Preliminary Study on Speech in Noise Training in Children with Sensory Processing Disorder and Hyperacusis

Abstract

Hyperacusis is the perception of certain everyday sounds as too loud or painful. Past research suggests that some individuals with Sensory Processing Disorder (SPD) may also have a comorbid hyperacusis. The aim of this preliminary study was to explore if hyperacusis symptoms in children with SPD change following Speech in noise training (SPINT). This was a retrospective cross-sectional study. Data were included for 28 children with SPD and sound intolerance (12/28 were female, mean age was 8.7 ± 1.9 years old). Patients were assessed using the Persian Buffalo Model Questionnaire-Revised version (P-BMQ-R) that measures various behavioural aspects of auditory processing disorder and word in noise test (WINT) before and after SPINT. After SPINT the subscales of DEC, TFM with its Noi, and Mem, subcategories, APD, ΣCAP, and Gen of P-BMQ-R questionnaire significantly improved (P < 0.05), however, the changes in subscales of Var, INT and ORG were not statistically significant (P > 0.05). In addition, SPINT led to better performance in WINT in both ears (P < 0.05). This preliminary study showed promising result for the effect of SPINT on improving behavioural indicators of APD (as measured via P-BMQ-R and WINT) and decreasing hyperacusis symptoms (as measured via Noi).

Introduction

Sensory processing is the ability to manage inputs of sensory systems (e.g. visual, auditory, vestibular, proprioceptive) to facilitate congruous responses within the environment [1]. Individuals with SPD, have difficulties in detecting, modulating, interpreting or organizing sensory stimuli as stated in the most recent version of the diagnostic classification of mental and developmental disorders of infancy and early childhood-revised [2]. SPD is prevalent among individuals with Autism Spectrum Disorder (ASD) [3–5], Attention Deficit Hyperactivity Disorder (ADHD) [6, 7], Learning Disability (LD) [8, 9] and Auditory Processing Disorder (APD) [10]. The prevalence of SPD is estimated to be 5–16% of school-aged children [11] and 60–90% of children with neurodevelopmental conditions [12, 13] as well as 90% of individuals with ASD [14–17]. Over-responsivity for auditory stimuli, a concept similar to hyperacusis, can be viewed among individuals with sensory modulation disorder which is a main component of Sensory Processing Disorder (SPD) [18].

Hyperacusis is defined as an intolerance of certain everyday sounds, which are perceived as too loud or uncomfortable and can cause distress and difficulties in the individual’s daily activities [19] including speech perception in noise (SPIN) problems [20]. There are different theories which explain hyperacusis is caused by increased neural synchrony of the tonotopic structure of the auditory cortex as well as problems with the limbic system and auditory pathways [21–24]. In pediatric population, the estimated prevalence varies between 3.2% [25] and 17.1% [26].

Several procedures are used to evaluate hyperacusis including measurement of uncomfortable loudness levels (ULLs) and questionnaires as well as behavioural observation strategies include clinical interviews with children and their parents [25, 27–32]. Self-report questionnaires are used to evaluate hyperacusis which very few of them are available in the literature about childhood hyperacusis [18, 33]. There are different methods for management of hyperacusis based on psychological and auditory approaches. Cognitive Behavioural Therapy (CBT) is a psychological intervention which teaches the patient skills and knowledge required to modify their noise-related thoughts, emotional reaction and safety-seeking behaviours [34]. Some of the auditory-based treatment methods comprise Auditory Integration Training (AIT) [35] and sound therapy [27]. These methods aim to desensitize the auditory system [36].

Hyperacusis can cause listening difficulties in background noise [20]. Speech in noise training (SPINT) is one of the methods of improving speech perception in noise (SPIN) by practicing listening to speech sounds, understanding, memorizing and repeating them in various and gradually increased exposure to background noise. It was found that SPINT can be improved SPIN by increasing neural firing, improve timing representation and optimization of up-down pathways [37–39]. SPINT leads to the improving of SPIN by involving auditory pathways and since hyperacusis is caused by problems within auditory pathways [21–24] and also SPIN problems exists in people with hyperacusis, SPINT may be effective on it. However, the effects of this intervention on hyperacusis symptoms have not been investigated in previous studies. The aim of this preliminary retrospective study was to explore if hyperacusis symptoms in children with SPD change following SPINT.

Methods

Ethical Approval

Patient consent was waived as this was a retrospective analysis of available clinical data. This study was approved by the Ethics Committee of The Hamadan University of Medical Science (IR.UMSHA.REC.1401.678).

Participants

Data were imported from records of children who underwent routine assessment and management for auditory processing issues in Private Ordibehesht Audiology clinic in Tehran, Iran, including results of their audiological investigations, the outcomes of questionnaires (filled by their parents) and reports of their intervention sessions. The data for a total of 28 children (12 females and 16 males), mean age 8.7 ± 1.9 years old, diagnosed with SPD who were seen between June 2021 and March 2022 due to listening difficulties in noise and hyperacusis were included. They were referred from the occupational therapy clinic. Edinburgh Handedness Inventory was used as a screening tool for handedness and Intelligence Quotient (IQ) of children was measured with modified Persian version of the Wechsler Intelligence Scale for Children [40].

Procedure

All children took part in an auditory processing assessment session that lasted for about 1 h. Assessment comprised: (1) interviewing with parents about auditory processing difficulties and hypersensitivity to sound symptoms of their children. In interview with parents, the relationship that their children have with sounds was investigated by asking them do you think that your child is too sensitive to every day’s sounds? Is there any sound that your child dislikes? Is there any sound that your child considers painful? Is there any sound that scares your child [25]?. (2) Pure Tone Audiometry (PTA) (3) Filling out Persian Buffalo Model Questionnaire-Revised version (P-BMQ-R) [41] by parents. (4) Word In Noise Test (WINT) [42]. After that, intervention was delivered based on Speech In Noise Training (SPINT) [43]. All the tests were performed in an acoustic insulated chamber with a maximum ambient noise at the level of < 30 dBA. The stimuli in WINT and SPINT were provided by Acer Nitro5 AN515 Laptop and Audacity Software and Sennheiser HD 280 Pro Monitor Headphones. The intensity levels of the stimulus were calibrated using a Bruel & Kjaer 2250 L sound level meter and a 4153 D&K artificial ear attached to a weight or a headphone adapter. The tests and intervention performed at the most comfortable level (MCL) with the presentation level of 55–60 dB SPL [44].

Persian Buffalo Model Questionnaire-Revised Version (P-BMQ-R)

To investigate behavioural aspects of auditory processing, Persian Buffalo Model Questionnaire-Revised version (P-BMQ-R) [41] was used. The P-BMQ-R has 48 questions and its subscales comprise: [1] Decoding (DEC), [2] Tolerance Fading Memory (TFM) (with its three subcategories: Noi; noise, Mem; memory, Var; various), [3] Integration (INT), [4] Organization (ORG), [5] non- specific Auditory Processing Disorder score (APD) that are not specific to any one of the four main subscales (subscales 1 to 4). The sum of the subscales 1 to 5 is denoted Central Auditory Processing (ΣCAP) [6]. The last subscale which is not included to the total score is General (Gen) which is a group of general behaviors that may or may not be associated with APD.

According to the authors’ knowledge, currently there is no validated Persian questionnaire for investigating of hyperacusis in children. Therefore, in this study, the Noi subcategory ( in TFM category) was used as a measure of hyperacusis. It has 4 questions which 3 of them assess symptoms of hypersensitivity to sound, distractibility especially in noise and speech understanding in noise [41]. In addition of first question which is related to hyperacusis, the second and third questions are mentioned in Parent Hyperacusis Questionnaire (P-HQ) [33] and Hyperacusis Impact Questionnaire (HIQ) [19].

Each item has three response choices: “Yes” (1 point) which indicates that this may be a problem for the child, “No” (0 point) indicates that this is not a problem for the child, and “Non-Appropriate” (N/A) (Subtracting one point from the total points of associated category) indicates that the question is not appropriate for the child. P-BMQ-R questionnaire is similar to the original version [45].

Word In Noise Test (WINT)

WINT has two lists of 35 recorded monosyllable words. In this test, level of background multi-talker-babble noise is fixed, and five words are presented at seven signal-to-noise ratios from 24 to 0 dB in 4 dB decrements. Each list is scored based on the number of correct responses at each speech in noise ratios (SNRs) and finally the SNR 50% score is calculated [42]. Persian version of WINT was used in this study which is executed and calculated similarly [46].

Speech In Noise Training (SPINT)

The intervention sessions were done in a quiet room. SPINT with white noise [43] was used in intervention sessions (two 45-minute sessions per week for 12 weeks). Training tasks included recorded mono-syllable word [47], double-syllable word and sentences [38] in different level of noise. It began with monosyllabic words and low level of noise, and gradually, with improvement of child’s performance (seven correct answers out of ten items), the level of exercises became more difficult. The different SNR levels in training included + 20 dB, + 10 dB, + 6 dB, + 4 dB, + 2 dB, 0 dB, − 2 dB and − 4 dB [39]. Then double-syllable word and sentences were presented in different levels of noise. In total, the exercises consisted of twenty lists with thirty items in each list.

Data Analysis

Descriptive statistics (means and standard deviations) were used to describe the sample and summarize the data. The data was screened for normality using kolmogove- smirnov statistics. According to the normality of data, paired t-test was used to compare P-BMQ-R and WINT scores before and after SPINT. Effect size is a statistical analysis that identifies the size of the improvement seen between the pre- and post-intervention tests based on the number of standard deviations difference. Using Cohen’s d to calculate the effect size for each of the subscales of P-BMQ-R and scores of WINT, results would indicate how much improvement was made after subjects completed SPINT. Effect sizes of 0.2 or less are considered small improvements, between 0.2 and 0.5 are medium improvements, between 0.5 and 0.8 are large improvements, and those above 0.8 are very large improvements [48]. Negative effect sizes indicate that the mean of post-treatment scores is less (better) than pre-treatment scores. Data was analysed using the SPSS package (version 23.0). Differences were considered to be statistically significant if p < 0.05.

Results

In interview with parents, it was revealed that all of children were too sensitive to every day’s sounds and there were some sounds which children disliked, considered painful and frightening.

Their IQ was over 85 and all children were right-handed. There was no history of head trauma, epilepsy, seizures, and no use of neuroleptics. All children were native Farsi (Persian) speakers. For all children, hearing thresholds were ≤ 20 dB at frequencies of 250, 500, 1000, 1500, 2000, 3000, 4000, 6000 and 8000 Hz.The results of P-BMQ-R showed all of children had abnormal scores in DEC, TFM (and its subcategories of Noi, Mem, and Var,), APD, ΣCAP and Gen subscales. Also, about 86% and 67% of children had abnormal scores in INT and ORG subscales, respectively.

Table 1 shows pre and post comparison of the scores on P-BMQ-R and its subscales. The results indicate that following SPINT there seemed to be improvements in the scores of certain subscales of P-BMQ-R comprising DEC, TFM (and its subcategories of Noi and Mem), APD, and Gen (p < 0.05). The sum score of subscales 1 to 5, ΣCAP, also improved. This indicates that SPINT may be effective in improving decoding and phonetic skills, auditory short-term memory, following/performing instructions and listening skills especially in the presence of noise as well as reducing hypersensitivity to sound. However, subscales scores of INT and ORG and the subcategory of Var in TFM subscale did not change significantly (p > 0.05). This indicates that SPINT may not be effective in improving dichotic listening and auditory sequencing in this population.

Table 1.

Comparison of the score on subscales of P-BMQ-R and its sum pre and post SPINT training using paired t-test in SPD children (n = 28)

Variable		Pre	Post	t	df	p value
		Mean ± SD	Mean ± SD
DEC		2.96 ± 1.29	2.07 ± 1.08	6.89	27	< 0.001
	Noi.	3.32 ± 0.66	0.75 ± 0.44	23.75	27	< 0.001
TFM	Mem.	2.75 ± 1.26	2.07 ± 1.01	4.38	27	< 0.001
	Var.	2.07 ± 0.97	1.96 ± 0.99	1.80	27	0.083
TFM (Totally)		8.14 ± 2.17	4.82 ± 1.82	19.42	27	< 0.001
INT		1.96 ± 1.10	1.89 ± 1.06	1.44	27	0.161
ORG		0.92 ± 0.76	0.89 ± 0.73	1.00	27	0.362
APD		2.32 ± 0.98	1.92 ± 0.89	3.03	27	0.005
ΣCAP		16.28 ± 4.90	11.64 ± 4.27	14.79	27	< 0.001
Gen.		3.64 ± 0.95	3.50 ± 1.03	2.12	27	0.043

DEC; decoding, Noi.; noise, Mem.; memory, Var.; various, TFM; tolerance fading memory, INT; integration, ORG; organization, APD; auditory processing disorder, CAP; central auditory processing, Gen.; general, P-BMQ-R; Persian Buffalo Model Questionnaire-Revised version, SPD; Sensory Processing Disorder

The scores of WINT significantly decreased in both ears (p < 0.001) after SPINT (Table 2). It seems that SPINT is effective in improving speech understanding in noise conditions.

Table 2.

Word In Noise Test results, in children with sensory processing disorder (n = 28)

Variable	Pre	Post	t	df	p value
	Mean ± SD	Mean ± SD
Right Ear Score in WINT	1.72 ± 0.32	1.12 ± 0.26	15.60	27	< 0.001
Left Ear Score in WINT	1.75 ± 0.31	1.13 ± 0.27	17.81	27	< 0.001

According to the results of Cohen’s d, in all SPD children, SPINT had very large effect on TFM and its Noi, subcategory as well as ΣCAP indicators in P-BMQ-R. Also, it had large effect on DEC subscale and Mem subcategory of TFM as well as medium effect on APD indicators. In addition, SPINT had small effect on Var, subcategory of TFM, INT, ORG and Gen categories of P-BMQ-R (Table 3).

Table 3.

Results of Cohen’s d measures for effect size for each of the subscales of P-BMQ-R in SPD children (n = 28)

Variable		Effect Size dCohen	Confidence Interval for dCohen
DEC		-0.748	-1.514 to 0.018
	Noi.	-4.582	-5.992 to -3.172
TFM	Mem.	-0.596	-1.353 to 0.162
	Var.	-0.112	-0.854 to 0.629
TFM (Totally)		-1.658	-2.516 to -0.799
INT		-0.065	-0.806 to 0.676
ORG		-0.04	-0.781 to 0.701
APD		-0.427	-1.177 to 0.322
ΣCAP		-1.01	-1.796 to -0.223
Gen		-0.141	-0.883 to 0.6

DEC; decoding, Noi.; noise, Mem.; memory, Var.; various, TFM; tolerance fading memory, INT; integration, ORG; organization, APD; auditory processing disorder, CAP; central auditory processing, Gen.; general, P-BMQ-R; Persian Buffalo Model Questionnaire-Revised version, SPD; Sensory Processing Disorder

Results of Cohen’s d measures for effect size for scores of WINT demonstrated in all SPD children, SPINT had very large effect on SPIN (Table 4).

Table 4.

Results of Cohen’s d measures for effect size for scores of Word In Noise Test in SPD children (n = 28)

Variable	Effect Size dCohen	Confidence Interval for dCohen
Right Ear Score in WINT	-2.058	-2.974 to 1.142
Left Ear Score in WINT	-2.133	-3.061 to 1.205

SPD; Sensory Processing Disorder

Discussion

The findings of the current retrospective study revealed that SPINT with white noise, in addition to improving behavioral auditory processing indicators, leads to decrease in hyperacusis symptoms in SPD children based on the scores of Noi, and P-BMQ-R subcategory, also word in noise perception improved after SPINT.

Effect of Speech in Noise Training on Behavioural Auditory Processing Indicators and Hyperacusis

SPINT, which is directly focused on improving listening skills in noise, is one of the main components of the auditory training program for individuals with hearing impairment, APD, learning and language disorders [37–39]. This technique helps subjects to desensitise to noise and learn strategies to extract more information from the auditory signal in challenging listening conditions [49]. Different types of speech materials including words (monosyllabic and double-syllable), sentence, phrase and text can be used in SPINT with specific SNRs [37–39].

Kaul and et al. (2016) examined the results of the auditory training based on the Buffalo model including phonemic training, dichotic training, speech in noise training and localization training on 20 children diagnosed with APD [50]. After intervention, a significant improvement in the results of the speech in noise test in the right ear and in the scores of all BMQ components. The changes were statistically significant in the DEC category and Mem subcategory. In the study Kumar et al. (2021), researchers evaluated the effect of the word-in-noise training program using monosyllabic and trisyllabic words in the presence of speech and multi-talker noise in the SNR range of + 20 to -4 dB on processing and cognitive skills (working memory) tests in 20 children with APD (10 children in the experimental group and 10 in the control group) results showed that SPINT has a significant effect on improving the scores of SPIN, temporal processing (GIN test and duration pattern) and memory ( reverse, descending and ascending digit span test) [39]. The results of present study are consistent with the study by Kaul et al. although the rehabilitation material was different, and the intervention was based on SPINT. Also, the results of this study are aligned with the study conducted by Kumar et al. with the difference that other auditory processing tests were not used in this study.

SPINT can be a method to improve listening skills. Other listening methods including AIT and sound therapy which, aim to desensitize the auditory system are methods for management of hyperacusis and tinnitus. In the study conducted using AIT, the results of AIT were examined on 17 autistic children and adolescents aged 4–21 years. The experimental group received the AIT protocols, and the control group only listened to unfiltered music. The results showed that in the experimental group, there was a significant improvement in hypersensitivity to sound, attention, repetitive behaviors and hyperactivity [51]. Vernon and Press, used pink noise for treatment of patients with complaint of hyperacusis who should listen to pink noise through headphones for 2 h at comfortable listening level, 20 of 30 participants who received the noise cassette used it, and 13 showed improvement in hyperacusis symptoms based on questionnaire [52]. Desensitation methods in SPD children are performed with more consideration. In such a way it starts by enriching the auditory system with pleasant sounds and them continues with annoying sounds which are collected, downloaded and trained to listen first at a low volume and then increase the volume each week [53]. Danesh et al. founded this method effective in reducing the negative reaction to sound in ASD children [53].

In study on individuals with SPIN problems and hyperacusis, poor results in the TOAE suppression test was obtained which is dysfunction in auditory efferent system (the medial olivocochlear bundle (MOCB)) [54]. It is possible that in the current retrospective study, SPINT by affecting the efferent system and strengthening inhibitory system reduced hyperacusis [37–39]. Since SPINT leads to strengthening neural timing [37–39], it may help reduce hyperacusis by decreasing neural synchrony in the auditory cortex. Also, it can be assumed that the use of white noise in the SPINT by involving non-classical auditory pathways like sound therapy methods, can desensitize the limbic system and help reduce the symptoms of hyperacusis.

Effect of Speech in Noise Training on Word in Noise Test

SPINT involves listening to speech stimuli at different noise levels. The importance of SPINT, especially as a method of desensitization to noise, has been reported in various researches.

In 1971, for the first time Katz and Burge used SPINT to improve speech perception skills in noise [55]. They trained 49 children with learning disabilities during 8 sessions (30 min). At the end of the intervention sessions, their ability to SPIN had increased. Studies show that SPINT changes the neural coding of sound structure at the cortical and subcortical levels of the brain and causes neuroplasticity. Also, it was found that after the intervention based on SPINT, significant improvement was obtained in behavioral and electrophysiology results in people with APD. In 2012, Song et al. investigated the effect of SPINT and its biological mechanisms using auditory brainstem responses to a speech syllable recorded in silence and noise, in 60 participants with a mean age of 24.7 years (28 trained people and 32 control people) before and after SPINT [38]. The results showed that after 6 months of training, the subjects showed a significant improvement in SPIN. Also, the subcortical responses in the noise after rehabilitation showed the improvement of the decoding of pitch-related cues.

In 2019, Jutras et al. reported that children with APD had a significant increase in noise tolerance after SPINT. They investigated the effect of SPINT on speech perception test scores, electrophysiological measures, as well as listening behaviors and life habits in the real life environment of the child and found that after SPINT, SPIN and electrophysiological components improved. In addition, the improvement was not only limited to noisy training conditions and was extended to other challenging environments of daily life [56].

In 2020, Kumar, Sing and Hussain designed a computerized speech rehabilitation program in noise and investigated its effect on SPIN, auditory processing tests and auditory electrophysiological responses of APD children with difficulties in SPIN. This computer program included words in noise training, in such a way that monosyllabic and trisyllabic words were presented as target words in the presence of speech-like noises, as well as multitalker babble in different SNRs in the range of + 20 to -4 dB. Compared to children in the control group, experimental group showed improvements in SPIN and auditory temporal processing after intervention. Also, after SPINT, the latency of evoked responses in silence and noise showed a significant decrease. The results showed that the organized SPINT program is effective in improving auditory behavioral skills and electrophysiological responses [37].

The results of present study are aligned with the studies conducted in this field ( [37–39, 55, 56], although the rehabilitation materials used in them were monosyllable words in multitalker noise [55], sentences in multitalker noise [37–39], monosyllabic and trisyllabic words in speech and multitalker babble noise [56]. In the current study, the materials were monosyllabic words, two-syllable words and sentences in white noise. White noise contains every frequency within the human range of hearing, generally 20 Hz to 20,000 kHz, with equal energy across the spectrum [57].

However, this study did not have a control group. Future studies with randomised controlled designed are needed to further investigate effect of SPINT on hyperacusis in this patient population. Another limitation of this study was the lack of specific Persian hyperacusis questionnaire for children. Therefore, there is a need to develop and validate a measure for assessment of hyperacusis in children to be used in future studies. Also, other limitation is the lack of parental satisfaction information and clinical observations based on the how SPINT is clinically significant on hyperacusis symptoms. Then, we recommend in further studies relevant questionnaires for clinical observations be used. Furthermore, it is recommended to investigate this treatment method in different population using questionnaires specific to each age group. More research needs to be done using different syllable and speech materials with different types of multi-talker noises.

Conclusion

This preliminary study showed promising result for the effect of SPINT on improving behavioural indicators of auditory processing disorder (as measured via P-BMQ-R and WINT) and decreasing hyperacusis symptoms (as measured via Noi). However, our result was only exploratory and future studies with randomized controlled design are required to assess effectiveness of SPINT on hyperacusis.

Funding

This research received no specific grant from any funding agency, commercial organization, or not-for-profit sector.

Declarations

Conflict of Interest

The authors state that there is no conflict of interest.

Ethical Approval

This study was approved by the Ethics Committee of The Hamadan University of Medical Science (IR.UMSHA.REC.1401.678).

Informed Consent in the manuscript

Patient consent was waived as this was a retrospective analysis of available clinical data.

Competing Interest

The authors state that there is no conflict of interest.