Limitations of the various methods for collecting dietary intake data

Area of Science:

• Nutritional Science
• Epidemiology
• Biostatistics

Background:

• Dietary intake assessment is vital for public health monitoring and individual health studies.
• National food consumption data and household surveys are useful for population trends but insufficient for individual-level analysis.
• Accurate individual dietary data is essential for analytical studies linking diet and health outcomes.

Purpose of the Study:

• To critically review various dietary assessment techniques for individuals.
• To quantify the accuracy and errors associated with different methods.
• To highlight the importance of validation methods in dietary research.

Main Methods:

• Review of individual dietary assessment techniques including food records, questionnaires, and biological markers.
• Quantitative estimation of errors (coefficients of variation) for different methods.
• Discussion of validation techniques like the doubly labeled water method and 24-hour urine nitrogen output.

Main Results:

• Food portion estimation errors can reach 50% for foods and 20% for nutrients.
• Twenty-four-hour recall methods show wide coefficients of difference (4-400%) compared to observed intakes.
• Systematic bias and underreporting by subjects are significant potential error sources, observed even in athletes.

Conclusions:

• Dietary assessment methods for individuals contain substantial random and systematic errors.
• Precision can be improved by increasing observations or individuals, but power is lost.
• Independent validation methods (doubly labeled water, urine nitrogen) are essential for accurate dietary assessment in free-living individuals.