Short Oral Presentation Nutrition Society of Australia 50th Annual Scientific Meeting 2026

Children’s self-selected portion sizes: associations with liking, familiarity, perceived healthiness and anthropometric measures (146209)

Zhen Yu 1 , Georgie Russell 1 , Kathleen E. Lacy 1 , Sze-Yen Tan 1
  1. Deakin Institute for Physical Activity and Nutrition, School of Exercise and Nutrition Science, Deakin University, Geelong, Victoria, Australia

The ability to self-select appropriate portion sizes across different foods, including adjusting portions according to energy density and hunger levels, is important for maintaining diet quality, energy balance and attendant weight outcomes. (1) Yet, little is known about the factors associated with children decide on food portion sizes, and how portion size selection relates to adiposity. This study aimed to examine factors associated with portion size selection of cereals and snacks among children aged 5-12-years. A total of 568 children who participated in the Children’s Eating and Weight Study (CHEWS 2024) were included in this cross-sectional study. Children’s 1) objectively measured weight, height and waist circumference, 2) self-selected portions of six test foods varying in healthiness (4 breakfast cereals and 2 snacks), and self-rated familiarity, perceived healthiness, liking, and 3) self-reported hunger levels were assessed. Linear regression models were used to examine associations of food perceptions and anthropometric measures with selected portion size, while two- and three-way interactions between food familiarity and perceptions were examined using multilevel mixed-effects linear regression models with current hunger levels as a covariate. Children’s portion selections were associated with food type, familiarity and health perceptions but not anthropometry. For cereals, larger portions were consistently chosen when familiarity was high, and the magnitude of differences varied by perceived healthiness (unhealthy +275 kJ, neutral +190 kJ, healthy +103 kJ; all p < 0.01). For snacks, larger portions were only selected when foods were both familiar (moderate +330 kJ, high +782 kJ; all p < 0.01) and perceived as healthy, with liking predicting larger portions (neutral liking +516 kJ, like +1017 kJ; both p < 0.01) but only for healthy snacks. No significant three-way interactions were found, and portion size was not consistently associated with BMI z-scores or waist circumference. In terms of children’s hunger/fullness levels at the time of participation, almost half reported feeling “Just right,” with fewer children indicating they were either very hungry or already full. Children’s portion size selection was associated with familiarity, liking, and perceived healthiness, but did not predict adiposity. Distinct patterns between cereals and snacks highlight the need for research beyond breakfast foods, using repeated measures across a variety of foods and contexts. This suggests that children’s portion size decisions may be shaped by their prior experiences with foods, as familiarity, liking, and perceived healthiness appear to influence the amount they choose to serve. (2) These potentially modifiable factors may therefore represent useful targets for future interventions seeking to influence children’s portion size selection.

  1. Russell CG & Russell A (2025) Nutr Rev 83(9), 1784–1799
  2. McCrickerd K (2018) Nutr Bull 43(4), 388-99