National Cancer Institute Home at the National Institutes of Health | www.cancer.gov
Please wait while this form is being loaded....

Using Intake Biomarkers to Evaluate the Extent of Dietary Misreporting

Summary: A study assessing dietary measurement error using two self-report instruments (a food frequency questionnaire and 24-hour recall) and unbiased biomarkers of energy and protein intake (doubly labeled water and urinary nitrogen) found that men and women underreported both protein and energy intakes on both instruments. These findings have important implications for nutritional epidemiology and dietary surveillance.

Reference: Subar AF, Kipnis V, Troiano RP, Midthune D, Schoeller DA, Bingham S, Sharbaugh CO, Trabulsi J, Runswick S, Ballard-Barbash R, Sunshine J, Schatzkin A. Using intake biomarkers to evaluate the extent of dietary misreporting in a large sample of adults: the OPEN study. Am J Epidemiol 2003;158(1):1-13.

Nutritional epidemiologists and nutritionists rely on self-reported dietary information when they conduct research and surveillance activities. This information is often derived from food frequency questionnaires (FFQs), which attempts to measure a person's usual intake over a defined time (usually a year). Although it is well known that this information contains errors, the qualitative and quantitative nature of these errors is not well understood. As a result, to evaluate measurement error in FFQs and correct for it, some investigators have incorporated "reference" instruments, such as 24-hour dietary recalls (24HRs), which record in detail everything consumed in the previous 24 hours, or food records. However, several studies using dietary intake biomarkers have found significant bias (generally underreports) in these instruments as well. Understanding the structure and extent of dietary measurement error is essential to interpreting findings from nutritional epidemiology and surveillance studies.

The Observing Protein and Energy Nutrition (OPEN) Study examined dietary measurement error in FFQs and 24HRs by using doubly labeled water (DLW) and urinary nitrogen as unbiased biomarkers of total energy expenditure and protein intake. A total of 484 people aged 40-69 and living in the Washington DC suburb of Montgomery County, MD, participated in the September 1999 to March 2000 study. Over the course of three visits with study staff, OPEN participants completed two FFQs, two 24HRs, and several other health-related questionnaires. They were dosed with doubly labeled water, provided several spot urine samples, completed two 24-hour urine collections, and had their height and weight measured.

Similar to the results of previous studies, OPEN found underreporting of both energy and protein intakes. On the 24HRs, 9% of men and 7% of women were defined as underreporters; on the FFQs, 35% of men and 23% of women were underreporters. On average, compared to their total energy expenditure, men underreported energy intake by 12-14% on 24HRs and 31-36% on FFQs; compared to the protein biomarker, they underreported protein intake by 11-12% on 24HRs and by 30-34% on FFQs. Compared to their total energy expenditure, women underreported energy intake by 16-20% on 24HRs and 34-38% on FFQs; compared to the protein biomarker, they underreported protein intake by 11-15% on 24HRs and by 27-32% on FFQs.

Several trends also emerged from the data:

  1. as total energy expenditure or protein intake increased, so too did underreporting;
  2. underreporting tended to increase with body mass index; and
  3. underreporting of energy was somewhat greater than that of protein, suggesting greater underreporting of fat, carbohydrate, and alcohol.

These findings have important implications for nutritional epidemiology and dietary surveillance, particularly for the use of these instruments to measure intake of key dietary components, such as fat and energy. The authors conclude that work to further elucidate the issues examined in this study will help support the nutritional epidemiology and dietary surveillance research that is central to answering questions about diet's relationship to disease.

Learn more about the Risk Factor Monitoring and Methods Branch's activities related to biomarkers.

<< Previous Summary | Next Summary >>

Last Modified: 11 Apr 2014