Alzheimer’s cognitive testing improved with multivariate analysis

Written by Lauren Pulling

Researchers from the Hunter Medical Research Institute and the University of Newcastle (both NSW, Australia) have developed a novel multivariate approach to improve methods of cognitive testing in Alzheimer’s disease (AD) patients. Using the new model, the researchers were able to analyze the cognitive performance of patients over time and identify five distinct patient groups.
At present, cognitive impairment in AD patients is measured using the Mini-Mental State Examination (MMSE) test. Patients’ responses to five types of questions are monitored and then scored using an algorithm. However, critics have suggested that the MMSE test is not entirely valid as scores can be affected by factors such as educational background.

In the new study, published in Future Science OA, the researchers present a new multivariate approach for cognitive testing. 3717 patients, taken from the Coalition Against Major Diseases dataset, were analyzed and could subsequently be identified as belonging to one of five patient groups: cognizant, inattentive, forgetful, distant and absent. These groups were demonstrated to be distinct in terms of both characteristics and prognostics. Additionally, the team identified that only three out of the five MMSE question categories were essential for grouping: registration, attention and recall.

The researchers, led by Pablo Muscato (University of Newcastle), commented: “This project [was] motivated by the observation that the CAMD database is an aggregation of several trials that employed varying cognitive test versions and are therefore generally incompatible or not directly comparable among each other in terms of the calculated patients’ overall cognitive performance.

“Our goal [was] to resolve which test categories contribute to the patient group differentiation the most and to build a model that can be employed to reliably assign a patient to an AD cognitive group associated with certain impairment characteristics and clinical prognosis, independent of the questionnaire version.”

In addition, the team analyzed transitions between groups. First author Inna Tischenko (Hunter Medical Research Institute and the University of Newcastle) explained: “We analyzed the cognitive performance of AD patients over a time frame of up to 4 years and obtained hints that multivitamins might slow progression only in the group that presents a mild impairment in recalling.” Following analysis of AD progression, the team also indicated the possible existence of subgroups.

Sources: Tishchenko I, Riveros C, Moscato P. Alzheimer’s disease patient groups derived from a multivariate analysis of cognitive test outcomes in the Coalition Against Major Diseases dataset. Future Science OA doi:10.4155/fsoa-2016-0041 (2016) (Epub ahead of print);

You can read the full article for free here.