Associations Of Exposure To Heavy Metals With The Indicator Of Liver Injury And Potential Mediating Effects Assessments In Community Residents

The liver is an organ with metabolic function.Liver disease is one of the major disease burdens in China.Risk factors for liver injury include viruses,alcohol,and drugs.Previous studies suggested that exposure to environmental pollutants such as heavy metals might also induce liver injury.Serum alanine transaminase(ALT)is the most commonly used sensitive indicator of liver injury.In epidemiological studies with large sample sizes,abnormally elevated ALT level that cannot be explained by viral hepatitis,hemochromatosis,or alcohol consumption is commonly used as a screening tool for non-alcoholic fatty liver disease.In the living environment,tobacco smoke,traffic exhaust,kitchen fumes,and food intake are common sources of heavy metal for the general population.After heavy metals enter the human body through ingestion,inhalation,and dermal contact,they are enriched and metabolized in the liver,which can directly damage the liver.Experimental studies have shown associations between exposure to certain heavy metals and liver injury,but the epidemiological evidence is sparse and the conclusions are inconsistent.In addition,most of the previous studies were cross-sectional design,and studies on the association between long-term exposure to heavy metal and liver injury in the general population was scarce.The mechanism of liver injury induced by heavy metals is still unclear.Studies have shown that systemic inflammation,oxidative stress,and the transforming growth factor-β1(TGF-β1)may play important roles in the occurrence and development of liver injury.However,whether systemic inflammation,oxidative stress,and TGF-β1 are involved in liver injury caused by heavy metals remain unclear.Based on these,this study was conducted among community adults based on the Wuhan-Zhuhai cohort established by this research group.First of all,information such as the demographic characteristics and living habits was collected through the questionnaire,and the morning urine samples were collected to determine the concentrations of heavy metals/metalloid which were must measured for soil pollution,including cadmium(Cd),arsenic(As),lead(Pb),chromium(Cr),copper(Cu),nickel(Ni),and zinc(Zn)(GB 156182018).The population distribution of urinary metals and the influence of common external exposure sources on urinary heavy metal levels were assessed.The serum ALT level in fasting blood samples was further measured,and elevated ALT was determined based on ALT level and questionnaire information(excluding viral hepatitis,hemochromatosis,or alcohol consumption).Serum ALT level and elevated ALT were used as indicators of liver injury to explore associations between urinary heavy metal levels and liver injury.Based on these,the potential roles of systemic inflammation,oxidative stress,and TGF-β1 in associations between heavy metal exposure and liver injury were explored through interaction and mediation analysis,to provide clues for further study of the mechanism of liver injury caused by heavy metals.The study mainly includes the following three parts:Part Ⅰ.The distribution and exposure sources of urinary heavy metals in community populationObjective:To describe the concentrations and distribution of Cd,As,Pb,Cr,Cu,Ni,and Zn of community residents,and to assess sources of internal exposure to heavy metals.Method:The subjects of this study were participants of the Wuhan-Zhuhai cohort,which was established from 2011-2012.The community residents aged 18 to 80,who did not suffer from serious diseases and had lived locally for 5 years or more were recruited in Wuhan and Zhuhai.The cohort was followed up every 3 years,and two follow-up visits have been completed.This part was based on the baseline population of the cohort.The concentrations of urinary Cd,As,Pb,Cr,Cu,Ni,and Zn were determined by inductively coupled plasma mass spectrometry and corrected by urinary creatinine.The correlation of heavy metals was evaluated using Spearman rank correlation coefficient.The distribution of urinary metals in different characteristics was assessed using the multiple linear regression model.Linear mixed models were conducted to assess the potential sources of heavy metal exposure,including smoking,cooking,traffic exposure,and dietary intake,and the relative contributions of exposure sources were assessed by the percentage change in R2 of the multiple linear regression model.Results:1.The average age of the 4097 subjects at baseline was 52.76 years old,and the average BMI was 23.98 kg/m2,of which 67.90%were female.The median concentrations of urinary Cd,As,Pb,Cr,Cu,Ni,and Zn were 0.10,2.68,0.24,0.13,0.69,0.19,and 26.65μg/(mmol urinary creatine),respectively.2.The Spearman correlation coefficients among urinary heavy metals ranged from 0.14 to 0.52(P<0.05).Compared with other research populations,the concentrations of urinary heavy metals in this study were at medium levels.3.In the multiple linear regression model,compared with younger subjects(≤55 years old),older subjects(>55 years old)had higher levels of urinary Cd,As,Cu,and Zn(P<0.05);compared with males,females had higher levels of urinary Cd,As,Pb,Cr,Cu,and Ni,but lower level of urinary Zn(P<0.05);compared with overweight subjects(BMI≥24 kg/m2),non-overweight subjects(BMI<24 kg/m2)had higher urinary Cd level(P<0.05);compared with non-smokers,smokers had higher levels of urinary Cd,Pb,Cu,and Zn(P<0.05).4.Smoking(yes/no,no smoking as reference),passive smoking(yes/no,no passive smoking as reference),cooking(yes/no,no cooking as reference),traffic exposure(hours/day),aquatic products intake(times/month),and pickled products intake(times/month)were positive associated with levels of some heavy metals.In order of contribution of exposure sources:Cd,smoking>cooking>passive smoking;As,aquatic products>passive smoking;Pb,smoking>pickled products;Cr,pickled products>traffic exposure;Cu,smoking>traffic exposure;Zn,smoking>traffic exposure>pickled products(all P<0.05).Conclusions:There were differences in the distribution of urinary heavy metals among participants with different ages,genders,BMI,and smoking status.Smoking,passive smoking,cooking,traffic exposure,aquatic products intake,and pickled products intake may be important sources of internal exposure to some heavy metals.Part Ⅱ.Associations of urinary heavy metal levels with the indicator liver injury in community populationObjective:To assess associations of urinary heavy metals levels with ALT level and elevated ALT in cross-sectional and longitudinal study.Method:In the cross-sectional study,4022 subjects were included in the analyses,excluding those without serum ALT concentration,on the basis of 4097 subjects in part Ⅰ.The criteria for elevated ALT including:(1)excluding viral hepatitis,hemochromatosis,or alcohol consumption;(2)ALT>30 U/L for male,or ALT>19 U/L for female.In a single metal model,linear mixed models and logistic regression models were used to evaluate associations of the levels of urinary heavy metals with serum ALT level and the risk of elevated ALT,respectively.The linear/nonlinear associations were evaluated by the restrictive cubic spline model.And sensitivity analyses were performed by excluding smokers,subjects whose urinary creatine concentrations exceeded the reference range,or patients with kidney disease,or using another criterion for abnormally elevated ALT(ALT>50 U/L for male,or ALT>40 U/L for female).Stratified analyses were conducted according to the basic characteristics.In mixed exposure models,weight quantile sum(WQS)regression model and Bayesian kernel machine regression(BKMR)model were used to evaluate the association of mixed exposure of urinary heavy metals with ALT level and the risk of elevated ALT.In the longitudinal study,4022 subjects(7061 observations)who participated in the baseline survey and follow-up survey were retained.Intraclass correlation coefficients(ICCs)were used to assess the temporal variability of urinary heavy metals at 3-year and 6years follow-up.The longitudinal associations between urinary heavy metal and ALT level were evaluated using linear mixed models.Cox regression models were used to assess the Hazard ratio(HR)between urinary metal(ICC≥0.40 at 3-year intervals)levels at baseline and the risk of elevated ALT.Results:1.The cross-sectional study included 4022 subjects(females 67.93%).After excluding 611 subjects with viral hepatitis,hemochromatosis,and alcohol consumption,1226 subjects were identified as elevated ALT(the prevalence was 35.94%).In a single metal model,when each 1-unit increased in log-transformed levels of urinary Cd,Pb,Cr,Cu,and Zn,the log-transformed level of serum ALT increased by 0.027(95%CI:0.004,0.049),0.058(0.038,0.078),0.040(0.020,0.060),0.042(0.020,0.063),and 0.036(0.011,0.059)-units,respectively,and the risk of elevated ALT increased by 15.9%(95%CI:3.8%,29.5%),24.5%(13.1%,37.1%),22.1%(11.0%,34.4%),22.9%(10.8%,36.5%),and 25.3%(11.5%,40.7%),respectively(all PFDR<0.05).Linear and positive associations were observed by restricted cubic models.And such positive associations did not change appreciably in sensitivity analyses.Stratified analyses showed that associations of urinary Cd,Pb,Cr,and Zn with ALT level were stronger in non-smokers,the association of urinary Pb with ALT level was stronger in overweight subjects(BMI≥24 kg/m2),and the association of urinary Cd with elevated ALT was stronger in drug users(all Pmodification<0.05).2.WQS regression model showed that WQS index of urinary heavy metals was positively associated with ALT level(β=0.075,95%CI:0.049,0.100),where Pb had the largest weight.WQS index was positively associated with the risk of elevated ALT(OR=1.243,95%CI:1.100,1.403),where Cr had the largest weight.BKMR model indicated that the ALT level and the risk of elevated ALT increased significantly with the increase of urinary heavy metal levels.There was a statistical interaction between urinary Cr and Zn in the association with the risk of elevated ALT(Pinternation<0.05).3.After a 3-year interval,the ICCs of urinary Cd,As,and Zn were 0.76,0.42,and 0.44,respectively,and other heavy metals were less than 0.40.After a 6-year interval,the ICC of urinary Cd was 0.68,and other heavy metals were less than 0.40.The stability of urinary Cd level was fair to good at intervals of 3 and 6 years.4.In the longitudinal study,each 1-unit increase in log-transformed levels of urinary Cu and Zn was associated with an additional rate of annual ALT increase for 0.62%(95%CI:0.08%,1.16%)and 0.74%(0.05%,1.44%),respectively.And such associations did not change appreciably in sensitivity analyses.In overweight subjects(BMI≥24 kg/m2),the risk of elevated ALT was significantly increased in the highest quartile of urinary As level,with the lowest quartile of urinary As level as reference(HR=1.467,95%CI:1.031,2.086).Conclusions:It is suggested that exposure to Cd,Pb,Cr,Cu,and Zn may be associated with liver injury,and exposure to As in overweight subjects may increase the risk of liver injury.Part Ⅲ.Potential effects assessments of systemic inflammation,oxidative stress,and TGF-β1 in associations between urinary heavy metals levels and liver injuryObjective:To assess the mediating effects of systemic inflammation,oxidative stress,and TGF-β1 in associations of urinary heavy metals with ALT level and the risk of elevated ALT.Method:This part was based on 4022 subjects at baseline.Plasma C-reactive protein(CRP),a biomarker of systemic inflammation,and TGF-β1 as well as urinary 8-iso-prostaglandin F2α(8-iso-PGF2α),a biomarker of lipid peroxidation,were determined by enzyme-linked immunosorbent assay.Urinary 8-hydroxy-deoxy-guanosine(8-OHdG),a biomarker of DNA oxidative damage,was determined by high-performance liquid chromatography with electrochemical detection.Linear mixed models were used to evaluate associations of urinary heavy metals with CRP,8-iso-PGF2α,8-OHdG,and TGF-β1,as well as CRP,8-isoPGF2α,8-OHdG,and TGF-β1 with serum ALT.Logistic regression models were used to evaluate associations of CRP,8-iso-PGF2α,8-OHdG,and TGF-β1 with elevated ALT.The interactions of urinary heavy metals and CRP,8-iso-PGF2α,8-OHdG,and TGF-β1 in associations with ALT level and elevated ALT were evaluated.Mediating models were used to assess the mediating effects of CRP,8-iso-PGF2α,8-OHdG,and TGF-β1 in associations of urinary metals with serum ALT level as well as the risk of elevated ALT.Results:1.Urinary Cd,Cu,and Zn levels were positively associated with plasma CRP(all P<0.05).Plasma CRP was positively associated with serum ALT level as well as the risk of elevated ALT(P<0.05).There was a statistical interaction between urinary Cd and plasma CRP in the association with the risk of elevated ALT(ORinteraction=1.130,Pinteraction<0.05).The mediating models showed that CRP partly mediated the associations of urinary Cd,Cu,and Zn with serum ALT as well as the risk of elevated ALT,with the proportion of mediation ranging from 2.67%to 6.74%(Pmediation<0.05).2.Urinary Cd,Pb,Cr,Cu,and Zn were positively associated with 8-iso-PGF2α as well as 8-OHdG(P<0.05).In addition,8-iso-PGF2α was positively associated with ALT level(P<0.05).There was a statistical interaction between urinary Zn and 8-OHdG in the association with the risk of elevated ALT(ORinteraction=1.092,Pinteraction<0.05).The mediating models showed that 8-iso-PGF2α rather than 8-OHdG partly mediated the associations of urinary Cd,Cr,Cu,and Zn with ALT level,with the mediating proportion of 22.95%,13.73%,15.53%,and 22.47%,respectively(Pmediation<0.05).3.Urinary Pb and Cr were negatively associated with TGF-β1,and urinary Zn was positively associated with TGF-β1(P<0.05).In addition,TGF-β1 was negatively associated with serum ALT level as well as the risk of elevated ALT(P<0.05).There was a statistical interaction between Cr and TGF-β1 in the association with ALT level(βinteraction=-0.024,Pinteraction<0.05).The mediating models showed that TGF-β1 partly mediated the associations of urinary Pb and Cr with serum ALT level as well as the risk of elevated ALT,with the proportion of mediation ranging from 3.92%to 11.84%(Pmediation<0.05).Conclusions:Systemic inflammation may be involved in the mechanism underlying Cd,Cu,and Zn exposure-associated liver injury;lipid peroxidation may be involved in the mechanism underlying Cd,Cr,Cu,and Zn exposure-associated liver injury;and TGF-β1 may be involved in the mechanism underlying Pb and Cr exposure-associated liver injury.