Neurodevelopmental Origins of Social Competence in Very Preterm Children

Abstract

Social problems are an important feature of the preterm behavioral phenotype but are diverse and multidimensional. A model of social competence comprised of the three interrelated components of social cognition, social interaction, and social adjustment is useful in conceptualizing these problems. Weaknesses in social cognition in very preterm (VPT) children, although rarely studied, are found on tasks involving interpretation of social cues. Difficulties in social interaction in VPT infants and young children are documented by structured observations of their interactions with adults. Problems in social adjustment are endorsed on parent and teacher ratings of VPT infants and youth and on self-reports of VPT adults. These multiple deficits likely stem from early disruptions in neural development, are related to other consequences of preterm birth, and subject to postnatal environment influences. Further research is needed to more fully account for the effects of VPT birth on the development of social competence.

1. Introduction

Common behavior problems following very preterm (gestational age [GA] <32 weeks, VPT) birth include anxiety and other internalizing symptoms, attention deficits, and problems with peer relationships and prosocial behaviors, referred to as the “preterm behavioral phenotype” [1–3]. Diagnoses of these and other behavior problems range from 25% to 28% in very low birth weight (<1500 g, VLBW) children and are 3–4 times higher in extremely preterm (EPT, GA <28 weeks) and/or extremely low birth weight (<1000 g, ELBW) children compared to full-term (FT) controls [4]. Social problems are a pervasive component of the preterm phenotype but are diverse and multi-dimensional [2,3,5,6]. Whereas ratings of these problems on scales such as the Child Behavior Checklist [7] and the Strengths and Difficulties Questionnaire (SDQ) [8] provide information on social problems and social adjustment, they fail to assess other aspects of social competence, such as weaknesses in social cognition, or to fully characterize children’s social relationships.

2. A framework for conceptualizing social competence

A useful framework for conceptualizing social competence in VPT children more broadly is one comprised of the three hierarchically related components of social cognition, social interaction, and social adjustment [9–12]. As illustrated in Figure 1, social cognition, also referred to as social information processing, includes the mental processes involved in social interactions. In seminal formulations [13,14], an interaction begins when the child encounters a social cue, such as another child’s verbalization, tone of voice, or facial expression. The child draws from a store of social knowledge and past emotional experience in interpreting the meaning of the cue and in considering different options for responding to it. Following interpretation of the cue and consideration of social goals and the consequences of alternative response options, the child decides how to respond and enacts the decision in a behavioral response.

Figure 1.

Three component model of social competence (adapted from Yeates et al. [12])

Social cognition consists of several distinct cognitive operations, including affect perception, emotion processing, mentalizing, and executive functioning [2,12,15]. Brain regions involved in affect perception include multiple regions within the temporal cortex, inferior occipital cortex and fusiform gyrus, and intraparietal sulcus. Regions implicated in further affective processing, including emotion regulation, decision making, and sensitivity to reward and punishment, include the amygdala, ventral striatum, and orbitofrontal cortex. The ability to appreciate others’ perspectives and intentions, referred to as “theory of mind,” involves a mentalizing network comprising medial prefrontal, superior temporal and temporal-parietal regions and the anterior cingulate gyrus. Cognitive regulation of social cognition, or executive functioning, is also required and involves the dorsal medial, ventral prefrontal, and orbitofrontal cortices. The cerebellum is additionally implicated in social information processing, with underlying brain systems viewed more as interconnected frontal-striatal-cerebellar, fronto-limbic, and fronto-parietal networks than as collections of isolated brain regions [3].

Social interaction in the model refers to the manner in which the child interacts with social partners as described in terms of prosocial vs. antisocial approaches by the child, whether the child seeks out and engages in interactions, or is more likely to withdraw from or avoid them, and the age appropriateness of the child’s interactions. Examples include initiating and sustaining interactions, staying on topic, taking turns, cooperating, joint problem solving, and demonstrating awareness of others’ feelings and intentions. Interactions take place in a situational context that makes specific demands on the child and can only be ascertained in terms of the “back and forth” with others [10]. The interactions are assessed in terms of specific behaviors, such as the degree of social initiation or responsiveness to a social partner, and not in terms of impressions of children’s day-to-day social functioning.

Social adjustment pertains to the ways in which social interactions impact on social functioning and is the most directly observable component of social competence. Manifestations of social adjustment include peer acceptance, rejection, and victimization and the extent and quality of friendships. The child’s own responses to social interactions is a further indicator of social adjustment, as reflected in satisfaction or dissatisfaction with social relationships, by personal agency or defeat and by comfort or distress in relating to others. Social adjustment thus entails judgements of the child’s social functioning by the child or others [10].

Support for the three-component model of social competence is provided by studies demonstrating its utility in distinguishing different aspects of social problems and by evidence for associations between model components [12]. A review of the literature on social outcomes in VPT children reveals evidence for deficits in all three components [11]. Links between model components have been confirmed by evidence that deficits in aspects of social cognition are associated with difficulties in social adjustment [16].

3. Developmental manifestations of problems in social competence

2.1. Infancy (birth–2 years)

Problems in social competence following VPT birth are evident within the first year of life. A study comparing infants with <1600 g birth weight to FT controls found that the preterm group made slower progress from 6–36 months in initiating social interactions with their mothers [17]. Another study demonstrated that young infants with VPT birth were less likely than FT infants to restore a broken social interaction with another adult and follow the gaze of that adult to a non-attended object [18]. Across the first 2 years of life, VPT/VLBW infants also display higher rates of insecure attachment with their parents than FT infants and have more difficulty transitioning between activities and regulating and persisting in interactions with primary caregivers [19,20].

By 2 years of age social deficits in VPT infants are evident on parent ratings. A study comparing VPT infants at 2 years’ corrected age to FT controls found that the VPT group had more problems in behavioral and emotional self-regulation, less empathy, and lower levels of task motivation, social interactions, and prosocial peer relationships [21]. Deficits in social-communication behaviors in preterm infants are also manifest in high rates of positive screens for autistic traits [24]. A study of 2-year-old EPT infants documented rates of positive screens for autism spectrum disorder (ASD) on the Modified Checklist for Autism in Toddlers (M-CHAT) of 41% [22]. Another investigation found that approximately a quarter of a sample of VLBW infants screened positive for ASD on the M-CHAT, with similar proportions scoring more that 2 SDs below the mean on the Socialization and Communication scales of the Vineland Adaptive Behavior Scales [23]. Associations of positive screens with sensory-motor and physical impairments suggest that these elevated rates may partially reflect the more generalized neurodevelopmental problems to which VLBW infants are susceptible [22].

2.2. Early childhood (3–5 years)

A study following VPT infants from 2 to 4 years of age found that those with more extreme prematurity had problems in emotional self-regulation and task persistence that were similar to problems identified at 2 years [20]. Another longitudinal investigation revealed higher rates of emotional and peer problems and lower rates of prosocial behavior on the SDQ in 3-year-old VPT infants compared to controls [24], with similar outcomes at 5 years [25]. Additional studies of outcomes at 3 to 5 years of age report that VPT/VLBW cohorts have more peer and emotional problems, less positive play with peers and less synchronous interactions with their mothers, more difficulty naming facial expressions of emotions, and higher parent ratings of social problems on the CBCL [5,6,26–28].

2.3. Middle childhood (6–10 years)

Studies dating back to the 1980s indicate parent and teacher ratings of more limited social skills, poorer self-regulation, and more social withdrawl in middle school-age VPT/VLBW children compared to FT controls [4,16]. In a study using the SDQ, parents and/or teachers reported more emotional and peer relationship problems in 6-year-old VPT children compared to FT controls [29]. In another study of young school-age children, an EPT/ELBW group received lower teacher ratings of peer relations, self-management, and academic engagement than FT children [30]. Studies of VPT/EPT samples additionally reveal more symptoms of ASD and higher rates of ASD diagnoses compared to FT groups [31–33].

Problems in social competence in VPT children of middle school age are also evident on assessments of social cognition and social interaction. Jaekel and colleagues [34] coded video-taped interactions of VPT/VLBW children and FT controls with their mothers while engaged in a collaborative problem-solving game. At ages 6 and 8 years the VPT/VLBW group showed less task persistence and were less interactive than controls. In one of the few studies of social cognition, a group of 8- to 11-year-old VLBW children had more difficulty than controls in identifying the feelings of characters in video-taped social scenarios and in explaining how they recognized these feelings [33]. The VLBW group also gave less appropriate descriptions than controls of animated movements of geometric forms and were less sensitive to the intentions represented by the animations [32].

2.3. Adolescence (11–18 years)

Adolescence is a unique period of development characterized by increases in cognitive and emotional self-regulation, greater dependence on peers for socialization, and heightened sensitivity to reward and social-affective stimuli [35]. The mentalizing neural network continues to develop during this period and supports an enhanced awareness of mental states and intentions [36]. A well-recognized feature of adolescence is the disparity between the more rapidly developing social-affective-reward systems relative to top-down cognitive regulation of behavior, a mismatch associated with increased risk-taking as hypothesized by the dual systems model [37]. Despite growth promoting aspects of adolescents’ greater sensitivity to reward and affective stimuli, these changes increase vulnerability to stress and the possibility of making decisions with negative adaptive consequences [36,38].

VPT/EPT cohorts continue to have more parent-reported problems in social functioning during this developmental period [38,39], even when controlling for IQ or excluding youth with intellectual disabilities and neurosensory disorders [40,41]. Self-reports of social functioning in EPT adolescents indicate fewer friendships, less time spent with friends, more dissatisfaction with their peer network, and more bullying compared to FT peers [11,42]. EPT adolescents also have more ASD symptoms and higher rates of ASD diagnoses than FT controls [43].

2.4. Adulthood (18+ years)

Although parent interviews have documented weaknesses in socialization in VLBW 20-year-olds [39], social problems in VPT/VLBW adults are based in large part on self-reports. A individual participant meta-analysis found more symptoms of avoidant personality problems in preterm adults with VLBW/ELBW than in FT adults [44], results consistent with self-reports of more shyness, lower sociability, lesser risk-taking, and more symptoms of ASD in VPT/VLBW adults compared to controls [45].

A meta-analysis of social relationships in preterm and low birth weight adults revealed that they were less likely than FT controls to have experienced romantic partnerships, sexual intercourse, or parenthood and that the likelihood of these experiences was proportional to the degree of prematurity [46]. The findings parallel results from investigations demonstrating fewer friendships in VLBW young adults and a lower likelihood of marriage or co-habilitation, sexual intercourse, or parenthood in ELBW adults 26–36 years of age compared to controls [47,48].

3. Correlates of social problems in VPT infants and young people

Across age levels, social problems in VPT infants and young people are more pronounced for those with more extreme prematurity, perinatal complications, early neurodevelopmental impairments, less advantageous sociodemographic status or family circumstances, and lower scores on tests of early development and cognitive ability [1,3,6,11,17,22,23,27]. Male gender is associated with more social problems in some studies [22,23,27,34] though not uniformly [21]. Parenting that is more negative, overly controlling, or intrusive is associated with poorer social outcomes, whereas more sensitive parenting that maintains the infant’s or child’s interests is related to better outcomes [20,27,34]. Parent mental health problems are other potential predictors of social competence in VPT cohorts [27,49]. More problems in social competence are found even when controlling for social risk [6,20,21,24–27,30,34] or scores on developmental or cognitive assessments [25,26,34].

Neuroimaging abnormalities are also associated with social problems in VPT cohorts. Ratings of white matter abnormalities on MRIs conducted at term equivalent age predict lower levels of self-regulation and social functioning in infants and young children [11,20,21,27]. Other studies suggest associations of social problems with abnormalities in specific brain regions and neural circuits. Rogers et al. [50] documented associations of white matter abnormalities in the right orbitofrontal cortex on neonatal MRIs with peer problems in 6-year-old VPT children. They also found associations of smaller hippocampal volumes with peer problems in females, and of reduced bifrontal diameter with more problems in prosocial behavior in males. Another study of 6-year-olds found associations of lower connectivity of the superior frontal gyrus and lateral orbitofrontal cortex with poorer social reasoning in EPT children, and of lower connectivity in fronto-subcortical and parieto-subcortical circuits with poorer social reasoning and prosocial behavior in a group with intrauterine growth restriction [51]. Studies of VPT adolescents report associations of poorer social adjustment with smaller volumes of the left caudate in males [52] and of social problems with increased volumes of the fusiform gyrus, a structure that typically decreases in size during the transition from childhood to adolescence [41]. In an MRI study of a sample of 20-year-old VPT young adults and controls, Johns et al. [39] reported associations of increased negative connectivity between the amygdala and posterior cingulate with lower parent reports of social participation and social functioning at 16 and 18 years of age.

Other features of the preterm behavioral phenotype, including elevated internalizing symptoms and higher ratings of attention deficits, often accompany social problems in VPT/EPT cohorts [1,6,23,26,27], whereas symptoms of externalizing and conduct problems are absent or less pronounced [2,5,25,29]. Most studies fail to report the magnitude of associations between components of the preterm behavioral phenotype. However, one investigation revealed significant correlations of internalizing symptoms in a VPT group at 2 years with emotional symptoms and peer problems at 5 years [26] and another found that symptoms of ASD correlated with internalizing problems [23].

4. Developmental implications

Problems in peer relationships and social adjustment in the general child population have adverse long-term implications for physical and mental health, schooling, employment, and independent functioning [11,16,46]. Research confirms similar consequences of social problems in VPT samples. A longitudinal study of VPT infants found that poorer self-regulation at 2 and 4 years of age was related to more mental health and educational problems at age 9 [53]. Additional investigations of VPT/EPT children document associations of inhibitory control at 20 months with deficits in academic achievement at 8 years, social-emotional problems at age 5 with a psychiatric diagnosis at age 7 years, and peer relationship problems at age 6 years with inattention at age 11 [31,54,55].

Social problems in VPT adolescents also predict longer-term adult outcomes. In a sample of VPT and term-born adolescents, youth with more social problems at age 15 years had higher psychiatric morbidity at age 19 [41]. Another prospective study found that peer victimization in ELBW adolescence predicted higher rates of depression, anxiety, avoidant personality, antisocial personality, and attention deficit hyperactivity disorder at ages 22–26 years, and higher rates of panic and obsessive-compulsive disorders at ages 29–36 years [56].

5. Potential mechanisms of effect

Social problems following VPT birth are most likely to originate in a large part from early disruptions in neural development [3,41]. VPT infants are prone to brain insults and developmental derangements involving neural systems underlying social competence. Studies of VPT young people have documented decreased volumes in the orbitofrontal cortex, fusiform gyrus, amygdala, and hippocampus, as well as for diminished integrity in white matter tracts linking these regions [3]. An MRI study of 6-year-olds found weaker structural connections of fronto-subcortical and orbital-medial networks in EPT children than in controls, as well as weaker cortical-cortical connections in the EPT group in regions surrounding the temporal cortex [51]. Distinct patterns of connectivity between the amygdala and other regions involved in social and emotional functioning have also been identified in VPT young adults [39].

Other potential sources of deficits in social competence in VPT cohorts include the impact of early stress related to intensive neonatal medical care on the endocrine system and the adverse effects of preterm birth on other aspects of neurodevelopment and parenting [3,44,45,48]. Early exposures of preterm infants to medical management may affect brain development, dopamine function, susceptibility to environmental stressors, and even personality traits [3,41,45]. The negative effects of preterm birth on aspects of behavior and cognition not considered strictly social in nature, such as motor skills, language, attention, and anxiety, may also contribute to problems in social functioning [1–3,27]. Social difficulties in VPT infants and children are additionally associated with non-optimal parenting styles and parent mental health problems [27,49], and VPT youth experience fewer and less satisfactory friendships and more bullying than their peers [42,56]. Research on adolescent development indicates that social isolation increases risks for adjustment problems and may alter the structure of the frontal cortex, suggesting that social “threat” may further exacerbate deficits in social functioning among preterm youth [3,36].

6. Conclusions

Research on social outcomes of VPT birth documents problems in all three components of social competence, including: (a) poor performance on tests requiring affect recognition and interpretation of real and simulated social interactions; (b) difficulties in social interaction observed in direct observations of child-adult play or problem-solving sessions; and (c) deficits in social adjustment based on parent and teacher ratings of emotional and peer relationship problems. Socialization difficulties in the VPT population are related to other adverse outcomes of preterm birth but are evident when controlling for cognitive or developmental status and are more closely related to other features of the preterm behavioral phenotype (i.e. internalizing symptoms and inattention) than to externalizing disorders.

Despite the potential for non-optimal postnatal experiences to contribute to deficits in social competence in VPT youth, these deficits most likely have neurodevelopmental origins. Brain abnormalities on neonatal MRIs are related to difficulties in self-regulation, social-emotional adjustment, and peer interactions in infancy and early childhood. VPT infants and young children have weaknesses in socialization and self-regulation that persist over time and predict subsequent problems in attention, academic achievement, and mental health. Social consequences are evident even into adulthood in the form of lower sociability, more limited social relationships, and less risk taking.

Although findings reveal continuity across age in self-regulatory deficits and in parent ratings of problems in social interaction and social adjustment, research on social cognition is sparse. Only single studies of middle school-age VLBW samples have examined affect processing or social reasoning, and little if any research has been conducted on reward sensitivity or on peer interactions as assessed by direct observations or peer perceptions. More research is also needed on the relation of the three components of the social competence model depicted in Figure 1 to better understand the nature and sources of social problems [33]. Research on social interactions has been largely limited to infants and young children and investigations of adults have relied primarily on self-reports. Further research is needed to identify early biomarkers of social risk and examine the manner in which social problems evolve over time and how the development of social competence relates to other behavior and cognitive problems, children’s environmental exposures and experiences, and the emergence of mental health disorders [3,11,27,38,39,57]. Studies of peer perceptions of VPT youth and adults and other more ecologically valid assessments of social interactions would provide insight into their social relationships not available in self-or parent and teacher ratings. Further research is also needed to test the efficacy of interventions to improve social outcomes in this and other at-risk populations [11,58,59].

Practice points.

• Infants, youth, and adults with VPT birth are at risk for social and emotional problems, including difficulties in social interactions, peer relationships, and behavioral self-regulation.

• Full appreciation of the extent and nature of these problems entails awareness of all three components of social competence, including social cognition, social interaction, and social adjustment.

• Weaknesses in social competence are related to other consequences of VPT birth and are subject to postnatal influences but most likely emerge due to early disruptions in neural development.