Predictor model for mild cognitive impairment in Parkinson's disease using visual rating scales

Early detection of cognitive impairment in Parkinson's disease (PD) may reduce the risk of developing dementia. The assessment of brain atrophy may be a biomarker to differentiate between PD patients with Mild Cognitive Impairment (PD-MCI) from the unimpaired patients (PD-CU). The study aims to investigate the detection of MCI in PD through the visual rating scales of atrophy.

Twenty-nine PD patients underwent a cognitive assessment and a structural magnetic resonance scan. Based on the cognitive assessment, PD patients were classified as PD-CU (n = 16) and PD-MCI (n = 13). Furthermore, 12 healthy adults (HC) were included as a control group. Brain atrophy was evaluated by two raters using a protocol of visual rating scales and the inter-rater agreement was computed using the weighted Kappa. ANCOVAs and post-hoc tests were used to determine the difference in atrophy between the three groups. Logistic regression analyses were performed to identify the best predictors in the differentiation between the groups.

The agreement between the raters was good to excellent. Higher scores of both the Medial Temporal lobe Atrophy (MTA) and the Occipital lobe (OC) scales were found in PD-MCI as compared to the PD-CU patients and HC. Lastly, MTA seemed a good predictor for differentiating the cognitive profile between the PD groups, while the OC between PD-MCI and HC.

Visual assessment of MTA and OC may be a useful index to support the presence of cognitive impairment in PD patients, particularly as OC atrophy also discriminates between HC and cognitively impaired PD patients.