Premium
Uncovering the multicausality of Alzheimer’s disease: A systems modeling approach
Author(s) -
Uleman Jeroen,
Melis René JF,
Hoekstra Alfons,
Quax Rick,
Rikkert Marcel GM Olde
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.041105
Subject(s) - disease , causal model , computer science , psychological intervention , multitude , complex system , psychology , association (psychology) , cognitive psychology , artificial intelligence , data science , cognitive science , medicine , psychiatry , pathology , philosophy , epistemology , psychotherapist
Abstract Background In the onset and progression of Alzheimer’s disease (AD), many factors and processes are involved. Our understanding of the multicausal and complex nature of AD may benefit from computational modeling approaches that can be used for knowledge synthesis at the level of the whole system. We have developed a systems dynamics (SD) model of AD. Method We have applied the group model building methodology to develop from expert input and literature review a causal loop diagram (CLD) that graphically describes the relationships between important risk factors and proposed causal mechanisms in non‐familial AD from midlife onwards. This CLD was implemented computationally as a SD simulation model. Result For the resulting CLD, experts identified 38 variables and 144 connections between them. The experts modeled hypotheses about their relations with a multitude of important processes ranging from amyloid‐β accumulation and cerebral endothelial dysfunction, via physical activity and sleep to social functioning. The SD model was parametrized using empirical information taken from the Global Alzheimer’s Association Interactive Network (GAAIN) and is able to simulate preventive interventions on potentially modifiable risk factors. Conclusion The CLD and SD model described and quantified hypotheses about relationships and feedback loops implicated in AD onset in a comprehensive way. Our approach showed the potential of systems‐oriented simulation models for studying the complex etiology of AD.