Recently revised criteria for mild cognitive impairment (MCI) will likely lead to far fewer diagnoses of Alzheimer disease (AD), according to one prominent critic of the new criteria. In a paper in the Feb. 6 online edition of the Archives of Neurology, he argues that the new definition means “you can interpret the same patient in different ways now because there is no clear boundary between the two conditions. It's totally arbitrary.”
But at least one clinician, who helped revise the criteria, disputes the assertion. “Essentially there has not been a change in clinical criteria of MCI,” said Ronald Petersen, MD, PhD, director of the Alzheimer's Disease Research Center at the Mayo Clinic in Rochester, Minnesota. [For the Neurology Today coverage of the criteria, published in 2011 in several papers in the Alzheimer's Association journal, Alzheimer's & Dementia, see “New Guidelines for Diagnosing Alzheimer Disease: First Update Since 1984”: http://bit.ly/xjt96T.]
While MCI is now a ubiquitous term, criteria for it were promulgated for the first time only in 1995, and limited the diagnosis to those patients with impairment in memory alone. In 2004, after almost a decade of experience with the diagnosis, experts expanded the criteria to include impairments in other cognitive domains, while requiring “essentially normal” functional activities. “The sole remaining difference after the criteria were revised in 2004 is that function is preserved,” said John C. Morris, MD, director and principal investigator of the Alzheimer's Disease Research Center at Washington University in St. Louis School of Medicine
The critical change in the new criteria developed in 2011, according to Dr. Morris, was in the third criterion for MCI: “Preservation of independence in functional activities.” As stated, it allows “mild problems” in daily activities such as paying bills or shopping, and includes those who require “minimal aids or assistance” to perform them. “In my mind that no longer equates to independence,” he said.
According to Dr. Petersen, however, such individuals are “functionally normal,” and though they may be experiencing inefficiency, they “can still perform independently, much the same as the 2004 criteria.”
The consequence, according to Dr. Morris, is that a large fraction of individuals previously classified as having mild AD could now be classified as having MCI instead. To quantify that assertion, Dr. Morris analyzed clinical and cognitive data collected from over 17,000 patients enrolled in federally funded Alzheimer's disease centers and maintained by the National Alzheimer's Disease Coordinating Center. Patients had diagnoses that spanned the range from normal cognition to MCI to probable AD.
Functional performance was assessed with the Clinical Dementia Rating (CDR) scale. This multipart instrument measures both cognitive loss and functional impairment. On this scale, which takes into account multiple functional categories, impairment is rated from none (0) to severe (3). A score of 0.5 indicates very mild functional impairment, and is consistent with the revised MCI criteria for maintenance of functional independence. A CDR score of 1 (mild impairment) could also be interpreted as consistent with MCI, Dr. Morris said.
Dr. Morris asked what proportion of those patients in the database with a diagnosis of AD (reflecting both cognitive and functional impairments) could be reclassified as having MCI based on their degree of functional independence under the revised criteria. He found that 99.8 percent of all individuals with a CDR score of 0.5, and 92.7 percent of those with CDR score of 1, could be reclassified as MCI.
“The categorical distinction between MCI and the milder stages of AD dementia has been compromised by the revised criteria,” he concluded.
But several other experts noted that MCI has been a difficult stage to define. “We often debate the construct of MCI,” said Dr. Petersen. Of the criteria for MCI, functional ability “is the most difficult aspect,” because it is the most difficult to assess. “There are no scales of function that adequately capture the mild end of the spectrum,” he said.
The difficulty in making the judgment of MCI versus AD, he said, is a consequence of making an earlier diagnosis. “The field has moved to identify people with clinical cognitive impairment in its very early stage — that's what MCI has done for the field. But it has also interjected an element of uncertainty since not all subjects with MCI have early AD.” That uncertainty, he said, can be reduced by using biomarkers as part of the diagnostic process, which was a major focus of the 2011 revision, although currently the use of such markers is beyond the reach of most community neurologists.
Clinically, he said, the criteria “have not changed” in their language regarding functional independence, but still depend on whether the patient has the ability to engage in normal activities or not. He noted that the individuals reclassified as having MCI in Dr. Morris's analysis were all judged to have dementia by expert clinicians at the Alzheimer's disease centers. “The exclusion of dementia is a criterion for MCI; so in this instance the CDR of 0.5 does not discriminate between MCI and dementia.”
Furthermore, Dr. Petersen said, the CDR is not well suited for making distinctions between MCI and AD, and was designed before the MCI concept existed. “It is too coarse a tool to answer these questions.” In the Mayo Clinic Study of Aging, one third of the MCI subjects have a CDR of 0, implying that the CDR is insensitive at the MCI stage, he explained. “Furthermore, two large studies on MCI at Harvard and Mayo have had to modify the CDR due its lack of sensitivity to discriminate.”
Marilyn S. Albert, MD, PhD, professor of neurology and psychiatry at Johns Hopkins University, headed the panel that revised the MCI criteria. “The boundaries between normal and MCI, and between MCI and dementia, are fuzzy ones. We all acknowledge that. I just don't see that eliminating that boundary is useful. There was unanimity in our group that people who have MCI have some mild functional difficulty. We'd all been seeing that. But I think everyone felt they could make that judgment. I think that's the heart of the issue.”
Dr. Morris pointed out that in his view, an essential reason it is hard to draw such lines is that in most cases, MCI is due to AD, and the process is one of continuous decline. “When there is underlying AD, my tenet is that there is no longer a need for the term MCI. We can just call it what it is: symptomatic Alzheimer's disease.”
But that is problematic as well, according to Deborah Blacker, MD, ScD, professor of epidemiology at Harvard School of Public Health, associate professor of psychiatry at Harvard Medical School, and director of the Gerontology Research Unit at Massachusetts General Hospital. Dr. Blacker said the revised criteria haven't shifted the line between MCI and AD so much as recognized that, given where it has historically been set, “there actually is functional impairment where people used to say there wasn't.”
More importantly, she added, “the earlier you go, the more at risk you are for errors of diagnosis.” The consensus in the field is that it's safer not to give the diagnosis of AD when there is some doubt, “even though we know some of them do have Alzheimer's disease,” Dr. Blacker added.
“Most of us feel we are pretty good at figuring out which is which, but we're by no means perfect. Most of us feel that's a dangerous thing to get wrong,” she continued, at least until some treatment is available to change the course of the disease. There are social and legal consequences of an AD diagnosis, and while the neurologist may recognize the inherent uncertainty in distinguishing what is clinical AD from what is not, that is not a subtlety that is recognized as much beyond the clinic.
Dr. Morris countered: “In my experience patients and their families, even in the early stages, would like to know the diagnosis. They ask for it. A more nebulous diagnosis such as MCI is more anxiety producing than the diagnosis of Alzheimer's disease. At least if you have a diagnosis, you can begin to deal with it.” In the earlier stages, he noted, the patient is better able to make decisions such as assigning durable power of attorney.
Dr. Petersen disagreed, however. He noted that a 2010 survey of 420 AAN members documented that 80-90 percent of practicing neurologists prefer the use of the term MCI in this instance.
For the community neurologist, the distinction may not be as important as for the academic researcher, Dr. Blacker said. Probably the most informative sign of cognitive decline is that the patient has come to the clinic with a complaint of memory problems. “Seeking treatment is in some ways the biggest predictor of all,” she said, “since it means the patient has already decided there is a problem. Something's going on or they wouldn't be there. And then in some ways, the rules don't matter.”