OUP user menu

Hypokinesia without decrement distinguishes progressive supranuclear palsy from Parkinson's disease

Helen Ling, Luke A. Massey, Andrew J. Lees, Peter Brown, Brian L. Day
DOI: http://dx.doi.org/10.1093/brain/aws038 1141-1153 First published online: 6 March 2012


Repetitive finger tapping is commonly used to assess bradykinesia in Parkinson's disease. The Queen Square Brain Bank diagnostic criterion of Parkinson's disease defines bradykinesia as ‘slowness of initiation with progressive reduction in speed and amplitude of repetitive action’. Although progressive supranuclear palsy is considered an atypical parkinsonian syndrome, it is not known whether patients with progressive supranuclear palsy have criteria-defined bradykinesia. This study objectively assessed repetitive finger tap performance and handwriting in patients with Parkinson's disease (n = 15), progressive supranuclear palsy (n = 9) and healthy age- and gender-matched controls (n = 16). The motion of the hand and digits was recorded in 3D during 15-s repetitive index finger-to-thumb tapping trials. The main finding was hypokinesia without decrement in patients with progressive supranuclear palsy, which differed from the finger tap pattern in Parkinson's disease. Average finger separation amplitude in progressive supranuclear palsy was less than half of that in controls and Parkinson's disease (P < 0.001 in both cases). Change in tap amplitude over consecutive taps was computed by linear regression. The average amplitude slope in progressive supranuclear palsy was nearly zero (0.01°/cycle) indicating a lack of decrement, which differed from the negative slope in patients with Parkinson's disease OFF levodopa (−0.20°/cycle, P = 0.002). ‘Hypokinesia’, defined as <50% of control group's mean amplitude, combined with ‘absence of decrement’, defined as mean positive amplitude slope, were identified in 87% of finger tap trials in the progressive supranuclear palsy group and only 12% in the Parkinson's disease OFF levodopa group. In progressive supranuclear palsy, the mean amplitude was not correlated with disease duration or other clinimetric scores. In Parkinson's disease, finger tap pattern was compatible with criteria-defined bradykinesia, characterized by slowness with progressive reduction in amplitude and speed and increased variability in speed throughout the tap trial. In Parkinson's disease, smaller amplitude, slower speed and greater speed variability were all associated with a more severe Unified Parkinson's Disease Rating Scale motor score. Analyses of handwriting showed that micrographia, defined as smaller than 50% of the control group's mean script size, was present in 75% of patients with progressive supranuclear palsy and 15% of patients with Parkinson's disease (P = 0.022). Most scripts performed by patients with progressive supranuclear palsy did not exhibit decrements in script size. In conclusion, patients with progressive supranuclear palsy have a specific finger tap pattern of ‘hypokinesia without decrement’ and they do not have criteria-defined limb bradykinesia. Similarly, ‘micrographia’ and ‘lack of decrement in script size’ are also more common in progressive supranuclear palsy than in Parkinson's disease.

  • hypokinesia
  • bradykinesia
  • repetitive finger tap
  • micrographia
  • progressive supranuclear palsy


Bradykinesia is a sine qua non for the diagnosis of Parkinson's disease. In clinical practice, the term bradykinesia is often used interchangeably with the terms akinesia and hypokinesia. Nevertheless, bradykinesia literally describes slowness in movements, akinesia means absence or poverty of expected spontaneous voluntary movement including slow reaction time (Golbe and Ohman-Strickland, 2007), and hypokinesia refers to small amplitude movements. Bradykinesia, akinesia and hypokinesia are closely related but not necessarily well correlated in individual patients and each component of motor abnormality probably has a different underlying mechanism (Berardelli et al., 2001). Both bradykinesia and hypokinesia in Parkinson's disease improve with levodopa therapy, whereas reaction time is thought to be related to non-dopaminergic deficit (Velasco and Velasco, 1973; Berardelli et al., 1986; Jahanshahi et al., 1992). Bradykinesia is explicitly defined in the Queen Square Brain Bank criteria for the diagnosis of Parkinson's disease as ‘slowness of initiation of voluntary movement with progressive reduction in speed and amplitude of repetitive action’ (Gibb and Lees, 1988). The term sequence effect is used to describe the progressive reduction in amplitude and speed of sequential movements, which is a key feature of Parkinson's disease. If the amplitude and speed of sequential movements progressively decline until the movement ceases, this is known as motor arrest (Marsden, 1989; Kim et al., 1998; Iansek et al., 2006). The pathophysiology and levodopa response of the sequence effect are unclear.

Here, we investigate whether bradykinesia, as defined above, is also a feature of progressive supranuclear palsy (PSP). PSP is characterized by vertical supranuclear gaze palsy, early gait instability with falls characteristically in a backwards direction, axial rigidity and bulbar dysfunction. In their seminal paper describing the nine original PSP cases, Steele et al. (1964) provided brief accounts of elements of bradykinesia in only two cases, one of whom had slowness in walking and the other had awkwardness in performing rapid repetitive movements. As a consequence of their findings the authors concluded that PSP was a distinct clinico-pathological entity that was unlikely to be confused with Parkinson's disease. However, more recent literature closely links bradykinesia and parkinsonism with PSP and many movement disorder specialists consider PSP to be an example of atypical parkinsonism. In two large post-mortem series, early bradykinesia was reported in 88% and 75% of patients with pathologically confirmed PSP (Litvan et al., 1996a; Williams et al., 2005). In line with this view, ∼6% of cases with a clinical diagnosis of Parkinson's disease turn out to have tau pathology compatible with PSP at post-mortem examination (Hughes et al., 2002). These and other findings (Morris et al., 2002) have led to the delineation of two common clinical phenotypes: classical PSP, termed Richardson's syndrome and PSP-parkinsonism (Williams et al., 2005). PSP-parkinsonism closely resembles Parkinson's disease and is characterized by asymmetric symptoms at onset, tremor and a moderate initial therapeutic levodopa response.

Critically though, it is unclear whether the movement disorder described in the above literature adheres to the Queen Square Brain Bank definition of bradykinesia. Our clinical observations over the last 10 years suggest that most patients with PSP do not exhibit slowness or progressive reduction in amplitude and speed during finger tapping or handwriting.

Micrographia or small handwriting was first noted by Pick (1903) and has been associated with focal cerebral lesions (Pick, 1903; Scolding and Lees, 1994; Derkinderen et al., 2002; Kim et al., 2005; Kuoppamaki et al., 2005), post-encephalitic parkinsonism (Froment, 1921), Parkinson's disease (McLennan et al., 1972) and Huntington's disease (Iwasaki et al., 1999). Micrographia characterized by small handwriting with further progressive reduction in size can be observed in 15% of patients with Parkinson's disease (McLennan et al., 1972). The relationship between micrographia and bradykinesia remains controversial (McLennan et al., 1972). It is also not known if the handwriting in Parkinson's disease differs from PSP.

In this study we have looked for differences in the form of bradykinesia and handwriting in Parkinson's disease and PSP. Importantly, we objectively assessed the performance of repetitive finger tap movements. Repetitive finger tapping was selected as it is more severely impaired in patients with Parkinson's disease than hand opening and closing and hand pronation and supination elements of the motor section of Part III of the Unified Parkinson's disease rating scale (UPDRS; Agostino et al., 1998, 2003). Both finger tapping and writing are simple and commonly used bedside assessments and any distinctive features identified for each condition would provide helpful diagnostic clinical clues.

Materials and methods


Fifteen patients with Parkinson's disease, nine with PSP and sixteen healthy controls of similar age and gender ratio (Table 1) participated in this study. Patients were recruited from the movement disorder clinic in the National Hospital for Neurology and Neurosurgery, Queen Square, London. All patients with Parkinson's disease fulfilled the United Kingdom Queen Square Brain Bank diagnostic criteria (Hughes et al., 1992). All patients with PSP fulfilled the National Institutes of Neurological Disorders and Stroke (NINDS) Society for PSP diagnostic criteria (Litvan et al., 1996a). Patients with Parkinson's disease were included in the study if they were taking levodopa treatment with predictable motor fluctuations but were excluded if they had hand dystonia or if their tremor or dyskinesia were severe enough to interfere with their motor performance in the experiments. Exclusion criteria that applied to all subjects included significant medical co-morbidity, cognitive impairment (Mini-Mental State Examination Score < 28), depression (Beck depression score ≥ 21) (Beck et al., 1961) and disabilities that might restrict finger movements. All participants were assessed by the Edinburgh Handedness Inventory (Oldfield, 1971). The UPDRS was performed in all patients (Fahn and Elton, 1987). The PSP Rating Scale (Golbe and Ohman-Strickland, 2007) and the Frontal Assessment Battery (Dubois et al., 2000) were performed in patients with PSP. Patients' daily intake of anti-parkinsonian medications including levodopa, dopamine agonist, monoamine oxidase type B inhibitor, catechol-O-methyl transferase inhibitor and amantadine was recorded. Total daily levodopa equivalent dose was calculated for each patient according to published conversion formulae (Tomlinson et al., 2010). The study was conducted with the understanding and written consent of all participants and was approved by the Camden and Islington Community Research Ethics Committee of the National Research Ethics Service.

View this table:
Table 1

Demographic and clinical data

Controls (n = 16)Patients with PSP (n = 9)Patients with Parkinson's disease (n = 15)P-value
Age (years)68.9 ± 4.570.9 ± 8.365.0 ± 9.20.14**
Gender9M : 7F5M : 4F9M : 6F0.97#
Handedness13R : 3L7R : 2L14R : 1L0.51#
Edinburgh Handedness Inventory61.9 ± 56.455.6 ± 77.378.0 ± 50.90.62**
Disease duration (years)NA4.5 ± 3.3 (1.0–12.0)10.8 ± 7.5 (2.0–26.0)0.01*
Total daily levodopa equivalent dose (mg/day)NA255.6 ± 194.4874.8 ± 323.9< 0.00*
    INANA3.1 ± 2.5
    II-ONNANA7.7 ± 3.9
    II-OFFNANA17.2 ± 8.7
    III-ONNANA24.4 ± 9.3
    III-OFFNA41.6 ± 14.136.3 ± 9.7
    IVNANA6.7 ± 4.3
Hoehn and Yahr
    ONNANA2.1 ± 0.4
    OFFNANA2.8 ± 0.6
PSP Rating ScaleNA39.4 ± 2.4NA
Frontal Assessment BatteryNA14.4 ± 2.4NA
  • Mean ± SD; *Student t-test; **ANOVA; #Pearson Chi-Square test; NA = not applicable.


Participants were instructed to repeatedly tap their index finger and thumb as rapidly and as widely as possible for 15 s. The participants were instructed to relax the third, fourth and fifth digits in a semi-extended position so that the index finger-thumb movements were not restricted. The beginning and the end of the 15-s finger-tapping trial were signalled by a buzzer. Infrared-emitting diodes were fixed to eight designated regions on digits and the back of the hand, and motion was recorded in 3D (Coda Cx1, Charnwood Dynamics) (Fig. 1). Three 15-s trials were performed consecutively by each hand with 60 s rest in-between. Hand order was pseudo-randomized across participants. Patients with Parkinson's disease were tested during the ‘OFF’ condition (PD-OFF) in the morning after 12 h of overnight withdrawal of levodopa therapy, followed by a second experiment during an ‘ON’ condition (PD-ON) in the afternoon 1 h after taking levodopa. Only two patients with PSP were receiving levodopa treatment and both underwent overnight withdrawal of medication for 12 h prior to testing.

Figure 1

Infrared-emitting diodes fixed to eight designated regions. Motion was recorded in 3D (Coda) and key parameters were measured for each cycle of finger tap.

Handwriting task was performed after the tapping experiments by all participants and was repeated during the ‘ON’ condition by patients with Parkinson's disease. The participant was asked to copy a standardized print of Times New Roman, 34 pt font size, 11-word sentence on unlined A4 paper, three times (Fig. 2). No instructions were provided to the participants regarding the required size or speed of their script. The letters ‘a’ in the third (W3) and 10th words (W10) were selected and measurements were obtained using Microsoft Paint® program. The script size (cm2) of the selected letter was determined by the product of height and width outlined by the upper, lower, left and right margins of the loop in the letter. The size of W3 and W10 were plotted separately against successive sentence trials (1 to 3). Progressive reduction in size was represented by two slopes of the fitted linear regression line across the scatter-plots: script slope 1 from W3 and script slope 2 from W10.

Figure 2

Handwriting performed by a healthy 65-year-old female (mean script size = 0.86 cm2, slope 1 = 0.28, slope 2 = −0.06).

Kinematic parameters

Amplitude (mm), cycle duration (ms) and mean speed (mm/s) were measured for each cycle of finger tap from one finger-thumb separation to the next (Fig. 1) using custom scripts written in Matlab. Mean speed, designed to be sensitive to both amplitude and cycle duration, was the mean rate of change in aperture regardless of whether the aperture was opening or closing. Thus, mean speed decreased when the cycle duration increased independently of amplitude, when amplitude decreased independently of duration, and when both occurred simultaneously. If amplitude increased at the expense of cycle duration, or vice versa, the mean speed tended to stay constant. Close and open velocities (mm/s) were the peak velocities of aperture closure and opening within a cycle. To eliminate potential confounding factors of different hand size and finger length across participants, distance (mm) measured was converted into the degree (°) of angle separation between index finger and thumb. The conversion was obtained by the product of distance (mm) and k-value (°/mm), calculated by the linear regression slope of maximum finger-thumb separation angle against maximum finger-thumb separation distance of each hand of the participant. The separation angle was calculated as the angle between the straight line segments joining the index fingertip marker and the thumb marker to the marker placed at marker 3.

Progressive changes in amplitude, duration and speed across a 15-s finger tap trial were represented by the slope of the fitted linear regression line across the scatter-plot of the kinematic parameter against the tap cycle. The slope of change in amplitude was used to assess progressive hypokinesia or ‘decrement’. The slope of change in speed that encompassed both amplitude and duration was used to assess progressive slowing of movement or ‘fatigue’ (Fig. 3). Measurement of regularity of amplitude and speed across a tap trial was represented by the coefficient of variation, which was computed by the residual standard deviation about the linear regression line divided by the mean value. High amplitude or speed coefficient of variation values represent irregularities of these kinematic parameters.

Figure 3

Kinematic parameters during the first 15-s right finger tap trial in a Parkinson's disease patient during the ‘OFF’ condition represented by red circle plots (UPDRSIII-OFF = 32) and a patient with PSP represented by blue triangle plots (UPDRSIII = 69; Frontal Assessment Battery = 14). The slopes (S) for amplitude, duration and speed, and coefficient of variation (CV) for speed are shown. Lack of decrement and fatigue in the patient with PSP is reflected by the positive amplitude slope and speed slope. The speed coefficient of variation in the patient with Parkinson's disease is three times greater than PSP, suggesting high irregularity in speed.

Group parameters including amplitude, cycle duration, maximum close velocity, maximum open velocity, mean speed, slopes and coefficient of variations were summarized by computing the mean parameter value for all tap cycles across three finger tap trials of both hands for all subjects.

Statistical analysis

Comparisons of continuous variables were carried out by univariate Analysis of Variance (ANOVA) with gender, age and disease duration as covariates. Student t-test was used to compare disease duration and total daily levodopa equivalent dose between the two patient groups. Tukey honestly significant difference post hoc analysis was used to determine differences between groups (controls, PD-OFF and PSP). Paired t-tests were used to compare variables of patients with Parkinson's disease in ON versus OFF states. Chi-square test was used for discreet variables. Spearman's correlation was used to study correlation between group parameters and clinimetric scores. Statistical significance was determined when P ≤ 0.05. SPSS version 17.0 was used for statistical analysis.


The demographic features and clinimetric scores are listed in Table 1. Age was closely matched between groups. There were slightly more male participants compared to female in each group and the majority of participants were right-handed. All the patients with Parkinson's disease were receiving dopamine replacement therapies and all of them had derived good or excellent sustained therapeutic benefit. There were significant improvements in the UPDRS II, III, Hoehn and Yahr scores in patients with Parkinsinson's disease one hour after taking levodopa (OFF versus ON; paired t-test: P < 0.001 in all). The mean total daily levodopa equivalent dose in the Parkinson's disease groups were greater than that of the PSP group (t-test, P < 0.001). Eight patients with PSP were taking amantadine but only two were receiving levodopa therapy. The patients with PSP who were not receiving levodopa had either failed to respond to levodopa or had had a negative therapeutic response to an acute levodopa challenge (Steiger and Quinn, 1992).

Five of nine patients with PSP had evidence of midbrain atrophy on their most recent MRI. One PSP patient died 6 months after participating in this study and his pathological diagnosis was confirmed to be PSP at post-mortem. The mean disease duration in Parkinson's disease was longer than in PSP (t-test, P = 0.01).

The mean bradykinesia subscore, which included the sum of UPDRS motor scores for finger tap, hand opening and pronation/supination movements, also improved after levodopa therapy (OFF: 2.06 ± 0.54; ON: 1.76 ± 0.66; paired t-test, P = 0.009).

Repetitive finger tap movements

Spatial and temporal variables (amplitude, duration, peak velocities and mean speed) were measured for each tap cycle and used to characterize different aspects of motor performance. The analyses focused on mean performance, progressive changes in performance (slope of linear regression line of variable against cycle number), and regularity of performance (coefficient of variation) achieved over a 15-s trial. As shown in Table 2, one-way ANOVA revealed significant differences between the three groups for all measures apart from slope of cycle duration. In the following sub-sections we dissect out these group differences with post hoc analyses and use this to describe the important performance characteristics of each group in turn.

View this table:
Table 2

Mean parameter measurements (SD) of control, PD-OFF and PSP subgroups and P-values from one-way ANOVA adjusting for age, gender and duration

Controls (n = 16)PSP (n = 9)PD-OFF (n = 15)F(1, 19)P-value
Average number of tap cycles/15 s50.04 (10.7)52.22 (10.3)41.54 (9.7)8.530.009*
Average parameters
    Amplitude (°)45.91 (8.7)18.65 (6.3)37.82 (16.0)18.53< 0.001*
    Duration (ms)295.80 (65.6)288.34 (67.7)356.51 (80.6)7.800.012*
    Close velocity (°/s)−928.22 (215.2)−386.28 (134.4)−737.46 (385.0)11.390.003*
    Open velocity (°/s)788.91 (167.9)327.72 (106.1)584.40 (297.0)10.340.005*
    Speed (°/s)330.13 (64.6)142.90 (49.8)224.08 (93.1)8.330.009*
Average coefficients of variations
    Amplitude0.09 (0.03)0.27 (0.13)0.14 (0.08)9.690.006*
    Duration0.09 (0.03)0.28 (0.18)0.17 (0.10)4.26< 0.001*
    Speed0.09 (0.02)0.236 (0.9)0.167 (0.06)6.880.017*
Average slopes values
    Amplitude slope (°/cycle)−0.12 (0.12)0.01 (0.17)−0.20 (0.21)4.450.048*
    Duration slope (s/cycle)0.77 (0.75)1.86 (2.58)1.49 (2.39)0.160.70
    Speed slope (°/s/cycle)−1.52 (0.81)−0.39 (0.79)−1.71 (1.59)7.810.012*
  • Mean (SD); Covariates appearing in the model above are evaluated at the following values: gender = 0.58 (0 = female; 1 = male); age = 67.87 years; disease duration = 8.43 years; F(degrees of freedom) = f-value from one-way ANOVA; *P < 0.05 by one-way ANOVA.

Healthy subjects

Linear regression analysis did not show a significant correlation of any performance variables with age or gender. There was a modest effect of hand dominance in that mean cycle duration was longer for the non-dominant hand (dominant hand: 289.38 ± 64.6 ms; non-dominant hand: 302.23 ± 67.5 ms; P = 0.003) but no other performance parameters differed between the two hands.

The slope of the dominant hand's mean speed was significantly more negative in the third trial compared to the first (Trial 1 speed slope = −1.03°/s/cycle; Trial 3 speed slope = −1.46°/s/cycle; P = 0.043), indicating an increase in physiological fatigue. All other parameters showed a similar, but non-significant, slight decline in performance in progressive trials.

Patients with progressive supranuclear palsy

The performance of patients with PSP was characterized by strikingly small amplitudes of finger-thumb separation distance with a lack of performance decrement during a trial and with excessive variability of performance between cycles.

The small mean amplitude in the PSP group (mean = 18.65°) was less than half that of healthy subjects (mean = 45.91°, P < 0.001) and the PD-OFF group (mean = 37.82°, P < 0.002) (P < 0.001 in both cases) (Table 2 and Fig. 4A). The amplitude slope in the PSP group had a positive value of 0.01°/cycle, indicating a lack of amplitude decrement throughout the 15-s finger tap trial. This value differed significantly from the negative slope in the PD-OFF group (−0.2°/cycle, P = 0.02) (Table 2 and Fig. 4B). The possibility that the very small tapping amplitude in PSP as a group might have masked the detection of small degree of decrement was further explored. After adjusting for mean amplitude, there was no difference in amplitude slope between the PSP group and controls, indicating an absence of decrement in PSP (P = 0.36, Table 3).

Figure 4

(A) Mean amplitude, duration and speed of control, PSP and PD-OFF groups and P-values by post hoc analysis. Error bars represent 95% confidence intervals. *P < 0.05 indicates statistical significance and #P = 0.05–0.10 indicates borderline significance by Tukey HSD post hoc analysis. (B) Mean slope values for amplitude, duration and speed of control, PSP and PD-OFF groups and P-values by post hoc analysis. Error bars represent 95% confidence intervals. Dotted reference lines represent zero, values below which represent progressive downward negative slope across the 15-s finger tap trial. *P < 0.05 indicates statistical significance and #P = 0.05–0.10 indicates borderline significance by Tukey HSD post hoc analysis.

View this table:
Table 3

P-values for the comparisons of slope values between PD-OFF, PSP and control groups after adjusting for mean amplitude, duration and speed respectively

PD-OFF versus ControlsPSP versus ControlsPSP versus PD-OFF
F (1, 26)P-valuesF (1, 20)P-valuesF (1, 18)P-values
Amplitude slope (adjusted for mean amplitude)4.410.046*0.880.364.450.048*
Duration slope (adjusted for mean duration)0.470.5013.750.067#3.200.09 #
Speed slope (adjusted for mean speed)2.890.070 #1.700.2083.330.085 #
  • General linear model univariate analysis (*significant, P < 0.05, #borderline significant, P = 0.05-0.10).

  • Covariates appearing in the model PD-OFF versus controls are evaluated at the following values: gender = 0.58 (0 = female; 1 = male), age = 66.99 years, mean amplitude for amplitude slope model = 42.34°, mean duration for duration slope model = 325.18 s, mean speed for speed slope model = 278.81°/s.

  • Covariates appearing in the model PSP versus controls are evaluated at the following values: gender = 0.56 (0 = female; 1 = male), age = 69.6 years, mean amplitude for amplitude slope model = 36.09°, mean duration for duration slope model = 293.12 s, mean speed for speed slope model = 262.73°/s.

  • Covariates appearing in the model PSP versus PD-OFF are evaluated at the following values: gender = 0.58 (0 = female; 1 = male), age = 67.19 years, disease duration = 8.43 years, mean amplitude for amplitude slope model = 31.08°, mean duration for duration slope model = 330.95 s, mean speed for speed slope model = 193.64°/s.

  • F (degrees of freedom) = f-value from univariate analysis of variance.

A greater number of tap cycles were achieved by patients with PSP during a 15 s trial (mean = 52.22 cycles) when compared to the PD-OFF group (mean = 41.54 cycles; P = 0.046), but not controls (mean = 50.03 cycles, P = 0.87) (Table 2).

Although the PSP group cycle duration was similar to controls, the markedly reduced amplitude led to an overall reduction in close and open velocities and mean speed in the PSP group when compared with the PD-OFF group (close velocity: P = 0.014; open velocity: P = 0.021; mean speed: P = 0.035) and controls (P < 0.001 in all; Table 2 and Fig. 4A). This probably does not indicate an intrinsic slowing of movement as such, but simply stems from the digits moving through smaller amplitude with approximately the same cycle duration.

In the PSP group there was greater variability of performance from one cycle to the next as reflected in the highest coefficient of variation values. They were greater than controls for amplitude coefficient of variation, cycle duration coefficient of variation and mean speed coefficient of variation (P < 0.001 in all cases) and were also greater than the PD-OFF group for amplitude coefficient of variation (P = 0.001) and mean speed coefficient of variation (P = 0.009).

Among the PSP group, there was no correlation between mean amplitude and clinical markers of disease severity including disease duration (P = 0.40), total daily levodopa equivalent dose (P = 0.72), UPDRS motor score (P = 0.64), Hoehn and Yahr (P = 0.57), PSP staging (P = 0.40) or Frontal Assessment Battery scores (P = 0.15).

Patients with Parkinson's disease

When compared with controls, the main finding in the PD-OFF group was slowness of movement coupled with greater variability of speed between tap cycles. When compared with the PSP group, the PD-OFF group exhibited larger amplitude movements, a smaller number of tap cycles and greater decrement of performance across a 15-s trial.

The PD-OFF group amplitude (P = 0.10) tended to be smaller than that in healthy subjects while cycle duration (P = 0.062) tended to be more prolonged, but only with borderline significance. However, the combination of both these trends led to a highly significant lower mean speed of the PD-OFF group compared with controls (P = 0.001; Fig. 4). Similarly, peak open velocity of the PD-OFF group was less than controls (P = 0.033), although there was no difference in peak close velocity between the two groups. In addition, coefficient of variation of mean speed in the PD-OFF group was significantly greater than that of controls (P = 0.004), suggesting proportionally greater irregularities between cycles.

Both amplitude and speed slopes in the PD-OFF group, reflecting the progressive change in performance, were more strongly negative when compared to those of the PSP group (amplitude slope: P = 0.018; speed slope: P = 0.028). However, the negative amplitude and speed slopes of the PD-OFF group were numerically, but not significantly, greater than healthy subjects. In patients with Parkinson's disease with severe parkinsonism, slope measurements may be underestimated due to poor performance during the tap trial, which would render their slope values lower than patients with milder disease severity who do not exhibit a ‘floor’ effect. After adjusting for differences in mean amplitude, the amplitude slope in the PD-OFF group became significantly more strongly negative than the PSP group and healthy controls (PD-OFF versus PSP, P = 0.048; PD-OFF versus controls, P = 0.046, Table 3). There was a trend for a more negative speed slope in the PD-OFF group when compared with controls after adjusting for mean speed (P = 0.07, Table 3). These findings demonstrate progressive decrement and possibly fatigue in the PD-OFF group and represents sequence effect in Parkinson's disease.

A more severe UPDRSIII-OFF score was correlated with a smaller mean amplitude (Spearman's coefficient: −0.79, P < 0.001), slower mean speed (Spearman's coefficient: −0.68, P = 0.005) and greater variability in speed (Spearman's coefficient: 0.75, P = 0.001; Fig. 5). There was no correlation between performance decrement (slopes for amplitude, duration and speed), disease duration, total daily levodopa equivalent dose, UPDRSIII or Hoehn and Yahr stage (P > 0.05 in all cases).

Figure 5

Among PD-OFF subgroup, a more severe UPDRSIII-OFF score was correlated with smaller mean amplitude, slower mean speed and greater variability in speed.

Levodopa therapy improved the total number of tap cycles (OFF = 41.5 ± 9.7 cycles/15 s; ON = 45.9 ± 9.5 cycles/15 s, P = 0.04), peak open velocity (OFF = 584.4 ± 297.0°/s; ON = 639.9 ± 269.0°/s; P = 0.04), mean speed (OFF = 224.1 ± 93.1°/s; ON = 255.6 ± 86.4°/s; P = 0.006) and speed coefficient of variation (OFF = 0.167 ± 0.07; ON = 0.150 ± 0.08; P = 0.014). However, levodopa therapy did not significantly improve performance decrement or fatigue (amplitude slope: OFF = −0.20 ± 2.1°/cycle; ON = −0.17 ± 2.1°/cycle; speed slope: OFF = −1.71 ± 1.6°/s/cycle; ON = −1.78 ± 1.4°/s/cycle).

When analysis of the effect of levodopa was limited to the more affected hand of patients with Parkinson's disease, more robust ON versus OFF differences were observed. In addition to the improvements described above, improvement was also observed in mean cycle duration (OFF = 370.8 ± 103.6 ms; ON = 321.1 ± 93.2 ms; P = 0.005) and there was a trend towards improvement in performance decrement (amplitude slope: OFF = −0.20°/cycle; ON = −0.15°/cycle; P = 0.07).

Hypokinesia without decrement

Each 15-s finger tap trial in the PSP and PD-OFF groups was analysed. Hypokinesia was defined as a mean amplitude of <23°, i.e. 50% of the control group's mean amplitude. Hypokinesia was observed in 70% of the finger tap trials in the PSP group, 24% in the PD-OFF group and 2% of the control group. The remaining 30% of the PSP finger tap trials had small mean amplitude of 27.8 ± 3.7° and a positive mean amplitude slope of 0.05°/cycle. The 24% of the PD-OFF group finger tap trials with hypokinesia were performed by four patients who had severe parkinsonism with a mean UPDRSIII-OFF score of 46.4 and a long mean disease duration of 17.5 years. All four patients had good levodopa response and an average improvement in UPDRS motors score by 14.3 1 h after intake of levodopa therapy. Despite severe hypokinesia with a mean amplitude of 11.4 ± 5.6°, decrement was still evident with a negative mean amplitude slope of −0.037 ± 0.1°/cycle (versus control, P = 0.05). When lack of decrement, defined as a positive amplitude slope, was combined with hypokinesia, 87% of finger tap trials in the PSP group, 12% in the PD-OFF group and none in the control group were noted to exhibit both features.

Handwriting in Parkinson's disease and progressive supranuclear palsy

The scripts from one patient with PSP and two patients with Parkinson's disease were discarded from the analysis as they were written in capital letters. The mean script size of the PSP group (0.50 ± 0.46 cm2) was numerically, but not statistically, smaller than the PD-OFF group (0.78 ± 0.38 cm2; P = 0.29) and controls (0.79 ± 0.20 cm2; P = 0.07). Progressive changes in script size were assessed from the slopes of the linear regression lines separately fitted for W3 (script slope 1) and W10 (script slope 2) across the three successive sentences. There was less of a decrement over successive sentences for W10 in the PSP group (mean script slope 2; 0.06 ± 0.09) than in the PD-OFF group (−0.08 ± 0.30) after adjusting for age, gender, disease duration and mean script size (P = 0.02). A similar trend was found in mean script slope 1 (PSP: −0.004 ± 0.21; PD-OFF: −0.103 ± 0.21; P = 0.16). After levodopa therapy, six patients with Parkinson's disease exhibited a mean increase of 0.26 cm2 in script size, however, the overall script size did not achieve statistical significance between PD-OFF and PD-ON groups (P = 0.28, Fig. 6). Decrements in script size persisted in patients with Parkinson’s disease despite levodopa therapy as shown by the negative script slope 1 (OFF: −0.10, ON: −0.05; P = 0.48) and script slope 2 (OFF: −0.08, ON: −0.09; P = 0.82).

Figure 6

(A) Handwriting performed by a 58-year-old right-handed patient when in the ‘OFF’ condition (UPDRS = 18, mean script size = 0.66 cm2, slope 1 = −0.01, slope 2 = −0.16). (B) Handwriting by the same patient with Parkinson's disease when in the ‘ON’ condition (UPDRS = 13, mean script size = 1.14 cm2, slope 1 = −0.08, slope 2 = −0.46).

Micrographia was determined as present when the mean script size was <0.40 cm2, i.e. half the mean script size of the control group, and the lack of progressive micrographia was defined by a positive mean script slope. Micrographia was more frequent in the PSP group (n = 6, 75%) than in the PD-OFF group (n = 2, 15.4%; P = 0.022) and controls (n = 1, 6.3%; P = 0.003). The script size numerically improved in the two patients with Parkinson's disease who had micrographia but their ON script size still did not exceed 0.40 cm2. A positive script slope 2 was more frequent in the PSP group (n = 5, 62.5%) than in control (n = 1, 6.3%; P = 0.007) and, possibly, in PD-OFF (n = 3, 23.1%; P = 0.09) groups. A positive script slope 1 was more frequent in PSP (n = 6, 75%) than in PD-OFF patients (n = 3, 23.1%; P = 0.03), but it did not differ from control subjects (n = 8, 50%; P = 0.23). The patients with the smallest script size in the PSP and PD-OFF groups were also noted to have the most severe UPDRSIII score in their group (minimum script size in PSP = 0.14 cm2, UPDRSIII = 69, Fig. 7; minimum script size in PD-OFF = 0.11 cm2, UPDRSII I = 50).

Figure 7

An example of micrographia from a patient with advanced PSP who had the smallest script size in the PSP group (UPDRSIII = 69, mean script size = 0.14 cm2, slope 1 = 0.23, slope 2 = −0.01).

There were more patients with PSP (n = 5, 62.5%) who had both hypokinesia (<23°) and micrographia (<0.40 cm2) than the control (0; P = 0.001) and the PD-OFF (n = 1, 7.7%; P = 0.014) groups. In PSP, the finger tap amplitude slope was strongly correlated with script slope 2 (Spearman's coefficient = 0.88, P = 0.004). No correlation was found between script findings and markers of disease severity in either Parkinson's disease or PSP groups.


Bradykinesia in Parkinson's disease

We made objective recordings of the sequence effect during repetitive finger tap movements and found that a progressive decrement in amplitude was present in Parkinson's disease but not in PSP. We also confirmed that the characteristic finger tap pattern in Parkinson's disease consists of slowness with variability in speed and progressive decrement in performance (Agostino et al., 1994, 1998). Although levodopa improved most tapping parameters in Parkinson's disease, it did not improve the sequence effect of progressive deterioration in cycle duration and speed. However, there was a borderline improvement in decrement in treated Parkinson's disease when only the more affected hand was studied. We conclude that the sequence effect in Parkinson's disease may be relatively independent of dopaminergic regulation. A recent study using a Modified Purdue Pegboard Test showed that sequence effect in Parkinson's disease did not respond to levodopa medication (Kang et al., 2010). In another study, reduced stride length (hypokinesia) improved with either levodopa or visual cues, but the progressive reduction of stride length (sequence effect) only improved with cueing (Iansek et al., 2006). We found that the variability of speed was significantly greater in the PD-OFF group when compared with controls, and that it improved with levodopa therapy, suggesting that the mechanisms underlying the temporal regularity of movements and the sequence effect are likely to be different.

Hypokinesia without decrement in progressive supranuclear palsy

The most striking finding in the present study was the very small index finger-to-thumb separation amplitude during repetitive finger tapping in PSP. The average amplitude of finger separation in PSP was less than half of that in controls and the PD-OFF group. Patients with PSP also had a greater number of tap cycles when compared to the PD-OFF group. The greater number of tap cycles was most probably related to the small amplitude as the digits moved through a smaller distance allowing more cycles to be performed within a given time. While small amplitude in the PD-OFF group was correlated with more severe UPDRS motor score, there was no correlation between amplitude and markers for disease severity in the PSP group. Thus, the differences in disease duration between Parkinson's disease and PSP could not account for the reduced mean amplitude in the PSP group. Furthermore, it could not be explained by medication status because all patients were tested after 12-h withdrawal of anti-parkinsonian medication.

The second key finding was the lack of progressive reduction in amplitude. This is compatible with our clinical impression that most patients with PSP do not exhibit decrement during repetitive finger tapping. The positive amplitude slope of 0.01°/cycle in the PSP group was similar to controls but differed significantly from the negative slope of −0.2°/cycle in the PD-OFF group. It is possible that a lack of decrement in PSP might be due to a floor effect caused by severe hypokinesia. However, even among the subgroup of patients with Parkinson's disease with severe hypokinesia (amplitude <23°), we noted a mean negative amplitude slope of −0.037°/cycle. Furthermore, when comparisons were performed after adjustment for any differences in mean amplitudes between groups, we found that the mean amplitude slope in the PD-OFF group was more negative than PSP and control groups, while there was no difference between the PSP group and controls (Table 3). These findings support a minimal or lack of performance decrement and sequence effect in PSP that is incompatible with the Queen Square Brain Bank definition of bradykinesia for the clinical diagnosis of Parkinsonism.

Pathophysiological mechanisms

Severe neuronal loss in the substantia nigra pars compacta is observed in both Parkinson's disease and PSP with greater involvement in the ventromedial and dorsal tiers in PSP (Fearnley and Lees, 1990; Hardman et al., 1997). In PSP, substantial damage also occurs in the zona reticulata of the nigra (Hardman et al., 1996), the internal segment of the globus pallidus, the subthalamic nucleus of Luys, the dentate nucleus, superior cerebellar peduncle and to a lesser degree, the striatum and thalamus (Demirci et al., 1997; Bryant et al., 2010).

In Parkinson's disease, the subthalamic nucleus and globus pallidus interna are affected functionally with increased neuronal discharges as a result of disruption of the basal ganglia circuit (Wichmann and DeLong, 2003). Functional compensatory change in the putamen has also been reported in Parkinson's disease, which has been proposed to contribute to the diminished levodopa response later in the disease course (Halliday, 2007). The cerebellum may play a role in motor sequencing (Garraux et al., 2005). Greater activity of both cerebellar hemispheres was found in functional imaging in patients with Parkinson's disease during automated movements when compared with healthy controls, suggesting that the cerebellum might contribute to the compensatory pathway in Parkinson's disease (Wu and Hallett, 2005).

It has been postulated that movement size is regulated by phasic signals from globus pallidus interna to the supplementary motor area and premotor cortex (Alexander and Crutcher, 1990). Severe hypokinesia in PSP, therefore, might be due to the extensive pathological damage to the globus pallidus interna and subthalamic nucleus (Hauw et al., 1994; Litvan et al., 1996b). There is also loss of cholinergic neurons in the putamen and loss of pyramidal neurons in the premotor cortex (Halliday, 2007), which could also influence the nature of the motor deficit. Finally, potential compensatory mechanisms via the cerebellar outflow pathway are cut off in PSP due to damage of the superior cerebellar peduncle (Tsuboi et al., 2003; Whitwell et al., 2011). The putamen appears to have a role in movement timing and it might contribute to the variability in performance in PSP and Parkinson's disease (Garraux et al., 2005).

Sequence effect is reflected by the impairment of scaling of motor sequences and contributes to prolonged movement time in Parkinson's disease (Benecke et al., 1987; Agostino et al., 1994). Its pathophysiology in Parkinson's disease is still poorly understood but it is likely to be independent of dopaminergic pathways. It appears that the sensorimotor apparatus in patients with Parkinson's disease is set smaller but the capacity to achieve the correct amplitude is intact and can be overcome by visual guidance (Hallett, 2003). These findings may not be relevant in PSP where the pathological lesion is more extensive and where visual cueing is an ineffective strategy to improve gait. The lack of levodopa response in sequence effect in Parkinson's disease was also supported by our findings.

In Parkinson's disease, rigidity and tremor are thought to contribute to slowness in limb movements (Berardelli et al., 2001; Quencer et al., 2007). On the other hand, patients with PSP who have more axial symptoms and sometimes no detectable rigidity of the limbs on examination might arguably manifest less degree of slowness on repetitive finger tapping.

Finger tap assessments

‘Hypokinesia without decrement’ was identified in 87% of finger tap trials in the PSP group and only 12% in the PD-OFF group. This finding might be particularly useful in patients with PSP-parkinsonism, where the physical signs can mimic Parkinson's disease. The remaining PSP finger tap trials also had a small mean amplitude of 27.8°, but not quite making the cut off value of 23° for hypokinesia. Small finger tap amplitudes can be easily recognized by careful bedside examination. Small degrees of decrement may however be difficult to detect in patients with Parkinson's disease with severe motor impairment who have small amplitude finger movements on initiation of finger movements. These patients are readily differentiated from PSP by their sustained levodopa response and relatively long disease duration. The patients with Parkinson's disease with severe hypokinesia in the present study had mean disease duration of 17.5 years, whereas the mean duration from diagnosis to death in PSP is 7 years (Williams et al., 2005). In addition to decrement, delayed initiation of voluntary movements and motor arrests during repetitive finger tapping in Parkinson's disease may also have clinical usefulness (Fahn and Elton, 1987; Marsden, 1989).

The average number of tap cycles performed in 15 s was 50 in controls, 52 in PSP, 42 in PD-OFF and 46 in PD-ON. Therefore, to detect the differences reported above would require a tap trial of ∼50 finger-thumb tap cycles. The modified Movement Disorder Society UPDRS (Goetz et al., 2008) proposed a 10-tap trial, which would take an average of 3.8 s (15 s/42 taps × 10 taps) for PD-OFF subjects to perform. We postulated that a tap trial consisting of only 10 taps would be too brief for the sequence effect to emerge in either treated or untreated Parkinson's disease. To investigate this, we have performed an additional analysis on our data by arbitrarily assessing only the first 20 taps of the first trials performed by both hands after adjusting for disease duration, age and gender (Supplementary material). With a 20-tap trial, PSP can still be differentiated from both PD-OFF and control groups by having amplitudes of less than half the expected size. Mean speed in the PD-OFF group was slower than controls (P = 0.007). However, after 20 taps, the amplitude slope (mean = +0.04) and speed slope (mean = +0.21) in the PD-OFF group were both positive, indicating the lack of decrement and fatigue at that time point and the slope values did not differ between PD-OFF, PSP and control groups. This analysis indicated that 20-tap trials were not adequate to detect either decrement or fatigue in Parkinson's disease. We propose that repetitive finger tapping with 50 tap cycles is required to detect criteria-defined bradykinesia in treated and untreated patients with Parkinson's disease.

Handwriting in Parkinson's disease and progressive supranuclear palsy

Micrographia was more common in PSP (75%) than in Parkinson's disease (15%). Decrement in script size was less common in the PSP group than in the PD-OFF group. These findings were similar to the hypokinesia without decrement in repetitive finger tapping found in PSP. Five of the six patients with PSP who had micrographia, also manifested hypokinesia on repetitive finger tapping. Despite the similarities in the findings of finger tapping and handwriting in the PSP group, the correlations between the parameters of these two kinematic tasks are inexact. ‘Fast micrographia’ characterized by microscopically small letters performed at a normal or slightly faster than normal speed may be a physical sign related to pallidal damage (Kuoppamaki et al., 2005) and has been associated with some cases of PSP. Our clinical impression suggests an increase in writing speed in some patients with PSP. Nevertheless, it is uncertain if the ‘fast’ speed represents a shorter performance time due to the reduced stroke size or an intrinsic increase in writing speed. The present study did not time the handwriting task so we were unable to verify this.

McLennan et al. (1972) reported micrographia in 15% of patients with Parkinson's disease and, in 16 out of 30 cases, a significant and sustained improvement in script size was noted after levodopa therapy. Our findings also showed the same percentage of micrographia in Parkinson's disease, and, after levodopa therapy, six patients exhibited marked improvement in script size but decrements in script size persisted. Copying scripts, writing on parallel lines and verbal reminders to write ‘big’ can serve as external cues to correct a reduction in script size (Oliveira et al., 1997; Kim et al., 2005; Bryant et al., 2010). Abnormally increased dependence on external visual feedback has been noted in patients with Parkinson's disease (Demirci et al., 1997). The mechanism of micrographia is poorly understood but the hypothesis of a ‘tuned-down’ sensorimotor apparatus might explain the reduction in motor scaling during sequential motor tasks such as finger tapping and handwriting (Demirci et al., 1997).

Strengths and limitations of present study

The 3D motion assessment used in this study proved particularly useful in tremulous patients who would have had difficulties maintaining their finger separation in a 2D plane. This method captures accurate and diverse measurements of the finger tap trials. Pilot studies were conducted on healthy volunteers and it was determined that healthy elderly participants become tired after prolonged tap sequences of more than 15–20 s. In order to minimize the confounding factor of physiological fatigue, the trial duration was, therefore, limited to 15 s. The study was specifically designed to study motor execution, not processing or reaction time. We accept that certain aspects of the quantitative measurements made here may not be applicable to subjective bedside observation. Further studies are warranted to apply our findings in a clinical context. However, very small finger tapping amplitudes can be easily identified during neurological examination. The relative persistence of the sequence effect despite levodopa therapy in Parkinson's disease makes it an especially useful physical sign to help distinguish Parkinson's disease and PSP. Future prospective studies on patients with early PSP or the PSP-parkinsonism subtype will determine if this specific finger tap pattern can be used as a reliable early diagnostic sign to distinguish between Parkinson's disease and PSP. Further kinematic studies in PSP may also improve our understanding of the pathophysiological mechanisms of severe hypokinesia in PSP.

An inherent limitation of clinical studies of this kind is the lack of pathological confirmation of diagnosis given that ∼20% of suspected PSP cases and 10% of suspected Parkinson's disease cases are found at post-mortem to have a different pathology (Froment, 1921; Ling et al., 2010). Finally, it should be noted that patients with prominent tremor were excluded from our study. In tremor-predominant Parkinson's disease, motor flurries can potentially interrupt normal self-paced movements, confounding clinical interpretation (Bajaj et al., 2010).


Patients with PSP have small finger separation amplitude without progressive decrement on repetitive finger tapping and do not have criteria-defined limb bradykinesia. The severe hypokinesia irrespective of disease severity and the lack of a sequence effect help distinguish patients with PSP from those with Parkinson's disease. Similarly, micrographia and lack of decrement in script size are also more common in PSP than in Parkinson's disease.


This work was supported by the PSP (Europe) Association [6AJV to H.L.]; the Medical Research Council [G0501740 to B.L.D.]; and the Wellcome Trust [084870/Z/08/Z to B.L.D.].

Supplementary material

Supplementary material is available at Brain online.


The authors are indebted to the patients and healthy volunteers who participated in this study. We are grateful to Hans-Leo Teulings, Karen Shaw, Omar Mian, Dorothy Cowie, Amy Peters and Daniel Voyce for their helpful advice and assistance in this study.

Parkinson's disease OFF levodopa medication
Parkinson's disease ON levodopa medication
progressive supranuclear palsy
Unified Parkinson's disease rating scale


View Abstract