(LB01-28-26) The Relative Influence of Determinants of Meal Energy Intake Across Multiple Studies

Objectives: Studies demonstrate the robust influence of energy density, portion size, eating rate, and variety on energy intake at meals. Much research has also focused on the influence of macronutrients (% energy from carbohydrates, protein, or fat) as determinants of consumption. To compare the relative contribution of each of these factors on energy intake at meals, we combined data from 10 controlled studies conducted in our laboratory.

Methods: The data is comprised of measured intake from 355 participants across 1478 laboratory lunch meals (46% women; BMI Mean: 23.7 kg/m², Range: 18.5-39.1; age 26.2 y, 18-68). Across studies, meals varied in energy density served (1.4 kcal/g, 1.0-2.1), portion size (803 g, 400 -1213), eating rate (40.3 g/min, 20.3 – 205.1), variety (single food n=469, or multiple foods n=1009), proportion of energy from protein (%PRO; 19%, 11-28), carbohydrates (%CHO; 49%, 31-64), and fat (%FAT; 31%, 20-56), and proportion of meal weight from fiber (%Fiber; 1.8%, 0.4-10.5). Hierarchical and backwards regression analyses were used to assess the influence and variance in energy intake explained by each variable. The Bonferoni adjusted significance level was p< 0.008.

Results: The base model accounted for sex, BMI, age, and pre-meal hunger and liking ratings (all p< 0.001, model R²=0.15). Across meals, higher energy density, portion size, eating rate, and variety, and lower %PRO were each associated with greater energy intake (all p< 0.001), and explained 18%, 14%, 3%, 3%, 3% of the variance respectively. Thus, together energy density, portion size, eating rate, variety, and %PRO explained over 40% of the variance in energy intake. Adding %CHO, %Fat, and %fiber to the model revealed that these factors were not significantly associated with energy intake (all p >0.07) and accounted for < 1% variability. This theory-driven hierarchical regression approach aligned with the data-driven backwards regression, which retained energy density, portion size, eating rate, variety, and %PRO as significant predictors of energy intake.

Conclusions: Overall, this secondary analysis of over 1400 laboratory meals confirms that energy density, portion size, eating rate, and variety are primary determinants of energy intake at meals, and additionally supports the influence of protein on energy consumption.

Funding Sources: NIH R01 DK059853 and R37 DK039177 (PI: Rolls).