Perceptual and premotor factors of unilateral neglect E. Bisiach, MD; G. Geminiani, MD; A. Berti, MD; and M.L. Rusconi, MD
Article abstract-Right-brain-damaged patients showing unilateral neglect underwent a specifically devised line-bisection task that allowed uncouplingof the direction of visual attentionfrom that of hand movement. This made it possible to isolate and separatelyassess perceptual and premotor factors of the disorder.Comparison of experimentaland radiologic data suggestedthat premotor factorswere more pronounced in patientswith lesions involving the frontal lobes than in patientswith lesions confined to postrolandic areas. The technique employed is compatible with bedside examination and provides data useful for standard assessment of neglect symptomatology for both clinical and experimental purposes. NEUROLOGY 1990;40:1278-1281
First envisaged by Watson et al,l “unilateral directional hypokinesia” (DH) implies a powerful, although elusive, theoretic construct relative to the neural implementation of space representation. With this term, we refer to the reluctance of some brain-damaged patients to initiate and carry out motor activities toward the contralesional side of egocentric space, irrespective of the side of the limbs involved in such activities. As one of the more striking manifestations of unilateral neglect, the phenomenon is more frequent and severe in right- compared with left-brain-damaged patients. DH must be distinguished from motor neglect, which is a reluctanceto move the contralesional limbs irrespective of the direction of movement and is not related to functional impairment due to disorders other than hemineglect. The concept of DH implies the existence of a mechanism subserving spatial aspects of motor ideation and preparation, which is relatively independent of the sensory-based spatial representation of the organism’s environment. The theoretic significance of such a construct is self-evident. It prefigures a (longitudinally) componential structure of space representation, orthogonal to the (transversely) componential structure resulting from different input modalities and from the different frames of reference (eg, retinotopic versus body-centered) in which unilateral neglect may manifest i t ~ e l f .The ~ - ~construct underlying DH is therefore heuristicallypowerful in that it motivates research that might significantlyimprove our knowledge of the brain mechanisms involved in space representation. In particular, as was first suggested by Watson et all and later by Mesulam: it might help to differentiate varieties of neglect specific to anterior and posterior brain damage. On closer examination, however, the construct at ~~
~~~
~~
issue proves less manageable than at first glance, insofar as it is difficult to conceivejust how, if at all, it might be relatively independent of earlier stages of space representation, purportedly uninvolved in the initiation of any particular motor activity. Furthermore, as noted by Coslett et a l 6 , empirical results allegedly in support of DH as an independent element of the neglect syndrome’ cannot rule out the possibility that such phenomena originate from dysfunction of the above-envisaged(sensory-based and behaviorally noncommitted) earlier stages of space representation. This ambiguity is still present in the investigation carried out by Coslett et a1.6 These authors asked 4 left hemineglect patients to bisect lines placed to the left or right of their body’s sagittal midplane, while controlling their performance on a T V screen which was also placed to either the left or right of that plane. By so doing, they disconnected the side of space on which the performing hand was lying and the side on which visual attention was directed; they did not uncouple, however, the direction of hand movements and the direction of the image of the movements projected on the TV screen. Thus, whatever may be the best explanation of the results, these might bear upon the issue of the multisensory-in this instance visual and somatosensory-representation of space rather than upon the distinction of input and output components in unilateral neglect. This hypothetical dissociation must be sought by uncoupling the direction of a sensory (eg, visual) stimulus and that of the movement performed by the subject in order to drive that stimulus onto a particular target. We did so by asking left hemineglect patients to perform a bisection task in 2 different conditions: a congruent (canonical) condition, in which a pointer had to be directly moved along the line, and a noncongruent
~
From the Istituto di Clinica Neurologica, UniversitA di Milano,and Iatituto Neurologic0 C. Beata, Milano,Italy. Supported by a grant from the Ministera d e b Pubblica Istruzione to Dr. E. Bisiach. Received November 17,1989. Accepted for publication in final form January 29,1990. Address correspondenceand reprint requests to Dr. Moardo Bisiach, Istituto di Clinica Neurologica d e b Universits. di Milano,Via Francesco Sfom 35,20122 Milano,Italy.
1278 NEUROLOGY 40 August 1990
Table 2. Mean displacements (mm) of the subjective midpoint in the congruent and in the noncongruent conditions and differences between the 2 conditions in the 15 patients and in the control group*
Table 1. Sex, age, and clinical data relative to brain-damaged subjects Pt. no. Sex Age
1
2 3 4 5 6 7 8 9 10 11 12 13 14 15
M M M F F F
M F F F
M F F F
M
Etiology
Duration* (days)
Locus of the lesion (CT)
4
FP PO P FP PO FP FP Caud. n., int. cps. T post. FTP FTPO (Unavailable) TP FP TPO, thalamus
68 Ischemia 53 Tumor 45 Ischemia 64 Ischemia 46 Ischemia 54 Ischemia 65 Ischemia 26 Ischemia 36 Ischemia 45 Ischemia 48 Ischemia 78 Ischemia 53 Hemorrhage 78 Tumor 56 Ischemia
-
1 30 9 20 28 16 22 8 20 5 11 -
56
Pt. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
* Duration from ictus to being studied. F Frontal. P Parietal. T Temporal. 0 Occipital.
Controls
Congruent Nonconquent condition condition Difference 143.0 34.0 81.5 111.5 10.0 26.5 18.5 22.5 148.0 21.0
62.0 220.5 267.5 159.0 52.0
125.0 25.0 60.0 81.5 129.5 1.5 1.5 -1.0 130.0 - 18.0 125.0 184.0 172.0 104.5 30.5
- 18.0
-9.0 -21.5 -30.0 119.5 -25.0 - 17.0 -23.5 - 18.0 -39.0 63.0 -36.5 -95.5 -54.5 -21.5
-4.5
-4.6
-0.1
( 4.0)
( -t 4.9)
( -t 2.6)
*
%t 12.6 26.5 26.4 26.9 94.3 91.9 104.4 12.2 185.7 16.6 35.7 34.3 41.3
(ns) (ns) (0.01)
(ns) (0.001) (0.02) (ns) (0.01) (ns) (ns) (0.01) (ns) (0.05) (0.01) (0.01)
2.2 (ns)
* The levels of significanceof the differences (paired t tests) are shown within parentheses.
t When negative, the differences between the result obtained in the noncongruent and in the congment conditions are expressed as a percentage of the error scored in the congruent condition.
I I
I
I
I
I I
I
Figure. Apparatus used for the investigation. The bisection task may be performed by either handling the pointer (triangle), as in the congruent condition of the experiment, or the portion of string marked by the rectangle, as in the noncongruent condition.
condition, in which the pointer had to be indirectly moved by means of a pulley device, so that its lateral displacement required a movement of the patient’s hand in the opposite direction. Methods. Subjects. The 2 experimental groups constituted a continuous series of 15 right-handed, right-brain-damaged patients showing left hemineglect on a cancellation task: and 10 normal subjects matched to the former for age and educational level. Table 1summarizes data relative to the patient group: age, nature and length of the illness, and locus of the lesion assessed on the basis of CT examination. Apparatus. The pulley device used in the experiment (figure) was on a table and centered on the sagittal midplane of the subject’sbody. The distance between the pivots of the 2 pulleys was 75 cm. The subjects moved the pointer (triangle) either by holding it between the thumb and the forefinger of their right hand (which in the brain-damaged group was the ipsilesional, unaffected hand)-the congruent condition-or by holding the portion of the string nearer to their body at exactly the point (rectangle)that could be aligned with the pointer in the sagittal midplane of the subject’s body-the noncongruent condition. Black lines 60 cm long and 3 mm thick printed on strips of paper were placed as shown in the figure.A strip of tissue stretchedover the nearer side of the device hid from sight the thumb and forefingerholding the pointer or the string. Procedure. The congruent and the noncongruent conditions
were examined in an ABAB experimental design, starting from the congruent condition. Each condition comprised 8 trials, 4 of which started with the pointer tip at the leftmost and 4 at the rightmostpoint of the line accordingto a random, fixed schedule. There were no time limits for the setting of the pointer tip at the subjective midpoint of the line and corrections were allowed. Deviations of the subjectivefrom the actual midpoint of the lines to be biseded were scored in millimeters; positive and negative values were arbitrarily assignedto rightward and leftward deviations, respectively.
Results. We will consider 3 theoretical possibilities. If neglect were due exclusivelyto a disorder of the (broadly conceived) sensory analysis required by the task, the bisection error should not be influenced by the condition (congruent or noncongruent) of the latter: the same degree of rightward deviation-positive score-of the subjective midpoint with respect to the objective should occur in both conditions. Conversely, if it were due exclusivelyto premotor factors (that is, to DH), the same degree of deviation should occur in the 2 conditions, but in opposed directions: rightward-positive score-in the congruent and leftward-negative score-in the noncongruent condition. If sensory and premotor factors contributed in exactly the same extent to the disorder, no deviation should occur from the actual midpoint in the noncongruent condition, whatever the amount of rightward deviation in the congruent. Table 2 shows the mean displacementsof the subjective midpoint in the congruent and noncongruent conditions of the task,relative to the performance of each patient. Trials starting with the pointer to the left and right were collapsed. The 4th column of the table reports the differences between the scores of the nonAugust I990 NEUROLOGY 40 1279
congruent and congruent conditions. This table also shows the results obtained in the control group as a whole. In 13 of 15 patients, the amount of rightward displacement of the subjective midpoint is less in the noncongruent than in the congruent condition (p < 0.004; binomial test). As shown by table 2, the difference was statistically significant in 6 of these 13 patients. No significant difference was found in the control group. In the 13 patients under consideration, the differences ranged from -9.0 mm to -95.5 mm (mean, -31.5 mm). However, the mean displacement in the noncongruent condition occurred to the left of the objective midpoint only in patients 8 and 10. Table 2 also shows that in patients 5 and 11 the mean displacement in the noncongruent condition (129.5 mm and 125.0 mm, respectively)was much further to the right than in the congruent condition (10.0mm and 62.0 mm, respectively). In both these patients the difference between the 2 conditions was significant.
Discussion. The results of the investigation lend firm support to the hypothesis that DH is one of the factors underlying unilateral neglect. The well-defined prediction of this hypothesis, namely the leftward displacement of the subjective midpoint in the noncongruent compared with the congruent condition of our experiment, was present in the great majority of patients (13 of 15), and was individually significant in 6 of them. This is all the more striking since there was virtually no difference in the results of the 2 conditions in the control group. In the noncongruent condition, perceptual and premotor factors of unilateral neglect are placed in opposition: the former pushes the subjective midpoint ipsilesionally, ie, to the right, and the latter contralesionally, ie, to the left. (The terms “perceptual” and “premotor” are used here somewhat loosely.) Both kinds of factors may coexist in the same patient (table 2); indeed, both appeared to be operative in each of the 13 patients in whom the negative values of the difference between the results obtained in the congruent and the noncongruent conditions unveiled the presence of premotor factors. In most of these patients, however, perceptual factors prevailed to a varying extent. This is shown by the fact that, with the exception of patients 8 and 10, the rightward bisection error found in the congruent condition was reduced but not reversed in direction in the noncongruent condition. Further data emerge in tables 1 and 2 which, with all the reservationsdue to the small number of patients, are undoubtedly suggestive and stimulating. The 1st paragraph of the Results section can be rephrased as follows: if the differencesbetween the results obtained in the congruentand noncongruentconditions are expressed as percentages of the error found in the congruent condition, values of 0%, loo%, and 200% would respectively correspond to exclusively perceptual, mixed (50:50), and exclusively premotor nature of the observed disorder. In patients 2 , 3 , 9 , and 13 (ie, in patients whose CT showed exclusively postrolandic lesions) this percentage (meanvalue, 25.2%)is lower than 1280 NEUROLOGY 40 August lee0
in patients 1,4,6,7,10, and 14, who all suffered from a lesion that included the frontal lobe (mean value, 74.3%;tables 1 and 2). The highest value (185.7%)was present in patient 10, who suffered from frontotemporoparietal ischemia; this patient’s bisection error thus appears to be due almost exclusively to premotor factors, ie, to DH. A relatively high value (41.3%)was also present in patient 15, who had a temporoparietooccipital and thalamic ischemia, whereas patient 8, with ischemia involving the caudate nucleus and the internal capsule, reached the value of 104.4%. Unfortunately, we did not test patients with exclusively frontal lesions; nonetheless, our observations fit Mesulam’s5conjecture strikingly well: “Whereas the salient manifestations of parietal neglect may well be sensory and those of frontal neglect motor, the dichotomy is unlikely to be absolute. Thus, patients with parietal neglect may also show reluctance to project the contralateral limb into the neglected hemispace, while those with frontal neglect may show such diminished reaction to contralateral visual stimuli that the presence of hemianopia may be suspected in the acute phase.” (p 318; emphasis added. Mesulam’s hypothesis may be slightly modified to include the ipsilesional limb in the reluctance to project into the neglected hemispace.) Our results demonstrate that the perceptual/premotor dichotomy of unilateral neglect exists, although further investigation must assess the extent to which the above suggestion is accurate relative to the neural basis of such a dichotomy. This implies that the neural representation of egocentric space is not only functionally differentiated as regards retinotopic versus corporeal coordinate^^-^ and personal versus extrapersonal environment?lo but also as regards perceptual versus premotor aspects. Patients 5 (parieto-occipital ischemia) and 11 (frontotemporoparietal ischemia) were outlyers. In both, the rightward bisection error was significantly greater in the noncongruent than in the congruent condition. We cannot explain this paradoxic behavior. In conclusion, uncoupling the direction of visual attention from that of hand movement in a line-bisection task has allowed the individuation and comparative assessment of independent perceptual and premotor factors of unilateral neglect. The latter correspond to the notion of Watson et all of unilateral directional hypokinesia. This result enriches insight into the functional complexity of space representation. Our data are too scanty to address the neural mechanisms corresponding to the functionally identified perceptual and motor factors of neglect, but are in harmony with Mesulam’s5 earlier speculations. Further research would be greatly facilitated by the simplicity of the required apparatus and by the ease with which it can be used for bedside examination. This is a point of practical importance in the comparative assessment of perceptual and premotor factors of neglect, especially in conditions of time pressure, as during right intracarotid amobarbital injection,which has recently been found to cause contralateral neglect probably dependent-at least in part-on frontal lobe dysfunction.”
References 1. Watson RT, Miller BD, Heilman KM. Nonsensory neglect. Ann Neurol1978;3:505-508. 2. Bisiach E,Capitani E, Porta E. Two basic properties of space representationinthebrain.JNeurolNeurosurgPsychiatry1985;48:141-144. 3. Gazzaniga MS, Ladavas E. Disturbances in spatial attention following lesion or disconnection of the right parietal lobe. In: Jeannerod M, ed. Neurophysiologicd and neuropsychological aspects of spatial neglect. Amsterdam North-Holland, 1987203-213. 4. Rapcsak SZ,Watson RT, Heilman KM. Hemispace-visual field interactions in visual extinction. J Neurol Neurosurg Psychiatry 1982501117-1124. 5. M e s h MM. A cortical network for directed attention and unilateral neglect. Ann Neurol1981;10309-325.
6. Coslett HB, Bowers D, Fitzpatrick E, Haws B, Heilman KM. Hemispatial hypokinesia and hemisensory inattentionin neglect. Brain (in press). 7. Heilman KM, Bowers D, Coslett HB, Whelan H, Watson RT. Directional hypokinesia: prolonged reaction times for leftward movements in patients with right hemisphere lesions and neglect. Neurology 1985;35855-859. 8. Bisiach E,Luzzatti C, Perani D. Unilateral neglect, representational schema and consciousness. Brain 1979;102:609-618. 9. Rizzolatti G, Matelli M, Pavesi G. Deficitsin attention and movement following the removal of postarcuate (area6) and prearcuate (area 8) cortex in macaque monkeys. Brain 1983;106655-673. 10. BisiachE,Perani D, Vallar G, Berti A. Unilateral neglect: personal and extra-personal.Neuropsychologia 1986;24759-767. 11. Spiers PA, Schomer DL, Blume HW, et al. Visual neglect during intracarotid amobarbital testing. Neurology (in press).
Abundant REM sleep in a patient with Alzheimer’s disease Donald L. Bliwise, PhD; German Nino-Murcia, MD; Lysia S. Forno, MD; and Chirane Viseskul, MD
Article abstract-In patients with Alzheimer’s disease (AD), greatly diminished REM sleep might be expected because of the cholinergic deficit in this disease and because cholinergic agonists stimulate REM sleep in humans and animals. We present here an unusual case of neuropathologically verified AD with abundant REM sleep. We suggest 4 possible explanations for this phenomenon: (1) selective cell loss in caudal midbrain/rostral pontine structures known to control sleep; (2) development of narcolepsy; (3) unrecognized affective disorder; (4) disruption of circadian timekeeping system. NEUROLOGY 1990;401281-1284
Alzheimer’s disease (AD) is characterized by widespread cholinergic deficiency in the brain.’ Because the occurrence of REM (rapid eye movement) sleep can be stimulated by direct, local application of cholinergic agonists in animals and infusion of cholinergic precursors in humans,2REM sleep should be greatly reduced in AD. Indeed some, but not all,3sleep studies of AD support this prediction, but none suggest that REM sleep should be abundant. In this paper, we present a patient with neuropathologically verified AD who was followedpolysomnographically over 3 years whose sleep studies showed a pattern strikingly opposite to prediction. The patient showed an atypical pattern of unusually short REM latency and abundant REM sleep. Case report. A 77-year-old retired man living at home presented to the Sleep Disorders Clinic accompanied by his wife after referral from his family physician for an evaluation of overwhelming and pervasive daytime sleepiness. The patient recalled alwaysbeing a sound sleeper and denied any history of cataplexy, hypnagogic hallucinations, or sleep paralysis. He
denied any tendency to nap, but his wife of 19 years noted a 2year history of falling asleep throughout the daytime at inappropriate times (during meals, when conversing).The patient averaged about 7 hours of sleep a night (bedtime at midnight, awakening at 7:OO AM). His wife noted only very mild, primarily continuous, snoring and absence of movements during sleep; however, the patient was enuretic almost every night. The wife also noted occasional periods of mental confusion, but both she and the family physician felt that the problem was secondary to the marked hypersomnolence. Review of systems indicated chronic morning phlegm (35 pack/year smoking history), frequent diarrhea, and poor urinary sphincter control. There was mild weight gain of 20 pounds over the preceding 2 years. The patient did not consume alcohol. An older sister (age 87) was still alive in a nursing home but had suffered a series of small strokes over the preceding few years. Medications included acetaminophen with codeine and diphenoxylate (for chronic diarrhea), piroxicam (for arthritis in both hands), and hydrochlorothiazide. Physical examination revealed bilateral finger clubbing. Rales were present but there was no evidence of consolidation.
From the Sleep Disorders Center (Drs. Bliwise and Nino-Murcia),and Department of Pathology (Dr. Fomo),Stanford University School of Medicine, Stanford; the Department of Pathology (Dr. Fomo),VA Medical Center, Palo Alto,and the Department of Pathology (Dr. Viseskul), VA Medical Center, Livermore, CA. Supportedby NIA AG-06066, NIA AG-04558, NIMH MH-40041, and the VA Medical Research Program. Received November 24,1989. Accepted for publication in final form January 29,1990. Address correspondenceand reprint request to Dr. Donald L. Bliwise, Sleep Disorders Center, Stanford University School of Medicine, Stanford, CA 94035.
August 1990 NEUROLOGY 40 1281
Article abstract-Right-brain-damaged patients showing unilateral neglect underwent a specifically devised line-bisection task that allowed uncouplingof the direction of visual attentionfrom that of hand movement. This made it possible to isolate and separatelyassess perceptual and premotor factors of the disorder.Comparison of experimentaland radiologic data suggestedthat premotor factorswere more pronounced in patientswith lesions involving the frontal lobes than in patientswith lesions confined to postrolandic areas. The technique employed is compatible with bedside examination and provides data useful for standard assessment of neglect symptomatology for both clinical and experimental purposes. NEUROLOGY 1990;40:1278-1281
First envisaged by Watson et al,l “unilateral directional hypokinesia” (DH) implies a powerful, although elusive, theoretic construct relative to the neural implementation of space representation. With this term, we refer to the reluctance of some brain-damaged patients to initiate and carry out motor activities toward the contralesional side of egocentric space, irrespective of the side of the limbs involved in such activities. As one of the more striking manifestations of unilateral neglect, the phenomenon is more frequent and severe in right- compared with left-brain-damaged patients. DH must be distinguished from motor neglect, which is a reluctanceto move the contralesional limbs irrespective of the direction of movement and is not related to functional impairment due to disorders other than hemineglect. The concept of DH implies the existence of a mechanism subserving spatial aspects of motor ideation and preparation, which is relatively independent of the sensory-based spatial representation of the organism’s environment. The theoretic significance of such a construct is self-evident. It prefigures a (longitudinally) componential structure of space representation, orthogonal to the (transversely) componential structure resulting from different input modalities and from the different frames of reference (eg, retinotopic versus body-centered) in which unilateral neglect may manifest i t ~ e l f .The ~ - ~construct underlying DH is therefore heuristicallypowerful in that it motivates research that might significantlyimprove our knowledge of the brain mechanisms involved in space representation. In particular, as was first suggested by Watson et all and later by Mesulam: it might help to differentiate varieties of neglect specific to anterior and posterior brain damage. On closer examination, however, the construct at ~~
~~~
~~
issue proves less manageable than at first glance, insofar as it is difficult to conceivejust how, if at all, it might be relatively independent of earlier stages of space representation, purportedly uninvolved in the initiation of any particular motor activity. Furthermore, as noted by Coslett et a l 6 , empirical results allegedly in support of DH as an independent element of the neglect syndrome’ cannot rule out the possibility that such phenomena originate from dysfunction of the above-envisaged(sensory-based and behaviorally noncommitted) earlier stages of space representation. This ambiguity is still present in the investigation carried out by Coslett et a1.6 These authors asked 4 left hemineglect patients to bisect lines placed to the left or right of their body’s sagittal midplane, while controlling their performance on a T V screen which was also placed to either the left or right of that plane. By so doing, they disconnected the side of space on which the performing hand was lying and the side on which visual attention was directed; they did not uncouple, however, the direction of hand movements and the direction of the image of the movements projected on the TV screen. Thus, whatever may be the best explanation of the results, these might bear upon the issue of the multisensory-in this instance visual and somatosensory-representation of space rather than upon the distinction of input and output components in unilateral neglect. This hypothetical dissociation must be sought by uncoupling the direction of a sensory (eg, visual) stimulus and that of the movement performed by the subject in order to drive that stimulus onto a particular target. We did so by asking left hemineglect patients to perform a bisection task in 2 different conditions: a congruent (canonical) condition, in which a pointer had to be directly moved along the line, and a noncongruent
~
From the Istituto di Clinica Neurologica, UniversitA di Milano,and Iatituto Neurologic0 C. Beata, Milano,Italy. Supported by a grant from the Ministera d e b Pubblica Istruzione to Dr. E. Bisiach. Received November 17,1989. Accepted for publication in final form January 29,1990. Address correspondenceand reprint requests to Dr. Moardo Bisiach, Istituto di Clinica Neurologica d e b Universits. di Milano,Via Francesco Sfom 35,20122 Milano,Italy.
1278 NEUROLOGY 40 August 1990
Table 2. Mean displacements (mm) of the subjective midpoint in the congruent and in the noncongruent conditions and differences between the 2 conditions in the 15 patients and in the control group*
Table 1. Sex, age, and clinical data relative to brain-damaged subjects Pt. no. Sex Age
1
2 3 4 5 6 7 8 9 10 11 12 13 14 15
M M M F F F
M F F F
M F F F
M
Etiology
Duration* (days)
Locus of the lesion (CT)
4
FP PO P FP PO FP FP Caud. n., int. cps. T post. FTP FTPO (Unavailable) TP FP TPO, thalamus
68 Ischemia 53 Tumor 45 Ischemia 64 Ischemia 46 Ischemia 54 Ischemia 65 Ischemia 26 Ischemia 36 Ischemia 45 Ischemia 48 Ischemia 78 Ischemia 53 Hemorrhage 78 Tumor 56 Ischemia
-
1 30 9 20 28 16 22 8 20 5 11 -
56
Pt. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
* Duration from ictus to being studied. F Frontal. P Parietal. T Temporal. 0 Occipital.
Controls
Congruent Nonconquent condition condition Difference 143.0 34.0 81.5 111.5 10.0 26.5 18.5 22.5 148.0 21.0
62.0 220.5 267.5 159.0 52.0
125.0 25.0 60.0 81.5 129.5 1.5 1.5 -1.0 130.0 - 18.0 125.0 184.0 172.0 104.5 30.5
- 18.0
-9.0 -21.5 -30.0 119.5 -25.0 - 17.0 -23.5 - 18.0 -39.0 63.0 -36.5 -95.5 -54.5 -21.5
-4.5
-4.6
-0.1
( 4.0)
( -t 4.9)
( -t 2.6)
*
%t 12.6 26.5 26.4 26.9 94.3 91.9 104.4 12.2 185.7 16.6 35.7 34.3 41.3
(ns) (ns) (0.01)
(ns) (0.001) (0.02) (ns) (0.01) (ns) (ns) (0.01) (ns) (0.05) (0.01) (0.01)
2.2 (ns)
* The levels of significanceof the differences (paired t tests) are shown within parentheses.
t When negative, the differences between the result obtained in the noncongruent and in the congment conditions are expressed as a percentage of the error scored in the congruent condition.
I I
I
I
I
I I
I
Figure. Apparatus used for the investigation. The bisection task may be performed by either handling the pointer (triangle), as in the congruent condition of the experiment, or the portion of string marked by the rectangle, as in the noncongruent condition.
condition, in which the pointer had to be indirectly moved by means of a pulley device, so that its lateral displacement required a movement of the patient’s hand in the opposite direction. Methods. Subjects. The 2 experimental groups constituted a continuous series of 15 right-handed, right-brain-damaged patients showing left hemineglect on a cancellation task: and 10 normal subjects matched to the former for age and educational level. Table 1summarizes data relative to the patient group: age, nature and length of the illness, and locus of the lesion assessed on the basis of CT examination. Apparatus. The pulley device used in the experiment (figure) was on a table and centered on the sagittal midplane of the subject’sbody. The distance between the pivots of the 2 pulleys was 75 cm. The subjects moved the pointer (triangle) either by holding it between the thumb and the forefinger of their right hand (which in the brain-damaged group was the ipsilesional, unaffected hand)-the congruent condition-or by holding the portion of the string nearer to their body at exactly the point (rectangle)that could be aligned with the pointer in the sagittal midplane of the subject’s body-the noncongruent condition. Black lines 60 cm long and 3 mm thick printed on strips of paper were placed as shown in the figure.A strip of tissue stretchedover the nearer side of the device hid from sight the thumb and forefingerholding the pointer or the string. Procedure. The congruent and the noncongruent conditions
were examined in an ABAB experimental design, starting from the congruent condition. Each condition comprised 8 trials, 4 of which started with the pointer tip at the leftmost and 4 at the rightmostpoint of the line accordingto a random, fixed schedule. There were no time limits for the setting of the pointer tip at the subjective midpoint of the line and corrections were allowed. Deviations of the subjectivefrom the actual midpoint of the lines to be biseded were scored in millimeters; positive and negative values were arbitrarily assignedto rightward and leftward deviations, respectively.
Results. We will consider 3 theoretical possibilities. If neglect were due exclusivelyto a disorder of the (broadly conceived) sensory analysis required by the task, the bisection error should not be influenced by the condition (congruent or noncongruent) of the latter: the same degree of rightward deviation-positive score-of the subjective midpoint with respect to the objective should occur in both conditions. Conversely, if it were due exclusivelyto premotor factors (that is, to DH), the same degree of deviation should occur in the 2 conditions, but in opposed directions: rightward-positive score-in the congruent and leftward-negative score-in the noncongruent condition. If sensory and premotor factors contributed in exactly the same extent to the disorder, no deviation should occur from the actual midpoint in the noncongruent condition, whatever the amount of rightward deviation in the congruent. Table 2 shows the mean displacementsof the subjective midpoint in the congruent and noncongruent conditions of the task,relative to the performance of each patient. Trials starting with the pointer to the left and right were collapsed. The 4th column of the table reports the differences between the scores of the nonAugust I990 NEUROLOGY 40 1279
congruent and congruent conditions. This table also shows the results obtained in the control group as a whole. In 13 of 15 patients, the amount of rightward displacement of the subjective midpoint is less in the noncongruent than in the congruent condition (p < 0.004; binomial test). As shown by table 2, the difference was statistically significant in 6 of these 13 patients. No significant difference was found in the control group. In the 13 patients under consideration, the differences ranged from -9.0 mm to -95.5 mm (mean, -31.5 mm). However, the mean displacement in the noncongruent condition occurred to the left of the objective midpoint only in patients 8 and 10. Table 2 also shows that in patients 5 and 11 the mean displacement in the noncongruent condition (129.5 mm and 125.0 mm, respectively)was much further to the right than in the congruent condition (10.0mm and 62.0 mm, respectively). In both these patients the difference between the 2 conditions was significant.
Discussion. The results of the investigation lend firm support to the hypothesis that DH is one of the factors underlying unilateral neglect. The well-defined prediction of this hypothesis, namely the leftward displacement of the subjective midpoint in the noncongruent compared with the congruent condition of our experiment, was present in the great majority of patients (13 of 15), and was individually significant in 6 of them. This is all the more striking since there was virtually no difference in the results of the 2 conditions in the control group. In the noncongruent condition, perceptual and premotor factors of unilateral neglect are placed in opposition: the former pushes the subjective midpoint ipsilesionally, ie, to the right, and the latter contralesionally, ie, to the left. (The terms “perceptual” and “premotor” are used here somewhat loosely.) Both kinds of factors may coexist in the same patient (table 2); indeed, both appeared to be operative in each of the 13 patients in whom the negative values of the difference between the results obtained in the congruent and the noncongruent conditions unveiled the presence of premotor factors. In most of these patients, however, perceptual factors prevailed to a varying extent. This is shown by the fact that, with the exception of patients 8 and 10, the rightward bisection error found in the congruent condition was reduced but not reversed in direction in the noncongruent condition. Further data emerge in tables 1 and 2 which, with all the reservationsdue to the small number of patients, are undoubtedly suggestive and stimulating. The 1st paragraph of the Results section can be rephrased as follows: if the differencesbetween the results obtained in the congruentand noncongruentconditions are expressed as percentages of the error found in the congruent condition, values of 0%, loo%, and 200% would respectively correspond to exclusively perceptual, mixed (50:50), and exclusively premotor nature of the observed disorder. In patients 2 , 3 , 9 , and 13 (ie, in patients whose CT showed exclusively postrolandic lesions) this percentage (meanvalue, 25.2%)is lower than 1280 NEUROLOGY 40 August lee0
in patients 1,4,6,7,10, and 14, who all suffered from a lesion that included the frontal lobe (mean value, 74.3%;tables 1 and 2). The highest value (185.7%)was present in patient 10, who suffered from frontotemporoparietal ischemia; this patient’s bisection error thus appears to be due almost exclusively to premotor factors, ie, to DH. A relatively high value (41.3%)was also present in patient 15, who had a temporoparietooccipital and thalamic ischemia, whereas patient 8, with ischemia involving the caudate nucleus and the internal capsule, reached the value of 104.4%. Unfortunately, we did not test patients with exclusively frontal lesions; nonetheless, our observations fit Mesulam’s5conjecture strikingly well: “Whereas the salient manifestations of parietal neglect may well be sensory and those of frontal neglect motor, the dichotomy is unlikely to be absolute. Thus, patients with parietal neglect may also show reluctance to project the contralateral limb into the neglected hemispace, while those with frontal neglect may show such diminished reaction to contralateral visual stimuli that the presence of hemianopia may be suspected in the acute phase.” (p 318; emphasis added. Mesulam’s hypothesis may be slightly modified to include the ipsilesional limb in the reluctance to project into the neglected hemispace.) Our results demonstrate that the perceptual/premotor dichotomy of unilateral neglect exists, although further investigation must assess the extent to which the above suggestion is accurate relative to the neural basis of such a dichotomy. This implies that the neural representation of egocentric space is not only functionally differentiated as regards retinotopic versus corporeal coordinate^^-^ and personal versus extrapersonal environment?lo but also as regards perceptual versus premotor aspects. Patients 5 (parieto-occipital ischemia) and 11 (frontotemporoparietal ischemia) were outlyers. In both, the rightward bisection error was significantly greater in the noncongruent than in the congruent condition. We cannot explain this paradoxic behavior. In conclusion, uncoupling the direction of visual attention from that of hand movement in a line-bisection task has allowed the individuation and comparative assessment of independent perceptual and premotor factors of unilateral neglect. The latter correspond to the notion of Watson et all of unilateral directional hypokinesia. This result enriches insight into the functional complexity of space representation. Our data are too scanty to address the neural mechanisms corresponding to the functionally identified perceptual and motor factors of neglect, but are in harmony with Mesulam’s5 earlier speculations. Further research would be greatly facilitated by the simplicity of the required apparatus and by the ease with which it can be used for bedside examination. This is a point of practical importance in the comparative assessment of perceptual and premotor factors of neglect, especially in conditions of time pressure, as during right intracarotid amobarbital injection,which has recently been found to cause contralateral neglect probably dependent-at least in part-on frontal lobe dysfunction.”
References 1. Watson RT, Miller BD, Heilman KM. Nonsensory neglect. Ann Neurol1978;3:505-508. 2. Bisiach E,Capitani E, Porta E. Two basic properties of space representationinthebrain.JNeurolNeurosurgPsychiatry1985;48:141-144. 3. Gazzaniga MS, Ladavas E. Disturbances in spatial attention following lesion or disconnection of the right parietal lobe. In: Jeannerod M, ed. Neurophysiologicd and neuropsychological aspects of spatial neglect. Amsterdam North-Holland, 1987203-213. 4. Rapcsak SZ,Watson RT, Heilman KM. Hemispace-visual field interactions in visual extinction. J Neurol Neurosurg Psychiatry 1982501117-1124. 5. M e s h MM. A cortical network for directed attention and unilateral neglect. Ann Neurol1981;10309-325.
6. Coslett HB, Bowers D, Fitzpatrick E, Haws B, Heilman KM. Hemispatial hypokinesia and hemisensory inattentionin neglect. Brain (in press). 7. Heilman KM, Bowers D, Coslett HB, Whelan H, Watson RT. Directional hypokinesia: prolonged reaction times for leftward movements in patients with right hemisphere lesions and neglect. Neurology 1985;35855-859. 8. Bisiach E,Luzzatti C, Perani D. Unilateral neglect, representational schema and consciousness. Brain 1979;102:609-618. 9. Rizzolatti G, Matelli M, Pavesi G. Deficitsin attention and movement following the removal of postarcuate (area6) and prearcuate (area 8) cortex in macaque monkeys. Brain 1983;106655-673. 10. BisiachE,Perani D, Vallar G, Berti A. Unilateral neglect: personal and extra-personal.Neuropsychologia 1986;24759-767. 11. Spiers PA, Schomer DL, Blume HW, et al. Visual neglect during intracarotid amobarbital testing. Neurology (in press).
Abundant REM sleep in a patient with Alzheimer’s disease Donald L. Bliwise, PhD; German Nino-Murcia, MD; Lysia S. Forno, MD; and Chirane Viseskul, MD
Article abstract-In patients with Alzheimer’s disease (AD), greatly diminished REM sleep might be expected because of the cholinergic deficit in this disease and because cholinergic agonists stimulate REM sleep in humans and animals. We present here an unusual case of neuropathologically verified AD with abundant REM sleep. We suggest 4 possible explanations for this phenomenon: (1) selective cell loss in caudal midbrain/rostral pontine structures known to control sleep; (2) development of narcolepsy; (3) unrecognized affective disorder; (4) disruption of circadian timekeeping system. NEUROLOGY 1990;401281-1284
Alzheimer’s disease (AD) is characterized by widespread cholinergic deficiency in the brain.’ Because the occurrence of REM (rapid eye movement) sleep can be stimulated by direct, local application of cholinergic agonists in animals and infusion of cholinergic precursors in humans,2REM sleep should be greatly reduced in AD. Indeed some, but not all,3sleep studies of AD support this prediction, but none suggest that REM sleep should be abundant. In this paper, we present a patient with neuropathologically verified AD who was followedpolysomnographically over 3 years whose sleep studies showed a pattern strikingly opposite to prediction. The patient showed an atypical pattern of unusually short REM latency and abundant REM sleep. Case report. A 77-year-old retired man living at home presented to the Sleep Disorders Clinic accompanied by his wife after referral from his family physician for an evaluation of overwhelming and pervasive daytime sleepiness. The patient recalled alwaysbeing a sound sleeper and denied any history of cataplexy, hypnagogic hallucinations, or sleep paralysis. He
denied any tendency to nap, but his wife of 19 years noted a 2year history of falling asleep throughout the daytime at inappropriate times (during meals, when conversing).The patient averaged about 7 hours of sleep a night (bedtime at midnight, awakening at 7:OO AM). His wife noted only very mild, primarily continuous, snoring and absence of movements during sleep; however, the patient was enuretic almost every night. The wife also noted occasional periods of mental confusion, but both she and the family physician felt that the problem was secondary to the marked hypersomnolence. Review of systems indicated chronic morning phlegm (35 pack/year smoking history), frequent diarrhea, and poor urinary sphincter control. There was mild weight gain of 20 pounds over the preceding 2 years. The patient did not consume alcohol. An older sister (age 87) was still alive in a nursing home but had suffered a series of small strokes over the preceding few years. Medications included acetaminophen with codeine and diphenoxylate (for chronic diarrhea), piroxicam (for arthritis in both hands), and hydrochlorothiazide. Physical examination revealed bilateral finger clubbing. Rales were present but there was no evidence of consolidation.
From the Sleep Disorders Center (Drs. Bliwise and Nino-Murcia),and Department of Pathology (Dr. Fomo),Stanford University School of Medicine, Stanford; the Department of Pathology (Dr. Fomo),VA Medical Center, Palo Alto,and the Department of Pathology (Dr. Viseskul), VA Medical Center, Livermore, CA. Supportedby NIA AG-06066, NIA AG-04558, NIMH MH-40041, and the VA Medical Research Program. Received November 24,1989. Accepted for publication in final form January 29,1990. Address correspondenceand reprint request to Dr. Donald L. Bliwise, Sleep Disorders Center, Stanford University School of Medicine, Stanford, CA 94035.
August 1990 NEUROLOGY 40 1281
