Background Community- or population-based longitudinal studies with brain donation end points offer an opportunity to examine correlations between pathology and clinical state prior to death. Discriminating the earliest signs of dementing disorders, such as Alzheimer's disease (AD), is imperative to determine early interventions and treatments. Objectives The principle aims of the study were:To define pathological boundaries between findings that are clinically relevant and those that are not.To identify known risk factors for dementia and assess their predictive capabilities on pathological outcome. To determine whether there are clinical predictors that define pathological outcome when made years in advance of illness and death.Methods and design Using the first 89 donated brains from The University of Manchester Longitudinal Study of Cognition in Normal Healthy Old Age, immunohistochemical techniques were employed to assess the neuropathological profile of the cohort including CERAD and Braak stage. The presence of dementia risk factors was assessed by questionnaire and correlated with presence and severity of pathology. Cognitive test scores collected many years prior to death were correlated with extent of AD pathology at death.Results The neuropathology of the examined cohort correlated well with the established literature and affirms that the cohort is typical and representative. Cognitive impairment in life correlated strongly with all pathologies found at post-mortem. Cardiovascular and genetic risk factors were found to be the main influences that drive the pathogenesis of neurodegenerative diseases such as AD. The Memory Circle test had the ability to independently distinguish between individuals who were considered none/low risk of AD and medium/high risk of AD. The mean length of time between conduction of the test and death was 21.3 years (± 3.5) placing the predictive testing period within the preclinical and prodromal AD phase.Conclusions and implications Results suggest it may be possible to predict the ultimate development of AD pathology over 20 years before death using a simple cognitive test. The present work may facilitate early interventions, therapeutics and treatments for AD by identifying at risk and minimally affected (in pathological terms) individuals in whom the prospects of halting or even reversing disease would become realistic.