1097-0258/asset/SIM_left.gif?v=1&s=1b631772c3897aa95941da3609d901cd1d389e83)
1097-0258/asset/olbannerright.gif?v=1&s=6d257623b3308a7485294c87b3b5e1e665484099)
Research Article
Dichotomizing continuous predictors in multiple regression: a bad idea
Article first published online: 11 OCT 2005
DOI: 10.1002/sim.2331
Copyright © 2005 John Wiley & Sons, Ltd.
1097-0258/asset/cover.gif?v=1&s=64ebf4a6597e744f418c952845cddf175ccc795f "Statistics in Medicine")
Additional Information
How to Cite
Royston, P., Altman, D. G. and Sauerbrei, W. (2006), Dichotomizing continuous predictors in multiple regression: a bad idea. Statist. Med., 25: 127–141. doi: 10.1002/sim.2331
Publication History
- Issue published online: 10 DEC 2005
- Article first published online: 11 OCT 2005
- Manuscript Accepted: 11 MAY 2005
- Manuscript Received: 4 JAN 2005
Funded by
- Volkswagen-Stiftung
- Abstract
- References
- Cited By
Keywords:
- continuous covariates;
- dichotomization;
- categorization;
- regression;
- efficiency;
- clinical research
Abstract
In medical research, continuous variables are often converted into categorical variables by grouping values into two or more categories. We consider in detail issues pertaining to creating just two groups, a common approach in clinical research. We argue that the simplicity achieved is gained at a cost; dichotomization may create rather than avoid problems, notably a considerable loss of power and residual confounding. In addition, the use of a data-derived ‘optimal’ cutpoint leads to serious bias. We illustrate the impact of dichotomization of continuous predictor variables using as a detailed case study a randomized trial in primary biliary cirrhosis. Dichotomization of continuous data is unnecessary for statistical analysis and in particular should not be applied to explanatory variables in regression models. Copyright © 2005 John Wiley & Sons, Ltd.