Mild Cognitive Impairment and Functional Status

The Greeks knew that we get the answers to the questions we ask. In 2005, the answer to the question, "Who has mild cognitive impairment (MCI)?" is "MCI is the class of cognitively impaired persons who do not meet criteria for dementia." However, nature requires that we put our questions to her carefully. We have not yet achieved this in the case of MCI. MCI is surely a heterogeneous condition. Not all cases progress to dementia over time. A significant fraction revert back to "normal," and others develop non-Alzheimer's disease (AD)-related conditions.1,2 Although memory loss is by definition essential to "dementia," it is apparently not essential to MCI. Instead, we increasingly distinguish between "amnestic" and "nonamnestic," including "dysexecutive," MCI.3 Although it was tempting to concentrate initially on "isolated" memory impairment as the harbinger of future AD, 35% of "nondemented" memory-impaired cases have equally severe executive impairments, whereas an equal number have isolated executive impairments that memory assessment alone would not detect.4 Moreover, the very definition of "impairment" varies from study to study. What MCI cases seem to share is our conviction that they do not meet criteria for "dementia." To this extent, MCI is well represented by the Canadian Health and Aging study's concept of "cognitive impairment no dementia" (CIND).5 Unfortunately, this leaves us at the mercy of dementia's definition. For example, if "dementia" is a Clinical Dementia Rating (CDR) score of 1.0 or greater, as Bennett et al.6 suggest in this issue of the Journal, then cognitively impaired persons with CDR scores less than 1.0 have MCI. However, the CDR, the Diagnostic and Statistical Manual, Fourth Edition (DSM-IV), and the International Classification of Diseases, 10th Edition (ICD-10), all appear to be biased in favor of AD, at the expense of many non-AD disorders.7 If the definition of dementia is altered, say by the DSM-V, then the prevalence of CIND will change as well. It seems to me that three conditions are necessary for the diagnosis of dementia. There must be acquired cognitive impairment(s). There must the functional disability. The disability must be related to the cognitive impairment(s) that are observed. Unfortunately, neither the dementia literature nor the MCI literature has provided cogent guidance regarding the specific functional capacities, or functional status measures, that should be used to determine the presence of "dementia," and hence CIND. There is ambiguity at every step. Even the standards against which functional and cognitive "impairment" should be measured are uncertain. Most psychologists, acting perhaps out of their forensic heritage, use age-specific means to define cognitive "impairment." However, young adult means may be more relevant when attempting to correlate cognition with disability, and hence "dementia."8 Who wants to know whether the patient can drive as well as the average 82-year-old? We want to know if she can drive. At the First International MCI Key Consensus Symposium in Stockholm, Sweden (April 2003), it was suggested that MCI be defined in part by the absence of impairment in activities of daily living (ADLs) and no more than "mild" impairment in instrumental activities of daily living (IADLs).3 Likewise, Bennett et al.6 failed to find ADL impairment in their "MCI" cases. Nonetheless, they found a high percentage of IADL impairment, which could in turn be associated with changes in cognition and motor function. Geriatricians have long studied functional impairments in elderly populations. It appears that functional capacity is lost hierarchically. Normally, IADLs are lost before ADLs,9 and there are even hierarchical patterns of functional losses within these domains.10 Bennett et al.6 distinguish between IADLs and "complex" IADLs. ADLs and IADLs are widely used in epidemiological studies and frequently used by governments and policy-makers to distribute healthcare resources. The Health Insurance Portability and Accountability Act (HIPAA), for example, defines "chronic-illness" as 90 days or more of impairment in at least two ADL domains. The American Association of Retired Persons, using data from the 1994 National Long-Term Care Survey (NLTCS), estimates that full-time caregiving (36–41 h/wk) is required to accommodate this level of disability.11 As many as six (of eight) IADLs would have to be compromised to result in a care-giver burden comparable with "mild" ADL impairment. HIPAA's definition of chronic illness is too restrictive for dementia case finding and is likely to underestimate the prevalence of clinically diagnosable dementia. According to the NLTCS, only 2.3 million elderly persons in the United States report one or more ADL impairments, versus 4.4 million who report one or more IADL impairments. The latter figure is similar to dementia prevalence rates from elderly community samples.12 This suggests that one or more IADL impairments may serve as a better threshold for dementia case finding than would a similar degree of ADL impairment. The fact that many clearly demented persons receive less than full-time supervision and that ADL impairment develops after the appearance of both IADL impairment and clinical dementia, according to dementia "staging" measures such as the Global Deterioration Scale and the Functional Assessment for the Staging of Alzheimer's Disease, supports such a conclusion. Thus, it seems to me that impairments in IADLs, rather than ADLs, are most relevant to dementia's incident diagnosis. Hence, MCI cases should free of any cognitively mediated IADL impairment. However, as they are currently operationally defined, CIND/MCI cases are not free of IADL impairment.13 Bennett et al.6 report that 52 of 106 (49.1%) "nondemented" older persons need help with housework, 36.8% need help with shopping, and 38.6% report having not read a single article in the previous 6 months. Are these "normal" levels of functional capacity? Statistically, they are normal in Bennett et al.'s6 elderly sample, but these normal age-specific functional levels can nonetheless be associated with cognitive and extrapyramidal motor changes. Some authors discount the importance of functional status in the definition of MCI, because it does not reliably distinguish MCI from AD.14 This is a circular argument. If the cognitive correlates of IADL impairment defined dementia, then "MCI" cases with IADL impairment would be demented, and the failure of disability to distinguish dementia from nondementia would be resolved. The problem is that not all cognitive domains are equally well related to IADLs. My colleagues and I have shown that measures of executive function are stronger cross-sectional15 and longitudinal16,17 correlates of IADLs than are measures of general cognition and memory. Other authors include certain visuospatial measures, particularly nonverbal executive measures, in this class.18 Such cases risk being diagnosed with "mild" cognitive impairment, regardless of their functional status, merely because impairment in a single cognitive domain is not sufficient for the diagnosis of "dementia" according to the DSM-IV or ICD-10. Are such cases truly without dementia? I think not. Let me illustrate this with some data from the Freedom House Study (FHS). My colleagues and I have reported that the cross-sectional mean performance±standard deviation on the Executive Interview (EXIT25)19 for noninstitutionalized retirees (mean age 77.9) is 12.5±4.6.15 For comparison, consider that a score of only 15 of 50 perfectly predicts the capacity of elderly pulmonary patients to competently manage their inhalers20 and that the EXIT25 has been shown to explain 45% of the variance in patients' capacity to give informed consent for medical care using a rational reasons standard—56% of the variance in an "appreciative" standard.21 Thirty-eight percent of FHS subjects fail the EXIT25 at 15 of 50, consistent with Bennet et al.'s6 high prevalence of IADL impairment in their sample. However, the threshold that is most commonly used to identify "mild" cognitive impairment is 1.5 standard deviations below an age-specific mean. In the FHS, this would translate to an EXIT25 score of 20 of 50 (the EXIT25 is positively scaled). That score is above the mean for retirees at an assisted living level of care22 and comparable with that of institutionalized elderly patients, including those with schizophrenia.23 It has previously been shown that the executive impairment of elderly retirees is comparable with that of schizophrenic patients when the two groups are matched to the services they receive from their respective institutions.24 Should such debilitating (in the sense of IADLs) executive impairment be understood as "mild" merely because it does not meet CDR or DSM criteria for "dementia"? The MCI literature has come about as far as it can without specifying the details of its case definition. To move forward The class of functional impairments most relevant to incident dementia needs to be defined. The functional status measures best suited to assess these domains need to be identified, and the thresholds for dementia case finding empirically determined. It will be interesting to see whether the practice of using age-specific norms will survive once they have been applied to functional outcomes. If we drop this practice, the prevalence of "dementia" will greatly increase. If we retain it, we will be forced to defend ludicrous levels of functional capacity as "normal" and justify a high prevalence of cognitively disabled "nondemented" persons. The cognitive domains most relevant to these disability thresholds need to be empirically defined. These should be considered "essential" to dementia case definitions. "Optional" cognitive impairments, perhaps less strongly related to disability and hence dementia, may be useful in discriminating the preclinical stages of specific disorders from others. MCI thus becomes cognitive impairment in the "optional" domains in the absence of both "essential" cognitive impairments and functional limitations. I predict that we will find executive impairment to be essential to dementia. Memory impairment will be found to be optional, albeit useful, in predicting conversion to AD, specifically. The subset of amnestic persons with IADL impairment is also probably executively impaired (relative to young adults) and arguably "demented" already.17 Isolated executive impairment would become a dementing condition.25,26 This will effectively increase the prevalence of all-cause dementia in society, … but by then, we will be asking different questions. Financial Disclosure: None. Author Contributions: I was entirely responsible for the concept, design, and preparation of this manuscript. Sponsors' Role: There is no sponsor for this paper.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.