Neuroanatomical substrate of noise sensitivity

&NA; Recent functional studies suggest that noise sensitivity, a trait describing attitudes towards noise and predicting noise annoyance, is associated with altered processing in the central auditory system. In the present work, we examined whether noise sensitivity could be related to the structural anatomy of auditory and limbic brain areas. Anatomical MR brain images of 80 subjects were parcellated with FreeSurfer to measure grey matter volume, cortical thickness, cortical area and folding index of anatomical structures in the temporal lobe and insular cortex. The grey matter volume of amygdala and hippocampus was measured as well. According to our findings, noise sensitivity is associated with the grey matter volume in the selected structures. Among those, we propose and discuss particular areas, previously linked to auditory perceptual, emotional and interoceptive processing, in which larger grey matter volume seems to be related to higher noise sensitivity. HighlightsNoise sensitivity is related enlarged primary auditory areas in the left hemisphere.The bilateral hippocampus and temporal pole are enlarged in noise sensitivity.The volume of the right anterior insula is increased in noise sensitivity.Noise sensitivity is related to the morphology of auditory‐limbic brain areas.


M A N U S C R I P T
A C C E P T E D ACCEPTED MANUSCRIPT emotional valence. A sustained amygdala activation can be evoked by unpredictable auditory 85 stimulation, and this activation is coupled with anxiety-like behaviours (Herry et al., 2007). 86 The hippocampus is also involved on auditory information processing and contributes to 87 sensory gating, which is an inhibition of irrelevant, repetitive sensory input (Cromwell et al., 88 2008). Moreover, the amygdala-hippocampal complex displays a unidirectional coupling 89 during processing of emotionally important stimuli, so that amygdala detects a stimulus' 90 salience and then influences dynamics of the hippocampal response to it (Zheng et al., 2017). 91 In turn, hippocampus-dependent memory representations of stimulus emotional significance 92 can influence amygdalar function (Phelps, 2004). 93 Both the hippocampus and amygdala have rich connections with auditory areas of the brain. 94 Amygdala receives inputs from the auditory cortex and less processed information directly 95 from the thalamus. Through its connections to the inferior colliculus, the amygdala may 96 potentially influence the processing of an auditory stimulus even before it reaches the cortex 97 (Marsh et al., 2002). The hippocampus, in turn, does not have direct connections with the 98 primary and secondary auditory cortical areas (Mohedano-Moriano et al., 2007), but it is 99 largely interconnected with auditory associative areas either directly or via pathways coming 100 through the amygdala, insula, and other cortical areas, such as the temporal pole (Pascual et 101 al., 2015). The hippocampus responds to sounds or the sound deprivation (e.g., in hearing 102 loss) with the neuroplastic changes in its functional and structural organization (Kraus and 103 Canlon, 2012). Moreover, the volume of amygdala and hippocampus is known to decrease in 104 chronic stress (Abdalla and Geha, 2017), and small hippocampus is predictive for pathological 105 stress responses (Gilbertson et al., 2002). In relation to noise sensitivity, an increase and 106 decrease in amygdalar and hippocampal volumes could be expected alike. A larger volume of 107 these structures could indicate increased activation of amygdala during sound processing, M A N U S C R I P T

A C C E P T E D ACCEPTED MANUSCRIPT
from which an enlargement of both amygdala and hippocampus could follow, as they are 109 functionally tight. In turn, a decrease in volume of hippocampus and amygdala could result 110 from emotional stress noise-sensitive people experience in response to noises. 111 In addition to the auditory cortex, amygdala, and hippocampus, an important role in stimulus 112 evaluation is played by insula. A recent study found that the insula is related to symptoms of a 113 distress caused by tinnitus but not to the characteristics of tinnitus itself, such as its loudness 114 (Leaver et al., 2012). Further, in misophonia (an affective disorder characterized by negative 115 emotions towards specific sounds, such as chewing or swallowing) the activation of bilateral 116 anterior insula increased parallelly with higher subjective misophonic distress caused by a 117 triggering sound (Kumar et al., 2017). Other studies propose that anterior insula is involved in 118 anticipation of aversive bodily states and negative emotions (Phelps et al., 2001). Moreover, 119 insula, along with the amygdala and the hippocampus, can have an influence on autonomic 120 functions. Shepherd and colleagues (2016) observed differences in the dynamics of heart rate 121 in response to emotional stimuli and heart rate variability between noise-sensitive and noise-122 resistant groups. Changes in heart rate serve as indices of noise sensitivity affecting 123 integration between central and autonomic nervous systems (Thayer and Lane, 2000). Hence, 124 we expected that noise sensitivity could be related to the structure of the insular cortex that is 125 involved in regulating autonomic functions and plays a major role in the interoceptive feeling.

126
As the insular cortex and, specifically, its anterior part was found enlarged in relation to 127 distress caused by sound sensitivities, such as tinnitus and misophonia (Leaver et al., 2012), 128 we could expect the same pattern of structural change to occur in relation to noise sensitivity.

129
In this study, we measured grey matter volume and morphology (cortical area, cortical 130 thickness and cortical folding) in selected regions of interest from both cerebral hemispheres, M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT automatic parcellation and labelling of cortical and subcortical structures (Dale et al., 1999;133 Fischl et al., 1999). These measures were used to explore whether noise sensitivity is related 134 to changes in the brain morphology and what the direction of that relationship is.

137
The experimental procedure for this study was included in the research protocol "Tunteet"

146
From the "Tunteet" dataset we selected those participants that underwent anatomical MR 147 scanning and whose images were successfully parcellated with FreeSurfer (N=121). Two of 148 them were excluded from the analysis due to brain abnormalities detected by a 149 neuroradiologist. Thirty-eight subjects decided not to complete online questionnaires (see the 150 section below), and thus their data could not be studied. Additionally, one participant was an 151 outlier with more than three standard deviations lower NSS than the mean and was excluded 152 from the analysis. The final set consisted thus of 80 participants: 39 males and 41 females 153 with an age range from 19 to 52 years (Mage = 28.8; SD = 7.8).

M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT
Noise sensitivity was assessed using the Weinstein's Noise Sensitivity Scale (Weinstein, 156 1978). The questionnaire consists of 21 statements to rank on a 6-point Likert scale ranging 157 from "agree strongly" to "disagree strongly". Fourteen items were reverse-scored. The total 158 sum represents noise sensitivity score (NSS), and a higher score corresponds to higher Volumes of each ROI were proportionally adjusted for the intracranial volume to control for 184 differences in head size. Cortical thickness of each ROI was corrected for mean cortical 185 thickness. We took into consideration that age is known to decrease volume, thickness, 186 surface area, and folding of cortical structures (Lemaitre et al., 2012;Thambisetty et al., 2010;187 Toga et al., 2011). Moreover, in our data age positively, but non-significantly, correlated with 188 NSS (r = 0.207, P = 0.066). According to that, age was included in the statistical models to 189 assure that the observed effects are not explained by age differences.

190
To test the effect of noise sensitivity on each morphological measure, we first applied a

M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT
Thus, the analysis revealed that GM volume, but not cortical thickness, folding or area, was 216 significantly affected by noise sensitivity. Hence, we focused our further analysis on 217 investigating the relationship between NSS and cortical anatomy in each of the ROI using GM 218 volume measures only. For that, we applied partial correlations controlling for the effects of Grey matter volume is corrected for subjects' age and the intracranial volume. Noise sensitivity score is corrected for age. P-values are uncorrected. LH -left hemisphere; RHright hemisphere; Amg -amygdala.

232
This study aimed to explore morphological markers associated with noise sensitivity. We 233 focused our research on the brain areas involved with auditory processing, attributing 234 emotions to sounds, detecting their salience and regulating bodily functions in response to 235 auditory events. Our data suggest that noise sensitivity is related to changes in GM volume M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT over the selected areas. In particular, we propose that higher noise sensitivity may be related 237 to enlarged GM volumes in the bilateral temporal pole, the left Heschl's sulcus, the right 238 anterior insula, and bilateral hippocampus. However, we point out that the observed 239 associations did not survive a correction for multiple comparisons and are only suggested as 240 candidate areas for an involvement with noise sensitivity. The potential roles of the left 241 Heschl's sulcus, the right anterior insula, as well as the bilateral hippocampus and temporal 242 pole in noise sensitivity are further discussed. 243 We expected to observe noise sensitivity-related changes to the volume of the auditory cortex  The temporal pole was another structure that we found to be potentially associated with 267 noise sensitivity in both hemispheres. The temporal pole is an anterior-most part of the 268 temporal lobe. It is thought to belong to the paralimbic brain and is attributed with multiple 269 cognitive functions, one of which is an integration of higher-order processed stimuli and GM volume and noise sensitivity in our study. Besides the auditory system, the temporal pole 278 receives input from visual and olfactory systems and serves as a structure of sensory-279 emotional coupling for these modalities as well (Olson et al., 2007). The activation of the 280 temporal pole induced by auditory, visual, or olfactory information seems to follow a 281 dorsal/ventral segregation with auditory stimuli activating its dorsal part (Olson et al., 2007).

282
However, the parcellation approach used in our study did not allow us to determine more 283 precisely which part of the temporal pole was specifically enlarged. Some studies report that M A N U S C R I P T

A C C E P T E D ACCEPTED MANUSCRIPT
noise sensitivity overlaps with other environmental sensitivities, such as odour intolerance, 285 and it is debated whether they are concomitant or independent (Shepherd et al., 2015).

286
Further investigation of the structure of the temporal pole and its functional involvement with 287 sensory intolerances could be beneficial for understanding whether environmental 288 sensitivities are specific to a single sensory domain.

289
Noise sensitivity was positively associated with the volume of the left and right hippocampus. and right hippocampus and noise sensitivity score, we may speculate that noise sensitivity is 302 related to the ability to form the associations between negative emotional experience and 303 noise. 304 We predicted that noise sensitivity could be related to the morphology of the insular cortex.  (Barrett et al., 2004;Pollatos et al., 2007). 333 Hence, based on the observation of a larger volume of the right anterior insula in association 334 with noise sensitivity, we may speculate that noise-sensitive individuals might have an 335 increased awareness of their inner state and as a consequence might react stronger to the 336 stress effects caused by noise. This would lead them to exhibit more negative attitudes 337 towards noise than resistant individuals do. However, these anatomy-based speculations 338 should be followed up by studies on bodily awareness in noise-sensitive individuals.

339
The amygdala could be one of the structures that are associated with noise sensitivity based  We are tempted to conclude that the differences in the brain morphology related to noise 361 sensitivity are use-dependent. However, we cannot rule out a potential contribution of genetic 362 factors. Perhaps, noise-sensitive individuals are born with a predisposition for larger volumes 363 of the primary auditory cortex, anterior insula, and hippocampus, leading them to be more 364 prone to evaluate aversively environmental (auditory) stimuli. Noise sensitivity has 365 previously been shown to aggregate in families, and twin analyses provided an estimate of 366 heritability of 36% (Heinonen-Guzejev et al., 2005). Moreover, in a rare genetic disorder 367 called Williams syndrome, in which noise sensitivity is often comorbid, there is a structural 368 and functional augmentation of the left auditory cortex that cannot be explained by training 369 but by genetics (Wengenroth et al., 2010). Hence, at least in a clinical population, it is possible 370 that structural brain differences are pre-existent. Whether this could be the case for noise-371 sensitive but healthy individuals is a question requiring further investigation.

372
Taken together, in our exploratory study we propose that based on the observation of a 373 change in the GM volume, several brain structures should be investigated further for their role 374 in noise sensitivity. Namely, we suggest that enlargements in the left Heschl's sulcus, bilateral 375 temporal pole, right anterior insula as well as bilateral hippocampus could be related to high 376 noise sensitivity. We call for confirmatory investigations. Another interesting direction for 377 future research is to address whether anatomical and functional connections between these 378 brain areas are affected in noise sensitivity.