American Journal of Medical Genetics 37:346-351 (1990)
Cerebral Palsy in Multiple Births in Western Australia: Genetic Aspects Beverly Petterson, Fiona Stanley, and Diane Henderson Department of Anatomy and H u m a n Biology, University of Western Australia (B.P.) and National Health and Medical Research Council Unit in Preventive Medicine and Epidemiology (F.S.), Western Australia, Australia; Los Angeles Children$ Hospital, Los Angeles (D.H.)
A study of cerebral palsy in multiple births was undertaken to test genetic involvement and assess the impact of the special conditions of pregnancy and parturition in these cases. Complete ascertainment of cerebral palsy in multiple gestations that occurred in Western Australia between 1956 and 1985 was obtained from the Western Australian Cerebral Palsy Register. There were 74 twins and 5 triplets. Data on sex, birth order, motor handicap, outcome in co-twins and triplets, zygosity, and pedigree information was obtained from the Register, hospital records, and, where possible, by interview of the parent(s1 of the propositi. There was a significantly higher (P = 0.0026) concordance rate in MZ than in DZ twin pairs. However, pedigree studies showed no other relatives with a motor handicap similar to that of the propositi. This is consistent with a multifactorial cause in at least some of the cases. The sex ratio of affected twins was found to be 2.1 compared to 1.3 for singletons and all 5 affected triplets were boys. The trend of increasing sex ratio with increasing plurality was significant at the 1%level.
tance of a variety of pre- and perinatal conditions and the degree to which genetic factors are involved have been debated since the time of Little [1862] and Freud [18971. Previous work has suggested that the recurrence risk for CP among sibs of unselected cases is low, about 1-2% [Ingram, 1964; Bundey, 1985; Gustavson et al., 19691. However, this is 5 times the prevalence in all births. In selected CP cases in which there are no perinatal risk factors, the recurrence risk for sibs is often much higher. It has been reported as 1 in 9 where the motor handicap is spastic and symmetrical [Bundey and Griffiths, 19771, and calculated to be 1 in 8 in ataxia with mental retardation [Gustavson et al., 19691. Genetic factors may contribute to this increased risk factor either in the form of single gene defects [Bundey, 1985; Adler, 19611, by chromosome microdeletions or rearrangements, or by giving rise to a constitutional vulnerability to other risk factors in a multifactorial cause [Polani, 19841. By using a polygenic model [Falconer, 19651, a heritability of approximately 40% can be calculated from the population birth prevalence relative to that in sibs of propositi. If genetic factors are significant one might expect to find a higher concordance rate in MZ compared to DZ twins. The occurrence of CP in multiple births affords an opportunity to assess the impact of the special conditions of pregnancy and parturition in these cases and also to test the hypothesis of genetic involvement. With these 2 aims, a descriptive study of all cases of CP in multiple births in Western Australia between KEY WORDS: motor handicap, twins, concordance rate, sex ratio 1956 and 1985 was undertaken. All cases born during the study period were ascertained and data on the preand perinatal courses of each individual, as well as their family histories, were collected. We also attempted to INTRODUCTION establish the zygosity and outcome in co-sibs for all There is little doubt that cases of cerebral palsy (CP) multiple gestations in the study group. are etiologically heterogeneous. The relative imporThis report presents part of this descriptive study and is primarily concerned with a n assessment of the degree to which genetic factors are involved in the cause of the Received for publication October 16, 1989; revision received group of disorders which come under the heading of CP. March 29, 1990. Subsequent reports will address the obstetric and periAddress reprint requests to Beverly Petterson, Department of natal risk factors and the contribution of low birth Anatomy and Human Biology, University of Western Australia, weight. Nedlands Western Australia 6009.
0 1990 Wiley-Liss, Inc.
Cerebral Palsy in Muliple Births
DEFINITIONS Cerebralpalsy (CP)is defined as a group of disorders of movement and posture due to a defect or lesion of the immature brain [Bax, 19641. It excludes progressive neurological disorders and those occurring after a documented postnatal insult. Live birth is the complete expulsion or extraction from its mother of a product of conception, irrespective of duration of pregnancy, which after separation shows signs of life. Neonatal death is the death of a liveborn infant within 28 days of birth. Stillbirth (SB) is the complete expulsion or extraction from its mother of a product of conception of at least 20 weeks gestation or 400 g birthweight which, after separation, did not show any sign of life. Perinatal death includes stillbirths and neonatal deaths.
METHODS Data Collection Case ascertainment was from the Western Australian Cerebral Palsy Register, a complete population-based collection of all diagnosed individuals with CP obtained by using multiple sources of ascertainment [Stanley, 19791. Data used for calculating sex ratios for CP in singletons and numerators for birth prevalence of CP in twins and triplets were also available from the register, as were the sex, birthweight, gestational age, and updated diagnoses of motor handicap of cases. Total annual numbers of multiple births in Western Australia were obtained from the Western Australian Midwives Notification of Births System [Moore, 19881. Outcome in co-twins and triplets and details of birth order were obtained from hospital records of birth or interview with parent(s1 of the proposita. These data were collected by 2 of the authors (B.P., D.H.) using standard data collection and parent interview schedules. Evidence used for assigning monozygosity in like-sex twin pairs included polymorphism coincidence, available from medical records for one pair, and birth records of monochorionic placentation for 3 pairs. For the rest we relied on parents’ and doctors’ assessment of the physical likeness of the twin pairs. We asked specifically about eye color, hair type and color, and facial features, and generally the difficulty of telling the twins apart. This method has been shown to be 95% accurate for normal twin pairs [Hrubec and Robinette, 19841. However, in this study the motor handicap may have affected facial features or body form in one twin of a discordant pair, which may have resulted in a n over-estimation of dizygosity in this category. Because of the vast distances in Western Australia, personal interview of all parents was not possible. Therefore, a structured telephone interview was used to obtain pedigree information. The questionnaire was developed with the help of the Australian Bureau of Statistics. Specific questions on the incidence of motor handicap, epilepsy, intellectual handicap, and congenital malformations and birth losses in parents, grandparents, aunts, uncles, and cousins of the proposita were
347
included. The technique was pilot-tested and results compared well with pedigrees which were available from medical records. The pedigree interview is available from the primary author (B.P.) on request.
RESULTS A total of 79 CP individuals from multiple births in Western Australia in the 30 year period from 1956 was found. There were 74 twins and 5 triplets. Comparisons of birth prevalence of CP in singletons, twins, and triplets in Western Australia are shown in Table I. Both twins and triplets show a marked and significant increase in relative risk compared to singletons. Mean birthweights and gestational ages are also significantly lower in both twins and triplets than in singletons. The outcome in all co-triplets and all but 2 co-twins was established and the results are shown in Tables I1 and 111. Among the twins there were 6 pairs concordant for CP, 24 pairs in which the co-twin died either pre- or perinatally, and 36 pairs in which the co-twin showed no major malformation or neurological impairment. Of the 10 co-triplets there were 3 SB, 1 termination, and 6 normal individuals. There were no sets of triplets showing concordance for CP. Medical records of birth gave no indication of zygosity in these cases. Of the 3 sets in which there were 2 living children of like sex, 2 were definitely not identical and we were unable to contact the other family (Table 111,family 3). It is interesting to note that all triplet sets were born between 1980 and 1985. Only one set was conceived spontaneously. Two sets were the result of hormonal stimulation of ovulation and 2 were the result of in vitro fertilization. Of the affected twins, 2 were conceived after hormonal stimulation of ovulation and one after i n vitro fertilization. The birth outcome in relation to birth order within twin pairs is shown in Table IV. There was no significant difference between the first- and second-born twins in relation to numbers of individuals with CP or those who died perinatally. Pedigree information was obtained from 59 of the 74 families in the study, including all 6 twin pairs concordant for CP. This showed no obvious patterns of monogenic inheritance, although we may have missed autosoma1 recessive cases. The relative rates of concordance and discordance in MZ and DZ twin pairs are shown in Table V. Discordant in this context means twin pairs in which there was one CP and one normal twin. Any pairs in which there was a pre- or perinatal death were eliminated from this analysis as were several pairs in which it was not possible to establish zygosity. The excess of concordance above expected values is highly significant ($ = 7.4, P = 0.0026). The motor handicap in each concordant pair, listed in Table VI, was very similar. Medical records showed no clear identification of syndromes. Birthweights and gestational ages in each concordant pair are also listed. The mean birthweights and gestational ages for all affected twins, and for MZ and DZ twins separately, are included for comparison. The concordant MZ twins are not significantly different in these respects from all affected MZ twins.
348
Petterson et al. TABLE I. Birth Prevalence of CP by Plurality (Western Australia 1966-1985)*
N
Plurality Singleton
877
Twins
50
Triplets
5
Birth prevalence per 1,000 livebirths 2.4 6.3 32
(95% CI) (2.22-2.54) (4.51-8.03) (3.7-59.5)
GA mean SEM range 37.5 0.21 25-46 35.1 0.53 25-42 28.6 1.50 25-34
BW mean SEM range 2,778 45.5 565-4,825 2,050 83.5 765-4,048 1,249 176.2 740-1,822
* GA, gestational age in weeks; BW, birthweight in grams; CI, confidence interval; SEM, standard error of mean.
TABLE 11. Outcome in Twin Pairs Ascertained for CP (Western Australia 1956-1985) CPICP CPISB CPlneonatal death CPlnormal CPlunknown
6 pairs 15 pairs 9 pairs 36 pairs 2 Dairs
TABLE 111. Outcome in Triplets Ascertained for CP (Western Australia 1956-1985) Spontaneous
1.
CP
2'
Termination
3.
CP
4.
0
5.
0
If. 2SB3
CP
Hormonal stimulation of ovulation
stimulation , & 0 Hormonal % of ovulation
0
.
CP
B
I n uitro fertilization
O I n vitro fertilization
CP
TABLE IV. Outcome in First- and Second-Born Twins in Twin Pairs Ascertained for CP (Western Australia 1956-1985) Twin I 35 9 44
CP SB or neonatal death Total
Twin I1 39 15 54
TABLE V. Presumed Zygosity of Surviving Twin Pairs Ascertained for CP (Western Australia 1956-1985)* Concordant Monozygotic Dizveotic * x : = 7.4; P =
6 0
~
0.0026.
Discordant 9 21
The outcome in twin pairs in which the co-twin died either pre- or perinatally is shown in Table VII. Hospital records contained little evidence of zygosity in these cases and the numbers of like and unlike sex pairs, 15 and 5, respectively, are not significantly different from the ratio of 2.6:l for all twin pairs born in Western Australia from 1981 to 1986 (V. Waddell, unpublished, 1989). The outcome by sex in all multiple gestations in the study group is shown in Table VIII. Table IX shows the ratio of maledfemales in singletons, twins, and triplets, The excess ofmales in both twins and singletons with CP compared with the expected proportion in the whole population, corrected for gestational age distribution, is highly significant. The trend of a n increasing sex ratio with increasing plurality is significant a t the 1%level. (xl for trend = 6.83).
DISCUSSION This study attempted to assess the contribution of genetic factors to the cause of CP in all twins and triplets with CP born in Western Australia between 1956 and 1985. Pedigree studies showed no obvious cases of monogenic inheritance. However, a marked excess of concordance among MZ twin pairs was found, which is consistent with a genetic predisposition. The use of the Western Australian Cerebral Palsy Register ensured complete ascertainment of CP in multiple births throughout the State [Stanley, 19791. This avoided biases imposed by using either hospital-based selection, where the milder cases are often missed, or a twin registry where twins with handicaps and especially twin pairs in which one twin has died pre- or perinatally are significantly underrepresented. Our data show a greater birth prevalence for CP in both twins and triplets than in singletons, a phenomenon reported in several studies. This has been partly attributed to the difference in birthweight distribution between multiple and singleton births and partly to the added complications of pregnancy and delivery in multiple births [Alberman, 1964; Griffiths, 1967; Russell, 1960; Illingworth and Woods, 1960; Durkin et al., 19761. The idea that multiple conception and pregnancy itself may be a risk for CP is less popular [Stanley, 19841.
Cerebral Palsy in Muliple Births
349
TABLE VI. Motor Handicap Birthweight and Gestational Age in Twin Pairs Concordant for CP (Western Australia 1956-1985) Pair No. 1. ~
Pair order 1 2
Sex d d
. . . . .-
2.
d d
3.
d d d d
4.
1 2 1 2 1 2 1 2 1 2
P
5.
P 6.
.-Motor handicap Spastic diplegia Spastic diplegia-some ataxia Spastic diplegia Spastic diplegia Ataxia, spastic quadriplegia Ataxia, spastic quadriplegia" Ataxia, mixed spastic dyskinetic Ataxia, mixed spastic dyskinetic Spastic quadriplegia Spastic quadriplegiab Spastic quadriplegia Spastic quadriplegia
d d
Mean Mean Mean Mean
(SEMI for (SEMI for @EM) for (SEMI for
Severity Moderate Moderate
35
Mild Mild
32
Mild Moderate
34
Severe Severe
40
Severe Severe
36
Mild Mild
29
concordant MZ CP twins all C P twins MZ CP twins DZ C P twins
.
~~~~
GA
BW 2,260 2,130 1,814 1,758 2,098 1,985 2,098 3,090 2,353 1,871 1,445 1,570
34.3 (1.52) 35.1 (0.53) 33.9 (0.64) 36.3 (0.82)
~~~
~~
2,039 (123.1) 2,050 (83.5) 1,961 (97.5) 2,144 (141.9)
"Epileptic, died a t 3 years from respiratory arrest. Died at 9 years from pneumonia.
TABLE VII. Perinatal Deaths in Like- and Unlike-Sex Twin Pairs, Ascertained for C P (Western Australia 1956-1985) Co-twin
Neonatal
Stillbirth
(Macerate& uawraceous)
Like sex Unlike sex Unknown sex Total
6 3
9 2 4
(1) (1) (1)
TABLE VIII. Outcome by Sex in Multiple Gestations Ascertained for CP (Western Australia 1956-1985) Twins.~ Outcome_-
Sex
d
P
?
d
Triplets" P
50 13 15
24 7 21
0 4 2b
5 1 2
0 1 4
?
~
CP Perinatal death Normal
0 1 0
1 triplet terminated. 'Outcome unknown assumed normal. a
Differences in CP and mortality rates between firstand second-born twins have also been reported [Ellis et al., 1979; Russell, 19601. However, our data, in common with other studies [Durkin et al., 1981; McCarthy et al., 19811, show no significant effect of birth order on outcome. Co-twins and triplets of CP cases show a high morbidity and mortality rate during the pre- and perinatal periods. Of the 68 twin pairs in which CP was ascertained there were only 36 (53%)where the co-twin was normal. These findings are similar to those reported by other investigators [Alberman, 1964; Griffiths, 1967; Russell, 19601.
Total
15 5 4 24
The proportion of twin pairs concordant for CP was 6/68 (8.8%), which does not differ significantly from the studies of Alberman 2/45 (4.4%)[Alberman, 19641 and Griffiths 4/78, (5.1%)[Griffiths, 19671. All 6 concordant pairs were MZ. There were 9 MZ and 15 DZ discordant pairs. The excess of concordant MZ pairs above expected values is highly statistically significant. The motor handicap in each concordant pair was very similar, but pedigrees of these families showed no other relatives with similar motor handicaps and there was no clear identification of syndromes. This, of course, does not eliminate the possibility of autosomal recessive inheritance in some cases. It is also possible that the combination of environmental stress with a susceptible genotype caused the motor handicap. However, there is another possible contributing factor to this excess of concordant MZ twin pairs. This is related to the observation that anastomoses may exist between placental circulations in MZ twins. This mechanism has certainly been implicated in cases of CP in a n individual with a papyraceous or macerated co-twin [Durkin et al., 1976; Melnick, 1977; Moore et al., 1969; Yoshioka et al., 19791. However, the similarity of motor handicap in each concordant twin pair in this study may suggest a less random causal process. In addition, there
350
Petterson et al. TABLE IX. Sex Ratios in CP bv Pluralitv (Western Australia 1956-1985) ~~
~
Singletons (720 males, 554 females) Twins (50 males, 24 females) Triplets (5 males, 0 females)
~
Proportion male
(95%CI)
Male/female
Expected“ malelfemale
0.56
(0.537-0.593)
1.3
1.1
0.68
(0.557-0.780)
2.1
1.0
1.o
(0.48- 1.00)
1.2
“Expected sex ratios according to GA distribution, calculated by using sex ratios of all liveborn singletons and multiples by GA in Western Australia 1985-1987.
was no evidence of a n increase in monozygosity among the 24 twin pairs in which the co-twin died either pre- or perinatally, a situation which is thought more likely to be due to vascular problems of this nature [Melnick, 1977; Mannino et al., 19771. One other observation which emerged from the study was a n increasing sex ratio with increasing plurality. It is interesting to relate this trend to the increasing birth prevalence of CP with plurality. Multiple births overall have lower sex ratios than singletons [Bulmer, 1970; Moore, 19881.Sex ratios for both singleton and multiple births are marginally higher in the Western Australian population for preterm than for term births [Moore, 19881. However, a comparison of expected sex ratios, according to the gestational age distribution of both the singleton and multiple CP populations, shows that low gestational age is not the only factor contributing to the male excess. It would seem that males fare worse than females under the adverse environmental conditions of increasing plurality. That changing environmental conditions can differentially affect males and females was also shown by the greater fall in mortality and CP rates for males than for females in Western Australia during the 1970’s [Stanley, 19811. Sex ratios in other animals also tend to change with changes in the environment [Austin and Edwards, 19811. In conclusion the data presented support the hypothesis that genetic factors play a part in the cause of CP. However, one should take care in extrapolating the twin data to singletons. It may be that genetic factors contribute to a n individual‘s susceptibility to adverse environmental conditions and this will no doubt be more important in multiple than in singleton births.
REFERENCES Adler E (1916): Familial cerebral palsy. J Chronic Dis 13:207-214. Alberman E (1964): Cerebral palsy in twins. Guy’s Hosp Rep 113:285-295. Austin CR, Edwards RC (1981):“Mechanisms of Sex Differentiation in Animals and Man.” New York: Academic Press. Bax MCO (1964):Terminology and classification of cerebral palsy. Dev Med Child Neurol 6:295-307. Bulmer MG (1970):“The Biology of Twinning in Man.” Oxford: Clarendon Press. Bundey S (1985): “Genetics and Neurology.” London: Churchill Livingstone, pp 251-253. Bundey S, Griffths MI (1977):Recurrence risks in families of children with symmetrical spasticity. Dev Med Child Neurol 19:179-191. Durkin MV, Kaveggia EG, Pendleton E, Neuhauser G, Opitz J M (1976): Analysis of etiologic factors in cerebral palsy with severe
mental retardation I. Analysis of gestational, parturitional and neonatal data. Eur J Pediatr 123:67-81. Ellis RF, Berger GB, Keith L, Depp R (1979): The Northwestern University multihospital twin study 11.Mortality of first versus second twins. Acta Genet Med Gemellol (Roma) 28:347-352. Falconer DS (1965): The inheritance of liability to certain diseases, estimated from the incidence among relatives. Ann Hum Genet 29:51-76. Freud S (1897): “Infantile Cerebral Paralysis” (translation by Russin LA, 1968, of Die Infantile Cerebrallahmungen. Wein: A. Holder). Florida: University of Miami Press. Griffiths M (1967): Cerebral palsy in multiple pregnancy. Dev Med Child Neurol 9:713-731. Gustavson KH, Hagberg B, Sanner G (1969): Identical syndromes of cerebral palsy in the same family. Acta Paediatr Scand 58: 330-340. Hrubec Z, Robinette D (1984): The study of human twins in medical research N Engl J Med 310:435-441. Illingworth RS, Woods G (1960): The incidence of twins in cerebral palsy and mental retardation. Arch Dis Child 35:333-335. Ingram TTS (1964):A study ofthe importance ofdevelopmentalmalformation amongst patients suffering from congenital cerebral palsy in the present series. In Ingram TTS (ed): “Paediatric Aspects of Cerehral Palsy.” London: E & S Livingstone, pp. 411-445. Little WJ (1862): On the influence of abnormal parturition, difficult labour, premature birth and asphyxia neonatorum on the mental and physical conditions of the child, especially in relation to deformities. Trans Obstet SOCLond 3:293. Mannino FL, Jones KL, Benirschke K (1977): Congenital skin defects and fetus papyraceus. J Pediatr 91:559-564. McCarthy BJ, Sachs BP, Layde PM, Burton A, Terry JS, Rochat R (1981): The epidemiology of neonatal death in twins. Am J Obstet Gynecol 141:252-265. Melnick M (1977):Brain damage in survivor after in-utero death of monozygous co-twin. Lancet 2:1287. Moore CM, McAdams AJ, Sutherland J (1969): Intrauterine disseminated intravascular coagulation: A syndrome of multiple pregnancy with a dead twin fetus. J Pediatr 74523-528. Moore DJ (1988):“PerinatalStatistics in W.A. 4th Annual Report of the Western Australian Midwives Notification System 1986.” Perth: Public Health Department of Western Australia. Polani PE (1984): Preface. In Stanley FJ, Alberman E (eds): “The Epidemiology of the Cerebral Palsies.” London: Spastics International Medical Publications, pp ix-x. Russell EM (1960): Cerebral palsied twins. Arch Dis Child 36: 328-336. Stanley FJ (1979): An epidemiological study of cerebral palsy in Western Australia 1956-1975. I: Changes in total incidence of cerebral palsy and associated factors. Dev Med Child Neurol 21:701-713. Stanley FJ (1981): Improved male neonatal outcome in Western Australia. Early Hum Dev 5:179-185. Stanley F (1984): Prenatal risk factors in the study of the cerebral palsies. In Stanley F, Alberman E (eds): “The Epidemiology of the Cerebral Palsies.” Oxford: Spastics International Medical Publications, pp 69-86. Yoshioka H, Kadomoto Y, Mino M, Morikawa Y, Kasabuchi Y, Kusunoki T (1979): Multicystic encephalomalacia in liveborn twin with a stillborn macerated co-twin. J Pediatr 95:798-800.
Cerebral Palsy in Muliple Births APPENDIX A. Reference: Waddell V (1989): Perinatal Statistics in Western Australia (UnFkb)Midwives Files 1981-1986* No. of Twin Pairs Year of birth -~ 1981 1982 1983 1984 1985 1986
Like sex 169 178 167 167 184 199 1.064
Unlike sex 66 53 69 79 81 66 414
1 Sex Unknown
1
*These unpublished data may be cited in B. Petterson’s paper. Dr. Vivienne Waddell, Health Department of Western Australia, Epidemiology Branch.
351
Cerebral Palsy in Multiple Births in Western Australia: Genetic Aspects Beverly Petterson, Fiona Stanley, and Diane Henderson Department of Anatomy and H u m a n Biology, University of Western Australia (B.P.) and National Health and Medical Research Council Unit in Preventive Medicine and Epidemiology (F.S.), Western Australia, Australia; Los Angeles Children$ Hospital, Los Angeles (D.H.)
A study of cerebral palsy in multiple births was undertaken to test genetic involvement and assess the impact of the special conditions of pregnancy and parturition in these cases. Complete ascertainment of cerebral palsy in multiple gestations that occurred in Western Australia between 1956 and 1985 was obtained from the Western Australian Cerebral Palsy Register. There were 74 twins and 5 triplets. Data on sex, birth order, motor handicap, outcome in co-twins and triplets, zygosity, and pedigree information was obtained from the Register, hospital records, and, where possible, by interview of the parent(s1 of the propositi. There was a significantly higher (P = 0.0026) concordance rate in MZ than in DZ twin pairs. However, pedigree studies showed no other relatives with a motor handicap similar to that of the propositi. This is consistent with a multifactorial cause in at least some of the cases. The sex ratio of affected twins was found to be 2.1 compared to 1.3 for singletons and all 5 affected triplets were boys. The trend of increasing sex ratio with increasing plurality was significant at the 1%level.
tance of a variety of pre- and perinatal conditions and the degree to which genetic factors are involved have been debated since the time of Little [1862] and Freud [18971. Previous work has suggested that the recurrence risk for CP among sibs of unselected cases is low, about 1-2% [Ingram, 1964; Bundey, 1985; Gustavson et al., 19691. However, this is 5 times the prevalence in all births. In selected CP cases in which there are no perinatal risk factors, the recurrence risk for sibs is often much higher. It has been reported as 1 in 9 where the motor handicap is spastic and symmetrical [Bundey and Griffiths, 19771, and calculated to be 1 in 8 in ataxia with mental retardation [Gustavson et al., 19691. Genetic factors may contribute to this increased risk factor either in the form of single gene defects [Bundey, 1985; Adler, 19611, by chromosome microdeletions or rearrangements, or by giving rise to a constitutional vulnerability to other risk factors in a multifactorial cause [Polani, 19841. By using a polygenic model [Falconer, 19651, a heritability of approximately 40% can be calculated from the population birth prevalence relative to that in sibs of propositi. If genetic factors are significant one might expect to find a higher concordance rate in MZ compared to DZ twins. The occurrence of CP in multiple births affords an opportunity to assess the impact of the special conditions of pregnancy and parturition in these cases and also to test the hypothesis of genetic involvement. With these 2 aims, a descriptive study of all cases of CP in multiple births in Western Australia between KEY WORDS: motor handicap, twins, concordance rate, sex ratio 1956 and 1985 was undertaken. All cases born during the study period were ascertained and data on the preand perinatal courses of each individual, as well as their family histories, were collected. We also attempted to INTRODUCTION establish the zygosity and outcome in co-sibs for all There is little doubt that cases of cerebral palsy (CP) multiple gestations in the study group. are etiologically heterogeneous. The relative imporThis report presents part of this descriptive study and is primarily concerned with a n assessment of the degree to which genetic factors are involved in the cause of the Received for publication October 16, 1989; revision received group of disorders which come under the heading of CP. March 29, 1990. Subsequent reports will address the obstetric and periAddress reprint requests to Beverly Petterson, Department of natal risk factors and the contribution of low birth Anatomy and Human Biology, University of Western Australia, weight. Nedlands Western Australia 6009.
0 1990 Wiley-Liss, Inc.
Cerebral Palsy in Muliple Births
DEFINITIONS Cerebralpalsy (CP)is defined as a group of disorders of movement and posture due to a defect or lesion of the immature brain [Bax, 19641. It excludes progressive neurological disorders and those occurring after a documented postnatal insult. Live birth is the complete expulsion or extraction from its mother of a product of conception, irrespective of duration of pregnancy, which after separation shows signs of life. Neonatal death is the death of a liveborn infant within 28 days of birth. Stillbirth (SB) is the complete expulsion or extraction from its mother of a product of conception of at least 20 weeks gestation or 400 g birthweight which, after separation, did not show any sign of life. Perinatal death includes stillbirths and neonatal deaths.
METHODS Data Collection Case ascertainment was from the Western Australian Cerebral Palsy Register, a complete population-based collection of all diagnosed individuals with CP obtained by using multiple sources of ascertainment [Stanley, 19791. Data used for calculating sex ratios for CP in singletons and numerators for birth prevalence of CP in twins and triplets were also available from the register, as were the sex, birthweight, gestational age, and updated diagnoses of motor handicap of cases. Total annual numbers of multiple births in Western Australia were obtained from the Western Australian Midwives Notification of Births System [Moore, 19881. Outcome in co-twins and triplets and details of birth order were obtained from hospital records of birth or interview with parent(s1 of the proposita. These data were collected by 2 of the authors (B.P., D.H.) using standard data collection and parent interview schedules. Evidence used for assigning monozygosity in like-sex twin pairs included polymorphism coincidence, available from medical records for one pair, and birth records of monochorionic placentation for 3 pairs. For the rest we relied on parents’ and doctors’ assessment of the physical likeness of the twin pairs. We asked specifically about eye color, hair type and color, and facial features, and generally the difficulty of telling the twins apart. This method has been shown to be 95% accurate for normal twin pairs [Hrubec and Robinette, 19841. However, in this study the motor handicap may have affected facial features or body form in one twin of a discordant pair, which may have resulted in a n over-estimation of dizygosity in this category. Because of the vast distances in Western Australia, personal interview of all parents was not possible. Therefore, a structured telephone interview was used to obtain pedigree information. The questionnaire was developed with the help of the Australian Bureau of Statistics. Specific questions on the incidence of motor handicap, epilepsy, intellectual handicap, and congenital malformations and birth losses in parents, grandparents, aunts, uncles, and cousins of the proposita were
347
included. The technique was pilot-tested and results compared well with pedigrees which were available from medical records. The pedigree interview is available from the primary author (B.P.) on request.
RESULTS A total of 79 CP individuals from multiple births in Western Australia in the 30 year period from 1956 was found. There were 74 twins and 5 triplets. Comparisons of birth prevalence of CP in singletons, twins, and triplets in Western Australia are shown in Table I. Both twins and triplets show a marked and significant increase in relative risk compared to singletons. Mean birthweights and gestational ages are also significantly lower in both twins and triplets than in singletons. The outcome in all co-triplets and all but 2 co-twins was established and the results are shown in Tables I1 and 111. Among the twins there were 6 pairs concordant for CP, 24 pairs in which the co-twin died either pre- or perinatally, and 36 pairs in which the co-twin showed no major malformation or neurological impairment. Of the 10 co-triplets there were 3 SB, 1 termination, and 6 normal individuals. There were no sets of triplets showing concordance for CP. Medical records of birth gave no indication of zygosity in these cases. Of the 3 sets in which there were 2 living children of like sex, 2 were definitely not identical and we were unable to contact the other family (Table 111,family 3). It is interesting to note that all triplet sets were born between 1980 and 1985. Only one set was conceived spontaneously. Two sets were the result of hormonal stimulation of ovulation and 2 were the result of in vitro fertilization. Of the affected twins, 2 were conceived after hormonal stimulation of ovulation and one after i n vitro fertilization. The birth outcome in relation to birth order within twin pairs is shown in Table IV. There was no significant difference between the first- and second-born twins in relation to numbers of individuals with CP or those who died perinatally. Pedigree information was obtained from 59 of the 74 families in the study, including all 6 twin pairs concordant for CP. This showed no obvious patterns of monogenic inheritance, although we may have missed autosoma1 recessive cases. The relative rates of concordance and discordance in MZ and DZ twin pairs are shown in Table V. Discordant in this context means twin pairs in which there was one CP and one normal twin. Any pairs in which there was a pre- or perinatal death were eliminated from this analysis as were several pairs in which it was not possible to establish zygosity. The excess of concordance above expected values is highly significant ($ = 7.4, P = 0.0026). The motor handicap in each concordant pair, listed in Table VI, was very similar. Medical records showed no clear identification of syndromes. Birthweights and gestational ages in each concordant pair are also listed. The mean birthweights and gestational ages for all affected twins, and for MZ and DZ twins separately, are included for comparison. The concordant MZ twins are not significantly different in these respects from all affected MZ twins.
348
Petterson et al. TABLE I. Birth Prevalence of CP by Plurality (Western Australia 1966-1985)*
N
Plurality Singleton
877
Twins
50
Triplets
5
Birth prevalence per 1,000 livebirths 2.4 6.3 32
(95% CI) (2.22-2.54) (4.51-8.03) (3.7-59.5)
GA mean SEM range 37.5 0.21 25-46 35.1 0.53 25-42 28.6 1.50 25-34
BW mean SEM range 2,778 45.5 565-4,825 2,050 83.5 765-4,048 1,249 176.2 740-1,822
* GA, gestational age in weeks; BW, birthweight in grams; CI, confidence interval; SEM, standard error of mean.
TABLE 11. Outcome in Twin Pairs Ascertained for CP (Western Australia 1956-1985) CPICP CPISB CPlneonatal death CPlnormal CPlunknown
6 pairs 15 pairs 9 pairs 36 pairs 2 Dairs
TABLE 111. Outcome in Triplets Ascertained for CP (Western Australia 1956-1985) Spontaneous
1.
CP
2'
Termination
3.
CP
4.
0
5.
0
If. 2SB3
CP
Hormonal stimulation of ovulation
stimulation , & 0 Hormonal % of ovulation
0
.
CP
B
I n uitro fertilization
O I n vitro fertilization
CP
TABLE IV. Outcome in First- and Second-Born Twins in Twin Pairs Ascertained for CP (Western Australia 1956-1985) Twin I 35 9 44
CP SB or neonatal death Total
Twin I1 39 15 54
TABLE V. Presumed Zygosity of Surviving Twin Pairs Ascertained for CP (Western Australia 1956-1985)* Concordant Monozygotic Dizveotic * x : = 7.4; P =
6 0
~
0.0026.
Discordant 9 21
The outcome in twin pairs in which the co-twin died either pre- or perinatally is shown in Table VII. Hospital records contained little evidence of zygosity in these cases and the numbers of like and unlike sex pairs, 15 and 5, respectively, are not significantly different from the ratio of 2.6:l for all twin pairs born in Western Australia from 1981 to 1986 (V. Waddell, unpublished, 1989). The outcome by sex in all multiple gestations in the study group is shown in Table VIII. Table IX shows the ratio of maledfemales in singletons, twins, and triplets, The excess ofmales in both twins and singletons with CP compared with the expected proportion in the whole population, corrected for gestational age distribution, is highly significant. The trend of a n increasing sex ratio with increasing plurality is significant a t the 1%level. (xl for trend = 6.83).
DISCUSSION This study attempted to assess the contribution of genetic factors to the cause of CP in all twins and triplets with CP born in Western Australia between 1956 and 1985. Pedigree studies showed no obvious cases of monogenic inheritance. However, a marked excess of concordance among MZ twin pairs was found, which is consistent with a genetic predisposition. The use of the Western Australian Cerebral Palsy Register ensured complete ascertainment of CP in multiple births throughout the State [Stanley, 19791. This avoided biases imposed by using either hospital-based selection, where the milder cases are often missed, or a twin registry where twins with handicaps and especially twin pairs in which one twin has died pre- or perinatally are significantly underrepresented. Our data show a greater birth prevalence for CP in both twins and triplets than in singletons, a phenomenon reported in several studies. This has been partly attributed to the difference in birthweight distribution between multiple and singleton births and partly to the added complications of pregnancy and delivery in multiple births [Alberman, 1964; Griffiths, 1967; Russell, 1960; Illingworth and Woods, 1960; Durkin et al., 19761. The idea that multiple conception and pregnancy itself may be a risk for CP is less popular [Stanley, 19841.
Cerebral Palsy in Muliple Births
349
TABLE VI. Motor Handicap Birthweight and Gestational Age in Twin Pairs Concordant for CP (Western Australia 1956-1985) Pair No. 1. ~
Pair order 1 2
Sex d d
. . . . .-
2.
d d
3.
d d d d
4.
1 2 1 2 1 2 1 2 1 2
P
5.
P 6.
.-Motor handicap Spastic diplegia Spastic diplegia-some ataxia Spastic diplegia Spastic diplegia Ataxia, spastic quadriplegia Ataxia, spastic quadriplegia" Ataxia, mixed spastic dyskinetic Ataxia, mixed spastic dyskinetic Spastic quadriplegia Spastic quadriplegiab Spastic quadriplegia Spastic quadriplegia
d d
Mean Mean Mean Mean
(SEMI for (SEMI for @EM) for (SEMI for
Severity Moderate Moderate
35
Mild Mild
32
Mild Moderate
34
Severe Severe
40
Severe Severe
36
Mild Mild
29
concordant MZ CP twins all C P twins MZ CP twins DZ C P twins
.
~~~~
GA
BW 2,260 2,130 1,814 1,758 2,098 1,985 2,098 3,090 2,353 1,871 1,445 1,570
34.3 (1.52) 35.1 (0.53) 33.9 (0.64) 36.3 (0.82)
~~~
~~
2,039 (123.1) 2,050 (83.5) 1,961 (97.5) 2,144 (141.9)
"Epileptic, died a t 3 years from respiratory arrest. Died at 9 years from pneumonia.
TABLE VII. Perinatal Deaths in Like- and Unlike-Sex Twin Pairs, Ascertained for C P (Western Australia 1956-1985) Co-twin
Neonatal
Stillbirth
(Macerate& uawraceous)
Like sex Unlike sex Unknown sex Total
6 3
9 2 4
(1) (1) (1)
TABLE VIII. Outcome by Sex in Multiple Gestations Ascertained for CP (Western Australia 1956-1985) Twins.~ Outcome_-
Sex
d
P
?
d
Triplets" P
50 13 15
24 7 21
0 4 2b
5 1 2
0 1 4
?
~
CP Perinatal death Normal
0 1 0
1 triplet terminated. 'Outcome unknown assumed normal. a
Differences in CP and mortality rates between firstand second-born twins have also been reported [Ellis et al., 1979; Russell, 19601. However, our data, in common with other studies [Durkin et al., 1981; McCarthy et al., 19811, show no significant effect of birth order on outcome. Co-twins and triplets of CP cases show a high morbidity and mortality rate during the pre- and perinatal periods. Of the 68 twin pairs in which CP was ascertained there were only 36 (53%)where the co-twin was normal. These findings are similar to those reported by other investigators [Alberman, 1964; Griffiths, 1967; Russell, 19601.
Total
15 5 4 24
The proportion of twin pairs concordant for CP was 6/68 (8.8%), which does not differ significantly from the studies of Alberman 2/45 (4.4%)[Alberman, 19641 and Griffiths 4/78, (5.1%)[Griffiths, 19671. All 6 concordant pairs were MZ. There were 9 MZ and 15 DZ discordant pairs. The excess of concordant MZ pairs above expected values is highly statistically significant. The motor handicap in each concordant pair was very similar, but pedigrees of these families showed no other relatives with similar motor handicaps and there was no clear identification of syndromes. This, of course, does not eliminate the possibility of autosomal recessive inheritance in some cases. It is also possible that the combination of environmental stress with a susceptible genotype caused the motor handicap. However, there is another possible contributing factor to this excess of concordant MZ twin pairs. This is related to the observation that anastomoses may exist between placental circulations in MZ twins. This mechanism has certainly been implicated in cases of CP in a n individual with a papyraceous or macerated co-twin [Durkin et al., 1976; Melnick, 1977; Moore et al., 1969; Yoshioka et al., 19791. However, the similarity of motor handicap in each concordant twin pair in this study may suggest a less random causal process. In addition, there
350
Petterson et al. TABLE IX. Sex Ratios in CP bv Pluralitv (Western Australia 1956-1985) ~~
~
Singletons (720 males, 554 females) Twins (50 males, 24 females) Triplets (5 males, 0 females)
~
Proportion male
(95%CI)
Male/female
Expected“ malelfemale
0.56
(0.537-0.593)
1.3
1.1
0.68
(0.557-0.780)
2.1
1.0
1.o
(0.48- 1.00)
1.2
“Expected sex ratios according to GA distribution, calculated by using sex ratios of all liveborn singletons and multiples by GA in Western Australia 1985-1987.
was no evidence of a n increase in monozygosity among the 24 twin pairs in which the co-twin died either pre- or perinatally, a situation which is thought more likely to be due to vascular problems of this nature [Melnick, 1977; Mannino et al., 19771. One other observation which emerged from the study was a n increasing sex ratio with increasing plurality. It is interesting to relate this trend to the increasing birth prevalence of CP with plurality. Multiple births overall have lower sex ratios than singletons [Bulmer, 1970; Moore, 19881.Sex ratios for both singleton and multiple births are marginally higher in the Western Australian population for preterm than for term births [Moore, 19881. However, a comparison of expected sex ratios, according to the gestational age distribution of both the singleton and multiple CP populations, shows that low gestational age is not the only factor contributing to the male excess. It would seem that males fare worse than females under the adverse environmental conditions of increasing plurality. That changing environmental conditions can differentially affect males and females was also shown by the greater fall in mortality and CP rates for males than for females in Western Australia during the 1970’s [Stanley, 19811. Sex ratios in other animals also tend to change with changes in the environment [Austin and Edwards, 19811. In conclusion the data presented support the hypothesis that genetic factors play a part in the cause of CP. However, one should take care in extrapolating the twin data to singletons. It may be that genetic factors contribute to a n individual‘s susceptibility to adverse environmental conditions and this will no doubt be more important in multiple than in singleton births.
REFERENCES Adler E (1916): Familial cerebral palsy. J Chronic Dis 13:207-214. Alberman E (1964): Cerebral palsy in twins. Guy’s Hosp Rep 113:285-295. Austin CR, Edwards RC (1981):“Mechanisms of Sex Differentiation in Animals and Man.” New York: Academic Press. Bax MCO (1964):Terminology and classification of cerebral palsy. Dev Med Child Neurol 6:295-307. Bulmer MG (1970):“The Biology of Twinning in Man.” Oxford: Clarendon Press. Bundey S (1985): “Genetics and Neurology.” London: Churchill Livingstone, pp 251-253. Bundey S, Griffths MI (1977):Recurrence risks in families of children with symmetrical spasticity. Dev Med Child Neurol 19:179-191. Durkin MV, Kaveggia EG, Pendleton E, Neuhauser G, Opitz J M (1976): Analysis of etiologic factors in cerebral palsy with severe
mental retardation I. Analysis of gestational, parturitional and neonatal data. Eur J Pediatr 123:67-81. Ellis RF, Berger GB, Keith L, Depp R (1979): The Northwestern University multihospital twin study 11.Mortality of first versus second twins. Acta Genet Med Gemellol (Roma) 28:347-352. Falconer DS (1965): The inheritance of liability to certain diseases, estimated from the incidence among relatives. Ann Hum Genet 29:51-76. Freud S (1897): “Infantile Cerebral Paralysis” (translation by Russin LA, 1968, of Die Infantile Cerebrallahmungen. Wein: A. Holder). Florida: University of Miami Press. Griffiths M (1967): Cerebral palsy in multiple pregnancy. Dev Med Child Neurol 9:713-731. Gustavson KH, Hagberg B, Sanner G (1969): Identical syndromes of cerebral palsy in the same family. Acta Paediatr Scand 58: 330-340. Hrubec Z, Robinette D (1984): The study of human twins in medical research N Engl J Med 310:435-441. Illingworth RS, Woods G (1960): The incidence of twins in cerebral palsy and mental retardation. Arch Dis Child 35:333-335. Ingram TTS (1964):A study ofthe importance ofdevelopmentalmalformation amongst patients suffering from congenital cerebral palsy in the present series. In Ingram TTS (ed): “Paediatric Aspects of Cerehral Palsy.” London: E & S Livingstone, pp. 411-445. Little WJ (1862): On the influence of abnormal parturition, difficult labour, premature birth and asphyxia neonatorum on the mental and physical conditions of the child, especially in relation to deformities. Trans Obstet SOCLond 3:293. Mannino FL, Jones KL, Benirschke K (1977): Congenital skin defects and fetus papyraceus. J Pediatr 91:559-564. McCarthy BJ, Sachs BP, Layde PM, Burton A, Terry JS, Rochat R (1981): The epidemiology of neonatal death in twins. Am J Obstet Gynecol 141:252-265. Melnick M (1977):Brain damage in survivor after in-utero death of monozygous co-twin. Lancet 2:1287. Moore CM, McAdams AJ, Sutherland J (1969): Intrauterine disseminated intravascular coagulation: A syndrome of multiple pregnancy with a dead twin fetus. J Pediatr 74523-528. Moore DJ (1988):“PerinatalStatistics in W.A. 4th Annual Report of the Western Australian Midwives Notification System 1986.” Perth: Public Health Department of Western Australia. Polani PE (1984): Preface. In Stanley FJ, Alberman E (eds): “The Epidemiology of the Cerebral Palsies.” London: Spastics International Medical Publications, pp ix-x. Russell EM (1960): Cerebral palsied twins. Arch Dis Child 36: 328-336. Stanley FJ (1979): An epidemiological study of cerebral palsy in Western Australia 1956-1975. I: Changes in total incidence of cerebral palsy and associated factors. Dev Med Child Neurol 21:701-713. Stanley FJ (1981): Improved male neonatal outcome in Western Australia. Early Hum Dev 5:179-185. Stanley F (1984): Prenatal risk factors in the study of the cerebral palsies. In Stanley F, Alberman E (eds): “The Epidemiology of the Cerebral Palsies.” Oxford: Spastics International Medical Publications, pp 69-86. Yoshioka H, Kadomoto Y, Mino M, Morikawa Y, Kasabuchi Y, Kusunoki T (1979): Multicystic encephalomalacia in liveborn twin with a stillborn macerated co-twin. J Pediatr 95:798-800.
Cerebral Palsy in Muliple Births APPENDIX A. Reference: Waddell V (1989): Perinatal Statistics in Western Australia (UnFkb)Midwives Files 1981-1986* No. of Twin Pairs Year of birth -~ 1981 1982 1983 1984 1985 1986
Like sex 169 178 167 167 184 199 1.064
Unlike sex 66 53 69 79 81 66 414
1 Sex Unknown
1
*These unpublished data may be cited in B. Petterson’s paper. Dr. Vivienne Waddell, Health Department of Western Australia, Epidemiology Branch.
351
