2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine, a Carcinogen in High- Temperature-Cooked Meat, and Breast Cancer Risk

Abstract

Although intake of well-done red meat has been associated with an increased risk of breast cancer (1), it is unclear what component(s) of well-done red meat is associated with this risk. Meats cooked to well-done at high temperatures contain heterocyclic amines (HCAs), such as 2-amino-3,4,8-trimethylimidazo[4,5-f]quinoxaline (DiMeIQx), 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx), and 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) (2–9). The amounts of these compounds vary according to cooking technique, temperature, cooking time, and type of meat (10,11). Although PhIP administered orally can induce mammary gland carcinomas in rats (12–14), the association of HCAs with human breast cancer is unclear. Two studies have investigated the association between meat-cooking methods and breast cancer. One study (15) did not obtain information on the degree of meat doneness, from which levels of HCA can be estimated, and the other study (16) used HCA estimates from laboratory-cooked meat samples from one country and subjects from a different country. We used a newly created database [for details, see (10) and (11)] to estimate HCAs in a breast cancer case–control study of 41 836 cohort members participating in the Iowa Women's Health Study (1). Of the selected subjects, 273 case patients (60% of all women with breast cancer diagnosed from 1992 through 1994) and 657 control subjects (75% of randomly selected cohort members who were alive and free of cancer on January 1, 1992, and participated in the 1992 follow-up survey) participated in this study. All subjects completed a self-administered food-frequency questionnaire that included validated questions on frequency of intake and cooking techniques of 15 meat items. The participants reported their usual preference for level of doneness by using a series of color photographs that represented increasing levels of doneness of a hamburger patty and beef steak (four photographs for each) as well as bacon (three photographs) (10,11). We estimated HCA intake by use of our database (10,11,17) and the responses from the food-frequency questionnaire. First, we estimated gram consumption by frequency, portion size, cooking technique, and doneness level. Second, we derived HCA intake by multiplying grams of meat by the HCA concentration measured for each cooking technique/doneness level for that meat type and summed across the three meats. To estimate the doneness levels, we added the gram amounts for “rare/medium,” “well-done,” and “very well-done” steak, hamburger, and bacon. Dietary intakes of each HCA and each type of red meat were lower among the control subjects than among the case patients, with the case patients consuming 50% more PhIP (Table 1). Odds ratios (ORs) were computed with unconditional logistic regression (18), with test of trends based on continuous variables (Table 2). The association between HCA and risk was determined for each HCA individually and with adjustment for the other HCAs. ORs for DiMeIQx, MeIQx, and PhIP are presented in two ways: increments of 10 ng/day and categorically. ORs associated with an increase of 10 ng/day in daily consumption of HCAs provide the relative potency of the different HCAs and make it easier to compare the results between studies with different populations, where the amounts consumed differ. We observed an increased risk of breast cancer across increasing quintiles of PhIP consumption; ORs were 1.0 (referent), 1.1 (95% confidence interval [CI] = 0.6–1.8), 1.2 (95% CI = 0.7–1.9), 1.4 (95% CI = 0.8–2.3), and 1.9 (95% CI = 1.1–3.4) adjusted for the intake of MeIQx and DiMeIQx, with a P for trend of <.001 (Table 2). MeIQx and DiMeIQx were not associated with risk of breast cancer in these analyses. The levels of HCA intakes are measured with error. If measurement errors are the same for case and control subjects, this generally creates a bias toward the null. However, if the errors are different (HCA levels are underestimated for case subjects and overestimated for control subjects), it can lead to bias away from the null. We investigated whether HCAs found in well-done red meats could explain the well-done meat association with breast cancer reported by Zheng et al. (1). In separate models, the ORs were 1.03 (95% CI = 1.01–1.05; P for trend <.001) per 10 ng of PhIP (Table 2) and 1.27 (95% CI = 1.05–1.60; P for trend = .01) per 10 g of very well-done red meat. With the two variables adjusted for each other, the OR remained 1.03 (95% CI = 1.01–1.06; P for trend = .02) per 10 ng of PhIP. In contrast, the OR for very well-done meat decreased to 1.04 (95% CI = 0.80–1.52; P for trend = .73) per 10 g of meat, indicating that PhIP may be more strongly associated with risk than very well-done red meat. When the effects of both PhIP intake and grams of very well-done meat were jointly estimated, the errors of those estimates are larger than when these effects were estimated separately, creating larger CIs. The main adv