Finger strength, individuation, and their interaction: Relationship to hand function and corticospinal tract injury after stroke

Abstract

Objective: The goal of this study was to determine the relative contributions of finger weakness and reduced finger individuation to reduced hand function after stroke, and their association with corticospinal tract (CST) injury.

Methods: We measured individuated and synergistic maximum voluntary contractions (MVCs) of the index and middle fingers, in both flexion and extension, of 26 individuals with a chronic stroke using a robotic exoskeleton. We quantified finger strength and individuation, and defined a novel metric that combines them - "multifinger capacity". We used stepwise linear regression to identify which measure best predicted hand function (Box and Blocks Test, Nine Hole Peg Test) and arm impairment (the Upper Extremity Fugl-Meyer Test).

Results: Compared to metrics of strength or individuation, capacity survived the stepwise regression as the strongest predictor of hand function and arm impairment. Capacity was also most strongly related to presence or absence of lesion overlap with the CST.

Conclusions: Reduced strength and individuation combine to shrink the space of achievable finger torques, and it is the resulting size of this space - the multifinger capacity - that is of elevated importance for predicting loss of hand function.

Significance: Multi-finger capacity may be an important target for rehabilitative hand training.

Keywords: Corticospinal tract (CST) injury; Finger individuation; Finger strength; Hand function; Multi-finger strength; Neurorehabilitation; Stroke.

Figure 1

The FINGER robotic movement therapy device. FINGER allows subjects to move their index and middle fingers through a naturalistic curling motion, and can assist or resist this movement. Two force sensors on each finger are used to calculate joint torques during movement or while the robot holds the fingers in an isometric position.

Figure 2

Example finger torques measured by the robot. Subjects were instructed to flex their fingers against the stationary robot in the following order: index, middle, both together. The solid red and green lines are the flexion MCP torques generated by the subject with their index and middle fingers, respectively. Red, green, and blue dashed boxes identify the flex index, flex middle, and flex both flexion events.

Figure 3

LEFT: An example multi-finger capacity plot of a subject’s MVC test performance. Results from a single subject’s unimpaired (flexion in red and extension in blue) and impaired hand (flexion in magenta and extension in cyan) are shown. Convex hulls fit to these four torque traces are shown with striped lines; and the area of these convex hulls indicate flexion capacity and extension capacity. RIGHT: Individuation is indicated by the angle of the circular sector fit to each hull area, increasing from 0 to 1 as the sector angle increase from zero to 90°. Average strength is indicated by the radius of the sector. In this case, the subject’s impaired side has reduced strength and individuation in both flexion and extension.

Figure 4

Multi-finger capacity plots for all subjects (N=26) shown in order of increasing normalized impaired flexion capacity (FC). MCP torques for the index and middle fingers are plotted on the x and y axes, respectively, providing a way to simultaneously visualize and quantify finger strength (distance from origin), individuation ability (ability to move along the x and y axes independently), and capacity (the product of strength and individuation). Flexion scores for strength (FS), individuation (FI), and capacity (FC) are shown in the top right corner of each plot, followed by Fugl-Meyer (FM) score. Results are shown for both flexion (1^st quadrant) and extension (3^rd quadrant) and both subject’s unimpaired and impaired hands. For the unimpaired hand, flexion is shown in red and extension in blue. For the impaired hand, flexion is shown in magenta and extension in cyan. The convex hull boundaries of these four torque traces are shown with a striped line of the same color. By plotting both unimpaired and impaired hands, we can readily see how stroke affects flexion strength (radius from origin), flexion individuation (angle of sector) and flexion capacity (convex hull area).

Figure 5

LEFT: Flexion strength (MCP torque of the impaired hand normalized to that of the unimpaired hand) versus extension strength (also MCP torque normalized to the unimpaired hand). RIGHT: Strength (normalized MCP torque) versus individuation for both flexion and extension of the impaired hand.

Figure 5

LEFT: Flexion strength (MCP torque of the impaired hand normalized to that of the unimpaired hand) versus extension strength (also MCP torque normalized to the unimpaired hand). RIGHT: Strength (normalized MCP torque) versus individuation for both flexion and extension of the impaired hand.

Figure 6

LEFT: Box plots for injury groups below thalamus (injury below the level of the thalamus, N = 6), % CST > 0 (% CST overlap greater than zero, N = 17), and % CST = 0 (% CST overlap equal to zero, N = 3). RIGHT: Correlations for flexion strength (normalized to unimpaired hand), individuation, and capacity (normalized to unimpaired hand) vs. % CST overlap for injuries above the thalamus (% CST ≥ 0, N = 20).

Figure 6

LEFT: Box plots for injury groups below thalamus (injury below the level of the thalamus, N = 6), % CST > 0 (% CST overlap greater than zero, N = 17), and % CST = 0 (% CST overlap equal to zero, N = 3). RIGHT: Correlations for flexion strength (normalized to unimpaired hand), individuation, and capacity (normalized to unimpaired hand) vs. % CST overlap for injuries above the thalamus (% CST ≥ 0, N = 20).

Figure 7

Flexion strength (normalized to unimpaired hand), individuation, and capacity (normalized to unimpaired hand) vs. Fugl-Meyer score (FM, LEFT) and normalized Box and Blocks (BB, RIGHT). All three flexion metrics of finger performance (strength, individuation, and capacity) correlate strongly with Fugl-Meyer and Box and Blocks scores; they also correlate with Nine-Hole-Peg Test scores (not shown, see Table 3).

Figure 7

Flexion strength (normalized to unimpaired hand), individuation, and capacity (normalized to unimpaired hand) vs. Fugl-Meyer score (FM, LEFT) and normalized Box and Blocks (BB, RIGHT). All three flexion metrics of finger performance (strength, individuation, and capacity) correlate strongly with Fugl-Meyer and Box and Blocks scores; they also correlate with Nine-Hole-Peg Test scores (not shown, see Table 3).