Exp Brain Res (2004) 157: 162–173DOI 10.1007/s00221-004-1829-x

RESEARCH ARTICLES

Stella M. Michaelsen . Stéphane Jacobs .Agnès Roby-Brami . Mindy F. Levin

Compensation for distal impairments of grasping in adults withhemiparesis

Received: 16 May 2003 / Accepted: 13 December 2003 / Published online: 19 February 2004# Springer-Verlag 2004

Abstract Previous studies have shown that patients witharm and hand paresis following stroke recruit an additionaldegree of freedom (the trunk) to transport the hand duringreaching and use alternative strategies for grasping. Thefew studies of grasping parameters of the impaired handhave been case studies mainly focusing on describinggrasping in the presence of particular impairments such ashemi-neglect or optic ataxia and have not focussed on therole of the trunk in prehension. We hypothesized that thetrunk movement not only ensures the transport of the handto the object, but it also assists in orienting the hand forgrasping when distal deficits are present. Nineteen patientswith chronic hemiparesis and seven healthy subjectsparticipated in the study. Patients had sustained a strokeof non-traumatic origin 6–82 months previously (31±22 months) and had mild or moderate to severe armparesis. Using a whole hand grasp, subjects reached andgrasped a cylinder (35 mm) that was placed sagittally (T1)or at a 45° angle to the sagittal midline in the ipsilateralworkspace (T2), both at about 90% arm’s length (10 trialsper target). Eight infrared emitting diodes were placed onbony landmarks of the hand, arm and trunk and kinematicdata were recorded by an optical motion analysis system(Optotrak) for 2–5 s at 120 Hz. Hand position andorientation were recorded by a Fastrack Polhemus system.

Our results show that during goal-directed prehensiontasks, individuals with hemiparesis oriented the hand morefrontally for grasping and used more trunk anteriordisplacement or rotation to transport the hand to the targetcompared to healthy subjects. Despite these changes, themajor characteristics of reaching and grasping such as gripaperture size, temporal coordination between hand trans-port and aperture formation and the relative timing of gripaperture were largely preserved. For patients with moresevere distal impairments, the amount of trunk displace-ment was also correlated with a more frontal handorientation for grasping. Furthermore, in healthy subjectsand patients without distal impairments, the trunk move-ment was mostly related to proximal arm movementswhile in those with distal impairments, trunk movementwas related to both proximal and distal arm movements.Data support the hypothesis that the trunk movement isused to assist both arm transport and hand orientation forgrasping when distal deficits are present.

Keywords Grasping . Hand orientation . Coordination .Hemiplegia . Compensatory movement

Introduction

The ability to reach and grasp is a necessary component ofmany daily-life functional tasks. A major variableaffecting hand transport is the final hand posture ororientation required for grasping and is related tocomfortable end postures (cf. Rosenbaum et al. 1996,2001). Despite the large number degrees of freedom (DFs)of the upper limb, the final posture of the arm and hand isremarkably stable for a given object orientation (Desmur-get and Prablanc 1997) and location (Gréa et al. 2000). Forprehension, aside from the transport of the hand to thetarget, two other components have to be controlled: thehand orientation component permitting the alignment ofthe finger-thumb opposition axis with that of the object(Gentilucci et al. 1991, 1996; Paulignan et al. 1997), andthe grasp component consisting of the selection and

S. M. MichaelsenSchool of Rehabilitation, University of Montreal ResearchCentre, Rehabilitation Institute of Montreal,6300 Darlington,Montreal, Quebec, H3S 2J4, Canada

S. Jacobs . A. Roby-BramiHôpital Raymond-Poincaré, CNRS UMR 8119,45 rue des Saints Pères,75270 Garches Cedex 06, France

M. F. Levin (*)Research Centre of Rehabilitation Institute of Montreal, CRIR,6300 Darlington,Montreal, Quebec, H3S 2J4, Canadae-mail: levinm@poste.umontreal.caTel.: +1-514-3402780Fax: +1-514-3402154

control of the finger grip aperture according to the size andthe shape of the object (Jeannerod 1984; Paulignan andJeannerod 1996). Based on the traditional visuomotorchannels hypothesis, transport, orientation and graspingcomponents are planned separately but coordinated in time(Jeannerod 1984; Hoff and Arbib 1993). In this view, thereis a clear separation between the control of proximal jointsassociated with hand transport and of distal joints relatedto hand orientation and grasping. Visual information aboutextrinsic and intrinsic properties of the object is used tocontrol respectively the proximal musculature to place thehand in the correct spatial location and the distal forearmand hand musculature to orient the hand and fingers.However, studies showing that unexpected changes inobject location (Paulignan et al. 1991) or distance(Jakobson and Goodale 1991; Chieffi and Gentilucci1993) can affect grip size and that changes in object sizecan affect hand transport (Castiello et al. 1998) havequestioned the idea of separate control channels. Indeed, ithas been suggested that hand transport and orientation arecontrolled together (Desmurget et al. 1996) and that theyvary with reaching direction (Roby-Brami et al. 2000).This is supported by findings that scaling of joint rotationto movement distance or direction is shared among mostDFs of the arm (Roby-Brami et al. 2003a).

After stroke and consequent hemiparesis affecting theupper limb, the ability to produce functional arm and handmovements is impaired (Wade et al. 1983). In order tocompensate the upper limb impairment, participants withhemiparesis can use alternative strategies to improvefunctional arm and hand use. For example, when the activerange of arm motion is decreased, individuals cantransport the hand to the object by using the trunk(Roby-Brami et al. 1997; Cirstea and Levin 2000;Michaelsen et al. 2001). Studies have shown that thisadditional trunk recruitment is compensatory because itallows the patient to bring the hand to the target ratherprecisely even when active movements at the affectedelbow and shoulder are restricted or impossible (Cirsteaand Levin 2000; Michaelsen et al. 2001). Further evidencethat the trunk plays a compensatory role in reaching stemsfrom findings that patients who use altered strategies oftrunk recruitment do not have a primary deficit in trunkcontrol (Esparza et al. 2003) and that arm extension maybe improved and trunk use diminished when appropriateinterventions such as guided practice (Cirstea et al. 2003)and trunk restraint (Michaelsen et al. 2001) are applied. Ithas also been suggested that individuals with hemiparesisnot only use compensatory trunk displacement to reachtoward and place the hand in the correct spatial locationbut, faced with distal muscle weakness and a lack of fingercontrol, they develop new strategies of grasping. Windingfingers around the object, sliding the hand along a surfaceand downward grasping are some examples of differentcompensatory grasping strategies (Roby-Brami et al.1997).

The way in which the additional trunk DFs areintegrated into the reaching pattern in patients withhemiparesis suggests that the CNS takes into account the

biomechanical restrictions of the limb in motor planningand forms a new coordinative structure including thetrunk. Coordinative structures are defined as the self-organization of functional ensembles of different DFs, inwhich each DF may participate in a motor task accordingto its potential contribution to that task (Gelfand andTsetlin 1971; Kugler et al. 1980). Levin et al. (2002)showed that in patients with hemiparesis, trunk DFs areintegrated into the reaching pattern for targets placedwithin the reach of the arm in a similar way as for reachesbeyond the reach in healthy individuals (Wang andStelmach 1998, 2001; Rossi et al. 2002). Thus, in additionto shoulder, arm and hand DFs, a model of reaching inpatients with hemiparesis should include DFs related tomovement of the trunk.

Many studies have addressed reaching deficits of thehemiparetic arm (Trombly 1992; van Vliet et al. 1995;Roby-Brami et al. 1997; 2003a, 2003b; Michaelsen et al.2001; Levin et al. 2002) but few have analysed thegrasping component. Exceptions are Gentilucci et al.(2000) and Binkofski et al. (1998), who studied onlypatients without post-stroke hand deficits and Steenbergenet al. (2000), who identified grasping deficits inadolescents with hemiparesis due to cerebral palsy.Although previous studies have examined the relationshipbetween the direction of reaching and hand transport andorientation in adults with hemiparesis (Roby-Brami et al.2003a), it is unknown how changes in arm transport affectthe parameters of grasping in these patients. Thus, the firstgoal of this study was to describe parameters of graspingand the coordination of reaching and grasping in patientswith different degrees of hand impairment due to a stroke.The second goal was to determine the relationshipsbetween compensatory movements of the trunk andclinical arm and hand motor impairments. We hypothe-sized that the compensatory trunk movements used byindividuals with hemiparesis for hand transport are alsoused to orient the hand for grasping when distal (hand)impairments are present. Thus, the third goal of the studywas to determine the role of the trunk in compensatinggrasping as well as reaching deficits. Preliminary datahave been presented in abstract form (Michaelsen et al.2002).

Materials and methods

Nineteen patients (52±19 years) with chronic hemiparesis (ten menand nine women) and seven neurologically healthy subjects (53±24 years) participated in the study after signing a consent formapproved by local hospital Ethics Committees. We did not study alarger number of healthy subjects since reaching and graspingparameters to stationary targets have already been well described.We included a small group, however, in order to have task specificcomparative data for individuals with hemiparesis. Patients hadsustained a stroke of non-traumatic origin 6–82 months previously,mean 31±22 months, and had mild or moderate to severe upper limbparesis (Fugl-Meyer score ≥26/66 on the upper limb subscale). Allpatients were able to reach and had gross prehension ability. Patientshad no hemispatial neglect or apraxia and were able to understandsimple instructions. Those with shoulder pain or other neurological

163

Tab

le1

Dem

ograph

icdata

andresults

ofclinical

testing(FM

Fug

l-Meyer

Scale,BBTBox

andBlocksTest)forparticipantswith

hemiparesis.B

BTvalues

show

ageandsexno

rmsfor

thedo

minant/n

on-dom

inantor

non-do

minant/d

ominanthand

saccordingto

thedo

minance

oftheaffected

hand

(Rrigh

t,Lleft,A

affected

arm,L

Aless-affectedarm,n

tnot

tested,C

VAcerebrov

ascularaccident,MCAmiddlecerebral

artery,AVM

arteriov

enou

smalform

ation)

SSex/age

(years)

Tim

esince

stroke

(mon

ths)

Site

ofstroke

FM

score

arm

(36)

FM

score

wrist(10)

FM

score

hand

(14)

BBT

(blocks/min)

(A/LA)

BBT

(age

and

sexno

rms)

Gripstreng

th(kgF

)(A

/LA)

Wristextension

streng

th(kgF

)(A

/LA)

1F34

6RMCA,ischem

ic36

1014

43/63

80/85

9.5/32

11.1/15.3

2F39

6LMCA,ischem

ic36

1014

38/65

85/83

19/31

8.4/17

.53

M60

7LMCA,ischem

ic34

1014

35/61

71/70

14/50

2.7/15

.74

F49

43Lparietal/sub

cortical

ischem

ic30

1014

43/72

82/78

14/30

3.2/11.2

5M

6732

LMCA,ischem

ic29

1014

31/43

68/67

13/38

8.6/14

.56a

F79

32LCVA

1810

1329

/58

65/64

6.5/20

2.2/16

.27a

M72

14RCVA

1710

70/38

63/64

14/42

3.2/14

.68a

F25

19Lparietal,hemorrhagic

169

1219

/58

86/81

nt2.0/14

.39

F69

82LCVA

309

1134

/59

72/71

5/22

3.4/8.4

10M

6015

Ltempo

roparietal

259

60/59

70/71

0/26

3.0/9.0

11F62

11LCVA

248

146/74

76/74

10/21

4.1/11.5

12M

6820

RCVA

168

1210

/41

67/68

7/33

4.8/12

.713

F23

26LMCA,ischem

ic31

76

5/76

88/83

11/26

5.1/9.4

14M

7448

Ltempo

ral,ischem

ic24

414

24/53

66/64

19/35

7.4/12

.615

M61

10Lthalam

us,cortical/sub

cortical

155

80/41

71/70

2/28

3.5/15

.516

aM

2040

RAVM,parieto-occipital,hemorrhagic

124

46/66

86/88

7/42

0/10

.717

M57

58Rinternal

capsule,

ischem

ic25

03

0/60

74/75

8/34

0/13

.718

aM

5351

LCVA

220

20/62

79/77

3/50

0/18

.519

aF23

64LAVM,parietal

andsubcortical

hemorrhagic

220

00/50

88/83

6/28

3.0/8.3

X±S

D-

52±1

9-

--

-17

±17/58

±11

-9.0±

0.5/33

.0±9

.04.0±

3.0/13

.1±3

.1

a Participantswho

madereachesto

Target1on

ly

164

or orthopaedic conditions affecting the performance of the task andthose with elbow flexion contracture of more than 5° were excluded.No distinction was made between right- and left-sided lesions sincedifferences in upper limb movement kinematics due to the side ofdamage are generally seen only when task accuracy demands arehigh (Pohl et al. 1997). In our study the task involved whole-handprehension and had low accuracy demands.

Clinical evaluation

Evaluation of arm and hand motor impairments was done by atrained physical therapist using four clinical tests. Residual upperlimb movements were assessed with the upper limb section of theFugl-Meyer Scale (Fugl-Meyer et al. 1975; Berglund and Fugl-Meyer 1986). This section consists of four subitems: I—arm(shoulder/elbow/forearm), II—wrist, III—hand, IV—coordination/speed and also measures the presence or absence of cutaneoussensation and proprioception. To distinguish between proximal anddistal deficits, the scores of the arm, wrist and hand are presentedseparately. For the wrist, the test measures stability in extension,alternating flexion/extension and circumduction for a maximumwrist motor function score of 10. Patients were divided into twogroups according to their scores on the wrist subitem of the Fugl-Meyer scale. According to this classification, seven participants hadno wrist impairment (FM wrist=10; patients 1–7) and 12 had amoderate to severe wrist impairment (FM wrist ≤9, patients 8–19).The hand section evaluates the ability to perform mass flexion, massextension and five different grasps for a total score of 14 points.Manual dexterity was also evaluated using the Box and Blocks Test(BBT) that measures the number of 2.5 cm3 cubes transported in1 min from one side of a box to another. The test was done twice andthe best score was retained for each hand (Mathiowetz et al. 1985).Grip strength was measured with a Jamar dynamometer. Finally,isometric force of the wrist extensors was measured with a hand-held dynamometer (Nicholas, MMT, Lafayette instruments—model01160). For strength testing, the maximal value of three trials of theaffected hand was expressed as a percentage of the maximal forceproduced for the same movement by the contralateral hand.Demographic data and individual scores on upper limb motorfunction tests, including age and sex norms for dominant and non-dominant hands on the BBT test, are presented in Table 1.

Reach-to-grasp task

Reach-to-grasp movements were made from the sitting position totwo targets placed in front of the participant. The trunk was notrestrained. The participant’s hip and knee joints were flexed to 90°with the feet supported on the floor. Reaching and grasping were

done with the affected upper limb of participants with hemiparesisand with the dominant limb of healthy participants. At the start ofthe task, the reaching arm was resting on a support placed on a tableso that the shoulder was in ~10° extension and ~20° abduction(where 0° for each direction is defined as the arm positionedvertically beside the body), the elbow was flexed to ~70° (where thefully outstretched position is 180°), the forearm was pronated andthe wrist was in the neutral position between flexion and extension(Fig. 1A). The contralateral arm rested alongside the body. Reachingand grasping were recorded to an object (35 mm diameter by 95 mmheight) placed at two different locations. Target one (T1) was placeddirectly in the midline of the body and Target two (T2) waspositioned at the same distance as T1 but was displaced by 45°lateral to the midline towards the ipsilateral side (Fig. 1B). Bothtargets were placed within a comfortable range for grasping. Thiswas determined for each subject according to the length of their armwhen the elbow was in full extension (180°) and the handcomfortably grasped the cylinder (about 90% arm’s length, seeMark et al. 1997). Movements were made with full vision. The taskwas to reach and grasp the cylinder with the whole hand in responseto an auditory signal at a self-selected speed and to hold the hand inthe final position (without lifting or displacing the object) until asound signalled the end of the trial. Blocks of ten trials wererecorded for each target location and counterbalanced acrosssubjects.

Data acquisition and analysis

Kinematic data from the hand, arm and trunk were recorded by anoptical motion analysis system (Optotrak 3010, Northern Digital,Waterloo) for 2–5 s at 120 Hz. Eight infrared emitting diodes(IREDs) were placed on bony landmarks of the hand, arm and trunk:(1) distal phalanx of the index—lateral to the nail, (2) distal phalanxof the thumb medial to the nail, (3) head of the first metacarpal bone,(4) radial styloid process, (5) lateral epicondyle, (6) homolateralacromion process, (7) contralateral acromion process, (8) middle ofthe sternum.In addition, the hand orientation was recorded by a Fastrack

Polhemus System, in order to obtain rotational data in three planes.This system uses electromagnetic fields generated by a transmitter ofa remote sensor with a 60 Hz recording frequency. The electro-magnetic sensor was placed on the dorsum of the hand with the mainaxis along the middle part of the third metacarpal bone.To determine the duration (MT) of the whole movement (reaching

and grasping), we used the tangential velocity of the wrist markercomputed from the magnitude of the velocity vector obtained bynumerical differentiation of the x, y, and z positions. The beginningand end of movement were defined as the times at which thetangential velocity rose above or fell and remained belowrespectively 5% of the peak tangential velocity of the marker.

Fig. 1A, B Experimental setup. Seated individuals reached towardsa target placed at arm’s length in the midline (T1) or in the ipsilateralworkspace (T2). C Use of Euler angles to measure hand orientation.xyz indicate the reference frame of the sensor fixed on the back of

the hand. The x axis is aligned with the middle of the thirdmetacarpal bone. Orientation in space is measured by three orderedrotations: azimuth, elevation and roll are rotations around the z, yand x axis respectively. Positive values are clockwise

165

Angular displacement of the wrist was determined by computingthe angle between the vectors joining the head of the first metacarpalbone and the radial styloid markers, and the radial styloid and lateralepicondyle markers (where 0° corresponds to a straight line—neutral position). For the elbow displacement, the angle was formedby the vector joining the radial styloid and lateral epicondylemarkers and the vector joining the lateral epicondyle and homolat-eral acromion markers. Shoulder flexion/extension was calculated asthe angle between the vectors joining the elbow and ipsilateralshoulder markers, and the sagittal plane through the vertical axis ofthe ipsilateral shoulder joint. Since the trunk moved forward duringthe reach, this implies that the axis moved during the task resultingin an underestimation of this angle. Shoulder horizontal adductionwas defined as the angle formed by the vector between the lateralepicondyle and the homolateral acromion markers and that betweenthe vector joining the two acromion markers projected on thehorizontal plane. Trunk rotation was determined by the anglebetween the vector joining the two shoulders (from the acromionmarkers) and the frontal axis in the horizontal plane (where 0°corresponds to a straight line—neutral position). Ranges of motionwere calculated for each trial as the difference between thebeginning and end of movement.Hand orientation was defined by Euler angles as follows: hand

azimuth is the orientation of the hand in the horizontal plane,elevation is an upward-downward angle in the vertical plane and rollis a rotation around the longitudinal axis of the hand. Positive valuesof hand azimuth are in the clockwise direction where 0 is defined asthe sagittal axis (Fig. 1C). For hand orientation, values of azimuth,roll and elevation were calculated at the end of the movement. Gripaperture was calculated as the 3D distance (x, y, z coordinates) inmillimetres between the markers on the index finger and thumb.The spatial coordination between the transport and grasp

components was analysed by computing the time to peak velocityof the wrist marker (TPV), the time to maximal hand aperture(TMA) and the hand closure distance. The latter was expressed asthe distance in millimetres from the beginning of grip closure (atpeak hand aperture) until the maximal closure coinciding withgrasping. Temporal coordination was expressed as the delay inmilliseconds between absolute values of TPV and TMA (TPV-TMAdelay). Each variable was expressed as a percentage of MT.

Statistical analysis

For our first goal, we determined the grasping strategies used byindividuals with hemiparesis by comparing the average values of thegrasping parameters (wrist angle; hand orientation: roll, elevation,azimuth; maximal grip aperture; and hand closure distance) formovements made to T1 and T2: (1) in the same group of subjectsand (2) between the two groups of subjects with 2-factor (target,group) ANOVAs. To determine the temporal coordination betweenhand transport and grasping, the same statistical comparisons weremade for the timing of critical components of the transport phase(TPV), the grasping phase (TMA) and the delay between them(TMA-TPV).For our second goal, to determine the relationships between

compensatory trunk movements and hand and arm motor impair-ments, Pearson Product Moment correlations were calculatedbetween clinical status indicators [Fugl-Meyer scores, dexterity(BBT), grip strength and wrist extension strength] and mean-by-subject values of kinematic parameters [hand orientation (elevation,azimuth, roll), wrist extension angle, trunk anterior displacementand trunk rotation].For the third goal, to determine the role of the trunk in

compensating grasping, we used simple regression analysis betweentrunk displacement and hand orientation (azimuth) on a trial-by-trialbasis. The correlation (r) and the slope of the regressions were usedto estimate the strength of the relationships. To address the relatedquestion of the role of the trunk for compensating both grasping andreaching, we used multiple regression analysis in which thedependent variable was trunk displacement and the independentvariables were the spatial kinematic parameters of transport andgrasp (joint angles and hand orientation). Regression analyses wereperformed separately on data from healthy subjects, participantswith hemiparesis without wrist control deficits (patients 1–7) andthose with wrist control deficits (patients 8–19).When homogeneity of variance requirements (Levene’s test) were

not met, non-parametric statistics were substituted (Kruskal-WallisANOVA). Initial p values of <0.05 were used for all tests.

Fig. 2 Stick figures for reachesto Targets 1 and 2 viewed fromabove (x–y plane) for a healthyindividual (left) and for twoparticipants with hemiparesis(middle, subject 3, and right,subject 11). Circles representpositions of index finger andthumb at maximal grip aperture(filled circles) and at the time ofgrasping (open circles). Trunk,arm and hand configurations areshown at the initial position(dotted lines), at maximal gripaperture (dashed lines) and atthe end of the reach (solid lines)

166

Results

Distal deficits ranged from mild to severe and, in somepatients, the impairment was different in the wrist andhand. For example, two patients (patients 11 and 14) hadnormal motor scores for the hand but lower scores for thewrist. Inversely, patient 7 had normal wrist control and amarked impairment of the hand. Even those patientsclassified as having normal motor function at the wrist andhand according to the clinical scale (patients 1–5) haddeficits in dexterity, grip strength and wrist extensionstrength. For example, wrist extension strength on theaffected side was only 31%±21 (range 0–73%) of thestrength of the contralateral side. Despite this distalweakness, all patients were able to reach and grasp thecylinder when it was placed in the midline (T1). However,a subgroup of patients (n=6) who could reach T1 were not

able to reach and grasp the object when it was placed inthe ipsilateral workspace (T2). This inability was notexplained by distal deficits since three of these six patientshad good distal recovery (patients 6, 7 and 8). Figure 2shows the body configurations required for successfulreaching of T1 and T2 in healthy subjects (left side ofpanels). Those patients who were unable to grasp T2 couldnot attain the body configuration of shoulder abductioncombined with elbow extension (compare right to middlepanels for T1 and T2). Specifically, they lacked about 50%elbow extension compared to those who could reach T2(Table 2). Consequently, comparisons of kinematic databetween targets were done only on the subgroup ofpatients (n=13) who could reach and grasp both targets.

Table 2 Kinematic and hand orientation data of reaching and grasping movements to two targets (T1, T2) placed at different locations.Values are given for the end of movement (TPV time to peak tangential velocity of the wrist, TMA time to maximal aperture of the hand)

T1 (n=7) T2 (n=7)

HealthyWrist extension (deg) 37 (9) 39 (6)Elbow extension (deg) 67 (08) 71 (06)Shoulder horizontal adduction (deg) 82 (11) 62 (11)Hand orientation (deg)Roll 75 (9) 78 (9)Elevation 7 (4) 8 (4)Azimuth −11 (3) 21 (5)***

Movement time (s) 1.24 (0.39) 1.14 (0.31)Wrist peak velocity (mm/s) 887 (154) 813 (185)***TPV (s) 0.35 (0.05) 0.30 (0.03)***Hand maximal aperture (mm) 84 (10) 81 (10)TMA (s) 0.58 (0.07) 0.47 (0.08)***Trunk displacement (mm) 37 (15) 16 (4)Trunk rotation (deg) 6 (4) 3 (2)

T1a (n=6) T1 (n=13) T2 (n=13)HemipareticWrist extension (deg) 23 (14) 29 (9) 29 (9)Elbow extension (deg) 24 (11)** 42 (20)* 45 (18)*Shoulder horizontal adduction (deg) 51 (9) 68 (21) 48 (15)***Hand orientation (deg)Roll 65 (26) 70 (23) 69 (22)Elevation −8 (21) −1 (8)* −6 (10)*Azimuth −40 (26) −28 (17)* −7 (20)*,***

Movement time (s) 2.09 (0.45) 1.87 (0.52)* 1.96 (0.66)*Wrist peak velocity (mm/s) 600 (140) 672 (241)* 561 (243)*,***TPV (s) 0.59 (0.16) 0.48 (0.14)* 0.41 (0.09)*Hand maximal aperture (mm) 78 (32) 75 (24) 77 (13)TMA (s) 1.21 (0.43) 1.01 (0.42)* 1.03 (0.58)*Trunk displacement (mm) 320 (71)** 146 (87)* 138 (116)*Trunk rotation (deg) 13 (3) 12 (3)* 8 (3)*

*Significant difference (p<0.05) between healthy and stroke groups, **significant difference (p<0.05) between stroke subgroups,***significant difference (p<0.05) between targets (T1, T2)

aData for the subgroup of patients with hemiparesis who were only able to reach and grasp T1 are shown separately

167

Description of grasping strategies used by individualswith hemiparesis

Just prior to grasping at the end of the reach, the wrist ofparticipants with hemiparesis was in a less extendedposition (by approximately 10°) compared to the healthysubjects for both targets (group main effect F(1,18)=4.58,p=0.046; Table 2). The parameters of roll, elevation andazimuth described the orientation of the hand. From aninitial palm down posture, hand roll increased as themovement progressed in the sagittal direction for bothgroups (Fig. 3A). For healthy subjects, as the hand reachedforward, it began to rotate upward and reached 75±9° atthe end of the movement (Table 2). For participants withhemiparesis, the values were similar to those of healthysubjects (70±23°; p>0.05). Similar values for roll wereobtained for reaches to T2 with no differences betweentargets or groups. However, individuals with hemiparesisoriented their hand more downward with respect to thehorizontal plane (less elevation) for both targets at the endof the reach just prior to grasping compared to healthysubjects (Table 2; F(1,18)=19.27, p=0.0004).

In healthy subjects, hand azimuth was negative (i.e. themain axis of the hand was slightly oriented toward the left)for T1 (Fig. 3B) and positive for T2 (not shown). Inparticipants with hemiparesis, hand azimuth was morenegative (i.e. the main hand axis was oriented morefrontally, Fig. 3D) for both targets compared to healthysubjects: the difference between groups was 17° for T1(Levene’s test, p<0.01; H=4.94, p=0.03) and 28° for T2(Levene’s test, p<0.05; H=8.85, p=0.003). For both

groups, the mean hand azimuth changed with the directionof the movement (toward T1 or T2, p=0.0001).

Temporal coordination between reaching and grasping

In all healthy subjects and most participants with hemi-paresis, the grip aperture typically increased progressivelyto a maximum as the hand moved forward in the sagittalplane and then decreased as the hand approached the target(Fig. 4A, C). The maximal grip aperture (MA) and thetime to maximal grip aperture (TMA), expressed aspercentages of the MT, were not different between groups(p>0.05; Fig. 5A, B) and MA occurred during thedeceleration phase of the wrist movement (Fig. 4B, D).Similarly, hand closure distance did not differ betweengroups (50±52 mm for healthy subjects and 49±35 mm forparticipants with hemiparesis; p>0.05). However, twopatients (8 and 17) kept the grip aperture almost constantduring the hand transport phase, and two patients (10 and19), who had a larger than average grip aperture, suddenlyopened the hand at the end of the transport phase.

The temporal delays between TPV and TMA expressedas a percentage of MT were not significantly differentbetween groups for movements to T1. These delays,however, were significantly longer (by 14%) for reaches toT2 by the patient group (Levene’s test, p=0.02; H=3.77,p=0.05; Fig. 5D).

Trunk movement and relationship to arm and handmotor impairment

In general, individuals with hemiparesis used greater trunkdisplacement to reach T1 (Levene’s test, p=0.004;H=10.81, p=0.001) and T2 (Levene’s test, p=0.02;H=12.77, p=0.0004) and greater trunk rotation for eachtarget (ANOVA, F(1,18)=17.87, p=0.0005) compared tohealthy subjects (Table 2, Fig. 2). Trunk displacement androtation correlated differently with motor impairmentsbased on Fugl-Meyer scores according to the direction ofthe target (Table 3). For T1, while there were nocorrelations between the severity of the clinical impair-ment and the amount of trunk rotation, the greater theimpairment (lower Fugl-Meyer scores), the greater thetrunk anterior displacement. For T2, individuals withgreater upper limb impairment used more trunk rotationbut not trunk anterior displacement.

Relationship between trunk movement and handorientation as a function of wrist impairment

In healthy subjects, there were no correlations betweentrunk displacement or rotation and hand orientation foreither T1 or T2 respectively. For example, the correlationbetween trunk movement and hand azimuth was r=0.11,slope=0.04 for T1 (Fig. 6A), and r=0.08, slope=0.24 forT2 (not shown). In individuals with hemiparesis, since

Fig. 3 Examples of mean hand roll (A, C) and azimuth (B, D)during a reaching movement toward Target 1 in one healthy subject(A, B) and one participant with hemiparesis (C, D; subject 14).Arrows show the direction of movement

168

trunk movement correlated differently with arm impair-ment severity for reaches to each target (Table 3), we onlyexamined the relationship between trunk displacement forreaches to T1 and trunk rotation for reaches to T2. Thiswas done separately for individuals with and without wristcontrol deficits. For T1, trunk displacement was relatedmore strongly to azimuth in individuals with wrist deficits(r=−0.52, Fig. 6B) than in those without (r=0.21, p<0.05)and this relationship was negative (with deficits, slope=−0.11; without deficits, slope=0.02). For T2, again only inpatients with wrist deficits, a significant relationship wasfound between trunk rotation and hand azimuth (r=−0.45,p<0.001, slope=−2.86) but not in those without (r=0.11, n.s., slope=0.08).

Once establishing that trunk movement varied withhand orientation, we were interested in finding outwhether the trunk movement was used both to compensatethe reaching and grasping deficits together. Multipleregression analysis was used to describe the relationshipbetween variables related to reaching (elbow extension,shoulder horizontal adduction, shoulder flexion) andgrasping (azimuth) with trunk movement. In healthyparticipants and participants without wrist control deficits,azimuth alone accounted for a small percentage of the totalvariance of the model (<10%) while most of the variancecould be explained by proximal arm movements (shoulderand elbow movements in healthy subjects and primarilyelbow movements in patients without wrist motor deficits;Fig. 7). In contrast, deficits in active range of both distal

Fig. 4 Examples of mean pre-hension movements made by arepresentative healthy subject(A, B) and a participant withhemiparesis (C, D, subject 12).A The spatial relationship be-tween transport (forward dis-tance moved by the hand) andgrasp (aperture or distance be-tween thumb and index mar-kers). B Grip aperture (thinlines) and wrist tangential velo-city profiles (thick lines). Notethe double labelling in B and D

Fig. 5 Group mean (SD)values of A maximal grip aper-ture (MA), B time to maximalgrip aperture (TMA), C time topeak velocity (TPV), and Ddelay between TPV and TMA.Values in B, C and D areexpressed as percentages of themovement time (MT) for reach-to-grasp movements made toTarget 1 (solid bars, T1) andTarget 2 (hatched bars, T2) inhealthy subjects and in partici-pants with hemiparesis due tostroke. The asterisk below thehorizontal bar indicates a sig-nificant difference at the p<0.05level

169

(r2=0.31) and proximal joints explained most of thevariance of the model in the participants with wrist motordeficits (Fig. 7).

Discussion

Our results show that during goal-directed reaching,individuals with hemiparesis orient the hand morefrontally for grasping and use more trunk anteriordisplacement and rotation to transport the hand to thetarget compared to healthy subjects. Although this hasbeen shown previously (Roby-Brami et al. 1997; 2003a,2003b; Michaelsen et al. 2001; Levin et al. 2002), a newfinding is that despite the changed pattern of joint andsegment recruitment, those patients who were able tograsp in our study preserved the major characteristics ofreaching and grasping such as grip aperture size, thetemporal coordination between hand transport and aper-

Table 3 Pearson product mo-ment correlations between clin-ical scores and kinematic vari-ables for reaches to Target 1(T1) and Target 2 (T2). Onlysignificant correlations are indi-cated. Grip strength and roll arenot shown since no significantrelationships were found withthese variables (FM Fugl-Meyerscore, BBT Box and BlocksTest)

*p<0.05, **p<0.005

FM score(arm)

FM score(wrist)

FM score(hand)

Wrist extensionstrength (%)

BBT (%)

T1 (n=19) - - - - -Elevation - - 0.46* - -Azimuth - 0.75** 0.63** - 0.57*Wrist extension - 0.46* 0.54* - -Trunk displacement −0.76** - −0.46* −0.50* −0.56*Trunk rotation - - - - -

T2 (n=13) - - - - -Elevation - - - - -Azimuth 0.68* 0.80** 0.65* - 0.67*Wrist extension - - - - -Trunk displacement - - - - -Trunk rotation −0.71** −0.74** −0.71* - −0.74*

Fig. 6 Linear regression between hand azimuth and trunk anteriordisplacement for reaches to Target 1 in healthy subjects (A) and inpatients with wrist motor deficits (B) (patients 8–19, Table 1; n=12).R value was significant only for subjects with hemiparesis. Notedifference in abscissae scaling in A and B

Fig. 7 Results of multiple regression analysis between trunkmovement (independent variable) and hand azimuth, elbow exten-sion, shoulder horizontal adduction and shoulder flexion (dependentvariables). Horizontal bars describe the contribution of each degreeof freedom to the total variance in the model for each group ofsubjects

170

ture formation and the relative timing of grip aperture (atleast for the midline target). Despite the fact that the handwas oriented more frontally for grasping (see also Roby-Brami et al. 1997, 2003a) the ability to modify the handorientation according to the reaching direction was alsopreserved. Furthermore, hand azimuth changed while wristextension remained unchanged for reaches to both targets,suggesting that other degrees of freedom may have alsocontributed to hand orientation.

Changes in reaching direction affect the orientation ofboth proximal and distal arm segments or joints(Desmurget and Prablanc 1997; Roby-Brami et al. 2000,2003a). Our finding that, in patients, trunk movement wassignificantly inversely correlated with the hand azimuthsuggests that movements of the trunk contributed to handorientation and that individuals with more severe hemi-paresis (less elbow and wrist extension) made more use ofthis compensatory strategy (Table 3; Figs. 2, 6). Indeed,multiple regression analysis revealed that in individualswithout distal motor deficits, increased trunk recruitmentwas correlated with proximal arm movements while inpatients with distal motor deficits, trunk recruitment wascorrelated with both proximal and distal arm movements(Fig. 7). Based on this analysis, we propose that thedamaged nervous system solves the motor deficit problemby recruitment of trunk DFs, in order to both transport thehand to the target and to achieve a functional handorientation for grasping when distal impairments arepresent (Figs. 6, 7).

Previous studies have suggested that the increased trunkrecruitment for reaching in patients with hemiparesis iscompensatory and not, in itself, a direct consequence ofthe lesion (see “Introduction”). It is possible that theincreased trunk recruitment was due to greater taskdifficulty when the reaching task also had a graspingcomponent as has been shown in healthy subjects (Mackeyet al. 2000). In contrast, other studies have shown that inhigh accuracy tasks, the trunk may be involved in thetransport phase of the reach and not in accuracy controlrelated to the hand (Saling et al. 1996; Seidler andStelmach 2000). Thus, the role of the trunk in tasks withhigh accuracy demands is controversial.

It is also possible that movement speed may affect trunkrecruitment. However, in healthy subjects, Seidler andStelmach (2000) showed that trunk displacement was notrelated to arm movement speed or temporal constraintsimposed on a reaching and grasping task. Although thespeed of reaching and grasping was not systematicallyvaried in our study, it seems unlikely that speed was afactor related to trunk recruitment.

Functional synergies

The relative preservation of the coordination between thedifferent components of reaching and grasping after astroke can be explained by the concept of functionalsynergy. Many studies have shown that there are aredundant number of degrees of freedom to produce any

given movement (Bernstein 1967; Feldman and Levin1995; Latash and Anson 1996). However, the optimizationof coordination can emerge naturally from task demands(Turvey et al. 1978). In healthy subjects during reaching,movements of the arm and trunk are co-ordinated together.For example, when pointing movements are made totargets placed beyond the arm’s reach involving trunkdisplacement, the influence of the trunk movement on thehand trajectory is actively neutralized in the early parts ofthe reach by compensatory rotations in the arm joints. In astudy by Rossi et al. (2002), trunk displacement only madea substantive contribution to hand transport towards theend of the reach. For a similar reaching task, thecontributions of trunk movement to hand transport weresignificantly greater throughout the reach in participantswith hemiparesis (Figs. 5, 6 in Levin et al. 2002). Theseresults show that in participants with hemiparesis, thetrunk is implicated in hand transport at an earlier phase ofthe reach when compared to healthy subjects and supportthe idea that the CNS chooses specific combinations ofDFs for performance of a behaviour in a task-specific way(the formation of coordinative structures; Kugler et al.1980). In this formulation, in the presence of motordeficits, additional movement components are added toform a new coordinative structure to achieve the functionalgoal. In participants with hemiparesis, the constraintsimposed on reaching by deficits in proximal joints(shoulder and elbow) and on grasping by deficits in distaljoints (wrist and hand) lead to the addition of the trunk intothe coordinative structure for both reaching and grasping.The correlational data between trunk movement and handorientation obtained in our study provides additionalevidence for the involvement of the trunk in a newcoordinative structure for reaching and grasping followingstroke. It should be pointed out, however, that thisconclusion is based on only correlational data of end-state positions of the trunk, arm and hand. Confirmation ofthe hypothesis should be done in which the dynamicrelationship between trunk and proximal and distal armdisplacements is examined during the course of themovement. Additional confirmation could also be ob-tained by comparing trunk use during reaches with andwithout a grasping component.

Cortical control of reaching and grasping

We investigated the kinematics of a reach-to-grasp move-ment without special regard to the consequences ofdifferences in lesion site. Many cortical and subcorticalregions may be involved in the control and performance ofgrasping. Pathways that control motoneurons of proximaltrunk and distal hand muscles are anatomically segregated(Lawrence and Kuypers 1968). Also, reaching andgrasping movements may be differentially controlled bythe corticospinal system (Lemon et al. 1995) as well as byseparate cortical areas such as the posterior parietal area(reaching neurons, Kalaska et al. 1997), area 6 (prehensionneurons, Rizzolatti et al. 1988) and the anterior intrapari-

171

etal area (object size, shape and orientation, Sakata et al.1999; Murata et al. 2000).

Frontal areas (primary motor cortex, supplementarymotor area and premotor area) contribute directly to thecontrol of hand movements (Colebatch et al. 1991). Morerecently the role of the posterior parietal area in graspinghas also been described (Matsumura et al. 1996; Chapmanet al. 2002; Binkofski et al. 1998; Luaute et al. 2002). Incase studies of patients with bilateral posterior parietallesions, isolated grasping deficits can be present (Jeanner-od et al. 1994) or not (Gréa et al. 2002) when grasping astationary object. Specific lesions involving the anteriorportion of the intraparietal sulcus in the posterior parietalcortex can affect grip aperture formation in the absence ofparesis (Binkofski et al. 1998). In our patient group, lesionsites were not homogeneous and clear relationshipsbetween the lesion sites and clinical impairments werenot seen. Patients with middle cerebral artery lesions thatnormally affect upper limb motor areas had good distalrecovery (for example patients 1, 2, 3 and 5). The degreeand distribution of distal impairments were variable.Results of clinical tests showed for example that patients11, 12 and 14 had high scores for hand control despite alow degree of wrist control. The heterogeneity of thedistribution of motor deficits in the arm, wrist and handmade it difficult to relate lesion types to functional deficits.

Even though participants with hemiparesis had a largerange of impairments (scores of 26–66 on the Fugl-Meyerupper limb scale), those with more severe impairmentcould not participate in this study since grasping was notpossible at all. The number of patients who recoverfunctional hand use is small (Lai et al. 2002). In this study,to be able to include a larger number of patients, we chosea task that was relatively easily done with whole handgrasping and did not require lifting the object. Tasksrequiring precision grasping and lifting may reveal othercompensatory strategies to preserve grip and load forces(Steenbergen et al. 1998).

Acknowledgements The authors wish to thank Jill Tarasuk,Philippe Archambault, Ruth Dannenbaum-Katz and Sheila Schnei-berg for their valuable contributions to this work. Financial supportfor SMM was provided by the Physiotherapy Foundation of Canada,Centre de Recherche Interdisciplinaire en Réadaptation de la Régionde Montréal (CRIR) and CAPES-Brazil. Research support was alsoprovided by the Heart and Stroke Foundation of Canada and bycollaborative grants to MFL and ARB from the Fonds de laRecherche en Santé du Québec (FRSQ) and INSERM-MRC.

References

Berglund K, Fugl-Meyer AR (1986) Upper extremity function inhemiplegia. A cross validation study of two assessmentmethods. Scand J Rehabil Med 18:155–157

Bernstein NA (1967) The co-ordination and regulation of move-ments. Pergamon Press, Oxford

Binkofski F, Dohle C, Posse S, Stephan KM, Hefter H, Seitz RJ,Freund H-J (1998) Human anterior intraparietal area subservesprehension. A combined lesion and functional MRI activationstudy. Neurology 50:1253–1259

Castiello U, Bennett K, Chambers H (1998) Reach to grasp: theresponse to a simultaneous perturbation of object position andsize. Exp Brain Res 120:31–40

Chapman H, Gavrilescu M, Wang H, Kean M, Egan G, Castiello(2002) Posterior parietal cortex control of reach-to-graspmovements in humans. Eur J Neurosci 15:2037–2042

Chieffi S, Gentilucci M (1993) Coordination between the transportand the grasp components during prehension movements. ExpBrain Res 94:471–477

Cirstea MC, Levin MF (2000) Compensatory strategies for reachingin stroke. Brain 123:940–953

Cirstea MC, Ptito A, Levin MF (2003) Arm reaching improvementswith short-term practice depend on the severity of the motordeficit in stroke. Exp Brain Res 152:476–488

Colebatch JG, Deiber MP, Passingham RE, Friston KJ, FrackowiakRS (1991) Regional cerebral blood flow during voluntary armand hand movements in human subjects. J Neurophysiol65:1392–1401

Desmurget M, Prablanc C (1997) Postural control of threedimensional prehension movements. J Neurophysiol 77:452–464

Desmurget M, Prablanc C, Arzi M, Rossetti Y, Paulignan Y,Urquizar C (1996) Integrated control of hand transport andorientation during prehension movements. Exp Brain Res110:265–278

Esparza DY, Archambault P, Winstein CL, Levin MF (2003)Hemispheric specialization in the coordination of arm and trunkmovements during pointing in patients with unilateral braindamage. Exp Brain Res 148:488–497

Feldman AG, Levin MF (1995) The origin and use of positionalframes of reference in motor control. Behav Brain Sci 18:723–744

Fugl-Meyer AR, Jääsko L, Leyman L, Olsson S, Steglind S (1975)The post-stroke hemiparetic patient. I. A method for evaluationof physical performance. Scand J Rehab Med 7:13–31

Gelfand IM, Tsetlin ML (1971) On mathematical modeling ofmechanisms of central nervous system. In: Gelfand IM,Gurfinkel VS, Fomin SV, Tsetlin ML (eds) Models of thestructural-functional organization of certain biological systems.MIT Press, Boston

Gentilucci M, Castiello U, Corradini ML, Scarpa M, Umilta C,Rizzolatti G (1991) Influence of different types of grasping onthe transport component of prehension movements. Neuropsy-chologia 29:361–378

Gentilucci M, Daprati E, Gangitano M, Saetti MC, Toni I (1996) Onorientating the hand to reach and grasp an object. Neuroreport7:589–592

Gentilucci M, Bertolani L, Benuzi F, Negrotti A, Pacesi G,Gangitano M (2000) Impaired control of an action aftersupplementary motor area lesion: a case study. Neuropsychol-ogia 38:1398–1404

Gréa H, Desmurget M, Prablanc C (2000) Postural invariance inthree-dimensional reaching and grasping movements. ExpBrain Res 134:155–162

Gréa H, Pisella L, Rossetti Y, Desmurget M, Tilikete C, Grafton S,Prablanc C, Vighetto A (2002) A lesion of the posterior parietalcortex disrupts on-line adjustments during aiming movements.Neuropsychologia 40:2471–2480

Hoff B, Arbib MA (1993) Models of trajectory formation andtemporal interaction of reach and grasp. J Mot Behav 25:175–192

Jakobson LS, Goodale MA (1991) Factors affecting higher-ordermovement planning: a kinematic analysis of human prehension.Exp Brain Res 86:199–208

Jeannerod M (1984) The timing of natural prehension movements. JMot Behav 16:235–254

Jeannerod M, Decety J, Michel F (1994) Impairment of graspingmovements following a bilateral posterior parietal lesion.Neuropsychologia 32:369–380

Kalaska JF, Scott SH, Cisek P, Sergio LE (1997) Cortical control ofreaching movements. Curr Opin Neurobiol 7:849–59

172

Kugler PN, Kelso JAS, Turvey MT (1980) On the concept ofcoordinative structures as dissipative structures: 1. Theoreticallines of convergence. In: Stelmach GE, Requin J (eds) Tutorialsin motor behavior. North Holland, Amsterdam

Lai S-M, Studenski S, Duncan PW, Perera S (2002) Persistingconsequences of stroke measured by the stroke impact scale.Stroke 33:1840–1844

Latash ML, Anson JG (1996) What are “normal movements” inatypical populations? Behav Brain Sci 19:55–68

Lawrence DG, Kuypers HG (1968) The functional organization ofthe motor system in the monkey. II. The effects of lesions of thedescending brain-stem pathways. Brain 91:15–36

Lemon RN, Johansson RS, Westling G (1995) Corticospinal controlduring reach, grasp, and precision lift in man. J Neurosci15:6145–6156

Levin MF, Michaelsen S, Cirstea C, Roby-Brami A (2002) Use ofthe trunk for reaching targets placed within and beyond thereach in adult hemiparesis. Exp Brain Res 143:171–180

Luaute J, Rode G, Rossetti Y, Morel C, Ferraton B, Boisson D(2002) Mode de récupération et fonctionnalité de la préhensionchez l’hémiplégique vasculaire. In: Pelissier J, Benaim C,Enjalbert M (eds) La préhension et l’hémiplégie vasculaire.Masson, Paris, pp 31–39

Mackey DC, Bertram CP, Mason AH, Marteniuk RG, MackenzieCL (2000) The effect of task complexity on trunk-assistedreaching. J Sport Exerc Psychol 22:S73

Mark LS, Nemeth K, Gardner D, Dainoff MJ, Paasche J, Duffi M,Grandt K (1997) Postural dynamics and preferred criticalboundary for visually guided reaching. J Exp Psychol HumPercept Perform 23:1365–1379

Mathiowetz V, Volland G, Kashman N, Weber K (1985) Adultnorms for the box and block test of manual dexterity. Am J OccTher 39:386–391

Matsumura M, Kawashima R, Naito E, Satoh K, Takahashi T,Yanagisawa T, Fukuda H (1996) Changes in rCBF duringgrasping in humans examined by PET. Neuroreport 29:749–752

Michaelsen SM, Luta A, Roby-Brami A, Levin MF (2001) Effect oftrunk restraint on the recovery of reaching movements inhemiparetic patients. Stroke 32:1875–1883

Michaelsen SM, Roby-Brami A, McKinley P, Levin MF (2002)Compensation for distal impairments in prehension in stroke.Society for Neuroscience Abstracts, Orlando, November

Murata A, Gallese V, Luppino G, Kaseda M, Sakata H (2000)Selectivity for the shape, size, and orientation of objects forgrasping in neurons of monkey parietal area AIP. J Neurophy-siol 83:2580–2601

Paulignan Y, Jeannerod M (1996) The visuomotor channelshypothesis revisited. In: Wing AM, Haggard P, Flanagan JR(eds) Hand and brain, the neurophysiology and psychology ofhand movements. Academic Press, San Diego, pp 265–282

Paulignan Y, Mackenzie C, Marteniuk R, Jeannerod M (1991)Selective perturbation of visual input during prehensionmovements. 1. The effects of changing object position. ExpBrain Res 83:502–512

Paulignan Y, Frak VG, Toni I, Jeannerod M (1997) Influence ofobject position and size on human prehension movements. ExpBrain Res 114:226–234

Pohl PS, Winstein CJ, Onla-or S (1997) Sensory-motor control inthe ipsilesional upper extremity after stroke. Neurorehabilita-tion 9:57–69

Rizzolatti G, Camarda R, Fogassi L, Gentilucci M, Luppino G,Matelli M (1988) Functional organization of inferior area 6 inthe macaque monkey. II. Area F5 and the control of distalmovements. Exp Brain Res 71:491–507

Roby-Brami A, Fuchs S, Mokhtari M, Bussel B (1997) Reachingand grasping strategies in hemiparetic patients. Motor Control1:72–91

Roby-Brami A, Bennis N, Mokhtari M, Baraduc P (2000) Handorientation for grasping depends on the direction of thereaching movement. Brain Res 869:121–129

Roby-Brami A, Bennis N, Jacobs S, Levin MF (2003a) Handorientation for grasping and upper-limb joint rotation patternsin healthy subjects and hemiparetic stroke patients. Brain Res969:217–229

Roby-Brami A, Feydy A, Combeaud M, Biryukova EV, Bussel B,Levin MF (2003b) Motor compensation and recovery forreaching in stroke patients. Acta Neurol Scand 107:369–381

Rosenbaum DA, Van Heugten CM, Caldwell GE (1996) Fromcognition to biomechanics and back: the end-state comforteffect and the middle-is-faster effect. Acta Psychol 94:59–85

Rosenbaum DA, Meulenbroek RJ, Vaughan J, Jansen C (2001)Posture-based motion planning: applications to grasping.Psychol Rev 108:709–734

Rossi E, Mitnitski A, Feldman AG (2002) Sequential control signalsdetermine arm and trunk contributions to hand transport duringreaching in humans. J Physiol 538:659–671

Sakata H, Taira M, Kusunoki M, Murata A, Tsutsui K, Tanaka Y,Shein WN, Miyashita Y (1999) Neural representation of three-dimensional features of manipulation objects with stereopsis.Exp Brain Res 128:160–169

Saling M, Stelmach GE, Mescheriakov S, Berger M (1996)Prehension with trunk assisted reaching. Behav Brain Res80:153–160

Seidler RD, Stelmach GE (2000) Trunk-assisted prehension:specification of body segments with imposed temporalconstraints. J Mot Behav 32:379–389

Steenbergen B, Hulstijn W, Lemmens IH, Meulenbroek RG (1998)The timing of prehensile movements in subjects with cerebralpalsy. Dev Med Child Neurol 40:108–114

Steenbergen B, van Thiel E, Hulstijn W, Meulenbroek RG (2000)The coordination of reaching and grasping in spastic hemipar-esis. Hum Mov Sci 19:75–105

Trombly CA (1992) Deficits of reaching in individuals with lefthemiparesis: a pilot study. Am J Occup Ther 46:887–897

Turvey MT, Shaw RE, Mace W (1978) Issues in the theory ofaction: degrees of freedom, coordinative structures andcoalitions. In: Requin J (ed) Attention and performance VII.Erlbaum, Hillsdale, NJ, pp 557–595

van Vliet P, Sheridan M, Kerwin DG (1995) The influence offunctional goals on the kinematics of reaching following stroke.Neuroreport 19:11–19

Wade DT, Hewer RL, Wood VA, Skilbeck CE, Ismail HM (1983)The hemiplegic arm after stroke: measurement and recovery. JNeurol Neurosurg Psychiatry 46:521–524

Wang J, Stelmach GE (1998) Coordination among the bodysegments during reach-to-grasp action involving the trunk.Exp Brain Res 123:346–350

Wang J, Stelmach GE (2001) Spatial and temporal control of trunk-assisted prehensile actions. Exp Brain Res 136:231–240

173