Serum Insulin and Leptin Levels in Valproate-associated Obesity
Address correspondence and reprint requests to Dr. V. Pylvänen at Department of Neurology, University of Oulu, FIN-90220 Oulu, Finland. E-mail: firstname.lastname@example.org
Summary: Purpose: Weight gain is an important adverse effect of valproate (VPA) therapy, and it is associated with hyperinsulinemia and hyperandrogenism in women with epilepsy. Leptin is considered a signaling factor regulating body weight and energy metabolism. In human subjects, obesity is in general associated with elevated serum leptin levels, suggesting decreased sensitivity to leptin. The present study aimed at evaluating the role of insulin and leptin in VPA-related obesity.
Methods: Body mass index (BMI) was calculated, and serum insulin and leptin levels were measured in 81 patients with epilepsy taking VPA and in 51 healthy control subjects.
Results: Forty (49%) of the patients taking VPA and 25 (49%) of the control subjects were obese. The mean insulin levels were higher in VPA-treated patients than in the control subjects despite similar BMI values, when all subjects were included in the comparison. Both obese male and female patients taking VPA had higher serum insulin levels than the respective control subjects with similar BMI values. Serum insulin levels also were higher in lean male and lean female patients compared with the lean control subjects of same sex. Serum leptin levels did not differ between the VPA-treated patients and the control subjects.
Conclusions: Both obese and lean patients taking VPA for epilepsy have hyperinsulinemia, suggesting development of insulin resistance. This may be one of the factors leading to weight gain during VPA treatment. However, the results of the present study do not suggest an independent role for leptin in the pathogenesis of VPA-related obesity.
Valproate (VPA) is a branched fatty acid that is structurally unrelated to any other antiepileptic drug (AED). It has a broad spectrum of antiepileptic efficacy in the prevention of both partial and generalized seizures (1). However, the use of VPA may be associated with adverse effects, among which weight gain is one of the most common (2–4). In women with epilepsy, VPA-related weight gain is associated with hyperinsulinemia and low serum concentrations of insulin-like growth factor binding protein-1 (IGFPB-1), which may promote hyperandrogenism and the development of polycystic ovaries (PCOs) (4,5). However, the pathogenetic mechanisms leading to weight gain during VPA medication are still poorly defined.
Hyperinsulinemia in general is known to be associated with obesity, dyslipidemia, and insulin resistance. These conditions are risk factors for type 2 diabetes, hypertension, and eventually coronary heart disease (6). Furthermore, hyperinsulinemia is related to the PCO syndrome in obese women (7). This may result in hyperandrogenism, hirsutism, and menstrual disorders.
Leptin is a product of the OB gene, which was first detected in mice. Mutations in this gene lead to obesity in rodents. Leptin is considered to be a signal factor that regulates body weight and energy expenditure (8–10). In rats leptin has been shown to reduce food intake and increase energy expenditure (9). It is likely that leptin regulates body weight through neuropeptide Y, which stimulates food intake and decreases thermogenesis in hypothalamus (11). A strong correlation has been observed between serum leptin concentrations, body mass index (BMI), and body fat mass in humans (12). This suggests that obesity may be associated with a decreased sensitivity to leptin. Conversely, lean as well as obese human subjects with insulin resistance have high concentrations of serum leptin. However, a similar degree of insulin resistance is associated with higher levels of circulating leptin in obese than in lean subjects (13). It has been suggested that VPA-related obesity may be associated with elevated serum insulin and leptin levels in women with epilepsy (4,14). The aim of the present study was to evaluate the role of insulin and leptin in VPA-related obesity in men and women with epilepsy.
MATERIALS AND METHODS
The study was carried out in the Outpatient Departments of Neurology at the Oulu and Helsinki University Hospitals, Finland, with the approval of the local Ethics Committees. Written informed consent from all the patients and control subjects was obtained.
Eighty-one consecutive patients (46 men, 35 women) taking VPA monotherapy for epilepsy participated in the study. None of the patients had any other regular medication in addition to VPA, had diseases other than epilepsy, were alcohol abusers, or were pregnant or lactating. The characteristics of the patients are presented in Table 1. Epilepsy type was classified according to recommendations of the International League Against Epilepsy (15). Fifty-one healthy volunteers (24 men, 27 women) served as control subjects. The mean age of the control subjects was 37.1 ± 7.1 years. The proportion of obese and nonobese subjects recruited in the control group was similar to that in the VPA-treated group of patients with epilepsy.
Table 1. The clinical characteristics of the patients studied
|81||29.6 ± 10.9||58||23||12.1 ± 7.9||6.5 ± 4.9||1126 ± 371||438 ± 145|
The medical history of each patient was obtained by interview and examination of hospital records. At clinical examination, weight, height, and hip, and waist circumferences were measured and BMI (weight in kilograms divided by the square of height in meters) and waist–hip ratio (WHR, waist circumference in centimeters divided by hip circumference in centimeters) were assessed. Patients with a BMI >25 kg/m2 were considered obese. Blood samples for serum insulin, leptin, and drug-concentration assays were obtained at 8 a.m. after an overnight fast.
Serum insulin concentrations were measured with an enzyme-linked immunosorbent assay (15). The sensitivity of the assay was 0.5 mU/L, and the intraassay and interassay coefficients of variation were <7.5 and 9.3%, respectively. Serum leptin concentrations were analyzed with radioimmunoassay by using kits from Linco Research, Inc. (St. Louis, MO, U.S.A.). The sensitivity of the assay was 0.5 μg/L, the intraassay variation was 3.9%, and the interassay variation was 4.7%. Serum VPA concentrations were assayed with a fluorescence polarization immunoassay system (TDX; Abbott Diagnostic Division, Irving, TX, U.S.A.). The sensitivity of the VPA assay was 0.7 mg/L, and the intraassay and interassay coefficients of variation were 2.3 and 3.1%, respectively.
Data were analyzed by using the Mann–Whitney U test, as some of the data were unevenly distributed.
The main results are presented in Table 2.
Table 2a. Age, body mass index (BMI), waist/hip ratio (WHR), serum insulin and leptin levels in female patients with valproate treatment
|BMI (kg/m2)||28.0 ± 2.1||22.2 ± 1.6||25.0 ± 3.5||27.8 ± 3.8||21.9 ± 2.0||23.9 ± 3.9|
|WHR||0.87 ± 0.05||0.79 ± 0.04||0.80 ± 0.06||0.85 ± 0.07||0.82 ± 0.06||0.80 ± 0.06|
|Insulin (mU/L)||9.8 ± 6.0a||6.7 ± 2.9b||8.2 ± 4.8c||5.4 ± 2.4||4.1 ± 3.0||4.5 ± 2.6|
|Leptin (μg/L)||27.9 ± 14.9||12.4 ± 6.4||20.0 ± 13.7||26.3 ± 12.2||11.4 ± 4.8||16.8 ± 11.6|
Table 2b. Age, body mass index (BMI), waist/hip ratio (WHR), serum insulin and leptin levels in male patients with valproate treatment
|BMI (kg/m2)||28.2 ± 1.8||22.6 ± 1.0||25.4 ± 3.1||28.2 ± 1.7||22.2 ± 1.2||26.1 ± 3.2|
|WHR||0.97 ± 0.07||0.90 ± 0.05a||0.94 ± 0.07||0.96 ± 0.06||0.95 ± 0.05||0.95 ± 0.05|
|Insulin (mU/L)||11.7 ± 5.9b||8.05 ± 7.1c||10.2 ± 6.75d||5.3 ± 2.0||5.7 ± 9.8||5.3 ± 5.8|
|Leptin (μg/L)||8.8 ± 4.4||4.2 ± 3.1||6.1 ± 4.2||6.9 ± 3.6||3.1 ± 0.9||5.5 ± 3.45|
Forty (49%) of the patients taking VPA and 25 (49%) of the control subjects were obese. The mean BMI was similar in the patients and the control subjects. The obese patients had higher WHR (0.92 ± 0.08) than the lean patients (0.86 ± 0.07; p < 0.001), consistent with centripetal obesity. Serum insulin levels were higher in the obese (11.3 ± 6.0 mU/L) than in the lean patients (7.4 ± 5.4 mU/L; p < 0.001) taking VPA.
Serum leptin levels were higher in the obese patients (8.8 ± 6.59 μg/L) than in the lean patients (4.2 ± 3.1 μg/L; p < 0.01), and they also were higher in the obese control subjects than in the lean control subjects (6.9 ± 3.6 μg/L and 3.1 ± 0.9 μg/L, respectively; p < 0.01). Serum leptin levels were higher in the female patients (20.0 ± 13.7 μg/L) than in the male patients (6.1 ± 4.2 μg/L; p < 0.001).
The mean serum insulin levels were higher in VPA-treated patients (9.2 ± 6.0 mU/L) than in control subjects (5.0 ± 4.5 mU/L; p < 0.001) despite similar BMI values, when all subjects were included in the comparison. The obese male patients taking VPA had higher serum insulin levels than the obese male control subjects (p < 0.001), although these two groups had BMI values of the same magnitude. Similarly a significant difference was seen in serum insulin levels between the obese female patients treated with VPA and the obese control women (p < 0.01). Furthermore, serum insulin levels were higher both in lean male (p < 0.05) and lean female (p < 0.01) patients compared with the lean control subjects of same sex and similar BMI values.
There was no statistically significant difference in serum leptin levels between the patients treated with VPA and the control subjects (Table 2).
Half of the VPA-treated patients (49%) in the present series were observed to be obese. Moreover, both lean and obese patients taking VPA had higher fasting serum insulin concentrations than did the control subjects of same sex with a similar degree of obesity. Conversely, serum leptin concentrations did not differ between patients taking VPA for epilepsy and the control subjects. The frequency of VPA-related obesity observed in the present study was close to that reported in earlier surveys (2–5). VPA-related obesity was centripetal (i.e., the WHR was high), which is consistent with our previous report (5). Hyperinsulinemia together with visceral obesity implies that obese VPA-treated patients may be insulin resistant.
The pathogenic mechanisms underlying VPA-related weight gain have remained open. It has been suggested that VPA inhibits β-oxidation (17), competes with palmitate for binding to albumin (18), or is involved in the regulation of energy expenditure and appetite in the hypothalamus through γ-aminobutyric acid (GABA) (19). However, the results of the present and previous studies show that VPA-related obesity is associated with hyperinsulinemia (4,5).
The mechanisms leading to hyperinsulinemia in patients with VPA-related obesity have not been defined. VPA could directly stimulate the secretion of insulin from the pancreatic β cells. It also has been suggested that VPA, a fatty acid, could increase the flow of fatty acids into the liver, enhancing gluconeogenesis, eventually resulting in stimulated insulin secretion (6,18). Because VPA-related obesity is usually visceral and the women with VPA-related obesity have hyperandrogenism and PCOs, hyperinsulinemia during VPA treatment most likely reflects insulin resistance. Obese persons in general have insulin resistance leading to hyperinsulinemia in their peripheral circulation. However, VPA-treated obese patients had higher insulin levels than did untreated obese persons, indicating that the insulin resistance is more severe in VPA-related obesity than in obesity in general.
No increase in fasting serum insulin levels was observed after starting VPA medication in a short-term longitudinal study in eight patients. However, none of those patients gained a significant amount of weight during the short follow-up period (20). Conversely, in a recent report, serum insulin concentrations decreased to normal 2 months after replacing VPA with lamotrigine (LTG) in women with VPA-related PCOs and hyperandrogenism, whereas the weight loss in these women was gradual and progressive during the first 12 months after tapering off VPA. Furthermore, in the present study, serum insulin levels were higher in the lean patients taking VPA than in the lean control subjects. This implies that the hyperinsulinemia seen in obese persons taking VPA is not merely a consequence of insulin resistance induced by weight gain. It is possible that the development of insulin resistance may be one of the factors leading to weight gain in some patients (5).
In the present study both the obese patients taking VPA and the obese control subjects had higher serum concentrations of leptin than did the lean subjects. Furthermore, the female subjects had higher serum leptin levels than the male subjects among both VPA-treated patients and control subjects. These data are consistent with results previously reported in human subjects (12). Serum leptin levels did not differ between the obese patients taking VPA and the obese control subjects or between the lean patients and the lean control subjects, despite the higher serum insulin levels in VPA-treated patients. This is in accordance with the reports indicating that insulin may have a trophic effect on the adipocytes, eventually increasing the serum leptin levels, rather than having a direct effect on OB gene expression (13,21). This suggests that VPA itself does not have any effect on the leptin-associated regulation of appetite and energy expenditure, but actually the leptin activation is similar in obese VPA-treated subjects to that seen in otherwise obese subjects. This has been observed earlier in female patients taking VPA in a longitudinal study (14).
In summary, the high serum insulin levels and centripetal obesity observed in the obese VPA-treated patients suggest that VPA-related obesity is associated with the development of insulin resistance. Both insulin resistance (6) and visceral obesity (22) are independent risk factors for coronary heart disease. The findings do not provide evidence for an independent role of leptin in the pathogenesis of VPA-related obesity.
Acknowledgment: We are indebted to Mrs. Anja Heikkinen and Mrs. Sirpa Anttila for performing the laboratory analyses.