Estimation of Cognitive Performance Based on Premorbid Intelligence in Parkinson's Disease

The estimation of premorbid intelligence (PI) is needed for an accurate diagnosis.

This study aimed to estimate the cognitive performance taking into account the PI in Parkinson's disease (PD) compared to healthy controls (HC); and to analyze the discrepancies between the current and the predicted cognitive performance based on the PI.

Semantic fluency, verbal and visual memory, and executive functions were assessed in 39 PD and 162 HC. A linear regression model was used to analyze the discrepancies between the predicted cognitive performance and the current raw scores through PI variables (Word Accentuation Test (WAT), Pseudo-Words (PW) Reading subtest from PROLEC-R, age, and years of education). ROC analyses were performed to assess their diagnostic properties.

Significant differences were found in the raw cognitive scores between patients and HC [semantic fluency (t = 6.07; p < 0.001), verbal memory (t = 6.63; p < 0.001), and executive functions (t = 2.57; p = 0.013), and in visual memory (t = 1.97; p = 0.055 marginally significant)]. Compared to HC, PD patients presented higher discrepancies between the predicted cognitive performance and the raw scores in semantic fluency, verbal memory, visual memory, executive functions (AUC = 0.78, 0.78; 0.64, 0.61, respectively).

The magnitude of the discrepancies scores between the current and the predicted cognitive performance based on PI indicates the presence of cognitive decline in the specific cognitive domain in PD patients. This study underlines the usefulness of premorbid measures and variables, such as WAT, PW, age, and years of education, to more accurately estimate the cognitive performance in PD.