Evidence for altered upper extremity muscle synergies in chronic stroke survivors with mild and moderate impairment

Ask your therapist for how this will be incorporated into your stroke protocols.
http://journal.frontiersin.org/article/10.3389/fnhum.2015.00006/full?
Jinsook Roh^1,2*, William Z. Rymer^2,3,4 and Randall F. Beer^2,3

• ¹Department of Kinesiology, Temple University, Philadelphia, PA, USA
• ²Department of Physical Medicine and Rehabilitation, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA
• ³Sensory Motor Performance Program, Rehabilitation Institute of Chicago, Chicago, IL, USA
• ⁴Department of Biomedical Engineering, Northwestern University, Chicago, IL, USA

Previous studies indicate that motor coordination may be achieved by assembling task-dependent combinations of a few muscle synergies, defined here as fixed patterns of activation across a set of muscles. Our recent study of severely impaired chronic stroke survivors showed that some muscle synergies underlying isometric force generation at the hand are altered in the affected arm. However, whether similar alterations are evident in stroke survivors with lesser impairment remains unclear. Accordingly, we examined muscle synergies underlying spatial patterns of elbow and shoulder muscle activation recorded during an isometric force target matching protocol performed by 16 chronic stroke survivors, evenly divided across mild and moderate impairment levels. We applied non-negative matrix factorization to identify the muscle synergies and compared their structure across groups, including previously collected data from six age-matched control subjects and eight severely impaired stroke survivors. For all groups, EMG spatial patterns were well explained by task-dependent combinations of only a few (typically 4) muscle synergies. Broadly speaking, elbow-related synergies were conserved across stroke survivors, regardless of impairment level. In contrast, the shoulder-related synergies of some stroke survivors with mild and moderate impairment differed from controls, in a manner similar to severely impaired subjects. Cluster analysis of pooled synergies for the 30 subjects identified seven distinct clusters (synergies). Subsequent analysis confirmed that the incidences of three elbow-related synergies were independent of impairment level, while the incidences of four shoulder-related synergies were systematically correlated with impairment level. Overall, our results suggest that alterations in the shoulder muscle synergies underlying isometric force generation appear prominently in mild and moderate stroke, as in most cases of severe stroke, in an impairment level-dependent manner.

Introduction

Broadly speaking, disturbances of motor function following stroke can be attributed to three primary impairments: reduced corticospinal drive to agonist muscles (weakness), altered reflex activity (e.g., spasticity), and impaired motor coordination (Twitchell, 1951; Brunnstrom, 1970; Knutsson and Mårtensson, 1980). In many patients, when weakness and spasticity are treated effectively, or resolve spontaneously, motor dysfunction remains severe (Landau, 1980; Hesse et al., 1996; Dewald et al., 2001). Accordingly, understanding the mechanisms that underlie impaired motor coordination following stroke is essential for the design of effective rehabilitation protocols.

Behavioral and stimulation-based studies in motor systems suggest that normal neuromuscular coordination can be achieved by activating a relatively limited number of muscle synergies, each of which represents a pattern of muscle activation with distinct spatial (and in some formulations, temporal) characteristics (Tresch et al., 1999; Ivanenko et al., 2004; Cheung et al., 2005; d’Avella and Bizzi, 2005; d’Avella et al., 2006, 2008; Torres-Oviedo et al., 2006; Torres-Oviedo and Ting, 2007; Kargo et al., 2010; Overduin et al., 2012). Muscle synergies have been identified as building blocks for a variety of motor tasks in humans, including postural responses (Krishnamoorthy et al., 2003; Weiss and Flanders, 2004; Torres-Oviedo and Ting, 2007), locomotion (Clark et al., 2010; Monaco et al., 2010), hand shaping and signing (Santello et al., 1998; Weiss and Flanders, 2004; Ajiboye and Weir, 2009), isometric force generation in the upper extremity (Roh et al., 2012), and reaching movements performed under different biomechanical constraints (Sabatini, 2002; d’Avella et al., 2006, 2008; Cheung et al., 2009a; Muceli et al., 2010). Furthermore, muscle synergies generalize across different task constraints, as shown in recent animal studies of postural responses involving different perturbation types and postures (Torres-Oviedo et al., 2006) as well as voluntary motor behaviors (Roh et al., 2011). Similarly, muscle synergies have been shown to be quite robust in intact humans (Valero-Cuevas, 2000; Ivanenko et al., 2004; d’Avella et al., 2006, 2008; Chvatal et al., 2011; Hug et al., 2011; Roh et al., 2012). The use of muscle synergies may facilitate control of task-level variables (Ting and Macpherson, 2005; Torres-Oviedo et al., 2006; McKay and Ting, 2008), and/or simplify the generation of motor behaviors by reducing the dimensionality of the control problem associated with mechanically redundant musculature (Grillner, 1985; Bizzi et al., 1991; Tresch et al., 1999; Fetz et al., 2000; Tresch et al., 2002; Miller, 2004; Bizzi et al., 2008; d’Avella and Lacquaniti, 2013).

Several recent studies have examined how stroke impacts the modular control of voluntary limb movements (Cheung et al., 2009b, 2012; Clark et al., 2010).The initial study, which focused on reaching movements performed by a group of chronic stroke survivors with predominantly mild impairment, concluded that stroke altered the recruitment patterns of normal muscle synergies, rather than altering synergy internal structure (Cheung et al., 2009b). A subsequent study involving subjects with a diverse range of impairment levels confirmed preservation of normal muscle synergies in mildly impaired stroke survivors, but reported evidence of merging and fractionation of normal synergies in more impaired stroke subjects (Cheung et al., 2012). Similarly, Clark et al. (2010) found that fewer muscle synergies were required to reconstruct locomotor muscle activation patterns in more impaired stroke survivors, reflecting an apparent merging of synergies identified in healthy subjects. Overall, these studies suggest that alterations in muscle synergy structure are mainly evident in stroke survivors with severe impairment.

Muscle synergies are potentially shaped by biomechanical or task constraints, independent of putative neural constraints (Todorov, 2004; Kutch and Valero-Cuevas, 2012). Thus, differences in task performance between stroke survivors and healthy controls are a potential confounding factor when comparing the number and structure of muscle synergies underlying voluntary limb movements. Accordingly, we developed an isometric protocol that provides an opportunity to more closely match task variables (i.e., limb posture and required force level) across healthy and impaired individuals. Our initial study, confined to chronic stroke survivors with severe motor impairment, examined muscle synergies underlying isometric force generation at the hand (Roh et al., 2013). In contrast to the results for dynamic tasks (Cheung et al., 2009b, 2012; Clark et al., 2010), we found preservation of the number of muscle synergies in the affected arm of severely impaired stroke survivors and relatively stereotyped alterations in specific muscle synergies related to the activation of shoulder muscles.

As an extension of our previous study (Roh et al., 2013), the current study focused on higher functioning stroke survivors. Specifically, we hypothesized that alterations in muscle synergy structure are also evident in stroke survivors with mild or moderate impairment. To evaluate this hypothesis, subjects with mild and moderate motor impairment completed a 3-D isometric force target matching protocol identical to Roh et al. (2013). Synergies underlying shoulder and elbow muscle activations were identified using non-negative matrix factorization, and compared with those previously identified in age-matched healthy subjects and severely impaired stroke survivors.