作者: K. B. Michels , S. Greenland , B. A. Rosner
DOI: 10.1093/OXFORDJOURNALS.AJE.A009430
关键词:
摘要: Body mass index (BMI) has become the most commonly used of body composition in epidemiologic research. It displaced weight, height, and other measures composition. In this paper, authors show that use BMI alone does not always capture adequately joint relation size to health outcomes, such often represents implausible restrictions on relation. Use height or weight will be needed describe control confounding by these variables.