武汉社区人群尿中23种金属与平均血小板体积的关系

Association between 23 urinary metals and mean platelet volume among a community-dwelling population in Wuhan, China

目的 探讨武汉社区人群尿中23种金属浓度与平均血小板体积（mean platelet volume, MPV）的相关关系。方法 2011年4—5月采用分层整群随机抽样方法抽取了3 053名在武汉市汉阳和汉口地区生活不少于5年且年龄介于18～80岁之间的社区居民，在清晨空腹状态下采集了外周静脉血和尿样。采用电感耦合等离子体质谱仪测定尿中铝、钛、钒、铬、锰、铁、钴、镍、铜、锌、砷、硒、铷、锶、钼、镉、锡、锑、钡、钨、铊、铅和铀共23种金属离子，采用全自动生化分析仪测定血液中MPV。将MPV介于7.0～11.0 fl的调查对象定义为正常组，将MPV〉11.0 fl的定义为异常升高组。在排除了变量缺失与患有心血管疾病的调查对象后，对2 203名调查对象进行统计分析。应用广义线性回归模型分析尿金属与MPV的相关关系，将尿金属浓度分为四分位后，采用多变量Logistic回归模型评估不同分位调查对象MPV异常升高的风险。采用阳性错误发现率（FDR）控制由多重假设检验产生的假阳性率。结果 尿中砷（P50=2.431 μg/mmol肌酐）和钼（P50=4.035 μg/mmol肌酐）与MPV升高具有正相关关系[β（95%CI）值分别为0.119（0.043～0.196）和0.119（0.042～0.195），经FDR校正的P值均为0.018；与尿金属浓度为第一分位的调查对象相比，第四分位的OR（95%CI）值分别为1.830（1.382～2.423）和1.496（1.125～1.989），经FDR校正的P值分别为〈0.001和0.014]，而尿中铝（P50=2.706 μg/mmol肌酐）和铊（P50=0.046 μg/mmol肌酐）与MPV升高具有负相关关系[β（95%CI）值分别为-0.115（-0.195～-0.034）和-0.307（-0.386～-0.228），经FDR校正的P值分别为0.029和〈0.001；与尿金属浓度为第一分位的调查对象相比，第四分位的OR（95%CI）值分别为0.566（0.412～0.779）和0.302（0.219～0.416），经FDR校正的P值分别为0.008和〈0.001]。尿中铁（P50=6.716 μg/mmol肌酐）、锑（P50=0.014 μg/mmol肌酐）和铀（P50=0.003 μg/mmol肌酐）与MPV异常升高存在正相关[与尿金属浓度为第一分位的调查对象相比，第四分位的OR（95%CI）值分别为1.866（1.395～2.496）、1.507（1.111～2.043）和1.452（1.063～1.984），经FDR校正的P值分别为〈0.001、0.022和0.012]，而尿中钨（P50=0.010 μg/mmol肌酐）和铅（P50=0.265 μg/mmol肌酐）与MPV异常升高具有负相关关系[与尿金属浓度为第一分位的调查对象相比，第四分位的OR（95%CI）值分别为0.551（0.417～0.726）和0.534（0.394～0.725），经FDR校正的P值均〈0.001]。此外，多金属模型的输出结果还表明，MPV异常升高与尿中铬（P50=0.120 μg/mmol肌酐）、硒（P50=0.646 μg/mmol肌酐）呈正相关关系，而与尿中镍（P50=0.193 μg/mmol肌酐）存在负相关关系[与尿金属浓度为第一分位的调查对象相比，第四分位的OR（95%CI）值分别为1.578（1.054～2.363）、1.718（1.159～2.549）和0.535（0.373～0.767），P值分别为0.017、0.028和0.002]。结论 武汉社区人群铝、铬、铁、镍、砷、硒、钼、锑、钨、铊、铅和铀暴露与MPV异常升高存在相关关系。

Objective To investigate the potential association between 23 urinary metals and mean platelet volume （MPV） among a community population in Wuhan.Methods A total of 3 053 community residents who lived in the sampling buildings for more than 5 years, aged from 18 to 80 years, were recruited using a stratified, cluster sampling approach in Wuhan city, China. Blood and urine samples were obtained from participants in the morning under fasting conditions. Urinary metals, including aluminum, titanium, vanadium, chromium, manganese, iron, cobalt, nickel, copper, zinc, arsenic, selenium, rubidium, strontium, molybdenum, cadmium, tin, antimony, barium, tungsten, thallium, lead and uranium, were measured by inductively coupled plasma mass spectrometry. The MPV contents were determined using a fully automated clinical chemistry analyzer. Participants with missing data on covariates or cardiovascular disease were excluded. According to the reference intorvals of MPV for Chinese adults, the participants were classified into normal （7.0-11.0 fl） and high MPV （〉11.0 fl） subgroups. Data from 2 203 participants were used to evaluate the associations between urinary metals and MPV levels using generalized linear regression models, and the risk of abnormal elevation of MPV using multivariable logistic regression models. The false discovery rate （FDR）-corrected P-value from 23 hypothesis tests was used to adjust for multiple testing.Results After adjusting for potential confounders, urinary concentrations of arsenic （P50=2.431 μg/mmol creatinine） and molybdenum （P50=4.035 μg/mmol creatinine） were significantly associated with increased MPV levels and the risk of abnormal elevation of MPV. In contrast, urinary aluminum （P50=2.706 μg/mmol creatinine） and thallium （P50=0.046 μg/mmol creatinine） were associated with decreased MPV levels, but also the risk of abnormal elevation of MPV. The regression coefficients and 95% CIs were 0.119 （0.043-0.196） for arsenic （FDR-adjusted P=0.018）, 0.119 （0.042-0.195） for molybdenum （FDR-adjusted P=0.018）, -0.115 （-0.195--0.034） for aluminum （FDR-adjusted P=0.029）, and -0.307 （-0.386- -0.228） for thallium （FDR-adjusted P〈0.001）, respectively. When comparing the extreme quartiles for arsenic, molybdenum, aluminum and thallium, adjusted OR and the 95%CIs were 1.830 （1.382-2.423）, 1.496 （1.125-1.989）, 0.566 （0.412-0.779） and 0.302 （0.219-0.416）, respectively, and FDR-adjusted P-values were 〈0.001,〈0.014,〈0.008 and〈0.001, respectively. Moreover, significant associations were found between an increased risk of abnormal MPV elevation with urinary iron （P50=6.716 μg/mmol creatinine） , antimony （P50=0.014 μg/mmol creatinine） and uranium （P50=0.003 μg/mmol creatinine） , and a decreased risk with urinary tungsten （P50=0.010 μg/mmol creatinine） and lead （P50=0.265 μg/mmol creatinine） . When comparing the extreme quartiles for iron, antimony and uranium, the respective adjusted OR （95%CI） were 1.866 （1.395-2.496）, 1.507 （1.111-2.043） and 1.452 （1.063-1.984）, and the respective FDR-adjusted P-values were 〈0.001,〈0.022 and〈0.012. The respective adjusted OR （95%CI） were 0.551 （0.417-0.726） and 0.534 （0.394-0.725）, and the respective FDR-adjusted P-values were〈0.001 and〈0.001, when comparing the extreme quartiles for tungsten and lead. Based on multi-metal models, urinary chromium （P50=0.120 μg/mmol creatinine） and selenium （P50=0.646 μg/mmol creatinine） were associated with increased risk of abnormal MPV, while urinary nickel （P50=0.193 μg/mmol creatinine） was associated with decreased risk of abnormal MPV. When comparing the extreme quartiles for chromium, selenium and nickel, adjusted OR （95% CI） were 1.578 （1.054-2.363）, 1.718 （1.159-2.549） and 0.535 （0.373-0.767）, respectively, and the FDR-adjusted P-values were 0.017, 0.028 and 0.002, respectively.Conclusion In the general population of Wuhan city, exposure to aluminum, chromium, iron, nickel, arsenic, selenium, molybdenum, antimony, tungsten, thallium, lead and uranium were all associated with abnormal MPV elevation.