Bridging the gap: A conversion framework for CDR-SB and MoCA scores in Alzheimer's disease and related dementia.

Background: Accurate assessment of cognitive impairment is essential to effective Alzheimer's disease (AD) management and research. However, the absence of validated methods to translate scores between widely used instruments-such as the Clinical Dementia Rating Scale Sum of Boxes (CDR-SB) in trials and the Montreal Cognitive Assessment (MoCA) in clinical practice-poses a significant barrier. This limits data harmonization, impedes cross-study comparability, and complicates the integration of clinical and research evidence. Bridging this gap is critical for consistent staging, longitudinal monitoring, and data-driven decision-making in AD and related dementias.

Objectives: To develop and validate bidirectional score conversion tables between CDR-SB and MoCA using a large, diverse cohort spanning the full spectrum of cognitive function.

Design: Retrospective, cross-sectional analysis using equipercentile equating with log-linear smoothing. Optimal smoothing parameters were selected by minimizing mean squared error, Akaike Information Criterion, and Bayesian Information Criterion. Concordance was assessed using Spearman's rank correlation and Bland-Altman plots.

Setting: National Alzheimer's Coordinating Center (NACC), aggregating standardized assessments from 35 U.S.-based Alzheimer's Disease Research Centers.

Participants: 23,717 individuals (59,871 visits) with same-day CDR-SB and MoCA assessments from January 2015 to September 2024, spanning normal cognition, mild cognitive impairment (MCI), and dementia.

Intervention: None; this was a secondary analysis of existing data.

Measurements: Primary measures included CDR-SB (0-18; higher = greater impairment) and MoCA (0-30; higher = better cognition). Bidirectional crosswalk tables were derived using equipercentile equating.

Results: CDR-SB and MoCA scores showed strong inverse correlation (Spearman's ρ = -0.68; p < 0.001). Crosswalk tables demonstrated good agreement across the cognitive spectrum and performed consistently in the full cohort and an AD-specific subgroup.

Conclusions: This study provides the first validated, bidirectional CDR-SB-MoCA crosswalk, supporting data harmonization and consistent interpretation of cognitive severity across research and clinical settings.