CHAPTER 5
Selective Mutism and
Communication Disorders
Exploring Co-morbidity
The essence of Selective Mutism (SM) is the absence of speaking in certain situations. Communication disorders such as speech, language and fluency problems may accompany SM. The communication disorder may also contribute to the development of the selectively mute behaviour. An example would be Peter in Chapter 10, whose cleft palate speech may have influenced his willingness to talk in public. The presence of more than one disorder in the same person is called co-morbidity, although the term multi-morbidity has been used recently (Smith 2009). This chapter will explore the existence of communication disorders in children with SM.
Prevalence of communication disorders
It is not entirely clear how many children with SM have a communication disorder. Some studies (e.g. Black and Uhde 1995) have found this occurred in 10 per cent of cases; others (e.g. Krohn et al. 1992; Kumpulainen et al. 1998) have reported between 30 per cent and 40 per cent of children affected; while more studies (Dow et al. 1995; Kolvin and Fundudis 1981; Kristensen 2000) have shown that around half of all children with SM will have some type of communication disorder. The discrepancy between 10 per cent and 50 per cent may occur for several reasons, including small sample size, the population the sample was drawn from (i.e. clinical or community), the study’s inclusion criteria, and how the diagnosis was made (e.g. case notes, parental reports, and/or standardized assessments). Furthermore, correctly identifying a communication disorder in children who are usually silent is challenging, although some suggestions will be explored later in this chapter. Currently, there is no consensus about how to test these children’s speech and language skills, although it is interesting that studies (e.g. Dow et al. 1995; Kolvin and Fundudis 1981; Kristensen 2000) that involved standardized assessments have identified more cases. A recognized formal and comprehensive approach to assessing the communication skills of this population of children is long overdue.
Types of communication disorders
Research into the communication disorders that accompany SM is a work in progress. Early studies have tended to focus on numbers of children affected while some recent papers have provided more detail.
Speech and language
General descriptions are usually given, such as: speech/articulation difficulties, that is, pronunciation problems (Kolvin and Fundudis 1981; Krohn et al. 1992) and, language problems (Black and Uhde 1995; Kumpulainen et al. 1998; Wong 2010); and speech and language disorder, that is, difficulty using and/or understanding words and grammar (Steinhausen and Juzi 1996; Wong 2010). Not surprisingly, when standardized assessments are administered, more specific information is recorded, such as: phonological delay (Kristensen 2000), which is an immature sound system in word production, for example, ‘tar’ for ‘car’, ‘pider’ for ‘spider’; expressive and/or receptive language delay/disorder (Dow et al. 1995; Kristensen 2000); and problems with expressive narrative discourse (Klein et al. 2012; McInnes et al. 2004). Cases of stammering are also mentioned (e.g. Schwartz et al. 2006; Wright et al. 1985). For a discussion on stammering, see also Chapter 7. Few research studies have focused just on the communication skills of children with SM when they are speaking – for instance, at home. One exception is an Australian study conducted by the author (Cleator 1998) involving five subjects. Four of the five subjects (three boys and one girl) were found to have a communication disorder, which included: articulation errors; phonological delay/disorder; receptive language disorders; and expressive language problems involving syntax (grammar) and semantics (word meaning). There is still much research needed into this area but, for now, while any communication disorder may be present there is a tendency for problems involving pronunciation and expressive language to be more common.
Children who have a history of speech and/or language problems can also be at risk for difficulties acquiring literacy skills (Harrison et al. 2009; Serry, Rose and Liamputtong. 2008). Some studies (e.g. Manassis et al. 2007; McInnes and Manassis 2005) have begun to examine this issue as part of their research. To date, there has been a focus on phonemic awareness (the ability to hear, identify and manipulate the sounds in words), which is an important predictor of literacy success. For example, Manassis and colleagues (2007) assessed phonemic awareness as part of a Canadian study that looked at language, cognition and anxiety. They found that the children with SM did not perform as well in this area (and all other language measures) when compared with children in the comparison groups (i.e. those with anxiety but without SM, and a control group). Recognizing that a child with SM may be experiencing additional difficulties with reading, writing and spelling is important because these skills, which can be a source of anxiety at school, are often used during treatment. Currently, literacy remains a largely unexplored area of research in SM.
Voice
Throughout the literature on SM there are references to these children’s vocal quality, although actual voice disorders seem not to be reported. Descriptions vary, usually reflecting the setting where the child is speaking. While some parents say their child’s voice is loud at home, away from home it’s a different matter. Some researchers (Cleator 1998; McInnes and Manassis 2005) report problems with prosody such as volume, intonation and pitch; while others (Kolvin and Fundudis 1981) mention children who say they are afraid of hearing the sound of their voice, or their voice sounds funny (Black and Uhde 1992). Precisely what gives rise to these impressions is difficult to say, but they are not uncommon, so much so that Hayden (1980), in a classification of four types of SM, created a category which she named ‘speech phobic mutism’ specifically for children who were frightened of hearing their voice. There is no doubt that anxiety may impact on vocal quality (Kagan et al. 1987), although the role anxiety plays in the voice of children with SM remains unexplored. Notably, the recent edition of the Diagnostic and Statistical Manual of Mental Disorders (APA 2013), hereafter DSM-5, has classified SM as an anxiety disorder.
Pragmatics
Of all the communication disorders found in SM, problems with the use of language in social contexts, sometimes referred to as pragmatics, requires the most attention because it is this ability that affects children’s inclusion in, or exclusion from, social and academic life. Although it is often difficult to unravel communication behaviours that are attributable to an actual pragmatic disorder from those associated with selectively mute behaviour, these children’s communication patterns will reveal both verbal and non-verbal pragmatic problems. For example, opportunities to practise skills such as greeting people, saying ‘thank you’ and ‘sorry’, and answering questions, may be absent and possibly undeveloped in a child who is usually silent outside the home.
Non-verbal pragmatic skills may be affected as well, although there seems to be wide individual variation among reported cases. For instance, while some children use gesture to communicate, others avoid eye gaze and have limited facial expression (Lebrun 1990) or appear frozen with fear (McInnes and Manassis 2005) and so anxious they seem physically unable to interact at all (Kanehara et al. 2009).
Having said that, there may be a marked discrepancy between social interactions at home and at school, which is important diagnostically for ruling out or clarifying the presence of a pragmatic disorder. For example, while some children will be inhibited at school, at home they may display impulsive and/or oppositional verbal behaviour (Cunningham et al. 2004) with some parents reporting that they are overly talkative (Lebrun 1990). By contrast, other children may continue to be reticent at home, often reflecting their family’s culture of quietness (APA 2013; Steinhausen et al. 2006). This is one of the problems in researching SM because there is considerable individual variation (Cline and Baldwin 2004). Studies often do not include large numbers of cases because the condition is rare, which results in conflicting findings that make it difficult to generalize to a wider population.
Speech and language development and onset of selectively mute behaviour
Most parents of children with SM report no problems with their child’s early speech and language development (i.e. during the first year). It is when the child starts using language to communicate during the second year that delay is sometimes mentioned (Kolvin and Fundudis 1981; Toppelberg et al. 2005). For example, Kolvin and Fundudis (1981) reported that the 24 children with SM in their English study began using phrases five and a half months later (27.3 months compared with 21.9 months) than the 102 matched control group; and Cleator (1998) found that four of the five subjects in her study had a history of delayed speech and language development.
Of further interest in the (admittedly small) study conducted by the author was the parents’ observation that, for no apparent reason, the speech and language development in three of the four subjects appeared to stop between the ages of 12 and 18 months and, although it had resumed, had never become normal.
Parents usually state that the selectively mute behaviour started in the pre-school years, which coincide with a period of rapid growth in a child’s speech and language development. It is at this time that communication skills are at their most vulnerable and problems such as stammering are most likely to develop. It is possible that some form of disruption or stress during this period may impinge on the child’s developing communication skills. For instance, separating from a caregiver is a challenging experience for most children who are often shy and withdrawn when they start daycare or pre-school, with some experiencing separation anxiety (Black and Uhde 1995; Steinhausen and Juzi 1996) after being left on their own, often for the first time. For most children, these feelings pass. However, for some, remaining silent may help them reduce feelings of anxiety and perhaps signal the beginning of the selectively mute behaviour.
With transition to school, the child’s coping skills will once again be challenged. Whereas at home a child can speak spontaneously, it is the teacher who controls most of the talking in the classroom. To compound the problem, differences exist between language spoken at home and at school. For example, children have to learn how to answer the teacher’s questions, and stand up in front of the class and give news. They also have to learn to master narrative discourse such as telling stories, sharing experiences and giving explanations (Sage 2004). Not only will being silent reduce opportunities to learn and practise these linguistic skills (Klein et al. 2012; McInnes and Manassis 2005) but there will be social implications, too. For instance, the child will miss out on chances to develop relationships through activities involving language such as negotiating with peers, discussing events or programmes on TV, and telling jokes. Some children may even be teased (Cline and Baldwin 2004; APA 2013) or picked on by other children because they are unable to protect themselves verbally (Manassis 2009). It is not difficult to see how a child with SM and a communication disorder will be doubly disadvantaged.
A communication disorder as an area of vulnerability
Why might children develop a fear of speaking? One answer is that, for some, talking is an area of vulnerability because of a communication disorder. This view is not new. As early as 1912, Gutzman (1912, cited in Kratochwill 1981) proposed that SM may develop as a reaction to being teased, criticized or not understood because of a speech and/or language problem. More recently, others (e.g. Elizur and Perednik 2003; Klein et al. 2012) have supported the idea that a communication disorder could be a factor in the development of the selectively mute behaviour. Certainly, children who have difficulties expressing themselves verbally and whose speech is sometimes unintelligible are likely to feel worried about speaking, especially at school. Such children may remain silent in an attempt to reduce feelings of anxiety (Anstendig 1998). Indeed, the development of anxiety along with social isolation, withdrawal and reticence has been found among children with language problems (McInnes and Manassis 2005), especially those with expressive language delay (Irwin, Carter and Briggs-Gowen 2002).
Meanwhile, other researchers (Elizur and Perednik 2003; Kristensen 2000) have gone further and suggested that the communication disorder could be part of an underlying neurodevelopmental immaturity that makes the selectively mute child vulnerable to everyday challenges. For example, Kristensen (2000), in a Norwegian study of 54 children with SM, found that 68.5 per cent met the criteria for a diagnosis of developmental delay/disorder (motor, language) compared with 13 per cent of the 108 children in the control group. The observation that multiple factors may be implicated in the development of the selectively mute behaviour led Steinhausen and Juzi (1996) to propose a vulnerability model for the development of the disorder. There will be important treatment implications if this is the case, because the silent behaviour may be concealing other co-morbid conditions such as a disabling communication disorder (Kristensen 2000). By attempting to treat just the selectively mute behaviour, clinicians will be missing other problems which require attention if a full recovery is to be possible (Elizur and Perednik 2003). In this connection, it is worth mentioning the issue of gender. SM seems to occur more often among girls (Kumpulainen et al. 1998; Steinhausen and Juzi 1996) whereas communication disorders are more commonly found among boys (McLeod and McKinnon 2007). It therefore seems likely that more boys than girls with SM will have a communication disorder and clinicians should be mindful of this. The need for future studies to identify the sex of subjects with co-morbid problems is recommended.
Before proceeding further, it is relevant to look at two groups where there is an increased incidence of SM: children from migrant families and twins.
SM and children from migrant families
Studies (Elizur and Perednik 2003; Toppelberg et al. 2005) have clearly demonstrated that children from migrant families are more at risk for developing SM than children from non-migrant backgrounds. Some children either do not learn the native language prior to school entry or have difficulty mastering the new language, especially if they have a communication disorder. In many ways, early second language acquisition resembles a communication disorder with the likelihood of mispronunciations, grammatical errors and prosodic deviation (e.g. speaking with an accent). It is easy to see how concerns about being teased for the way they speak might make a child feel shy and self-conscious about talking, especially in front of peers (McInnes and Manassis 2005). Some children may even stop speaking for a while (Toppelberg et al. 2005), although the duration and even existence of what has sometimes been referred to as ‘the silent period’ has been questioned by some (Gibbons 1985). The migration process is complex so it would be misleading to suggest that the increased incidence of SM among children from migrant backgrounds is due solely to issues involving second language acquisition. Nevertheless, it is entirely understandable how learning a second language, together with starting school, could be challenging experiences for some children.
SM and twins
While twins represent around 1.1 per cent of all live births, the occurrence of SM among twins in the general population is difficult to calculate. Cline and Baldwin (2004), using the numbers of twins featured in the published (English) literature on SM, estimated a minimum incidence of around 4 per cent. Twins with SM usually speak to each other, but sometimes not at home with family members (Wallace 1986). Obviously, twins have a different language experience from singletons. For example, two-way communication is regularly extended to include a third party, thereby reducing occasions for one-on-one conversational exchanges; and twins spend a lot of time speaking to one another, possibly limiting the scope of their language experience.
While twins’ language acquisition usually follows a similar pattern to that of singletons, delay between the ages of two and four has been reported, although this has usually resolved by the time the twins reach the ages of seven and eight (Crystal 1997). In cases where there is a communication delay or disorder, twins may inadvertently reinforce each other’s non-standard speech and language behaviour, with some continuing to use these features beyond an age when they are appropriate. Indeed, the development of so-called secret languages may, in fact, emerge from these communication patterns (Crystal 1997). It is possible that some of the different language behaviours observed among twins may become a means for maintaining the self-sufficiency of the twinship and contribute to the formation of SM in some cases.
Pathways and a biopsychosocial model
So far, this chapter has focused on SM and communication disorders. It would be misleading to imply the presence of just one co-morbid condition, because any number may be present (Kristensen 2000; Manassis et al. 2007). For example, in a Japanese study conducted by paediatrician Kanehara et al. (2009), the following co-morbid conditions in 23 cases of SM were found: anxiety disorders (73.9%); developmental disorder (60%); other disorders such as bedwetting. (The study was published prior to the classification of SM as an anxiety disorder in DSM-5.) This was a clinical sample and the ‘symptom burden’, as Kristensen (2000) has called it, among cases in the wider community may not be as heavy. There is no doubt that some cases of SM have a greater number of co-morbid conditions than others, and with this comes complexity. In such cases the term multi-morbidity seems more appropriate (Smith 2009).
The complexity often found in cases of SM has led some researchers to try to harness the elusive nature of the disorder by creating sub-groups. Some studies (Cohan et al. 2008; Kristensen and Torgersen 2001) have included the presence, or not, of a communication disorder. For example, Cohan and colleagues (2008) identified three groups of children: (1) those who were mildly anxious when pressured to speak; (2) those who were just anxious; and (3) those who were anxious and had delayed communication skills. Cohan and associates suggested that this latter group could be ‘the most severely impaired’ (Cohan et al. 2008, p.780). More research is required to explore this issue.
Children with SM are a diverse population, so creating sub-groups has its difficulties. Another approach is to view the behaviour – that is, not talking – as a symptom of a wide range of disturbances, and the diagnosis of SM an umbrella term for grouping together children with differing pathologies (Anstendig 1998; Kumpulainen et al. 1998). Indeed, the presence of multi-morbidity in SM has led several writers (Elizur and Perednik 2003; McInnes and Manassis 2005) to suggest that there may be different pathways leading to the development of the silent behaviour. In order to conceptualize these pathways, a broad framework is helpful, especially clinically. One such framework is the biopsychosocial model proposed by Engel (1980) which examines biological, psychological, and social factors (see Figure 5.1, page 92).
Figure 5.1 The biopsychosocial/cultural model
Recently, the model has been extended to include cultural factors, making its application to SM all the more pertinent because of the often compelling role cultural factors play, especially among children from migrant backgrounds. Applying a biopsychosocial model should clearly demonstrate where the ‘symptom burden’ (Kristensen 2000, p.254) lies. For example, a child with neurodevelopmental immaturity will have more biological features than a child from a migrant background who will likely have more social/cultural factors. This will assist in preparing each child’s unique formulation during assessment.
Assessment and treatment
Children with SM are best managed by a multi-disciplinary team (Dow et al. 1995; McInnes et al. 2004). This could include a paediatrician, clinicians with a background in mental health such as a psychiatrist and/or clinical psychologist, and a speech and language therapist.
Assessment
Assessment usually involves a full case history from parents so that issues such as the child’s developmental milestones, medical history, patterns of socialization and progress at school can be explored. If the history reveals the existence of other conditions such as developmental and/or behavioural problems, referral to an appropriate specialist can be made (Keen, Fonseca and Wintgens 2008). As far as the child’s speech and language development is concerned, it is not adequate to rely solely on parental reports, and referral to a speech and language therapist is essential (Dow et al. 1995; Wong 2010), especially when the parent interview reveals a family history of communication disorders (Cohan et al. 2008; McLaughlin 2011). If the child is from a migrant background, it is necessary to make sure sufficient time has passed to allow familiarity with a new language to develop. If the child has had at least six months’ exposure to a new language but is still mute, a bilingual speech and language therapist who speaks both languages, or an experienced interpreter, should be found to conduct the assessment (Toppelberg et al. 2005). The child’s hearing should also be tested (Cleator 1998; Wong 2010) and an oromuscular assessment carried out, although some children with SM are hesitant about opening their mouths (Cleator 1998; Lebrun 1990), so sensitivity is required.
Conducting a comprehensive speech and language assessment of children who are usually silent is complicated. Nevertheless, an assessment is possible if it is conducted under certain conditions. Administering standardized language comprehension tests is usually straightforward because the child does not have to speak. However, assessing the speech and expressive language of children who are usually silent involves some creativity, especially on issues such as who does the testing and where.
Some successful methods include:
•obtaining from the family audio- and/or videotapes of the child speaking at home for later transcription and analysis (Cleator 1998; McInnes and Manassis 2005)
•training parents to administer tests either at home or in the clinic using a one-way mirror (Cunningham et al. 2004; Klein et al. 2012; McInnes and Manassis 2005)
•conducting the assessment in the setting where the child usually speaks and feels comfortable, namely their home (Cleator and Hand 2001; Cohan et al. 2008).
Whether a child will speak during assessment in a clinical setting is unpredictable. Klein and colleagues (2012) looked at this issue and found that all 33 children in their American study spoke during testing when they were alone with their parents but only 18 spoke when the tests were administered by a speech and language pathologist. The study also found that children scored higher when tests were administered by parents. It is not always possible to obtain an accurate picture of the child’s true communication skills during the first contact because their performance may be influenced by anxiety and behavioural inhibition, so further attempts may be necessary. A number of researchers (Dow et al. 1995) have suggested tests that could be used with this population of children, while others (e.g. McInnes and Manassis 2005) have provided information for conducting a comprehensive, standardized assessment of these children’s speech and language skills. Furthermore, there is no doubt that social media (e.g. Skype, Vine) will make an increasing contribution not only to their assessment but to their treatment, too.
Treatment
Treatment of the selectively mute behaviour will ideally involve a multi-disciplinary team that includes members of the child’s family and their teacher. Obviously, the teacher, who is often most affected by the selectively mute behaviour, will play a critical role during treatment. Unfortunately, little research has focused on exploring that role from a teacher’s perspective.
Therapy for the communication disorder may begin and continue simultaneously with treatment of the selectively mute behaviour (Klein et al. 2012; McInnes and Manassis 2005). Having said that, some clinicians (McInnes and Manassis 2005; Smayling 1959) have found that focusing just on the communication disorder can have positive effects on reducing the selectively mute behaviour itself. This finding is intriguing. It is beyond the scope of this chapter to explore what mechanisms could be involved here but, it is hoped, further clinical experience and research will shed light on the issue.
Even if the child does not have a diagnosable communication disorder, a case may be made for involving a speech and language therapist in developing the child’s essential pragmatic skills so they gain confidence speaking socially (McLaughlin 2011). This may include practising routines such as offering an opinion, asking questions, interrupting politely and taking conversational turns in an acceptable way (Dow et al. 1995). Even after the elimination of selectively mute behaviour, support and encouragement is likely to be necessary so the children consolidate their gains and continue to develop communication skills as they increase the number of people they speak to and where they speak. Thus, maintenance of a level of communicative competence and generalization of speaking to other settings will require ongoing support for some children.
A final word on intervention
Early intervention is essential in SM (Kanehara et al. 2009; Keen et al. 2008; Schwartz et al. 2006). Unfortunately, there are many instances reported in the literature (Kumpulainen et al. 1998; Steinhausen and Juzi 1996) where a substantial time-lag exists between the onset of the selectively mute behaviour and beginning of treatment. For example, Schwartz and colleagues (2006) found a period of 11.5 months between the time when parents voiced their concerns to their doctor and when a diagnosis was made. They also found that 69.7 per cent of the children in their American study had never been diagnosed correctly nor referred for intervention. One of the reasons may be acceptance of shyness and reticence in children at the time the behaviour is usually first noticed (i.e. aged between three and five years).
Clearly, most children overcome being silent but a small number do not. Regrettably, at the moment there appears to be no way of distinguishing between the two groups. It is always important to remember that these are vulnerable children for whom facing life outside the home is challenging. Rather than adopting a ‘wait and see’ policy, a proactive approach would be to refer any child who shows signs of assuming a silent position for assessment as soon as possible (Schwartz et al. 2006). Personnel working in the area of childcare are ideally placed to make this referral, and they are also in a position to refer children with a communication disorder. For, while a communication disorder is neither necessary nor sufficient for the development of SM, in some cases it may be one of a number of co-morbid conditions that lays the foundation for the development of the selectively mute behaviour.
Acknowledgements
I would like to acknowledge the many helpful discussions with Dr Peter Jenkings, Professor Vicki Reed, Professor Mark Onslow and especially Dr Linda Hand. Also, the assistance of Christine Monie, librarian.