Imagine you are at a party or at a busy restaurant and you are trying to follow what your conversation partner is telling you. Despite the music, conversations of the other guests, and the sound of clattering dishes in the background, you are able to understand most of the story. You are able to direct your attention to your conversation partner, make use of contextual information, and fill in the gaps of bits of the conversation you may have missed. Of course, it’s important for the auditory system to accurately encode sounds. However, you also need to be able to make sense of the input received. In other words, successful speech perception—especially in challenging environments—relies on cognitive processes to supplement the “bottom-up” auditory information transmitted by the peripheral auditory system. 

Cognitive processing is even more important for individuals with hearing loss than for individuals with normal hearing, because auditory information is deficient as a result of cochlear pathology. To compensate for degraded input signals, a listener will need to rely more on cognitive processes to reconstruct the message and fill in the gaps of anything they may have missed. Just as clients vary in terms of their hearing abilities, they differ in terms of their cognitive abilities. And, just as differences in audiometric thresholds can affect treatment success, so can individual differences in cognition. Such differences may partially explain why two clients with the same audiogram respond differently to hearing aid signal processing or vary in their ability to understand speech in complex environments. 

Over the past decade, hearing researchers have become increasingly interested in the role of cognition in auditory perception (e.g., Arlinger et al, 2009). Although there is a growing body of research in this area, the standard audiometric evaluation still emphasizes the peripheral auditory system by emphasizing pure-tone air and bone conduction thresholds and speech in quiet. A similar emphasis on peripheral auditory abilities occurs for hearing aids, where the initial hearing aid response is based on the pure-tone audiogram and adjusted according to user feedback (Anderson et al, 2017). While clinicians are interested in more individualized clinical treatment, most are not yet considering cognitive abilities in diagnostic and rehabilitation decisions. Should we be? And if the answer is yes, what do clinicians need to understand about cognition? 

Cognitive Abilities Are Important to Audiology Practice 

While we sometimes refer to cognition as a single entity, it includes multiple abilities (see TABLE 1). Although all of the abilities listed impact perception to some extent, working memory has been a focus of cognitive hearing science work on speech recognition. The role of working memory is summarized in a model of perception, the Ease of Language Understanding (ELU) model (Rönnberg et al, 2013). In the ELU model, auditory and/or visual inputs are decoded according to their lexical representation. 

TABLE 1. Multiple Cognitive Abilities

COGNITIVE ABILITY

WHAT IS IT?

Working Memory

Holding information in memory while processing the same or new information at the same time. Working memory capacity refers to the ability to simultaneously store and process information (Daneman and Carpenter, 1980; Baddeley, 1992). This process is integral to speech recognition because it allows the listener to retain stored information (e.g., recently spoken words or phrases) while processing incoming information. 

Processing Speed

Rapidly responding to a simple stimulus presented through any sensory modality. During speech recognition, listeners process speech sound constantly, extract syllables/words out of the continuous acoustical stream, temporarily store them in memory, and put new information into context for understanding spoken sentences. The unfolding of this process over milliseconds to seconds taxes processing speed (e.g., Crowder, 1981; McClelland and Elman, 1986).

Selective Attention

Focusing on and reacting to certain stimuli while ignoring irrelevant information. In this context, attention refers primarily to the ability to selectively focus on the talker of interest while ignoring any competing sounds, as well as switching attention among multiple talkers in a conversation. 

General Sequential Reasoning

Reaching a solution to a problem by taking multiple steps based on rules and conditions.

Lexical Knowledge

Understanding the correct meaning for a word. Unlike other cognitive abilities, which show age-related declines, lexical knowledge is preserved with age (Verhaeghen, 2003). 

 

A clear and audible signal is easily interpreted for meaning. However, degraded signals (as in the case of hearing loss or environmental distortions such as noise and reverberation) are not easily matched to stored lexical representations. This lexical “mismatch” forces engagement of working memory to decode the distorted input. Because working memory has a finite capacity, it may not be possible to accurately decode all information without overloading the system. And, in clinical terms, that overload means misunderstood or missed information, a gap in communication, and even frustration or disengagement by the listener. 

Measuring Cognitive Ability in the Clinic

At the present time, there are few options for working memory tests that can be readily used in clinical test batteries. Two currently available tests are the Word Auditory Recognition and Recall Measure (WARRM Smith et al, 2016), and the Reading Span Test (RST) (Daneman and Carpenter, 1980; Rönnberg et al, 1989; for its short version, see Ng et al, 2013). 

WARRM is a new test that simultaneously assesses auditory word recognition and working memory capacity. In this test, monosyllabic words are presented in quiet. The individual is asked to repeat back the word and to make a judgment about the first letter of the word (to engage cognitive processing) after each word. In addition, he or she has to recall the words after a certain number of words have been presented. Because the WARRM requires minimum extra resources for implementation, it may be relatively easy to incorporate into an audiometric battery. 

In the RST, the individual reads visually presented sentences and makes a semantic processing judgment. After a predefined number of sentences are presented, the individual is asked to recall keywords from the sentences. The accuracy of word recall is taken as the measure of working memory capacity. The RST takes approximately seven to 15 minutes to administer depending on the version used (i.e., short or long version). This test may be suitable for clients with significant hearing loss, as it will not be affected by hearing difficulty. For clients with visual impairment, similar tests are available in an auditory format (e.g., Pichora-Fuller et al, 1995). 

To triage individuals according to working memory, a reasonable approach is to consider individuals with scores that are substantially above or below the test’s median score (typically 35–40 percent for the RST and about 3.5 words for the WARRM) as having high or low working memory capacity. With individuals whose scores are close to the median score and could plausibly be considered to fall in either group, other information (e.g., age and reported communication difficulty) can be considered in clinical decisions.

An alternative metric is a cognitive screening test, which assesses and quantifies global cognitive functioning in older adults. While there are many options (see Shen et al, 2016 for a recent review), tests that include measures of working memory and executive functions are most likely to provide information that is helpful in explaining speech recognition difficulty. For example, the Montreal Cognitive Assessment (MoCA) (Nasreddine et al, 2005) includes questions that assess working memory. Therefore, the MoCA or a test with a similar format could be a good option for clinical use, particularly for clinicians whose caseloads are dominated by the older adults likely to experience declines in cognitive ability.     

Even if it isn’t feasible to assess cognition directly, it can still be part of a clinical decision matrix. Because poor speech-in-noise recognition is related to low working memory (e.g., Souza and Arehart, 2015), adults who report difficulty in noise disproportionate to their pure-tone audiogram may have lower-than-average working memory (Note, however, that this relationship might not apply for individuals likely to experience poor speech-in-noise recognition due to peripheral distortions, such as those with severe-to-profound loss or auditory neuropathy). Clients with a prior diagnosis of mild cognitive impairment are also likely to experience detrimental consequences of impaired cognition in difficult listening situations. 

Considering Cognition in Treatment Decisions

Working memory plays a crucial role in successful speech recognition in noise. In a review of 20 studies, Akeroyd (2008) found that working memory capacity explained listeners’ abilities to understand speech in noise better than a variety of other cognitive measures. Specifically, individuals with low working memory had more difficulty in noisy situations than individuals with high working memory, even when adequate amplification was provided. Individuals with low working memory also have more difficulty recognizing speech in reverberant rooms, such as auditoriums (Reinhart and Souza, 2016). 

Accordingly, individuals with low working memory should be encouraged to use noise-reducing technology, including directional microphones, remote microphones, TV headsets, and loop receivers (even if not a hearing aid wearer). They may benefit from visual supplements, such as caption phones, television captions, and video calls in unfavorable listening environments. Although the individual (and some audiologists) may feel they don’t need such accommodations, remember that cognition is a limited-capacity system. The more cognitive resources must be devoted to decoding the signal, the fewer resources are available for other cognitive operations: remembering, reflecting, and responding. Individuals with low working memory can also benefit from extra counseling and/or aural rehabilitation training that supports good communication strategies and environmental modification. 

Working memory also affects the response to hearing aid processing. The relationship appears to be tied to susceptibility to an altered acoustic signal (the “mismatch” scenario posed by the ELU model). The relationship is strongest for fast-acting WDRC, where a dozen studies have shown that adults with lower working memory capacity receive more benefit from slow-acting (and less distorting) WDRC than from fast-acting WDRC (e.g., Gatehouse et al, 2006; Lunner and Sundewall-Thoren, 2007; Ohlenforst et al, 2015; Souza and Sirow, 2014). 

Adults with higher working memory are able to benefit from the improved audibility supported by fast-acting WDRC systems; and in general show more constant benefit as WDRC speed is changed. Altering the acoustic signal through frequency lowering or digital noise reduction is more weakly related to working memory, but the available evidence suggests that extreme signal processing should be avoided in low working memory individuals (Souza et al, 2015). 

A reasonable approach would be to set processing parameters (such as frequency compression ratios) for sufficient audibility, but avoid over-processing in an attempt to “improve” the signal. The audiologist should consider whether the default processing parameters meet those goals and undertake necessary modifications if not. Product choices that allow for adjustment of processing parameters (i.e., the ability to change WDRC speed and/or adjust frequency lowering settings) facilitate such adjustments, especially when cognitive status is not known a priori, or changes over time.  

Including Cognition in Counseling

While researchers continue to acknowledge the importance of cochlear dysfunction in its own right, they also recognize that cochlear dysfunction can impact cognitive processing: the more deficient the signal at the level of the cochlea, the more compensation effort is required from the cognitive system to make sense of degraded input. Listeners with more limited cognitive abilities—who may not have sufficient resources to support that compensation—are at a disadvantage. Within this view, cognitive abilities also become relatively more important when the listening environment is difficult (e.g., noisy, distorted, or reverberant). The end result is a broader view of communication that incorporates both peripheral and cognitive abilities. 

If cognition has been measured directly, results and implications can be discussed with the client. With many of our clients concerned about age-related changes in cognition and/or memory, the idea that working memory has been measured and found to be “poor” is not a productive approach. Instead, we have found it helpful to present working memory scores, if available, as one of many varied individual abilities, much as the pure-tone audiogram is discussed as occurring within a range of possible hearing abilities with the focus on addressing challenges associated with that audiogram. Individuals falling below the passing score on a cognitive screening test who are not already receiving medical follow-up for cognitive concerns should be appropriately referred.

Whether the individual’s working memory has been measured or not, a discussion of the role of cognition in communication can facilitate counseling. Clients (and their families) often approach audiology treatment with emphasis on the device. They expect that the “right” hearing aid will be recommended, and consequently they will hear well. This passive view of the device as the sole solution is at odds with reality, and certainly with the role of cognition in difficult listening environments. The clinician who focuses all of his or her attention (and appointment time) on the role of the device does the client a disservice. As we know, no hearing aid will make speech intelligible in all environments, and the client who believes otherwise is sure to be disappointed. A single focus on the hearing aid also diminishes the listener’s role in actively managing his or her own experience. 

A more realistic approach is to discuss with the client that communication is an aggregate of information gathered by the auditory system (assisted by the hearing aid) and the way in which the transmitted information is interpreted and processed by the cognitive system. In our own research and clinical experience, we’ve counseled many clients in this way. Our clients understand and accept that listening involves both ears and brain. Many individuals raise the issue themselves, noting that listening requires effort and that they are aware that they are compensating for poor hearing. 

Formal qualitative research confirms that older adults are aware of and concerned about the role of cognition in perception (Preminger and Laplante-Levesque, 2013). When individual cognitive ability is introduced into a treatment planning discussion in an informed and supportive way, clients are not confused or concerned. In fact, we’ve found that a discussion that includes cognitive abilities and their role in communication facilitates a better client-clinician partnership and, often, increases confidence in the clinician’s knowledge and support.

Looking Ahead

As we look to the future, hearing aids (or other sound amplification devices) are (or will be) available through direct purchase (PCAST, 2015), reducing the need to see an audiologist simply to purchase a device. Our clients are living longer and presenting with more varied health and cognitive abilities than ever before. These changes are occurring in the context of a new understanding of (and ability to assess) the whole person: their needs, limitations, strengths, and weaknesses; and a desire to use the wealth of signal processing and assistive device choices to customize the solutions to the individual. 

A broader understanding of the interaction between peripheral and cognitive abilities is central to these efforts. There are still research and development needs— more sensitive and time-efficient cognitive tests; a more precise understanding of which signal processing parameters will optimize performance for different peripheral-cognitive profiles; and more information about how listeners learn and adapt to those parameters. Nonetheless, audiology practice is a dynamic, evolving field supported by clinicians who can embrace a broader view of perception—which includes cognition—to serve their clients.