Mental State Language Use in Children with Down Syndrome and the Role of Caregivers

Abstract

Children with Down syndrome (DS) have both strengths and difficulties in speech, language, and social communication. Mental state language—the ability to discuss others’ perspectives such as their thoughts, feelings, and intentions—represents a foundational social communicative skill that is delayed in many children with DS, even into the school-age years. The purpose of this article is to review the evidence base on mental state language development in school-age children with DS, focusing in particular on assessment and intervention. We discuss assessment procedures that are both age appropriate and developmentally appropriate for this population. We also present preliminary data highlighting the role of caregivers in supporting mental state language development in school-age children with DS through shared storytelling. We propose that interventions aimed at supporting mental state language development in DS should include a focus on caregiver–child shared storybook reading, even in the school-age years. Therefore, we discuss key considerations for clinicians when teaching caregivers strategies for supporting mental state language and social communication in children with DS.

The ability to understand and talk about others’ mental states is critical to successful social interactions. Mental state language involves talking about one’s own and others’ emotions, thoughts, desires, intentions, and other mental states. This skill set involves perspective-taking, or putting oneself in another’s mental shoes, which enables the individual to interpret other people’s social cues and respond appropriately. Thus, it facilitates age-appropriate peer interaction and is important for developing communicative competence.^1,2 On the other hand, difficulties in this area of communicative development can contribute to problems interacting with peers, social isolation, and behavior that reduces learning and independent living opportunities. Individuals with Down syndrome (DS) are often described as highly social and indeed may show relative strengths in early social–emotional understanding.³ However, the phenotype associated with DS includes both relative strengths and difficulties that contribute to communicative competence.^4-6 For example, perspective-taking and expressive language can be areas of relative difficulty for many individuals with DS.³

Due to the array of speech and language delays present in children with DS, clinicians may not often focus on social-cognitive aspects of communication (i.e., emotional competence, theory of mind, social problem-solving, inferencing, etc.) when implementing assessments and planning intervention. However, these skills are critical to developing communicative competence, and success in developing these skills may lead to greater opportunities for community-based inclusion.⁷ In the literature on typical development, there is a wealth of evidence pointing to short- and long-term effects of social-cognitive communication skills on emotion regulation, school readiness, friendship development, peer relations, classroom adjustment, language development, and general academic achievement.^8-14 Thus, clinicians should consider this skill set in children with DS during assessment,¹⁵ particularly in the school-age years. In this population, the school-age years represent a time in which mental state language is still emerging¹⁶ and when both academic achievement and functional skills are important goals. However, traditional social-cognitive communication assessments (e.g., pragmatic language tests) require high-level language and processing skills that often result in floor effects and thus may provide limited information in populations with intellectual disability such as DS. In consideration of the DS phenotype, we propose that mental state language represents a foundational skill that should be assessed at least in part through naturalistic language sampling.¹⁷ The purpose of this article is to inform clinicians of the evidence base on mental state language development with a focus on assessment and intervention in school-age children with DS.

DEVELOPMENT OF MENTAL STATE LANGUAGE

Overview of Typical Development

Typically developing toddlers begin using mental state words to communicate desire, usually to describe their own wants and needs (e.g., want juice),^18,19 soon after they begin combining words. By around age 3, they begin using basic emotional state words (e.g., happy, sad),²⁰ first to describe themselves and soon thereafter to describe others.^21,22 By around 4 years of age, they begin using more complex emotion vocabulary and start describing cognitive states (e.g., think, know).^18,19,22,23 Later in development, children are expected to interpret others’ mental states from narratives (i.e., inferencing) and use mental state language when telling narratives.^24-26

Current Evidence in DS

The evidence base on mental state language development in children with DS is quite limited, but the available literature does show that it is delayed even in early childhood. For example, Beeghly and Cicchetti²⁷ observed that young children with DS ages 2 to 7 years used significantly less mental state language during mother–child play than mental age–matched typically developing children. In general, most of the children with DS used earlier developing mental state words to refer to basic emotions (e.g., happy, sad) or physiological states (e.g., tired, hungry) with minimal referencing of more abstract states such as cognitive states (e.g., think, know). Similarly, Tager-Flusberg²⁸ found that children with DS ages 3 to 7 years used the same amount of mental state language during mother–child play as expressive language-matched peers with autism spectrum disorder, again suggesting that mental state language is delayed in young children with DS.

Mental state language appears to continue to be delayed in school-age children and adolescents with DS.^16,29-31 These delays may be tied, at least in part, to delayed expressive morphosyntax (i.e., mean length of utterance [MLU]).^16,29,30 In the largest study to date, our research group examined mental state language in 37 school-aged children with DS (6–11 years old) during narrative storytelling of a wordless picture book.¹⁶ There was pronounced variability within the sample, with several children not yet using any mental state language or only using a small number of references to mental states (average = 9% of utterances; most commonly used words = happy, sad, see, cry). As a benchmark, recent findings from our laboratory suggest that in younger typically developing preschoolers matched by nonverbal developmental level, on average 17% of utterances spoken during the same narrative storytelling task contained mental state language.³⁰ Interestingly, however, the two participant groups did not differ in the number of different mental state words used. These findings suggest that school-age children with DS have ample mental state vocabulary for their developmental level but do not use these words as frequently or spontaneously (i.e., without prompting). Therefore, interventions prompting the use of mental state language may prove beneficial. Importantly, however, these results reflect mental state vocabulary expectations based on developmental level, meaning that overall mental state vocabulary is probably still delayed in school-age children with DS relative to age-based expectations.

In our sample of school-age children with DS, mental state language was not significantly associated with age or nonverbal IQ, showing that it is not simply developing at a slower rate, nor is it a product of general cognitive delays due to intellectual disability. Instead, more mental state language use was associated with stronger general language skills (expressive vocabulary and MLU) and nonverbal emotion knowledge.¹⁶ Emotion knowledge involves the ability to recognize and label others’ emotions from their facial expressions and the social situation (e.g., a child who dropped his ice cream cone probably feels sad). This study’s findings suggest that while language development in general plays a key role, it is not the only factor influencing the ability to use mental state language in children with DS. Nonverbal emotion knowledge also plays a role. Of importance, emotion knowledge appears to be a relative strength for many individuals with DS³² and thus could be an effective starting point when teaching mental state vocabulary.

ASSESSMENT CONSIDERATIONS FOR MENTAL STATE LANGUAGE IN DS

Early in typical development, most assessments of mental state language use involve play-based interactions, usually during caregiver–child play, which is an age-appropriate context for early childhood. Importantly, however, for children with DS who show delayed mental state language well into the school-age years, there are additional considerations for assessment. The choice of language sampling context should be driven by both the individual’s age and developmental level. For example, caregiver–child play is appropriate for early childhood. For school-age children who use phrase speech, narrative storytelling (e.g., in Channell’s¹⁶ study) is more age appropriate and also provides more opportunities to talk about other people’s mental states.

Additionally, when assessing mental state language in children with DS, it is also important to assess their receptive language skills, both vocabulary knowledge and grammatical understanding. This information will be valuable when planning intervention to tailor the complexity of the language input provided to the child. Furthermore, the inclusion of a mental state language inventory would aid the clinician in determining which mental state words the child has already acquired and which ones to target (e.g., starting with basic emotional state words before moving to more challenging cognitive state words).

Finally, as highlighted by Channell’s¹⁶ study, clinicians should consider emotion knowledge in their assessment battery alongside structural language skills when targeting mental state language and perspective-taking in individuals with DS. Emotion knowledge can be measured nonverbally, which is helpful in minimizing language and other task demands that can limit the ability of children with DS to demonstrate their skills. An example of a nonverbal emotion knowledge assessment involves receptively recognizing another person’s emotional expression from silent video clips, pointing to the schematic face representative of the emotion choice, and, when possible, verbally labeling the emotion in a multiple-choice format (see the article by Channell et al³² for procedural details). This information could be used to better determine the child’s level of social–emotional development as a starting point for the selection of particular emotional state words to target during intervention.

HOW TYPICALLY DEVELOPING CHILDREN LEARN TO TALK ABOUT MENTAL STATES

Caregiver–child talk about mental states provides an avenue for discussing others’ perspectives, their emotions, and causes and consequences of their behavior. Therefore, caregiver mental state language used during conversations and shared storytelling with their children is associated with typically developing children’s emotion knowledge, perspective-taking skills, and mental state language development.^11,33,34 For example, caregiver talk about mental states during picture description tasks with their young children (2–4 years old) is predictive of child mental state language use approximately 6 to 9 months later, above and beyond growth in general language skills.^35,36

One theory about the role of caregiver mental state language is that caregiver talk about others’ mental states facilitates the development of children’s social–emotional understanding (e.g., perspective-taking; emotion knowledge),^2,35,37-40 which in turn leads to children’s use of mental state language.¹¹ The age range in which these relationships have been observed in typically developing children falls roughly within the preschool years (i.e., ~3–5 years).

EXPLORATION OF HOW CHILDREN WITH DS LEARN TO TALK ABOUT MENTAL STATES

Preliminary Data

Our research group recently conducted an exploratory study to apply the conceptual model from typical development to children with DS and their caregivers. See Fig. 1 for a visual representation of how we applied this conceptual model to our sample with DS. We collected data on 33 mother–child dyads of school-age children with DS from the same cohort reported by Channell.¹⁶ The 6- to 11-year age range was selected to target the developmental levels in which children with DS were expected to start using mental state language on their own. See Table 1 for descriptive information about the participant sample. Below, we describe our procedure and report preliminary results in which we explored concurrent relationships within the conceptual model.

Figure 1.

Model tested in Down syndrome: maternal mental state language facilitates child mental state language through the development of child emotion knowledge.

Table 1.

Sample Descriptive Information (N = 33 Mother–Child Dyads)

Descriptive variable	Mean (SD)	Range
Child age	8.66 (1.70)	6.00–11.83
Child nonverbal IQ (Leiter-3)	59.85 (8.94)	41–75
Child EVT-2 standard scoresa	67.20 (12.59)	44–94
Child mean length of C-unit in morphemes	3.58 (1.78)	1.18–8.17
Child mental state language (% C-units)	8.00 (6.16)	0–23.16
Child number of different mental state words	3.91 (3.69)	0–14
Child emotion knowledge (% correct)	77.86 (20.26)	16.67–100.00
Maternal age	45.20 (5.55)	28.97–54.95
Maternal mental state language (% C-units)	22.65 (7.86)	6.25–39.81
	Frequency (percentage)
Child sex	20 (61%) female
	13 (39%) male
Child race/ethnicity (caregiver reported)	24 (73%) White, non-Hispanic
	5 (15%) Black
	1 (3%) Hispanic
	3 (9%) more than one race
Maternal race/ethnicity (self-identified)	27 (82%) White, non-Hispanic
	5 (15%) Black
	1 (3%) Hispanic
Maternal level of education	5 (15%) completed some college or technical college degree
	9 (27%)bachelor’s degree
	19 (58%) completed some graduate work or graduate/professional degree

Abbreviations: C-unit, Communication unit (utterance); EVT-2, Expressive Vocabulary Test, 2nd edition; SD, standard deviation.

^an = 30 due to examiner failure to establish basal/ceiling.

PROCEDURE

Mothers were given wordless picture books and were instructed to “read” them with their children just like they would at home. This shared storybook interaction was recorded, and transcripts of the mother–child narratives were coded for maternal use of mental state language (i.e., proportion of maternal utterances [C-units] containing mental state language; intercoder agreement = 96–100%). Children’s use of mental state language was assessed in a separate examiner–child narrative storytelling task, and their social–emotional understanding was assessed by a nonverbal emotion knowledge task.¹⁶

RESULTS

We tested for concurrent mediation (see Fig. 1) using a statistical bootstrapping method that quantifies effects of the relationships between variables rather than inferring causality.^41-43 Additional methodological details are available from the first author by request. The “ab” path (Fig. 1) was significant, meaning that the combined relationships among maternal mental state language, child emotion knowledge (the intervening/mediating variable), and child mental state language use produced a significant effect. Furthermore, the addition of child emotion knowledge into the model resulted in the direct relationship between maternal and child mental state language becoming nonsignificant (i.e., full mediation). In other words, maternal mental state language facilitates child mental state language indirectly, by increasing children’s emotion knowledge. These results support the application of the conceptual model from typical development to DS. However, these data are preliminary in nature, and additional research is needed. Notably, we do not yet have longitudinal evidence and thus cannot infer causal relationships. Regardless, these results highlight the key role of maternal language input in the emergence of mental state language use by children with DS, even as late as the school-age years. Mothers of school-age children with DS have reported that shared storybook reading with their children is a frequent activity in which they already engage at home.^44,45 Thus, it is a ripe context for increasing conversations about mental states.

INTERVENTION IMPLICATIONS

The literature on toddlers and young children with DS has shown that increasing maternal responsiveness strategies in the language input that mothers provide improves children’s language outcomes.^46,47 However, this effect has not yet been tested in school-age children with DS or specifically with mental state language. Importantly, our research finding that mental state language use in older children with DS is associated with their mothers’ language input suggests that children with DS could benefit from caregiver-implemented language interventions even during the school-age years, a time when mental state language is starting to develop in this population. When teaching caregivers of children with DS strategies to support their children’s social communication, there are several key considerations which are discussed below.

Goal of Shared Enjoyment

First, it is important to emphasize to caregivers that the goal is making shared storybook time an activity of shared enjoyment rather than one focused on teaching specific literacy skills. This works because many of the strategies designed to promote child engagement during shared storybook “reading” also tend to promote children’s skill development. For example, rather than simply reading the text, the caregiver could make the story more conversational in nature (e.g., through dialogical reading strategies) which will increase the likelihood of child engagement.^48-50 This, in turn, will give caregivers more opportunities to be responsive to their child’s language and communication attempts. To illustrate the effectiveness of this procedure with children with DS, a recent pilot study found that providing embedded prompts for mothers of preschoolers with DS during shared storybook reading resulted in higher levels of participation by the children, including more overall talk and more diverse vocabulary use.⁵¹ These results serve as preliminary evidence supporting the use of caregiver–child shared storybook interaction as an avenue for increasing child engagement and language skills specifically in DS.

Talk About Mental States

In addition to increasing child participation, shared storybook interactions offer the opportunity to expose children to diverse vocabulary words and practice social thinking. For example, caregivers could label character emotions and connect them to character actions (e.g., “The little boy feels sad because he dropped his ice cream cone.” or “The boy was scared, so he ran away.”). Doing this will introduce both mental state and causal language. In addition to exposing children to mental state vocabulary, caregivers could prompt their children to use mental state language (e.g., “How would you feel if that happened to you?” or “What do you think he will do?”). Such prompting should eventually help promote children’s spontaneous talk about mental states.

Although we do not yet have evidence for caregiver-implemented shared storybook interventions increasing mental state language in children with DS (see O’Toole et al⁵²), results from interventions with other populations, such as fragile X syndrome, are particularly promising. For example, a caregiver-implemented shared storybook intervention for school-age children with fragile X syndrome has been successful in increasing children’s mental state and inferential language use. This intervention focused on responsive strategies of expanding, recasting, and asking Wh-questions.⁵³ Another clinician-implemented shared storybook intervention (i.e., RECALL)⁵⁴ used prompting strategies with specific types of Wh-questions and found that children with autism spectrum disorder increased spontaneous on-topic responding to both fact-based and inferential Wh-questions, illustrating the importance of including both types of Wh-questions. Because there are known differences in narrative language and inferential language use between DS and other neurodevelopmental disorders (e.g., fragile X syndrome)^29,55,56 and in the caregiver–child relationship,^57-60 more evidence is required to determine how best to apply these interventions to DS. Based on our preliminary data on caregiver–child interactions reported earlier, discussing characters’ emotions may be an important starting point for these families. In doing so, mothers can build upon the phenotypic strength of emotion knowledge in DS while modeling and prompting language used to talk about emotions and other mental states.

Matching the Child’s Developmental Level

Another consideration for caregiver-implemented interventions with school-age children who have DS is the variability in which language skills and other behaviors present across the DS phenotype.^4,6,61 Thus, it is important for caregivers to learn prompting and questioning strategies that match the developmental level of the child. This includes receptive language comprehension level, both vocabulary knowledge and grammatical understanding, as well as nonverbal emotion knowledge.

For caregivers of children with DS who primarily use augmentative and alternative communication (AAC), a caregiver-implemented emotion communication intervention developed by Na and colleagues^62,63 may be a good starting point. This intervention encourages caregivers to communicate about character emotions during a shared storybook interaction and provides access to emotion-based vocabulary through AAC. The initial evidence provided by Na and colleagues⁶³ using single-subject design supports this intervention’s efficacy in increasing conversations about emotions in caregivers of children with DS, thus supporting the application of this type of caregiver–child intervention to DS.

In addition to considering the child’s language comprehension level, it is also important to consider the child’s level of attention and task persistence (e.g., when it becomes challenging) in teaching caregivers these strategies. To illustrate this point, McDuffie and colleagues employed a storybook intervention implemented by caregivers of children with fragile X syndrome.^64,65 This program started by decreasing challenging behaviors in their children so that they could attend to the storybooks. Although children who have DS do not show as many challenging behaviors as those with fragile X syndrome,^66,67 many caregivers report that their children with DS show difficulty with task persistence.⁶⁸ Tailoring the storybook interaction to the current developmental level of the child should also help maintain the child’s attention, keeping the child engaged.⁵¹

Cultural and Familial Practices

Finally, sociocultural considerations are always an important element within the scope of practice for speech-language pathologists.⁶⁹ There are several cultural dimensions relevant to caregiver-implemented shared storybook interventions for children with DS, including beliefs about disability, parenting practices, and educational practices.⁷⁰ There are also cross-cultural and familial differences in the frequency and nature of caregiver–child shared storybook activities,⁷¹ in emotion communication,^72-75 and in how to facilitate social competence.⁷⁶ Furthermore, having a child with a disability also impacts family functioning, and thus, cultural and familial dynamics are important considerations in service delivery.^77,78

Almost 20 years ago, Kaderavek and Justice⁷¹ argued that when implementing caregiver-child shared storybook reading interventions, speech-language pathologists should consider each family’s reading practices and cultural values surrounding storytelling. For example, some families may find it easier to engage in storytelling conversations through family reminiscing rather than shared book reading. Wordless picture books may also be beneficial because they provide more flexibility for caregivers to introduce the story in their preferred style.⁷⁹ Over the two decades since, the research surrounding cross-cultural differences in caregiver–child storytelling has further highlighted the need for supporting family-centered and culturally sensitive shared storytelling practices.^80-83 Clinicians should encourage caregivers to implement strategies within their preferred shared storytelling style that promote the child’s engagement in the interaction and ultimately promote social-cognitive and communication skills.⁸⁴ As an example, Na et al⁶² developed an assessment tool to capture familial preferences regarding emotion communication within and across cultures, and they applied their tool to their AAC-based caregiver–child emotion communication intervention. Their interview-style checklist is available in multiple languages. Family expectations of emotion expression are captured by questions about how caregivers feel about expressing their own emotions in front of their child. Questions specific to caregiver input preferences involve probing about how they talk about emotions in general and whether this talk occurs during book reading or other contexts.⁶² This type of assessment tool could aid clinicians in determining sociocultural preferences when designing caregiver–child intervention strategies for all families, including those who have a child with DS.

CONCLUSIONS

Social-cognitive aspects of communication are important for clinicians to consider when assessing functional skills in school-age children with DS. Mental state language is often delayed in children with DS and can be assessed through narrative storytelling. Nonverbal emotion knowledge and structural language are associated with mental state language development in children with DS and also should be assessed and targeted as needed. Children develop emotion knowledge and learn to use mental state language within the cultural-familial context through caregiver–child interactions, especially shared storytelling. Thus, when providing intervention services, clinicians should involve caregivers by teaching them strategies for supporting social communication during activities such as shared storybook “reading,” even for school-age children with DS. When teaching caregivers these strategies, there are several key points for clinicians to consider, including the following:

1. Make the goal shared enjoyment rather than reading or teaching specific skills. Doing so will increase child engagement and opportunities for learning.

2. Label character emotions and connect them to character actions. This will lead to caregiver–child conversations about emotions and their causes and consequences and will promote child emotion knowledge and mental state language use. Responsive strategies of expanding, recasting, and asking different types of Wh-questions (fact-based and inferential) may also be helpful.

3. Match the specific strategies (e.g., type and frequency of input) to the developmental level of the child. Child assessment should include receptive language (mental state and general vocabulary; grammatical understanding) as well as emotion knowledge and should consider level of attention and task persistence.

4. Encourage strategies within the caregiver’s shared storytelling style. Assessment should include gathering information about the family’s cultural preferences regarding parenting and education as well as shared book reading or other avenues for storytelling and emotion communication.

By taking these points into consideration, caregiver–child shared storybook “reading” provides key opportunities for facilitating mental state language and social communication development in children with DS, even in the school-age years.

Learning Outcomes:

As a result of this activity, the reader will be able to (1) discuss the importance of assessing social communication skills such as mental state language in children with Down syndrome; (2) explain how to assess mental state language in children with Down syndrome; (3) explain the role of caregivers in facilitating children’s expression of emotion knowledge and their mental state language development; (4) summarize four key considerations for clinicians when teaching caregivers strategies for increasing mental state language and related skills in children with Down syndrome.