【作者】 孙斌；

【导师】 潘安；

【作者基本信息】 华中科技大学 ， 流行病与卫生统计学， 2022， 博士

【摘要】 在过去的几十年间,男性精液质量下降已成为一个重大的公共卫生问题,引起全世界的广泛关注。研究表明工业化和城镇化的高速发展所造成的环境污染是导致男性生殖健康损害的重要危险因素,而酚类化合物是其中一类常见的环境污染物。毒理学研究和流行病学研究均发现暴露于高浓度的酚类化合物环境中,可造成实验动物睾酮水平下降、精子形成障碍和精子DNA损伤,以及可引起男性精子活力降低、精子总数下降和精子浓度降低等不良影响。精子线粒体拷贝数和精子端粒长度是精子DNA的重要遗传表征,任何影响都可能引起精子正常功能的改变;且已有研究报道,精子线粒体拷贝数的增多可引起精子活力、精子浓度和精子总数降低;端粒长度异常增多,可诱发精子细胞产生异常。因人类在日常生活中通常会同时接触多种酚类化合物,而不是单独暴露;但过往研究普遍仅关注单一类别的酚类化合物与男性生殖健康的关系。此外,过往研究大多仅采集单次样本评估个体酚类化合物暴露水平和男性精液质量参数;而酚类化合物暴露水平和男性精液质量参数均存在较大的时间变异,仅采用单次采样可能会引起暴露或结局的错误分类,导致关联性结果的不一致。且既往研究的样本量相对较小,以及研究对象多招募自不孕不育诊所或生殖中心,故研究结果的外推性受限。因此,本研究依托湖北省人类精子库为研究现场,招募捐精筛查的普通育龄男性为研究对象,在基线及其随后的90天重复采集个体的尿液和精液样本,并检测个体内等体积混合尿液中双酚类化合物、二苯甲酮和对羟基苯甲酸酯浓度以反映在精子形成期间内的平均暴露水平,同时检测精液质量参数以及提取精子DNA进行精子线粒体拷贝数和精子端粒长度检测,探讨尿液中酚类化合物暴露与男性精液质量参数、精子线粒体拷贝数和精子端粒长度之间的关联。目的:本研究基于湖北省人类精子库建立的队列研究数据,第一、二、三部分分别探讨尿液中双酚类化合物、二苯甲酮和对羟基苯甲酸酯浓度与男性精液质量参数、精子线粒体拷贝数和精子端粒长度的关联,第四部分探讨酚类化合物混合暴露对精液质量参数、精子线粒体拷贝数和精子端粒长度影响的联合效应。方法:在2017年4月至2018年7月期间,本研究共招募到1487名研究对象,在排除484名因疾病或样本不足以完成实验室检测的研究对象后,最终本研究共纳入1003名男性,共提供了5948份重复采集的精液样本并完成精液质量参数检测,以及完成个体内等体积混合尿液样本中酚类化合物检测。在1003名完成尿液样本检测的男性中,成功提取了907名男性的精子DNA,并完成精子线粒体拷贝数和端粒长度的测定。本研究采用标准化问卷收集研究对象基本人口学信息、生活方式和疾病史等资料。尿液中双酚类化合物、二苯甲酮和对羟基苯甲酸酯浓度采用超高效液相色谱-串联质谱仪检测。精液质量参数分析根据世界卫生组织第五版精液检查手册进行。精子线粒体拷贝数和精子端粒长度使用实时荧光定量聚合酶链式反应进行测定。本研究将双酚类化合物、二苯甲酮和对羟基苯甲酸酯浓度按其四分位数划分为四组,以最低四分位数组为参考组,使用混合线性模型分别探讨个体混合尿液中双酚类化合物、二苯甲酮和对羟基苯甲酸酯浓度与多时点重复测量的精液质量参数之间的关联;使用广义线性模型分别探讨个体混合尿液中双酚类化合物、二苯甲酮和对羟基苯甲酸酯浓度与精子线粒体拷贝数和精子端粒长度之间的关联。对于在混合线性模型或广义线性模型中有统计学显著性差异的结果,进一步采用限制性立方样条(Restricted cubic spline,RCS)探讨潜在的非线性剂量反应关系。最后,采用加权分位数之和(Weighted quantile sum,WQS)回归模型、贝叶斯核机器回归(Bayesian kernel machine regression,BKMR)模型和Quantile g-computation(QG-C)模型评估不同种类酚类化合物混合暴露对男性生殖健康指标的影响。结果:本研究第一部分发现,尿液中双酚A(Bisphenol A,BPA)、双酚F(Bisphenol F,BPF)和双酚S(Bisphenol S,BPS)的检出率分别为96.41%、70.49%和98.40%,几何均数浓度(95%置信区间;Confidence interval,CI)分别为2.79(2.58,3.01)、0.90(0.75,1.07)和0.51(0.48,0.55)ng/m L。在混合线性模型中调整了研究对象的年龄、体质指数(Body mass index,BMI)、禁欲天数、尿肌酐、合并尿样数、文化程度、吸烟状态、饮酒状态、家庭每月收入水平、婚姻状态以及精液采样季节后,与尿BPA浓度处于最低四分位数组的男性相比,BPA浓度处于最高四分位数组的男性精子前向运动活力、精子总活力和精液体积分别降低了6.46%(95%CI:-12.01%,-0.56%)、6.24%(95%CI:-11.35%,-0.83%)和10.26%(95%CI:-16.83%,-3.15%);与尿BPS浓度处于最低四分位数组的男性相比,BPS浓度处于最高四分位数组的男性精子前向运动活力和精子总活力分别降低了6.82%(95%CI:-12.24%,-1.07%)和6.20%(95%CI:-11.22%,-0.89%)。在调整尿肌酐和潜在混杂因素后,广义线性模型结果显示与尿BPF浓度处于最低四分位数组的男性相比,BPF浓度处于最高四分位数组的男性精子线粒体拷贝数增加了18.39%(95%CI:2.86%,36.27%)。上述关联均在RCS中进一步得到验证。本研究第二部分发现,尿液中二苯甲酮-1(Benzophenone-1,BP-1)、二苯甲酮-4(Benzophenone-4,BP-4)、二苯甲酮-6(Benzophenone-6,BP-6)和4-羟基二苯甲酮(4-hydroxy benzophenone,4-OH-BP)的检出率分别为99.60%、89.03%、90.43%和97.46%,几何均数浓度(95%CI)分别为0.20(0.18,0.22)、4.17(3.70,4.70)、2.09(1.89,2.31)和0.10(0.09,0.11)ng/m L。在混合线性模型中,与尿BP-6浓度处于最低四分位数组的男性相比,BP-6浓度处于最高四分位数组的男性精子前向运动活力和精子总活力分别下降了6.99%(95%CI:-12.68%,-0.93%)和6.24%(95%CI:-11.52%,-0.64%)。在广义线性模型中,与尿BP-4浓度处于最低四分位数组的男性相比,BP-4浓度处于最高四分位数组的男性精子端粒长度增加了10.31%(95%CI:3.92%,17.08%);与尿4-OH-BP浓度处于最低四分位数组的男性相比,4-OH-BP浓度处于最高四分位数组的男性精子端粒长度增加了6.98%(95%CI:0.62%,13.75%)。本研究第三部分发现,尿液中对羟基苯甲酸甲酯(Methyl paraben,Me P)、对羟基苯甲酸乙酯(Ethyl paraben,Et P)和对羟基苯甲酸丙酯(Propyl paraben,Pr P)的检出率分别为100%、99.80%和99.20%,几何均数浓度(95%CI)分别为13.26(12.05,14.58)、1.21(1.08,1.35)和0.86(0.75,0.99)ng/m L。在混合线性模型中,并未发现Me P、Et P和Pr P与男性精液质量参数间的统计学显著性关联;仅发现与尿Me P浓度处于最低四分位数组的男性相比,Me P浓度处于最高四分位数组的男性精子前向运动活力和精子总活力分别下降了5.33%(95%CI:-10.93%,0.61%;P=0.078)和4.75%(95%CI:-9.93%,0.73%;P=0.088)。在广义线性模型中,与尿Pr P浓度处于最低四分位数组的男性相比,Pr P浓度处于最高四分位数组的男性精子线粒体拷贝数增加了15.12%(95%CI:-0.27%,32.88%;P=0.055)。本研究第四部分发现,在WQS回归模型中,在调整了尿肌酐和潜在的混杂因素后,WQS指数每增加一个四分位数,精子前向运动活力和精子总活力分别下降了2.88%(95%CI:-5.77%,0.10%;P=0.058)和2.65%(95%CI:-5.31%,0.09%;P=0.058),其中BPS是引起精子前向运动活力和精子总活力下降的主要危险因素(权重分别为0.34和0.32);WQS指数每增加一个四分位数,精子线粒体拷贝数增加了11.72%(95%CI:-1.20%,26.33%;P=0.078)以及精子端粒长度增加了5.08%(95%CI:0.75%,9.60%;P=0.021),其中BPF和BP-4分别是引起精子线粒体拷贝数和精子端粒长度增加的主要危险因素(权重分别为0.49和0.31)。在BKMR模型中,与所有酚类化合物暴露浓度处于中位数水平相比,当酚类化合物暴露浓度处于第60百分位数水平及以上时,精子线粒体拷贝数和精子端粒长度呈增加的趋势,而精子前向运动活力呈下降的趋势。将其他各酚类化合物暴露浓度固定在中位数水平时,BPF和Pr P与精子线粒体拷贝数呈正向的暴露-反应关系,BP-4和4-OH-BP与精子端粒长度呈正向的暴露-反应关系。在QG-C模型中,酚类化合物混合暴露每增加一个四分位数,精子前向运动活力和精子总活力分别下降了3.94%(95%CI:-7.98%,0.28%;P=0.067)和3.58%(95%CI:-7.19%,0.18%;P=0.062);且提示BPF和BP-4分别是引起精子线粒体拷贝数和精子端粒长度增加的主要危险因素。结论:本研究发现酚类化合物混合暴露与精子前向运动活力和精子总活力下降有关,其中BPS为主要危险因素;而酚类化合物混合暴露与精子线粒体拷贝数和精子端粒长度增加有关,其中BPF和BP-4分别是引起精子线粒体拷贝数和精子端粒长度增加的主要危险因素。但仍需进一步的研究来验证上述结果,并深入探讨酚类化合物暴露引起男性生殖健康损害的潜在生物学机制。更多还原

【Abstract】 Over the past few decades,a decline in human semen quality has become a major public health problem that has attracted widespread attention worldwide.Studies have shown that environmental pollution caused by rapid industrialization and urbanization is an important risk factor for male reproductive health,and phenols are one of the common environmental pollutants.Toxicological and epidemiological studies have found that exposure to higher concentrations of phenols can cause reduced testosterone levels,impaired sperm formation and sperm DNA damage in experimental animals,as well as adverse effects such as reduced sperm motility,decreased total sperm count and reduced sperm concentration in men.Sperm mitochondrial DNA copy number and sperm telomere length are important genetic characteristics of sperm DNA,and any alterations can affect normal sperm function;and it has been reported that increased sperm mitochondrial DNA copy number are associated with decreased sperm motility,sperm concentration and total sperm count;and increased sperm telomere length can affect spermatogenesis.Humans are usually exposed to multiple phenols simultaneously in daily life,rather than individually.And previous studies generally focused on the relationship between a single class of phenols and male reproductive health.In addition,it is noted that previous studies generally measured urinary phenol levels and sperm quality parameters at a single time point,which could overlook the high within-individual variability.Another factor likely contributing to inconsistent results across studies is the relatively small sample size,which may be underpowered to generate precise risk estimates.Therefore,by employing repeated measurements in a group of healthy young men who volunteered as potential sperm donors at the Hubei Province Human Sperm Bank,the present study repeatedly collected urine and semen samples from individuals at baseline and the following 90 days.Concentrations of bisphenols,benzophenones and parabens in within-individual pooled urine were measured to reflect the average exposure during spermatogenesis,while semen quality parameters were examined and sperm DNA were extracted to measure sperm mitochondrial copy number and telomere length.Objectives: The present study aimed to explore the associations of urinary concentrations of bisphenols,benzophenones and parabens with semen quality parameters,sperm mitochondrial copy number and sperm telomere length in Parts 1,2 and 3,respectively,and the overall effect of mixed exposure to phenols on semen quality parameters,sperm mitochondrial copy number and sperm telomere length in Part 4.Methods: Between April 2017 and July 2018,1487 men were recruited,and after excluding 484 men due to illness or insufficient samples to complete laboratory determination,a total of 1003 men were ultimately included in this study,who provided5948 repeated semen samples.Urinay concentrations of phenols were determined by an ultra-performance liquid chromatography coupled with a triple quadrupole mass spectrometer.Semen analyses were performed according to the 5th World Health Organization laboratory manual.Real-time quantitative polymerase chain reaction was used to measure sperm mitochondrial DNA copy number and sperm telomere length.907 of 1003 participants completed sperm mitochondrial copy number and sperm telomere length determination.Urinary concentrations of bisphenols,benzophenones and parabens were divided into four groups according to their quartiles,and the lowest quartile group was used as the reference group.Mixed-effect models were used to examine the associations of urinary concentrations of bisphenols,benzophenones and parabens with semen quality parameters.General linear models were applied to explore the associations of urinary concentrations of bisphenols,benzophenones and parabens with sperm mitochondrial DNA copy number and sperm telomere length.We further applied restricted cubic spline(RCS)functions to explore potential non-linear dose-response relationships.Weighted quantile sum(WQS)regression,Bayesian kernel machine regression(BKMR)model and Quantile g-computation(QG-C)were used to evaluate the relationship between phenol mixture exposure and male reproductive health indicators.Results: In Part 1,the detection rates of bisphenol A(BPA),bisphenol F(BPF)and bisphenol S(BPS)were 96.41%,70.49%,and 98.40% in 1003 pooled urine,respectively.Their geometric mean(95%CI)of BPA,BPF,and BPS concentrations were 2.79(2.58,3.01),0.90(0.75,1.07)and 0.51(0.48,0.55)ng/m L,respectively.After adjusting for age,body mass index(BMI),abstinence period,urinary creatinine,number of pooled urine samples,education levels,smoking,drinking,monthly income,marital status,and semen sampling season in mixed-effect model,compared with men in the first quartile of BPA,those in the fourth quartile had a lower progressive sperm motility,total motility,and semen volume of 6.46%(95%CI:-12.01%,-0.56%),6.24%(95%CI:-11.35%,-0.83%)and 10.26%(95%CI:-16.83%,-3.15%),respectively;Compared with men in the lowest quartile of BPS,those in the highest quartile had a lower progressive sperm motility and total motility of 6.82%(95%CI:-12.24%,-1.07%)and 6.20%(95%CI:-11.22%,-0.89%),respectively.After adjusting for urinary creatinine and potential confounders in general linear model,compared with men in the first quartile of BPF,those in the fourth quartile had a higher sperm mitochondrial DNA copy number of 18.39%(95%CI: 2.86%,36.27%).The abovementioned associations were further confirmed in RCS.In Part 2,the detection rates of benzophenone-1(BP-1),benzophenone-4(BP-4),benzophenone-6(BP-6)and 4-hydroxy benzophenone(4-OH-BP)were 99.60%,89.03%,90.43% and 97.46%,respectively.Their geometric mean(95%CI)of BP-1,BP-4,BP-6and 4-OH-BP were 0.20(0.18,0.22),4.17(3.70,4.70),2.09(1.89,2.31)and 0.10(0.09,0.11)ng/m L,respectively.In mixed-effect model,compared with men in the first quartile of BP-6,those in the fourth quartile had a lower progressive sperm motility and total motility of 6.99%(95% CI:-12.68%,-0.93%)and 6.24%(95% CI:-11.52%,-0.64%),respectively.In general linear model,compared with men in the first quartile of BP-4 and4-OH-BP,those in the fourth quartile had a higher sperm telomere length of 10.31%(95%CI: 3.92%,17.08%)and 6.98%(95% CI: 0.62%,13.75%),respectively.In Part 3,the detection rates of methyl paraben(Me P),ethyl paraben(Et P)and propyl paraben(Pr P)were 100%,99.80% and 99.20%,respectively.Their geometric mean(95%CI)of Me P,Et P and Pr P were 13.26(12.05,14.58),1.21(1.08,1.35)and 0.86(0.75,0.99)ng/m L.In mixed-effect model,no significant associations between Me P,Et P and Pr P and semen quality parameters were found;We just found men in the fourth quartile of Me P had a lower progressive sperm motility and total motility of 5.33%(95%CI:-10.93%,0.61%;P=0.078)and 4.75%(95%CI:-9.93%,0.73%;P=0.088),compared to men in the first quartile.In general linear model,compared with men in the first quartile of Pr P,those in the highest quartile had a higher sperm mitochondrial DNA copy number of 15.12%(95%CI:-0.27%,32.88%;P=0.055).In Part 4,after adjusting for urinary creatinine and potential confounders in WQS regression model,one quartile increase in the WQS index was associated with a lower progressive sperm motility and total motility of 2.88%(95%CI:-5.77%,0.10%;P=0.058)and 2.65%(95%CI:-5.31%,0.09%;P=0.058),respectively;while BPS was the most heavily weighted chemicals(weights 0.34 and 0.32,respectively);and one quartile increase in the WQS index was associated with a higher sperm mitochondrial DNA copy number and sperm telomere length of 11.72%(95%CI:-1.20%,26.33%;P=0.078)and5.08%(95%CI: 0.75%,9.60%;P=0.021),respectively;while BPF and BP-4 were the most heavily weighted chemicals for sperm mitochondrial DNA copy number and sperm telomere length(weights 0.49 and 0.31,respectively).In the BKMR model,sperm mitochondrial copy number and sperm telomere length showed an increasing trend,while progressive sperm motility showed a decreasing trend when the phenol mixtures was at the 60 th percentile or above compared to their 50 th percentile.When all the other phenols were fixed at the median exposure level,BPF and Pr P showed a positive exposure-response relationship with sperm mitochondrial copy number,and BP-4 and4-OH-BP showed a positive exposure-response relationship with sperm telomere length.In QG-C,one quartile increase in phenol mixtures was associated with a lower progressive sperm motility and total motility of 3.94%(95%CI:-7.98%,0.28%;P=0.067)and 3.58%(95%CI:-7.19%,0.18%;P=0.062),respectively;and suggested that BPF and BP-4 were the main risk factors for sperm mitochondrial DNA copy number and sperm telomere length.Conclusion: The joint effect of phenol mixtures was associated with decreased progressive sperm motility and total motility,and increased sperm mitochondrial DNA copy number and sperm telomere length.BPS,BPF and BP-4 were identified as the most important risk factors with sperm motility,sperm mitochondrial DNA copy number and sperm telomere length,respectively.However,further studies are needed to confirm these results and explore the underlying biological mechanisms.更多还原