Objective
We sought to estimate the association between maternal seizure disorder and adverse pregnancy outcomes.
Study Design
We performed a retrospective cohort study of singleton, nonanomalous pregnancies. Women with self-reported seizure disorder were compared to women without medical problems. The primary outcome was intrauterine growth restriction (IUGR) <10th percentile. Secondary outcomes included IUGR <5th percentile, stillbirth, preeclampsia, and preterm delivery. A sensitivity analysis was performed using women who reported using antiepileptics to estimate the impact of disease severity on pregnancy outcomes.
Results
Of 47,118 women, 440 reported a seizure disorder. Women with seizure disorder were not at increased risk of IUGR <10th percentile (adjusted odds ratio, 1.11; 95% confidence interval, 0.82–1.50), IUGR <5th percentile, stillbirth, preeclampsia, or preterm delivery. The results were similar in the sensitivity analysis of women taking antiseizure medications.
Conclusion
Our results suggest women with a seizure disorder are not at increased risk of IUGR, stillbirth, preeclampsia, or preterm delivery.
Seizure disorders are characterized by recurrent seizure activity and affect 0.3-0.8% of all pregnancies. It has been suggested that changes in placental blood flow during seizure activity or in the postictal phase may lead to hypoxia and acidosis. Fetal heart rate changes during a maternal seizure indicative of hypoxia, including bradycardia and reduced variability, have been reported. It is biologically plausible, therefore, that a maternal seizure disorder could lead to adverse pregnancy outcomes, such as intrauterine growth restriction (IUGR) and stillbirth, through recurrent hypoxic insults to the fetus.
Several studies report an increased risk of adverse pregnancy outcomes in women with seizure disorders; however, this risk is based primarily on birth certificate data and case-control studies. Given the prevalence of seizure disorders in pregnant women, validated and more accurate estimates of associated pregnancy risks would assist clinicians in caring for these patients.
Therefore, we aimed to estimate the association between maternal seizure disorder and IUGR, as well as other adverse pregnancy outcomes including severe IUGR, stillbirth, preeclampsia, and preterm delivery.
Materials and Methods
We performed a retrospective cohort study of all consecutive singleton pregnancies undergoing routine second-trimester ultrasound anatomic survey at Washington University Medical Center at 15 0/7 through 22 0/7 gestational weeks from 1990 through 2009. The Washington University Human Research Protection Office approved the study prior to its initiation.
Dedicated research nurses collected the data prospectively, primarily from medical records. In addition, at the time of second-trimester ultrasound, each patient provided detailed information regarding medical history, pregnancy history, and medication use. Each patient who presented for ultrasound was also given a form to be completed after delivery regarding pregnancy outcomes including pregnancy complications, delivery complications, and neonatal outcomes. If the form was not returned within 4 weeks of the expected date of delivery, the coordinator called the patient to obtain the information. Of patients, 92% delivered at our institution; if the patient did not deliver at Washington University, the coordinator contacted the referring physician to obtain and verify outcome data.
Patients were included in the study if they had a confirmed singleton gestation delivered >20 weeks and complete outcome data were available. Gestational age was assigned based on the first day of a woman’s last menstrual period. If the dating was not consistent with a first- or second-trimester ultrasound (±7 days in the first trimester or ±10 days in the second trimester), the gestational age was reassigned. Exclusion criteria included any other maternal medical problems (ie, chronic hypertension and pregestational diabetes) and congenital anomalies diagnosed prenatally. Women with a self-reported seizure disorder were compared to women with no seizure disorder.
The primary outcome was IUGR defined as birthweight <10th percentile for gestational week at delivery based on the Alexander growth standard. Secondary outcomes included birthweight <5th percentile for gestational week at delivery based on the Alexander growth standard, stillbirth defined as fetal death >20 weeks, preeclampsia as determined by the delivering physician, and preterm birth <37 weeks. Because seizure disorder was determined by maternal self-report, a sensitivity analysis was performed comparing the primary and secondary outcomes of women with a self-reported history of a seizure disorder and reported antiepileptic drug (AED) use to women without a seizure disorder. Therapy was determined by provider preference. In 2005, the trend of lamotrigine, oxcarbazepine, levetiracetam, and zonisamide use increased, while that of carbamazepine, phenytoin, divalproex sodium, and phenobarbital decreased. To account for changes in AED therapies with the availability of new AEDs, the data were divided into 2 periods using the year 2005 as a cutpoint, and a sensitivity analysis was performed. Finally, women with a self-reported seizure disorder not on medication were compared to women with a self-reported seizure disorder on medication to assess the impact of AED use on pregnancy outcome.
Baseline characteristics of the 2 groups were compared using the Student t test or Mann-Whitney U test for continuous variables and the χ 2 or Fisher exact test for categorical variables as appropriate. The incidences of the primary and secondary outcomes were compared between groups and the unadjusted relative risks with 95% confidence intervals (CIs) were estimated. Univariable analyses were performed to identify potential confounding factors. Logistic regression models were developed using backwards stepwise regression to estimate the independent effect of seizure disorder for each outcome adjusting for confounding factors that were identified historically and in the unadjusted analyses. Differences in the explanatory model were tested using the likelihood ratio test or Wald test. Statistically significant variables were included in the final models and adjusted odds ratios (aORs) with 95% CI were estimated ( P < .05 was considered significant). Statistical analyses were performed using software (STATA 10.0, Special Edition; StataCorp, College Station, TX).
Results
Of 76,453 women with singleton pregnancies undergoing routine second-trimester ultrasound, complete outcome data were available for 66,956 (87.6%). Of these, 15,611 women were excluded for medical comorbidities and 4227 were excluded for fetal congenital anomalies, leaving 47,118 remaining in the analysis. Of these, 440 (0.9%) reported a history of a seizure disorder, of whom 256 (0.5%) also reported AED use ( Figure ).
Women with a self-reported seizure disorder were similar to those without a seizure disorder with respect to nulliparity, history of a preterm delivery, and history of stillbirth. Women with a seizure disorder were younger, more likely to report tobacco or alcohol use during the pregnancy, and more likely to be white ( Table 1 ). The distribution of AED use is reported in Table 2 . The most commonly used AEDs were phenobarbital, carbamazepine, and phenytoin. Approximately 19% of women with a seizure disorder reported using multiple agents.
| Characteristics | Seizure disorder (n = 440) | No seizure disorder (n = 46,678) | P value |
|---|---|---|---|
| Maternal age, y | 28.7 ± 5.8 | 30.3 ± 6.3 | < .01 |
| Race | |||
| Black | 90 (20.5%) | 10,705 (22.9%) | < .01 |
| White | 302 (68.6%) | 28,317 (60.7%) | |
| Other | 48 (10.9%) | 7662 (16.4%) | |
| Nulliparous | 183 (41.6%) | 18,070 (38.7%) | .22 |
| Previous preterm birth (<37 wk) | 24 (5.5%) | 2532 (5.4%) | .97 |
| Previous stillbirth a | 8 (1.8%) | 946 (2.0%) | .76 |
| Tobacco exposure | 83 (19.0%) | 4906 (10.6%) | < .01 |
| Alcohol exposure | 63 (14.5%) | 8599 (18.6%) | .03 |
| Antiepileptic medication | No. of patients reporting use |
|---|---|
| Carbamazepine | 96 (37.5%) |
| Phenytoin | 83 (32.4%) |
| Phenobarbital | 45 (17.8%) |
| Divalproex sodium | 42 (16.4%) |
| Lamotrigine | 15 (5.9%) |
| Gabapentin | 10 (3.9%) |
| Topiramate | 8 (3.1%) |
| Levetiracetam | 5 (2.0%) |
| Oxcarbazepine | 3 (1.2%) |
| Zonisamide | 2 (0.8%) |
| Multiple agents | 48 (18.8%) |
Women with a self-reported history of seizure disorder did not have an increased risk of IUGR <10th percentile compared to women without a seizure disorder. Maternal age, black race, smoking, alcohol use, and history of a preterm delivery were considered in the regression models. The risk of IUGR <10th percentile remained similar after adjusting for maternal age, tobacco use, and black race (aOR, 1.11; 95% CI, 0.82–1.50). The risk of IUGR <5th percentile (aOR, 0.82; 95% CI, 0.56–1.20) and stillbirth (relative risk, 0.99; 95% CI, 0.32–3.08) were also similar between the study group and the women without seizure disorders ( Table 3 ). The risk of IUGR <5th percentile remained similar after adjusting for maternal age, tobacco use, and black race. In the additional secondary outcome analyses, there was not an increased risk of preeclampsia or preterm birth. The risk of preeclampsia remained similar after adjusting for maternal age, tobacco use, and black race (aOR, 1.04; 95% CI, 0.73–1.49). The risk of preterm delivery remained similar after adjusting for maternal age, tobacco or alcohol use during the pregnancy, black race, and history of a prior preterm delivery (aOR, 1.06; 95% CI, 0.81–1.38).
| Outcome | Seizure disorder (n = 440) | No seizure disorder (n = 46,678) | Unadjusted relative risk (95% CI) | Adjusted odds ratio (95% CI) | P value |
|---|---|---|---|---|---|
| IUGR a <10th percentile | 44 (10.00%) | 3733 (8.00%) | 1.25 (0.94–1.66) | 1.11 (0.82–1.50) c | .49 |
| IUGR a <5th percentile | 25 (5.68%) | 2885 (6.19%) | 0.92 (0.63–1.35) | 0.82 (0.56–1.20) d | .31 |
| Stillbirth b | 3 (0.68%) | 321 (0.68%) | 0.99 (0.32–3.08) | NA | NA |
| Preeclampsia | 33 (7.60%) | 3125 (6.90%) | 1.10 (0.79–1.53) | 1.04 (0.73–1.49) c | .81 |
| Preterm delivery <37 wk | 49 (11.14%) | 5000 (10.71%) | 1.04 (0.80–1.36) | 1.06 (0.81–1.38) e | .70 |
a Based on Alexander growth curve;
c Adjusted for maternal age, currently smoking, and black race;
d Adjusted for maternal age and currently smoking;
e Adjusted for maternal age, currently smoking, black race, alcohol exposure, and history of preterm birth.
In the sensitivity analysis comparing only women with a history of a seizure disorder and reported AED use to women with no seizure disorder, no difference was found in the risk of IUGR, stillbirth, preeclampsia, or preterm delivery ( Table 4 ). There were no differences in outcomes of the sensitivity analyses when the data were divided in the year 2005. Women with a seizure disorder on medication were at lower risk of IUGR <5th percentile and preeclampsia when compared to women with a seizure disorder not on medications ( Table 5 ).
