IF: 0.644

Meta-analysis of the Relationship Between Plasma Homocysteine Levels and Carbamazepine Monotherapy in Patients with Epilepsy


Yin Ying 1 , Hongchun Li 2 , Jiye Jin 3 , Xuliang Lu 4 , Pinpin Feng 1 , Yunfeng Sun 1 , *


1 Department of Pharmacy, Tongde Hospital of Zhejiang Province, Hangzhou, Zhejiang, 310012, China

2 Department of Gerontology, Tongde Hospital of Zhejiang Province, Hangzhou, Zhejiang, 310012, China

3 Department of Medical Rehabilitation, Tongde Hospital of Zhejiang Province, Hangzhou, Zhejiang, 310012, China

4 Department of Pharmacy, Fuzhou General Hospital of Nanjing Military Region, Fuzhou, Fujian 350025, China

How to Cite: Ying Y, Li H, Jin J, Lu X, Feng P, et al. Meta-analysis of the Relationship Between Plasma Homocysteine Levels and Carbamazepine Monotherapy in Patients with Epilepsy, Iran Red Crescent Med J. 2017 ; 19(11):e14182. doi: 10.5812/ircmj.14182.


Iranian Red Crescent Medical Journal: 19 (11); e14182
Published Online: November 26, 2017
Article Type: Review Article
Received: April 12, 2017
Revised: May 28, 2017
Accepted: July 29, 2017




Background: Numerous studies have shown that long- term treatment with anticonvulsants may be an important risk factor for the onset of atherosclerosis, or worsening of its symptoms.

Context: The present study aimed at investigating the relationship between carbamazepine (CBZ) monotherapy and plasma homocysteine (Hcy) levels in patients with epilepsy.

Evidence Acquisition: Studies concerning homocysteine levels in carbamazepine monotherapy patients with epilepsy, which were published in VIP, Wanfang, CNKI, Cochrance Library, PubMed, Web of Science, and EMBASE, were included in March 2016. The quality of the controlled clinical trials (CCT) selected for this study was assessed using the Newcastle-Ottawa scale (NOS), and the relevant data of the included studies were extracted through RevMan5.2 software.

Results: In this meta-analysis, 22 eligible studies were enrolled including 9 Chinese and 13 English studies. The study included a total of 1604 cases including 575 cases of patient group and 1029 control group cases. Results of the meta-analysis showed that plasma homocysteine level in patients with epilepsy with long- term treatment of CBZ was significantly higher than that in the healthy control group [SMD = 1.55, 95% CI: [1.09, 2.01], P < 0.00001]. Moreover, there was significant heterogeneity in the estimates according to I2 test (P < 0.00001; I2 = 93%). Further subgroup analysis showed that no significant difference was present when the study participants were grouped by region and age, however, the risk of heterogeneity in the West Asian group (I2 = 58%, P = 0.07) was diminished when compared with overall groups (I2 = 93%, P < 0.00001). The results of sensitivity analysis by Stata12.0 showed good stability. The funnel plot method and Begg method were used to detect publication bias, and the results showed a substantially symmetrical funnel plot, Pr > |Z|= 0.091 > 0.05 (no statistical significance), suggesting no significant publication bias in the study. Loss factor of safety (Nfs) 0.05 equaled 7269.16 (P = 0.05), meaning that addition of about 7269 negative results were required to overthrow the conclusion of this study.

Conclusions: The seizures significantly increased plasma homocysteine levels in the patients, thus it is appropriate to add folic acid, vitamin B12, and vitamin B6 to reduce the seizures. Moreover, homocysteine may be beneficial for those patients with epilepsy who take carbamazepine.


Epilepsy Carbamazepine Homocysteine Meta-Analysis

Copyright © 2017, Iranian Red Crescent Medical Journal. This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/) which permits copy and redistribute the material just in noncommercial usages, provided the original work is properly cited.

1. Background

Epilepsy is a clinical syndrome with transient brain dysfunction, which is caused by sudden and abnormal discharges of brain neurons, induced by a variety of reasons. This syndrome is common in clinical neurological diseases, with the characteristics of high incidence, long duration, and easy relapse, etc. The global number of epileptic patients is about 50 million; the domestic epidemiological data show that the clinical prevalence of epilepsy is about 7% and the incidence is about 25,100,000 to 45,100,000. There are about 10 million patients with epilepsy in China, and the number is increasing at the rate of 450,000 per year (1). Sener et al. (2) found that epilepsy itself will not increase the homocysteine level. However, a current study found that long-term treatment of antiepileptic drugs would cause metabolic abnormalities in patients, causing the increase of homocysteine (Hcy) concentration (3), and they further indicated that about 10% to 40% of patients with epilepsy might have hyperhomocysteine (Hhcy) (4) concurrently. Eikelboom et al. (5) showed that Hhcy caused vascular endothelial dysfunction, thrombosis, cholesterol, disorder of triglyceride synthesis and metabolism, and the activation of monocytes, which all can contribute to sclerosing disease including the occurrence and progression of glomerulosclerosis. In the recent years, Hhcy has been considered as an independent risk factor for diseases (6) such as atherosclerosis, cardiovascular, and cerebrovascular diseases. Therefore, antiepileptic drugs may be one of the causes to increase the risk of the occurrence of these diseases. Antiepileptic drugs are now widely used by patients with epilepsy. As one of the first-line antiepileptic drugs, carbamazepine is the first choice and most commonly used in curing complex partial seizures (7). There is controversy about the effect of carbamazepine (CBZ) on the level of Hcy, and there are few related studies on this topic in the world. The underling mechanism is not clear, so further studies are needed to reach a more accurate and comprehensive conclusion. Here, we, used meta-analysis to investigate the relationship between carbamazepine monotherapy and plasma homocysteine levels to provide the basis for clinical rational drug use.

2. Methods

2.1. Search Strategy

We retrieved studies from databases of VIP, Wanfang Database, CNKI and Cochrance Library, National Library of Medicine (PubMed), Chinese Science Citation Database (Web of Science), Excerpta Medica Database (EMBASE) and other databases in March 2016. We used the words “carbamazepine”, “homocysteine”, “epilepsy”, and “epilep*” as search terms. We also looked up the bibliographies of the retrieved literatures one by one to find eligible studies. The language of the literatures was limited to Chinese and English.

2.2. Inclusion and Exclusion Criteria and Outcome Indicator

Inclusion Criteria were as follow: (1) controlled clinical trial studies; (2) the patient group consisting of patients with epilepsy who received carbamazepine monotherapy, while the control group consisted of healthy volunteers not taking antiepileptic drugs; and (3) only Chinese and English studies were included in the study. Exclusion Criteria were as follow: (1) non-clinical trial studies; (2) non-rigorous experimental design such as significant proportion imbalance between the patient group and the control group; (3) studies with incomplete summarization and data as well as repeated data; and (4) patients with congenital metabolic defects of Hcy, folic acid, and vitamin B; (5) patients with hypertension, diabetes, coronary heart disease, stroke, liver and kidney diseases, etc.; (6) those taking B vitamins as adjunctive drugs in drug treatment process; (7) smokers, long-term alcoholics and vegetarians; (8) those with tumor and recent surgery trauma;(9) patients with peripheral vascular disease or peripheral vascular embolism diseases and endocrine diseases such as hypothyroidism; (10) patients with mental illnesses such as depression and schizophrenia; and (11) those who were taking thiazide diuretics or azathioprine and other similar drugs (8). Interventions: The patient group consisted of epileptic patients receiving carbamazepine monotherapy, while the control group consisted of healthy individuals who received no pharmacological intervention. Outcome Indicator: The plasma homocysteine levels in patients with epilepsy receiving carbamazepine monotherapy or the healthy group without pharmacological intervention in the same period; the core data were continuous variables [presented as mean ± standard deviation].

2.3. Data Screening and Quality Evaluation
2.3.1. Data Screening

Two independent researchers separately screened the studies according to the inclusion and exclusion criteria. (1) Initial Screening: previewing the title and abstract of the studies; excluding those that did not match the research content; summarization; animal or in vitro trials; report of adverse reactions; repeatability and so on.; (2) reading the full text of the studies that might have been qualified to be included in the study and analyzing them one by one to determine whether they met the inclusion criteri; (3) Objection Handling: for any objection or disagreements arising in the screening process, the 2 researchers held discussion.

2.3.2. Quality Evaluation

The Newcastle-Ottawa Scale (NOS) was used to evaluate the quality of each eligible article including the appropriateness of the case, the representativeness of the case, the selection of the control, and the determination of the control. The same method was used by the 2 researchers to determine the exposure factors of the patient group and control group, and the no response rate. The quality evaluation criteria were as follow: (1) the section of the patient group and the healthy control group had the highest score of 4 points; (2) the comparability of the patient group and the control control group had the highest score of 2 points; and (3) the exposure factor measurement had the highest score of 3 points. The highest total score was 9 points; those studies with > 6 points were of high-quality studies. If there were any disagreements, the 2 researchers settled them through discussion.

2.4. Data Extraction

Two researchers independently extracted and recorded the relevant data; if there was any disagreement, the 2 researchers would have reached a consensus before extraction. The data presented included the first author, time of publication, the region of the study, the information of the patient group, and the control group (carbamazepine dose and duration of treatment, number of participants, and the plasma homocysteine levels of the 2 groups).

2.5. Statistical Analysis

Statistical analysis was performed using RevMan 5.2 software based on the data provided in the eligible studies. The heterogeneity test of the findings was performed using χ2 test and I2 test. When I2<50%, it presents homogeneity in each study, and thus the fixed-effect model should be used for combined statistical analysis; when I2 ≥ 50%, it indicates that there is heterogeneity among the studies, and the random-effect model shall be used for statistical analysis. The measuring results of this study are continuous variables, which calculated the standardized mean difference (SMD) and 95% confidence interval (95% CI) to create the forest plot. If there is heterogeneity risk, sensitivity analysis and other methods shall be employed to analyze the causes of heterogeneity, and the subgroup analysis should be undertaken. We used funnel plot method, Begg method, or Fail-safe number to conduct publication bias analysis on the Meta-analysis.

3. Results

3.1. Literature Retrieval Results

According to the retrieval strategy (Figure 1), we retrieved 722 Chinese and English studies including 94 Chinese (67 from CNKI, 16 from VIP, and 11 from Wanfang) and 628 English studies (120 from PubMed, 75 from Web of Science, and 433 from EMBASE). There were 657 studies left after excluding the duplicate articles. After rejecting those that did not meet the inclusion criteria, 79 studies were left. Eventually, by reading their full texts, we selected 22 studies (9 in Chinese and 13 in English). See Table 1 for the basic parameter data included in the study.

Flowchart Depicts the Process of Selection of Studies to be Included in This Analysis
Figure 1. Flowchart Depicts the Process of Selection of Studies to be Included in This Analysis
Table 1. Basic Parameter Data Included in the Study
StudyYearEthnicityDurationCBZ GroupControl Group
DoseNHcy, μmol/LNHcy, μmol/L
Schwaninger et al. (9)1999European≥ 1 moUnknown1521.7 ± 9.6359.5 ± 0.5
Verrotti et al. (10)2000EuropeanAfter 1 yUnknown2814.1 ± 8.1637.9 ± 4.5
Apeland et al. (11)2001European6 moUnknown4213.8 ± 1.6429.3 ± 0.4
Achilleas et al. (12)2006European20 wunknown207.6 ± 1.71726.4 ± 1.6
Kurul et al. (13)2007West-Asian> 1 yunknown118.13 ± 5.28107.66 ± 2.34
Vurucu et al. (14)2008West-Asian≥ 6 mounknown296.38 ± 1.73625.52 ± 2.53
Mintzer et al. (15)2009European6 moUnknown1511.1 ± 4.21611.5 ± 5.3
Chuang et al. (16)2011East-Asian> 2 yUnknown4113.31 ± 8.41609.41 ± 2.65
El-Farahaty et al. (17)2014North-Africa> 2 y10-17 mg/kg/d148.63 ± 2.99342.7 ± 0.5
Karabiber (3)2002West-Asian> 12 moUnknown3016.0 ± 13.1299.2 ± 2.7
Yoo and Yung (18)1999East-Asian< 5 yUnknown2312.7 ± 1.71037.9 ± 1.2
Sener et al. (2)2005West-Asian2 yUnknown1919.6 ± 21.41111.5 ± 11.4
Eldeen et al. (19)2012North-Africa1 yUnknown85.5 ± 1.9305.5 ± 1.4
Xia (20)2012East-Asian6 mo0.6g/d3511.31 ± 2.66749.79 ± 2.23
Li et al. (21)2013East-Asian> 1 yUnknown4010.005 ± 2.9494406.755 ± 1.5726
Liu et al. (22)2012East-Asian2 mo0.3 g/d4524.5 ± 5.95010.0 ± 4.0
Ming (23)2014East-AsianOne CourseUnknown3221.92 ± 11.543212.74 ± 5.29
Xuefeng (24)2012East-Asian5 mo ~ 3 yUnknown1913.56 ± 4.34406.56 ± 2.11
Wang et al. (25)2007East-Asian2 mo ~ 13 yUnknown3021.1 ± 6.44010.20 ± 4.3
Caizhi et al. (26)2008East-Asian2 mo ~ 4.5 yUnknown3112.78 ± 3.06366.17 ± 1.34
Zhang (27)2013East-Asian≥ 32 wUnknown2322.64 ± 10.452010.30 ± 4.43
Shan et al. (28)2014East-Asian1 moUnknown2520.77 ± 5.233010.47 ± 2.43
3.2. Basic Features Included in the Study

We included 22 eligible studies in total, all of which were cohort studies containing 1,604 study participants, among which there were 575 patient groups and 1,029 healthy control groups. After conducting Newcastle-Ottawa quality evaluation on the included studies, we found that 10 of the studies scored 7 points in the quality evaluation and 12 scored 6 points.

3.3. Meta-analysis Results
3.3.1. Heterogeneity Test

A heterogeneity test was performed among the 22 studies included in the final study. Heterogeneity among each clinical trial was I2 = 93%, P < 0.00001, which was statistically analyzed using the random-effect model. The results revealed that the overall difference was statistically significant. The plasma homocysteine level of the patient group receiving carbamazepine monotherapy was higher than that of the healthy control group (overall effect SMD = 1.55, 95%, CI: [1.09, 2.01], P < 0.00001; Figure 2). The results of the analysis were different. Heterogeneity existed among various studies, thus it is necessary to conduct subgroup analysis to explore the source of heterogeneity.

Forest Plot of Changes of Plasma Homocysteine Levels in Patients with Epilepsy Receiving Carbamazepine Monotherapy
Figure 2. Forest Plot of Changes of Plasma Homocysteine Levels in Patients with Epilepsy Receiving Carbamazepine Monotherapy
3.3.2. Subgroup Analysis

The studies were divided into 4 subgroups (Figure 3) according to regional difference including Europe, North Africa, West Asia and East Asia. They could also be divided into 2 subgroups according to age difference of the participants; namely, < 18-year-old and ≥ 18-year-old (Figure 4). Subgroup analysis revealed a lower risk of heterogeneity in the West Asian group (I2 = 58%, P = 0.07) than in the overall groups (I2 = 93%, P < 0.00001) (Table 2).

Subgroups Grouped by Regions of the Studies
Figure 3. Subgroups Grouped by Regions of the Studies
Subgroups Grouped by Age of the Subjects
Figure 4. Subgroups Grouped by Age of the Subjects
Table 2. Subgroup Analysis Results Included in the Study
SubgroupNumber of StudyStandardized Mean Difference95%CIHeterogeneity Test, %P Value
European54.121.61 - 6.6294< 0.00001
West-Asian117.415.30 - 9.52580.07
East-Asian42.34-0.51 - 5.1895< 0.00001
North-Africa22.95-2.86 - 8.7797< 0.00001
Age, y
< 1883.451.34 - 5.5594< 0.00001
≥ 18146.785.12 - 8.4494< 0.00001
3.3.3. Sensitivity Analysis

We used Stata12.0 to perform sensitivity analysis to evaluate the stability of the meta-analysis. The results revealed that all study results were between the ceiling and floor, in other words, the overall trend of the SMD of the study was consistent with good stability (Figure 5).

3.3.4. Publication Bias Analysis

There are many methods to evaluate publication bias in the meta-analysis. The funnel plot method is a commonly used bias detection method. In this study, the funnel plot is plotted with the standard mean difference as abscissa and standard error as ordinate (Figure 6). The results showed that the funnel diagram is basically symmetrical. The Stata 12.0 Software was used for Begg test, the results of which showed the Pr > = 0.091 > 0.05, which indicated that the results were not statistically significant, and suggesting no significant publication bias in the study.

3.3.5. Fail-safe Number Analysis

When the results of the meta-analysis are statistically significant, the analysis on fail-safe number is used to calculate the minimum number of studies with negative results that is enough to overthrow the conclusion of the meta-analysis and to eliminate the possibility of bias. The greater the fail-safe number, the more stable the results of the meta-analysis, and the less likely the meta-analysis is biased (29). In this study, the fail-safe number (Nfs) (30) 0.05 is 7269.16 (P = 0.05), which needs about 7,269 negative results to overthrow this conclusion that plasma homocysteine levels of patients who received carbamazepine monotherapy significantly increased.

4. Discussion

Epilepsy is mainly treated by drug therapy. A vast majority of patients with epilepsy need long-term or even life-long treatment of antiepileptic drugs to control seizures. However, recent studies have found that long-term use of antiepileptic drugs could induce metabolic abnormalities in patients, leading to elevated Hcy concentration (3). Homocysteine is a non-essential amino acid containing sulfenyl, which is an important intermediate in the metabolic processes of methionine and cysteine. It has been found that homocysteine can damage the vascular endothelial cells, resulting in the dysfunction of cell selective permeability, the deposition of cholesterol and triglycerides in the vascular wall, the proliferation of smooth muscle cell, inhibition of cellular NO synthesis, promotion of cellular NO degradation, and the spasm of vascular smooth muscle. It will also affect the platelet aggregated coagulation factors’ activity through the promotion of thromboxane and prostacyclin formation, thereby promoting the occurrence of atherosclerosis and thrombosis formation, leading to cardiovascular and cerebrovascular diseases (31, 32). Moreover, Gu (33) and Castro et al. (34) found that plasma Hcy level was the main risk factor of atherosclerosis and cardiovascular diseases. As a first-line antiepileptic drug, carbamazepine is widely used to control various types of epileptic seizures, and thus long-term use of carbamazepine by patients with epilepsy may be one of the important risk factors of atherosclerosis and cardiovascular diseases.

Vitamin B12, vitamin B6, and folic acid are cofactors in the process of homocysteine metabolism. It is generally assumed that Hcy level and blood vitamin B12, vitamin B6 and folic acid levels were negatively correlated in nonlinearity, which means lower concentration of vitamin B12, vitamin B6, and folic acid caused higher homocysteine in serum (35). Consuelo et al. (36) found that taking a very low dose of folic acid caused a significant decrease in the plasma Hcy level of patients with epilepsy who had a history of long-term use of antiepileptic drugs. However, some studies found no effect of carbamazepine monotherapy on folate levels (11, 37). Tumer et al. (38) found that the mechanism of function of carbamazepine was different from enzyme-induced antiepileptic drugs which induce liver enzymes to accelerate the metabolism of folic acid. Instead, it interferes with intestinal folate absorption or directly affects the function of coenzyme in the process of folic acid metabolism, resulting in folic acid level decrease. Michael Linnebak et al. (39) have indicated that by appropriate oral intake of vitamin B12, vitamin B6, and folic acid and reducing methionine in diet, the blood Hcy level can be decreased.

Heterogeneity evaluation is a must in meta-analysis, which is of great importance to ensure the quality of meta-analysis (40). In the present study, heterogeneity analysis was performed using sensitivity analysis and subgroup analysis. The subgroup analysis based on the differences in age and region of the population revealed a lower heterogeneity risk in the West Asian group than in the overall population. The differences of vitamin B12, vitamin B6, and folic acid intake of people between different regions may be important factors in influencing the plasma Hcy level; however, a more accurate and comprehensive conclusion shall be further studied. Publication bias is one of the important factors affecting the quality of meta-analysis. There are many methods to evaluate publication bias in meta-analysis including funnel plot method, Begg method, Egger method, trim method, fail-safe number method, and so on. Currently, the most commonly used bias detection methods are funnel plot method and fail-safe number method, but these 2 methods cannot make quantitative detection on publication bias, while the Begg method and Egger method can be used to quantify the bias (29). In addition, considering the fact that the funnel plot method can only be fully effective when the number of studies is large and that the study only included 22 studies, the application of funnel plot was limited to a large extend (41). Therefore, the Begg method was used for quantify detection.

This study showed that plasma homocysteine levels in patients with epilepsy, who received carbamazepine monotherapy for long periods (≥ 1 month), were significantly higher than that in the healthy control group. To reduce the risk of carbamazepine-induced hyperhomocysteinemia, it is recommended that patients with long-term use of carbamazepine monitor their blood homocysteine levels regularly for early detection and intervention. Once an increase in Hcy level is found, appropriate folic acid, vitamin B12, and vitamin B6 supplements are necessary to reduce homocysteine level, which may be beneficial for those patients with epilepsy who take carbamazepine. However, because of the mere inclusion of published Chinese and English studies, the large difference of treatments among the groups, the unclear carbamazepine dosage, and other limitations, a more reliable and comprehensive conclusion needs to be verified based on more high-quality studies with larger sample sizes.



  • 1. Ding D, Hong Z, Wang WZ, Wu JZ, de Boer HM, Prilipko L, et al. Assessing the disease burden due to epilepsy by disability adjusted life year in rural China. Epilepsia. 2006;47(12):2032-7. doi: 10.1111/j.1528-1167.2006.00802.x. [PubMed: 17201700].
  • 2. Sener U, Zorlu Y, Karaguzel O, Ozdamar O, Coker I, Topbas M. Effects of common anti-epileptic drug monotherapy on serum levels of homocysteine, vitamin B12, folic acid and vitamin B6. Seizure. 2006;15(2):79-85. doi: 10.1016/j.seizure.2005.11.002. [PubMed: 16414291].
  • 3. Karabiber H, Sonmezgoz E, Ozerol E, Yakinci C, Otlu B, Yologlu S. Effects of valproate and carbamazepine on serum levels of homocysteine, vitamin B12, and folic acid. Brain Dev. 2003;25(2):113-5. [PubMed: 12581807].
  • 4. Khanna S, Kapoor P, K. Pillai K, Vohora D. Homocysteine in Neurological Disease: A Marker or a Cause?. CNS Neurol Disord Drug Targets. 2011;10(3):361-9. doi: 10.2174/187152711794653797.
  • 5. Eikelboom JW, Lonn E, Genest JJ, Hankey G, Yusuf S. Homocyst(e)ine and cardiovascular disease: a critical review of the epidemiologic evidence. Ann Intern Med. 1999;131(5):363-75. [PubMed: 10475890].
  • 6. WAN H, YU H, GUO H. Effects of Pravastatin on level of plasma homocysteine and degree of carotid atherosclerosis in patients with carotid atherosclerosis. J Clin Neurol. 2011;2:24.
  • 7. Jinheng L. Methods and Applications of Clinical Therapeutic Drug Monitoring. 2003.
  • 8. Hu XW, Qin SM, Li D, Hu LF, Liu CF. Elevated homocysteine levels in levodopa-treated idiopathic Parkinson's disease: a meta-analysis. Acta Neurol Scand. 2013;128(2):73-82. doi: 10.1111/ane.12106. [PubMed: 23432663].
  • 9. Schwaninger M, Ringleb P, Winter R, Kohl B, Fiehn W, Rieser PA, et al. Elevated plasma concentrations of homocysteine in antiepileptic drug treatment. Epilepsia. 1999;40(3):345-50. [PubMed: 10080517].
  • 10. Verrotti A, Pascarella R, Trotta D, Giuva T, Morgese G, Chiarelli F. Hyperhomocysteinemia in children treated with sodium valproate and carbamazepine. Epilepsy Res. 2000;41(3):253-7. [PubMed: 10962216].
  • 11. Apeland T, Mansoor MA, Strandjord RE. Antiepileptic drugs as independent predictors of plasma total homocysteine levels. Epilepsy Res. 2001;47(1):27-35.
  • 12. Attilakos A, Papakonstantinou E, Schulpis K, Voudris K, Katsarou E, Mastroyianni S, et al. Early effect of sodium valproate and carbamazepine monotherapy on homocysteine metabolism in children with epilepsy. Epilepsy Res. 2006;71(2-3):229-32. doi: 10.1016/j.eplepsyres.2006.06.015. [PubMed: 16889940].
  • 13. Kurul S, Unalp A, Yis U. Homocysteine levels in epileptic children receiving antiepileptic drugs. J Child Neurol. 2007;22(12):1389-92. doi: 10.1177/0883073807307081. [PubMed: 18174557].
  • 14. Vurucu S, Demirkaya E, Kul M, Unay B, Gul D, Akin R, et al. Evaluation of the relationship between C677T variants of methylenetetrahydrofolate reductase gene and hyperhomocysteinemia in children receiving antiepileptic drug therapy. Prog Neuropsychopharmacol Biol Psychiatry. 2008;32(3):844-8. doi: 10.1016/j.pnpbp.2007.12.018. [PubMed: 18234410].
  • 15. Mintzer S, Skidmore CT, Abidin CJ, Morales MC, Chervoneva I, Capuzzi DM, et al. Effects of antiepileptic drugs on lipids, homocysteine, and C-reactive protein. Ann Neurol. 2009;65(4):448-56. doi: 10.1002/ana.21615. [PubMed: 19296463].
  • 16. Chuang YC, Chuang HY, Lin TK, Chang CC, Lu CH, Chang WN, et al. Effects of long-term antiepileptic drug monotherapy on vascular risk factors and atherosclerosis. Epilepsia. 2012;53(1):120-8. doi: 10.1111/j.1528-1167.2011.03316.x. [PubMed: 22085257].
  • 17. El-Farahaty RM, El-Mitwalli A, Azzam H, Wasel Y, Elrakhawy MM, Hasaneen BM. Atherosclerotic effects of long-term old and new antiepileptic drugs monotherapy: a cross-sectional comparative study. J Child Neurol. 2015;30(4):451-7. doi: 10.1177/0883073814551388. [PubMed: 25342306].
  • 18. Yoo JH, Hong SB. A common mutation in the methylenetetrahydrofolate reductase gene is a determinant of hyperhomocysteinemia in epileptic patients receiving anticonvulsants. Metabolism. 1999;48(8):1047-51. [PubMed: 10459572].
  • 19. Eldeen ON, Abd Eldayem SM, Shatla RH, Omara NA, Elgammal SS. Homocysteine, folic acid and vitamin B12 levels in serum of epileptic children. Egypt J Med Human Genetics. 2012;13(3):275-80. doi: 10.1016/j.ejmhg.2012.05.002.
  • 20. Xia C. Effects of Carbamazepine and Oxcarbazepine on the Risk Factors of Vascular Diseases in Adult Epileptic Patients. Nanchang University; 2012.
  • 21. Li J, Zhang M, Chen X, Yang Y, Lu J, Luo XP, et al. Effects of Carbamazepine on Carotid Artery Intima Media Thickness in Epileptic Patients and Relative Risk Factors. China Pharm. 2013;20:25.
  • 22. Mintzer S, Skidmore CT, Abidin CJ, Morales MC, Chervoneva I, Capuzzi DM, et al. Effects of antiepileptic drugs on lipids, homocysteine, and C‐reactive protein. Ann Neurol. 2009;65(4):448-56.
  • 23. Ming L. Effects of Carbamazepine on Blood Homocysteine Levels and Folic Acid and Vitamin B12 Levels in patients with epilepsy. Contemp Med. 2014;10:135-6.
  • 24. Xuefeng X. A Deep Analysis on Effects of Commonly Used Antiepileptic Drugs on Homocysteine Levels in Children with Epilepsy. Health Must Read. 2012;11(10):357.
  • 25. WANG X, ZHANG T, ZHAO X, GAO J, Liu XQ, CHI Z. Effect of anti-epileptic drugs on the blood levels of homocysteine, folate and vitaminB12 [J]. J Shandong Univ Health Sci. 2007;4:8.
  • 26. Caizhi H, Liya M, Bin H. Effect of Antiepileptic Drugs on the Level of Serum Homocysteine in Epilepic Children. Modern Prevent Med. 2008;35(23):4737-8.
  • 27. Zhang QM, Jie X. Effects of Valproate and Carbamazepine on the levels of blood homocysteine,folic acid,vitamin B_(12) in patients with epilepsy. J Clin Neurol. 2012;25(4):291-2.
  • 28. Shan Q, Wang S, Li P. Effects of multi-vitamin supplement therapy on plasma homocysteine level in patients with epilepsy. J Chongqing Med Univ. 2014;39(2):197-202.
  • 29. Xu T, Li X, Wang WZ, Hu P, Du F. Detection of publication bias in meta-analysis of dichotomous variable—Egger test and Begg test. J Evid Based Med. 2009;19(3):181-4.
  • 30. Mai JZ, Li H, Fang JQ, Liu XQ, Rao XX. Estimation of fail-safe number in meta-analysis. J Evid Based Med. 2006;6:297-303.
  • 31. Upchurch GJ, Welch GN, Fabian AJ, Freedman JE, Johnson JL, Keaney JJ, et al. Homocyst(e)ine decreases bioavailable nitric oxide by a mechanism involving glutathione peroxidase. J Biol Chem. 1997;272(27):17012-7. [PubMed: 9202015].
  • 32. Yan QC, Xiong AH, Xiao X, Luo YT. Significance of plasma 8-iso-prostaglandin F2alpha level in acute myocardial ischemia and intervention effect of N-acetylcysteine: a study in rats. AcaD J First Med College PLA. 2003;23(6):605-7.
  • 33. Gu Q, Li Y, Cui ZL, Luo XP. Homocysteine, folate, vitamin B12 and B6 in mothers of children with neural tube defects in Xinjiang, China. Acta Paediatr. 2012;101(11):e486-90. doi: 10.1111/j.1651-2227.2012.02795.x. [PubMed: 22860981].
  • 34. Castro R, Rivera I, Blom HJ, Jakobs C, Tavares de Almeida I. Homocysteine metabolism, hyperhomocysteinaemia and vascular disease: an overview. J Inherit Metab Dis. 2006;29(1):3-20. doi: 10.1007/s10545-006-0106-5. [PubMed: 16601863].
  • 35. Xiaoguang WQ. Homocysteine detection and its clinical application. Chinese J Lab Med. 2006;29(3):193-5.
  • 36. Fernandez-Miranda C, De la Pena P, Penas M, Saiz R, GOmez P, Gomez D. Hyperhomocysteinemia and treatment with antiepileptic drugs. Effects of different doses of folic acid. Medicina clinica. 2005;124(14):521-4.
  • 37. Semmler A, Moskau-Hartmann S, Stoffel-Wagner B, Elger C, Linnebank M. Homocysteine plasma levels in patients treated with antiepileptic drugs depend on folate and vitamin B12 serum levels, but not on genetic variants of homocysteine metabolism. Clin Chem Lab Med. 2013;51(3):665-9. doi: 10.1515/cclm-2012-0580. [PubMed: 23382314].
  • 38. Türner L, Serdaroglu A, Hasanoglu A, Biberoglu G, Aksoy E. Plasma homocysteine and lipoprotein (a) levels as risk factors for atherosclerotic vascular disease in epileptic children taking anticonvulsants. Acta Paediatrica. 2002;91(9):923-6.
  • 39. Linnebank M, Moskau S, Semmler A, Widman G, Stoffel-Wagner B, Weller M, et al. Antiepileptic drugs interact with folate and vitamin B12 serum levels. Ann Neurol. 2011;69(2):352-9. doi: 10.1002/ana.22229. [PubMed: 21246600].
  • 40. Huijuan WLD. The identification and solution of heterogeneity in Meta-analysis. Acad J Second Mil Med UniV. 2006;27(4):449-50.
  • 41. Biljana M, Jelena M, Branislav J, Milorad R. Bias in meta-analysis and funnel plot asymmetry. Stud Health Technol Inform. 1999;68:323-8. [PubMed: 10724898].