Brain structural reorganization mediates the relationship between hearing loss and cognitive impairment

doi:10.21203/rs.3.rs-4845256/v1

Download PDF

Research Article

Brain structural reorganization mediates the relationship between hearing loss and cognitive impairment

https://doi.org/10.21203/rs.3.rs-4845256/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Presbycusis is a common disease in the elderly linked to cognitive decline. Studies have shown structural alterationsin brain regions associated with hearing and cognitive function in presbycusis patients. However, the neural mechanisms linking brain structure with hearing loss and cognitive impairment remain unclear. This study investigated neuroanatomical alterations associated with auditory and cognitive functions. A total of 67 presbycusis patients and 68 normal-hearing controls underwent magnetic resonance imaging, auditory assessment, and cognitive assessment. Presbycusis patients were divided into presbycusis cognitive preservation (presbycusis-CP) and presbycusis cognitive impairment (presbycusis-CI) subgroups according to Montreal Cognitive Assessment scores. Atrophy of gray matter volume (GMV) was observed in presbycusis patients. The results indicated that GMV of the hippocampus, superior frontal cortex (SFC), and posterior cingulate cortex (PCC) mediated the connection between hearing loss and cognitive impairment. Interestingly, the hippocampus mediated the relationship between the GMV of the SFC and PCC in presbycusis and with the strongest mediating effect (56.14%) in presbycusis-CI patients. Notably, this was not valid in normal-hearing controls and presbycusis-CP patients. Overall, presbycusis patients, especially presbycusis-CI patients, show a specific reorganization pattern of auditory and cognitive-related brain areas and that the hippocampus may be a specific target for presbycusis-related cognitive impairment.

brain area reorganization

cognitive impairment

hippocampus

MRI

Presbycusis

Presbycusis, which typically involves bilateral sensorineural hearing loss at high frequencies, presents as a decline in sound information processing and challenges with phonetic perception in noisy environments (Cope et al., 2015). Approximately 60% of people over age 65 and 90% of people over age 85 suffer from presbycusis (Amieva et al., 2015), making presbycusis one of the three major chronic health diseases after arthritis and hypertension (Incidence, 2016). Previous studies have reported a correlation between hearing loss and cognitive impairment (Rabbitt, 1968, Lin et al., 2011, Ren et al., 2019). The latest report also indicates that there is a correlation between hearing loss and unhealthy factors such as cognitive decline and dementia. Middle aged and elderly patients with hearing loss are the largest potential risk factor for dementia, accounting for 8% of all dementia cases (Lin et al., 2024). For example, a longitudinal study reported that, compared with normal-hearing elderly, the risk of dementia more than doubled in elderly individuals with mild hearing loss and more than tripled in those with moderate hearing loss (Lin et al., 2011). Similarly, a systematic review and meta-analysis found that decreased cognitive ability, cognitive impairment, and increased likelihood of dementia were significantly associated with presbycusis (Loughrey et al., 2018). A 2020 Lancet Committee report identified hearing loss as a modifiable risk element for the progression of dementia (Livingston et al., 2020). Hearing loss has been identified as the most modifiable risk factor for dementia. In challenging situations, or when the auditory system is damaged, the cortical resources of the brain can be affected, which may alter the brain's response to cognitive demands (Slade et al., 2020). Despite these strong associations, the neural mechanism of cognitive impairment in presbycusis patients remains unclear.

Magnetic resonance imaging (MRI) has been extensively used to understand the pathogenesis of various neuropsychiatric disorders (Mudar and Husain, 2016, Uchida et al., 2018). Gray matter atrophy is a common structural abnormality observed in neuropsychiatric conditions and has also been reported in presbycusis. A prior MRI study of presbycusis patients found that high-frequency (Delano et al., 2020) hearing loss was negatively correlated with auditory cortical and temporal lobe gray matter volume (GMV) (Qian et al., 2017, Eckert et al., 2019). Isler et al. reported that deficient vowel processing in presbycusis is affected by changes in auditory cortex atrophy (Isler et al., 2021). A meta-analysis found an overall pattern of temporal lobe gray matter atrophy in presbycusis patients compared to healthy controls (Eckert et al., 2019). Several investigations have reported that cognitive damage in presbycusis involves other cognitive-related cortical regions. A surface-based morphometric MRI study reported reduced GMV of the precuneus and posterior cingulate cortex (PCC), which are vital components of the default network (DMN), in presbycusis patients (Ren et al., 2018). Another MRI study found that dysfunction of the cochlear amplifier in presbycusis patients was related to decreases in cognitive ability, such as disposal speed and language working memory, as well as decreases in GMV in the anterior and posterior cingulate cortex, sensorimotor area, and ParaHippocampal area of non-auditory brain regions (Belkhiria et al., 2019). In addition, an MRI study proposed several hypotheses to explain the potential relationship between hearing impairment and cognitive impairment: some hypotheses suggest that this relationship is supported by general neurodegeneration during aging, Other hypotheses suggest that there is a causal relationship between auditory impairment, sensory deprivation, and cognitive impairment. In addition to structural changes in the cortex, elderly individuals with clinically significant pure tone hearing loss also exhibit functional differences in auditory processing compared to young individuals with normal pure tone thresholds (Slade et al., 2020). Studies have also shown that a decrease in the suprathreshold auditory nerve response in presbycusis is related to a decrease in processing speed and cortical thickness in the temporal and posterior cingulate cortex (Delano et al., 2020). A functional MRI (fMRI) study indicated that age-related gray matter atrophy in the speech-responsive temporal lobe region was closely related to increased activity in the left middle frontal gyrus (MFG), suggesting that decreased GMV of the temporal lobe, which is associated with speech recognition, could result in increased reliance on the middle frontal gyrus in the cognitive control system (Eckert et al., 2008). A resting-state functional MRI (rs-fMRI) study found that a decrease in directional functional connectivity from the left inferior parietal lobe to the left hippocampus in presbycusis patients was significantly negatively related to scores on the Trail Making Test Part B (TMT-B), indicating that functional changes in the inferior parietal lobe may be associated with reduced cognitive ability in presbycusis (Chen et al., 2020).

As summarized above, numerous studies have reported changes in the structure of brain regions related to hearing or cognitive performance in presbycusis patients. However, few studies have focused on the internal relationship among the brain regions where morphological alterations occur. It is also possible that hearing loss and cognitive impairment could have a simultaneous impact on morphological alterations in brain regions (Giroud et al., 2017, Rigters et al., 2017, Belkhiria et al., 2019), which previous studies did not consider. To identify the impact of cognitive impairment on morphological alterations in brain regions, the present study divided presbycusis patients into two subgroups, namely presbycusis cognitive preservation (presbycusis-CP) and presbycusis cognitive impairment (presbycusis-CI), to help avoid potential confounding effects.

The primary purpose of this study was to explore specific brain region reorganization associated with hearing and cognition in presbycusis. To achieve this, we compared whole-brain structural differences between presbycusis patients and normal-hearing controls. All participants completed general cognitive ability and hearing assessments. To clarify the effect of cognition on brain region reconstruction in presbycusis, we divided presbycusis patients into presbycusis-CP and presbycusis-CI subgroups according to Montreal Cognitive Assessment (MoCA) scores. We hypothesized that: (1) GMV changes in brain regions may be involved in and mediate the processes of hearing loss and cognitive impairment, and (2) presbycusis patients may show specific brain structural reorganization related to both hearing and cognition.

2.1 Participants

A total of 135 older adults (59 males, 76 females) with a mean age of 62.71 ± 4.91 years participated in this study, including 67 presbycusis patients and 68 normal-hearing (NH) controls. For the presbycusis group, the inclusion criteria were pure tone average (PTA) at frequencies of 0.5, 1, 2, and 4 kHz > 25 decibels (dB) HL (Halai et al., 2015). The exclusion criteria were: (1) a history of head trauma/injury, tinnitus, seizures, and/or self-reported central nervous system diseases; (2) conductive hearing loss and/or otologic disease; (3) contraindications for MRI; (4) unprotected exposure to loud noise and/or ototoxic medications, ear surgery, and/or recent ear infections; and (5) use of a hearing aid. This study was approved by the Shandong University institutional review board (approval no. 2016-KY-059) and all participants provided written informed consent.

2.2 Auditory assessment

Auditory assessment was performed in a sound attenuation chamber. First, we performed tympanometry to assess the optimum middle ear status using a GSI Tympstar. Next, a GSI AudioStar Pro audiometer was used to measure the pure tone (PT) audiometry threshold under the standard frequencies of 0.125, 0.25, 0.5, 1, 2, 4, and 8 kHz.

Distortion product otoacoustic emissions (DPOAEs) are the otoacoustic emissions resulting from two primary pure tones with a particular frequency ratio (Adank, 2012b). The DPOAE amplitude of each 2f1-f2 frequency point was recorded by a SmartOAE (Intelligent Hearing Systems, USA) at the detected frequencies of 0.5, 0.75, 1, 1.5, 2, 3, 4, 6, and 8 kHz. The intensity of the two original pure tone signals was set as f1 = 65 dB SPL and f2 = 55 dB SPL, respectively, and the frequency ratio f1/f2 = 1.22.

The speech recognition threshold (SRT) was determined using HOPE software according to the SRT testing guidelines (Schlauch et al., 1996). The software played five spondee words at random and reduced the sound of each word by 5 dB until the subjects could no longer accurately identify the word. Each subjects' SRT was determined by subtracting the initial sound intensity from the number of words recognized during the test.

2.3 MRI data acquisition

The 3T scanner (Philips, Achieva TX) was adopted with an eight-channel phased-array magnetic head coil as a receiver. Participants were instructed to keep their eye closed and stay awake during scanning and wore foam pads and earplugs to reduce head movement and scanner noise. Three-dimensional T1-weighted TFE images were obtained as anchor points using the following scanning parameters: Repetition Time (TR) = 8.1 ms, Echo Time (TE) = 3.7 ms, thickness = 1 mm, Field Of View (FOV) = 24 cm × 24 cm, and Flip Angle (FA) = 8◦. Images were reconstructed with isotropic voxels of 1 mm × 1 mm × 1 mm.

2.4 MRI data preprocessing

Voxel-based morphometry (VBM) was used to analyze structural MRI data using Statistical Parametric Mapping 12 package (Ashburner and Friston, 2005) and CAT 12 (Ashburner, 2007) software. The preprocessing procedures included bias field correction, segmentation, spatial normalization using DARTEL templates in Montreal Neuroscience Institute (MNI) space, and registration to MNI standard space. Finally, Gaussian smoothing kernels were applied to smooth the normalized gray matter images. A two-sample t-test was used to compare GMV of NH and presbycusis groups after smoothing with age, sex, education, and total intracranial volume included as regression covariables. The cluster size was set as a minimum 20 voxels. False Discovery Rate (FDR) correction was applied for multiple comparisons. P < 0.05 indicates a statistically significant difference.

2.5 Cognitive assessment

All participants completed the following cognitive tests in the same order within 60 min: (1) MoCA, which was used to measure common cognitive function; (2) Auditory Verbal Learning Test (AVLT, Chinese version) to assess language learning; (3) Stroop Color-Word Interference Test (Stroop) as a measure of attention; (4) Symbol Digit Modalities Test (SDMT) to evaluate the information processing speed; and (5) Trail-Making Test (TMT) (components A and B) to assess cognitive processing speed and executive function. For TMT-A, participants were asked to connect 25 numbers (ranging from 1 to 25) on the test paper using sequential lines from the smallest to largest. For TMT-B, participants were asked to quickly draw an alternating line between a circle and a square on the same size paper while connecting the numbers in ascending order. Testing results were recorded according to TMT-A and TMT-B completion times.

2.6 Statistical analysis

Descriptive statistics for demographic data were computed using SPSS 26.0 (SPSS Inc., Chicago, IL, USA, 2019). The Kolmogorov-Smirnov test was used to verify the normality of the distribution of the data. The two-tailed t-test was used to calculate differences in age, education, and hearing test results between the presbycusis and NH groups. We used the χ² test to evaluate differences in gender, hypertension, diabetes, hyperlipidemia, smoking, and alcohol abuse between the presbycusis and NH groups.

The PT thresholds were divided into low- and high-frequency ranges through factor analysis (Dekeyzer et al., 2017). The corresponding frequency variables of common factors (low-frequency: 0.125, 0.25, 0.5, and 1 kHz; high-frequency: 2, 4, and 8 kHz) were determined according to the rotated component matrix (Supplementary Table A). As the DPOAE amplitudes were significantly correlated with PT thresholds, their frequency-division was in line with PT.

Pearson and Spearman correlation coefficients were used to explore the potential relationships between audiological outcomes, cognitive assessment scores, and GMV of brain regions that differed between the two groups. Brain structures significantly associated with both audiological assessments and cognitive assessment scores were selected as regions of interest (ROIs).

Mediation analysis was performed to better understand the extent of interactions. Several hearing indicators were chosen as independent variables, including high-frequency PT (H-PT), PTA, SRT, and high-frequency DPOAE (H-DPOAE). The cognitive assessment scores chosen as dependent variables included MoCA, AVLT, SDMT, Stroop, TMT-A, and TMT-B. ROIs associated both with audiological and cognitive assessment scores were selected as mediating variables. The effects of hearing indicators (independent variable) on the cognitive outcome score (dependent variable) were described using a mediation model with ROIs (mediator variables).

To investigate variation in the above mediating effect size (magnitude of the mediation effect) in presbycusis patients with different cognitive levels, the presbycusis group was subdivided into a presbycusis-CP group (MoCA ≥ 26; n = 28) group and presbycusis-CI (MoCA < 26; n = 39) group based on MoCA scale scores for further analyses. A two-sample t-test was used to compare GMV between the NH, presbycusis-CP, and presbycusis-CI groups. Pearson and Spearman correlation coefficients were computed to explore potential relationships between audiological outcomes, cognitive outcome scores, and the GMV of ROIs. Mediation analysis was then used to explore the connections between ROIs.

3.1 Participant demographic and clinical characteristics

The characteristics of the presbycusis and NH groups are summarized in Table 1. As shown, there were no significant group differences in age, sex, education, hypertension, diabetes, hyperlipidemia, smoking, or alcohol abuse. However, there were significant group differences in PTA, DPOAE, and SRT (p < 0.001). In addition, the presbycusis group performed significantly worse on several cognitive tests, including the MoCA, AVLT, SDMT, TMT-A, and TMT-B tests (p < 0.05). There was no significant group difference in Stroop test score (p = 0.074).

Table 1

Participants’ demographic and clinical data.
Characteristics	NH group	Presbycusis group	All participants	p
Characteristics	(n = 68)	(n = 67)	(n = 135)	NH vs. PC
Age	61.51 ± 4.96	62.84 ± 4.80	62.17 ± 4.91	0.118
Education	11.59 ± 3.01	10.67 ± 3.27	11.13 ± 3.17	0.093
Sex (Male/Female)	24/44	35/32	59/76	0.057
Hypertension (Y/N)	19/49	25/42	44/91	0.202
Diabetes (Y/N)	7/61	8/59	15/120	0.784
Hyperlipidemia (Y/N)	8/60	6/61	14/121	0.573
Smoking (Y/N)	4/64	4/63	8/127	0.983
Alcohol abuse(Y/N)	3/65	3/64	6/129	0.698
PTA	12.81 ± 4.66	35.58 ± 10.96	24.11 ± 14.17	< 0.001
SRT	13.12 ± 4.97	35.98 ± 13.33	24.46 ± 15.22	< 0.001
DPOAE	1.72 ± 3.74	-2.65 ± 5.90	-0.45 ± 5.38	< 0.001
MoCA	25.74 ± 3.37	23.39 ± 5.19	24.57 ± 4.51	0.002
AVLT	52.72 ± 10.99	44.87 ± 13.12	48.82 ± 12.68	< 0.001
SDMT	31.75 ± 11.26	25.12 ± 12.63	28.46 ± 12.37	0.002
Stroop	139.84 ± 35.20	152.73 ± 47.09	146.21 ± 41.87	0.074
TMT-A	61.46 ± 24.55	79.60 ± 38.08	70.46 ± 33.14	0.001
TMT-B	170.24 ± 69.00	210.25 ± 82.41	1.72 ± 3.74	0.003

Data are presented as means ± standard deviations. PTA: pure tone average in four frequencies; SRT: speech reception threshold; DPOAE: distortion product otoacoustic emission; MoCA: Montreal cognitive assessment; AVLT: Auditory verbal learning test; SDMT: Symbol digit modalities test; TMT: Trail-making test.

3.2 Group differences in GMV

Figure 1 and Table 2 indicate significant variation in GMV between the NH and presbycusis groups (FDR corrected, p < 0.05, cluster size > 20 voxels). The presbycusis group showed decreased GMV in the middle cingulate cortex, hippocampal/parahippocampal area, supplementary motor area, superior frontal cortex (SFC), medial orbital, postcentral gyrus, fusiform gyrus, lingual gyrus, insula, and inferior temporal gyrus compared with the NH group.

Table 2

The difference brain region of GMV between NH and presbycusis groups.
Brain region	MNI coordinate x y z			T-value	Cluster size
L inferior temporal gyrus	-49.5	-7.5	-34.5	3.6047	41
L fusiform gyrus	-27	-34.5	-22.5	3.6865	128
R posterior cingulate	10.5	-61.5	13.5	5.1438	3228
L hippocampus/parahippocampa regions	-18	-18	-10.5	3.8430	68
R superior frontal gyrus	10.5	33	-10.5	3.7699	214
L lingual gyrus	-4.5	-82.5	-4.5	3.4177	32
L insula	-45	3	-3	3.8027	71
L inferior frontal gyrus, orbital part	-48	13.5	15	4.4835	1542
R hippocampus	16.5	-33	0	4.2915	320
L hippocampus	-15	-34.5	1.5	4.3113	335
R insula	37.5	7.5	10.5	3.4429	51
L superior frontal gyrus	-13.5	51	31.5	3.5985	129
R middle cingulate gyrus	10.5	-31.5	36	3.8693	77
L cuneus	-7.5	-76.5	34.5	3.3990	27
L supplementary motor area	-12	-4.5	46.5	4.2202	490
L inferior parietal	-54	-27	39	3.3972	63
L middle cingulate cortex	-12	-30	45	3.4090	58
R postcentral gyrus	21	-33	60	3.9964	164

NH, normal hearing; MNI, Montreal Neurological Institute; L, left; R, right.

3.3 Association between GMV and audiological/cognitive assessments

The correlation between GMV and audiological/cognitive assessment scores is illustrated in Fig. 2A. In the presbycusis group, there was a positive correlation between the GMV of hippocampus and MoCA score (r = 0.281, p = 0.025). Furthermore, the GMV of the posterior cingulate was negatively correlated with TMT-A score (r = -0.271, p = 0.030) and positively correlated with MoCA score (r = 0.300, p = 0.016).

For all participants, the GMV of the hippocampus was negatively correlated with H-PT (r = -0.259, p = 0.003), PTA (r = -0.229, p = 0.008), SRT (r = -0.203, p = 0.020), and TMT-A (r = -0.190, p = 0.029), but positively correlated with MoCA (r = 0.238, p = 0.006), AVLT (r = 0.188, p = 0.031), and SDMT (r = 0.224, p = 0.010). Furthermore, the GMV of the SFC was negatively correlated with H-PT (r = -0.259, p = 0.003), PTA (r = -0.268, p = 0.002), and SRT (r = -0.282, p = 0.001), but positively correlated with H-DPOAE (r = 0.201, p = 0.021) and MoCA (r = 0.252, p = 0.004). The GMV of the posterior cingulate was negatively correlated with H-PT (r = -0.227, p = 0.009), PTA (r = -0.229, p = 0.008), SRT (r = -0.201, p = 0.021), and TMT-A (r = -0.189, p = 0.036), but positively correlated with MoCA (r = 0.229, p = 0.008) and SDMT (r = 0.171, p = 0.049).

3.4 Association between audiological and cognitive assessments

The correlation between audiological and cognitive assessment scores is shown in Fig. 2B. In the presbycusis group, H-PT was negatively correlated with AVLT (r = -0.346, p = 0.005) and positively correlated with TMT-A (r = 0.317, p = 0.011). PTA was negatively correlated with MoCA (r = -0.269, p = 0.032), AVLT (r = -0.308, p = 0.013), and SDMT (r = -0.261, p = 0.037), but positively correlated with TMT-A (r = 0.252, p = 0.044) and TMT-B (r = 0.306, p = 0.014). SRT was negatively correlated with MoCA (r = -0.263, p = 0.036). H-DPOAE was negatively correlated with TMT-A (r = -0.261, p = 0.037) and positively correlated with MoCA (r = 0.264, p = 0.035).

In the NH group, H-PT was negatively correlated with AVLT (r = -0.371, p = 0.002) but positively correlated with Stroop (r = 0.319, p = 0.010). PTA was negatively correlated with MoCA (r = -0.326, p = 0.008), AVLT (r = -0.485, p < 0.001), and SDMT (r = -0.292, p = 0.018), but positively correlated with TMT-B (r = 0.360, p = 0.003). SRT was negatively related to MoCA (r = -0.361, p = 0.003), AVLT (r = -0.493, p < 0.001), and SDMT (r = -0.271, p = 0.029), but positively related to TMT-B (r = 0.341, p = 0.005). DPOAE was negatively related to TMT-A (r = -0.252, p = 0.043) and TMT-B (r = -0.389, p = 0.001), but positively correlated with AVLT (r = 0.308, p = 0.013) and SDMT (r = 0.381, p = 0.002). H-DPOAE was negatively correlated with Stroop (r = -0.343, p = 0.005), TMT-A (r = -0.266, p = 0.032), and TMT-B (r = -0.383, p = 0.002), but positively correlated with AVLT (r = 0.394, p = 0.001) and SDMT (r = 0.255, p = 0.040).

For all participants, H-PT was negatively related to MoCA (r = -0.259, p = 0.001), AVLT (r = 0.414, p < 0.001), and SDMT (r = -0.258, p = 0.003), but positively related to TMT-A (r = 0.377, p < 0.001) and TMT-B (r = 0.347, p < 0.001). PTA was negatively correlated with MoCA (r = -0.336, p < 0.001), AVLT (r = 0.406, p < 0.001), and SDMT (r = -0.309, p < 0.001), but positively correlated with TMT-A (r = 0.349, p < 0.001) and TMT-B (r = 0.390, p < 0.001). SRT was negatively related to MoCA (r = -0.338, p < 0.001), AVLT (r = 0.378, p < 0.001), and SDMT (r = -0.280, p = 0.001), but positively correlated with TMT-A (r = 0.334, p < 0.001) and TMT-B (r = 0.360, p < 0.001). H-DPOAE was positively correlated with MoCA (r = 0.303, p < 0.001), AVLT (r = 0.347, p < 0.001), and SDMT (r = 0.284, p = 0.001), but negatively related to TMT-A (r = -0.356, p < 0.001) and TMT-B (r = -0.380, p < 0.001).

3.5 Regions of interests

According to the correlation analysis for GMV and auditory/cognitive assessment scores in the presbycusis group, the hippocampus, SFC, and posterior cingulate gyrus were significantly associated with hearing loss and cognitive assessment scores. Thus, the hippocampus, SFC, and posterior cingulate gyrus were selected as the three ROIs for subsequent analyses (Fig. 3).

3.6 Mediation analysis in the elderly adults

The correlation analysis indicated an association between GMV, audiological assessment, and cognitive assessment. Next, we tested the mediating effect of GMV changes on the relationship between hearing loss and cognitive impairments in elderly individuals (Fig. 4).

As shown in Fig. 4A, there was an obvious indirect influence of the hippocampus on the relationship between H-PT (β = -0.1848, 95% CI = [-0.4321, -0.0062]), PTA (β = -0.0111, 95% CI = [-0.0289, -0.0005]), SRT (β = -0.0096, 95% CI = [-0.0239, -0.0001]), and MoCA. A significant indirect effect of the hippocampus on the relationship between the SRT (β = -0.0259, 95% CI = [-0.0630, -0.0001]) and SDMT was also observed.

The analysis further revealed an indirect effect of the SFC on the relationship between the H-PT (β = -0.2013, 95% CI = [-0.4480, -0.0313]), PTA (β = -0.0134, 95% CI = [-0.0290, -0.0015]), SRT (β = -0.0128, 95% CI = [-0.0282, -0.0005]), H-DP (β = 0.0307, 95% CI = [0.0038, 0.0678]), and MoCA (Fig. 4B). There was also a significant indirect influence of the posterior cingulate on the relationship between the H-PT (β = -0.1605, 95% CI = [-0.4178, -0.0023]) and MoCA (Fig. 4C). Figure 4D indicates the indirect effect of the hippocampus on the relationship between the SFC (β = 0.1577, 95% CI = [0.0795, 0.2583]) and posterior cingulate.

3.7 Subgroup analyses

3.7.1 Group differences in GMV

As the GMV of the above three ROIs showed a significant mediation effect in the overall sample, the presbycusis group was divided into presbycusis-CP and presbycusis-CI subgroups to investigate whether brain structural remodeling was present in presbycusis patients with different cognitive levels. The demographic characteristics of the presbycusis subgroups are shown in Supplementary Table B. Both the GMV of the hippocampus and posterior cingulate were significantly decreased in the presbycusis-CI subgroup compared with the NH group. In addition, the GMV of the superior frontal cortex showed a significant decrease in the presbycusis-CI subgroup compared with the presbycusis-CP subgroup (Supplementary Table C).

3.7.2 Association between hearing loss, cognitive impairment, and GMV in subgroups

Figure 5 presents the findings for hearing loss, cognitive impairment, and GMV of ROIs for the presbycusis subgroups. In the presbycusis-CP group, TMT-B scores were positively related with the GMV of the hippocampus (r = 0.148, p = 0.038), SRT was negatively correlated with the GMV of the superior frontal cortex (r = -0.404, p = 0.045), and DPOAE was negatively correlated with the GMV of the posterior cingulate (r = -0.398, p = 0.049). In the presbycusis-CI group, L-PT was positively correlated with the GMV of the hippocampus (r = 0.370, p = 0.026), SDMT was positively correlated with the GMV of the hippocampus (r = 0.339, p = 0.043), and Stroop score was negatively correlated with the GMV of the hippocampus (r = -0.349, p = 0.037).

3.7.3 Mediation analysis between the hippocampus, SFC, and posterior cingulate

The mediation analysis indicated significant indirect effects of the hippocampus on the relationship between the SFC and posterior cingulate in presbycusis patients (β = 0.1574, 95% CI = [0.0510, 0.3053]) and the presbycusis-CI subgroup (β = 0.2001, 95% CI = [0.0117, 0.4015]). However, there was no mediation effect in the NH group or the presbycusis-CP subgroup. The magnitude of the mediation effect was 52.74% in the presbycusis group and 56.14% in the presbycusis-CI (Fig. 6). As the education level differed significantly between the two subgroups (Supplementary Table B), we performed a moderating effect analysis in the presbycusis-CI subgroup to evaluate the influence of education level on the mediation model. As shown in Supplementary Fig. 1, the moderated mediation effect was negative and the confidence interval contained 0 (β = -0.0341; 95% CI = [-0.019, 0.1487]). Therefore, we conclude that education level does not regulate the indirect effect of GMV of the superior frontal cortex on GMV of the posterior cingulate cortex via GMV of the hippocampus.

The results of the present study indicate GMV atrophy in presbycusis patients compared with NH controls in brain areas including the inferior temporal gyrus, inferior frontal gyrus, and limbic system. The results further suggest that hippocampal GMV atrophy is associated with both hearing loss and cognitive impairment in presbycusis-CI patients. In all subjects, the GMV of the hippocampus, SFC, and PCC mediated the relationship between hearing loss and cognitive impairment. Interestingly, hippocampus GMV mediated the relationship between GMV of the SFC and PCC in both the presbycusis and presbycusis-CI groups, with the strongest mediation effect in presbycusis-CI patients. However, this mediation model was not supported in NH and presbycusis-CP patients.

In addition to GMV atrophy in the inferior temporal gyrus, inferior frontal gyrus, and lingual gyrus in presbycusis patients, our results indicated GMV atrophy in the cingulate gyrus, parahippocampal regions, and hippocampus of the limbic system. Previous research has shown that the inferior temporal gyrus and inferior frontal gyrus are closely related to language comprehension (Adank, 2012b) and significant activation of the inferior temporal gyrus has been reported during auditory semantic processing (Halai et al., 2015). In presbycusis patients, the directional connection between the MFG and lingual gyrus is reduced (Xing et al., 2021), and the lingual gyrus is considered to be related to selective attention. Abnormal functional connectivity of the lingual gyrus in this study suggests that hearing loss may lead to auditory attention dysfunction. The PCC is a central part of the DMN and is crucial for regulation of attention (Gusnard and Raichle, 2001). Our finding of a close association between decreased GMV of the PCC and hearing loss is consistent with a previous SBM-based study reporting GMV atrophy in the right posterior cingulate cortex in presbycusis patients (Ren et al., 2018).

Presbycusis-Cl patients exhibited atrophy of the hippocampus and PCC, while GMV of the hippocampus was correlated with both hearing loss and cognitive impairment. The hippocampus, which is located in the temporal lobe, is a central component of the limbic system and is involved in memory formation (Lövblad et al., 2014) as well as learning and spatial navigation (Dekeyzer et al., 2017). The hippocampus receives sensory information through the entorhinal cortex and projects directly from CA1 to the auditory cortex (Wible, 2013), although it is not part of the classic auditory pathway. There is increasing evidence that GMV of the hippocampus is closely related to hearing. A cross-sectional study reported significant hippocampal shrinkage in presbycusis patients compared to older adults with normal-hearing, and that the hippocampal volume was smaller with worse hearing (Uchida et al., 2018). A longitudinal study reported significantly accelerated hippocampus atrophy in presbycusis patients, despite a higher baseline volume (Xu et al., 2019). A recent rs-fMRI study of presbycusis patients reported reduced directional functional connectivity from the left inferior parietal lobule (IPL) and right supplementary motor area (SMA) to the left hippocampus, which was negatively correlated with executive function (TMT-B) (Chen et al., 2020). In light of such evidence, the hippocampus may participate in auditory processing and hearing loss could trigger alterations in hippocampal structure and function (Zhang et al., 2022). In the present study, hippocampal GMV was linked to hearing loss and cognitive impairment in presbycusis-CI patients, but not in presbycusis-CP patients, suggesting different reorganization patterns of hippocampal volume in the two subgroups. Thus, hippocampal volume may serve as an imaging marker of early cognitive impairment in patients with presbycusis.

Several longitudinal studies of elderly populations have demonstrated that hearing loss is independently correlated with cognitive decline, with the severity of hearing loss indicating the likelihood of increasing cognitive impairment (Valentijn et al., 2005; Tay et al., 2006; Lin et al., 2013). Likewise, a significant positive relationship between hearing loss and cognitive impairment was observed in presbycusis patients and all participants in our study. Importantly, we identified that three brain regions strongly associated with cognition, namely the hippocampus, SFC, and PCC, mediated the relationship between hearing and cognitive function. Moreover, the GMV of the SFC and PCC was negatively correlated with auditory ability (Fig. 5). The SMA/Pre-SMA is situated in the posterior area of superior frontal gyrus; although it is not part of the auditory processing model, this region is significantly activated during auditory information perception and processing (Vagharchakian et al., 2012; Adank, 2012a). Speech recognition activates the superior temporal gyrus (STG) in the temporal lobe as well as the extra-auditory region, including the fusiform face area (FFA) (von Kriegstein et al., 2006). The PCC is also reported to be significantly activated when recognizing familiar and unfamiliar voices (Arnott et al., 2008). A study has shown that the PCC is not only involved in memory and navigation, but also has effective connections with the temporal pole, inferior temporal visual cortex, superior temporal sulcus cortex involved in auditory and semantic processing, and hippocampal system. This connection links it with hippocampal plot memory (T E R et al., 2022).

Strikingly, our results revealed that the GMV of the hippocampus mediated the relationship between the GMV of the SFC and PCC in both the presbycusis and presbycusis-CI groups. Previous work has shown that activity of the auditory cortex during speech listening is modulated by recording and successfully encoding the content heard into recent episodic memory. This top-down regulation of the auditory cortex may be due to both the feed words and feedback between the memory system, the system maintaining auditory attention, and the modal input cortex (Dhanjal et al., 2013). Related brain regions may also be involved in this process, such as the hippocampus (associated with episodic memory), PCC (the DMN node associated with attention), and the SMA/Pre-SMA (closely related to information processing in the dorsal auditory pathway) (Rauschecker, 2011). In individuals with mild cognitive impairment, the earliest change in nerve fibers has been observed in the medial temporal lobe (MTL), particularly the limbic system, which is closely linked to the hippocampus (Braak and Braak, 1991). The functional connectivity of the MTL with the DMN is also decreased in patients with Alzheimer's Disease (AD), and the MTL's functional connectivity becomes increasingly complex as AD progresses (Grajski et al., 2019). Based on the structural and functional characteristics of auditory brain tissue, these features endow neurodegeneration with fragility, extensive and mutually causal interactions between "edge" and "center" hearing impairments, as well as the auditory features of recent typical neurodegenerative dementia (AD, Lewy body disease, and frontotemporal dementia) (S C J J et al. 2020). Among individuals at high risk of AD, auditory information processing deficits caused by hearing loss may occur earlier than the more common cognitive decline (George et al., 2011). This finding supports our results that the top-down regulation mode is more affected when cognitive impairment is severe, resulting in the strongest mediation effect in presbycusis-CI patients, but no mediation effect in NH participants or the presbycusis-CP group. Therefore, specific reorganization patterns associated with hearing and cognition may exist among the hippocampus, SFC, and PCC in presbycusis patients, particularly presbycusis-CI patients, in neural circuits associated with hearing and cognition.

This study is subject to several limitations. First, although we observed a relationship among gray matter structure, hearing loss, and cognitive impairment in presbycusis-CI patients, we cannot draw any causal conclusions due to the cross-sectional design of this study. Second, considering the small sample size, the effect of GMV mediating the relationship between hearing loss and cognitive impairment appeared in all participants, while a significant correlation was only observed in presbycusis patients. Third, although educational level was not well matched between the presbycusis-CI and presbycusis-CP subgroups, the moderating effect analysis revealed that the influence of education level on the mediation model was not significant. Nevertheless, the size of the subgroups should be expanded in future studies.

Overall, the findings of this study have revealed that the GMV of the hippocampus, SFC, and PCC play a mediating role between hearing loss and cognitive impairment, providing further evidence for hearing loss-related brain reorganization. Importantly, we found that the GMV of the hippocampus mediated the effect between SFC and PCC GMV in presbycusis-CI patients. Thus, the hippocampus may be a vulnerable target for presbycusis-related cognitive impairment.

Acknowledgments：

We would like to thank all patients and healthy volunteers for their supportive participation in this study.

Conflict of Interest: The authors declare no conflict of interest.

Corresponding author: Fei Gao and Xiao Li, Department of Radiology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, 250021, Shandong, China. Email: [email protected], or [email protected]

Funding：

This work was supported by the National Natural Science Foundation of China (grant numbers 61701342, 81601479), Taishan Scholars Project of Shandong Province (grant number tsqn20240525), Tianjin Natural Science Foundation (grant number 19JCQNJC13100), Shandong Provincial Natural Science Foundation of China (grant numbers ZR2021MH030, ZR2021MH355), and the Academic Promotion Programme of Shandong First Medical University (grant number 2019QL023).

Conflicts of interest/Competing interests:

The author declare no competing interests.

Ethics approval:

This research protocol has been approved by the ethics committee

Consent to participate:

all participants have obtained written informed consent.

Availability of data and material：

The datasets analyzed in the current study are not publicly available. However, the datasets are available from the corresponding author upon reasonable request.

Code availability (Not applicable)

Author Contributions：

Author contributions included conception and study design (Y.W., J.L., W.L., S.L.,M.S. and F.G.), data collection or acquisition (Y.H., S.L., Y.A., M.S., F.R., and F.G.), statistical analysis (Y.W., J.L.,Y.H., S.L. and F.L.), interpretation of results (Y.W., J.L., W.L., Y.H. and S.L), drafting the manuscript work or revising it critically for important intellectual content (W.Y., J.L., W.L., W.M., L.X., Z.F., F.G. and H.W.) and approval of final version to be published and agreement to be accountable for the integrity and accuracy of all aspects of the work (All authors).

Adank P (2012a) The neural bases of difficult speech comprehension and speech production: Two Activation Likelihood Estimation (ALE) meta-analyses. Brain Lang 122:42-54.
Adank P (2012b) Design choices in imaging speech comprehension: an Activation Likelihood Estimation (ALE) meta-analysis. Neuroimage 63:1601-1613.
Amieva H, Ouvrard C, Giulioli C, Meillon C, Rullier L, Dartigues JF (2015) Self-Reported Hearing Loss, Hearing Aids, and Cognitive Decline in Elderly Adults: A 25-Year Study. Journal of the American Geriatrics Society 63:2099-2104.
Arnott SR, Heywood CA, Kentridge RW, Goodale MA (2008) Voice recognition and the posterior cingulate: an fMRI study of prosopagnosia. J Neuropsychol 2:269-286.
Ashburner J (2007) A fast diffeomorphic image registration algorithm. Neuroimage 38:95-113.
Ashburner J, Friston KJ (2005) Unified segmentation. Neuroimage 26:839-851.
Belkhiria C, Vergara RC, San Martin S, Leiva A, Marcenaro B, Martinez M, Delgado C, Delano PH (2019) Cingulate Cortex Atrophy Is Associated With Hearing Loss in Presbycusis With Cochlear Amplifier Dysfunction. Frontiers in Aging Neuroscience 11.
Braak H, Braak E (1991) Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol 82:239-259.
Chen YC, Yong W, Xing C, Feng Y, Haidari NA, Xu JJ, Gu JP, Yin X, Wu Y (2020) Directed functional connectivity of the hippocampus in patients with presbycusis. Brain Imaging Behav 14:917-926.
Cope TE, Baguley DM, Griffiths TD (2015) The functional anatomy of central auditory processing. Pract Neurol 15:302-308.
Dekeyzer S, De Kock I, Nikoubashman O, Vanden Bossche S, Van Eetvelde R, De Groote J, Acou M, Wiesmann M, Deblaere K, Achten E (2017) "Unforgettable" - a pictorial essay on anatomy and pathology of the hippocampus. Insights Imaging 8:199-212.
Delano PH, Belkhiria C, Vergara RC, Martinez M, Leiva A, Andrade M, Marcenaro B, Torrente M, Maass JC, Delgado C (2020) Reduced suprathreshold auditory nerve responses are associated with slower processing speed and thinner temporal and parietal cortex in presbycusis. PLoS One 15:e0233224.
Dhanjal NS, Warren JE, Patel MC, Wise RJ (2013) Auditory cortical function during verbal episodic memory encoding in Alzheimer's disease. Ann Neurol 73:294-302.
Eckert MA, Vaden KI, Dubno JR (2019) Age-Related Hearing Loss Associations With Changes in Brain Morphology. Trends in Hearing 23.
Eckert MA, Walczak A, Ahlstrom J, Denslow S, Horwitz A, Dubno JR (2008) Age-related effects on word recognition: Reliance on cognitive control systems with structural declines in speech-responsive cortex. Jaro-Journal of the Association for Research in Otolaryngology 9:252-259.
George A, Gates, Melissa L, Anderson, Susan M, McCurry, Patrick M, Feeney, Eric B, Larson (2011) Central auditory dysfunction as a harbinger of Alzheimer dementia. Archives of otolaryngology--head & neck surgery.
Giroud N, Lemke U, Reich P, Matthes KL, Meyer M (2017) The impact of hearing aids and age-related hearing loss on auditory plasticity across three months - An electrical neuroimaging study. Hearing Research 353:162-175.
Grajski KA, Bressler SL, Alzheimer's Disease Neuroimaging I (2019) Differential medial temporal lobe and default-mode network functional connectivity and morphometric changes in Alzheimer's disease. Neuroimage Clin 23:101860.
Gusnard DA, Raichle ME (2001) Searching for a baseline: Functional imaging and the resting human brain. Nature Reviews Neuroscience 2:685-694.
Halai AD, Parkes LM, Welbourne SR (2015) Dual-echo fMRI can detect activations in inferior temporal lobe during intelligible speech comprehension. Neuroimage 122:214-221.
Incidence GBDDI (2016) Global, regional, and national incidence, prevalence, and years lived with disability for 310 diseases and injuries, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet 388:1545-1602.
Isler B, Giroud N, Hirsiger S, Kleinjung T, Meyer M (2021) Bilateral age-related atrophy in the planum temporale is associated with vowel discrimination difficulty in healthy older adults. Hear Res 406:108252.
Lin FR, Metter EJ, O'Brien RJ, Resnick SM, Zonderman AB, Ferrucci L (2011) Hearing Loss and Incident Dementia. Archives of Neurology 68:214-220.
Lin FR, Yaffe K, Xia J, Xue QL, Harris TB, Purchase-Helzner E, Satterfield S, Ayonayon HN, Ferrucci L, Simonsick EM, Health ABCSG (2013) Hearing loss and cognitive decline in older adults. JAMA Intern Med 173:293-299.
Lin FR (2024) Age-Related Hearing Loss.The New England journal of medicine(16),1505-1512.
Livingston G et al. (2020) Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet 396:413-446.
Loughrey DG, Kelly ME, Kelley GA, Brennan S, Lawlor BA (2018) Association of Age-Related Hearing Loss With Cognitive Function, Cognitive Impairment, and Dementia A Systematic Review and Meta-analysis. Jama Otolaryngology-Head & Neck Surgery 144:115-126.
Lövblad, Schaller K, Vargas MI (2014) The fornix and limbic system. Semin Ultrasound CT MR 35:459-473.
Mudar RA, Husain FT (2016) Neural Alterations in Acquired Age-Related Hearing Loss. Front Psychol 7:828.
Qian ZJ, Chang PD, Moonis G, Lalwani AK (2017) A novel method of quantifying brain atrophy associated with age-related hearing loss. Neuroimage Clin 16:205-209.
Rabbitt PM (1968) Channel-capacity, intelligibility and immediate memory. Q J Exp Psychol 20:241-248.
Rauschecker JP (2011) An expanded role for the dorsal auditory pathway in sensorimotor control and integration. Hear Res 271:16-25.
Ren F, Ma W, Li M, Sun H, Xin Q, Zong W, Chen W, Wang G, Gao F, Zhao B (2018) Gray Matter Atrophy Is Associated With Cognitive Impairment in Patients With Presbycusis: A Comprehensive Morphometric Study. Front Neurosci 12:744.
Ren FX, Luo JF, Ma W, Xin Q, Xu L, Fan ZM, Ai Y, Zhao B, Gao F, Wang HB (2019) Hearing Loss and Cognition Among Older Adults in a Han Chinese Cohort. Frontiers in Neuroscience 13.
Rigters SC, Bos D, Metselaar M, Roshchupkin GV, de Jong RJB, Ikram MA, Vernooij MW, Goedegebure A (2017) Hearing Impairment Is Associated with Smaller Brain Volume in Aging. Frontiers in Aging Neuroscience 9.
S C J J ,R C M ,S R W (2020) Hearing and dementia: from ears to brain..Brain : a journal of neurology(2).
Schlauch RS, Arnce KD, Olson LM, Sanchez S, Doyle TN (1996) Identification of pseudohypacusis using speech recognition thresholds. Ear Hear 17:229-236.
Slade K ,Plack J C ,Nuttall E H (2020) The Effects of Age-Related Hearing Loss on the Brain and Cognitive Function.Trends in Neurosciences(10),810-821.
Tay T, Wang JJ, Kifley A, Lindley R, Newall P, Mitchell P (2006) Sensory and cognitive association in older persons: findings from an older Australian population. Gerontology 52:386-394.
T E R ,Sylvia W ,Gustavo D (2022) The human posterior cingulate, retrosplenial, and medial parietal cortex effective connectome, and implications for memory and navigation.Human brain mapping(2),629-655.
Uchida Y, Nishita Y, Kato T, Iwata K, Sugiura S, Suzuki H, Sone M, Tange C, Otsuka R, Ando F, Shimokata H, Nakamura A (2018) Smaller Hippocampal Volume and Degraded Peripheral Hearing Among Japanese Community Dwellers. Frontiers in Aging Neuroscience 10:319.
Vagharchakian L, Dehaene-Lambertz G, Pallier C, Dehaene S (2012) A temporal bottleneck in the language comprehension network. J Neurosci 32:9089-9102.
Valentijn SAM, van Boxtel MPJ, van Hooren SAH, Bosma H, Beckers HJM, Ponds R, Jolles J (2005) Change in sensory functioning predicts change in cognitive functioning: Results from a 6-year follow-up in the Maastricht Aging Study. Journal of the American Geriatrics Society 53:374-380.
von Kriegstein K, Kleinschmidt A, Giraud AL (2006) Voice recognition and cross-modal responses to familiar speakers' voices in prosopagnosia. Cereb Cortex 16:1314-1322.
Wible CG (2013) Hippocampal physiology, structure and function and the neuroscience of schizophrenia: a unified account of declarative memory deficits, working memory deficits and schizophrenic symptoms. Behav Sci (Basel) 3:298-315.
Xing C, Chen YC, Tong Z, Xu W, Xu JJ, Yin X, Wu Y, Cai Y (2021) Aberrant brain functional hubs and causal connectivity in presbycusis. Brain Imaging Behav 15:453-463.
Xu W, Zhang C, Li JQ, Tan CC, Cao XP, Tan L, Yu JT, Alzheimers Dis N (2019) Age-related hearing loss accelerates cerebrospinal fluid tau levels and brain atrophy: a longitudinal study. Aging-Us 11:3156-3169.
Zhang L, Wang J, Sun H, Feng G, Gao Z (2022) Interactions between the hippocampus and the auditory pathway. Neurobiology of Learning and Memory 189.

No competing interests reported.

Download PDF

Reviews received at journal
16 Sep, 2024
Reviewers agreed at journal
23 Aug, 2024
Reviewers invited by journal
22 Aug, 2024
Editor assigned by journal
22 Aug, 2024
Submission checks completed at journal
05 Aug, 2024
First submitted to journal
01 Aug, 2024

You are reading this latest preprint version

Brain structural reorganization mediates the relationship between hearing loss and cognitive impairment

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

2.1 Participants

2.2 Auditory assessment

2.3 MRI data acquisition

2.4 MRI data preprocessing

2.5 Cognitive assessment

2.6 Statistical analysis

Results

3.1 Participant demographic and clinical characteristics

3.2 Group differences in GMV

3.3 Association between GMV and audiological/cognitive assessments

3.4 Association between audiological and cognitive assessments

3.5 Regions of interests

3.6 Mediation analysis in the elderly adults

3.7 Subgroup analyses

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1