The relationship between lower limb proprioception and balance performance in children with cerebral palsy: a multisystem and multi-joint approach

Abstract

BACKGROUND: Balance problems are a frequent focus of rehabilitation approaches in children with cerebral palsy (CP). Although proprioceptive impairments, affecting 46-90% of children with CP, are commonly observed, their role in postural control remains insufficiently understood. AIM: Given the multisystemic nature of postural control, this study aimed to investigate the association between lower-limb proprioception (hip, knee, ankle) and distinct postural control systems in children with CP compared to typically development (TD). DESIGN: Cross-sectional case-control study. SETTING: Different outpatient settings and the community. POPULATION: Children aged 5-12 years with TD (N.=30, 8.6 +/- 2.2y; boys: 15) and spastic CP (N.=32, 8.4 +/- 2.1y; boys: 16; GMFCS-I/II/III: 24/7/1; uni-/bilateral: 17/15). METHODS: Postural control was assessed using the age-specific Kids-Balance Evaluation Systems Test-extended version (Kids-BESTest-2). Hip, knee, and ankle proprioception were assessed using passive-ipsilateral Joint-Position-Reproduction (JPR). Joint Reproduction Error (JRE,degrees) was calculated from 3D kinematics for the joints of the dominant (JREd) and nondominant (JREnd) leg. Associations between Kids-BESTest-2 (total and domain scores, %), JREtotal (all joints) and summed joint JREankle, JREknee and JREhip (JREd+JREnd) were analyzed using general linear models with correction for multiple testing. RESULTS: JREtotal was not significantly associated with Kids-BESTest-2 scores in either group (P>0.01). In CP, higher JREhip (reflecting poorer proprioception) was significantly a ssociated w ith l ower total (qp2=0.24, P <0.001) a nd 3 o ut o f 5 d omain s cores (qp2=0.14-0.22, P<0.004), reflecting poorer balance performance. Similarly, JREknee predicted scores for `Transitions and Anticipatory Postural Adjustments' (qp2=0.13, P=0.008). Across both groups, poorer hip (JREhip) and ankle (JREankle) proprioception were significantly associated with poorer performance for `Transitions and Anticipatory Postural Adjustments' (qp2=0.16, P=0.003) and `Reactive postural responses' (qp2=0.12, P=0.005), respectively. CONCLUSIONS: Postural control deficits in children with CP affect multiple systems, and each system is associated with joint-specific proprio-ceptive impairments (particularly at the hip), rather than overall proprioceptive accuracy. These findings suggest task-dependent and pathology-dependent associations, the latter likely reflecting limitations in compensatory proprioceptive reweighting across lower-limb joints. CLINICAL REHABILITATION IMPACT: These findings underscore the need for a system-and joint-specific approach to better understand postural control deficits in children with CP and tailor interventions that enhance proprioceptive input during task-specific balance training.