Recent research continues to advance our understanding of the boundary between normal aging and dementia. In addition to identifying changes associated with both normal and abnormal aging, researchers have identified a group of individuals whose cognitive deficits place them in an intermediate position on this continuum. The diagnosis of mild cognitive impairment (MCI) is given when older individuals have greater than expected memory impairment for their age but do not meet criteria for dementia (132). Thus, MCI denotes a transitional state between the cognitive changes associated with normal aging and the earliest clinical features of dementia (137,200). This “preclinical” phase is associated with neuropathologic and cognitive disturbances, which gradually rise to a level sufficient for a diagnosis of probable dementia (154). The original formal MCI criteria (56,136) include: (a) significant memory complaints, such as a chronic forgetting of important information, corroborated by an informant; (b) memory impairment on standardized tests relative to ageand education-matched healthy controls (at least 1.5 standard deviation below the mean); (c) otherwise normal cognitive function; (d) normal activities of daily living (ADLs); and (e) failure to meet criteria for dementia. This profile is now considered to indicate the amnestic subtype of MCI (132).

Compared to age-matched peers, individuals with a diagnosis of MCI generally present with subjective memory difficulties of insidious onset. In fact, subjective cognitive complaints can be a harbinger of dementia even before the development of demonstrable cognitive deficits, particularly in highly educated individuals (90). As time progresses, the memory problem becomes more frequent and significant. Despite increasing memory problems, social and occupational functioning is relatively preserved. Nonmemory cognitive domains are relatively spared in MCI, as originally defined, but in practice, they may fall slightly below age- and education-based normative data. For example, speed of processing and cognitive flexibility may be subtly impaired (132).

As more researchers have become involved in studying MCI, it has become clear that it is not a homogeneous disorder, and there appear to be several cognitive subtypes (148). Currently, there are four classifications of MCI, which are based on the presence or absence of memory problems and difficulties in other cognitive domains. These are amnestic MCI-single domain, amnestic MCI-multiple domains, nonamnestic MCI-single domain, and nonamnestic MCI-multiple domains (24). At one time, it was believed that the amnestic variant of MCI would progress to Alzheimer’s disease (AD), whereas the other MCI forms would progress to other forms of dementia such as Lewy body or vascular dementia (136). However, more recent studies have shown that cognitive subtypes of MCI are not necessarily predictive of conversion to specific types of dementia, and both amnestic MCI and nonamnestic MCI frequently develop into AD (53,114). In a community-based study of 980 dementia-free people with diagnoses of all four cognitive subtypes of MCI, many of the individuals continued to demonstrate cognitive decline and eventually dementia. All forms of MCI, except nonamnestic MCI-multiple domains, converted to AD within 6 years of follow-up, whereas many individuals in the nonamnestic MCI-multiple domains group progressed to non-AD forms of dementia (24). More longitudinal studies of the various MCI clinical profiles are needed to diagnose individuals as early as possible in the disease course so that, when further treatments become available, they can be used before there is too much irreversible brain damage.

This chapter will focus on the amnestic variant of MCI. Current estimates indicate conversion rates from MCI to AD of between 6% and 25% per year (133), with most studies falling in the 10% to 15% per year range (19,46,56,133,136,188). This is significantly greater than the 1% to 2% per year rate at which AD develops in the normal elderly population. Petersen et al. (136) report that up to 50% of individuals diagnosed with MCI will convert to AD within 4 years, and Morris et al. (124) report a conversion rate of >80% in individuals followed for 9.5 years.

Given the likelihood of progression to AD in individuals with MCI, early identification and diagnosis have important implications for treatment (2).
Although it is not yet clear who will or will not convert to AD, several potential risk factors have been identified. Prominent among the risk factors is the presence of the apolipoprotein (ApoE) ε4 allele. ApoE is a marker initially studied as a risk factor for cardiovascular disease. It is involved in the normal regulation of phospholipid metabolism and cholesterol and may play a role in neural repair. Identified cognitive correlates of ApoE genotype include subtle decreases on verbal memory tested over time in longitudinal studies (33) and failure to benefit from cueing at recall (134). Neuroimaging associations with ApoE include mesial temporal lobe changes such as decreased volume of the hippocampus (82) and entorhinal cortex (58,82,95,96,103), as well as decreased medial temporal gray matter (201). The use of magnetic resonance imaging (MRI) or other imaging techniques such as functional neuroimaging may be useful to distinguish MCI patients who will develop AD from those who will not (27).