P1‐091: Aß deposition in the nasal cavity of a transgenic mouse model of Alzheimer's disease

was not found in older adults for that task. Furthermore, this study found that increased variability in performance diminished the statistical likelihood of finding a difference in performance between training paradigms (Jackson et al., 2011). The current study sought to expand upon the previously reported findings regarding older adults, with a particular focus on how a risk factor for dementia, such as PiB binding, might be related to both visuospatial and verbal memory. Methods: Twelve cognitively healthy older adults (M 1⁄4 70.2, SD 1⁄4 7.8) were recruited from a longitudinal aging study in Melbourne, Victoria, Australia. As part of their participation in this study, individuals were assessed every three months for one year, plus one follow up at two years. They also underwent PET imaging using the PiB ligand. Participants were identified as either PiB positive (N1⁄4 5) or PiB negative (N1⁄4 7), based on a Standardized Uptake Value Ratio (SUVR) of 1.6, following the methods of Villemagne and colleagues (2008). Cognitive Assessment. During the cognitive assessment portion of data collection, participants completed three learning trials for a word-list learning task (International Shopping List task; ISLT) and five learning trials for a visuospatial learning and memory task (Continuous Paired Associate Learning task; CPAL). Participants also completed a 24-hour delayed recall trial for both tasks under both training paradigms. This study utilized a within-subject cross-over design for both tasks. Tasks were counterbalanced across four conditions and participants were pseudo-randomly assigned to condition. Massed versus spaced presentation. Massed and spaced training trials were administered to all participants for both the CPAL and ISLT. Massed learning trials of the CPAL and ISLT were administered in immediate succession. Spaced learning trials were administered with 15-minute delays between each new trial. During the spaced paradigm delay, participants were asked to sit in a resting state, without deliberate visual stimulation.Results: Data required a log10 transformation for normalization. An ANCOVA [between-subjects factors: condition (massed, spaced)], covarying for performance on final learning trial, identified a significant difference between delayed recall performance following massed and spaced training on the visuospatial task (F (1,21)1⁄4 6.02, p1⁄4 .02), but not the word-list learning task (F (1,21)1⁄4 .99, p 1⁄4 .33). Cohen’s d effect sizes were calculated for both tasks: d 1⁄4 1.99 (visuospatial task), and d1⁄4 .56 (word-list learning task). A separate ANCOVA comparing performance on delayed recall by PiB status [between-subjects factors: condition (massed, spaced), PiB (positive, negative)], covarying for performance on final learning trial, identified non-significant interactions between PiB status and group for the visuospatial (F (1,19) 1⁄4 1.22, p1⁄4 .28) and word-list learning tasks (F (1,19)1⁄4 .74, p1⁄4 .40). A power analysis, using the effect sizes calculated for the visuospatial andword-list learning tasks for the group as a whole, alpha set to .05, and power set to .80, identified a needed sample size of 15 for the visuospatial task and 83 for the word-list learning task. This statistical evidence supports the argument that, although the spacing effect did exist for each of the tasks in this sample, the study itself was underpowered to detect differences between PiB groups.Visual comparisons between PiB groups on the visuospatial memory task and word-list learning task revealed subtle differences in performance based on group; however, large standard deviations influenced the ability to detect statistical significance. A plot of the difference score between performance on the delayed recall trial following massed and spaced training was constructed and regression lines were fit to the data. The slopes for each regression line (visuospatial task and word-list learning task) were compared; as SUVR increased, participants performed worse on the visuospatial task (slope 1⁄4-13.22), yet performed better on the word-list learning task (slope 1⁄4 4.89).Conclusions: Significant differences in delayed recall performance following massed and spaced training were identified for this small sample for the visuospatial task, but not the word-list learning task. When comparing delayed recall performance by PiB status there were no statistical differences between PiB groups on the delayed recall trial following massed and spaced training. However, comparing the regression lines for the visuospatial and word-list learning tasks, plotted against SUVR, the slopes of the lines indicate distinct differences in how individuals with higher SUVR performed on each of these tasks. Specifically, as SUVR increased, participants tended to perform worse on the visuospatial task, yet better on the word-list learning task, following spaced training. This data suggests a performance benefit for verbal information that is presented in a spaced manner, while visuospatial information may be better presented under massed conditions in an effort to improve delayed recall. These results provide evidence for distinct differences in recall between individuals at varying levels of risk for future Alzheimer’s disease dependent upon the type of information presented. This study has clear limitations in terms of sample size and the known variability in performance inherent in older adults. Therefore, future studies with larger sample sizes are recommended. However, despite these limits, the ability to detect a difference in recall, dependent upon the type of information presented, as individuals vary in risk for Alzheimer’s disease, is an important step in maximizing the learning capabilities for individuals in early stages of cognitive decline.