88 Single Trial of Biber Figure Learning Test Captures Subjective Cognitive Decline

Abstract

Objective:

The Biber Figure Learning Test (BFLT) is a serial figure learning assessment previously been shown to be sensitive to various biomarkers of the aging brain. BFLT is an extensive assessment requiring about 30 minutes for administration. In this study, we investigated BFLT’s associations with subjective cognitive decline (SCD), an early marker for preclinical Alzheimer’s Disease (AD), and examined whether alternative BFLT indices could be utilized to considerably shorten the length of assessment without decreasing its sensitivity to SCD.

Participants and Methods:

Participants were 50 cognitively normal older adults (8% Hispanic, 92% Non-Hispanic; 78% White, 16% Black; 64% female; mean age =72.7 (SD =6.2); mean education =17.05 (SD =2.09)). SCD was measured using a 20-item age-anchored dichotomous questionnaire that assessed complaints of cognitive functioning, and the BFLT was administered in full. Pearson correlations were conducted between SCD and BFLT scores including: Trial 1 Learning (T1), Trials 1 to 2 Total Learning (T1T2), Trials 1 to 3 Total Learning (T1T3), Trials 1 to 5 Total Learning (Total Learning), Immediate Recall, Delayed Recall, Proactive Interference (Trial B – Trial 1), Retroactive Interference (Immediate Recall – Trial 5), and Total Discrimination (calculated as [Recognition Total Correct + 0.5]/16) − ([Total False Alarms + 0.5]/31]). A Fishers Exact Test was utilized to compare the correlational strength between SCD and each of the BFLT scores. Lastly, demographically adjusted (age, gender, and education) regression models were conducted to examine SCD as an individual predictor for the various BFLT scores.

Results:

SCD was negatively associated with BFLT T1 (r =-0.406, p =0.003), T1T2 (r =-0.331, p =0.019), T1T3 (r =-0.323, p =0.022), Total Learning (r =-0.283, p =0.046), Immediate Recall (r =-0.322, p =0.023), Delayed Recall (r =-0.318, p =0.025), and Retroactive Interference (r =-0.388, p =0.005) and positively associated with Proactive Interference (r =0.308, p =0.029). There was no significant difference in correlational strength between any of these BFLT scores and SCD. Adjusting for demographics, SCD predicted Immediate Recall (B =-0.273, p =0.029), Total Learning (B =- 0.253, p =0.040), T1 (B =-0.412, p =0.002), T1T2 (B =-0.326, p =0.010), T1T3 (B =-0.299, p =0.017), Proactive Interference (B =0.292, p =0.050), and Retroactive Interference (B =- 0.330, p =0.025).

Conclusions:

Eight of the nine assessed BFLT scores were strongly correlated with age-anchored SCD and age-anchored SCD predicted seven of the nine assessed BFLT indices when adjusted for demographics. Although additional work is needed, these findings suggest SCD’s sensitivity to changes in visuospatial learning and memory, supporting its use as an early marker for preclinical AD. Likewise, our results suggest that an abbreviated version of the BFLT could be utilized that shortens testing time and reduces participant fatigue without a decrease in clinical relevance. Through ongoing longitudinal work, we hope to further disentangle the relationship between SCD and visuospatial learning and memory as measured through the BFTL and to examine how associations between SCD and the BFLT assessment change over time.