Phthalates are ubiquitous in modern materials. The U.S. National Health and Nutrition Examination Survey (NHANES) showed that most U.S. adults have detectable levels of phthalate metabolites in their urine. This is troubling because phthalates can disrupt normal endocrine signaling, and may therefore drive hormonally-mediated outcomes including breast cancer, testicular cancer, prostate cancer, cardiometabolic disorders, male and female infertility, and abnormal fetal development. Despite urgent need for epidemiologic studies of phthalate health effects, few such studies exist. This lack of evidence is unsurprising because few epidemiology studies collect urine, which is the only reliable medium for measuring phthalate metabolites. Several specific drugs (e.g., propranolol, ranitidine, and mesalamine) include products that use phthalates as excipients to delay or prolong release of active ingredients from capsules. Users of these medications have phthalate burdens up to 50-fold greater than individuals with only background exposure. Medication users may therefore help to define highly exposed groups in epidemiologic studies of the health effects of phthalate exposure. We will develop predictive models for longitudinal phthalate exposure by capitalizing on existing medication data and urinary phthalate metabolite measurements from a Women's Health Initiative (WHI)- nested breast cancer case-control study (500 cases and 1,000 controls). These predictive models will then be applied to the entire WHI cohort to multiply impute phthalate metabolite levels in all WHI participants (n>160,000). We will then use the multiply imputed phthalate data to conduct the largest-yet study of breast cancer risk as a function of longitudinal phthalate exposure. Our models and the multiple imputation framework may be extended to existing epidemiologic study populations with medication data to permit cost- and time- efficient investigation into the health effects of phthalate exposure?without having to enroll and follow new cohorts with appropriate biological specimen collection.
The proposed project will develop a tool for predicting exposure to phthalates. Phthalates are used in many consumer products, and might play a role in cancer, infertility, and birth outcomes. Our prediction tool will allow existing studies?many of which cannot measure phthalates directly?to estimate phthalate exposure using available data, and then examine the many potential health effects of phthalate exposure.
| Broe, A; Pottegård, A; Hallas, J et al. (2018) Association between use of phthalate-containing medication and semen quality among men in couples referred for assisted reproduction. Hum Reprod 33:503-511 |
