Effects of Add-on Melatonin Administration on Antioxidant Enzymes in Children with Epilepsy Taking Carbamazepine Monotherapy: A Randomized, Double-blind, Placebo-controlled Trial


Address correspondence and reprint requests to Dr. Madhur Gupta C/O Dr. Kamlesh Kohli at Department of Pharmacology, Lady Hardinge Medical College, Shaheed Bhagat Singh Marg, New Delhi-110003, India. E-mail: madhurgupta@hotmail.com


Summary: Purpose: Melatonin has been shown to exhibit antioxidant, antiexcitotoxic, and free radical–scavenging properties in various animal models. The study was designed to assess its effects on the blood levels of antioxidant enzymes in children with epilepsy receiving carbamazepine (CBZ).

Methods: In a double-blind, randomized, parallel-group, placebo-controlled trial, we assessed the effect of add-on melatonin (6–9 mg/day for 14 days) on the antioxidant enzymes glutathione peroxidase (GPx) and glutathione reductase (GRd) in 31 children with epilepsy receiving CBZ monotherapy, who were seizure free at least for the last 6 months. The interaction of melatonin with CBZ and its active metabolite, carbamazepine-10, 11-epoxide (CBZ-E), also was studied.

Results: An increase in GRd activity was noted in the melatonin group as compared with a decrease of the same enzyme in the placebo group. Changes in GPx activity failed to reach statistical significance. No significant changes were found in the serum levels of CBZ and CBZ-E in either group.

Conclusions: The study suggests that melatonin exerts antioxidant activity in patients with epilepsy receiving CBZ therapy.

Evidence has accumulated about the involvement of reactive oxygen species (ROS) in epilepsy (1). The free radicals generated cause a cascade of neurochemical events leading to neurodegeneration and cell death. Free radicals also are produced by antiepileptic drugs (AEDs) like carbamazepine (CBZ) (2).

The neuromodulator melatonin has been shown to reduce oxidative stress in various animal models, because of its antioxidant as well as free radical–scavenging properties (3), and to antagonize a mutagenic effect of CBZ in some systems (4). Melatonin also exhibits anticonvulsant activity in different seizure models (5), and at least one clinical report suggests its potential efficacy in a child with severe myoclonic epilepsy (6).

No randomized controlled trials have assessed the effect of add-on melatonin on antioxidant enzymes in pediatric epilepsy. The objectives of the present study were (a) to compare the effect of add-on melatonin with placebo on antioxidant enzymes in children with epilepsy receiving CBZ monotherapy; and (b) to study the pharmacokinetic interaction between melatonin and CBZ and its active metabolite carbamazepine-10,11-epoxide (CBZ-E).


The study was a double-blind, randomized, placebo-controlled trial. Children with epilepsy, aged between 3 and 12 years of either sex, seen in the seizure clinic at the Kalawati Saran Children's Hospital, Lady Hardinge Medical College, New Delhi, between April 2002, and February 2003, were enrolled. All patients were assessed and screened for inclusion/exclusion criteria (n = 45). The institutional scientific and ethical committee approved the study protocol, and the written informed consent was obtained from the accompanying parent/relative. Patients who were receiving CBZ monotherapy in the last 6 to 9 months, had a confirmed diagnosis of epilepsy limited to partial or generalized seizures as classified according to the Increasing Current Electroshock Seizure (ICES), and were seizure free at least for the last 6 months were included. Patients with a history of psychiatric or progressive neurologic disorder or a chronic hematologic, cardiac, hepatic, renal, or thyroid disorder were excluded.

Blood samples (5 ml) were collected just before the morning dose of CBZ. A randomization code list was prepared by a statistician not connected to the study. Patients were then randomly divided into two groups: one group received add-on melatonin (n = 16), while the other received placebo (n = 15), 1 h before bedtime. Melatonin tablets of 3 mg strength (Aristo Pharmaceuticals Ltd, Mumbai, India) were used. The placebo tablets, identical in shape, size, color, and packaging, were prepared for the study by Aristo Pharmaceuticals. The dose of melatonin was 6 mg (two tablets) for children younger than 9 years/weighing <30 kg, and 9 mg (three tablets) for children older than 9 years/weighing >30 kg. The doses of melatonin administered were based on those used by Jan and Donnell (7) in pediatric sleep disorders. The duration of melatonin treatment was 14 days. The duration of follow-up was 4 weeks. Blood samples for estimation of CBZ, CBZ-E, and the antioxidant enzymes glutathione peroxidase (GPx) and glutathione reductase (GRd) were drawn before melatonin/placebo treatment and on day 14 of treatment.

The patients were called for regular follow-ups at weekly intervals. Clinical laboratory tests (liver function tests, hemoglobin, etc.) were performed at baseline and at each visit. For each patient, a daily diary was provided with the instruction to record immediately any side effects or unusual symptoms observed.

Biochemical analysis

Peripheral blood (5 ml) was co1lected before the first daily dose (trough sample) in a glass test tube and centrifuged at 3,500 rpm, and the sera were separated and stored in deep freezer at –80°C until assay of GRd, CBZ, and CBZ-E. Heparinized blood was collected and stored at –80°C for the assay of GPx in whole blood. CBZ and CBZ-E were measured by high-performance liquid chromatography (HPLC) (8). All enzyme-activity assays were performed on the Ames (Technicon) RA 50 chemistry analyzer. GPx activity was measured in whole blood samples at 340 nm as nicotinamide adenine dinucleotide phosphate, reduced form (NADPH) oxidized by using the assay kit with cumene hydroperoxide as the substrate (Randox Laboratories Ltd., Crumlin, Co. Antrim, United Kingdom). One activity unit was defined as 1 μmol NADPH oxidized per minute per liter of hemolysate. GRd activity was determined in the serum samples at 340 nm by using a commercial kit (Randox Laboratories Ltd.). One activity unit was defined as 1 μmol NADPH formed per minute per liter of serum.

Statistical analysis

Descriptive statistics were calculated for all the outcome variables and expressed as mean ± SD or median (range), as appropriate. The Wilcoxon rank-sum test was used to compare the effects of add-on melatonin versus placebo groups, and the Wilcoxon signed rank test was used to compare pre- and posttreatment values within each group because the antioxidant enzymes and the serum levels of CBZ and CBZ-E were not normally distributed, and the total size of the completers' sample is <30. The χ2 test was used to compare the categoric variables with two different groups. No power calculation was done because this was an explorative study. A p value <0.05 was considered significant. All the data were analyzed by using STATA 7.0 (intercooled version).


A total of 31 patients met the entry criteria. Sixteen patients were randomly allocated to melatonin, and 15, to placebo. One patient in the melatonin group, who showed computed tomography (CT) signs of tuberculous meningitis, was transferred to another hospital, and was therefore not included in the analysis. Two patients in the placebo group were lost to follow-up after the first week, hence their data also were not included in the analysis. Thus 15 patients in the melatonin group and 13 patients in the placebo group were finally assessed.

The patients in the two groups did not differ significantly with respect to median age, sex, and weight (Table 1). The hematologic and biochemical investigations carried out at weekly intervals were within normal limits in all the patients. No adverse event warranting discontinuation of the therapy was observed. The parents of 13 of 15 children in the melatonin group reported a perceptible increase in the children's appetite, as compared with four of 13 in the placebo group.

Table 1. Demographic characteristics of the study groups

Carbamazepine +
placebo group
(n = 13)
Carbamazepine +
melatonin group
(n = 15)

  1. Data are expressed as mean (SD).

  2. ap < 0.05, significant.

Age (yr)a8.3 (3.1)8.1 (2.6)0.83
Weight (kg)a21.0 (6.4)21.6 (5.2)0.73
Age at onset of6.6 (2.6)5.3 (2.3)0.32
seizures (yr)a
Types of seizures
Complex partial seizures910 
Generalized tonic–clonic seizures330.37
Simple partial seizures12 
Type of epilepsy Idiopathic
Localization related56 
 Localization related56 

An 8.1% decline was noted in the median levels of GRd in the placebo group as compared with a 30.7% increase in the melatonin group, the difference between groups being statistically significant (p = 0.0001; Table 2). As far as GPx is concerned, a 12.9% decrease in the placebo group, and a modest increase of 10.14% in the melatonin group occurred, but the between-group difference failed to reach statistical significance (p = 0.09).

Table 2. Antioxidant enzymes, serum levels of carbamazepine, and carbamazepine-10,11-epoxide in the add-on melatonin versus add-on placebo groups
Study variablesCarbamazepine + placebo (n = 13)Carbamazepine + melatonin (n = 15)p Value (between groups)
  1. Data are expressed as median (range).

  2. ap < 0.05.

Glutathione peroxidase(U/L)
 Pretreatment  10,988 (4,551–17,794)   8,487 (1,025–15,785)0.14
 Posttreatment 0.40
 p Value  8,200 (3,157–17,056)  9,223 (3,198–14,808)0.09
 Percentage change0.180.36 
 −12.9 (−48.3–40.5)+10.1 (−30.5–212)  
Glutathione reductase(U/L)
 Pretreatment 74.5 (42.0–186.0) 82.5 (44.0–198.0)0.28
 Posttreatment68.5 (42.0–94.0)117.5 (44.0–224.0)  0.002a
 p Value0.01a0.002a 
 Percentage change−8.1 (−52.0–1.4)+30.7 (−20.2–150.3)   0.0001a
Serum carbamazepine levels (μg/ml)
 Pretreatment4.7 (0.8–8.9) 4.7 (2.4–14.3)0.77
 Posttreatment5.2 (0.7–14.9)5.1 (2.9–10.9)0.87
 p Value0.670.12 
 Percentage increase  0.98 (−47.0–302.5)+5.7 (−56.0–70.2)0.95
 Pretreatment1.41 (0.93–2.86)0.7 (0.3–2.3)  0.04a
 Posttreatment0.96 (0.04–3.51)0.7 (0.2–1.4) 0.43
 p Value1.000.93 
 Percentage change  7.1 (−22.7–29.0) −6.7 (73.4–200.0)0.74

The percentage changes in the serum levels of CBZ and CBZ-E in both groups were not statistically significant (Table 2).


To our knowledge, this is the first randomized controlled trial to show that in children with epilepsy receiving CBZ therapy add-on melatonin leads to an increase in antioxidant enzyme activity. In particular, an increase in GRd activity was observed in the melatonin-treated children, whereas a decrease in the activity of this enzyme was observed in the placebo group. One possible explanation for our findings is that CBZ-triggered ROS accumulation (2) resulted in disruption of the glutathione redox cycle at the enzyme level, and that this effect was antagonized by melatonin. Free radicals generated in epileptic processes (9) also have been reported to inactivate GRd and GPx, and in primary cortical neuron cultures, melatonin has been shown to counteract the changes in glutathione induced by kainic acid. These neuroprotective effects of melatonin may result from sparing of GRd and GPx, which decreased in kainic acid–treated, but not in kainic acid/melatonin–treated animals (10). As only seizure-free patients were recruited in our study, however, it is unlikely that melatonin acted by counteracting a metabolic effect related to seizure activity.

Melatonin reportedly stimulates several antioxidative enzymes including GRd and GPx. At pharmacologic as well as at physiological levels, melatonin has been shown either to stimulate gene expression for the antioxidant enzymes or to increase their activity. In the nucleus, melatonin binds to retinoid Z receptor (RZR) RZRβ, RZRα, and retinoic acid–related orphan receptor, RORα1, which are proposed to be involved in melatonin-modulated transcriptional regulation in peripheral tissues. RZR/ROR response elements have been found in some important genes, some related to oxidative stress. However, to date, no RZR response elements have been described in the promoters of antioxidant enzyme genes (11,12).

In the present study, no significant changes in the serum concentrations of CBZ as well as its metabolite CBZ-E were observed after add-on melatonin therapy, thereby suggesting that a pharmacokinetic interaction is unlikely to occur.

Melatonin, used in human studies at doses of 1 to 300 mg, has been shown to be safe, with no serious adverse effects observed (13,14). It is our hypothesis that simultaneous scavenging/preventing the free radical generation by add-on melatonin may offer a potentially useful approach to minimize damage caused by oxidative stress in CBZ-treated patients.


Acknowledgment:  We acknowledge Aristo Pharmaceuticals Ltd., Mumbai, India, for supplying melatonin as well as placebo tablets for the study. The study was supported partially by a financial grant from the Indian Council of Medical Research (ICMR), New Delhi, India.