Skip to Main Content
An official website of the United States government
Epidemiology and Genomics Research Program

Methodology - NCS Dietary Assessment Literature Review

Overview

This review examines studies conducted mainly in industrialized, developed countries and published in English between 1982 and December 2003 for pregnant or lactating women, infants, toddlers, and preschoolers. For school age children and adolescents, our search built on a recent review of the literature through 2000 (1) and focused primarily on literature published since that time. Studies were identified in National Library of Medicine PubMed database and the CABI Publishing Nutrition Abstracts and Reviews database using a number of comprehensive search strings containing relevant key words. A number of key word search strings were developed to search the databases for relevant documents.

In addition to the key word search strings, author searches were conducted in both databases to identify relevant publications by key investigators with expertise in maternal and child nutrition. These searches were supplemented by cross-referencing from reference lists from reviewed literature. A Reference Manager Database was developed to manage the search and review process. In addition, specific criteria were used to determine whether to include a citation in the literature review, and a careful process for reviewing each article was followed.

Measurement of Validity

Validity is the ability of an instrument to measure what it is intended to measure. Most dietary assessment methods are intended to measure usual or customary food and/or supplement intake over a defined period of time. Because true usual diet is difficult if not impossible to measure, investigators assess relative or criterion validity. Relative validity compares a new measurement method with one or more established methods believed to have a greater degree of demonstrated or face validity (2). The measurement error in the new instrument or method is examined and calibrated with the reference method. This type of validity assessment will fail to detect systematic reporting error or bias if both the new and reference method have correlated error (3). Alternatively, the new method or instrument can be validated against an independent, external criterion reference measurement, such as a biomarker of intake. The doubly labeled water (DLW) method to measure total energy expenditure is an example of a biomarker that can be used to independently validate self-reported energy intake. The error in the DLW method is independent of self-reported intake error, thus allowing true reporting bias to be detected (3,4).

This review examines studies in which the relative or criterion validity of one diet or supplement assessment method is evaluated by comparisons with measurements obtained from a reference method. The reference methods vary by study and target age group and include established dietary assessment methods, biological markers of habitual intake, and techniques such as direct observation of intake.

Definitions and Abbreviations

Definitions of and abbreviations for dietary assessment methods and reference methods discussed in this report.(5,6)

Food Record (FR)

Food records are used to record food intake at the time of consumption, over a number of days that are not necessarily sequential. Most studies ask respondents to enter the information on hard copy form, although tape-recording, bar-coding, and electronic weighing also have been used to collect descriptive and quantity information.

Weighed FR: The respondent weighs on a small scale all food and beverages consumed.

Estimated FR: The respondent estimates all food consumed using household measures or portion size estimating aides.

Diet History (DH) Diet History questionnaires are a retrospective assessment method ascertaining a respondent's "usual" food intake by collecting descriptive detail and amount information about each food. DHs may include questions on meal patterns, lists of common foods and groups of generic food. DH questionnaires are typically administered by a trained interviewer either in-person or by telephone, but also can be self-reported.
24-Hour Recall (24HR) The 24HR is a retrospective assessment method in which an interviewer prompts a respondent to recall and describe all foods and beverages consumed in the preceding 24 hours or the preceding day. The interview may be conducted in-person or by telephone and may be paper and pencil or computer assisted. Portion size estimating aides assist the respondent to recall amounts consumed. The methodology for conducting the 24HR has evolved during the last two decades. Among the methods reported are: 3-pass method, 5-pass method, U.S. Department of Agriculture protocol, University of Minnesota protocol, Bogalusa Heart Study protocol.
Food Frequency Questionnaire (FFQ)

The food frequency questionnaire is a retrospective method asking respondents to report their usual frequency of consumption of each food from a list of foods for a specific period (several months or a year). Food lists vary by the purpose of the study and study population. Frequency of consumption categories also vary by questionnaire but usually include per day, week, or month.

Semi quantitative FFQ: In this type of FFQ, portion size information is collected; portion sizes are specified as standardized portions or choice (range of portions).

Non-quantitative FFQ: Portion size information not collected.

NCI Health Habits and History Questionnaire (HHHQ): Semi-quantitative FFQ developed at the National Cancer Institute under the direction of Gladys Block.

Harvard FFQ (HFFQ): FFQ developed at Harvard University by Walter Willett and colleagues. Portion size information is included as part of the food item rather than as a separate listing.

NCI Diet History Questionnaire (DHQ): Semi-quantitative FFQ, using an embedded question approach, developed at the NCI under the direction of Amy Subar and Fran Thompson (7,8).

Propensity Questionnaire Comprehensive FFQ-type questionnaire designed to supplement other dietary assessment method. Information on portion size information is not collected. May provide information on infrequently consumed foods (9).
Direct Observation (DO) Intakes are watched and recorded by trained observers.
Doubly Labeled Water Method (DLW) for total energy expenditure (TEE) The DLW method is used to measure energy expenditure in free-living subjects. This method involves the administration of water containing enriched quantities of the stable isotopes deuterium (2H) and oxygen-18 (18O). The label of "doubly" labeled comes from the fact that both the hydrogen and oxygen are labeled. The oxygen-18 is eliminated from the body in the form of carbon dioxide (C18O2) and water (H218O), and the deuterium is eliminated in water (2H2O). The difference in elimination rate between these two isotopes is a measure of carbon dioxide production. Carbon dioxide production can then be used to calculate energy expenditure by use of standard equations for indirect calorimetry (3). The DLW method has been shown to be accurate to 1%, with within-subject precision of 5 to 8% (10). Because the method is expensive and analysis requires specialized, expensive equipment, it cannot be considered routine. However, the method is widely available and is being applied to dietary assessment validations with sample sizes ranging from 20 to 500 (10).
Test Method (TM) Dietary assessment method being validated.
Reference Method (RM) Method against which the TM is being compared and validated.

Appendix to Methodology

Review Process
Explains how the Westat master's level registered dietitians carry out the citation review process.

Citation Relevancy Criteria
The citation review initial screen criteria for title and/or abstract necessary to order the full text article.

Reference Manager Database
Reference Manager is a commercially available citation management system. It was used to track all of the references - from identification to article retrieval to evaluation.

Completed Search Strings
Strings used to search for citations, which are imported into Reference Manager database.