Reifying Relevance in Mild Cognitive Impairment: An Appeal for Care and Caution

Abstract

Reifying Relevance in Mild Cognitive Impairment:An Appeal for Care and Caution Janice E. Graham (bio) and Karen Ritchie (bio) Keywords Alzheimer’s disease, construction, dementia, market forces, mild cognitive impairment We thank the reviewers for their thoughtful comments that probe shadowy areas in our argument, and we welcome this opportunity to elucidate our position. First, we are not repudiating the natural and social facts of pathologic brain degeneration and the physical and cognitive impairments that manifest in people affected by dementia disorders. We are, however, questioning the epistemology, politics, and persistence by some to harden the construct of mild cognitive impairment (MCI) prematurely as a "natural" precursor to dementia. We suggested insufficiencies in viewing MCI as prodromal to Alzheimer's disease (AD; Graham and Ritchie 2006, 36), and pointed to evidence for the prognostic irrelevance of some MCI subcategories (35). The commentators have read our manuscript as suggesting that AD and MCI are different kinds, and that a diagnosis of AD is more "real" than that of MCI. But AD is also fraught with classification issues, as we describe in the section of our paper that deals with consensus committees, published practice guidelines, and the determination of treatment efficacy. Like MCI, AD is a heterogeneous condition, but unlike MCI, AD is a constructed practical kind that does some "work" beyond essentialist (neural degeneration, brain atrophy) claims. Its criteria identify people with well-defined dementia symptoms and multiple cognitive and functional impairments who are in need of, and benefit from, support. Even with these criteria, however, there have been great attempts to achieve definitional accuracy in dementia diagnoses. Practical refinements to the differential diagnosis of dementia based on solid research evidence have made way for more specific recognitions of vascular dementias, dementia of the Lewy body type, and other subtypes beyond AD. So AD too has been subject to more careful, more cautious operationalization in recent years, with research supporting multiple factors and the comorbidity of several types (kinds) of dementia in any one individual. Our paper takes up concerns that are not resolved for dementia and suggests that the nascent category of MCI is still so insufficiently defined as to be ambiguous, and as such, is in danger of identifying individuals with what will be, and has [End Page 57] already been, linked to a debilitating and deeply stigmatizing and fatal disorder (AD). We are not contesting the reality that some people begin to display early signs of cognitive loss, nor are we suggesting that symptoms be ignored; as researchers, we stand behind the careful tracking of symptoms and signs in patients and research participants as an essential clinical and research practice. In a era where the public has been made wary of the potentially fatal effects of new treatments whose more careful scrutiny in specific populations would have saved lives (e.g., Fontanarosa, Rennie, and DeAngelis 2004; Psaty and Furberg 2005), and made all the more popular by what Kramer coined "diagnostic bracket creep" (1993, 15), we are questioning the drivers and value of a premature diagnostic classification whose definitional accuracy is neither sufficient nor effective for specific identification of individuals. That matters of fact are constructed does not take away from their natural fact. Some people have subjective reports or show objective evidence of cognitive decline while having preserved activities of daily living—the construction of this evidence indicates they are not normal and they are not demented. Based on such evidence, where then does the balance of probability for MCI lie? A multidisciplinary international group of experts gathered in Stockholm in 2003 to discuss MCI concepts (Winblad et al. 2004). They acknowledged that the heterogeneous etiology of MCI contributes to "some confusion concerning the specific boundaries of the condition", stating that this "not normal, not demented" condition is nonetheless "useful both clinically and as a research entity" (p. 241). Importantly, they concluded that MCI needs "better definition" (p. 246). It was at this stage that we first wrote our manuscript, and in the subsequent two years we have eagerly awaited new studies, and the recognition of flexible, practical subtypes to evolve. We state "that MCI cannot be considered to be a separate clinical entity" (Graham and...