DEVELOPMENT OF INTELLIGENCE AND MEMORY IN CHILDREN ...

Brain (1992), 115, 315-329

DEVELOPMENT OF INTELLIGENCE AND MEMORY IN CHILDREN WITH HEMIPLEGIC CEREBRAL PALSY THE DELETERIOUS CONSEQUENCES OF EARLY SEIZURES by FARANEH VARGHA-KH ADEM , ' ELIZABETH ISAACS,1 SIEBEREN VAN DER WERF,1 STEPHANIE ROBB3 and JOHN WILSON2 (From the Departments of ^Neurology and Developmental Paediatrics, Institute of Child Health, the ^•Department of Neurology, Hospital for Sick Children and the 3Newcomen Centre, Guy's Hospital, London, UK) SUMMARY Although substantial data exist regarding the consequences of early lateralized cerebral lesions on intelligence and language ability, little is known about the development of other cognitive functions after such lesions. We examined the development of both verbal and nonverbal memory in 82 hemiplegic children, grouped according to hemispheric side of injury and presence or absence of seizure disorder. The control group consisted of 41 age-matched normal children, most of them siblings of the patients. Measures were obtained of intelligence and of immediate and delayed recall for prose passages, word paired associates and geometric designs. Electrophysiological and neuroradiological measures were available for a majority of the patients. The scores of the hemiplegic children on the IQ and memory tests did not exhibit the pattern seen in brain-damaged adults, in that the children's deficits showed no relation to hemispheric side of damage. Indeed, early cerebral damage to either hemisphere, even if extensive, resulted in relatively few and mild deficits if the damage was unaccompanied by seizure activity. By contrast, early lateralized lesions that were accompanied by a seizure disorder resulted in both a high incidence and degree of deficit that was unrelated to lesion side.

INTRODUCTION

The incidence of cerebral palsy of perinatal origin is approximately two in 1000 births (Brett, 1983). Of these cerebral palsy cases, approximately one-third have hemiplegia (Paneth and Kiely, 1984; Perlstein and Hood, 1954). The hemispheric brain injury responsible for the hemiplegia is presumed to be a thrombotic, vasospasmic or embolic episode in the middle cerebral and/or the internal carotid artery. Although the etiology of the circulatory event is not well understood, it is believed to occur sometime between the end of the second trimester of pregnancy and the early postnatal period (Brett, 1983; Hagberg and Hagberg, 1984). Typically, the pregnancy and delivery are uneventful, and it is usually not until after the first few months of life that a weakness of one set of limbs is noted. Between 30% and 40% of such hemiplegic cases develop a cerebral seizure disorder (Perlstein and Hood, 1954; Holt, 1961), which commonly responds to anticonvulsant regimes (Holt, 1961) but not invariably (Menkes, 1974). Correspondence to: Dr F. Vargha-Khadem, The Wolfson Centre, Mecldenburgh Square, London WC1N 2AP, UK. © Oxford University Press 1992

316

F. VARGHA-KHADEM AND OTHERS

In adults, the syndrome most akin to hemiplegic cerebral palsy of perinatal origin also results from cerebro-vascular accidents in the territory of the middle cerebral artery. Such strokes in the left hemisphere of adults often lead not only to hemiplegia but also to impairments in various cognitive functions, including different aspects of language (Brust et al., 1976; Chiarello and Church, 1986; Adams and Victor, 1989), verbal memory (Beauvois and Lhermitte, 1975; Coughlan, 1979; Risse et al. ,1984; Cappa et al., 1990), calculation ability (Grafman et al., 1982; Rosselli and Ardila, 1989; see also Gerstmann's Syndrome, e.g. Roeltgen et al., 1983), reading and writing (Hecaen and Kremin, 1977; Basso et al., 1978) and manual gestures (Kimura and Archibald, 1974). Similarly, hemiplegia following right hemisphere strokes in adults is often accompanied by impairments in visuo-perceptual and spatial functions (Kase, 1988) as well as in nonverbal memory (Boiler and De Renzi, 1967). Examination of the cognitive consequences of early lateralized lesions (i.e. before age 1) has been limited mainly to assessment of IQ and language. IQ studies have yielded variable results (Woods, 1980; Vargha-Khadem et al., 1985; Riva and Cazzaniga, 1986; Aram and Ekelman, 1988) but little evidence of the laterality effect (e.g. selective lowering of verbal IQ after left hemisphere damage and of performance IQ after right) commonly found in adult patients with unilateral lesions. As for language studies, it is clear that early left hemisphere lesions, unlike those in adulthood, do not cause prolonged dysphasia and lead to at least gross sparing of speech and language (Basser, 1962; Alajouanine and Lhermitte, 1965; Wood and Carey, 1979; Rankin et al., 1981; Vargha-Khadem et al., 1985). In contrast to the relative abundance of data on the effects of early lateralized lesions on intelligence and language ability, there is a dearth of information on the later development of other cognitive abilities. An important question which has not been addressed is whether the escape of function is specific to verbal IQ and language. Do other aspects of cognitive performance typically subserved by the left hemisphere, such as verbal memory, verbal fluency, reading and writing, and manual gestures show the same pattern of sparing and reorganization? Similarly, what aspects of cognitive performance normally dependent on the right hemisphere are spared or impaired following early right hemisphere lesions? Our purpose in the present study was to begin to fill this information gap by examining the development of both verbal and nonverbal memory in hemiplegic children with and without seizures. METHODS Subjects Forty-two patients with left hemisphere lesions, 40 patients with right hemisphere lesions, and 41 normal control subjects were studied. The patients had all presented as children with neurological problems to the neurology departments of two paediatric hospitals, either for initial diagnosis or routine follow-up management. The criterion for the selection of patients was the existence of a measurable degree of hemiparesis, independent of severity, resulting from a unilateral brain lesion sustained pre- or perinatally. The unilateral nature of the injury was determined by neurological examination, electroencephalograms (EEG) and computed tomography (CT) or magnetic resonance imaging (MRT). Patients with bilateral involvement and those with indications of other confounding disorders were excluded from the sample. The perinatal origin of the brain injury was determined by a neurologist on the basis of (1) a normal

DEVELOPMENT OF MEMORY IN CEREBRAL PALSY CHILDREN

317

course of pregnancy, delivery and postpartum period; (2) a unilateral motor, coordination or body size deficit as documented by a physician during the first 6 mths of life; (3) an abnormally early demonstration of hand preference (i.e. preferential use of one hand and neglect of the other). Patients were excluded if there was evidence that the hemiparesis followed a late postnatal episode (i.e. later than 6 mths of age) (Michaelis et ai, 1980; Cohen and Duffner, 1981). The patient groups with left or right hemisphere lesions were further subdivided on the basis of presence or absence of clinical seizures of cerebral origin, exclusive of febrile seizures during the first 5 yrs of life. (Only a small minority had started their first seizure after the age of 3 yrs.) Also excluded were patients with uncontrolled seizure disorders, i.e. those seizing frequently on a daily basis and/or those who had suffered episodes of status epilepticus. (Patients with seizures occurring on a daily basis were excluded to avoid bias from the obfuscating effect of a recent attack, whereas patients with episodes of status epilepticus were excluded because of the possibility of bilateral hemispheric involvement and insult as a result of such episodes.) Finally, patients with IQ scores below 70 were excluded from the sample. The latter restriction resulted in the elimination of one subject only. The resulting groups were designated LNS (n = 28) for left hemisphere lesions with no seizure disorder, LS (n = 14) for left hemisphere lesions with seizure disorder, RNS (n = 24) for right hemisphere lesions with no seizure disorder; and RS (n = 16) for right hemisphere lesions with seizure disorder. All patients in the seizure groups were on anticonvulsant drug therapy. Before psychological testing, the adequacy of the blood levels for anticonvulsants was verified for all patients except two. In some cases, the medication eliminated ail seizures, but in others the seizures continued to occur as frequently as one per week. The normal control group, designated NC, consisted of 23 siblings of patients, plus 18 subjects selected from a larger pool who were considered comparable with the patients with respect to age, IQ and socioeconomic status. All subjects were of normal intelligence with a range from dull to bright normal, and none was clinically dysphasic. Table 1 shows the number of subjects in each group, and their mean full-scale, verbal and performance IQs, as well as their age at time of testing. In view of the wide age range (6—21 yrs), the groups were further subdivided by age, with the dividing point placed at 12.0 yrs, for statistical analysis.

TABLE 1. AGE AND IQ SCORES OF SUBJECT GROUPS MEANS AND STANDARD DEVIATIONS IQs Age (yrs/mlhs) Group LNS RNS LS RS NC

n

Mean

28 24 14 16 41

12.9 11.3 11.8 11.6 11.8

SD 3.7 3.5 2.6 3.6 3.3

Verbal

Full scale

99.8 100.7 83.7 89.3

SD 14.0 16.3 13.5 12.0

107.2

11.1

Mean

Mean 101.3 103.4 86.2 90.1 104.3

Performance SD

Mean

SD

16.1 16.5 13.4 14.0 12.0

98.6 97.9 83.9 90.3

14.0 15.3 12.7 12.4

109.6

12.0

Materials and procedures Psychological measures Intelligence was measured with the Wechsler Intelligence Scale for Children—Revised (Wechsler, 1976). The test was administered in full, to provide measures of both verbal and performance intelligence (VIQ, PIQ). Memory was measured with the Wechsler Memory Scale (WMS), Form I (Wechsler and Stone, 1945). Age corrections similar to those employed for individuals over the age of 20 yrs were used, as suggested by Taylor (personal communication, 1981). Three subtests were selected to provide measures of both immediate and delayed recall, as follows:

318

F. VARGHA-KHADEM AND OTHERS

(1) Logical memory. For children over the age of 12 yrs, the two standard Wechsler stories constituting the Logical Memory subtest (Form 1) were used, containing 23 and 22 units of information, respectively. For children below the age of 12 yrs, two children's stories developed by Taylor (Kimura and McGlone, 1979) were substituted for the standard Wechsler stories. Each of these contained 18 units of information. Whichever set was used, recall of each story was obtained immediately after its presentation and an average score calculated for the two stories. Following a 90 min interval filled with other tasks, delayed recall of each story was obtained and an average score again calculated (Milner, 1975). (2) Paired associate learning. This subtest was administered as described in the WMS manual and scored for immediate recall after each of the three presentations of the 10 paired associates. The set of 10 pairs consists of six pairs of related items (e.g. up-down) and four pairs of unrelated items (e.g. cabbage-pen). These two subsets were scored separately. Recall was also measured after a 90 min filled delay. Delayed recall was scored out of 10. (3) Visual reproduction. The immediate reproduction of geometric designs was scored as described in the WMS manual. In addition, following 40 min filled with interpolated tasks, subjects were again asked to draw the geometric designs from memory (Milner, 1975). Neurological measure To quantify the severity of the hemiplegia, all subjects were administered a calibrated test of hand grip strength, measured with a Kiddie Hand Dynamometer. Three readings of grip strength (in kilograms) were obtained from each hand, and the mean difference in power between the two hands was calculated. The hand ipsilateral to the hemispheric side of lesion was always tested first. To quantify severity of motor loss, results were rated in the following way. The difference in power between the two hands was calculated on the basis of norms for each hand at each age level (Lafayette Co., 1989). A difference that was more than 2 SDs above the normal differential was rated as severe motor loss (1), 1.51 to 2 SDs above normal as moderate loss (2), 1 to 1.5 SDs above normal as mild loss (3) and less than 1 SD above normal as normal (4). Neuroradiological measure Computed tomography scans were obtained for 35 patients and the abnormalities rated according to the following five-point scale (Vargha-Khadem et al., 1985): (1) moderate to marked ventricular dilation and gross loss of substance seen on three or more cuts; (2) moderate to marked ventricular dilation with minimal to moderate loss of brain substance seen on fewer than three cuts; (3) moderate cerebral atrophy indicated by moderate to marked ventricular dilation seen on three or more cuts; (4) minimal hemisphere atrophy indicated by minimal ventricular dilation—asymmetry seen on fewer than three cuts; (5) normal. Magnetic Resonance Imaging was performed on 12 other patients. The samefive-pointrating scale designed for the CT scans was adjusted to allow assessment of the abnormalities on the MRIs. Patients with CT or MRI ratings of 1 or 2 all had porencephalic cysts. Electrophysiological measure Eelectroencephalography recordings were obtained for 58 of the patients. Each recording was analysed and rated by an electrophysiologist for nature of abnormality (diffuse, focal, paroxysmal), type of epileptiform activity, severity, condition during which abnormality occurred and changes relative to previous recordings. Results of these analyses were then compiled and rated according to a four-point severity scale ranging from 1 (severe) to 4 (normal). For purposes of correlations with psychological analyses, the ratings used were those for the EEG recording closest to the time of psychological testing.

RESULTS

Psychological Statistical analyses of the psychological test results were performed in two steps. In the first step, the normal control group was excluded. The effects of hemispheric side of lesion, presence or absence of seizures, age at testing, and test condition were evaluated on each of the tests (FSIQ, VIQ, PIQ, MQ, logical memory: immediate recall; logical

DEVELOPMENT OF MEMORY IN CEREBRAL PALSY CHILDREN

319

memory: delayed recall; visual reproduction: immediate recall; visual reproduction: delayed recall; paired associate learning: related pairs; paired associate learning: unrelated pairs; paired associate learning: third trial; paired associate learning: delayed recall). In no case was the effect of hemispheric side of lesion found to be significant (see Table 2 and Appendix). The scores for this variable were therefore collapsed and a

TABLE 2. FIRST SET OF ANALYSES OF VARIANCE Variables MQ (younger group) Full-scale IQ Performance IQ and Verbal IQ

Significant effects Seizure Seizure Seizure

F 13.95 17.65 17.47

d.f. ,34 ,78 ,74

P