On the nature and measurement of spasticity

The problem of the measurement of spasticity has been addressed by many groups in numerous publications for at least half a century. A recent literature search on the keyword ‘spasticity’ returned 6846 papers since 1954. Of these, 335 papers were directly or indirectly concerned with measurement. The paper by Condliffe, Clark and Patten (2005) in this issue, once again tackles the difficult problem of the validity and reliability of spasticity measurement. Traditional measures have focused on the quantification of phasic and tonic stretch reflex magnitudes despite the fact that magnitude measures do not generally correlate with the clinical impression of the severity of spasticity. The paper by Condliffe et al. (2005) goes beyond traditional measures by including a measure of stretch reflex threshold excitability while simultaneously including more common measures of EMG burst duration, burst intensity and torque. The question arises as to what exactly do we need to know in order to characterize a disorder. A large number of measures, classified as being electrophysiological, mechanomyographic, kinematic or functional, have been described for spasticity. How many measures do we need? Is it valid to focus on the quantification of a variable just because it is easy to obtain or that it may have been used before? This persistence leads to a rather confusing situation in which the clinician has more information than he or she knows what to do with. Adding to the confusion is that some measures may vary on repeated testing while others may not. A valid measure of spasticity is one that is chosen within the context of a theory describing the physiological mechanisms underlying the control of posture and movement in healthy individuals and possible impairments of these mechanisms leading to motor disorders. It has been well founded that in healthy subjects, the regulation of stretch reflex thresholds may be a major mechanism underlying the control of posture and movement (Feldman and Levin, 1995). According to this theoretical approach, in healthy subjects, the range of regulation of stretch reflex or muscle activation thresholds is larger than the biomechanical range of the joint. The whole repertoire of available motor tasks includes full muscle relaxation to full muscle activation in all parts of the physiological range. The ability to regulate muscle force in all parts of the physiological range may be impaired in patients with central nervous system lesions because of a narrowing of the limits of regulation of stretch reflex thresholds. Thus, in this theoretical perspective, the stretch reflex threshold measure expressed in angular coordinates achieves content validity, at least for muscles around a single joint, and provides the clinician/researcher with specific information about the behaviour of the central nervous system. In the case of spasticity, the threshold measure used by Condliffe et al. (2005) is based on this theoretical perspective in which spasticity is inextricably related to disordered motor control and both are described by problems in threshold regulation (Levin and Feldman, 1994; Levin et al., 2000). More specifically, this theoretical approach suggests that deficits in the range of regulation of stretch reflex thresholds underlie motor disorders. Studies supporting this theory have shown that patients with hemiparesis were unable to regulate stretch reflex thresholds of elbow flexors in the required range (Levin et al., 2000) leading to spasticity, weakness, postural instability and abnormal muscle activation of individual muscles in specific joint ranges. Threshold control theory is also consistent with the current understanding of stretch reflexes as task-adjustable, broadly tuneable, parameterized structures. In addition to the description of valid measures, researchers and clinicians have struggled to develop measures that are reliable, responsive and useful. All of these criteria, however, have been difficult to obtain in a single measure. For example, the ubiquitous Ashworth Scale (AS; Ashworth, 1964) in its original or modified form (MAS) has shown only poor to good inter-rater reliability (Blackburn et al., 2002; Bohannon and Smith, 1987; Pandyan et al., 1999) but experts agree that the scale may not be measuring the characteristic that essentially differentiates spasticity from other disorders of tone (cf. Pandyan et al., 1999; Sanger et al., 2003). While the scale does a good job in determining the amount of resistance felt to Clinical Neurophysiology 116 (2005) 1754–1755 www.elsevier.com/locate/clinph