6

Subgroups of Mild Cognitive Impairment and Plasma Phospho-Tau217 Predict Dementia with or Without Alzheimer Pathology. A Four Year Follow up of Mild Cognitive Symptoms

Susanna Vestberg, Department of Psychology, Lund University, Lund, Sweden
Pontus Tideman, Dep. of Clinical Sciences, Clinical Memory Research Unit, Lund University, Malmoe, Sweden
Shorena Janelidze, Dep. of Clinical Sciences, Clinical Memory Research Unit, Lund University, Malmoe, Sweden
Sebastian Palmqvist, Dep. of Clinical Sciences, Clinical Memory Research Unit, Lund University, Malmoe, Sweden
Oskar Hansson, Dep. of Clinical Sciences, Clinical Memory Research Unit, Lund University, Malmoe, Sweden



Objective:

The study aim is to predict dementia after four years in patients with mild cognitive symptoms using a classification into four subgroups of Mild Cognitive Impairment (MCI) and one of Subjective Cognitive Decline (SCD), alone or combined with plasma tau phosphorylated at threonine 217 (p-tau217).

Participants and Methods:

Patients (N=303) with mild cognitive symptoms, no dementia from the BioFINDER study were included. Females: n=140, (46.2%), age: 71.14 ± 5.5 years, education: 11.7 ± 3.5 years. The cognitive domains Verbal ability, Visuospatial construction ability, Episodic memory and Executive functions were assessed using two tests per domain. Premorbid cognitive level was estimated with the Swedish version of the National Adult Reading Test (NART-SWE). The raw score from each test was converted to a z-score according to published norms. The mean z-score from each cognitive domain were used in the classification into four MCI groups: amnestic MCI (single or multiple domains: aMCI/maMCI), non-amnestic MCI (single or multiple domains: naMCI/mnaMCI), and SCD. MCI was classified when a mean z-score of a cognitive domain was -1.5 or lower compared to the NART z-score.  Plasma p-tau217 concentration was measured using assay on a Meso-Scale Discovery platform. The five subgroups were predictors as were plasma p-tau217, controlling for age, sex and education, in multinomial logistic regression analyses. Primary outcome was diagnosis after four years: Alzheimer’s Disease (AD) (n=81), other dementia (n=44), and no dementia (n=178; reference category).

Results:

The five NART based subgroups predict 74.3 % of the patients correct (R2 0.516, p<0.001, with covariates included): no dementia=92.1 %, AD=63 %, other dementia=22.7 %. Plasma p-tau217 predict 66.3 % correct (R2 0.382, p<0.001, with covariates included); no dementia=86.5 %, AD=58 %, other dementia=0 %. Combining NART-based subgroups and plasma p-tau217 predict 76.2 % correct (R2 0.625, p<0.001); no dementia=88.8 %, AD=69.1 %, other dementia=38.6 %, χ2(16, N = 303) = 229.439, AIC: 379.459, p< .001. According to Wald statistics, maMCI and plasma p-tau217 were the strongest predictors for subsequent AD, whereas maMCI and mnaMCI were the strongest predictors for subsequent dementia of other aetiologies.

Conclusions:

The classification into SCI or one of four MCI groups based on a restricted neuropsychological test-battery, is clinically meaningful in assessing the outcome in a medium-long time frame for patients with mild cognitive symptoms. In combination with plasma p-tau217, the prediction is strengthened, especially of progression to dementia without AD pathology. Neuropsychological tests and blood tests are methods that can easily be implemented in primary care to meet the increasing numbers of patients with incipient dementia.

Category: Neurodegenerative Disorders

Keyword 1: neuropsychological assessment
Keyword 2: dementia - Alzheimer's disease
Keyword 3: dementia - other cortical