From the desk of Gus Mueller
I met up with an old friend of mine the other day, and learned that he had just purchased hearing aids. I recall him telling me some time ago that he was having problems at work, a management job which consisted of attending a lot of meetings. He has a mild high-frequency loss in the 3000-4000 Hz range; someone most of us would consider a "borderline" candidate. I asked him how he was doing with the hearing aids. "I'm hearing things I've never heard before" was his first reply. I then specifically asked if the hearing aids were helping him understand speech better. His reply was interesting: "I don't really know that I'm understanding any better," he said, "but I can tell you this, at the end of the day I'm not as tired."
Listening, especially when comprehension is important, is tiring, even for someone with normal hearing. It can be a lot more tiring for someone with a hearing loss. The effort expended can potentially interfere with performing other simultaneous tasks. When we fit hearing aids, we normally think about improving audibility, restoring appropriate loudness perceptions and improving speech understanding. We don't think (directly) much about reducing listening effort, or even test to see if the use of amplification is indeed helping in this area.
Researchers, however, have been thinking about listening effort and listening fatigue as it relates to hearing loss, the use of hearing aids, and special features of hearing aids. We found one of the best of this group to join us for this month's 20Q.
If you think of Vanderbilt University when you hear the name "Ben Hornsby" it's little wonder, as that has been his home for the past 20 years. He is currently an Assistant Professor in the Department of Hearing and Speech Science at Vanderbilt. In addition to his own research, he is actively involved in teaching and research projects with both AuD and PhD graduate students in Audiology.
If you've followed Dr. Hornsby's work over the years you know that much of it focuses on identifying factors responsible for the individual variability associated with the benefit from and satisfaction with amplification, and developing methods to minimize the perceptual and psychosocial consequences of hearing loss. You've seen his many publications on these topics; Dr. Hornsby has also served in numerous journal editorial positions—currently, he is the Amplification Section Editor for Ear and Hearing.
Ben even thinks about "effort" and "fatigue" when he's not in the lab, as he spends his time away from the office rock climbing (as in mountains). We were able to convince him to put down his carabiners and quickdraws for a weekend, so that he could join us for this 20Q. The study of listening effort is intriguing, and Ben's excellent review provides a good summary of what we know, and the numerous areas that still needs to be researched.
Gus Mueller, Ph.D.
To browse the complete collection of 20Q with Gus Mueller articles, please visit www.audiologyonline.com/20Q
Dr. Mueller interviewed Dr. Hornsby on a few of these topics last year at Vanderbilt - you can view their conversation here.
1. The topic of "listening effort" sure has gotten a lot of press lately. What exactly do people mean by this?
Well, listening effort is hard to define but to butcher, I mean borrow, a quote from the late Supreme Court Justice Potter Stewart, "I know it when I see it." By this I mean many of us have probably had direct experiences in situations where "listening is effortful" or you have to "mentally work" in the listening situation. A standardized definition of listening effort, however, has not been agreed upon, so it's hard to say what people really mean. One way we might think about listening effort during speech tasks is in terms of the amount of cognitive resources that are being allocated to process and understand speech (e.g., Zekveld, Kramer & Festen, 2011). Anecdotal reports, and more formal research, suggest that individuals with hearing loss have to "work harder" cognitively to process auditory stimuli (like speech) than people without hearing loss (Hetu, Riverin, Lalande, Getty, & St. Cyr, 1988; Nachtegaal, et al., 2009).
2. Is this related to how much hearing loss a person has?
Maybe. Traditional audiologic measures, such as pure tone threshold testing and measures of speech recognition, give us valuable information about auditory function and processing abilities. For example, measures of speech recognition tell us how much someone understands when speech is presented in noise at a conversational level. However, these measures don't tell us how much "effort" was exerted to achieve that level of understanding. Certainly, in some situations listening effort and speech recognition abilities will be related. For example, in general as listening conditions worsen and speech recognition decreases, listening effort increases (Mackersie & Cones, 2011; Zekveld, et al., 2011). However, at any given performance level both subjective and objective measures of listening effort can vary widely between individuals (Hornsby, 2011; Picou, Ricketts, & Hornsby, 2011).
3. Are complaints of increased listening effort a common thing that people with hearing loss report?
Both anecdotal and formal research suggest that the need for increased concentration and attention when listening is a fairly common complaint for people with hearing loss and one that may have significant negative consequences. For example, here are some anecdotal reports from people with hearing loss I found recently via an Internet search of blogs and message boards:
- "Listening IS exhausting!"
- "On a typical day, it can be tiring to put in the effort to listen, especially when you deal with all kinds of people all day long,..."
- "I go to bed most nights with nothing left. It takes so much energy to participate in conversations all day that I'm often asleep within minutes."
- In regards to the consequences of listening with hearing loss one person reported "...at the end of every day I am physically exhausted."
4. Yep, those comments all sound familiar. You mentioned there was formal research on the topic?
There is. Most of the evidence supporting these anecdotal reports comes from subjective interviews and survey studies. Research in this area suggests that increasing listening effort may be a coping strategy used by people with hearing loss to compensate for communication and auditory processing difficulties. To maintain optimal communication, individuals with hearing loss may report a need for increased concentration, attention and focus, compared to individuals without hearing loss. Common consequences of sustained listening effort appear to be increases in subjective reports of stress, tension and fatigue (e.g., Hetu, et al., 1988).
5. Not to be callous, but I have normal hearing and I still get stressed, feel tense, and get fatigued at times. Is this really a serious problem?
Well it certainly can be. There is evidence to suggest that these subjective effects have functional consequences that are increased in persons with hearing loss. For example, Nachtegaal, et al. (2009) found that individuals with hearing loss had a higher "need for recovery" from work. In other words following a day's work, individuals with hearing loss needed more time to feel "recovered" from the effort exerted during the workday. Likewise, Kramer, Kapteyn, and Houtgast (2006) found that the increased effort and the resulting mental stress experienced by workers with hearing loss impacted their job performance. Individuals with hearing loss were more likely to miss work due to "fatigue, strain or burnout" than individuals without hearing loss working at a similar or the same job (in some cases in the same building!). Also, although we've been talking about effects on adults, children with hearing loss are also at risk for serious consequences due to increased and sustained listening effort in school settings (Hicks & Tharpe, 2002; Pittman, 2011).
6. Well it sounds like it's worth "listening" to someone talk about "listening effort". Is listening effort something we can reliably measure?
In recent years there has been a resurgence in research exploring ways to reliably and efficiently measure listening effort. Methods for quantifying listening effort generally fall into one of three broad categories: subjective measures, physiologic measures, and psychophysical or behavioral measures.
7. They all sound like interesting areas. Can we go through them one at a time?
Sounds good to me—let's start with subjective measures. Our brain is a powerful processor and, in some cases, it may allow us to perceive differences in conditions that are too subtle to be detected by more common objective tests, like speech recognition. As I mentioned before, open-ended and structured interviews and surveys have been very useful in identifying the subjective impact of increased listening effort. In addition, validated and study specific questionnaires and rating scales also have been used to quantify listening effort. The Speech, Spatial and Qualities of Hearing Scale (SSQ), developed in 2004 by Gatehouse and Noble, is an example of a validated questionnaire that has been used to assess hearing disability in several areas. One area assessed by the SSQ is the "quality of the hearing experience" and includes items such as ease of listening and the ability to attend to or ignore sounds. This tool was sensitive to the benefits of bilateral compared to unilateral hearing aid fittings for reducing listening effort (Noble & Gatehouse, 2006). In addition, several researchers have included study-specific subjective ratings scales to assess "ease of listening" and listening effort and in several cases found the subjective ratings more sensitive than other behavioral or objective measures for detecting changes in listening effort (e.g., Hallgren, Larsby, Lyxell, & Arlinger, 2005).
8. Hmm... if subjective measures are more sensitive, why not just use those?
Subjective ratings provide important information, but alone they provide only a partial picture of the potential impact of listening effort. Subjective data can be highly variable, in part, because each person's interpretation of a question or survey item may vary. For example, two individuals listening in the same situation and with similar speech recognition abilities may respond very differently in response to the question "How effortful was listening in that situation?" The differences in responses may be due to true differences in cognitive processing demands or may simply reflect different "thresholds" for considering something "effortful". In addition, subjective responses are influenced by individual biases and so it is not always clear whether the responses accurately reflect the consequences of hearing loss on an individual. Likewise, standardized questionnaires may not address situations or settings that are specific to issues for a given individual. Together, it is clear that additional measures would be helpful to characterize the impact of hearing loss on listening effort.
9. So other, more objective, measures are needed?
Right. Researchers have examined a variety of physiologic markers in an attempt to identify correlates of cognitive processing demands associated with increased listening effort. For example, Hicks and Tharpe (2002) measured cortisol levels in school aged children with and without hearing loss. Cortisol is a hormone produced by the adrenal gland in response to stress. Changes in cortisol levels are associated with changes in fatigue and effort.
Other researchers have monitored pupil size (pupillometry) in an attempt to quantify cognitive processing load. As cognitive processing demands increase across a variety of tasks (e.g., visual search tasks, mental arithmetic), so does pupil diameter. Pupillometry also has been shown to be sensitive to variations in cognitive processing demands resulting from changes in signal-to-noise ratio during auditory speech recognition tasks (Zekveld, et al., 2011).
Evoked potential measures, such as the P300, and physiologic recordings of heart rate and electromyographic (EMG) activity also have been used to quantify cognitive processing demands/listening effort (Strauss, Corona-Strauss, & Froehlich, 2008; Mackersie & Cones, 2011).
Each of these methods has advantages and disadvantages. They all may provide important information regarding physiologic correlates of perceived effort, although further work is needed to confirm the utility of these measures when examining listening effort in persons with hearing loss.
10. All interesting, but these tests don't really focus on perception. I thought that was our main concern?
It usually is, and that is why there is long history of using behavioral psychophysical methods to assess cognitive processing demands, including listening effort. A common tool from the cognitive psychology literature that is used to quantify changes in cognitive processing demands is the dual task paradigm. In these tasks participants have to complete a primary task (like speech recognition) and a secondary task (like watch for a light or remember a string of numbers) at the same time. To do well on each task requires some allocation of cognitive resources to each task. We then systematically vary the difficulty of the primary task (e.g., word recognition is assessed at multiple noise levels) while monitoring performance on the secondary task. The underlying assumption is that changes in secondary task performance are due to shifts in the allocation of cognitive resources to or from the primary task as the primary task difficulty is varied (e.g., Sarampalis, Kalluri, Edwards, and Hafter, 2009).
11. Are these more objective types of measures consistent with the subjective measures you described?
Actually, many times they are not. In fact, it is common in studies that assess both subjective and objective listening effort to find that the measures are not strongly correlated. For example, Gosselin and Gagne (2011) used a dual task paradigm that combined sentence and tactile recognition and response times, to objectively measure listening effort in younger and older adults with normal hearing. They also had people subjectively rate the effort required to complete all their tasks and found no correlation between the subjective and objective measures of listening effort. Findings like this are not uncommon in the literature and suggest that subjective and objective measures of listening effort may be assessing different things.
12. It certainly makes sense that hearing loss will lead to increased listening effort, so I have to ask the obvious question: "Do hearing aids help?"
Like many situations dealing with hearing loss, the answer is "sometimes". Hearing aids are complex devices and offer many advanced features like directional processing and digital noise reduction (DNR). But let's start with the basic situation comparing unaided and aided listening effort in the absence of any advanced features. In these situations the evidence from subjective measures is very strong and shows that hearings aids can reduce subjective perception of listening effort. In fact, when comparing unaided and aided subjective ratings of listening effort they often mirror speech recognition abilities. That is, in those aided conditions where speech understanding is improved, subjective ratings of "ease of listening" also tend to improve (e.g., Humes et al., 1999a).
13. If it's that straightforward, then why not just measure speech recognition?
Well, although there is a general relationship between some subjective measures of listening effort and measures of speech recognition ability, they don't appear to be measuring the same thing (Humes, et al., 1999). In fact in some conditions (e.g., varying SNRs) speech recognition scores may be very similar yet measures of listening effort could show significant differences. This is similar to what we see when comparing sound quality measures and speech recognition. It is well known that subjective preferences for sound quality don't depend solely on speech recognition.
For example, in a study I did with Todd Ricketts, we found that individuals with hearing loss had a strong subjective preference for DNR processing even though speech recognition did not improve (Ricketts & Hornsby, 2005). In addition, as I mentioned earlier, measures of speech recognition don't give us information about the cognitive effort being expended to obtain a given level of understanding. In fact, some objective measures of listening effort may actually be more sensitive than speech recognition measures for detecting subtle differences between unaided and aided listening conditions (Gatehouse & Gordon, 1990).
14. What do these objective measures tell us about benefits from hearing aids, in terms of reducing listening effort?
Unfortunately, there has been very little work examining unaided and aided listening effort using objective, psychophysical measures and actual hearing aids (and no physiologic work). Results from the few studies are promising but somewhat mixed. For example, Gatehouse and Gordon (1990) used vocal reaction times during speech recognition testing, with and without hearing aids, as a measure of cognitive processing load (listening effort). They found that during speech recognition tasks, as the cognitive processing demands required for understanding increased, vocal response times also increased. Interestingly, they found that even in conditions where unaided and aided speech recognition was similar, vocal response times were faster aided. This is consistent with the suggestion that listening effort is reduced when using hearing aids, even when speech recognition remains the same.
In a more recent study, Hallgren and colleagues (2005) found either minimal (in quiet) or no difference (in noise) on objective measures of listening effort, as measured by a reduction in reaction times during cognitively demanding speech tasks completed both with and without hearing aids. It appears that the way we measure listening effort can impact our findings and the optimal method for objectively measuring listening effort remains unclear. More work in this area clearly is needed.
15. So far you've been talking about basic amplification. What about adding a feature such as digital noise reduction?
Great question, thanks for bringing it up. As you maybe know, this had been the topic of some recent research. Recall that Gatehouse and Gordon (and others) suggested that objective measures of listening effort may be more sensitive than objective measures of speech recognition, at least in some cases. Well, DNR processing is a case where there often are strong subjective preferences for having the DNR active in noise even though speech recognition doesn't improve (or may even get worse)—I already mentioned the work that Todd Ricketts and I did a few years back. The reason for this preference is not entirely clear but some have suggested it could reflect a decrease in listening effort.
16. And this has been documented in research?
To some extent, yes. Sarampalis et al. (2009) looked at this recently in listeners with normal hearing. In one experiment they used a dual task paradigm and in another they used a memory recall task to assess cognitive processing demands when listening with and without DNR processing active. They found that in some adverse SNRs, using DNR processing reduced cognitive processing demands (as shown by faster reaction times and better memory recall). In other conditions, when the SNR was not as adverse, there was no benefit.
Pittman (2011) indirectly assessed the effects of DNR processing on listening effort in children with hearing loss by examining short term word learning with and without DNR active. The underlying assumption was that DNR processing would degrade word learning due to the gain reduction imposed by the processing when active in noise. In contrast to this expectation, she found word learning improved for older children (11-12 years), but not for younger children (8-9 years) with hearing loss when listening with the DNR active. This finding is consistent with a reduction in cognitive processing demands/listening effort due to DNR processing. The results of these studies suggest DNR processing may provide benefits, in addition to listening comfort, even in the absence of improvements in speech recognition. Although more work using actual hearing aids and individuals with hearing loss is needed.
17. It would seem that this extra effort would lead to fatigue. Has that also been studied?
Not very much, however, I think this is a pretty important area. There is a huge body of literature on fatigue, its association with multiple chronic diseases (e.g., cancer, heart disease, HIV) and its negative effect on quality of life. Despite this recurrent link between chronic disease and fatigue, the relationship between hearing loss and fatigue has received little research attention. It's generally assumed that increases in listening effort during lab-based speech recognition studies are associated with subjective complaints such as fatigue, and effort in everyday settings. While intuitive, there is little objective evidence to validate this link.
Like listening effort, fatigue can be measured both subjectively and objectively. Objective methods include both physiologic and psychophysical paradigms. When assessing fatigue objectively it is common to look for changes in performance/physiologic processing over time. Psychophysical vigilance tasks, tasks which require sustained attention and concentration for optimal performance, have been used frequently used to objectively quantify fatigue associated with cognitively demanding situations (e.g., Lieberman, 2007). We have recently begun using these types of tasks to objectively quantify the effects of hearing loss on fatigue resulting from sustained speech processing demands.
Some initial findings, presented at AAS meeting earlier this year, suggest that unaided listening during a demanding speech task can lead to objectively measured fatigue (Hornsby, 2011).
18. You lost me there. How did you objectively measure fatigue?
Okay, I'll be more specific. We used a dual task paradigm to objectively measure listening effort, with and without hearing aids, during a speech task over about an hour period. The task was to listen to strings of words in noise, remember the last five words, and at the same time remain vigilant for a visual cue. When they saw the cue participants pressed a button as quickly as possible. This gave us measures of word recognition, word recall and visual reaction times. We looked at how these different performance measures changed over time. Our hypothesis was that listening effort (as measured by word recall ability and visual reaction time speed) would need to increase over time in order to maintain optimal performance on the speech task.
When listening without hearing aids we did see a systematic increase in visual reaction times as the speech task progressed. Word recall and recognition, however, stayed the same. This suggested that as participants became fatigued they began to shift cognitive resources from the reaction time task to the speech processing tasks. An important finding was that this fatigue was reduced when listening with hearing aids. Given the significant negative consequences of fatigue, while preliminary, these findings are encouraging.
19. Sounds encouraging. Any tips on how I might use this information for my hearing aid fittings?
Well, you probably already fit all your patients with DNR, but you certainly could use some of these data to support your counseling regarding "Why hearing aids work." I think you also might find that some of these data apply to those patients you've worked with who have very mild hearing loss; the ones whose hearing is almost "too good" to amplify. We've all seen these cases where some of these patients love their hearing aids, yet it's difficult for us to document the benefit using traditional speech recognition testing or outcome measures. It could be that for these patients we are reducing listening effort and overall fatigue, and that is important for them. A good thing!
20. So what's your take home message for us clinicians?
Well, for clinicians interested in quantifying listening effort for use as an outcome measure, currently the only option is to assess listening effort subjectively. For example, you could use a validated subjective scale, such as the SSQ. Alternatively, several researchers have shown that simple rating scales (e.g., 0-10 or 0-100 "ease of listening" ratings) can be used to quantify listening effort (Humes, et al., 1999; Hallgren, et al., 2005). Objective measures appropriate for clinical use are, as of yet, unavailable. That said, given the potential sensitivity of both subjective and objective measures of listening effort to advanced signal processing strategies, development of clinical tools may be something on the horizon.
Gatehouse, S., & Gordon, J. (1990). Response times to speech stimuli as measures of benefit from amplification. British Journal of Audiology, 24(1), 63-8.
Gatehouse, S., & Noble, W. (2004). The Speech, Spatial and Qualities of Hearing Scale (SSQ). International Journal of Audiology, 43(2), 85-99.
Gosselin, P. A., & Gagne, J.-P. (2010). Use of a dual-task paradigm to measure listening effort. Canadian Journal of Speech Language Pathology and Audiology, 34(1), 43-51.
Hallgren, M., Larsby, B., Lyxell, B., & Arlinger, S. (2005). Speech understanding in quiet and noise, with and without hearing aids. International Journal of Audiology, 44(10), 574-83.
Hetu, R., Riverin, L., Lalande, N., Getty, L., & St-Cyr, C. (1988). Qualitative analysis of the handicap associated with occupational hearing loss. British Journal of Audiology, 22(4), 251-64.
Hicks, C. B., & Tharpe, A. M. (2002). Listening effort and fatigue in school-age children with and without hearing loss. Journal of Speech, Language, and Hearing Research, 45(3), 573-84.
Hornsby, B. (2011, March). Effect of hearing aid use on mental fatigue. Paper presented at the American Auditory Society annual meeting, Scottsdale, AZ.
Humes, L. E., Christensen, L., Thomas, T., Bess, F. H., Hedley-Williams, A., & Bentler, R. (1999). A comparison of the aided performance and benefit provided by a linear and a two-channel wide dynamic range compression hearing aid. Journal of Speech, Language and Hearing Research, 42(1), 65-79.
Humes, L. E. (1999). Dimensions of hearing aid outcome. Journal of the American Academy of Audiology 10(1), 26-39.
Kramer, S. E., Kapteyn, T. S., & Houtgast, T. (2006). Occupational performance: comparing normally-hearing and hearing-impaired employees using the Amsterdam Checklist for Hearing and Work. International Journal of Audiology, 45(9), 503-12.
Lieberman, H. R. (2007). Cognitive methods for assessing mental energy. Nutritional Neuroscience, 10(5-6), 229-42.
Mackersie, C. L., & Cones, H. (2011). Subjective and psychophysiological indexes of listening effort in a competing-talker task. Journal of the American Academy of Audiology, 22(2),
Nachtegaal, J., Kuik, D. J., Anema, J. R., Goverts, S. T., Festen, J. M., & Kramer, S. E. (2009). Hearing status, need for recovery after work, and psychosocial work characteristics: Results from an internet-based national survey on hearing. International Journal of Audiology, 48(10), 684 - 691.
Noble, W., & Gatehouse, S. (2006). Effects of bilateral versus unilateral hearing aid fitting on abilities measured by the Speech, Spatial, and Qualities of Hearing Scale (SSQ). International Journal of Audiology, 45(3), 172-81.
Picou, E. M., Ricketts, T. A., & Hornsby, B. W. (2011). Visual cues and listening effort: individual variability. Journal of Speech, Language, and Hearing Research, 54(5), 1416-30.
Pittman, A. (2011). Children's performance in complex listening conditions: Effects of hearing loss and digital noise reduction. Journal of Speech, Language, and Hearing Research, 54(4), 1224-39.
Ricketts, T. A., & Hornsby, B. W. (2005). Sound quality measures for speech in noise through a commercial hearing aid implementing digital noise reduction. Journal of the American Academy of Audiology, 16(5), 270-7.
Sarampalis, A., Kalluri, S., Edwards, B., & Hafter, E. (2009). Objective measures of listening effort: Effects of background noise and noise reduction. Journal of Speech, Language, and Hearing Research, 52, 1230-40.
Strauss, D. J., Corona-Strauss, F. I., & Froehlich, M. (2008). Objective estimation of the listening effort: towards a neuropsychological and neurophysical model. Conference Proceedings IEEE Engineering in Medicine and Biology Society, 1777-80.
Zekveld, A. A., Kramer, S. E., & Festen, J. M. (2011). Cognitive load during speech perception in noise: the influence of age, hearing loss, and cognition on the pupil response. Ear and Hearing, 32(4), 498-510.