1 INTRODUCTION

Increased risk for emotional and behavioral challenges is frequently reported in pediatric chronic diseases such as epilepsy, sickle cell disease, and cystic fibrosis, as well as a variety of genetic conditions (e.g., neurofibromatosis type 1 [NF1], Turner syndrome, and fragile X syndrome) (Chong et al., 2016; Hijmans et al., 2009; Lesniak-Karpiak, Mazzocco, & Ross, 2003; Smith, Modi, Quittner, & Wood, 2010). For children with chronic health conditions, a complex array of factors may contribute to emotional problems such as anxiety, depression, and social withdrawal (Schwartz, Tuchman, Hobbie, & Ginsberg, 2011). Investigating common factors that contribute to emotional distress across disorders, as well as contributing factors that are unique to each specific syndrome, may lead to more effective approaches to preventing, identifying, and treating emotional distress. For instance, identification of a key factor (e.g., chronic pain) contributing to emotional distress across two separate conditions may facilitate application of an intervention designed for one patient group to individuals with the other condition. Conversely, where research identifies an important risk factor that is distinct to a condition (e.g., side effects of a specific medication used to treat the disease), prevention and treatment approaches can be designed to address that specific element of vulnerability.

When investigating emotional functioning in children with genetic syndromes, it can be especially challenging to distinguish the relative influences of the neurodevelopmental differences that arise from the genetic pathology from the influences of the medical and psychosocial sequelae of the disorder (e.g., beliefs/expectations about medical risk, disruption to school and family activities). For this reason, studying disorders that have a shared molecular genetic pathology, but that differ in terms of clinical presentation, is instructive. This approach allows investigators to consider the relative influence of genetic, medical, and environmental factors on emotional and behavioral presentation. The current study focused on two genetic disorders that are caused by alterations in a common molecular signaling pathway, but whose most prominent medical features are fairly distinct.

Gene mutations within the Ras-mitogen-activated protein kinase (RAS-MAPK) signaling cascade cause a spectrum of genetic syndromes known as “RASopathies.” NF1 and Noonan syndrome (NS) are the two most common RASopathies. These two syndromes can either be inherited from an affected parent, or they can occur for the first time in a child as the result of a de novo mutation. The features of NF1 and NS can vary substantially from one person to the next, and even among family members with the same genotype. Both syndromes are associated with heightened risk for chronic medical conditions, including skin manifestations, cardiac disease, skeletal problems, and growth delays. NF1 is caused by pathogenic variants in the NF1 gene and is characterized predominately by cutaneous features (e.g., café-au-lait macules, axillary freckling), neurofibromas, Lisch nodules, optic pathway gliomas, and skeletal anomalies. NF1 is estimated to occur in 1:3,000 births (Friedman, 1999; Lammert, Friedman, Kluwe, & Mautner, 2005). Pathogenic variants in one of more than 10 different genes can cause NS (Aoki, Niihori, Inoue, & Matsubara, 2016), which has an estimated incidence of 1:1,000 to 1:2,500 births (Mendez & Opitz, 1985). Prominent features of NS include congenital heart disease, craniofacial characteristics, and short stature (Roberts, Allanson, Tartaglia, & Gelb, 2013).

Among the organ systems most commonly impacted in all RASopathies is the central nervous system. Disruption to molecular, cellular, and neural systems as a result of gene mutations in the RAS-MAPK pathway can lead to a multitude of neurobiological differences, including neuroanatomical abnormalities, aberrant white matter microstructure, impaired synaptic plasticity, and altered neurotransmitter function (Johnson et al., 2018; Loitfelder et al., 2015; Mainberger, Langer, Mall, & Jung, 2016; Shilyansky et al., 2010). While most individuals with NF1 or NS have intact intellectual function, there is increased vulnerability to a variety of well-documented neurodevelopmental disorders, such as attention deficit hyperactivity disorder (ADHD), autism spectrum disorder (ASD), and learning disabilities (Eijk et al., 2018; Lehtonen, Howie, Trump, & Huson, 2013; Pierpont, 2016). These types of learning and behavioral challenges can adversely impact various important life outcomes, including academic performance, employment opportunities, financial and functional independence, interpersonal relationships, and quality of life (Cohen, Levy, Sloan, Dariotis, & Biesecker, 2015; Smpokou, Tworog-Dube, Kucherlapati, & Roberts, 2012).

With respect to emotional functioning, a number of studies have found that internalizing symptoms (e.g., depression, anxiety, and loneliness) occur with greater frequency and severity than externalizing symptoms (e.g., aggression, rule-breaking behaviors) in both children with NF1 (Allen, Willard, Anderson, Hardy, & Bonner, 2016; Coutinho et al., 2017; Graf, Landolt, Mori, & Boltshauser, 2006) and children with NS (Alfieri et al., 2014). In studies involving individuals with NF1, internalizing symptoms have been found to correlate with overall disease severity in some studies (Martin et al., 2012), but not in others (Rietman et al., 2018). The number of stressful life events and the extent of pain interference in the person's life have been associated with internalizing symptoms in NF1, although the direction of this relationship is not always clear (Martin et al., 2012; Wolters et al., 2015). In terms of psychological predictors, there is emerging evidence that symptoms of ADHD are associated with more internalizing symptoms in children with RASopathies. Mautner, Kluwe, Thakker, and Leark (2002) found significantly higher parent-rated symptoms of anxiety and depression among children with comorbid presentations of NF1 and ADHD than among children with NF1 only.

The current study utilized broadband scales to evaluate the emotional and behavioral functioning of children with NF1 in comparison to children with NS. A primary goal of the study was to assess whether emotional symptoms were similar in frequency and severity across these two clinical groups, and whether these symptoms were higher as compared to children living in a similar psychosocial environment who did not have a genetic pathology (i.e., unaffected siblings). Based on previous work, we hypothesized that both groups would have higher rates of internalizing symptoms than their siblings without a neurogenetic disorder. Secondary goals of the study were to assess the consistency of parent and self-report assessments of emotional symptomatology, and to further evaluate the connection between the presence of ADHD symptomatology and the severity of anxiety and depressive symptoms in these populations.

2 MATERIALS AND METHODS

3 RESULTS

3.4 Clinical significance of internalizing symptomatology

Ratings of anxiety and depression symptoms were available on both parent and child-report forms. Parent ratings across all groups in this study were generally indicative of higher symptom severity than child ratings for both anxiety and depressive symptoms (Figure 1). For anxiety symptoms, approximately half of parents in both clinical groups rated their children as having symptoms that were outside of normal limits. Notably, unaffected siblings were also rated by their parents as having elevated frequency and severity of anxious symptomatology relative to the normative population. Several of the parent ratings for unaffected siblings indicated a severe level of anxiety (clinically significant range; defined as >2 SD from the normative mean score), even though no children from this group rated themselves as having severe symptoms. In terms of depressive symptoms, about one-third of children in each group had scores outside of normal limits based on parent report. The proportion of children rated in the clinically significant range was higher in the clinical groups than in the unaffected sibling group. Child self-report ratings of depression across all three groups were minimal overall.

4 DISCUSSION

This study compared ratings of emotional and behavioral functioning of children with two relatively common genetic syndromes and their unaffected siblings. NF1 and NS both arise from disruptions to cellular signaling within a shared biochemical pathway, but they differ considerably in terms of their predominant medical features. Results derived from parent ratings confirm previous reports indicating that children with these conditions are at heightened risk for behavioral health issues. Past research using broadband parent- and teacher-report scales has frequently found that problems tend to be most evident on scales measuring attention problems and social problems (Cipolletta, Spina, & Spoto, 2018; Descheemaeker, Ghesquiere, Symons, Fryns, & Legius, 2005; Loitfelder et al., 2015; Pierpont, Tworog-Dube, & Roberts, 2015). These important behavioral concerns have been the focus of increasing research interest, although the volume of studies to date has been significantly greater for NF1 than for NS.

With regard to ADHD symptomatology, increasing evidence now supports the view that difficulties with inattention and/or hyperactivity/impulsivity arise from the neurodevelopmental effects of the genetic pathology in some affected individuals (Green, Naylor, & Davies, 2017). Cohort studies that include an unaffected sibling group, including the current study, have repeatedly found that children with NF1 or NS demonstrate notably higher rates of ADHD symptoms relative to the general population as well as relative to their siblings who are raised in a similar environment but do not have the genetic syndrome (Dilts et al., 1996; Hyman, Shores, & North, 2005; Pierpont et al., 2015). Results of the current study suggest that ADHD symptoms are generally equivalent in terms of frequency and severity across the clinical groups. We observed similar findings across groups when considering multiple parameters, including mean scores on the BASC-2 Attention Problems and Hyperactivity subscales, the percentage of children who met symptom criteria for ADHD based on parent report (i.e., ADHD-RS), and the number of children in each clinical group who were taking ADHD-related medications. Social challenges were also apparent within the sample (e.g., parent ratings indicated elevations on the Withdrawal scale, which measures children's tendencies to have difficulty making friends, avoid others or prefer to be alone). These social challenges have been more thoroughly explored using focused measures of social development and social competence both in the current cohort (Pierpont et al., 2018) and in other groups of children and adults with RASopathies (e.g., Eijk et al., 2018; Garg et al., 2017; Lewis, Porter, Wiliams, North, & Payne, 2016; Plasschaert et al., 2015; Walsh et al., 2013).

While attention-related and social difficulties may be the most prevalent behavioral challenges in RASopathies, our results suggest that internalizing symptoms such as depression, anxiety, and physical complaints also constitute an area of vulnerability. Greater frequency and severity of internalizing symptoms as compared to externalizing symptoms has been a common theme in studies of RASopathies (Allen et al., 2016; Coutinho et al., 2017; Graf et al., 2006; Martin et al., 2012). Therefore, analyses in the current study focused on developing a better understanding of the nature of these internalizing symptoms, particularly symptoms of depression and anxiety.

To illustrate our findings within the context of previous literature, results from the scales measuring anxiety and depression symptomatology were plotted in Figure 2 relative to observations from other studies that utilized broadband parent-report questionnaires that assessed these symptoms, including the BASC-2 or the Child Behavior Checklist (CBCL). Consistent with our study, nearly all studies have found heightened symptoms of anxiety and depression for these clinical groups relative to the normative sample mean. However, the present investigation contains slightly higher mean scores than most studies. This moderately higher magnitude of internalizing symptoms in our cohort as compared to past studies could be related to a few factors, including sample characteristics or differences in psychometric properties of the measures used (BASC-2 vs. CBCL). The most apparent discrepancy in scores on anxiety and depression measures across studies appears to be between our study and that of Klein-Tasman et al. (2014), which enrolled preschool-aged children with NF1 (ages 3–6). Klein-Tasman and colleagues found no significant parent-rated group mean differences between the NF1 group and a typically developing comparison group for anxiety or depressive symptomatology on the BASC-2. This sample was notably younger than our current cohort, which may account for the differences in the frequency and severity of internalizing problems. Research indicates that the mean age of onset of depressive disorders occurs between 11 and 14 years old; the mean age of onset of anxiety symptomology tends to vary based on the subtype of anxiety, with the earliest mean age of onset (i.e., separation anxiety and specific phobias) occurring in middle childhood and more generalized forms occurring later (Merikangas, Nakamura, & Kessler, 2009). Hence, it might be expected that greater symptoms of depression and anxiety would be apparent in our somewhat older cohort.

Important to our central question of how emotional and behavioral functioning is associated with RAS pathway dysregulation, findings from the current study demonstrated a similar pattern of symptom endorsement for children with NF1 and with NS. Specifically, scores on anxiety and depression scales were outside the normative range for almost half of each clinical group, with varying degrees of impairment noted within each subgroup. Past research on NS has reported a frequency of clinically elevated symptoms of anxiety/depression of somewhere between 16 and 32% (Alfieri et al., 2014; Pierpont et al., 2015). The current results corroborate earlier studies reporting that many children with RASopathies have “subclinical” symptoms based on questionnaire measures that may not be severe enough to warrant a diagnosis (e.g., Perrino et al., 2018), but may nevertheless impact a child's well-being.

An important aim of this study was also to understand how parent and child perspectives of emotional and behavioral symptomatology were reflected in their ratings on widely used behavioral questionnaires. Notably, while ratings by children and their parents in this study were correlated, they also diverged significantly in terms of the severity of the symptoms reported. Parents in the study rated their children as having significantly greater symptomatology than children reported in their self-ratings. This finding is consistent with other studies of children with NF1 (Krab et al., 2009; Rietman et al., 2018), as well as research that has looked more broadly at children with chronic illness and also healthy children with a sibling with chronic illness (Sharpe & Rossiter, 2002).

Parents of children with a chronic disease may view their child(ren) as especially vulnerable, or may be overly sensitive to negative outcomes, thus overestimating symptomatology (Anthony, Gil, & Schanberg, 2003). Conversely, it may be that parents' insights of their children are fairly accurate, but that children in these scenarios have less insight into their own behaviors and emotions than adults do. Children with neurodevelopmental disorders, who often have co-occurring difficulties with social cognition and insight, may be particularly apt to under-report symptoms or to tend to try to respond the way they think people want them to rather than accurately reporting their symptoms. For example, Verhoeven, Wingbermuhle, Egger, Van der Burgt, and Tuinier (2008) remarked that individuals with NS in their study were “in general remarkably friendly, cooperative and very willing to please. These are all aspects of a social[ly] desirable attitude” (p. 194). This tendency to respond in a socially desirable manner may have resulted in children with NS tending to endorse fewer symptoms of distress as compared with ratings by their parents.

Among our NS cohort, parent ratings of anxiety symptomatology indicated clinically significant symptoms in 8% of the sample and moderate elevations (subclinical anxiety) in 26%. This rate is closely matched to the frequency of anxiety reported in a recent study that utilized diagnostic evaluations, which reported rates of 8% and 30% for clinical and subclinical levels of anxiety, respectively (Perrino et al., 2018). In contrast to parent ratings, child self-report measures in our study only revealed rates of 5% (clinical) and 14% (subclinical). Notably, we observed that some children appeared to have poor patience/focus when completing the questionnaires, and may have rushed through them in an attempt to complete them rapidly. Furthermore, some of the questions on the BASC-2 seemed too broad or vague for children to identify with them. Another research group that used an anxiety-specific self-report questionnaire reported that children with NF1 rated themselves as having significantly higher anxiety than children from the community without NF1 (Pasini et al., 2012). Thus, when working with children with RASopathies, it may be important that any self-report measurement tools pose questions that are highly relatable to daily life or more focused on concrete, specific behaviors.

Assuming that parents and teachers are fairly accurate raters of children's difficulties, how do we understand the sources of heightened emotional difficulties in children with RASopathies? While we did not observe a consistent association between major medical complications (e.g., presence of tumors, cardiac disease) and emotional symptomatology in this cohort, our results do suggest the possibility that prematurity is a risk factor for difficulties with emotion regulation, which is consistent with the literature on neurodevelopmental effects of preterm birth (Cassiano, Gaspardo, & Linhares, 2016; Clark, Woodward, Horwood, & Moor, 2008). Previous work has found an association of preterm birth and other birth complications with lower adaptive functioning and increased behavior problems in children with NS and other RASopathies (Pierpont et al., 2010; Pierpont & Wolford, 2016). More broadly, anxiety has been found to have a modest correlation with the severity of disease in children with NF1, but this was not tied to the visibility of physical symptoms (Pasini et al., 2012). It is clear that the severity of physical clinical manifestations of the disease does not fully explain emotional challenges experienced by some children. Secondary experiences of medical involvement, such as pain interference, however, may have a more robust association with emotional symptoms in adolescents with NF1 (Wolters et al., 2015). We did not directly measure the extent of pain, but this is an area that is receiving increasing recognition as a problem interfering with quality of life in individuals with RASopathies (Leoni et al., 2019; Wolkenstein, Zeller, Revuz, Ecosse, & Leplege, 2001; Wolters et al., 2015), and should be examined more closely in future research.

Our study identified a robust association of emotional symptoms with other aspects of neurodevelopmental function in this cohort, specifically with ADHD symptomatology. In line with results from Mautner et al. (2002), we found that ADHD symptoms, especially in children with NS, seem to be associated with symptoms of depression and anxiety. This association suggests that severity of behavioral dysregulation may present a greater risk factor for emotional functioning challenges than the medical syndrome diagnosis itself, and is consistent with literature indicating increased emotional challenges among children with ADHD in the general population (Bowen, Chavira, Bailey, Stein, & Stein, 2008; Schatz & Rostain, 2006). There are several potential explanations as to why this pattern may exist. First, symptoms of anxiety and depression may arise from effects of altered brain connectivity or neurotransmitter function. Research has found widespread connectivity differences across various brain regions in individuals with NF1 that are associated with cognitive, social, and behavioral sequelae (Loitfelder et al., 2015). Alternatively (or additionally), it could be that children have internalized their experiences in school and other environments (e.g., experiences of struggling academically, not being accepted, or getting into trouble due to their behaviors), resulting in greater emotional challenges. Finally, in the current study, ratings of depressive symptoms tended to be higher among children with NF1 who had an affected parent than those without an affected parent. This finding further suggests that heritable neurobiological vulnerabilities may contribute to mood difficulties, or to the perception of these difficulties by their parents.

An unexpected finding of the current study was the fact that differences from the normative sample with respect to anxiety symptoms were found not only in the clinical groups, but also among unaffected siblings. Almost half of the children with NF1 and NS in the current study were rated by their parents as having symptoms of anxiety that were outside of normal limits. A similar pattern of symptomology was also found in the unaffected sibling group. This is one of the first studies that have noted increased anxiety among unaffected siblings of children with RASopathies. Although the source of anxiety among the unaffected siblings in this study is unclear, there is some research to suggest that internalizing behaviors, including anxiety and depression, are common among siblings of children with a chronic illness (e.g., Sharpe & Rossiter, 2002). While some studies have found siblings of children with neurodevelopmental disorders such as Williams syndrome (Leyfer, Woodruff-Borden, & Mervis, 2009) or ASD (Hastings & Petalas, 2014) to be at no greater risk of anxiety or internalizing problems than the general population, other research suggests that siblings of children with disabilities (e.g., ASD, cystic fibrosis, and cerebral palsy) may be at increased risk of depression symptoms (Breslau & Prabucki, 1987; Shivers, Deisenroth, & Taylor, 2013). Differences in methodology, severity of illness/disability associated with different conditions, and sibling age may help to explain these mixed results across studies. There is also some research to suggest that family-related variables, such as maternal history of anxiety, may predict higher anxiety symptoms in female siblings of children with disabilities (e.g., Shivers et al., 2013). Thus, potential explanations for sibling risk of internalizing symptoms include heritable, gender-based, and environmental risk factors.

5 CONCLUSIONS

The current study found that parent ratings of emotional functioning were outside the normative range for almost half of the children with NF1 or NS. Consistent with previous research, children in both clinical groups were rated by their parents as having greater problems as compared to the normative population across most areas of emotional and behavioral functioning, except externalizing problems (e.g., acting in an overly aggressive or antisocial manner). Overall parent ratings in this study were generally indicative of higher symptom severity as compared to self-report results. While child ratings were minimal overall, parent ratings across all groups were significantly correlated with child self-report ratings. This finding suggests that while children may endorse fewer symptoms than indicated by their parents' observations, they may have some insight that these symptoms exist. Intergroup comparisons of emotional functioning allowed for novel findings in the current study, which revealed that higher parent reporting of depressive symptoms appears to be unique to the children with RASopathies, whereas high rates of anxiety were found across both the clinical groups and unaffected siblings. As higher rates of depressive symptoms were endorsed among children with a RASopathy, especially those with an affected parent, these symptoms may be related to neurobiological changes as a result of RAS-MAPK pathway-related mutations.

While parent reporting of internalizing symptoms in their children with RASopathies did not significantly differ based on the presence or absence of the more commonly associated medical complications (e.g., tumors, cardiac problems), there were differences noted based on the presence or absence of ADHD symptoms. Parent ratings of anxiety and depression were generally higher among children who had notable ADHD symptoms than among those without ADHD symptoms. As symptoms of ADHD may be more recognizable, at least at younger ages, than internalizing symptoms, future research may examine whether there is a decrease in internalizing problems in children with RASopathies who participate in early intervention to manage ADHD-related symptoms. Children with RASopathies, even those with subclinical symptoms, may benefit from referral to mental health providers.

CONFLICT OF INTEREST

The authors declare no potential conflict of interest.

AUTHOR CONTRIBUTIONS

R.H., M.S.C., C.M., and E.P. conceptualized the study and developed the study design. R.H., A.F., M.E.P., S.B., K.S., C.M., and E.P. participated in the acquisition of data. The initial draft of the manuscript was written by A.M. and E.P. A.M., R.S., and E.P. completed statistical analysis and drafting of figures. All authors edited and provided final approval of the manuscript.