Jim Lacey

NCI, DCEG, HREB

PLCO (Learn more about this study)

2006-0158

Jul 1, 2006

Breast cancer rates and relative risk according to recognized breast cancer risk factors in the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial

Population-based studies with large numbers of women in particular breast cancer risk groups, such as those defined by race/ethnicity, exogenous hormone use, cancer in relatives, first birth at late ages, or obesity, offer excellent opportunities to better understand etiology. We evaluated the validity and generalizability of breast cancer rates and risk factor associations in the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial, a large cohort of postmenopausal women. From 1993 to 2001, 71,751 cancer-free female PLCO study enrollees completed a baseline risk factor questionnaire. We identified incident breast cancers from self-reports, medical records, and state cancer registry and death index linkages. Poisson regression produced age-specific incidence rates and rate ratios (RR). Through mid-2003, 1958 women developed breast cancer. Compared to U.S. SEER rates, PLCO age-specific incidence rates were slightly higher at ages below 65 but otherwise equivalent. African-Americans' and Asian/Pacific Islanders' rates were higher in PLCO than SEER. Compared to RRs from other large studies, decreased RRs for later menarche, earlier menopause, and lower parity and increased RRs for menopausal hormone therapy use were similar to expectation, whereas elevated RRs for family history, height, weight, and body mass index (BMI) were lower than expected. Incidence rates and individual RRs that compare favorably to external benchmarks signify good validity and generalizability for PLCO data. With extensive biologic specimens available for approximately half of the cohort, large numbers of women in at-risk groups, and the ability to evaluate associations with combined risk factors, appropriate future studies can be devised.