Influence of the Confounding Factors Age and Sex on MicroRNA Profiles from Peripheral Blood
Influence of the Confounding Factors Age and Sex on MicroRNA Profiles from Peripheral Blood
复制标题
DOI：
10.1373/clinchem.2014.224238
复制
发表时间：
2014-09-01
影响因子：
9.3
通讯作者：
Keller, Andreas
中科院分区：
文献类型：
--
作者：
Meder, Benjamin;Backes, Christina;Keller, Andreas
背景：从血液样本中测量的microRNAs(MiRNAs)是很有前途的微创生物标志物候选，已经在几个病例对照研究中得到了广泛的研究。然而，年龄和性别作为混杂变量的影响在很大程度上仍不清楚。方法：我们系统地研究了年龄和性别对109名生理上未受影响的人的miRNAs的影响，这些人的血液经微阵列技术表征(阶段1)。我们还调查了来自不同机构的一个独立队列，该队列由58名平均年龄相似但年龄分布较小的生理未受影响的个体组成。结果：在第一阶段中，我们检测到了318个与年龄显著相关的miRNAs，在调整了35个miRNAs的多重测试后，这些miRNAs仍然具有统计学意义。在性别方面，144个miRNAs表现出明显的失调。在这里，没有miRNA在多次检测调整后仍然显著。在阶段2的高通量数据集中，我们通常观察到较少数量的显著关联，主要是由于较小的队列大小和年龄分布的影响。然而，我们发现了7个与年龄相关的miRNAs，其中5个与阶段1一致。结论：纳入病例对照研究的个体的年龄分布需要仔细考虑，而性别可能较少混淆。为了支持miRNA到临床应用的转换，我们提供了一个基于Web的应用程序(http://www.ccb.uni-saarland.de/mirnacon))，以测试单个miRNA或miRNA签名受到影响的可能性。(C)2014年美国临床化学协会
BACKGROUND: MicroRNAs (miRNAs) measured from blood samples are promising minimally invasive biomarker candidates that have been extensively studied in several case-control studies. However, the influence of age and sex as confounding variables remains largely unknown.METHODS: We systematically explored the impact of age and sex on miRNAs in a cohort of 109 physiologically unaffected individuals whose blood was characterized by microarray technology (stage 1). We also investigated an independent cohort from a different institution consisting of 58 physiologically unaffected individuals having a similar mean age but with a smaller age distribution. These samples were measured by use of high-throughput sequencing (stage 2).RESULTS: We detected 318 miRNAs that were significantly correlated with age in stage 1 and, after adjustment for multiple testing of 35 miRNAs, remained statistically significant. Regarding sex, 144 miRNAs showed significant dysregulation. Here, no miRNA remained significant after adjustment for multiple testing. In the high-throughput datasets of stage 2, we generally observed a smaller number of significant associations, mainly as an effect of the smaller cohort size and age distribution. Nevertheless, we found 7 miRNAs that were correlated with age, of which 5 were concordant with stage 1.CONCLUSIONS: The age distribution of individuals recruited for case-control studies needs to be carefully considered, whereas sex may be less confounding. To support the translation of miRNAs into clinical application, we offer a web-based application (http://www.ccb.uni-saarland.de/mirnacon) to test individual miRNAs or miRNA signatures for their likelihood of being influenced. (C) 2014 American Association for Clinical Chemistry