When Does Forgetting Become Accelerated? Operational Normative Benchmark For The Rey Auditory Verbal Learning Test

Background: Accelerated Long-Term Forgetting (ALF) involves rapid loss of information over extended delays despite preserved learning and initial retention. ALF is increasingly recognized as a neuropsychological marker in preclinical Alzheimer’s Disease, reflecting disruptions in consolidation processes that are not detected by standard neuropsychological assessments. However, its assessment in clinical settings remains limited due to the absence of standardized tools and normative data. This study aims to establish normative benchmarks for one-week forgetting on the Rey Auditory Verbal Learning Test (RAVLT) by assessing retention loss between the standard-delay and one-week recall in healthy adults. Methods: A systematic review and meta-analysis (pre-registered on OSF, 10.17605/OSF.IO/W89QZ) were conducted to identify studies assessing ALF in healthy adults using the RAVLT. Standardized mean changes were calculated using the raw-score standardization approach (SMCR) to quantify pre-post performance changes in healthy adults. Sampling variances were computed assuming a pre-post correlation of r = 0.5, with sensitivity analyses conducted for r = 0.3 and r = 0.7. Effects were pooled using random-effects meta-analysis (restricted maximum likelihood). Between-study heterogeneity was quantified using τ² and I², and 95% prediction intervals were estimated. Small-study effects were explored using Egger’s regression test. Results: Across 16 studies (n = 1,565), healthy adults showed a consistent decline in performance over one week (SMCR = −1.27, 95% CI [−1.45, −1.09], p < .001). Results were highly consistent across assumed pre-post correlations (r = .3-.7). Between-study heterogeneity was substantial (I² = 79.51-87.64%). The 95% prediction interval ranged from approximately −1.94 to −0.60, indicating variability in the magnitude of forgetting across samples. Egger’s test did not indicate significant funnel plot asymmetry (p = .10). Conclusion: Healthy adults show a quantifiable decline over one week, indicating that forgetting can occur beyond standard delays even without pathology. Prediction intervals offer a clinically useful range of expected performance, helping to distinguish between normal and abnormal forgetting. The observed decline provides a quantitative normative benchmark for comparing patient-related forgetting and supports the use of extended-delay RAVLT recall in clinical practice. These findings endorse the application of ALF assessment and the development of data-driven thresholds within clinical populations.