You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

 
Poster C60, Thursday, November 9, 10:00 – 11:15 am, Harborview and Loch Raven Ballrooms

Graph Theoretical Approaches Show a Relationship Between Resting State Functional Connectivity in Younger and Older Adults and Phonological Aspects of Language Production

Victoria Gertel1, Kerem Oktar2, Michele Diaz1;1Pennsylvania State University, 2Pomona College

Older adults often experience difficulties producing fluent speech compared to younger adults, including increased word retrieval failures and slower speech rates (Burke et al., 2008). Resting state fMRI (rsfMRI), a task-free technique that is used to investigate intrinsic network function, was used in this study. While used extensively to examine some networks, and to compare older and younger adults, few studies have used it to investigate the language network. Prior rsfMRI studies have demonstrated that neural networks strongly correlated during task are strongly correlated during rest as well (Greicius et al., 2003; Rosazza & Minati, 2011). Additionally, resting state functional connectivity (RSFC) is predictive of behavior during task and symptom severity in clinical populations (Mennes et al., 2010; Wang et al., 2007). Here we used rsfMRI to investigate how functional connectivity in the language network in older and younger adults predicts behavior. In examining RSFC within the language network, we hypothesize that older adults will exhibit an overall decline in the RSFC in language network regions compared to younger adults and that age and RSFC will significantly predict accuracy and reaction time on categorically related and phonologically related picture word interference task conditions. Twenty younger adults (mean age = 23.70, 10 females) and 19 older adults (mean age = 67.32 14 females) were analyzed. Regions of interests (ROIs) in the left-hemisphere were selected based on prior research and anatomical atlases (Desikan et al., 2006; Hickok & Poeppel, 2007). During a picture word interference task, participants named a target picture that was overlaid with a distractor word. The two task conditions analyzed included those with the target picture and distractor word being categorically or phonologically related. Accuracy and reaction time data during these two conditions were used as the behavioral metric. Independent Samples t tests showed group differences between older and younger adults in overall network connectivity, with older adults having less network connectivity. A graph theoretical approach was used to determine degree and centrality of the language network across older and younger adults. Weighted, binary graphs with a threshold of r = 0.2 were created. Nodes were first described as connected or not connected. Weights were applied to the connected nodes only (Dosenbach et al., 2007). We used MANOVAs to compare two aspects of network dynamics: centrality and degree. There were no significant age differences in centrality, however, there was a marginally significant effect of age for degree (p = .07). Several regions contributed to this effect including IFG, aMTG, pMTG, aInsula, STS, and STG. Relating these effects to behavior, we found a significant effect of degree in anterior insula, which predicted RTs during phonological trials (p < .05, r2 = .129, younger degree = 3.6, older degree = 2.2). Control analyses in the visual cortex revealed no significant age differences. Our results suggest that while the overall structure of the left hemisphere language network is resistant to age-related decline, connectivities between nodes may decrease with age. This decline may be related to phonological aspects of language production.

Topic Area: Speech Motor Control and Sensorimotor Integration

Back to Poster Schedule