Background: Meat intake is associated with risk of certain cancers, but the biological mechanism underlying this association has yet to be fully established. Exposure to polycyclic aromatic hydrocarbons (PAH) and heterocyclic amines (HCA), two classes of carcinogens formed during the cooking of meat, may contribute to this relationship. Epidemiological studies of these specific dietary chemicals and cancer suggest an association, but results have been inconsistent. The standard practice is to use food frequency questionnaires (FFQ) and carcinogen databases to estimate dietary intake of these chemicals, yet little is known about how these methods compare to other available methods, particularly biological markers.
Methods: We compared exposure estimates for the PAH benzo[a]pyrene (BaP) and pyrene and the HCA 2-amino-1-methyl-6-phenylimidazo[4,5b]pyridine (PhIP) derived from the following three accepted approaches: a meat-specific FFQ administered twice and combined with a dietary carcinogen database; diet diaries longitudinally collected three times, combined with the same database; and urinary PhIP and 1-hydroxypyrene-glucuronide (1-OHPG) measured in overnight urine samples longitudinally collected at three times concurrent with the diaries. This study was completed in 54 non-smoking controls from an ongoing case-control study of colorectal adenoma in Baltimore, MD.
Results: For PAH, the strongest between-method association was observed between the FFQ-database derived BaP estimates and urinary 1-OHPG measurements (p<0.001). The diary-database derived BaP was also statistically significantly associated with urinary 1-OHPG after adjusting for cotinine (p=0.03). For HCA, the only statistically significant between-method association was between the diet diary-database derived PhIP estimates and urinary PhIP, when restricting the analyses to values above zero or the detection limit (p=0.02). No association was observed between the other methods.
Conclusions: The three different approaches for estimating dietary exposure to PAH were moderately correlated, providing confidence in the different methods. In contrast, there was generally a lack of association between the PhIP exposure estimates produced by the different methods. This may be due to the small number of sampling days or short-comings of the FFQ and the database. This research highlights the need to improve or validate dietary exposure assessment methods. Improved methods could be used to clarify the meat-cancer link, helping to shape new policies and guidelines on behavioral changes.
Nicole Cardello Deziel received her bachelor of science degree in public health at the University of North Carolina at Chapel Hill and her master of health science at the Johns Hopkins Bloomberg School of Public Health. Nicole recently completed her doctorate in environmental health engineering at the Bloomberg School. Her research was funded, in part, by a Center for a Livable Future Student Innovation Grant. Currently, Nicole is a senior environmental health scientist at Westat, the coordinating center for the National Children’s Study, and is also an instructor at Georgetown University. Other work experiences include research positions at the Environmental Protection Agency (EPA) as well as teaching assistantships for graduate and undergraduate environmental and occupational health courses. Nicole served on the EPA’s Voluntary Children’s Chemical Evaluation Program Peer Consultation Panel as well as on an expert panel at an EPA workshop, Characterizing and Presenting Chemical Exposure Assessment Results.
