Increased listening effort and decreased speech discrimination at high presentation sound levels in acoustic hearing listeners and cochlear implant users

Abstract

The sounds we experience in our everyday communication can vary greatly in terms of level and background noise depending on the environment. Paradoxically, increasing the sound intensity may lead to worsened speech understanding, especially in noise. This is known as the "Rollover" phenomenon. There have been limited studies on rollover and how it is experienced differentially across aging groups, for those with and without hearing loss, as well as cochlear implant (CI) users. There is also mounting evidence that listening effort plays an important role in challenging listening conditions and can be directly quantified with objective measures such as pupil dilation. We found that listening effort was modulated by sound level and that rollover occurred primarily in the presence of background noise. The effect on listening effort was exacerbated by age and hearing loss in acoustic listeners, with greatest effect in older listeners with hearing loss, while there was no effect in CI users. The age- and hearing-dependent effects of rollover highlight the potential negative impact of amplification to high sound levels and therefore has implications for effective treatment of age-related hearing loss.

Figure 1.. Schematic of experimental paradigm.

Speech perception task initializes with a listening phase (left) of 4000 ms where minimal word pairs are presented via headphones after an initial 500 ms of silence. Listeners fixate on a grey screen with a red cross during this phase. The response phase (middle) begins when the red cross turns green and the two words presented are shown on the screen. Participants either press 1 or 2 to indicate which word was heard second in the listening phase. The reset phase (right) occurs after the participant provides a keyboard press lasting 6000 ms to allow pupils to return to baseline.

Figure 2.. Behavioral performance on experimental task segregated by type of contrast.

A) Mean performance of participants for YNH (blue), ONH (orange), OHI (purple) listeners, and OCI (green) users as a function of sound level when speech perception task with temporal contrast word pairs presented in quiet (left) and in noise (right). B) Same as A but for task performed with non-temporal contrast word pairs. Error bars indicate ± 1 standard error.

Figure 3.. Behavioral performance on experimental task.

A) Mean performance of participants for YNH (blue), ONH (orange), OHI (purple) listeners, and OCI (green) users as a function of sound level when speech perception task was performed in quiet. B) Same as A but for task performed in babble noise. Error bars indicate ± 1 standard error. For detailed statistics, refer to Supplementary Table 1 and 2.

Figure 4.. Pupillometry data for normal-hearing listeners.

A) Mean pupil dilation as a percentage of dynamic range across the timescale of each trial of experimental task performed in quiet (orange) and in the presence of noise (cyan) at 65 dB SPL for YNH listeners. B) Same as A but for ONH listeners. C and D) Same as A and B respectively but at 85 dB SPL. Shaded error bars indicate ± 1 standard error. Black bars indicate significantly different pupil size between quiet and noise conditions at the p < 0.05 level across the timescale shown.

Figure 5.. Pupillometry data for OHI listeners.

A) Mean pupil dilation as a percentage of dynamic range across the timescale of each trial of experimental task performed in quiet (orange) and in the presence of noise (cyan) at 65 dB SPL for OHI listeners. B) Same as A but at 85 dB SPL. Shaded error bars indicate ± 1 standard error. Black bars indicate significantly different pupil size between quiet and noise conditions at the p < 0.05 level across the timescale shown.

Figure 6.. Pupillometry data for older CI users.

A) Mean pupil dilation as a percentage of dynamic range across the timescale of each trial of experimental task performed in quiet (orange) and in the presence of noise (cyan) at 65 dB SPL for OCI listeners. B) Same as A but at 85 dB SPL. Shaded error bars indicate ± 1 standard error. Black bars indicate significantly different pupil size between quiet and noise conditions at the p < 0.05 level across the timescale shown.