Robin Bliss1, Janice Weinberg, Thomas Webster, Veronica Vieira. 1. Department of Environmental Health, Boston University School of Public Health, Boston, MA, USA ; Orthopedic and Arthritis Center for Outcomes Research, Brigham and Women's Hospital/Harvard Medical School, Boston, MA, USA.
Abstract
BACKGROUND: In epidemiologic studies researchers are often interested in detecting confounding (when a third variable is both associated with and affects associations between the outcome and predictors). Confounder detection methods often compare regression coefficients obtained from "crude" models that exclude the possible confounder(s) and "adjusted" models that include the variable(s). One such method compares the relative difference in effect estimates to a cutoff of 10% with differences of at least 10% providing evidence of confounding. METHODS: In this study we derive the asymptotic distribution of the relative change in effect statistic applied to logistic regression and evaluate the sensitivity and false positive rate of the 10% cutoff method using the asymptotic distribution. We then verify the results using simulated data. RESULTS: When applied to a logistic regression models with a dichotomous outcome, exposure, and possible confounder, we found the 10% cutoff method to have an asymptotic lognormal distribution. For sample sizes of at least 300 the authors found that when confounding existed, over 80% of models had >10% changes in odds ratios. When the confounder was not associated with the outcome, the false positive rate increased as the strength of the association between the predictor and confounder increased. When the confounder and predictor were independent of one another, false positives were rare (most < 10%). CONCLUSIONS: Researchers must be aware of high false positive rates when applying change in estimate confounder detection methods to data where the exposure is associated with possible confounder variables.
BACKGROUND: In epidemiologic studies researchers are often interested in detecting confounding (when a third variable is both associated with and affects associations between the outcome and predictors). Confounder detection methods often compare regression coefficients obtained from "crude" models that exclude the possible confounder(s) and "adjusted" models that include the variable(s). One such method compares the relative difference in effect estimates to a cutoff of 10% with differences of at least 10% providing evidence of confounding. METHODS: In this study we derive the asymptotic distribution of the relative change in effect statistic applied to logistic regression and evaluate the sensitivity and false positive rate of the 10% cutoff method using the asymptotic distribution. We then verify the results using simulated data. RESULTS: When applied to a logistic regression models with a dichotomous outcome, exposure, and possible confounder, we found the 10% cutoff method to have an asymptotic lognormal distribution. For sample sizes of at least 300 the authors found that when confounding existed, over 80% of models had >10% changes in odds ratios. When the confounder was not associated with the outcome, the false positive rate increased as the strength of the association between the predictor and confounder increased. When the confounder and predictor were independent of one another, false positives were rare (most < 10%). CONCLUSIONS: Researchers must be aware of high false positive rates when applying change in estimate confounder detection methods to data where the exposure is associated with possible confounder variables.
Authors: Tyler J Lane; Rebbecca Lilley; Sheilah Hogg-Johnson; Anthony D LaMontagne; Malcolm R Sim; Peter M Smith Journal: J Occup Rehabil Date: 2018-06
Authors: Michael J Rosen; Rebekah Karns; Jefferson E Vallance; Ramona Bezold; Amanda Waddell; Margaret H Collins; Yael Haberman; Phillip Minar; Robert N Baldassano; Jeffrey S Hyams; Susan S Baker; Richard Kellermayer; Joshua D Noe; Anne M Griffiths; Joel R Rosh; Wallace V Crandall; Melvin B Heyman; David R Mack; Michael D Kappelman; James Markowitz; Dedrick E Moulton; Neal S Leleiko; Thomas D Walters; Subra Kugathasan; Keith T Wilson; Simon P Hogan; Lee A Denson Journal: Gastroenterology Date: 2017-01-26 Impact factor: 22.682
Authors: Cecilie Kyrø; Anja Olsen; Rikard Landberg; Guri Skeie; Steffen Loft; Per Åman; Max Leenders; Vincent K Dik; Peter D Siersema; Tobias Pischon; Jane Christensen; Kim Overvad; Marie-Christine Boutron-Ruault; Guy Fagherazzi; Vanessa Cottet; Tilman Kühn; Jenny Chang-Claude; Heiner Boeing; Antonia Trichopoulou; Christina Bamia; Dimitrios Trichopoulos; Domenico Palli; Vittorio Krogh; Rosario Tumino; Paolo Vineis; Salvatore Panico; Petra H Peeters; Elisabete Weiderpass; Toril Bakken; Lene Angell Åsli; Marcial Argüelles; Paula Jakszyn; María-José Sánchez; Pilar Amiano; José María Huerta; Aurelio Barricarte; Ingrid Ljuslinder; Richard Palmqvist; Kay-Tee Khaw; Nick Wareham; Timothy J Key; Ruth C Travis; Pietro Ferrari; Heinz Freisling; Mazda Jenab; Marc J Gunter; Neil Murphy; Eilo Riboli; Anne Tjønneland; H B as Bueno-de-Mesquita Journal: J Natl Cancer Inst Date: 2013-12-07 Impact factor: 13.506