In the world of dementia, and particularly of Alzheimer's disease, few phrases are more likely to polarise opinion than ‘mild cognitive impairment’ (MCI). After something of a cottage industry of publications covering the definition, epidemiology and treatment of MCI, many researchers now feel that this is a redundant concept, as what is required is a term that accurately and reliably represents the earliest stages of Alzheimer's disease. As such, there has been a greater emphasis on the use of specific neuropsychological assessments, neuroimaging and other biomarkers with the hope that these will reduce the numbers of MCI patients who fail to evolve into any form of dementia.1

However, the year 2012 saw the publication of new criteria for the definition of MCI in Parkinson's disease (PD), based on a literature review and expert consensus.2 Are such criteria likely to be useful in aiding our understanding of the progression of cognitive decline in PD and identifying people at high risk of developing dementia associated with Parkinson's disease (PDD)? Or do they merely add an unwanted element of complexity, and divert research energy from early detection of the underpinning pathophysiological …