Brain Correlates of Verbal Working Memory in Children with Hearing Loss: Auditory Experience Matters

Abstract

Childhood hearing loss is a significant risk factor for language and academic delays.1–5 At the same time, the current generation of children with hearing loss show enormous variability in outcomes, with some children with hearing loss performing at the level of children with normal hearing, while others fall significantly behind.5 There has been a notable interest in recent years to determine the locus of variability in outcomes in children with hearing loss, including children with mild-to-severe hearing loss (i.e., children who are hard of hearing) who wear hearing aids. Researchers have pinpointed various factors that account for individual differences in children who are hard of hearing, including the quality and quantity of hearing intervention (e.g., amount of hearing aid use, amount that the hearing aid adequately restores access to speech),3–6 as well as cognitive factors such as working memory and attention.7,8 However, a comprehensive understanding of how these factors interact is still in its infancy, and the neural mechanisms underlying these interactions is unknown.www.shutterstock.com, audiology, hearing impaired, neural dynamics.Figure 1: Differences in brain activity between children with and without hearing loss. There was an increase in activity in the rightfrontal and parietal cortices in children who are hard of hearing (CHH) relative to children with normal hearing (CNH). Activity inthese regions correlated with verbal intelligence across groups. Audiology, hearing impaired, neural dynamics.Figure 2: Impact of hearing aid use on neural activity. Increased hearing aid (HA) use, especially more than ~8.5 hours of use,was correlated with more typical patterns of brain activity serving working memory processing. Audiology, hearing impaired, neural dynamics.On the heels of recent work using electroencephalography (EEG) and functional magnetic resonance imaging in people with severe-to-profound hearing loss,9–12 our lab is taking a new approach to understanding the impact of auditory experience on cognitive development by quantifying the real-time neural dynamics underlying these language and cognitive processes using magnetoencephalography (MEG). MEG is a noninvasive neuroimaging technique that records the small magnetic fields that naturally emanate from active neuronal populations in real time while a participant performs a task. MEG has a good combination of spatial accuracy (4-5 mm) and temporal precision (1 ms), which allows us to investigate the complex neural dynamics underlying cognitive and language processes as they evolve. In other words, these data provide a quantitative measure of neural activity throughout the brain during each stage of a behavioral process. This tool holds significant promise in determining the brain mechanisms and cognitive subprocesses that underlie differences in behavioral performance between children with normal hearing (CNH) and children who are hard of hearing (CHH), as well as between individual CHH. CURRENT STUDIES We had two goals in the current studies.13,14 The first goal was to identify the impact of hearing loss on the neural dynamics underlying verbal working memory processing groupwise between CHH and CNH, and the second was to determine the impact of the amount of hearing aid use on individual variability in brain activity within the hard-of-hearing group only. We analyzed data from a total of 30 children ages 7-15 years old, including 14 children with mild-to-severe hearing loss (9 females) and 16 children with normal hearing (8 females), who underwent neuroimaging with MEG while they performed a letter-based version of the Sternberg working memory task, as well as neuropsychological testing. During the verbal working memory task, participants were initially presented visually with six consonants for 2.0s (encoding period). The letters disappeared for 3.0s (maintenance period), then a probe letter appeared (retrieval period). Participants were asked to respond with a button press whether the probe letter was one of the previous six letters. The two groups did not significantly differ in reaction time or accuracy on this task. Finally, participants completed the Weschler Abbreviated Scale of Intelligence (WASI-II). Scores from the WASI-II were used to determine each participant’s verbal intelligence, nonverbal intelligence, and overall IQ. Parents also filled out questionnaires regarding the child’s hearing aid use that included questions such as, “How many hours per day does your child wear their hearing aids during the school year Mon-Fri? What about Sat-Sun?” Finally, degree of hearing loss was calculated in the children who are hard of hearing using their most recent audiogram. We used the MEG data to create whole-brain images of encoding- and maintenance-related neural responses separately, then compared the images groupwise to probe whether children who are hard of hearing showed differences in the neural patterns underlying each phase of verbal working memory compared to their normal hearing peers.13 We found that children who are hard of hearing exhibited significant elevations in right frontal activity during encoding, as well as elevations in maintenance-related right parietal activity relative to children with normal hearing. For context, the right prefrontal cortex is homologous to one of the most important regions for encoding verbal stimuli into memory storage,15,16 while the right parietal cortex is critical for working memory maintenance.17 Thus, increases in activity in these regions in the hard-of-hearing group was likely compensatory, which enabled the children to perform the task as well as their normal hearing peers. Crucially, we also found that activity in these two regions significantly correlated with verbal intelligence across groups, suggesting a tight link between working memory-related neural activity and language ability, in line with previous work (Figure 1).7,13,18,19 Our second goal was to determine whether working memory--related neural activity was sensitive to amount of hearing aid use in the children who are hard of hearing only.14 To this end, we performed whole-brain correlations between hours of hearing aid use and encoding-related and maintenance-related brain images separately, controlling for degree of hearing loss. We found significant correlations between hearing aid use and bilateral occipital and right precentral encoding-related activity, as well as maintenance-related activity in the right frontal cortex, above and beyond any effects of degree of hearing loss. In all these regions, more hearing aid use was correlated with a “normalization” of neural activity. Specifically, occipital encoding-related activity is known to be important for effective working memory encoding, and greater hearing aid use was associated with an increase in occipital responses. On the contrary, precentral and right prefrontal responses are generally considered atypical or compensatory in these types of tasks, and increased hearing aid use was associated with a decrease in activity in these regions (Figure 2). Critically, the “break even” point in these data, or the amount of hearing aid use that was related to more typical neural patterns, was at least 8.5 hours per day.14 CLINICAL IMPLICATIONS These data provide striking evidence that children with mild-to-severe hearing loss exhibit compensatory alterations in brain physiology during the performance of cognitive tasks. More consistent hearing aid use was associated with normalized neural patterns in children who are hard of hearing. It is important to note that brain differences were found despite there being no differences in behavioral performance between groups. These tasks were performed in a controlled laboratory setting, so it is possible that real-life environments, where there are multiple cognitive demands or distractions, may lead to an exhaustion of these compensatory resources and eventual behavioral decrements in children who are hard of hearing. Expanding on this idea, variability in the neural patterns that serve language and cognitive processes, especially relative to the difficulty of a given task, may explain why some studies find behavioral differences between those with and without hearing loss, while others find no such differences. Even more remarkable is the relationship between hearing aid use and neural activity in children with hearing loss. We found significant relationships between the neural patterns serving each phase of working memory and hearing aid use, such that consistent hearing aid use was related to more prototypical neural dynamics. These data suggest that the amount of quality auditory experience has substantial effects on their underlying brain physiology, and that at least 8.5 hours of hearing aid use seems to be beneficial, at least from a neural perspective. This aligns well with previous work suggesting that at least 8-10 hours of hearing aid use relates to better behavioral outcomes in these children.5,8 More broadly, our studies provide preliminary evidence that neuroimaging may hold promise in uncovering the mechanisms underlying differences in language and cognitive development between children with and without hearing loss, and may help explain the variability in outcomes between individuals with hearing loss.