Movement Disorders Vol. 23, No. 9, 2008, pp.

12621268 2008 Movement Disorder Society

Quantitative Measures of Fine Motor, Limb, and Postural Bradykinesia in Very Early Stage, Untreated Parkinsons Disease
Mandy Miller Koop, MS,1 Nicole Shivitz, MS,2 and Helen Bronte-Stewart, MD, MSE2,3*
2

Department of Mechanical Engineering, Stanford University, Stanford, California, USA Department of Neurology and Neurological Sciences, Stanford University, Stanford, California, USA 3 Department of Neurosurgery, Stanford University, Stanford, California, USA

Abstract: Few studies have characterized the motor control abnormalities of very early stage Parkinsons disease (PD), when symptoms are mild and usually unilateral. However, this group is the most targeted for potential disease-modifying therapeutics. We have validated several quantitative measures of bradykinesia with the Unied Parkinsons Disease Rating Scale motor disability score (UPDRS III) and have found these useful in studies of advanced PD. In this study, we asked if quantitative measures of nger, forearm, and postural movement velocity could detect bradykinesia in 20 patients with very early stage, untreated PD. The results revealed evidence of signicant nger and forearm bradykinesia of the patient groups more affected side when compared to the nondominant side of 19 age-matched control subjects (P 5 0.001 and P < 0.001, respectively). Furthermore, the patient groups forearm movement velocity on the more affected side was

signicantly slower than their less affected side (P 5 0.005), highlighting the importance of using an outcome measure that is lateralized in studies of very early stage PD. In contrast to our previous study that revealed signicant postural bradykinesia in patients with advanced PD, we did not detect postural bradykinesia in patients with very early stage, untreated PD. Based on these ndings, we suggest that the use of quantitative, lateralized measures of bradykinesia would be useful in studies of very early stage, untreated PD. These measures may improve a study by: increasing efciency and objectivity of the evaluation, decreasing cost, and decreasing the number of subjects needed for statistical signicance. 2008 Movement Disorder Society Key words: early stage Parkinsons disease; bradykinesia; motor control; quantitative digitography; postural instability; dynamic posturography

The diagnosis of Parkinsons disease (PD) relies on the clinical observation of two of three motor signs: tremor, rigidity, and/or bradykinesia, which manifest after the pathology of the disease is well established.1 Currently there is no cure for PD, but a number of clinical trials are assessing potential disease-modifying therapeutics that may intervene and slow the progression of the disease.25 A recent study concluded that the early, untreated PD group is the most targeted

*Correspondence to: Dr. Helen M. Bronte-Stewart, Department of Neurology and Neurosciences Institute, Stanford University, 300 Pasteur Dr, Rm A343, Stanford, CA 94305-5327 E-mail: hbs@stanford.edu Received 23 October 2007; Revised 6 March 2008; Accepted 18 March 2008 Published online 7 May 2008 in Wiley InterScience (www. interscience.wiley.com). DOI: 10.1002/mds.22077

group for these trials.6 However, these trials face two main challenges. First, recruiting such patients is difcult. Typically, the onset of movement abnormalities is gradual, and experienced on only one side of the body. As such, patients may not seek medical attention until they are in need of symptomatic treatment, which then disqualies them from these clinical trials. Second, many trials assess potential putative neuroprotective therapeutics by tracking the progression of motor signs and/or the need for introduction of symptomatic therapy. Motor signs are traditionally measured with the Unied Parkinsons Disease Rating Scale motor disability score (UPDRS III). The UPDRS III is a noncontinuous, coarse-grained scale that may have limited resolution for detecting small changes in disease progression in very early stage PD. Therefore, a large number of subjects are needed to adequately power the

1262

BRADYKINESIA IN EARLY, UNTREATED PD statistical tests, and the assessment of performance can be subjective. Our group has supplemented the UPDRS III with quantitative measures of motor control that are sensitive to improvements from both medication and Deep Brain Stimulation (DBS).711 These measures included quantication of bradykinesia, which has been shown to be a good indicator of overall motor disability when measured by repetitive nger tapping.12 We measured movement velocity during a repetitive ne motor task, upper limb task, and during a discrete whole body weight shifting task while standing. These tools are very useful although there are others emerging as we detailed in a previous publication.9 In this study, we have turned our attention to a group of patients with very early stage, untreated PD to investigate whether our quantitative measures could detect bradykinesia in this patient group. We evaluated a group of 20 early stage, untreated PD patients with quantitative measures of nger, forearm, and postural movement velocity, in conjunction with the UPDRS III. We hypothesized that quantitative measures of movement velocity could detect ne motor, limb, and/ or postural bradykinesia.

1263

subsequently had a positive response to dopamine replacement therapy and none have exhibited signs that suggest atypical Parkinsonian syndromes. Control subjects were recruited from the Stanford University campus. The age-matched control group consisted of 19 subjects (6 men and 13 women, mean age 55.5 6 11.1 years). Control subjects completed a comprehensive questionnaire that screened for neurological and vestibular disease. Study Protocol All subjects gave informed consent to participate in this study, which was approved by the Stanford University Institutional Review Board. The subjects were assessed in the Stanford Human Motor Control and Balance Laboratory using quantitative measures of nger movements with Quantitative Digitography (QDG), forearm movements with angular velocity sensors, and postural movements with computerized dynamic posturography (CDP). These quantitative systems have been described previously and are summarized below.7,911 In all assessments, prior to testing, the movement tasks were demonstrated, and the subjects performed a few cycles of the movement in order to minimize short-term practice effects. This study focused on a comparison of nger, forearm, and postural movement velocities of the nondominant (ND) side of control subjects, to the MA and LA sides of the PD subjects. Quantitative Digitography Using a Musical Interface Digital (MIDI) Keyboard, we have shown that the measurement of mean keystrike velocity (Vel), during a repetitive alternating nger tapping (RAFT) task is a sensitive measure of bradykinesia in PD.9 Vel is recorded as an integer in the range 1 to 127, and it is a monotonically increasing function of keystrike velocity.9 Subjects were tested in a seated position and instructed to perform RAFT as fast as possible with their index and middle ngers for 30 seconds, rst with their right hand and then with their left. Auditory feedback was provided through headphones and subjects were instructed to watch and listen as they performed RAFT. Movements were selfpaced. Details of data acquisition and of the MIDI keyboard apparatus have been described previously.7,9 Quantitative Repetitive Wrist Pronation-Supination A motion analysis system (Motus Bioengineering Inc, Benicia, CA) was used to measure the angular velocity of forearm movement during a repetitive wrist

METHODS Subjects Twenty patients (16 men and 4 women, mean age 60.0 6 10.0 years) with very early stage, untreated PD met the following inclusion criteria: (1) a diagnosis of PD and (2) no prior use of any dopaminergic medication. The diagnosis was made at the time of evaluation in 14 of the 20 subjects. The other six had been diagnosed within 1.5 years prior to their evaluation. Detailed questioning established that all patients had less than a two-year history of symptoms. All patients stated that their symptoms were present on one side or were much more severe on one side of their body (more affected, MA) compared to the other side (less affected, LA). The patient groups mean UPDRS III score was 18.1, and their mean lateralized bradykinesia score (UPDRS III Items 2326) was MA-5.0 and LA2.0. No patient had requested or had been advised to start any dopaminergic medication. All patients had a comprehensive history and neurological examination, which excluded any other focal neurological decits or dementia. Seven of the 20 patients had a uorodopa PET scan that supported the diagnosis of PD, and 14 of the 20 have been subsequently followed for an average of 1.8 years (0.184.3 years). All 14 patients have

Movement Disorders, Vol. 23, No. 9, 2008

1264

M.M. KOOP ET AL. ing qrWPS was calculated over the 30-second trial for each hand. As documented in a previous study, angular velocity data were screened to conrm that movements were not dominated by an action tremor.11 For a measure of postural bradykinesia, CoG MV scores were averaged across eight directions to create a mean CoG MV for each subject. Statistical Analysis A one-way ANOVA with a subsequent Tukey Test was used to determine differences between the control group (ND side) and the PD patient group (MA and LA sides) for ne motor and limb movement tasks (QDG and qrWPS) A Students t-test was used to determine differences in postural MV between the control group and the patient group. A Mann-Whitney Rank Sum Test was used to determine signicance between the PD patient groups MA and LA UPDRS III scores. RESULTS Fine Motor Bradykinesia Figure 1 shows the performance of 30 seconds of RAFT by a control subjects non-dominant hand (Fig. 1A), and a patients LA side, and MA side (Fig. 1B,C, respectively). The patient was a 45-year-old man with a 14-month history of PD symptoms who was diagnosed 11 months prior to the evaluation. His symptom complaints were limited to his left side (MA side). A clinical exam revealed left-sided rigidity and bradykinesia, and mild postural tremor on the right and left upper extremities. UPDRS III nger tapping (FT) subscores (Item 23) were 0 for the LA and 2 for the MA hand. However, in contrast to the normal UPDRS III FT score for his LA hand, Figure 1B reveals that the performance of RAFT with his LA hand was impaired compared to the control ND hand (Fig. 1A). The performance was more irregular and the Mean Vel was lower. The performance of his MA hand (Fig. 1C) was much slower than either his LA hand or the control subject. When the group of early stage PD patients was compared to age-matched controls there was a signicant reduction in Mean Vel in the MA hand of the patients compared to the control groups ND hand (P 5 0.001), as shown in Figure 2. The performance of the LA side appeared to be better than that of the MA side, but worse than that of the controls ND side, although neither difference reached

pronationsupination task. We have shown that the root mean square of the forearm angular velocity data (Vrms) is a sensitive measure of bradykinesia in PD.10 Subjects were tested seated with their arm elevated and elbow exed at 908, with an angular velocity sensor attached to the dorsum of their hand. Patients were instructed to perform qrWPS as fast as possible while rotating their wrist all the way from pronation to supination for 30 seconds. Movements were self-paced. Details of data acquisition and the Motus motion analysis system have been described previously.10 Computerized Dynamic Posturography Computerized dynamic posturography (CDP) (NeuroCom, Clackamas, OR) is a quantitative method for assessing upright balance.13 Subjects stood on a force plate that measured vertical forces. The height of the subject and the vertical forces were used to calculate the center of gravity (CoG) location.13 CoG velocity was measured by requiring the subjects to make voluntary movements by leaning their head, trunk, arms, and legs as one axis about their ankles in different directions in response to a cue. Their goal was to lean in one direction until a cursor, visible on a computer screen at eye level, aligned inside a target. Subjects were instructed to lean towards the target on the screen as quickly and as accurately as possible. Trials where subjects used other movement strategies to accomplish the movement task, such as anterior exion over a stable pelvis, were stopped and subjects were reinstructed and asked to perform the task again. Each subject performed eight, 8-second trials, leaning once in the forward, backward, left, right, and the four diagonal directions. The value of the CoG angle (deg) was the angular displacement of CoG location from a vertical axis passing through the center of foot support.13 CoG movement velocity (MV, degree/second) was the CoG displacement measured between 5 and 95% of the angle at which the subjects movement had stopped (reached zero velocity) or diverged from the target, divided by the time interval. Calculations of CoG MV from vertical force plate measurements assume that dynamical force contributions are negligible, an assumption warranted in this slow moving task. Data Analysis For assessment of ne motor control, the keystrike velocity during RAFT was measured for each of the two ngers used (index/middle) and averaged together to produce a mean velocity measurement (Mean Vel) for each hand. For limb motor control, the (Vrms) dur-

Movement Disorders, Vol. 23, No. 9, 2008

BRADYKINESIA IN EARLY, UNTREATED PD

1265

patients LA and MA forearms (Fig. 3B,C, respectively). The patient was a 72-year-old, left-handed man diagnosed with PD 8 months prior to the evaluation, with symptoms of right-sided (MA) bradykinesia and mild resting tremor for less than 2 years. The patient did not report symptoms on his left side. His UPDRS III repetitive wrist pronation-supination scores (Item 25) were 0 and 2 on LA and MA sides, respectively. Comparison of Figure 3(A,B) supports the UPDRS III (Item 25) ndings and suggests that there was little difference between the performance of the controls ND and the patients LA forearm, and Figure 3C shows the performance of the patients MA forearm (Fig. 3C) was much slower compared to both the controls ND and patients LA sides. Figure 4 compares the mean Vrms values of the control groups ND side and the PD patient groups MA and LA sides. The Vrms of the patient groups LA side was very similar to that of the control subjects (P > 0.05). The patient groups Vrms was signicantly worse on the MA side compared to that of the control groups ND side (P < 0.001). Also, the patient groups MA Vrms was signicantly worse than that of their LA (P 5 0.005), which was supported by their mean UPDRS III Item 25 score that also showed a signicant difference between the MA and LA sides (1.3 and 0.5, respectively, P 5 0.015).

FIG. 1. Example trace of repetitive alternating nger tapping task (QDG) for 30 seconds from (A) an age-matched control subjects ND hand, (B) very early stage PD patients LA side, (C) patients MA side. The upper note velocity is depicted on the upper y-axis and the lower note velocity on the lower y-axis (which is positive in both directions), and the x-axis depicts time in seconds.

signicance (P 5 0.076, P 5 0.245, respectively). The UPDRS III FT score for the PD patient group supported this nding, and did not nd a signicant difference between the average FT score for the LA (0.7) and MA (1.6) sides (P 5 0.337). Limb Bradykinesia Figure 3 shows the angular velocity traces from a control subjects ND forearm (Fig. 3A), and from a

FIG. 2. Very early stage PD patient group showed a signicant impairment in mean nger MV in the MA side (48.7 6 23.9) compared to the control groups ND hand (76.7 6 18.0, P 5 0.001). Mean nger MV for the patient groups LA side (64.8 6 25.9) showed a nonsignicant trend of being worse than the control groups ND hand (P 5 0.245). Bar graphs with error bars show mean velocity scores and 6 one standard deviation.

Movement Disorders, Vol. 23, No. 9, 2008

1266

M.M. KOOP ET AL. control groups ND side (data not shown, P > 0.05). Unlike the other quantitative measures, the postural movement task does not have a direct corresponding test in the UPDRS III. However, the UPDRS III does assess voluntary postural movements by rating a patients ability to rise from a seated position to a standing position (Item 32). The UPDRS III sit to stand test showed no voluntary postural impairments in 19 of the 20 patients (score 5 0). The remaining patient received a score of one. Thus, in early stage, untreated PD there appears to be no signicant postural bradykinesia as measured quantitatively by this postural movement task.

DISCUSSION In a group of patients with very early stage, untreated PD, quantitative measures of ne, limb, and postural movement velocity revealed signicant ne motor and limb bradykinesia in the clinically MA upper extremity of patients with respect to the ND side of age-matched controls. However, postural movement velocity was not measurably impaired. All patients had a disease duration of less than 1.5 years, and 70% were diagnosed at the time of the evaluation. This suggests that in very early stages of PD, nger and forearm bradykinesia are evident, but postural movement velocity is not yet affected. In contrast, we have shown

FIG. 3. Example trace of forearm angular velocity of repetitive pronation-supination for 30 seconds from (A) an age-matched control subjects ND hand, (B) very early stage PD patients LA side, (C) patients MA side. The y-axis shows the forearm angular movement velocity in degrees/second (which is positive for pronating movements and negative for supinating movements), and the x-axis shows time in seconds.

Postural Bradykinesia A different outcome was seen for postural bradykinesia. Figure 5A shows that the mean CoG MV of the PD patient group was not signicantly different from that of control subjects (P 5 0.208). When the movement velocities between groups were compared in each direction there was again no signicant difference between patients and control subjects (P > 0.05 without correcting for multiple comparisons, Fig. 5B). Even when separating the patient group into MA and LA sides, there was no signicant difference in their postural movement velocity compared to the

FIG. 4. Very early stage PD patient group showed a signicant impairment in mean forearm movement velocity in the MA side (431.8 6 191.48/second) compared to the age-matched control groups ND forearm (707.6 6 130.48/second, P < 0.001). However, the patient groups LA side (603.2 6 169.08/second) was similar to the mean Vrms of the ND forearm of the control group (P > 0.05), and was signicantly better than the mean Vrms of the patients affected forearm (P 5 0.005). Bar graphs with error bars show mean velocity scores and 6 one standard deviation.

Movement Disorders, Vol. 23, No. 9, 2008

BRADYKINESIA IN EARLY, UNTREATED PD

1267

FIG. 5. Postural bradykinesia is not evident in the very early stage PD patient group. (A) Patient groups mean postural movement velocity was not signicantly different compared to control group (2.9 6 1.28/second, 3.4 6 1.48/second, P > 0.208). (B) Directional analysis: mean postural movement velocities separated into four cardinal directions, which revealed no statistical difference between the very early stage PD patient group and the age-matched controls (P > 0.05, not corrected for multiple comparisons). Bar graphs with error bars show mean velocity scores and 6 one standard deviation.

that patients in more advanced stages of PD (mean duration from diagnosis of 8.9 years) exhibited a signicantly slower postural movement velocity off medication than normal controls.11 Thus postural movements may only become bradykinetic in later stages of disease. This does not mean that other aspects of postural instability, such as postural reaction time and the sensory organization process involved in postural control, are not present earlier.8,11,14 For instance, one study found evidence of postural instability after a retropulsion pull in almost 41% of PD patients who were examined within 3 years of diagnosis.15 However, this was much more prevalent in patients older than 70 years of age and was not corrected for other co-morbidities.

There was a signicant difference between the PD patient groups LA and MA sides in quantitative measures of forearm bradykinesia (Vrms) and UPDRS III (Item 23), but not between the MA and LA sides in quantitative measures of nger bradykinesia (Mean Vel) or UPDRS III (Item 25). This nding could be explained by: (1) either the patient groups MA side showed more forearm bradykinesia resulting in a larger discrepancy between the MA and LA sides compared to nger bradykinesia, or (2) nger velocity in the patient groups LA side was beginning to show impairment before forearm velocity on the LA side, which resulted in a smaller difference between MA and LA sides in measures of nger bradykinesia. Previously, we examined RAFT in sensory deprived conditions as this has been shown to unmask bradykinesia in motor tasks in PD.7,16 However, since the postural movement task required sensory feedback to perform the test, we provided sensory feedback in the forearm and nger movement tasks to prevent introduction of bias in the results. A longitudinal study of these patients is ongoing to determine the rate of progression of ne versus limb bradykinesia. The statistical difference in the lateralized forearm bradykinesia does emphasize the importance of using a lateralized outcome measure in clinical trails of potential disease-modifying agents in early stage PD, so that the effect of the agent is not diluted. Finally, postural bradykinesia, as measured in this study, is recognized as a motor sign in advanced disease. Clinical trials aimed at slowing the disease progression could use the LA side and/or measures of postural bradykinesia as a marker of disease progression as neither are signicantly affected in very early stage untreated disease. Based on these ndings, integrating simple, lateralized, quantitative measures into clinical trials that study very early stage, untreated PD might reduce subjectivity in the outcome measure, reduce testing variability, and increase the number of times each patient could be tested. The tests might even be administered from home and data could then be streamed to a central location.17 It would be interesting to expand this small trial to larger numbers of newly diagnosed, or at risk populations to see if a simple RAFT task using Quantitative Digitography (QDG) and/or a qrWPS task using angular velocity sensors would be able to detect bradykinesia even before the patient was aware of the decit. It is important to note that while these quantitative tests separate the control group from the early, untreated Parkinsons disease patient group nicely, they are not sufciently sensitive and specic to be used on an individual basis for diagnostic purposes. However, we conclude

Movement Disorders, Vol. 23, No. 9, 2008

1268

M.M. KOOP ET AL.

8. Bronte-Stewart HM, Minn AY, Rodrigues K, Buckley EL, Nashner LM. Postural instability in idiopathic Parkinsons disease: the role of medication and unilateral pallidotomy. Brain 2002;125: 21002114. 9. Taylor Tavares AL, Jefferis GS, Koop M, et al. Quantitative measurements of alternating nger tapping in Parkinsons disease correlate with UPDRS motor disability and reveal the improvement in ne motor control from medication and deep brain stimulation. Mov Disord 2005;20:12861298. 10. Koop MM, Andrzejewski A, Hill BC, Heit G, Bronte-Stewart HM. Improvement in a quantitative measure of bradykinesia after microelectrode recording in patients with Parkinsons disease during deep brain stimulation surgery. Mov Disord 2006;21:673 678. 11. Shivitz N, Koop MM, Fahimi J, Heit G, Bronte-Stewart HM. Bilateral subthalamic nucleus deep brain stimulation improves certain aspects of postural control in Parkinsons disease, whereas medication does not. Mov Disord 2006;21:10881097. 12. Agostino R, Berardelli A, Curra A, Accornero N, Manfredi M. Clinical impairment of sequential nger movements in Parkinsons disease. Mov Disord 1998;13:418421. 13. Nashner LM. Computerized dynamic posturography. Chicago: Mosby-Year Book; 1993. 14. Horak FB, Nutt JG, Nashner LM. Postural inexibility in parkinsonian subjects. J Neurol Sci 1992;111:4658. 15. Kang GA, Bronstein JM, Masterman DL, Redelings M, Crum JA, Ritz B. Clinical characteristics in early Parkinsons disease in a central California population-based study. Mov Disord 2005;20:11331142. 16. Georgiou N, Iansek R, Bradshaw JL, Phillips JG, Mattingley JB, Bradshaw JA. An evaluation of the role of internal cues in the pathogenesis of parkinsonian hypokinesia. Brain 1993;116:1575 1587. 17. Goetz CG, Kubota K, Stebbins GT, et al. Parkinsons disease athome testing battery: reliability of data collection and transmission of objective motor data from home to a central study center. Mov Disord 2006;21 (Suppl 15):S429.

that all three quantitative measures of movement velocity reported in this study are useful to track the progression of bradykinesia in a PD population.
Acknowledgments: This study was supported in part from grants from the Ruth and Robert Halperin Foundation, the Vincent Coates Foundation, and from the Bio-X translational research initiative at Stanford University Medical Center. We thank Wendy Cole, Stephanie Louie, Bruce Hill, and Anna Frenklach for helpful comments on the manuscript.

REFERENCES
1. Braak H, Del Tredici K, Rub U, de Vos RA, Jansen Steur EN, Braak E. Staging of brain pathology related to sporadic Parkinsons disease. Neurobiol Aging 2003;24:197211. 2. The NINDS NET-PD Investigators. A randomized clinical trial of coenzyme Q10 and GPI-1485 in early Parkinson disease. Neurology 2007;68:2028. 3. The Parkinson study group. Effect of lazabemide on the progression of disability in early Parkinsons disease. Ann Neurol 1996;40:99107. 4. The Parkinson Study group. A controlled trial of rasagiline in early Parkinson disease: the TEMPO study. Arch Neurol 2002;59:19371943. 5. The Parkinsons Study group. A controlled, randomized, delayedstart study of rasagiline in early Parkinson disease. Arch Neurol 2004;61:561566. 6. Kieburtz K. Issues in neuroprotection clinical trials in Parkinsons disease. Neurology 2006;66 (Suppl 4):S50S57. 7. Bronte-Stewart HM, Ding L, Alexander C, Zhou Y, Moore GP. Quantitative digitography (QDG): a sensitive measure of digital motor control in idiopathic Parkinsons disease. Mov Disord 2000;15:3647.

Movement Disorders, Vol. 23, No. 9, 2008