Introduction

Alzheimer's disease (AD), the commonest cause of dementia, is estimated to affect over 400 000 people in the UK,1 with numbers around the world set to rise dramatically as the population ages. Distinguishing AD from other causes of cognitive impairment is important for numerous reasons, including provision of appropriate care and access to medications. As and when disease-modifying therapies become available, the priority for earlier and more accurate diagnosis will increase still further.

An important starting point is the definition of AD. This is not as straightforward as it might seem: concepts of AD have evolved rapidly over recent years, with new criteria proposed as recently as 2011 continuing to provoke debate. Ultimately, a diagnosis of AD requires pathological confirmation. However, even a pathological diagnosis comes with a degree of uncertainty2 and the pathological phenotype of AD is itself heterogeneous.3

Diagnosis in life has until relatively recently depended on very broad clinical criteria dating back to the 1980s, mandating that an individual be ‘demented’, that is, have progressive impairment of memory and at least one other cognitive domain sufficient to impact on activities of daily living.4 The new criteria for diagnosis of AD have built on a large body of work demonstrating that biomarkers of AD—including MRI, cerebrospinal fluid (CSF) and positron emission tomography (PET)—may improve diagnostic sensitivity and specificity. The criteria also reflect a growing move to try to achieve earlier diagnosis. The new criteria propose that AD can be diagnosed with different degrees of certainty based on the extent of the cognitive deficits and presence/absence of biomarker abnormalities. These criteria were reviewed recently in this journal.5

For the purposes of this article and with the aim of providing a practical guide to the differential diagnosis of AD, we will first consider the ‘typical’ clinical …