New Tool Helps Predict Transition to MCI

Megan Brooks

March 24, 2015

Researchers have developed a new scoring system that can help identify cognitively normal elderly individuals at risk for mild cognitive impairment (MCI).

"This risk scale may be an inexpensive and easy way for doctors to identify people who should undergo more advanced testing for memory issues or may be better candidates for clinical trials," study author Ronald C. Petersen, MD, PhD, from Mayo Clinic in Rochester, Minnesota, said in a statement.

The long-running population-based Mayo Clinic Study of Aging (MCSA) has identified several factors associated with risk for MCI, including age, education, sex, APOE genotype, parkinsonism, diabetes, depressive symptoms, cardiovascular disease, stroke, and slow gait.

Dr Ronald C. Petersen

Using these risk factors, the MCSA investigators developed an algorithm to predict the risk of transitioning from cognitively normal to MCI. They describe it in an article published online March 18 in Neurology.

"The basic model includes factors that could be easily assessed from the medical records, such as years of education, history of stroke or diabetes, and smoking," author Michelle M. Mielke, PhD, also from the Mayo Clinic, told Medscape Medical News. "Obtaining this information would only take a few minutes and the risk score could be calculated prior to seeing the patient. This information might be beneficial for doctors to identify which individuals might be at greatest risk of MCI and should be followed more closely. This could include brief cognitive screens and questions of the patient and caregiver with regards to changes in cognition," she explained.

Dr Michelle M. Mielke

"The augmented model," Dr Mielke noted, "includes information that can be obtained at a clinic visit, including a cognitive screen and questions on memory and behavior to both the patient and caregiver. This would take some additional time. However, if a patient scores high in the basic model, this additional information would be beneficial in determining whether the patient should be referred to more advanced testing for memory issues, including an extensive cognitive profile and neuroimaging."

The researchers tested their models in 1449 randomly selected people from Olmsted County, Minnesota, who were aged 70 to 89 years and cognitively normal at baseline. Participants were evaluated at baseline and every 15 months for 4.8 years. During this time 401 people (27.7%) developed MCI.

The basic model had a C statistic of 0.60 (0.58 for women and 0.62 for men). Women in the highest risk quartile had a 2.1-fold higher risk for MCI relative to their peers in the lowest quartile. Men in the highest quartile had a 3.0-fold high risk for MCI than men in the lowest quartile.

The performance of the augmented model was "significantly better," with a C statistic of 0.70 (0.69 for women, 0.71 for men). Both men and women in the highest vs lowest sex-specific quartiles of the augmented model's risk scores had a 7.2-fold higher risk for MCI. In this model, the Short Test of Mental Status was the strongest predictor of MCI risk. A nonzero score on the Clinical Dementia Rating–Sum of Boxes and measures from the Neuropsychiatric Inventory were also associated with an increased risk for MCI, the authors note.

While APOE ε4 carrier status improved the estimation of risk scores at the individual level, "the C statistic for the model that incorporated this genetic feature was unchanged from the augmented clinical model. This finding highlights a need to conduct further population-based studies to determine whether additional biomarkers improve risk stratification for MCI, and ultimately dementia," they say.

Dr Mielke cautioned, though, that this tool is "not yet ready for prime time. The present study developed the scoring system. However, it needs to first be validated in other populations before it should be used by clinicians."

The "Olmsted County MCI Risk Score"?

Alan B. Zonderman, PhD, from the National Institutes of Health, Baltimore, Maryland, and Timo Grimmer, MD, from Technische Universität Munich, Germany, comment on this research in a linked editorial.

"For medical practice, this risk score provides a tool with which to advise patients and relations about likely prognoses," they write. "For clinical research, this risk score may provide a simple and inexpensive way to identify at-risk candidates for primary prevention trials."

For now, however, "clinicians are well advised to utilize the risk score with caution, particularly when applying it to less-educated persons with non-European origins. Various trials in other specialties demonstrated the utility of studying at-risk samples, most notably in coronary heart disease. Similar principles should apply to MCI and dementia," Dr Zonderman and Dr Grimmer note.

The Mayo Clinic team, they add, has "shown considerable humility by neglecting to name their new risk index after their study, town, or clinic. With hopes that it will fulfill its initial promise, we renamed it for them" — The Olmsted County MCI Risk Score.

The study was supported by the National Institutes of Health, the Robert Wood Johnson Foundation, the Robert H. and Clarice Smith and Abigail van Buren Alzheimer's Disease Research Program, and the Rochester Epidemiology Project. The researchers made several disclosures; all are listed with the original article.

Neurology. Published online March 18, 2015. Abstract Editorial


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.