Longitudinal Investigation of Early Motor Development in Neurofibromatosis Type 1

Abstract

Objective

Research indicates that children with neurofibromatosis type 1 (NF1) have weaknesses in fine and gross motor development in early childhood; however, little is known about the stability and developmental trajectory of motor functioning. We investigated (1) whether motor difficulties are evident and stable in the preschool period in children with NF1 and (2) whether there are particular patterns of motor development in this population.

Methods

Participants with NF1 and a control group of unaffected siblings were enrolled at ages 3–8 years and were assessed yearly. Motor functioning was assessed longitudinally using the Scales of Independent Behavior-Revised Motor Scale and the Differential Ability Scales-II Copying subtest. Wilcoxon sign tests were used to compare motor functioning at 3 or 4 years to 5 or 6 years old for children with NF1 seen during both time periods (N = 27). Linear mixed model growth curve analyses were used to compare trajectories for both children with NF1 (N = 62) and unaffected siblings (N = 37).

Results

Children with NF1 made relative gains in raw scores, but not standard scores, across measures. Growth curve analyses revealed a significant effect of NF1 status on gross motor, fine motor, and copying scores, as well as an age by NF1 status effect on fine and gross motor scores.

Conclusions

Motor difficulties are evident early in life in children with NF1. Though children with NF1 clearly acquire motor skills over time, they continue to fall behind unaffected siblings, with the gap potentially widening over time. Further implications are discussed.

Introduction

Neurofibromatosis type 1 (NF1) is the most common autosomal dominant genetic disorder, affecting about 1 in 3,500 births (Huson & Hughes, 1994). In addition to a variety of medical vulnerabilities, children with NF1 show elevated rates of attention (Isenberg et al., 2013), executive functioning (Casnar & Klein-Tasman, 2017; Plasschaert et al., 2016), social (Huijbregts, 2012), and learning (North et al., 1994) difficulties. Motor functioning difficulty has also been identified across developmental periods, including toddlerhood (Lorenzo et al., 2011), childhood (Rietman et al., 2017), and adulthood (Jensen et al., 2019). However, no research to date characterizes the development of motor difficulties using a longitudinal design. Such research is crucial to understanding when motor problems develop and whether they persist, which would inform clinical care and time points for early intervention.

Motor Functioning in NF1

Children with NF1 as a group demonstrate significantly lower motor functioning in comparison to same-aged peers across several domains of motor ability. Motor development in early childhood is crucial to monitor, as motor abilities affect functioning in a variety of contexts including exploration of environments, playing, and prewriting skills. Children with NF1 show weaknesses in balance and upper limb coordination (Iannuzzi et al., 2016), sensorimotor skills (Coutinho et al., 2016), strength and agility (Johnson et al., 2010), problems with visuospatial skills (Dilts et al., 1996), and visual-motor integration (Gilboa et al., 2014).

Although it is increasingly clear that motor functioning is an area of weakness for children with NF1, the literature about motor abilities in NF1 in early childhood is quite sparse and little is known about the development of these difficulties. One study found that toddlers with NF1 have significantly poorer motor development than age-matched controls, but none of the children showed motor functioning in the significantly delayed range (index score <70; Lorenzo et al., 2011). Our group (with a sample that overlaps with this study) has also found that young children with NF1 have problems with both mildly challenging (Imitating Hand Positions and Visuomotor Precision) and complex fine motor tasks (Differential Ability Scales-II Copying), compared with both normative data and to a control group of unaffected children (Casnar et al., 2014). However, there are no known longitudinal studies documenting motor development in NF1 to determine whether young children with NF1 eventually “catch up” to their unaffected peers.

The Present Study

Aim 1 is to characterize the degree to which motor functioning is stable in early childhood by examining whether motor functioning in NF1 at 3 or 4 years old is consistent with motor functioning at 5 or 6 years old. Aim 2 is to characterize the developmental trajectory of motor functioning in children with NF1 within the early childhood period (3–8 years old) in comparison to their unaffected siblings. Elucidating the stability and trajectory of motor difficulties in early childhood may provide insight into the critical time points for early intervention. For example, if mild signs of motor difficulty at the beginning of early childhood develop into a more severe difficulty by the end of early childhood, then clinicians should be more rigorous in their treatment approach in the early childhood years. In contrast, if difficulties emerge over time, clinicians should make sure to reassess motor functioning at a later age if early motor functioning challenges are not evident. Comparison of children with NF1 to their unaffected siblings presents some advantages over a community-based control group, including controlling for some shared environmental factors, socioeconomic status (SES), and other familial variables that may otherwise contribute to group differences. Given that previous literature has suggested that motor difficulties are present across many age groups in NF1, we hypothesize that motor difficulty will be stable over time. We also hypothesize that the development of motor skills in children with NF1 will be delayed in comparison to unaffected siblings. The present study uses both a performance-based measure, which offers a snapshot of the child’s ability under controlled conditions, as well as caregiver-report, which provides a more functional and everyday picture of functioning. These approaches provide complementary clinically useful information (Ten Eycke & Dewey, 2016).

Methods

Participants

For a summary of participant enrollment, please refer to Supplementary Figure 1. The study from which these data were drawn utilized a longitudinal design; participants were enrolled between ages 3 and 8 with the plan of approximately yearly assessment from age of enrollment until age 8. For Aim 1, data from 27 children with a confirmed clinical diagnosis of NF1 who had a first visit at either age 3 or 4 and another visit at either age 5 or 6 were examined (6 additional participants were seen at age 3 or 4 but did not return at age 5 or 6). All children in Aim 1 are represented in Aim 2. For Aim 2, all available participant time points were used. The Differential Ability Scales-Second Edition (DAS-II) Early Years Copying sample had four fewer participants (two in the NF1 sample and two in the unaffected sibling sample) due to behavioral challenges. The sample for unaffected siblings is smaller than the NF1 sample as only unaffected siblings in the 3–8 age range (both older and younger) were recruited. One caregiver participated, with the caregiver remaining consistent across the study period. Participant demographic information at their first visit for both Aim 1 and Aim 2 are described in Table I.

Procedure

Recruitment took place at several Midwest Neurofibromatosis Clinics and through the National Neurofibromatosis Research Registry. Inclusion criteria included: physician diagnosis of NF1, age between 3 and 8 years, and English as the first and main language spoken in the home. Exclusion criteria included: any comorbid conditions not commonly associated with NF1 and a significant surgery within 6 months; no participants were excluded. For the unaffected sibling group, inclusion criteria included: age between 3 and 8 years and English as the first and main language spoken in the home. Informed consent was obtained prior to the study visit or at the study visit. Participants were administered a 4-hr age-appropriate neuropsychological battery (with appropriate breaks), including measures from a variety of domains, by trained study team members. This study was approved by the University of Wisconsin-Milwaukee, Children’s Hospitals of Wisconsin, and University of Chicago Medical Center’s Institutional Review Boards.

Measures

The Scales of Independent Behavior—Revised (SIB-R; Bruininks et al., 1996) is a caregiver interview that examines motor, social interaction and communication, personal living, and community living skills. The SIB-R Motor Scales demonstrate acceptable to excellent internal consistency across age groups (Gross Motor: r = .62–.92; Fine Motor: r = .74–.89; Motor Scale: r = .84–.89), excellent test–retest reliability (Gross Motor: r = .96; Fine Motor: r = .92; Motor Scale: r = .96), and good to excellent validity (Bruininks et al., 1996). The Motor scale, which is a composite of the Gross-Motor and Fine-Motor subscales, in addition to Gross-Motor and Fine-Motor subscales were used to assess caregiver-report of motor functioning. The Gross Motor scale includes 19 items related to balance, coordination, strength and endurance. The Fine Motor scale includes 19 items related to precise movements and hand–eye coordination using small muscles. For each item, caregivers indicate whether the child can do the task completely without help on a 4-point scale from 0 (never or rarely does the task without help) to 3 (does the task very well without help). Standard scores are provided for the Motor scale. No standard scores are available at the subscale level thus, subscale raw scores were used. Higher scores represent stronger abilities.

The DAS-II Early Years (Elliott, 2007) is a performance-based measure of cognitive abilities including verbal reasoning, nonverbal reasoning, and spatial abilities in children. The DAS-II demonstrates excellent internal consistency (r = .79–.92), test–retest reliability (r = .73–.94), and validity (Elliott, 2007). The General Conceptual Ability (GCA) standard score was used to examine overall cognitive function. The DAS-II Copying subtest T-score and ability score (which is akin to a raw score) were used to examine motor abilities. The DAS-II Copying subtest is a paper and pencil task that requires that the child copy simple line drawings of increasing complexity and thereby examines fine motor coordination and visuospatial skills. The Copying subtest has a total of 20 items; however, this is an adaptive test with blocks of items administered depending on participant performance. Higher scores represent stronger abilities.

Demographic information was collected using a background questionnaire developed by the study investigators. Occupation and education level of the participant’s caregivers were used to calculate SES according to the Hollingshead Four-Factor Index of Social Status (Hollingshead, 1975). SES scores range from 8 to 66 with higher scores reflecting higher SES.

Analytic Approach

Data were analyzed using IBM SPSS version 25 and RStudio version 3.4.4. Only results that survived false discovery rate (Benjamini & Hochberg, 1995) correction at q < .05 were interpreted to correct for multiple comparisons.

Aim 1

Mann–Whitney U-tests and Spearman’s rho correlations were conducted using data from the first assessment year (i.e., ages 3 or 4) to examine whether there were any significant associations of demographic characteristics (age, sex, NF etiology classification, or SES) with motor functioning (see Relations to Demographics section). Wilcoxon sign-rank tests were conducted comparing motor functioning at 3 or 4 years to motor functioning at 5 or 6 years old for children with NF1 seen during both time periods (N = 27) using the following dependent variables: DAS-II Copying T-scores and ability scores, SIB-R Motor Scale standard scores, SIB-R Gross Motor raw scores, and SIB-R Fine Motor raw scores.

Aim 2

First, Mann–Whitney U-Tests and Spearman’s rho correlations were conducted to examine whether there were any significant differences on the DAS-II Copying or SIB-R scales based on demographic characteristics using data from the first assessment year (see Relations to Demographics section). Next, Mann–Whitney U-tests and chi-square tests of independence were conducted to assess differences between the participants who had one visit during early childhood and those who returned for a subsequent visit(s). Finally, linear mixed model (LMM) growth curve analyses were conducted using lme and lme4 packages. Each model was centered around age 4 using mean raw scores (SIB-R Gross Motor: 32.83; SIB-R Fine Motor: 30.59; DAS-II Copying: 67.31) with precise age (e.g., 5.15 years) and NF1 status (NF1, sibling) as fixed effects and participant ID as a random effect using the Restricted Maximum Likelihood Approach. LMM analyses used raw scores for SIB-R Fine (NF1 N = 62; unaffected siblings N = 37) and Gross Motor scales (NF1 N = 62; unaffected siblings N = 37) and ability scores for the DAS-II Copying scale (NF1 N = 58; unaffected siblings N = 34). Trajectories of the NF1 and unaffected sibling groups were compared. Although the present sample is smaller than preferred for such analyses, Curran et al. (2010) indicate that the guidelines for running such analyses are vague, and such statistical methodology has indeed been used with small samples (Kuckertz et al., 2020; van Leuteren et al., 2020; Teigset et al., 2018). Growth curve analyses flexibly account for missing data and inconsistent spacing between time points. This methodology allows for analysis of overall trends in the development of motor functioning through the preschool years and into the early school age years.

Results

Aim 1

Relations to Demographics

For NF1 participants at the first assessment, SIB-R Fine Motor (q < .001) and Gross Motor (q = .008) raw scores were significantly correlated with age. SES, NF1 etiology, and sex were not significantly related to the motor functioning study variables.

Results

Standard scores on the SIB-R Motor Scale significantly decreased from age 3 or 4 to age 5 or 6 (Z = −2.37, q = .022), whereas SIB-R Gross Motor raw scores (Z = 3.63, q < .001), SIB-R Fine Motor raw scores (Z = 4.44, q < .001), and Copying Ability scores (Z = 4.35, q < .001) increased over time (see Table II). DAS-II Copying T-scores did not significantly change.

Aim 2

Relations to Demographics

SES, NF1 etiology, and sex were not significantly related to the motor functioning study variables for the NF1 sample. For unaffected sibling participants, SIB-R Fine Motor (q < .001), Gross Motor (q < .001), and DAS-II Copying (q < .001) raw scores increased significantly with age. Sex was not significantly related to the study variables. Paradoxically, SES was negatively correlated with DAS-II Copying T-score (q = .043) in the unaffected sibling sample.

Attrition

In total, 15 of 62 children with NF1 (24%) and 15/37 unaffected siblings (40%) had at least one visit from ages 3 to 8 years but did not return for a subsequent visit. No significant differences were found on most study variables (sex, NF1 etiology, SES, DAS-II GCA, DAS-II Copying, and SIB-R Motor standard score) between participants with 1 visit and those with more than 1 visit. For both children with NF1 and unaffected siblings, SIB-R Fine Motor raw scores at visit 1 were significantly lower for those who returned for subsequent visits (NF1 U = 186.0, q = .008; Unaffected siblings U = 57.5, q = .017). For unaffected siblings (but not participants with NF1), SIB-R Gross Motor raw scores were also significantly lower at visit 1 for those who returned for subsequent visits (U = 74.0, q = .008).

Growth Curve Analyses Results

Model summaries can be found in Table III and Loess Lines for each model can be found in Figures 1–3. For each model, age had a significant effect on performance: SIB-R Fine Motor (q < .001), Gross Motor (q < .001), and DAS-II Copying (q < .001). There was a significant effect of NF1 Status on SIB-R Fine Motor (q = .025), Gross Motor (q < .001), and DAS-II Copying (q < .001) scores. There was a significant age by NF1 status interaction on SIB-R Fine (q < .001) and Gross Motor (q < .001), but not DAS-II Copying Scores. The Loess Lines of the SIB-R Fine Motor and Gross Motor raw scores depict a growing gap in functioning with increasing age between the NF1 and unaffected sibling group. The Loess Lines of the DAS-II Copying Ability scores demonstrate consistently poorer performance in the NF1 group in comparison to unaffected siblings across time.

Discussion

Although motor functioning is known to be an area of difficulty for children with NF1 (Rietman et al., 2017), there is sparse literature about the motor functioning of young children with NF1 and there have been no investigations documenting longitudinal patterns of motor development in this population. In this study, we used both caregiver-report and performance-based measures of motor development. We first analyzed motor functioning longitudinally within the early childhood period by comparing motor performance of children with NF1 at 3 or 4 years old with their subsequent performance at 5 or 6 years old. Our findings show that young children with NF1 are making gains in both their gross and fine motor abilities based on raw score analysis within the preschool period. However, examination of standard scores indicate that their level of functioning in comparison to same-aged peers (using normative data) either does not improve over time (i.e., on DAS-II Copying) or worsens (i.e., on the SIB-R Motor scale). Further, we expanded on these analyses by calculating LMM growth curves of motor performance across a broader age range (3–8 years old) to gain a better understanding of the patterns of motor development in comparison to a group of unaffected siblings. Based on the visualization of the data using Loess Lines comparing the participants with NF1 to their unaffected siblings over a somewhat broader developmental period, it is evident that although young children with NF1 are acquiring motor skills, they fall behind unaffected siblings in early childhood. On the performance-based measure or visuospatial skills (DAS-II Copying), the NF1 group’s scores were consistently lower than the unaffected group’s scores, though the pattern of gains in skills was similar across time. In the case of the SIB-R Fine and Gross motor scales, the gap in ratings widens over time beginning around age 5.5–6.5 years.

For exploratory purposes, we also examined the data at each age point by comparing children with NF1 and their unaffected siblings on each measure (see Supplementary Table 1). These findings parallel the findings found in the growth curves, in that differences between the two groups emerged at varying time points across measures. In this case, the performance-based task (DAS-II Copying) identified difficulties earlier than the caregiver-report; it is possible that caregivers (SIB-R) may not observe difficulties until demands increase and developmental differences start diverging more noticeably.

Because there are no published data using imaging techniques longitudinally in young children with NF1, we must look to the unaffected literature for considerations of brain-behavior relationships. Beginning at 5 years old, children’s brains begin expanding in the prefrontal cortex at a rate of about 1 mm/year (Sowell et al., 2004), which typically coincides with identification of more subtle motor functioning problems (Piek et al., 2012). The prefrontal cortex is an integral component of different aspects of motor ability, such as sensory-motor processes (Piek et al., 2012). Thus, it may be the case that it is at this developmental juncture that children with NF1 and their unaffected siblings have differing neural developmental patterns, which may lead to observable difficulties on every day, caregiver-report and performance-based measures. Further longitudinal investigations using neuroimaging techniques are indeed warranted to delineate these neurodevelopmental differences and developmental patterns.

Implications

Our data suggest that motor difficulties for individuals with NF1 begin early in life and continue to worsen at least through early childhood, in comparison to normative data and unaffected siblings, with slightly different patterns of emergence of difficulties for different indicators of functioning. Difficulties were present based on both caregiver-report and performance-based measures; thus, both modalities may be appropriate tools for assessing these difficulties in this developmental period. Focused assessments or screenings of motor abilities are recommended early in life for children with NF1. Clinicians working with young children with NF1 are encouraged to routinely screen for motor difficulties using either or both modalities even early in the child’s development and throughout the early childhood period or to refer families to specialists who are able to do so. Early intervention is recommended for fine motor, gross motor, and visuospatial skills at least as early as 5–6 years of age; it is at this juncture that a gap in some motor abilities begins to develop between children with NF1 and unaffected siblings. Difficulties in comparison to unaffected siblings on performance-based tasks may be evident as early as 4 years, gross motor challenges based on caregiver-report by 5.5 years, and fine motor difficulties are evident to caregivers by about 6.5 years. It is possible that challenges could be identified even earlier; this study did not include younger children. To date, there are no well-validated treatments for poor motor functioning specific to NF1, thus we must look to other populations. Generally, preschool aged children benefit from individual therapy (e.g., occupational therapy, physiotherapy) that is tailored to the specific motor deficit or delay present in the child (Piek et al., 2012).

Limitations and Future Directions

Growth curve analyses allow for the inclusion of participants with only one data point, and one limitation of this study is that a number of participants in this sample were not seen multiple times. To verify that the present findings were evident when only longitudinal participants were included, Aim 2 analyses were also run including data of participants with only two or more visits, and the models produced similar findings to those presented in the present Results section, with lower residual variance. Further, replication with a larger sample is also needed. Another limitation of the present study is that our design did not include traditional physical and occupational therapy clinical measures of motor development (e.g., grip strength, gait). However, caregiver and performance-based measures often used clinically in psychological settings are related to these measures (Jewell et al., 2010). This study did not include blinding procedures for test administrators of the DAS-II, and there were no specific strategies to minimize caregiver bias in their reports on the SIB-R. However, caregivers reported on specific motor abilities (e.g., “turns knob and opens a door,” “stands alone and walks for at least 6 feet,”), rather than providing general, subjective and comparative impressions of motor ability; as such, this is a behaviorally anchored interview. Last, a natural extension of this work would be to assess motor development in even younger children and to track motor development in individuals with NF1 into the school age years and beyond. Based on previous literature, it is clear that motor dysfunction is common in NF1 and may be tied to developmental cascades related to the functioning of the NF1 gene (Summers et al., 2015); thus, since NF1 is a progressive neurodevelopmental condition, it would be helpful to track patterns over time and identify early risk factors that increase the likelihood of motor impairment later in life.

Conclusions

Although it has been evident that individuals with NF1 are at an increased risk of motor difficulties (Torres Nupan et al., 2017), the longitudinal, developmental nature of these concerns had not been previously documented. The findings in this study suggest that fine motor, gross motor, and visuospatial skill difficulties begin in the preschool years for individuals with NF1. Further, the gap in functional ratings of motor abilities between NF1 abilities and those of unaffected siblings persists and may even widen with increasing age. In particular, children with NF1 displayed significantly different developmental trajectories than their unaffected siblings on caregiver reports of fine and gross motor functioning. Early screening for motor difficulties and intervention is highly recommended for young children with NF1.

Supplementary Data

Supplementary data can be found at: https://dbpia.nl.go.kr/jpepsy.

Acknowledgments

The authors would like to thank the participants and their families for participating and returning year after year. The authors would like to thank Scott Hunter, James Tonsgard, Pamela Trapane, Dawn Siegel, Donald Basel, and Robert Listernick for referring participants with NF1 for this research. The authors would also like to thank Ryan Sullivan and G. Nathanael Schwarz for their consultation regarding statistical analysis. We also acknowledge the Child Neurodevelopment Research Lab team for their work and collaboration in the laboratory.

Funding

This work was supported by grants from the NF Midwest; NF MidAtlantic; NF Northeast; University of Chicago CTSA (UL1 RR024999); and the University of Wisconsin–Milwaukee Research Growth Initiative.

Conflicts of interest: None declared.

References