Predictors of early weight loss in a diet and exercise weight loss trial

This analysis examined baseline demographic, psychological, weight‐related, dietary and eating behavior variables as predictors of weight loss during the first two weeks of a randomized diet and exercise intervention. Overweight or obese premenopausal women [n=105; mean±SD, age: 33.8±7.5 y, body mass index (BMI): 30.1±3.1 kg/m 2 ] were assigned one of two energy‐restricted diets that varied in macronutrient composition, started a daily walking regimen, and attended weekly nutrition education classes. The relationships between the predictor variables and weight loss outcome were initially examined using univariate linear regressions. Age, ethnicity, waist‐to‐hip ratio, 2‐year weight fluctuations, and cognitive restraint were significantly associated with weight loss ( p <0.05) and were entered in the multivariate model. In the multivariate regression model, increased age was a significant predictor of weight loss ( p =0.003) as was decreased weight fluctuation over the previous two years ( p =0.001). In a separate regression analysis by diet group, similarity between baseline food intake and randomly assigned diet was found to predict weight loss ( p =0.013). Understanding predictors of early weight loss may facilitate the development of more effective long‐term interventions. Grant Funding Source : The General Mills Bell Institute of Health and Nutrition.