Regression-based normative data for neuropsychological assessment in Buenos Aires, Argentina: RAVLT, trail making test, digit span, and verbal fluency

Abstract

Objective: To develop regression-based normative data for a set of widely used neuropsychological tests and to provide a freely accessible normative data calculator for use in Buenos Aires, Argentina. Methods: Participants were drawn from a large clinical dataset and selected to reflect preserved global cognition, intact memory performance, and functional independence. Regression-based norms were developed for the Rey Auditory Verbal Learning Test, Verbal Fluency, Trail Making Test, and Digit Span tasks. For each model, predictors included age, sex, and years of education, as well as polynomial terms and interaction effects; final models were obtained via stepwise backward elimination procedures. Model assumptions were evaluated, and predictive performance was estimated using 10-fold cross-validation. Results: 393 participants met inclusion criteria. Age was negatively associated with verbal memory, fluency, and executive function, whereas higher educational attainment was associated with better performance. Women outperformed men in verbal memory and phonological fluency. Demographic variables explained between 2% and 23% of variance across tests. An open-access calculator was developed to facilitate individualized normative estimation in clinical settings. Conclusions: This study provides locally developed, regression-based norms for neuropsychological assessment in Buenos Aires – Argentina, addressing the lack of culturally relevant norms and limitations of traditional methods.