Presentation

Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Poster Slams

Components of limb apraxia distinguish patients with different primary progressive aphasia variants

Poster E24 in Poster Session E, Saturday, October 8, 3:15 - 5:00 pm EDT, Millennium Hall

Haley C. Dresang1,2, Rand Williamson1, Hana Kim3, Argye E. Hillis3, Laurel J. Buxbaum1,4; 1Moss Rehabilitation Research Institute, 2University of Pennsylvania, 3Johns Hopkins University, 4Thomas Jefferson University

INTRODUCTION: Primary progressive aphasia (PPA) is a clinical syndrome primarily characterized by a progressive language deficit and relative sparing of other cognitive domains[1,2]. However, limb apraxia-- a disorder of skilled movements not attributable to sensory, motor, or language deficits-- frequently co-occurs[3,4]. Our prior research indicates that limb apraxia after left hemisphere stroke may be characterized by deficits in two components of gesture: (1) hand posture (HP, the shape and movement trajectory of the hand and wrist) and (2) kinematics (KIN, the amplitude and timing of movement[5]). Additionally, following left-hemisphere stroke, HP errors are associated with semantic processing of actions and manipulable objects (“action semantics”) and posterior temporal lesions, while KIN errors are associated with meaningless gestures and lesion to pre/postcentral gyri and the inferior parietal lobe[6,7]. However, these gesture components have never been examined in PPA. We examined how each component contributes to limb apraxia in patients across different variants of PPA. METHODS: Participants were 21 adults with PPA (4 semantic [svPPA], 6 nonfluent [nfvPPA], 11 logopenic variants [lvPPA]). Participants were excluded if they had hearing loss, uncorrected visual loss, or history of other neurological disease. Participants completed verb naming[8] and two video-based gesture imitation tasks: (1) meaningless gestures and (2) meaningful transitive gestures (e.g., pantomimed hammering). HP and KIN gesture components each received a 0 or 1 per trial, following established procedures by trained, reliable coders[9]. Generalized logistic mixed-effects models[10] examined a three-way interaction between PPA variant (svPPA, nfvPPA, lvPPA), task condition (meaningful, meaningless gestures), and gesture component (HP, kinematics) on gesture imitation accuracy, controlling for dementia severity (Dementia Rating Scale[11]), with random effects of subject and item. RESULTS: There was a significant three-way interaction between PPA variant, task condition, and component on gesture accuracy (χ²=24.95, p=0.001). Pairwise comparisons with Tukey corrections[12] revealed that HP was more impaired than kinematics for all PPA variants in the meaningless condition (p’s<0.05), but only svPPA patients showed this pattern for meaningful gestures (β=1.98, p=0.007). Although KIN was less impaired than HP at the group level, five participants (4 lvPPA, 1 nfvPPA) showed the opposite pattern, such that KIN was more impaired than HP on meaningful gestures. None of the svPPA showed superiority of HP over KIN. An exploratory follow-up analysis showed that the integrity of verb naming was correlated with a greater HP over KIN advantage (R2=0.60, p<0.001). CONCLUSION: The finding that HP was more impaired than the KIN component for meaningless gesture imitation across all PPA variants is consistent with evidence that HP is the most sensitive measure of limb apraxia in patients with left-hemisphere stroke[5]. However, svPPA was the only variant that also showed relative deficits in HP for meaningful gestures, and no svPPA participants performed better on HP than KIN components. These findings, along with the observed correlation of verb naming and HP advantage, suggest that degraded action semantic representations or retrieval processes in svPPA are tightly linked to HP accuracy. Similar to evidence from stroke[6,7], these findings highlight the important interplay between semantics and gesture deficits in PPA.

Topic Areas: Disorders: Acquired, Signed Language and Gesture