Department of Nutrition, Harvard TH Chan School of Public Health, Boston, MA, USA.
Division of General Medicine Research Section, Beth Israel Deaconess Medical Center, 1309 Beacon Street, 2nd Floor, Brookline, MA, 02446, USA.
Eur J Epidemiol. 2020 Feb;35(2):93-97. doi: 10.1007/s10654-019-00578-3. Epub 2019 Nov 24.
In recent years, epidemiologists have increasingly sought to employ genetic data to identify 'causal' relationships between exposures of interest and various endpoints - an instrumental variable approach sometimes termed Mendelian randomization. However, this approach is subject to all of the limitations of instrumental variable analysis and to several limitations specific to its genetic underpinnings, including confounding, weak instrument bias, pleiotropy, adaptation, and failure of replication. Although the approach enjoys some utility in testing the etiological role of discrete biochemical pathways, like folate metabolism, examples like that of alcohol consumption and cardiovascular disease demonstrate that it must be treated with all of the circumspection that should accompany all forms of observational epidemiology. Going forward, we urge the elimination of randomization or causality in reports of its use and suggest that Mendelian randomization instead be termed exactly what it is - genetic instrumental variable analysis.
近年来，流行病学家越来越倾向于利用遗传数据来确定感兴趣的暴露与各种终点之间的“因果”关系——这种工具变量方法有时被称为孟德尔随机化。然而，这种方法受到工具变量分析的所有限制，以及其遗传基础特有的几个限制，包括混杂、弱工具偏差、多效性、适应和复制失败。尽管这种方法在测试离散生化途径（如叶酸代谢）的病因作用方面具有一定的实用性，但像饮酒与心血管疾病这样的例子表明，必须谨慎对待所有形式的观察性流行病学。今后，我们敦促在报告其使用情况时消除随机化或因果关系，并建议将孟德尔随机化恰当地称为——遗传工具变量分析。
