Date of Award

Fall 2019

Document Type

Open Access Dissertation


Epidemiology and Biostatistics

First Advisor

Susan E. Steck


Introduction: Advanced glycation end-products (AGEs) are implicated in the pathogenesis of chronic diseases including cancer. AGEs are produced endogenously and are also consumed from foods. High amounts of AGEs are found in animal products and processed foods, and AGE formation is accelerated during cooking at high temperatures. Using data from the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO), we assigned and quantified NƐ-carboxymethyl-lysine (CML)-AGE content in food and investigated the association between dietary AGE intake and breast cancer (BrCa) risk. We assessed whether the risk was modified by race/ethnicity, obesity status, fruit and vegetable intake, BrCa invasiveness and tumor subtype. We also investigated the association between post-diagnosis CML-AGE intake and mortality from all-causes, BrCa and cardiovascular disease (CVD) using data from the Women’s Health Initiative (WHI). Methods: The PLCO enrolled women aged 55 to 74 years into a randomized controlled trial. Our study sample included only women enrolled in the intervention arm who were cancer-free at baseline and completed a baseline questionnaire and dietary questionnaire (DQX). CML-AGE values were assigned and quantified to foods in the DQX using a published AGE database. The WHI enrolled postmenopausal women aged 50 to 79 years into a clinical trial and observational study (OS). Our study sample included women diagnosed with invasive BrCa during follow-up who completed a food frequency questionnaire (FFQ) post-diagnosis, had energy intakes between 600 kcal/day and 5000 kcal/day and had AGE data in the WHI database. Descriptive analyses estimated means and percentages while Pearson correlation estimated correlation coefficient of CML-AGE intake with dietary factors linked to AGEs. In the PLCO, Cox proportional hazards modelling was used to estimate the hazard ratios (HRs) and 95% confidence interval (CIs) of BrCa and was stratified by race/ethnicity, obesity status, fruit and vegetable intake, invasiveness of disease and hormone receptor status to assess effect modification. In the WHI, HRs and 95%CI for risk of mortality from all-causes, BrCa and CVD were assessed and was stratified by fruit and vegetable intake and hormone receptor status. Results: After a median 11.5 years of follow-up, 1,592 women were diagnosed with BrCa in the PLCO. The average CML-AGE consumption among all the women was 6,105 ± 2691 kilounits (KU)/1000 kcal and was highest among non-Hispanic black (NHB) compared to non-Hispanic white (NHW) and Other race/ethnicity.

Significant positive correlations were observed between CML-AGE intake and dietary sources of animal protein, monounsaturated fatty acids, polyunsaturated fatty acid and saturated fatty acids while the correlations between CML-AGE and fructose, carbohydrates and plant protein were weaker. There was an increased risk of BrCa across levels of CML-AGE intake. Higher CML-AGE intake was associated with increased risk of BrCa among NHW but was not statistically significant for other race groups. Increased CML-AGE intake was associated with increased risk of in situ and hormone receptor positive BrCa. In the WHI, after a median 15.1 years of follow-up, 642 deaths were reported including 198 BrCa-specific and 129 CVD-specific deaths. The average post- diagnosis CML-AGE consumption among all women was 6,647 ± 2279 KU/1000 kcal and ranged from 830 KU/1000 kcal to 20,480 KU/1000 kcal. Compared to the lowest level of CML-AGE intake, increased mortality risk was observed in highest level of CML-AGE intake for mortality from all-causes, BrCa and CVD. Increased post-diagnosis CML-AGE intake was positively associated with BrCa and CVD mortality among low consumers of fruits and vegetables and with all-cause mortality among both hormone receptor positive and negative BrCa. Conclusion: The strong positive correlation observed between CML-AGE and fat and protein intake reflect the high AGE levels found in animal products especially those cooked at high temperature. High intake of dietary CML-AGE may contribute to an increased risk of BrCa and the associations were stronger in NHW women, and hormone receptor positive and in situ BrCa. High intake of dietary CML-AGE after BrCa diagnosis in postmenopausal women may contribute to an increased risk of mortality from all-causes, BrCa and CVD.

Available for download on Thursday, December 16, 2021

Included in

Epidemiology Commons