A paper estimating the effects of limiting fast food meals with toys to under 550 calories has been retracted after concerns arose regarding the scientists’ use of an outdated model for estimating weight changes in kids.
The paper estimated that kids who eat fast food twice a week would avoid gaining two pounds a year if calorie limits are imposed on meals with toys. However, everyone we spoke to, and the notice, indicated that their estimate was inaccurate.
Here’s the notice for “Modeling Potential Effects of Reduced Calories in Kids’ Meals with Toy Giveaways”:
Childhood Obesity is officially retracting the article entitled, “Modeling Potential Effects of Reduced Calories in Kids’ Meals with Toy Giveaways,” by Maysoun Y. Freij, Randall L. Sell, Anne K. Bozack, Linda J. Weiss, and Ana C. Garcia, which was published in the February 2014 issue, volume 10, number 1, pages 58-63. doi: 10.1089/chi.2013.0082
After the article was published, Childhood Obesity received a letter to the editor, which questioned the model used by the authors to predict potential weight gain or loss. Specifically, the letter’s authors reported that one component of the model was outdated and not applicable to predicting weight gain or loss in children. The letter’s authors went on to state that the erroneous model construction led to predicted weight differentials that were approximately 10 times higher than those that would be predicted by accepted models in obesity science, and thus major portions of the article by Freij and colleagues were invalid.
Dr. Freij was contacted by the editorial office of Childhood Obesity and the concerns expressed within the letter to the editor shared. Dr. Freij concurs with retracting this article. Childhood Obesity is also publishing the letter, “Order of Magnitude Misestimation of Weight Effects of Children’s Meal Policy Proposals,” by Andrew W. Brown, Kevin D. Hall, Diana Thomas, Nikhil V. Dhurandhar, Steven B. Heymsfield, and David B. Allison in its entirety.
Childhood Obesity is dedicated to the highest standards of scientific research and assurance of accurate information in the work published by the Journal.
Here’s what we heard from author Maysoun Freij:
Although the data and calculations in our study were correct, concerns focused on estimates – which had just recently become available – regarding the effects of caloric reductions in the weight of children. We opted to retract the article, because we felt uncomfortable with the dissemination of a model that we knew (based on the new estimates) had certain deficiencies. We are now updating the work but have not yet decided whether to publish a revision.
I was pleasantly surprised to learn that the authors had chosen to retract their paper in response to our letter pointing out a flaw in their calculations. Our letter was prompted when we noticed the erroneous calculations upon reading the original article. The erroneous calculations were, to the best of our ability to discern, merely an honest error on the authors’ part. Yet the error was large enough that it invalidated the authors’ conclusions. I believe that the journal and the authors did the right thing to ensure the integrity of the scientific record and I commend them for it.