Abstract
Background: Selective mutism (SM) research from China remains scanty. This study describes the sociodemographics and experiences of children and families with SM in China and identifies correlates of SM in China. Methods: This case-control study enrolled 172 cases with SM and 179 controls by online surveys with closed and open-ended responses. Multilevel logistic regression was performed to examine the association between potential correlates and SM. Open-ended responses were analyzed thematically. Results: Only 51.2% of SM cases were diagnosed by a professional, and 31.1% of SM cases that received treatment were guided by a professional. Child SM was associated with having parent(s) with an introverted personality, being born to a mother aged > 35, and having a sibling. Conclusion: Many children with SM in China have not received professional treatment or interventions. Novel policies such as access to special education resources, SM training for therapists, and school support are needed to enhance the early detection and treatment of SM in China.
Plain Language Summary
Selective mutism (SM) is an anxiety disorder characterized by a consistent failure to speak in specific social settings (e.g., school) despite speaking normally in other situations. However, SM research from China remains scanty at best. In partnership with the Selective Mutism Association of China, we conducted what is arguably the first nationwide study of SM in China. Findings suggest that many children with SM in China have not received effective professional treatment or interventions. Results from our case-control analysis highlight hereditary and social environmental factors that may be contributing to childhood onset of SM in China.
Background
Selective mutism (SM) is an anxiety disorder characterized by a consistent failure to speak in particular public settings (e.g., school), while being able to speak normally in other situations (e.g., home; APA, 2013; Muris & Ollendick, 2015). The disorder is listed among the anxiety disorders in the Diagnostic and Statistical Manual of Mental Disorders, fifth Edition (DSM-5; APA, 2013). To diagnose this disorder, the DSM-5 stipulates that speech inhibition should last for at least 1 month. The diagnosis cannot be established during the first month at school, where the absence of speech may be attributed to children’s nervousness in an unfamiliar environment instead of a mental disorder. Moreover, according to the DSM-5, the lack of speech is not merely the result of a lack of familiarity with or discomfort with the spoken language needed in the social situation. In addition, the failure to speak is not better explained by a communication disorder and does not occur exclusively during the course of autism spectrum disorder, schizophrenia, or another psychotic disorder (APA, 2013).
The prevalence of SM among children is estimated between 0.03% and 1%, and the mean age of onset is between 2 and 5 years, thus indicating that SM is a relatively rare childhood disorder (APA, 2013; Muris & Ollendick, 2015). SM has a mean duration of 8 years, after which the key symptoms such as the total lack of speech in certain settings usually dissipate, but the related difficulties may persist long term (Remschmidt et al., 2001). The difficulties associated with SM include speech problems, fears, and social phobia, which can interfere with children’s daily function, and educational and occupational achievement (APA, 2013; NHS, 2019).
International literature suggests that several sociodemographic and environmental factors may be associated with SM among children. For example, Koskela et al. (2020) stated that compared to a child of a mother who was married or in a relationship, the child of a single mother was twice as likely to develop SM. A telephone interview study revealed that the prevalence of SM among immigrant children was 2.2%, while the prevalence among children in native families was 0.5%, indicating that immigrant status is a potential contributing factor to SM development (Elizur & Perednik, 2003). However, there is only limited evidence of the associations between these factors and children’s SM, and some of these findings remain controversial. For example, while McGrath et al. (2014) suggested an association between maternal age and SM among their children, the study by Koskela et al. (2020) reported no significant association between these factors. Evidence on the marital status of parents with a child with SM is heterogeneous as well. Although divorce has been frequently reported by families with SM children, convincing evidence on the association between parental divorce and children’s SM onset is still lacking (Elizur & Perednik, 2003; Muris & Ollendick, 2015).
Several other sociodemographic and environmental variables might also be associated with children’s risk for SM, including parental educational attainment and major life events such as relocation (Black & Uhde, 1995; Morris et al., 2017; Sonego et al., 2013). Nevertheless, there is a dearth of knowledge on the relationship between these factors and the onset of SM. Chavira et al. (2007) has reported that shyness was more common in parents of children with SM, but there have been no other studies exploring the association between SM and parents’ personality traits. Introversion in parents, in particular, could be an important trait to investigate further. Prior research found that personality traits are at least partly inherited in anxiety disorders (Brandes & Bienvenu, 2006). Children with behavioral inhibitions tend to be introverted and shy when faced with unfamiliar situations. This highlights the potential correlation between introversion in parents and SM in offspring, a relationship that has yet to be explored in the existing literature.
In terms of treatment, the international literature suggests that behavioral or cognitive-behavioral interventions and pharmacotherapy with selective serotonin reuptake inhibitors are effective for treating SM (Muris & Ollendick, 2015). Individualized treatment plans with the combined effort of teachers, parents, and clinics have been recommended to help children overcome SM (Dow et al., 1995). However, there are scanty quantitative data on how families with SM children cope with SM and how they perceive the effectiveness of SM treatments.
Selective Mutism in China
China has the second-largest number of children in the world. In 2015, there were 271 million children under the age of 17 years in China, accounting for 20% of the population of China and 15% of the world’s child population (UNICEF, 2018). Therefore, although SM is a relatively rare condition, China likely has many children living with SM due to its large base population. Due in part to such concerns, the Selective Mutism Association of China (SMAC; https://www.selectivemutism.org.cn/) was launched in 2018 (SMAC, 2021). The SMAC is a non-governmental organization composed of professional therapists and parents. The goal of the SMAC is to spread knowledge about SM, raise public awareness, and help SM patients and families receive timely SM support and treatment (SMAC, 2021).
Children in China have potentially unique triggers for SM. Using local language dialects for communication at home is one potential trigger, as there are diverse dialects and most are quite different from Mandarin, the official language used in public and official settings. Additionally, parental divorce is another potential trigger for SM, as the divorce rate in China has been rising in recent years, and parental divorce can increase children’s anxiety levels (Tahirović & Demir, 2017; Zhang et al., 2014) and the likelihood of a child growing up in a single parent household. Moreover, it is estimated that 69 million children in China are left behind in the care of their extended family and friends as their parents move to other cities for employment (UNICEF, 2020). Furthermore, the unique One-Child Policy implemented between 1980 and 2015 makes China the country with the highest number and percentage of only-children globally (Lin et al., 2021). Being a left-behind child and being an only child may also be a trigger for SM since parental companionship and siblings play a vital role in children’s cognitive development and mental health (Brody, 2004; Tahirović & Demir, 2017).
Nonetheless, there remains limited knowledge about SM in China. To the best of our knowledge, there has been virtually no quantitative SM research in China, and no systematic data on the experiences and treatment of children and families dealing with SM in China has been published.
Rationale and Objectives of the Study
To better understand SM in China and its correlates, this nationwide case-control survey study aims to: (1) describe the sociodemographics and experiences of children and families with SM in China and (2) identify sociodemographic and environmental correlates of SM in China.
Methods
SM Parental Survey Recruitment
This case-control study was conducted in partnership with SMAC. We recruited case participants (i.e., children suspected of ever having SM) by disseminating the online survey questionnaire in a WeChat group organized by SMAC (WeChat is a Chinese messaging and social media application). In addition, we recruited control participants (i.e., children not suspected of ever having SM) by using the survey service offered by Wenjuanxing (WJX), a professional online questionnaire survey platform. The Wenjuanxing sample database covers more than 2.6 million respondents who register under real names and have diverse socio-demographics (J. Wang et al., 2020; Wenjuanxing, 2022).
SM Case Definition
The DSM-5 is the latest updated guideline for diagnosing mental disorders developed by the American Psychiatric Association (APA) in 2013 (APA, 2013). In our survey, we translated the DSM-5’s diagnostic criteria of SM into Chinese, and parents were asked if they felt their child ever met the criteria of SM. Possible responses were yes, no, and unsure. Parents recruited from the SMAC WeChat group and answered “yes” or “unsure” were defined as parents of cases. The cases involved children suspected of having selective mutism (SM) based on their parents’ understanding of the children’s behavior according to the DSM-5 definition.
Inclusion Criteria
(1) Currently residing in China; (2) Having at least one child currently between the ages of 2 and 17 years who currently resides in China; (3) Being able to provide informed consent.
Only participants who met all of the above criteria were enrolled in the survey. Before filling in the questionnaire, each participant completed an informed consent form.
Recruitment of cases started on March 28th, 2022 and ended on April 18th, 2022. Recruitment of controls began on April 13th, 2022 and ended on April 18th, 2022. Each of the study participants (i.e., parents of children) reported information for one or two children in their family. We stipulated that each family should only complete one questionnaire.
Measures
Online questionnaires collected information about parent and child sociodemographics, and whether or not children exhibited symptoms of SM. In addition, the survey for cases included more detailed questions about each child’s SM diagnosis, severity, and impact of SM on parental and child stress.
Selective Mutism Questionnaire (SMQ)
The SMQ is a 17-item parent-report measure of children’s speaking behavior developed by Bergman et al. (2008). The frequency of a child’s speaking behavior is rated by the parent using a scale ranging from 0 (never), 1 (seldom), 2 (often), to 3 (always). Lower scores represent higher speech inhibition (Bergman et al., 2008). In typically developed children (i.e., children without SM) and children with SM, the mean scores were ≥2.5 and 0.5, respectively (Oerbeck et al., 2020), respectively. Based on the 17-item SMQ, Letamendi et al. (2008) published a 13-item SMQ, where four other items with low interpretability or relatively high secondary cross-loadings were dropped. The 13-item SMQ had internal consistency ranging from moderate to high and good convergent and incremental validity (Letamendi et al., 2008). We translated the 13-item SMQ into Chinese and used it in the survey. To ensure the accuracy of the translated questionnaire, the initial translation from English to Chinese was conducted by a bilingual individual fluent in both languages. To validate the accuracy of the translation, a back-translation method was used. A separate translator, who was unfamiliar with the original questionnaire, independently translated the text back into English. The back-translated text was then compared with the original questionnaire to identify and resolve any discrepancies.
We examined the dimensional structure of the translated 13-item SMQ by conducting an Exploratory Factor Analysis (EFA) using principal axis factoring with the rotation method of Varimax with Kaiser Normalization. We then assessed the internal consistency by computing Cronbach’s alpha estimates for the total scale and different factors.
Table 1 shows the items and factor loadings for the translated 13-item SMQ. Factor 1 (School) contained five items with value loadings from 0.675 to 0.853. Factor 2 (Outside of School) included three items with value loadings from 0.624 to 0.809. Factor 3 (Home and Family) included five items with value loadings from 0.639 to 0.816. Table 2 shows Cronbach’s alpha coefficients for each factor and the total SMQ scale. Cronbach’s alpha values for each factor ranged from .60 to .86, indicating acceptable to good reliability. Furthermore, the Cronbach’s alpha for the total scale was .80, indicating good internal consistency of the translated SMQ.
Items and Factor Loadings for the Translated 13-Item SMQ.
Note. Bold font indicates the primary factor loading for each item.
Cronbach’s Alpha for the Translated 13-Item SMQ.
In addition, the SMQ contains items regarding the stress of children’s current SM behavior for children and parents. A different scale is used for these items (0 = Not at all, 1 = Slightly, 2 = Moderately, 3 = Extremely; Bergman et al., 2008). We used two items that assessed how stressful the child’s SM behavior was for the parent and child.
Parental Stress
Parental stress was measured by participants’ five-point response (from “strongly disagree” to “strongly agree”) to the statement “having children leaves little time and flexibility in my life.” This measure came from one item from the parental stress scale developed by Berry and Jones (1995).
Statistical and Thematic Analyses
To describe the sociodemographics and experiences of children and families with SM in China, a descriptive analysis was performed. t-Tests were conducted to assess differences in speech inhibition between children living with SM who did and did not receive treatment. In addition, a thematic analysis was conducted to analyze treatment experiences by coding open-ended qualitative responses and grouping the codes into themes. The procedures of the thematic analysis included: (1) familiarization of the data, (2) creating initial codes, (3) generating themes, (4) reviewing themes, (5) defining and naming themes, and (6) finalizing themes and generating conclusions.
To identify sociodemographic and environmental correlates of SM in China, a multilevel binary logistic regression with random intercepts was performed, with individual children clustered within families. Multicollinearity was not detected among the potential correlates. The measures of association between the examined factors and SM onset were presented as odds ratios (OR) with 95% confidence intervals (CI). Two-sided p-values < .05 were considered statistically significant.
Ethical Assessment
The research protocol was approved by the ethics review board of Xi’an Jiaotong-Liverpool University (#21-02-43).
Results
Sociodemographics and Experiences of Children and Families with SM in China
The sample included 172 cases of children suspected of ever having SM and 179 controls who were not suspected of ever having SM. Most participants were female (89.0% of cases; 79.3% of controls) and most households had two parents with a bachelor’s degree (63.4% of cases; 74.3% of controls). Parents of cases were more likely to be over 40 years old, compared to parents of controls (23.8% vs. 8.4%). Additional sociodemographic information is presented in Table 3.
Characteristics of Children and Their Parents.
Sibling(s) aged between 2 and 17, currently residing in China.
SM Severity, Treatment, and Impact
Among 172 case children with suspected SM, 51.2% were diagnosed by a professional and 167 (97.1%) had symptoms that had been detected by the participants. The mean values, standard deviations, and ranges of the mean scores of the 13-item SMQ were 1.19 ± 0.43 (0.31–2.46) among 94 children who had received treatment and were 1.16 ± 0.40 (0.15–2.00) among 78 case children who had not received treatment. There was no significant difference in the mean values of SMQ mean scores between these two groups (p = .674).
Among 167 children whose symptoms had been previously detected by participants, 89 (53.3%) had received treatment; 52 (31.1%) had received treatment under the instruction of a professional; the percentages of children who received very effective, effective, and ineffective professional treatment based on parents’ self-assessment were 20.2%, 53.9%, and 25.8%. Based on the thematic analysis of answers to an open-ended question regarding SM children and their families’ treatment experiences (see Table 4 for illustrative quotes), it appeared that few educational institutions in China provide special support for children with SM (“There is no government educational institution that specially supports SM children. Therefore, the implementation of relevant measures depends on luck to see if you can meet good teachers”). In addition, some participants indicated that there is a lack of professionals experienced with SM (“The therapist has no experience with selective mutism.”). Among participants whose child had received professional SM treatment, multiple participants cited the use of Cognitive-Behavioral Therapy, which several regarded as an effective method (“Doing Cognitive-Behavioral Therapy and making significant progress.”). Sandplay therapy, which is a psychotherapy method where clients are asked to play in a specially proportioned sandbox (Kalff, 1991), was another relatively common treatment method for SM among the participants. One participant felt that Sandplay might be of benefit, but apparently stopped because of cost: “I personally think the sandplay therapy can play some role, but it is expensive, and finally, I chose not to use it..” Another participant concluded that Sandplay helped their child relax, but had limited effectiveness on SM: “Sandplay therapy. I feel that it is not effective and does not target the symptoms, but it’s relaxing to some degree.” Lastly, a third participant reported that Sandplay had no obvious effect on SM after multiple sessions: “Sandplay therapy for about eight times, no obvious effect.”
Thematic Analysis on Treatment Experiences.
The open-ended question was “please describe what kind of treatment your child has received, children’s treatment experiences, and your views on it.” The original answers were in Chinese.
In terms of SM as a source of stress, the parents of 167 children with SM (97.1%) reported themselves as being moderately or extremely stressed by their child’s SM; while 137 (79.7%) reported that their children were moderately or extremely stressed by SM. Compared to parents of children without SM, the parents of SM children were more likely to agree or strongly agree that having children leaves little time and flexibility in their lives (50.6% vs. 35.8%).
Sociodemographic and environmental correlates of SM in China
Table 5 presents unadjusted and adjusted measures of association between SM and select correlates. Results of adjusted multilevel regression analysis indicated that compared to children with no introverted parent, children of parent(s) with an introverted personality had greater odds of having SM (Adjusted Odds Ratio (AOR): 15.64, 95% Confidence Interval (CI): [6.18, 39.58] if both parents were introverted; AOR 2.25, 95% CI: [1.07, 4.75] if one parent was introverted). Children born to mothers aged > 35 at time of birth had significantly greater odds of SM (AOR: 4.74, 95% CI: [1.12, 20.12]). Children with a sibling had significantly greater odds of SM (AOR: 2.45, 95% CI: [1.36, 4.41]) than those without a sibling.
Participant Characteristics and Correlates of Childhood Selective Mutism.
Sibling(s) aged between 2 and 17, currently residing in China.
p < .05. **p < .01. ***p < .001.
Discussion
To the best of our knowledge, this is the first quantitative study to describe the sociodemographics and experiences of children and families with SM in China. Our findings revealed that only about half of SM cases were diagnosed by a professional, and less than one in three SM cases received treatment guided by a professional. In addition, there appears to be a paucity of trained professionals experienced with SM and special support from educational institutions. Although children with special needs have been increasingly educated in mainstream classrooms within China, mainstream public schools often lack the resources and organization needed to meet such students’ needs (S. Q. Xu et al., 2018). The relatively large size of classrooms (often ≥ 40 students in urban settings; Y. Wang et al., 2015) constrains the time regular classroom teachers can devote to students with special needs. Teachers tasked with supporting students with special needs often have limited special education training and may need to take on such responsibilities in addition to other job duties, but without additional financial compensation (X. Xu & Olli-Pekka, 2015). In more developed regions, private professional therapists can be hired to implement SM treatment programs in school settings, but such services may be prohibitively expensive for many if not most families. Due to the fact that many children with SM do not exhibit any intellectual disability, there is a very real risk that educators in mainstream schools may incorrectly attribute SM behavior to shyness or introversion, rather than a recognized anxiety disorder (Perednik, 2016). In fact, anecdotal evidence from the SMAC suggests that many children in China living with SM demonstrate relatively high academic achievement. Greater access to special education resources and SM interventions are needed in China.
Results from our study also indicated that SM is a substantial source of stress for both parents and their children in China. Part of this stress may be exacerbated by parent-teacher relations, wherein parents are expected to be personally engaged with their child’s studies and provide at-home support as the teacher requests (Y. Wang et al., 2015). It is conceivable that homework and at-home “make-up” assignments may be greater for students with SM who are unable to participate in school activities that require verbal communication. Given that speech becomes increasingly salient for developing social bonds and communication skills as the child ages, it is likely that SM-attributed stress will only increase over time unless SM symptoms are resolved. In the interest of the well-being of the child and their parents, it is imperative that SM treatment begin as soon as possible for the child (Johnson & Reitman, 2023; Perednik, 2016; Reitman et al., 2023). Chinese language resources for therapists and parents of children with SM are growing, but have been relatively sparse to date with a few exceptions (SMAC, 2022).
Correlates of SM
Previous research from outside Asia has suggested that genetics and parental personality have a strong influence on SM onset (Muris & Ollendick, 2015). Findings from our sample from China corroborate these previous studies, with SM cases having significantly greater odds of introverted parent(s). However, contrary to our original expectations, having a sibling was positively associated with SM. This association could be confounded by unmeasured effects of socioeconomic status, but could not be assessed in the present study. As more families have multiple children following the end of the “one-child policy,” future research should examine how sibling dynamics may or may not be influencing SM onset in China.
Advanced maternal age was strongly associated with SM in the current study, but we can only speculate upon the mechanism of association. A large nationwide study from Finland found advanced maternal age to be significantly associated with SM in unadjusted models, but was no longer significantly associated with SM after the psychopathology of parents was controlled for (Koskela et al., 2020). It is conceivable that parental psychopathologies can be confounding the association between advanced maternal age and child SM, but it is also possible that maternal psychopathologies are mediating the effects of advanced maternal age on child SM (Muraca & Joseph, 2014). Longitudinal studies are needed to disentangle the relationship between maternal age, maternal psychopathologies, and SM in offspring.
Limitations
The study has several limitations. First, although case-control studies are efficient for studying rare conditions, there are several drawbacks inherent to this design. Due to its retrospective nature, case-control studies can only be used to identify associations between exposures and outcomes instead of the causation between them (Sedgwick, 2014; Tenny et al., 2022). Therefore, future studies (ideally longitudinal cohort studies complemented by qualitative methods such as observation and personal interviews) are needed to investigate the causal relationships between the identified correlates and children’s SM. The potential for recall bias is another serious weakness of case-control designs, and we are unable to rule out potential differential exposure misclassification. It is possible that parents of SM cases were more likely to over-report known exposures of SM.
Selection bias also could not be ruled out. For example, a majority of participants were female and had a Bachelor’s or above degree, and over one in three participants resided in Beijing, Shanghai, Guangzhou, and Jiangsu, the most economically developed regions in China. Hence, in our samples, those who were male and had a relatively lower socioeconomic status might be under-represented. This may have caused an overestimate of SM children receiving treatment, as children from higher socioeconomic status families usually have relatively greater access to fee-based medical and social services. In addition, potential for self-selection bias should also be noted. Parents of children with serious SM might opt out or drop out of the survey due to greater time constraints. This could result in children with more severe SM and their families being underrepresented. Future SM studies in China should attempt to recruit samples across more diverse sociodemographic subgroups.
Lastly, the introverted personalities of parents were only assessed with a single, self-reported item. Future research should consider using validated, multi-item personality tests to assess the personalities of parents.
Conclusion
Access to SM treatment in China remains low, and is likely due to lack of awareness and insufficient resources at medical and educational institutions. This is problematic because SM has adverse implications for childhood development and appears to be a concerning source of stress for children and their parents. Given the severe dearth of SM research from China, it is incumbent for Chinese researchers, educators, clinicians, and policymakers to collaboratively develop novel approaches to more effectively detect and treat SM.
Footnotes
Acknowledgements
Many thanks to the participants for making this study possible. Much appreciation also to Yvonne Lu for her feedback on an earlier version of this manuscript.
Ethical Considerations
The study protocol was approved by the Xi’an Jiaotong-Liverpool University ethical review board (protocol # 21-02-43).
Consent to Participate
Consent to participate was obtained from all study participants.
Author Contributions
XH drafted the manuscript, conducted the analysis, and provided substantial contribution to interpretation of data and revising the manuscript for important intellectual content. JR provided substantial contribution to data collection, interpretation of data, and revising the manuscript for important intellectual content. SWP conceptualized the study, provided substantial contribution data collection, co-drafted the manuscript, provided substantial contribution to data collection, interpretation of data, and revising the manuscript for important intellectual content. All authors approved the final version of the manuscript to be published and accept accountability for all aspects of the manuscript.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this study was provided by Xi’an Jiaotong-Liverpool University.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
Data not publicly available.