Compositional data analysis for physical activity, sedentary time and sleep research

作     者：Dumuid, Dorothea Stanford, Tyman E. Martin-Fernandez, Josep-Antoni Pedisic, Zeljko Maher, Carol A. Lewis, Lucy K. Hron, Karel Katzmarzyk, Peter T. Chaput, Jean-Philippe Fogelholm, Mikael Hu, Gang Lambert, Estelle V. Maia, Jose Sarmiento, Olga L. Standage, Martyn Barreira, Tiago V. Broyles, Stephanie T. Tudor-Locke, Catrine Tremblay, Mark S. Olds, Timothy 

作者机构：Univ South Australia Sch Hlth Sci Adelaide SA Australia Univ Adelaide Sch Math Sci Adelaide SA Australia Univ Girona Dept Informat Matemat Aplicada & Estadist Girona Spain Victoria Univ Inst Sport Exercise & Act Living Melbourne Vic Australia Flinders Univ S Australia Sch Hlth Sci Adelaide SA Australia Univ Palackeho Dept Math Anal & Applicat Math Olomouc Czech Republic Pennington Biomed Res Ctr 6400 Perkins Rd Baton Rouge LA 70808 USA Childrens Hosp Eastern Ontario Res Inst Hlth Act Living & Obes Res Ottawa ON Canada Helsingin Yliopisto Dept Food & Environm Sci Helsinki Finland Univ Cape Town Dept Human Biol Cape Town South Africa Univ Porto Fac Desporto Porto Portugal Univ Los Andes Fac Med Bogota Colombia Univ Bath Dept Hlth Bath Avon England Syracuse Univ Dept Exercise Sci Syracuse NY USA Univ Massachusetts Dept Kinesiol Amherst MA 01003 USA 

出 版 物：《STATISTICAL METHODS IN MEDICAL RESEARCH》 (医学研究统计方法)

年 卷 期：2018年第27卷第12期

页      面：3726-3738页

核心收录：

学科分类：0710[理学-生物学] 1204[管理学-公共管理] 02[经济学] 0202[经济学-应用经济学] 020208[经济学-统计学] 1001[医学-基础医学(可授医学、理学学位)] 07[理学] 0714[理学-统计学（可授理学、经济学学位）] 070103[理学-概率论与数理统计] 0701[理学-数学] 

基　　金：Australian Government Research Training Program Scholarship National Heart Foundation Spanish Ministry of Economy and Competitiveness under the project CODA-RETOS [MTM2015-65016-C2-1(2)-R] Coca-Cola Company 

主　　题：Compositional data analysis physical activity sedentary behaviour sleep multicollinearity 

摘      要：The health effects of daily activity behaviours (physical activity, sedentary time and sleep) are widely studied. While previous research has largely examined activity behaviours in isolation, recent studies have adjusted for multiple behaviours. However, the inclusion of all activity behaviours in traditional multivariate analyses has not been possible due to the perfect multicollinearity of 24-h time budget data. The ensuing lack of adjustment for known effects on the outcome undermines the validity of study findings. We describe a statistical approach that enables the inclusion of all daily activity behaviours, based on the principles of compositional data analysis. Using data from the International Study of Childhood Obesity, Lifestyle and the Environment, we demonstrate the application of compositional multiple linear regression to estimate adiposity from children s daily activity behaviours expressed as isometric log-ratio coordinates. We present a novel method for predicting change in a continuous outcome based on relative changes within a composition, and for calculating associated confidence intervals to allow for statistical inference. The compositional data analysis presented overcomes the lack of adjustment that has plagued traditional statistical methods in the field, and provides robust and reliable insights into the health effects of daily activity behaviours.