【作者】 赵新；

【导师】 徐顺清；

【作者基本信息】 华中科技大学 ， 公共卫生， 2020， 硕士

【摘要】 背景:中国快速的工业革命和城市化进程,加重了重金属对环境的污染,使得重金属污染物对人类健康的潜在影响受到越来越多的关注。越来越多的研究表明,宫内暴露于重金属污染物砷（As）、镉（Cd）、铅（Pb）易对胎儿宫内发育产生不利影响,甚至会对子代发育产生持续性影响。但以往多研究单个重金属暴露与体格发育的关系,对于多个金属混合暴露下与儿童体格发育的关系有待阐明。此外,宫内暴露于重金属已被证明会影响子代的免疫功能,以往的研究中对特异性免疫相关的淋巴细胞和单核细胞略有提及,但是缺乏对非特异性免疫的研究。本研究基于大型前瞻性出生队列,检测了4410名孕妇产前尿液中9种金属的浓度,包括钒（V）、铬（Cr）、锰（Mn）、钴（Co）、镍（Ni）、砷（As）、镉（Cd）、铅（Pb）、铊（Tl）,探讨产前金属暴露对3岁儿童体格发育及非特异性免疫指标的影响。第一部分:产前金属暴露对学龄前儿童体格发育的影响目的:探讨产前金属暴露对学龄前儿童体格发育的影响。方法:本研究基于前瞻性出生队列,研究人群选自2012年9月至2015年4月在武汉市招募并提供分娩尿液样本的4410名孕妇及其子女。以分娩前尿液（分娩前3天内采集）中金属的浓度作为儿童宫内暴露标志物,利用电感耦合等离子质谱（Inductively coupled plasma-mass spectrometry,ICP-MS）测定孕妇分娩尿样中的9种金属浓度,同时测定尿肌酐浓度,用于对尿金属浓度的校正。收集儿童3岁龄体检时身高和体重,计算体质指数（Body mass index,BMI）,并根据2006年世界卫生组织（World Health Organization,WHO）发布的儿童生长发育标准将3岁儿童的体格指标基于月龄进行校正,分别为体重-Zscore、身高-Zscore、BMI-Zs core。由于母亲分娩前尿液中各种金属浓度均呈左偏态分布,因此对金属浓度自然对数转换后再纳入模型。采用广义线性回归模型分析产前单金属暴露与3岁儿童标准化后的体格发育指标的关联,并按儿童性别分层分析,对金属暴露与性别交互项判断金属对儿童体格发育的影响是否具有性别差异。并进一步分析多金属混合暴露下与儿童体格发育的关系。结果:9种金属的检出率在94%以上,其中钴（Co）、砷（As）、镉（Cd）100%检出。分娩尿金属浓度经尿肌酐校正后的中位数分别为:砷（As）:28.50,镍（Ni）:4.93,铅（Pb）:3.44,铬（Cr）:1.89,锰（Mn）:1.59,钒（V）:1.41,钴（Co）:1.40,镉（Cd）:0.58,铊（Tl）:0.53μg/g。本研究人群产前尿金属水平高于澳大利亚、美国等发达国家孕妇;但低于缅甸、孟加拉国、中国贵屿等金属污染地区孕妇。产前尿金属水平调整可能的混杂因素并对模型中的P值进行FDR校正后,单金属暴露模型发现产前钴（Co）与儿童身高-Zscore正相关(%ch ange=6.00%,95%CI:2.43%,9.57%,P_-FDR=0.01),性别分层后在女孩中仍然正相关(%change=7.28%,95%CI:2.46%,12.10%,P_-FDR=0.03),但是男孩中关系没有统计学意义(P_-FDR=0.55);Tl发现与女孩BMI-Zscore正相关(%chang e=8.68%,95%CI:3.00%,14.35%,P_-FDR=0.02),但是男孩中没有统计学意义(P_-FDR=0.85)。多金属暴露模型与单金属暴露模型结果相似,钴（Co）与儿童身高-Zscore正相关(%change=5.68%,95%CI:1.69%,9.67%,P_-FDR=0.04);铊（T l）发现与女孩BMI-Zscore正相关(%change=9.81%,95%CI:3.00%,16.62%,P_{-F DR}=0.04),但是男孩中没有统计学意义(P_-FDR=0.63)。性别差异的P_-interaction均无统计学意义。结论:武汉市孕产妇广泛暴露于多种金属,砷（As）的暴露浓度最高,铊（Tl）的浓度最低,产前尿金属水平高于澳大利亚、美国等发达国家孕妇,但低于缅甸、孟加拉国、中国贵屿等金属污染地区。没有发现特别强的产前金属暴露与3岁儿童体格的关系,产前暴露钴（Co）可能与3岁儿童身高正相关;产前暴露于铊（Tl）可能与女孩BMI正相关。产前金属对儿童生长发育的影响,还需要进一步对出生后生长轨迹的影响进行研究。第二部分:产前金属暴露对学龄前儿童非特异性免疫指标的影响目的:探讨产前金属暴露对学龄前儿童非特异性免疫指标的影响。方法:本研究基于前瞻性出生队列,在2013年10月至2016年4月于武汉市招募并提供分娩尿液样本（分娩前3天内采集）的孕妇子代中,检测601名3岁儿童静脉血中各类白细胞计数。利用全自动生化仪检测儿童体检时静脉血中各类白细胞计数,产前金属暴露水平的获得与处理同第一部分,儿童各类白细胞计数均呈右偏态分布,经自然对数转化后纳入模型。采用广义线性回归模型分析产前金属暴露与学龄前儿童非特异性免疫水平的关联,并按儿童性别分层分析,金属与性别交互项判断金属对非特异性免疫指标的影响是否具有性别差异。另外将金属浓度按四分位数进行分层,以最低四分位数为参照组,将金属以分类变量纳入模型重复以上分析,并用每组的中位数重新赋值,重新赋值的变量以连续型变量的形式纳入模型,计算趋势性P_-trend,以此明确剂量反应关系。采用阳性错误发现率（False discovery rate,FDR）控制由多重假设检验产生的假阳性率。结果:金属作为连续型变量的模型中,校正混杂因素后,产前镉（Cd）暴露每增加一个自然对数单位,儿童淋巴细胞计数（LYMPH）降低14.00%(95%CI:-22.38%,-5.74%,P_-FDR=0.01)。性别分层后,产前钴（Co）暴露每增加一个自然对数单位,男孩淋巴细胞计数（LYMPH）降低14.85%(95%CI:-24.14%,-4.25%,P_-FDR=0.05),此负向关联在女孩中无统计学意义(P_-FDR=0.33),性别差异无统计学意义(P_-interaction=0.78);产前镉（Cd）暴露每增加一个自然对数单位,男孩淋巴细胞计数（LYMPH）降低15.32%(95%CI:-25.87%,-3.77%,P_-FDR=0.05),此负向关联在女孩中无统计学意义(P_-FDR=0.91),性别差异无统计学意义(P_{-in teraction}=0.78);产前钒（V）暴露每增加一个自然对数单位,男孩嗜酸性粒细胞计数（EOS）和嗜酸性粒细胞百分比（EOS%）分别降低23.05%(95%CI:-35.9%,-7.61%,P_-FDR=0.05)、19.71%(95%CI:-30.63%,-7.07%,P_-FDR=0.03),女孩中无统计学意义上的关联(P_-FDR=0.64、0.45),性别差异显著P_-interaction=0.072、0.027。金属作为分类变量的模型中,锰（Mn）第四组与第一组相比,嗜酸性粒细胞（EOS%）降低21.75%(95%CI:-32.29%,-9.55%,P_-FDR=0.01),第三组、第二组与第一组相比无统计学意义,但线性趋势显著(P_-trend=0.009),其他金属的暴露水平与各白细胞计数均无统计学意义上关联。多金属混合暴露下,产前锰（Mn）暴露每增加一个自然对数单位,男孩白细胞计数（WBC）增加7.08%(95%CI:2.35%,12.03%,P_-FDR=0.03),女孩中无统计学意义上的关联(P_-FDR=1.00);女孩单核细胞百分比（MONO%）随锰（Mn）暴露的增加而降低9.47%(95%CI:-14.91%,-3.68%,P_-FDR=0.02),而男孩中此关联无统计学意义(P_-FDR=0.87),以上关联均无性别差异(P_-interaction均>0.05)。结论:现有环境暴露水平下,产前暴露于钒（V）、钴（Co）、镉（Cd）可能与学龄前3岁男孩白血球计数减少有关,其中钒（V）与嗜酸性粒细胞计数（EOS）、钴（Co）和镉（Cd）与淋巴细胞计数（LYMPH）存在负向关联。钒（V）与嗜酸性粒细胞百分比（EOS%）的关联有性别差异。产前高水平的锰（Mn）可能与学龄前3岁儿童嗜酸性粒细胞百分比（EOS%）降低有关。多金属混合暴露下,产前锰（Mn）可能与男孩白细胞计数（WBC）增加,女孩单核细胞百分比（MONO%）降低有关。这些结果表明产前暴露于钒（V）、钴（Co）、镉（Cd）、锰（Mn）对学龄前儿童的非特异性免疫水平可能会产生影响,特别是免疫抑制。更多还原

【Abstract】 Background:China’s rapid industrial revolution and urbanization process have increased the pollution of heavy metals to the environment,and the potential impact of heavy metal pollutants on human health has received increasing attention.More and more studies have shown that intrauterine exposure to heavy metal pollutants arsenic（As）,cadmium（Cd）,and lead（Pb）is prone to adversely affect fetal intrauterine development,and may even have a sustained impact on the development of offspring.However,in the past many studies have studied the relationship between single heavy metal exposure and physical development,and the relationship between multiple metal exposure and children’s physical development remains to be elucidated.In addition,intrauterine exposure to heavy metals has been shown to affect the immune function of the offspring.In previous studies,specific immune-related lymphocytes and monocytes were slightly mentioned,but there is a lack of research on non-specific immunity.Based on a large prospective birth cohort,this study detected the concentrations of9 metals in the prenatal urine of 4,410 pregnant women,including vanadium（V）,chromium（Cr）,manganese（Mn）,cobalt（Co）,and nickel（Ni）,Arsenic（As）,cadmium（Cd）,lead（Pb）,thallium（Tl）,to explore the effects of prenatal metal exposure on the physical development and non-specific immune indicators of 3-year-old children.Part 1:The effect of prenatal metal exposure on physical development of preschool childrenObjective:To explore the effect of prenatal metal exposure on physical development of preschool children.Methods:This study is based on a prospective birth cohort.The study population was selected from 4,410 pregnant women and their children who were recruited in Wuhan from September 2012 to April 2015 and provided childbirth urine samples.The concentration of metals in urine before delivery（collected within 3 days before delivery）was used as a marker of intrauterine exposure in children,and 9 kinds of urine samples from pregnant women were measured by inductively coupled plasma mass spectrometry（ICP-MS）The concentration of metal,and the determination of urine creatinine concentration at the same time,is used to correct the concentration of urine metal.Collect the height and weight of children at the age of 3,and calculate the body mass index（BMI）.Based on the child growth and development standards issued by the World Health Organization（WHO）in 2006,the physical indicators of the 3-year-old children are Months of age were corrected for weight-Zscore,height-Zscore,and BMI-Zscore.Because the concentration of various metals in the mother’s urine before delivery was left-skewed,the natural logarithm conversion of the metal concentration was included in the model.A generalized linear regression model was used to analyze the association between prenatal single-metal exposure and 3-year-old children’s standardized physical development indicators,and stratified analysis according to children’s gender.The metal exposure and gender interaction terms were used to determine whether the effect of metal on children’s physical development was gender.difference.And further analyze the relationship between multi-metal exposure and children’s physical development.Results:The detection rate of 9 metals is more than 94%,of which 100%is detected for cobalt（Co）,arsenic（As）and cadmium（Cd）.The median values of urinary creatinine corrected for urinary creatinine were arsenic（As）:28.50,nickel（Ni）:4.93,lead（Pb）:3.44,chromium（Cr）:1.89,manganese（Mn）:1.59 Vanadium（V）:1.41,Cobalt（Co）:1.40,Cadmium（Cd）:0.58,Thallium（Tl）:0.53μg/g.The prenatal urine metal levels of this study population were higher than those of pregnant women in developed countries such as Australia and the United States;but lower than those of pregnant women in metal-contaminated areas such as Myanmar,Bangladesh,and Guiyu,China..After adjusting for possible confounding factors and adjusting the P value in the model by FDR,the single-metal exposure model found a positive correlation between prenatal cobalt（Co）and child height-Zscore(%change=6.00%,95%CI:2.43%,9.57%,P_-FDR=0.01),after sex stratification,it is still positively correlated among girls(%change=7.28%,95%CI:2.46%,12.10%,P_-FDR=0.03),but there is no statistical relationship among boys Academic significance(P_-FDR=0.55);Thallium（Tl）found positive correlation with girls BMI-Zscore（%change=8.68%,95%CI:3.00%,14.35%,P-FDR=0.02）,but there was no statistical significance among boys(P_-FDR=0.85).The results of the multi-metal exposure model and the single-metal exposure model are basically the same.Cobalt（Co）is positively correlated with the child height-Zscore(%change=5.68%,95%CI:1.69%,9.67%,P_-FDR=0.04);It was positively correlated with girls’BMI-Zscore(%change=9.81%,95%CI:3.00%,16.62%,P_-FDR=0.04),but there was no statistical significance among boys(P_-FDR=0.63).P_-interaction of gender differences was not statistically significant.Conclusion:Pregnant women in Wuhan are widely exposed to various metals,with the highest concentration of arsenic（As）and the lowest concentration of thallium（Tl）.Prenatal urine metal levels are higher than pregnant women in developed countries such as Australia and the United States,but lower than those in metal-contaminated areas such as Myanmar,Bangladesh,and Guiyu,China.No particularly strong relationship was found between prenatal metal exposure and the physical fitness of 3-year-old children.Prenatal exposure to cobalt（Co）may be positively correlated with the height of 3-year-old children;prenatal exposure to thallium（Tl）may be positively correlated with girl BMI.The influence of prenatal metals on the growth and development of children needs further research on the influence of postnatal growth trajectories.Part 2:The effect of prenatal metal exposure on non-specific immune indicators of preschool childrenObjective:To explore the effect of prenatal metal exposure on non-specific immune indexes of preschool children.Methods:This study is based on a prospective birth cohort.601 pregnant women who were recruited in Wuhan from October 2013 to April 2016 and provided childbirth urine samples（collected 3 days before childbirth）and their children had a medical record at 3 years of age 601 For mother and baby.Using automatic biochemical instrument to detect various types of venous blood leukocyte counts during physical examination in children.A generalized linear regression model was used to analyze the association between single-metal exposure in utero and non-specific immunity levels of preschool children,and stratified analysis according to children’s gender.The metal×gender interaction term was used to determine whether the effect of heavy metals on innate immune cells had gender differences.In addition,the metal concentration is stratified according to quartiles,with the lowest quartile as the reference group,the metal is included in the model as a categorical variable,and the above analysis is repeated,and the median value of each group is re-assigned,and the re-assigned variable is continuous The form of sexual variables is included in the model,and the trend Ptrend is calculated to clarify the dose-response relationship.A false discovery rate（FDR）is used to control the false positive rate generated by multiple hypothesis testing.Results:In the model of metal as a continuous variable,after adjusting for confounding factors,for every natural log unit of intrauterine cadmium（Cd）exposure,the lymphocyte count（LYMPH）of children decreased by 14.00%(95%CI:-22.38%,-5.74%,P_-FDR=0.01).After sex stratification,for every natural log unit increase in intrauterine cobalt（Co）exposure,the boy’s lymphocyte count（LYMPH）decreased by14.85%(95%CI:-24.14%,-4.25%,P_-FDR=0.05),This negative correlation was not statistically significant in girls(P_-FDR=0.33),and there was no statistically significant difference in gender（P-interaction=0.78）;for each additional natural log unit of intrauterine cadmium（Cd）exposure,boys’lymph The cell count（LYMPH）decreased by15.32%(95%CI:-25.87%,-3.77%,P_-FDR=0.05),this negative association was not statistically significant among girls(P_-FDR=0.91),and the gender difference was not Statistical significance(P_-interaction=0.78);for each additional natural logarithmic unit of intrauterine vanadium（V）exposure,the boy’s eosinophil count（EOS）and eosinophil percentage（EOS%）decreased by 23.05%(95%CI:-35.9%,-7.61%,P_-FDR=0.05),19.71%(95%CI:-30.63%,-7.07%,P_-FDR=0.03),no statistically significant association among girls(P_-FDR=0.64,0.45),gender interaction P_-interaction=0.072,0.027.In the model with metal as a categorical variable,the fourth group of manganese（Mn）had a reduction in eosinophils（EOS%）of 21.75%(95%CI:-32.29%,-9.55%,P_-FDR=0.01),the third group and the second group are not statistically significant compared with the first group,but the linear trend is significant(P_-trend=0.009),and the exposure levels of other metals are not statistically related to each white blood cell count.Under mixed multi-metal exposure,for every natural log unit increase in intrauterine manganese（Mn）exposure,boy white blood cell count（WBC）increased by 7.08%(95%CI:2.35%,12.03%,P_-FDR=0.03),among girls No statistically significant association(P_-FDR=1.00);the percentage of girls’monocytes（MONO%）decreased with the increase in manganese（Mn）exposure by 9.47%(95%CI:-14.91%,-3.68%,P_-FDR=0.02),and this association was not statistically significant in boys(P_-FDR=0.87),and there was no gender difference in the above associations(P_-interaction>0.05).Conclusion:At current environmental exposure levels,intrauterine exposure to vanadium（V）,cobalt（Co）,and cadmium（Cd）may be associated with a decrease in the white blood cell count of preschool-age 3-year-old boys.),Cobalt（Co）and cadmium（Cd）are negatively correlated with lymphocyte count（LYMPH）.There is a gender difference in the association between vanadium（V）and the percentage of eosinophils（EOS%）.High levels of manganese（Mn）in utero may be related to a decrease in the percentage of eosinophils（EOS%）in preschool-age 3-year-old children.Under mixed metal exposure,intrauterine manganese（Mn）may be associated with an increase in white blood cell count（WBC）in boys and a decrease in the percentage of mononuclear cells（MONO%）in girls.These results indicate that intrauterine exposure to vanadium（V）,cobalt（Co）,cadmium（Cd）,and manganese（Mn）may have an impact on the non-specific immune levels of preschool children,especially immunosuppression.更多还原