Search

Constructing a graphical model of the relationship between physical activity and cognitive function based on a systematic review of prospective evidence Citation: Thomas Campbell, Jessica Whyte, Karl Ferguson, Breda Cullen Background: Elucidating modifiable factors that may confer protection against cognitive decline is an important area of investigation. Physical activity (PA) represents one such factor and has generated substantial evidence regarding its protective benefit. However, in epidemiological research the presence of multiple covariates has posed challenges to obtaining unconfounded estimates of associations between exposure and outcome variables. Causal diagrams, or Directed Acyclic Graphs (DAGs), have emerged as promising tools for addressing these challenges and improving the planning and specification of statistical analyses. This review aims to utilise a recently developed research method for constructing DAGs in a robust and systematic way: Evidence Synthesis for Constructing Directed Acyclic Graphs (ESC-DAGs) (Ferguson, 2019). The method will be applied to observational studies of the relationship between PA and cognitive decline, and the resulting DAG will be used to inform empirical analysis of UK Biobank data (a very large cohort study), in a separate paper. Review Question(s): (a) What clinical and sociodemographic factors, and health behaviours, do the included studies report as being associated with cognitive function/impairment and/or physical activity? (b) How do the studies vary with regard to the covariates included, how these are adjusted for, and the conclusions of the study? Reporting framework: This systematic review will be conducted and reported according to Conducting Systematic Reviews and Meta-analyses of Observational Studies of Etiology (COSMOS-E) (Dekkers et al., 2019). Searches A systematic search of literature will be conducted using the following electronic databases: • Medline on the Ovid platform • Embase on the Ovid platform • PsycINFO on the EBSCOhost (EBSCO) platform In the event that the searches performed on these databases are not sufficiently sensitive (as would be indicated by known relevant papers not being captured), the CINAHL database on EBSCOhost and Web of Science platform may be incorporated to the search strategy. The search terms will incorporate medical subject headings (MeSH terms) which will be adapted for compatibility within each database. The terms used will be based on the following key terms: - Physical activity OR Exercise OR Physical fitness OR Activity level AND - Cognitive Function OR Cognitive Dysfunction OR Cognitive Decline OR Cognitive Aging OR Cognitive Impairment OR Neuropsychological Tests OR Alzheimer’s Disease OR Dementia AND - Longitudinal studies OR Prospective OR Cohort Where applicable the search terms will be modified to include truncation and wildcard characters. Searches will be restricted to English Language publications from 01.01.2005 to 01.07.2020. Additional hand-searching of the reference lists of recent relevant review papers will also be applied. The sensitivity of the search strategy will be tested by checking whether a sample of papers known to be relevant are detected by the search. Types of study to be included Inclusion criteria Longitudinal cohort studies (prospective design). Exclusion criteria Cross-sectional studies, case-control studies, retrospective studies, intervention studies & review papers. Condition or domain being studied Cognitive function as demonstrated by performance on objective cognitive tests. Where the outcome is measured categorically as impaired versus unimpaired, the threshold for impairment will be as defined by study authors but will at a minimum require performance to be one standard deviation below the mean of a healthy comparison group. Population Community dwelling adults, free of cognitive impairment before the period of exposure to physical activity began and aged ≥45