Listening with an Ageing Brain – a Cognitive Challenge

Abstract

Hearing impairment has been recently identified as a major modifiable risk factor for cognitive decline in later life and has been becoming of increasing scientific interest. Sensory and cognitive decline are connected by complex bottom-up and top-down processes, a sharp distinction between sensation, perception, and cognition is impossible. This review provides a comprehensive overview on the effects of healthy and pathological aging on auditory as well as cognitive functioning on speech perception and comprehension, as well as specific auditory deficits in the 2 most common neurodegenerative diseases in old age: Alzheimer disease and Parkinson syndrome. Hypotheses linking hearing loss to cognitive decline are discussed, and current knowledge on the effect of hearing rehabilitation on cognitive functioning is presented. This article provides an overview of the complex relationship between hearing and cognition in old age.

1. Introduction

Successful communication in complex listening situations requires not only the detection of the target signal and the segregation of the scenario into different sound sources. The listener must also track who is speaking, grasp the meaning of the statement, memorize and compare it with already existing knowledge, suppress irrelevant interfering signals, formulate an own response in parallel and execute it at the right time. Longer conversations in groups require the integration of new information with already expressed contents of each speaker while again and again the attention switches between the persons involved.

This means that in order to assess and use the information contained in spoken language, a fluent and swiftly functioning integrative system of perceptual and cognitive processes is required. Both the auditory and cognitive systems are subject to typical aging processes; and with higher age, the incidence of neurodegenerative diseases increases, sometimes having a considerable influence on the ability to communicate. In recent years, hearing disorders have increasingly become the focus of scientific research as a potentially modifiable risk factor for neurocognitive impairment in an aging society. In this review, hypotheses on the causal relationship will be presented as well as specific auditory impairments in the context of the most common neurodegenerative disorders of the elderly. Finally the effect of hearing rehabilitation will be discussed.

2. Cognition and speech understanding

2.1 Definition and domains

Cognition (Latin: cognoscere =to recognize, to experience, to perceive) is a collective term for processes of reception, processing, storage, and retrieval of information as well as their results (knowledge, attitudes, beliefs, expectations). These processes can take place consciously, e. g. when solving tasks, or unconsciously, e. g. when forming opinions ¹ . Human cognitive skills include, among others, processes of perception, attention, learning and memory, thinking, but also recognition of emotions, and control of one’s own behavior. The ability to use these skills to solve problems, adapt to new situations, and interact effectively with the environment is referred to in psychology as “intelligence” (Latin intelligentia =cognition, intellect). While Cattell’s intelligence model distinguished only between fluid intelligence (innate, experience-independent ability to reason and solve problems) and crystalline intelligence (predominantly culture-dependent ability to apply acquired knowledge), nowadays the Cattell-Horn-Carroll (CHC) model is considered the one that most comprehensively describes the structure of intelligence ² . It includes 16 factors from the areas of acquired knowledge, thinking ability, processing speed, memory, sensory processing, psychomotor skills, and kinesthetics and serves as the basis for the most widely used intelligence tests.

For the diagnosis of neurocognitive disorders, the “Diagnostic and Statistical Manual of Mental Disorders – DSM-5” ³ defines 6 cognitive domains on which the diagnostic criteria are based and which can be assessed in standardized neuropsychological testing ( Table 1 ).

Table 1 Cognitive domains for the diagnosis of neurocognitive disorders in DSM-5 ³

Cognitive domain	Subdomains
Complex attention	Permanent attention	Didived attention	Selective attention	Processing rate
	Executive functions	Planning	Decision making	Working memory	Exploiting feedback/correting errors	Acting against habits/behavioral inhibition	Mental flexibility
		Learning and memory	Immediate memory*	Short-term memory (including free recall, recall with cue stimuli, and recognition)	Ultra long-term memory (semantic and autobiographical)	Implicit (procedural) learning
			Speech	Speech production (including naming, word finding, word fluency, grammar and syntax)	Speech comprehension
Perceptive-motor				Visuo perception	Visuo construction	Perceptive-motor	Practice	Gnosis
			Social cognition	Recognizing emotions	Theory of mind (ability to observe the Erkennen von Emotionen Theory of Mind (ability to pay attention to another person's state of mind or experience)

*is sometimes included in the term of working memory.

2.3 Normal cognitive aging

Cognitive processes are subject to chronological aging processes to varying degrees and are highly associated with the loss of everyday functioning, onset of dementia, and general mortality ^{4 5} . It is well known that basal, knowledge-independent “fluid” functions show a greater age decline than lifelong acquired “crystalline” knowledge, which can still show growth into old age ⁶ . A persons’ intelligence is seen as the result of function or knowledge build-up, loss, and compensation mechanisms. This means in order to maintain cognitive performance as fluid abilities are lost, we rely more and more on already established, automated crystalline processes to accomplish tasks ^{6 7} . Research continues to address the extent to which training can counteract functional loss and the importance of individual cognitive domains in this process. In a large cross-sectional study on 48,537 subjects and evaluation of normative values of standardized IQ and memory tests, Hartsthorne and Germine were able to show that there is greater heterogeneity with regard to the time of maximum functional capacity between the individual domains than previously assumed ⁸ : short-term memory and processing speed reach maximum values already in the teenage years, working memory peaks in young adulthood with the onset of decline in the 30ies. Peak performance in e. g. vocabulary and emotion recognition, on the other hand, is reached only in the middle age and maintained over a much longer period of several years. As an explanation for individual performance differences, however, non-specific age effects, such as a general slowing, must be considered in addition to these domain- and function-specific changes. Recent long-term studies indicate that approximately 30-50% of individual differences in age progression are due to a “general factor” ⁹ .

In addition to significantly reduced general processing speed and working memory compared to younger people, loss of executive functions and episodic memory occur in older age ^{10 11} . Morphologically, changes are seen in the middle temporal lobe (episodic memory) and the prefrontal/striatal system (executive functions) ¹² . Neurodegenerative diseases such as Alzheimer’s disease or Parkinson’s syndrome affect these areas to varying degrees and lead to specific functional deficits.

2.4 Cognitive reserve

People of about the same age with similar central changes, e. g., in the context of a neurodegenerative disease but also in the course of normal aging, may nevertheless vary considerably in their clinical symptoms and cognitive performance. To explain this observation, the concept of cognitive reserve was introduced ¹³ . It refers to the ability to compensate for newly occurred damage and to maintain existing functions by using alternative neuronal networks ¹⁴ . Both congenital and acquired or environmental factors (e. g., intelligence, educational level, physical activity, recreational and social activities) play a role. Differences in cognitive reserve are also considered as an explanation for the individual impact of sensory impairment (e. g., hearing loss) in higher age ¹⁵ .

2.5 Information processing model and cognitive concepts in relation with hearing and speech comprehension

From a cognition-psychology perspective, spoken communication can be understood as a process of information processing: The incoming stimulus is perceived by the sensory system, processed, and finally leads to a reaction ( Fig 1 , adapted from ¹⁶ ). This complex process depends the properties of the incoming stimulus (bottom-up) and is also influenced by cognitive processes (top-down). In the theoretical model of Wingfield and Tun ¹⁷ ( Fig 2 ), the interactive roles of peripheral, central, cognitive, and linguistic factors to speech understanding are illustrated in more detail: In the periphery, the sensory system must receive the spectral and temporal cues of the speech signal and pass them on to the central auditory pathway for further processing with as little interference as possible. In the next stage of central auditory processing (perceptual system), binaural information is encoded in addition to spectral and temporal features of the speech signal (especially signal onset and duration). The so-called “object formation”, i. e. the ability to recognize a target signal and to follow it in the presence of competing background noise or speakers, also occurs at this level. This is followed by the linguistic operations of sound analysis and lexical recognition at the word level. Based on syntactic (position of a word in a sentence) and semantic (word meaning) prior knowledge, sentences are captured. The comparison with contextual information (speaker, situation, object, time, etc.) finally enables comprehension within the conversation ¹⁸ . The single processing steps are influenced by cognitive abilities or processes such as memory functions (prior knowledge, working memory) and general processing speed, attention, and executive functions (top-down). At the same time, the characteristics of the stimulus (e. g., speech rate, accent, type and number of noise sources, reverberation etc.) determine subsequent processing (bottom-up). Auditory and cognitive processes are so closely intertwined that a sharp separation of “peripheral” and “central” auditory functions does not adequately capture the complexity of speech processing ¹⁹ . The typical complaint of the elderly – to hear but to understand poorly – is merely a clinical symptom of normal age-related changes in all sections of this system from the periphery to the cortex, which may be additionally impaired by neurodegenerative diseases.

Fig 1.

Generalized model for bottom-up and top-down processing of auditory information (adapted from [16]). The stimulus is first coded into neural information in the periphery, relevant information is selected and then interpreted in the next step. Finally, it is stored in memory while the answer is formulated at the same time. The quality and content of the stimulus influence further processing (bottom-up), information that has already been extracted or recorded content can lead to changes in the processing of subsequent stimuli (top-down).

Fig 2.

Wingfield and Tun's information processing model [17]. Sensory, perceptual and cognitive systems interact when processing auditory information. The mixed input signal must first be broken down into relevant (target signal) and irrelevant information (interfering signal). The attention filter determines the extent to which the individual signal components are further processed. Word recognition is first achieved via several intermediate steps, and finally, after further linguistic operations, discourse comprehension. The information processing process can be influenced at all levels by both cognitive and acoustic factors.

3. Age-related hearing loss

3.1 Prevalence and socio-economic consequences

In 2019, according to the WHO, about 1.5 billion people worldwide were affected by hearing loss ²⁰ , and 430 million (about 5.5% of the world population) had at least moderate hearing impairment. The WHO expects this number to increase to 700 million people with moderate or higher levels of hearing loss in the better hearing ear by 2050, out of a projected total of 2.5 billion people affected. The individual development of hearing throughout the life span depends on various protective and damaging factors ²¹ . In addition to genetic, biological, and environmental influences, individual lifestyle (nicotine abuse, diet, noise exposure) also plays a role. Age-related hearing loss (ARHL) represents the greatest socio-economic burden over a lifetime due to its high prevalence in the population. According to current estimates, in 2019 approximately 42% of all people affected by hearing loss were at least 60 years old ²⁰ , and the proportion of moderate or higher levels of hearing loss increases exponentially with higher age (prevalence at 60-69 years: 15.4%; more than 90 years: 58.2%). WHO estimates the annual costs due to hearing loss to be approximately $ 980 billion. In recent years, age-related hearing loss has been increasingly identified as a potential risk factor for neurocognitive disorders ^{22 23 24 25} . Positive effects of audiological rehabilitation with hearing aids for the course of these disorders ^{26 27 28} as well as health-related quality of life ²⁹ are seen. Nevertheleess, in Europe, hearing aids are used by only about 33% of the approximately 57 million people with hearing loss in need of care, although they are widely available ^{20 30} .

3.2 Age-related changes of the peripheral auditory system

Age-related degenerative processes affect both outer and inner hair cells, supporting cells, stria vascularis, and spiral ganglion cells ^{31 32 33 34 35 36} . The pure-tone audiogram typically shows a loss of high frequency hearing ^{36 37 38} . For medical expert reports, DIN EN ISO 7029:2017 should be consulted, which allows estimation of normal hearing for ages between 20 and 80+ years ³⁹ ( Fig 3 ). The current 3 ^rd version is based on data from healthy men and women published after 2000. Compared with previous versions, the average hearing loss is lower for all age groups, reflecting changes in living and working conditions.

Fig 3.

Average hearing threshold progression for men and women aged 20-80+ according to DIN EN ISO 7029:2017:06 (according to [39]). The 50th percentile of the respective age group is shown.

Based on experience from animal models regarding the underlying etiology, Dubno et al. ³⁷ classified audiometric phenotypes of age-related hearing loss. A low-grade hearing loss up to 1 kHz and rather flat high-frequency hearing loss in indicative of metabolically-induced atrophy and degeneration of the stria vascularis, whereas a steeply declining hearing threshold between 2 and 8 kHz with normal low-frequency hearing indicates a sensory disorder (hair or supporting cell damage).

In the same study ³⁷ , approximately 11% of pure-tone audiograms were classified as “older normal”, with an average hearing loss of no more than 20 dB HL in the high-frequency range. Nevertheless, elderlies with normal pure-tone audiograms also report hearing difficulties and tinnitus ^{40 41} . For this hidden hearing loss (HHL), different pathophysiological mechanisms have been discussed in recent years ^{42 43 44} . In addition to disturbances of the afferent synapse of the inner hair cells (cochlear synaptopathy ^{35 42 43 45 46} ), demyelination processes (temporary loss of cochlear Schwann cells ⁴⁷ or in the context of demyelinating neuropathy ⁴⁸ ), and persistent dysfunction of the outer hair cells ^{49 50} have been described. These changes lead to impaired transmission of temporal and spectral fine structure ⁵¹ , especially of rapid signal changes as well as signal duration. The phonetic contrasts necessary for accurate word recognition decrease, which manifests in reduced speech understanding, especially in noisy environments, even before high-frequency hearing loss is detected in the puretone audiogram.

Amplitude changes of wave I of the early auditory evoked potential elicited by suprathreshold stimulation ^{34 46 51} or an altered SP/AP amplitude ratio in electrocochleography ⁵² are discussed as electrophysiologic markers of the disturbed cochlear function.

Age-related changes of the central auditory system

3.2.1 Structural-morphological as well as neurochemical changes

Aging processes affect the entire central auditory pathway from the cochlear nucleus to the auditory cortex (see ^{53 54} for a comprehensive overview). Throughout the lifespan, the human cortex is subject to remodeling processes that become visible and measurable due to modern imaging techniques such as magnetic resonance imaging. MR spectroscopy also allows metabolic and neurochemical changes to be detected. In healthy adults, there is a general brain volume reduction with increasing age ^{55 56 57} . Volume changes in gray matter ^{58 59 60} and white matter ^{60 61 62} as well as cortex thickness ^{58 63} have been discribed. Regions particularly affected include the temporal lobe, hippocampus ^{60 64 65} and prefrontal cortex ^{59 61 66 67} . Lin et al. ⁶⁸ demonstrated that hearing loss accelerates volume decline in both the total volume and the right temporal lobe. Further studies showed gray matter reductions beyond the age norm in the superior and medial temporal gyrus ⁶⁹ and superior and medial frontal gyrus ^{69 70 71} , primary auditory cortex ^{72 73} , and occipital lobe and hypothalamus ⁷⁰ . Diffusion-weighted MR images also showed changes in myelination, fiber density, and axonal parameters in the superior olive complex, lemniscus lateralis, and inferior colliculus ^{69 74} . MR spectroscopy has demonstrated dysfunction of GABAergic neurotransmission in the central auditory system of patients with presbycusis ^{54 75 76} .

This means that, on one hand, structural changes in the central auditory pathway already occur in the course of normal aging, which can have a negative effect on speech understanding; on the other hand, age-related hearing loss additionally leads to impairment of further areas in the association cortex ⁷⁷ .

3.2.2 Changes of central-auditory processing and perception

Structural and neurochemical changes in the central hearing pathway lead to impaired encoding of temporal characteristics of speech. As part of normal aging processes, there are changes in neural timing and precision in speech processing ¹⁸ with implications for comprehension of speech both in quiet and in noise. In general, the ability to perceive rapid temporal changes in the speech signal decreases. That is, older people need larger differences or temporally longer features (voice onset time, vowel duration, pauses etc.) to distinguish individual speech sounds ⁷⁸ . If the speech signal is additionally spectrally altered, these difficulties increase, as several studies with vocoded speech have shown ^{79 80} . This is particularly relevant with regard to cochlear implant fitting. Impaired neural encoding of signal onset is also thought to be the cause of greater difficulty for elders to understand speech with altered speed, stress, or rhythm. For example, research by Gordon-Salant et al. demonstrated that older normal-hearing subjects have significantly greater problems with understanding fast speakers or speech with a foreign accent ^{81 82} .

The ability to separate single speech streams, i. e. to follow a speaker in the presence of noise or competing speakers, also declines with age and has been demonstrated in a multitude of studies ^{83 84 85 86} . This has been attributed to impaired processing of temporal fine structure as well as perception of brief amplitude changes in the envelope of the speech signal (“listen to the dips”) ⁸⁷ . In addition, age effects have been demonstrated in the binaural processing of speech signals ^{88 89 90} .

A comprehensive review of age-related electrophysiological changes in the central auditory pathway can be found in ⁹¹ . Early auditory evoked potentials, especially the so-called frequency following response (FFR) after stimulation with both tone and speech signals, objectify impaired temporal processing at the brainstem level. Depending on the experimental design, late auditory evoked potentials allow both the differential detection of disturbed temporal processing of auditory stimuli at the cortical level, independent of attention and cognition (N1-P2) and the assessment of cognitive processes if the potentials are measured in an event-related manner (P300, N200). Therefore, the latter can also be used to distinguish between normal aging processes, mild cognitive impairment, and Alzheimer’s dementia ⁹² .

3.2.3 Central presbycusis

In English-speaking countries, the described age-related disorders of central processing and perception of auditory information with age-appropriate pure-tone hearing threshold are summarized under the term of central auditory processing disorder (CAPD) or central presbycusis ^{93 94} . The disorder is considered to have multifactorial causes, correlations with age-related cognitive disorders are seen, clinically, a sharp separation between cognitive and auditory processing is hardly possible.

The German Society of Phoniatrics and Pediatric Audiology defines auditory processing and perception disorders in more detail: According to the current guideline, the diagnosis of auditory processing and perception disorders should only be made if, at age-appropriate pure-tone hearing thresholds, there are deficits in analysis, differentiation, and identification of time, frequency, and intensity changes of acoustic or auditory speech signals as well as processes of binaural interaction (e. g., for sound localization, lateralization, noise suppression, and summation) and dichotic processing that cannot be better explained by other disorders, such as attention deficits, general cognitive deficits, cross-modality mnestic disorders ⁹⁵ . The deficits in the auditory domain must be significant compared to language-independent cognitive performance. At the same time, there is a high comorbidity to e. g. disorders of attention. Clinically, it must then be decided, taking into account all findings, which disorder is leading. With regard to differentiation from infantile auditory processing and perception disorder and in view of the usually modality-spanning aging processes, it seems to make sense to rather use the term of “central presbycusis” for disorders of central auditory processing newly occurring in older age.

3.3 Influence of cognitive processes on speech comprehension

In order to follow a conversation successfully and participate in it, listeners and speakers must not only perceive what is being said and understand the single words even under unfavorable complex conditions (background noise, reverberation, high speech rate, accent, etc.), but also grasp the content in context, compare it with their own prior knowledge, and formulate a response. On the cognitive level, this requires, among other things, keeping one’s attention on the target signal, storing it in working memory, and matching it with long-term memory – as quickly as possible in order to be able to follow the rest of the conversation. Working memory, executive functions, and processing speed are therefore seen as the most important cognitive factors for speech comprehension, especially in noise ⁹⁶ , and a large number of studies have investigated them (e. g., executive functions and attention ^{97 98} , processing rate and working memory ⁸⁷ ). The importance of auditory and cognitive factors and their interaction for the quality of speech comprehension has been increasingly taken into account in the last 10-20 years, so that the term of “cognitive hearing science” has been established ⁹⁹ .

3.3.1 Inhibition control

In the information processing model by Wingfield and Tun ¹⁷ ( Fig 2 ), the “attention filter” symbolizes the ability to selectively follow a single signal in the presence of noise or competing speakers and thus to suppress further processing of the non-selected speech streams very early in the process. Disruption of inhibition control, e. g., in the course of normal aging, limits this ability and may thereby impair speech comprehension.

At the word level, perceived phonemes must be matched with the mental lexicon. The success of this lexical process depends on the frequency of occurrence of a word within a language as well as the number of words, with overlapping phonemes (neighborhood density). The Neighborhood Activation Model ¹⁰⁰ theorizes that the more frequently a word occurs within a language (high frequency) and the fewer words with overlapping phonemes (low neighborhood density), the easier it is to recognize the word correctly. Accordingly, words with high neighborhood density have more competitors that must be suppressed by the listener to enable correct word retrieval. Research on the neighborhood density effect has shown that in older adults, there is a significant relationship between measures of inhibition control and speech comprehension in noise (e. g., ¹⁰¹ ). In addition, with increasing age, frequently occurring competing words are more intrusive, i. e. they are more often misidentified as a target signal ^{102 103} .

3.3.2 Working memory

In cognition psychology, working memory is understood as limited resource that allows information to be kept and processed in immediate memory ¹⁰⁴ . In phonological analysis, working memory is considered to play a significant role as an interface to long-term memory. To explain why in some situations speech understanding is effortless while in others increased listening effort is required, Rönnberg et al. developed the Ease of Language Understanding (ELU) model (see ¹⁰⁵ for a comprehensive review). The incoming multimodal signal is quickly and automatically matched (within 180-200 ms ¹⁶ ) with the mental lexicon. If a minimum number of matching phonological attributes is found, the implicit lexical process proceeds rapidly, and the signal is understood. If no match is found, semantic and episodic long-term memory must be explicitly accessed with the aid of working memory to enable language processing. If the input signal is difficult to understand – e. g., due to hearing impairment or unfavorable acoustic environment – it must be held longer in working memory and more cognitive resources must be expended to understand what is being said. Listening effort increases ¹⁰⁶ . In particular, a significant dependence on working memory capacity has been shown for speech comprehension in noise, independent of age ^{107 108} .

3.3.3 Significance of the context

Phonological matching can be facilitated by the aid of contextual information, allowing partial compensation for the deficits caused by hearing impairment. Benichow et al. ¹⁰⁹ , for example, demonstrated that although hearing loss had a significant effect on speech understanding in noise, it decreased with increasing probability of the target word to occur in the context of the sentence. At the same time, both age and cognitive performance (especially working memory as well as processing speed) were significant predictors of speech understanding independent of the amount of contextual information.

Increasing deficits in inhibition control with age may, in turn, contribute to wrongly identify acoustically unintelligible words as utterances that are probable within context ^{110 111 112} . A recent study by van Os et al. ¹¹³ revealed that older subjects are also able to rationally adjust their response behavior within a trial and, for example, rely more on the acoustic information than the context when the context offered is misleading.

3.3.4 Listening effort

If cognitive resources must be used to understand disturbed speech signal, they are lacking for other processes such as encoding what is heard into memory. The so-called “Framework for Understanding Effortful Listening” (FUEL) ¹¹⁴ describes successful speech comprehension as dependent on the quality of the acoustic stimulus, the demand of the task, and the listener’s motivation to exert the effort necessary to achieve it. Increased listening effort may not only deplete available cognitive resources more rapidly, but also reduce the listener’s motivation to exert that effort at all – even if the utterance itself was correctly understood.

4. Hearing disorders in frequent neurodegnerative diseases in higher ages

4.1 Neurocognitive disorders

Neurocognitive disorders (NCD) are disorders that are associated with a subjective or objective loss of previously existing cognitive abilities in at least one of the 6 cognitive domains of complex attention, executive function, learning and memory, language, perceptual-motor, social cognition (cf. Table 1) and do not only occur exclusively in the context of delirium or can be explained by another existing mental disorder (such as major depression, schizophrenia) ³ . The DSM-5 distinguished between mild (minor) and severe (major) forms, which are seen on a continuum of cognitive and functional impairment. In minor NCD, moderate cognitive performance impairments are present but do not affect the ability to perform activities of daily living independently, although greater effort or compensatory strategies may be required. In major NCD cognitive performance has significantly declined and impairs independence in performing activities of daily living. The impairment in everyday activities can me mild (only instrumental activities such as household, handling money), moderate (limitations in basic activities of daily life like eating, dressing), or severe (complete dependence). The major NCD is intended to replace the widely used, and sometimes stigmatizing, term of dementia. Specific pathophysiological processes are known for the majority of neurocognitive disorders, allowing further specification of both minor and major NCD ( Table 2 ).

Table 2 Specific etiology of neurocognitive disorder NCD) in DSM-5 ³ .

Minor/major NCD due to
Alzheimer’s disease
Fronto-temporal lobe degeneration
Lewy body dementia
Vascular disease
Cranio-cerebral trauma
Substance/drug consumption
HIV infection
Prion disease
Parkinson’s syndrome
Huntington disease
Other medical factors
Multiple etiologies
Not specified

4.1.1 Socio-economic relevance

Neurocognitive disorders predominantly affect older age, so a global increase in the number of cases is expected with demographic change. Based on data from the Global Burden of Disease Study of 2019 ¹¹⁵ , the number of dementia patients worldwide was estimated to 55.4 million in 2019, and projections expect and increase to 152.8 million affected people in 2050 ¹¹⁶ . In some regions, however, decreases in new cases were observed: A recent analysis of the incidence rate over the last 25 years for Europe and North America showed a decrease in the incidence of dementia 13% per decade ²⁵² . According to the German Alzheimer Society, approximately 1.8 million people in Germany were affected by dementia at the end of 2021, the vast majority (1.7 million) were over 65 years of age ¹¹⁷ , and women were twice as likely to develop the disease than men. The number of newly diagnosed cases in the 65+ age group was estimated at 430,000 ¹¹⁷ . It is expected to increase to 2.8 million affected persons by 2050. At the same time, due to demographic change, the number of working-age individuals caring or paying for the care of dementia patients will decrease significantly ¹¹⁸ .

In view of this major social challenge, prevention is of particular importance. An expert consortium recently identified 12 potentially modifiable risk factors, which together explain almost 40% of all dementias ( Table 3 ). Hearing loss is the most important risk factor in middle age.

Table 3 Modifiable risk factors for the development of dementia according to ²³

Time	Risk factor	Relative risk	Attributable risk
Younger age (<45 years)	Education	1.6	7.1%
Middle age (45-65 years)	Hearing loss	1.9	8.2%
	Cranio-cerebral trauma	1.8	3.4%
	Hypertension	1.6	1.9%
	Excessive alcohol consumption (>24g/d)	1.2	0.8%
	Obesity (BMI ≥ 30)	1.6	0.7%
Higher age ( 65 Jahre)	Smoking	1.6	5.2%
	Depression	1.9	3.9%
	Social isolation	1.6	3.5%
	Physical inactivity	1.4	1.6%
	Air pollution	1.1	2.3%
	Diabetes	1.5	1.1%

*the attributable risk indicates the percentage by which one can reduce the incidence of disease if one completely eliminates the risk factor; BMI = Body-Mass-Index.

Societal changes such as improved education as well as adjustments in individual lifestyles could therefore contribute to a significant reduction in the risk of dementia and thus improve the quality of life in older age. For example, Norton et al. ¹¹⁹ estimated that even a prevalence reduction of 10-20% of each risk factor per decade could reduce the number of global Alzheimer patients by 8.8-16.2 million in 2050.

The national dementia strategy paper, adopted in 2020, seeks to address the increasing societal demands of dementia and aims to improve the lives and care of people with dementia in Germany. However, a concrete package of measures for the implementation of prevention strategies based on the above-mentioned risk factors is missing to date ¹²⁰ .

4.2 Alzheimer’s disease

Alzheimer’s disease (AD) is the most common cause of major NCD, accounting for an estimated 2/3 of all cases ¹²¹ . It is a progressive neurodegenerative disease with characteristic biological changes, primarily associated with memory impairment, leading to dementia ¹²¹ . The biological feature is the increasing deposit of β-amyloid and tau proteins in the brain of affected individuals. Approximately, 95% of the cases occur sporadically and usually after the age of 65 years (“late onset Alzheimer’s disease”, LOAD), in less than 5% of the cases, the first symptoms appear before the age of 60 years (“early onset Alzheimer’s disease”, EOAD) ¹²² . The sporadic form usually progresses slowly over years to decades, whereas more rapid courses are often observed in EOAD. The most important genetic risk factor for the sporadic disease is the so-called ApoE-4 allele of the gene for apolipoprotein E, which is involved in lipid metabolism and plays a role in amyloid deposit. For the early onset of the disease, 3 genes (presenilin-1, presenilin-2, amyloid precursor protein) have been identified so far as risk factors ¹²¹ , which occur in a familial cluster in about 1% of all AD patients. In the course of the disease, β-amyloid accumulates between the nerve cells, initially in the form of oligomers, later as amyloid plaques, leading to a disturbance of nerve cell function and the associated development of clinical symptoms. Since about 20 years, subtypes of β-amyloid can be detected in CSF and used as biomarkers for AD (Aβ ₄₂ and Aβ ₄₂ /Aβ ₄₀ ratio). In addition to extracellular amyloid deposits, intracellular deposits of defective tau proteins are typically found as neurofibrillary bundles or “tangles”. Total tau and phosphor-tau concentrations can be determined in the CSF. The first one indicates nonspecific nerve cell damage and may also be elevated in other neurodegenerative diseases or strokes. Phospho-tau (pTau), on the other hand, is significantly elevated exclusively in AD. The German S3 Dementia Guideline therefore recommends the combined measurement of Aβ ₄₂ , total tau, and pTau to differentiate neurodegenerative and other causes in unclear dementias ¹²³ .

The amyloid deposits can also be visualized by positron emission tomography (amyloid PET).

The leading clinical symptom is slowly progressive disturbances primarily of learning and memory, but also of attention as well as spatial and temporal orientation ^{121 122} . Radiologically, in addition to a general brain volume reduction, atrophy of the medial temporal lobe, especially the hippocampus, is typically found ¹²⁴ . In approximately 10% of the cases, the disease manifests with atypical symptoms such as loss of visuospatial abilities (posterior parietal atrophy, Benson syndrome) ¹²⁵ or as frontal or logopenic variants ^{126 127} , both of which resemble typical fronto-temporal dementias. Parieto-temporal metabolic disorders can be visualized by fluorodeoxyglucose PET (FDG-PET) and assist in confirming the diagnosis. Cognitive function loss is usually accompanied by neuropsychiatric symptoms, such as apathy, agitation, anxiety, sleep disturbances, and depression.

Alzheimer’s disease is nowadays understood as a continuum, as the biological processes begin years to decades before the onset of the first symptoms and result in cognitive changes as the disease progresses. Based on the biological markers, it is possible to identify patients as affected by Alzheimer’s disease already at the preclinical stage or at the stage of mild cognitive impairment (minor NCD or mild cognitive impairment, MCI).

4.2.1 Hearing loss and Alzheimer’s disease

Already in 1993, Sinha et al. ¹²⁸ reported the involvement of the auditory system in Alzheimer’s disease. Amyloid plaques and intracellular neurofibrils were detected in the medial geniculate corpus and inferior colliculus, primary auditory cortex, and auditory association areas. A functional feature of temporo-parietal changes in AD is considered to be a disturbance in auditory scene analysis, i. e., the ability to identify auditory objects – e. g., a speaker – and to follow them even in the presence of noise ^{129 130 131 132 133} . For example, Goll et al. ¹²⁹ demonstrated that Alzheimer patients were significantly worse at discriminating spectrally and temporally altered environmental sounds compared to healthy individuals with comparable peripheral auditory thresholds when non-verbal working memory was taken into account, while the ability to perceive pitch and timbre remained the same. Coeberg et al. ¹³⁴ also found significantly more auditory agnosia for environmental sounds in patients with mild Alzheimer’s disease compared to healthy individuals, with 37% of patients showing impairment in recognition and 57% in naming test sounds. The mean hearing threshold of the patients affected by agnosia was significantly higher, independent of age. This means, peripheral hearing loss in combination with Alzheimer’s pathology increases the likelihood of the occurrence of further central auditory deficits (in this study, an odds ratio of 13.75 versus healthy subjects).

Already in 1986, Uhlmann et al. ¹³⁵ described a correlation between peripheral hearing and significantly faster cognitive performance loss in AD. In a long-term study of 639 cognitively healthy individuals at study inclusion ¹³⁶ , an increase in dementia risk of 20% was shown for each 10 dB increase in mean hearing threshold. Broken down by degree of hearing loss, the hazard ratios were 1.89 for low, 3.00 for moderate, and 4.94 for severe hearing loss. A meta-analysis of 33 studies confirmed the association of peripheral hearing and cognitive function ¹³⁷ . The cognitive performance of patients with hearing loss was lower than that of hearing healthy individuals, regardless of whether the hearing loss was treated or not. Nevertheless, the difference between individuals with treated hearing loss and hearing healthy individuals was more than half. Hearing loss had a negative effect on all cognitive domains investigated (attention, processing speed, working memory, long-term memory, executive functions, semantic and lexical knowledge), but the effect size was small (accounting for 4-6% of variance).

A similar relationship has been shown for central hearing impairment. As early as 1996, Gates et al. ¹³⁸ reported a 6-fold higher risk of dementia for patients with central hearing impairment, and further large longitudinal and cross-sectional studies came to similar conclusions ^{139 140 141 142 143} . Central hearing impairment in particular has therefore been discussed as a possible harbinger of later dementia ^{133 138 140 144} . A recent meta-analysis ¹⁴⁵ concluded that although a number of subjective audiometric methods for assessing central auditory processing (including speech in noise, dichotic hearing/binaural processing, time-compressed speech) can discriminate well between normal cognitive aging and mild cognitive impairment or AD, a reliable differentiation between MCI and AD has not yet been possible. Moreover, whether in the preclinical phase of AD without cognitive impairment these investigations can contribute to an earlier diagnosis than by the currently known neurological and biological markers remains open ¹⁴⁶ .

Auditory, event-related potentials could potentially close this gap. In a study of 26 patients with a positive family history of AD, it was shown that carriers of mutations in the presenilin-1 and APP genes already show significant changes in central auditory potentials even before cognitive deficits become clinically manifest ¹⁴⁷ . The latency delay of late auditory-evoked potentials N100, P200, N200, and P300 demonstrated in this study was taken as an electrophysiological sign of slower cortical information processing. A later meta-analysis by Morrison et al. ⁹² , evaluating studies published between 2005 and 2017 on auditory-evoked potentials in patients over 60 years of age, concluded that P300 and N200 are appropriate electrophysiological markers for distinguishing normal cognitive aging, mild cognitive impairment, and AD.

4.3 Parkinson’s syndrome (PS)

Parkinson’s syndrome is the most common neurodegenerative disease after Alzheimer’s disease ^{148 149} . According to a recent epidemiological estimate based on health insurance data of 3.7 million insured persons, approximately 420,000 people in Germany were affected in 2015 ¹⁵⁰ , the standardized prevalence amounted to 511.4/100,000.

The incidence increases with higher age: while about 50/100,000 of the 65-year-old people are affected, about 400/100,000 patients are found in the age group of 85 years and older ¹⁵¹ . Due to demographic change, but also earlier detection, the number of people affected by PS in the EU is expected to increase to about 4.25 million by 2050 ¹⁵² . Parkinson’s syndrome (PS) comprises an etiologically and phenotypically heterogeneous group of disorders. In addition to idiopathic Parkinson’s syndrome (IPS, about 75% of all cases), a distinction is made between genetic forms as well as Parkinson’s syndromes in the context of other neurodegenerative diseases (atypical PS, multisystem atrophy, Lewy body-type dementia, progressive supranuclear gaze palsy, corticobasal degeneration) and symptomatic (secondary) Parkinson’s syndrome (drug-induced, posttraumatic, toxic, metabolic, inflammatory, tumor-related) ^{153 154 155 156} . In addition to the cardinal motor symptoms (akinesia/bradykinesia, resting tremor, rigor, and postural instability), a wide variety of accompanying sensory, autonomic, psychological, and cognitive symptoms may occur and significantly impair quality of life ^{157 158} . Cognitive disorders mainly affect executive functions, such as planning, anticipatory thinking, working memory, and difficulties in switching attention between different tasks.

The incidence of so-called Parkinson’s dementia is estimated in international cross-sectional studies to be between 20-44%, which corresponds to a 3-6-fold higher risk of disease for Parkinson patients compared to non-affected individuals ^{159 160} . In a German cross-sectional study of 873 patients with idiopathic Parkinson’s syndrome, 28.6% of the patients met the diagnostic criteria for dementia according to DSM-5, with the frequency increasing significantly with higher age as well as disease stage ¹⁵⁸ . The British CamPalGN study followed 142 patients newly diagnosed with IPS between 2000 and 2002 ¹⁶¹ , 46% of this population developed dementia within the 10-year follow-up period, again including age at diagnosis and disease stage as significant prognostic factors.

4.3.1 Hearing loss and Parkinson’s syndrome

Hearing loss is discussed as another non-motor accompanying symptom of PS ^{162 163 164 165 166} . Several studies have shown that hearing impaired people suffer more frequently from PS ^{162 167} . In pure-tone audiometry, predominantly high-frequency losses ^{168 169 170 171} are found that exceed the extent of merely presbycusis ^{169 172 173 174 175} . A British case-control study of 55 patients with PS and early onset (≤55 years) found unilateral or bilateral hearing thresholds deviating from the age norm in 64.7% of patients and 28% of the age- and sex-matched control group ¹⁶⁹ . No difference was found in brainstem audiometry between the two groups in this study, so the authors assumed pure cochlear involvement. The suggestion of dopamine-dependent cochlear dysfunction is supported by evidence of reduced DPOAE amplitudes that improved with levodopa substitution ¹⁷² ; in this study, DPOAE dysfunction correlated with the clinical severity of Parkinson’s syndrome. Another study group found additional significant lateral differences. Cochlear function measured by DPOAE and pure-tone audiometry was not only worse in Parkinson patients than in the control group of the same age, but also significantly more pronounced on the ipsilateral ear of motor symptoms ¹⁷⁶ .

Beyond tone audiometric changes, difficulties in the perception of rhythms and tonal differences ^{177 178} have been reported.

A number of studies have demonstrated changes in the morphology, latency, and interpeak intervals of early auditory brainstem response (ABR) in PS patients ^{168 179 180} . Similarly, reduced amplitudes and prolonged latencies of vestibular evoked potentials (VEMP) were found ^{179 181 182} . The event-related potential P3 is suitable to detect stage and progression of Parkinson’s syndrome. The subject is offered sequences of repetitive standard stimuli that are rarely interrupted by a deviant stimulus (so-called oddball paradigm). The evoked potential (P300, P3a, P3b) is dependent on attention and working memory and therefore seems to be suitable to assess the impairment of executive functions in PS ^{183 184 185 186 187} . With increasing severity, there is a reduction in amplitude as well as prolongation of latency, so that patients with and without Parkinson’s dementia can be distinguished electrophysiologically ^{188 189} .

Although auditory stimuli and music are used for the treatment of Parkinson-related gait disorders and postural instability ^{190 191 192} , the importance of auditory rehabilitation for Parkinson patients is not discussed in therapy studies.

5. Correlation of hearing loss and cognitive impairment

The importance of cognitive processes for speech comprehension, especially in challenging listening situations, is well established. Age-related deficits lead to restrictions in communication ability, social isolation and, associated with this, to psychological stress and reduced quality of life. The question of a possible causal relationship between hearing loss and reduced cognitive abilities up to manifest dementia has increasingly become the focus of scientific research in recent years (see comprehensive reviews in e. g. ^{53 146 166 193 194 195 196} ). The analysis of already published study results is complicated by the great heterogeneity of the collected data, both in terms of audiological and cognitive parameters, as well as in terms of the studied groups, recorded influencing factors, and duration of follow-up.

Usually, the pure-tone hearing threshold is used for the assessment of (peripheral) hearing loss, but already here, there are differences in the grouping of the included subjects, depending on the method used to differentiate between subjects with and without hearing loss.

On the basis of 3 long-term studies ^{136 197 198} (at least 5 years of follow-up) of subjects without cognitive impairment with tone audiometrically determined hearing threshold, the Lancet Commission ^{24 25} calculated a relative risk of 1.9 for developing dementia in the presence of hearing impairment (defined as hearing loss greater than 25 dB HL in the pure-tone audiogram) in middle age (55 years and older) compared with normal hearing subjects. Hearing loss in middle age has been identified as the most important modifiable risk factor for developing dementia.

Few studies explicitly address the relationship between central auditory disorders and dementia or cognitive deficits in old age. A meta-analysis by Dryden et al. ¹⁹⁹ identified 25 studies that investigated the relationship between cognitive performance and speech understanding in noise as a measure of central hearing impairment. For both the subset of studies that included only peripherally normal hearing subjects (16 articles) and studies that also included subjects with at most moderate hearing loss (up to 70 dB HL, 9 studies), the overall correlation (r=0.31 [normal hearing], r=0.32 [up to moderate hearing loss]) of cognitive function and speech understanding in noise was weak. Broken down by cognitive domains, the strongest correlation was seen for processing speed (r=0.39), followed by inhibition control (r=0.34), working memory (r=0.28), and episodic memory (r=0.26), whereas global measures of crystalline intelligence showed a significantly weaker correlation (r=0.18).

Wayne and Johnsrude ¹⁹⁴ state that the use of global cognitive screening tests such as the Montreal Cognitive Assessment (MoCa ²⁰⁰ ), the Mini-Mental State Test (MMST ²⁰¹ ), and the Modified Mini-Mental State Test (3MS ²⁰² ) in normal aging individuals shows little variability, and thus may underestimate the impact of hearing loss on cognitive function.

At the same time, the presence of hearing impairment may interfere with the performance in cognitive tests and lead to an overestimation of the cognitive deficit present, especially when instructions are given verbally, as shown by several studies in normal-hearing, cognitively healthy subjects with simulated hearing loss ^{203 204 205} . Therefore, special versions of cognitive screening instruments for hearing-impaired people have been developed, which should be used preferentially in the future (refer to Völter et al. ²⁰⁶ for a comprehensive overview).

5.1 Explanatory models for the interaction of hearing and cognition

In order to explain the relationship between (age-related) hearing loss and cognitive decline, a number of models are discussed, which will be briefly described below. A comprehensive review is provided by Wayne and Johnsrude ¹⁹⁴ .

5.1.1 Model 1: Cognitive load on perception hypothesis

Declining cognitive capacity places increasing load on perception so that no longer sufficient resources are available for the processing of sensory information. This leads to an audiometrically measurable hearing impairment ^{207 208} . A study by Kiely et al. ²⁰⁹ seems to confirm this theory. After analyzing longitudinal data from a total of 4221 subjects, the authors concluded that, in addition to age and hypertension, a score of less than 24 on the Mini-Mental State Test was among the independent predictors of annual hearing threshold deterioration. Ex post, it remains unclear to what extent the hearing impairment itself affected the test result, because the test used was presented verbally ( Fig 4a ).

Fig 4.

Explanatory models for the connection between age-related hearing loss and cognitive function loss: A) Cognitve load on perception hypothesis: Loss of cognitive function leads to a measurable hearing impairment via the disturbed processing of sensory information B) Information degradation hypothesis: Age-related hearing loss degrades the information available for further processing. Temporarily cognitive resources are used to compensate, which are then no longer available for other cognitive processes. This process is potentially reversible by providing hearing aids which improve the information available. C) Sensory deprivation hypothesis: The sensory deprivation associated with presbycusis leads to permanent structural brain changes and permanent loss of cognitive function D) Common cause hypothesis: Common endogenous and exogenous causes lead to both a loss of cognitive function and presbycusis.

5.1.2. Model 2: Information degradation hypothesis

This model assumes that reduced or impaired peripheral hearing triggers an upward cascade in which cognitive resources are applied to compensate for the hearing impairment, rendering them unavailable for other cognitive processes ^{207 210} . Evidence for this assumption is high; for example, several studies have shown that the ability to recall words or sentences deteriorates during a demanding perceptual experiment in elderly subjects ^{17 211} . The associated increased listening effort has negative effects on working memory and inhibition control ¹⁷ . The cognitive loss in this model is reversible – it is assumed that if peripheral input is improved, e. g., by compensating for hearing loss with hearing aids, at least partial recovery of cognitive performance is possible ( Fig 4b ).

5.1.3 Model 3: Sensory deprivation hypothesis

This model assumes that a lasting shift in resources to compensate for perceptual deficits leads to a permanent loss of cognitive function. Neuroplastic remodeling in central auditory areas and neurovascular and neurophysiological changes similar to those seen in dementia are postulated as possible mechanisms ^{106 212 213 214} . For congenital or early acquired hearing loss, the associated neuroplastic changes are already well established ^{215 216} , but cognitive performance is little affected ²¹⁷ . Sensory deprivation alone is thus insufficient as an explanatory model for cognitive loss in old age ( Fig 4c ).

5.1.4 Model 4: Common cause hypothesis

General age-related neurodegeneration processes could have negative consequences for both cognitive performance and sensory perception ^{207 208} . For example, the decrease in processing speed is discussed as one such common factor ²¹⁸ . In addition to genetic causes ²¹⁹ , cerebrovascular disease ²²⁰ and general loss of physical functioning have been considered as possible mechanisms ( Fig 4d ).

5.1.5 Multifactorial model

None of the above assumptions alone can explain all observed changes in older age; a combination of several effects is most likely. Wayne and Johnsrude ¹⁹⁴ therefore postulated a multifactorial model illustrating the interdependence of sensory and cognitive processes ( Fig 5 ).

Fig 5.

Multifactorial model of the connection between age-related hearing loss and cognitive function loss (adapted and expanded from [194]. Aging processes affect both the sensory and the cognitive system. Age-related hearing loss leads to a sensory deficit with impaired perception. Compensatory mechanisms increase access to cognitive resources which are already reduced by aging. The communication disorder resulting from the perceptual disorder promotes loneliness and social isolation, which has negative psychosocial consequences (e.g. depression) and potentially increases frailty. Cognitive performance decreases due to multiple loads.

Age-related neurodegenerative changes increase cognitive demands and, in combination with sensory deficits, lead to impaired perception. Compensation for perceptual deficits increases cognitive load, which can lead to declines in mental performance. Other sensory deficits (e. g., impaired vision or balance) amplify the impairment. The communication disorder caused by the hearing loss promotes social isolation and loneliness and with it depression and frailty – the latter being further risk factors for cognitive decline independent of hearing loss ^{53 221} .

6. Can treatment of hearing loss reduce cognitive impairment?

Due to the widespread availability of hearing aids, treatment of age-related hearing loss is perceived as an achievable target for dementia prevention. However, testing the effectiveness of such an intervention presents unique challenges. For example, in the context of an observational study, it is difficult to monitor the quality of hearing aid fitting as well as the duration of daily use. The latter is now facilitated by the possibility of data logging by the hearing aid. A recent study on datasets of more than 15,000 hearing aid users was able to objectify the considerable inter- but also intraindividual variance in daily hearing aid use ²²² . At the same time, factors such as socio-economic status, education level, social environment, communication behavior, and access to health care play a role in both hearing aid use and risk of cognitive decline, making independent assessment of the impact of hearing rehabilitation difficult. Large epidemiologic aging studies in the past have partially included hearing threshold but not systematic hearing aid use (e. g., for the German-speaking countries ²²³ ).

A multicenter, randomized-controlled longitudinal intervention study initiated in 2018 in the USA including more than 800 70-84-year-old individuals without dementia with low to moderate hearing loss comparing the efficacy of hearing aid provision with health education alone with parallel collection of audiologic data as well as cognitive performance over a 3-year period (ACHIEVE study, ²²⁴ ) intends to address the issue, but completion is not expected until late 2022 at the earliest.

Regarding the different intervention options, currently most data are found on conventional hearing aid fitting, in recent years increasingly also on cochlear implantation.

6.1 Provision of hearing aids

The Lancet Commission ²⁴ cites 3 recent studies to support the possible preventive effect of hearing aid use. A prospective study demonstrated a correlation between increased incidence of dementia in subjects with self-reported hearing loss within the 25-year observation period only if they did not use hearing aids ²²⁵ . The cross-sectional study of Ray et al. ²²⁶ also found cognitive deficits only in the subgroup of hearing impaired subjects who did not use their hearing aids, but the groups studied varied considerably in age and severity of hearing loss. The long-term study by Maharani et al. ²²⁷ found a slowing of age-related functional loss in episodic memory after the onset of hearing aid use.

In a comprehensive systematic analysis of long-term studies published between 1990 and 2020 on the relationship between hearing aid use and cognitive function ²²⁸ , the authors concluded that to date, based on the current body of studies, no definitive conclusion on the preventive effect of hearing aid use can be drawn. The methodology of the existing studies is extremely heterogeneous, of particular importance is the generally short follow-up period with regard to the rather slow age-related cognitive function loss. In addition to the aforementioned study by Maharani et al. ²²⁷ , the authors were able to identify only 1 other study in which subjects were followed-up for at least 10 years, which did not reveal any differences between intervention group (with hearing aids) and control group for any cognitive measures ²²⁹ . In addition, a common problem in comparative studies was large hearing threshold differences between intervention and control groups. Furthermore, hearing aid compliance was poorly reported or not reported at all in 9/17 studies, leaving it unclear to what extent subjects used the hearing aid adequately. The greatest potential benefit of hearing aid provision appeared to be in the area of executive function – after all, 6/11 studies found improvement ²²⁸ . Two out of 4 studies found significant improvement with hearing aid use on screening tests (MMST). However, it was not reported whether the hearing impaired version was used, so it cannot be excluded that due to hearing impairment in baseline testing, cognitive function loss was overestimated and the improvement found by using the hearing aids was only due to a better understanding of the verbally presented tasks.

6.2 Cochlear implantation

It is well established that elderly patients with severe hearing loss or deafness benefit from cochlear implantation in terms of speech understanding and quality of life (e. g., ^{230 231 232 233 234} ). Compared to normal-hearing individuals, the incoming signal is already highly degraded by the signal processing of the cochlear implant, which requires a greater input of cognitive resources to understand speech in the first place. Assuming that aging processes of the central auditory pathway affect CI recipients to the same extent as normal-hearing individuals, older CI users are at an even greater disadvantage because impaired temporal processing further deteriorates the already degraded signal ²³⁵ . As in normal-hearing individuals, working memory function affects speech comprehension ^{236 237} , and linguistic context can be used to some extent to improve speech comprehension ²³⁸ .

In recent years, a number of studies have been published explicitly addressing the alteration of (global) cognitive functions by cochlear implantation ^{239 240 241 242 243 244 245 246 247 248 249 250 251} . Similar to studies on hearing aid users, the neurocognitive test batteries chosen varied widely, although tests suitable for hearing impaired people were increasingly used ^{244 245 246 247 248 251 252} . The follow-up period was relatively short (12 months) in most studies, probably because the long-term studies in questions were initiated only in recent years. Four research groups reported results after 18 ²⁵¹ , 24 ²⁴⁶ , at least 25 ²⁴² , and 60 months ²⁴⁰ . Positive effects, especially on executive functions, were already reported within the short follow-up period. A limiting factor is the small number of cases – mostly<20 patients have been included ^{241 242 243 244 252} . The largest number of participants with simultaneous use of a neurocognitive test battery adapted for hearing impaired subjects has been studied so far by Völter et al. ^{246 247 248} . During a follow-up period of at least 24 months, 71 elderly CI patients (mean age at implantation 66.03 years) showed significant improvements in executive functions (attention, working memory, inhibition) already after 6 months compared to preoperative performance, and after 12 months, memory and word fluency had also significantly improved. After 24 months, there was an improvement in processing speed; inhibition control (flanker) was no longer significantly better, and there were no changes in mental flexibility throughout the study period. Preoperatively, the performance of 12 of the 71 subjects was below 68% confidence interval in 3 or more subtests; after 12 months, this was the case in only 3/71 subjects. By the end of the study, 5/71 subjects had deteriorated in more than 2 subtests. Cognitive performance had no significant effect on speech comprehension at rest.

A similar result was already reported by Mosnier et al. ²³⁹ in their investigation of 94 CI users aged 65-85 years: Of 37 subjects with preoperatively worse cognitive function, 81% improved within the first 12 months, and performance remained stable in 19%. Regarding dementia development, the follow-up study by the same research group is particularly interesting ²⁴⁰ . 80 subjects of the original 94 included were still alive 5 years after implantation, 70 of whom could be followed-up. Before cochlear implantation, 31 subjects had cognitive performance in the range of mild cognitive impairment. Of these, 32% recovered to normal function, 6% developed dementia, and 61% remained stable. Of the 38 subjects with preoperatively normal function, none developed dementia during the follow-up period, but in 32% of the cases, cognitive performance was in the range of mild cognitive impairment after 5 years. A correlation with the achieved speech comprehension could not be proven.

Overall, all studies published so far show a clear positive, at least stabilizing, mostly even improving effect of cochlear implantation.

7. Outlook

Sensory and cognitive deficits are closely linked via complex bottom-up and top-down processes. The consequences of both normal and pathological aging processes will inevitably pose major challenges to our society in the future. The realization that a number of risk factors can be modified already in young and middle ages offers opportunities for prevention. In particular, the consistent treatment of hearing loss must become an even greater focus of health education, also in view of the threat of social isolation and depression as further risk factors for cognitive decline, in order to increase the alarmingly low rate of care, even in industrialized countries. It is essential to take into account the special needs of the elderly, both with regard to the operation of hearing systems (fine motor requirements when changing batteries vs. using rechargeable batteries, simple operating structure/coupling with external systems) and the fitting process (possibly longer habituation phase, slower processing speed, lower differentiation acuity when comparing different settings). Appropriate compensation for the increased time required for consultation and repeated adjustment would increase the incentive for providers to devote the necessary attention to this patient group. The higher costs of care would be offset by a significantly improved quality of life and longer cognitive function preservation in the case of successful adaptation, which could lead to a reduction in the costs of care and thus to a reduction in the burden on society as a whole. To validate the success of the fitting, further long-term studies are required that record in detail both cognitive function and hearing performance as well as the type and extent of use of hearing systems and apply measurement methods that are methodologically adapted to possible cognitive and sensory deficits.