Prevalence of cognitive impairment in motor neuron diseases: alternative norming methods lead to considerably diverse estimates.

Abstract

Objectives: Cognitive impairment (CI) is frequently observed in motor neuron diseases (MNDs), and the Edinburgh Cognitive and Behavioral ALS Screen (ECAS) represents the most widely used measure to evaluate it. For the Italian ECAS, two norming approaches are currently available: the "regression-based" (R) and the "2 standard deviation-based" (2SD). In this study, we therefore aimed at comparing those two methods for the detection of CI in MNDs. Methods: Raw ECAS scores from 160 MND patients were corrected using both the R and the 2SD methods. According to Strong's criteria, patients were then categorized as either cognitively normal (CN), with CI (ALSci), with behavioral impairment (ALSbi), or both (ALScbi). Results: Using the R approach, the frequency of below-cutoff performances was 3.12% for the ECAS total, 4.37% for the ALS specific, and 3.75% for the ALS nonspecific score. Using the 2SD method, the prevalence increased to 25.62% for the ECAS total, 21.25% for the ALS specific, and 23.12% for the ALS nonspecific score. The same increase was observed across all single tasks except for the digit span backward. The R-based frequency of Strong's diagnoses was 7.50% for ALSci, 15.62% for ALSbi, and 3.14% for ALScbi, which became 24.38% for ALSci, 8.75% for ALSbi and 10% for ALScbi with the 2SD method. Conclusions: The norming approach has a significant impact on the estimated prevalence of CI in MNDs, with the R method representing the most conservative one. These findings highlight the need for harmonized protocols in future studies evaluating CI in MNDs.