Objectives: Speech comprehension under “cocktail party” scenarios deteriorates with age even in the absence of measurable hearing loss. Musical training is suggested to counteract the age-related decline in speech-in-noise (SIN) perception, yet which aspect of musical plasticity contributes to this compensation remains unclear. This study aimed to investigate the effects of musical experience and aging on SIN perception ability. We hypothesized a key mediation role of auditory working memory in ameliorating deficient SIN perception in older adults by musical training. Design: Forty-eight older musicians, 29 older nonmusicians, 48 young musicians, and 24 young nonmusicians all with (near) normal peripheral hearing were recruited. The SIN task was recognizing nonsense speech sentences either perceptually colocated or separated with a noise masker (energetic masking) or a two-talker speech masker (informational masking). Auditory working memory was measured by auditory digit span. Path analysis was used to examine the direct and indirect effects of musical expertise and age on SIN perception performance. Results: Older musicians outperformed older nonmusicians in auditory working memory and all SIN conditions (noise separation, noise colocation, speech separation, speech colocation), but such musician advantages were absent in young adults. Path analysis showed that age and musical training had opposite effects on auditory working memory, which played a significant mediation role in SIN perception. In addition, the type of musical training did not differentiate SIN perception regardless of age. Conclusions: These results provide evidence that musical training offsets age-related speech perception deficit at adverse listening conditions by preserving auditory working memory. Our findings highlight auditory working memory in supporting speech perception amid competing noise in older adults, and underline musical training as a means of “cognitive reserve” against declines in speech comprehension and cognition in aging populations.

No abstract

BackgroundVisuospatial processing requires wide distribution or narrow focusing of attention to certain regions in space. This mechanism is described by the zoom lens model and predicts an inverse correlation between the efficiency of processing and the size of the attentional scope. Little is known, however, about the exact timing of the effects of attentional scaling on visual searching and whether or not additional processing phases are involved in this process.MethodElectroencephalographic recordings were made while participants performed a visual search task under different attentional scaling conditions. Two concentric circles of different sizes, presented to the participants at the center of a screen modulated the attentional scopes, and search arrays were distributed in the space areas indicated by these concentric circles. To ensure consistent eccentricity of the search arrays across different conditions, we limited our studies to the neural responses evoked by the search arrays distributed in the overlapping region of different attentional scopes.ResultsConsistent with the prediction of the zoom lens model, our behavioral data showed that reaction times for target discrimination of search arrays decreased and the associated error rates also significantly decreased, with narrowing the attentional scope. Results of the event-related potential analysis showed that the target-elicited amplitude of lateral occipital N1, rather than posterior P1, which reflects the earliest visuospatial attentional processing, was sensitive to changes in the scaling of visuospatial attention, indicating that the modulation of the effect of changes in the spatial scale of attention on visual processing occurred after the delay period of P1. The N1 generator exhibited higher activity as the attentional scope narrowed, reflecting more intensive processing resources within the attentional focus. In contrast to N1, the amplitude of N2pc increased with the expansion of the attentional focus, suggesting that observers might further redistribute attentional resources according to the increased task difficulty.ConclusionThese findings provide electrophysiological evidence that the neural activity of the N1 generator is the earliest marker of the zoom lens effect of visual spatial attention. Furthermore, evidence from N2pc shows that there is also a redistribution of attentional resources after the action of the zoom lens mechanism, which allows for better perform of the search task in the context of low attentional resolution. On the basis of the timing of P1, N1, and N2pc, our findings provide compelling evidence that visuospatial attention processing in the zoom lens paradigm involves multi-stage dynamic processing.

No abstract