Parental age and birth order in Chinese children with congenital heart disease

To order reprints of this article go to:
http://journals.bmj.com/subscriptions/
To subscribe to Journal of Medical Genetics go to:
Downloaded from hwmaint.jmg.bmj.com on 30 January 2010
Journal of Medical Genetics, 1982, 19, 441-443
Parental age and birth order in Chinese children with
congenital heart disease
JOHN S H TAY, WILLIAM C L YIP, AND R JOSEPH
From the Department ofPaediatrics, National University of Singapore, Singapore.
SUMMARY Parental age and birth order were studied in 100 Chinese children with congenital
heart disease (proven by cardiac catheterisation) and in 100 controls. A higher incidence of congenital
heart disease was present in the children with higher birth orders. No relationship was
found between the incidence and the paternal or maternal ages. Using the method of multiple
regression analysis this birth order effect was significant (p <0 01) and independent of parental age.
This finding provides indirect evidence of environmental influence in the causation of congenital
heart disease, which is known to be inherited in a multifactorial manner. Family planning to limit
the size of the family may possibly contribute to the reduction of the incidence of congenital heart
disease.
Parental age and birth order effects have been
demonstrated in a number of genetic diseases and
congenital malformations. Certain chromosomal
disorders such as trisomy 21 (Down's syndrome),
trisomy 13 (Patau's syndrome), trisomy 18 (Edwards's
syndrome), and Klinefelter's syndrome have been
shown to be related to maternal age.1 Some autosomal
dominant disorders such as Apert's syndrome,
Marfan's syndrome, and achondroplasia are related
to paternal age.2 A number of congenital malformations
have been shown to be related to birth order.3
Congenital dislocation of the hip and congenital
pyloric stenosis occur more frequently in first born
children, and anencephaly and spina bifida are
more common among first born children as well
as in high birth orders.4
The rationale of studying parental age and birth
order effects is that the results of such studies may
shed some light on pathogenesis. The presence of a
parental age or birth order effect is presumptive
evidence of an environmental influence in the
causation of the disease.' This kind of information
may also be of help in genetic counselling.
The purpose of this study was to investigate the
parental age and birth order effects in Chinese
children with congenital heart disease.
Materials and methods
The material consisted of 100 consecutive Chinese
children with congenital heart disease proven by
Raceived for publication 9 April 1982.
441
cardiac catheterisation. The parental ages (at the
birth of the child) and the birth order of the children
with congenital heart disease were noted. The
parental age and birth order of 100 consecutive
normal Chinese children born in Singapore during
the same period were similarly recorded.
To distinguish the separate effects of paternal age,
maternal age, and birth order, which are usually well
correlated with one another, the method of Smith,5
using multiple regression analysis, was used.
A computer programme was written specially for
the Apple II microcomputer to compute all the
required calculations.
The method and terminology of Smith5 was
followed closely. The essence of this method is that,
given the matrix of the coefficients of regression
(for the controls),
rmm Pmf Pmb
| 1fm Pff Pfb
-Pbm Pbf 13bb
(where 3mf is the coefficient of regression of mother's
age on father's age and so on), it is possible to test
the hypothesis 'that only the father's age has any
appreciable real influence on the condition, the rises
in the mother's age and birth rank being statistical
consequences of this' (as well as similar hypotheses
involving the mother's age and the birth order).5 The
variance-covariance matrix is given by
rt1nrmnImf ltnib
I "'fill 'ff 'Ifh
i bm t/bf t'bb I
Downloaded from hwmaint.jmg.bmj.com on 30 January 2010
442
(where Vmm denotes the variance of m, the maternal
age, in the sample, Umf, the covariance of m and f,
and so on).
The expected rise in the mother's age (and
similarly for the father's age and the birth order) is
3mf Af, where Am is the observed rise in the mother's
age, and the difference between observed and
expected is Am-3mf Af. The error variance of An,
is vmm/n, where n is the number of observations in the
sample, the error variance of Af is uff/n, and their
covariance is /mf /n. Hence, the error variance of
Am-3mf Af iS given by
("imnm 2-2mf",mf 13 Imf!lff)/n.
Smith's method also allows estimation of the
expected independent changes (dm, df, db) in the
variables m, f, and b. The expected increase in the
mean maternal age m owing to dm, df, db is dm+ Pmf
df+rmfdb. By setting this equal to the observed
increase Am, we have
dnm+pmfdf+P3mbdb - Am.
In all, there are three linear equations for the
three unknowns dmi di., and db, which are written
in matrix form ,d- A where d is the column vector
with elements [dmi df, db]T and A similarly is given
by [Am, Af, Ab]T. This has the solution d -1A.
The error variance matrix is given by var d -IV
(I1)T/n.
The values of dm, df, and db are compared with the
size of their respective standard errors to determine
whether there are independent paternal age, maternal
age,'or birth order effects.
Results
The parental ages and birth order of the 100 Chinese
children with congenital heart disease and the 100
controls are shown in table and the correlation
coefficients listed in table 2. The birth order of the
children with congenital heart disease was significantly
higher than the controls (p<0.01), but there
was no significant difference in the paternal and
maternal ages of the patients and controls (p >0 05).
As expected, the parental ages and birth order were
TABLE 1 Parental age and birth order in congenital
heart disease.
Father's AMother's Birth
age (yr) age (yr) order
Congenital heart
disease (n = 100) Mean 31*51 27-66 2.78
SD 6-10 5.41 2-24
Controls (n = 100) Mean 30-37 26-55 2-09
SD 4.52 3.83 1-01
Observed rise in parental age
or birth order 1.14 1.11 0.69
Significance of difference p>0*05 p>0.05 p<0-01
JSHTay, WCL Yip, andR Joseph
TABLE 2 Correlation coefficients.
Congenital Controls
heart disease
Father's age and mother's age 0.7219 0.5809
Father's age and birth order 0.5612 0-4588
Mother's age and birth order 0.4789 0-4696
all well correlated with one another for both groups
of children, with correlation coefficients ranging
from 0 46 to 0 72 (p <0 0001).
The matrix for the regression coefficients (controls)
are shown below:
0000
p = 0.6863
0-1234
0 4916
1*0000
0-1020
1-7871]
2.0628
I.0000 .
The variance-covariance matrix for the children
with congenital heart disease is as follows:
r29 - 2368
v - 23 8014
5*- 8032
23-8014
37- 1817
7-6689
5.8032]
7-6689
5.0218 .
The error variance matrix is given by
r 04195 -0 1029 -0.0760-
vard- -0 1029 0.4188 --0.0451
-0-0760 -- 0 0451 0.0700].
The values of dm,, df, and db, their standard errors,
and the significance of the parental age and birth
order effects are shown in table 3. There was no
significant paternal or maternal age effect (p>0.05)
but the birth order effect was significant (p<0-01).
The various types of congenital heart disease are
shown in table 4.
TABLE 3 Significance ofparental age and birth order
eftects.
Factor acting independently Standard p value
error (SE)
Maternal age- 0.020 (0.03 x SE) 0.648 p>0.05
Paternal age-O0 348 (0.54 x SE) 0*647 p>0.05
Birth order 0 * 730 (2*75 x SE) 0.265 p<0O01
TABLE 4 Types of congenital heart disease.
Diagnosis Number
Ventricular septal defect 22
Fallot's tetralogy 19
Pulmonary stenosis+ventricular or atrial septal defect 11
Persistent ductus arteriosus 8
Transposition of great vessels ±other lesions 8
Endocardial cushion defect 7
Double outlet right ventricle 6
Pulmonary atresia ±ventricular septal defect 5
Pulmonary valvular stenosis 3
Coarctation of aorta + other lesions 3
Others (including Ebstein's anomaly, aorto-pulmonary
window, persistent truncus arteriosus, total
anomalous pulmonary venous drainage, and cor
triatriatum) 8
Total 100
Downloaded from hwmaint.jmg.bmj.com on 30 January 2010
Parental age and birth order in Chinese children with congenital heart disease
Discussion
This is the first study, as far as we are aware, of the
parental age and birth order effects in Chinese
children with congenital heart disease. A moderate
birth order effect is demonstrated.
Nora6 and Nora et a17 demonstrated a multifactorial
inheritance for congenital heart disease.
They showed that the risk of recurrence of a congenital
heart defect in another child in a family
varies with the frequency the lesion is found in the
general population and the type of lesion in the first
affected child, the recurrence risk ranging from about
1 to 4 %. Applying Falconer's model8 for the
estimation of heritability to their data, it can be
shown that the heritability of various types of
congenital heart disease is in the order of 50 to 80 %.
This measure can give some indication of the
relative importance of heredity and environment
in the causation of the disease.
The presence of a birth order effect in congenital
heart disease is indirect evidence of the presence of
an environmental influence in the causation of the
disease and is thus consistent with what is known
about the multifactorial inheritance of congenital
heart disease. While the exact mechanism of the
birth order effect is at present unclear, this finding
may nevertheless have practical applications. The
increase in the incidence of congenital heart disease
with the higher birth orders would suggest that
family planning to reduce the size of families may
possibly contribute towards the lowering of the
incidence of congenital heart disease.
References
Emery AEH. Methodology in medical genetics. An introduction
to statistical methods. Edinburgh, London, New
York: Churchill Livingstone, 1976.
2 Jones KL, Smith DW, Harvey MAS, Hall BD, Quan L.
Older paternal age and fresh gene mutation: data on
additional disorders. J Pediatr 1975;86:84-8.
3 Carter CO. The inheritance of common congenital
malformations. Prog Med Genet 1965 ;4:59-84.
4 Fedrick J. Anencephalus: variation with maternal age,
parity, social class and region in England, Scotland and
Wales. Ann Hum Genet 1970;34:31-8.
5 Smith CAB. Note on the estimation of parental ageeffects.
Ann Hum Genet 1972;35:337-42.
6 Nora JJ. Multifactorial inheritance hypothesis for the
etiology of congenital heart disease. Circulation 1968;38:
604.
7 Nora JJ, Vargo TA, Nora AH, et al. Dexamphetamine:
a possible environmental trigger in cardiovascular malformations.
Lancet 1970;i :1290.
Falconer DS. The inheritance of liability to certain
diseases, estimated from the incidence among relatives.
Ann Hum Genet 1965;29:51-76.
Requests for reprints to Professor John Tay,
University Department of Paediatrics, Singapore
General Hospital, Outram Road, Singapore 0316.