Cancer Control Research5R03CA119697-02
Gatz, Margaret J.
A TWIN STUDY OF COGNITIVE OUTCOMES IN OLDER ADULT CANCER
DESCRIPTION (provided by applicant): This project addresses the longer term cognitive impact of cancer and its treatment on older adults through use of the population-based Swedish Twin Registry. All twins aged 65+ who have survived any malignant cancer [excluding brain cancer] for three or more years have been identified through the nationwide Swedish Cancer Registry. From a previous NIA-funded study, we have collected data about cognitive outcomes for these twins, including telephone screening for cognitive impairment and diagnoses for dementia based on complete clinical evaluations. For the proposed project, we will link information about cancer with information about cognitive function. Funding will permit analyzing these previously collected data for epidemiologic purposes and for planning a future larger application. The Cancer Registry indicates the cancer diagnosis and the center where treatment took place for all patients. The registry does not indicate type of treatment. Thus, we propose to go to the cancer treatment centers and obtain records for each survivor. From the records, we will abstract detailed information about date, type, length, and dose of treatment. This is the main activity that cannot take place without funding. This information will be used to test the hypothesis that cancer survivors will have poorer cognitive functioning and will be more likely to be clinically diagnosed as demented than older adults who have not been treated for cancer, controlling for measured covariates (education, alcohol use, and physical exercise) and hypothesized mediators (cardiovascular disease, depressive symptoms, and fatigue), as well as unmeasured familial effects (genetic influences and shared environmental influences early in life). Secondly, we will examine how specific treatments, for example, chemotherapy and radiotherapy, relate to impairment. Hypotheses will be tested comparing those who have been diagnosed with cancer to both (1) co-twins, that is, within each pair, the twin sibling who has not been diagnosed with cancer, and (2) an unrelated comparison group consisting of individual twins who were not diagnosed with cancer and are not related to the individuals with cancer. The co-twin study includes over 500 twin pairs discordant for cancer, with average survival of nearly 12 years since diagnosis. For the unrelated comparison design, we will match three individuals from the twin registry to each of over 700 cancer survivors. Data on key outcomes and relevant covariates will come from the Inpatient Discharge Registry; questionnaires collected when the twin registry was formed over 30 years ago; a telephone interview of the entire twin registry in 1998-2001; and clinical evaluations for dementia conducted after telephone screening. This study capitalizes on a uniquely available population-based dataset of older adult cancer survivors. Through abstracting data from cancer treatment records and linking them to existing datasets, we will be able to examine possible longer term effects, along with likely covariates, with optimal comparison groups.