Abnormal language-related oscillatory responses in primary progressive aphasia

Abstract

Patients with Primary Progressive Aphasia (PPA) may react to linguistic stimuli differently than healthy controls, reflecting degeneration of language networks and engagement of compensatory mechanisms. We used magnetoencephalography (MEG) to evaluate oscillatory neural responses in sentence comprehension, in patients with PPA and age-matched controls. Participants viewed sentences containing semantically and syntactically anomalous words that evoke distinct oscillatory responses. For age-matched controls, semantic anomalies elicited left-lateralized 8–30 Hz power decreases distributed along ventral brain regions, whereas syntactic anomalies elicited bilateral power decreases in both ventral and dorsal regions. In comparison to controls, patients with PPA showed altered patterns of induced oscillations, characterized by delayed latencies and attenuated amplitude, which were correlated with linguistic impairment measured offline. The recruitment of right hemisphere temporo-parietal areas (also found in controls) was correlated with preserved semantic processing abilities, indicating that preserved neural activity in these regions was able to support successful semantic processing. In contrast, syntactic processing was more consistently impaired in PPA, regardless of neural activity patterns, suggesting that this domain of language is particularly vulnerable to the neuronal loss. In addition, we found that delayed peak latencies of oscillatory responses were associated with lower accuracy for detecting semantic anomalies, suggesting that language deficits observed in PPA may be linked to delayed or slowed information processing.

The power of neural oscillations to inform sentence comprehension: A linguistic perspective

“…One result which does not easily fit under the prediction account is the gamma power increase in sentences with incongruent endings compared to congruent ones, reported in Wang et al (2018a). Mixed gamma findings like this one, or the absence of gamma effects in Kielar et al (2015Kielar et al ( , 2018 and Wang, Jensen, et al (2012) for example, might be at least partially accounted for by the fact that finding and/or interpreting effects of cognitive manipulations on gamma power can be challenging. This is because there are multiple potential pitfalls that have to be taken into account in the analysis procedures.…”

“…In a series of studies which examined the link between neural oscillations and semantic processing, Kielar et al (2014Kielar et al ( , 2015Kielar et al ( , 2018 have consistently found an alpha/beta power decrease in response to semantic anomalies in a sentence, and in two out of three studies (Kielar et al, 2014(Kielar et al, , 2015, they also reported a delta/theta power increase in response to semantic anomalies. One reason Kielar et al (2018) may not have observed an increase in theta could be their relatively small sample size: 15 compared to 28 and 20 participants in their other studies. Five further studies (Bastiaansen & Hagoort, 2015;Davidson & Indefrey, 2007;Hagoort et al, 2004;Hald, Bastiaansen, & Hagoort, 2006;Wang, Zhu, & Bastiaansen, 2012) also reported a theta increase in response to semantic anomalies.…”

“…Further evidence for the importance of beta oscillations for syntactic processing comes from a series of studies run by Kielar and colleagues. They have consistently found alpha/beta decreases in response to verb tense violations (Kielar et al, 2014;Kielar et al, 2015;Kielar et al, 2018). In all three studies, Kielar et al have used a sentence acceptability judgment task, in which they asked participants to indicate that sentences were unacceptable in case they contained a grammatical violation or semantic anomaly.…”

Taking the sub-lexical route: brain dynamics of reading in the semantic variant of Primary Progressive Aphasia

“…This is especially relevant when the effects might occur later with respect to stimuli onset, as in the current setting, where differences between controls and patients in early, low-level visual processes are not expected, while divergence in later stages is predicted by both behavioral evidence and neuroanatomical findings. Furthermore, increased latency in response to linguistic stimuli have been observed in svPPA patients, and analyses of oscillatory behavior might be more sensitive than ERPs in detecting abnormalities in slow neural responses with increased trial-by-trial variability and thus decreased phase-locking (Kielar et al, 2018). Second, we aimed at providing functional interpretation of our findings.…”

“…MEG has only been deployed recently to investigate PPA (e.g., Kielar et al, 2018), but has already shown its potential as a tool to study the neurophysiological signatures of network-level alterations due to neurodegenerative disorders. It can be instrumental in identifying syndromespecific changes in the spectral properties of oscillatory responses (Ranasinghe et al, 2017;Sami et al, 2018).…”