Age-related hearing loss (ARHL), or presbycusis, is a common condition involving gradual hearing loss due to aging (National Institute on Deafness and Other Communication, 2023). The increasing prevalence of ARHL globally is resulting in a major public health burden (Haile et al., 2021; World Health Organization, 2021). In the United States alone, a third of adults between 65 to 74 years and half over 75 years have ARHL (National Institute on Deafness and Other Communication, 2023). It is estimated that by 2030, ARHL will have an economic impact of 30 billion dollars (Stucky et al., 2010). Research in the past decade has shown that not only do individuals with ARHL experience typical auditory processing challenges, such as recognition of speech-in-noise (SiN), but they also experience non-auditory cognitive challenges beyond those seen with typical aging (CHABA, 1988; Lin, 2011; Humes, 2013; Powell et al., 2022; Shende and Mudar, 2023). Emerging work, including studies conducted by our group, has shown cognitive alterations even in those with mild ARHL in cognitive control assessed using visual modality (Shende et al., 2021, 2022).

Cognitive control refers to various mental operations that allow prioritization of information to accomplish target goals (Diamond, 2013; Mackie et al., 2013). An important cognitive control process is “value-directed strategic processing,” which refers to the preferential processing of salient information of higher value (Castel, 2007; Nguyen et al., 2019). This process is critical for efficient information processing in all sensory modalities. For instance, we strategically attend to warning signs related to hazardous conditions when driving on a freeway in a snowstorm compared to general information signs. In laboratory settings, value-directed strategic processing has been typically examined using visual word-list learning tasks (Castel, 2007; Castel et al., 2009; Murphy et al., 2024). In these tasks, words are paired with different point values ranging from low- to high-, such as 1–16 (1: low-value; 16: high-value), and participants are asked to recall as many words as they can with the goal of scoring maximum points. Typically, participants who perform well on these tasks recall more words of higher value, suggesting preferential value-based processing of information (Castel et al., 2007; Nguyen et al., 2019, 2020; Schwartz et al., 2023). It is important to note that the pairing of words with point values distinguishes strategic processing tasks from episodic learning tasks that use a list learning paradigm to measure recall irrespective of value as a measure of episodic learning (Castel et al., 2011; Elliott et al., 2020).

Value-directed strategic processing subsumes other attentional processes, including selective attention (Anderson, 2013). In the context of ARHL, selective attention has been long examined (Barr and Giambra, 1990; Craik, 2007; Shinn-Cunningham and Best, 2008; Van Gerven and Guerreiro, 2016; Dai et al., 2018), especially in the context of speech recognition (Humes and Dubno, 2010; Nuesse et al., 2018; Fuglsang et al., 2020; Shechter Shvartzman et al., 2022). On various tasks, such as coordinate response measure (Humes et al., 2006, 2017), attentive matrices (Bonmassar et al., 2022), n- back with distracting stimuli (Guerreiro and Van Gerven, 2017), Stroop (Guerreiro and Van Gerven, 2017; Gillingham et al., 2018; Ren et al., 2019; Huber et al., 2020; Chavant and Kapoula, 2022), and dichotic listening tasks (Shinn-Cunningham and Best, 2008; Torrente et al., 2022), studies have found alterations in selective attention in older adults with ARHL relative to younger and older adults with normal hearing. These findings suggest that changes in individuals with ARHL go beyond typical age-related declines in attention. Considering that these alterations have been found even on visual tasks, these findings suggest a more modality-general change in attentional processing in this population.

In contrast to the large evidence body on selective attention changes, value-directed strategic processing has remained largely unexplored in ARHL, with the exception of a study by our group (Shende et al., 2022). In our study, we examined behavioral changes in value-directed strategic processing in individuals with mild untreated ARHL relative to normal hearing (NH) controls with comparable age and education. We used an in-house developed value-directed list learning task where words were paired with binary values (10: high-value; 1: low-value) differentiated by letter case (Upper vs. Lower Case). See (Nguyen et al., 2019) for details of task development. Our results revealed that our untreated mild ARHL group recalled fewer high-value words relative to the NH controls, suggesting behavioral alterations in value-directed strategic processing beyond typical age-related cognitive control changes. Given that we used a visual task, these behavioral changes in individuals with ARHL offered preliminary evidence for modality-general alterations in value-directed strategic processing. Whether there are underlying neural alterations linked to these behavioral changes as one would expect based on theoretical postulations, such as information degradation (CHABA, 1988; Schneider and Pichora-Fuller, 2000), sensory deprivation (Lindenberger and Baltes, 1994; Baltes and Lindenberger, 1997), and common cause hypotheses (CHABA, 1988; Lindenberger and Baltes, 1994; Baltes and Lindenberger, 1997), remains unknown.

Event-related encephalography (EEG) is a useful technique to capture real-time temporal unfolding of value-directed strategic processing from the neurophysiological standpoint since it taps into neural activity time-locked to specific events (Sur and Sinha, 2009). Particularly measures linked to event-related spectral perturbations (ERSPs) capture both phase-locked and non-phase-locked spectral activity in the EEG signal (Makeig, 1993; Makeig et al., 2004). Using a value-directed strategic processing task, our group examined ERSPs in the theta and alpha bands in both cognitively healthy controls (Nguyen et al., 2019, 2020) and older adults with mild cognitive impairment (Nguyen et al., 2022). Overall, we found oscillatory activity in theta and alpha bands linked to strategic processing. Specifically, we found greater theta synchronization (more positive power relative to baseline) during processing of low-value relative to high-value words in frontal electrodes (Nguyen et al., 2019, 2020, 2022). Extant literature suggests that frontal theta activity is linked to cognitive control in general [see (Cavanagh and Frank, 2014) for a review] and is typically recorded at Fz and FPz electrodes [see (Mitchell et al., 2008)]. In particular, studies have linked frontal theta to general processing (Eschmann et al., 2020), working memory (Itthipuripat et al., 2013), and inhibitory control (Nigbur et al., 2011; Cavanagh and Frank, 2014), all of which are integral to value-directed strategic processing. We also found greater alpha desynchronization (more negative power relative to baseline) during processing of high- relative to low-value words in centroparietal and parietal electrodes in our study on value-directed strategic processing (Nguyen et al., 2019, 2020). Numerous studies have suggested that alpha desynchronization reflects information processing (Benedek et al., 2014; Fumuro et al., 2015) and selective attention (Klimesch et al., 1997, 2007).

In the context of ARHL, there is an emerging body of work on ERSPs (Petersen et al., 2015; Price et al., 2019) in relation to SiN recognition (Price et al., 2019) and auditory working memory tasks (Petersen et al., 2015). Findings suggest changes in alpha band power in those with ARHL relative to controls, specifically increased alpha power with mild and moderate degrees of hearing loss (Petersen et al., 2015). Studies have also found association between listening in challenging/noisy environments and parietal alpha power (Dimitrijevic et al., 2019; Wisniewski et al., 2021). In cochlear implant users, lower alpha desynchronization has been associated with increased listening effort (Dimitrijevic et al., 2019). In comparison to alpha oscillation, examination of theta in ARHL has received very little attention with some exceptions (Griffiths et al., 2020). To the best of our knowledge, none have examined theta and alpha oscillations linked to cognitively demanding tasks, such as value-directed strategic processing tasks, in this population.

The goal of the current study was to examine differences in theta and alpha power in individuals with unaided mild ARHL relative to NH controls of comparable age and education using our visual value-directed strategic processing task. Guided by the literature, including emerging work on neural oscillations in ARHL, and our previous studies on value-directed strategic attention in cognitively healthy and impaired populations, we hypothesized that the mild ARHL group would show lower theta synchronization and alpha desynchronization relative to the control group. To clarify, given the lack of studies on theta ERSPs in ARHL, our hypothesis of lower theta synchronization in the ARHL relative to the control group was guided by findings of our previous study (Nguyen et al., 2022), as well as expectation of poorer cognitive control in ARHL given that studies have shown poorer performance on cognitive control tasks in those with ARHL relative to NH controls (Lin, 2011; Powell et al., 2022; Shende and Mudar, 2023). We also hypothesized greater theta synchronization for low- versus high-value words and greater alpha desynchronization for high- versus low-value words. It is important to mention that we focused on individuals with unaided mild ARHL to understand whether mild alterations in hearing, which typically go untreated, impact neural activity linked to strategic processing. Furthermore, we intentionally chose to use a visual paradigm to minimize any confounds related to unaided hearing loss, and to ascertain whether neural changes, if observed, are modality-general in those with ARHL.

The current study examined ERSPs corresponding to a visual value-directed strategic processing task in older adults with ARHL relative to NH controls of similar age and education. Behaviorally, the ARHL group recalled fewer total words across all lists combined than the NH group. On EEG measures, two major findings related to group differences emerged: (1) the ARHL group had greater theta synchronization (100–200 ms) relative to the NH group during early processing periods; and (2) furthermore, the ARHL group had differences in theta synchronization (400–500 ms) for high- compared to low-value words, unlike the NH group. There were also some similarities observed in the two groups. Both groups showed greater alpha desynchronization for high- compared to low-value words (700–800 ms) although the magnitude of these differences was slightly more in the NH group. Also, both groups showed value-based differences in theta (200–300 ms; 700–1,000 ms) and alpha (600–1,000 ms).

Task-related behavioral data combined across all lists showed that the mild ARHL group recalled significantly fewer total number of words relative to the NH controls. These results could be attributed to just differences in episodic list learning between the groups. A few others have also found poorer episodic memory/list learning in individuals with ARHL relative to normal hearing controls, including on visual tasks of episodic memory (Rönnberg et al., 2011; Brewster and Rutherford, 2021; Tarawneh et al., 2022). Studies have shown that in ARHL, the brain undergoes functional reorganization (Campbell and Sharma, 2013; Campbell and Sharma, 2014; Lin et al., 2014; Armstrong et al., 2019), which can negatively affect the ability to retain information in memory (Loughrey et al., 2019). This is known as maladaptive plasticity, which has been found even in individuals with milder ARHL (Campbell and Sharma, 2014; Sharma and Glick, 2016; Slade et al., 2020). Changes in strategic processing likely contributed to the difficulties in list learning in individuals with ARHL relative to controls. Significant group difference in total points earned, with the ARHL group scoring lower total points than the NH controls, supports this speculation. To encourage strategic processing and learning of high-value words, participants were instructed to remember as many words as possible at the end of each list, with the goal of scoring a maximum number of points. However, participants were not given explicit instructions to focus on only high-value words since we wanted to examine inherent strategic processing ability. Our findings suggest that older adults with mild ARHL have compromised strategic processing compared to NH controls when point-values earned is used as a metric.

Task-related ERSP data revealed differences in neural processing between ARHL and NH groups in early time periods. Contrary to our hypothesis, we observed greater theta synchronization between 100 and 200 ms post-stimulus onset in the prefrontal electrode cluster in the ARHL group compared to NH controls. Theta synchronization in frontal regions has been linked to proactive and reactive control (Cooper et al., 2015; Messel et al., 2021; Pscherer et al., 2023). Proactive control allows one to adapt thoughts and behavior in anticipation of task goals, including any interference, whereas reactive control is a correction mechanism that involves detection and resolution of interference after its onset (Braver, 2012; Stuphorn and Emeric, 2012; Botvinick and Braver, 2015). Perhaps higher theta synchronization in the ARHL group relative to controls was related to the engagement of additional neural resources to support proactive control of task performance. This compensatory mechanism in the ARHL group to better support overall processing related to the task was still not sufficient to match their behavioral performance to NH controls, since they recalled fewer total number of words and earned lower total points. We have to be cautious in linking our ERSP findings to behavioral data given that behavioral data in our value-directed list learning paradigm includes cognitive processes beyond what is linked to strategic processing captured by the ERSP findings. In general, word recall on list learning tasks involves initial processing, learning/encoding, and subsequent retrieval. While our instructions promoted strategic processing, when EEG data were collected, participants were not asked to provide any overt response after each word in the list was presented. The task was designed this way intentionally to examine real-time strategic processing with minimal confounds from other cognitive processes (e.g., retrieval) or motoric response from a button press. We elicited recall at the end of each list to demonstrate that the behavioral results are consistent with value-directed strategic processing and learning behavioral literature (more high-value words recalled compared to low-value words). While behavioral data in terms of recall at the end of the list provides some insight into strategic processing, given that such recall involves cognitive processes related to word encoding and retrieval, not captured by the ERSPs, these findings need further validation. Future studies should consider how this task could be modified to collect ERSP data that can reveal possible associations between top-down cognitive processes of strategic processing and subsequent behavioral recall of high and low-value words to verify these findings. Higher theta synchronization in the ARHL relative to the control group between 100–200 ms could also point toward alterations in bottom-up sensory processing. Studies examining electrophysiological activity involving visual tasks have associated processing within 200 ms post-stimulus onset to visual awareness (Koivisto and Grassini, 2016) and sensory processing (Loughrey et al., 2023; Luck, 2023). Higher theta synchronization in the ARHL group relative to controls could be related to hyperexcitability due to cross-modal sensory enhancement. This aligns with evidence from electrophysiological studies (Puschmann et al., 2014; Slade et al., 2020; Madashetty et al., 2024), including a recent study that examined neural correlates of visual working memory tasks in older adults with ARHL (Loughrey et al., 2023). Our theta finding in the ARHL group is consistent with a few other studies (Petersen et al., 2015; Griffiths et al., 2020) and furthermore adds to the rapidly emerging body of work showing that sensory and cognitive processing interact more than has been previously acknowledged (Rönnberg et al., 2022a,b).

Group differences in information processing based on value (interaction between group and value) were observed at later time points (400-500 ms). Specifically, the ARHL group had greater theta synchronization for high- compared to low-value words. This was not observed for the NH controls. We did not expect this given that our previous studies have not found similar differences between theta synchronization for high- versus low-value words in cognitively healthy adults (Nguyen et al., 2020) and those with mild cognitive impairment (Nguyen et al., 2022). Whether these findings are specific to the ARHL group needs further validation. With theta synchronization linked to proactive control (as mentioned earlier), it could be that our participants with ARHL proactively engaged their top-down processes to strategically attend to and process high-value over low-value words as part of a compensatory neural mechanism as has been postulated by others (Loughrey et al., 2023). Studies have reported increased neural engagement of higher-order cognitive functions on non-auditory cognitive tasks in individuals with mild degrees of ARHL (Slade et al., 2020; Zhao et al., 2022; Loughrey et al., 2023). Given that theta is also linked to inhibitory control (Nigbur et al., 2011; Cavanagh and Frank, 2014; Cavanagh and Shackman, 2015), alternately, our findings might suggest that the ARHL group did the unexpected, i.e., used more neural resources linked to inhibitory control during the processing of high-value words (in other words, inhibited processing of high-value words), suggestive of maladaptive plasticity. However, our behavioral data does not support these findings, although further research, including modification of the current task, is necessary to truly connect our behavioral and ERSP findings. Regardless, our theta findings point toward potential cross-modal changes in those with mild ARHL, considering that the NH group did not demonstrate theta differentiation for high- versus low-value words. This aligns with emerging studies that have reported cross-modal plasticity in those with mild to moderate ARHL, especially in the visual cortical areas (Campbell and Sharma, 2014; Puschmann et al., 2014; Sharma and Glick, 2016; Ponticorvo et al., 2022). Further examination is needed to better delineate cross-modal reorganization in the context of visual strategic processing in ARHL. In general, our findings show that ARHL influences theta oscillations at earlier time points on a visual cognitive control task. Our findings align with other studies, both human (Heinrichs-Graham et al., 2022; Ryan et al., 2023) and animal (Johne et al., 2022), which have shown that hearing loss modulates oscillatory activity in the theta band. However, adequate comparison of our findings with others is difficult since we used a visual task while most have examined theta activity in relation to auditory tasks, primarily SiN recognition (Doelling et al., 2014; Hyafil et al., 2015; Etard and Reichenbach, 2019).

Some similarities were observed in both groups in strategic processing in the alpha band between 700 and 800 ms at the centroparietal electrode cluster. As hypothesized, both ARHL and NH groups had greater alpha desynchronization for high- than low-value words suggesting similar alpha modulation for processing information of varying values (i.e., high vs. low) in later time points. Alpha desynchronization in parietal regions is considered to reflect the engagement of attentional processes during selective attention (Foxe and Snyder, 2011; Deng et al., 2019; Trajkovic et al., 2023). The effect size of these alpha findings (more alpha desynchronization for high- than low-value words) was marginally larger in size within the NH group (Cohens’ d = 1.2) relative to the ARHL group (Cohen’s d = 1.12), suggesting a more robust modulation in the control group. We also found some similarities in our groups in theta oscillations. Greater frontal theta synchronization was observed for low-value relative to high-value words between 200 and 300 ms, and 700–1,000 ms. This suggests that both ARHL and NH groups strategically allocate more cognitive control resources to inhibit the processing of low-value information to prioritize the processing of high-value information, and these findings are consistent with our hypotheses and with our previous studies that used the same task in cognitively healthy (Nguyen et al., 2019, 2020) and cognitively impaired (Nguyen et al., 2022) populations. Both groups also showed greater alpha desynchronization for high- versus low-value words, between 400 and 1,000 ms in the centroparietal electrode cluster and between 200–300 ms and 600–1,000 ms in the parietal cluster, indicating greater attentional allocation for processing the high-value versus low-value words. The similarities between ARHL and NH groups, especially during the later time windows (700 ms and beyond) is similar to the findings of ERP studies involving P3 and late positive potential (LPP) on visual paradigms (Loughrey et al., 2023; Zhao et al., 2023). Perhaps, by the later stages of processing, those with ARHL are able to “catch up” to their age-matched NH peers by utilizing compensatory neurocognitive resources.

This study has several limitations. We analyzed ERSPs from a subset of electrodes that were defined a priori based on past literature. Future studies that use data-driven approaches, such as principal component analyses, would be useful to corroborate findings and advance knowledge related to neural oscillatory changes in older adults with mild ARHL. Future studies should examine ERSP changes with varying measures (e.g., power, phase coherence), as well as utilize complementary techniques such as source localization with EEG or functional magnetic resonance imaging to further unravel the neural mechanisms that underlie strategic processing in this population. Additionally, our study had a small sample size, and replicability from future studies is critical before the generalization of results. Finally, we did not develop the value-directed strategic processing task to distinguish between the processes of encoding, storage, and retrieval, and their examination is critical to understanding changes in value-directed strategic processing in ARHL.

In summary, the current study examined ERSPs underlying visual value-directed strategic processing in older adults with mild ARHL relative to NH controls of comparable age and education. Behavioral data revealed observable differences in total recall of words and total points earned, which points towards cognitive changes in those with ARHL. Changes in theta band were observed during early time periods in ARHL, but these were not specific to value-based processing, indicating more general changes in proactive control and cross-modal sensory enhancement. During the mid-time points (400–500 ms), the ARHL group showed greater theta modulations for high-value compared to low-value information, while these differences were not significant in controls. This might indicate adaptive changes in those with ARHL during strategic processing. In later time widows (after 700 ms), theta and alpha modulations were similar in both groups, perhaps indicating that the ARHL group does “catch up” to the NH controls. Our study adds to the emerging body of work on neural oscillatory changes underlying a cognitive control task in individuals with ARHL. Given that neural changes typically precede behavioral changes, these ERSPs can be used to assess the benefits of intervention (e.g., hearing aids, auditory and cognitive intervention programs) before behavioral changes are observable, especially given that our sample included individuals with untreated mild ARHL.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

The studies involving humans were approved by Institutional Review Board of the University of Illinois Urbana-Champaign (protocol # 17067). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

SS: Data curation, Formal analysis, Investigation, Methodology, Project administration, Visualization, Writing – original draft, Writing – review & editing. SJ: Visualization, Writing – review & editing. RM: Conceptualization, Funding acquisition, Methodology, Resources, Supervision, Writing – review & editing.