Influence of amyloid and diagnostic syndrome on non‐traditional memory scores in early‐onset Alzheimer’s disease
Justin Bushnell, Dustin B. Hammers, Liana G. Apostolova, Joel H. Kramer, Maria C. Carrillo, Brad C. Dickerson, Gil D. Rabinovici, David G. Clark- Psychiatry and Mental health
- Cellular and Molecular Neuroscience
- Geriatrics and Gerontology
- Neurology (clinical)
- Developmental Neuroscience
- Health Policy
- Epidemiology
Abstract
Background
The Rey Auditory Verbal Learning Test (RAVLT) is a neuropsychological test used for assessing episodic memory impairment. The sequence of the RAVLT consists of five learning trials, an intrusion list recall, a subsequent recall (short delay), and a delayed recall (approximately thirty minutes after the short delay recall). The RAVLT has been studied extensively in various presentations of dementia, but there is limited research for early‐onset Alzheimer’s disease (EOAD). We analyze the influence of amyloid and diagnostic syndrome on traditional and novel RAVLT scores in EOAD.
Method
We transcribed RAVLT recordings from 303 subjects in the Longitudinal Early‐Onset Alzheimer’s Disease Study (LEADS). Subjects were group by amyloid status (control/EOnonAD vs EOAD) and syndrome: posterior cortical atrophy (PCA), primary progressive aphasia (PPA), amnestic, and non‐amnestic presentations. The traditional method of scoring the RAVLT is to count the total number of correct words recalled for each task (i.e., raw score). Two other count‐based scores were calculated by examining serial position effects (SPEs). SPEs describe the tendencies to remember words in the beginning (primacy) or end (recency) of the list. A relational SPE score, known as J‐curve, is examined here and is defined as primacy minus recency. Three timing‐based scores were analyzed: duration, stopping time, and speed. The seven scores were entered separately into linear mixed effects models as dependent variables. Amyloid status, syndrome, and nuisance variables (age, sex, education) were entered as independent variables.
Result
Compared with amyloid‐negative subjects, amyloid‐positive subjects showed negative effects on raw score, primacy and recency. The presence of amyloid increased stopping time by 5%. Duration and speed were not found to be associated with amyloid. Significant differences among clinical syndromes were observed with recency, duration, and stopping time.
Conclusion
RAVLT measures are sensitive to the effects of amyloid and syndrome in early‐onset dementia. Stopping time is a novel score that may complement existing measures for describing amyloid‐positive individuals. Further work is needed to quantify the predictive value of these scores.