(P31-040-23) Identifying Bio-Behavioral Predictors of Food Intake in Older Adolescents

Objectives: The purpose of this study was to identify the bio-behavioral predictors of daily food intake in older adolescents.

Methods: Baseline data was utilized from our current 6-month randomized-controlled breakfast trial in 130 healthy, ‘breakfast skipping adolescents (age: 18.95 ± 1.42 years; and BMI 24.49 ± 3.32kg/m²). Throughout the 7-day baseline period, the following procedures were performed. The participants wore a continuous glucose monitor (LibrePro, Abbott) throughout the 7-day period to assess 24-h glucose variability and area under the curve. For 3 days, participants completed visual analog scale questionnaires assessing morning appetite, satiety, and mood in conjunction with food intake via ad libitum food packouts. The Pittsburg Sleep Quality Index Score (PSQI) questionnaire was also completed to assess perceived sleep quality. Pearson correlation analyses were initially performed to identify which biobehavioral factors were associated with daily food intake and macronutrient composition. Of those that were correlated, regression analyses were performed to determine predictors. BMI was included as a covariate. Data were analyzed using R version 4.1.1.

Results: Upon waking fullness, prospective food consumption (PFC), and 24-h glucose variability predicted daily energy intake (R² 0.063; P< 0.05). Upon waking fullness, hunger, desire to eat, and PFC predicted the total amount of food consumed throughout the day (R² 0.040; P< 0.05). Upon waking PFC and 24-h glucose variability predicted daily protein intake (R² 0.032). Upon waking fullness, PFC, hunger, and 24-h glucose variability predicted total carbohydrate intake (R² 0.068) as well as daily sugar intake (R² 0.064). Lastly, 24-h glucose variability predicted total fat intake (R² 0.046).

Conclusions: Although variability in daily glucose concentrations in combination with self-reported appetite predicted daily food intake in older adolescents, these factors accounted for < 10% of the food intake variability. Thus, additional research including a more comprehensive assessment of biobehavioral factors is warranted. Overall, understanding predictors of food intake is important for informing lifestyle interventions in this population.

Funding Sources: National Institutes of Health