Main sensitivity drivers in indoor acoustic comfort for autistic individuals Marco Caniato 1 Free University of Bozen – Faculty of Science and Technology Piazza Università, 1, 39100 Bolzano BZ, Italy Arianna Marzi 2 Free University of Bozen – Faculty of Science and Technology Piazza Università, 1, 39100 Bolzano BZ, Italy Andrea Gasparella 3 Free University of Bozen – Faculty of Science and Technology Piazza Università, 1, 39100 Bolzano BZ, Italy ABSTRACT The present study analyzes the acoustic stress induced on autistic individuals, studied by means of questionnaires administrated to parents and professional caregivers. The acoustics comfort domain has a great impact on autistic people; indeed, variations of spatial acoustic parameters can influence the sensitivities of each individual and the perceived stress is influenced by severity of autism, co- morbidities and age. The study highlights that there are clear differences between the indirect evaluations performed by parents or professional caregivers, in relation to the specific identification of stress sources. Acoustics is identified as the one causing the major stress, especially caused by noises and particular sounds such as voices, animal sounds and impacts, and its dependence on the severity of autism is evidenced. 1. INTRODUCTION

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by impaired social and communication skills along with repetitive and restrictive behaviour. The clinical presentation is highly heterogeneous, including individuals with severe comorbidities and intellectual disability (ID) as well as individuals with above-average intelligence quotient (IQ) and high levels of academic and occupational functioning. ASD affects 1–1.5% of individuals and both common and rare variants contribute to its aetiology [1]. Over-reactivity to sensory stimuli and tendency to engage sensory stimulation and/or motor activities are common comorbidities for ASD individuals. According to some researchers, these disturbances in perception are caused by brainstem abnormalities. Studies of autonomic and vestibular responses have provided support for such abnormalities. Communication is strikingly

1 mcaniato@unibz.it 2 amarzi@unibz.it.com 3 Andrea.gasparella@unibz.it

we i nllaieey. inter.noise 21-24 AUGUST SCOTTISH EVENT CAMPUS GLASGOW

impaired so that the function of the hearing system has been discussed [2]. Indeed, the disorder diagnosis by many clinicians includes behavioural descriptive of different environment perception [3]. In a multi-sensory world, persons are constantly exposed to information reaching them through our different senses. The brain has to synthesize a mix of sensory information into one coherent multi- sensory perception [4], [5]. What happens if the brain is overloaded in integrating this mix of sensory input? This is one of the main issues related to ASD individuals. In addition to conventional impairments in communication, social behavior and repetitive movements, abnormalities in sensory processing are often reported [6]. Several studies focused on the physiologic and neurologic analysis to understand the auditory perception of ASD people, but only few focused on the acoustic perception. Other more complete studies are devised by the present authors on this topic, specifically on the 4 indoor comfort domains. This research explored the perception of people in the spectrum among the 4 domains of comfort highlighting how the acoustic environment is the most critical for this type of users both in health care facility and in households. In the present paper, a focus on the acoustic perception issues is presented, also performing an analysis of influence on gender, severity of the autism, age and presence of other disorders in co-morbidities. In order to investigate this topic a dedicated questionnaire was designed. Due to the frequent communication disorder that may affect the individuals, the questionnaire was filled by proxy respondents, namely parents and caregivers. The results collected were then analysed to check if the reported perception could vary based on the proxy respondents or personal data of the ASD individuals. This survey was conducted in accordance with the Declaration of Helsinki and the Ethical Committee of the Free University of Bozen. 2. MATERIALS AND METHOD

In order to investigate the ASD users’ indoor comfort, questionnaires on the environment perception are a common tool among experts and researchers [7], [8]. Unfortunately, questionnaires may not be suitable to investigate the indoor comfort of people with communication difficulties, who would need support in their daily activities and life. For this reason, thanks also to the contribution of a psychologist expert in this field, it was possible to develop a dedicated questionnaire for parents and caregivers (proxies) and focused on the perception of the ASD individuals on the visited environments (home and care facility). Due to the possibility of recording hyposensitivity besides hypersensitivity, and the possibility that altered perception occurs only in specific cases/contexts, as it is the case with persons on the autistic spectrum [9], the scale was designed to register if increased or reduced perception occurred, as well as if regularly or sporadically [10] - [11]. Through this questionnaire, it was possible to collect anonymized information on the perception in the acoustic comfort domain as well as on severity of autism, and on the presence of other comorbidities. Figure 1 summarize the ad hoc perception scale built-in for the survey.

Figure 1: Perception designed for the questionnaire

After preliminary instructions and clarifications, respondents were asked to indicate whether they were caregivers or parents. In each questionnaire, the respondent was required to report the main information about the individual with ASD they were referring to, such as gender (male, female or

Yes Legend of perception persistent aspects of perception (CL-minorly.2-averagely, 3-extremely) sporadically (S1-sporacically minorly.s2-sporadically averagely, S3-sporadically extremely) Hyposensitive (H) Absent (0)

other), age, severity of ASD and comorbidities (Obsessive-Compulsive Disorder - OCD, depressive disorder, anxiety disorder or other psychiatric or neurocognitive disorders). The indication of the severity of autism and presence of co-morbidities was asked according to The Diagnostic and statistical manual of mental disorders of the American Psychiatric Association [12] which identifies three main levels: • A – low severity, requiring support • B – medium severity, requiring substantial support • C – high severity, requiring very substantial support. In addition, questions regarding the date of questionnaire completion and the environment where the described individual was (family house, apartment, assisted facility, etc.) were included. The main part of the questionnaire was then devoted to the registration of the individual’s special sensitivity to the thermo-hygrometric, acoustical, visual and indoor air environments, through multiple-choice scales. Different subjects were considered by parents and caregivers.

2.1. Data Collection In total 138 questionnaires were collected during the survey. Table 1 shows the number of questionnaires collected divided by grade of autism severity and the number of co-morbidities. Intellectual disability was the co-morbidity detected more often. Table 1: Personal data of the individuals in the spectrum reported by the two proxy respondents

Overall Parents Caregivers

Severity of autism A 20 27 B 22 24 C 16 29

OCD 12 8

ID 29 33

Co-morbidities

AD 10 9

Other 13 19 The gender and the age of the respondents of the survey are reported in Table 2, the majority of subjects are males and the majority of questionnaire collected regard people in young age and young adults. Table 2: Distribution of gender among the ages of the ASD individuals

Age Overall

Females Males Other

7-10 1 15 3 10-17 15 26 0 18-29 11 36 0 30-39 12 6 0 40-49 1 5 0 50+ 3 4 0

2.2. Data Analysis In order to compare properly the perception of the ASD users in the two environments, a qualitative preliminary analysis based on the comparison of frequency distributions, followed by a statistical analysis were conducted. The methodological analysis is structured as follows:

1) a preliminary percentage analysis was performed in order to highlight whether a difference could exist in the perception detected by the two types of proxy respondents. 2) a Mann Whitney test [13] was performed to verify the possibility of obtaining reports of perceptions different between the two categories of proxy respondents. In this case the null hypothesis in the Mann-Whitney test is that the two datasets are drawn from the same ASD population and that therefore their probability distributions should be equal; 3) finally, a further analysis with Mann Whitney test is performed to evaluate if there are correlations between the acoustic perceptions and influence of personal data of the ASD individuals (age, presence of comorbidities and severity of autism). 3. RESULTS AND DISCUSSIONS

Figure 2 shows the results of the aggregated perception depicted by both proxy respondents on behalf of ASD individuals. From the overall results it can be seen how the least stimulus perceived to provide discomfort to people on the spectrum are stimuli produced by low noise levels. Then, by looking particularly to the “averagely” and “extremely” perceptions, retrieved by the dedicated scale in the survey (fig. 1) it can be noticed that the overall acoustic environment is the most detected, followed almost equally by high noise levels (38% averagely and 19% extremely), outside high noise (27% averagely and 14% extremely), rumble sounds (18% averagely and 18% extremely) and particular noises (28% averagely and 14% extremely). Very low percentage of hyposensitivity and sporadic sensitivity were detected, proving how ASD individuals seem to be hypersensitive to acoustic environment perception.

A COUSTIC - O VERALL

Hyposensitive

not at all

minorly

averagely

extremely sporadic

100%

50%

0%

ACOUSTIC ENVIRONMENT HIGH NOISE LEVELS LOW NOISE LEVELS OUTSIDE HIGH NOISE RUMBLE SOUNDS IN THE ROOM

PARTICULAR NOISES

Figure 2:Overall perception of the acoustic environment of ASD individuals

3.1. Influence of the proxy respondents on ASD perception reports Figure 3 highlights that the acoustic domain has a very strong impact on ASD individuals, with a high percentage of parents and caregivers indicating “averagely” perception (39 % by parents and 36 % by caregivers) and “extremely” (24 % by parents and 19 % by caregivers). This demonstrates that sounds can strongly disturb the quietness and cause severe discomfort for people on the spectrum. On the other hand, negligible percentages of hypo-sensitive individuals were present. The most disturbing elements indicated by the participants were high noise levels from indoors or outdoors, rumbling sounds and particular sounds such as voices, impacts, animals sounds and calls. In general, parents and caregivers mainly disagreed on sensitivities for high noises from outside, rumbling sounds and particular noise, for which parents indicated a higher percentage of “extremely” perceptions. Recommendations in the literature suggest avoiding loud noises as well as echo and background noises in building designs for ASD people [14]. However, these are usually based on single case

studies or very small samples. By means of the presented analysis, based on a large sample, it is possible to generalize these findings and identify also specific noise sources to be avoided.

A COUSTIC - C AREGIVERS

A COUSTIC - P ARENTS

100%

100%

Hyposensitive

not at all

minorly

averagely

extremely sporadic

Hyposensitive

not at all

minorly

averagely

extremely sporadic

50%

50%

0%

0%

ACOUSTIC ENVIRONMENT HIGH NOISE LEVELS LOW NOISE LEVELS OUTSIDE HIGH NOISE RUMBLE SOUNDS IN THE ROOM

PARTICULAR NOISES

ACOUSTIC ENVIRONMENT HIGH NOISE LEVELS LOW NOISE LEVELS OUTSIDE HIGH NOISE RUMBLE SOUNDS IN THE ROOM

PARTICULAR NOISES

Figure 3:Perception of the acoustic environment reported by the two proxy respondents on behalf of ASD individuals: prantes (left) and caregivers (right)

Table 3 shows the statistical results as regards parents and caregivers. It can be highlighted how different perception is reported by the two proxy respondents on rumbling sounds and perception of particular noises (voices, impacts, animal call etc.). These differences could be due to two main reasons: (i) different environments and time (moment of the day and duration) of observation they are referring to; (ii) different background and sensitivity of parents and caregivers. Table 3: Results of the Mann-Whitney test comparison on the answers of parents and caregivers, considering the overall sample and the control sample. (Np: Number of parents; Nc: Number of caregivers). P-values results presented in columns refer to the comparison between parents and caregivers’ answers in relation to the overall sample. In the cases where a statistically significant difference was found, the mean difference between the samples is reported

Question P-value

Number of answers (N P = number of questionnaires filled in by parents; N C = number of questionnaires filled in by caregivers) N P = 58;

N C = 80 Acoustic environment 0.241 High noise levels in the environment where she/he is 0.131 Low noise levels in the environment where she/he is 0.603 High noise levels coming from outside 0.201

Rumbling sounds 0.069*

µ P -µ C = 0.353

Particular noises (voices, impacts, animal calls, etc.) 0.072*

µ P -µ C = 0.325

* = test significant at 10 % significance level; ** = test significant at 5 % significance level; *** = test significant at 1 % significance level;

µP = average value reported in the parents’ sample; µC = average value reported in the caregivers’ sample

3.2. Influence on acoustic of personal data Table 4 shows the outcomes of the Mann Whitney test evaluating the statistical differences between: 1. Gender (second column); 2. Autism severity: A -low severity and higher levels (levels B and C) (third column); 3. Absence or presence of co-morbidities (fourth column). The following observations can be made: 1. Gender influenced only the sensitivities to low noise levels. 2. The severity of autism clearly influenced acoustic perception and all the specific stimuli regarding this environmental comfort domain, with the exception of rumbling sounds. The mean difference analysis shows that an increase in the severity of autism means a higher sensitivity to these stimuli. 3. No dependence of the presence of co-morbidities was found to be related to the acoustic environment perception.

Table 4: Results of the Mann-Whitney test comparison on the answers regarding gender, severity of autism and presence of co- morbidities considering the overall sample. P-values results presented in columns are referred to the comparison between Males

versus Females, A versus B+C severity of autism and individuals with no co-morbidities versus individuals with co-morbidities. In the cases where a statistically significant difference was found, the mean difference between the samples is reported

P-value

Question

Gender Level of

Co-

autism

morbidities

Number of answers

(N F = number of questionnaires regarding females; N M = number of questionnaires regarding males;

N A = number of questionnaires regarding individuals with ASD Level A; N B+C = number of questionnaires regarding

N F = 42;

N A = 48;

N NC = 53 ;

individuals with ASD Level B or C;

N M = 94

N B+C = 90

N C = 85

N NC = number of questionnaires regarding individuals with no co-morbidities; N C = number of questionnaires

regarding individuals with co-morbidities)

0.001***

Acoustic environment 0.167

0.192

µ A -µ B+C =

-0.628

< 0.001***

High noise levels in the environment where she/he is 0.292

0.133

µ A -µ B+C =

-0.839

< 0.001***

0.003***

Low noise levels in the environment where she/he is

0.111

µ F -µ M =

µ A -µ B+C =

0.444

-0.379

0.021**

High noise levels coming from outside 0.352

0.308

µ A -µ B+C =

-0.451

Rumbling sounds 0.388 0.123 0.661

< 0.001***

Particular noises (voices, impacts, animal calls, etc.) 0.987

0.162

µ A -µ B+C =

-0.673

* = test significant at 10 % significance level; ** = test significant at 5 % significance level; *** = test significant at 1 % significance level;

µ F = average value reported in the female sample; µ M = average value reported in the male sample;

µ A = average value reported in the sample with ASD level A; µ B+C = average value reported in the sample with ASD level B or C;

µ NC = average value reported in the sample without co-morbidities; µ B+C = average value reported in the sample with co-morbidities

Table 5 depicts how answers were statistically influenced by the age of the individuals participating in the survey. The three thresholds of 18, 30 and 40 provided statistically valuable samples that made it possible to analyze the potential variations in perception according to age. These thresholds are reported respectively in the second, third and fourth column. In each column, the statistical comparison of the samples under and over the corresponding threshold is reported. Only in two cases an age dependency was detected by Mann-Whitney. In the first case, a progressive change of perception to high noise levels stimulus can be identified with the increase of the age threshold. Indeed, the difference in perception between people under-and over-18 persists beyond the age of 30 for high noise levels. On the other hand, the perceptual differences decrease when considering people under- and over-40. Moreover, in this case, the significance level is higher when considering the 30-year-old threshold. For these reasons, people in the age range of 30-39 can be identified as the ones who mostly manifested differences in the perception of the indoor environment for the high noise levels issues. Then, as for the case of rumbling sound, a difference in responses distribution between under- and over-40-year-olds was detected. Consequently, with this stimulus, the difference in perception is attributable to the over 40 group: when aggregated with under-18s and under-30s, its influence is attenuated.

Table 5: Results of the Mann-Whitney test comparison on the answers regarding subjects under and over different age thresholds considering the overall sample. P-values results presented in columns are referred to the comparison between age thresholds. In the cases where a statistically significant difference was found, the mean difference between the samples is reported

Question

Threshold age 18 30 40

N. of answers (N U = number of questionnaires regarding subjects under the age threshold; N O = number of questionnaires regarding subjects over the age threshold) N U = 59; N O = 79 N U = 107;

N O = 31 N U = 124 ;

N O =14 Acoustic environment 0.102 0.270 0.461

High noise levels in the environment where she/he is 0.097* µ U -µ O = -0.269

0.032** µ U -µ O = -0.306

0.115

Low noise levels in the environment where she/he is 0.244 0.488 0.742 High noise levels coming from outside 0.683 0.347 0.213

Rumbling sounds 0.995 0.129 0.062* µ U -µ O = -0.281 Particular noises (voices, impacts, animal calls, etc.) 0.130 0.711 0.752

* = test significant at 10 % significance level; ** = test significant at 5 % significance level; *** = test significant at 1 % significance level;

µ U = average value reported in the sample under the age threshold; µ O = average value reported in the sample over the age threshold 4. CONCLUSIONS

The perception of acoustic domain perceived by individuals on the autism spectrum disorder was investigated in this paper. A survey was used to investigate these issues, which were completed by parents and caregivers, due to the limited autonomy of many individuals. Results were studied by means of descriptive and statistical analyses. As main output, it can be seen how acoustics was considered a paramount source of stress. Loud or rumbling noises were identified as particularly disturbing, as well as some particular sounds such as voices, impacts and animal sounds. Acoustic sensitivity was found to be very impactful for all ages. In addition, stimuli linked to acoustics were observed to be more significant with higher severities of autism. No influence on the presence of comorbidity was identified in these analyses, while gender resulted to influence the sensitivity on “low noise levels”. 5. ACKNOWLEDGEMENTS

This work was financed by the European Interreg SENSHome project, ITAT 1088 CUP: I54I18000310006. The authors want to thank Michele Borghetto for his precious help in defining the correct questions form and all the partners of the European Project SENSHome, who helped us with their contributions. This paper was also partially financed by the SCORELINE project, Italy financed by the Free University of Bozen, CUP I55F21001090005. The study was conducted after the approval of the Ethics Committee of the Free University of Bozen-Bolzano. The procedure was implemented in order to comply with the Declaration of Helsinki. 6. REFERENCES

[1] J. Grove, et al., Identification of common genetic risk variants for autism spectrum disorder, Nature Genetics. 51 (2019) 431–444. https://doi.org/10.1038/s41588-019-0344-8. [2] U. Rosenhall, V. Nordin, M. Sandström, G. Ahlsen, C. Gillberg, Autism and Hearing Loss, Journal of Autism and Development Disorderd. 29 (1999). https://doi.org/0162-3257/99/1000- 0349$16.00/0. [3] A. Talay-Ongan, K. Wood, Unusual Sensory Sensitivities in Autism: A possible crossroads, International Journal of Disability, Development and Education. 47 (2000) 201–211. https://doi.org/10.1080/713671112.

[4] M.A. Boudjarane, M. Grandgeorge, R. Marianowski, L. Misery, É. Lemonnier, Perception of odors and tastes in autism spectrum disorders: a systematic review of assessments, Autism Research. 10 (2017) 1045–1057. [5] N.R. Zürcher, O. Rogier, J. Boshyan, L. Hippolyte, B. Russo, N. Gillberg, A. Helles, T. Ruest, E. Lemonnier, C. Gillberg, Perception of social cues of danger in autism spectrum disorders, PloS One. 8 (2013) e81206. [6] L. de Boer-Schellekens, M. Eussen, J. Vroomen, Diminished sensitivity of audiovisual temporal order in autism spectrum disorder, Frontiers in Integrative Neuroscience. 7 (2013). https://doi.org/10.3389/fnint.2013.00008. [7] CEN/TR 16798-2: 2019, Energy Performance of Buildings—Ventilation for Buildings—Part 2: Interpretation of the Requirements in EN 16798-1—Indoor Environmental Input Parameters for Design and Assessment of Energy Performance of Buildings Addressing Indoor Air Quality, Thermal Environment, Lighting and Acoustics (Module M1-6), European Committee for Standardization Brussels, Belgium, 2019. [8] B. Noble, N. Isaacs, S. Lamb, The impact of IEQ factors on people on the autism spectrum, in: International Conference of the Architectural Science Association, 2018: pp. 27–33. [9] A. Neklyudova, K. Smirnov, A. Rebreikina, O. Martynova, O. Sysoeva, Electrophysiological and Behavioral Evidence for Hyper-and Hyposensitivity in Rare Genetic Syndromes Associated with Autism, Genes. 13 (2022) 671. [10] E. Back, H. Hunt, A. Lindell, The perception and pose of emotional expressions in adolescents with autism, (2010). [11] V. Khullar, H.P. Singh, M. Bala, IoT based assistive companion for hypersensitive individuals (ACHI) with autism spectrum disorder, Asian Journal of Psychiatry. 46 (2019) 92–102. [12] American Psychiatric Association, Diagnostic and statistical manual of mental disorders - Edition Fifth, American Psychiatric Association. 21 (2013) 591–643. [13] P.E. McKnight, J. Najab, Mann ‐ Whitney U Test, The Corsini Encyclopedia of Psychology. (2010) 1–1. [14] M. Mostafa, Architecture for Autism: Autism ASPECTSS TM in School Design, ArchNet-IJAR: International Journal of Architectural Research. 2 (2014).