Effects of new antiepileptic drugs on circulatory markers for vascular risk in patients with newly diagnosed epilepsy



Although it is well documented that long-term therapy with older antiepileptic drugs (AEDs) leads to an increase in risk for atherosclerosis, there has been only limited information regarding the vascular risk in patients who are treated with new AEDs. We therefore conducted a prospective longitudinal study to assess the potential effects of new AEDs on the circulatory markers for vascular risk in patients with newly diagnosed epilepsy. We recruited adult patients with epilepsy who began to receive monotherapy with one of the new AEDs, including levetiracetam (LEV), oxcarbazepine (OXC), and topiramate (TPM). Circulatory markers of vascular risk were measured twice before and after 6 months of AED monotherapy. A total of 109 patients completed the study (LEV, n = 40; OXC, n = 40; TPM, n = 29). Six months of monotherapy resulted in significant increases in low-density lipoprotein cholesterol (LEV, from 90.2 to 98.5 mg/dl, 9.2% increase, p = 0.025; OXC, from 96.5 to 103.2 mg/dl, 7.0% increase, p = 0.049), homocysteine (LEV, from 7.9 to 10.4 μm, 31.6% increase, p = 0.001; OXC, from 8.7 to 11.5 μm, 32.2% increase, p < 0.001; TPM, from 8.3 to 12.3 μm, 48.2% increase, p < 0.001), apolipoprotein B (LEV, from 63.6 to 77.4 mg/dl, 21.7% increase; OXC, from 67.0 to 83.2 mg/dl, 24.2% increase; TPM, from 66.7 to 84.4 mg/dl, 26.5% increase; all p < 0.001), and apolipoprotein B/apolipoprotein A1 ratio (LEV, from 0.51 to 0.61, 19.6% increase; OXC, from 0.52 to 0.67, 28.8% increase; TPM, from 0.50 to 0.67, 34.0% increase; all p < 0.001). Serum apolipoprotein A1 and folate were significantly decreased in TPM (from 139.1 to 132.1 mg/dl, 5.0% decrease, p = 0.014) and OXC (from 8.1 to 6.4 ng/ml, 21.0% decrease, p = 0.046) groups, respectively. There were no significant changes in total cholesterol, triglyceride, high-density lipoprotein cholesterol, lipoprotein(a), and vitamin B12 in all three groups. Our findings suggest that treatment with some new AEDs might be associated with alterations in circulatory markers of vascular risk, which could contribute to the acceleration of atherosclerosis and increased risk of vascular diseases.

Recent epidemiologic studies showed significant associations between epilepsy and comorbid cardiovascular or cerebrovascular diseases (Ding et al., 2006). There is also emerging evidence that prolonged use of some antiepileptic drugs (AEDs) is related to increased risk for atherosclerosis (Lopinto-Khoury & Mintzer, 2010). Older AEDs such as carbamazepine, phenytoin, and phenobarbital are recognized to accelerate atherosclerosis, since these AEDs are potent inducers of the hepatic cytochrome P450 (CYP) system that is extensively involved in the synthesis and metabolism of cholesterols. The inducing AEDs are also implicated in the predisposition to atherosclerosis by altering other markers of vascular risk including homocysteine, folate, lipoprotein(a) [Lp(a)], and C-reactive protein (CRP; Hamed et al., 2007; Mintzer et al., 2009; Chuang et al., 2012). Because of known detrimental effects of older AEDs on atherosclerosis, new AEDs are often suggested as first-line agents in the treatment of epilepsy. Because new AEDs are less likely to induce hepatic enzymes, it seems plausible that their effects on increased vascular risk may be lower than those of older AEDs (Mintzer & Mattson, 2009). However, limited data are available to date, making the harmful effects of new AEDs on vascular risk largely unknown.

In the present study, we sought to determine the potential effects of new AEDs on vascular risk. Specifically, we evaluate whether 6 months of monotherapy with levetiracetam (LEV), oxcarbazepine (OXC), or topiramate (TPM) leads to alterations in circulatory markers of vascular risk in patients with newly diagnosed epilepsy.



In four tertiary referral epilepsy centers, adult patients with newly diagnosed epilepsy (≥18 and ≤60 years old), who began to receive monotherapy with LEV, OXC, or TPM, were prospectively recruited from March 2010 to September 2011. The choice of AEDs was based on each clinician's decision. AEDs were administered at a starting dose of 500 mg/day (LEV), 300 mg/day (OXC), and 50 mg/day (TPM), and were then increased slowly to reach the maximum tolerable dose that provided good to excellent seizure control. The study was completed after a 6-month period of monotherapy with one of three AEDs. Patients were not eligible if they had a history of exposure to any AEDs, cerebrovascular/cardiovascular disease, diabetes mellitus, hypertension, chronic renal/hepatic disease, chronic alcohol consumption, or concomitant administration of vitamin supplements and lipid-lowering agents. Patients were excluded from the analysis in which other AED was added as polytherapy or initial AED was changed to another because of inadequate seizure control or serious adverse events of the drugs. Patients who had not taken AED for >2 weeks during the study period were further excluded. The local ethics committees approved this study, and all participants gave written informed consent.

Laboratory test and statistical analysis

For all participants, venous blood samples were collected twice at baseline and at the end of 6-month treatment, after an overnight fasting. Laboratory tests for vascular risk markers included total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), non-HDL cholesterol (non-HDL-C, calculated as TC minus HDL-C), Lp(a), homocysteine, folate, vitamin B12, apolipoprotein B (apoB), and apolipoprotein A1 (apoA1).

Patients were divided into three groups according to AED used. Between-group differences in demographics and clinical data were tested using analysis of variance (ANOVA), chi-square test, or Fisher's exact test. Paired t-test was used to assess changes in vascular risk markers before and after treatment within each group. ANOVA was then used to assess differences in AED-induced changes in vascular risk markers among the groups. Post hoc test with Bonferroni correction was used for pairwise comparisons if necessary.


A total of 109 patients completed the study, and consisted of LEV (n = 40), OXC (n = 40), and TPM groups (n = 29). Demographics, clinical data, baseline, and follow-up values as well as summary results for changes in vascular risk markers are presented in Table 1. The three groups did not differ in age, sex, and body mass index. There was a difference in epilepsy syndrome among the three groups (LEV, 26 partial vs. 14 generalized epilepsy; OXC, 40 partial vs. 0 generalized epilepsy; TPM, 23 partial vs. 6 generalized epilepsy; p < 0.001). Because the choice of AEDs was based on each clinician's decision, OXC was prescribed exclusively to patients with partial epilepsy. Mean daily dose for each AED was 1,207 mg (range, 500–3,000) in LEV group, 957 mg (range, 600–1,200) in OXC group, and 157 mg (range, 50–400) in TPM group.

Table 1. Demographics, clinical data, and comparison of changes in vascular risk markers in three treatment groups
 LEV (n = 40)OXC (n = 40)TPM (n = 29)p-Value (between groups)
Age (years)35.3 ± 2.437.9 ± 2.339.4 ± 1.40.468
BMI (kg/m2)22.1 ± 0.523.3 ± 0.423.3 ± 0.70.193
Sex (female:male)21:1914:2614:150.284
Epilepsy syndrome (partial:generalized)26:1440:023:6<0.001
AED dose (mg/day)1,207 ± 61 (range, 500–3,000)957 ± 39 (range, 600–1,200)157 ± 14 (range, 50–400) 
Measured variableBaseline valueFollow-up valueBaseline valueFollow-up valueBaseline valueFollow-up valuep-Value for baseline value (between groups)
TC (134–270 mg/dl)156.5 ± 5.1162.1 ± 5.5165.2 ± 4.5171.0 ± 5.1168.7 ± 7.3171.9 ± 7.60.277
HDL-C (35–70 mg/dl)47.9 ± 1.948.3 ± 1.649.9 ± 2.351.4 ± 2.150.0 ± 3.155.9 ± 4.00.759
TG (49–284 mg/dl)89.4 ± 8.1106.6 ± 10.5111.4 ± 13.1113.9 ± 14.7115.7 ± 15.0119.0 ± 14.80.249
LDL-C (81–189 mg/dl)90.2 ± 4.098.5 ± 4.696.5 ± 3.8103.2 ± 4.395.1 ± 6.497.8 ± 6.30.568
Non-HDL-C (mg/dl)108.7 ± 4.8113.8 ± 5.3115.4 ± 4.8119.6 ± 5.0118.7 ± 7.6116.0 ± 6.70.442
Homocysteine (5–15 μm)7.9 ± 0.510.4 ± 0.88.7 ± 0.511.5 ± 1.08.3 ± 0.612.3 ± 1.10.490
ApoA1 (115–224 mg/dl)127.4 ± 4.1129.7 ± 4.1133.6 ± 3.2128.8 ± 3.3139.1 ± 5.0132.1 ± 5.10.137
ApoB (60–130 mg/dl)63.6 ± 2.777.4 ± 3.867.0 ± 2.483.2 ± 3.466.7 ± 3.984.4 ± 4.90.640
ApoB/ApoA10.51 ± 0.020.61 ± 0.030.52 ± 0.030.67 ± 0.040.50 ± 0.040.67 ± 0.050.940
Lp(a) (0–30 mg/dl)14.0 ± 2.715.7 ± 3.018.8 ± 3.419.1 ± 3.714.4 ± 3.317.0 ± 3.80.475
Vitamin B12 (160–970 pg/ml)758.3 ± 50.2718.6 ± 46.1752.8 ± 47.8696.2 ± 54.7760.3 ± 67.9680.7 ± 44.00.995
Folate (1.5–16.9 ng/ml)6.9 ± 0.96.8 ± 1.08.1 ± 1.46.4 ± 0.910.2 ± 2.28.7 ± 1.50.318
Measured variableChange in variablep-Value (within group)Change in variablep-Value (within group)Change in variablep-Value (within group)p-Value (between groups)
  1. LEV, levetiracetam; OXC, oxcarbazepine; TPM, topiramate; BMI, body mass index; AED, antiepileptic drug; TC, total cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride; LDL-C, low-density lipoprotein cholesterol; Non-HDL-C, non-HDL cholesterol; apoA1, apolipoprotein A1; apoB, apolipoprotein B; Lp(a), lipoprotein(a).

  2. All values are given as means ± SE or number.

  3. The values in the parentheses of the first column indicate reference ranges for each parameter.

  4. a

    TPM versus LEV, p = 0.046.

  5. Statistical analyses were conducted using SPSS software (version 19.0; SPSS Inc., Chicago, IL, U.S.A.). A p-value < 0.05 is considered statistically significant in all analyses.

TC (mg/dl)5.5 ± 3.10.0805.7 ± 3.10.0713.2 ± 4.90.5180.877
HDL-C (mg/dl)0.4 ± 1.10.6931.5 ± 1.40.2875.9 ± 3.20.0760.116
TG (mg/dl)17.2 ± 8.90.0602.5 ± 14.00.8593.3 ± 13.90.8130.623
LDL-C (mg/dl)8.3 ± 3.60.0256.7 ± 3.30.0492.7 ± 4.30.5270.574
Non-HDL-C (mg/dl)5.1 ± 3.40.1344.3 ± 3.40.220−2.6 ± 5.20.6150.348
Homocysteine (μm)2.5 ± 0.70.0012.8 ± 0.6<0.0014.0 ± 0.8<0.0010.301
ApoA1 (mg/dl)2.4 ± 2.20.300−4.8 ± 2.80.091−7.0 ± 2.70.0140.031a
ApoB (mg/dl)13.9 ± 2.6<0.00116.2 ± 2.0<0.00117.7 ± 2.3<0.0010.529
ApoB/ApoA10.10 ± 0.02<0.0010.15 ± 0.03<0.0010.17 ± 0.02<0.0010.127
Lp(a) (mg/dl)1.7 ± 1.00.1200.4 ± 2.00.8622.6 ± 1.40.0870.640
Vitamin B12 (pg/ml)−39.7 ± 46.20.395−56.6 ± 42.00.186−79.7 ± 61.00.2020.855
Folate (ng/ml)−0.1 ± 0.50.840−1.7 ± 0.80.046−1.5 ± 1.60.3620.443

There were no differences in baseline values for all vascular risk markers among the groups. LDL-C levels were increased in LEV (p = 0.025) and OXC groups (p = 0.049) with 6 months of monotherapy, whereas TC, HDL-C, TG, and non-HDL-C levels were unchanged. A 6-month period of monotherapy increased homocysteine levels in all groups (p ≤ 0.001), decreased apoA1 levels in TPM group (p = 0.014), and increased apoB levels in all groups (p < 0.001). As expected, apoB/apoA1 ratio was increased in all groups (p < 0.001). There were no changes in Lp(a) and vitamin B12 levels in all groups. Folate levels were decreased in the OXC group (p = 0.046), but not in LEV and TPM groups. ANOVA showed no differences in changes in vascular risk markers among three groups, except for reduced apoA1 in TPM versus LEV (p = 0.046).


This prospective longitudinal study shows that monotherapy with LEV, OXC, or TPM resulted in overall increased atherogenic markers such as apoB, apoB/apoA1 ratio, homocysteine, and LDL-C.

Patients with epilepsy who are treated with older AEDs have a higher chance of increased levels of TC, TG, HDL-C and LDL-C, although the results are not uniform across the studies (Mintzer & Mattson, 2009; Cheng et al., 2010; Lopinto-Khoury & Mintzer, 2010). The mechanism by which older AEDs cause hypercholesterolemia is presumed to be mediated by induction of CYP enzymes. It is therefore unexpected that LDL-C levels were mildly but significantly increased in patients under OXC and LEV monotherapy, since OXC and LEV are known to have mild and no induction of CYP enzymes, respectively. More interesting is that all three groups showed marked increases in apoB levels and apoB/ApoA1 ratio, which can also provide sensitive measures of atherogenicity and act as indicators for assessment of vascular risk (Davidson, 2009).

Homocysteine, a nonessential amino acid with prothrombotic properties, has been implicated as an important risk factor for vascular diseases (Humphrey et al., 2008). It is frequently suggested that older AEDs, mostly being strong CYP inducers, increase homocysteine by way of deficient cofactors for homocysteine metabolism, such as folate and vitamin B12. However, only a few studies have addressed the issue of whether new AEDs cause hyperhomocysteinemia. Our study showed significant increases in homocysteine levels in patients under monotherapy with OXC and TPM, in accordance with recent cross-sectional studies that found increased homocysteine in epilepsy patients on OXC and TPM (Belcastro et al., 2010; Linnebank et al., 2011). Because OXC and TPM are mild inducers of hepatic enzymes, especially when used in higher doses, it seems plausible that these AEDs can lead to hyperhomocysteinemia. The mechanism by which OXC and TPM cause hyperhomocysteinemia is currently unknown; however, folate deficiency induced by AEDs has been suggested as a possible mechanism (Belcastro et al., 2010). In our study, folate was found to be marginally decreased in the OXC group but not in TPM group, consistent with a recent cross-sectional investigation (Linnebank et al., 2011). Moreover, an unexpected but intriguing finding is that treatment with LEV, a noninducing AED, also resulted in a significant elevation of homocysteine with unchanged folate and vitamin B12 levels. Our finding seems to contrast with those of previous studies showing that most patients under monotherapy with LEV had normal homocysteine levels (Belcastro et al., 2010; Linnebank et al., 2011). A possible explanation for this difference is that hyperhomocysteinemia was defined as the level above the reference range in their cross-sectional studies, whereas significant increases in homocysteine levels were mostly within the physiologic range in our longitudinal study. In addition, when comparing our study with recent works (Belcastro et al., 2010; Linnebank et al., 2011), we consider that relatively low doses and short-term use of AEDs in our patient population might account for the differences in increased homocysteine levels between the studies. Taken together, we speculate that increased atherogenicity caused by new AEDs is not entirely dependent on the CYP system or deficient cofactors for homocysteine metabolism but on additional mechanisms that remain to be elucidated.

An advantage of our study is to use a prospective longitudinal design. The large majority of previous studies investigating relationships between AEDs and atherosclerosis used a cross-sectional design. Clinical complexities and considerable confounders within the patient population under cross-sectional study can obscure the pure effects of AEDs on vascular risk markers. This difference might be responsible for some inconsistencies across the studies. We also acknowledge several limitations in our study. First, the small number of patients recruited to each AED group and lack of control subjects are insufficient to draw a firm conclusion. Second, we did not measure some important markers such as C-reactive protein (CRP) and uric acid as well as carotid artery intima media thickness, an established surrogate marker for atherosclerosis. Third, a possible interaction between the epilepsy syndromic category (partial vs. generalized) and metabolic changes caused by each AED was not contemplated in our study (Morrell et al., 2002). Finally, our study investigated relatively short-term effects of AEDs, and therefore, the changes we observed may be merely temporary in nature. Future study with a complete constellation of circulatory markers and a long-term use of AEDs would be needed to corroborate our findings.

In summary, our findings suggest that treatment with some new AEDs might be associated with alterations in circulatory markers of vascular risk, which could contribute to the acceleration of atherosclerosis and increased risk of vascular diseases. Because changes in vascular risk markers were small and mostly within the physiologic range, caution should be taken when interpreting our results with regard to increased atherogenicity.


The authors are grateful to the participants for taking part in the present study. This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology of Republic of Korea (grant number 2012001850 for DWK, 20110005418 for JHK).


None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.