Hearing loss and dementia - A PROBLEMATIC NARRATIVE?

Right across the spectrum of medical sectors, it is not unusual to find recently suggested statistical correlations – called links or associations – being presented as strong evidence that treatment of one condition will affect another.

It is hard to escape the impression that, at the first genuine sniff of such links, from the C-suites of manufacturers to the newsdesks of the media, the apparatus is set in motion, campaigns are launched, words are massaged, and plausibility solidifies into collective convictions. Bold statements, as often as not voiced by scientists, issue forth, and even if there are initial reminders that studies have found no mechanisms, no proof of causality, the absence of such a fundamental expectation of thought and knowledge of the external world quickly seems to have no significance.

Since before this magazine reported in March 2019 (Audio Infos UK issue 126) on an industry-led Hearing Loss and Dementia round table, there was already a buzz about the potential impact on cognitive decline of treating hearing loss. The buzz is now a very loud noise.

Just this April, the hearing device giant GN became the latest player in audiology to launch a campaign based on taken-as-known links between hearing health and dementia, or between hearing treatment and dementia. Listen to This is a campaign that, through its BrainWorks innovation unit, aims to elevate the importance of hearing health as part of overall wellbeing, in particular cognitive health“, and the launch release quotes George Vradenburg, Chair and Co-founder of UsAgainstAlzheimer’s, as saying”hearing aids themselves are an important means of reducing the risk of cognitive decline“. This is a reverberation working its way round the room, and many of us help it to continue echoing. Do we just trust the research without examining its quality?

Photo by 愚木混株 cdd20 on Unsplash

While GN Group carefully attributes to UsAgainstAlzheimer’s its Linkedin declaration that”hearing aids could reduce the rate of cognitive decline by a remarkable 48%“, the fact is that many members of scientific community have been circumspect about endorsing such claims. 

Some scientists, however, have gone to lengths to point out that not only do studies so far not justify the strength of current claims, but that the dementia message could feasibly backfire on the hearing industry and the profession, wiping out progress made by years of positive messaging on quality of life.

Hearing Loss and Dementia: Where to From Here? was published in the May/June issue of Ear and Hearing, the official journal of the American Auditory Society. Its authors, Kevin Munro (NIHR Senior Investigator and Ewing Professor of Audiology in the Manchester Centre for Audiology and Deafness, University of Manchester, UK), and Piers Dawes (Professor of audiology, Director of the University of Queensland Centre for Hearing Research (CHEAR)) argue that addressing hearing loss in terms of dementia prevention may be”inappropriate on the grounds of both relevance at individual level and lack of clear evidence of benefit“. They urge the audiology community to deliver”a positive message focusing on the known benefits of addressing hearing loss in terms of improved communication, quality of life, and healthy aging“, rather than resort to”scare tactics“.

There are clearly some crucial questions to be answered on the science behind dementia’s links with hearing loss. We are grateful for this interview with Professor Kevin Munro, the only audiologist to be awarded Senior Investigator status by the UK’s National Institute for Health and Care Research (NIHR).

Audiology News UK Question: Since the publication of outcomes from the ACHIEVE trial, agents in the hearing health sphere, particularly within the industry, are more confidently claiming that hearing aid use can prevent, delay, or reduce dementia. Is there actually any solid scientific evidence that hearing treatment helps dementia?

©Bernadette Delaney Photography
Professor Kevin Munro

Response: There is currently very little high-quality evidence supporting the use of hearing loss interventions to prevent, delay or reduce dementia.

Dawes and Völter (2023) reviewed studies investigating the impact of hearing loss interventions on cognitive decline and incident cognitive impairment. The findings were mixed. When investigating cognitive impairment, three cochlear implant studies reported positive findings, which is encouraging, but methodological limitations (e.g., lack of adequate control group for comparisons) precluded any conclusions about cognitive benefit of cochlear implants.

For acoustic hearing aids, four studies reported lower incidence of cognitive impairment in hearing aid users compared to non-users, and five reported no difference between groups. When investigating cognitive decline instead of impairment, four studies reported lower rates of decline in hearing aid users compared to non-users and three reported no difference between groups. An important methodological consideration is that all these studies compared people who chose to use a device or not. This self-selection can bring biases if, for example, those with more difficulties in everyday life are more likely to use hearing devices.

The ACHIEVE trial, to which you refer, is an excellent example of a large, multi-centre randomised controlled trial (Lin et al., 2023). It is a comprehensive, high-quality study, impressive in rigour, scope and potential impact. ACHIEVE has overcome many of the limitations of the observational studies reviewed by Dawes and Völter (2023). The hearing intervention did not reduce 3-year cognitive decline in the total cohort of cognitively healthy older adults with hearing loss (although see below for my comments on a secondary analysis of ACHIEVE data).

PD
Professor Piers Dawes

Question: Is there even strong evidence of a causal impact of hearing loss – not treatment – on cognitive decline?

No, not currently. There is strong evidence for an association (in statistical language, a correlation) between hearing loss and cognitive decline, but there is a well-used adage that”correlation does not imply causation“. It is a logical fallacy that arises from thinking because two things co-occur, one must have caused the other. In this case, the correlation doesn’t automatically mean that hearing loss causes cognitive decline, or that cognitive decline causes hearing loss. There could be something independent that causes them to co-occur. For example, if poor cardiovascular health causes both to occur, perhaps a healthier lifestyle (e.g., more physical activity, healthier diet and quitting smoking) will benefit both hearing and cognition. In fact, correlations can be entirely spurious.

There is a website that automatically generates nonsense associations, such as the correlation between consumption of mozzarella cheese and the number of civil engineering doctorates awarded, or the number of people who drown and the number of films Nicolas Cage has appeared in each year.

Question: While the main finding of the ACHIEVE trial was that a comprehensive hearing intervention did not mitigate decline in older adults, it is mostly the positive results of a secondary analysis by the ACHIEVE team that are fuelling so many claims that hearing intervention can help dementia. This analysis concluded that”a hearing intervention might reduce cognitive change over 3 years in populations of older adults at increased risk for cognitive decline but not in populations at decreased risk for cognitive decline“. The analysis even suggests a reduction in potential cognitive changes by up to 48% in the at-risk group. Your point-of-view article urges caution regarding positive reactions to this secondary analysis. Why?

You are correct that a secondary analysis was reported as showing a reduction in cognitive decline in a subgroup of participants who were allegedly at high risk of cognitive decline. This finding was statistically significant, but despite the hype and media coverage, the effect size was small (Cohen’s d = 0.25), and the real-world significance is unclear. It is also unclear if the family-wise error rate (i.e., a finding falsely identified as significant when multiple comparisons are being tested) was controlled across all the tests, meaning that the statistical result could be a chance finding. We need to be cautious that an unchecked desire for the finding to be true could hinder rational thought and scientific rigour. For example, the benefit of the hearing intervention in reducing cognitive decline would be expected to be greatest in those with most hearing loss, but this was not the case. Also, the well-known correlation between sensory function and cognition (and possible benefits of hearing interventions in improving cognitive outcomes) have been reported in the general population, and not only among sub-groups at high risk of cognitive decline. We elaborate on why this finding should be treated with caution in our point-of-view article (Dawes and Munro, 2024). Of course, time will tell if the finding is robust and of real-world importance.

Question: Those who are enthusiastic about associations between hearing loss and cognitive decline might point out that there are other studies besides ACHIEVE that can be used to support claims of a direct benefit from intervention. In broad terms, why do other studies fail to offer such solid evidence?

Unlike ACHIEVE, most studies have used an inferior design where adults who happen to use hearing devices are compared with adults who don’t happen to wear hearing aids. It is challenging to definitively know if any benefits can be attributed direct to the hearing intervention or if there are differences between the two groups that could explain the finding. In addition, the quality of evidence from observational studies is sometimes low because of”self-inflicted“avoidable biases. For example, if an observational study compares interventions for hearing loss, a self-inflicted bias would be to have the groups unmatched for hearing loss. Observational trials that do not emulate the design of a randomised control trial (RCT) introduce avoidable biases on top of confounding biases.

Question: A research review by The Lancet produced a figure of a”potential“8% reduction in all dementia cases if hearing loss could be eliminated or entirely mitigated. There are ifs involved here. Why is that word”potential“potentially misleading?

Actually, it is important to include”potentially“because it is not known if the association is causal. There are many potentially modifiable risk factors for dementia including hypertension, obesity, diabetes, and hearing loss. The mechanisms that underlie the associations are unknown and are still disputed. What Livingston and colleagues (2020) are saying is if hearing loss is shown to cause dementia, 8% of dementia cases could be avoided if all hearing loss could be entirely eliminated or successfully treated. That’s a big ask – to eliminate all hearing loss or entirely mitigate it. The reality, therefore, is likely to be less than 8%.

Hearing loss is reported to be the highest potentially modifiable risk factor for dementia. However, the largest number of people living with dementia are in low and middle-income countries where the 8% attributable fraction may be different if the number of people exposed to the risk factor is not the same as the UK and similar high-income countries. For example, some low and middle-income countries may have a higher proportion of the population who smoke, and this will increase the percent attributable fraction to smoking (reported as 1.6 in Livingston et al, 2020).

It is also important to point out that the 8% dementia risk at the population level is not the strength of the personal dementia risk to an individual. According to Livingstone et al (2020), the chances of dementia co-occurring in an individual are somewhat similar (relative risk between 1.4 and 1.9) across a range of potentially modifiable factors including those in early life (less education), mid-life (hearing loss, traumatic brain injury, hypertension and obesity) and later life (smoking, depression, social isolation, physical inactivity, and diabetes).

I recently viewed a website designed by professionals to educate the public about their hearing and the homepage stated,”hearing loss is the single greatest risk factor for dementia.“When I asked family and friends (presumably the target audience of the website) what this statement meant to them, the majority said their chance of dementia is higher if they have a hearing loss than any other health condition. The public assumed the statement was about personal risk and not at the population level. My family and friends interpreted hearing loss not just as a risk, but actually causing dementia. This illustrates that the topic is ripe for problematic narratives and discarded nuance.

Question: In your article, you refer to a paper and editorials in The Lancet Public Health journal that were”quietly retracted“when a blunder in the analysis was discovered. Codes for hearing aid users and nonusers had been mixed up, leading to reporting of a positive finding when, in fact, the study referred to really showed that risk of dementia was higher for hearing aid users than non-hearing aid users.”It remains to be seen if the revised findings and interpretation will be reported,“your article comments. Don’t such incidents and, indeed, the readiness of media to disregard lack of evidence and make research froth into fact, point to a more general need for rigour on how scientific information is conveyed to public and professionals?

In this example, the journal editor was happy to hype up the result of the Jiang et al. (2023) study when it looked newsworthy and, understandably, the article received lots of attention. What subsequently transpired was a postdoctoral researcher in a different country closely examined the data and raised concerns that something was awry. Thanks to the tenacity of the researcher, the authors and editor acted, eventually, but it took some time (to see how the story unfolded, see Retraction Watch [2024]). Given the fanfare proclaiming the importance of the study findings, it is concerning that the retraction announcement was relatively subdued because the hype may continue to go unchecked. In any case, the scientific literature should be a balanced and transparent record of all research outcomes, with articles valued according to the importance of the research question and the quality of the science, rather than the newsworthiness of the findings. Perhaps the authors and editors could be encouraged to publish a revised version of the manuscript with equal prominence to that of the original publication, in which the corrected results are reported, and the alternative interpretation is discussed.

Question: Perhaps hiding behind what can easily be passed off
as good intentions, aren’t those who build entire sales campaigns on dubitable evidence guilty of using scare tactics?

I do not encourage negative campaigns where it is implied that addressing dementia is more important than addressing hearing loss. This is hype and a misrepresentation of research. If the publicity given to links between hearing loss and dementia risk and benefits of hearing interventions in preventing dementia is shown to lack a firm foundation, then policymakers, and funders of hearing research and clinical services may lose interest in hearing. There is a strong, convincing, and factually correct, positive message that should be used for campaigns (elaborated in my answer to your next question, below). And in fact, it has been shown from behaviour change research that positive messaging and/or reframing may be more impactful than negative messaging or scare tactics in terms of promoting healthy behaviours.

Question: Your article underlines the seriousness of hearing loss itself as a”major public health challenge“and suggests ample risks – such as prolonging stigma – to the message that addressing dementia is more important than addressing hearing loss. How might such an emphasis backfire for the industry, the profession, and the patient?

The fact that we are discussing this issue means it has already detracted from what is important. Hearing loss ranks third for Years Lived with Disability, first for sensory disorders, and first for those over age 70 (GBD 2019 Hearing Loss Collaborators, 2021). Addressing hearing loss is an important component of healthy ageing. Those who enter older age in good health have a very different experience to people with multiple health issues. Improving and maintaining the health (and hearing health) of older adults is a strong social responsibility. This is the message I would like to see amplified and shared.

Question: You point out that there is a significant degree of complexity to possible causal relationships between cognition and sensory function, so what does this mean for type, timelines, and financing of studies from this point onwards? Put another way, with large, costly, and long-running RCTs needed, what chance is there of strong evidence of causality being delivered in the next ten years?

Although observational studies are not the magic bullet (because they will always have unmeasured or residual confounds), their quality can be improved using a framework called “Target Trial Emulation” (Fu, 2023). This is where one should plan what the protocol for a target RCT should look like (in terms of design and analysis) and emulate this in the design and analysis of an observational study. Key components include aligning: (i) eligibility criteria in both arms of the trial, (ii) treatment strategies, and (iii) follow up (all of which would automatically occur at the time of randomisation in an RCT). The statistical analysis plan should also specify the methods that will be used to adjust for confounds, how missing data will be dealt with and the methods to be used to estimate effect sizes. Also, the design and analysis can be benchmarked against the results of a good RCT. If the findings from the RCT can be accurately replicated in the observational data, more confidence is gained that the design and analysis were adequate to adjust for confounds. Observational studies that are of sufficiently high quality can supplement findings from ACHIEVE and other RCTs (especially since some RCT designs are ethically problematic because they involve withholding treatment).

Question: Until further studies bring solid evidence of causal impact, might audiology and patients benefit more if practitioners catch their breath and just maintain a dementia-friendly focus on the positive effects of hearing treatment on quality of life and improved ageing, rather than giving nebulous guarantees about individually unmeasurable percentages of immunity from dementia?

As we have said previously, encouraging policymakers, health professionals and patients to address hearing loss in terms of dementia prevention may be inappropriate on the grounds of relevance to the individual as well as a lack of clear evidence of benefit (Dawes and Munro, 2024). Hearing loss and its relevance to healthy ageing is important it its own right. A socially responsible approach involves a positive focus on the known benefits of addressing hearing loss in terms of improved communication, well-being and quality of life, and facilitating healthy ageing.

The authors declare no conflict of interest. Both authors are supported by the NIHR Manchester Biomedical Research Centre.

—————————————————————————————————————

Refrain from saying hearing aids prevent dementia/ cognitive decline

A viewpoint article from the August 2023 issue of the American Journal of Audiology¹ by Sonova-related audiology experts (Maren Stropahl, Sigrid Scherpiet, and Stefan Launer) reflected concerns about the impetus of the dementia narrative in hearing care, urging those in hearing care to “refrain from saying ‘hearing aids prevent dementia/ cognitive decline’ or ‘hearing aids are an intervention for dementia/cognitive decline’”.

The article concludes that audiologists should ”adapt the narrative to focus on the relationship between hearing and the brain“and, additionally,” focus on how hearing aids can facilitate a socially active life and enable well-being and healthy living“.

”Avoid negative framing of the message around the association between hearing loss and cognitive decline, that is, do not use a narrative that poses a threat to elderly people with hearing loss,“ the authors propose.

Confusingly, however, this same article argues that there is powerful evidence that studies do show strong evidence of the link. Despite the published results of the ACHIEVE trial stating that hearing intervention ”did not reduce 3-year cognitive decline“, the AJA article claims that ACHIEVE, and another longitudinal study, ENHANCE, “provide strong evidence on top of the published evidence stemming from retrospective observational studies. The results from ACHIEVE and ENHANCE indicate that comprehensive hearing care, ‘best practice audiology,’ including good counseling, diagnostics, and well-controlled and assessed fitting of hearing aids can mitigate the risk of cognitive decline over a 3- year period in older adults with hearing impairment”.

^{1 https://doi.org/10.1044/2024_AJA-23-00176}

References

Dawes P, Munro KJ. (2024). Hearing loss and dementia: where to from here? Ear and Hearing, 45, 529-36, doi: 10.1097/AUD.0000000000001494

Dawes P, Völter C. (2023). Do hearing loss interventions prevent dementia? Zeitschrift für Gerontologie und Geriatrie, 1-7. doi: https://doi.org/10.1007/s00391-023-02178-z

Fu EL. (2023). Target trial emulation to improve causal inference from observational data: what, why and how? Journal of the American Society of Nephrology 34, 1305-14. Doi: 10.1681/ASN.0000000000000152

GBD 2019 Hearing Loss Collaborators. (2021). Hearing loss prevalence and years lived with disability, 1990-2019: findings from the Global Burden of Disease Study 2019. Lancet, 397, 996-1009. doi:10.1016/s0140-6736(21)00516-x

Jiang F, Mishra SR, Shrestha N, Ozaki A, Virani SS, Bright T, . . . Zhu D. (2023). Association between hearing aid use and all-cause and cause-specific dementia: an analysis of the UK Biobank cohort. The Lancet Public Health, 8(5), e329-e338. doi:https://doi.org/10.1016/S2468-2667(23)00048-8

Lin FR, Pike JR, Albert MS, Arnold M, Burgard S, Chisolm T, . . . Glynn NW. (2023). Hearing intervention versus health education control to reduce cognitive decline in older adults with hearing loss in the USA (ACHIEVE): a multicentre, randomised controlled trial. The Lancet, 402, 786-97. https://doi.org/10.1016/S0140-6736(23)01406-X

Livingston G, Huntley J, Sommerlad A, Ames D, Ballard C, Banerjee S, . . . Cooper C. (2020). Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. The Lancet, 396, 413-46. doi:https://doi.org/10.1016/S0140-6736(20)30367-6

Retraction Watch 2024 (https://retractionwatch.com/2024/01/04/we-should-have-followed-up-lancet-journal-retracts-article-on-hearing-aids-and-dementia-after-prodding/