Background: Population ageing over the first half of this century is likely to lead to dramatic increases in the prevalence of dementia. This will affect all regions of the world, but particularly developing regions. Dementia projections have been used extensively to support policy. It is therefore important these projections are as accurate as possible.
Discussion: In this paper we provide a commentary on studies projecting the future prevalence of dementia for the world or for individual continents. We identify some important limitations of the methods used in published projections and provide recommendations to improve the accuracy of future projections, and allow for the checking of the accuracy of the predictions.
Summary: Accurate projections of dementia incidence, at both the global and local level, are essential for healthcare planners.