Objective: We tested the performance of the SelfCog® to classify PD patients as MCI.

Background: MCI is a strong risk factor of dementia in PD patients. Differentiating patients with MCI from cognitively normal subjects requires an in-depth evaluation by a neuropsychologist that lasts more than an hour, which could be advantageously replaced by an automated rapid assessment.

Method: PD patients (n=41) fulfilling UK PDS Brain Bank criteria were selected consecutively in a Movement Disorder clinic based on disease duration ≥5y, age <75y, French as native language, absence of dementia (MMS>24). Patients were classified into PD-nonMCI (n=24, age mean 67.12) or PD-MCI (n=17, age mean 64.59) by using a comprehensive cognitive assessment (Level II from the MDS Task Force). All PD patients and 150 healthy controls were tested with the SelfCog®, which assesses motor, language, memory, visuospatial and executive functions in 15 min. During the test, participants must press a button in response to visual cues. Inverse efficiency scores (IES) were calculated by dividing response time by accuracy for each function. A global cognitive score was calculated by averaging each cognitive IES. Linear regression analyses were performed to test for differences in IES between the PD patient groups and 53 controls matched for age and education. IES were transformed to z-scores based on normative data established from the control cohort. Concordance of MCI diagnoses by the SelfCog®, using a range of standard deviation (SD) cutoff scores, were compared to diagnoses based on MDS criteria.

Results: PD-MCI global (p<0.001), executive (p<0.0001) and language (p<0.01) IES were higher than controls. PD-MCI executive IES was higher than that of PD-nonMCI (p<0.01). When using a cutoff of 1 SD below norms for at least 2 cognitive z-scores, we were able to classify PD patients into MCI or nonMCI with 82% sensitivity and 63% specificity.

Conclusion: Despite the small number of subjects analysed, the SelfCog® was able to differentiate PD-MCI from controls and PD-nonMCI in only 15 minutes. It allowed to classify PD subjects with good sensitivity. Longitudinal follow-up is necessary to better assess its specificity since 2 patients classified as MCI with the SelfCog® and nonMCI according to the MDS criteria were diagnosed with dementia 2 years later.

« Back to 2022 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/selfcog-a-digitized-cognitive-battery-to-classify-parkinson-disease-pd-patients-with-or-without-mild-cognitive-impairment-mci/