Now You’re With Me, Now You’re Not: Between-Session and Within-Session Intra-Individual Variability in Attention in Aphasia

• ¹ Boston University, United States

Introduction: Patients with aphasia (PWA) have been found to exhibit impaired performance on attention tasks relative to healthy controls, even when tasks contain only non-linguistic stimuli (e.g. Robin & Rizzo, 1989). Additionally, our previous work has shown that PWA exhibit increased day-to-day, or between-session, intra-individual variability (BS-IIV) on non-linguistic attention tasks (Villard & Kiran, 2015). The current study examines BS-IIV, as well as moment-to-moment, or within-session, intra-individual variability (WS-IIV), in performance on a series of non-linguistic and linguistic attention tasks. We theorize that BS-IIV and WS-IIV may be critical metrics for understanding attention processing in PWA. Methods: Data from 17 PWA and 17 age-matched healthy controls have been analyzed; data is expected from a total of 20 in each group. Each participant completed a series of five computerized experimental tasks of incrementally increasing demands/complexity. Task 1 measured non-linguistic sustained visual attention; Task 2 added visual distractor stimuli; Task 3 added non-linguistic target auditory stimuli, requiring integration of two modalities; Task 4 added lexical decision demands; and Task 5 added lexical-semantic processing demands. This series of five tasks was administered four times, on four different non-consecutive days. We hypothesized that increased task complexity would elicit increased BS-IIV and WS-IIV in attention for PWA but not for controls. Data Analysis: In order to examine BS-IIV and WS-IIV in reaction time (RT), coefficients of variation (COVs, or the standard deviation over the mean) were calculated for each participant, on each task. Specifically, a COV representing BS-IIV in RT (BS-COV) was calculated for each participant, each task, using data collected across the four sessions. Similarly, a COV representing WS-IIV in RT (WS-COV) was calculated for each participant, each task, using data collected within a single session. Higher COVs indicate higher degrees of variability in RT. Results: A 2 (Group) x 5 (Task) R-ANOVA showed that BS-COVs were higher for PWA than for controls (F (1, 32) = 4.79, p < .05). WS-COVs were also found to be higher for PWA than for controls (F (1, 32) = 7.61, p < .01). Additionally, WS-COVs were consistently higher than BS-COVs for both groups. An R-ANOVA investigating the effect of Task on BS-COV within the PWA group revealed a result approaching significance (F (4, 64) = 2.48, p = .053), such that more complex tasks elicited higher levels of variability than less complex tasks. A significant effect of Task on WS-COV within the PWA group was also observed (F (4, 64) = 9.47, p < .0001), such that the most complex task (Task 5, the only task requiring true language processing), elicited higher WS-COVs than any other task (p < .05, see Figure); such an effect of complexity was not observed for controls. Discussion: PWA showed increased IIV in attention relative to controls, both from moment to moment and from day to day. Furthermore, for PWA, increased task demands – in particular, language processing – elicited increased IIV. Findings may have implications for both assessment and treatment in aphasia, as testing and therapy require consistent attention to linguistic stimuli.