Dissociable electrophysiological measures of natural language processing reveal differences in speech comprehension strategy in healthy ageing

Abstract

Healthy ageing leads to changes in the brain that impact upon sensory and cognitive processing. It is not fully clear how these changes affect the processing of everyday spoken language. Prediction is thought to play an important role in language comprehension, where information about upcoming words is pre-activated across multiple representational levels.However, evidence from electrophysiology suggests differences in how older and younger adults use context-based predictions, particularly at the level of semantic representation. We investigate these differences during natural speech comprehension by presenting older and younger subjects with continuous, narrative speech while recording their electroencephalogram. We use linear regression to test how distinct computational measures of 1) semantic dissimilarity and 2) lexical surprisal are processed in the brains of both groups.Our results reveal dissociable neural correlates of these two measures that suggest differences in how younger and older adults successfully comprehend speech. Specifically, our results suggest that, while younger and older subjects both employ context-based lexical predictions, older subjects are significantly less likely to pre-activate the semantic features relating to upcoming words. Furthermore, across our group of older adults, we show that the weaker the neural signature of this semantic pre-activation mechanism, the lower a subject's semantic verbal fluency score. We interpret these findings as prediction playing a generally reduced role at a semantic level in the brains of older listeners during speech comprehension and that these changes may be part of an overall strategy to successfully comprehend speech with reduced cognitive resources.

Amplified cortical tracking of word-level features of continuous competing speech in older adults

“…In Broderick et al (2020), the greatest age-related differences were shown for semantic dissimilarity, whereas our goodness-of-fit results showed relatively weak contributions from this feature (compared to surprisal and word audibility) that did not differ significantly between the younger and older age groups. However, we did observe greater frontal TRF deflections in the older group for dissimilarity, with significant group differences around 250 and 600 ms, suggesting an increased gain for this feature in the older population.…”

“…Studies comparing envelope-related cortical responses have revealed a pattern of amplified envelope representations in older populations (Presacco et al, 2016;Decruy et al, 2019;Zan et al, 2020), potentially reflecting changes in the utilization of cognitive resources during speech comprehension. More recently, Broderick et al (2020) compared higher-level representations of speech in younger and older populations. They estimated EEG responses to 5-gram surprisal, reflecting the predictability of words given the preceding sequence of four words, as well as semantic dissimilarity, reflecting the contribution of each word to the semantic content of a sentence.…”

“…These findings demonstrate that representations of higher-level features of speech may indeed reveal robust effects of age. However, because Broderick et al (2020) did not report behavioral measures related to speech comprehension, nor measures of subjective speech perception difficulties among their participants, it is unclear whether these metrics would correlate with the reported EEG-based findings. Moreover, participants in that study were presented with clear speech without any distractors (e.g., competing speakers), making it unclear how speech representations differ in complex listening scenarios where speech perception difficulties are most commonly reported.…”

“…After fitting the full three-feature model as described above, we computed the unique contribution of each feature to the overall model fit using procedures described in Broderick et al (2020). Briefly, on each cross-validation fold, we estimated each feature's contribution to the overall fit by comparing the goodness-of-fit for the full model to a null model, in which that feature's contribution was eliminated.…”

“…Recently, Broderick et al (2020) compared representations of semantic dissimilarity and 5-gram lexical surprisal derived from responses to clean speech in younger and older adults. They showed that although younger adults exhibited robust responses to each feature, older adults only showed strong responses to lexical surprisal (albeit with a delayed peak response), with a nearly absent response to semantic dissimilarity.…”