Print this page

NORC researchers find behavioral response to health interventions leads to less weight change than often predicted

  • November 26, 2014

New International Journal of Obesity paper offers an algorithm to more accurately predict body weight change outcomes.

cycling sizedEmily Dhurandhar, Ph.D., Kathryn Kaiser, Ph.D., and David B. Allison, Ph.D., of the University of Alabama at Birmingham Nutrition Obesity Research Center recently published new findings on realistic predictions of weight change for interventions that changed food intake or physical activity levels in free-living adults in the International Journal of Obesity.

The paper, titled “Predicting adult weight change in the real world: a systematic review and meta-analysis accounting for compensatory changes in energy intake or expenditure,” quantified the range of compensation in energy intake or expenditure observed in 28 randomized controlled trials in which adult participants could compensate both behaviorally and metabolically for the controlled intervention. The researchers compared observed weight changes to predictions from models that account only for metabolic compensation.

“We found evidence that people compensate for changes in food intake or physical activity considerably (e.g., if someone increases food intake once he or she starts an exercise regimen), such that weight change from these interventions is much less than expected when they are implemented in free-living people,” said Dhurandhar, study co-lead author. 

Based on their results, the investigators provided an algorithm that they say is a more accurate alternative for predicting outcomes in free-living populations as compared to other commonly used prediction models that operate under the assumption of no behavioral compensation in energy intake or expenditure. 

“Our study was important because the potential impact of many public health and clinical interventions is justified by prediction models that don’t account for the full range of compensation, and as a result, these models likely overestimate the potential impact of interventions,” Dhurandhar said.

“Now that there are better data to estimate population-level effects of different types of interventions on body weight, policymakers can be better informed,” said Kaiser, study co-lead author.