CHILDHOOD HEMIPLEGIA: IS THE SIDE OF LESION INFLUENCED BY A FAMILY HISTORY OF LEFT-HANDEDNESS?

Robert Goodman

d rn d 0’ m

% .. E E 4 8 a

d

406

Studies of childhood hemiplegia have consistently demonstrated an excess of right hemiplegia. Combining Uvebrant (1988) and Jarvis et al.3 (1985) recent studies with the eight older studies reiiewed by Uvebrant (1988). 56 per cent of 1716 hemiplegic children had right hemiplegia, i.e. right hemiplegias out- numbered left hemiplegias by 1 - 3 : 1. The excess of right hemiplegia may be largely or entirely confined to children with congenital rather than acquired hemi- plegia (Uvebrant 1988).

Why should the developing left hemisphere be more susceptible to damage? One possibility is that there is a link with dominance, either because dominance for hand control and language directly renders the developing left hemisphere more vulnerable, or because some aspect of early hemispheric asymmetry, such as a lateralized gradient in maturational rate (Best 1988), in- dependently increases the likelihood of left-hemisphere dominance and left- hemisphere injury. In either case, the preponderance of right over left hemi- plegia should be reduced or reversed among individuals who would have become left-handed, ambidextrous or weakly right-handed had they not been hemiplegic.

If it were possible to identify the ‘natural dominance’ of hemiplegic

children (i.e. the dominance pattern that would have emerged had the hemiplegia not supervened), it would be a relatively straightforward matter to test this possibility: children who were destined to have a dominant left hemisphere would be over-represented among children with right hemiplegia, while children who were destined to have a dominant right hemisphere would be over-represented among the smaller number of children with left hemiplegia. A direct test for a link between dominance and vulnerability may become feasible when more is known of the molecular genetics of cerebral lateralization. For the present, however, only indirect tests are possible.

One such indirect test involves a comparison of hemiplegic children with and without a family history of left- handedness. Individuals are more likely to become left-handed (or only weakly right- handed) if they have a family history of left-handedness-an empirical finding that has formed the basis for various genetic theories, of which Annett’s right- shift theory is the most widely accepted (see Annett 1985, Bishop 1990~). Con- sequently, hemiplegic children with left- handed relatives will, as a group, be less constitutionally predisposed to strong left-hemisphere dominance than hemi- plegic children without left-handed relatives. If ‘natural dominance’ does

influence the laterality of hemiplegia, this will result in a relatively low proportion of right hemiplegia among hemiplegic children with left-handed relatives.

Contrary to the hypothesis being tested here, the greater vulnerability of the left hemisphere may have nothing to do with dominance. For example, the asymmetry of the aortic arch may make the left carotid circulation more vulnerable to emboli, irrespective of the pattern of cerebral ’ specialization. Another ‘dominance is irrelevant’ hypothesis is that the left hemisphere is at greater risk because most babies are born in the left occipital position (Churchill 1968), though this hypothesis is not supported by the empirical evidence (Uvebrant 1988). I f the excess of right hemiplegia has nothing to do with dominance, there is no reason to suppose that a family history of left- handedness will affect the ratio of left and right hemiplegias.

The hypothesis that the dominant hemisphere is more vulnerable to early damage was examined as part of a large study of the causes and consequences of childhood hemiplegia.

Patients and method The London Hemiplegia Register ob- tained information on 463 children from Greater London who were aged between one and 16 years and who had a clinical diagnosis of hemiplegia, whether con- genital or acquired. Children were ascertained from multiple sources, including hospital and community paediatricians, orthopaedic surgeons and neurosurgeons, hospital and community physiotherapists, special schools and voluntary organizations. 15 1 children aged between six and 10 years were examined personally. Of these, 106 (70.2 per cent) had purely unilateral motor signs and 43 (28-5 per cent) had pre- dominantly unilateral motor signs (with varying degrees of possible or definite involvement of the contralateral side). In two of the cases (1.3 per cent) examined personally there was no evidence of a hemiplegia at the time of the assessment, so both these children were excluded from further analysis. 312 children were not examined personally, but there is no reason to doubt that the overwheiming

majority were correctly diagnosed as hemiplegic. Right hemiplegia was present in 263 (57 per cent) of the 461 cases. 61 per cent of the subjects were male. Right hemiplegia was present in 55.9 per cent of male cases and 58.9 per cent of female cases.

Parents-nearly always the mothers- completed a questionnaire on 420 children. Items related to a family history of left-handedness were completed on 414 (90 per cent) of the 461 eligible children. Parents were asked: ‘Is there anyone in the immediate family who is fully or partly left-handed?’. Those who responded in the affirmative were asked: ‘Please say who they are (brother, mother, aunt, etc.)’. Parents reported that 180 children had no left-handed relative-the family history negative (FH-) group. These children were compared with 216 children who were family history positive (FH +): 132 children who had at least one first-degree left-handed relative and 84 children who had no first-degree, but at least one second-degree, left-handed relative. In addition, seven children were reported to have only third-degree left-handed relatives-these children were excluded from further analysis because the group was too small to be considered alone and it was not obvious whether children with such a weak family history should be grouped with FH+ or FH- children. The parents of a further 11 children reported that the children did have left-handed relatives, but failed to specify who they were. Though these 11 children were excluded from further analysis, their inclusion in the FH+ group would not substantially have altered the findings reported here (with nine of the 11 having right hemiplegia).

Considering the first-degree relatives of the 396 FH+ and FH- children, left- handedness was reported to be present in 8.1 per cent of mothers, 11 - 9 per cent of fathers and 11.8 per cent of siblings.

Some of the analyses were performed after eliminating children with evidence of possible or definite motor involvement of the contralateral side. For the 149 hemiplegic children who were examined personally, this was based on a detailed neurological examination. For the

2 2 P Q

407

t remaining children, exclusion was based on a questionnaire report from parents that the arm and leg on the ‘good’ side were not completely normal but were partly affected by weakness, stiffness or clumsiness. (Parental report and neurological examin- ation corresponded fairly well for the 149 children for whom both were available: phi coefficient = 0.43, p<O.OOI.) 75 children

5 were eliminated on this basis and the remaining 321 children were designated the

c 0 ‘purely unilateral’group’. 2 Congenital hemiplegia was defined as a

hemiplegia in which the motor deficit z emerged gradually over the first year or

two of life with no clearly definable onset .- m and with no likely or definite brain insult

after the neonatal period. Hemiplegia was considered acquired if the onset was sudden and associated with an acute illness or accident, or with clinical or radiological evidence of an acquired lesion. I t was also considered to be acquired if the motor deficit emerged gradually in the second half of the first year after an earlier post-neonatal brain insult (e.g. the gradual emergence of a hqmiplegia from about six months of age in a child who had had severe bacterial meningitis at three months).

E x s w

k s E

4

*

; m

c u -I

0 - VI .-

TI m

cl) u - n X

.- E

Results Of the 396 hemiplegic children from the London Hemiplegia Register who could be classified as F H + or FH-, 227 (57 .3 per cent) had a right hemiplegia. The right side of the body was affected in 195 (57-5 per cent) of 339 children with congenital hemiplegia. compared with 32 (56.1 per cent) of 57 children with acquired hemiplegia (x2=0.003, df 1 , NS).

The preponderance of right hemiplegia was confined to the FH + group. Whereas 139 (64 per cent) of the 216 FH + children had a right hemiplegia, only 88 (49 per cent) of the 180 FH- children had a right hemiplegia (x2 = 8.97, df 1, p = 0.003).

The analysis was repeated after excluding 75 children with evidence of possible or definite motor signs on the contralateral side. Of the 175 F H + children with purely unilateral motor involvement, 113 (65 per cent) had a right hemiplegia; of the 146 FH- children with purely unilateral motor involvement, 73

408 (50 per cent) had a right hemiplegia

(x2=6*35, ef 1, p=0*012). There was no evidence of a ‘dose

response’. Right hemiplegia was present in 82 (62 per cent) of the 132 F H + children who had at least one affected first-degree left-handed relative and in 57 (68 per cent) of the 84 children who had no first-degree but at least one second- degree left-handed relative (x2 = 0.51, df 1, NS). I t was of little consequence which first-degree relative was affected: right hemiplegia was present in 61 per cent of those with a left-handed mother, in 58 per cent of those with a left-handed father, and in 66 per cent of those with a left-handed sibling.

The link between a positive family history of left-handedness and a pre- ponderance of right hemiplegia was most marked among the children with an acquired hemiplegia. Among the 57 children with an acquired hemiplegia, 21 (84 per cent) OC the 25 FH + children had a right hemiplegia, compared with 11 (34 per cent) of the 32 FH- children (x2= 12.09, df 1 , p=O*OOl). Among the children with a congenital hemiplegia, 118 (62 per cent) of the 191 FH + children had a right hemiplegia, compared with 77 (52 per .cent) of the 148 FH- children (x2=2*86, df 1 , p=0*09) .

The link between family history and side of hemiplegia was stronger in girls. Of the 78 FH + girls, 57 (73 per cent) had a right hemiplegia, compared with 32 (44 per cent) of the 7 3 FH- children ( x 2 = 12.14, df 1 , p < O - O O l ) . Of the 138 F H + boys, 82 (59 per cent) had a right hemiplegia, compared with 56 (53 per cent) of the 107 FH- boys (x2=0.96, df 1, NS).

Finally, it was important to ask whether children with right hemiplegia were more likely to have at least one left- handed relative simply because they had substantially more relatives. In fact, there was no significant difference in the number of first-degree relatives. Children with right hemiplegia had an average of 3.43 first-degree relatives, while children with left hemiplegia had an average of 3.25 first-degree relatives ( I test, p=O.12). Since no information was collected on the number of second-degree relatives, it was not possible to compare the size of the extended families.

Discussion Contrary to the prediction, hemiplegic children with a family history of left- handedness were significantly more, rather than less, likely to have a right hemiplegia. This is strong evidence against the hypothesis that right hemiplegia is more common than left hemiplegia because the dominant hemi- sphere is more vulnerable to early damage.

I t is difficult to know what inter- pretation to place on the unexpected association between a family history of left-handedness and right hemiplegia. The association was highly significant, with only a one in 300 possibility of such a strong association occurring by chance. Furthermore, the findings emerged from hypothesis testing rather than from ‘data dredging’, i.e. the highly significant findings were not drawn from a much larger pool of largely negative findings generated by correlating everything with everything else.

Two possible causes for a spurious association between a family history of left-handedness and right hemiplegia need to be considered. First, as Bishop (1980) pointed out, the presence of at least one left-handed relative is more likely when the proband comes from a large family, so the association between a family history of left-handedness and right hemiplegia could have been the result of children with right hemiplegia being drawn from larger families. In fact, there was no significant difference in family siLe between children with left and right hemiplegia, at least as far as first-degree relatives were concerned. A spurious association would also have arisen if parents with a strongly left-handed child (such as a child with right hemiplegia) are more likely to notice and remember other left-handed people, including left-handed relatives. Though this may be relevant to the reported rate of left-handedness in second-degree relatives-with parents of left-handed children being more likely to notice left-handed ‘in-laws’-it seems most unlikely that parents’ awareness of their own, their spouse’s or their other children’s handedness is much affected by this source of bias. I t is improbable, therefore, that biased recall accounts for

the association between a family history of left-handedness in a first-degree relative and right hemiplegia.

Though the association between right hemiplegia and family history of left- handedness was highly significant and was not apparently due to greater family size or biased recall, there are several reasons for supposing that the association occurred by chance rather than as the result of a genuine causal link. First, there was no ‘dose response’: chiIdren without a family history of left-handedness had the lowest proportion of right hemiplegia, children with a mild family loading for left-handedness had the highest proportion of right hemiplegia, and children with a strong family loading for left-handedness had an intermediate level of right hemiplegia. Second, since an excess of right hemiplegia is more typical of congenital than acquired hemiplegia (Uvebrant 1988), a true association between a family history of left- handedness and side of hemiplegia ought to have been more obvious among those with congenital than acquired hemi- plegia-contrary to the findings of this study. Third, there is no obvious reason why a genuine effect of family history of left-handedness on the side of hemiplegia should largely be confined to girls. Finally, since the presence or absence of a family history of left-handedness is a relatively weak predictor of the ‘natural dominance’ that the child would have developed had hemiplegia not supervened (Bishop 1990b), it is hard to see how it could be such a strong predictor of side of hemiplegia among girls and children with acquired lesions.

On balance, therefore, it seems likely that a family history of left-handedness does not influence the laterality of the hemiplegia. In passing, it is worth noting that a family history of left-handedness does not seem to affect the rate of hemiplegia either-the proportion of left- handed parents and siblings is close to the observed .proportion in the general population (Bishop 1990b).

Despite all the reservations discussed, there is still an outside possibility that the association between right hemiplegia and a family history of left-handedness is a genuine finding. If so, the association 409

a- m

v) e B c

U. m r c s L 0

0 2

I

- “? .-

U C

L

410

could potentially be explained in terms of Geschwind and Galaburda’s (1985) theory of the nature of left-hwkdness. Follow- ing their line of reasoning,” a family history of left-handedness would, on I am very grateful to all the parents for finding time average, be associated with a slower rate to complete the questionnaires, to Carole Yude, Bob

Adak and Suzanne Pemberton for helping to collect Of development Of the lef t hemisphere* and code the questionnaires. and to the Wellcome and this slower development could Trust and The Spastics Society for their generous

ment is what puts males at greater overall risk of brain damage). Accepted for publication 20th September 1993. Acknowledget,lents

potentially result not only in a tendency support. for the right hemisphere to take the Author’s Appointments dominant role, but also in greater Robert Goodman, M.R.C.P.. hZ.R.C.Psych.,

Senior Lecturer, Institute of Psychiatrv, vulnerability of the left hemisphere to Dcrespigny Park, London SES 8AF, uK; Honorary injury cust as Taylor and Ounsted (1972) Senior Lecturer, Institute of Child Health, London. have argued that slower brain develop- Correspondence to frrst address.

SUMMARY Family histories of left-handedness were obtained for 396 children with congenital or acquired herniplegia. As in other studies, right-sided hemiplegia was substantially more common than left- sided hemiplegia. The excess of right hemiplegia was entirely accounted for by children with left- handed relatives. This is contrary to what would be expected if the preponderance of right hemiplegia reflected a greater vulnerability of the dominant hemisphere to early damage. Though the observed association between right hemiplegia and a family history of left-handedness may well have occurred by chance, it could potentially reflect some heritable aspect of cerebral asymmetry (such as slow development of the left hemisphere) that predisposes an individual both to left-handedness and to left-hemisphere injury.

RESUME Heniiplegie de I‘enfance: le cbte de la lesion esr-il influence par une histoire familiale de gaucherie Les histoires familiales de gaucherie ont eti recueillies chez 396 enfants avec hemiplegie congenitale ou acquise. Comme dans les autres etudes, l’hemiplegie droite etait significativement plus frequente que I’hemiplegie gauche. L’exces d’hemiplegie droite etait entierement explique par les enfants de familles de gauchers. Cela est a I’oppose de ce qui pouvait ttre attendu si la preponderance des hemiplegies droites traduisait une plus grande vulnerabilite de I’hemisphere dominant vis a vis des lesions precoces. Bien que I’association observee entre I’hemiplegie droite et une histoire familiale de gaucherie puisse t tre due au hasard, elk pourrait auusi bien traduire u n aspect hereditaire d’asymetrie cerebrale (tel qu’un plus lent developpement de I’hemisphere gauche) predisposant u n individu a la fois 5 la gaucherie et a une lesion de I’hemisphere gauche.

ZUSAMMENFASSUNG Hemiplegie im Kindesalter: 1st die Seite der Lasion durch jamiliare Linkshandigkeir beein flupt? Bei 396 Kindern mit kongenitaler oder erworbener Hemiplegie fand sich eine familiare Linkshandigkeit. \Vie in anderen Studien war die rechtsssitige Hemiplegie deutlich haufiger als die linksseitige. Der GroRteil der rechtsseitigen Hemiplegie entfiel auf Kinder mit linkshandigen Angehorigen. Dies steht im Gegensatz zu dem, was man erwarten wurde, wenn die rechtsseitige Hemiplegie Ausdruck einer groaeren Vulnerabilitat der dominanten Hemisphare durch fruhzeitige Schadigungen ware. Obwohl die beobachtete Relation zwischen rechtsseitiger Hemiplegie und familiarer Linkshandigkeit zufallig aufgetreten sein konnte, konnte sie vielleicht ,auch auf eine erbliche cerebrale Asymmetrie hinweisen (wie langsame Entwicklung der linken Hemisphare), die

. sowohl zur linkshandigkeit als auch zur Schadigung der linken Hemisphare pradisponiert.

RESUMEN Hemiplejia en la infancia. El lado de la lesion esla influenciado poi una historia familiar de zurderia? Se registraron historias familiares de zurderia en 396 ninos con hemiplejia congenita o adquirida. Como en otros estudios. la hemiplejia del lado derecho era mas frecuente que la del lado izquierdo. Este exceso de hemiplegia derecha estaba en relacibn enteramente con la existencia de parientes con zurderia. Esto va en contra de lo que carbria esperar si la predominancia de hemiplejia derecha reflejara una mayor vulnerabilidad del hemisferio dominante frente a una lesion precoz. Si bien la asociacibn de la hemiplejia derecha con una historia familiar de zurderia puede ser una casualidad. podria reflejar potencialmente algun aspect0 heredado de asimetria cerebral (como un desarrollo mas lento del hemisferio izquierdo) que predisponga a un individuo a la zurderia y a la lesi6n del hemisferio izquierdo.

References Annett, M. (1985) Lefi. Righi. HandandBrain: The

Right Shill Theory. London: Lawrence Erlbaum Associates.

Rest, C. T. (1988) T h e emergence of cerebral asymmetries in early human development: a literature review and an neuroembryological model.' In Molfese, D. L., Segalowitz, S. J. (Eds.) Brain Lareralizarion in Children: Developmental Implicarions. New York: Guilford Press.

Bishop. D. V . M. (1980) '%leastiring familial sinistrality.' Cortex, 16, 31 1-313.

- ( 19900) Handedness and Developmenral Disorder. Clinics in Developmental Medicine, No. 110. London: Mac Keith Press.

- (19906) 'On the fu t i l i t y of using familial sinistrality 10 subclassify handedness groups.' Cortex, 26, 153-155.

Churchill, J . A. (1968) 'A study of hemiplegic

cerebral palsy.' Developmenial Medicine and Child Neurology, 10, 453-459.

Geschwind, S., Galaburda, A. hl. (1985) 'Cerebral fareralization: biological mechanisms. associ- ations. and pathology. 1: A hypothesis and a program for research.' Archives of Neurology,

Jarvis, S. N., Holloway, J. S.. Hey, E. N. (1985) 'Increase in cerebral palsy in normal birthweight babies.' Archives of Disease in Childhood. 60,

Taylor, D. C., Ounsted, C. (1972) 'The nature of gender differences explored through ontogenctic analysis of sex ratios in disease.' In Ounsted. C.. Taylor, D. C. (Eds.) Gender Differences: Their Onrogeny and Significance. Edinburgh: Churchill 1.ivingstone.

Uvebrant, P. (1988) 'Hemiplegic cerebral palsy: aetiology and outcome.' Aria Paediarrica Scandinavica, Suppl. 345.

42, 428-459.

1 1 13-1 121.

- Y a 0 v a- m

2

411