Elastic, viscous, and mass load effects on poststroke muscle recruitment and co-contraction during reaching: a pilot study

Abstract

Background: Resistive exercise after stroke can improve strength (force-generating capacity) without increasing spasticity (velocity-dependent hypertonicity). However, the effect of resistive load type on muscle activation and co-contraction after stroke is not clear.

Objective: The purpose of this study was to determine the effect of load type (elastic, viscous, or mass) on muscle activation and co-contraction during resisted forward reaching in the paretic and nonparetic arms after stroke.

Design: This investigation was a single-session, mixed repeated-measures pilot study.

Methods: Twenty participants (10 with hemiplegia and 10 without neurologic involvement) reached forward with each arm against equivalent elastic, viscous, and mass loads. Normalized shoulder and elbow electromyography impulses were analyzed to determine agonist muscle recruitment and agonist-antagonist muscle co-contraction.

Results: Muscle activation and co-contraction levels were significantly higher on virtually all outcome measures for the paretic and nonparetic arms of the participants with stroke than for the matched control participants. Only the nonparetic shoulder responded to load type with similar activation levels but variable co-contraction responses relative to those of the control shoulder. Elastic and viscous loads were associated with strong activation; mass and viscous loads were associated with minimal co-contraction.

Limitations: A reasonable, but limited, range of loads was available.

Conclusions: Motor control deficits were evident in both the paretic and the nonparetic arms after stroke when forward reaching was resisted with viscous, elastic, or mass loads. The paretic arm responded with higher muscle activation and co-contraction levels across all load conditions than the matched control arm. Smaller increases in muscle activation and co-contraction levels that varied with load type were observed in the nonparetic arm. On the basis of the response of the nonparetic arm, this study provides preliminary evidence suggesting that viscous loads elicited strong muscle activation with minimal co-contraction. Further intervention studies are needed to determine whether viscous loads are preferable for poststroke resistive exercise programs.

Figure 1.

Experimental loads. (A) Cart-and-rail apparatus used for the experimental task. Participants reached 15 cm by pushing forward against the elastic load (far), the viscous load (center), and the mass load (near). Each cart was fitted with infrared markers and a force transducer. (B) Representative data for force profiles collected from a single control participant (x-axis shows time in milliseconds; y-axis shows force in newtons) during forward reaching against elastic, viscous, and mass loads. Shaded areas under the acceleration components of the curves were used to determine equivalent loads.

Figure 2.

Electromyography (EMG) data processing. (A) Representative data collected in one trial from a single control participant reaching against a viscous load. From top to bottom: channels of raw surface EMG data collected from the anterior deltoid muscle, posterior deltoid muscle, long head of the biceps muscle, short head of the biceps muscle, lateral head of the triceps muscle, and long head of the triceps muscle and force, velocity, and position data collected during forward reaching against a viscous load. Red cursors indicate the onset and offset of movement. (B) Representative data collected from a single control participant for the 6 muscles of interest plotted as the average of 10 trials of normalized, full-wave-rectified EMG data (±95% confidence interval) at 30% maximum voluntary isometric contractions (MVIC) for elastic, viscous, and mass loads. The dependent measure of agonist muscle activation (EMG impulse) was calculated as the area under the curve. Trials were aligned at movement onset (vertical dashed line). Scale bars (solid black lines) represent 500 ms for time on the x-axis and 20% MVIC on the y-axis. For traces indicated by asterisks, the vertical scale bar corresponds to 50% MVIC. (C) Representative example of normalized full-wave-rectified EMG data for anterior deltoid (blue trace) and posterior deltoid (green trace) muscle activation from a control arm (left panel) and a paretic arm (right panel) during reaching against a mass load. The red trace indicates the lowest EMG signal between the 2 muscles at each time point; the area under the red curve represents the measure of co-contraction. The left panel represents little co-contraction; the right panel represents significant co-contraction. The vertical dashed line indicates the onset of movement, when the forward reach velocity exceeded 0.02 m/s. Scale bars (solid black lines) represent 500 ms for time on the x-axis and 10% MVIC on the y-axis.

Figure 3.

Effect of arm type on normalized electromyography (EMG) impulses. Bar graphs show mean and standard error (SE) of the normalized EMG impulses for the agonist anterior deltoid and triceps muscles (top row) and coactivity for the shoulder and elbow (bottom row) by group collapsed across load types. Shaded bars represent the paretic arm (PAR [dark blue]) and the nonparetic arm (N-PAR [light blue]) of participants with stroke; white bars represent matched control (CON) participants. Significant differences are indicated by asterisks: *P<.05, **P<.001.

Figure 4.

Effect of load type by arm on normalized electromyography (EMG) impulses. Bar graphs show mean and standard error (SE) of the normalized full-wave-rectified EMG impulses for the agonist anterior deltoid (first row) and triceps (second row) muscles, shoulder coactivity (third row), and elbow coactivity (fourth row) by group and load type. Bars are ordered by load type (M=mass, E=elastic, and V=viscous) for the paretic (PAR [dark blue]), nonparetic (N-PAR [light blue]), and respective matched control (CON [white]) groups. Significant differences are indicated by asterisks: *P<.05. Note differences in the scaling of the y-axis.