Population studies have reported the incidence of epilepsy in both sexes is 44 cases per 100,000 person years. The incidence in females, at 41 cases per 100,000 person years, is less than that for males, at 49 cases per 100,000 person years.  The Rochester epilepsy study also found that the prevalence of epilepsy was slightly higher in males than females (6.5 vs 6.0 per 1000 persons).  As these higher rates in males may be attributable to the higher frequency of some major etiologies of seizures in men (eg, cerebrovascular disease, head trauma, alcohol-related seizures), it may be that increasing rates of such conditions in women may result in less difference between the sexes. The risk for recurrent seizure is similar between males and females,  as is the likelihood of ultimate remission of epilepsy. 
Although most epilepsy syndromes are equally or more commonly found in males than in females, childhood absence epilepsy and the syndrome of photosensitive epilepsy are more common in females.  In addition, some genetic disorders with associated epilepsy (eg, Rett syndrome and Aicardi syndrome) and eclamptic seizures in pregnancy can only occur in females. The age-adjusted incidences for simple and complex partial and generalized tonic-clonic seizures are higher for men than women. 
For more information, see the following:
Sex Hormones and Epilepsy
Cortical excitability is known to be affected by pituitary and gonadal hormones. Estrogens can activate seizures and interictal discharges when directly applied to the cerebral cortex or infused intravenously. This effect is at least in part due to altered calcium permeability of the cell membrane, reduction of chloride influx through the gamma-aminobutyric acid (GABA)-A receptor, and action of estrogen as a glutamate agonist in the hippocampus. Progesterone decreases cortical excitability, by enhancing GABA effects, and increases electroshock seizure threshold in experimental models.
Menarche begins with increases in ovarian hormones, the development of cyclic fluctuations, and an overall marked increase in pituitary luteinizing hormone (LH) and follicle-stimulating hormone (FSH). These hormones also fluctuate in pregnancy, during the perimenopausal period, and with the introduction of exogenous hormones (eg, oral contraceptives, hormone replacement therapy). All of these hormonal alterations can result in changes in epilepsy and seizure threshold, although this effect has marked variability among individuals.
Menarche and epilepsy
Many women report cyclic exacerbations of seizures, and patients with childhood-onset epilepsy may note that seizures that occurred sporadically before menarche become somewhat more predictable after puberty. Reviews have identified 25% of women had their first seizure in association with menarche.  Certain genetically determined epilepsies (ie, juvenile absence epilepsy [JAE] and juvenile myoclonic epilepsy [JME]) will present around puberty, whereas some nongenetic partial epilepsies may worsen, causing them to come to medical attention at this time. Childhood absence epilepsy  and benign rolandic epilepsy  may remit at puberty. Other researchers have noted that if epilepsy presents within the year after menarche, it is common for seizures to maintain variability with cyclic hormone changes. 
At menarche, pituitary gonadotropins (FSH and LH) and ovarian steroids (estrogen and progesterone) increase in overall concentration. Although epilepsy may be expressed at this time in part due to this increase in estrogen, cyclic increases in estrogen relative to progesterone appear as a likely trigger for breakthrough seizures.  Different seizure types may be exacerbated by different hormonal patterns of exacerbation. For example, partial seizures more commonly worsen in the follicular phase, whereas absence seizures more likely increase in the luteal phase. 
Catamenial epilepsy is a general term applied to any exacerbation in seizures with the menstrual cycle. True catamenial epilepsy requires reproducible and consistent increase or change in character of seizures at the same point in a regular menstrual cycle. Anovulatory cycles can make the relationship of seizure to hormones difficult to discern.
Perceived rates of catamenial epilepsy tend to be much higher than true catamenial epilepsy. Although 10-75% of women in one study noted a cyclic variation in seizure frequency, it was not reported whether this occurred during regular, ovulatory cycles, or was confirmed by physiologic hormonal level changes.  The breadth of the reported range itself indicates that seizure exacerbations are influenced by multiple factors, rather than exclusively hormonal changes. The physiologic processes of menstruation and ovulation can transiently worsen many medical conditions, and it is likely that, for some women, physical effects that are only secondarily related to hormones will increase seizures. Nonetheless, between one third and one half of women report increased cyclic seizures related primarily to hormonal factors.  .
Potentially, variations in antiepileptic drug (AED) levels may contribute to seizures related to decreases in circulating estrogen and progesterone premenstrually. Induction of hepatic isozymes at this time could reduce effective levels of hepatically metabolized AEDs.
Herzog et al described 3 patterns of hormonally based catamenial epilepsy.  Significantly higher numbers of seizures were found when estrogen increased faster than progesterone: perimenstrually (catamenial type 1) and preovulatory (catamenial type 2). In patients with anovulatory cycles, seizures more commonly occurred in the second half of the cycle (catamenial type 3) when progesterone was lower than normal due to failure to develop a corpus luteum. 
Management of patients with catamenial exacerbations must begin with a clear definition of the clinical situation. An accurate diagnosis of epilepsy is necessary, and may indicate a need for video-electroencephalogram (EEG) monitoring. Paroxysmal events mimicking epilepsy (ie, nonepileptic events, or "pseudoseizures") may also worsen at cyclic intervals.
Clear identification of ovulation at regular intervals is helpful in determining whether there is any specific higher risk time in the menstrual cycle. A scrupulously maintained calendar listing seizures and menses for several cycles can suggest a cyclic pattern of seizures. This can be confirmed by physiologic measures including basal body temperature and serum hormone levels. Other factors that may contribute to poor seizure control such as sleep deprivation, suboptimal medication for epilepsy type, and simple noncompliance must be eliminated, if possible.
Optimization of AED dosing is often helpful, and increasing doses each cycle for the duration of the expected exacerbation may be the best-tolerated option. Some AEDs, however, are poorly suited to this strategy, particularly if they exhibit nonlinear pharmacokinetics or extensive enzyme induction.
Intermittent dosing with short-acting benzodiazepines in catamenial exacerbations may be considered, although these agents are typically ineffective for long-term use owing to tolerance. Clobazam, however, has been studied in a double-blind cross-over study in 18 patients with catamenial epilepsy. Fourteen patients reported better control on clobazam than placebo.  Temporary use of adjunctive therapy such as acetazolamide has anecdotally been advocated but without prospective supportive evidence. 
Management strategies using exogenous hormones have not been consistently successful. If improved control is incidentally noted with hormonal contraceptives, and pregnancy is not desired, then longer-acting contraceptive agents, such as medroxyprogesterone (Depo-Provera), may be helpful in seizure control. Persistent menstrual irregularity or other suggestions of spontaneous anovulatory cycles warrants evaluation by an endocrinologist.
Some researchers have had limited success with progesterone to control true catamenial epilepsy. The metabolites of natural progesterone, especially allopregnanolone, appear to have anticonvulsant properties. Continuous progestational supplements reduced catamenial seizures as they suppressed cycles,  and intermittent progesterone suppositories reduced seizures up to 60%, but half of women on these supplements discontinued treatment due to intolerable side effects.  Adverse effects of synthetic and, to a lesser degree, natural progesterones include sedation, breast tenderness, depression, increased appetite and weight, and breakthrough menstrual bleeding. In addition, unopposed progesterone has been implicated in breast cancer, lipid elevations, and hypercoagulability. 
Synthesized novel neurosteroids such as ganaxolone, a synthetic analog of allopregnanolone, have proven promising but also have prohibitive production costs.  Overall, natural progesterones are better tolerated than synthetic agents, but no evidence substantiates the efficacy of one agent over the other. Further evidence is needed before any hormonal agent can be recommended for epilepsy management.
Infertility and Epilepsy
Infertility, polycystic ovary syndrome, and sexual dysfunction are discussed in this section.
Overall, women with epilepsy have lowered fertility compared with women in the general population.  Menstrual disorders are estimated to occur in 1 of 3 women with epilepsy compared with 1 in 7 in the general population.  Oligomenorrhea and abnormal cycle length (< 23 d or >35 d) occur in up to one third of women with epilepsy. At least one third of menstrual cycles in women with generalized seizures are anovulatory. 
Seizures themselves can result in abnormal reproductive hormone variations. Fluctuations of luteinizing hormone (LH) and pulsatile release of prolactin and sex steroids have been observed in temporal relation to some seizures.  While some clinicians incorrectly assume that elevated prolactin indicates true seizure and low prolactin indicates nonepileptic events, the American Academy of Neurology guideline for the use of serum prolactin level documents that this is not a consistent enough relationship for diagnostic purposes. 
The type of epilepsy may also have a significant influence on reproductive hormones and function. The reciprocal feedback of the temporal and limbic structures with the hypothalamus can alter secretion of hypothalamic, pituitary, and gonadal hormones. A structural neurologic abnormality of the amygdala or mesial temporal lobe may cause both epilepsy and altered hypothalamic hormonal function. Different hypothalamic gonadotropins responses result from right versus left temporal lobe seizure activity have been observed. 
Observed lower birth rates in women with epilepsy may be due to social inhibitions. There may be fear of rejection due to seizures during social occasions, development of relationships, or intercourse. Some studies have indicated that reproduction rates in married women with epilepsy are not less than those in married women without epilepsy, although marriage rates are lower and occur at older ages, when fertility may be less for other reasons.  Others have reported that the reduced fertility rate in married women with epilepsy is only partly explained by these social factors. 
Fear of effects of epilepsy or its treatments on a pregnancy may cause women with epilepsy to seek contraception, abortions, and even sterilization. In recent years, attention has been focused on historical state sterilization programs for reproductive age women with mental handicaps, often including those with epilepsy. Such programs are no longer enforced, although caregivers of women with epilepsy not infrequently request healthcare providers to perform surgical sterilization or prescribe long-acting contraception. The American College of Obstetrics and Gynecology (ACOG) has issued a committee opinion that mental capacity in itself does not justify sterilization, and that any such plan must be agreed upon with the appropriate caregivers or agents to preserve the patient's best interests and, to the maximum extent possible, her autonomy. 
For more information on this topic, see Infertility.
Polycystic ovary syndrome
Polycystic ovary syndrome (PCOS) and polycystic ovaries (PCOs) have received much attention in the epilepsy literature. Polycystic ovary morphology has been found by ultrasonography in 20-30% of healthy premenopausal women.  In contrast, using National Institutes of Health (NIH) diagnostic criteria (ovulatory dysfunction, clinical manifestation of hyperandrogenism, exclusion of other endocrinopathies), the cumulative prevalence of PCOS is only 6.6%. 
Long-term effects of PCOS include infertility, insulin resistance, obesity, dyslipidemia, hirsutism, and increased risk for endometrial cancer. PCOS may have similar cardiovascular risks of metabolic syndrome, because the 2 disorders are both associated with abdominal obesity, insulin resistance, and dyslipidemia. PCOS has been found to be more prevalent in women with epilepsy than in the general population. 
PCOS has been reported in 41% of women with idiopathic generalized epilepsy and in 26% of women with localization-related epilepsy.  Because idiopathic generalized epilepsy has often been treated with valproate, it remains controversial whether this medication or the seizure type (or perhaps a combined effect) is the critical trigger for development of PCOS.
Treatment of PCOS includes the antiestrogenic drug [AED] clomiphene, which is also used as a fertility-enhancing agent. Use of this agent in PCOS patients with epilepsy has been anecdotally reported to be associated with a reduction in seizures, but no evidence supports this agent as a primary antiepileptic agent. 
A retrospective series found multiple cysts by ultrasonography on the ovaries of 43% of women taking valproate for epilepsy,  adding to the ongoing controversy of the relative contributions of valproate, weight gain, and epilepsy itself to the development of PCOS. Harden's extensive review of the literature concluded PCOS has multiple etiologies, but because valproate is often a confounding variable, "…it seems prudent that clinicians educate themselves on the prevalence of PCOS in epilepsy before recommending the use of valproate, or any AED to female patients." 
For more information on this topic, see Polycystic Ovarian Syndrome.
Studies indicate up to one third of women with epilepsy self-report sexual dysfunction.  Just as with infertility, it is likely that multiple social and psychologic reasons underlie this difference. However, some studies indicate that physiologic factors may play a significant part. An outpatient study noting difficulty with sexual arousal also found increased occurrence of vaginal dryness and vaginismus.  Another study of physiologic responses noted decreased vaginal blood flow in women with epilepsy compared with controls. 
Treatment of sexual dysfunction must begin with identification of potential psychologic factors and appropriate counseling or therapy. Physiologic factors should be identified by an overall review of health and medications. Simplifying polytherapy to monotherapy or changing an antiepileptic drug to another better-tolerated agent may be considered.
Contraception Issues in Epilepsy
Women taking cytochrome P450 enzyme-inducing antiepileptic drugs (AEDs) have a potential 6% failure rate per year for oral contraceptive pills.  These AEDs increase hepatic metabolism of steroid hormones and increase their binding to sex hormone–binding globulin (SHBG) and other serum proteins, both effects that reduce the availability of hormonal contraception. The more potent enzyme inducers (carbamazepine, phenytoin, phenobarbital, primidone) are the most likely to interfere with contraception. Milder inducers (oxcarbazepine,topiramate) appear not to alter contraceptive efficacy significantly when administered at low doses. Lamotrigine, which has one of the most complex interactions with hormones, also potentially reduces the efficacy of contraception.
In addition, progesterone and its derivatives have been shown to significantly reduce lamotrigine levels, potentially increasing the risk of seizures. This effect is easily seen during pregnancy with a more than 50% drop in lamotrigine levels due to normal gestational increased progesterone.  Contraceptives with low doses of estrogen (eg, ethinyl estradiol) or progesterone (eg, norgestrel, norethindrone) may be poorly effective with these AEDs. Triphasic contraceptives, which contain 1 week of very low-dose estrogen immediately following the placebo week, effectively provides no contraceptive benefit until day 14 of the cycle. Such contraceptive regimens may also be particularly ineffective in women taking enzyme-inducing AEDs or lamotrigine. Alternative contraception or adjunctive methods should be considered in these patients.
Although neurologists and obstetricians should be familiar with these interactions, a 1996 survey indicated that both specialties were predominantly unaware of these effects and thus unable to provide appropriate contraception counseling to women with epilepsy.  Krauss et al found that 27% of neurologists and 21% of obstetricians reported oral contraceptive failures in their patients taking AEDs, but only 4% of neurologists and none of the obstetricians knew the effects of the 6 most common AEDs on oral contraceptives. 
Subdermal levonorgestrel implants (Norplant) have also been shown to have reduced efficacy in women taking enzyme-inducing AEDs.  It is likely that other forms of hormonal contraception (eg, transdermal patch, Depo-Provera) have potentially reduced efficacy with these drugs, but no literature supports this conjecture. Although the American Academy of Neurology Practice Parameter for Management Issues for Women with Epilepsy stated that increasing the estrogenic component of a contraceptive to at least 50 mcg will improve contraceptive effect, reproductive specialists disagreed, arguing that the progestin component has a greater effect in preventing ovulation than the estrogen component.  The literature is not consistent in identifying whether the estrogen or progestin component is clinically more important in pregnancy prevention. 
Whether gonadotropin-releasing hormone (GnRH) analogues and other nonovarian hormones have altered efficacy in epilepsy is unknown. Adjunctive contraception either by nonhormonal methods or changing antiepileptic therapy to those without hormonal interactions is a reasonable consideration. No impairment of hormonal contraception has been reported with ethosuximide, felbamate, gabapentin, levetiracetam, pregabalin, tiagabine, valproate, or zonisamide.  No adverse effecton contraception has been reported with implantable stimulators for epilepsy or with epilepsy surgery.
Reproductive counseling and folic acid are discussed in this section.
All physicians treating women of reproductive potential must discuss pregnancy with these patients, and when appropriate, their caregivers. There are few medical conditions in which pregnancy and childbirth do not complicate management. It is always preferable to discuss epilepsy management options before conception occurs. A candid discussion of whether or when pregnancy is desired can help determine the timing of diagnostic tests and medication changes. The Physician's Discussion Checklist for Women with Epilepsy contains helpful clinical printable practice aids for physicians of women with epilepsy in their reproductive years.
There are several ways to optimize therapy before consideration of pregnancy. First, establish whether the female patient requires antiepileptic therapy at all. Review of history and electroencephalographic (EEG) results may show that the diagnosis is unsubstantiated (due to lack of historical documentation or inadequate diagnostic testing) or that the patient is now seizure free for several years. Second, determine if the current regimen is appropriate for the epilepsy syndrome or seizure type. If drug choice or dosing is in question, referral to a comprehensive epilepsy center for a second opinion should be considered. Third, advocate simplification of treatment to monotherapy, or consider whether epilepsy surgery may be beneficial. EEGs and video-EEGs, scrupulously conducted by qualified laboratories and interpreted by qualified epileptologists, may be helpful.
Low serum or red blood cell folate levels have been associated with spontaneous abortion and developmental anomalies in animals and humans, especially neural tube defects (NTDs). [41, 42] All women of reproductive age, with or without epilepsy, should be on folate supplementation, even if pregnancy is not planned in the near future. Blood folate levels decline in women taking older antiepileptic drugs (AEDs). Although folate levels have not been shown to decrease in all AEDs, folate supplementation (see folic acid) is recommended for all patients with epilepsy, both men and women. In addition, a reported association with the MTHFR (a polymorphism of methylenetetrahydrofolate reductase) genotype appears to confer increased susceptibility to development of malformations when AED exposure occurs in utero. 
The American Academy of Neurology (AAN) updated their guidelines addressing the care of women with epilepsy in 2009. [44, 45] recommends use of folate supplementation at a dosage of no less than 0.4 mg daily for all reproductive-aged women before conception and throughout gestation.  Prenatal multivitamins contain at least 0.8 mg of folate, and individual folate tablets are available as 1-mg doses. Folate in doses up to 15 mg daily is safe and usually well-tolerated.  Higher doses likely offer more protection against development of NTDs, and most epileptologists recommend between 1 and 5 mg daily.
The AAN states women who have personal or family history of NTD births should receive 4 mg daily.  This is supported by the recommendation of the American College of Obstetrics and Gynecology that women with "health risks, including epilepsy" should take 5 mg folate from 3 months before conception through 3 months after conception, followed by 0.4-1.0 mg folate daily through lactation.  Note, however, that supplementation with as little as 1 mg daily can be expected to decrease phenytoin serum concentration and anticonvulsant effect,  thus, increased dosing of this AED may be necessary to maintain therapeutic effect.
Pregnancy and Epilepsy
The following topics will be reviewed in this section: seizures in pregnancy, teratogenicity of antiepileptic drugs (AEDs), pregnancy registries, prenatal care recommendations, and breastfeeding.
Seizures in pregnancy
Approximately 35% of women with epilepsy have more seizures during pregnancy, 10% have fewer, and 55% remain the same.  Multiple factors may contribute to worsening seizures. Physical and emotional stress may alter AED compliance, gastrointestinal absorption, or sleep habits. Compliance likely decreases during pregnancy more commonly than women report to their doctors, in part due to concern for the effects of AEDs on the developing baby.  Failure of patients to communicate noncompliance or seizure recurrence not only poses risk to mother and child, but it may put others indirectly at risk if the mother continues to drive a motor vehicle.
Physiologic changes of pregnancy that can alter both seizure threshold and AED pharmacokinetics include increases in sex hormones, increased volume of distribution, altered plasma protein binding, decreased gastric motility, increased cardiac output, increased renal elimination, and altered hepatic metabolism. The fetal hepatic metabolism may also be inadequate to process these drugs, which circulate from the mother through the placenta.
Frequent AED serum level monitoring should be performed at least at the beginning of every trimester and monthly from the eighth month through the eighth postpartum week of highly metabolized AEDS.  Free levels should be monitored in patients treated with highly protein-bound drugs, as the unbound fraction accounts for therapeutic effect. Dose adjustment in pregnant women, as in all epilepsy patients, should not be based on serum levels alone, but must consider seizure control, adverse effects, as well as the expected alteration of efficacy due to changing weight and metabolism.
Monitored fetal distress has been documented in labor during both generalized and partial maternal seizures. [51, 52] Avoiding situations that precipitate seizures (eg, changing AED, sleep deprivation, noncompliance) is of paramount importance. Thus, it is also recommended that AED changes should be completed at least 6 months before conception, and that AEDs should not be changed during pregnancy solely for the purpose of reducing teratogenicity. There is no "best" AED for women of reproductive age. Both the American Academy of Neurology (AAN) and American College of Obstetrics and Gynecology (ACOG) recommend choosing the AED most appropriate for control of the patient’s seizure type, with a goal of monotherapy.
For more information on this topic, see Seizure Disorders in Pregnancy.
Teratogenicity of AEDs
Although genetic influences, medical conditions, or lack of prenatal care may contribute to an increased incidence of fetal malformations in epilepsy, several lines of evidence suggest that AEDs play a primary role. Rates of fetal malformations are higher in pregnant women treated with AEDs than in those who are not treated with these agents. Higher AED levels have been associated with higher rates of malformation, and polytherapy carries a higher teratogenic risk than monotherapy. These findings suggest that optimal therapy for a pregnant woman with epilepsy should be to maintain monotherapy in the lowest effective dose. This must be balanced by the need to control convulsive seizures and other types that risk physical injury to the fetus or the infant. 
Although most AEDs are still classified as pregnancy category C (unknown risk to developing fetus), those with less hepatic metabolism and protein binding may be reasonable choices in pregnancy. In 2011, the FDA changed topiramate from category C to category D because of an observed increased the rate of oral clefts. Barbiturates and benzodiazepines used in epilepsy are category D in pregnancy. Drugs listed as pregnancy category D have known risk to the fetus, while those that are category X are contraindicated in pregnancy. Valproate remains category D for epilepsy, but in 2013 was downgraded to category X for migraine use.
Polypharmacy is usually not desirable during pregnancy as it has been associated with increased risk of malformations and results in even more complex metabolically induced pharmacokinetic changes as the pregnancy progresses. Thus, it is preferable to optimize the efficacy of a single agent by maintaining stable, unbound AED levels than to add another AED to the regimen. Adding or changing to another AED during pregnancy also invokes the risk of possible allergic reaction, worse seizure control, or unanticipated side effects. Of course, isolated situations occur in which changing therapy may unfortunately be necessary (eg, rash, life-threatening toxic effects).
The mechanisms underlying teratogenicity of AEDs remain unclear. Some AEDs may result in malformations by generating free radical (arene oxide) metabolites that bind with RNA and alter DNA synthesis and organogenesis.  Susceptibility to such oxidative damage may be genetically determined.  Impaired folate absorption or conversion to tetrahydrofolate (THF) (the biologically active form) is implicated in development of neural tube defects.
Studies have amply documented the protective effect of folic acid in pregnancies not exposed to AEDs. However, no conclusive evidence suggests that folic acid supplementation lowers the risk of neural tube defect in women on AEDs for epilepsy. Specific recommendations for the amount of folate are variable, but the general consensus among epileptologists is that 1-4 mg daily is appropriate for pregnant women with epilepsy. Although one study found that lower serum levels of folic acid predicted a higher risk of malformation in children of mothers who were treated for AED,  data from the North American Antiepileptic Pregnancy Registry indicated that malformation rates in women treated with valproate or lamotrigine are not lower even with folate supplementation of 3 mg daily. [57, 58]
Malformation rates in the general population range from 2% to 3% in most sources, and 4% to 6% in women with epilepsy taking AEDs.  Minor malformations such as facial dysmorphisms and digital hypoplasia occur in 6-20% of pregnancies with in utero exposure to AEDs.  These malformations are often subtle and disappear with maturation through the first year of life. Evidence for medication-specific syndromes (eg, fetal hydantoin syndrome, fetal barbiturate syndrome) is lacking; because similar minor anomalies have been reported with in utero exposure to multiple different AEDs, fetal anticonvulsant syndrome is a more accurate term.
The most common major malformation in infants of mothers with epilepsy is orofacial clefting, occurring in 13.8 infants per 1000 births compared with 1.5 infants per 1000 births in the general population.  Neural tube defects have been previously reported in 0.5-1% of carbamazepine exposures  and 1-2% of valproate exposures.  These data, however, were confounded by polypharmacy; retrospective reporting, which did not include aborted pregnancies; and predated widespread folic acid supplementation. Questions about the validity of these numbers and a need to obtain prospective data on pregnancy exposures to the growing number of AEDs since 1993 prompted the development of multiple pregnancy registries.
Although additional research is needed to confirm findings, the Neurodevelopmental Effects of Antiepileptic Drugs(NEAD) study found that children exposed in utero to higher dosages of valproate and carbamazepine have lower motor and adaptive functioning.  Children exposed in utero to valproate may also be at increased risk for attention-deficit/hyperactivity disorder.
Recommendations from a joint Task Force of the Commission on European Affairs of the International League Against Epilepsy (CEA-ILAE) and the European Academy of Neurology (EAN) indicate that valproate should not be used as first-line treatment in female children, in female adolescents, in women of childbearing potential, and pregnant women unless alternative treatments are ineffective or not tolerated. 
For more information on this topic, see Antiepileptic Drugs.
Multiple prospective registries track outcomes of AED-exposed pregnancies worldwide. Although the enrollment methodologies and lengths of follow-up vary, significant findings have emerged. The North American AED Pregnancy Registry (NAAPR) identified at least a doubled risk of major malformations with either phenobarbital  or valproate monotherapy. 
Phenobarbital was the first AED in monotherapy identified by the NAAPR with a documented increased risk of major malformations (6.5% compared with background rate of 1.62%, or a relative risk of 4.2).  The United Kingdom Epilepsy and Pregnancy Register and the Australian Registry of Antiepileptic Drugs in Pregnancy also found increased risk of major defects with valproate monotherapy.  The number of pregnancies exposed to phenobarbital monotherapy was too small in other registries to establish the risk found by the NAAPR. Dose-related teratogenicity has been reported in the European epilepsy and pregnancy registry (EURAP). 
The US Food and Drug Administration (FDA) reported to healthcare providers preliminary evidence in the NAAPR of an increased rate of cleft lip and cleft palate with lamotrigine exposure in the first 3 months of pregnancy, but this has not been confirmed by other registries.  Current updated information from the NAAPR can be found at Antiepileptic Drug Pregnancy Registry.
Although pregnancy registries have confirmed concerns regarding the teratogenicity of specific AEDs, evidence that epilepsy itself confers an increased risk of malformation is conflicting, except in cases of inherited syndromes and congenital neurologic lesions in the mother.  It is likely that an interdependent combination of AEDs, underlying epilepsy diagnosis, nutritional issues, and genetic factors contribute to the overall risk of increased malformations in infants born to mothers with epilepsy. Lindhout et al found that, contrary to data from older studies, seizures especially during the first trimester are associated with a marked increased risk of malformations.  Whether the rate of malformation is increased, independent of AEDs, in well-controlled maternal epilepsy is unknown.
Prenatal care recommendations
Patients who are treated with valproate or carbamazepine should be offered prenatal testing with alpha-fetoprotein (AFP) levels at 14-16 weeks' gestation and level II (structural) ultrasonography at 16-20 weeks' gestation.  If appropriate, amniocentesis for alpha-fetoprotein and acetylcholinesterase levels should be obtained to determine presence of neural tube defects.
In addition to folate supplementation throughout pregnancy, the AAN also recommends prescribing 10 mg daily oral vitamin K in the last month of pregnancy to minimize risk of hemorrhagic disease of the newborn due to reduced vitamin K–dependent clotting factors in women treated with AEDs. This latter recommendation remains controversial, with the neurology and pediatric literature tending to recommend maternal preterm supplementation, whereas the obstetrical literature recommends postpartum parenteral supplementation of the infant. Controversy persists whether maternal oral supplementation offers any reduction in risk, and so recommendations to patients should be discussed between the treating neurologist and obstetrician to avoid contradictory management.
AEDs cross into maternal milk to varying extents. Highly protein-bound AEDs have low concentrations in breast milk, whereas AEDs that have no protein binding are present in similar concentrations between milk and serum. As long as AEDs are not changed after delivery, this is simply an enteral continuation of the same parenteral AED exposed to while in utero. There is a potential for barbiturate withdrawal in infants of women taking phenobarbital or primidone during pregnancy who do not breastfeed. Withdrawal effects of other AEDs in infants are not known.
Breastfeeding is strongly recommended to promote maternal-infant bonding and to reduce the risk of infectious and immune disorders later in life. The mother must be aware that the infant may not like the flavor imparted to milk at certain times after taking her AED, and she should be advised to consider timing feeding or pumping milk to optimize nursing.
Most AEDs have an extensive transplacental transfer rates and low to moderate excretion into breast milk. Levetiracetam, however, can have higher excretion into breast milk.  The benefits of breastfeeding likely outweigh the risk of most AED exposure in the neonate; however, some drugs with age-specific side effects may warrant consideration. For example, whether lamotrigine has the same high risk of rash with exposure from breast milk or whether valproate has the same increased risk of hepatotoxicity is unknown. Presumably, because the infant tolerated the drug in utero, the risk for these age-specific toxicities is likely to be low.
The mother must be advised that her metabolism and clearance of AEDs will remain higher while she is lactating. Her levels may increase when she stops breastfeeding, requiring a dose adjustment.
Menopause and seizures and bone health are discussed in this section.
Menopause and seizures
Just as hormonal changes associated with menarche or pregnancy can affect epilepsy, menopause can alter seizure control. Women who have had reproducible catamenial patterns are more likely to experience improved seizure control after menopause. However, the perimenopausal time results in erratic fluctuations in gonadal steroids, which can temporarily worsen seizures.  Once hormone levels stabilize, such effects should improve, but exogenous hormones and the increasing risk of cerebrovascular disease may obscure this benefit. Postmenopausal estrogen replacement has been reported to exacerbate seizures in some women with epilepsy. 
Antiepileptic drugs (AEDs) may decrease bone mineral density (BMD) and result in osteopenia, osteoporosis, and fractures. Although these risks are present in both men and women treated with AEDs for more than several years, postmenopausal women are especially susceptible due to the added risk factor of hormonal depletion. A prospective study of women older than 65 years found that those taking AEDs were twice as likely to develop hip fractures.  Patients with epilepsy have been reported to suffer osteopenic fractures due to falls and tonic events associated with convulsions.
Several mechanisms have been proposed for bone loss with AED treatment. Cytochrome p-450 enzyme-inducing agents (phenytoin, phenobarbital, primidone, carbamazepine) increase vitamin-D metabolism, leading to decreased calcium absorption in the intestine, and increased parathyroid hormone, causing bone calcium stores to be mobilized. Reports suggest that non–enzyme-inducing AEDs, such as valproate, may also result in decreased bone mineral density, although to a lesser degree.  One study found increased fracture risk related to trauma in seizures that could not be ascribed simply to bone fragility. 
Testing for bone loss can be performed by using chemical or structural methods. Measuring biochemical serum levels for low calcium and phosphate, reduced vitamin D and its metabolites, and elevated serum alkaline phosphatase (ALP) and parathyroid hormone (PTH) levels may provide insight into risks for bone loss. However, such testing should not be used as the sole measure of bone health, because structural integrity and fracture resistance may be impaired even when laboratory tests are in the normal range. Markers of bone formation and resorption can also be monitored. Severity of these abnormalities can correlate with duration of AED exposure as well as the number and types of AEDs used. 
Bone mineral density testing is the preferred method to quantify risk of osteoporosis. Dual energy x-ray absorptiometry (DXA or DEXA) can be performed centrally (hip, spine) or peripherally (finger, wrist, heel, tibia) and are not always directly comparable, due to the different normal composition of these bones. The test defines a T-score with zero or greater as normal bone mass, and negative numbers define osteopenia (T score between -1 and -2.5) or osteoporosis (T <-2.5). Testing bone structure by bone mineral density should be conducted in all patients taking AEDs longer than 5 years and repeated at biannual intervals to assess adequacy of therapy. 
Therapy to prevent bone loss should begin well before menopause and is a consideration for any patient (male or female) treated with AEDs. Calcium supplements are most helpful when used in conjunction with vitamin C (which promotes absorption of calcium) and vitamin D. The recommended daily allowance of calcium varies from 1000 to 1500 mg in adults, and increased doses are suggested for those with evidence or risk of bone loss. The recommended daily intake of vitamin D is 400-800 IU, but doses up to 15,000 IU may be necessary in patients with evidence of bone loss.  Judicious sunlight exposure to the skin helps provide vitamin D for calcium metabolism. Weight-bearing exercise (even walking) also significantly reduces the risk of bone loss in the general population. Pharmacologic intervention with bisphosphonates or other agents is recommended for T scores of -2 or less. Such therapy should usually be managed by an internist or endocrinologist.