Files
Abstract
Verbal memory deficits have long been tied to both healthy and pathological aging. This dissertation experiment examined the interactions of contextual language components (syntactic and semantic), overall cognitive status, retention duration, and lateralization of cognitive resources on verbal memory encoding and recognition, within a community sample of older adults. Individuals completed neuropsychological measures of cognition and subsequently completed an fMRI task involving categorizing short phrases by language context. Immediate recall, delayed recall, and delayed recognition for target nouns of these phrases were used as measures of verbal memory retention. FMRI BOLD activation was measured in predicted regions of interest during verbal memory encoding and recognition then correlated with behavioral variables. Syntactic specific encoding included bilateral pars opercularis (PO) and middle temporal gyrus (MTG). Semantic specific encoding including left hemisphere PO, pars triangularis (PT), middle frontal gyrus (MFG), middle orbitofrontal gyrus (MOFG), and MTG. Semantic specific encoding also included right hemisphere MFG, MOFG, and MTG. Syntactic specific recognition activation was not significant. Semantic specific recognition activation was observed within left MFG and bilateral MTG. Semantic specific encoding activation wassignificantly related to cognitive status in left MOFG, left PO and left PT, as well as right MTG and PO at liberal thresholds. Semantic encoding demonstrated a recall benefit predominantly after a delayed duration. Number of total activated voxels did not vary across hemispheres, and scope of activation across left and right regions of interest are discussed. Results suggest benefit of bilateral resources on semantic processing with age, which in turn benefits verbal memory within older adults. They further suggest that semantic manipulation of phrases had a greater relationship with overall cognition in comparison to syntactic manipulation of phrases. The current findings are discussed framed by the processing model of verbal memory (Craik & Lockhart, 1972) and the HAROLD model of bilateral dedifferentiation (Cabeza, 2002).