Re‐examining the impact of adjusting for education and sex the RBANS delayed memory score: Implications on the cut‐offs used to detect cognitive impairment in prodromal Alzheimer’s disease. Data from the EPAD‐LCS study.

Abstract

AbstractBackgroundthe RBANS is a multilingual brief cognitive battery commonly used in both clinical practice and research. Its delayed memory index has been used as entry criteria for clinical trials in early stages of Alzheimer’s disease (AD). Although several studies reported a significant effect of education and sex in RBANS scores, the published normative data only provide adjustments by age. We aimed at exploring the impact of sex and education in the age‐adjusted scores and studying the percentage of change in the classification of memory impairment, along with the association with AD biomarker status after adjusting for education and sex.MethodsWe analyzed data from the EPAD baseline Longitudinal Cohort Study (n = 1,716 [Clinical Dementia Rating [CDR] = 0 n = 1,233; CDR = 0.5 n = 483]), see Table 1. We used multiple regression to assess the impact of years of education and sex on the age‐adjusted delayed memory index within the CDR = 0 sample. The coefficients obtained were used to generate sex‐, education‐ and age‐adjusted scores. The whole sample was classified as having/not having memory impairment (<85 points) with published and full adjusted scores. Within CDR = 0.5 group, the concordance between presence of memory impairment and pathologic amyloid‐β (A+; cut‐off = 1000 pg/ml) and p‐tau levels (T+; cut‐off = 27 pg/ml) in cerebrospinal fluid was assessed.ResultsEducation and sex (being a woman) had significant positive effects (p<0.001) on the age‐adjusted delayed memory scores (Partial eta2 = 0.035 and 0.010). We observed a change of up to 10 points after sex and education adjustment (mean[SD] change = 0.83 [2.6]). The use of fully adjusted scores increased the number of individuals classified as having memory impairment by 1.7%. Within the CDR = 0.5 group, these scores identified 2.5% individuals with memory impairment that were considered within the normal range using the regular norms, and increased the concordance between memory impairment and A+T+ biomarkers by almost 5% (regular norms = 65.2%, adjusted norms = 69.7%).ConclusionAlthough small, education and sex exert a relevant effect in the delayed memory score of the RBANS and improve clinical and biomarker concordance of the conclusions on the cognitive status derived from them. Adding education and sex adjustments to the RBANS measures may improve the selection of individuals for clinical trials.