Related Experiment Videos
Serving size issues in estimating dietary exposure to food substances
C J Lewis1, A M Beloian, E A Yetley
1Clinical Nutrition Branch, Food and Drug Administration, Washington, D.C. 20204.
Journal of the American Dietetic Association
|December 1, 1988
Summary
Estimating dietary intake using mathematical models can be flawed. This study found that standard serving sizes significantly underestimate cola consumption for frequent, high-volume drinkers, highlighting database biases.
Area of Science:
- Nutrition Science
- Dietary Exposure Assessment
- Food Consumption Databases
Background:
- Mathematical models are crucial for estimating dietary exposure to food substances.
- Assumptions regarding consumption frequency, serving size, and their relationship are vital for model accuracy.
Purpose of the Study:
- To examine assumptions about cola consumption frequency and serving size.
- To assess the impact of these assumptions on dietary exposure models.
- To identify potential biases in food consumption databases.
Main Methods:
- Analysis of cola beverage intake data from two national surveys (NFCS 1977-78 and NHANES II).
- Examination of the relationship between consumption frequency and serving size.
- Comparison of serving sizes across different demographic groups and surveys.
Main Results:
- Cola consumption frequency explained 55% to 83% of intake variability.
- Serving size was not linearly related to consumption frequency.
- Upper-level consumers had both higher frequency and larger serving sizes.
- The model underestimated intake for high-level consumers by over 1% (p < .01).
- Significant differences in cola serving sizes were observed by sex and between surveys.
Conclusions:
- Standard serving sizes in dietary models can lead to significant underestimation of intake.
- Systematic biases in food consumption databases can cause considerable estimation differences.
- Careful consideration of database characteristics is essential for accurate dietary exposure assessment.