Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Apr 9.
Published in final edited form as: J Mot Behav. 2026 Jan 2;58(1):70–80. doi: 10.1080/00222895.2025.2605977

Effects of symmetry and age on bilateral upper-limb coordination in children with unilateral cerebral palsy

Cassandra Kemmel-Bartletti a, Karin Goodfriend b, Meghann Sytsma b, Sheila Schindler-Ivens c, Samuel T Nemanich a,*
PMCID: PMC13061373  NIHMSID: NIHMS2144501  PMID: 41482733

Abstract

Bilateral movements are important for daily function and are impaired in children with unilateral cerebral palsy (UCP). Prior work has established that coordination is dependent on symmetry demands between the limbs. We examined how symmetry impacted bilateral coordination, and as an exploratory aim, compared age-related differences in upper-limb (UL) coordination in children with UCP to children with typical development (TD). Using an instrumented cycling device, participants performed bilateral UL cycling in symmetric or asymmetric patterns. Metrics of inter-limb (phase error) coordination were compared between groups (TD or UCP) and with respect to age (young: 7–9 years; older: 13–15 years). Children with UCP had poorer inter-limb coordination in both symmetric (50.88°) and asymmetric (93.22°) bilateral tasks compared to TD children. Younger children, regardless of group, had poorer inter-limb coordination for the asymmetric (32.54°) bilateral task only. Regression modelling revealed UCP, age, and unilateral coordination ability were significantly associated with asymmetric bilateral coordination. The results indicate that bilateral coordination skills reflect atypical processes related to UCP and typical age-related motor development, particularly for more complex asymmetric tasks. Altogether, this research confirms children with UCP have impaired bilateral coordination and highlights potential for future work to study developmental changes in this population.

Keywords: bilateral coordination, cerebral palsy, motor development, age

Introduction

Children with unilateral cerebral palsy (UCP) experience upper limb (UL) motor impairments that significantly limit function and independence at every stage of development (Eliasson & Gordon, 2000; Eliasson et al., 1991; Gordon et al., 2013; Hung et al., 2004). A considerable body of work has focused on understanding the mechanisms underlying disordered motor control in children with UCP. Movement coordination, defined as producing appropriate timing and amplitude of a single (unilateral) or both (bilateral) ULs, is critical for motor control. Bilateral coordination is particularly important for performing daily activities (Wang et al., 2021). Children with UCP exhibit impaired coordination during bilateral tasks involving force production (Islam et al., 2011; Smits-Engelsman et al., 2011), opening a drawer (Hung et al., 2004), and marching (Sidiropoulos et al., 2021).

Therefore, the primary objective of this study was to evaluate bilateral UL coordination in children and adolescents with UCP and typically-developing (TD) children. Since bilateral skill may be dependent on the (a)symmetry of bilateral movements, we examined both symmetric and asymmetric bilateral tasks (Barral et al., 2006; Kantak et al., 2017). We hypothesized that children with UCP would demonstrate impaired coordination relative to their typically developing peers. In addition, we hypothesized to observe greater impairments in asymmetric than symmetric coordination, given the greater demands on sensorimotor processing required for asymmetric movements. Finally, we considered the ability of the neuromuscular system to control multiple joints within and between ULs and how this could affect unilateral and bilateral coordination. Specifically, we examined whether unilateral coordination alone could account for changes in bilateral coordination among children with UCP and typically-developing children.

A secondary aim of this study was to examine age-related differences in bilateral UL coordination. A well-established aspect of typical motor development is gradual improvement in bilateral coordination across childhood and adolescence (Barral et al., 2006; de Boer et al., 2012; Muetzel et al., 2008; Rudisch et al., 2018). For example, compared to children younger than 6 years of age, older children and adolescents demonstrate greater spatial and temporal coupling of the limbs, a skill strengthened by experience-dependent plasticity and motor practice in adolescence (de Boer et al., 2012; Rudisch et al., 2018). In children with UCP, the location, size, and timing of the brain injury collectively lead to lower motor performance at all ages relative to children with typical development. However, age-related improvements in gross motor (Jahnsen et al., 2025) and hand function (Eliasson et al., 2006; Klingels et al., 2012) have been reported among children with UCP, changes that could be attributed to injury- and experience-dependent neuroplasticity (Kirton & Deveber, 2013; Staudt, 2010). However, the relationship between age and bilateral UL coordination in children with UCP, compared to a typically-developing group of children, is not well understood. Elucidating this relationship is important because, as highlighted in a recent scoping review of bimanual motor function in UCP (Gascon et al., 2025), understanding age-related differences in neuromotor development in children with UCP is important to distinguish therapeutic gains from maturation. Such knowledge may also provide prognostic information that can help children and families affected by UCP to anticipate developmental trajectories. Although differences in bilateral coordination are observed across several studies, it remains unclear whether the ability to coordinate joint motion between the ULs improves with age in this population. We hypothesized that older children will show better bilateral coordination, consistent with ongoing motor development across childhood.

Materials and Methods

Participants

Participants included 10 individuals with UCP (4 female) and 10 individuals with typical development (TD) (4 female), aged 7–15 years. Within each group, participants were categorized as young (7–9 years) or old (13–15 years) based on a natural gap in the sample, with no observations available for children aged 10–12. This categorization resulted in 6 young and 4 old participants in each group (UCP/TD).

Participants with UCP were recruited from outpatient referrals from the Department of Physical Medicine and Rehabilitation at Children’s Hospital of Wisconsin. They had a history of brain injury occurring before the first year of age. All were ambulatory, with Gross Motor Function Classification Scale (GMFCS) scores of I or II and Manual Ability Classification Scale (MACS) scores from I to III. Four had right-sided and 6 had left-sided hemiparesis. All were free of neurological conditions other than UCP and communication impairments that could affect their understanding of the study procedures. Additionally, none had received botulinum toxin injections or undergone surgical interventions affecting the UL within the past three months.

TD participants were recruited from the community through advertisements on university websites and social media accounts. They had little to no mirror movements, as assessed by the Woods and Teuber Mirror Movements assessment (Woods & Teuber, 1978). Groups were directly age- and sex-matched, and did not have different levels of physical activity (p ≥ .155); physical activity was assessed by an unstandardized, parent-report demographics item of the number of hours the child participated in moderate and vigorous physical activity per week. Scores on clinical tests of UL function—Box and Blocks (Liang et al., 2021), Speed Stacks (Hung & Gordon, 2013), Mirror Movements Assessment (Woods & Teuber, 1978), and ABILHAND-Kids (Arnould et al., 2004)—were significantly different between groups, with the UCP group showing lower performance across all assessments. See Table 1 for details on participant characteristics.

Table 1.

Participant Characteristics

Characteristic UCP (n=10) TD (n=10) p value
Young (n=6) Old (n=4) Young (n=6) Old (n=4)
Age, yr, mean (SD) 11.00 (3.20) 10.80 (3.26) .913
8.20 (0.84) 14.67 (0.58) 8.33 (0.82) 14.50 (0.58)
Side of hemiparesis/Non-dominant, n
 Right 2 2 1 0
 Left 4 2 5 4
Sex, n
 Male 4 2 4 2
 Female 2 2 2 2
GMFCS, n
 I 3 2 ---
 II 3 1 ---
MACS, n
 I 1 2 ---
 II 5 1 ---
 III 0 1 ---
Lesion Type, n
 Perinatal stroke 4 4 ---
 Intraventricular hemorrhage 2 0 ---
Box and Blocks, blocks (MA/ND) 22.75 (10.59) 55.30 (12.37) <.001
26.36 (7.65) 17.33 (13.17) 50.11 (8.63) 63.09 (14.15)
Speed Stacks (sec) 19.68 (7.61) 8.60 (2.14) .005
20.81 (8.42) 16.39 (5.71) 9.31 (2.50) 7.54 (0.94)
Mirror Movements Score (MA/ND) 0.83
(Range = 0 – 2.67)		 0.00
(Range = 0 – 0.67)		 .077
1.17
(Range = 0 – 2.67)		 0.75
(Range 0 – 2)		 0.11
(Range 0 – 0.33)		 0.17
(Range 0 – 0.67)
ABILHAND-Kids, mean (SD) 32.40 (6.98) 41.11 (1.69) .006
29.50 (7.23) 36.75 (4.11) 40.40 (2.08) 42.00 (0.00)
Vigorous Physical Activity, n
 Less than 1 hour/week 0 0 1 1
 1–4 hours/week 3 3 1 1
 5–10 hours/week 2 1 4 2
 10 or more hours/week 1 0 0 0
Moderate Physical Activity, n
 Less than 1 hour/week 0 0 0 0
 1–4 hours/week 2 2 1 2
 5–10 hours/week 3 1 4 1
 10 or more hours/week 1 1 1 1
Open in a new tab

n: Number; GMFCS: Gross motor function classification scale; MACS: Manual ability classification scale; MA: more-affected side, ND: non-dominant

Equipment

A custom split-crank cycling device (Figure 1A) was used to assess bilateral and unilateral coordination of the ULs (Cleland et al., 2019; Nemanich & Schindler-Ivens, 2025). The device comprised a left and right handle, each connected to a crankshaft via a crank arm. Unlike a conventional cycling device, the shaft was split into left and right halves, with no mechanical connection between the handles. As a result, movement of one handle had no mechanical effect on the other. On certain trials (see Procedures), a detachable coupler was inserted to provide a mechanical connection between the handles. Each crankshaft was connected to a rotary optical encoder through a chain and sprocket assembly, allowing measurement of instantaneous position and velocity. A 16-bit analog-to-digital converter and data acquisition software (Micro 1401 mk II, Spike2, Cambridge Electronic Designs, UK) housed on a laptop computer sampled these signals at 2.4 kHz.

Figure 1.

Figure 1.

Open in a new tab

Cycling device (A) and experimental protocol (B).

Procedures

All study procedures were approved by the Marquette University Institutional Review Board (Approval #4238) and complied with federal laws and guidelines. Informed consent was obtained from two parents or caregivers before any study procedures began. Additionally, participants aged 15 and older provided their own written consent, while those under 15 gave verbal assent.

The experimental procedures are summarized in Figure 1B. Participants were seated in front of a table that held the cycling device. Their feet rested on the floor or a stool, and the table height was adjusted so the center of rotation of the crankshaft was at the same height as their shoulders. Before data collection, participants performed two bouts of bilateral cycling with a coupler inserted between the left and right crankshafts. The coupler created a mechanical connection between the handles, converting the device into a conventional cycling system with a single crankshaft. The purpose of these trials was to familiarize participants with symmetric and asymmetric cycling. In one bout, the coupler was inserted so the crankshafts were aligned in the same position (in-phase), and in the other, they were positioned 180° out-of-phase.

After familiarization, the coupler was removed, and data were recorded while participants performed four different cycling movements: two unilateral and two bilateral. The unilateral movements involved cycling with the more-affected arm (non-dominant arm for TD) or the less-affected arm (dominant arm for TD). Bilateral movements involved cycling with the arms moving in-phase (symmetric) or 180° out-of-phase (asymmetric). One 30-second trial of each task was performed, totalling two minutes of cycling. Rest was provided between trials, and the order in which tasks were introduced was counterbalanced. A metronome set to 45 beats per minute provided an external pacing cue. Moreover, before each trial of data collection, participants were shown a video of an actor performing the desired cycling movement. The video provided verbal instructions about how to perform the cycling (“Pedal both handlebars with your arms together, so that each side is pedalling at the same time” for the symmetric condition, “Pedal both handlebars with your hands apart, like you are pedalling a bike”, for asymmetric condition, and “Grab the left/right pedal with your left/right hand. Pedal using just your left/right side.”). For all conditions, the video provided an additional verbal cue to, “Please cycle at a smooth and even pace while trying to match the beat of the metronome.”

Data Processing and Dependent Variables

Digitized voltage-time signals from the optical encoders were converted to MATLAB data files using built-in and custom scripts (The MathWorks Inc., Natick, MA). Data were transformed into position-time (degrees) and velocity-time (degrees/sec) series, visually inspected, and low-pass filtered at 10 Hz with a 4th-order Butterworth filter. For unilateral task conditions, the coefficient of variation of velocity (COVvel) for each limb was calculated as a measure of unilateral coordination.

Dependent variables related to bilateral coordination were computed as previously described (Cleland et al., 2019; Nemanich & Schindler-Ivens, 2025). In brief, we calculated the continuous relative phase of the limbs (CRP) as the difference in position between the left and right crankshaft as a function of time (Nemanich & Schindler-Ivens, 2025). We then computed phase error (E) as the difference between CRP and 0° for the symmetric condition and the difference between CRP and 180° for the asymmetric condition.

E=‐CRP360*360+CRP;symmetric‐CRP‐180360*360+CRP‐180;asymmetric

Modular arithmetic (⌊⌋) was used to account for wrap-around effects (Jones, 1964). A modulus of 360° (i.e. the denominator within the brackets) was used because the crankshaft position “wraps around” and resets every full revolution. With this approach, E could assume any value within ±180°, where the sign indicated the leading limb. A negative value meant the more-affected (nondominant) limb was leading, while a positive value meant the less-affected (dominant) limb was leading. When the limbs were ideally phased, E was zero. The mean (μE) and standard deviation (σE) of the absolute value of phase error were calculated as measures of bilateral coordination during the bilateral cycling conditions.

Statistical Analyses

For bilateral cycling, two-way ANOVAs tested the effects of group (UCP/TD) and task (symmetric or asymmetric), with a covariate of age (young vs. old), within each task on μE and σE. To determine a relationship between general coordination (measured as unilateral cycling performance) and bilateral coordination, a series of Pearson correlations, with a Bonferroni correction for multiple comparisons, were performed. These correlations were between bilateral asymmetric cycling μE, unilateral cycling variability of the more-affected/non-dominant and less-affected/dominant limbs, and age (in years). We then determined, using linear regression, whether bilateral coordination was more dependent on group, age, or on general coordination. Predictor variables were entered into the model in the following order: group (UCP or TD), age (in years), and unilateral cycling variability of more-affected/non-dominant limb and unilateral cycling of less-affected/dominant limb. Regression values were evaluated by the significance of individual predictors and R2 changes. Effects from ANOVAs and regression models were considered significant at p ≤ 0.05. The magnitude of differences between groups, ULs, or conditions were also reported to estimate effect sizes from the small sample size. All analyses were conducted in SPSS (v. 29).

Results

Bilateral Coordination

Representative examples of cycling performance during bilateral symmetric and asymmetric conditions are shown in Figure 2 (upper panel) for a TD and UCP child. Performance in the child with UCP was poorer than the TD child across both tasks (compare gold to blue traces). For both children, performance declined with asymmetric compared to symmetric cycling. The mean cycling speed of participants did not significantly differ from the desired rate of 45 rpm for both asymmetric (mean = 39.4 rpm, p = 0.098) and symmetric cycling (mean =44.2 rpm, p = 0.743).

Figure 2.

Figure 2.

Open in a new tab

Performance on bilateral cycling tasks in two 9 year old children. Upper panels show individual traces of cycling position, velocity, and phase error (E) in a representative participant with unilateral cerebral palsy (UCP) and typical development (TD). Individual and group (UCP/TD) mean phase error (μE, A) and standard deviation of phase error (σE, B) across symmetric and asymmetric conditions. Data are stratified by age group (young/old)

UCP vs. TD

Group analysis supports the representative data. As shown in the bottom panel of Figure 2, μE and σE were significantly higher in the UCP group (mean = 72.05°, 95% CI [57.10°, 87.00°]; 103.69°, 95% CI [84.06°, 123.32°]) than in the TD group (36.36°, 95% CI [22.04°, 50.68°]; 62.58°, 95% CI [38.55°, 86.60°]). Regarding task effects, μE and σE were significantly higher during the asymmetric (73.88°, 95% CI [56.83°, 90.95°]; 83.13°, 95% CI [66.03°, 100.24°]) than the symmetric task for all participants (34.52°, 95% CI [24.24°, 44.81°]; 42.74°, 95% CI [30.16°, 55.33°]). Statistical values and ANOVA output are provided in Table 2.

Table 2.

Factorial ANOVA Analyses of Group (UCP/TD) Differences in Bilateral Task with Covariate of Age (Young/Old)

μE
σE
Variables F df p F df p
Group (UCP/TD) 17.70 1,17 <.001 19.95 1,17 <.001
Task (ASYM/SYM) 22.76 1,17 <.001 29.27 1,17 <.001
Age (Young/Old) 3.69 1,17 .072 3.51 1,17 .078
Group * Task 0.31 1,17 .587 0.02 1,17 .899
Age * Task 6.20 1,17 .023 7.93 1,17 .012
Open in a new tab

Young vs. Old

The differences in bilateral phasing error in young and old children are shown in Figure 2. An exploratory analysis of age effects as a covariate showed that there was an interaction of task and age. Younger children had significantly greater μE and σE than older children during the asymmetric task (mean difference, MD = 30.28°, η2 = .264; MD = 31.26°, η2 = .318), but not during the symmetric task (MD = 2.97°; MD = 3.41°), regardless of group. Statistical values and ANOVA output are provided in Table 2.

Unilateral Coordination

The relationship between unilateral variation in velocity and bilateral mean phase error is shown as scatterplots in Figure 3. There were significant, moderate to strong (r = 0.60–0.80, p < .006) linear relationships between the COVvel of the more-affected/non-dominant UL and μE (Figure 3A and C).There was a significant, moderate (r = 0.47, p = .04) linear relationship for the less-affected/dominant UL and asymmetric μE, but a non-significant, weak (r = 0.32, p = .17) relationship with symmetric μE (Figure 3B and D).

Figure 3.

Figure 3.

Open in a new tab

Relationship between unilateral (COVvel) and bilateral coordination (μE) for symmetric and asymmetric cycling conditions. Pearson r correlation coefficients provided for strength of correlations.

Linear regression

Group (UCP/TD) and age (in years) were the strongest predictors of asymmetric bilateral phase error (μE), indicated by the greatest R2 changes. Although not as strong, unilateral COVvel of the more-affected/non-dominant limb was also a significant predictor. In contrast, symmetric bilateral phase error was only significantly predicted by group (See Table 3).

Table 3.

Linear Regression Analyses to Predict Asymmetric and Symmetric μE

Asymmetric
Symmetric
Predictors Condition Index b t p R2 Change b t p R2 Change
 Group (UCP/TD) 3.97 −25.26 −2.70 .017 .296* −26.58 −3.82 .002 .584*
 Age 6.77 −4.74 −2.61 .020 .296* −0.05 −0.04 .969 .025
 Unilateral More-Affected/Non-Dom 10.34 1.00 3.90 .001 .161* 0.36 1.91 .076 .080
 Unilateral Less-Affected/Dom 23.06 −0.52 −1.81 .091 .044 −0.08 −0.39 .700 .003
Open in a new tab

Overall model for asymmetric μE: F(4,15) = 14.78, p < .001, R2 = .80; symmetric μE: F(4,15) = 8.42, p < .001, R2 = .69

*

Significant R2 change

Discussion

In this study, we examined UL coordination in children with UCP and TD children. We found that children with UCP showed poorer bilateral coordination, measured by mean and standard deviation of phase error, compared to their TD peers. We also observed age-related improvements in asymmetric but not symmetric bilateral coordination. Coordination was significantly explained by UCP, age, and overall coordination ability. These findings support and extend previous research of UL bilateral coordination impairments in UCP continues to develop as the nervous system matures, even in children with UCP, and have important implications for interpreting clinical studies involving cohorts of children with UCP with wide age ranges.

Bilateral coordination is impaired in children with UCP

Children with UCP demonstrated poorer bilateral UL coordination during cycling, as indicated by higher values of μE and σE. These findings are consistent with the broader understanding that children with UCP experience impaired UL motor control during bilateral movements (Eliasson & Gordon, 2000; Eliasson et al., 1991; Hung et al., 2004; Smits-Engelsman et al., 2011). Our results further demonstrate a clear effect of UL symmetry on bilateral coordination; both groups of children performed better during symmetric cycling than during asymmetric cycling. Th A recent study similarly found that asymmetric movements are more challenging for children with UCP (Sidiropoulos et al., 2021), and highlight the importance of task constraints and demands on bilateral performance. Therefore, there appears to be consistent motor behavior for all children and adolescents corresponding to better performance on symmetric bilateral tasks, which aligns with previous studies showing a developmental preference for symmetric movements and reduced ability to produce asymmetric patterns in children (de Boer et al., 2012; Fagard & Corroyer, 2003; Ringenbach & Amazeen, 2005; Robertson, 2001). However, the neural control mechanisms underlying these behaviors likely differ between children with UCP and TD. To achieve symmetric control, a shared motor command produces synchronized activation of homologous muscle groups of the same effector (e.g. left shoulder flexion-right shoulder flexion). Alternatively, to achieve asymmetric control, competing motor commands are needed to activate non-homologous muscle groups (e.g. left shoulder flexion-right shoulder extension), leading to less stable coordination (Swinnen & Wenderoth, 2004).

Mirror movements, or mirroring, are defined as involuntary but similar movement patterns produced during unilateral movement and might further explain the preference for symmetric coordination. Mirroring is prevalent in young children but declines in frequency and intensity with age. However, in children with UCP, UL mirroring movements are more common and thought to reflect differences in corticospinal plasticity after early unilateral brain injury (Klingels et al., 2016; Rich et al., 2020; Riddell et al., 2019; Zielinski et al., 2017). Although typically considered a motor deficit, mirroring may have enhanced symmetric bilateral coordination by enabling simultaneous activation of homologous muscles. Likewise, this same mirroring could impair performance during asymmetric coordination, where different patterns of motor output are required for each limb. However, analysis of mirroring data showed no evidence that mirroring aiding symmetric coordination, but rather mirroring was associated with poorer symmetric and asymmetric coordination. Our assessment of mirroring predominantly involved individual finger movement and was not directly related to the movements required for bilateral cycling. Thus, additional studies comparing mirroring to functional bimanual tasks involving the same UL segments could offer further insight into how mirroring relates to coordination.

Another possible explanation is that our task differed from the typical roles of the limbs used during everyday bimanual activities. For example, in a drawer-opening task, studies have shown that when the more-affected limb is assigned the more challenging role (e.g., opening the drawer), the coordination of the less-affected limb performing a simpler task (e.g., reaching into the drawer) improves compared to when the roles are reversed (Hung et al., 2004). This suggests that task-specific demands influence interlimb coordination. In the present study, assigning the same role to each limb during cycling may have led to comparable bilateral coordination. Future studies could further test this idea by systematically manipulating the demands placed on each limb during cycling—for example, requiring one arm to cycle at twice the speed of the other. Such experiments could reveal whether asymmetries emerge under unequal task demands, thereby advancing our understanding of the roles each UL plays in movement coordination in children.

Bilateral coordination is greater in older children

Based on evidence that bilateral motor coordination does not fully develop until late adolescence or early adulthood (Barral et al., 2006; Leversen et al., 2012; Marion et al., 2003; Nemanich & Schindler-Ivens, 2025), we initially hypothesized that older children would exhibit better bilateral coordination than younger children. Our results of an interaction between age and cycling condition supported this hypothesis: there was a robust relationship between age and asymmetric, but not symmetric, bilateral coordination. Computational models of human bilateral coordination suggest that in-phase (symmetric) and anti-phase (asymmetric) movements are the most stable modes of coordination (Haken et al., 1985; Robertson, 2001). While both the symmetric and asymmetric cycling tasks required independent control of each UL, coordination was more stable in the symmetric (in-phase) condition, as indicated by lower phasing errors. Prior research has shown that young children prefer symmetric, in-phase rather than asymmetric, anti-phase bilateral motor patterns, likely due to underdeveloped sensorimotor integration necessary for performing more complex asymmetric movements (Robertson, 2001). This preference may help explain the absence of age effects for bilateral coordination in the symmetric task.

While this age-related improvement in bilateral coordination was expected in TD children, a novel finding of our study was that these age-related differences in UL coordination were also observed in children with UCP. This suggests a common developmental trajectory for motor coordination in children, where inter-limb coordination improves with age, regardless of motor impairments. Previous research on the relationship between age and UL motor function in children with UCP has been limited and somewhat conflicting. A prospective cohort study found that bilateral hand function, as measured by the Assisting Hand Assessment (AHA), reached a stable plateau by 7 years of age (Eliasson et al., 2006). Furthermore, some studies have proposed that unimanual capacity, rather than age, may be a stronger predictor of bimanual coordination (Chaleat-Valayer et al., 2015; Sakzewski et al., 2010). These results led to a conclusion that motor impairments are relatively static in children with UCP. In contrast, other studies suggested that hand dexterity improved with age in children with UCP, particularly in those with mild to moderate impairments (Holmefur et al., 2010; Klingels et al., 2018). Our results are aligned with these studies that together offer an alternative conclusion that bilateral coordination may change with age-related development and experience-dependent plasticity, particularly for complex inter-limb tasks.

A key difference between the present study and prior work is the motor demands of the tasks used to assess function. Most previous studies focused on tasks requiring coordination and manual dexterity to grasp and manipulate objects. In contrast, the cycling task in our study placed minimal demands on fine motor skills. The distinct neuroanatomical substrates involved in UL coordination and dexterity may help explain these differences. Specifically, hand dexterity is strongly associated with the integrity of the descending corticospinal system (Duque et al., 2003; Fuelscher et al., 2021), whereas motor coordination is linked to cortico-cerebellar circuits that integrate sensorimotor information to execute and update motor plans (Diedrichsen et al., 2007). Furthermore, gross motor control of the limb can be regulated by alternative descending motor pathways, such as the reticulospinal tract (McPherson et al., 2018). We speculate that whereas hand dexterity is substantially affected by decreased corticospinal input in UCP, UL coordination (and its improvement with age) might be preserved though alternate descending motor pathways and relatively intact cortico-cerebellar loops. However, such an explanation would require neurophysiological testing to confirm and thus remains a potential area of future research.

Contribution of unilateral coordination to bilateral coordination

Correlations between unilateral and bilateral coordination indicated there were stronger relationships between bilateral μE and COVvel in the more-affected/non-dominant UL than the less-affected/dominant UL (Figure 3). We further explored these relationships using linear regression to determine if unilateral coordination potentially confounded the differences in bilateral coordination observed between children with UCP and TD and between age. After accounting for UCP and age, unilateral coordination of the more-affected/non-dominant UL explained an additional 16% and 8% of the variation in asymmetric and symmetric μE respectively, whereas the less-affected/dominant UL explained 4.4% and 0.3% of the variation in asymmetric and symmetric μE respectively (Table 3). However, it is worth noting these variances were still smaller than the variance explained by UCP/TD and age (about 60%). In children with UCP, unilateral impairments correlate with bilateral coordination (Chaleat-Valayer et al., 2015; Sakzewski et al., 2010) and variations in performance of each UL bilateral coordination have been observed in discrete bimanual coordination tasks (Hung et al., 2004; Rudisch et al., 2016). Since our task was continuous, with each UL performing the same motor pattern, the role of each UL is comparable. Thus, our results suggest impairments of cyclic continuous bilateral movements in UCP are related more to unilateral coordination of the more-affected UL than the less-affected limb. The lack of collinearity between group (UCP/TD), age, and unilateral coordination indicates these variables each explain unique aspects of bilateral coordination ability, highlighting the complex nature of a skill such as bilateral UL cycling.

Implications for future research and neurorehabilitation

An important and challenging consideration for future research is how to evaluate bilateral movement coordination, particularly in terms of spatiotemporal constraints, movement complexity, and the number of motor effectors involved. This highlights the need for objective and quantifiable assessments that can distinguish movements of each UL independently. While various laboratory and clinical assessments can achieve this goal and measure different aspects of bilateral coordination (Gascon et al., 2025), they may lack clinical applicability. Therefore, future studies should select assessments that provide specificity about movement properties while remaining feasible for clinical use.

Our results also suggest that age is a critical factor to consider when developing personalized neurorehabilitation approaches. Age stratification should be considered in randomized trials to account for heterogeneity in outcomes, particularly as they relate to neurodevelopment and experience.

Limitations

These conclusions should be considered in the context of several limitations. First, the statistically significant results and moderate effect sizes may not generalize to a larger and more heterogeneous population of children with varied motor skills. Second, the cross-sectional design of the study limits our ability to draw conclusions about developmental changes in bilateral coordination within individual children over time. Finally, our findings are specific to continuous, rhythmic bilateral movements of the ULs and may not generalize to the more complex and varied bilateral tasks required in daily life.

Conclusions

Using a novel bilateral cycling task, we found that bilateral coordination is impaired in children with UCP but was improved in older children, suggesting age-related development of bilateral motor control even in the presence of motor impairments. Our findings also show that bilateral coordination is influenced by the symmetry of the task and the functional role of each limb, wherein the more-affected UL in children with UCP could coordinate with the less-affected limb during symmetric bilateral movements but was less coordinated when moving in a more complex asymmetric pattern. These results highlight the importance of considering age, task symmetry, and limb demands when assessing motor function. Continued research to track the progression of bilateral motor skills in children with UCP will be essential for evaluating the impact of neurorehabilitation interventions and understanding motor development over time.

Acknowledgements

The authors thank the Marquette University Occupational Therapy Doctoral students for their assistance with data collection and analysis. The authors acknowledge the use of artificial intelligence (AI) to generate the graphic of the pedalling device in Figure 1A. AI was used solely to create an illustrative image from real photograph taken in the laboratory of the device. This work was supported by the National Pediatric Rehabilitation Resource Center (C-PROGRESS), the National Institutes of Health, including the Eunice Kennedy Shriver National Institute of Child Health and Human Development, the National Institute of Neurological Disorders and Stroke, the National Institute of Biomedical Imaging and Bioengineering, and by the American Academy of Cerebral Palsy and Developmental Medicine.

Footnotes

Declaration of Interest

The authors report there are no competing interests to declare.

References

  1. Arnould C, Penta M, Renders A, & Thonnard JL (2004). ABILHAND-Kids: a measure of manual ability in children with cerebral palsy. Neurology, 63(6), 1045–1052. 10.1212/01.wnl.0000138423.77640.37 [DOI] [PubMed] [Google Scholar]
  2. Barral J, Debu B, & Rival C (2006). Developmental changes in unimanual and bimanual aiming movements. Developmental Neuropsychology, 29(3), 415–429. 10.1207/s15326942dn2903_2 [DOI] [PubMed] [Google Scholar]
  3. Chaleat-Valayer E, Bard-Pondarre R, Ganne C, Roumenoff F, Combey A, & Bernard JC (2015). Relation between unimanual capacities and bimanual performance in hemiplegic cerebral-palsied children: impact of synkinesis. European Journal of Paediatric Neurology, 19(2), 193–201. 10.1016/j.ejpn.2014.12.002 [DOI] [PubMed] [Google Scholar]
  4. Cleland BT, Gelting T, Arand B, Struhar J, & Schindler-Ivens S (2019). Impaired interlimb coordination is related to asymmetries during pedaling after stroke. Clinical Neurophysiology, 130(9), 1474–1487. 10.1016/j.clinph.2019.05.025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. de Boer BJ, Peper CE, & Beek PJ (2012). Development of Temporal and Spatial Bimanual Coordination During Childhood. Motor Control, 16(4), 537–559. 10.1123/mcj.16.4.537 [DOI] [PubMed] [Google Scholar]
  6. Diedrichsen J, Criscimagna-Hemminger SE, & Shadmehr R (2007). Dissociating timing and coordination as functions of the cerebellum. Journal of Neuroscience, 27(23), 6291–6301. 10.1523/Jneurosci.0061-07.2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Duque J, Thonnard JL, Vandermeeren Y, Sébire G, Cosnard G, & Olivier E (2003). Correlation between impaired dexterity and corticospinal tract dysgenesis in congenital hemiplegia. Brain, 126, 732–747. 10.1093/brain/awg069 [DOI] [PubMed] [Google Scholar]
  8. Eliasson AC, Forssberg H, Hung YC, & Gordon AM (2006). Development of hand function and precision grip control in individuals with cerebral palsy: A 13-year follow-up study. Pediatrics, 118(4), E1226–E1236. 10.1542/peds.2005-2768 [DOI] [PubMed] [Google Scholar]
  9. Eliasson AC, & Gordon AM (2000). Impaired force coordination during object release in children with hemiplegic cerebral palsy. Developmental Medicine and Child Neurology, 42(4), 228–234. 10.1017/S0012162200000396 [DOI] [PubMed] [Google Scholar]
  10. Eliasson AC, Gordon AM, & Forssberg H (1991). Basic co-ordination of manipulative forces of children with cerebral palsy. Developmental Medicine and Child Neurology, 33(8), 661–670. 10.1111/j.1469-8749.1991.tb14943.x [DOI] [PubMed] [Google Scholar]
  11. Fagard J, & Corroyer D (2003). Using a continuous index of laterality to determine how laterality is related to interhemispheric transfer and bimanual coordination in children. Developmental Psychobiology, 43(1), 44–56. 10.1002/dev.10117 [DOI] [PubMed] [Google Scholar]
  12. Fuelscher I, Hyde C, Efron D, & Silk TJ (2021). Manual dexterity in late childhood is associated with maturation of the corticospinal tract. Neuroimage, 226. 10.1016/j.neuroimage.2020.117583 [DOI] [PubMed] [Google Scholar]
  13. Gascon HD, Maldonado Numata SA, Emond S, Nemanich S, & Robert MT (2025). Age-related differences in behavioral outcomes of bimanual functional motor tasks in children and adolescents with cerebral palsy: a scoping review. Disability and Rehabilitation, 47(15), 3781–3793. 10.1080/09638288.2024.2435528 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gordon AM, Bleyenheuft Y, & Steenbergen B (2013). Pathophysiology of impaired hand function in children with unilateral cerebral palsy. Developmental Medicine and Child Neurology, 55 Suppl 4, 32–37. 10.1111/dmcn.12304 [DOI] [PubMed] [Google Scholar]
  15. Haken H, Kelso JA, & Bunz H (1985). A theoretical model of phase transitions in human hand movements. Biological Cybernetics, 51(5), 347–356. 10.1007/BF00336922 [DOI] [PubMed] [Google Scholar]
  16. Holmefur M, Krumlinde-Sundholm L, Bergstrom J, & Eliasson AC (2010). Longitudinal development of hand function in children with unilateral cerebral palsy. Developmental Medicine and Child Neurology, 52(4), 352–357. 10.1111/j.1469-8749.2009.03364.x [DOI] [PubMed] [Google Scholar]
  17. Hung YC, Charles J, & Gordon AM (2004). Bimanual coordination during a goal-directed task in children with hemiplegic cerebral palsy. Developmental Medicine and Child Neurology, 46(11), 746–753. 10.1017/s0012162204001288 [DOI] [PubMed] [Google Scholar]
  18. Hung YC, & Gordon AM (2013). Motor learning of a bimanual task in children with unilateral cerebral palsy. Research in Developmental Disabilities, 34(6), 1891–1896. 10.1016/j.ridd.2013.03.008 [DOI] [PubMed] [Google Scholar]
  19. Islam M, Gordon AM, Skold A, Forssberg H, & Eliasson AC (2011). Grip force coordination during bimanual tasks in unilateral cerebral palsy. Developmental Medicine and Child Neurology, 53(10), 920–926. 10.1111/j.1469-8749.2011.04040.x [DOI] [PubMed] [Google Scholar]
  20. Jahnsen RB, Weedon-Fekjar H, Myklebust G, & Storvold GV (2025). Gross Motor Development by Age and Functional Level in Children with Cerebral Palsy from 6 Months to 17 Years-A Norwegian Population-Based Registry Study. Journal of Clinical Medicine, 14(1). 10.3390/jcm14010178 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Jones BW (1964). Modular arithmetic (1st ed.). Blaisdell Publishing Corporation. [Google Scholar]
  22. Kantak S, Jax S, & Wittenberg G (2017). Bimanual coordination: A missing piece of arm rehabilitation after stroke. Restorative Neurology and Neuroscience, 35(4), 347–364. 10.3233/RNN-170737 [DOI] [PubMed] [Google Scholar]
  23. Kirton A, & Deveber G (2013). Life after perinatal stroke. Stroke, 44(11), 3265–3271. 10.1161/strokeaha.113.000739 [DOI] [PubMed] [Google Scholar]
  24. Klingels K, Feys H, De Wit L, Jaspers E, Van de Winckel A, Verbeke G, De Cock P, & Molenaers G (2012). Arm and hand function in children with unilateral cerebral palsy: a one-year follow-up study. European Journal of Paediatric Neurology, 16(3), 257–265. 10.1016/j.ejpn.2011.08.001 [DOI] [PubMed] [Google Scholar]
  25. Klingels K, Jaspers E, Staudt M, Guzzetta A, Mailleux L, Ortibus E, & Feys H (2016). Do mirror movements relate to hand function and timing of the brain lesion in children with unilateral cerebral palsy? Developmental Medicine and Child Neurology, 58(7), 735–742. 10.1111/dmcn.12977 [DOI] [PubMed] [Google Scholar]
  26. Klingels K, Meyer S, Mailleux L, Simon-Martinez C, Hoskens J, Monbaliu E, Verheyden G, Verbeke G, Molenaers G, Ortibus E, & Feys H (2018). Time Course of Upper Limb Function in Children with Unilateral Cerebral Palsy: A Five-Year Follow-Up Study. Neural Plasticity, 2018. 10.1155/2018/2831342 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Leversen JS, Haga M, & Sigmundsson H (2012). From children to adults: motor performance across the life-span. PloS One, 7(6), e38830. 10.1371/journal.pone.0038830 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Liang KJ, Chen HL, Shieh JY, & Wang TN (2021). Measurement properties of the box and block test in children with unilateral cerebral palsy. Scientific Reports, 11(1). 10.1038/s41598-021-00379-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Marion SD, Kilian SC, Naramor TL, & Brown WS (2003). Normal development of bimanual coordination: Visuomotor and interhemispheric contributions. Developmental Neuropsychology, 23(3), 399–421. 10.1207/S15326942dn2303_6 [DOI] [PubMed] [Google Scholar]
  30. McPherson JG, Chen A, Ellis MD, Yao J, Heckman CJ, & Dewald JPA (2018). Progressive recruitment of contralesional cortico-reticulospinal pathways drives motor impairment post stroke. Journal of Physiology-London, 596(7), 1211–1225. 10.1113/Jp274968 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Muetzel RL, Collins PF, Mueller BA, A MS, Lim KO, & Luciana M (2008). The development of corpus callosum microstructure and associations with bimanual task performance in healthy adolescents. Neuroimage, 39(4), 1918–1925. 10.1016/j.neuroimage.2007.10.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Nemanich S, & Schindler-Ivens S (2025). Age-Related Development of Bilateral Coordination of the Upper Limbs in Children and Adolescents. Journal of Motor Behavior, 57(1), 12–20. 10.1080/00222895.2024.2396114 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Rich TL, Nemanich S, Chen CY, Sutter EN, Feyma T, Krach LE, & Gillick BT (2020). Ipsilateral Corticospinal Tract Excitability Contributes to the Severity of Mirror Movements in Unilateral Cerebral Palsy: A Case Series. Clinical EEG and Neuroscience, 51(3), 185–190. 10.1177/1550059419899323 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Riddell M, Kuo HC, Zewdie E, & Kirton A (2019). Mirror movements in children with unilateral cerebral palsy due to perinatal stroke: clinical correlates of plasticity reorganization. Developmental Medicine and Child Neurology. 10.1111/dmcn.14155 [DOI] [PubMed] [Google Scholar]
  35. Ringenbach SDR, & Amazeen PG (2005). How do children control rate, amplitude, and coordination stability during bimanual circle drawing? Ecological Psychology, 17(1), 1–18. 10.1207/s15326969eco1701_1 [DOI] [Google Scholar]
  36. Robertson SD (2001). Development of bimanual skill: The search for stable patterns of coordination. Journal of Motor Behavior, 33(2), 114–126. 10.1080/00222890109603144 [DOI] [PubMed] [Google Scholar]
  37. Rudisch J, Butler J, Izadi H, Birtles D, & Green D (2018). Developmental Characteristics of Disparate Bimanual Movement Skills in Typically Developing Children. Journal of Motor Behavior, 50(1), 8–16. 10.1080/00222895.2016.1271302 [DOI] [PubMed] [Google Scholar]
  38. Rudisch J, Butler J, Izadi H, Zielinski IM, Aarts P, Birtles D, & Green D (2016). Kinematic parameters of hand movement during a disparate bimanual movement task in children with unilateral Cerebral Palsy. Human Movement Science, 46, 239–250. 10.1016/j.humov.2016.01.010 [DOI] [PubMed] [Google Scholar]
  39. Sakzewski L, Ziviani J, & Boyd R (2010). The relationship between unimanual capacity and bimanual performance in children with congenital hemiplegia. Developmental Medicine and Child Neurology, 52(9), 811–816. 10.1111/j.1469-8749.2009.03588.x [DOI] [PubMed] [Google Scholar]
  40. Sidiropoulos AN, Santamaria V, & Gordon AM (2021). Continuous inter-limb coordination deficits in children with unilateral spastic cerebral palsy. Clinical Biomechanics, 81. 10.1016/j.clinbiomech.2020.105250 [DOI] [PubMed] [Google Scholar]
  41. Smits-Engelsman BCM, Klingels K, & Feys H (2011). Bimanual force coordination in children with spastic unilateral cerebral palsy. Research in Developmental Disabilities, 32(5), 2011–2019. 10.1016/j.ridd.2011.04.007 [DOI] [PubMed] [Google Scholar]
  42. Staudt M (2010). Reorganization after pre- and perinatal brain lesions. Journal of Anatomy, 217(4), 469–474. 10.1111/j.1469-7580.2010.01262.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Swinnen SP, & Wenderoth N (2004). Two hands, one brain: cognitive neuroscience of bimanual skill. Trends in Cognitive Sciences, 8(1), 18–25. 10.1016/j.tics.2003.10.017 [DOI] [PubMed] [Google Scholar]
  44. Wang TN, Howe TH, Liang KJ, Chang TW, Shieh JY, & Chen HL (2021). Bimanual motor performance in everyday life activities of children with hemiplegic cerebral palsy. European Journal of Physical and Rehabilitation Medicine, 57(4), 568–576. 10.23736/S1973-9087.21.06504-7 [DOI] [PubMed] [Google Scholar]
  45. Woods BT, & Teuber HL (1978). Mirror movements after childhood hemiparesis. Neurology, 28(11), 1152–1157. [DOI] [PubMed] [Google Scholar]
  46. Zielinski IM, Green D, Rudisch J, Jongsma ML, Aarts PB, & Steenbergen B (2017). The relation between mirror movements and non-use of the affected hand in children with unilateral cerebral palsy. Developmental Medicine and Child Neurology, 59(2), 152–159. 10.1111/dmcn.13204 [DOI] [PubMed] [Google Scholar]

RESOURCES