The developmental impact of sex chromosome trisomies on emerging executive functions in young children: Evidence from neurocognitive tests and daily life skills

Abstract Sex chromosomal trisomies (SCT) are associated with impairments in executive functions in school‐aged children, adolescents, and adults. However, knowledge on preschool development of executive functions is limited but greatly needed to guide early intervention. The current study examined emerging executive functions in young children with SCT. Participants were 72 SCT children and 70 population‐based controls, aged 3–7 years, who completed a neurocognitive assessment of both global executive function (MEFS) and verbal executive function skills (NEPSY Word Generation). Caregivers completed the Behavior Rating Inventory of Executive Function (BRIEF) questionnaire to capture real‐world behavioral manifestations of impairments in executive functions. Results showed that impairments were significantly more prevalent in SCT than in controls and already present from 3 years, specifically verbal executive functions and working memory. Broader more pronounced impairments were found in older children with SCT. Age was significantly related to executive functions, but specific domains showed different relations with age. For example, deficits in planning and organizing remained evident with older age in SCT whereas it declined with age in controls. Impairments in executive functions were present across different levels of intelligence. Already at an early age, impairments across executive functions should be considered part of the neurodevelopmental profile of SCT, which appear more prominent at later age. Future studies should investigate developmental pathways of executive functions in SCT, given its relevance in cognitive, social, and emotional development. Executive functions should be screened and monitored in children with SCT and could be an important target of preventive intervention.


| INTRODUCTION
With a high prevalence of 1-2 in 1000 births, sex chromosomal trisomies (SCT) are one of the most common chromosomal aneuploidies. 1,2 Karyotypes that result from SCT are 47, XXY (Klinefelter syndrome) and 47, XYY (XYY syndrome) in males and 47, XXX (Trisomy X syndrome) in females. Recent technological advances allow for safe and earlier screening for genetic syndromes and are expected to lead to an increase of the number of prenatally diagnosed children with SCT. 3 This calls for more knowledge on the developmental impact of SCT which is needed to improve genetic counseling and clinical care for children with these conditions. Also, studying genetic conditions such as SCT from pregnancy on provides a unique opportunity to prospectively examine early neurocognitive development and its link to later developmental outcome.
Having an extra X or Y chromosome not only impacts physical development but also neurodevelopmental and psychological functioning. 4 This is not surprising given the high density of genes on the sex chromosomes that are essential for brain development, 5 putting children with SCT at increased risk for neurodevelopmental problems (i.e., impairments of growth and development of the brain, that may lead to differences in brain functioning and thus emotion and cognition amongst other domain). So far, neuroimaging studies have shown that brain architecture and functioning appears different in individuals with SCT compared to peers from the general population (XXY 6 ; XXX/XXY/XYY 7 ). Furthermore, underlying information processing difficulties are also found in individuals with SCT with a quarter of the group showing difficulties of clinical relevance (for a review see Reference [8]). Amongst other domains, impairments are found across executive functions while intellectual functioning is usually within the typical range (although at the lower end, particularly for verbal IQ) (XXY and XXX 9 ; XXY 10 ). Of relevance to the current study are studies showing neuroanatomical and functional differences in the (pre)frontal cortex in individuals with an extra X chromosome, 11,12 an area strongly involved in executive functions. 13 The term executive functions (EF) refers to a set of interrelated cognitive skills essential to learn, cope, and manage daily life. 14 Executive functions are responsible for purposeful, goal-directed, and problem-solving tasks and behavior. Several components can be identified, including attention, inhibition, monitoring, flexibility, working memory, planning, and fluency. 15 Proper executive functions are crucial when it comes to positive childhood development: executive functions promote mental and physical health; predict success in school and in life; and support cognitive, social, and psychological development. 14 On the other hand, impairments across executive functions are involved in many neurodevelopmental disorders, including attention deficit hyperactivity disorder (ADHD 16 ), autism spectrum disorder (ASD 17 ), and intellectual disabilities. 18 Until now, studies that have examined executive functions in individuals with an extra X or Y chromosome showed, on average, reduced executive function performance compared to controls from the general population (for review see 19 ). Children with SCT show more impairments across executive functions, including attention, inhibition, mental flexibility, working memory, and planning/problem solving. 10,18,[20][21][22][23] In daily life, parents of children with an extra X chromosome (XXX and XXY) report difficulties in (sub)domains of behavioral regulation and meta-cognition. 10,18 A substantial part of the SCT group shows significant executive function difficulties, that are present across studies and assessments. For example, 19% to 57% of children with SCT (both Dutch and American) show a clinical score on sustained attention tasks. 22 Furthermore, impairments in executive functions in children with SCT have been linked to increased externalizing behavior problems, 23 increased social difficulties 24 as well as increased symptoms of ASD, 25 psychotic symptoms such as disorganized thought, 26 and ADHD symptoms. 9 It is thus not surprising that parents frequently mention that their child's executive dysfunction, amongst others, is a major barrier to learning and academic development. 27 Previous studies that examined executive functions in the SCT population included school-aged children, adolescents, and adults.
There have been very limited systematic studies on executive functions in early childhood, specifically before 6 years of age and prior to starting the early school years. However, the preschool period (the period between 3 and 6 years of age) is of particular interest when it comes to executive functions, given its development accelerates in the preschool years. 28 This acceleration is partly due to increased connectivity between neural networks in the brain within this period, 29 as well as changes at the contextual level (such as social experience 30 ) and other cognitive abilities (increasing memory capacity, increasing language abilities and accelerated information processing 31 ). Studying this important window in child development in individuals with SCT may help to understand the impact of an extra X or Y chromosome on the developing brain. Differences with typically developing peers are to be expected, given that a high density of genes on the sex chromosomes are essential for brain development, 5 putting children with SCT at increased risk for neurodevelopmental problems including impairments across executive functions. Also, early identification of these difficulties may reveal risk markers in the development of children with an extra X or Y chromosome, that could prove helpful in identifying targets for early intervention to improve outcomes later in life.
Assessment of executive functions usually relies on a combination of a direct assessment of information processing skills as well as structured behavioral observations in daily life. There is growing evidence that executive functions represent diverse but also united constructs in early childhood. 32,33 This has also led to new techniques to measure executive functions in young children, such as the Minnesota Executive Function Scale (MEFS App™) that provides a standardized performance-based assessment of global executive function skills, designed for children ages 2 and up. 34 It integrates three basic executive functions (working memory, inhibitory control, and cognitive flexibility) into a single graded scale. Because the assessment is sensitive to age and performance, following an adaptive testing protocol, it provides the opportunity to assess and follow to the development of emerging executive functions; with emerging meaning still-developing, not yet stable. 35 In addition to the neurocognitive assessment of executive functions, structured observations of behavioral problems in daily life are also crucial. Parents are vital observants in providing information on the behavior of their child to gain insight in the developing functions. To illustrate, a child that has difficulties with cognitive flexibility may experience difficulties with a changing caregiver or shift in routine. Using standardized parental rating systems is a wellaccepted evidence-based method in the assessment of social, emotional, and behavioral functioning. 36 In this study, both neurocognitive tasks and structured observations are used to provide information on executive functions in young children with SCT.
The current study is, to the best of our knowledge, the first to examine emerging executive functions in a large, international cohort of children with SCT between the ages of 3-7 years old, compared to population-based controls. As the three trisomy karyotypes (i.e., XXX, XXY, XYY) are characterized by similar neurocognitive impairments during childhood, 19,37,38 we grouped them into a single sex chromosome trisomy group. The primary goal of the current study was to investigate how executive functions present across different ages in young children with SCT, expressed in terms of information processing skills as well as behavioral observations. Given that executive functions in early childhood are considered a unitary construct, we examined executive functions by using a single performance measure that is appropriate for a large age-span. In addition, a verbal fluency task was used to examine verbal executive functions specifically.
This task was chosen as the language domain is an evident vulnerability in children with SCT, 39 already at a young age, 40,41 and we wanted to examine emerging executive functions in the context of both verbal and non-verbal based information processing. Furthermore, the behavioral report allowed for examination of smaller subdomains of executive functions that could inform on specific vulnerabilities of young children with SCT. Based on earlier research with older children and adults, we hypothesized that even pre-school age children with SCT already experience difficulties with executive functions.

| Participants
The current study is part of a large ongoing international longitudinal study (the TRIXY Early Childhood Study, at Leiden University in the  Table 1 to present a copy of the karyotyping report of the child that was provided by their clinician at time of diagnosis. Children from the control group were not subjected to genetic screening. Given the low prevalence of SCT ($1 in 1000) in the general population, we decided that the burn of blood draw for testing for SCT in our control group outweighed their potential utility. We reviewed the possible risk of having a child with undiagnosed SCT in our control group minimal and acceptable. The majority of the children with 47, XXY (57%, n = 17) did not receive testosterone replacement therapy at any given time in their development. Parental education level was assessed according to the Hollingshead criteria and ranged from category 1 (no formal education) to 7 (graduate professional training). 42 When the child was raised by two parents (95%), educational level was averaged over both parents. Parental education level varied from 4 to 7 (median 6) in the SCT group and from 2 to 7 (median 5) in control group. All participants were Dutch-or English-speaking. Children had no history of traumatic brain injury, severely impaired hearing or sight, or colorblindness.
To examine the developmental impact of SCT, children were divided into two age groups: 3-to 4-year-olds and 5-to 7-year-olds (see Table 1 for demographic variables). Groups were split at the age of 5 to ensure equal-enough sample sizes in order to maximize statistical power. This split also optimized the available data regarding the questionnaire data (e.g., 3-to 4-year-olds filled out a different version of the BRIEF compared to the 5-to-7-year-olds). The two age-groups were similar with respect to distribution of karyotype (χ 2 [2,72] = 2.088, p = 0.352) and recruitment strategy (χ 2 [2,72] = 0.185, p = 0.912). Differences between research groups (SCT vs. controls) were investigated within the two age-groups in terms of age, gender, and parental education level. Within the 3-to 4-year-old group, the SCT group included significantly more boys but was similar to the control group with respect to age and parental education level. In the 5-to 7-year-old group, children with SCT were significantly older than controls but groups were similar in terms of gender and parental education level.

| Ethics and procedure
This study was approved by the Ethical Committee of the Leiden University Medical Center, Leiden, the Netherlands, and the Colorado Multiple Institutional Review Board in the United States. A team of researchers, consisting of child psychologists, research associates, and graduate students, were trained and supervised by professionals in the field of child psychology, certified and specialized in neuropsychological assessment. All primary caregivers signed a written informed consent prior to assessment. Children were tested either in a quiet room at the University (SCT: 53%, controls: 43%) or at home (SCT: 47%, controls: 57%) using written protocols detailing all procedures and verbal instructions to standardize assessments. Researchers from Leiden University were responsible for project and data-management (i.e., training and supervision of researchers, processing, and scoring of data). The primary caregiver (92% female) of the child completed the questionnaire in either Dutch or English using the online survey software Qualtrics (http://www.qualtrics.com/).

| Executive function skills
Global executive function skills were measured with the Minnesota Executive Function Scale (MEFS App™): a standardized performancebased assessment of global executive function skills, designed for children ages 2 and up, that is administrated on a touch-screen tablet. 34 Administration time is usually 2-6 min (average of 4 min). The reliability and validity are high and the app has been used in general and clinical populations. 34 The MEFS App™ is a comprehensive executive function measure that goes down to 2 years of age and spans throughout adulthood and provides a single graded scale based on the combined assessment of working memory, inhibitory control, and cognitive flexibility. The MEFS assessment has increasing difficulty and is sensitive to age and performance, according to an adaptive testing protocol based on the responses of the child. Children are asked to sort cards into two boxes according to one rule and then switch to sorting the same cards again using an opposite or conflicting rule (see Figure 1). It requires children to switch between rules and inhibit one's automatic response. Furthermore, working memory is required to remember the current rule(s) for each trail. Because of its adaptive testing protocol, the MEFS provides a sensitive assessment of each individual child and his/her global executive function skills. After finishing the task, a total score (0-100) is calculated based on an T A B L E 1 Demographic variables of the SCT group and the control group Parental education level a median (range) 6 (4-7) 6 (2-7) 0.586 6.5 (4-7) 6 (2-7) 0.461 6 (4-7) 6 (2-7) 0.965

| Verbal executive function skills
To assess verbal executive function skills a measure of verbal fluency was used. Verbal fluency is commonly described as a measure of executive function in the context of verbal information 43 because it requires goal-directed behaviors such as cognitive flexibility, strategic planning, and error-monitoring. 14 For this study, the subtest 'word generation' of the NEPSY-II Developmental Neuropsychological Assessment was used. 44 In this subtest children are asked to generate words within two specific categories ('animals' and 'food/beverages') as many as possible within a 60-s period for each category. Administrated answers were afterwards coded to yield either 0 points for an incorrect answer or 1 point for a correct unique answer. Higher scores represent higher levels of verbal fluency. Either summed raw scores or scaled scores were used in analyses (see Section 2.4). Scaled scores were derived from the manual, using the appropriate norm group depending on the language spoken by the child (Dutch or English).

| Executive functions in daily behaviors
The Behavior Rating Inventory of Executive Function (BRIEF) was used as an assessment tool of everyday executive functions. 45

| Statistical analyses
Some data were missing due to outliers or technical dysfunction.
Resulting from this, sample sizes varied from 60 to 72 SCT children and 67 to 69 controls per analysis. Data were analyzed using IBM SPSS version 25. Demographic characteristics were compared with independent sample t-tests and Chi-square tests. As preliminary analysis, to examine whether recruitment site was relevant to executive function outcomes, t-tests were used to examine mean group differences within the SCT group (Dutch vs. US) and mean group differences within the control group (Dutch or US-referenced norms).
For each executive function measure, the following analyses steps were taken. First, to test the hypothesis that executive functions were dependent on age, a linear approach using correlation analyses was used to maximize statistical power. Because raw scores were used in these analyses, they needed to be corrected for recruitment site. We used PROCESS, a bootstrapping, nonparametric resampling procedure (for further information see References [52,53]), to control for the potential role of recruitment site in the SCT group. Subsequently, if significant effects of age on EF measures were found, subsequent ANOVAs per age-group were performed to identify group differences at specific ages using standardized norms of the EF measures. These ANOVAs were carried out as post-hoc tests to analyze the differences between SCT and control within the different age-groups (3-to 4-year-olds and 5-

| Role of age in executive functions
The results from the correlation analysis between the variable 'age' and executive function parameters are shown in Table 2. Recruitment site was included as a covariate in the analyses but there were no significant interaction effects (see Appendix B and Table B1 for Table 2). For the other domains, the strength of correlations did not differ significantly between the SCT and control group.
These results indicate that age is an important factor in executive function (problems), with differential presentation across ages, as is visible in Figure 2.

| Age-specific group differences in executive functions
As age appeared to be associated with executive functions and showed differential patterns across ages and domains (for an illustration of these specific relations see Figure 2), specific age-groups (3-4 year-olds and 5-to 7-year-olds) were examined to identify impairments in executive functions at specific ages.

| 3-to 4-year-olds
There were significant group differences found for executive functions between 3-to 4-year-old children with SCT and controls (see Table 3). Children with SCT had significantly lower verbal executive function skills than the control group (medium effect size

| 5-to 7-year-olds
There were significant group differences for executive functions between 5-to 7-year-old children with SCT and controls (for results see Table 3).     To evaluate the relevance of these findings, we reran previous analyses within separate IQ-groups (children with SCT and below average IQ vs. children with SCT and average IQ) as compared to controls (see Table 4 for the results). In both IQ groups, significant group differences were found between SCT and controls on almost all parameters, showing that difficulties with executive function were found across the range of intelligence levels and were not limited only to those children with below average IQ (also see Appendix C for additional analysis).

| Role of karyotype and recruitment strategy
In terms of the comparability of karyotypes, there were no significant group differences found between the three different karyotypes on Also critical to discussion of results is the acknowledgement of the variability in the SCT group and marked overlap with the control group such that many participants in the SCT group had scores that were similar or even improved compared to some individuals in the control group. However, statistical analyses of group differences are important as they help to delineate specific domains affected by SCT in order to understand how to support and to develop treatments for the proportion of individuals with SCT whose challenges in these areas are clinically significant such that they affect daily functioning and quality of life.
Our results revealed that children with SCT are at increased risk for problems with emerging executive functions, from as early as 3 years old, and that those problems appear more pronounced at an The increased risk for emerging executive function difficulties in children with SCT indicates that their ability to show purposeful, goaldirected, and problem-solving behavior is affected, from as early as 3 years old. The impact for these children is significant, given that preschool executive functions are vital for school readiness, 55 putting children with SCT at a substantial disadvantage at school entry. Furthermore, executive functions continue to be an important factor throughout childhood with regards to academic success, given that early executive functions also predict math and reading competence. 56 Next to school readiness and academic success, adequate executive functions also impact psychological well-being, considering that impairments in executive functions has been linked to various symptoms of psychopathology in the general population, including both internalizing and externalizing behavioral problems. 57  Our study results also underline the importance of a developmental approach with regards to neurocognition in early childhood. Albeit we studied these children cross-sectionally, our results showed that increasing age is associated with more prominent and broader executive function difficulties in children with SCT. Deviations from controls were already evident from 3 years of age, but children in the 5-to-7-year-group showed more pronounced executive function diffi- suggesting that those at risk might benefit more from stimulation than children without additional risks. However, Scionti et al. 64 65 We support the initiative of Aksglaede et al. 65 who call for a randomized and placebocontrolled trial with an adequately powered cohort sample: one of which is currently underway (PI Davis, NCT03325647).

| CONCLUSION
In sum, the present study showed that when it comes to emerging (e.g., still-developing) executive functions, many (but not all) children with SCT experience reduced performance and everyday functioning, which seems to be present from a young age (3 years

CONFLICT OF INTEREST
The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.

APPENDIX A
See Table A1.

APPENDIX B
Moderation effect of recruitment site in the SCT group Analysis description: Bias-corrected bootstrapping analyses (PROCESS) were conducted to test for a moderating effect of the recruitment site (Dutch or US) on the relations between age and executive function parameters in the SCT group. There were no significant interaction (e.g., moderation) effect of recruitment site, revealing that the relation between age and executive function outcomes did not differ across sites. See Table B1 for the exact results.

APPENDIX C
Role of estimated IQ in the group differences between SCT and controls on executive function outcomes To address the robustness of our results concerning the role of IQ, we also ran a MANOVA with the executive function parameters as dependent variables, research group as independent variable and included IQ as covariate. These results again showed a multivariate effect of research group (p = 0.004), next to a multivariate effect of IQ (p < 0.001). Thus, caregivers of children with T A B L E A 1 Group differences on executive function parameters (standardized scores) in the SCT and control group.  SCT reported significant more daily life executive functions problems (while controlling for IQ), specifically emotional control (p = 0.017) and working memory (p = 0.022). Also, children from the SCT group performed significantly less well than controls on verbal executive functions (p = 0.017). Group differences were non-significant for inhibit (p = 0.969), shift (p = 0.199), and plan/ organize (p = 0.331).