ABSTRACT More breast cancer patients are treated with anti-estrogen agents, such as aromatase inhibitors (AIs), than with any treatment other than surgery. Due to their effectiveness, recent guidelines for postmenopausal women advise extending AI therapy to 10-years duration. Therefore, exposure to AIs among the more than 3.5 million breast cancer survivors will soon increase. A major concern of breast cancer survivors is therapy-induced cognitive dysfunction. About 75% of women suffer cancer-related cognitive impairment during chemotherapy treatment, which persists for months to years. There are theoretical and empirical reasons to believe that cognitive impairment may also occur with estrogen deprivation. Surprisingly, there are few empirical human studies of AI effects on cognition; none with cognition measured pre-cancer. Leveraging 18 years of data from the Health and Retirement Studies―a nationally-representative, longitudinal survey augmented by Medicare medical and pharmaceutical claims, we propose a longitudinal study designed to: (1) Characterize breast cancer survivors’ cognitive functioning post-initiation of adjuvant AI therapy over a 60-month period using a validated neuropsychological measure and compare their performance to those of (i) women with incident breast cancer who did not receive AI therapy and (ii) non-cancer controls. Cross-sectional analyses corresponding to roughly one-, three-, and five years post-AI therapy initiation will provide estimates of early and late effects of AIs; (2) Contrast the longitudinal trajectory of cognitive functioning preceding and following cancer diagnosis and treatment among women treated with an AI to those of (i) cancer survivors who did not receive AI therapy and (ii) cancer-free controls, over an 18-year span. We anticipate a gradual decline in cognitive functioning among cancer-free women over the study period. The AI group, however, is expected to exhibit a steeper downward slope after AI initiation, indicating an adverse AI effect on cognition above and beyond normal cognitive aging; and (3) Quantify the tradeoff between overall and disease-free life years and cognitive functioning among women with incident breast cancer undergoing different adjuvant therapies. Using findings from Aim 2 combined with parameter estimates from analyses of time to disease recurrence and time to death, we will calculate overall and disease-free life years by cognitive functioning for various scenarios of patients’ sociodemographic characteristics and cancer treatment(s). Results from these analyses will provide rigorous evidence regarding the potentially adverse effects of AI exposure on cognition that may serve as the basis for clinical interventions. They will also generate population-level estimates of survival and cognitive decline attributable to AI and other cancer-treatments (beyond normal cognitive aging) that could be used to better inform patients about “real world” tradeoffs of different breast...