Memory Complaints Predict Future Cognitive Impairment

Pauline Anderson

September 24, 2014

Patients reporting memory problems are at increased risk for subsequent mild cognitive imapirment (MCI) or dementia, a new study suggests.

The study also showed that patients with these subjective memory complaints (SMCs) harbor Alzheimer disease–like brain pathology, even if they had no observable cognitive impairment.

In light of these findings, physicians should ask older patients about their memory, said lead author Richard J. Kryscio, PhD, professor, statistics, and associate director, Center on Aging, University of Kentucky, Lexington.

"If patients are complaining about their memory and are older, then my advice to the typical clinician would be to start monitoring their cognition," he said. "If it doesn't look in the next 2 or 3 years like they are doing any worse than at baseline, I would sort of dismiss it, but if it's starting to go south, then you're looking ahead at something likely to happen."

At that point, some clinicians may want to intervene diagnostically, for example with MRI or spinal tap, to get more information, he said.

The study is published online September 24 in Neurology.

BRAiNS Study

The new analysis included 531 cognitively intact participants from the longitudinal Biologically Resilient Adults in Neurological Studies (BRAiNS) at the University of Kentucky's Alzheimer's Disease Center. As part of the study, they were offered APOE genotyping and agreed to donate their brain upon death.

The study cohort was relatively young (mean age, 73.2 years), mostly female (63.1%), and highly educated; had a high proportion of APOE ε4 carriers (30.3%); and included many current (9.2%) and former (43.5%) smokers.

At annual visits, participants completed multiple tests of memory, language, and executive and visuospatial function. They were also asked if they had noticed any change in memory since their last visit.

On the basis of results of these assessments, researchers classified participants into 4 cognitive states: no serious impairment (NSI), SMC, MCI, or dementia. They looked at probabilities of moving into an impaired state and the timing of this change and adjusted for intervening death.

The researchers found that 55.7% of the study participants went from having no serious impairment to entering the SMC state. SMCs represented a higher risk for a subsequent diagnosis of cognitive impairment, either MCI or dementia, vs dying, when compared with other states (odds ratio, 2.8; 95% confidence interval, 1.9 - 4.2; P < .0001).

The change occurred on average at age 81.5 years, or about 8.3 years after enrollment into the study.

Investigators looked at several risk factors that might affect the time to transition from one clinical state to another. These included APOE carrier status, female sex, low education, smoking history, family history of dementia, high blood pressure, use of hormone replacement therapy (HRT), and type 2 diabetes.

The analysis showed that the presence of type 2 diabetes and current smoking actuallly decreased the odds of transitioning into the SMC state. Although it seems counterintuitive, smoking "is a competing risk of premature death" and so decreases the at-risk period for SMC to occur, said Dr. Kryscio. "Half of the participants were smokers at one point or another and they tended to die earlier."

Diabetes is a bit more difficult to explain, he said. "We're still in a sense scratching our heads about that." It could be that participants who had diabetes volunteered for the study because they were already worried about their memory. "We didn't ask them about their memory at baseline," said Dr. Kryscio.

While high blood pressure promoted transition from NSI directly into MCI, being female reduced the risk for this transiton. The gender effect "is a kind of strange phenomenon," commented Dr. Kryscio. "Several others have seen this in their data as well — that if you're a male you're more likely to become cognitively impaired; however, if you're female and become mildly cognitively impaired, you are more likely to move on to dementia."

This could somehow be linked to longevity in women compared with men, he said.

Once in the SMC state, those with the APOE ε4 risk allele had increased odds of moving into MCI or dementia (OR, 2.2; P = .036).

Transition Timing

Various risk factors affected the timing of transitions. For example, the transition from SMC directly to dementia for HRT users was lenghtened by 9.7 years. However, Dr. Kryscio said he's "leery" about this result because it's probably due to small numbers. He also noted that participants were asked about their HRT use at baseline, when they were about 72 years old, and were not questioned about it after that.

From enrollment, it took on average 9.4, 11.0, and 16.8 years for patients to enter states of MCI, dementia, and death, respectively. The time to death was shortened by almost 5 years for former smokers and almost 9 years for current smokers.

That these various risk factors can lengthen or shorten transition from memory complaints to impairment suggests that there's a period during which interventions may be helpful. Although nothing definitive has been proven, there's some suggestion that a regular exercise regimen can have a positive cognitive effect, said Dr. Kryscio.

Almost half (45.7%) of the cohort has now died. With autopsy information, researchers have classified these participants as having or not having SMC, and with and without impairment (MCI or dementia).

This analysis showed that 42.9% of SMC participants died with no observable clinical impairment. Among these unimpaired participants, the mean neuritic plaque counts in the medial temporal lobe and neocortex were significantly higher than for SMC-negative participants (P < .03 for both).

Dr. Kryscio said his group is among the first to provide this kind of evidence of a connection between memory complaints and AD pathology.

A limitation of the study was that SMC assessment involved asking just 1 simple question. No informant was involved, and patients with SMC were not probed further. APOE ε4 carriers and those with a family history of AD were over-represented in the study.

In addition, while SMCs are often associated with depressive symptoms, objective ratings of depression were unavailable in this study; however, participants with clinical depression were excluded from BRAiNS. Another study limitation was that risk factors, such as smoking, diabetes and high blood pressure, were recorded only at baseline and could have changed throughout the study.

Meaningful Trend

Ron Petersen, MD, PhD, director, Alzheimer's Disease Research Center, Mayo Clinic, Rochester, Minnesota, praised the research.

"It's a good study done by good group of people and adds more evidence that this trend toward subjective concerns is meaningful and could be important in terms of predicting what's going to happen down the road," he told Medscape Medical News.

Although the study doesn't provide much new in terms of results, it does have some unique features, said Dr. Petersen. In addition to the modeling used to predict transitioning between states, a novel aspect of the study is the state-of-the-art autopsy evaluations.

These analyses found an association between SMCs and elevated neuritic plaques in the neocortex and medial temporal lobe, "which are Alzheimer's kinds of pathologic features," noted Dr. Petersen.

"I think what they are implying — and it's up to the reader to believe it or not — is that these people could have actually been suspicious of themselves, and that suspicion was ultimately founded in them having more pathology at time of death."

Biomarker studies — including one that Dr. Petersen's group is carrying out — are starting to show that "normal" people who have subjective memory complaints have a higher prevalence of Alzheimer's biomarkers than other "normal" people who don't have these complaints.

"This new study takes it beyond biomarkers and actually looks at the pathology in these people after they died," noted Dr. Petersen.

The study was supported by the National Institutes of Health, the National Institute on Aging (NIA), and the National Center for Advancing Translational Sciences (NCATS). Dr. Kryscio is the associate editor for biostatistics for Neurology. Dr. Kryscio is partially supported by grants from the NIA, NCATS, the National Institute on Drug Abuse, and the National Institute of Neurological Disorders and Stroke and receives a consulting fee from the Alltech Corporation.

Neurology. 2014;83:1359-1365. Published online September 24, 2014. 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.