Abstract
Dieting is theorized as a risk factor for loss-of-control (LOC)-eating (i.e., feeling a sense of lack of control while eating). Support for this association has largely relied on retrospective self-report data, which does not always correlate with objectively assessed eating behavior in youth. We hypothesized that during a laboratory-based LOC-eating paradigm, children and adolescents who reported current (at the time of the visit) dieting would consume meals consistent with LOC-eating (greater caloric intake, and intake of carbohydrates and fats, but less intake of protein). Participants were presented with a buffet-style meal and instructed to "Let yourself go and eat as much as you want." Current dieting (i.e., any deliberate change to the amount or type of food eaten to influence shape or weight, regardless of how effective the changes are) was assessed via interview. General linear models were adjusted for fat mass (%), lean mass (kg), height, sex, protocol, race and ethnicity, pre-meal hunger and minutes since consumption of a breakfast shake. Of 337 participants (M(age) 12.8 ± 2.7y; 62.3 % female; 45.7 % non- Hispanic White and 26.1 % non-Hispanic Black; M(BMIz) 0.78 ± 1.11), only 33 (9.8 %) reported current dieting. Current dieting was not significantly associated with total energy intake (F = 1.63, p = .20, η(p)(2) = 0.005), or intake from carbohydrates (F = 2.45, p = .12, η(p)(2) = 0.007), fat (F = 2.65, p = .10, η(p)(2) = 0.008), or protein (F = 0.39, p = .53, η(p)(2) = 0.001). Contrary to theories that dieting promotes LOC-eating, current dieting was not associated with youth's eating behavior in a laboratory setting. Experimental approaches for investigating dieting are needed to test theories that implicate dieting in pediatric LOC-eating.