Portion size has sustained effects over 5 days in preschool children: a randomized trial

Abstract

Background: Although short-term studies have found that serving larger portions of food increases intake in preschool children, it is unknown whether this portion size effect persists over a longer period or whether energy intake is moderated through self-regulation.

Objectives: We tested whether the portion size effect is sustained in preschool children across 5 consecutive days, a period thought to be sufficient for regulatory systems to respond to the overconsumption of energy.

Methods: With the use of a crossover design, over 2 periods we served the same 5 daily menus to 46 children aged 3-5 y in their childcare centers. In 1 period, all foods and milk were served in baseline portions, and in the other period, all portions were increased by 50%. The served items were weighed to determine intake.

Results: Increasing the portion size of all foods and milk by 50% increased daily consumption: weighed intake increased by a mean ± SEM of 143 ± 21 g/d (16%) and energy intake increased by 167 ± 22 kcal/d (18%; both P < 0.0001). The trajectories of intake by weight and energy across the 5-day period were linear and the slopes did not differ between portion conditions (both P > 0.13), indicating that there were sustained increases in intake from larger portions without compensatory changes over time. Children differed in their response to increased portions: those with higher weight status, lower ratings for satiety responsiveness, or higher ratings for food responsiveness had greater increases in intake from larger portions (all P < 0.03).

Conclusions: This demonstration that preschool children failed to adjust their intake during prolonged exposure to larger portions challenges the suggestion that their self-regulatory behavior is sufficient to counter perturbations in energy intake. Furthermore, overconsumption from large portions may play a role in the development of overweight and obesity, as the magnitude of the effect was greater in children of higher weight status. This trial was registered at www.clinicaltrials.gov as NCT02963987.

Keywords: eating behavior; energy intake; obesity; portion size; preschool children.

Copyright © American Society for Nutrition 2019.

Figures

FIGURE 1

Mean ± SEM cumulative (A) weight consumed and (B) energy intake over 5 consecutive days during which the portion sizes of all foods and milk served to 46 preschool children were either 100% or 150% of the baseline amounts. Daily meals consisted of morning snack, breakfast, lunch, afternoon snack, dinner, and evening snack, each of which is represented by a point on the graph. Through the use of a mixed linear model, a significant portion size effect was found for cumulative intake of food and milk starting at afternoon snack on day 1 (indicated by *; P = 0.0002) and for cumulative energy intake starting at lunch on day 1 (indicated by **; P = 0.001). The increase in intake by weight and energy from larger portions accumulated over time.

FIGURE 2

Mean ± SEM daily intake across time according to the (A) weight and (B) energy consumed by 46 preschool children during which the portion sizes of all foods and milk served were either 100% or 150% of the baseline amounts. The trajectories from a random coefficients model were linear for both portion conditions, and the slopes did not differ between conditions for intake by either weight (P = 0.13) or energy (P = 0.14). This indicates that increasing the portion sizes led to a significant and sustained increase in intake by weight (P < 0.0001) and energy (P < 0.0001) across all 5 d.

FIGURE 3

Influence of the energy density of 82 food items on the magnitude of the portion size effect for each item. During different 5-d periods, 46 preschool children were served portion sizes of all foods and milk in either 100% or 150% of baseline amounts. The portion size effect for each food item was defined as the difference in the amount consumed from the 2 portions, expressed as a percentage of the difference in the amount served. The energy density of the items significantly affected the magnitude of the portion size effect for that item, according to a linear mixed model with repeated measures (P < 0.0001). For every 1 kcal/g increase in item energy density, there was a mean 7.4 ± 1.0% increase in the portion size effect.

FIGURE 4

The influence of children's body size, as assessed by sex-specific BMI-for-age percentile, on the relation between portion size and mean daily energy intake. In different 5-d periods, 46 preschool children were served 100% and 150% of baseline portions of all foods and milk. Results from analysis of covariance through the use of mixed linear models with repeated measures showed that the slope of the regression line for the 150% portions was significantly greater than the slope for the 100% portions (adjusted P = 0.0005), even after adjustment for energy requirements. Thus, children with a higher BMI-for-age percentile showed a greater response to increased portions than children with a lower BMI-for-age percentile.

FIGURE 5

Influence of scores on the (A) satiety responsiveness, (B) food responsiveness subscales, and (C) enjoyment of food of the Children's Eating Behaviour Questionnaire on the relation between portion size and intake in 46 preschool children who were served 100% and 150% of baseline portions of all foods and milk in different 5-d periods. Results from analysis of covariance through the use of mixed linear models with repeated measures showed that for all 3 subscales, the slope of the regression line for the 150% portions was significantly steeper than the slope for the 100% portions (all adjusted P < 0.03). Thus, children with higher scores for satiety responsiveness had smaller increases in intake from larger portions than did children with lower scores. Also, children with higher scores for food responsiveness and enjoyment of food had greater increases in intake from larger portions than did children with lower scores.