ABSTRACT
The Sunnybrook Memory Study is an ongoing prospective study that commenced in 1991. It was designed to examine the usefulness of neuropsychological, behavioral, and genetic measures in the prediction of Alzheimer's disease (AD) in nondemented patients. We attempted to optimize the selection of participants most likely to progress to AD by including patients identified by their family physicians as having memory problems but no other neurological disorders. Because family physicians are the first point of contact with the health care system by older individuals and their families, we wanted to ensure that our findings would be generalizable to patients of family physicians. Thus, our first selection criterion to define participants for this study was judgment by the family physician that cognitive impairment was present, regardless of whether this was confirmed by subsequent neuropsychological testing. Our second criterion was that participants not meet the criteria for dementia, which we verified by a thorough diagnostic work-up, including specialist medical and neuropsychological assessments. In an attempt to reduce the number of participants with dementia required to undergo the diagnostic work-up, we also requested that family physicians not refer their patients with dementia. In addition, all referred patients underwent an initial screening examination and those who scored below established cut-points on the Dementia Rating Scale (DRS) and the Mini-Mental State Examination (MMSE) were excluded. Those remaining participants who scored above the cut-points underwent a rigorous diagnostic examination to exclude dementia. Finally, because we were specifically interested in the prediction of progression to probable AD, we excluded referred patients who showed evidence of any conditions affecting the brain (e.g., stroke, Parkinson's disease, etc.).
