Diagnostic and screening power of neuropsychological testing in detecting mild cognitive impairment in Parkinson's disease.

Prevalence of mild cognitive impairment (MCI) in Parkinson's disease (PD) is variable likely due to methodological differences in classification criteria and lack of consensus about neuropsychological tests used for cognitive profiling. The main objective of our study was to identify the most suitable neuropsychological tests and determine their screening and diagnostic cutoff scores for PD-MCI. A series of 104 consecutive PD patients performed an extensive neuropsychological evaluation. Individual test values were converted into Z-scores using relative published normative data. According to published criteria, PD patients were categorized as PD-CNT (PD without cognitive impairment), PD-MCI (patients performing -1.5 SDs below the mean score in at least one cognitive domain), and PDD. We used receiver operating characteristic (ROC) curves and K-means clustering analyses to calculate the best discriminating power of each neuropsychological tests in detecting PD-MCI. PD patients were categorized as follows: 55 PD-CNT (53 %), 34 PD-MCI (33 %), and 15 PDD (14 %). PD-MCI had lower education, longer disease duration and greater frequency of hallucinations than PD-CNT. We found that only the Trail Making test, Rey-Osterrieth Complex Figure Test (ROCF) copy, Frontal Assessment Battery (FAB), Digit Span Backward, and Rey's word auditory verbal learning test (RVLT) immediate recall reached significant screening and diagnostic validity in predicting PD-MCI (AUC 0.705-0.795) with cutoff scores calculated by ROC analyses lying within normal range for normative data. Specific neuropsychological tests covering verbal memory, attention/set-shifting, and visual-spatial deficits are the best predictors of MCI in PD if valid cutoff scores are used. These results have consequences for cognitive diagnosis and potentially in establishing the rate of PD cognitive decline.