Abstract
BACKGROUND: Previous studies have demonstrated a strong inverse association between cancer and risk of Alzheimer's disease (AD). This study aimed to further investigate this association by examining measures of cognitive performance and neuroimaging. METHODS: Neuropsychological (NP) test batteries consisting of quantitative measures of memory and executive function and volumetric brain magnetic resonance imaging (MRI) scans measuring brain and white-matter hyperintensity volumes were administered to 2,043 dementia-free participants (54% women) in the Framingham Heart Study (FHS) Offspring cohort from 1999-2005. History of cancer was assessed at examination visits and through hospital records. Linear regression was used to examine the association between cancer history and NP/MRI variables. RESULTS: There were 252 and 1,791 participants with and without a previous history of cancer, respectively. Cancer survivors had an average time between diagnosis and NP/MRI exam of 9.8 years. History of any invasive cancer was associated with better executive function (Beta=0.16, p=0.04) but not memory function. Non-invasive cancer was not associated with any change in cognitive performance. Patients with prostate cancer had larger frontal brain volumes (Beta=4.13, p=0.03). Cancer history was not associated with any other MRI measure. CONCLUSIONS: We did not find any strong evidence linking cancer to cognitive or neuroimaging biomarkers that would explain a lower risk of subsequent AD, although a previous FHS study demonstrated a strong inverse association between cancer and risk of AD. Future work should examine the association between cancer and other biomarkers of AD as well as more sensitive metrics of AD-related brain aging markers.