Hearing and dementia: from ears to brain

Abstract

The association between hearing impairment and dementia has emerged as a major public health challenge, with significant opportunities for earlier diagnosis, treatment and prevention. However, the nature of this association has not been defined. We hear with our brains, particularly within the complex soundscapes of everyday life: neurodegenerative pathologies target the auditory brain, and are therefore predicted to damage hearing function early and profoundly. Here we present evidence for this proposition, based on structural and functional features of auditory brain organization that confer vulnerability to neurodegeneration, the extensive, reciprocal interplay between 'peripheral' and 'central' hearing dysfunction, and recently characterized auditory signatures of canonical neurodegenerative dementias (Alzheimer's disease, Lewy body disease and frontotemporal dementia). Moving beyond any simple dichotomy of ear and brain, we argue for a reappraisal of the role of auditory cognitive dysfunction and the critical coupling of brain to peripheral organs of hearing in the dementias. We call for a clinical assessment of real-world hearing in these diseases that moves beyond pure tone perception to the development of novel auditory 'cognitive stress tests' and proximity markers for the early diagnosis of dementia and management strategies that harness retained auditory plasticity.

Keywords: Alzheimer’s disease; Lewy body disease; dementia; frontotemporal dementia; hearing.

Figure 1

Processes and interactions in ‘peripheral’ and ‘central’ hearing. The functional organization of the auditory processing hierarchy and the interplay of hearing with more general cognitive functions. Ellipses indicate the broad domains of peripheral hearing (blue; anatomically, the peripheral hearing apparatus which receives incoming sounds, cochlea and auditory nerve), precognitive auditory processing (green; chiefly the auditory brainstem), auditory cognition (yellow; auditory cortex and its cerebral connections) and general cognitive functions (red; see Fig. 2 for neuroanatomy). Listed within the ellipses are some key stages in the analysis of auditory information: ‘peripheral’ and ‘central’ hearing processes lie on a functional and anatomical continuum, with reciprocal connections between successive processing stages (black arrows). This organization implies that pathologies (such as neurodegenerative proteinopathies) predominantly targeting auditory cognitive (and general cognitive) processing stages may have cascading effects at other processing stages. Certain additional functional properties that operate across auditory processing stages, such as non-linear signal coding and plasticity, are likely to be particularly vulnerable to the effects of neurodegenerative pathologies (see text). External red and blue arrows here signify general mechanisms by which hearing dysfunction of any cause may promote cognitive decline, and the converse; these mechanisms are likely to be mutually reinforcing and may additionally compound more specific effects of auditory brain dysfunction, with the potential to establish pathophysiological ‘vicious cycling.’

Figure 2

The auditory brain in health and neurodegenerative disease. (A) Major anatomical regions that mediate the processes underpinning hearing (Fig. 1) as spheres overlaid in a left lateral view of the brain. These regions are anatomically and functionally linked into large-scale, distributed networks. The colour convention follows that in Fig. 1 (green, precognitive auditory processing in brainstem pathways, enclosed by the grey filled outline; yellow, auditory cognition in auditory cortices; red, general cognitive processes in connected cerebral regions); note, however, that there is no simple, one-to-one correspondence between particular brain regions and individual ‘tiers’ of the processing hierarchy outlined in Fig. 1. Brain regions are designated as follows: ATL = antero-mesial temporal lobe (also encompassing amygdala and hippocampus); CN = cochlear nucleus (ventral and dorsal); HG = Heschl’s gyrus (medial portion contains primary auditory cortex); IC = inferior colliculus; IFG = inferior frontal gyrus (closely associated with insular cortex, deep to the cerebral surface); IPL = inferior parietal lobe; ITC = inferior temporal cortex; MGB = medial geniculate body; MTG = middle temporal gyrus; OFC = orbitofrontal cortex; PFC = prefrontal cortex; SO = superior olive (its main projection in the lateral lemniscus has several additional, small associated nuclei); STG = superior temporal gyrus; TPJ = temporo-parietal junctional cortex. Also shown in grey filled outline is the cingulate gyrus, projected from the medial surface of each cerebral hemisphere: this signifies linked deep medial prefrontal and parietal cortices that also participate importantly in integrative and modulatory cognitive processes relevant to hearing. (B) Key components of the brain networks implicated in hearing that are also predominantly targeted in representative neurodegenerative proteinopathies. These patterns of brain degeneration anticipate the differential involvement of particular auditory functions and therefore distinctive functional hearing profiles or ‘auditory phenotypes’ of these disorders (see text and Table 1). Although the neuroanatomical patterns shown correspond to the distribution of most severe regional brain atrophy in each disease, dysfunction predates atrophy and additional connected brain regions may also be implicated in the pathogenesis of auditory symptoms. AD = typical Alzheimer’s disease; nfvPPA = non-fluent agrammatic variant primary progressive aphasia; svPPA = semantic variant primary progressive aphasia.

Figure 3

A pathophysiological synthesis of hearing impairment and dementia. This figure schematizes proposed relations between development of peripheral hearing loss (blue), changes in auditory cognition (gold) and general cognitive function (red) and underlying neurodegeneration (black), based on emerging epidemiological and pathophysiological evidence. Hearing loss can be considered a potential causal risk factor for cognitive decline (Risk), a proximity marker for incipient dementia (Proximity) or a feature of the established dementia syndrome (Phenotype), according to the time window in which it occurs; the mechanisms of these effects are distinct but likely to be interdependent. Alzheimer’s disease has been the major focus of epidemiological studies assessing the risk of developing dementia in association with hearing loss (Lin et al., 2011; Taljaard et al., 2016; Livingston et al., 2017; Loughrey et al., 2018), though the distinction from cerebrovascular and other pathologies is problematic; midlife hearing loss may account for ∼10% of all cases of dementia, and has been proposed to have a direct potentiating effect (arrow) on the evolution of neurodegeneration. Though the mechanism of this linkage is unclear, animal models suggest it could occur via cellular effects such as oxidative stress or altered gene expression (Frenzilli et al., 2017; Park et al., 2018), changes in neural circuit function (Oxtoby et al., 2017; Bidelman et al., 2019) or a complex interaction between aberrant circuit activity and protein spread (Griffiths et al., 2020). However, a direct causal effect has not been established: for example, peripheral hearing function was not associated with brain amyloid deposition (a relatively specific preclinical marker of Alzheimer’s disease) in a large cohort of cognitively healthy older people (Parker et al., 2020) and such an effect would still not account for the majority of cases of dementia with hearing alterations. Here we suggest that alterations in ‘central’ hearing or auditory cognition may constitute an early warning signal of incipient dementia, due to the computational demands imposed by listening in challenging everyday acoustic environments. In support of this idea, predominantly central auditory deficits (involving, for example, dichotic listening) have been shown to predict CSF tau levels and regional atrophy profiles consistent with Alzheimer’s disease pathology in cross-sectional studies (Tuwaig et al., 2017) and longitudinal development of a clinical syndrome compatible with Alzheimer’s disease (Gates et al., 2011), while large genetic and neuropathological surveys have suggested changes in hearing (in particular, speech-in-noise perception) may be a preclinical marker of neurodegeneration (Brenowitz et al., 2020a, b). We emphasize that deficits of peripheral and central hearing and more general cognitive functions are likely to interact strongly, with ‘vicious cycling’.