Metabolic and Hormonal Disturbances in Women with Epilepsy on Antiepileptic Drug Monotherapy
Disclosure: The authors have reported no conflicts of interest.
Address correspondence and reprint requests to Dr. Hyang Woon Lee, Associate Professor, Epilepsy and Sleep Disorder, Department of Neurology and Ewha Medical Research Institute, College of Medicine, Ewha Womans University, 911-1, Mok-dong, Yangcheon-gu, Seoul, 158-710, Korea, E-mail: email@example.com
Summary: Purpose: Women with epilepsy (WWE) tend to have hormonal and metabolic abnormalities, raising concerns about an increased risk of cardiovascular disorders. This study was performed to determine whether epilepsy itself and/or antiepileptic drug (AED) medication cause metabolic abnormalities.
Methods: WWE in premenopausal state aged 18 to 45 years old, currently on AED monotherapy for more than six months, were recruited for this study. The subjects checked their oral temperature each morning, and tested serum levels for lipid profiles, insulin, glucose, and leptin. A HOMA-index was used as a marker for insulin resistance.
Results: Of the 54 total patients, 18 women were diagnosed with primary generalized epilepsy (PGE) and the other 36 were diagnosed with localization-related epilepsy (LRE). Among the subjects, 19 women were on carbamazepine (CBZ), 12 on valproate (VPA), 12 on lamotrigine (LTG), and 11 on topiramate (TPM). Body mass index increased and HDL-cholesterol decreased in patients on VPA monotherapy compared with CBZ, LTG, or TPM (p = 0.046 and 0.002). Metabolic syndrome was more frequently associated with VPA-treated patients (41.7%) than CBZ (5.3%), LTG (0%), or TPM group (0%) (p = 0.005). There were no differences in hormonal and metabolic indices between PGE and LRE groups.
Conclusions: WWE on VPA monotherapy are more obese and more frequently suffer from metabolic syndrome. LTG or TPM may be safer when prescribed to the patients with high risk of cardiovascular disease.
Women with epilepsy (WWE) have hormonal or metabolic irregularities more often than healthy women of the same age (Morrell, 2003). Various metabolic and hormonal abnormalities have been reported in epileptic women, which include obesity, amenorrhea, oligomenorrhea, abnormally prolonged or shortened menstrual cycles, anovulation, polycystic ovary syndromes (PCOS), hyperandrogenism, and dyslipidemia (Isojärvi et al., 2001; Rättyä et al., 2001; Luef et al., 2002a; Verrotti et al., 2002; Morrell et al., 2003). The etiology of these hormonal or metabolic disturbances is still unknown. This might be due to neuroendocrine effects caused by the seizure or epilepsy itself, or it may be the side effect of chronic use of antiepileptic drugs (AEDs).
Valproic acid (VPA) is generally thought to induce hormonal and metabolic changes leading to low fertility rates, and increased risk of atherosclerotic vascular diseases (Isojärvi et al., 2001; Luef et al., 2002a; Morrell et al., 2003). The mechanism underlying these abnormalities is not certain, but insulin resistance and subsequent increase of serum insulin are considered to affect lipolysis in adipose tissue and cause altered lipid profiles as well as leptin resistance (Luef et al., 2002b; Pylvänen et al., 2002). On the other hand, several reports suggest that there is a high prevalence of reproductive endocrine disorders including menstrual irregularity, amenorrhea, and PCOS in WWE, whether or not treated with AEDs (Dana-Haeri et al., 1983; Herzog et al., 1986; Herzog, 1993, 1997). One study showed a significant increase in total and LDL cholesterol among CBZ-treated patients while triglycerides tended to be higher in VPA-treated group (Luef et al., 2003). Another study revealed that VPA monotherapy is more commonly associated with weight gain and higher androgen levels compared with lamotrigine (Morrell et al., 2003). There have been few studies, however, on newer AEDs so far.
Major risk factors of metabolic syndrome include abdominal obesity and insulin resistance. The diagnosis of metabolic syndrome is based on abdominal obesity, hypertension, high serum glucose and triglycerides, and low high density lipoprotein-cholesterols (HDL-chol). Patients diagnosed with metabolic syndrome are known to have a higher risk of type 2 diabetes mellitus (DM), cardiovascular diseases, and cerebrovascular diseases (Grundy et al., 2005), which can cause a significant impact on individual patients.
In this study, we investigate whether different epileptic syndromes or specific AEDs cause the hormonal and metabolic abnormalities. We evaluated various metabolic indices in Korean WWE on AED monotherapy longer than 6 months.
SUBJECTS AND METHODS
Epileptic women in premenopausal state aged 18 to 45 years with AED monotherapy for longer than six months were enrolled for this study. Clinical assessments, which included previous disease or medication history, seizure classifications, neurological examinations, EEG, brain CT, or MRI, were performed in all patients. The AEDs, types of seizures and epileptic syndromes, and the frequency of seizures were reviewed in all of the patients.
Based on clinical information, the epileptic syndromes were classified into primary generalized epilepsy (PGE) and localization-related epilepsy (LRE). To assess the effects of AEDs, only patients with monotherapy were included in the analysis, and AED groups were classified into four groups based on their medications; carbamazepine (CBZ), valproic acid (VPA), lamotrigine (LTG), and topiramate (TPM). Patients were excluded if they (1) had any endocrine or medical disorders (e.g., hypertension, heart diseases); (2) were pregnant or breast-feeding within the previous six months; (3) were menopausal or on hormonal treatments; (4) had a history of taking medications that affect lipid or hormonal metabolism; and (5) had any progressive neurological disorders other than epilepsy.
The subjects recorded their menstrual and seizure information in seizure diaries for more than three months or more than three menstrual cycles. Collected data included basal oral temperatures taken each morning, and the frequencies, types, and durations of clinical seizures. The subjects checked their oral temperatures while they were in bed, as soon as they awoke in the mornings. One or two days after menstruation began, the subjects went in for laboratory tests, which included serum triglycerides (TG), total cholesterol (chol), high and low density lipoprotein cholesterol (HDL-, LDL-chol), insulin, glucose, leptin, and the levels of AED. We also checked their weights, heights, waist, and hip circumferences. All the subjects filled out questionnaires describing their ordinary menstrual cycle, menstrual duration, history of amenorrhea, any treatment for ovulatory failures, degree of hirsutism, and other medical or family history. The degree of hirsutism was based on the Ferriman–Gallwey score (Ferriman et al., 1961). Menstrual irregularities were defined as oligomenorrhea, amenorrhea, or polymenorrhea (Palter et al., 1996). Informed consent was obtained from all patients and the study protocol was approved by our local Institutional Review Board based on the ethical principles of the Declaration of Helsinki (1964).
We followed the criteria of the American Heart Association/National Heart, Lung, and Blood Institute (AHA/NHLBI) to make the metabolic syndrome diagnosis. The diagnosis required the fulfillment of more than three of the following conditions: (1) waist circumference ≥ 80 cm (31 inches, Asian women), (2) TG ≥ 150 mg/dL or took antihyperlipidemic drugs, (3) HDL < 50 mg/dL, (4) systolic blood pressure ≥ 130 mmHg, or diastolic blood pressure ≥ 85 mmHg, or taking antihypertensive drugs, or (5) fasting blood glucose ≥ 100 mg/dL or taking antidiabetic drugs (Grundy et al., 2005).
Using the fasting glucose and insulin levels, we calculated the Homeostasis model assessment (HOMA)-index for each patient. The equation was as follows (Reilly et al., 2004):
We preformed a t-test and a one-way-ANOVA (SPSS 11.0, SPSS, Chicago, IL, U.S.A.) for metabolic changes, according to the types of epileptic syndromes and AEDs. A chi-square analysis was preformed for differences between each AED group.
Effects of epileptic syndromes
A total of 54 female patients were recruited; 18 of the women were diagnosed with PGE and the other 36 were LRE. Among the patients, 19 women had been on CBZ, 12 were on VPA, 12 on LTG, and 11 on TPM monotherapy.
Clinical characteristics and metabolic indices, according to the types of epileptic syndromes, are summarized in Tables 1 and 2. There were no differences in the rates of patients with menstrual irregularity between PGE and LRE groups. Heights, weights, body mass index (BMI), waist and hip circumferences, and waist to hip ratios (W/H ratio) were not different between the two groups. No significant differences in lipid profiles and other metabolic indices were found between PGE and LRE groups (Table 2).
Table 1. Clinical characteristics of epileptic women with different epileptic syndromes
|Number of patients (n)||18||36|| |
|Age (year)||24.8 (15–48)||30.6 (16–48)||a0.027|
|Irregular menstruation (%)|| 6 (33.3%)|| 11 (30.1%)|| 0.817|
|Menstrual cycle length (days)||29–39||29–35|| 0.144|
|Duration of AED medication (months)||31.5 ± 21.1||38.8 ± 26.5|| 0.501|
Table 2. Metabolic indices for epileptic women with different epileptic syndromes
|Weight (kg)||59.81 ± 8.51||56.23 ± 7.03||0.169|
|Height (m)||1.60 ± 0.04||1.58 ± 0.06||0.438|
|Waist (cm)||74.28 ± 10.20||74.83 ± 8.70||0.784|
|Hip (cm)||95.21 ± 6.19||93.88 ± 6.88||0.505|
|W/H ratio||0.78 ± 0.09||0.80 ± 0.07||0.868|
|BMI (kg/m2)||23.49 ± 3.52||22.47 ± 3.17||0.348|
|TG (mg/dL)||81.00 ± 42.23||93.67 ± 72.26||0.839|
|Chol (mg/dL)||168.37 ± 31.03||177.75 ± 42.70||0.362|
|LDL-chol (mg/dL)||89.85 ± 30.43||97.07 ± 31.85||0.498|
|HDL-chol (mg/dL)||65.78 ± 23.90||62.97 ± 19.49||0.966|
|Fasting-insulin (mg/dL)||14.30 ± 22.54||7.69 ± 6.80||0.511|
|Leptin (mg/L)||10.37 ± 7.88||9.36 ± 5.04||0.750|
|HOMA-index||2.89 ± 4.25||1.67 ± 1.55||0.713|
|Insulin resistance (%)||4 (22.2%)||5 (13.9%)||0.193|
|Metabolic syndrome (%)||2 (11.1%)||4 (11.1%)||0.687|
Effects of antiepileptic drug medications
Menstrual irregularity was twice as frequent in patients on VPA monotherapy than patients on CBZ, LTG, or TPM (p = 0.089). Menstrual cycles were more irregularly prolonged in VPA-treated patients, compared with CBZ, LTG, or TPM (p = 0.015) (Table 3). BMI was higher in VPA-treated patients, than CBZ, LTG, or TPM (p = 0.046). HDL-chol decreased in VPA-treated patients, compared with CBZ, LTG, and TPM (p = 0.002). Total or LDL-chol levels were different in each AED groups. The mean HOMA-index was two-fold higher in VPA-treated patients than CBZ, LTG, or TPM group, which was not statistically significant (p = 0.178). While the occurrence of insulin resistance was not different in these AED groups, metabolic syndrome was more frequently found in VPA-treated patients (41.7%), than CBZ (5.3%), LTG (0%), and TPM (0%) (p = 0.005) (Table 4).
Table 3. Clinical characteristics for epileptic women who were taking different antiepileptic drugs
|Number of patients (n)||19||12||12||11|| |
|Age (year)||30.5 ± 15.6||28.8 ± 12.2||26.3 ± 11.5||29.0 ± 18.9|| 0.542|
|Epilepsy type|| |
| PGE|| 0|| 9|| 5|| 5|| |
| LRE||19|| 3|| 6||10|| |
|Irregular menstruation (%)||4 (21.1%)||7 (58.3%)||3 (25.0%)||3 (27.3%)|| 0.089|
|Menstrual cycle length (days)||28–32||30–43||28–35||27–34||a0.015|
|Duration of AED medication (months)||38.8 ± 23.8||37.3 ± 25.8||24.8 ± 6.2||22.1 ± 14.4|| 0.059|
Table 4. Metabolic indices for epileptic women who were taking different antiepileptic drugs
|Weight (kg)||54.7 ± 5.9 ||60.3 ± 7.1 ||56.6 ± 7.8 ||56.9 ± 6.4 || 0.112|
|Height (m)||1.58 ± 0.05||1.57 ± 0.04||1.61 ± 0.06||1.60 ± 0.04|| 0.171|
|Waist (cm)||73.4 ± 8.6 ||78.1 ± 11.2||73.7 ± 5.6 ||72.4 ± 4.19|| 0.414|
|Hip (cm)||93.3 ± 5.5 ||97.7 ± 7.4 ||92.9 ± 6.2 ||91.2 ± 7.4 || 0.185|
|W/H ratio||0.79 ± 0.07||0.80 ± 0.11||0.80 ± 0.07||0.79 ± 0.08|| 0.911|
|BMI (kg/m2)||22.03 ± 2.45 ||24.43 ± 3.61 ||21.68 ± 2.14 ||21.19 ± 2.82 ||a0.046|
|TG (mg/dL)||99.9 ± 90.4||98.3 ± 56.6||78.0 ± 46.8||83.6 ± 32.1|| 0.624|
|Chol (mg/dL)||178.0 ± 35.2 ||165.3 ± 29.5 ||190.7 ± 44.3 ||172.2 ± 49.7 || 0.354|
|LDL-chol (mg/dL)||89.2 ± 18.1||94.0 ± 25.3||106.0 ± 32.9 ||87.1 ± 7.6 || 0.175|
|HDL-chol (mg/dL)||62.6 ± 11.0||49.3 ± 10.7||70.6 ± 16.2||65.5 ± 12.2||a0.002|
|Fasting-insulin (mg/dL)||7.1 ± 3.1||19.6 ± 29.0||8.7 ± 7.6||9.55 ± 4.51|| 0.181|
|Leptin (mg/L)||8.5 ± 7.9||14.3 ± 6.0 ||7.9 ± 3.4||6.92 ± 4.02|| 0.292|
|HOMA-index||1.53 ± 0.77||4.07 ± 5.54||1.77 ± 1.53||1.53 ± 0.77|| 0.178|
|Insulin resistance (%)||3 (15.8%)||4 (33.3%)||1 (8.3%)||1 (9.1%)|| 0.181|
|Metabolic syndrome (%)||1 (5.3%) ||5 (41.7%)||0 (0%) ||0 (0%) ||a0.005|
In the present study, markers for hormonal and metabolic indices were measured in Korean WWE aged 18–45, who were currently on AED monotherapy for longer than 6 months. BMI increases and metabolic syndrome was more frequently found in epileptic women on VPA monotherapy compared with other AEDs including CBZ, LTG, and TPM. Menstrual period was more variably prolonged in VPA-treated women. Metabolic indices were not altered in LTG and TPM-treated patients. There were no differences in hormonal or metabolic profiles observed between PGE and LRE groups.
WWE are known to be prone to hormonal abnormalities. However, the underlying mechanisms are still debatable. To determine whether these are an iatrogenic effect caused by the AED or a direct impact of epilepsy on the endocrine control centers of the brain, particularly the hypothalamic–pituitary axis (Herzog, 1997; Bauer et al., 2002; Mikkonen et al., 2004). The prevalence of PCOS, which is one of the most common reproductive endocrine disorders in epileptic women, is higher in epileptic women than in the general population, even for women who are not taking AEDs (Luef et al., 2002a). The study showed that partial seizures, of left temporal origin, were associated with the development of PCOS. PGE was found as a risk factor for ovulatory failure (Morrell et al., 2002). In our study, no significant difference in hormonal and metabolic features was observed between PGE and LRE groups. Iatrogenic effects of AEDs on reproductive endocrine functions can be mediated by alterations in protein binding and hormonal metabolisms (Morrell, 2003; Morrell et al., 2003). The incidence of menstrual disturbances was reported to be higher in patients with VPA monotherapy (45%) compared with VPA and CBZ combined therapy (25%), or CBZ monotherapy (19%) (Isojärvi et al., 1993). Anovulatory menstrual cycle was more frequently found in patients with PGE than LRE, and in patients on VPA monotherapy than other AEDs (Morrell et al., 2002). The incidences of women with normal menstrual cycles were 88% in LTG- and 78% in VPA-treated patients in the same study. They asserted that VPA therapy might be related to the disturbance of menstrual cycle. In contrast, menstrual irregularity was not different between CBZ- and VPA-treated women (17% versus 12%) in another study (Luef et al., 2002b). In our study, menstrual irregularity tended to be more frequent in VPA-treated women (58.3%), compared with other AEDs including CBZ (21.1%), LTG (25.0%), and TPM (27.3%).
The previous study revealed that BMI and mean body weight increased in women on VPA therapy (Isojärvi et al., 1996). One explanation is that VPA is a derivative of a fatty acid, and fatty acids stimulate the release of pancreatic insulin. VPA affects the release of insulin directly or indirectly by competing with serum fatty acids for binding sites with the albumin, leading to increased serum insulin levels (Luef et al., 2003). Another explanation is that weight gain induced by VPA might give rise to insulin resistance, and then hyperinsulinemia might occur (Morrell et al., 2003; Isojärvi et al., 1996). Internal insulin can induce serum glucose to enter into cells and act as an antilipolytic hormone, and then affect the fatty tissues directly. This results in increase in appetite and weight through induced hypoglycemia (Luef et al., 2003). Hyperinsulinemia exceedingly promotes the production of an androgen hormone and dyslipidemia (Morrell et al., 2003). Obesity related to VPA therapy might increase serum leptin released by fatty tissues as well as serum insulin. The effects of leptin in human beings suppress appetite and accelerate the metabolism. The increase of serum leptin produces leptin resistance, which impairs the normal weight and the energy consumption control mechanism (Luef et al., 2002b; Pylvänen et al., 2002). In our study, fasting insulin, leptin, and HOMA-index showed a nonsignificant tendency to increase in VPA-treated patients, compared with CBZ, LTG, or TPM.
In the previous study on lipid metabolism, serum TG levels were higher and cholesterol levels were lower in VPA-treated patients than CBZ (Luef et al., 2002a). Compared with LTG, VPA was found to decrease HDL-chol and elevate TG (Morrell et al., 2003). Serum TG levels were high in obese men taking VPA therapy; but cholesterol levels were high in CBZ-treated patients, because CBZ accelerate the cholesterol production (Pylvänen et al., 2003). However, our results showed HDL-chol was decreased in VPA-treated patients, whereas there were no significant differences in other lipid profiles between different AED groups.
Insulin resistance and/or metabolic syndrome have been suggested as one of the unrecognized risk factors for cardiovascular disease or stroke (Reilly et al., 2004). Compared to American women, the incidence of metabolic syndrome has been reported relatively low in Korean women; 22% in United States (Reilly et al., 2004) and 7.4% in Korea (Ryu et al., 2005). In this study, metabolic syndrome was found in 41.7% of the VPA-treated women, whereas it was only in 5.3% in the CBZ-treated group. None developed metabolic syndrome in LTG or TPM-treated patients, suggesting the usefulness of those drugs in patients with high risk of cardiovascular diseases.
In summary, our results demonstrate that disturbances of the menstrual cycle, obesity, and changes of the metabolic indices can be caused by long-term use of specific AED rather than specific epileptic syndrome. Menstrual and metabolic disturbances were more frequently found in VPA-treated patients than CBZ, in Korean WWE. LTG and TPM cause no significant alterations on hormonal or metabolic profiles. Special attention should be paid to the potential of causing hormonal and metabolic disturbances, when prescribing long-term AED therapy.
Acknowledgments: The authors thank Bosil Kim, R.N. for her contribution in collecting data from the patients and for revision of the manuscript.