The Authors Reply

Victoria L. Stevens, Eric J. Jacobs, Alpa V. Patel, Marjorie L. McCullough, Peter T. Campbell, Susan M. Gapstur
2015 American Journal of Epidemiology  
the younger age group might have resulted in selection bias. Because the number of excluded cancer patients roughly matched the number included in the analyses (approximately 22,000 vs. 25,000 patients), the probable bias could be significant. Furthermore, we reviewed some studies (2, 3) with results contrary to those of the study by Stevens et al. and found that in those studies, there were wider age ranges for participants respectively,. Therefore, the differences between the results of these
more » ... studies and those of the study by Stevens et al. might have resulted from the interaction of age (younger or older than 50 years, for instance) and weight cycling. Consequently, we think that limiting the participants of the study to individuals older than 50 years of age and, more seriously, excluding all prevalent cases regardless of the time of diagnosis and primacy and recency of the exposure and outcome might have compromised the results of the study (1). ACKNOWLEDGMENTS Conflict of interest: none declared. REFERENCES 1. Stevens VL, Jacobs EJ, Patel AV, et al. Weight cycling and cancer incidence in a large prospective US cohort. Am J Epidemiol. 2015;182(5):394-404. 2. Nagle CM, Marquart L, Bain CJ, et al. Impact of weight change and weight cycling on risk of different subtypes of endometrial cancer.
doi:10.1093/aje/kwv247 pmid:26454258 fatcat:i3c725hjp5hxlk4nntr4mnd3wm