Neuropsychologia, Vol. 29, No. 11. pp. 1129-1135, 1991 Printed in Great Britain.

002s3932/91 %3.00+0.00 0 1991 Pergamon Press plc

USE OF LEFT VS RIGHT HAND IN RESPONDING TO LATERALIZED STIMULI IN UNILATERAL NEGLECT*

IAN ROBERTSON

MRC Applied Psychology Unit, 15 Chaucer Road, Cambridge CB2 2EF, U.K.

(Received 9 April 1991; accepted 3 July 1991)

Abstract-Previous research suggests that contralesional limb activation may reduce the degree of visual neglect shown by hemi-inattentive subjects. The present study examines whether minimal assisted-left-limb activation (pressing a response key with the hemiplegic hand, aided by the ipsilesional hand) resulted in lowered response latencies on the neglected side in a group of six patients showing unilateral left visual neglect. The study was carried out using a simple computerized test of speed of detection of lateralized stimuli, with responses being made on a key located at the body midline.

In five of the subjects, there was no evidence of relatively faster response times to contralesional stimuli when the contralesional limb was involved in the responses. In one of these five subjects, there was no hemiplegia, and hence full use of the contralesional limb.

Only in one subject did an interaction appear between the side of presentation of the stimuli and limb used. In this case, a small but significant tendency appeared for the subject to make relatively faster responses to left-sided stimuli when using the left hand to make responses.

INTRODUCTION

JOANETTE and BROUCHON [7] described a 64year-old woman who suffered a right brain CVA and who tended to point to stimuli on her left as if she had seen them on her right. Interestingly, an interaction appeared between the side of space upon which the stimulus appeared and the arm which was used. Only when the right arm was used in response to a left-sided stimulus did the allesthesia appear. When the other arm was used in response to the same stimulus on the same side, there was no allesthetic response.

A subsequent series of cases [S] found that it was not only allesthetic problems which revealed such an interaction. In a standard stimulus identification procedure, neglect was less severe when the contralesional limb was used to point to the target stimuli than when the ipsilesional limb was used.

In a single case study, HALLIGAN and MARSHALL [S] also found that use of the left arm for a cancellation and line bisection task respectively resulted in less neglect.

These findings are in line with the theoretical propositions of RIZZOLATTI and CAMARDA [9], who propose that spatial attention is based upon a series of circuits largely independent from one another which programme motor plans in a spatial framework. Spatial attention is not seen as a supraordinate function controlling whole-brain activity, but as a property intrinsically linked to the premotor activity and distributed among a range of centres.

Subsequently, however, HALLICAN et al. [6] showed in a series of experiments that the

*This work was cdrried out during the author’s tenure as a Medical Research Council (U.K.) Travelling Fellowship to the University of Rome. The work was carried out with the support and facilities of the Research Center, Clinica S Lucia, via Ardeatina, Rome, Italy.

1129

1130 I. ROBERTSON

advantage of contralesional arm use in reducing neglect was better explained by a spatiomotor cueing process than by a hemispheric activation hypothesis. More specifically, they found that the advantage of left arm use in line bisection was eliminated by having the subjects begin the task on the right side of the line, i.e. with the arm crossed beyond the body midline.

The present study attempted to further evaluate the activation hypothesis by engaging neglect patients in a task requiring a central response-namely pressing a response key at the body midline, in response to lateralized stimuli. If limb activation is the crucial variable, then contralesional limb use, even at a central location, should result in reduced neglect. However, if the main effect of contralesional activation is, as HALLIGAN et al. [6] suggest, one of spatiomotor cueing, then central responses to lateralized stimuli should result in no advantage for one limb over the other.

The present study used a consecutive sample of patients showing left unilateral neglect, all of whom were hemiplegic except one. The reason for choosing such a sample pertains to the possible rehabilitation uses of the limb-activation hypothesis: if contralesional limb use does indeed reduce perceptual neglect, then this might be used therapeutically. However, given that few neglect patients have full contralesional limb use, then the therapeutic applicability would be small if the effect were confined to a nonhemiplegic sample.

Subjects

METHOD

Six subjects were studied, all having suffered right CVA’s. All subjects were given the letter cancellation (‘h’) (version by DULLER et al., [3]) as well as Albert’s line cancellation test [l] and all showed visual neglect, defined by performance on these tests. The subjects had the following characteristics.

Subject 1. Sixty-six-year-old man, right-handed, 13 weeks post infarct, left homonymous hemianopia, left hemiplegia, 9 years of formal education, with following CT scan report: ‘area of profound hypodensity in right Sylvian area, margins badly defined, with oedema and compression of the ventricle.’ This patient made 14 errors (70% error) on Albert’s test-all on the left-and cancelled only the rightmost 5 out of 104 stimuli (95% error) on a simple letter cancellation task.

Subject 2. Seventy-five-year-old right-handed male, 27 weeks post infarct, left hemiplegia, upper left quandrantanopia, 5 years of formal education. No CT scan available. No errors on Albert’s test, 30 errors on letter cancellation, 28 on the left (29% error).

Subject 3. Seventy-one-year-old right-handed woman, 30 weeks post infarct, 3 years offormal education. No CT scan carried out. Visual fields full. This patient made 11 errors, all left, on Albert’s test (55% error), and omitted 7 letters, again all left, from the letter cancellation test (7% error). At the time of testing, this patient had full use of her left arm, with only a residual weakness. She also had full left leg function.

Subject 4. Seventy-eight-year-old left-handed female, 22 weeks post infarct, left hemiplegia, left homonymous hemianopia, 3 years of formal education. No CT scan available. Two errors on Albert’s test (10% error), both left, and 26 on letter cancellation, all on the left (25% error).

Subject 5. Seventy-seven-year-old right-handed man, 33 weeks post infarct, 5 years of formal education, no CT scan available, left hemiplegia, left homonymous hemianopia. No errors on Albert’s test, 16 on letter cancellation (15% error), all on the left.

Subject 6. Sixty-five-year-old right-handed man, 13 weeks post infarct, 5 years of formal education. CT scan showed a right frontoparietal infarct. He showed a left hemiplegia, left homonymous hemianopia, had two errors on Albert’s test (20% error), 18 on letter cancellation(l7% error), all on the left.

Subject details are summarized in Table 1.

Procedure

Subjects were seated in front of an Apple Macintosh 2E microcomputer, their eyes at a distance of 35 cm from the screen. They were asked to fixate a point marked at the centre of the screen, which was adjusted to the subject’s eye level. A chin rest was used to aid fixation. The experiment was prepared using the PsychLab software.*

Fixation was checked by the experimenter with the aid of a mirror. In the basic task, random numbers between 0

*PsychLab software developed by Daniel Bub and Teren Gum of the Montreal Neurological Institute, Canada.

LATERALIZED STIMULI IN UNILATERAL NEGLECT 1131

Table 1. Subject characteristics

Time post Albert’s infarct test: Lett. cant. Visual

Subject Age Sex (weeks) Hemiplegia % error % error fields

1 66 M 13 Left 70 95 LHH*

2 75 M 27 Left 0 29 ULQt 3 71 F 30

4 78 F 22 5 77 M 33 6 65 M 13

*Left homonymous hemianopia. tUpper left quadrantanopia.

Residual left weakness

Left Left Lett

55 7 Full

10 25 LHH 0 15 LHH

20 17 LHH

and 9 were presented randomly to the left and right 2.3 degrees eccentric to the fixation point. This was within the intact visual fields of all hemianopic subjects.

Numbers subtended a vertical angle of 0.80 degrees and a horizontal angle of 0.46 degrees. Numbers were presented for 3 set or until the subject pressed the space bar on the keyboard. There was an interval 3 set between presentations.

There were 12 blocks each of 10 left and 10 right presentations, half of the blocks for the left hand, and other half of the blocks for the right.

Within each block of 20 presentations, 5 ‘catch’ trials were randomly presented. These consisted of central presentation of an asterisk. 0.46 degrees horizontal visual angle by 0.46 degrees vertical. Patients were told not to respond to these, and any responses were recorded. Stimuli within blocks were randomized, as were the blocks themselves.

The subject simply had to press the space bar on the keyboard as soon as number was detected to the left or right. The computer recorded the response latency.

Prior to the experiment, a practise round of 10 left, 10 right, and 5 catch trials was given for each hand.

Nature of responses made with contralesional limb

Only one subject had full use of her left arm and hand-Subject 3 above. She made her response by pressing the space bar of the computer at the body midline, and had no difficulty doing this with either hand.

All other subjects had minimal shoulder movement. In these latter five cases, responses were made by the index finger of the left hand, while this hand was supported at the wrist by the right hand. Subjects were encouraged to respond ‘as much as possible with the left arm and hand, using the right arm as necessary’.

Given the nature of the response, and given the absence of EMG recordings of muscle activity, it was impossible to determine to what extent the left arm was involved in the responses made, though in order to sustain the response position, some activation of at least the contralesional shoulder was necessary in all cases. Hence the response required of the contralesional limb was not simply that of moving the index finger, rather there was a requirement to sustain the contralesional limb in the position required to make the response with the finger, though ofcourse in all cases, except Case 3, this was done with the help of the ipsilesional arm and hand.

RESULTS

Separate two-way (side of stimulus vs hand used) ANOVA’s were carried out for each subject in turn. Responses of over 5000 msec were omitted from the analysis.

Table 3 shows the means and standard deviations of the responses in each condition for each of the subjects. Table 2 shows the omissions on each side for each subject.

Subject 1 showed 16 omissions on the left in total when using his right hand, and the same number when using his left hand. He showed no right-sided omissions. Looking at latency of response to the remaining stimuli two-way analysis of variance showed a main effect for side of stimulus (F= 108.1; P<O.OOOl), no effect for hand used (F=0.39; ns), and an interaction between side of stimulus and hand used (F=4.84; P~0.03). Figure 1 shows these results graphically.

1132 I. ROBERTSON

Table 2. Means and standard deviations of latencies of response to lateralized stimuli, left vs right hand (all values in msec)

Subject Right hand: Right hand: Left hand: Left hand: left stimulus right stimulus left stimulus right stimulus

1 2167 (1473)

2 1223

(886) 3* 1638

(893) 4 2317

(951) 5 io54’

(520) 6 1393

(1125)

570

(90) 999

(1050) 1146

(352) 936

(415) 705

(193) 673

(460)

1812 (1280) 1530

(757) 1680

(634) 2916

(799) 2066

(928) 995

(529)

770

(214) 1367

(963) 1072

(339) 1764

(657) 1588

(86;;)

(245)

*This subject had full use of her left hand and could respond without help from the right hand.

Table 3. Number of omissions (i.e. responses not given within 5000 msec) for each experimental condition

Subject Right hand: left stimulus

Right hand: Left hand: Left hand: right stimulus left stimulus right stimulus

1 16 0 16 0 2 2 0 4 2 3* 0 0 0 0 4 13 0 11 2 5 0 0 0 0 6 2 0 2 0

*This subject had full use of her left hand and could use it to respond without help from the right hand.

3000

0 Left hand Right hand

Fig. 1. Latency of response, left vs right hand, Subject I

LATERALIZED STIMULI IN UNILATERAL NEGLECT 1133

Subject 2 omitted two stimuli on the left when responding with his right hand, four when responding with the left hand. He omitted two on the right when responding with the left hand, none on the right when responding with the right hand.

Looking at latency of response to the remaining stimuli, two-way analysis of variance showed no main effect for side of stimulus (F=2.28; P=O.137), a significant effect for hand used (F= 5.5; P=O.O22), and no interaction between side of stimulus and hand used (F= 0.43; ns)

Subject 3 omitted no stimuli when responding with either hand. Looking at latency of response to the remaining stimuli, two-way analysis of variance showed a main effect for side of stimulus (F= 2 1.38; P < O.OOOl), no significant effect for hand used (F= 0.012; ns), and no interaction between side of stimulus and hand used (F=0.28; ns).

Subject 4 omitted 13 stimuli on the left when responding with his right hand, 11 when responding with left hand. He omitted two on the right when responding with the left hand, none on the right when responding with the right hand.

Looking at latency of response to the remaining stimuli, two-way analysis of variance showed a main effect for side of stimulus (F= 108.2; P<O.OOOl), a significant effect for hand used (I;= 60.3; P<O.OOOl), and no interaction between side of stimulus and hand used (F= 1.08; P=O.30).

Subject 5 omitted no stimuli when responding with either hand. Looking at latency of response to the remaining stimuli, two-way analysis of variance showed a main effect for side of stimulus (F= 10.02; P=O.O025), a significant effect for hand used (F= 57.73; P<O.OOOl), and no interaction between side of stimulus and hand used (F=0.27; ns).

Subject 6 omitted two stimuli on the left when responding with his his right hand, two when responding with the left hand. He omitted none on the right when responding with either hand.

Looking at latency of response to the remaining stimuli, two-way analysis of variance showed a main effect for side of stimulus (F= 19.53; P-C O.OOOl), no significant effect for hand used (F=2.76; P=O.O12), and no interaction between side of stimulus and hand used (F= 1.98; P=O.165).

DISCUSSION AND CONCLUSIONS

Of six subjects, only one (Subject 1) showed a small but significant tendency to show faster latencies to contralesional stimuli when using the contralesional limb. None of the other subjects showed such an effect, including Subject 3, who had full use of her left hand and arm.

Subjects 2, 4 and 5 all showed slower responding to left stimuli with their left hands compared with their right hands, though responding to right stimuli was also slowed and there were no statistically significant interactions between side of responding and side of stimulus. This suggests that use of the hemiplegic side in responding may cause a general slowing of responding, as opposed to a selective facilitation of left responding, among many patients. One possible reason for this is the effort required to use the hemiplegic arm reduces the attention paid to the visual stimulus and hence slows all responding.

These results tend to support HALLIGAN et d’s. [6] view that limb activation per se does not result in reduced neglect in most cases. The results of Subject 1 present the only possible caveat to this conclusion. While it is possible that this subject was still cued to attend more to the neglected side, by virtue of holding his left hand at the central location on the computer keyboard, the degree of spatial cueing was considerably less than that involved in making a

1134 I. ROBERTSON

normal left-handed response in left hemispace when there is no hemiplegia. Hence it may be the case that for some cases, the contralesional limb use effect cannot be reduced to a phenomenon of spatiomotor cueing.

On the other hand, scrutiny of the reaction times for Case 1 in Table 2 shows that left- handed responses to right-sided stimuli are on average 200 msec slower than right-handed responses to right-sided stimuli. In contrast, left-handed responses to left-sided stimuli are almost 150 msecfuster than right-handed responses to left stimuli. This result argues for a role of stimulus-response compatibility in determining response times. By this account, HALLICAN et d’s position appears to be supported by the data, i.e. the presence of a left stimulus and a left-handed response together cause significant spatiomotor cueing to the contralesional side.

As Halligan et al. point out, however, the activation hypothesis may relate to the hemispace within which responses are made, rather than to the limb used. Neither their experiments, nor the present one, rule out this possibility. Research by COSLETT et al. [2] suggests however, that only some neglect patients show such hemispace effects, with others showing responses dominated by the side of space in which stimuli appear, rather than by the side of space in which responses have to be made. Perhaps if the patients in this study had been divided according to such a distinction, then clearer results may have emerged.

Assuming for a moment that such a hemispace-based activation process exists, it may be that it relates only to specific types of activation of the contralesional limb. Such movements may be restricted to ones where the perceptual and motor elements of the stimulus-response cycle are more directly linked than they are in the present study, for example in a pointing task. If that were the case, then the present equivocal results may relate to the fact that the response in question was only conceptually related to the stimuli, and not spatially related.

Returning to Subject 1, and for a moment rejecting the spatiomotor cueing hypothesis to explain his results, the apparent minimal effectiveness of this nonspecific contralesional limb activation might point to a more general alerting effect produced by contralesional limb activation. If, as some have suggested, low arousal exacerbates neglect [4], then it is conceivable that the act of having to mobilize the hemiplegic limb may, in some cases, sufficiently increase arousal so as to reduce the degree of neglect shown. This hypothesis is however not at present strongly supported by either the studies of HALLIGAN et al. [6], or by the present data.

Recent research carried out on a case of left visual neglect by the present author casts some light on both the hemispace-based activation phenomenon as well as on the general activation hypothesis. Comparing the effect of variously positioned hand movements on letter-cancellation performance, the results suggest a privileged role for left-limb activation in left hemispace in reducing neglect, a phenomenon which is not evoked by right-arm activation in left hemispace, nor by left-limb activation in right hemispace (unpublished observation).

This suggests that Halligan et al. may be correct in attributing such phenomena to a spatiomotor cueing process, but also suggests that the effect is not only limited to some subjects, but is also limited to where the response is made by the contralesional limb. It may be the case therefore that stimulus-response compatibility is required for the spatiomotor cueing effect to manifest itself.

The possible therapeutic applications of this phenomenon appear to have some promise, according to data recently collected by the present author (unpublished observations), where unilateral limb activation in three subjects was associated with relatively enduring changes on performance on visual scanning tasks. All subjects had some contralesional arm or hand

LATERALIZED STIMULI IN UNILAlERAL NEGLECT 1135

use, however, and it may be the case that the therapeutic effects of this phenomenon are confined to a subset of these neglect subjects who have some contralesional limb use.

Acknowledgements-The author would like to thank Profs Luigi Pizzamigho and Pierre-Luigi Zoccolotti, at the Research Center, Clinica S Lucia, Rome, Italy for their generous help and support in carrying out this study. Many thanks also to the Clinic Director, Dr Amadio, as well as all the staff in the Research Center.

1. 2.

3.

4. 5.

6.

I.

8.

9.

10.

REFERENCES ALBERT, M. L. A simple test of visual neglect. Neurology 23, 6588665, 1973. COSLETT, H. B., BOWERS, D., FITZPATRICK, E., HAWS, B. and HEILMAN, K. M. Directional hypokinesia and hemispatial inattention in neglect. Brain 113, 475486, 1990. DILLER, L., BEN-YISHAY, Y., GERSTMAN, L., GOODKIN, R., GORDON, W. and WEINBERG, J. Studies in Cognition and Rehabilitation in Hemiplegia. Rehabilitation Monograph No. 50. New York University Medical Center, New York, 1974. FLEET, W. S. and HEILMAN, K. M. The fatigue effect in unilateral neglect. Neurology 36, 258, 1986. HALLIGAN, P. W. and MARSHALL, J. C. Laterality of motor response in visuo-spatial neglect: a case study. Neuropsychologia 21, 1301-1307, 1989. HALLIGAN, P. W., MANNING, L. and MARSHALL, J. C. Hemispheric activation vs spatio-motor cueing in visual neglect: a case study. Neuropsychologia 29, 165-116, 1991. JOANETTE, Y. and BROUCHON, M. Visual allesthesia in manual pointing: some evidence for a sensory-motor cerebral organization. Brain Cognition 3, 152-165, 1984. JOANET~E, Y., BROUCHON, M., GAUTHIER, L. and SAMSON, M. Pointing with left versus right hand in left visual field neglect. Neuropsychologia 24, 391-396, 1986. RIZZOLATTI, G. and CAMARDA, R. Neural circuits for spatial attention and unilateral neglect. In Neurophysiological and Neuropsychological Aspects of Neglect, M. JEANNEROD (Editor), North Holland, Amsterdam. ROBERTSON, I. The rehabilitation ofattentional and hemi-inattentional disorders. In Cognitive Neuropsychology and Cognitive Rehabilitntion. G. HUMPHREYS and J. RIDDOCH (Editors). Lawrence Erlbaum Associates. London, (in press).