Most Cases of Mild Cognitive Impairment Do Not Become Dementia

Pauline Anderson

April 14, 2009

April 14, 2009 ( updated with commentary April 17, 2009 ) — The number of patients with mild cognitive impairment (MCI) who progress to dementia is at least half of what it was previously believed to be, new research suggests. A large meta-analysis showed that the cumulative risk over 10 years ranged between 30% and 50%, depending on whether the studies that were analyzed used a definition of MCI that included subjective memory complaints.

Until now, the prevailing opinion was that the progression rate from MCI to dementia was about 10% per year, or a 100% conversion to dementia over 10 years.

This research suggests that instead of always being an invariable transitional state between normal aging and dementia, MCI is a condition in which some patients stay static and some even improve, principal investigator Alex J. Mitchell, MD, from the University of Leicester, in the United Kingdom, told Medscape Psychiatry.

The review is published in April issue of Acta Psychiatrica Scandinavica.

Still Not Good News

But even with this study's more optimistic estimates of dementia risk, doctors should be careful when interpreting this information for patients, said Dr. Mitchell. "Let's say a patient is in a high-risk group, and the risk is 50% over 10 years, that's still not good news," he said. "Telling someone they have half the risk of developing a condition that is greatly feared is not generally good news, even if it was worse news before."

In a search of various databases, researchers found 41 studies that fit their criteria for inclusion in the analysis. Among other things, the studies had to last a minimum of 3 years and include at least 20 subjects.

Researchers excluded case-control studies with no longitudinal period of observation. They also excluded studies of Lewy-body dementia or frontotemporal dementia because of the lack of data. The 41 studies included 25 that were conducted using strict Mayo Clinic criteria for MCI and 16 with non-Mayo criteria. The researchers also categorized studies into those carried out in a specialist setting and population studies carried out in the community. Three of the studies that used the Mayo Clinic definition of MCI were randomized trials. Of the remaining 22 studies, 14 were clinical studies conducted in specialist centers and 8 were population-based studies. Of the 16 studies conducted using non-Mayo Clinic criteria, 4 took place in specialist centers and 12 were population studies.

The mean observation period was 4.57 years for studies using a Mayo Clinic MCI definition and 4.59 years for those with a non-Mayo definition.

Across all studies, researchers analyzed the rates of conversion from MCI to dementia. After adjustment for sample size, in the studies adhering to the Mayo Clinic MCI definition, the cumulative risks over approximately 5 years in specialist settings were: 39.2% for progression to dementia; 33.6% for progression to Alzheimer's disease (AD); and 6.2% for progression to vascular dementia (VaD). In population studies, the respective cumulative risks for dementia, AD, and VaD were: 21.9%, 28.9%, and 5.2%.

Higher Risk in Clinical Setting

The cumulative risks were higher in specialist settings because patients are typically referred to these clinics by primary-care physicians concerned about memory problems, whereas patients in the community setting are sought out by researchers conducting population-based cognition studies, said Dr. Mitchell.

"You would be at high risk by the very nature of that process of being seen in a clinical setting," he said.

The respective cumulative risks in studies using the non-Mayo Clinic definition of MCI were, as expected, generally lower than those using this definition. The cumulative rate averaged 32.3% for those with Mayo Clinic–defined MCI and 24.1% for those with non-Mayo criteria. The cumulative proportion that progressed to dementia rarely exceeded 50%, even in long-term studies. In the 6 studies with an observational period of 5 years or more, the cumulative conversion rate was only 38.2%.

As for the annualized conversion rate (ACR), in studies adhering to the Mayo Clinic definition of MCI, after sample size was corrected for, these rates in specialized settings were: 9.6% for dementia, 8.1% for AD, and 1.9% for VaD.

The adjusted ACR in community settings was 4.9% for dementia, 6.8% for AD, and 1.6% for VaD. Again, these rates were generally lower in studies using non–Mayo Clinic criteria

Across all 41 studies, the overall ACR was 6.7% for progression to dementia, 6.5% for AD, and 1.6% for VaD.

The relative risk for progression was 15.9 (MCI Mayo type) and 6.2 (MCI non-Mayo type) for dementia and 9.5 (MCI Mayo type) and 4.7 (MCI non-Mayo type) for AD when compared with healthy elderly patients.

Modeling the development of dementia over time showed a "plateau" at 10 years of 50%, said Dr. Mitchell.

Challenging Accepted Progression Rate

This study challenges the previous accepted annual progression rate of 10%, which, if true, would mean that all surviving MCI sufferers would have dementia within 10 years.

"The chances of developing dementia with MCI, at least for the foreseeable future — which is the length of a study and in this case is up to 10 years — is somewhere between 30% and 50%, depending on the relative risk," said Dr. Mitchell. "Before, most people thought that the cumulative risk was basically 100%."

The cumulative risk is roughly arrived at by using a simple model of multiplying the annual rate of progression to dementia by the number of years studied. However, the actual numbers do not always add up, because, as Dr. Mitchell explained, the risk continues after the end of the study; there are outcomes other than dementia; and ACR and cumulative-risk statistics have different weightings.

Those previous rates were arrived at using short-term studies or small samples, said Dr. Mitchell. He added that it is only now that there are sufficient studies with large samples that followed subjects for long enough to make a solid meta-analysis possible.

MCI Has Subgroups

The study suggests that instead of being a homogenous transitional phase between normal aging and dementia, MCI includes several subgroups, some of which go on to develop dementia, some of which stay static, and some of which even improve, said Dr. Mitchell.

"This paper is just another piece of evidence that this MCI group is not just 1 explanation (for memory problems), that there are different things going on; the groups may look the same at certain points, but they will have slightly different trajectories," he said.

As an example, Dr. Mitchell used patients with MCI whom experts presume are destined to develop AD. "If you're able to examine brain pathology in life, you'll find that many individuals with MCI are not in a pre-Alzheimer's group at all, that there are other explanations for their memory decline." One common cause for such memory decline is depression, and when the depression is treated, the memory decline is reversed, he said. Another possible explanation for memory problems is "silent" cerebral infarction. In many cases, patients suffer a stroke and do not present to the hospital or, if they do, their stroke is not picked up on magnetic resonance imaging, said Dr. Mitchell. "It's only when you test their memory that you find they have these memory complaints."

Progression from MCI to dementia is fraught with "gray areas," said Dr. Mitchell. For 1 thing, while patients who complain of memory problems are considered at increased risk to develop dementia, just because a patient does not complain of memory problems does not mean these problems do not exist.

Such patients may not want to talk about their memory problems or prefer to minimize them out of embarrassment. "Many of these people are functioning reasonably well and many are still working," said Dr. Mitchell. "They want to be doing as well as they can for as long as possible."

He pointed out that other factors that were not measured in the meta-analysis — including biological and psychological factors — could pose additional risks in terms of progression to dementia.

While these results are more favorable than previously thought, they should be interpreted carefully, said Dr. Mitchell. "I would cautiously say that this is not as bad news as it once was, but I wouldn't go so far as to say that all patients do well."

A good reason to be cautious is that patients with MCI are at higher risk for mortality and other negative outcomes than the general population, said Dr. Mitchell. He and his colleagues will investigate these risks more closely in future research.

Results Should Be Considered With Caution

Asked by Medscape Psychiatry to comment on the meta-analysis, Ronald Petersen, MD, PhD, principal investigator of the Mayo Clinic Study of Aging, said the findings should be considered with a little caution, as some of the studies included relatively young patients.

"A study that includes people in their 40s and 50s in the overall analysis will get very different conversion rates than a study that includes people in their 70s and 80s. If you're evaluating a set of criteria, that's an important factor to take into account," he said.

He noted the 50% conversion rate indicated in the study is perhaps "a little too optimistic. If I diagnose someone who is 75 years old with MCI and they don't have depression, the conversion rate is a lot higher than 50% over 10 years."

However, Dr. Petersen added, if the study findings result in physicians having more discussions about cognitive problems with their patients, then that is a good thing. "That would be wonderful, and I wouldn't detract from that message."

The authors report no relevant conflicts of interest.

Acta Psychiatr Scand. 2009;119:252-265. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: