Related Experiment Video
Updated: Sep 18, 2025
Concept Development and Use of an Automated Food Intake and Eating Behavior Assessment Method
Published on: February 19, 2021
Automated registration of snacking behavior in 3- to 7-year-Old children using SnackBox technology
Femke J de Gooijer1, Marlou Lasschuijt2, Edith J M Feskens2
1Division of Human Nutrition and Health, Department of Agrotechnology and Food Sciences, Wageningen University and Research, P.O. Box 17, Wageningen, 6700 AA, the Netherlands; OnePlanet Research Centre, Bronland 10, Wageningen, 6708 WE, the Netherlands.
Insights
The SnackBox sensor shows potential for tracking children's snacking habits, offering insights into dietary patterns. While not yet precise enough for exact intake measurement, it could aid in understanding early eating behaviors.
Area of Science:
- Pediatric Nutrition
- Dietary Assessment Technology
- Behavioral Science
Background:
- Accurate measurement of children's eating habits is crucial for effective dietary guidance and understanding diet-health links.
- Traditional parent-reported methods for assessing children's food intake are often biased and labor-intensive.
- Emerging sensor technologies offer a promising avenue for passively capturing children's dietary behaviors.
Purpose of the Study:
- To evaluate the efficacy of the SnackBox, a novel sensor device, for measuring children's dietary intake.
- To assess the accuracy and reliability of the SnackBox in a home environment for children aged 3-7 years.
- To explore snacking patterns and the impact of different vegetable offering strategies.
Main Methods:
- The study involved 43 children (ages 3-7) using the SnackBox sensor at home over four weeks.
- The SnackBox features three weighing stations to automatically record the timing and quantity of snack consumption.
- Vegetable intake data from the SnackBox were compared against pre-weighed portions and post-consumption leftovers (gold standard).
Main Results:
- The SnackBox demonstrated a strong correlation (rs = 0.83) and good reliability (ICC = 0.83) in measuring vegetable intake.
- A mean error of 12% was observed, with higher errors in younger children and for specific vegetables.
- Time-series analysis revealed variations in snacking patterns among children and across different vegetable offering strategies.
Conclusions:
- The SnackBox shows promise as a low-input tool for assessing children's snacking behaviors.
- Current error rates necessitate further refinement for precise dietary intake quantification.
- The technology holds potential for improving the understanding of early-life dietary habits and informing targeted nutritional interventions.
Background:
Accurate measurement of children's eating habits is essential for dietary guidance and understanding diet-health relationships. Traditional methods rely on parent reports, which are prone to bias and burdensome. Sensor technologies offer promise for passively capturing children's dietary behaviors. This study evaluates such a sensor, the SnackBox, in children aged 3-7 years at home.
Methods:
Forty-three children (23 girls, mean age 5.7 ± 1.1 years) used the SnackBox with parental assistance. This sensor device has three weighing stations to store snack containers or beverages bottles and automatically measures timing and quantity of consumption. Children's snack vegetable intake (cucumber, carrot, bell pepper) was recorded at home on eight school-day afternoons over four weeks. SnackBox data were compared to pre-weighed portions and post-consumption left overs (gold standard). Time-series data were analyzed to explore snack patterns across children and across different vegetable offering strategies.
Results:
The SnackBox recorded slightly higher average vegetable intake (111 ± 56 g) than the reference method (96 ± 49 g), with a mean error of 12 % (6 g) and mean absolute error of 38 % (28 g). A strong correlation (rs = 0.83, p < .001) and good reliability (ICC = 0.83, 95 % CI: 0.79-0.86) were observed. Error increased with higher intake, in younger children, and varied by vegetable. Case study results suggest differences in snacking patterns (e.g., eating duration, breaks, switching behaviors) between children and offering strategies.
Conclusion:
The SnackBox shows promise for measuring snacking behaviors in children with minimal input, though current error rates limit its use for precise intake assessment. With refinement, it could potentially enhance understanding of early dietary habits and support tailored interventions.
