ABSTRACT‍

‍Coffee consumption is associated with a reduced risk of several diseases but uncertainty remains about the influence of coffee consumption on the risk of dementia. We performed a dose-response meta-analysis to summarize the prospective data on coffee consumption and associated risk of dementia and Alzheimer’s disease. We identified studies by searching PubMed (from January 1966) and Web of Science (from January 1945) through 4 October 2018 and by scrutinizing the reference lists of pertinent publications. Two researchers independently reviewed the literature. Results were combined using a restricted cubic spline random-effects dose-response meta-analysis based on a one-stage approach. Eight relevant prospective studies were identified. These studies included 7486 dementia cases diagnosed among 328,885 individuals during an average follow-up of 4.9–25 years. Meta-analysis of all eight studies indicated no statistically significant association between coffee consumption and the risk of dementia and no deviations from a linear trend (p = 0.08). The relative risk of dementia per 1 cup/day increment of coffee consumption was 1.01 (95% confidence interval (CI) 0.98–1.05; p = 0.37). Meta-analysis of five studies that focused on Alzheimer’s disease revealed no association between coffee consumption and Alzheimer’s disease and no deviations from a linear trend (p = 0.79). The relative risk of Alzheimer’s disease per 1 cup/day increment of coffee consumption was 1.01 (95% confidence interval 0.95–1.07; p = 0.80). These results do not support an association between coffee consumption and an increased risk of overall dementia or Alzheimer’s disease specifically, but further research on the association of coffee consumption with dementia risk is needed.