Environmental and Genetic Risk Factors For Breast Cancer: The Sister Study

National Institute Of Environmental Health Sciences

Investigators

Linked publications & trials

Abstract

We are following 50,884 US women who were aged 35-74 and had a sister with breast cancer but did not have breast cancer themselves when they enrolled in 2003-2009. Enrollment data on potential risk factors and health status were collected using telephone interviews. Blood, urine, and environmental samples were collected in a home visit. To date, more than 3,500 participants reported a diagnosis of invasive or in situ breast cancer. The cohort is tracked annually for changes in vital status and major health outcomes. Detailed follow-up questionnaires on health outcomes, environmental and lifestyle exposures, and other topics are completed every 2-3 years. Medical records are retrieved for cancers and other priority conditions. Tumor tissue is obtained for breast cancers. The first Sister Study follow-up survey was completed in June 2012 with a response rate of 95%. The second and third detailed follow-up surveys were completed with better than 92% response. The fourth was launched in October 2017, with 85% responding as of July 2019. We have conducted several studies of dietary factors and breast cancer risk. One study failed to confirm a previously reported link between an estrogen-related diet and breast cancer. In another, we observed that high adherence to the Dietary Approaches to Stop Hypertension (DASH) diet was associated with reduced risk of breast cancer. Adherence to the Alternative Health Index was also associated with reduced breast cancer risk. We also found that women with high diet-dependent acid loads had a greater risk of developing breast cancer than those with low acid loads. This suggests that diets with decreased acid load, such as those abundant in fruits and vegetables and therefore rich in potassium salts, may protect against breast cancer. Additionally, we found that increasing consumption of red meat was associated with increased risk of invasive breast cancer whereas results suggested that replacing red meat with poultry could reduce breast cancer incidence. In the first of three studies of risk factors for obesity, we found that women who ate breakfast every morning were less likely to be or become obese, compared to women who irregularly ate breakfast (3-4 days a week). In the next, we found that chronic penicillin use may be associated with increased risk of obesity. Finally, we observed that exposure to artificial light at night while sleeping was associated with being or becoming overweight or obese, with stronger associations seen for those reporting lights or television on in the room. In 2014-15 we repeated the collection of biological and environmental samples and anthropometric measures from women diagnosed with breast cancer since enrollment and a random sample of women without breast cancer. Of approximately 3,800 participants invited, samples were collected from 2,436 (63%) including 1,229 women who had been diagnosed with breast cancer. So far, we have published three studies that made use of twice-collected biospecimens. In the first, we assessed concentrations of 16 trace elements from toenails collected at baseline and in 2014-15, finding that levels generally decreased over time, particularly for cadmium, chromium and lead. We used these results to better inform a study of trace elements and young-onset breast cancer, finding no evidence of associations for any of the measured elements, after correcting for differences in year of collection for cases versus controls. In another study, we observed that while both breast cancer cases and non-cases had higher serum 25-hydroxyvitamin D concentrations, on average, at the second time-point compared to the first, increases were greater in cases. Finally, we studied whether breast cancer survivors were more likely to develop cardiovascular disease risk factors (e.g. hypertension, high cholesterol, overweight) over time than women who did not have breast cancer, finding no difference. Breast cancer and ovarian cancer cases through 2014 and a random sample of the cohort have been genotyped as part of the multi-site Breast Cancer Association Consortium and Ovarian Cancer Association Consortium. Through these projects, Sister Study data have been included in several collaborative analyses, including: a study of predicted DNA methylation and breast cancer; a study of predicted DNA methylation and ovarian cancer; a study of vitamin D-related genes and ovarian cancer; a genome-wide association study (GWAS) of Anti-Mullerian hormone (AMH) levels; a GWAS of six solid tumor cancers; a Mendelian randomization study of obesity, insulin, glucose and breast cancer; and a study of a polygenic risk scores, breast cancer, and breast cancer subtypes. We additionally lead or are contributing to several other consortium projects, including the Premenopausal Breast Cancer Collaboration with primary collaborators Hazel Nichols (University of North Carolina) and Anthony Swerdlow and Minouk Schoemaker (Institute for Cancer Research, UK). To date more than 20 cohorts have joined this effort. In a paper published last year, we showed that higher BMI was associated with reduced risk of premenopausal breast cancer. We also showed that breast cancer risk is increased shortly after breast cancer for women who have given birth compared to women who do not have children; childbirth doesnt become protective for breast cancer until 20 more years after childbirth. Projects from other consortia include a study of analgesic use and ovarian cancer, a study of AMH and breast cancer risk prediction, and two studies of risk factors for biliary tract cancers. Dr. Alexandra White led efforts to identify environmental risk factors for breast cancer using publicly available data sets to look at participants estimated exposure to air toxics. We found that women who lived in areas of higher airborne cadmium, lead and mercury were at a higher risk of developing postmenopausal breast cancer. Other air pollutants, including methylene chloride, propylene dichloride, polycyclic organic matter, and styrene, were associated with a higher risk of breast cancer. Dr. White was recently awarded a grant through the Office of Research on Womens Health to study how trace elements (as measured in toenail samples) may be related to breast cancer risk. In addition, we have collaborated with Dr. Joel Kaufman (University of Washington) to study criteria air pollutants and breast cancer and a new collaboration considers the association between air pollution and cardiovascular disease in the Sister Study, an effort that will be expanded to include breast cancer. In an initiative led by Dr. Jack Taylor, we previously generated data on 450,000 CpGs for the non-Hispanic white women in the genotyping sample. We have now reported on several thousand sites associated with breast cancer risk, many of which may be useful in detecting early stage invasive breast cancer. We also identified CpGs associated with alcohol use and mothers age at birth. One of our most significant contributions has been in the field of epigenetic age, a biological measure of age quantified using specific DNA methylation markers. We observed a positive association between three epigenetic age clocks and breast cancer. In subsequent analyses, we found associations between epigenetic age and both shift work and air pollution. Dr. Taylors group plans a new DNA methylation study that will focus on African-American women and estrogen receptor negative breast cancer. Dr. Chandra Jacksons group has been using Sister Study data to study predictors and health effects of sleep. We found that women who experienced childhood trauma reported worse sleep. In another study, we showed that women who averaged less than 7 hours of sleep per night and who had difficulty falling or staying asleep were more likely to have metabolic disorders.