Mild Cognitive Impairment as a Clinical Entity and Treatment Target

M ILD COGNITIVE i m p a i r m e n t (MCI) has been actively investigated for the past decade. , 2 The term was coined in the late 1980s by the New York University group to identify individuals who were not cognitively normal for age and yet did not have overt dementia, and their outcomes were described in an article examining predictors of dementia by Flicker et al in the early 1990s. Flicker et al characterized MCI as equivalent to a Global Deterioration Scale rating of 3 and found that a diagnostic approach combining a careful, structured interview and the appropriate neuropsychological tests can discriminate those individuals with MCI likely to experience cognitive deterioration from those with a benign prognosis. Recently MCI has been given more specific criteria involving the features outlined in Figure 1 of the comments by Gauthier and Touchon abstracted from Petersen et al. In conjunction with increased attention to the early stages of development of Alzheimer disease (AD), these criteria served to catalyze intense interest in MCI as a possible prodromal stage of AD. Several studies on longitudinal populations were conducted, some retrofitting criteria to longitudinal studies of aging, and several others are either underway or completed and have demonstrated a progression from MCI to AD at an elevated rate over the base rate in the population. Most of these studies examined the amnestic subtype of MCI, in which memory impairment is a key feature. However, other studies, often encompassing a broader definition of MCI, have demonstrated a reversion to normal in some subjects, implying a lack of stability for the construct over time. This has led to a closer inspection of the underlying premises of MCI and a refinement of criteria to recognize that MCI is a heterogeneous condition. Several factors have emerged as contributing to the variability in the outcomes of several of these studies. Among these are the following factors: (1) criteria for MCI, (2) implementation of the criteria, (3) source of subjects, and (4) reference standards for normal performance. Each of these points deserves discussion as it pertains to the comments raised by Gauthier and Touchon.