182 IEEE TRANSACTIONS ON REHABILITATION ENGINEERING, VOL. 6, NO. 2, JUNE 1998 Stretch Reflex Sensitivity: Effects of Postural and Muscle Length Changes Jiping He, Senior Member, IEEE Abstract— In this study, a combination of clinical evalua- tion, laboratory testing, and model simulation of spasticity is performed under various postural conditions to investigate the changes in the sensitivity and specific mechanisms of spasticity. Fifty-nine multiple sclerosis patients participated in the study and received spasticity evaluation based on both the Ashworth scale and the pendulum test. Spasticity was found to increase in the pendulum test when the subjects were tested in a supine posture, compared to when they were sitting. Three patterns of stretch reflex response were seen for similar leg swing trajectories. While it was clear that the increased stretch of rectus femoris in the supine posture contributed to the increased spasticity, the results of modeling showed that other more complex factors were also involved. The supraspinal descending modulation associated with postural control may play a more dominant role in the severity of spasticity. The results suggest that the biomechanical test of spasticity should be performed for several different postures or ranges of movement with muscle activities monitored simulta- neously, so that the effect of various factors can be examined. The work also indicates that a neuromusculoskeletal model with detailed muscle dynamics and stretch reflex loops is a valuable tool for investigating the neural mechanisms of spasticity. Index Terms—Model, multiple sclerosis, muscle dynamics, sen- sory feedback, spasticity, stretch reflex. I. INTRODUCTION T HE spasticity of a patient is strongly affected by the intended activities and postures. It has been reported that the position of a patient during examination affects the evaluation of spasticity regardless of which quantitative measures are used [1]–[3]. This postural and task dependence of spasticity has raised two important questions. One is which postural position a patient should take to standardize spasticity evaluation. The other is how to interpret the difference in biomechanical characteristics and muscle responses due to postural changes and whether this information can help us understand the mechanism of spasticity for each individual patient. The first question relates to the effect of posture on biome- chanical properties of muscles and joints. When spasticity is evaluated using the pendulum test, a subject is in either the supine or the erect sitting position [4]–[6]. In the supine position, rectus femoris muscle is stretched and its static length is greater while the hamstring muscle group is relaxed, applying an increased extension force at the knee joint. When Manuscript received August 10, 1997; revised January 8, 1998 and February 14, 1998. This work was supported in part by a grant for Biomedical Engineering Research from the Whitaker Foundation. The author is with the Department of Chemical, Bio and Materials Engi- neering, Arizona State University, Tempe, AZ 85287 USA. Publisher Item Identifier S 1063-6528(98)03600-3. using imposed ankle joint flexion-extension to evaluate spas- ticity, the postural effect is represented by the stretching of gastrocnemius at the knee. Both muscle and joint viscoelastic properties will be affected. They will affect spasticity [3], [7]–[13]. The second question relates to interpretations of postural and task-dependent effects in underlying mechanisms of spasticity [14]. The starting lengths and the ranges of motion of several muscles may change when the testing posture is adjusted, affecting both the static and dynamic thresholds of a muscle’s stretch reflex [14], [15]. Stretch reflex sensitivity is also heavily modulated by various descending neural pathways that have strong task dependence. The demand for postural control is weaker in the supine posture than in the erect sitting or standing postures because the body center of mass is closer to the supporting surface in addition to a larger area. This may affect the descending control command to the spinal cord circuits. Since the effect of each pathway on spasticity may display different characteristics in stretch reflex response, a systematic analysis of variation patterns in stretch reflex response due to postural or task changes may reveal clinically useful information [14], [18]–[21]. To investigate the complex interaction between the factors influencing spasticity, several mathematical models have been developed over last three decades [5], [15], [22]. These models have different complexities and emphasize different aspects of the neuromusculoskeletal system. By combining model simulation and experimental approaches, it has allowed the investigation of various hypothesized mechanisms of spasticity [15]. In this paper, we report some results of investigating the two questions posted above using combined experiment and model simulation. From the results, we argue that when the pendulum test is performed in both the erect sitting and supine postures, the examination of muscle activities due to stretch reflex provides information on possible neural mechanisms of spasticity. The hypotheses formulated on these neural mechanisms could be examined by computer model simulations when both knee joint trajectory and muscle reflex activity are available. II. METHODS A. Subjects and the Evaluation of Spasticity Fifty-nine multiple sclerosis (MS) patients between the ages of 26 and 68 were evaluated for spasticity in this study. Twenty-three of the subjects were male and 36 female. Every subject signed a consent form for the study approved by the 1063–6528/98$10.00  1998 IEEE