close

Вход

Забыли?

вход по аккаунту

?

Prenatal and perinatal antecedents of febrile seizures.

код для вставкиСкачать
Prenatal and Perinatal
Antecedents of Febrde Seizures
Karin B. Nelson, MD, and Jonas H. Ellenberg, PhD
We examined prenatal and perinatal characteristics as possible ri& factors for febrile seizures in a large pediatric
population. Family history was among the few identified factors that made an important contribution to vulnerability
to febrile seizures; however, no more than 6% of febrile seizures could be attributed to a characteristic of family
history. Maternal illness, smoking history, and a few rare neonatal characteristics were associated with increases in
risk. No complication of labor or delivery was an important risk factor for febrile seizures.
Nelson KE3, Ellenberg JH. Prenatal and perinatal antecedents of febrile seizures.
Ann Neurol 1990;27:127-131
Two to five percent of children experience at least one
febrile seizure in the first 5 years of life El]. Several
earlier reports E2-41, some of them on highly selected
populations, some without controls or incompletely
described, suggested that abnormalities in perinatal
histories were relatively frequent in children with febrile seizures. Tsuboi IS] reported an excess of
difficult birth, postnatal asphyxia, and coiling of the
umbilical cord in a cohort of children with febrile seizures, based apparently on retrospectively reported
history.
In two methodologically appropriate studies, van
den Berg and Yerushalmy 161 reported an excess of
early fetal deaths in the pregnancy preceding the births
of children with seizures, febrile or nonfebrile, and
investigators in the British national cohort study [7]
found that of the pregnancy-related antecedents examined, the only one of even marginal statistical
significance as a risk factor for febrile seizures was
breech delivery.
In the hope of identifying factors susceptible to
preventive intervention, we investigated prenatal and
perinatal factors as antecedents of febrile seizures in a
large prospective study of a population unselected for
pediatric disorders. The natural history of febrile seizures in this population has previously been described
c1,8,91.
Methods
Febrile seizures were defined as in the 1980 National Institutes of Health consensus development statement on febrile
seizures [lo]: “A febrile seizure is an event in infancy or
childhood, usually occurring between three months and five
years of age, associated with fever but without evidence of
intracranial infection or defined cause. Seizures with fever in
children who have suffered a previous nonfebrile seizure
[including neonatal seizure] are excluded.”
The data for this investigation are from the Collaborative
Perinatal Project of the National Institute of Neurological
Disorders and Stroke. Information on sampling, study design, definitions, and procedures has been reported 111-1 31.
Approximately 400 characteristics of family history, maternal health, pregnancy, labor and delivery, and the nursery
period were evaluated for their association with febrile seizures. The history was given by the mother. These characteristics were previously examined as-possible risk factors for
cerebral palsy or nonfebrile seizure disorders 112, 131.
The statistical test used to evaluate fourfold contingency
tables of dichotomous antecedents for febrile seizures was
the chi-square test with 1 degree of freedom (using a continuity correction), except where otherwise indicated. Fisher’s exact test was used when cell size warranted it. Fourfold
comparisons of subgroups within larger contingency tables
were made only when the chi-square test for the larger table
was significant at the 0.05 level. For evaluation of possible
linear trends in the risk of febrile seizures for polychotomous
variables (e.g., number of maternal hospitalizations, cigarettes smoked per day) we used weighted least-squares regression analysis of the observed probabilities of risk, with
weights determined by the number of patients at each observation point.
Given the many chi-square tests made within this large
data set, some were likely to achieve nominal levels of statistical significance by chance alone. We did not adjust
significance levels using multiple comparison techniques,
which would lessen the chance of observing some clinically
important risk factors 1141. Rather, we chose to consider the
significance values simply as indicators for risk factors that
are candidates for confirmation in future studies. However,
~~
From the Neuroepidemiology Branch and Biometry and Field Studies Branch, National Institute of Neurological Disorders and Stroke,
Bethesda, MD.
Address correspondence to Dr Nelson, 7 5 5 0 Wisconsin Ave, Room
700, Bethesda, MD 20892.
Received Dec 7 , 1988, and in revised form Mar 13, May 9, and Jun
13, 1989. Accepted for publication Jun 19, 1989.
127
Table I. Prepregnancy and Pregnancy Risk Factors for Febrile Seizures
Factor
Seizure(s), mother
Any
Febrile
Remote nonfebrile
Current nonfebrile
Seizure medication during pregnancy
Phenytoin
Phenytoin and phenobarbital
Seizure(s), other family member
Seizure, father, any
Seizure(s) prior-born full sibling,
any
Other family factor
Motor defect, mother
Mental retardation, prior-born
half-sibling
Maternal illness
Glomerulonephritis in pregnancy
Confining maternal illness,
>3 in past year
Number of hospitalizations
since last menstrual periodb
0
1
2 or more
Observed
Frequency
(Rate/lO,OOO)
Observed
Risk (%I
Odds Ratiod
69
78
8.9
12.5
9.2
6.6
2.5 (2.0,
3.7 (2.4,
2.6 (1.8,
1.8 (1.2,
44
33
225
Population
Attributable
Risk (%>
3.2)
5.7)
3.8)
2.7)
3.0
7.6
8.1
2.1 (1.3, 3.6)
2.3 (1.3,4.1)
0.4
0.4
119
390
8.4
9.2
2.5 (1.7, 3.6)
2.7 (2.3, 3.2)
1.6
5.7
40
69
8.0
7.2
2.2 (1.3, 3.9)
2.0 (1.3, 3.1)
0.4
0.6
12
62
11.1
8.0
3.3 (1.5, 7.6)
2.2 (1.4, 3.4)
0.2
0.7
39
8,618
1,119
263
3.7
4.7
5.4
5,374
2,712
1,503
386
24
3.7
3.9
4.1
4.4
6.4
0.9
1.0
0.6
Cigarettes smoked per d a p C
0
1-10
11-20
21-40
41-61
‘A linear trend for proportions is indicated at the p 5 0.05 level with no indication of nonlinearity.
b A linear trend for proportions is indicated at the p 5 0.005 level with no indication of nonlinearity.
‘Factors associated with high risk were very low in prevalence and would not contribute many cases of febrile seizures.
dValuesin parentheses represent 95% confidence limits.
only those factors with a x2 statistic less than a tabled value of
p < 0.01 are identiiied in Tables 1and 2. (Most of the results
presented actually had a computed x2 with a tabled value o f p
-= 0.001.)
Odds ratios greater than 1.0 indicate a positive association
of the etiological factor with the outcome of febrile seizures.
Confidence intervals for the odds ratios were computed as
described by Fleiss 1157. Only those statistically significant
factors with odds ratios of approximately 2.0 or more are
presented in Tables 1and 2. Population attributable risk was
used as a measure of the proportion of the outcome (febrile
seizures) that might be prevented if the risk factor were
eliminated or modified { 151. Multivariate analyses of these
risk factors were not undertaken, the risk factors having in
general either very low prevalence or modest odds ratios or
negligible population attributable risk.
Results
Prenatal and pregnancy or delivery characteristics associated with at least a doubling of risk for febrile sei128 Annals of Neurology Vol 27 No 2 February 1990
zures were family characteristics and maternal illness
(see Table 1). Of factors identifiable before birth, only
mothers’ histories of febrile seizures or of glomerulonephritis were associated with an observed risk that
exceeded 10%; these were also the only factors with
odds ratios exceeding 3.0.
Nonfebrile maternal seizures, remote (>5 years before the birth of the child) or current ( 5 5 years before
the birth), maternal seizure medication during pregnancy, and any seizure in the father or an older sibling
were also associated with approximately doubling the
risk of febrile seizures, as were motor deficits in the
mother and mental retardation in an older half-sibling.
All statistically significant factors were associated with
small population attributable risks; maternal and s i b
ling seizures were the only characteristics associated
with population attributable risks of 3% or more.
The occurrence of more than three confining illnesses in the year prior to study entry was a risk factor.
Table 2. Neonatal Risk Factors for Febrile Seizures
Factor
Macrocephaly
Diminished cry
Respiratory distress syndrome
Degree of respiratory difficultyb.'
None
Slight
Moderate
Severe
Sepsis
Errors of metabolism
Observed Frequency
(Rate/10,000)
Population
Attributable
Observed
Risk (%)
88
127
18
7.3
7.4
9.8
9,620
219
129
31
14
2
3.8
4.9
6.3
5.7
13.1
25.0
Odds Ratio"
Risk (%)
2.0 (1.4, 2.9)
2.1 (1.5, 2.9)
2.8 (1.4, 5.7)
0.8
1.2
0.3
4.0 (1.9, 8.3)
9.7 (2.3.41.8)
0.3
0.1
~~
~
"Values in parentheses represent 95% confidence limits.
bA linear trend for proportions is indicated at the p zs 0.005 level with no indication of nonlinearity.
'Factors associated with high risk were very low in prevalence and would not contribute many cases of febrile seizures.
Glomerulonephritis was the only specific maternal illness, other than seizures during pregnancy, that was
associated with febrile seizures (odds ratio, 3.3). There
was a trend for increased risk of febrile seizures with
the number of hospitalizations since the last menstrual
period and the number of cigarettes smoked per day as
reported at the first prenatal interview. Number of
hospitalizations (>O vs 0 and >1 vs 11) and number
of cigarettes smoked per day (>0 vs 0, >20 vs 520,
and >40 vs 1 4 0 ) were both associated with less than
twofold increases in risk of febrile seizures in the child.
The population attributable risk associated with each
was very small: Even if it were possible to eliminate
glomerulonephritis, hospitalizations, and smoking, the
rate of febrile seizures in the population would probably not be much altered.
Socioeconomic status, using a scale of 0 (low) to 9.5
(high) based on a composite of scores for education
and occupation of the head of the household and total
family income, was inversely related to risk of febrile
) p < 0.0005).
seizures in white children ( ~ ~ (=421.3;
The same relationship was not evident for black children.
None of the large number of pregnancy or labor or
delivery factors examined, which included maternal
parity, number of prenatal visits, diabetes, prior infertility, prior pregnancy loss, x-ray exposure, asthma,
toxemia, anemia, kidney or bladder infections, vaginal
bleeding in pregnancy, abruptio placentae, placenta
previa, prolonged rupture of membranes, meconium,
chorionitis, breech presentation or breech delivery,
cesarean section, fetal bradycardia, low Apgar score,
and many other characteristics, was associated with as
much as a doubling of risk of febrile seizures.
Two uncommon neonatally recognized factors, sepsis and errors of metabolism, had odds ratios exceeding
4.0 (see Table 2). Because of their low prevalence,
these factors had very low population attributable
risks. There was a trend for increased risk of febrile
seizures with increased level of degree of respiratory
difficulty. However, respiratory distress (none vs any)
in the newborn period was associated with only modest increase in risk of febrile seizures. Several risk factors cited by other authors, included in Table 3, had
odds ratios slightly higher than 1.0 but none reached a
doubling of risk.
Discussion
The prospective study reported here had the advantage that the histories and follow-up were undertaken
for all members of the population, comparably for children who turned out to have febrile seizures and for
those who did not, before the occurrence of any seizure. In addition, this study included and defined a
large set of perinatal factors, hundreds in number, for
investigation.
Some relationships that were not seen or were weak
in the study of the total cohort of febrile seizures
might emerge in analysis of clinically defined subsets
such as lengthy, multiple, or focal febrile seizures
(complex seizures 191). This is a possible subject for
further investigation.
There was a trend toward decreasing risk with increasing socioeconomic status in white but not black
children. In the British national cohort study 171, socioeconomic class was not related to the risk of febrile
seizures.
Family history was the major identified contributor
to vulnerability to febrile seizures. Both febrile and
nonfebrile seizure histories in the mother, who was the
informant, were related to risk of febrile seizures in
the child. We could not distinguish febrile from non-
Nelson and Ellenberg: Prenatal and Perinatal Antecedents of Febrile Seizures
129
Table 3. Examination of Previously Reported Risk Factors
Factor
Observed Frequency
(Rate/lO,OOO)
Observed
Risk (%)
Odds Ratioa
Population
Attributable
Risk (%)
85
2,248
3,624
2,144
1,157
742
1,010
2,870
292
498
4.7
3.9
3.8
3.8
3.8
4.2
4.2
4.0
4.3
4.7
1.2
1.02
1.0
1.01
1.0
1.1
1.1
1.0
1.1
1.2 (1.0, 1.5)
1.1
1,022
500
90
44
191
168
4.9
4.7
4.7
4.6
6.7
4.7
1.3 (1.2, 1.5)
1.3 (1.0, 1.5)
1.2
1.2
1.8 (1.4, 2.5)
1.2
Maternal or obstetric
Maternal age (yr)b
10-14
15-19
20-24
25-29
30-34
35-49
Prior abortion or stillbirth
Cord complications
Breech delivery
Cesarean section
Infant characteristics
Birthweight 12,500 gm
1-minute Apgar score 0-3
5-minute Apgar score 0-3
Neonatal brain abnormality
Feeding problems
1-year brain abnormality
3.1
1.2
1.5
"No confidence l i t s or population attributable risks are reported for factors with tabledp value > 0.05. Values in parentheses represent 95%
confidence limits.
b ~ 2 ( 5for
) 2 x 6 contingency table greater than tabled value ofp = 0.8; relative risks, which compared each age group with 20-24 year olds, are
presented in place of odds ratios.
febrile seizure history in the father or siblings; the
magnitude of increase in risk in the child was comparable for any seizure in mother (odds ratio, 2.5), father
(odds ratio, 2.5) or sibling (odds ratio, 2.7). In addition
to family history of seizures, maternal motor deficit
and mental retardation in an older half-sibling were
related to risk of febrile seizures. Schuman and Miller
[lbf reported an association between febrile seizures
and mental retardation in siblings.
Caution is appropriate in interpreting population attributable risk and in generalizing population attributable risk from one population to another E17, 181. In
using population attributable risk to estimate the proportion of febrile seizures that might be prevented if
the identified predictors could be eliminated, we did
not find any factor that is potentially preventable and
possesses a high population attributable risk. The most
obviously preventable of the risk factors examined was
maternal smoking. Passive smoking is a known risk
factor for respiratory illnesses in young children [19].
While maternal smoking as reported in this study referred to smoking during pregnancy, it is possible that
the effect may have been caused by the postnatal environment. Elimination of smoking over two packs of
cigarettes per day might reduce the frequency of febrile seizures by only 0.2%, and elimination of all
smoking might reduce it by 3.1%.
Most of the febrile seizures in this population could
130 Annals of Neurology Vol 27 No 2 February 1990
not be attributed to any of the maternal, obstetric, and
neonatal risk factors among the large set examined in
this study. We confirm and extend the observation of
the British cohort study [-/I that pregnancy and birth
factors contribute little to risk of febrile seizures.
References
1. Nelson KB, Ellenberg JH. Predictors of epilepsy in children
who have experienced febrile seizures. N Engl J Med 1976;
295: 1029-1033
2. Millichap JG. Febrile convulsions. New York Macmillan,
1968:32-34
3. Lennox-Buchthal MA. Febrile convulsions, a reappraisal. Amsterdam: Elsevier, 1973:63
4. Wallace S. Aetiological aspects of febrile convulsions: pregnancy and perinatal factors. Arch Dis Child 1972;47:171-178
5. Tsuboi T. Seizures of childhood: a population-based and clinicbased study. Acta Neurol Scand 1986;74(supl 110):50-52
6. van den Berg B, Yerushalmy J. Studies on convulsive disorders
in young children: V. Excess early fetal deaths among pregnancies preceding the birth of children with febrile or nonfebrile
convulsions. J Pediatr 1974;84:837-840
7. Verity CM, Butler NR, Golding J. Febrile convulsions in a
national cohort followed up from birth: 11. Medical history and
intellectual ability at 5 years of age. Br Med J 1985;290:13111315
8. Ellenberg JH, Nelson KB. Febrile seizures and later intellectual
performance. Arch Neurol 1978;35:17-21
9. Nelson KB, Ellenberg JH. Prognosis in children with febrile
seizures. Pediatrics 1978;61:720-72 7
10. National Institutes of Health. Febrile seizures: a consensus of
their significance, evaluation and treatment. Pediatrics 1980;
66: 1009- 1012
11. Niswander KR, Gordon M. The women and their pregnancies.
Washington, DC: US Dept of Health, Education, and Welfare,
1972 (DHEW NO. 73-379)
12. Nelson KB, Ellenberg JH. Antecedents of seizure disorders in
early childhood. Am J Dis Child 1986;140:1053-1061
13. Nelson KB, Ellenberg JH. Antecedents of cerebral palsy: 1.
Univariate analysis of risks. Am J Dis Child 1985;139:10311038
14. Mantel N, Haenszel W. Statistical aspects of the analysis of data
from retrospective studies of disease. J Natl Cancer Inst 1959;
22~719-748
15. Fleiss J. Statistical methods for rates and proportions. 2nd ed.
New York: John Wiley, 1981:71-73
16. Schuman SH, Miller LJ. Febrile convulsions in families: findings
in an epidemiologic survey. Clin Pediatr (Phila) 1966;5:604-608
17. EllenbergJH. The extrapolation of attributable risk to new populations. Stat Med 1988;7:717-727
18. Greenland S, Robins JM. Conceptual problems in the definition
and interpretation of attributable fractions. Am J Epidemiol
1988;128(6):1185-1197
19. Eriksen MP, LeMaistre CA, Newell GR. Health hazards of
passive smoking. Annu Rev Public Health 1988;9:47-70
Nelson and Ellenberg: Prenatal and Perinatal Antecedents of Febrile Seizures 131
Документ
Категория
Без категории
Просмотров
4
Размер файла
379 Кб
Теги
perinatal, antecedent, prenatal, seizure, febrile
1/--страниц
Пожаловаться на содержимое документа