Abstract：
Previous studies on the association between metals and dyslipidemia are not completely consistent. There are few studies investigating the relationship between mixed metal exposure and dyslipidemia as well as the effects of metals on dyslipidemia in community-dwelling elderly. To evaluate the correlations and interaction effect between the urinary concentrations of metals and the risk of dyslipidemia in community-dwelling elderly. We designed a case-control study to assess the correlation between urine metals and dyslipidemia in elderly people in the Yinchuan. The urinary levels of 13 metals, including calcium, vanadium, iron, cobalt, zinc, copper, arsenic, selenium, molybdenum, cadmium, tellurium, and thallium, were measured by inductively coupled plasma-mass spectrometry (ICP-MS), and the blood biochemical analyzer was used to measure the blood lipid levels of 3384 senior individuals from four different areas of Yinchuan city. Logistic regression and restricted cubic splines (RCS) were used to explore the correlation and dose-response relationship between urinary metals and the risk of dyslipidemia. Least absolute shrinkage and selection operator (LASSO) regression was used to select metals, and then weighted quantile sum (WQS) regression was used to explore the weight of each metal in mixed metals. Bayesian kernel machine regression (BKMR) was used to explore the interactions between metals on dyslipidemia risk. (1) After selection by LASSO regression, in the multi-metal model, compared with the lowest quartile, the adjusted ORs (95%CI) of the highest quartiles were 0.47 (0.37-0.60) for Fe, 1.43 (1.13-1.83) for Zn, 1.46 (1.11-1.92) for As, 0.59 (0.44-0.80) for Se, 1.53 (1.18-2.00) for Mo, and 1.36 (1.07-1.73) for Te. (2) In the WQS regression model, Fe and Mo accounted for the largest weight in the negative and positive effects of dyslipidemia, respectively. (3) In the BKMR model, there may be a positive interaction between Te and Se on dyslipidemia. Among the mixed metals, Fe, As, Se, Mo, and Te were associated with the prevalence of dyslipidemia, with Fe and Mo contributing the most. There may be certain interactions between Te and Se.
摘要：
先前关于金属与血脂异常之间关联的研究并不完全一致。很少有研究调查混合金属暴露与血脂异常之间的关系以及金属对社区老年人血脂异常的影响。评价社区老年人尿金属浓度与血脂异常风险的相关性及交互作用。我们设计了一项病例对照研究，以评估银川市老年人尿金属与血脂异常之间的相关性。13种金属的尿液水平，包括钙,钒，铁,钴,锌，铜，砷,硒,钼,镉,碲,和铊,采用电感耦合等离子体质谱(ICP-MS),采用血液生化分析仪对银川市4个不同地区的3384名老年人进行血脂检测。采用Logistic回归和限制性三次样条(RCS)探讨尿金属与血脂异常风险的相关性和剂量-反应关系。使用最小绝对收缩和选择算子(LASSO)回归来选择金属，然后使用加权分位数和（WQS）回归来探索混合金属中每种金属的重量。贝叶斯核机回归(BKMR)用于探讨金属与血脂异常风险的相互作用。(1)通过LASSO回归选择后，在多金属模型中，与最低四分位数相比，Fe的最高四分位数的调整OR（95CI）为0.47（0.37-0.60），锌1.43（1.13-1.83），1.46(1.11-1.92)硒为0.59（0.44-0.80），1.53（1.18-2.00）适用于Mo，Te为1.36（1.07-1.73）。(2)在WQS回归模型中,Fe和Mo在血脂异常的负效应和正效应中占最大权重，分别。(3)在BKMR模型中，Te和Se对血脂异常可能存在正相互作用。在混合金属中，Fe,As,Se,Mo,和Te与血脂异常的患病率有关，Fe和Mo贡献最大。Te和Se之间可能存在某些相互作用。
