BACKGROUND: Many putative causes and risk factors have been associated with outcomes in Alzheimer's disease (AD) but all attempts at disease-modifying treatment have failed to be clinically significant. Efforts to address this "association-intervention" mismatch have tended to focus on the novel design of interventions.

OBJECTIVE: Here, we instead deal with the notion of association in depth. We introduce the concept of disease-associated process (DAP) as a flexible concept that can unite different areas of study of AD from genetics to epidemiology to identify disease-modifying targets.

METHODS: We sort DAPs using three properties: specificity for AD, frequency in patients, and pathogenic intensity for dementia before using a literature review to apply these properties in three ways. Firstly, we describe and visualize known DAPs. Secondly, we exemplify qualitative specificity analysis with the DAPs of tau protein pathology and autophagy to reveal their differential implication in AD. Finally, we use DAP properties to define the terms "risk factor," "cause," and "biomarker."

RESULTS: We show how DAPs fit into our collaborative disease ontology, the Alzheimer's Disease-Associated Processes and Targets (ADAPT) ontology. We argue that our theoretical system can serve as a democratic research forum, offering a more biologically adequate view of dementia than reductionist models.

CONCLUSION: The ADAPT ontology is a tool that could help to ground debates around priority setting using objective criteria for the identifying of targets in AD. Further efforts are needed to address issues of how biomedical research into AD is prioritized and funded.