Abstract
Some, but not all studies have reported that, among youth with disordered eating and high weight, the relative reinforcing value of food (RRV-F, i.e., how hard a person will work for a high-energy-dense food when another reward is available) is greater, and food-related inhibitory control (i.e., ability to withhold a response to food-related stimuli) is lower, compared to peers without disordered eating or overweight. In most studies, high RRV-F and low food-related inhibitory control have been studied separately, as independent factors, with each suggested to predict excess weight and adiposity (fat mass) gain. We hypothesized that the interaction of these factors would prospectively exacerbate risk for weight and adiposity (fat mass) gain three years later in a sample of healthy youth. At baseline, RRV-F was measured using a Behavior Choice Task with the rewards being standardized servings of chocolate candies, cheese crackers, or fruit snacks. Food-related inhibitory control was determined by performance in response to food and non-food stimuli during a Food Go/No-Go task. At baseline and 3-year visits, total body adiposity was measured by dual-energy X-ray absorptiometry (DXA) and body mass index (BMI) was obtained using measured weight and height. A linear regression was conducted with 3-year adiposity as the dependent variable. RRV-F, food-related inhibitory control, and the RRV-F x food-related inhibitory control interaction as independent variables. Baseline adiposity, age, height, sex, race/ethnicity, and days between visits were included as covariates for model predicting 3-year adiposity. Baseline BMI, age, sex, race/ethnicity, and days between visits were included as covariates for model predicting 3-year BMI. One-hundred and nine youth (mean 12.4±2.7y, mean 0.50±1.02 BMIz, 30.3% with overweight/obesity, 45.9% female, 51.4% non-Hispanic White), 8-17 years at baseline, were studied. Baseline food-related inhibitory control (β(unstandardized) = 0.33, p = .037, 95% CI [.02, 0.64]), but not baseline RRV-F (β(unstandardized) = -0.003, p = .914), 95% CI [-0.05, 0.05]) was significantly associated with 3-year adiposity such that those with the poorest food-related inhibitory control (great number of commision errors) had the greatest adiposity gain. The interaction between RRV-F and food-related inhibitory control did not predict 3-year adiposity (β(unstandardized) = -0.07, p = .648, 95% CI [-0.39, 0.25]). The pattern of findings was the same for models examining non-food related inhibitory control. Neither baseline food-related inhibitory control (β(unstandardized) = 2.16, p = .256, 95% CI [-1.59, 5.92]), baseline RRV-F (β(unstandardized) = 0.14, p = .660, 95% CI [-0.48, 0.75]), nor their interaction (β(unstandardized) = -1.18, p = .547, 95% CI [-5.04, 2.69]) were significantly associated with 3-year BMI. However, non-food related inhibitory control (β(unstandardized) = 0.54, p = .038, 95% CI [.22, 7.15]) was significantly associated with 3-year BMI. In summary, food-related inhibitory control but not RRV-F, was associated with changes in adiposity in a sample of children and adolescents. Among generally healthy youth, food-related inhibitory control may be a more relevant risk factor than food reinforcement for adiposity gain. Additional data are needed to determine how inhibitory control and reward systems, as well as other disinhibited eating behaviors/traits, may interact to promote excess weight gain over time in youth.