Direction-dependent differences in the quality and quantity of horizontal reaching in people after stroke

Abstract

Arm reaching is often impaired in individuals with stroke. Nonetheless, how aiming directions influence reaching performance and how such differences change with motor recovery over time remain unclear. Here, we elucidated kinematic parameters of reaching toward various directions in people with poststroke hemiparesis in the subacute phase. A total of 13 and 15 participants with mild and moderate-to-severe hemiparesis, respectively, performed horizontal reaching in eight directions with their more-affected and less-affected sides using an exoskeleton robotic device at the time of admission to and discharge from the rehabilitation ward of the hospital. The movement time, path length, and number of velocity peaks were computed for the mild group (participants able to reach toward all eight directions). In addition, the total amount of displacement (i.e., movement quantity) toward two simplified directions (mediolateral or anteroposterior) was evaluated for the moderate-to-severe group (participants who showed difficulty in completing the reaching task). Motor recovery was evaluated using the Fugl-Meyer assessment. The mild group showed worse values of movement parameters during reaching in the anteroposterior direction, irrespective of the side of the arm or motor recovery achieved. The moderate-to-severe group exhibited less movement toward the anteroposterior direction than toward the mediolateral direction at admission; however, this direction-dependent bias in movement quantity decreased, with the movement expanding toward the anteroposterior direction with motor recovery at discharge. These results suggest that direction-dependent differences in the quality and quantity of reaching performance exist in people after stroke, regardless of the presence or severity of hemiparesis. This highlights the need to consider the task work area when designing rehabilitative training.NEW & NOTEWORTHY Arm reaching, a fundamental function required for the upper extremities, is often impaired after stroke due to muscle weakness and abnormal synergies. Nonetheless, how aiming directions influence performance remains unclear. Here, we report that direction-dependent differences in the quality and quantity of reaching performance exist, surprisingly regardless of the presence or severity of hemiparesis. This result highlights the need to consider the task work area when designing rehabilitative training.

Keywords: cerebrovascular disorders; hemiparesis; reaching; robotics; upper extremity.

Graphical abstract

Figure 1.

Reaching task, participant groups, and movement parameters of reaching. A: starting position and target locations. The participants performed arm-reaching movement on a two-dimensional horizontal plane toward one of eight targets appearing in the medial (M), medial forward (MF), forward (F), lateral forward (LF), lateral (L), lateral backward (LB), backward (B), and medial backward (MB) directions from the starting position. B: classification of the participants into two groups. The participants were classified into the mild and moderate-to-severe groups according to the level of motor impairment severity, as evaluated using the Fugl-Meyer assessment for upper extremity (FMA-UE), a clinical assessment score for upper-extremity impairment. The numbers inside indicate the number of participants included in each category. C: hand paths and three movement parameters in a representative participant in the mild group. All right-side data, regardless of whether they were data for the more-affected or less-affected side, were flipped to correspond to the left-side data.

Figure 2.

Direction-dependent differences in movement parameters in the mild group. Path length (A), movement time (B), and number of velocity peaks (C). Group means for the more-affected and less-affected sides at admission (More-affected-T_A and Less-affected-T_A) and discharge (More-affected-T_D and Less-affected-T_D) are displayed in the left and right spider plots, respectively. The rightmost panels with line graphs represent the group means and standard errors of means for each movement parameter when reaching toward the medial (M), medial forward (MF), forward (F), lateral forward (LF), lateral (L), lateral backward (LB), backward (B), and medial backward (MB) directions.

Figure 3.

Patterns of joint movement in the mild group. The interjoint coordination patterns of the more-affected and less-affected sides at admission (More-affected-T_A and Less-affected-T_A) (A) and at discharge (More-affected-T_D and Less-affected-T_D) (B). The delta angle in the shoulder and elbow joints while reaching toward the medial (M), medial forward (MF), forward (F), lateral forward (LF), lateral (L), lateral backward (LB), backward (B), and medial backward (MB) directions is presented in the x- and y-axes, respectively. Thick and thin colored lines indicate the mean between participants and the mean within a participant, respectively. To control the difference in data length among trials within a participant and between the participants, the data in each trial were resampled to be expressed with 0–100% of movement. Note that the angle (0, 0) represents the starting arm position.

Figure 4.

Quantity of reaching movement in the moderate-to-severe group. The hand path while reaching with the more-affected and less-affected sides at admission (More-affected-T_A and Less-affected-T_A) (A) and at discharge (More-affected-T_D and Less-affected-T_D) (B). The hand path is superimposed among participants and visualized in a two-dimensional histogram colormap for display purposes. The center colormap represents the hand path when all qualified trials were included regardless of the target location, whereas the surrounding ones represent the hand path when a target appeared at each location: medial (M), medial forward (MF), forward (F), lateral forward (LF), lateral (L), lateral backward (LB), backward (B), and medial backward (MB). C: total amount of displacement toward the mediolateral (ML_total in the x-axis) and anteroposterior (AP_total in the y-axis) directions. The top and bottom show the absolute and relative (proportion of the total) values. Small dots in the left represent the individual data of the more-affected side at admission (right red) and discharge (red) as well as those of the less-affected side at admission (right blue) and discharge (blue). Large dots with error bars represent the means and standard errors of means across participants. Bar graphs in the bottom represent the mean proportion of the amount of movement toward the mediolateral direction. Small dots represent individual data.