Ramamoorthi R V, Rossano M G, Paneth N, Gardiner J C, Diamond M P, Puscheck E, Daly D C, Potter R C, Wirth J J
Department of Statistics and Probability, Michigan State University, East Lansing, MI 48824, USA.
Stat Med. 2008 Aug 15;27(18):3503-14. doi: 10.1002/sim.3236.
In studies of environmental effects on human health outcomes, it is often difficult to assess the effects of a group of exposure variables when the individual exposures do not appear to have statistically significant effects. To address this situation, we propose a method of U-scores applied to subsets of multivariate data. We illustrate the usefulness of this approach by applying it to data collected as part of a study on the effects of metal exposure on human semen parameters. In this analysis, profiles (pairs) of metals containing copper and/or manganese were negatively correlated with total motile sperm and profiles containing copper were negatively correlated with sperm morphology; profiles containing selenium and chromium were positively correlated with total motile sperm.
在关于环境对人类健康结果影响的研究中，当个体暴露似乎没有统计学上的显著影响时，通常很难评估一组暴露变量的影响。为了解决这种情况，我们提出了一种应用于多元数据子集的U分数方法。我们将这种方法应用于作为金属暴露对人类精液参数影响研究一部分收集的数据，以此来说明该方法的实用性。在这项分析中，含有铜和/或锰的金属组合（对）与总活动精子呈负相关，含有铜的组合与精子形态呈负相关；含有硒和铬的组合与总活动精子呈正相关。
