Skip to Main Content

An official website of the United States government

Principal Investigator
Name
Marilyn Tseng
Institution
Fox Chase Cancer Center
Email
About this CDAS Project
Study
PLCO (Learn more about this study)
Project ID
2008-0009
Initial CDAS Request Approval
Mar 27, 2008
Title
Measurement of Dietary Patterns in Nutritional Epidemiology
Summary
The use of principal components analysis (PCA) to describe and quantify patterns of food intake has become increasingly common in nutritional epidemiologic studies of cancer. A dietary pattern approach is appealing because nutrients and foods are consumed in complex combinations in our diets. Such an approach allows for the quantification of aggregate risk associated with common dietary patterns, and for more intuitive application to public health recommendations. The approach has also been criticized, however, for its lack of standard methodology and, accordingly, its reliance on subjective decisions and a posteriori interpretations. Whether results observed in previous studies on dietary patterns are robust to changes in methodological procedures, therefore, is uncertain. The proposed work will: (1) examine the robustness of dietary pattern analyses to differences in methodological procedures; and (2) examine associations of dietary patterns with all-cause and cause-specific mortality (cardiovascular, total cancer, and diabetes). We propose to use Diet History Questionnaire (DHQ) data from the PLCO cohort to address PCA-related methodological issues. Following consideration of these methodological issues, we will examine dietary patterns identified in the PLCO cohort in relation to all-cause mortality and mortality due to cardiovascular disease, cancer, and diabetes. We anticipate that our findings will strengthen the field of dietary pattern research and serve as a valuable resource for future research relating dietary patterns to other health outcomes in the PLCO and other study populations.
Aims

1. To examine the robustness of dietary patterns to differences in methodological procedures
We expect to identify at least two primary dietary patterns – a meat-starch-based dietary pattern, and a vegetable-based dietary pattern – in addition to culturally influenced region-specific pattern (e.g., southern).
2. To examine associations of dietary patterns with all-cause cardiovascular, cancer, and diabetes mortality
We expect that a meat/starch-based (‘western’) diet is associated with while a vegetable-based dietary pattern is inversely related to total and cause-specific mortality.

Collaborators

PLCO Center Representative, to be identified (2)
Barry Graubard (Division of Cancer Epidemiologyand Genetics)
Marilyn Tseng (Fox Chase Cancer Center)
Regina Ziegler (Division of Cancer Epidemiologyand Genetics)