Gaudrain, E. | bioRxiv Science

bioRxiv2023

Phase coding in phoneme processing slows with age

The comprehension of phonemes is a fundamental component of speech processing which relies on both, temporal fine structure (TFS) and temporal envelope (TE) coding. EEG amplitude in response to phonemes has been identified as indicator of speech performance in hearing aid users. Presbyacusis may also alter neuro-electric responses to phonemes, even with minimally or unaffected hearing thresholds. Elevated speech reception thresholds (SRT) in absence of pure-tone threshold (PTT) elevation suggest central processing deficits. We therefore collected audiometric data (PTT, SRT) and EEG during passive listening in 80 subjects, ranging in age from 18 to 76 years. We confirm phoneme-evoked EEG response amplitude (PEA) as indicator of speech comprehension. Specifically, PEA decreased with elevated SRT, PTT and increased age. As novel observation, we report the temporal delay of phoneme-evoked EEG responses (PED) to increase with age and PTT. The absolute duration of PED, its age-correlation, and the lack of PEA-lateralization combined with the frequency of phoneme stimuli used here suggest a predominantly thalamic generator of phoneme-evoked EEG responses. Hearing loss in extended high-frequencies affects PED more than PEA. In our sample, neural compensation for increased PTT came at the cost of decreased temporal processing speed. Most importantly, PED correlates with SRT and explains SRT-variance in quiet and in ipsilateral noise that PTT cannot. PED was a better predictor of TFS coding in quiet and of TE coding in ipsilateral noise. As PED reflects both TFS and TE coding, thalamic activity may provide integrated information at the gate of neocortex. Significance StatementIntact speech comprehension is essential for social participation which protects against depression and dementia. Age-related hearing loss is a growing problem in aging societies, as hearing deficits constitute the third most important modifiable risk factor for cognitive decline. This work uses electrical brain responses to phonemes in a cohort covering age 18 to 76 years. As the temporal delay of phoneme responses showed the most significant correlations with age and high-frequency thresholds, we demonstrated that speed of neural processing seems essential for speech comprehension. The observed neural signals likely originate from thalamus which receives feedback from neocortex and is embedded in cognitive processing. Developing objective markers for speech processing is key for ensuring cognitive fitness in aging.

systems biology↗

bioRxiv2023

Neural adaptation at stimulus onset and speed of neural processing as critical contributors to speech comprehension independent of hearing threshold or age

Loss of afferent auditory fiber function (cochlear synaptopathy) has been suggested to occur before a clinically measurable deterioration of subjective hearing threshold. This so-called "hidden" hearing loss is characterized by speech comprehension difficulties. We examined young, middle-aged, and older individuals with and without hearing loss using pure-tone (PT) audiometry, short-pulsed distortion-product otoacoustic emissions (DPOAE), auditory brainstem responses (ABR), auditory steady state responses (ASSR), speech comprehension (OLSA), and syllable discrimination in quiet and noise. After normalizing OLSA thresholds for PT thresholds ("PNOT"), differences in speech comprehension still remained and showed no significant dependence on age, allowing us to categorize participants into groups with good, standard, and poor speech comprehension. Listeners with poor speech comprehension in quiet exhibited smaller firing rate adaptions at stimulus onset (as measured by the difference between DPOAE threshold and pure-tone threshold) and delayed supra-threshold ABR waves I-V, suggesting high spontaneous rate low threshold fiber cochlear synaptopathy. In contrast, when speech comprehension was tested in noise, listeners with poor speech comprehension had larger DPOAEs acceptance rate, putatively resulting from altered basilar membrane compression (recruitment). This was linked with higher uncomfortable loudness levels and larger ASSR amplitudes. Moreover, performance in phoneme discrimination was significantly different below (/o/-/u/) and above the phase-locking limit (/i/-/y/), depending on whether vowels were presented in quiet or ipsilateral noise. This suggests that neural firing rate adaptation at stimulus onset is critical for speech comprehension, independent of hearing threshold and age, whereas the recruitment phenomenon counterbalances the loss in speech-in-noise discrimination due to impaired threshold. Significance StatementAge-related hearing loss is the third largest modifiable risk factor for cognitive decline. It has been suggested that the link between hearing loss and cognitive decline is not fully explained by hearing threshold loss. We here suggest that language comprehension deficits may be used as an early indication of future hearing loss and therefore cognitive decline. We found that, independent of age and pure-tone thresholds, speech comprehension in quiet and ipsilateral noise depend on different onset firing-rate adaptations of inner hair cells (measured by DPOAE threshold), along with cochlear synaptopathy of high spontaneous rate auditory nerve fibers and neural spiking synchronicity. These measures may be used as possible future indicators of risk for cognitive decline.

neuroscience↗