Serum-based biomarker development for dietary macronutrient densities and their association with breast and colorectal cancer risk in a cohort of postmenopausal females.

Background

Associations of the macronutrient composition of the diet with risks of postmenopausal breast and colorectal cancer (CRC) are uncertain, partly because of reliance on self-reported dietary data.

Objectives

We aimed to study biomarker development for several macronutrient component densities using serum and spot urine metabolomics and, when appropriate, to assess their associations with breast cancer and CRC risks in a Women's Health Initiative (WHI) cohort of postmenopausal U.S. females.

Design

AND

Methods

We explored linear biomarker equations for log-transformed macronutrient component densities using fasting serum metabolomic profiles, with and without spot urine, in a WHI feeding study (n=153). We used equations satisfying a cross-validated regression R2 (CV-R2) criterion to calculate potential biomarker values for 577 breast cancer cases and 181 colorectal cancer cases and their 1-1 matched controls. We used Cox regression with baseline stratification on matched pairs to examine dietary composition associations with cancer risk.

Results

Serum-based biomarker equations for macronutrient component densities had CV-R2 values as follows: polyunsaturated fatty acids (PUFA) 46.7%, monounsaturated fatty acids (MUFA) 36.3%, saturated fatty acids (SFA) 33.6%, carbohydrate 32.3%, and protein 29.4%. The inclusion of spot urine metabolites did not materially improve these values. In analyses including serum-based PUFA, MUFA, SFA and carbohydrate densities the breast cancer hazard ratios (95% CIs) for 20% increments in dietary densities were 0.98 (0.89, 1.07) for PUFA and 1.14 (0.97, 1.33) for MUFA. Corresponding CRC hazard ratios were 0.98 (0.81, 1.17) and 1.46 (1.10, 1.93). Analyses based on food frequency questionnaires differed from these estimates for breast cancer, but tended to agree for CRC. Intakes of dairy and meat products may help to explain observed associations.

Conclusions

In a population of U.S. postmenopausal females dietary MUFA density was associated with an elevation in CRC risk. Breast cancer associations with biomarker-based macronutrient densities require further development This study is registered with clinicaltrials.gov identifier: NCT00000611 https://clinicaltrials.gov/study/NCT00000611.