**5. Biomechanical quantification of parkinsonian rigidity**

In clinic, parkinsonian rigidity is examined and assessed using a numerical rating scale which is known as the Unified Parkinson Disease Rating Scale (Fahn & Elton, 1987; Goetz et al., 2008). However, the nature of this assessment tool is highly qualitative and subjective because it is largely dependent on examiners' individual interpretation and experience (Patrick et al., 2001; Prochazka et al., 1997). When the actual change in rigidity resulting from treatment is small, it may be challenging for the examiners to detect. This can limit ability for evaluation of treatment effectiveness especially in large multi-center clinical drug trials in which a large number of investigators are involved, because differences can exist between different examiners (i.e., inter-rater) and between assessments performed on different visits by a given examiner (i.e., intra-rater) with respect to the efficacy of treatment. Reliability studies have demonstrated varying degrees of inter-rater reliability with respect to rigidity component of clinical rating tools, ranging from low, moderate, very good to excellent (Martinez-Martin, 1993; Rabey et al., 1997; Richards et al., 1994; Van Dillen & Roach, 1988). A need for more accurate evaluations has been expressed to improve the management of symptoms in patients with Parkinson's disease (Obeso et al., 1996; Ondo et al., 1998; Ward et al., 1983).

During the past several decades, considerable efforts have been made aiming to quantify assessment of parkinsonian rigidity by means of biomechanical measures. A variety of quantitative methods have been developed to measure the dynamics of joint stiffness associated with rigidity (Lee et al., 2002; Prochazka et al., 1997; Teräväinen et al., 1989; Watts et al., 1986; Wiegner & Watts, 1986). The underlying approach is to measure the amount of imposed force resistance to externally generated passive movement about the examined joint. The passive movements applied in earlier studies were induced either by a torque motor (Fung et al., 2000; Mak et al., 2007; Shapiro et al., 2007; Watts et al., 1986; Xia et al., 2006) or generated by an examiner to closely resemble a clinical setting (Caligiuri, 1994; Endo et al., 2009; Patrick et al., 2001; Prochazka et al., 1997; Sepehri et al., 2007). Variables described in these previous studies included peak torque (Mak et al., 2007), impulse (i.e., an integral of torque with respect to time; Fung et al., 2000), work score which is calculated as a torque integral with respect to joint angular position (see Fig. 3; Fung et al., 2000; Mak et al., 2007; Shapiro et al., 2007; Teräväinen et al., 1989; Xia et al., 2006, 2009), elastic coefficient (Endo et al., 2009), and mechanical impedance calculated based on the force imposed and displacement of the movement (Patrick et al., 2001; Prochazka et al., 1997).

There are a few advantages of quantification by force or torque measures over quantification by EMG. Torque-based assessment of rigidity is more objective and reliable than EMGderived evaluation. However, there are limitations in estimation of using surface EMGs as its measures are susceptible to the placement of electrodes, condition of soft tissues and concerns of cross-talk. Biomechanical measures using torque can avoid the limitations inherent in EMG measures. In addition, non-neural contribution to parkinsonian rigidity is also included in torque measures but is not reflected in EMG recordings. Previous studies have shown that correlation is relatively weak between clinical degree of rigidity and EMG quantification of rigidity while correlation is found to be much stronger between clinical degree of rigidity and torque quantification of rigidity (Endo et al., 2009; Levin et al., 2009;

Physiological and Biomechanical Analyses of Rigidity in Parkinson's Disease 495

biomechanical measures have demonstrated that rigidity associated with extension movement is more evident as compared to rigidity during flexion movement in the upper limb including the wrist and elbow (Mera et al., 2009; Park et al., 2010; Xia et al., 2006). These authors used mechanical parameters, such as work, torque-angle slope and visco-elastic parameter, to evaluate the difference and distinction between the passive flexion and

Thirdly, parkinsonian rigidity has traditionally been considered to be independent of velocity in contrast to spasticity which is highly velocity-dependent (Lance, 1980). The notion of velocity-independency of rigidity might be anecdotal. As such, this view has been examined by a few recent studies. Lee et al. (2002) studied hypertonia at the elbow joint in patients with Parkinson's disease and patients with hemiparesis as compared to control subjects. Four different stretching velocities were applied, ranging from 40 to 160 °/s. The authors concluded that both rigidity and spasticity have approximately equal velocitydependent property. Quantitative measure of trunk rigidity in patients with Parkinson's disease also revealed a velocity-dependent feature (Mak et al., 2007). Our results on the effect of movement velocity on rigidity concurred with those reports. Velocity-dependency of rigidity was also demonstrated at the wrist joint of patients with Parkinson's disease (Xia et al., 2009) in which both slow velocity at 50 degree/second and fast speed at 280 °/s were applied. The results showed that the work done during the fast movement was significantly larger than the work associated with the slower movement. The accumulating evidence has

Fourthly, compared to the effect of movement velocity on quantitative analysis of parkinsonian rigidity, effect of displacement amplitude on rigidity has thus far sparsely been investigated, except for one study by Teräväinen et al. (1989). To determine the optimal angular velocity and displacement amplitude for detecting abnormal muscle tone, four movement amplitudes or central ranges of motion, ranging from ±15, ±20, ±25 to ±30 degrees, were applied to examine rigidity at the wrist joint in 29 patients with Parkinson's disease. The results showed that the larger movement amplitudes were more sensitive for detecting parkinsonian rigidity and had stronger correlation with the clinical scores of rigidity. Given the situation that some clinicians rotate the limb back and forth rapidly in the mid-range whereas others focus on the extremes of range of motion or the entire range of motion with slow stretches (Prochazka et al., 1997), it is important and significant to explore the influence of displacement amplitude on objective measurements of rigidity. We recently conducted a study aiming to examine the effect of displacement amplitude. Twenty four patients participated in the experiment under treated (On-medication) and untreated (Off-medication) conditions, with the more affected side of the wrist joint tested. Passive movements of wrist flexion and extension were imposed with two displacement amplitudes, ±30 degree and ±45 degree, respectively, at either 50 °/s or 280 °/s, and the order of movement pattern was presented in a random fashion. Figure 4 depicts and compares the torque-angle plots associated with two ranges of motion: 60 degree (Fig. 4A) and 90 degree (Fig. 4B), in a parkinsonian subject under the two medication conditions. The work score was calculated to quantify rigidity, and was normalized to the range of motion to validate the comparison. Clearly, there is a difference in the area of the torque-angle loop between the two displacement amplitudes or the ranges of motion. Figure 4B shows that the extreme joint position, the larger displacement amplitude,

pointed out the velocity-dependency of parkinsonian rigidity.

caused increase in rigidity work score.

extension movements.

Park et al., 2010; Teräväinen et al., 1989). Evidence indicates that torque measure has proven to be a more objective and robust way for assessing rigidity, compared to EMG evaluation of rigidity.

Fig. 3. Torque-angle relationship in a parkinsonian subject in the Off-medication (thicker line) and On-medication (thinner line) states. The subject's more affected side was tested. The upper traces represent imposed extension movements and the lower traces flexion movements. The wrist joint was externally rotated at 50 °/s between 30° flexion and 30° extension shown as a loop. The subject was instructed to remain relaxed. The work, used to quantify the degree of rigidity, was equivalent to the areas inside the loop of torque-angle plots in the respective medication states [from Xia et al. (2006) with permission].

Application of biomechanical measures has also enabled us to investigate more profoundly some of the characteristics associated with parkinsonian rigidity, thus further increasing our understanding of this motor symptom. Firstly, only through the measures of torque resistance and joint position, can lead-pipe nature of rigidity be examined and revealed (Mera et al., 2009; Xia & Rymer, 2004; Xia et al., 2006, 2011). The slope of torque-angle curve was used to quantify the degree of lead-pipe property. The smaller slopes represent higher degrees of constant and uniform resistance through the range of passive movement. The torque-angle slopes are smaller when patients were tested in the untreated conditions, and become greater in the treated conditions (Xia & Rymer, 2004; Xia et al., 2006).

Secondly, rigidity has been thought to be plastic with respect to direction of the movement (Berardelli et al., 1983; Delwaide, 2001). However, recent studies employing the

Park et al., 2010; Teräväinen et al., 1989). Evidence indicates that torque measure has proven to be a more objective and robust way for assessing rigidity, compared to EMG evaluation

Fig. 3. Torque-angle relationship in a parkinsonian subject in the Off-medication (thicker line) and On-medication (thinner line) states. The subject's more affected side was tested. The upper traces represent imposed extension movements and the lower traces flexion movements. The wrist joint was externally rotated at 50 °/s between 30° flexion and 30° extension shown as a loop. The subject was instructed to remain relaxed. The work, used to quantify the degree of rigidity, was equivalent to the areas inside the loop of torque-angle

Application of biomechanical measures has also enabled us to investigate more profoundly some of the characteristics associated with parkinsonian rigidity, thus further increasing our understanding of this motor symptom. Firstly, only through the measures of torque resistance and joint position, can lead-pipe nature of rigidity be examined and revealed (Mera et al., 2009; Xia & Rymer, 2004; Xia et al., 2006, 2011). The slope of torque-angle curve was used to quantify the degree of lead-pipe property. The smaller slopes represent higher degrees of constant and uniform resistance through the range of passive movement. The torque-angle slopes are smaller when patients were tested in the untreated conditions, and

Secondly, rigidity has been thought to be plastic with respect to direction of the movement (Berardelli et al., 1983; Delwaide, 2001). However, recent studies employing the

plots in the respective medication states [from Xia et al. (2006) with permission].

become greater in the treated conditions (Xia & Rymer, 2004; Xia et al., 2006).

of rigidity.

biomechanical measures have demonstrated that rigidity associated with extension movement is more evident as compared to rigidity during flexion movement in the upper limb including the wrist and elbow (Mera et al., 2009; Park et al., 2010; Xia et al., 2006). These authors used mechanical parameters, such as work, torque-angle slope and visco-elastic parameter, to evaluate the difference and distinction between the passive flexion and extension movements.

Thirdly, parkinsonian rigidity has traditionally been considered to be independent of velocity in contrast to spasticity which is highly velocity-dependent (Lance, 1980). The notion of velocity-independency of rigidity might be anecdotal. As such, this view has been examined by a few recent studies. Lee et al. (2002) studied hypertonia at the elbow joint in patients with Parkinson's disease and patients with hemiparesis as compared to control subjects. Four different stretching velocities were applied, ranging from 40 to 160 °/s. The authors concluded that both rigidity and spasticity have approximately equal velocitydependent property. Quantitative measure of trunk rigidity in patients with Parkinson's disease also revealed a velocity-dependent feature (Mak et al., 2007). Our results on the effect of movement velocity on rigidity concurred with those reports. Velocity-dependency of rigidity was also demonstrated at the wrist joint of patients with Parkinson's disease (Xia et al., 2009) in which both slow velocity at 50 degree/second and fast speed at 280 °/s were applied. The results showed that the work done during the fast movement was significantly larger than the work associated with the slower movement. The accumulating evidence has pointed out the velocity-dependency of parkinsonian rigidity.

Fourthly, compared to the effect of movement velocity on quantitative analysis of parkinsonian rigidity, effect of displacement amplitude on rigidity has thus far sparsely been investigated, except for one study by Teräväinen et al. (1989). To determine the optimal angular velocity and displacement amplitude for detecting abnormal muscle tone, four movement amplitudes or central ranges of motion, ranging from ±15, ±20, ±25 to ±30 degrees, were applied to examine rigidity at the wrist joint in 29 patients with Parkinson's disease. The results showed that the larger movement amplitudes were more sensitive for detecting parkinsonian rigidity and had stronger correlation with the clinical scores of rigidity. Given the situation that some clinicians rotate the limb back and forth rapidly in the mid-range whereas others focus on the extremes of range of motion or the entire range of motion with slow stretches (Prochazka et al., 1997), it is important and significant to explore the influence of displacement amplitude on objective measurements of rigidity. We recently conducted a study aiming to examine the effect of displacement amplitude. Twenty four patients participated in the experiment under treated (On-medication) and untreated (Off-medication) conditions, with the more affected side of the wrist joint tested. Passive movements of wrist flexion and extension were imposed with two displacement amplitudes, ±30 degree and ±45 degree, respectively, at either 50 °/s or 280 °/s, and the order of movement pattern was presented in a random fashion. Figure 4 depicts and compares the torque-angle plots associated with two ranges of motion: 60 degree (Fig. 4A) and 90 degree (Fig. 4B), in a parkinsonian subject under the two medication conditions. The work score was calculated to quantify rigidity, and was normalized to the range of motion to validate the comparison. Clearly, there is a difference in the area of the torque-angle loop between the two displacement amplitudes or the ranges of motion. Figure 4B shows that the extreme joint position, the larger displacement amplitude, caused increase in rigidity work score.

Physiological and Biomechanical Analyses of Rigidity in Parkinson's Disease 497

Finally, the phenomenon that rigidity can be reinforced by a concurrent ipsi- or contralateral voluntary activation has recently been further quantified using biomechanical measures (Hong et al., 2007; Powell et al., 2011). Figure 5 illustrates torque-angle traces of the entire cycle of flexion and extension movements when a subject with PD was tested in the Off-medication condition. Torque resistance was elevated by the presence of contralateral activation (Active condition) as compared to the Passive condition. There is an obvious difference in the contained area of torque-angle plots between the Passive and Active conditions. These studies aimed to provide quantitative data and objective evaluation of clinical assessment of rigidity as a component of the Unified Parkinson Disease Rating Scale (Fahn & Elton, 1987; Goetz et al., 2008). The type of voluntary activations applied in clinical examinations include a variety of motor acts such as tapping fingers, fist opening–closing or heel tapping. The use of reinforcing maneuvers was originated and first investigated by Jules Froment, a French neurologist, in the 1920's (Broussolle et al., 2007). Froment studied muscle tone at the wrist joint while the subject was in different positions, at rest in a sitting position, and standing in stable and unstable postures. In addition to clinical examination, Fremont also conducted experiments recording activity of forearm extensors using a myograph. He described an increased resistance to passive movements of a limb about a joint during the presence of a voluntary action of a contralateral body part. Due to his contributions to the study of parkinsonian rigidity, the activation or facilitation test has been referred to as the "Froment maneuver". The impact of facilitation test is significant as it has been formalized in the motor scale of the Unified Parkinson Disease Rating Scale. The maneuver is particularly used to detect increased muscle tone at an early stage of the disease when rigidity is not otherwise manifested

**6. Effect of anti-Parkinson medication on physiological and biomechanical** 

Rigidity generally responds well to anti-Parkinson medication. Several studies have examined the changes in muscle activation, joint torque resistance and torque-angle slope associated with rigidity reduction as a result of medication therapy (Kirollos et al., 1996; Mera et al., 2009; Powell et al., 2011; Xia & Rymer, 2004; Xia et al., 2006, 2009). Following a standard protocol, patients are tested initially in the Off-medication state, *i.e.*, 12 hours after the last dose of medication when the majority of the beneficial effects of medication therapy are eliminated (Defer et al., 1999). Twelve-hour overnight withdrawal of medication has been broadly used to examine the effect of medication on motor performance and on basal ganglia function (Brown & Marsden, 1999; Corcos et al., 1996; Jahanshahi et al., 2010; Robichaud et al., 2004; Tunik et al., 2004). After the initial tests are completed, patients are retested approximately one hour after taking their regular dose of medication in the Onmedication state. These studies have demonstrated that stretch-reflex and shortening reaction are diminished following the treatment (Powell et al., 2011; Xia & Rymer, 2004). The same effects are observed in the changes associated with torque resistance (Kirollos et al., 1996; Mera et al., 2009; Xia et al., 2006, 2009). Further, torque-angle curves associated with the On-medication test become steeper, manifesting the spring-like feature and the typical length-tension relationship (Gordon et al., 1966; Matthews, 1959; Rack & Westbury, 1969;

during the examination.

**measures of rigidity** 

Xia et al., 2006, see Fig. 2).

Fig. 4. Comparison of torque-position traces of passive flexion and extension movements of two ranges of motion: 60° (A) and 90° (B) at angular velocity of 50° /s from a subject with Parkinson's disease tested in the Off-medication state. The rigidity score, quantified by the integral of the torque with respect to angular position (Nm-deg), increased in response to the greater range of motion. Upper traces are associated with extension movements while lower traces are associated with flexion movements [from Powell et al. (in press) with permission].

Fig. 4. Comparison of torque-position traces of passive flexion and extension movements of two ranges of motion: 60° (A) and 90° (B) at angular velocity of 50° /s from a subject with Parkinson's disease tested in the Off-medication state. The rigidity score, quantified by the integral of the torque with respect to angular position (Nm-deg), increased in response to the greater range of motion. Upper traces are associated with extension movements while lower traces are associated with flexion movements [from Powell et al.

(in press) with permission].

Finally, the phenomenon that rigidity can be reinforced by a concurrent ipsi- or contralateral voluntary activation has recently been further quantified using biomechanical measures (Hong et al., 2007; Powell et al., 2011). Figure 5 illustrates torque-angle traces of the entire cycle of flexion and extension movements when a subject with PD was tested in the Off-medication condition. Torque resistance was elevated by the presence of contralateral activation (Active condition) as compared to the Passive condition. There is an obvious difference in the contained area of torque-angle plots between the Passive and Active conditions. These studies aimed to provide quantitative data and objective evaluation of clinical assessment of rigidity as a component of the Unified Parkinson Disease Rating Scale (Fahn & Elton, 1987; Goetz et al., 2008). The type of voluntary activations applied in clinical examinations include a variety of motor acts such as tapping fingers, fist opening–closing or heel tapping. The use of reinforcing maneuvers was originated and first investigated by Jules Froment, a French neurologist, in the 1920's (Broussolle et al., 2007). Froment studied muscle tone at the wrist joint while the subject was in different positions, at rest in a sitting position, and standing in stable and unstable postures. In addition to clinical examination, Fremont also conducted experiments recording activity of forearm extensors using a myograph. He described an increased resistance to passive movements of a limb about a joint during the presence of a voluntary action of a contralateral body part. Due to his contributions to the study of parkinsonian rigidity, the activation or facilitation test has been referred to as the "Froment maneuver". The impact of facilitation test is significant as it has been formalized in the motor scale of the Unified Parkinson Disease Rating Scale. The maneuver is particularly used to detect increased muscle tone at an early stage of the disease when rigidity is not otherwise manifested during the examination.
