Psychometric Validation of the New Misophonia Screening List—Child and Youth and AMISOS-Y (Dutch Child- and Parent-Report Versions) for Assessing Misophonia in Youth

Misophonia is a recently identified disorder of decreased tolerance to specific (mostly human) sounds and/or associated stimuli [1, 2]. When anticipating or perceiving so-called “trigger” sounds or stimuli (e.g., eating or mouth sounds), individuals with misophonia experience immediate and strong physiological and emotional reactions, which are often difficult to control, along with behavioral reactions [1, 3‐5]. Situations with potential triggers are commonly avoided, escaped, or encountered with anticipation anxiety [1, 3]. Misophonia typically develops during childhood or early adolescence [5‐8], and is associated with reduced quality of life and considerable educational, social and relational impairment in youth [6, 9].

While consensus on the definition of misophonia was recently reached by an international interdisciplinary group of experts through a Delphi-method [2], consensus regarding diagnostic criteria for misophonia is still lacking. As a disorder of decreased sound tolerance, misophonia might be understood in the context of audiology (by contrasting its features with audiological disorders such as hyperacusis and tinnitus; [10‐13]); and/or psychiatry (by contrasting its features with psychiatric disorders; [8, 14]). The absence of unified diagnostic guidelines complicates the research into etiology, assessment tools and treatment of misophonia, essentially leading to a vicious cycle.

As a first step in addressing this problem, several research groups have proposed diagnostic criteria for misophonia [1, 14‐16]. The Amsterdam UMC misophonia research group was the first to propose diagnostic criteria for misophonia as a psychiatric disorder [14], and published revised criteria in 2020 after a follow-up study of 575 adult patients with misophonia (full criteria in [1]). These Amsterdam UMC 2013/2020 criteria have been criticized for being too restrictive (e.g., children might not recognize that their feelings are excessive; [8, 16]), and for their assumption of psychiatric etiology and the absence of differentiation from hyperacusis, which according to Jastreboff and Jastreboff [13] might lead to biased samples of patients with misophonia. While it is important to consider these and other criticisms in the ongoing field of misophonia research, the proposed Amsterdam UMC 2020 criteria largely reflect the consensus definition of misophonia [2] and form an important starting point for harmonization and comparison of future international misophonia studies [17]. For instance, in the current absence of unified diagnostic guidelines, the Amsterdam UMC 2020 criteria might serve as a basis for development of questionnaires, which allows for a more straightforward and cross-cultural comparison between different misophonia studies.

This is the first study with the largest combined clinical (N = 95) and general population (N = 197) sample in the widest age range (8–18) thus far presenting an extensive iterative validation process of two parallel child- and parent-report questionnaires combining two aims: the Misophonia Screening List—Child and Youth to screen for misophonia symptoms in the general youth population (including suggested cutoff scores), and the AMISOS-Y to assess misophonia severity in children/adolescents with misophonia symptoms.

Instrument refinement resulted in an 11-item Misophonia Screening List—Child and Youth, and a 17-item AMISOS-Y (including seven questions assessing the nature of trigger sounds and responses, and 10 Likert-scale items). Subsequent reliability and principal component analyses revealed high internal consistency and a unidimensional structure of both questionnaires. For the Misophonia Screening List—Child and Youth we suggest a cutoff score of 10 for the child-reported version and 9 for the parent-reported version. Future studies might be conducted to investigate the cutoff scores in new community and clinical samples.

Evidence of concurrent and convergent validity was provided by the large correlations between total scores on our misophonia questionnaires and misophonia “diagnosis”, and between the child- and parent-report versions of our questionnaires. Furthermore, our hypotheses regarding divergent (discriminant) validity were mostly supported. We found small-to-moderate positive correlations between our misophonia questionnaires and measures of attention problems, internalizing problems, and sensory sensitivity, which suggests that these are related, but distinct constructs, and provides evidence that our instruments are able to discriminate well between these constructs. The finding that misophonia appears to be a distinct construct, does not imply correlations with other constructs need to be zero.

However, (partially) in contrast to our hypotheses, we only found significant correlations between total scores on our misophonia questionnaires and the Attention Deficit subscale (and not the total scale) of the Dutch ADHD questionnaire (AVL). This might reflect patterns of comorbidity found in previous studies, in which children with misophonia displayed more comorbidity with ADHD with predominantly inattentive presentation, than with ADHD with combined/predominantly hyperactive-impulsive presentation [6, 7]. Importantly, the AVL does not fully represent the DSM-5 ADHD diagnosis, as it is a parent-reported questionnaire (as opposed to a multi-informant questionnaire or clinical interview), and as some items are categorized differently than in the DSM-5 (e.g., "being easily distracted" is categorized as hyperactivity in AVL, and as attention deficit in DSM-5) [47, 66]. Future studies might aim to incorporate DSM-5 based clinical interviews, such as the SCID-5 Junior [56], for convergent/divergent validity analyses.

Furthermore, contrary to our hypothesis, the (parent-reported) CBCL-Internalizing subscale Somatic Complaints (measuring physical problems without known medical cause; [48]) was not significantly correlated with our parent-reported misophonia questionnaires. In contrast, in previous research increased parent-reported psychosomatic complaints were found in children with misophonia compared to peers [7]. This discrepancy might be due to the current study measuring broad psychosomatic complaints (e.g., nightmares, eye/skin problems, different types of pain), and the Siepsiak et al. [7] study measuring only psychosomatic pains (i.e., stomach pain, back pain, pain in other parts of the body), and excluding “physiological components of emotions” (e.g., stomach pains right before an exam). Possibly, our broad measurement of psychosomatic complaints might have resulted in different correlational patterns. Indeed, closer inspection of our Somatic Complaints scale revealed highest scores for the items ‘Headaches’ and ‘Stomachaches’. These findings suggest a potential relationship between (psychosomatic) pain and misophonia symptoms, which might be reflective of the distress caused by the disorder, as well as of the tendency of children to express psychological symptoms in a somatic manner [67].

While, contrary to our hypothesis, we found a significant (small) correlation between the (parent-reported) CBCL-Externalizing total problem scale and the parent-reported Misophonia Screening List—Child and Youth, this was not found for specific CBCL-Externalizing subscales, nor for the YSR-Externalizing problem scale and the other misophonia questionnaires. Possibly, this discrepancy might be due to increased rates of verbal and physical aggression in response to triggers found in younger children (7–12 years) compared to older children (13–18 years) [7], considering that the CBCL was filled in by parents for all participating children in this study, and the YSR only by children in the older age range (11+). Furthermore, in contrast to our hypothesis, we found (small) significant correlations between both child-reported misophonia questionnaires and the (child-reported) YSR-Social Problems scale, which implies that children reporting higher presence and severity of misophonia symptoms also report more social problems. This discrepancy might be attributed to the definition of social problems: these might imply difficulty in social skills, but also negative social consequences of misophonia symptoms (e.g., YSR-item 12: I feel lonely). In our hypothesis, we mainly focused on difficulty in social skills, however the social impairment caused by misophonia [6, 53, 68] might be reflected in our findings.

In short, the newly developed Misophonia Screening List—Child and Youth and AMISOS-Y (parallel child- and parent-reported versions) are easy-to-use and efficient questionnaires with excellent psychometric properties, with support for concurrent, convergent and divergent validity. In clinical practice, both questionnaires might be incorporated as part of a larger diagnostic process of screening, (differential) diagnostics, and severity assessment (e.g., according to Regier et al. [69]). First, the 11-item Misophonia Screening List—Child and Youth could be administered as standardized screening tool for use in general or potential at-risk populations, such as in children with autism spectrum disorder [70]. Scores above the cutoff (i.e., 10 for child report and 9 for parent report) indicate the presence of misophonia symptoms. Second, a clinical interview or evaluation should be performed, ideally by a team of medical professionals such as otolaryngologists, audiologists [16] and psychiatrists [31], to differentiate misophonia from related conditions (e.g., hyperacusis; [12, 13]) and provide a misophonia “diagnosis” (e.g., by assessment of proposed Amsterdam UMC 2020 criteria, in absence of unified diagnostic criteria). Third, (parents of) children with misophonia might then complete the AMISOS-Y to establish baseline misophonia severity, which could be compared to the norm group scores provided in this study (Table 1), or to the individual scores over time to assess natural course or treatment response.

Strengths of this study include our extensive validation process, resulting in two questionnaires with excellent psychometric properties and with clearly defined practical use. By comparing a carefully included group of clinically referred children/adolescents (with misophonia as main “diagnosis”, confirmed during an extensive intake process with expert psychiatrists/psychologists) with a control group, we were able to develop an innovative screening tool, with clear cutoff scores for parallel child- and parent-reported versions, which was not available until now. Furthermore, as measure of severity, the AMISOS-Y captures the multidimensional nature of misophonia symptoms by measuring a wide variety of triggers, emotional/behavioral responses and consequences of misophonia.

Several limitations should be considered. First, these questionnaires focus on misophonia per se, and do not incorporate questions assessing symptoms of related disorders (e.g., hyperacusis), and visual triggers or physiological responses to triggers (e.g., muscle tension), which are commonly reported in misophonia and mentioned in the consensus definition [2]. We recommend to screen for these symptoms in a clinical diagnostic interview instead (which was also our procedure). Second, the answer option ‘I don’t know/not applicable’ on the parent-reported versions was calculated by person-mean imputation, which could have resulted in inflated reliability coefficients [71]. However, since this answer option was used occasionally in our study (for < 20% of the items and respondents), the negative consequences of person-mean imputation can be considered negligible [71]. Interestingly, this answer option was used sparingly even in the control group (in which reporting of ‘Not applicable’ might have been expected). This might indicate that this answer option, while increasing the complexity of scoring the questionnaire, is not perceived inherently different from the answer option ‘No(t)’. Therefore, future replicating studies might consider removing this answer option from the parent-reported versions (and the parallel ‘Not applicable’ from the child-reported versions), which would result in a more straight-forward scoring of both questionnaires. Third, in the clinical group our misophonia questionnaires were assessed online (as part of the large RCT), and mostly at home, whereas in the control group assessment took place through paper-and-pencil questionnaires in the classroom. The latter might be associated with perception of reduced confidentiality, which could arguably lead to underreporting of symptoms in the control group. We aimed to minimize consequences by explicitly stating (verbal and on paper) in both groups that coded answers were only visible to the research team. In previous research, mode of administration did generally not influence questionnaire outcomes [72, 73], 74. Fourth, in the current study we did not include other (preliminarily validated) measures of misophonia in youth, such as the SMS-Adolescent [9], for validity assessment, however this might be an important topic for further investigations. Last, while our clinical sample may also be considered a strength, we employed strict in- and exclusion criteria and included only Dutch referred children/adolescents, which is not representative for the general population of children/adolescents with misophonia. Future studies should incorporate larger and more diverse population samples to increase generalizability and include different norm groups.

To conclude, this study presents two newly developed parallel child- and parent-reported questionnaires for screening and severity assessment of misophonia in children/adolescents, with sound psychometric properties. In the future, the misophonia research field should collaboratively establish diagnostic guidelines, in which these and additional questionnaires can be incorporated.

This paper presents an extensive iterative validation process of two innovative (parallel child-/parent-reported) questionnaires specifically designed to assess misophonia in youth: the Misophonia Screening List – Child and Youth for screening, and the AMISOS-Y to assess misophonia severity. On a large combined clinical and general population sample of children (aged 8-18) and parents, we performed reliability and principal component analyses, revealing a one-factor solution and excellent psychometric properties for both questionnaires. Screening cutoff scores were identified using ROC curve analyses. Evidence for concurrent, convergent, and divergent validity was found, with both questionnaires differentiating well between misophonia and related constructs. While future replication studies in community and clinical samples are needed, our results support incorporation of these new questionnaires for youth within a unified diagnostic framework