Advertisement

Impact of Time on Quality of Motor Control of the Paretic Upper Limb After Stroke

  • Joost van Kordelaar
    Affiliations
    Department of Rehabilitation Medicine, MOVE Research Institute Amsterdam, VU University Medical Center, Amsterdam, The Netherlands
    Search for articles by this author
  • Erwin van Wegen
    Affiliations
    Department of Rehabilitation Medicine, MOVE Research Institute Amsterdam, VU University Medical Center, Amsterdam, The Netherlands
    Search for articles by this author
  • Gert Kwakkel
    Correspondence
    Corresponding author: Gert Kwakkel, PhD, Chair of Neurorehabilitation, Department of Rehabilitation Medicine, MOVE Research Institute Amsterdam, VU University Medical Center, PO Box 7057, 1007 MB Amsterdam, The Netherlands.
    Affiliations
    Department of Rehabilitation Medicine, MOVE Research Institute Amsterdam, VU University Medical Center, Amsterdam, The Netherlands

    Amsterdam Rehabilitation Research Center, Reade Centre for Rehabilitation and Rheumatology, Amsterdam, The Netherlands
    Search for articles by this author
Published:October 24, 2013DOI:https://doi.org/10.1016/j.apmr.2013.10.006

      Abstract

      Objective

      To establish the time course of recovery regarding smoothness of upper limb movements in the first 6 months poststroke.

      Design

      Cohort study with 3-dimensional kinematic measurements in weeks 1, 2, 3, 4, 5, 8, 12, and 26 poststroke.

      Setting

      Onsite 3-dimensional kinematic measurements in stroke units, rehabilitation centers, nursing homes, and patients' homes.

      Participants

      Patients (N=44; 19 women, 25 men; mean age ± SD, 58±12y) with a first-ever unilateral ischemic stroke and incomplete upper limb paresis (27 left sided, 17 right sided) were included.

      Interventions

      Not applicable.

      Main Outcome Measures

      In each measurement, an electromagnetic motion tracker acquired hand and finger trajectories during a reach-to-grasp task. Movement duration was determined, and smoothness of hand transport and grasp aperture was quantified by normalized jerk. With the use of random coefficient analysis, the effect of progress of time on smoothness of hand transport and grasp aperture was investigated.

      Results

      During the first 5 weeks poststroke, there was a significant contribution of progress of time to reductions in movement duration and normalized jerk of hand transport and grasp aperture (P<.01).

      Conclusions

      The present longitudinal 3-dimensional kinematic study showed that smoothness of paretic upper limb movements improves in the first 8 weeks poststroke. This improvement suggests that motor control normalizes in the first 8 weeks poststroke and can be mostly explained by spontaneous neurologic recovery that occurs typically in the first weeks poststroke. Future 3-dimensional kinematic studies should investigate whether therapies starting early after stroke can improve the quality of motor control beyond spontaneous neurologic recovery.

      Keywords

      List of abbreviations:

      MD (movement duration), NJ (normalized jerk)
      Time poststroke is one of the most neglected features in explaining recovery of the paretic upper limb after stroke.
      • Kwakkel G.
      • Kollen B.J.
      • Twisk J.
      Impact of time on improvement of outcome after stroke.
      • Kwakkel G.
      • Kollen B.J.
      • Lindeman E.
      Understanding the pattern of functional recovery after stroke: facts and theories.
      Based on clinical assessment scales, several longitudinal studies
      • Kwakkel G.
      • Kollen B.J.
      • Twisk J.
      Impact of time on improvement of outcome after stroke.
      • Twitchell T.E.
      The restoration of motor function following hemiplegia in man.
      • Duncan P.W.
      • Goldstein L.
      • Matchar D.
      • Divine G.
      • Feussner J.
      Measurement of motor recovery after stroke. Outcome assessment and sample size requirements.
      have shown that most improvements in motor function and capacity occur during the first 10 weeks poststroke. These early time-dependent improvements are assumed to reflect processes of spontaneous neurologic repair
      • Kwakkel G.
      • Kollen B.J.
      • Lindeman E.
      Understanding the pattern of functional recovery after stroke: facts and theories.
      such as recovery of penumbral tissue,
      • Phan T.G.
      • Wright P.M.
      • Markus R.
      • Howells D.W.
      • Davis S.M.
      • Donnan G.A.
      Salvaging the ischaemic penumbra: more than just reperfusion?.
      alleviation of diaschisis,
      • Andrews R.J.
      Transhemispheric diaschisis. A review and comment.
      and reorganization of the dendritic spine architecture of cortical and corticospinal neurons.
      • Krakauer J.W.
      • Carmichael S.T.
      • Corbett D.
      • Wittenberg G.F.
      Getting neurorehabilitation right: what can be learned from animal models?.
      Since it is unclear whether spontaneous neurologic recovery leads to normalization of quality of motor control in the first weeks poststroke, there is a need for intensively repeated 3-dimensional kinematic measurements, preferably conducted during functional movements such as reaching and grasping.
      • Van Kordelaar J.
      • van Wegen E.E.H.
      • Nijland R.H.M.
      • et al.
      Assessing longitudinal change in coordination of the paretic upper limb using on-site 3-dimensional kinematic measurements.
      Several 3-dimensional kinematic studies
      • Alt Murphy M.
      • Willén C.
      • Sunnerhagen K.S.
      Kinematic variables quantifying upper-extremity performance after stroke during reaching and drinking from a glass.
      • Caimmi M.
      • Carda S.
      • Giovanzana C.
      • et al.
      Using kinematic analysis to evaluate constraint-induced movement therapy in chronic stroke patients.
      • Rohrer B.
      • Fasoli S.E.
      • Krebs H.I.
      • et al.
      Movement smoothness changes during stroke recovery.
      have shown that the quality of motor control after stroke is deficient, as reflected by a significant reduction in the smoothness of upper limb movements. The mechanism underlying this reduced smoothness is unknown but may involve segmentation of reaching movements caused by poor interjoint and intermuscular coordination.
      • Levin M.F.
      Interjoint coordination during pointing movements is disrupted in spastic hemiparesis.
      • Van Kordelaar J.
      • van Wegen E.E.H.
      • Kwakkel G.
      Unraveling the interaction between pathological upper limb synergies and compensatory trunk movements during reach-to-grasp after stroke: a cross-sectional study.
      However, longitudinal studies investigating the time course of smoothness during an early stage after stroke are lacking in the literature.
      The first aim of the present study was to investigate the time course of recovery in terms of smoothness of upper limb movements in the first 6 months poststroke. The second aim was to assess how progress of time contributes to normalization of smoothness of upper limb movements. We hypothesized that smoothness would normalize in a natural logistic pattern of recovery, as was found in previous studies using standard clinical assessment scales. Secondly, we hypothesized that longitudinal changes in smoothness would be time-dependent and only significant in the first 10 weeks poststroke, when spontaneous neurologic recovery occurs.

      Methods

      Participants

      Forty-four patients (19 women, 25 men) were included in the present study, which was part of the translational research program explaining plasticity after stroke (EXPLICIT-stroke)
      • Kwakkel G.
      • Meskers C.G.M.
      • van Wegen E.E.H.
      • et al.
      Impact of early applied upper limb stimulation: the EXPLICIT-stroke programme design.
      All included patients met the following criteria: (1) having experienced a first-ever ischemic stroke involving the territory of the medial or anterior cerebral artery as revealed by computerized axial tomography or magnetic resonance imaging scan; (2) no intravenous thrombolysis with recombinant tissue plasminogen activator (or alteplase), since thrombolysis influences the spontaneous neuronal processes in the brain and may therefore affect the time course of smoothness after stroke; (3) aged between 18 and 80 years; (4) able to sit independently without trunk support for at least 30 seconds; (5) showing motor deficits in the arm, hand, or both but nevertheless being able to grasp objects within the first 3 weeks poststroke; (6) no severe deficits of cognition, as indicated by a score of 23 or higher on the Mini-Mental State Examination; (7) no severe deficits of communication, as indicated by a score of 4 or higher on the Utrecht Communication Observation; (8) no complicating medical history such as cardiac, pulmonary, or orthopedic disorders; and (9) having provided written informed consent.
      According to the EXPLICIT-stroke protocol,
      • Kwakkel G.
      • Meskers C.G.M.
      • van Wegen E.E.H.
      • et al.
      Impact of early applied upper limb stimulation: the EXPLICIT-stroke programme design.
      all patients were screened and included in the first week poststroke at the stroke units of 11 hospitals in The Netherlands. Screening was performed by medical doctors, occupational therapists, and physical therapists who were employed at the stroke units and were trained to recruit patients according to the EXPLICIT-stroke protocol. Depending on the impairment level and site of inclusion, included patients underwent clinical measurements as well as functional magnetic resonance imaging, transcranial magnetic stimulation, 3-dimensional kinematics, and haptic robotics. Inclusion and intake were performed by the researcher who performed all clinical measurements. All 3-dimensional kinematic measurements were conducted by 1 researcher (J.V.K.). The EXPLICIT-stroke protocol (registered with The Netherlands National Trial Register as trial no. NTR1424) was approved by the local ethics committee.
      • Kwakkel G.
      • Meskers C.G.M.
      • van Wegen E.E.H.
      • et al.
      Impact of early applied upper limb stimulation: the EXPLICIT-stroke programme design.
      All patients underwent a clinical and 3-dimensional kinematic assessment of the upper paretic limb in weeks 1, 2, 3, 4, 5, 8, 12, and 26 poststroke.

      Clinical evaluation

      During the baseline session, the type of stroke was established with the Bamford classification.
      • Bamford J.
      • Sandercock P.
      • Dennis M.
      • Burn J.
      • Warlow C.
      Classification and natural history of clinically identifiable subtypes of cerebral infarction.
      The severity of the infarct was assessed with the National Institutes of Health Stroke Scale.
      • Goldstein L.
      • Bertels C.
      • Davis J.
      Interrater reliability of the NIH Stroke Scale.
      Motor impairments were assessed with the upper extremity section of the Fugl-Meyer Motor Assessment.
      • Fugl-Meyer A.R.
      • Jääskö L.
      • Leyman I.
      • Olsson S.
      • Steglind S.
      The post-stroke hemiplegic patient 1. A method for evaluation of physical performance.
      The ability to perform functional tasks with the paretic upper limb was assessed with the Action Research Arm Test.
      • Lyle R.C.
      A performance test for assessment of upper limb function in physical rehabilitation treatment and research.
      The ability to perform activities of daily living was assessed with the Barthel Index.
      • Collin C.
      • Wade D.T.
      • Davies S.
      • Horne V.
      The Barthel ADL Index: a reliability study.

      Kinematic setup

      Kinematic data were recorded with a portable electromagnetic motion tracking device (Polhemus Libertya). As depicted in figure 1, the electromagnetic source was placed at the edge of the table next to the participant at the side of paresis. The position and orientation data of the sensors have been shown to be accurate within a distance of 60cm from the source in multiple environments, including a motion laboratory, treatment room, and home situation.
      • Van Kordelaar J.
      • van Wegen E.E.H.
      • Nijland R.H.M.
      • et al.
      Assessing longitudinal change in coordination of the paretic upper limb using on-site 3-dimensional kinematic measurements.
      In addition, the reliability of the 3-dimensional kinematic parameters obtained with this device is good to excellent.
      • Van Kordelaar J.
      • van Wegen E.E.H.
      • Nijland R.H.M.
      • et al.
      Assessing longitudinal change in coordination of the paretic upper limb using on-site 3-dimensional kinematic measurements.
      These findings suggest that this device is suitable for use in a multicenter longitudinal study in patients with stroke as they move from the stroke unit to rehabilitation centers, nursing homes, and their home situation. All upper limb movements were measured relative to a global reference frame with its origin at the center of the magnetic source, with the x-axis directed forward, the y-axis directed upward, and the z-axis directed rightward (see fig 1). To standardize the 3-dimensional kinematic setup, the same portable electromagnetic motion tracker and portable table with a height of 76cm were used for all measurements.
      Figure thumbnail gr1
      Fig 1(A) Determination of the MRD. (B) Illustration of the task execution. During the task execution, the subject starts in the initial position (left panel). Subject reaches for the block (small black square) at the block position (middle panel) and places the block at the end position (right panel). The small rectangles on the subject (left panel) indicate the positions of the sensors. The large black square at the side of the subject indicates the position of the electromagnetic source. The dashed line represents the maximum reaching distance of the arm. Abbreviation: MRD, maximum reaching distance.
      Double-sided adhesive tape was used to attach the motion sensors to the thorax and to 6 segments of the paretic arm of each patient: scapula, upper arm, forearm, hand, thumb, and index finger. The configuration of the sensors is depicted in figure 1. This study focused on the forearm, hand, and finger sensors. A pointer device (ST8 stylusa) was used for anatomic calibration before each measurement to locate the basis of the third metacarpal bone relative to the hand sensor, as the reference position of the hand. The locations of the tips of the thumb and index finger were digitized with respect to their associated finger sensors. The sampling frequency during the motion recordings was 240Hz.

      Procedure

      While seated at the table with a height of 76cm, participants performed a functional reaching task with the paretic arm. Reaching and grasping form 2 major components in many activities of daily living that are performed with the upper limb. Furthermore, objects are often replaced after they have been grasped—for instance, during eating and cleaning. Therefore, we used a task that consisted of reaching, grasping, and the replacement of an object. Specifically, the task consisted of 2 parts: (1) a reach-to-grasp movement toward a block, followed by (2) a displacement of the block toward a target location. The reach-to-grasp movement started with the hand in front of the shoulder on the edge of the table, keeping the thumb against the index finger. Participants were asked to grasp and displace a cubic block of 5×5×5cm and 150g after the experimenter gave a verbal “go” signal. The block was placed in front of the shoulder at each participant's individual maximum reaching distance of the nonparetic arm (see fig 1).
      The reach-to-grasp movement ended successfully when the block had been grasped and had lost contact with the table. Immediately after this block lift, the reach-to-grasp movement proceeded to the second part of the movement, during which the block had to be displaced toward a target position located at the contralateral side (see fig 1). Participants were instructed to grasp the block between their thumb and index finger and not to slide their hand over the table but to move it through the air. After the “go” signal, participants were allowed to move their trunk away from the back of the chair if this was more comfortable, but they had to remain seated and were not allowed to slide or twist over the seat of the chair. Seven trials were recorded in each measurement.

      Data analysis

      The analysis focused on the first part of the experimental paradigm: the reach-to-grasp movement. The start of reach-to-grasp was defined as the moment at which the forearm sensor exceeded 5% of the maximum speed during the forward reach. The end of reach-to-grasp was given as the moment at which the block lost contact with the table and the displacement of the block started. The end of reach-to-grasp was therefore defined as the moment at which the forearm sensor exceeded 5% of the maximum speed during the displacement of the block. The time series for displacement of the hand and for grip aperture were determined between the start and end of reach-to-grasp and were filtered with a second-order Butterworth low-pass filter with a cutoff frequency of 20Hz. All data analyses were performed using custom-made algorithms in Matlab version R2006a.b

      Kinematic parameters

      Movement duration (MD) was used as an overall parameter to quantify reach-to-grasp performance and was defined as the time between the start and end of reach-to-grasp. The smoothness of hand displacement and grasp aperture was quantified by jerk measures. Mathematically, jerk is defined as the third derivative of a specific position variable. To quantify the amount of jerk in the hand displacement and grasp aperture, the jerk was squared and integrated over the total movement duration:
      Jhand=12tstarttendjerkhand2(t)dt
      (1)


      Jgrasp=12tstarttendjerkgrasp2(t)dt
      (2)


      where Jhand and Jgrasp represent the amount of jerk in the hand displacement and grasp aperture signal, respectively; tstart represents the moment of start of reach-to-grasp; tend represents the moment of end of reach-to-grasp; jerkhand(t) represents the third derivative of hand displacement; and jerkgrasp(t) represents the third derivative of grasp aperture.
      As Hogan and Sternad
      • Hogan N.
      • Sternad D.
      Sensitivity of smoothness to movement duration, amplitude, and arrests.
      point out, this jerk measure depends on movement length squared divided by the fifth power of movement duration, L2/MD5, even when the shape of the hand trajectory or grasp aperture curve is invariant. This dependency on movement length and MD is not desired, and therefore Jhand and Jgrasp need to be normalized, which is obtained by the following equations
      • Caimmi M.
      • Carda S.
      • Giovanzana C.
      • et al.
      Using kinematic analysis to evaluate constraint-induced movement therapy in chronic stroke patients.
      • Hogan N.
      • Sternad D.
      Sensitivity of smoothness to movement duration, amplitude, and arrests.
      :
      NJhand=12tstarttendjerkhand2(t)dt×MD5/Lhand2
      (3)


      NJgrasp=12tstarttendjerkgrasp2(t)dt×MD5/Lgrasp2
      (4)


      where NJhand and NJgrasp represent the normalized jerk for hand displacement and grasp aperture, respectively; Lhand represents the shortest distance between the start and end positions of the hand; and Lgrasp represents the difference in grasp aperture between the start and end of reach-to-grasp. The subsequent statistical analyses were performed on the normalized jerk measures (ie,NJhand andNJgrasp).

      Statistics

      Using random coefficient analysis (SPSS, version 20.0c), we modeled the longitudinal recovery profiles for MD, NJhand, and NJgrasp. To check whether our data met the assumptions for normality, we plotted the frequency distribution of the parameters and compared these plots visually with a normal distribution. We assessed how each parameter changed as a function of time poststroke using the following multivariate regression model that corrected for the covariates age and baseline value:
      Yij=(β0+b0i)+(β1j+b1ij)×Xij+β2x×COVxi+εij
      (5)


      where Yij is a 3-dimensional kinematic parameter for subject i at time point j, β0 is a fixed intercept and b0i is a random intercept for subject i, β1j is a fixed regression coefficient for time point j, b1ij is a random regression coefficient for subject i and time point j, and Xij is a dummy variable for subject i and time point j. The last time point (ie, 26wk poststroke) was used as the reference time point.β2x is a fixed regression coefficient corresponding to covariate x. COVxi is the value of covariate x for subject i. The covariates were (1) age at baseline, since older people may move more slowly than younger people; and (2) the baseline value of the kinematic parameter (Yi,baseline), in order to correct for between-subject variance. Both covariates were mean centered to obtain β1 coefficients that correspond to the mean age and baseline scores.εij specifies the residual value for subject i at time point j.
      The restricted maximum likelihood method was used in combination with a first-order homogeneous autoregressive covariance structure to fit our model (equation 5) onto the data. This covariance structure assumes that correlation between time points decreases with increasing intervals between time points, which is generally the case in longitudinal studies. For each kinematic parameter, the time window for change was defined as the series of time points with a significant β1 coefficient. The significance of each regression coefficient β was tested using the t statistic, which was given as β divided by its SE. The degrees of freedom were computed using the Satterthwaite's approximation. Level of significance was set 2-sided at P<.01.

      Results

      Table 1 shows the baseline characteristics of the 44 patients included (ie, 19 women, 25 men). The mean age ± SD was 58±12 years. All patients were able to reach and grasp objects and were measured once a week during the first 5 weeks poststroke and subsequently in weeks 8, 12, and 26 poststroke. In total, 293 of the 352 kinematic measurements were conducted. Figure 2 shows hand trajectories and the grasp aperture profiles at weeks 1, 5, and 26 poststroke of 1 patient. This example shows that most change in the smoothness of hand trajectories and grasp aperture occurs between week 1 and week 5 poststroke.
      Table 1Participant characteristics
      CharacteristicsValues
      N44
      Missing measurements at each time point
       Wk 123
       Wk 212
       Wk 34
       Wk 48
       Wk 50
       Wk 87
       Wk 124
       Wk 261
      Sex (F/M)19/25
      Age (y)58±12
      Paretic body side (L/R)27/17
      Type of stroke (Bamford)
       LACI31
       PACI9
       TACI4
      NIHSS total score4 (2–5)
      Cognitive disturbance
       Disorientation (NIHSS, item 1), no/yes44/0
       Inattention (NIHSS, item 11), no/yes40/4
      Impairments of vision
       Hemianopia (NIHSS, item 3), no/yes44/0
       Deviation conjugee (NIHSS, item 2), no/yes43/1
      FMA upper limb (0–66)41 (29–54)
      ARAT total score (0–57)21.5 (7.25–36)
      BI total score (0–20)15 (10–17)
      NOTE. Values are n, mean ± SD, or median (interquartile range).
      Abbreviations: ARAT, Action Research Arm Test; BI, Barthel Index; F, female; FMA, Fugl-Meyer Motor Assessment; L, left; LACI, lacunar anterior cerebral infarction; M, male; NIHSS, National Institutes of Health Stroke Scale; PACI, partial anterior cerebral infarction; R, right; TACI, total anterior cerebral infarction.
      Figure thumbnail gr2
      Fig 2Reaching trajectories of the hand and grasping profiles of 1 patient in weeks 1, 5, and 26 after stroke onset. The black square represents the position of the block. Each trace represents 1 repetition of the reach-to-grasp movement. Note that the improvement in smoothness of the reaching trajectories and grasp aperture profiles is larger between week 1 and week 5, compared with week 5 and week 26 poststroke.
      Frequency distributions of each kinematic parameter showed that MD was normally distributed, whereas NJhand and NJgrasp were not. Therefore, NJhand and NJgrasp were log-transformed to meet the assumptions for normality. Table 2 shows the regression coefficients for each kinematic parameter and each time point poststroke. The regression coefficients for each time point poststroke indicate the mean difference in the kinematic parameter between that time point and week 26. The regression coefficients for the covariates yield a correction factor for the difference between the actual and mean age and baseline value. Figure 3 depicts the longitudinal change in the regression coefficients. For MD, the regression coefficients for time poststroke decreased as a function of time poststroke. Up to week 5 poststroke, MD was significantly larger than MD at week 26 poststroke. The regression coefficients for log(NJhand) and log(NJgrasp) paralleled this decrease in MD as a function of time poststroke. In addition, up to week 5 poststroke, log(NJhand) and log(NJgrasp) were significantly larger than log(NJhand) and log(NJgrasp) at week 26 poststroke.
      Table 2Regression coefficients for recovery of MD and log-transformed values of normalized jerk of hand displacement and grasp aperture
      Time Points and CovariatesMD (s)Log(NJhand)Log(NJgrasp)
      Intercept1.17±.10 (<.001)
      P<.01.
      2.004±.063 (<.001)
      P<.01.
      3.668±.069 (<.001)
      P<.01.
      Wk 11.63±.14 (<.001)
      P<.01.
      1.074±.077 (<.001)
      P<.01.
      0.894±.087 (<.001)
      P<.01.
      Wk 21.08±.12 (<.001)
      P<.01.
      0.754±.068 (<.001)
      P<.01.
      0.612±.076 (<.001)
      P<.01.
      Wk 30.55±.12 (<.001)
      P<.01.
      0.396±.063 (<.001)
      P<.01.
      0.294±.071 (<.001)
      P<.01.
      Wk 40.41±.12 (.001)
      P<.01.
      0.323±.062 (<.001)
      P<.01.
      0.201±.071 (.005)
      P<.01.
      Wk 50.32±.11 (.004)
      P<.01.
      0.253±.057 (<.001)
      P<.01.
      0.181±.066 (.007)
      P<.01.
      Wk 80.10±.11 (.373)0.084±.055 (.129)−0.030±.064 (.643)
      Wk 120.04±.11 (.679)0.062±.046 (.176)0.005±.053 (.931)
      Wk 260
      This parameter is set to 0 because it is redundant.
      0
      This parameter is set to 0 because it is redundant.
      0
      This parameter is set to 0 because it is redundant.
      Age0.001±.01 (.850)0.001±.005 (.765)0.001±.005 (.878)
      Baseline score0.12±.05 (.040)0.10±.098 (.331)0.161±.095 (.099)
      NOTE. Values are mean ± SE (P).
      P<.01.
      This parameter is set to 0 because it is redundant.
      Figure thumbnail gr3
      Fig 3Change in MD and smoothness of hand transport (log(NJhand)) and grasp aperture (log(NJgrasp)) as a function of time poststroke. The contribution of time poststroke was significant for all 3-dimensional kinematic parameters until week 8 poststroke (gray).
      The regression coefficients for the covariates age and baseline were not significant for all kinematic parameters.

      Discussion

      To our knowledge, the present study is the first to show, by means of an intensively repeated 3-dimensional kinematic measures design, that the time courses of smoothness of reaching and grasping follow a natural logistic pattern of recovery poststroke. Importantly, the progress of time contributed significantly to improvements in smoothness until week 8 poststroke. This finding suggests that improvement in quality of motor control, like improvement in motor synergism,
      • Twitchell T.E.
      The restoration of motor function following hemiplegia in man.
      • Duncan P.W.
      • Goldstein L.
      • Matchar D.
      • Divine G.
      • Feussner J.
      Measurement of motor recovery after stroke. Outcome assessment and sample size requirements.
      • Prabhakaran S.
      • Zarahn E.
      • Riley C.
      • et al.
      Inter-individual variability in the capacity for motor recovery after ischemic stroke.
      is time dependent and parallels spontaneous neurologic recovery.
      • Langhorne P.
      • Bernhardt J.
      • Kwakkel G.
      Stroke rehabilitation.
      This finding is in agreement with those of earlier studies based on clinical assessment scales, which showed that the effect of progress of time, as a reflection of spontaneous neurologic recovery, is restricted to the first 6 to 10 weeks poststroke.
      • Kwakkel G.
      • Kollen B.J.
      • Twisk J.
      Impact of time on improvement of outcome after stroke.
      In addition, it is in line with a recent study
      • Alt Murphy M.
      • Willén C.
      • Sunnerhagen K.S.
      Responsiveness of upper extremity kinematic measures and clinical improvement during the first three months after stroke.
      showing that smoothness of hand trajectory is a responsive measure for capturing improvements in upper limb coordination early poststroke.
      Smooth coordination patterns are an important characteristic of skilled motor behavior.
      • Schmidt R.A.
      A schema theory of discrete motor skill learning.
      Moreover, computer simulations of reach-to-grasp behavior suggest that minimization of jerk is used as an important guiding principle by humans to select the optimal coordination strategy.
      • Rohrer B.
      • Fasoli S.E.
      • Krebs H.I.
      • et al.
      Movement smoothness changes during stroke recovery.
      • Todorov E.
      Optimality principles in sensorimotor control [review].
      Therefore, the improvements in smoothness early poststroke suggest that spontaneous neurologic recovery leads to optimization of the strategies to control the various degrees of freedom of the paretic upper limb and, with that, to an improved quality of motor control. These improvements in quality of motor control early poststroke further support the hypothesis that spontaneous neurologic recovery leads to restitution of motor function.
      Several mechanisms may be responsible for the observed improvements in smoothness of reach-to-grasp coordination. A previous study
      • Van Kordelaar J.
      • van Wegen E.E.H.
      • Nijland R.H.M.
      • Daffertshofer A.
      • Kwakkel G.
      Understanding adaptive motor control of the paretic upper limb early poststroke: the EXPLICIT-stroke program.
      by our group has shown that the ability to make dissociated shoulder and elbow movements during reach-to-grasp increases in the first 5 weeks poststroke. This improvement in the control over the degrees of freedom in the paretic upper limb may lead to more optimal coordination strategies, reflected by the improved smoothness of upper limb coordination patterns. The observed reductions in smoothness early poststroke may also reflect deficiencies in motor unit recruitment. These deficiencies include reduced discharge rates
      • Tang A.
      • Rymer W.Z.
      Abnormal force—EMG relations in paretic limbs of hemiparetic human subjects.
      • Gemperline J.J.
      • Allen S.
      • Walk D.
      • Rymer W.Z.
      Characteristics of motor unit discharge in subjects with hemiparesis.
      and spontaneous firing
      • Mottram C.J.
      • Wallace C.L.
      • Chikando C.N.
      • Rymer W.Z.
      Origins of spontaneous firing of motor units in the spastic-paretic biceps brachii muscle of stroke survivors.
      of motor units in the paretic upper limb and may lead to inaccurate control of the force output of the paretic muscles. Finally, patients with stroke are assumed to rely more on proprioceptive and visual feedback to make online corrections at the end of the reach-to-grasp movement.
      • Van Vliet P.M.
      • Sheridan M.R.
      Coordination between reaching and grasping in patients with hemiparesis and healthy subjects.
      Improvements in smoothness may therefore suggest that online corrections decrease during reaching movements. The above biological mechanisms underlying jerk support the hypothesis that the severity of jerk reflects the amount of noise in the voluntary motor control after stroke.
      • McCrea P.H.
      • Eng J.J.
      Consequences of increased neuromotor noise for reaching movements in persons with stroke.

      Study limitations

      The present study had some limitations. First, we excluded all patients who were unable to reach and grasp within 3 weeks poststroke. The present results can therefore only be generalized to patients with a mild hemiparesis of the paretic upper limb and with a favorable prognosis for recovery of upper limb function.
      • Nijland R.H.M.
      • van Wegen E.E.H.
      • Harmeling-van der Wel B.C.
      • Kwakkel G.
      Presence of finger extension and shoulder abduction within 72 hours after stroke predicts functional recovery: early prediction of functional outcome after stroke: the EPOS cohort study.
      • Stinear C.
      Prediction of recovery of motor function after stroke.
      Second, since we used 1 discrete reaching paradigm, it is not possible to determine whether the present results can be generalized to other functional or rhythmic upper limb tasks. Third, the random effects model that was used in the present study assumes missing values to be missing at random. However, in weeks 1 and 2 poststroke, a number of patients, in particular with a severe motor deficit at stroke onset, were not able to conduct the reach-to-grasp task, which resulted in nonrandomly missing values at these time points. An identical analysis of complete cases in the present study (n=28) yielded also significant regression coefficients for time poststroke up to week 4, 5, and 2 poststroke for MD, log(NJhand) and log(NJgrasp), respectively. This time window of the complete cases is shorter compared with the time window of the whole sample of patients. This latter finding suggests that the period of spontaneous neurologic recovery is relatively shorter in cases with milder deficits when compared with cases with more severe strokes. We argue that these complete cases mainly represent stroke patients with only a mild to moderate deficit of the upper paretic limb. Future studies should therefore also include less challenging paradigms, such as reaching tasks without a grasping component, in order to avoid missing values in the first weeks after stroke. Fourth, the present study did not investigate the spontaneous intrinsic cerebral recovery by measuring structural or functional changes in the brain. Future studies should investigate what changes in the brain may be responsible for improvements in the quality of motor control.
      • Kwakkel G.
      • Meskers C.G.M.
      • van Wegen E.E.H.
      • et al.
      Impact of early applied upper limb stimulation: the EXPLICIT-stroke programme design.
      Finally, the present study did not correct for the type or amount of therapy that the patients received, acknowledging that knowledge on the impact of early rehabilitation interventions on the neurologic repair of motor control is still lacking in the literature.
      • Langhorne P.
      • Bernhardt J.
      • Kwakkel G.
      Stroke rehabilitation.
      • Van Vliet P.
      • Pelton T.A.
      • Hollands K.L.
      • Carey L.
      • Wing A.M.
      Neuroscience findings on coordination of reaching to grasp an object: implications for research.

      Conclusions

      The results of the present study showed that until 8 weeks poststroke, progress of time poststroke contributes significantly to reductions in normalized jerk of hand and finger movements during reach-to-grasp. This finding suggests that spontaneous neurologic recovery in the first weeks poststroke in the brain leads to normalization of motor coordination of the paretic upper limb. Future studies with intensively repeated measures designs should investigate whether a very early start of therapies targeting upper limb function is able to modulate the quality of motor control beyond spontaneous neurologic recovery.
      • Kwakkel G.
      • Meskers C.G.M.
      • van Wegen E.E.H.
      • et al.
      Impact of early applied upper limb stimulation: the EXPLICIT-stroke programme design.

      Suppliers

      • a.
        Polhemus, 40 Hercules Dr, PO Box 560, Colchester, VT 05446.
      • b.
        Matlab version R2006a; MathWorks, 3 Apple Hill Dr, Natick, MA 01760-2098.
      • c.
        SPSS version 20.0; IBM Corp, 1 New Orchard Rd, Armonk, NY 10504-1722.

      References

        • Kwakkel G.
        • Kollen B.J.
        • Twisk J.
        Impact of time on improvement of outcome after stroke.
        Stroke. 2006; 37: 2348-2353
        • Kwakkel G.
        • Kollen B.J.
        • Lindeman E.
        Understanding the pattern of functional recovery after stroke: facts and theories.
        Restor Neurol Neurosci. 2004; 22: 281-299
        • Twitchell T.E.
        The restoration of motor function following hemiplegia in man.
        Brain. 1951; 74: 443-480
        • Duncan P.W.
        • Goldstein L.
        • Matchar D.
        • Divine G.
        • Feussner J.
        Measurement of motor recovery after stroke. Outcome assessment and sample size requirements.
        Stroke. 1992; 23: 1084-1089
        • Phan T.G.
        • Wright P.M.
        • Markus R.
        • Howells D.W.
        • Davis S.M.
        • Donnan G.A.
        Salvaging the ischaemic penumbra: more than just reperfusion?.
        Clin Exp Pharmacol Physiol. 2002; 29: 1-10
        • Andrews R.J.
        Transhemispheric diaschisis. A review and comment.
        Stroke. 1991; 22: 943-949
        • Krakauer J.W.
        • Carmichael S.T.
        • Corbett D.
        • Wittenberg G.F.
        Getting neurorehabilitation right: what can be learned from animal models?.
        Neurorehabil Neural Repair. 2012; 26: 923-931
        • Van Kordelaar J.
        • van Wegen E.E.H.
        • Nijland R.H.M.
        • et al.
        Assessing longitudinal change in coordination of the paretic upper limb using on-site 3-dimensional kinematic measurements.
        Phys Ther. 2012; 92: 142-151
        • Alt Murphy M.
        • Willén C.
        • Sunnerhagen K.S.
        Kinematic variables quantifying upper-extremity performance after stroke during reaching and drinking from a glass.
        Neurorehabil Neural Repair. 2011; 25: 71-80
        • Caimmi M.
        • Carda S.
        • Giovanzana C.
        • et al.
        Using kinematic analysis to evaluate constraint-induced movement therapy in chronic stroke patients.
        Neurorehabil Neural Repair. 2008; 22: 31-39
        • Rohrer B.
        • Fasoli S.E.
        • Krebs H.I.
        • et al.
        Movement smoothness changes during stroke recovery.
        J Neurosci. 2002; 22: 8297-8304
        • Levin M.F.
        Interjoint coordination during pointing movements is disrupted in spastic hemiparesis.
        Brain. 1996; 119: 281-293
        • Van Kordelaar J.
        • van Wegen E.E.H.
        • Kwakkel G.
        Unraveling the interaction between pathological upper limb synergies and compensatory trunk movements during reach-to-grasp after stroke: a cross-sectional study.
        Exp Brain Res. 2012; 221: 251-262
        • Kwakkel G.
        • Meskers C.G.M.
        • van Wegen E.E.H.
        • et al.
        Impact of early applied upper limb stimulation: the EXPLICIT-stroke programme design.
        BMC Neurol. 2008; 8: 49
        • Bamford J.
        • Sandercock P.
        • Dennis M.
        • Burn J.
        • Warlow C.
        Classification and natural history of clinically identifiable subtypes of cerebral infarction.
        Lancet. 1991; 337: 1521-1526
        • Goldstein L.
        • Bertels C.
        • Davis J.
        Interrater reliability of the NIH Stroke Scale.
        Arch Neurol. 1989; 46: 660-662
        • Fugl-Meyer A.R.
        • Jääskö L.
        • Leyman I.
        • Olsson S.
        • Steglind S.
        The post-stroke hemiplegic patient 1. A method for evaluation of physical performance.
        Scand J Rehabil Med. 1975; 7: 13-31
        • Lyle R.C.
        A performance test for assessment of upper limb function in physical rehabilitation treatment and research.
        Int J Rehabil Res. 1981; 4: 483-492
        • Collin C.
        • Wade D.T.
        • Davies S.
        • Horne V.
        The Barthel ADL Index: a reliability study.
        Int Disabil Stud. 1988; 10: 61-63
        • Hogan N.
        • Sternad D.
        Sensitivity of smoothness to movement duration, amplitude, and arrests.
        J Mot Behav. 2009; 41: 529-534
        • Prabhakaran S.
        • Zarahn E.
        • Riley C.
        • et al.
        Inter-individual variability in the capacity for motor recovery after ischemic stroke.
        Neurorehabil Neural Repair. 2008; 22: 64-71
        • Langhorne P.
        • Bernhardt J.
        • Kwakkel G.
        Stroke rehabilitation.
        Lancet. 2011; 377: 1693-1702
        • Alt Murphy M.
        • Willén C.
        • Sunnerhagen K.S.
        Responsiveness of upper extremity kinematic measures and clinical improvement during the first three months after stroke.
        Neurorehabil Neural Repair. 2013; 27: 844-853
        • Schmidt R.A.
        A schema theory of discrete motor skill learning.
        Psychol Rev. 1975; 82: 250-260
        • Todorov E.
        Optimality principles in sensorimotor control [review].
        Nat Neurosci. 2004; 7: 907-915
        • Van Kordelaar J.
        • van Wegen E.E.H.
        • Nijland R.H.M.
        • Daffertshofer A.
        • Kwakkel G.
        Understanding adaptive motor control of the paretic upper limb early poststroke: the EXPLICIT-stroke program.
        Neurorehabil Neural Repair. 2013; 27: 854-863
        • Tang A.
        • Rymer W.Z.
        Abnormal force—EMG relations in paretic limbs of hemiparetic human subjects.
        J Neurol Neurosurg Psychiatry. 1981; 44: 690-698
        • Gemperline J.J.
        • Allen S.
        • Walk D.
        • Rymer W.Z.
        Characteristics of motor unit discharge in subjects with hemiparesis.
        Muscle Nerve. 1995; 18: 1101-1114
        • Mottram C.J.
        • Wallace C.L.
        • Chikando C.N.
        • Rymer W.Z.
        Origins of spontaneous firing of motor units in the spastic-paretic biceps brachii muscle of stroke survivors.
        J Neurophysiol. 2010; 104: 3168-3179
        • Van Vliet P.M.
        • Sheridan M.R.
        Coordination between reaching and grasping in patients with hemiparesis and healthy subjects.
        Arch Phys Med Rehabil. 2007; 88: 1325-1331
        • McCrea P.H.
        • Eng J.J.
        Consequences of increased neuromotor noise for reaching movements in persons with stroke.
        Exp Brain Res. 2005; 162: 70-77
        • Nijland R.H.M.
        • van Wegen E.E.H.
        • Harmeling-van der Wel B.C.
        • Kwakkel G.
        Presence of finger extension and shoulder abduction within 72 hours after stroke predicts functional recovery: early prediction of functional outcome after stroke: the EPOS cohort study.
        Stroke. 2010; 41: 745-750
        • Stinear C.
        Prediction of recovery of motor function after stroke.
        Lancet Neurol. 2010; 9: 1228-1232
        • Van Vliet P.
        • Pelton T.A.
        • Hollands K.L.
        • Carey L.
        • Wing A.M.
        Neuroscience findings on coordination of reaching to grasp an object: implications for research.
        Neurorehabil Neural Repair. 2013; 27: 622-635