Partitioning macronutrient intake estimates from a food frequency questionnaire.

Area of Science:

• Nutritional epidemiology
• Dietary assessment methodology

Background:

• Food frequency questionnaires (FFQs) are widely used for dietary assessment.
• Assessing the relative validity of FFQs against detailed food records is crucial for nutritional research.

Purpose of the Study:

• To evaluate the relative validity of macronutrient estimates from a quantitative food frequency questionnaire (FFQ) compared to detailed food records.
• To identify sources of disagreement between the two dietary assessment methods.

Main Methods:

• The study utilized data from the 1984-1985 University of Michigan Food Frequency Study.
• 228 participants completed 16 days of food records and a quantitative FFQ.
• Macronutrient estimates (fat, protein, carbohydrate) were compared between methods.

Main Results:

• Significant differences were observed in both estimated group mean macronutrient intake and individual rankings between FFQs and food records.
• Frequency of consumption primarily influenced individual ranking disagreement, while serving size differences affected group mean intake.
• Adjusting for total calorie intake improved agreement in group mean intake but substantial disagreement in ranking persisted.
• Disproportionate reporting of macronutrient proportions by respondents limited the FFQ's ability to accurately rank individuals.

Conclusions:

• Food frequency questionnaires demonstrate limited relative validity for ranking individuals' macronutrient intake compared to food records, even after calorie adjustment.
• Discrepancies in reporting macronutrient proportions are a significant factor contributing to misclassification and reduced validity of FFQs.
• Further refinement of FFQ methodology is needed to improve the accuracy of individual-level dietary assessment.