Less Affected Upper Extremity Impairments Increase with Severity of the More Affected Upper Extremity in Children Following Perinatal Unilateral Stroke

Todd J. Levy¹, Rebecca N Ichord², and Lauren A Beslow²

¹ Center for Rehabilitation, Children’s Hospital of Philadelphia
² Division of Neurology, Children’s Hospital of Philadelphia, Departments of Neurology and Pediatrics, Perelman School of Medicine at the University of Pennsylvania

Corresponding author:
Todd J. Levy, MS, OTR/L, CBIST-AP
levytj@chop.edu
3500 Civic Center Blvd
Buerger Center, Room 4109
Philadelphia, PA 19104
215-590-7980

Abstract

Background: Children with hemiparesis rely more heavily on their dominant hand than typically developing peers. This study evaluates the motor performance of both upper extremities in children with unilateral perinatal stroke.

Methods: A retrospective analysis was conducted on a cohort of 33 children with unilateral perinatal arterial ischemic stroke (PAIS) confirmed by imaging. Data were collected from 2012 to 2023 in a multidisciplinary stroke clinic. Severity of the more affected limb paresis was categorized as mild, moderate, or severe. Motor dexterity was measured using the Box and Blocks Test (BBT) and the Nine-hole Peg Test (NHPT).

Results: Median age at testing was seven years (range 3-15 years). Children showed poor performance with the paretic limb and performed worse with their less affected limb compared to typically developing peers’ dominant limb on the BBT (mean Z-score = -2.1). ANOVA indicated that less affected limb BBT test scores declined with increasing severity of contralateral paresis (F (2,24) = 13.48, p = .0001). NHPT performance with the less affected limb was also reduced (mean Z-score = –1.4), with a non-significant trend toward poorer less affected limb performance with increasing severity of the more affected limb. Performance between limbs was positively correlated (r = .77 for BBT and r = .23 for NHPT). Nearly all participants performed worse than clinically meaningful benchmarks, with larger deviations for the more affected limb.

Conclusion: Children with PAIS demonstrate bilateral upper extremity impairments, and performance of the less affected limb declines with greater contralateral paresis.

Keywords: hemiparesis, pediatric stroke, unilateral cerebral palsy, less affected hand, contralateral hand, unaffected hand, motor outcomes

Introduction

Cerebral palsy (CP) is one of the most common motor disabilities in children, with a prevalence of 2.11 per 1000 births worldwide.^1,2 Unilateral cerebral palsy (UCP), a common subtype, has become more prevalent as rates of diplegic CP have declined.³ Paresis of one side is pathognomonic of UCP, yet performance deficits of both upper extremities have been reported, largely in studies of UCP of unknown or mixed etiology.^4–8 Some authors and clinicians describe the ipsilesional limb as less affected^4,8 while others use the terms affected/unaffected, paretic/nonparetic, or contralateral/ipsilateral. For clarity and consistency, we will use less affected (LA) to describe the upper extremity ipsilateral to the brain injury, and more affected (MA) to describe the upper extremity contralateral to the brain injury.

Nonetheless, the LA remains comparatively neglected in both clinical practice and research. A clearer understanding of each upper extremity, and of their spatial and temporal coordination, might improve the specificity and effectiveness of rehabilitation interventions. Moving towards this goal, it is essential to quantify motor performance of both upper extremities relative to normative data and whether observed differences rise to the level of clinically meaningful deficits. We addressed this by studying a more etiological homogenous population of children with hemiparesis following perinatal arterial ischemic stroke (PAIS). This choice is supported by converging evidence from network neuroscience and neonatal brain injury animal models, suggesting that functional outcomes may extend beyond the immediately injured area. For example, unilateral hippocampal injury in neonatal rats leads to severe, lasting effects on spatial learning and memory similar to bilateral injury in adults.^9,10 In this model, unilateral injury is associated with reduced long term potentiation and BDNF gene expression in the more affected hippocampus.¹⁰ In humans, functional impairments often appear disproportionate to neuroimaging findings, consistent with network disruption.¹¹ Recently, Craig et al.¹² demonstrated altered structural connectivity within the non-lesioned (less affected) sensorimotor networks of children following perinatal stroke, with differences in key motor areas compared to controls. Moreover, those changes were associated with clinical outcomes of the LA.

Together, findings from neonatal animal models and human neuroimaging suggest that focal injury can broadly disrupt motor systems. The successful use of the hands during everyday activities also depends on higher-order cognitive processes such as motor planning and coordination which are known to be vulnerable in UCP.^5,13–15 Prior to executing a goal-directed movement such as throwing a ball, or pouring a cup of water, the child must plan to move based on object weight and size, and other variables. Moreover, most daily activities are bimanual in nature,¹⁶ and coordinating the upper extremities in time and space increases the cognitive load. Occupational therapy (OT) is concerned with helping children accomplish everyday tasks, and therefore motor planning related to both extremities, and coordinating them together, have been common targets of intervention.^17–20 For children with severe paresis, rehabilitation of LA skills might be especially important, as many depend on that extremity.

Insights from adult stroke research provide further context. Impairments of the LA following unilateral stroke in adults are increasingly recognized,^21–23 and these deficits scale with the severity of the MA paresis.²⁴ For example, Johnson and Westlake²⁵ showed that adults have impaired gross and fine motor dexterity of the LA as measured by the BBT and the NHPT. Within the pediatric stroke literature, unilateral dexterity deficits of each hand have been noted in the context of a structural connectivity study¹² and a kinematic reaching study,²⁶ but not directly characterized. As the leading cause of UCP and a major contributor to lifelong disability,²⁷ PAIS provides a critical context for our investigation. Building on this background, we hypothesized that children with PAIS would show impaired dexterity with each hand and that LA performance would worsen with severity of MA paresis.

Materials and Methods

Ethics and Data Availability

All cases were consented participants in Children’s Hospital of Philadelphia’s (CHOP) Pediatric Stroke Registry which was approved by the hospital’s Institutional Review Board. Anonymized data are available by appropriately approved requests from qualified investigators.

Study Design and Inclusion Criteria

We performed a retrospective analysis of a single-center prospectively enrolled consecutive cohort of children with PAIS who received care in a multidisciplinary stroke clinic from 2012 to August 2023. We included children with a history of PAIS confirmed by magnetic resonance imaging or computed tomography based on the neuroradiology report and confirmed by expert review of the imaging by a pediatric neurologist. We included only patients with unilateral brain injury. We included both subtypes of PAIS, namely acute neonatal arterial ischemic stroke (AIS) (symptomatic presentation within the first 28 days of life) and presumed perinatal arterial ischemic stroke (recognized later in infancy or childhood when hemiparesis became apparent). Long-term motor outcomes after PAIS are multifactorial, reflecting interactions among injury characteristics, timing, and developmental factors. By limiting our cohort to unilateral PAIS, we examined outcomes in a more homogeneous group than previous UCP studies. For all comparisons of outcomes between hands, we used the first instance when data was collected on both hands within the same visit. In a single case, testing of each hand occurred on different days, one year apart. We excluded any outcome data collected following hemispherectomy or upper extremity surgical intervention undertaken to improve function. We reviewed three cases in which outcome testing occurred within three months of anti-spasticity injections. In all three cases, data were similar when collected within and outside the three-month window, thus we included those cases.

Data Collection

Data were abstracted from clinical follow-up records including OT and neurology reports. Study data were collected using Research Electronic Data Capture tools,^28,29 the Epic electronic medical record, and Excel, hosted at CHOP. All motor performance data were obtained well into the chronic phase of injury, between the ages of 3 and 15 years of age.

Clinical Data

We defined severity of MA paresis based on clinical impression of functional use of the MA at or near the time of outcomes testing, using a clinically derived categorization for descriptive stratification purposes. Our focus was on how severity of hemiparesis influences the performance of each upper extremity, including the LA. The OT notes were used to classify severity as mild, moderate, or severe. Classification was based on chart review of occupational therapy (OT) documentation, with emphasis on functional use of the MA during daily activities, rather than performance on the BBT or NHPT (Table 1). We did not use classification systems such as the Manual Ability Classification System (MACS) or the Bimanual Fine Motor Function (BFMF) scale, as both are based on bimanual performance and therefore do not specifically capture the severity of paresis of the MA. Accordingly, severity groupings were not intended to function as a validated classification system, but rather to support comparisons of motor performance across clinically meaningful levels of hemiparesis.

Table 1

Categorization of more affected limb paresis. The table presents the classification of severity for the more affected upper extremity paresis at the time of testing. Severity was categorized into mild, moderate, and severe based on clinical assessment criteria.

Outcomes

The BBT assessed unilateral gross motor dexterity using a grasp-carry-release sequence to transfer blocks (2.5 cm x 2.5 cm x 2.5 cm) between a partition. The number of blocks transferred in 60-seconds was recorded for each hand. Higher scores reflected better performance. Procedures were standardized as described by Jongbloed-Pereboom et al.³⁰ Construct validity and test-retest reliability (ICC = 0.98) has been demonstrated for children with UCP.³¹

The NHPT assessed unimanual fine motor dexterity using a pinch-carry-shift-place sequence to transfer small pegs (3.0 cm x 0.6 cm) between a well and a pegboard. The task requires isolated pinch, fine motor control, and either in-hand manipulation or compensatory strategies to align pegs with holes. The number of seconds required to complete the test was recorded for each hand. Higher scores reflected poorer performance. Standardized procedures were followed as described by Wang et al.,³² with the following exception: one trial for each hand was administered at each follow-up visit whereas ³² recorded the best of two trials. Test-retest reliability is high for healthy children (ICC = 0.95 for right hand, 0.92 for left hand).³³ When a child was unable to complete the NHPT with the MA, or when the tester (first author in most cases) judged that completing the test would impose undue burden, a floor score was assigned. For Z-score analyses, this was determined using a lower-fence technique (1.5 × the first quartile score of the cohort), corresponding to a Z-score of –50.4 when standardized relative to age norms. This value reflects a constructed floor for non-completion rather than an observed deviation in performance. For the percent-deviation metric introduced later, we instead used the lowest observed score for the cohort (162 s) to preserve clinical relevance and avoid inflating deviation beyond empirically recorded values.

Statistics

Excel^® (version 2022) was used for descriptive statistics and Pearson product-moment correlations. Stata^® (version 16.1, StataCorp, College Station, TX) was used for all other analyses.

For continuous dexterity data, we used Z-scores to compare each child’s MA performance to the nondominant hand of typically developing peers, and their LA performance to the dominant hand of typically developing peers. For children 3-10 years of age, we used BBT normative references stratified by sex and dominance.³⁰ For children over 10 years of age, we used BBT normative references stratified by sex and laterality (right/left hand).³⁴ Laterality was reported in the later study due to the finding that both right-handed and left-handed individuals performed better with their right hands. For the NHPT, Z-scores for each child were calculated based on normative references stratified by age and gender.³²

Group comparisons were based on mean or median normative Z-scores, depending on the distribution. Normality was assessed with Shapiro–Wilk tests. For paired comparisons, paired t-tests were used when distributions were normal, and Wilcoxon signed-rank tests were used when distributions were skewed. For independent group comparisons, one-way ANOVAs were conducted when distributions were normal, and Wilcoxon rank-sum tests were used when distributions were skewed. Specifically, separate one-way ANOVAs were used to analyze BBT scores by severity of MA paresis (mild, moderate, severe), and Wilcoxon rank-sum tests were used to analyze NHPT scores by severity of MA paresis (mild, moderate). Analyses for the MA and LA were conducted separately to reflect the study aim of characterizing performance of each limb relative to normative expectations, and examining whether severity of MA paresis was associated with performance of the LA. For ANOVAs, we report the F statistic; for paired t-tests, the t statistic; and for Wilcoxon tests, the Z statistic from the normal approximation as provided by Stata^®. We used Pearson correlations to examine the relationship between severity of MA paresis and outcomes of the LA. A p < .05 was considered statistically significant.

In addition, we examined each participant’s performance relative to benchmarks of important change. The study by Liang et al.³¹ provided a range of 5.3-6.5 blocks as the minimal clinically important difference (MCID) for the BBT in children with UCP; we adopted the upper bound as our benchmark (7 blocks). In the absence of a published MCID for the NHPT, we referenced the pediatric study by Mendoza-Sánchez et al.,³⁵ which reported a minimal detectable change (MDC) of 4 seconds in children with UCP, using this as a change threshold. Each participant’s deviation from age-normative means was expressed as a percentage relative to these benchmarks. To ensure validity, MCID-based analyses of the BBT were restricted to participants aged 6–17 years, reflecting the reference population.³¹ Because MCID and MDC are fixed external benchmarks rather than sample-dependent estimates, no inferential statistical tests were conducted on these metrics. Instead, we summarized distributions graphically to provide a clinically interpretable frame of reference.

Results

Subject Characteristics

We evaluated 166 individual stroke registry patients for eligibility, finding 41 with BBT and/or NHPT data. We excluded four patients who were not documented as having hemiparesis, three with evidence of bilateral brain injury, and one who underwent selective upper extremity neurectomy, resulting in 33 individuals for analyses. Eighteen individuals met criteria for acute neonatal AIS and 15 for presumed perinatal AIS. Participants ranged from 3 to 15 years of age (median 7 years), with most assessments completed between 3 and 6 years and fewer in later childhood (≥10 years). Demographic and injury characteristics are summarized in Table 2.

Table 2

Participant demographics and age at testing. (A) Participant demographics and age at testing. (B) Age distribution at Box and Blocks Test (BBT) and Nine-hole Peg Test (NHPT) testing.

(A) Participant demographics and age at testing

*Severity based on overall chart review at time of testing

(B) Age distribution at BBT and NHPT testing (years)

*One participant was tested on the more affected limb at 12 years of age and on the less affected limb at 13 years.

Outcomes

BBT data were available for 27 individuals (8 mild, 11 moderate, 8 severe), while NHPT data were available for 18 (9 mild, 8 moderate, 1 severe); thus, more individuals completed the BBT than the NHPT. Across multiple visits, severity of paresis remained consistent except for one individual whose severity increased from mild to moderate between BBT and NHPT assessment at ages three and four. The OT determined the appropriate measures based on clinical needs within the context of a multidisciplinary visit. Typically, the NHPT was not administered to those with severe injury due to task demands, as they were presumed to be unable to complete it with their MA.

Comparisons to Normative Values

Figure 1A and 1B show paired participant-level Z-scores for the BBT and NHPT, respectively, comparing performance of the more affected and less affected hands relative to normative references. Mean BBT Z-score was -5.4 (0.0 percentile rank) for the MA and -2.1 (1.79 percentile rank) for the LA. For the NHPT, mean Z-score was -29.4 (0.0 percentile rank) for the MA and -1.4 (8.1 percentile rank) for the LA. Although the mean NHPT Z-score for the LA was <1.5 SD below the normative reference mean, five out of 18 children scored >3.0 SD below the mean, with one score as low as -7.0 SD. When using their MA, all five of these children received the floor score because they were unable or presumed unable to complete testing, consistent with the task demands of the NHPT.

LA Compared to MA

A paired t-test (95% confidence interval) showed that the mean BBT score for the MA (M = -5.4, SD = 2.9) was significantly lower than the LA (M = -2.1, SD = 1.7) for our cohort (t (26) = -8.7, p < .0001). A Wilcoxon signed-rank test showed that the median NHPT score for the more affected limb (-27.7) was significantly lower (worse score) than the LA (Median = -0.6) for our cohort (Z = -3.48, p <.001).

Severity of Paresis

Figure 2A and 2B show mean gross motor dexterity as measured by the BBT for each hand, organized by severity of paresis. A one-way ANOVA revealed a main effect of severity on MA performance (F (2,24) = 14.80, p = .0001). Bonferroni post-hoc analysis indicated that scores of the mild group were significantly better than the moderate and severe group (p’s < .001). The moderate and severe groups did not significantly differ (p = 1.0). Similarly, a separate one-way ANOVA revealed a main effect of severity on LA performance (F (2,24) = 13.48, p = .0001). Bonferroni post-hoc analysis indicated that scores of the mild group were significantly better than the moderate group (p = .001) and the severe group (p < .001). The moderate and severe groups did not significantly differ (p = 1.0).

Figure 3A and 3B show mean fine motor dexterity as measured by the NHPT for each hand, organized by severity of paresis. Because NHPT data were only available for one severe case, only the mild and moderate groups were included in the analysis of severity. A Wilcoxon rank-sum test revealed a significant effect of severity for the MA. Median MA Z-scores were -6.2 (IQR -14.6 to -2.6) in the mild group and -50.4 (IQR -50.4 to -39.05) in the moderate group, (Z = 2.34, p <.02), reflecting the presence of floor scores in individuals who were unable to complete the task. There was no significant difference in LA scores between severity groups. Median LA Z-score was -0.3 (IQR -1.8 to -0.2) in the mild group and -0.75 (IQR -4 to 0.75) in the moderate group, (Z = 0.34, p = 0.74).

For the BBT and the NHPT, there were positive Pearson product-moment correlations between the MA and LA, r = .77 (p <.001) and r = .23 (p =.35), respectively.

Comparisons to Minimal Clinically Important Differences or Minimal Detectable Change

Figure 4A and 4B show percent deviation from the minimal clinically important difference (MCID) for the BBT and from the minimal detectable change (MDC) for the NHPT. On the BBT, the entire distribution for both hands fell below –100%. On the NHPT, the more affected hand showed deviations exceeding 200% (slower), with a median nearly 10 times greater than the MDC. For the less affected hand, the median deviation was 75%.

Discussion

Our findings demonstrate that children with hemiparesis following PAIS perform differently than typically developing peers when using either hand. While deficits in the more affected limb are expected, participants consistently showed underperformance of the less affected limb relative to age-based norms. This result, although less intuitive, aligns with patterns reported in UCP and adult stroke. Rich et al.⁸ observed slower Jebson Test of Hand Function test performance for the LA compared to the dominant hand of typically developing peers, Holmström et al.³⁶ found most children with UCP scored below age norms on the BBT with their dominant hand, and Groeschel et al.³⁷ reported lower scores for both hands in children with neonatal arterial ischemic stroke.

We applied complementary metrics to characterize motor performance: (1) Z-scores referenced to age norms and (2) percent deviation relative to clinical thresholds (MCID for BBT; MDC for NHPT). Expressing deviations relative to clinical thresholds as percentages provided a standardized way to interpret the extent of impairment beyond simple normative classifications. On the BBT, all children scored at least 100% below the MCID with both hands. On the NHPT, the MA limb showed profound impairment: all children performed more than 200% worse than the MDC, with a median deviation nearly 10 times higher than the threshold. For the LA, the median deviation was 75% worse than the MDC—typically within one MDC unit—but half of the cohort still fell outside this range, more of them worse and a small minority better. Because MDC reflects measurement error rather than clinical importance, these findings must be interpreted cautiously but suggest that differences in less affected performance may exceed test variability in a subset of children. This reinforces the need to attend to both hands when characterizing functional outcomes in children with PAIS.

Gross manual dexterity of each hand decreased with greater severity of hemiparesis. Children classified as mildly affected performed significantly better than those with moderate or severe paresis, though no differences were observed between the latter groups. Fine motor dexterity followed a similar pattern. The relatively favorable mean NHPT score (Z = -1.4) for the LA likely reflects the clinic’s tendency to administer the test primarily to children with mild paresis. Nonetheless, five out of 18 children scored >3.0 SD below the normative reference, with one score as low as -7.0 SD. Each of these five children received the floor score for their MA because they were unable or presumed unable to complete the test with their MA.

Positive correlations between LA and MA performance further support our hypothesis that motor skills of the two limbs are linked (significant for the BBT but not the NHPT). A compensatory role of the more affected cortex in our patients with moderate to severe paresis might come at some expense to the LA. In adults, involvement of the more affected motor cortex in paretic hand movement increases as a function of severity of paresis.³⁸ More broadly, bilateral deficits are consistent with phenomena such as apoptosis, diaschisis, and network-level disruption, which often produce impairments disproportionate to the size or location of visible injury.¹¹ Bilateral brain injury may therefore be under-detected and underreported in PAIS.^39,40

These results complement prior pediatric stroke studies. Kuczynski et al.²⁶ reported impaired reaching kinematics, but only modest associations with clinical measures, and did not compare LA performance to normative standards. Our study extends this work by showing age-normed Z-scores and the influence of severity on less affected hand function. Similarly, while Craig et al.¹² found structural connectivity changes correlated with unilateral motor outcomes, they did not report Z-scores for the BBT. By integrating both normative and clinically anchored metrics, our findings highlight that less affected impairments are measurable, related to severity, and likely of functional significance.

Further research is required to understand the underlying causes of LA performance deficits. The observed asymmetry between limbs, with consistently lower performance of the MA and relative preservation of the LA, may reflect patterns of motor reorganization following early unilateral brain injury. Evidence suggests that the non-lesioned hemisphere undergoes structural and functional reorganization and contributes to motor function after perinatal stroke Craig et al.¹² In some children, this may include bilateral or shared control of the upper extremities Zewdie et al.⁴¹ In the context of more severe paresis, increased reliance on the non-lesioned hemisphere to support the MA may contribute to the observed asymmetry across limbs, potentially reflecting competing demands on shared neural systems. While our data do not directly test these mechanisms, models of developmental motor reorganization may provide a physiological context for the observed differences in limb performance. Additional work to better characterize these mechanisms may further inform rehabilitation strategies, including how interventions are targeted across limbs and how timing and approach are optimized to support motor development.

Clinical Implications

While the more affected paretic hand naturally draws attention, our results highlight the importance of the LA, especially for children with more severe paresis who depend on it. Our findings support the importance of routine assessment of each hand in children with PAIS, as impairments of the less affected limb may be present, related to severity, and potentially functionally meaningful based on the individual’s goals or functional demands. The BBT and the NHPT are efficient for screening the LA. Even mild dexterity issues in the LA may warrant rehabilitation referrals for activities such as 1-handed typing, playing instruments, or sports.

Further research could refine intervention. Early evidence supports targeted rehabilitation for the LA to improve limb performance and overall functional independence in adults after stroke.⁴² For children, clinicians can leverage bimanual therapy (BMT) to impact both limbs and broad neural networks.

Understanding each upper extremity alongside higher order cognitive functioning may help contextualize intervention planning. While our study did not directly examine motor planning, prior work suggests that impairments in motor planning may influence bilateral motor performance, providing relevant context for interpreting our findings. Research shows children with UCP exhibit impairments of anticipatory motor planning^13,14 without the age-related improvements seen in typically developing children,⁴ suggesting that such deficits are unlikely to resolve spontaneously. Focusing on motor planning is another strategy. In adults with unilateral stroke, directing attention to the external effects of movement rather than to kinematics enhanced anticipatory planning and task success.⁴³ Gutterman et al.⁴⁴ demonstrated that right hemisphere lesions exert a particularly strong impact on anticipatory motor planning of the LA. In contrast, our cohort included more left hemispheric injuries. This is consistent with reports that Acute Neonatal AIS often shows left-hemisphere predominance, with a significant portion involving the left MCA territory.⁴⁵ Perforator strokes, a substantial subset of presumed perinatal AIS, also frequently involve the left hemisphere.⁴⁶ These findings are consistent with prior work demonstrating vulnerabilities in anticipatory motor planning in UCP and that the extent of impairment may depend on lesion laterality.

Hoare and colleagues have demonstrated associations between executive function and bimanual performance in children with unilateral CP.⁴⁷ Earlier, they had proposed interventions incorporating cognitive-strategy training, such as the Cognitive Orientation to daily Occupational Performance Approach (CO-OP), as well as intensive practice-based approaches such as BMT, may be relevant to enhance functional outcomes.^18,48 Hand–Arm Bimanual Intensive Therapy (HABIT), integrates cognitive–motor processes, challenging children to plan, and actively discover strategies for coordinating their hands during goal-directed activities. This aligns with BMT’s theoretical foundations emphasizing active learning.^49,50 Overall, therapies targeting motor planning may represent a promising area for future investigation and may be relevant for enhancing functional performance across the lifespan. Direct assessment and intervention of the less affected hand in pediatric and adult populations are warranted.

Limitations

This study is limited by a relatively small sample size, retrospective data collection, and differences in test availability and comparability across participants, as not all individuals completed both outcome measures. We could not examine potential differential effects on gross versus fine motor function due to our limited dataset, as not all participants completed both the BBT and NHPT. Although these assessments are commonly categorized as gross and fine motor assessments, respectively, they share overlapping components of upper limb coordination, complicating analysis. The severity of MA paresis was retrospectively assessed using clinical notes rather than a validated severity scale, which may have limited differentiation between moderate and severe categories. Additionally, the NHPT data distribution skewed towards mild paresis. A further limitation is that a floor effect was observed for the more affected limb, as some participants were unable to complete the task, resulting in clustering at the lowest scores and limiting interpretation of these data. Hand preference in children with PAIS often emerges early and may reflect compensatory use rather than typical dominance, shaped by severity of MA paresis, environmental influences, and other incompletely understood factors.

Our study focused on unilateral functional capacity, which may not predict participation in real-life activities involving diverse skills. Prior research indicates unimanual capacity and stereognosis largely explain the variance in bimanual performance in children with UCP.⁵¹ Our findings might relate to cognitive skills like knowledge application and organization, known to impact independence in self-care in UCP.⁵² Childhood stroke, with or without hemiparesis, has been linked to reduced school participation, with severity of hemiparesis only moderately correlated with home and community involvement.⁵³

The current study lacked direct measures of motor planning or other cognitive domains. The BBT and the NHPT involve relatively simple and repetitive tasks, requiring relatively little motor planning. Accordingly, we cannot determine whether observed performance differences reflect cognitive contributions, motor execution differences, or their interaction. Previous research found significant negative correlations between IQ and failure to plan an appropriate pattern of grasp but no correlation between grasp planning BBT performance for either UE in UCP.⁵

Normative reference values for pediatric dexterity have shifted over time, with reports showing a decline over recent decades.⁵⁴ As a result, comparisons against earlier norms may have skewed our cohort’s estimated performance downward. In this study, 2014 BBT norms were used for most comparisons, but three were compared against 1985 norms, which are based on faster expectations. In addition, NHPT reference data from 2015 were derived from the best of two trials,³² while our study used a single trial, potentially biasing results lower. The MCID and MDC values used for interpretation were derived from single studies of limited samples; therefore, these comparisons should be interpreted with caution.

Conclusion

Children with a history of PAIS show impairments in both limbs relative to typically developing peers and clinical thresholds. In the BBT, with a similar non-significant trend observed in the NHPT, LA motor performance worsened with increasing severity of more affected paresis. The goal of OT is participation in real-life activities that involve the interactions of myriad skills including social skills, cognitive skills, and motor planning skills as well as motor execution skills. Future work may further clarify how unilateral dexterity relates to bimanual performance in everyday activities across development and how these skills evolve over time in PAIS. In addition, examining associations between motor performance and cognitive domains may provide a more comprehensive understanding of functional outcomes in this population. Top-down approaches to rehabilitation are more likely to facilitate practice of these complex interactions. Fifty-two percent of the children in our cohort scored >2.0 SD below the normative reference on a test of gross motor dexterity when using their dominant LA. For those with severe paresis of the more affected limb, dexterity (gross and fine motor) of the less affected limb averaged >3.0 SD below the normative reference. The term less affected upper extremity is more fitting than unaffected upper extremity. Additional research can help inform intervention targeting the less affected UE as well as enhance BMT.

Author Contributions

T.J.L. conceptualized and designed the study, performed data collection and analysis, and drafted the manuscript. L.A.B. contributed to the statistical analyses. R.N.I. and L.A.B. contributed to study design, interpretation of results, and critical revision of the manuscript. All authors approved the final manuscript.

Conflicts of Interest

The authors declare no conflicts of interest related to this study.

References

1. Oskoui M, Coutinho F, Dykeman J, Jetté N, Pringsheim T. An update on the prevalence of cerebral palsy: a systematic review and meta-analysis. Developmental Medicine & Child Neurology. 2013;55(6):509–19.
2. Sadowska M, Sarecka-Hujar B, Kopyta I. Cerebral palsy: current opinions on definition, epidemiology, risk factors, classification and treatment options. Neuropsychiatric disease and treatment. 2020:1505–18.
3. Odding E, Roebroeck ME, Stam HJ. The epidemiology of cerebral palsy: incidence, impairments and risk factors. Disability and rehabilitation. 2006;28(4):183–91.
4. Krajenbrink H, Crichton A, Steenbergen B, Hoare B. The development of anticipatory action planning in children with unilateral cerebral palsy. Research in Developmental Disabilities. 2019;85:163–71.
5. Mutsaarts M, Steenbergen B, Bekkering H. Anticipatory planning deficits and task context effects in hemiparetic cerebral palsy. Experimental Brain Research. 2006;172:151–62.
6. Steenbergen B, Hulstijn W, Dortmans S. Constraints on grip selection in cerebral palsy: minimising discomfort. Experimental Brain Research. 2000;134:385–97.
7. Tomhave WA, Van Heest AE, Bagley A, James MA. Affected and contralateral hand strength and dexterity measures in children with hemiplegic cerebral palsy. The Journal of hand surgery. 2015;40(5):900–7.
8. Rich TL, Menk JS, Rudser KD, Feyma T, Gillick BT. Less-affected hand function in children with hemiparetic unilateral cerebral palsy: a comparison study with typically developing peers. Neurorehabilitation and neural repair. 2017;31(10-11):965–76.
9. Altemus KL, Almli CR. Neonatal hippocampal damage in rats: Long-term spatial memory deficits and associations with magnitude of hippocampal damage. Hippocampus. 1997;7(4):403–15.
10. van Praag H, Chun D, Black IB, Stäubli UV. Unilateral hippocampal ablation at birth causes a reduction in contralateral LTP. Brain research. 1998;795(1-2):170–8.
11. Jiang B, Hills NK, Forsyth R, Jordan LC, Slim M, Pavlakis SG, et al. Imaging predictors of neurologic outcome after pediatric arterial ischemic stroke. Stroke. 2021;52(1):152–61.
12. Craig BT, Kinney-Lang E, Hilderley AJ, Carlson HL, Kirton A. Structural connectivity of the sensorimotor network within the non-lesioned hemisphere of children with perinatal stroke. Scientific Reports. 2022;12(1):3866.
13. Crajé C, van Elk M, Beeren M, van Schie HT, Bekkering H, Steenbergen B. Compromised motor planning and motor imagery in right hemiparetic cerebral palsy. Research in Developmental Disabilities. 2010;31(6):1313–22.
14. Steenbergen B, Verrel J, Gordon AM. Motor planning in congenital hemiplegia. Disability and rehabilitation. 2007;29(1):13–23.
15. Sidiropoulos AN, Santamaria V, Gordon AM. Continuous inter-limb coordination deficits in children with unilateral spastic cerebral palsy. Clinical biomechanics. 2021;81:105250.
16. Bailey RR, Klaesner JW, Lang CE. Quantifying real-world upper-limb activity in nondisabled adults and adults with chronic stroke. Neurorehabilitation and neural repair. 2015;29(10):969–78.
17. Gordon AM. To constrain or not to constrain, and other stories of intensive upper extremity training for children with unilateral cerebral palsy. Developmental Medicine & Child Neurology. 2011;53:56–61.
18. Hoare B, Greaves S. Unimanual versus bimanual therapy in children with unilateral cerebral palsy: Same, same, but different. Journal of pediatric rehabilitation medicine. 2017;10(1):47–59.
19. Hoare BJ, Wallen MA, Thorley MN, Jackman ML, Carey LM, Imms C. Constraint-induced movement therapy in children with unilateral cerebral palsy. Cochrane Database of Systematic Reviews. 2019(4).
20. Novak I, Morgan C, Fahey M, Finch-Edmondson M, Galea C, Hines A, et al. State of the evidence traffic lights 2019: systematic review of interventions for preventing and treating children with cerebral palsy. Current neurology and neuroscience reports. 2020;20:1–21.
21. Hsu HY, Ke CW, Kuan TS, Yang HC, Tsai CL, Kuo LC. Impacts of sensation, perception, and motor abilities of the ipsilesional upper limb on hand functions in unilateral stroke: quantifications from biomechanical and functional perspectives. PM&R. 2018;10(2):146–53.
22. Hsu JE, Jones TA. Contralesional neural plasticity and functional changes in the less-affected forelimb after large and small cortical infarcts in rats. Experimental neurology. 2006;201(2):479–94.
23. Sunderland A, Bowers MP, Sluman S-M, Wilcock DJ, Ardron ME. Impaired dexterity of the ipsilateral hand after stroke and the relationship to cognitive deficit. Stroke. 1999;30(5):949–55.
24. Sainburg RL, Maenza C, Winstein C, Good D. Motor lateralization provides a foundation for predicting and treating non-paretic arm motor deficits in stroke. Progress in motor control: theories and translations. 2016:257–72.
25. Johnson BP, Westlake KP. Chronic poststroke deficits in gross and fine motor control of the ipsilesional upper limb. American journal of physical medicine & rehabilitation. 2021;100(4):345–8.
26. Kuczynski AM, Kirton A, Semrau JA, Dukelow SP. Bilateral reaching deficits after unilateral perinatal ischemic stroke: a population-based case-control study. Journal of neuroengineering and rehabilitation. 2018;15:1–13.
27. Kirton A, Deveber G. Life after perinatal stroke. Stroke. 2013;44(11):3265–71.
28. Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O’Neal L, et al. The REDCap consortium: building an international community of software platform partners. Journal of biomedical informatics. 2019;95:103208.
29. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. Journal of biomedical informatics. 2009;42(2):377–81.
30. Jongbloed-Pereboom M, Nijhuis-van der Sanden MW, Steenbergen B. Norm scores of the box and block test for children ages 3–10 years. The American Journal of Occupational Therapy. 2013;67(3):312–8.
31. Liang K-J, Chen H-L, Shieh J-Y, Wang T-N. Measurement properties of the box and block test in children with unilateral cerebral palsy. Scientific Reports. 2021;11(1):20955.
32. Wang Y-C, Bohannon RW, Kapellusch J, Garg A, Gershon RC. Dexterity as measured with the 9-Hole Peg Test (9-HPT) across the age span. Journal of Hand Therapy. 2015;28(1):53–60.
33. Wang Y-C, Magasi SR, Bohannon RW, Reuben DB, McCreath HE, Bubela DJ, et al. Assessing dexterity function: a comparison of two alternatives for the NIH Toolbox. Journal of Hand Therapy. 2011;24(4):313–21.
34. Mathiowetz V, Federman S, Wiemer D. Box and block test of manual dexterity: norms for 6–19 year olds. Canadian Journal of Occupational Therapy. 1985;52(5):241–5.
35. Mendoza-Sánchez S, Molina-Rueda F, Florencio LL, Carratalá-Tejada M, Cuesta-Gómez A. Reliability and agreement of the Nine Hole Peg Test in patients with unilateral spastic cerebral palsy. European Journal of Pediatrics. 2022;181(6):2283–90.
36. Holmström L, Vollmer B, Tedroff K, Islam M, Persson JK, Kits A, et al. Hand function in relation to brain lesions and corticomotor-projection pattern in children with unilateral cerebral palsy. Developmental Medicine & Child Neurology. 2010;52(2):145–52.
37. Groeschel S, Hertz-Pannier L, Delion M, Loustau S, Husson B, Kossorotoff M, et al. Association of transcallosal motor fibres with function of both hands after unilateral neonatal arterial ischemic stroke. Developmental Medicine & Child Neurology. 2017;59(10):1042–8.
38. Johansen-Berg H, Rushworth MF, Bogdanovic MD, Kischka U, Wimalaratna S, Matthews PM. The role of ipsilateral premotor cortex in hand movement after stroke. Proceedings of the National Academy of Sciences. 2002;99(22):14518–23.
39. Okumura A, Kato T, Kuno K, Hayakawa F, Watanabe K. MRI findings in patients with spastic cerebral palsy. II: Correlation with type of cerebral palsy. Developmental Medicine & Child Neurology. 1997;39(6):369–72.
40. Pagnozzi AM, Pannek K, Fripp J, Fiori S, Boyd RN, Rose S. Understanding the impact of bilateral brain injury in children with unilateral cerebral palsy. Human Brain Mapping. 2020;41(10):2794–807.
41. Zewdie E, Damji O, Ciechanski P, Seeger T, Kirton A. Contralesional corticomotor neurophysiology in hemiparetic children with perinatal stroke: developmental plasticity and clinical function. Neurorehabilitation and neural repair. 2017;31(3):261–71.
42. Maenza C, Wagstaff DA, Varghese R, Winstein C, Good DC, Sainburg RL. Remedial training of the less-impaired arm in chronic stroke survivors with moderate to severe upper-extremity paresis improves functional independence: a pilot study. Frontiers in Human Neuroscience. 2021;15:645714.
43. Kantak SS, Tessa J, William M. Differential effects of internal versus external focus of instruction on action planning and performance in patients with right and left hemispheric stroke. Human movement science. 2020;72:102654.
44. Gutterman J, Lee-Miller T, Friel KM, Dimitropoulou K, Gordon AM. Anticipatory motor planning and control of grasp in children with unilateral spastic cerebral palsy. Brain Sciences. 2021;11(9):1161.
45. Kirton A, Armstrong-Wells J, Chang T, Deveber G, Rivkin MJ, Hernandez M, et al. Symptomatic neonatal arterial ischemic stroke: the International Pediatric Stroke Study. Pediatrics. 2011;128(6):e1402–e10.
46. Calandrino A, Cipresso G, Battaglini M, Caruggi S, Bonato I, Massirio P, et al. Neonatal Perforator Stroke: Timing, Risk Factors, and Neurological Outcome from a Single-Center Experience. Neurology International. 2025;17(4):59.
47. Crichton A, Ditchfield M, Gwini S, Wallen M, Thorley M, Bracken J, et al. Brain magnetic resonance imaging is a predictor of bimanual performance and executive function in children with unilateral cerebral palsy. Developmental Medicine & Child Neurology. 2020;62(5):615–24.
48. Hoare B, Ditchfield M, Thorley M, Wallen M, Bracken J, Harvey A, et al. Cognition and bimanual performance in children with unilateral cerebral palsy: protocol for a multicentre, cross-sectional study. BMC neurology. 2018;18(1):63.
49. Charles J, Gordon AM. Development of hand–arm bimanual intensive training (HABIT) for improving bimanual coordination in children with hemiplegic cerebral palsy. Developmental medicine and child neurology. 2006;48(11):931–6.
50. Gordon AM, Schneider JA, Chinnan A, Charles JR. Efficacy of a hand–arm bimanual intensive therapy (HABIT) in children with hemiplegic cerebral palsy: a randomized control trial. Developmental Medicine & Child Neurology. 2007;49(11):830–8.
51. Sakzewski L, Ziviani J, Boyd R. The relationship between unimanual capacity and bimanual performance in children with congenital hemiplegia. Developmental Medicine & Child Neurology. 2010;52(9):811–6.
52. Russo RN, Skuza PP, Sandelance M, Flett P. Upper limb impairments, process skills, and outcome in children with unilateral cerebral palsy. Developmental Medicine & Child Neurology. 2019;61(9):1080–6.
53. Simon-Martinez C, Kamal S, Frickmann F, Steiner L, Slavova N, Everts R, et al. Participation after childhood stroke: Is there a relationship with lesion size, motor function and manual ability? European journal of paediatric neurology. 2021;35:16–26.
54. Waber DP, De Moor C, Forbes PW, Almli CR, Botteron KN, Leonard G, et al. The NIH MRI study of normal brain development: performance of a population based sample of healthy children aged 6 to 18 years on a neuropsychological battery. Journal of the International Neuropsychological Society. 2007;13(5):729–46.

Figures

Figure 1. Distribution of Z-scores relative to normative data: (A, left) Box & Blocks Test (BBT) for the more affected (MA) and less affected (LA), (B, right) Nine-hole Peg Test (NHPT) for the MA and LA. Each participant is represented by two points: one for the MA (gray dot) and one for the LA (gray triangle). Z-scores are shown on the Y-axis. The plots illustrate participant-level performance relative to typically developing peers. Participants are ordered according to the MA Z-score, from highest (best performing) to lowest (poorest performing), within each test. Of the five individuals whose LA score fell more than three standard deviations below the normative reference, all received the floor score (Z = -50) for the MA. In total, nine participants received the floor score for their MA because they were unable or presumed unable to complete the test with that hand.

Figure 2. Group mean Z-scores by hemiparesis severity: (A) BBT for the more affected hand, (B) BBT for the less affected hand. Bar graphs show mean Z-scores for participants with mild, moderate, and severe hemiparesis. The asterisk (*) indicates statistical significance, showing that the mild group performed better than both moderate and severe groups (p < 0.05). Error bars represent standard deviation. Performance did not differ significantly between the moderate and severe groups for either hand.

Figure 3. Group mean NHPT Z-scores by hemiparesis severity: (A) NHPT for the more affected (MA) hand, (B) NHPT for the less affected (LA) hand. Bar graphs show mean Z-scores for participants with mild and moderate hemiparesis. The asterisk (*) indicates statistical significance, showing that the mild group performed better than the moderate group (p < 0.05) for the MA. Error bars represent standard deviation (SD).

Figure 4. Motor performance relative to clinical thresholds. (A) BBT scores relative to minimal clinically important difference (MCID); (B) NHPT scores relative to minimal detectable change (MDC). Negative values indicate better performance than the threshold; positive values indicate worse performance. Boxes represent the interquartile range, the line indicates the median, and whiskers show the full range excluding outliers. Most participants performed worse than the clinical thresholds, with particularly large deviations for the MA compared to the LA.