Examining age-related differences in recollection and retrieval monitoring effects using single- and across-trial fMRI approaches

Poster Session C - Sunday, April 14, 2024, 5:00 – 7:00 pm EDT, Sheraton Hall ABC

Sarah Monier¹, Marianne de Chastelaine¹, Mingzhu Hou¹, Michael D. Rugg¹; ¹The University of Texas at Dallas

Most fMRI studies examining age differences in the neural correlates of episodic retrieval have employed across-trial general linear models (GLM) to derive estimates of BOLD activity elicited on ‘recollected’ versus ‘unrecollected’ test trials. Here, we re-analyzed a previously published dataset examining fMRI correlates of recollection and retrieval monitoring in 36 young and 64 older adults. Single-trial beta estimates of BOLD signal magnitude from 9 regions of interest (ROIs) were extracted. We estimated recollection effects by computing the differences in the mean beta estimates averaged over ‘recollected’ vs. ‘unrecollected’ trials. In addition, we transformed these recollection estimates into estimates of effect size to derive standardized retrieval-related metrics, thereby controlling for across participant differences in the gain of the fMRI hemodynamic response function (HRF). The raw and standardized recollection estimates were highly correlated (all rs > .900). These single-trial recollection estimates were then compared with those derived from the original analysis which employed an across-trial GLM approach. The correlations between the retrieval metrics estimated with the two approaches were mostly negligible (rs = -.03 - .32). Moreover, findings for age differences and the relationships between recollection effects and memory performance differed markedly between the two approaches in multiple brain regions. These findings call for further examination of the relative merits of the two approaches in investigating neural correlates of memory processing. Reasons for the divergent findings and their implications for the interpretation of previously reported age differences in fMRI correlates of episodic memory retrieval will be discussed.

Topic Area: METHODS: Neuroimaging