Dementia: hearing loss, a large contributor

 In people with hearing loss and high risk of developing dementia, hearing aids reduced cognitive decline in the ACHIEVE study (see  dementia hearing loss Lancet2023 in dropbox, or doi.org/10.1016/ S0140-6736(23)01406-X) 

 

Details: 

-- 1716 participants were recruited from the Atherosclerosis Risk in Communities (ARIC) study, which included a random sample of four community-based sites in the US (Forsyth County North Carolina, Jackson Mississippi, Minneapolis suburbs, and Washington County Maryland). there was also a de novo set of healthy community volunteers recruited in order to match the advancing age of the initial ARIC participants; these de novo participants were 70-84 years old and had untreated hearing loss but without substantial cognitive impairment, from 2017-2019

    -- the ARIC study is an ongoing longitudinal study of adults who were aged 45-64 at study onset, to help understand the risk factors for heart disease and stroke and its connection between cardiovascular and cognitive health 

    -- the de novo individuals were recruited through advertisements in local newspapers, radio, and the Internet.

-- all participants were randomly assigned and stratified by severity of hearing loss (pure tone audiogram <40db vs >40db)

-- 238 participants (32.5%) were from the ARIC study, 739 (76%) were in the de novo group

-- mean age 77, 54% female, 88% white/11% Black, education level 53% college or more/43% high school 

-- APOE e4 alleles in 25%, diabetes 20%, hypertension 67%, living alone 30%, income level <$25K in 15%/$25-50K in 30%/$50K-75K in 15%/>$100K in 18% 

-- average pure tone loss: 39.4 decibels

    -- ARIC group was older (79.2 versus 75.7 years), more likely to be female (62% versus 51%), had more risk factors for cognitive decline (eg, lower education, more cardiovascular risk factors, more likely to be living alone), and had lower baseline cognitive scores (especially global cognition, executive function, language, baseline memory, and lower use of hearing aids per day) and slightly lower baseline MMSE scores

-- participants were randomly assigned to a hearing intervention consisting of four one-hour sessions every 1-3 weeks with a study audiologist who provided bilateral hearing aids, fitted to prescriptive targets from real-ear measures, along with individualized sessions with instruction on device use, vs a control intervention of health education (four sessions with a health educator every 1-3 weeks, with a standardized session of didactic education tailored to each participant as well as activities, goal setting, optional extracurricular enrichment activities, and a 5-10 minute upper body extremity stretching program). All participants in each group returned for booster sessions every six months as well as 6-monthly follow-ups

 

-- Primary endpoint: three-year change in global cognition, determined by a comprehensive neurocognitive battery, at baseline and annually 

    -- this neurocognitive battery included word recall, digit symbol substitution, incidental learning, trail making, logical memory, digit span backwards, Boston naming (this is actually a validated test for word retrieval, not simply naming my fair city from a picture), word fluency, and animal naming 

    --  MMSE (Mini-Mental State Examination) was also administered at baseline and every six months

-- Secondary outcomes: three-year change in cognitive domain-specific latent factor scores (executive function, language, and memory); time until cognitive impairment occurred, as defined by a composite outcome of dementia and mild cognitive impairment from in-person assessments; and a three-point reduction in MMSE  (only baseline scores of MMSE were reported in this report)

-- prespecified sensitivity analysis: difference between outcomes in ARIC versus the de novo group; with exploratory prespecified outcomes of social, physical, and mental health at baseline and annually, and brain MRI scans at baseline in year three in one half of the sample of the cohort (the exploratory outcomes were not reported in this article)

 

 Results: 

-- three-year cognitive change (in standard deviation units): 

    -- hearing intervention group: -0.200 (-0.256 to -0.144) 

    -- health education group: -0.202 (-0.258 to -0.145) 

        -- no statistically significant difference 

 

-- But, significant differences in those in the ARIC group vs the de novo group in this secondary outcome. none of the de novo group assessments were statistically significant. assessing the difference between hearing intervention vs controls in ARIC:

    -- global cognition: 48% reduction in three-year cognitive change, which compared to controls led to a difference of 0.191 standard deviation units (0.022-0.36), p=0.027 

    -- language: 66% reduction, difference of 0.229 (0.050 to 0.408), p=0.12

    -- memory: almost significant 73% reduction, difference of 0.201 (-0.019 to 0.421), p=0.073

 

-- the differences between these two groups were greater in the per-protocol (i.e. actually using the hearing aids) versus the intention-to treat analyses  

 

-- here is the graph of many cognitive outcomes, noting the progressive differences between hearing aids and controls over time, but only in the ARIC group:

 

 

 

-- other prespecified sensitivity analyses: no difference 

-- no significant adverse events were attributed to the study in either group 

 

Commentary: 

-- as we all know, dementia is a major health and quality of life issue for an increasing number of people, as people are living to be older than times past 

-- it is predicted that more than 150 million individuals will be living with dementia by 2050, with most living in low-income and middle-income countries 

-- hence the importance of investigating and addressing potentially modifiable risk factors 

-- the Global Burden of Disease Study 2019 found hearing loss in 65% of adults older than 60yo globally, with the suggestion that hearing loss is the single largest potentially modifiable risk factor for dementia in both high income and low-to-middle income countries.

-- Interventions for hearing loss (eg, hearing aids) are dramatically underused: <10% of those in low-income countries, and <20-30% in high-income countries in those with hearing loss

-- Prior studies finding benefit of hearing aids have been observational, noting a decrease in subsequent cognitive decline and dementia 

    -- a pooled meta-analysis of 126,903 participants in 8 observational studies with follow-up from 2-25 years found a lower risk of cognitive decline in hearing aid users compared with those with untreated hearing loss 

 

-- the current study did find that there was a significant decrease in cognitive decline over a three-year period in older adults at increased risk for cognitive decline, but not so for populations at decreased risk: there was a 48% reduction in the three-year global cognitive decline found in the ARIC cohort

    -- however, it may well be that the battery of cognitive tests may miss subtle early decreases in cognitive decline in those not at increased risk 

    -- cognitive decline might well be decreased in those at lower risk if these persons were evaluated for a longer period of time, providing more time for larger numbers of people to have more significant, measurable cognitive decline (as well as having more high-risk cardiovascular/diabetes diagnoses)

    -- which does raise the question of whether an earlier intervention in those at low risk for cognitive decline would lead to significant cognitive benefit in the long-term (ie, perhaps we should not interpret the results of this study that the focus of the interventions should be only on those at high risk of dementia)

        -- one of the main benefits of improved hearing, even in those at lower risk of cognitive decline, is the increasing social isolation that accompanies decreased hearing, which may well anticipate subsequent cognitive decline, and may in and of itself be a major quality-of-life issue for those with decreased hearing. This may also be the case for those around them who may witness this decreased hearing, may be less able to engage these folks in conversations (and benefit from their input), and therefore may unwittingly add to the problem of their social isolation and decreased quality of life 

--further analyses assessing brain MRI and social engagement will be published subsequently 

 

-- Potential mechanisms that hearing loss could lead to increased cognitive decline: effects of hearing loss on cognitive load (information degradation hypothesis), effects on brain structure (sensory deprivation hypothesis), and reduced engagement in social and cognitively stimulating activities. Of note, these mechanisms are not mutually exclusive 

 

-- a few other side notes on dementia:

    -- there may well be a relationship between sustained PPI use and dementia (https://gmodestmedblogs.blogspot.com/2023/08/sustained-ppi-use-and-subsequent.html ), and this might also be a "modifiable risk factor" for dementia (there are a variety of potential mechanisms here, including PPI-related vitamin B12 malabsorption and the PPI’s profound effects on the microbiome/gut-brain axis, as noted in this blog)

        -- given the multitude of potential adverse clinical effects attributed to PPIs, including decreased vitamin and iron absorption, CKD, osteoporosis/fractures, community-acquired pneumonia, TB activation, C difficile infections (though some of these are better documented than others), there is a strong recommendations from the Am Gastroenterological Assn to de-prescribe PPIs (https://gmodestmedblogs.blogspot.com/2022/07/ppis-way-overprescribed.html), stressing that PPI prescriptions should be minimized, especially given the high frequency of unnecessary PPI prescribing (at least 60% of the time, and that does not include OTC)

        -- the issue for us in primary care is that once a person is doing well on PPIs, we often move on to other pressing issues of concern for the patient and, given our short times with patients, we may simply refill the PPIs. so, the best approach to GERD (except with errosive esophagitis) and dyspepsia is usually to start with calcium (tums) or perhaps H2 blockers or both (these do work quite well for large numbers of patients) and step-up to PPIs only if there is an inadequate response (though still try to step-down later). Studies have shown that step-up therapy is much more likely to avoid long-term PPIs than step-down therapy (which studies have shown to happen pretty infrequently)

    -- there was also a recent article suggesting blood pressure variability in the elderly over time is associated with dementia (dementia BP variability in elderly inc risk JAMA2023 in dropbox, or doi:10.1001/jamanetworkopen.2023.40249.

        -- short-term blood pressure variability has been shown to increase stroke risk

        -- this would argue strongly for using BP meds that have long half-lives (the best documented med overall for decreasing blood pressure variability in the short-term seems to be amlodipine, which is also associated with decreased strokes; ARBs are quite good (losartan at 100mg/d, others at any dose). Chlorthalidone also seems okay. ACE inhibitors do not seem to have 24-hour coverage (and are associated with increased stroke in several studies). HCTZ is awful as a solo agent: https://gmodestmedblogs.blogspot.com/2023/01/hypertension-hctz-vs-chlorthalidone.html .

        --this argument to use longer-acting antihypertensives is my extension of the issue of short-term blood pressure variability to long-term variability:

            -- it does seem logical that better short-term 24-hour control would be associated with decreased BP variation longterm, given that patients or clinicians checking BP at different times of the day may observe less variability with longer-acting meds, and therefore better 24-hr blood pressure control)

            -- and we know that the "gold standard" of good blood pressure control in terms of clinical outcomes is 24-hour ambulatory blood pressure monitoring: https://gmodestmedblogs.blogspot.com/2015/10/uspstf-guidelines-on-blood-pressure.html

             -- this suggests that we should be checking BPs in the elderly pretty often to make sure that blood pressure is well-controlled all the time (and home-based measurements may also be very useful).

    -- this study on hearing aids would suggest that other sensory deficits may be associated with increased dementia risk: vision loss is also associated with these same problems: https://www.nia.nih.gov/news/vision-impairment-associated-many-100000-u-s-dementia-cases

    -- social isolation itself is associated with dementia (https://www.cdc.gov/aging/publications/features/lonely-older-adults.html ), and I have seen several patients who had dramatic cognitive improvement after getting into adult day health programs vs staying at home alone when their children/grandchildren are away during the day)

 

-- the NY Times had a recent article on OTC hearing aids ( https://www.nytimes.com/2023/10/30/health/hearing-aids-dementia.html?smid=nytcore-ios-share&referringSource=articleShare ), which comments on the above study, as well as on a recent small study on self-fitting OTC hearing aids in those with mild to moderate hearing loss, finding that these OTC hearing aids were as good as the much more expensive audiologist-fitted ones (see https://gmodestmedblogs.blogspot.com/2023/05/hearing-aids-otc-ones-work-well.html ). However, on independent testing per the NY Times, only those costing $1000 or more are of adequate quality. still quite expensive, not covered or minimally covered by insurance (Medicare does not cover even audiologist-prescribed ones, except in some Advantage plans). The actual costs of making the hearing aids is in fact pretty small, so hopefully their cost will decrease and they will become accessible to the broad array of people who might benefit.

 

Limitations: 

-- as mentioned above, a longer study might well show benefit in those without pre-existing risk factors for cognitive decline. It may well be that the neurocognitive battery of tests was insufficiently sensitive to pick up small degrees of cognitive decline that might have been manifested in a longer study 

    -- another issue: many of the "high risk" factors for dementia (cardiovascular, diabetes) may become more manifest in the low-risk group over time

    -- there will be a follow-up study of this to assess long-term outcomes, as well as some of the social outcomes noted above  

-- the recruitment process used for patients in the de novo study, recruiting them through advertisements in local newspaper, radio, and Internet, may reflect an important selection bias, with fewer people seeing these advertisements who had more baseline cognitive decline; lower levels of education; not be literate or know English; not be tech-savvy enough; have more life stressors limiting their exposure to these advertisements; and generally be a less healthy group (the de novo group had a “healthy volunteer bias”). The ARIC study, in contrast, recruited randomly selected individuals from the communities

-- also, the participants in the ARIC group were recruited more than 30 years ago, also likely reflecting a substantial difference in participants who were recruited, as well as perhaps differences in the effects of repeated neurocognitive testing (the former group being more naïve to repeated cognitive testing than those recently recruited) 

-- perhaps the elephant in the room here is that for many people, the cost of hearing aids is prohibitive and there is no or minimal insurance coverage (for those with insurance) for huge numbers of patients

-- also this was not in unblinded study, clearly both the participants and the study technicians new which group participants were in, and this might have biased their approach to the patient

 

 So,

-- it is pretty clear that hearing deficits in the elderly have multiple bad effects: decreased social interactions/social isolation/quality of life, as well as increased dementia. 

-- as noted above, there are potentially other major potentially remediable causes of dementia, including extended PPI use, decreased vision, social isolation, blood pressure variability

    -- and, of course vitamin B12 deficiency, hypothyroidism, depression, syphilis, HIV, and atrial fibrillation (even without a history of strokes) should be considered/assessed as appropriate.

-- since there are many potentially modifiable conditions for dementia, it seems incumbent on us clinicians to screen for these conditions and treat appropriately

    -- my concern here is that we are probably much better at looking for potentially reversible medical causes instead of the more social ones (isolation, hearing/vision impairment)….

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

  

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

  -- click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

  -- or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them

if you would like access to the dropbox for articles, go to https://www.dropbox.com/scl/fo/vj803z91w1trd471h9fj8/h?rlkey=klpxdjpdhcdt3sahnpirzz730&dl=0

please feel free to circulate this to others. also, if you send me their emails (gmodest@bidmc.harvard.edu), i can add them to the list