Modifiable Risk Factors for Dementia: Causal Estimates on Individual-Level Data

Abstract

Background: The 2024 Lancet Commission report highlights 14 modifiable risk factors that if eliminated can prevent nearly half of dementia. The causal aspects remain however unclear, hence we aimed to establish causal estimates for these modifiable risk factors.

Methods: We used individual-level data from 408,788 European participants from the UK Biobank. We computed polygenetic risk scores of individual risk factors for each participant, exploiting independent instrumental variants from the largest genomic consortia for each risk factor. We conducted univariable and multivariable linear Mendelian randomization analyses and assessed genetic shapes by nonlinear approaches.

Findings: Genetically predicted high BMI (OR, 95% CI: 1.04, 1.02-1.07), smoking (1.18, 1.06-1.32), high systolic (1.14, 1.09-1.20) and diastolic (1.10, 1.02-1.19) blood pressure, type 2 diabetes (1.04, 1.00-1.09), high LDL cholesterol (1.12, 1.01-1.23) and high triglycerides (1.19, 1.01-1.41) were associated with increased risk of all-cause dementia. Results for Alzheimer’s disease and vascular dementia were in the same direction. Moreover, genetically predicted high physical activity was associated with reduced risk of Alzheimer’s disease (0.58, 0.33-0.99), and longer education was protective against all-cause dementia (0.72, 0.65-0.79), Alzheimer’s disease (0.71, 0.62-0.81), and vascular dementia (0.64, 0.52-0.77). Sensitivity analyses supported the main results, and nonlinear shapes were not detected.

Interpretation: These findings provide causal insights into modifiable risk factors for dementia, underscore the importance of more targeted interventions to mitigate these risk factors for dementia prevention, and give hope for maintaining brain health by timely treatment of high cholesterol and triglycerides, alongside hypertensive medication, smoking cessation and maintenance of normal weight.

Funding: The study was supported by grants from the Lundbeck Foundation (R278-2018-804), the Danish Heart Foundation, Innovation Fund Denmark (9084-00020B), and the Research Fund at Sygeforsikringen Danmark, all to RFS. JL was supported by a BRIDGE program post-doc fellowship from the Novo Nordisk Foundation.

Declaration of Interest: JL, IJR, and JQT have nothing to declare. RFS reports consultancies sponsored by Novo Nordisk. 

Keywords: Risk factor, Dementia, Alzheimer's disease, Mendelian randomization, prevention