A SYSTEMATIC REVIEW OF MENDELIAN RANDOMIZATION STUDIES INVESTIGATING CAUSAL ASSOCIATIONS BETWEEN RISK FACTORS AND DEMENTIA

Abstract

Very little is known about the causal nature of the associations between established risk factors, e.g. education and smoking and dementia risk. We have therefore conducted a systematic review of Mendelian randomization (MR) studies investigating causal links between risk factors and dementia or dementia-related outcomes. We searched five databases from inception to March 2016 using a search strategy including subject headings and free text terms relevant to dementia, cognition and MR. Forward and backward citation searches of included publications were also conducted. We included MR studies investigating the association between any risk factor and global cognitive function, all-cause dementia or dementia subtypes, excluding non-genetic studies, genetic studies other than MR and animal studies. Two reviewers independently screened titles and abstracts, reviewed full-texts and assessed the quality of included studies. Fourteen MR studies investigating a variety of risk factors (education, lifestyle factors, cardiovascular factors and related biomarkers, diabetes related and other endocrine factors, and telomere length) and dementia-related outcomes met our inclusion criteria. Most studies received an overall rating of good quality, however sample size and statistical power were rated low in half of the studies. Shorter telomeres were causally associated with an increased risk of Alzheimer's disease (AD, odds ratio per standard deviation decrease of telomere length = 1.36 (1.12, 1.67), p = 0.002). There was also some suggestion of a causal association with smoking quantity and metabolic risk factors including systolic blood pressure, fasting glucose, insulin sensitivity and high-density lipoprotein cholesterol. Causal associations between several risk factors and dementia-related outcomes were investigated in one study only. Moreover, survival and diagnostic bias or pleiotropic effects cannot be excluded. There was convincing genetic evidence to support a causal association between telomere length and AD. Inconclusive evidence for other risk factors or tentative evidence for smoking quantity and selected metabolic markers may result from insufficient statistical power. Further well designed MR studies are needed to investigate the causality of the associations between risk factors and dementia where randomised controlled trials are unethical or impractical.