Secular trends in the relationship between caloric intake, carbohydrate intake, and physical activity with obesity

The prevalence of obesity has increased dramatically over the past several decades. A number of factors contributing to obesity are well established; however, the etiology of obesity remains unclear. We set out to examine how caloric intake, carbohydrate intake, and physical activity associate with obesity, and whether this association has changed over time using data from the National Health and Nutrition Survey (NHANES). Linear regression was conducted to determine secular changes in the association between physical activity, caloric intake, and carbohydrate intake with BMI. The interaction of relative caloric intake (kcal/kg) and time on BMI was significant after adjusting for physical activity and carbohydrate intake, as the difference between the predicted BMI for any two relative caloric intakes increased over time (P<0.0001). The main effects of time on BMI was significant, indicating that a given relative caloric intake predicted a greater BMI over time (P<0.0001). The association between carbohydrate intake and time revealed significant main effects of time, as a given carbohydrate intake predicted a greater BMI over time (P<0.05). The interaction between exercise frequency (times/week) and time on BMI was significant for females; the difference between the predicted BMI for two physical activity levels increased over time (P=0.0006). The main effects of time on BMI was significant for both sexes, indicating that a given physical activity level predicted a greater BMI over time (P<0.0001). Our results indicate that neither caloric intake nor physical activity consistently predicted BMI over time. We speculate that additional factors may be influencing how physical activity and dietary factors relate with obesity.