Altered upper limb kinematics in individuals with dyskinetic cerebral palsy in comparison with typically developing peers – A statistical parametric mapping study

Dyskinetic cerebral palsy (DCP) is clinically characterized by involuntary movements and abnormal postures, which can aggravate with activity. While upper limb movement variability is often detected in the clinical picture, it remains unknown how movement patterns of individuals with DCP differ from typically developing (TD) peers. Do individuals with DCP show i) higher time-dependent standard deviations of upper limb joint angles and ii) altered upper limb kinematics in time and/or amplitude during functional upper limb tasks in comparison with TD individuals? Three-dimensional upper limb movement patterns were cross-sectionally compared in 50 individuals with and without DCP during three functional tasks: reach forward (RF), reach and grasp vertical (RGV) and reach sideways (RS). Mean and point-wise standard deviations of angular waveform of the upper limb joint angles were compared between groups to evaluate differences in time and/or amplitude using traditional and non-linear registration statistical parametric mapping. Thirty-five extremities from 30 individuals (mean age 17y4m, range 5–25 y; MACS level I(n = 2); II(n = 15); III(n = 16); IV(n = 2)) with DCP and twenty TD individuals (mean age 16y8m, range 8–25 y) were evaluated. The DCP compared to TD group showed higher point-wise standard deviations at the level of all joints, which was time-dependent and varied between tasks. Mean wrist and elbow flexion was higher for the DCP group during RF (0–83 % wrist; 57–100 % elbow), RGV (0–82 % wrist; 12–100 % elbow) and RS (0–43 % wrist; 70–100 % elbow). This is the first study exploring the movement patterns of individuals with DCP during reaching using quantitative measures. Analyzing these individual movement patterns by statistical parametric mapping (SPM) allows us to focus on both specific joint or on specific timing during the movement cycle. The individual information that this method yields can guide individual therapy aiming to improve reaching function in different parts of the movement cycle or evaluate intervention effects on upper extremity treatment. •We explored the variability of 3D upper limb waveforms in dyskinetic CP (DCP) with statistical parametric mapping (SPM).•This is the first study using non-linear registration SPM in 3D upper limb movements.•The DCP compared to typically developing group showed higher, task-dependent point-wise standard deviations in all tasks.•The DCP compared to typically developing group showed higher wrist/e