Constraints on grip selection in hemiparetic cerebral palsy: effects of lesional side, end-point accuracy, and context

Abstract

This study was concerned with selection criteria used for grip planning in adolescents with left or right hemiparetic cerebral palsy. In the first experiment, we asked participants to pick up a pencil and place the tip in a pre-defined target region. We varied the size of the target to test the hypothesis that increased end-point precision demands would favour the use of a grip that affords end-state comfort. In the second experiment, we studied grip planning in three task contexts that were chosen to let us test the hypothesis that a more functional task context would likewise promote the end-state comfort effect. When movements were performed with the impaired hand, we found that participants with right hemiparesis (i.e., left brain damage) aimed for postural comfort at the start rather than at the end of the object-manipulation phase in both experiments. By contrast, participants with left hemiparesis (i.e., right brain damage) did not favour a particular selection criterion with the impaired hand in the first experiment, but aimed for postural comfort at the start in the second experiment. When movements were performed with the unimpaired hand, grip selection criteria again differed for right and left hemiparetic participants. Participants with right hemiparesis did not favour a particular selection criterion with the unimpaired hand in the first experiment and only showed the end-state comfort effect in the most functional tasks of the second experiment. By contrast, participants with left hemiparesis showed the end-state comfort effect in all conditions of both experiments. These data suggest that the left hemisphere plays a special role in action planning, as has been recognized before, and that one of the deficits accompanying left brain damage is a deficit in forward movement planning, which has not been recognized before. Our findings have both theoretical and clinical implications.

Introduction

When individuals with hemiparetic cerebral palsy are asked to grasp an object with the impaired arm, a characteristic movement pattern is observed. The movement pattern is distinguished by an increased number of submovements (e.g., [51], [52]), increased variability of hand trajectories (e.g., [56], [57]), inappropriately coordinated force levels in the hand and fingers (e.g., [7], [8]), a stereotyped shoulder–elbow recruitment order (e.g., [24], [47]), and an increased level of trunk involvement (e.g., [1], [3], [55]). Movement patterns observed when the same subjects perform with the unimpaired hand also show some deviations from movement patterns exhibited by normal controls (e.g., [2], [10], [18]).

The cause of the movement pattern specific to hemiparetic cerebral palsy has been ascribed to damage to the mechanisms underlying movement execution. This ascription is reasonable considering that the primary motor cortex, which is crucially involved in movement execution, is often damaged in hemiparesis. Thus, it is unsurprising that hemiparesis has classically been referred to as an “upper motor neuron syndrome” (cf. Ref. [19]). As compelling as this analysis may be, however, because cortical areas such as the dorsolateral frontal, premotor cortex, and supplementary motor area are close to the primary motor cortex, it is also possible that when these more frontal cortical areas are damaged during the onset of hemiparesis, their associated functions—goal selection, sequencing, and motion planning [60]—are also affected. If this hypothesis is correct, the implication is that hemiparetic cerebral palsy reflects disorders of processes antecedent to, and so “higher than,” movement execution per se.

The present study was designed to test this hypothesis by evaluating the extent to which people with hemiparetic cerebral palsy show signs of abnormal motion planning. Because the study was behavioral, it only provided a test of the behavioral hypothesis offered here. Still, it is possible to speculate on the neural substrates of the behavior in interpreting the results, just as it was possible to motivate this behavioral inquiry by considerations of neuroanatomy and neurophysiology. We offer such speculations in the final section of the article.

In approaching the question of how motion planning works in individuals with hemiparetic cerebral palsy, it is important to focus on some type of motion that is sensitive to planning effects. A good candidate is prehension. Prehension requires motion planning to permit adaptive modification of grasps depending on what objects are grasped and for what purpose [25]. Initial grasps generally ensure that, when an object needs to be moved to a new position, the task ends with the upper extremity in a comfortable final posture (i.e., a posture for which the joints end at or near the middle of the range of motion) [4]. Rosenbaum et al. [30], [31], [34] found that the preference for ending near midrange applies to the termination of complex object transport maneuvers. In particular, these authors discovered that when people reached out and took hold of a cylinder that had to be moved to another location, they usually took hold of the cylinder with an awkward hand posture (i.e., with an underhand grip) if this enabled them to complete the cylinder transport with a comfortable posture (i.e., with an overhand grip). This “end-state comfort effect” has been observed in a wide variety of bar-transport and handle-rotation tasks [9], [29], [30], [31], [32], [34], [40], [41]. The end-state comfort effect indicates that subjects plan beyond the first grip and anticipate future states.

Rosenbaum et al. [34] proposed three explanations for the end-state comfort effect: (1) exploitation of gravity, (2) use of elastic energy, and (3) precision. According to the latter explanation, movements at or near the middle of the range of motion of a joint can be performed with greater precision than movements at or near extreme joint angles. Consistent with this hypothesis Rosenbaum et al. [34] showed that pronation–supination oscillations of the sort required to make corrective movements occur more rapidly at or near the middle of the forearm's pronation–supination range than at the extremes, and Rossetti et al. [36] showed less in pointing variability at mid-range joint angles than at extreme joint angles.

Recently, Steenbergen et al. [46] tested whether end-state comfort also functions as a selection criterion in grip planning for people with spastic hemiparesis. These authors anticipated that as a result of the reduced range of motion on the impaired side due to hyperactive stretch reflexes and hypertonia accompanying spasticity, there would only be a small joint range in which reliable position-error encoding exists. This prediction was tested both in a bar-positioning task and a bar-rotation task. The question was whether the observed grasping movements reflected constraint satisfaction with respect to postural comfort at the start of the movements, at the end of the movements, or during the movement. Constraint satisfaction that was defined with respect to postural comfort during the movement was operationalised as minimization of the sum of forearm pronation and supination (termed “total comfort” in what follows). The results indicated that for the impaired hand, end-state comfort was not a selection criterion for grip planning. Rather, the results suggested that grip selection was either determined by start-posture comfort or total comfort (i.e., the tendency to minimize forearm rotation). Limitations in the design of that study made it impossible to distinguish between the latter two hypotheses.

In the two experiments described here, we followed up on the earlier study of Steenbergen et al. [46] by seeking to identify the grip-planning selection criteria used by people with hemiparetic cerebral palsy. For this purpose, we studied bar-handling tasks in which we manipulated two experimental factors: end-point precision (Experiment 1) and task context (Experiment 2). We also examined differences between left- and right-brain damage participants to gain more insight into the hemispheric control of grip planning. Our interest in the differential effects of damage to the left or right side of the brain was prompted by earlier work suggesting that brain regions devoted to movement planning are primarily located in the left hemisphere. That work began with the classic observations concerning planning deficits in apraxia by Liepmann (see Freeman [11] for a review). More recently, the special role of the left hemisphere for motor planning has been confirmed by Schluter et al. [39], using PET, who showed activation in the prefrontal, premotor, and intraparietal areas of the left hemisphere but not the right hemisphere during choice reaction time tasks, irrespective of the hand performing the responses. The findings of Schluter et al. (2001) are consistent with the work on apraxia in suggesting that the areas responsible for movement planning are mainly located in the left hemisphere (e.g., [12], [14], [17], [49], [59]). Based on findings such as these, we hypothesized that movement planning, as indexed by the end-state comfort effect, would be less adversely affected by damage to the right hemisphere than by damage to the left hemisphere in individuals whose neurological conditions are otherwise comparable. We tested this prediction by comparing the grasping movements of adolescents with left and right hemiparetic cerebral palsy.