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

Using Biomarkers to Evaluate & Compare the Measurement Properties of Nutrition & Physical Activity Assessment Methods

Ross L. Prentice, PhD
Division of Public Health Sciences
Fred Hutchinson Cancer Research Center
and Department of Biostatistics
University of Washington
Seattle, Washington
rprentic@whi.org

What's the problem?

Despite strong biological evidence that supports hypotheses about the relationship of diet and chronic disease, few convincing or consistent associations between dietary consumption and chronic disease risk have emerged from epidemiological studies. One reason for the lack of evidence is that most observational studies rely on self-reported data from food frequency questionnaires (FFQs). Measurement error in self-reported dietary assessment is a significant problem and may hamper the ability of observational studies to detect associations between dietary intake and chronic disease risk.

Attempts to correct for measurement error in FFQs generally have used a second, less error-prone self-report reference instrument, such as a 24-hour dietary recall or food record, to calibrate findings from the original instrument. However, studies show that measurement error is correlated across self-report instruments, suggesting that such calibration may not be ideal.

[Return to top]

How has this research addressed the problem?

Over the past few years, Dr. Prentice and his colleagues, Drs. Marian Neuhouser, Lesley Tinker, Ying Huang, Johanna Lampe, Chongzhi Di, Shirley Beresford, Bette Caan, Linda Van Horn, and other Women's Health Initiative-associated colleagues, have explored, recovery-biomarker-based methods for calibrating self-report data for use in dietary association studies. Recovery biomarkers are specific biologic products that are directly related to intake and not subject to homeostasis or substantial interindividual differences in metabolism. In early work, Dr. Prentice used data from a subset of 544 women who were participating in the Women's Health Initiative Dietary Modification Trial (WHI-DM). The WHI-DM is a very large randomized controlled trial among 48,835 postmenopausal women that examined whether a low-fat dietary pattern reduced the incidence of breast and colorectal cancer, and secondarily, heart disease. The subset of women, who were enrolled in a WHI Nutritional Biomarkers Study (WHI-NBS), completed an FFQ, a doubly labeled water protocol (a biomarker for energy consumption), and a 24-hour urine collection (a biomarker for protein consumption). Collecting the two biomarkers in addition to the FFQ allowed the research team to characterize the measurement error distributions of energy and protein assessed by the FFQs and examine whether the measurement errors differed by the characteristics of the WHI-NBS participants.

The biomarkers used in this study confirmed that the FFQs did, indeed, underreport the consumption of energy and protein and overreport the percent of energy derived from protein. The extent of misreporting also was predicted by participant characteristics, including age, body mass index, and race and ethnicity. Energy underreporting was greater among overweight and obese women and younger women. Protein underreporting showed similar characteristics but to a lesser extent. Dr. Prentice and his team used these results to create regression calibration equations for energy, protein, and percent of energy from protein. The equations corrected for the self-report measurement error.

Dr. Prentice and his team then applied the regression calibration equations in a new study that assessed the association between energy and protein intake and type 2 diabetes risk. The intake data was taken from FFQs completed by women participating in the WHI-DM. The FFQ data of one group of women within the participants was biomarker-calibrated to correct for self-report measurement error. FFQ data from a second group was not biomarker-calibrated.

The research team found that uncalibrated data showed an association with increased risk of diabetes, but that the calibrated data showed a substantially greater risk of diabetes. Body weight, as an indicator of energy intake, appeared to be the major risk factor mediating the association between energy and protein intake and diabetes risk. The use of biomarker calibrated data similarly had a major influence on estimated associations between energy consumption and overall and site-specific cancer incidence, and coronary heart disease incidence.

The results of this study showed that the biomarker calibration approach has real promise for developing more accurate estimates of diet-disease risk associations. However, a major limitation of this approach has been the small number of nutrients for which suitable recovery biomarkers are available.

Currently, Dr. Prentice and his team are conducting a very large feeding study within their ongoing Nutrition and Physical Activity Assessment Study, to develop recovery biomarkers for additional nutrients and foods. The team has recruited 150 women, ages 60 to 80 years, who are already enrolled in the WHI. During the 2-week study, each woman will receive a highly controlled, individualized diet that approximates her usual consumption patterns. Blood and urine samples taken at the beginning and end of the study period will be assessed for biomarkers of energy, protein, sugars, whole grains, meat, fruits and vegetables, and fats and oils. The study also is using doubly-labeled water and indirect calorimetry to assess total energy expenditure and resting energy expenditure, thereby yielding an objective assessment of activity-related energy expenditure. The team will then develop calibration equations for each nutrient/food or physical activity measure that has a suitable recovery biomarker and apply the equations to data from FFQs and physical activity questionnaires.

The result will be more accurate assessments of nutrient/food intakes and physical activity that can be used in the 161,808 women in the WHI study population to examine associations between diet and physical activity and a broad range of clinical outcomes, such as cancer, heart disease, diabetes, frailty, and obesity.

[Return to top]

Significance of the research & results

Despite their accuracy, biomarkers are expensive and labor-intensive to use, thereby limiting their use to small studies. The development of these biomarker-based calibration equations means that the measures can be applied much more broadly in large population studies, thereby greatly enhancing their usefulness.

Reliable information on the health effects of dietary intakes and physical activity is essential for credible recommendations for individuals and for the formulation of national policies that can improve the food, nutrition, and physical activity environment. Dr. Prentice's research is helping to bridge the gap between the strong biological evidence supporting hypotheses about the relationships between diet, physical activity, and chronic diseases and the modest or null risk estimates that are commonly reported from epidemiologic studies on these associations.

[Return to top]

Recent related publications of interest

Neuhouser ML, Di C, Tinker LF, Thomson C, Sternfeld B, Mossavar-Rahmani Y, Stefanick ML, Sims S, Curb JD, Lamonte M, Seguin R, Johnson KC, Prentice RL. Physical activity assessment: biomarkers and self-report of activity-related energy expenditure in the WHI. Am J Epidemiol 2013 Mar 15;177(6):576-85. [View Abstract]

Tinker LF, Sarto GE, Howard BV, Huang Y, Neuhouser ML, Mossavar-Rahmani Y, Beasley JM, Margolis KL, Eaton CB, Phillips LS, Prentice RL. Biomarker-calibrated dietary energy and protein intake associations with diabetes risk among postmenopausal women from the Women's Health Initiative. Am J Clin Nutr 2011 Dec;94(6):1600-6. [View Abstract]

Prentice RL, Mossavar-Rahmani Y, Huang Y, Van Horn L, Beresford SA, Caan B, Tinker L, Schoeller D, Bingham S, Eaton CB, Thomson C, Johnson KC, Ockene J, Sarto G, Heiss G, Neuhouser ML. Evaluation and comparison of food records, recalls, and frequencies for energy and protein assessment by using recovery biomarkers. Am J Epidemiol 2011 Sep 1;174(5):591-603. [View Abstract]

Prentice RL, Huang Y, Kuller LH, Tinker LF, Horn LV, Stefanick ML, Sarto G, Ockene J, Johnson KC. Biomarker-calibrated energy and protein consumption and cardiovascular disease risk among postmenopausal women. Epidemiology 2011 Mar;22(2):170-9. doi: 10.1097/EDE.0b013e31820839bc. [View Abstract]

Prentice RL, Huang Y, Tinker LF, Beresford SA, Lampe JW, Neuhouser ML. Statistical Aspects of the Use of Biomarkers in Nutritional Epidemiology Research. Stat Biosci 2009 May 1;1(1):112-123. [View Abstract]

Prentice RL, Shaw PA, Bingham SA, Beresford SA, Caan B, Neuhouser ML, Patterson RE, Stefanick ML, Satterfield S, Thomson CA, Snetselaar L, Thomas A, Tinker LF. Biomarker-calibrated energy and protein consumption and increased cancer risk among postmenopausal women. Am J Epidemiol 2009 Apr 15;169(8):977-89. [View Abstract]

[Return to top]

Last Modified: 03 Sep 2013