Food Frequency Questionnaire at a Glance

Purpose

To obtain frequency and, in some cases, portion size information about food and beverage consumption over a specified period of time, typically the past month or year.

Description

A food frequency questionnaire (FFQ) consists of a finite list of foods and beverages with response categories to indicate usual frequency of consumption over the time period queried. To assess the total diet, the number of foods and beverages queried typically ranges from 80 to 120.

Usual portion size can be asked separately for each food and beverage. Alternatively, portion size can be combined with frequency information by asking respondents to translate usual consumption amount to number of specified units (e.g., How often do you eat a ½ cup of rice?). Some questionnaires include portion size images in an attempt to enhance reporting [glossary term:] accuracy.

In addition to foods and beverages, FFQs often ask about the frequency of intake and dosages of commonly consumed [glossary term:] dietary supplements.

FFQs are usually self-administered; interviewer-administration is done occasionally, for example, when literacy is low.

Completing a questionnaire aimed at capturing the total dietary intake usually requires 30 to 60 minutes.

An FFQ is sometimes loosely referred to as a [glossary term:] diet history. However, the term diet history can refer generally to any instrument that asks about diet in the past (Learn More about Diet History). It also is used to refer specifically to a meal-based FFQ or the Burke diet history, which includes a food frequency cross-check component, and its various adaptations.

To learn more about brief frequency questionnaires aimed at capturing particular aspects of the diet, see Screener Profile.

The National Cancer Institute’s Diet History Questionnaire is a freely available FFQ.

Utility of Data

• FFQs provide information on consumption of queried foods and beverages over the specified period.
  • Depending on the breadth of items queried, data can be used to assess total dietary intake and/or particular aspects of diet.
  • Depending on whether portion size is determined, information may represent either usual frequency of consumption only or total amount usually consumed. The term "food propensity questionnaire" is sometimes used to define an FFQ that does not include portion size.
  • FFQs may be better than [glossary term:] short-term instruments (e.g., 24-hour dietary recall [24HR]) at assessing intake of [glossary term:] episodically consumed foods because they attempt to directly capture usual intake over a period of time.
• Data can be linked to a database to yield information about total dietary intake.
  • If data are linked to a nutrient composition database, nutrient intake can be determined (Learn More about Food Composition Databases for Food Frequency Questionnaires and Screeners). However, [glossary term:] total nutrient intakes cannot be estimated unless dietary supplement intake data also are included. If not linked to a nutrient composition database, frequency data can be used as a qualitative indicator of diet (Learn More about Dietary Supplements and Estimating Total Nutrient Intakes).
  • If data are linked to a database that translates foods and beverages into food group equivalents, such as the Food Patterns Equivalents Database, food and beverages reported can be disaggregated into their component ingredients, which can then be converted to equivalent amounts of relevant guidance-based food groups (For more information, read a factsheet on FPED products and associated data files or its application to dietary analysis.). This provides information about the consumption of food groups of particular interest, such as the total amount of fruits, vegetables, or added sugars. The frequency of food groups consumed or pattern of food groups consumed can be used as indicators of particular aspects of diet.
• FFQs are sometimes used to describe a population's intake in [glossary term:] cross-sectional studies, although this is not recommended because of their large [glossary term:] systematic error, or [glossary term:] bias (a type of [glossary term:] measurement error) (see Choosing an Approach for Dietary Assessment and Key Concepts about Measurement Error).
• FFQs can be used to examine [glossary term:] associations between diet and health or other variables (see Choosing an Approach for Dietary Assessment).
  • FFQs are commonly used as the [glossary term:] main dietary assessment instrument in large [glossary term:] prospective studies.
  • Because of known [glossary term:] systematic error in an FFQ, a less biased short-term instrument (e.g., 24HR) may be administered in an [glossary term:] internal calibration sub-study to improve estimates of intake from an FFQ through [glossary term:] regression calibration (see Learn More about Regression Calibration), allowing for a more accurate estimate of the association between diet and another variable. Alternatively, data from an external source (called an [glossary term:] external calibration study) can be used (Learn More about Calibration).
• FFQs and screeners are the only practical methods for capturing diet in [glossary term:] retrospective case-control studies.
• FFQs are sometimes used to evaluate the effectiveness of interventions to change diet (see Choosing an Approach for Dietary Assessment). However, because of concern about potential [glossary term:] differential response bias due to the tendency of the intervention group to [glossary term:] misreport their diets to a greater extent than the control group, use of an FFQ (or any self-report dietary instrument) as the sole evaluation method should be carefully considered.

Limitations of Data

• Contains systematic error (see Key Concept about Measurement Error). This error can be at least partially mitigated through appropriate statistical modeling, using a less biased dietary assessment method as a [glossary term:] reference instrument, and through [glossary term:] energy adjustment.
• Detailed information about food preparation, specific food and beverages consumed, and brands, and contextual information about intake (e.g., which foods and beverages are consumed at the same meal) is lacking.
• Because an FFQ is composed of a pre-specified food list, any single FFQ may not reflect the eating patterns of a given population. Thus, the performance of a particular FFQ in a particular population may not reflect its performance in a different population.

Salient Features Compared to Other Self-Report Instruments

For an at-a-glance comparison of the major features of self-report instruments for assessing diet, including FFQs, see the Comparing Dietary Assessment Instruments table.

• Captures past [glossary term:] long-term diet (in contrast to 24HR and food record).
• Usually self-administered (in contrast to interviewer-administered 24HR).
• Relies on [glossary term:] generic memory (in contrast to 24HR, which relies on [glossary term:] specific memory, and food record, which does not rely on memory).
• Not affected by [glossary term:] reactivity (in contrast to food record) (Learn More about Reactivity).
• Requires ability to perform cognitively complex memory and averaging tasks (in contrast to 24HR and food record).
• Its major type of measurement error is [glossary term:] systematic versus [glossary term:] random (in contrast to 24HR).