From the Desk of Gus Mueller
Certainly one of the more intriguing topics in audiology over the past decade has been the study of the potential relationship between hearing loss and dementia. You hear a lot of different terms tossed out regarding this issue. Is it simply a relationship, or is it a link? An association, a correlation, or is there direct casual causation? And, could the use of hearing aids alter the course? We had a preliminary discussion on this topic here at 20Q a few years ago when Piers Dawes, PhD, reviewed some of the germinal research (you can read that article here). But more data are emerging each year, and it’s time to revisit the topic.
Research suggesting that hearing impairment in older adults is strongly associated with having dementia dates back to the 1980s, but little work was conducted in the area until around 2010, when we started to see reports from Frank Lin, MD, and his colleagues at Johns Hopkins, based on data from the Baltimore Longitudinal Study of Aging. Since that time, research from this group has set the pace of study in this interesting area. It is no coincidence, therefore, that our 20Q guest this month comes from this research team.
Nicholas Reed, AuD, is an Assistant Professor of audiology at Johns Hopkins School of Medicine. He is core faculty at the Cochlear Center for Hearing and Public Health at the Johns Hopkins Bloomberg School of Public Health. His research focuses on novel hearing care delivery models, over-the-counter amplification devices, the relationship between hearing loss and patient-provider communication, and the relationship between hearing loss and health care utilization patterns.
While Dr. Reed is just getting started in the world of research, his list of refereed articles reads more like someone who has been in the field for decades. His research is funded by an NIH-supported Kl2 award, and work from his research has been featured in the Journal of the American Medical Association (JAMA) and the Journal of the American Geriatrics Society (JAGS). His research spans many related areas, which is why he is coming back next month to address some other important areas of hearing health care.
Every now and then, we have a 20Q author who has a good story regarding how they are now working not too far from where they grew up (think Ruth Bentler). But Nick certainly wins the grand prize in this area. Not only did he grow up in Baltimore, obtain his AuD just outside of Baltimore at Towson State, but he spent his youth working at his grandparents’ fish market at Fell’s Point, less than a mile from the Johns Hopkins Hospital.
Gus Mueller, PhD
Browse the complete collection of 20Q with Gus Mueller CEU articles at www.audiologyonline.com/20Q
20Q: Hearing Loss and Dementia - Highlights from Key Research
After this course, readers will be able to:
- Explain how hearing loss in older adults is a public health issue, and list research topics in this area that have been investigated.
- Describe the difference between cognitive decline and dementia.
- Explain the association between hearing loss, hearing aid use, and cognitive decline/dementia based on the research to date, and how best to discuss this topic with patients in clinical practice.
1. I see that you are at the Cochlear Center at Johns Hopkins. What does your team study?
The name of our center might be a little misleading. The reason it is called the “Cochlear Center,” is because it was made possible by a generous contribution from Cochlear Corporation. However, our research spans well beyond the cochlea. In fact, we don’t actually study cochlear implantation at all. The Cochlear Center for Hearing and Public Health is unique in that it is the only center, to our knowledge, housed in a school of public health with a focus on hearing loss among older adults. Our Center is led by Frank Lin, MD, PhD, and includes faculty from audiology, otolaryngology, epidemiology, health services and policy, biostatistics, and psychology. We study how hearing impacts aging (e.g., cognitive decline, health resource utilization, dementia, physical decline, social isolation, etc.), whether hearing care intervention can mitigate the impact of hearing loss on these outcomes, and effective public health methods and solutions to address hearing loss. Our mission is to effectively train clinicians and researchers regarding the study of hearing loss in older adults from a public health perspective.
2. Why such a focus on aging?
I think many have always known that hearing loss was highly prevalent among older adults. However, the numbers are extraordinary. Using nationally representative data, approximately half of all adults over the age of 60 have a hearing loss based on a 4-frequency pure-tone average ( >25 dB HL in the better ear). This is roughly 38 million Americans. Given the aging demographics of the United States, this number will nearly double to 74 million by 2060.
Public health research has long focused on infectious and chronic diseases. We have seen serious success including vaccinations leading to eradication of some diseases, and success tackling smoking after research independently associating smoking and lung cancer was reported. Now, due in part to the high number of older adults in society, many in public health are focusing on healthy aging. Hearing loss plays a key role in healthy aging as it defines an aspect of how humans interact with the world around them.
3. What kind of research has come out of your Center?
Early on, Frank Lin, MD PhD, and Jennifer Deal, PhD, led multiple studies on the association of hearing loss and cognitive decline and dementia. These studies used large datasets and rigorous epidemiologic methodology to describe an independent association between hearing loss and cognition. Their findings laid the foundation for public health research focused on hearing loss. More recently, the interests of the faculty have gone in many directions within public health. Amber Willink, PhD, has examined access to hearing care and whether hearing care access modifies the relationship between hearing loss and health care spending. Carrie Nieman, MD MPH, has focused on projects to better understand access to hearing care among disadvantaged populations and how community-based programs can address hearing loss. Adele Goman, PhD, has led work describing the prevalence of hearing loss in the United States and the association of hearing loss and depression. Joshua Betz, MS has led work using advanced statistics to improve hearing loss identification from self-report questions. I have focused on how hearing loss impacts health care utilization and how persons with hearing loss interact with the health care system. This just skims the surface.
4. Why start with cognitive decline and dementia?
I already mentioned that the demographics of the United States is changing, such that the number of older adults is increasing. Unfortunately, with age comes a higher risk for cognitive decline and dementia. When focusing on healthy aging, cognitive outcomes become quite powerful on a personal level, as they place a significant burden on caregivers and relationships. I think many people can relate to an experience with a loved one with dementia. On a macro level, cognitive decline and dementia are also associated with quality of life and health care spending. I think the combination of increasing prevalence and the personal and health care implications of cognitive decline and dementia made it a strong first choice when Dr. Lin was initially beginning this work.
5. What is the difference between cognitive decline and dementia?
This is such a relevant question for this research area, and there is a difference. Cognitive decline represents a change in thinking and memory abilities in one or more cognitive domains (e.g., processing speech, executive function, working memory, language, complex attention, etc.). On the other hand, dementia (also referred to as ‘major neurocognitive disorder’ in the DSM 5) is characterized by major decline in one or more cognitive domains and the cognitive deficits are enough to interfere with independence in daily activities. The criterion of interference with independence is what separates a mild cognitive impairment from dementia. Notably, the DSM 5 specifies that the cognitive deficits associated with dementia shouldn’t be attributable to other mental disorder. A note for those of you not working in this area - DSM 5 is Diagnostic and Statistical Manual of Mental Disorders, 5th Edition published by the American Psychiatric Association. It is widely considered the go-to handbook for clinicians and health care professionals in regards to mental health and disorders.
6. So, the million-dollar question - does hearing loss cause cognitive decline and/or dementia?
The very word ‘cause’ can be a loaded term and can be easily misinterpreted. I think it is important to clarify that we are discussing whether hearing loss increases risk for dementia and the question of causality should be interpreted as whether hearing loss is directly contributing to the increased risk of dementia.
Nonetheless, causality can be difficult to prove. Many of you are familiar with the phrase ‘correlation is not causation.’ This is especially important to remember in science, as associations must be grounded in a methodologically rigorous approach. In epidemiology, we might use some of the criteria for a causal framework approach proposed by Austin Bradford Hill in 1965 (generally known as Hill’s Criteria). For example, in this approach we would consider the strength of the association, consistency of findings in the literature, temporality of the relationship (i.e., the exposure should come before the outcome), biologic gradient (i.e., dose-response relationships), and the biologic plausibility (i.e., plausible mechanisms linking the exposure and outcome must exist). Given the explosion of literature in this area, I think the association between hearing loss and dementia increasingly meets these criteria, albeit to different degrees, and I believe it is becoming safer and safer to say that hearing loss increases the risk for dementia.
7. That's a lot to take in. You mentioned plausibility and whether mechanisms exist to link hearing loss and dementia. Can you elaborate?
Sure. Three proposed mechanistic pathways linking hearing loss and cognitive function are cognitive load, structural changes to the brain, and social isolation. In very simplistic terms, the peripheral auditory system encodes a signal and sends it to the brain for processing. When that signal is fragmented, then it may require the brain to recruit other brain areas and divert more energy to decode the relatively poorer signal. This extra load on the brain may come at the expense of another area, perhaps working memory, or limit how well the brain can buffer against other pathologic changes that contribute to dementia (e.g., from microvascular disease or Alzheimer’s disease). Moreover, longitudinal MRI studies suggest hearing loss may also be associated with increased rates of brain volume loss, particularly in the lateral temporal lobe, suggesting that hearing impairment may actually also contribute to increased brain aging. Lastly, and perhaps most intuitive to audiologists, hearing loss is associated with social isolation. Importantly, social isolation has long been known as a strong predictor of dementia and cognitive decline.
8. You mentioned that a consideration in your research was that exposure should come before the outcome?
Yes, in epidemiology research this concept is referred to as “temporality.” One example of temporal evidence for cognitive decline and hearing loss is the Lin et al. (2013) study. This study looked at cognitive decline among 1,984 older adults (mean age = 77 years) from The Health, Aging and Body Composition Study over an 11-year period using the modified mini-mental state exam and the digit symbol substitution test. Cognitive measures were performed at years 1, 5, 8, 10, and 11. Pure-tone audiometry was performed at year 5 in the study and treated as a time-fixed variable. In the study, adults with hearing loss had a 41% and 32% greater annual rate of cognitive decline on the modified mini-mental exam and digit symbol substitution test, respectively. While year 1 measurement of hearing would be optimal, it’s important to consider the insidious and slow nature of peripheral hearing loss. It is highly unlikely the hearing loss showed up overnight in year 5. It is likely it was present in most adults prior to year 1. More importantly, sensitivity analyses suggest the outcome doesn’t change if the data prior to year 5 are excluded.
9. Are there similar findings in a study of dementia and hearing loss?
Absolutely. Using data from the Baltimore Longitudinal Study on Aging, Dr. Lin’s team conducted a time-to-event analysis followed 639 adults (age range 36 to 90 years) over a median timeframe of 12 years. At baseline, the adults completed audiometric measures and did not have dementia. Over time, persons with hearing loss had higher risk of developing dementia. Likewise, Jennifer Deal from our lab led a study using the Health, Aging and Body Composition Study data that examined 2034 adults (age range 70-79 years) over a 9-year period. In the study, moderate hearing loss was associated with an increased risk of incident dementia compared to those without hearing loss (hazard ratio: 1.55, 95% confidence interval: 1.10, 2.19). Lastly, Gallacher et al. (2012) reported a similar association between hearing loss and incident dementia (odds ratio: 2.67, 95% confidence interval: 1.38, 5.18) in 1,057 men over a 17-year period. Notably, these three studies were included in the Lancet Commission on dementia (Livingston et al., 2017).
10. I don’t know what hazard ratio means. You state numbers like 1.55. Is that good or bad?
Good point. I should clarify that the hazard ratio, like the odds ratio and risk ratio, is neither good or bad - it’s a measure of association. The hazard ratio reports the effect of an exposure (e.g., hearing loss) on an outcome (e.g., dementia) over time. In this case, the hazard ratio represents risk of dementia in the hearing loss (e.g., exposed) group divided by risk of dementia in the non-hearing loss (e.g., unexposed) group. Using the moderate hearing loss and dementia as an example from the Deal et al. study, we would interpret a hazard ratio of 1.00 as equal risk of dementia in those with moderate hearing loss compared to those without hearing loss over a given time period. A hazard ratio of 1.55 would be interpreted as those with moderate hearing loss had 55% higher risk of dementia compared to those without hearing loss over the follow-up period. A hazard ratio of 0.55 would mean those with moderate hearing loss had 45% less risk of dementia compared to those without hearing loss over the follow-up period.
Similar to the hazard ratio, odds ratios and risk ratios (also referred to as relative risk) are measures of association. It becomes tricky in that a risk ratio is often interpreted similarly. For example, a risk ratio of 1.87 would mean those with the exposure had 87% higher risk of the outcome compared to those without the exposure. What is important to note is that while odds ratios and risk ratios come from cumulative measures, the hazard ratio reports risk over time (generally from time-to-event or survival analysis) with multiple event measurements. Thus, a risk ratio doesn’t care about the timing of the outcome, just whether or not it occurred by the end of a study period while hazard ratio takes into account the total number of events and the timing of events. The hazard ratio reflects the difference in risk between the exposure groups at any point over the time period.
11. Thanks, that helps. Back to the studies - do other criteria support a causal link between hearing loss and cognitive decline and/or dementia?
We also have to consider something referred to as gradient or dose-response. Is there a monotonic increase of risk with increases of dose? Dose-response is important in our research as it makes sense that more of the exposure (i.e., more severe hearing loss) would be associated with higher risk of the outcome (i.e., dementia or cognitive decline). We do see this in the literature, in the Lin et al. (2013) study described above those with mild, moderate, and severe hearing loss had a hazard ratio for incident dementia of 1.89, 3.00, and 4.94 when compared to those with normal hearing. Similarly, in the Deal at al. (2016) study, those with mild hearing loss did not have a higher risk of incident dementia compared with normal hearing individuals while those with moderate hearing loss did have higher risk. In these studies, we see both a temporal and gradient relationship.
12. One of your criteria is that the literature on a given topic must be consistent. Is it?
I believe the literature is increasingly coalescing and demonstrating repeatable findings. Examples of this are the recent Lancet Commission on dementia which conducted a meta-analysis of hearing loss and dementia among three studies and reported a pooled risk ratio of 1.94 (confidence interval 1.10, 2.19). Importantly, the Lancet Commission found hearing loss accounted for 9% of attributable risk of dementia (the largest of any modifiable risk factor). This was the first time hearing loss had been featured prominently in a comprehensive report on dementia. Another example of some consensus is the Loughrey et al. (2017) systematic review and meta-analysis of hearing loss and cognitive impairment. These researchers reported significant associations between hearing loss and cognitive impairment in both cross-sectional and longitudinal studies.
Nonetheless, I think there will always be variability in the literature. This could be caused by different statistical methodology, different methods of measurement (i.e., pure-tone audiometry versus self-report), and/or selecting different populations to study that lack external validity and generalizability (i.e., a study of all PhD-holding individuals would not be useful for making inference to the entire United States population).
13. That’s quite convincing. Do you have any specific issues with the literature?
I think there are always issues with the body of literature and that no theory is bulletproof. To be honest, that’s what makes science fun. One area to consider is reverse causality, which in this case infers that cognitive decline increases risk of peripheral hearing loss.
The thought process here is that hearing tests are cognitively demanding and therefore an individual with cognitive decline would have poorer ability to complete a hearing test. Moreover, perhaps someone with dementia can’t complete an auditory test and therefore may present poorer hearing measures artificially. I think this does make sense when considering some auditory processing tasks like binaural integration. This also makes sense if hearing is being measured using self-report.
However, within the literature focusing on pure-tone audiometry and cognitive decline, I think it’s less likely. Firstly, although pure-tone audiometry could be fatiguing, it is difficult to imagine cognitive decline that has an effect on an upstream encoding task such as peripheral measurement of audiometry when one considers what the cognitive measures used really represent. For example, the digit symbol substitution test used in the Lin et al. (2013) study is not presenting a clinically meaningful cut-point but rather a scalar value among adults which may not represent a meaningful change in cognition. This means that while persons with hearing loss perform poorer on this cognitively demanding task, their performance on that task doesn’t represent an actual meaningful cognitive change that would affect a task like pure-tone audiometry. Second, and more important, longitudinal measures seem to suggest peripheral hearing loss predates cognitive decline. Likewise, in the dementia literature presented, hearing loss predates dementia which makes reverse causality unlikely.
14. It would seem that the specific cognitive tests used in the different studies are important?
This is an excellent point. It’s really important to consider when reading research studies how the exposure and outcome were measured. When an individual has hearing loss, a cognitive task that requires a lot of orally communicated material may not represent a cognitive issue so much as it represents the inability of a person with hearing loss to access the task information. It is important to look for cognitive tasks that do not rely on auditory input to complete the task, such as the aforementioned digit symbol substitution test.
15. I see. This is all fascinating. Now, if we have a possible causal association between hearing loss and dementia, it would make sense that hearing aids would help, right?
As an audiologist, I like to believe the answer is yes, but unfortunately, we don’t have the data to support this yet. It is plausible that hearing aids provide a better signal to reduce cognitive load, provide stimulation to the brain to prevent volume changes, and may reduce social isolation. However, one of the issues with the secondary data we've discussed is that many of the same variables associated with hearing aid use are also protective of cognitive decline and dementia. That is to say that socioeconomic variables such as income/wealth and education level are associated with higher odds of hearing aid ownership and lower risk of cognitive decline and dementia. This confounder makes it difficult to make conclusions based on the data.
16. Does the data available to us suggest anything?
Actually, I think a really interesting study was published recently that overcame some of the limitations of the previous work. Maharani et al. (2018) used a within-subject design and looked at the rate of cognitive decline among persons before and after they obtained hearing aids in the Health and Retirement Study. They found that the rate of cognitive decline (i.e., slope of decline) decreased following hearing aid use. Because these were the same individuals, many of the socioeconomic factors that could confound the relationship are truly controlled for in that they are, in theory, the same in the pre and post hearing aid timeframes (i.e., the education level of the individual doesn’t change).
17. If the data tells us that, can’t we just call it “case closed?”
It is still secondary data and could include selection bias, residual confounding and unobserved bias given that the study wasn’t designed to study that specific intervention. For example, perhaps this is just the outcome among individuals of higher socioeconomic status who can afford hearing aids or among individuals with higher health literacy who were more easily able to navigate the health care system to access hearing care. This limits the generalizability. We need well-designed, prospective randomized control trials which represent the gold-standard of understanding the impact of intervention.
I'm hopeful that more definitive research will emerge. In the United States, the Aging and Cognitive Health Evaluation in Elders (ACHIEVE) is underway. Its aim is to study the impact of best-practice hearing care versus a healthy aging education program (the 10 KeysTM, which includes one-on-one education sessions about topics relevant to healthy aging, such as diet and exercise, with a nurse) on change in function over a 3-year period. The study includes several secondary outcomes such as development of dementia, brain MRI scans, physical activity, social isolation, etc. The study will recruit 850 adults with mild to moderate hearing loss from four sites throughout the country (Hagerstown, MD, Jackson, MS, Minneapolis, MN, and Winston-Salem, NC) and randomize them to one of these interventions. However, the study is still approximately four years from disseminating these results.
18. That’s a long wait. Anything happening in the meantime in this area?
Agreed, it is a long time to wait. In the meantime, I would point to the work of Sara Mamo, AuD, PhD. Dr. Mamo is delivering basic hearing care to persons with dementia in outpatient clinical settings. In her pilot study, she found that caregivers perceived hearing care as beneficial (Mamo et al., 2017). However, more importantly, she also found that participants with higher symptom burden on depression and neuropsychiatric scales showed improvement on these measures 1-month post intervention.
19. Are you suggesting we treat hearing loss among persons with dementia?
I don’t think the research is settled but I am pretty bullish on Sara Mamo’s work. Hearing care is fairly unique in that it represents a low-risk, non-pharmacologic intervention with high-reward potential. I like to consider that addressing hearing loss may improve communication such that it reduces some of the behavioral outcomes (e.g., agitation, confusion) among persons with dementia. This doesn’t mean it cures dementia or anything like that, but rather, it removes hearing loss from the compounded burden of dementia and hearing loss. Audiology as a field may stand to offer a powerful strategy to improve dementia care.
I think it is important that audiologists really understand how this research translates to the clinical setting. It is important that audiologists are able to address their patients’ concerns and questions about this research. It is okay to confirm that hearing loss may increase risk for dementia, but it is really important that this isn’t presented to patients as a definitive 1:1 relationship such that any hearing loss definitely means that individual will develop dementia. That is fundamentally false. These are many other factors at play. Consider research that shows sugary soft drinks are associated with type-2 diabetes. It doesn’t mean having a sugary soft drink with lunch will definitively cause type-2 diabetes in an individual. Perhaps that individual also goes to the gym regularly which is protective from type-2 diabetes. Rather, the correct interpretation may be that the soft drink raised overall risk, but the risk is still relatively low considering other factors.
In addition, audiologists play an important role in educating the public about hearing loss and hearing care. It is vital audiologists understand that there is no definitive research showing hearing aids delay cognitive decline or dementia at this time. Therefore, it is inappropriate to tell a patient that hearing aids could do such. I think thoughtfully explaining what we know about hearing loss and cognitive decline and dementia and how hearing aids could help is okay, but it is important to stress it isn’t a known quantity.
20. That’s a lot to think about. Any final words?
We’ve been talking about hearing loss as it relates to cognition and dementia. However, hearing loss fits into healthy aging in numerous ways. Another important area of study is how hearing loss impacts health care outcomes and interactions with the health care system. This focuses on studying how hearing loss relates to areas such as health care expenses, health care satisfaction, and patient-provider communication. And of course, it is important to understand the impact of using hearing aids on these issues, from a cost-benefit analysis standpoint, relevant in value-based reimbursement models used by the Centers for Medicare and Medicaid Services. All great areas for us to discuss in next month's 20Q!
Deal, J.A., Betz, J., Yaffe, K., Harris, T., Purchase-Helzner, E., Satterfield, S.,...& Health ABC Study Group. (2016). Hearing impairment and incident dementia and cognitive decline in older adults: the health ABC study. Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences, 72(5), 703-709.
Gallacher, J., Ilubaera, V., Ben-Shlomo, Y., Bayer, A., Fish, M., Babisch, W., & Elwood, P. (2012). Auditory threshold, phonologic demand, and incident dementia. Neurology, 79(15), 1583-1590.
Lin, F.R., Yaffe, K., Xia, J., Xue, Q.L., Harris, T.B., Purchase-Helzner, E.,... & Health ABC Study Group, F. (2013). Hearing loss and cognitive decline in older adults. JAMA Internal Medicine, 173(4), 293-299.
Lin, F.R., Metter, E.J., O’Brien, R.J., Resnick, S.M., Zonderman, A.B., & Ferrucci, L. (2011). Hearing loss and incident dementia. Archives of Neurology, 68(2), 214-220.
Livingston, G., Sommerlad, A., Orgeta, V., Costafreda, S.G., Huntley, J., Ames, D.,... & Cooper, C. (2017). Dementia prevention, intervention, and care. The Lancet, 390(10113), 2673-2734.
Loughrey, D.G., Kelly, M.E., Kelley, G.A., Brennan, S., & Lawlor, B. A. (2018). Association of age-related hearing loss with cognitive function, cognitive impairment, and dementia: a systematic review and meta-analysis. JAMA Otolaryngology–Head & Neck Surgery, 144(2), 115-126.
Maharani, A., Dawes, P., Nazroo, J., Tampubolon, G., Pendleton, N., SENSE‐Cog WP1 group,... & Constantinidou, F. (2018). Longitudinal relationship between hearing aid use and cognitive function in older Americans. Journal of the American Geriatrics Society, 66(6), 1130-1136.
Mamo, S.K., Nirmalasari, O., Nieman, C.L., McNabney, M.K., Simpson, A., Oh, E.S., & Lin, F.R. (2017). Hearing care intervention for persons with dementia: A pilot study. The American Journal of Geriatric Psychiatry, 25(1), 91-101.
Reed, N. (2019). 20Q: Hearing loss and dementia - highlights from key research. AudiologyOnline, Article 25355. Retrieved from www.audiologyonline.com