Montefiore Einstein Robust Geriatric Normative Project (MERGER-NP): clinical decision rules for incorporating word reading into regression-based demographic norms and estimating premorbid cognitive functioning

Objective: This study expands on the Montefiore Einstein Robust Geriatric (MERGER) norms by providing normative data for two word reading tests (WTAR, AMNART). We also developed regression-based prediction formulas for these tests using demographics and a novel measure of cognitive literacy engagement, establishing base rates for discrepancies between actual and predicted scores. Discrepancy base rates for global cognition, derived from regression-based norms using word reading and demographic factors, were also examined to support detection of cognitive decline. Method: The MERGER sample included 420 community-dwelling older adults. Backward regression analyses predicted WTAR and AMNART scores. Base rates were calculated for discrepancies on the WTAR, AMNART, and the RBANS Global Cognition Index (GCI). One-way ANOVAs compared discrepancy scores across MERGER, mild cognitive impairment (MCI), and dementia groups. Clinically meaningful cutoffs were set at the 10% base rate, and ROC curves assessed diagnostic accuracy. Results: Normative data for WTAR and AMNART are presented. Cognitive literacy engagement significantly predicted both WTAR and AMNART scores, explaining small but significant variance. Base rate tables for discrepancies in word reading and GCI are provided. Mean discrepancies in the MERGER sample between actual and predicted word reading scores were near zero. ANOVAs showed MERGER participants had significantly smaller negative discrepancies than the MCI and dementia groups. For MCI detection, WTAR and AMNART discrepancies showed low sensitivity (17–24%) at 90% specificity. In the dementia group, sensitivities were fair (24–29%), improving in a subgroup diagnosed at their initial study visit (30–40%) with acceptable diagnostic accuracy. Conclusions: This study provides normative data and discrepancy base rates for word reading and global cognition, enhancing the clinical applicability of MERGER norms. These results support more precise interpretation of word reading abilities in older adults, aiding in the differentiation of typical versus atypical cognitive profiles and improving diagnostic confidence when evaluating potential cognitive decline.