivatives at the national level in China,covering a comprehensive variety of food types,with insightful perspectives.Spatial analysis was used to clarify the geographical distribution of DON exposure in China for the first time,and the influence of meteorological factors were analyzed.Duplicate diet study and biomonitoring were carried out in the same population.Through in-depth analysis of dietary intake data and biomarker data,a model indicating the correlation of internal and external exposure suitable for morning urine was constructed and applied.Part I: Risk assessment and influence factors analysis of DON exposureOBJECTIVE: To analyze the health risks and influence factors of DON exposure in Chinese population,and to provide scientific basis for food safety management.METHODS: The main food sources of human exposure to DON and its derivatives were identified through systemetic review.After sorting out the existing contamination data in the national surveillance database and generating data for the missing food categories,the contents of DON,3-Ac-DON and 15-Ac-DON in 17644 food samples including wheat flour,maize meal,rice,oats,beer and complementary foods for infants and young children analyzed during 2010-2017 were used for exposure assessment.According to the toxicity characteristics,population sensitivity and regional differences of DON exposure,risk assessment was conducted from a multi-dimensional perspective.Acute and chronic exposures were assessed for different scenarios.Deterministic and probabilistic methods were used to estimate the overall health risks for different subgroups.Spatial analysis was applied to characterize the geographical distribution of DON contamination and exposure in China,and to identify outlier areas.The regression model was used to analyze the influence of meteorological factors on DON exposure in the typical high-contamination area,and the prediction model of commercial wheat flour exceeding the maximal level was established.The time series analysis of wheat flour contamination was carried out by using LSTM.RESULTS: Among the 6 food categories involved,the contamination levels of wheat flour,maize meal and cereal-based complementary foods were relatively high,and the mean contents of DON plus its derivatives were 247.6 μg/kg,161 μg/kg and 95.6 μg/kg,respectively.Oats,rice and beer were slightly contaminated by DON,with mean contents of 13.6 μg/kg,10.5 μg/kg and 3.0 μg/kg,respectively.Across the years,the probability of DON intake exceeding ARf D from wheat products was 1.9%,while the probability increased to 9.1% in years with severe contamination.Among infants and young children aged 0-3 years,the probability of exposure over ARf D was 1.6% and 2.6% in those who consume complementary food types of biscuit and noodles,respectively.Chronic exposure assessment showed that the mean exposure to DON and its derivatives in people over 3 years old was 0.82 μg/kg bw/d,and 30.7% of individuals were exposed to higher levels than the TDI.The mean exposure of infants and young children aged 0-3 years was 1.09 μg/kg bw/d,higher than TDI.There is no maximal level for DON in cereal-based complementary foods yet in China.Assuming that a maximal level of 200 μg/kg is set,the over-limit ratios for the current noodle samples and biscuit samples would be 33.7% and 32.5%,respectively.After excluding the over-limit samples,the average exposure of DON in infants and young children eating these two types of complementary foods could be halved.DON exposure presented significant regional differences,with northern inhabitants higher than the southern.Spatial analysis showed that the geographical distribution of DON contamination in foods was different from that of human DON exposure.As to DON exposure,Inner Mongolia,Gansu,Shaanxi and Shanxi formed an aggregate area of high exposure,while the aggregate area of high contamination for DON in wheat flour consisted of Anhui,Zhejiang,Jiangsu and Shanghai in the southeast.The time series analysis of wheat flour in the highly contaminated area showed that the variation of DON level along with the sample collection time could be recognized through machine learning,but the data were not enough to build a reliable prediction model.The influence analysis of meteorological factors on DON contents in wheat flour on the market in Jiangsu Province showed that its content level was related to the temperature,humidity and sunshine hours of the producing area in May.A logistic regression model was established for predicting the risk of DON exceeding the maximal level for wheat flour by using the minimum temperature and relative humidity of the producing area,with the AUC=0.874.CONCLUSIONS: The acute health risk of DON in Chinese population is generally low,but the probability of DON poisoning from consuming wheat products should be paid attention to in years with adverse weather.Chronic exposure to DON and its derivatives in the general population is at acceptable levels,but children and people with high consumption of cereals may be exposed to levels that exceed TDI and carry some risks.The contamination rate and level of cereal-based complementary foods in China are both relatively high at present,which put some infants and young children at risk of acute or chronic health damage.A maximal limit of 200 μg/kg for this kind of food would effectively protect the health of infants and young children.The contamination level of wheat flour on the market is affected by the temperature,humidity and sunshine hours of the producing area in May.The risk of DON exceeding the maximal level for wheat flour in the sampling area can be predicted according to the minimum temperature and relative humidity of the producing area in May.Part II: Internal and external exposure conversion model construction for DONOBJECTIVE: To establish an conversion model of DON internal and external exposure based on morning urine biomarkers,and to provide support for human biomonitoring and internal exposure assessment for DON.METHODS: Anhui Province was selected as the study site considering the spatial distribution characteristics of DON contamination and DON exposure.Stratified cluster sampling method was used and a total of 249 respondents aged 3 to 80 years were recruited.The 24 h duplicate meal study and biomonitoring were carried out.The consumption of cereal related foods was accurately weighed,and the corresponding food samples were collected.The first morning urine of 50 m L was collected the next day for each respondent.High performance liquid chromatography tandem mass spectrometry(HPLC-MS/MS)was used to determine the contents of DON,3-AcDON,15-Ac-DON and DON-3-G in food samples,as well as DON and DOM-1 in urine samples before and after β-glucuronidase hydrolysis.An internal and external exposure conversion model suitable for morning urine data was constructed based on the accurate dietary intakes measured by duplicate diet study and the biomonitoring data,through establishing and solving age-specific correlation variables.RESULTS: The excretion forms of DON in human urine include free and glucuronic acid conjugated DON and DOM-1,with DON the predominant form.DON-Glc A accounted for 80.7% of the total DON.After β-glucuronidase hydrolysis,the positive rate of DON was 99.2%,and the mean level was 116.5 μg/L.Urinary total DON concentrations varied significantly among age groups,with higher levels in children and adolescents.The average dietary intake of DON and its derivatives obtained by duplicate diet method was 2.74 μg/kg bw/d.Based on the biomarker levels in morning urine samples,the average exposure calculated using the existing model reported in literatures was 3.95 μg/kg bw/d,which was higher than the actual dietary intake.The correlation variable c constructed according to the internal and external exposure data measured in this study is strongly correlated with age.The reference values of variable c for 3-6 years old,7-12 years old,13-17 years old,18-59 years old and ≥60 years old are 0.54,0.91,0.87,1.79 and 2.46,respectively.The average exposure calculated by the model constructed in this study was 3.32 μg/kg bw/d,which was closer to the actual dietary intake.CONCLUSIONS: Free DON and DON-Glc A are the main forms of DON in human urine,and their total amount is suitable as a biomarker of DON exposure.The existing internal and external exposure model is not accurate enough to estimate dietary intake using morning urine data.The conversion model constructed in this study is more suitable for morning urine biomonitoring data.Part III: Application of internal and external exposure conversion model and health effect exploration for DONOBJECTIVE: To conduct internal exposure assessment in Chinese population,and to explore the health effects of low level of DON exposure.METHODS: According to the spatial distribution characteristics of DON exposure,Henan and Sichuan provinces were selected as the study sites to carry out human biomonitoring of DON.A total of 599 respondents aged 3-80 years were recruited by stratified cluster sampling with a male to female ratio of 1:1,including 301 from Henan Province and 298 from Sichuan Province.A household diet survey was conducted for 3 consecutive days,and the weight of all cereal-related foods consumed in each survey day was weighed.After each survey day,first morning urine samples were collected the next day,mixed in equal proportion for three days and hydrolyzed with β-glucuronidase.The concentration of DON was determined by HPLC-MS/MS method.Fasten blood samples of 4 m L were collected at the end of the three-day investigation,and the serum was extracted by centrifuge.The levels of immunoglobulin IgA,cytokines IL-2,IL-6,COX2,and protein kinases PKR and HCK on the DON toxicity pathway were determined by ELISA,and the correlation between them and DON exposure was analyzed.RESULTS: The total DON detection rates of the urine samples hydrolyzed by β-glucuronidase in Henan and Sichuan provinces were 100% and 91.9%,respectively,and the average levels were 52.8 μg/L and 13.0 μg/L,respectively.Urinary total DON levels differed significantly between age groups,with higher concentrations in children and adolescents than in adults.According to the morning urine biomarker levels the average dietary intake of DON in Henan Province and Sichuan Province was 1.42 μg/kg bw/d and 0.38 μg/kg bw/d respectively,calculated by using the internal and external exposure conversion model constructed in the second part of this study,and the difference was significant.Serum PKR and HCK were positively correlated with DON exposure levels,and IgA was negatively correlated with DON exposure levels.The serum levels of HCK,PKR,IL-6 and COX-2 in Henan province were significantly higher than those in Sichuan province,while the level of IgA in Henan province was lower than that in Sichuan province.CONCLUSIONS: The urine DON detection rate in the three surveyed areas was 92%-100%,indicating that DON dietary exposure was very common in Chinese population.There were significant regional differences in the levels of DON exposure markers,which were closely related to the consumption of wheat products in the local population.DON exposure may influence some biological indexes in human serum.Summary In this study,through the combination of external exposure assessment based on food analysis and internal exposure assessment based on biomonitoring,the risk characteristics of DON and its derivatives in Chinese population were systematically analyzed.Risks of cereal-based complementary foods for infants and young children are identified.The geographical characteristics of DON contamination and exposure are clarified.A meteorological prediction model of DON in wheat flour exceeding the maximal level is established.The data of DON exposure biomarkers in multiple regions of China are obtained,and an internal and external exposure conversion model suitable for morning urine biomonitoring data is constructed.This study also suggests the possible health effects of DON exposure at low levels.The results of this study provide a scientific basis for the risk management of DON and promote the optimization and application of biomonitoring and internal exposure assessment methods for DON. |
PDF Full Text Request